Новости
PHARMACEUTICAL DOSAGE FORMS Tablets
SECOND EDITION, REVISED AND EXPANDED
In Three Volumes VOLUME 3
EDITED BY Herbert A. Lieberman
SECOND EDITION, REVISED AND EXPANDED
In Three Volumes VOLUME 3
EDITED BY Herbert A. Lieberman
Preface
Tablets are the most commonly prescribed dosage form. The reason for
this popularity is that tablets offer a convenient form of drug administration,
provide dosage uniformity from tablet to tablet. are stable over extended
and diverse storage conditions, and can be produced on high-speed
compression, labeling, and packaging equipment. As a result, tablet production
technology is constantly undergoing improvements that enhance
their ability to deliver, with precision. a desired drug in a dosage form
intended for immediate or extended therapeutic effect. In addition. the
growth of the generic industry as well as increased competition from both
foreign and domestic markets require that a tablet manufacturer have
greater concern regarding the economics of tablet production by introducing
less labor-intensive, higher-productivity manufacturing methods for making
the increasing number of tablet products available today. The changes in
the science and technology of tablet formulation, production, and quality
assurance to accomplish the above are reflected in the second edition of
the three-volume series Pharmaceutical Dosage Forms: Tablets.
The first volume in this series describes the many types of tablet
products, giving specific updated examples of typical formulations and
methods of manufacture. These include single- and multilayered tablets,
buccal and sublingual tablets, effervescent tablets. and diverse methods
for manufaeturtng them by wet and dry granulations and by direct
compression. In addition, medicated candy products are a form of drug
delivery that has appeared in the marketplace; no complete chapter on
this technology has been printed in any pharmacy text other than both
editions of this series on tablets.
To manufacture tablets a number of unit processes are required, such
as mixing, drying, size reduction, and compression. The economics of
tablet production today require an update of the technologies for each of
these pharmaceutical operations. The granulations and tablets produced
have particular characteristics that must be analyzed and understood in
order to produce superior tablets, particularly when new and sometimes
faster methods of manufacture are introduced. No drug dosage form
would be meaningful to the patient without the drug being bioavailable.
The chapter on bioavailability in tablet technology is updated in the second
edition of Volume 2. Finally, many advances in the specifications and
care of tablet tooling and problem solving caused by faulty compression
tools are expertly covered in the second volume.
Volume 3 in the series on tablets updates the special characteristics
that should be considered for optimizing tablet production. Particular
emphasis is given to design methods that should be considered when formuIattng
a tablet product. Discussions of specialized granule and tablet-coating
equipment are presented. discussing improvements or presenting new
equipment developed since the publication of the first edition. Aqueous film
coating is now firmly established in pharmaceutical coating processes, and
thus, a shift in emphasis on coating procedures has been made in the revised
chapter on coating. New coating pans and automation of aqueous
film- and sugar-coating methods are covered. Fluid-bed processes and
particle-coating methods, including theoretical considerations, are updated
to reflect current practices.
No text on tablet technology eould be considered complete without a
full theoretical and practical updated description of current methods for
formulating, manufacturing, and controlling the release of drug from sustained-
release tablet and particle dosage forms. A chapter on sustained
drug release through coating provides an updated and authoritative discussion
of this popular form of drug delivery. There is an enhanced emphasis
on the various polymers and their combinations used to attain sustained
drug activity. Pilot operations must reflect production methods in order
to minimize difficulties in transferring a product from preproduction to
production. Granules prepared by precompression, wet and dry granulation,
fluidized-bed granulation, and spray drying are compared. A new
method for preparing a granulation, namely the moisture-activated dry
granulation (MADG), is also suggested for more widespread pilot evaluation.
With the increasing emphasis on product uniformity from one tablet to
another. or from one batch to another, whether the product is made sequentially
or with long lag periods between batches, or whether the raw
material source is from several different manufacturers, the concept of
process validation is essential. An extensive chapter descrfbing the essential
considerations that should be evaluated in process validation has been
added to the revised edition of this volume. Although the chapter presents
a complete detailed description of many validation methods, it also
shows how less detailed approaches, some of which are commonly used in
the industry, are useful. Current tablet production methods are described
with sample control charts to help the readers improve their tablet production
methods. The importance of the several different functions of production
departments, their particular skills, and the need for coordinated
and cooperative work relationships are stressed in the chapter "Tablet
Production" so that the combined, partnership efforts of all production
personnel can lead to superior tablet production. Automation of tablet
compression and coating is also part of the chapter concerned with the
production of tablets.
In the discussion of stability, updated stability protocols to comply
with recent FDA guidelines are presented. A new covariance analysis and
statistical method for expiration date prediction are described. The chapter
"Quality Assurance" upgrades tablet testing for uniformity. dissolution,
assay limit, test methods, and compendia! requirements for tablets to comply
with current USP/NF requirements. Included are instructive figures
for new schematic sampling plans. an update of the restrictions on the use
of colors, and a recommended sampling method for raw materials. Thus,
with this third volume on tablets. all the parameters currently concerned
with the production of superior tablets are made current and discussed
extensively.
An updated and full coverage of the many topics concerned with tablets
requires highly knowledgeable authors for each of the many areas that must
be covered. To compile and update the pertinent information needed for
the various chapters in this book required a multiauthored text of technologists
with specific expertise and experience in their chosen subject matter.
Each of the authors was charged with teaching their subject in such a
fashion that the novice as well as the experienced reader will profit.
They were to offer basic scientific facts and practical information so that
all readers can learn theory and apply it toward the knowledge that each
needs to formulate, produce, and control tablet operations in a scientific
rather than an empirical manner.
With this third volume. the editors have finished their task of updating
the second edition on tablets. The editors are grateful to the authors for
their fine contributions and. particularly. their patient response to the
editors' suggestions for changes. The choice of the chapter topics, the
authors. and the format are the responsibilities of the editors. It is hoped
that these choices will prove fruitful to our readers by helping them solve
their tablet technology problems and thereby advance industrial pharmacy's
contribution toward improving both quality and efficiency in the manufacture
of tablets.
I. INTRODUCTION
The old saying that all progress is change but not all change is progress
is considered a truism by product manufacturers. Changes that are made
in the production process of an established product are thoroughly evaluated
from every point of view before implementation is allowed. The
very existence of the company depends on its ability to produce products
for sale. Therefore, any suggested change in the manufacture of an existing
product must be viewed with suspicion until it can be shown that
the change will indeed be advantageous to the company and not simply
represent change for the sake of change. A company's prestige, profitability,
and compliance with legal requirements are at stake when production
changes are considered. Therefore, changes in the pharmaceutical
industry must clearly be warranted before they are implemented. The
object of this chapter is to introduce to the pharmaceutical scientist how
changes can be accomplished in a pharmaceutical tablet production facility
that wID constitute progress for the company.
A. Unit Operations and Pharmaceutical Processing
A unit operation can be defined as a process designed to achieve one or
more changes in the physical and/or chemical properties of the raw material
( s) being processed.
Two or more unit operations that are designed to convert the basic
raw materials into the final product or at least to have significantly
improved the quality or value of the original raw material(s) describes a
manufacturing system. In a tablet-manufacturing system, some of the unit
operations may include (l) particle size reduction, (2) sieving or classification,
(3) mixing, (4) particle size enlargement, (5) drying, (6) compression,
(7) sorting, and (8) packaging.
The literature is well documented with various unit operations involved
in the manufacture of tableted products which may impact on a number of
final dosage-form quality features. The features include, but are not
limited to, such items as content uniformity, hardness, friability drug dissolution
properties, and bioavailability. The traditional responsibility of
the development pharmacist has been to identify and control such impacts.
This includes the establishment of specific operating limits for each unit
operation to ensure that the production system is under sufficient control
for the production of safe, effective, and reliable tablets. The scope and
purpose of this chapter is to examine system design considerations, and as
a result, the individual unit operations and their potential impact on product
quality will not be covered. The authors assume that each unit operation
has been thoroughly investigated and is under sufficient control to
do what it is purported to do. Specific effects of processing variables on
product quality are dealt with in the pertinent chapters in this book series.
B. Batch Versus Continuous Processing
Most pharmaceutical production operations are batch operations, whereby a
series of manufacturing steps are used to prepare a single batch or lot of
a particular product. The same quantity or batch of material, if processed
en mass through the various production steps to produce the final product,
will then typically be treated by the manufacturer and the U. S. Food and
Drug Administration (FDA) as one lot. A relatively few pharmaceutical
products are prepared by true continuous-processing procedures whereby
raw materials are continuously fed into and through the production sequence,
and the finished product is continuously discharged from the final
processing step(s). In continuous processing, one day's production, the
production from one work shift, the quantity of a critical raw material from
a given lot, or a combination thereof may define a single lot of manufactured
product. Development of continuous-processing procedures requires specially
designed and interfaced equipment, special plant layouts, and dedicated
plant space, which reduces plant flexibility. The equipment, special
plant design requirements, and dedicated space are all factors leading to
the high costs of setting up and maintaining a continuous-manufacturing
operation for a product. Unless the volume of product being produced is
very high, the cost of setting up a product-dedicated, continuous-production
operation will not usually be justified. Since pharmaceutical products,
even within a dosage-form class such as compressed tablets, differ materially
in many important aspects , such as drug dosage, excipients used, critical
factors, affecting product quality, manufacturing problems, and the
like, it is usually not feasible or possible to set up a continuous-processing
operation for one product and then apply it to several others. Thus true
continuous-processing operations are largely limited in the drug industry
to large-volume products, which often also have large doses (or high
weights per tablet), such as Tums, Gelusil, Aldomet, or Alka Seltzer, for
Which tons of material must be processed each day, on an ongoing basis.
Improved Table Production System Design
II. BENEFITS OF IMPROVED TABLET
PRODUCTION SYSTEMS
In either the batch or the continuous mode of operation, the objective of
the production function is to produce pharmaceutical tablets for sale.
Consequently, a production process should always be viewed as a candidate
for progressive change in an effort to maintain the company's products
in the marketplace. When looking at a process for possible changes,
one must be aware of the potential benefits for the company. The major
benefits that can be obtained, and which constitute the major valid reasons
for changing or redesigning a pharmaceutical tablet manufacturing
process, are the following [1- 5] :
Regulatory compliance
Current good manufacturing practices (CGMPs)
Occupational Safety and Health Administration (OSHA)
Environmental Protection Agency (EPA)
Increased production capacity and flexibility
Decreased product throughput time
Reduced labor costs
Increased energy savings
Broadened process control or automation for control, operator interface,
and reporting
Enhanced product quality
Enhanced process reliability
The benefits that can be derived from the redesign of a process are
dependent to a large extent on how "good" or "bad" the old process is.
With perhaps the exception of some CGMP, OSHA, or even EPA requirements,
the expected benefits to be derived from proposed process design
changes are generally reduced to a dollar figure in order to compare the
cost of implementing the change with the expected savings to be generated
by the change. Proposed changes on a good facility may not be costjustifiable,
whereas the same proposal on a bad facility can easily be costjustified.
What can be justified and what cannot is a function of what can
be referred to as the corporate policy or corporate personality at the time
the proposal is made. A pharmaceutical company, like any company, is a
part of the community in which it resides and as such has responsibilities
to many other societal groups. These include governments, stockholders,
employees, the community, the competition, and customers. The interaction
of all these groups with the company results in a corporate personality.
Therefore, companies have different standards for evaluating different
financial situations, different production philosophies, varying time
constraints for design and implementation, and varying technical support
available for design changes. All of these items will be given consideration
and will influence the decision-making process when a redesign proposal
is made.
III. PRODUCTION PROCESS DESIGN CONSIDERATIONS
A fact that should be well understood is that the production unit is not an
independent organization within the company. Any changes made in the
production unit impact on many other units that come in contact with the
production unit. However, with the rising costs of materials, labor, inflation,
and regulations being constantly added to the manufacture of
pharmaceuticals, the pharmaceutical industry cannot afford to neglect
change or the modification of existing processes simply because there are
established standards that may be difficult to change. Increased productivity
matched with a reduction in direct labor cost will probably be
the way of the future. This can only be accomplished by a conscious effort
on the part of those in process development. The development pharmacist
should therefore be familiar with ways in which the manufacture of
pharmaceutical products can be increased and labor reduced. Even though
the development pharmacist will not be thoroughly knowledgeable in all
areas of process design considerations, he or she should at least be aware
that those considerations exist and be able to interact with experts in
those fields to accomplish cost savings.
Major advances in efficiency have been made in tablet production over
the last 20-30 years. These advances have come about by the development
of methods; pieces of equipment, and instrumentation with which tablet production
systems have been able to (1) improve materials handling. (2) improve
specific unit operations, (3) eliminate or combine processing steps,
and (4) incorporate automated process control of unit operations and
processes.
A. Materials Handling
Materials-Handling Risks
Possibly the single largest contribution to the effectiveness of a manufacturing
facility is made by the facility's materials-handling capabilities.
In addition to a lack of efficiency, there are certain risks involved with
improper or inefficient materials handling. These riaks include increased
product costs, customer dissatisfaction, and employee safety liabilities.
Materials that are not handled efficiently can increase the cost of raw materials.
For example, penalty charges are assessed (demurrage) when
railroad cars are not loaded or unloaded according to schedule. If raw
materials are not moved as required by production schedules, delays are
incurred which can lead to situations in which machine time is wasted,
personnel time is wasted. in-process inventories are increased, and the entire
manufacturing process is slowed down. In addition, the improper
handling and storage of materials can lead to damaged, outdated, and lost
materials. Improper materials handling can place employees in physical
danger. An increase in employee frustration generated by constant production
delays due to poor materials handling can result in reduced morale.
Materials-Handling Objectives
A well-designed and efficiently operated materials-handling system should
impart to the manufacturing facility reduced handling costs, increased manufacturing
capacity. improved working conditions. and improved raw material
distribution to the appropriate manufacturing areas. To achieve an
efficient materials-handling system. as many of the basic general principles
of efficient materials handling need to be implemented as practical.
Improved Tablet Production System Design 5
Basic Principles of Materials Handling
SHORT DISTANCES. Raw materials used in the production process
should only be moved over the shortest possible distances. Moving materials
over excessive distances increases production time. wastes energy,
creates inefficiency, increases the possibilities of delays. and adds to the
labor costs if the material is moved by hand.
SHORT TERMINAL TIMES. When material is being transported, the
means of transportation should not have to spend time waiting On the material
to be picked up. Material should always be ready when the transportation
is available.
TWO-WAY PAYLOADS. The transportation mechanism for raw materials
should never be moved empty. For example. if tablet-packaging materials
are being transported from a warehouse to a packaging line by means of a
fork-lift truck. the return trip could be used, for example, for the removal
of accumulated scrap material or finished product from the production
line.
AVOID PARTIAL LOADS. The movement of quantities of raw materials
that require only a partial use of the total capacity of the transportation
system is a waste of time and energy. If the movement of such materials
cannot be avoided, then the partial loads should be doubled up to make
efficient use of the transportation system.
AVOID MANUAL HANDLING. The manual handling of raw materials is
the most expensive way of moving them. A great deal of direct costs can
be eliminated by mechanically moving material rather than having it moved
by the production personnel.
MOVE SCRAP CHEAPLY. The movement of scrap material is a nonproductive
function and scrap material should not be a primary work objective
for a transportation system. Therefore, scrap material should be moved as
inexpensiVely as possible. One way of achieving this is to handle the removal
of scrap on return loads after delivering needed materials.
GRAVITY IS THE CHEAPEST POWER. The movement of material from
top to bottom through multileveled facilities. if available. is the most efficient
way of moving material.
MOVE IN STRAIGHT LINES. A basic theorem of geometry states that
the shortest distance between two points is a straight line. Therefore,
the most efficient movement of material is along the straightest and most
direct path to its destination.
UNIT LOADS. Raw materials should be delivered to their ultimate
destination in loads that will be completely consumed by the manufacturing
process. This will eliminate the need for the excess raw materials to be
handled at the point of use and to be returned to the storage place.
LABEL THOROUGHLY. Thoroughly labeled raW materials will eliminate
or help to eliminate having the improper raw material delivered to a production
site. Much time and effort is wasted in correcting raw materials
errors.
Of course. not all of the basic materials-handling principles can be applied
or applied to the same extent in every situation. How a company designs
its materials-handling facility and its production process is based on several
different considerations. The type of process that is being designed,
batch or continuous; will have a great influence on the type of materialshandling
system used. The materials-handling system in a batch-ortented ,
multiproduct facility is significantly different from a single-product, continuous-
process. materials-handling system. The type and quantity of
product as well as the physical characteristics of the raw materials involved
will also affect the materials-handling system. For example, fluids for
granulating will have to be pumped, whereas solids will have to be handled
in another manner, such as with vacuum devices of conveyors. The type
of building that is available for the manufacturing process may exert the
single greatest influence on the materials-handling system. The lack of a
multistoried building will preclude the use of gravity for material flow.
Like all company decisions. the costs of purchase. installation. operation.
and maintenance as well as the useful life and the scrap value of the materials-
handling system must be weighed against its advantages.
Greene 16] quotes an old cliche which best summarizes materials
handling:
The best solution to a materials-handling problem is not to move
the material. If this is possible. then move it by gravity. And if
this is not possible, usually the next best way is to move it by
power. If people must move the material, it should be moved by
the cheapest labor possible.
Examples of Materials-Handling Improvement
Improvements in the use of materials-handling techniques have been incorporated
into the new product facilities of Merck Sharp & Dohme (MSD) and
Eli Lilly and Company (Lilly). Figure 1 shows a schematic of the granulation
and tableting area at MSD's new facility for the production of Aldomet.
There is virtually no human handling of materials during or between
processing steps. The movement of materials (solids and liquids) through
the system is accomplished by a sophisticated materials-handling system.
The system is built in a three-story building Which allows the entire
process to operate in three operational "eolumns , II The system incorporates
vertical drops to utilize gravity whenever possible and uses pumps,
vacuum, and bucket conveyors to move material upward whenever necessary.
In the fir-st processing column bulk raw materials are loaded into the
system on the third floor and flow by vertical drop through holding and
feeding hoppers into mixing equipment and- then into continuous-granulating
and drying equipment on the first floor. - The dried granulation is
then transported by vacuum back up to the third floor to start the second
processing column for milling and sizing. By vertical drop the sized granulatton
flows down through the lubrication hoppers and mixer and then to
the tablet compression machines on the first floor. The compressed tablets
are then transferred by bucket conveyor back up to the third floor and
the start of the third column. Here. the tablets are collected. batohed ,
and transferred to feed the air suspension film-coating equipment located
on the second floor. Once coated. the tablets are stored on the second
floor until required for packaging. During packaging operations the tablets
now by vertical drop from the second floor to the packaging lines
on the first floor.
The Aldomet facility was built to meet the need for increased manufacturing
capacity of Aldomet. The new Aldomet facility requires fewer
production people, thus keeping direct-labor costs low. Since the facUity
is continuously monitored by computer throughout each phase of the
process, Aldomet has a high degree of uniformity and reproducibility.
The continuous monitoring also bring about a conservation of energy. As
a result of the numerous and rigid in-process controls in the Aldomet
process. quality control and the quality assurance costs have been reduced.
Through the creation of a new building, MSD's compliance with the Good
Manufacturing Practices regulations was greatly facilitated [71.
A study of Figure 2. schematically representing Warner-Chilcott's redesigned
Ge1usil manufacturing system, shows that the dedicated singleprocess
system makes use of gravity, belts, and pneumatic devices to efficiently
move material through the process. The old manufacturing operation
was a semiautomated batch operation with a capacity limit of three
shifts employing 14 persons. The capacity of the redesigned system for
manufactur-ing' Ge1usil tablets is three times greater than that of the old
process. and the unit throughput time for Gelusil manufacture was reduced
to a third of the old process time. The output of the new extrusion
process, in one shift with two persons, equals the total daily output from
the old process. Obviously. another one of the benefits of the new process
was a reduction in required manpower, Which reduced the direct-labor
costs involved in producing Gelusil [5.8].
While the previous examples involved very large volumes of product
produced on a continuous basis. batch-oriented production systems have
also been made more efficient by materials-handling improvements. Figure
3 shows a schematic of E. R. Squibb IlL Sons Inc.'s (Squibb) production
process for Theragran M tablets. The vacuum transfer of the granulation
to the tablet compression machines has been accomplished by a
vertical drop from storage devices on a mezzanine above the tablet
machines.
Lilly's new dry products manufacturing facility (Fig. 4) has increased
their materials-handling efficiency by the use of (1) a driverless train
which transports bulk raw materials from the receiving dock to the staging
area, (2) vacuum powder transfer mechanisms, (3) direct transfer of inprocess
materials, and (4) the liberal use of vertical drops in the twostory
operation [9J.
B. Processing Step Combination or Elimination
The design of 9. new production facility should attempt to improve the old
process by eliminating or combining certain processing unit operations.
This could take the form of simply eliminating tasks that are no longer
necessary or by the utilization of new pieces of equipment that can perform
more than one of the unit operations required under the old
processing system.
Direct Compression
Many processing steps have been eliminated in the manufacture of some
pharmaceutical tablets as a result of the development of directly
A schematic of the Squibb Theragram M tablet production
(Courtesy of E. R. Squibb, New Brunswick, New Jersey.)
compressible excipients, wherein powdered drug can be directly mixed
with the excipient and then immediately compressed into a finiShed tablet.
Direct compression is the method of choice in tablet manufacture, when
the process may be employed to produce a high-quality finished product.
Direct compression offers the most expeditious method of manufacturing
tablets because it utilizes the least handling of materials, involves no
drying step, and is thus the most energy-efficient method, and is also the
fastest, most economical method of tablet production.
The reformulation of a product from a wet-granulated process to a
direct-compression process to eliminate several process steps is illustrated
in Table 1 by the comparison of Lederle Laboratories' (Leaderle) old and
new tablet-manufacturing processes. If reformulation to direct compression
cannot be accomplished, purchasing the critical raw material in granular
form could have the same effect from a processing standpoint. While the
cost of directly compressible raw material may be higher, the labor, time,
and energy cost savings realized by eliminating the granulation, drying,
Improved Tablet Production System Design
Table 1 Unit Processing of Solid Dosage Forms, Lederle Laboratories
11
Old tablet-manufacturing process
(wet granulation)
1. Raw materials
2. Weighing and measuring
3. Screening
4. Manual feeding
5. Blending (slow-speed planetary
mixer)
6. Wetting (hand addition)
7. Subdivision (comminutor)
8. Drying (fluid bed dryer)
9. SUbdivision (comminutor)
10. Premixing (barrel roller)
11. Batching and lubrication (ribbon
blender)
12. Manual feeding
13. Compression (Stokes rotary
press)
14. Solvent film coating (Wurster
Column)
15. Tablet inspection (manual)
New tablet-manufacturing process
(direct compression)
1. Raw materials
2. Weighing and measuring
{automatic weigher and recording
system)a
3. Gravity feeding
4. Blending (Littleford blender)
5. Gravity feeding from the
storage tanka
6. Compression (high-speed
rotary press)a
7. Aqueous coating (Hi-Coater)
aIn planning phase. To be installed later.
Source: Courtesy of Lederle Laboratories, Pearl River, New York.
and sizing of the raw material may more than justify the increased material
cost.
Unfortunately, there are many situations in which direct compression
does not lend itself to tablet production, for example, with low-dose drugs
or where segregation and content uniformity are a problem, or with highdose
drugs which are not directly compressible or which have poor flow
properties, or in the preparation of certain tablets or in many special
tablet-manufacturing operations. When direct compression is not a feasible
method of tablet manufacture, wet granulation is usually the method of
choice.
Equipment for the Elimination/Combination of
Processing Steps
Until relatively recently, wet granulation was a highly labor-intensive and
time-consuming process. Figure 5 illustrates the various steps involved in
batch wet-granulation processes in the 1960s and before. Typically, two
separate mixing devices were used for the dry solids blending and for wet
massing, with a manual transfer step in between. as shown in part A of
Figure 5. Following wet massing the entire batch was again manually
granulated by transfer to an appropriate piece of equipment after which
the granular material was further manually handled by being racked on a
series of drying trays. After oven drying the entire batch was again
manually removed from the trays and was dry sized by again being passed
through a mechanical screening device. Thereafter. the entire batch was
placed in a dry blender once again for incorporation of the lubricant.
after which the batch was placed in drums or containers to be held for
tablet compression. As can be seen in Figure SA. the components of the
batch or the entire batch was manually handled at least 10 or 11 times.
With the development of specialized high-shear powder and mass mixers
, and with the further development of fluid-bed dryers, the standard
wet-granulation batch-processing operation as it is being conducted in the
1980s is shown in part B of Figure 5. The number of transfer steps has
been reduced to three or four ~ the manual handling of the batch material
has been greatly reduced ~ and the time of granulation manufacture has
been cut from 2 days with an overnight drying step to as little as a few
hours. This procesaing can be further illustrated by a hypothetical
processing setup in Figure 6~ and by the process flow chart for the production
of Lasix in Figure 7.
New equipment is now being developed and gradually adapted by the
pharmaceutical industry which handles all of the step s in granulation
preparation in one piece of equipment. This is depicted in part C of
Figure 5. It should also be pointed out that not only is handling of materials
greatly reduced by the newer equipment used in the wet-granulation
operation, but the new process equipment is also capable of producing
granulation materials which have unique properties which could not
previously be routinely produced. Granular particles with a near spherical
shape can be routinely produced for they may be run under vacuum Conditions
that reportedly enhance the density and cohesiveness of the granular
particles. Some of the newer equipment can also be used to apply protective
coating to the particles during their manufacture. Some of these
special applications are described in the sections which follow. The reduction
of materials handling, Which is obvious from Figure 5, as the
process is currently practiced in the 1980s as compared to earlier times,
is due in large measure to the development of equipment Which can sequentially
undertake a series of processing steps which earlier were accomplished
by separate pieces of equipment. These new, multipurpose pieces
of equipment, which are capable of combining virtually all of the steps in
the wet-granulation process, are also described in the following sections.
Improved Tablet Production System Design 1'1
CONTINUOUS-BATCH POWDER MIXING AND MASSING EQUIPMENT.
The Littleford Lodige mixer (see Appendix) was one of the first highshear
powder blenders capable of rapidly blending pharmaceutical powders
and wet massing within the same equipment. With some formulations the
equipment may also be capable of producing agglomerated granular particles
which are ready for fluid-bed or other drying methods without further
processing. Figure 8 illustrates a conventional Lodige mixer and
describes the various assemblies of the unit. The unit consists of the
horizontal cylindrical shell equipped with a series of plow-shaped mixing
tools and one or more high-speed blending chopper assemblies mounted at
the rear of the mixer. For the addition of liquids, an injection tube terminating
in one or more spray nozzles is provided. The nozzle(s) is located
immediately above the chopper assembly.
In operation, the plow-shaped mixing tools revolve at variable speeds
from about 100 to 240 rpm and maintain the contents of the mixer in an
essentially fluidized condition. The plow device also provides a highvolume
rate of transfer of material back and forth across the blender.
When liquid granulating agents are added to dry powders the liquid enters
the mixer under pressure through the liquid nozzle immediately above the
chopper assembly or assemblies. Each chopper assembly consists of blades
mounted in a tulip shape configuration rotating at 3600 rpm. As the liquid
impinges on the powder in the area of the chopper it is immediately dispersed.
By controlling the duration of the mixing cycle, the particle size
of the granulation may be controlled. The choppers perform a secondary
function in this type of high-shear powder blender. It will be noted on
examination of part B of Figure 5 that the new processes involved in wet
granulation may not include a sieving operation. When the chopper blades
are operated during dry mixing dry lumps of powder are effectively dispersed.
Consequently, it will often be found that sieving is no longer an
essential prerequisite of powder blending when this type of equipment is
employed.
Using this type of high-shear powder mixing equipment, complete mixing
may be obtained in as little as 30-60 s, A temperature rise of 10-15°
may be expected if dry blending is continued over a period of 5-10 min.
When the Littleford Lodige blender is used for wet granulation the work
which must be done by the mixer will increase as the powder mass becomes
increasingly wet. This is often reflected in the readings on the
ammeter of the equipment because the increased work will result in an
amperage increase. Such readings may be very useful in helping to
identify the proper end point for the wet-granulation process.
The equipment illustrated in Figure 8 employs air purge seals. One of
the major difficulties in the operation of a high-shear powder mixer is
powder making its way behind the seals and even contaminating the bearing
assembly. This can lead to two difficulties: contamination of the next
product with materials previously run in the blender and early failure of
bearings and seals. A number of pharmaceutical manufacturers have found
the air purge seal to be a useful device and an effective mechanism for
overcoming this difficulty with this type of blender.
The Diosna mixer/granulator is another type of high-shear powder
mixer and processor (Fig. 9). The mixer utilizes a bowl (1) mounted in
the vertical position. The bowl is available in seven sizes between 25 and
100 L. The mixing bowl can be jacketed for temperature control. A highspeed
mixer blade (2) revolves around the bottom of the bowl. The blade
fits over the pin bar at the bottom of the mixing bowl which powers the
blade. The blade is specially constructed to discourage material from
getting under it. The speeds of the mixing blades vary with the size of
the mixer. The tip speed, however, is kept at 3-4/10 m s-1 on low and
6-8/10 m s-1 on high. The mixer also contains a chopper blade (3) at
either 1750 or 3500 rpm. The chopper functions as a lump and agglomerate
reducer. A pneumatic discharge port (4) may be specially ordered
for the unit to provide automatic discharge. The unit is provided with a
lid (5) and the larger units employ a counterweight (6) to assist in raising
and lowering the lid. The lid has three openings: one to accommodate a
spray nozzle, a second larger opening for an air exhaust sleeve, and a
third opening for a viewing port. The units are also equipped with an
ammeter (7) which may be employed to determine the endpoint of granulation
operations. Typical time sequences for the use of a Diosna mixer are
as follows: mixing 2 min or less; granulating, 8 min or less; discharge,
1 min, with discharge capable of being preset when the pneumatic discharge
system is in place.
Vosnek and Forbes [10] reported on the performance features of the
Diosna. They compared granulations prepared by a traditional method as
shown in part A of Figure 5 Wherein a pony mixer was used for wet
massing with granulations with the same product prepared in a Diosna
mixer followed by fluid-bed drying. Figure 10 shows a comparison of
particle size distribution of the two granulations. The Diosna equipment
produced a granulation with a more normal particle size distribution with a
smaller fraction of the granulation being 200 mesh or below (18 vs. 33%).
The shape of the particles produced by the Diosna were also more
Traditional Granulation
spherical. This is filustrated in Figure 11, where the particles on the
right side of the figure are from the Diosna, whereas those on the left
are from the conventional wet granulation process. Physical comparisons
of tablets prepared from granulations made using a range of granulating
agents as prepared by the Diosna and by the traditional granulation
equipment are shown in Table 2. The conclusion to be drawn from the
table is that the tablets produced with the Diosna equipment using the
streamlined manufacturing process were as good as and sometimes better
than tablets made by the traditional wet-granulation process. The Lilly
study summarized the advantages and disadvantages of the Diosna equipment
and process as shown in Table 3. The relatively high cost of the
equipment would be quickly offset by the efficiency and reduced handling
involved in this streamlined production method. The advantages of the
equipment obviously were thought to outweigh the disadvantages, since
Lfily has used this Diosna equipment as their primary powder-mixingI
processing equipment in their new dry products manufacturing facility
(see Fig. 4). The available Diosna models with their vessel capacity and
approximate working capacities are shown in Table 4.
Figure 12 describes the Littleford MGT mixerI granulator, which has
been developed by this company to more specifically meet granulation needs.
Since wet granular material may resist transfer by air conveyor systems
such as the Vacumax, the type of transfer provision shown in Figure 13
may be especially helpful. The need to raise the equipment to appropriate
working height in order to discharge directly into a bowl of a fluid-bed
dryer is not regarded as a major disadvantage, provided powder can be
conveniently charged into the unit when in a raised position.
Figure 14 illustrates the Oral mixer/granulator. Available from
Machines Collette, Inc., this equipment is a modification of the earlier
Oral industrial planetary mixers. The difference, however. between the
mixerI granulator and the standard planetary mixer is that the new unit
contains two mixing devices. A ll:lI'ge mixing arm (1) is shaped to the
rounded configuration of the bowl (2) and provides the large-scale mixing
motion in the powder , A smaller chopper blade (3) enters off-center from
the mixing arm and is located above it. The larger mixing blade and a
secondary chopper blade system is, therefore. similar to the Lodige and
Diosna units previously described. The difference, however, is that the
Gral unit has the configuration of a planetary top-entering mixer. The
mixing bowl may be loaded at floor level and then raised to the mixing
position by the hydraulic bowl elevator cradle (7). The bowl is brought
Improved Tablet Production System Design 27
into contact with a cover (4) providing a tight seal. An advantage of the
unit is that it may be discharged by its hydraulic port (5), whereas in
the raised position, offering sufficient space for a container to be placed
beneath the discharged port. The entire mixer unit does not have to be
elevated to provide this vertical discharge distance as is the case with the
previously two mentioned high-shear mixers. As with the other high-shear
mixers, all parts in contact with the product may be ordered in stainless
steel. Fluid may be injected into the mixer bowl. The equipment is
available with timer control (6). The Oral machines can also be supplied
with explosion-proof as well as standard electrics. The equipment is
available in five sizes ranging from 10 to 1200 L.
CONTINUOUS-BATCH MIXING, MASSING, GRANULATION, AND DRYING
EQUIPMENT. The ideal equipment for the preparation of granulations
by the wet- granulation process would be one unit which is capable of sequentially
dry mixing, wet massing. agglomerating. drying, and sizing the
material being processed. with no materials handling between steps. The
only interruption 1n such a continuous-batch process might be to stop the
unit in order to add the lubricant prior to discharging a final product
ready for compression. This type of approach is indicated in part C of
Figure 5. A number of equipment manufacturers and pharmaceutical scientists
have been working toward this goal over the last several decades.
However. it is only within the last several years that commercial equipment
has become available for continuous-batch wet granulation in one
unit.
Fluid-Bed Spray Granulators. The f'll'st equipment reported in the
pharmaceutical literature to provide continuous-batch wet granulation was
fluid-bed drying equipment which was modified by the addition of spray
nozzles or fluid injectors to provide additional liquid-binding and adhesive
agents to dry-powdered materials which were initially placed in the
equipment. Scientists working in European pharmaceutical companies were
the first to report on this approach [11-15]. However. several U. S.
companies are employing fluid-bed spray granulators (see Appendix).
Warner-Chilcott utilizes such a process in the manufacture of Pyridium
tablets. With the active ingredient, phenazopyridine, being a powerful
dye. the neW process also benefits from the fact that the system is totally
enclosed during the granulation stage [8]. Figure 15 presents a schematic
cross section of such a fluid-bed spray granulator. The airflow
necessary for fluidization of the powders is generated by a suction fan (2)
mounted in the top portion of the unit which is directly driven by an
electric motor. The air used for fluidization is heated to the desired temperature
by an air heater (5). after first being drawn through prefilters
to remove any impurities (6). The material to be processed is shown in
the material container just below the spray inlet (1). The liquid granuIating
agent is pumped from its container (3) and is sprayed as a fine
mist through a spray head (4) onto the fluidized powder. The wetted
particles undergo agglomeration through particle-particle contacts. Exhaust
filters (7) are mounted above the product retainer to retain dust
and fine particles. After appropriate agglomeration is achieved. the spray
operation is discontinued and the material is dried and discharged from the
unit. Figure 16 provides a schematic view of the automatic fluid-bed spraygranulator
system available from the Aeromatic Corporation, with its integrated
materials-loading and -handling systems. The unit described in
Figure 15 Schematic cross section of a fluid-bed spray granulator.
(Courtesy of Aeromatic, Somerville, New Jersey.)
Motor
Vacuum
Convey
Exhaust
1
Filter
Motor
I~'t\
Flap ~
Rclid'+---I
Panel+---I
Heate
Drum
Figure 16 Schematic of an automatic fluid-bed spray granulator (Streba
150) with integrated materials loading and materials handling. (Courtesy
of Aeromatic, Somerville, New Jersey.)
Improved Tablet Production System Design 29
Figure 16 is reportedly capable of processing approximately 200 kg per
batch. The advantages of such rapid wet massing, agglomeration, and drying
within one unit are obviously attractive. Exclusive of equipment clean-up,
the process may readily be sequentially completed within 60-90 min or less.
A number of pharmaceutical companies, both in Europe and in the
United States, are utilizing fluid-bed processing as a rapid continuousbatch
approach to wet granulation. However, there are a number of difficulties
that exist for the process which may account for the fact that
fluid-bed granulation processing has not been more widely accepted.
Fluid-bed systems as currently available may not provide adequate mixing
of powder components. Where potent drugs are employed in a fine state
of particle size reduction, there will be a tendency of these materials to
be separated from the powdered bed and collected on the filter of the unit.
Thus, even if a separate powder-mixing operation is undertaken to produce
adequate uniformity of drug distribution within the mix, thus demixing
action must be considered in any attempt to process potent drugs
in a fluid-bed spray granulator or when there is a considerable disparity
in the particle size or density of the materials being processed. Unlike
extrusion, wet screening, and the use of other more conventional mixing
equipment. the forces of agglomeration which cause the particles to be
brought together producing the agglomerates are relatively low in the
fluid-bed spray granulator. Accordingly, not all materials may lend themselves
to agglomeration in this continuous batch-processing equipment.
Low-density and hydrophobic materials can be particularly troublesome.
Another limitation of this type of equipment may be related to particles
containing granulattng agent on their surfaces which adhere to the filters
of the equipment. This creates two problems: a reduction in the effective
area of the filters, which may produce filter failure and a build up
of pressure within the unit and other difficulties during the granulation
operation, and problems connected with cleaning the unit following the
granulation process.
All production-size fluid-bed dryer equipment should contain explosion
relief panels. However, special attention should be given to safety precautions
if organic solvents are used in the fluid-bed spray granulation
process. A number of fatal accidents have occurred worldwide with the
use of fluid-bed dryers, as have a number of less serious explosions in
which only the installation was damaged or destroyed. Dust explosions
can also occur in a fluid-bed dryer with dry materials. However, when
working with solvents. especially those with flammable vapors. it might be
wise to Ill'st check with the supplier of the equipment to review whether
or not the operation is safe or advisable.
Other manufacturers of fluidized-bed drying equipment, such as Glatt
Air Techniques, have developed specialized equipment to provide continuousbatch
processing fluid-bed granulator capabilities. The number of scientific
papers which have appeared in this field in recent years attests to
the interest of the pharmaceutical industry in this continuous-batch
processing approach. For example, investigators have studied the mixing
of pharmaceutical raw materials in heterogeneous fluidized beds [11]. the
physical properties of granulations produced in a fluidized bed [12], the
effect of fluidization conditions and the amount of liquid binder on properties
of the granules formed [13], and the various process variables
influencing the properties of the final granulation [14,15].
Figure 17 A cutaway view of a double-cone mixer/dryer processor. (Courtesy of PaulO. Abbe, Ine , , Little Falls,
New Jersey.)
Fle~ible coltriecti(;Jn$ for
f:irt:ulat/ng heating mfJi:!ium
Agtf'iJttng.(i,eI'f'AenfS,
Th"edesigil.snd
numberb! the$ff
c##be chllltgedltJ
suitspeeific
requi/'8rnent$
J:leJtible connections for
alrcuf'ating besting medium
Flej:lblliCOnttilct16rt
fQrvltcuum lihs
~,.:;: ."., "x,... Ple1ll.b..•..,leClJnnect/on l!orJ;quid spray : 1
~
~
(')
~-
~
;:s
Q.
r
Figure 21 The Marumerizer, Model QJ-400.
tion , Process Division.)
(Courtesy of Luwa Corpora
Improved Tablet Production System Design 39
Figure 22 The Xtruder , Model EXDCS-I00. (Courtesy of Luwa Corporation,
Process Division.)
amount of granulation solution used and the total mixing time. However,
control of these two variables has proved to be less than satisfactory at
times.
The utilization of high-intensity mixers in the wet-granulation process
dramatically reduces the total time required to reach or exceed a proper
granulation end point. Therefore. the need of a better understanding and
control of wet granulation has increased in recent years as the use of
high-intensity mixers has increased. In recent years. more sophisticated
techniques for monitoring the granulation process have been investigated.
Power consumption [3 - 8], torque [9 -l1J. rotation rate [12]. and motor
load analyzer [13] measurements are indirect methods of monitoring the
granulation process. Bending moment [4J. beam deflection [14]. and
probe vibrational analysis [15 -16] techniques directly measure the density
and frictional changes that occur during wet massing. Most recently.
conductance [17] and capacitive-sensor [18 - 21] methods have been used
and found to respond to moisture distribution and granule formation in
high-intensity mixers.
40 Peck, Anderson, and Banker
While none of the systems studied have universal unity, all of the
various approaches have attempted to be responsive to the dynamic
variables of the process. be applicable to formulation research. and be
potentially useful in the control of manufacturing processes. The granulation
monitoring system shows promise as a research tool for gaining insight
into the granulation process and as a control system for obtaining
better reproduction of granulations and resultant tablets. The study of
the wet-granulation process remains an active research area [1- 2,
22-26].
Tableting Improvements
Pharmaceutical tablet production improvements have also been made by improving
the performance of a specific unit operation. The development of
high-speed tableting and capsule-filling machines has allowed many manufacturers
to increase their tablet and capsule production output by replacing
their older, much lower-capacity machines.
Since the introduction of high-speed rotary tablet presses tablet outputs
are now possible in the range of 8000 -12,000 tablets per minute.
Because of these high outputs, it has become advantageous. if not necessary,
to consider automatic control and monitoring of tablet weight. One
of the early attempts to automatically monitor tablet production with an online
system for tablet press weight control was the Thomas Tablet Sentinel
(TTS). This unit utilizes commonly available strain gage technology.
Strain gages, appropriately placed on a tablet machine, monitor the strain
that is incurred during the tablet compression step on the machine.
When a tablet press is in good operating condition, is properly set up,
and the punch and die tooling has been properly standardized the amount
of pressure or compressional force developed at each station is dependent
upon the amount of powder that was contained in each corresponding die.
Measurement of this compressional force is thus an indirect method of
monitoring the tablet weights being produced by the machine, and if this
force could then be used to initiate an automatic weight adjustment, continuous
monitoring of tablet production would be possible. The TTS system
shown in Figure 23 is able to do such a task. The schematic of the
setup of the strain gages and weight adjustment servomotor mechanism is
shown in Figure 24. Figures 23 and 24 illustrate the latest monitoring
system which is known as the TTS-II. The TTS-II is a fully automatic,
on-line. electromechanical tablet weight-control system which is capable of
continuously monitoring and controlling weights of tablets as they are
being produced on a single- or double-r-otary press. Figure 24 shows the
three elements of the TTS- II: (1) the sensing systems, (2) the weight
control system. and (3) the reject control system.
The control module may contain meters for monitoring tablets as they
are being produced within desired quality limits. The latest design also
includes a tablet-counting system, a defective tablet counter, and a printer
for recording when weight adjustments are made. The TTS-ll system
utilizes a sensing system of shield strain gages mounted on appropriate
pressure points on the machine. An electronic control system feeds power
to the strain gages and amplifies their output signal.
The third section of the TTS-II system consists of the weight -control
system, which functions according to signals received from the electronic
control portion of the system. Finally. an optional automatic rejectcontrol
system is available which will reject the tablets that may drift out
Figure 23 Thomas Tablet Sentinel II. (Courtesy of Thomas Bngfneermg , Ine , , Hoffman
Estates, illinois.)
3' '1:l
cs
~
(l)
~
.>-i s(
l) ....
~
~
l: c....
5'
;:s
~
to .... (l)
3
t:l
(l)
to
~.
;:s
II::l. ....
42
TABLET
THlCI: :.-.l _..__.'"
I I / ;- ,
I II \ ' y
I 1"--' , ~ '"'""<""," • ;__ ~..::: _ .. ~ LIQUID SUPPLY
I ~ ,:::~,.:::::......--:..:r ~ L1QUlDRETURN l ~ v~0~~;1 - --J-~: I i I
..... \
...... ~'-.- " ,~ .....
...... --..... / -"'- ) / - - - _--=::-- -...,/
230J460V
3 pH 60 Hz .....-
BALANCING DAMPER
Improved Tablet Production System Design 49
programmed, so that the operator may load the coater, set the temperature,
start the warm airflow, and then start the spray system. Based on
the requirements of the dosage form being coated (tablets, beads, or
small particles), the unit is set to coat for an appropriate preselected
period of time. This technique thus lends itself to the integration with a
miniprocessor, since all coating procedures are capable of being mechanically
or electronically controlled. The air-suspension approach may be
effectively employed with either nonaqueous or aqueous film coating.
About the same time that air-suspension coating was being perfected,
another device was being investigated. This was a side-vented or perforated
coating pan originally developed by Lilly and then sold as the
Acce1a-Cota by Thomas Engineering (Fig. 29). One of the major advantages
of this equipment is the one-way flow of air through the tablet bed
and out the perforations of the pans. This greatly reduces or eliminates
the "bounce-back" of atomized spray and particle spray drying of the
spray droplets that occurs, especially with solvent-based coating, with the
conventional pans. It also benefits coating because the greater air flow
through the bed facilitates drying. Again, the key to the advance of this
new coating pan was the availability of suitable spray systems which would
provide the new spray requirements while allowing for automation and thus
programmable coating sequences within the coating unit. The side-vented
coating pan has been successfully used for continuous coating of films and
sugar systems [24]. It has been reported that by the selection of proper
viscosity sugar Slurries, a complete sugar coat may be applied over an
8-h period. This must include such sequencing as spray, pause, and dry
cycles to allow for adequate sugar slurry spreading prior to drying. It
should be noted that the Pellegrini pan, which is not side vented, but
does have the capability of moving large amounts of air through the pan,
has been adapted to continuous sugar coating, and has also been used
with programmed control systems.
Automation and Coating
For the coating of tablets with sugar systems, it was recognized that it is
possible to spray, roll out the syrup, and dry. It is also known that this
sequence can be repeated to produce a suitable coating as long as an
adequate air supply is available and suitable spray systems can be used.
With these two elements in place, it is then possible to install timers,
mechanical or computer controllers, along with suitable solenoid valves, to
have a completely programmed system. In so doing an operation is developed
that is capable of being readily validated, since all important
parameters must first be defined and then precisely controlled by the
automated system without the variability of human error. Once under
automated control, a validated process is easy to confirm with appropriate
printout recording. cf
Associated with the concept of automation may be the desire to first
optimize a process, which would then ensure that the controlled process
is operating at its maximum efficiency. A side-vented. 48-in coating pan
has been used at Purdue University in the Industrial Pharmacy Laboratory
(Fig. 30) to study a number of coating variables. This pan is equipped
with the usual temperature-indicating dials and coating pan pressure differential
readout. However, it also includes the following features which
aid in the study of the coating process:
50 Peck, Anderson. and Banker
Figure 30 An instrumented 48-in Accela-Cota installation. (Courtesy of
Industrial Pharmacy Laboratory, Purdue University, School of Pharmacy
and Pharmacal Sciences, West Lafayette, Indiana.)
1. Variable heat input (1)
2. Variable air input (2)
3. Variable exit air control (3)
4. Measurement of dew point of input (4) or exit air (5)
5. Temperature detection for input (6) and exit air (7)
6. Chart 'recording of air velocity, input air temperature, and exit
air temperature
7. Associated electronics with future microcomputer attachment
capacity
With these measuring capabilities, studies have been conducted on bed
load, baffling configuration, tablet shape, and spray-pattern configurations.
The types of coating systems can be evaluated analytically with
this setup. By measuring and being able to accurately control spray
rstes and conditions (spray angles, degrees of atomization, etc.), airflow
rate through the pan, air temperature, and dew point into and from the
pan it should be possible to optimize not only coat quality features, but
coating and energy efficiency as well. As energy costs have risen rapidly,
this latter consideration has become increasingly important. The
optimization of the parameters involved with aqueous and sugar coating
Improved Tablet Production System Design 51
are currently being appropriately modeled and simulated, with the projected
results being verified in the new installation.
Completely automated coating systems are now available for the control
of both sugar- and film-coated products. An example of such a system is
the Compu-Coat II. This is the second generation of a computerized coating
system based on a new controller. This controller provides real time
monitoring and control and documentation of the coating pan, air handler,
and spray system while allowing complete back-up by an independent manual
system. This system is available for use with the Accela-Cota pans in
sizes of 24, 48, 60, and 66 in. The spray system is capable of handling
sugar and film coatings. It is possible to network up to eight coating systems
with a supervisor interface station, and to expand to 24 pans using a
production monitoring unit. The software of the Compu-Coat system may
control the coating formulations to be used, control process parameters,
and respond to alarm conditions. As indicated earlier, the system is of a
modular design and is adaptable to multipan use [25,26].
D. Increasing Role of Computer Process Control
Evolution and Growth
The digital computer control of processes, especially processes in the chemical
industries, have been studied and used for approximately 25 years.
By today's standards the original installations were slow and very space
consuming. However, several successful computer-controlled processes
were developed.
The evolution of the computer in process control systems can be seen
in Figure 31. In the early stages of process control (Fig. 31a) , instrumentation
monitoring of a process, input (I) devices, such as meters,
lights, recorders, etc., were mounted on a panel board to provide visual
and sound input to an operator. The operator could control the process
by manually activating devices on the panel, output (0) devices, such as
switches and valves. Process control record keeping was also manually
performed by the operator(s).
With the development of the computer. process control configurations
as shown in Figure 3lb. were used. At this point, computers were very
large and very expensive. The computers were used to enhance the monitoring
and alarm functions built into the panel board. Some predictive,
calculation, and report-writing work was also done by the computer.
However, the control of the process was still manually done by the operator.
As computer technology improved, the size and cost of the computer
decreased and actual computer control of the process was first initiated
in this configuration by having the computer monitor specific set
points on the panel board instrumentation. With the computer exerting
some of the control functions, the inputs required for the operator and
the manual activation required of the operator were reduced. Computergenerated
records also increased.
The initial successes in computer applications to process control did
not lead to a rapid growth in the computer process control field. Unfortunately,
a lack of understanding of the necessity of a good project
design. the size and cost of the first computers. and the lack of adequate
I/O devices to link the process with the computer caused a great many
frustrations on the part of both computer customers and computer suppliers.
As a result, the larger computer companies concentrated their
52
Manual ..
Records
(al
Peck. Anderson, and Banker
Ponel
Boord
000 0 -lnput/Outpu1 -Process
0000
Manuel ..
Records
[b]
Manual
Records..
Ie)
Ponel Input/Output- Process
Boord
000000
0000
o 000
~IComputerl
t
Auto Records
Panel
Board
o000 ---Ieomruter 1=lnpuf/Oufput-Process
0000
Auto Records
Visual ->.
X J ?~ Icom~w 1='"PUt/Output-Process
~ . [d] Auto Records
VislJOl <>:
X
Au'o
lei
/ [E] == InputlOufpuf - Process I
~_.......... ICompu'er 1---[E] == Input/Output-Process 2
A .R ~ rp:t = Input/Output - Process :3
uto ecords lCJ
Figu re 31 Process control systems.
Improved Tablet Production System Design 53
efforts in the data-processing field, which requires little I/O equipment.
and computer process control development was relegated to secondary
importance.
At the outset of computer control design the first emphasis was to
make the process fit the available computer devices. The "force-fit" approach
resulted in computers being advertised and sold to the management
function of companies who in turn then gave the computers to the engineering
function to find an application. The computer purchase was generally
done without an analysis of work to be done followed by the design
of the best system. With this type of computerization approach failures
occurred and the frustration between supplier and customer developed and
added to the technical hurdles hampering the development of computer control
of processes [25).
Interest in computer control of processes began to grow when the system-
engineering approach to control design was used. In developing control
designs, engineering groups began working with computer companies in
an effort to design control devices to fit the processes as opposed to
forcing processes to fit the devices. At the same time, major analog instrument
manufacturers began to manufacture total I/O packages of computer
hardware and software systems, enabling their products to be used
for process control. Companies also began to develop the more powerful
conversational software packages that enabled the process control engineers
to use the languages without extensive professional programming aid.
Another prime factor in the growing interest in computer control is the
current revolution in microelectronic technology [26]. In the early years
of process control, the designs called for the use of electromechanical devices
such as relays, timers, and solenoids in conjunction with analog instruments
such as temperature- and pressure-monitoring devices. All of
these components were mounted on large panel boards. The systems were
rarely fUlly automatic and significant operator interaction was required.
especially if the process went "out-of-spec."
At times, adding on to the control of a process was difficult simply because
there was no more room on the panel board. As with any electromechanical
system, wear and tear, dirt, and heat can take their toll if the
devices are not carefully and frequently maintained [25).
Figure 31 illustrates the improved configuration for computer process
control utilizing the increased capability of computers, their reduced size
and cost, and the increased availability of I/O control devices. Because of
the increased amount of control the computer was able to do. the amount of
reports generated by the computer also increased. The panel board was
essentially kept for visual display only; the operator received very little
input, was required to take very little manual action, and made few written
reports.
The development of solid-state electronic devices at first improved the
reliabUity of process control by eliminating mechanical devices. As development
continued, extensive flexibility in the process control could be developed
into the controlling devices. In addition, since 1960 the cost of a
computer divided by its computing power has dropped by a factor of more
than 100 [27].
As computer technology developed, the devices for interfacing the operator
and the computer developed to the point of eliminating the panel
board (Fig. 3lD). The televisionlike cathode ray tube (CRT) monitoring
console expands the visual displays available to the operator, and what
54 Peck, Anderson, and Banker
Figure 32 Basic devices required for a process control system.
little process control action is required by the operator can be made at the
CRT monitoring console. The manual action and report writing by the operator
are eliminated.
Basic Process Control Devices
The hardware required in computer process control can be divided into
five basic components, and is illustrated schematically in Figure 32.
SENSORS, First of all. there must be devices to measure the variables
of the process which must be considered in making control decisions.
These devices are called sensors and are needed because a process computer
cannot deal directly with the variables. For example. a computer
cannot measure temperature directly. Therefore. a sensor must measure
the temperature and convert (transduce) that information into an electrical
signal for electronic use [28, 29] •
I/O DEVICES. The second basic device for computer process control
is an I/O device. This device transforms the signals from the sensors
into signals that are usable by the computer. In other words, the I/O
device enables the computer to interface with the outside world. The I/O
device must also be used to take signals from the computer and convert
them into action via devices and instruments controlling the process.
COMPUTER. Programmable Controller, The third device is the computer
itself. If the process to be controlled consists of numerous sequential
(logical) steps, then the controlling device can be a first level
computer device called a logic controller. A logic controller can replace
the older relay-type control. As long as the process is purely logical J
the controller will monitor the process step-by-step to the end and then
go back and run the process over again in the exact same way. The programmable
controller is the modern logic controller, making the purchase
of relay logic systems nonadvantageous. Even though relay controllers
are reliable, every change in the process has to be wired into the system
by hand. With the modern controllers. a programming change in the controller
itself can handle a process change [30].
Improved Tablet Production System Design 55
Microcomputer. Unlike the logic control. which is a simple onloff type
of control, the next level of control requires that measurements be taken
during the process, such as pressures and temperatures, that an algorithm
take those measurements and make some sort of decision, and that
corrective action be implemented. This type of control is referred to as
feedback control and is normally the system used in computer control systems.
The computer device that can first be used in this type of situation
is the microcomputer. The capacity of a microcomputer is such that a
device can handle six to 12 feedback loops per process cycle.
Microcomputer systems are not very flexible unless the design engineers
learn the particular machine language. Very few microcomputers
operate in the higher level languages, such as FORTRAN, and microcomputers
generally do not operate with a universal computer language.
The basic language of each machine must be learned.
The flexibility of microcomputers is limited. If 10 processes require
control, then 10 microcomputers in the plant could control them. However,
the reality is that with time there could be four or five different microcomputer
systems controlling the 10 processes, and therefore four or five
different computer languages in use. The different languages add to the
confusion When a process change takes place because consideration must
now also be given to the compatibility of the languages in use.
The cost of the actual programmable controller and microcomputer devices
is similar, but the microcomputer will cost more in terms of time to
program.
Minicomputer. The number of feedback loops and the speed of the response
times required in the loops determine the size of the computing
machine needed. These factors determine when a control system design
will require the minicomputer. Short response times and numerous loops
require a minicomputer.
The distinction between microcomputers and minicomputers is very difficult.
There are microcomputers that can outperform a minicomputer.
ACTIVATORS. The fourth basic component of a computer process control
system is a set of process activators. Activator devices, such as electric
solenoids. electric motors, and hydraulic valves, take the command
signals from the computer via the I/O device and turn them into action in
the process [28].
PERIPHERALS. The last basic set of process control components allows
reports to be printed and information stored, information to be added to
the control process or the process changed, and the system to be visually
monitored. These devices are called peripherals, and the term comprises
such items as disk storage devices, typewriters, printers, card readers,
keyboards. and CRT screens.
Distributed Process Control Systems
DEVELOPMENT. The development of the microcomputer has made it
possible to place the 110 monitoring and controlling device close to the
point of use, which has allowed for a basic redesign of process control
systems (see Fig. 3lE). The control system can now be more sophisticated.
Plant optimization and management information can be handled by a "host"
56 Peck, Anderson, and Banker
microcomputer while simultaneously monitoring the process-controlling microcomputers
distributed throughout the entire process. This type of design
of computer process control is called distributed control.
Distributed control systems reduce installation costs by reducing the
amount of field wiring that must be done in the panel systems. The shorter
cable runs of the distributed systems also help eliminate the electrical
interferences that often distort low-voltage electrical signals. The distributed
system replaces the large panel board in favor of the much smaller
CRT display. Because of the low cost of microcomputers, each remote unit
can have an on-line backup unit which will continue the control function
should the first microcomputer fail, thus increasing the reliability of the
distributed system over the older systems. The repair of a distributed
microcomputer is relatively easy. In many cases, a device can be repaired
by the simple replacement of a printed circuit board [25,31].
To change to distributed systems has not come without criticism. Many
people feel that the distributed systems reduce the number of personnel
needed to monitor a control system below a safe limit and that in a situation
of plant upset, several operators working simultaneously on a panel
board are better able to correct the situation than one person on a single
CRT • In addition, many people feel that it is a necessity that a panel
board showing the schematic of the process be used in the monitoring of
that process [31].
CONFIGURATIONS. There are two basic approaches to designing a
distributed control system: (1) the loop approach, and (2) the unit operation
approach. In the loop approach, the individual computer is programmed
to control a specific number of selected functions within the
process. Similarly. a single computer could be placed in control of a
single function in several process loops. With the unit operation approach,
the computer is in control of a single unit operation. The unit operation
approach has the advantage of being inherently modular and the control of
a process can be instituted in a stepwise manner. In other words, the
most critical operation of a process could possibly be controlled by computer
first and the rest of the process added to the control system as
time and money allow.
The host minicomputer and the microcomputers it is monitoring can be
arranged in anyone of a number of operational configurations. Figure 33
illustrates some of the configurations used in distributed process control.
In each configuration. if the host computer fails, the microcomputers can
continue their control functions. If any of the microcomputers fail, the
others can continue to communicate to the host computer and continue
their control functions [31].
The major obstacles to computer process control today are
1. The continued need for smaller computer devices with high
capabilities
2. Better I/O interfacing devices
3. Better operator Imachine interfacing devices
Process Control Organizations
Along with the development of computer process control design. hardware,
and software has come the development of several important related organizations
that have enhanced. and can continue to enhance, its development:
Improved Tablet Production System Design
(1) the control engineering functions in manufacturing companies, (2)
major instrument- and equipment-manufacturing companies able to supply
both equipment and related control devices, (3) process control groups
within the major computer manufacturers, and (4) large engineering contractors
specializing in the design and installation of computer control
systems (Table 5) [26].
Application to Pharmaceutical Processing
Pharmaceutical manufacturing generally operates in a batch configuration
with a series or sequence of steps. Most batch operations have a large
number of simple steps that require the assistance of manufacturing personnel,
and that assistance is not generally overly challenging. The
automatic control of the sequential operation can in many situations improve
the following areas:
1. Product throughput time
2. Consistency of acceptable product batch-to-batch
3. Compliance with OSHA and EPA regulations
4. Conservation of energy
(a] Point to Point
(bl Multiple Drop or Data Bus
Ie) Hierarchical
57
(dl
[e] Star
Figure 33 Distributed process control configurations.
58 Peck. Anderson, and Banker
Table 5 Suppliers of Computer Systems for Process Control
Bailey Controls
29801 Euclid Avenue
Wickliffe; OH 44092
Fisher Controls Co.
205 South Center
Marshalltown, IA 50158
Foxboro Company
120 Norfolk Street
Foxboro, MA 02035
Honeywell, Inc.
Process Management Systems Div.
2222 W. Peoria Avenue
Phoenix, AZ 85029
Kaye Instruments
15 DeAngelo Drive
Bedford; MA 01730
Reliance Electric Company
Control Systems Division
350 West Wilson Bridge Road
Worthington, OH 43085
Taylor Instrument Company
P.O. Box 110
Rochester. NY 14692
5. Reduction of labor
6. Flexibility to change the batch process control as the process
changes
7. Stable plant operation with reduced operational errors and rejected
product
8. Reliable control of the process
From the above list of advantages of automatic control it is obvious
that the control system must (1) provide good performance characteristics,
(2) maintain that performance over a long period of time. and (3) be easy
to install, maintain, troublashoot , and repair [30,32].
With the development of the microcromputer and distributed control
design, several manufacturers of equipment that are used in the pharmaceutical
industry are supplying microcomputers and controllers with their
equipment. Not only does some of the equipment increase the efficiency
of a process as described earlier, but the automatic controller increases
the reliability of the operative control of the equipment and increases the
ability of the machine to be interfaced into a computer controlled process.
The individual control of a specific piece of equipment also offers the
ability of automating a single-unit operation without automating the entire
process. Table 6 gives a list of suppliers and their equipment that are
so controlled.
Implementation of Computer Control
NATURE OF THE COMPUTER CONTROL PROJECT. Reducing the frustrations
in implementing computer process control that were pointed out
earlier in this section requires a high level of cooperation and communication
between the various groups involved, especially between the function
that wants the control and the group charged with designing and installing
the control.
Improved Tablet Production System Design
Table 6 Computer-Controlled Pharmaceutical Processing Equipment
59
Processing equipment
Automatic particle inspection for ampuls and
vials
Automatic tablet checker
Automatic tablet-coating system
Automatic tablet sugar-coating module
Compu - Coat II coating system
Dataplus 2000 control system
Fette tableting presses
Fitzaire fluid-bed dryer
Flo-Coater fluid-bed spray granulator
Fluidaire fluid air dryer
Hata Rotary tableting presses
Hi-Coater tablet-coating system
Pharmakontroll rotary press monitor
Thomas tablet sentinel
Developing company
Eisai USA, Inc.
Fuji Electric Co., Ltd.
Graco, Inc.
Vector Corp.
Thomas Engineering
Glatt Air Techniques
Raymond Automation
Fitpatrick Co.
Vector Corp.
Fluidaire
Elizabeth-Hata
Vector Corp.
Emil Korsch oHG. Berlin
Thomas Engineering
The initial request or interest for some level of automatic process control
will probably come from a research and development, production, or
production support function within a company which thinks that a product
can be produced better and/or quicker if computer control of the process
were instituted.
Some engineering function of the company will then have the responsibility
for putting the desires of the production request into operation.
The engineering function responsibility is to work with the process
group to supply the mechanics of accomplishing the task required by the
process group. The responsibility of the process group is to supply the
engineering function with enough information about the process and the
influences of the process on the quality of the product so that the engineering
function can make the best decisions in securing the appropriate
mechanics to accomplish the task.
THE IMPORTANCE OF COMMUNICATION. At the very outset of the
control project design it is essential that communications be definite and
clear. The first thing to realize is that in most companies the production
and engineering functions speak different "languages ;"
The language problem stems from the fact that both the processing
group and the engineering group use terms unique to their functions.
The language probem can become worse when the topic of computers and
the associated technologies are discussed. The vocabulary of data
processing and process control is developing as fast as the technology it
60 Peck, Anderson. and Banker
describes. The processing group should be aware of the engineering
group1s understanding of processing terminology. Likewise, the engineering
group should be sensitive to a possible lack of understanding of engineering
terminology within the processing group. Differences in termin010gy
familiarity between groups should be minimized by education or communication
style so that the communication between groups during the
project can take place on a common level of understanding and be fully
productive.
A short glossary of computer terms [33] has been included (Table 7)
to exemplify the language problem that can exist and perhaps initiate a
familiarization of "computerese. If
The process group must first specify. in detail, how the process requiring
change is now being run. The engineering function wID reqgire
such information as the equipment presently being used, the number of
persons involved in the process, the exact details of each process step.
and the effects of process steps on the final product quality.
The next information required is how the process needs to be
changed. The process group does not need to indicate how the change
wID be accomplished, but should spell out in detail what changes are required
in the old process.
Because of the data-processing capabilities of a computer, reports
can be generated during the process that can be used to tell the process
operator what is going on in the process. Data from such reports can be
used for quality control records and other data could be used to make decisions.
To eliminate the frustrations of future reprogramming, project
cost overruns, specifying a computer with inadequate memory capacity,
creating unhappy groups that cannot get the reports they want, the specifics
of the reports to be generated during the process must be decided
before the engineering group can specify the mechanics of the process
control system. Such things as the type, number. frequency, and mode
(pictures, listings, hard copy, black and white CRT. color CRT) of the
reports need to be considered in deta:il and specified.
The report that deta:ils all of the process control objectives and requirements
is sometimes referred to as the functional specification (funet
spec). To reinforce the need for detail in writing this document. a good
rule-of-thumb is to ask for every statement made: Can anyone assume
anything in what has just been written? If assumptions can be made,
then the statement is not detailed enough. If assumptions can be made,
then the engineering function is forced to make decisions about a process
that they probably know very little about, especially as it pertains to
final product quality. Therefore, the people charged with writing the
funct spec must take great pains to detail the process so that engineering
does not make an erroneous assumption and jeopardize the quality of the
final product.
With the funct spec in hand the engineering function can gain an understanding
of the process and what is wanted, and can proceed to determine
what equipment and technology is available or has to be invented
to accomplish the goal established. The engineering function will also
place a price tag on the project. If the cost of the control project is too
great, the funct spec may be rewritten to scale down the size of the
project and, hopefully, the cost.
The communication of the "needs" and the costs of satisfying those
needs are often continued until the exact needs of the process group are
Improved Tablet Production System Design
Table 7 Glossary of Computer Terms
61
Access time
Address
Architecture
Batch processing
Binary
Bit
Byte:
Core memory
CPU
CRT display
Data
Diagnostics
Digital circuits
Disk memory
Downtime
EDP
Floppy disk
The time needed to retrieve information from the
computer.
The number indicating where specific information
is stored in the computer's memory
A purely conceptual term signifying the fundamental
design aspects of a computer system.
Literally, a batch of programs or data which has
been accumulated in advance and is processed
during a later computer run.
A number system based solely on ones and zeros
(base 2).
The smallest information unit that a computer can
recognize; 1 or O.
A byte is an arrangement of 8 bits.
Using small magnetic doughnuts, or cores, as its
storage element, core memory preceded semiconductor
memory in the evolution of computer technology
and is now largely obsolete.
(Central processing unit) The part of the computer
that controls the interpretation and execution
of the processing instructions.
(Cathode ray tube) A televisionlike screen which
may be used for viewing data while they are being
entered into or retrieved from a computer.
The raw information within a computer system.
Programs for detecting and isolating a malfunction
or mistake in the computer system; features that
allow systems or equipment to self-test for flaws.
Electronic circuits providing only two values as
an output, one or zero, and nothing else.
Memory using rotating plates on which to store
data and programs.
The time during which a computer is not operating
because of malfunctions, as opposed to the
time during which it is functionally available but
idle.
(Electronic data processing) The transformation
of raw data into useful data by electronic equipment;
sometimes referred to as ADP, or automatic
data processing.
A component similar to a 45 rpm record made of
flexible material and used for storing computer
data.
Table 7 (Continued)
GIGO (Garbage in, garbage out) A shorthand expression
meaning that if you put in incorrect or
sloppy data, you will get incorrect or sloppy
answers; originally contrived to establish that
errors are always the user's fault, never the
computer's.
Hard copy Computer output recorded in permanent form,
such as the paper copy produced by a printer.
Hardware The physical components of the computerprocessing
system, for example, mechanical,
magnetic, electrical, or electronic devices.
Ie (Integrated circuit) A digital electronic circuit
or combination of circuits deposited on Semiconductor
material; the physical basis of a computer's
intelligence and the key to the microelectronic
revolution in computers.
Input The data that is entered into the computer; the
act of entering data.
Instruction A group of bits that designates a specific computer
operation.
Intelligent terminal
Interface
K
Language
Language, BASIC
Language, COBOL
Language, FORTRAN
Language, Machine
Mainframe
Memory
A typewriter with a television screen attached to
it that can remember things, do calculations. type
letters, communicate with other intelligent terminals.
draw pictures. and even make decisions.
The juncture at which two computer entities meet
and interact with each other; the process of
causing two computer entities to intersect.
Equals 1024 bytes.
A set of words and the rules for their use that
is understood by the computer and operator alike
for programming use.
An easy-to-use high-level programming language
that is popular with the personal computer systems
(Beginners All-Purpose Symbolic Instruction
Code) •
A high-level programming language widely used
in business applications (Common Business
Oriented Language).
A high -level programming language designed to
facnitate scientific and engineering problem
solving (Formula Translation).
The language. patterns of ones and zeros. directly
intelligible to a computer.
A large computer. as compared to minicomputers
and microcomputers.
The selection of the computer Where instructions
and data are stored; synonymous with storage.
Improved Tablet Production System Design
Table 7 (Continued)
63
Memory cap acity
Microcomputer
Minicomputer
Nibble
Output
Peripheral
Program
Realtime
Semiconductor
Software
Storage
System
Terminal
Throughput
Timesharing
Turnaround time
Word
Word length
Word processor
The maximum number of storage positions (bytes)
in a computer memory.
A small computer in which the CPU is an integrated
circuit deposited on a silicon chip.
A computer that is usually larger, more powerful,
and costlier than a microcomputer but is not
comparable to a mainframe in terms of productivity
and range of functions.
A nibble is an arrangement of 4 bits.
The information generated by the computer.
A device, for example, a CRT or printer, used
for storing data. entering it into or retrieving it
from the computer system.
A set of coded instructions directing a computer
to perform a particular function.
The fmmedlate processing of data as it is entered
into the oomputer-,
A material having an electrical conductivity between
that of a metal and an insulator; it is used
in the manufacture of solid state devices such as
diodes. transistors, and the complex integrated
circuits that make up computer digital circuits.
A general term for computer programs, procedural
rules and, sometimes, the documentation
involved in the operation of a computer.
See Memory.
The computer and all its related components.
A peripheral device through which information is
entered into or extracted from the computer.
A measure of the amount of work that can be accomplished
by a computer during a given period
of time.
The method that allows for many operators using
separate terminals to use a computer simultaneously.
The time between the initiation of a computer job
and its completion.
A group of bits that the computer treats as a
single unit.
The number of bits in a computer word.
An automatic typewriter, often equipped with a
video display screen. on which it is possible to
edit and rearrange text before having it printed.
64 Peck, Anderson. and Banker
refined to within budget constraints. The costs of the new control system
must be justified by the improvements brought about by the new
process.
At this point in the project, the hardest and the most time-consuming
part has been completed. The remainder of the project, including specifying
and ordering the actual control devices, writing and checking the
computer control programs, and installing and checking the system, becomes,
in comparison. a minor part of the project providing the initial
background work has been done properly [31,32].
ROLE OF THE INDUSTRIAL PHARMACIST. In the introduction to this
chapter reference was made to the industrial pharmacist/scientist being
aware of certain areas from which process efficiencies could be developed
and, with such an awareness, being able to function with the engineering
group within a pharmaceutical firm to help improve the manufacturing
processes. As the development of computer process control grows and is
applied to pharmaceutical processes. the role of the pharmacist grows even
larger. The unique awareness of the impact and influence of a manufacturing
process on the safety and efficacy of a final drug dosage form is
the sole reponsibility of the industrial pharmacist. In the age of computer
control the industrial pharmacist will have to know in detail how each unit
operation in the manufacture of a drug product can influence the final
drug product and what parameters within a unit operation are important
and should be monitored and controlled to produce reliable drugs. The
indu strial pharmacist will play an important role in helping to develop the
increasing role of computer process control in the pharmaceutical industry.
IV. VALIDATiON
Validation as a concept and a system of procedures came into being in the
pharmaceutical industry to ensure the sternity of parenteral and other
atezile products. Basically, the system involved placing sensors and
measuring devices in all the critical locations or recording sites within
sterilizers to accurately record temperature (or other conditions impacting
on the attainment of sterility) so that following a sterilization cycle with a
particular product. in a particular container, under full production conditions,
examination of the records of the sterilization conditions would provide
absolute assurance that sternization was achieved in every product
unit so processed. Validation requires previously establishing, in such
sterilization applications. the permutations and combinations of conditions
(time, temperature, steam pressure, cycle conditions. ete .) which will
produce sternity with a very high degree of certainty. Thereafter. recorded
data from each production sternization run provide the documentation
that all units within all regions of the sterilizer were exposed to conditions
that have clearly been established and documented by careful previous
experimentation to produce sterility.
Currently. validation concepts are being recognized for their value in
helping to assure the quality features of classes of drug products other
than sterile products. Validation procedures are being applied to solid
oral dosage forms and especially to compressed tablets. Validation is
most important in tablet products in which manufacturing process conditions
are known to have the potential to affect important product quality
Improved Tablet ProducUon System Design 65
features, such as bioavailability. Thus, in the design of a new tablet
product and its method of manufacture, or as the design of process for
the manufacture of tablets is improved and evaluated, it will be necessary
to consider the concept of validation. It is increasingly recognized that
an adequate process for the pharmaceutical product is a validated process.
A definition currently in use by the Food and Drug Administration is:
"A validated manufacturing process is one which has been proved to do
what it is purported or represented to do. The proof of validation is obtained
through the collection or evaluation of data preferably from the
process and the developmental phase and continuing through into the production
phase" [34]. While the FDA definition is for a validated manufacturing
process, it suggests by reference to a developmental phase that
validation involves more than the actual production process. Indeed this
would be so for most tablet products.
Inherent in any discussion of validation is the assumption that the
product involved is well understood and that the essential design parameters
are under control. Validation, therefore, is not indep endent of product
formulation. It is not enough to start thinking about validation when
a formulation is turned over to a pilot plant or production group. It is
essential that the development pharmacist or pharmaceutical scientist be
knowledgeable not only in formulation and physical pharmacy, but also in
pharmaceutical processing and manufacturing operations. Otherwise an
experimental formula may be developed whereby the qualrty features of the
product are very sensitive to the manufacturing conditions and variables
routinely and normally encountered in practice, or involves such a large
number of variables impacting on assorted quality features that validation
becomes difficult or impossible.
There are pharmaceutical products currently being marketed that cannot
be Validated based on their design. Included in this group are some
sustained-release products, which involve a sequence of steps whereby
powdered drug is dusted on sugar beads with other mixed hydrophilic and
hydrophobic powders, employing a "binder" of polymers, fats, waxes, and
possibly shellac in a mixed solvent (sometimes including water plus organic
solvents), followed by addition of a retard coating, again using fats,
waxes, and polymers in varying stages of solution and dispersion. In
such systems, simply controlling rates of powder addition, Iiquid binder,
addition, spray-gun conditions, and geometry in the coating pan, or
other processing conditions will not lead to validation. When the binder
and retard dispersions vary as to extent of polymer, fat, wax hydration!
solvation in each batch, and are not and cannot be controlled (a formulation
factor), no degree of process manufacturing control can produce a
controlled product. Furthermore, as the number of variables which can
influence product quality increase (both process and formulation variables),
the number of interaction effects capable of producing further (and often
less predictable) effects increase in an exponential manner. Validation of
such systems becomes a true nightmare.
Another design parameter that must be studied in development is the
establishment of raw materials specifications. If raw materials specifications
are inadequate, no amount of process validation can ensure a satisfactory
product. Thus, an important responsibility of the development
group is to determine that the physical and chemical specifications for all
active and excipient materials have been adequately set. For example,
where polymers are employed, have molecular weight, degree of SUbstitution
66 Peck, Anderson, and Banker
of chemical groups, and polymer grade, including purity, been adequately
defined? Does the nature of the product or method of manufacture require
special specifications for the drug or excipients (such as a particular
particle size or particle size distribution for drugs Which are to be
directly compressed or which have dissolution or bioavailability problems)?
Are special precautions required for a drug having chemical instability or
reactivity potential (such as lower moisture content limits for excipients to
be combined with a hydrolyzable drug)? Unless such formulation or drugrelated
variables, which can by themselves influence product quality features
or do so by interacting with processing variables, are identified and
appropriate specification limits set in place, no level of subsequent process
control and validation can lead to a reproducible product.
Two approaches may be taken to determine the specifications that are
adequate for raw materials after the need for such critical specifications
has been identified. The first approach is to set the most rigid specification
far that property that could reasonably be achieved. The advantage
of this approach is that it is direct, reasonably simple. and requires minimal
experimentation. The disadvantage is that this approach may represent
overkill, and unnecessary costs or undue limitation of suppliers. The
second approach involves conducting well-designed experiments to determine
where the critical specification(s) should be set. Vary the specification
in question for the raw material, employing the most rigid value that
could reasonably be achieved as one specification, relax that specification
for another sample of the material to a value that might be expected to be
borderline. and use an intermediate specification for perhaps a third
sample of the material. Make product of each raw material, employing
manufacturing conditions in each case that would be expected to impact
least favorably with the critical product quality features being examined.
Review the data from the experiments and determine whether the most
rigid specification is actually necessary, or whether the intermediate specification
will suffice, or whether the least rigid specification is adequate.
This approach is most useful when the development pharmacist has an accurate
understanding of the product system in question and its ultimate,
exact method of manufacture, and can project the range of processing
variables the product may be subjected to and the effect of the direction
of such process variations on the quality feature in question. Ideally,
raw materials specifications will be appropriately set before a product is
placed in production for process validation. When appropriate standards
have not been established for raw materials specifications, and where such
variation in raw materials is significantly impacting on product quality features,
attempts to achieve process validation through control of the manufacturing
operation can become an exercise in futility.
Experienced development and production personnel of most companies
can recite case studies in which a new product could be made on a small
scale in the laboratory but could not be reliably reproduced in production.
A different type of frustration is expressed, depending on whether the account
is told by the development person or by the production person.
Production personnel have a powerful argument at their disposal against
accepting any product and/or manufacturing procedure that defies or does
not reasonably lend itself to Validation. It is imperative that R&D and
production groups work together to see that validation can be and is
achieved.
Improved Tablet Production System Design 67
As the validation concept is applied to pharmaceutical products of increasing
complexity of design and manufacture. it is very clear that validation
must be initiated at a very early stage in the design of any product.
not only from the traditional "process" orientation, but also from a "product"
orientation. Based on this philosophy. validation may be defined as a
systematic approach to ensuring product quality, by identifying process
variables that influence product qualtty features. in order to establish
processing methods and the necessary control of these methods, which
when followed will assure meeting all product quality specifications after assuming
that the formulation is reasonable and that the raw material specifications
are adequate.
A. Validation Priorities
Perhaps the initial step in validation is to consider which product most requires
a full validation treatment and which least. Validation efforts could
be centered on processes that involve a company's largest selling and most
commercially important products. Products containing drug(s) with a very
low dose. chemical instability. low solubility or poor wetting properties.
which can lead to content uniformity, stability. or bioavailability problems;
products with a history of production problems. often bordering on being
unacceptable. or which unaccountably produce occasional lots that are unacceptable
or borderline; sustained or controlled release products whose
processes are not under control should have a high validation priority. In
other words, the initial validation effort could be focused to provide the
greatest beneficial effect.
Another approach to validation priority setting is to review the documentation
and data currently available on present products and start validating
those products which can be done the quickest and easiest. For
example. a product that has a moderate dose (perhaps 25 to 150 or 200
mg) of a very soluble. stable drug. which is readily compressible and has
no bioavailability or other problems. and which has been made for years
by direct compression. with no history of a defective batch. while readily
meeting all product specifications. might be easily validated with historic
data. This approach can be advantageous in that validation concepts are
learned with the easier projects before more complex validations are
undertaken.
B. Steps in Validation
Step 1
Identify all the quality features and specifications which will or do define
product quanty and acceptability. final product as well as in-process
specifications (see Table 8 for examples of in-process product specifications
as well as process control parameters).
The first step in validation is thus to be certain that specifications
have been appropriately set to define the primary product quality features
and to establish their limitS of acceptability. Specifications may be set on
intermediate materials as well as on the finished product. For example. a
variety of specifications may be placed on the powder mixture or granulation.
in addition to the specifications which are placed on the final tablet
68 Peck. Anderson. and Banker
Table 8 Unit Operations and Possible Control Tests Associated with the
Manufacture of a Wet-Granulated Compressed Tablet
Process variable
Screening
Mixing
Granulating
Drying (oven)
Size reduction
Compression
Purpose
Ensure a set particle
size of raw material
Homogeneous mixture
Convert powders to
granules having
suitable flow and
compressive properties
Reduce moisture content
to proper level for compression
Size reduction
Manufacture of
compressed tablet
Test parameters
1- Mesh analysis
2. Bulk volume
1. Mesh analysis
2. Chemical content
uniformity analysis
3. Yield reconciliations
1- Weight per subpart
2. Amount of granulating
agent per subpart
3. Endpoint of batch
4. Mixing time
5. Granule size (wet)
1- Weight per tray
2. Thickness per tray
3. Relative humidity of
air
4. Velocity of air
5. Time
6. Loss on drying
7. Yield reconciliation
1. Bulk volume
2. Flow
3. Mesh analysis
4. Yield reconciliation
5. Loss on drying
1- Weight/10 tablets
2. Hardness
3. Thickness
4. Appearance
5. Disintegration
6. Dissolution
7. Weight variation
8. Content uniformity
9. Friability
10. Compressional force
II. Tablets per minute
12. Yield reconciliation
13. I\arl Fischer/loss on
dr-ying
Improved Tablet Production System Design 69
product. It is important that the list of specifications be sufficiently full
and complete to assure full eompendial , new drug application (NDA), or
company standards. Not all product specifications come under validation
treatment. Some product specifications bear on purity and generally are
not influenced by processing; they bear on raw material quality and
specifications. Other quality features may not be sensitive to processing
condition variations and thus may not be included in the validation treatment
or are of minor concern in such treatment. An example might be a
tablet which is formulated such that compressional force. under any projected
range of compressive loads, has no significant influence on tablet
disintegration or drug dissolution rate.
Quality features for tablets have been addressed throughout the chapters
in this series of books and will not be repeated here, other than to
note that they should usually include, as a minimum for an oral uncoated
tablet, weight variation, content uniformity. hardness, friability, disintegration-
dissolution specifications, and size uniformity.
Identification of critical product quality features for tablet products
must be treated as a separate undertaking for each and every product.
No standard or master approach exists. For example, not all tablet products
have fr.ability problems. In fact. the great majority do not, and
friability may then be discounted as a feature in the Validation program.
Rate of drug dissolution. as it relates to the rate and completeness of bioavailability,
and assurance of consistency of bioavailability from lot to lot
may be of great concern for some tablet products and of little or minor
concern for others.
Dissolution might reasonably be expected to be of great concern in situations
of potential bioavailability. such as with a very slightly soluble
acidic drug moiety which has an intermediate to large dose and is only well
absorbed in the upper gastrointestinal tract. In such a case, great attention
must be paid to such factors as methods of wet massing. how binders
are incorporated in the product, wet massing times. or the granulation
method employed. rate of compression, consistency of feeding granulation
to the dies, compressive force, particle size and particle size control,
methods of incorporating disintegrants, and other factors depending on the
method of tablet manufacture. On the other hand. the assured achievement
of content uniformity in such a large-dose product, through an extensive
validation program. might be totally unnecessary or of minimal
concern.
The concerns of product qualtty might be just the reverse with a low
dose of a highly soluble amine drug which is well absorbed along the
gastrointestinal tract. Bioavailability assurance through validation might
be completely unnecessary and pointless. but in comparison to the acidic
high-dose drug previously described, content uniformity may now become
the primary concern and focus of the validation efforts. In other situations.
validation of both bioavailability and content uniformity goals might
be warranted and in some situations neither of these quality features is
of major concern because they are readily achieved under all conceivable
permutations and combinations of processing conditions that could be imagined
for the product.
The situations given above are intended to point out that the critical
quality features are matters to be identified for each and every product.
These critical quality features may depend on the pharmacokinetic properties
of the drug (including but not limited to where and how the drug
70 Peck, Anderson, and Banker
is absorbed). physical chemical properties of the drug, drug dose, and
chemical stability of the drug.
Step 2
Review product performance against the proposed specifications. After describing
the quality features which must be achieved in the product and
the minimum acceptable values for each such feature, the specifications
developed for that product should be reviewed and agreed upon as being
appropriate and reasonable by knowledgeable people in research and development.
production. and quality assurance.
Another point that might be reviewed in this step is whether or not
the product in question has sets of specifications which constitute possibly
critical ncompeting objectives. n NelU'ly all tablet products do have competing
objectives. For example, if during the manufacture of a hard, nonfriable,
nondusting uncoated tablet these quality features arc enhanced.
then the rapid tablet disintegration and drug dissolution release rate quality
features would be expected to degrade and even become unacceptable.
Another type of competing objectives circumstance is encountered with
insoluble antacids. Most insoluble antacids exist as very fine powders
which have very poor compressibility. Their effectiveness is related to
their fine particle size and large surface. which is directly related to their
ability to neutralize gastric acid. Any steps which are taken to facilitate
compression by enhancing particle consolidation may be expected to reduce
the activity of such antacids.
Thus, processing procedures such as precompression, increased punch
dwell time, increased pressure, or procedures used to increase granule
bonding cannot benefit both sets of quality features simultaneously because
these processing effects will produce competing effects on the two sets of
quality features, no matter how the processing effects are controlled.
Fortunately. through proper formulation such competing objectives are not
difficult to handle for most drugs. However. when the occasional drug
comes along with such traits as a large dose, low solubility, poor wetting,
borderline bioavailability, and poor compression properties, the fact that
one is dealing with competing objectives cannot be ignored, and should be
carefully considered in the design of the validation protocol.
Step 3
Identify the methodology, processes. and pieces of equipment that are or
will be used in the manufacture of the product(s).
Once the product quality and acceptability standards have been established
in steps 1 and 2, the process that manufactures the product must
be thoroughly identified. If a tablet is made by direct compression, the
process wUl differ significantly in the unit operations involved as compared
with wet granulation. Not only must the processes that impact directly on
the manufacture of the product be identified but those systems that support
the process and thereby have an indirect impact on the process must
be identified. Table 9 lists some of the associated systems that support
tablet production and could have an impact on tablet validation considerations.
Improved Tablet Production System Design
Table 9 Support Systems Associated with
the Manufacture of Pharmaceutical Solid
Dosage Forms
Heating, ventilation, air conditioning (HVAC)
Periodic maintenance systems
Water systems
Compressed gas systems (air)
Cleaning systems
Vacuum systems
Electrical
Drainage
Dispensing of raw material
Paperwork systems
Personnel training
Computer process control
Special environmental protection systems
71
Step 4
Identify the potentially relevant and critical process variables. The information
base for this step in the total product validation is the list of
specifications and limits for the quality features which have been placed on
the product. together with the basic manufacturing process. formulation,
and raw materials specifications which were previously set. A knowledgeable
development pharmacist can usually look at the product specifications
and manufacturing method, and predict the process variables and process
steps which will have the greatest potential for impacting on the particular
specifications for the product. The processing steps and control of these
steps which are expected to impact most directly on primary product objectives.
such as bioavailability or drug dissolution rate, should receive
the closest attention. Examples of the types of processing variables which
should be considered for particular product specifications are binder concentration,
granule density, optimal mixing time, mill settings, and others. .
Step 5
Conduct the process validation experiments. Before validation experiments
can be performed. the various pieces of equipment used in the different
unit operations of the process must be certified to operate as they were
designed to. In addition, any auxiliary equipment operation-monitoring
72 Peck, Anderson, and Banker
instruments such as ammeters, wattmeters, rpm indicators, chart recorders,
and temperature indicators must be in working condition and calibrated.
While equipment certification should be a responsibility of a company's engineering
function, those involved in the validation project must be sure
that the certification is performed to ensure that the equipment used in the
validation studies will function reliably.
The next step is to undertake well-designed and controlled experiments
to establish the influence of the processing variables and combinations of
variables on the quality features of the product. A variety of approaches
can be taken, from full mathematical/statistical optimization, with the system
actually being modeled, to a relatively few determinations for a drug
having no bioavailability or manufacturing problems. Validation in the
latter case may simply involve the documentation of current process controls
and a historic review of the product's acceptance test records.
When the validation project becomes more complex. such as where competing
objectives exist, or where cause and effect relationships are not
obvious. the following general approach is suggested in order to develop a
validation protocol.
1. Establish preliminary limits on each processing procedure, above
and below which you never expect to go in practice. In some cases it will
be possible to set an exact single specification for a processing procedure.
such as a given mixing time. In other cases a range will be required and
it should be set as noted above. A range should be used when an exact
condition cannot be achieved, as in the volumes of granulating agent used.
wet-massing times, tablet machine and coating pan speeds, coating weights.
temperatures. or airflow rates. When such processing conditions are expected
to impact on quality features. and where exact control is not feasible.
project or determine the maximum range that might be reasonably expected
to be encountered in practice (remember, future production runs must be
inside these ranges or the product will be "outside validation II) •
2. Examine the listing of limits from the preceding step in light of the
most critical product quality features you are to achieve. Select the
processing variables and combination of variables along with their respective
limits which you would expect to impact least favorably on these quality
features and make the product under those conditions. Examine to see if
any extremes of the above processing conditions or combination of extremes
of processing conditions place the product outside the acceptance specifications
for any parameter.
Step 6
Review. monitor, and revalidate as needed. Once a process has been validated
it is necessary to continuously monitor the parameters controlling the
process. If instruments are used they must be calibrated on a routine
basis. Equipment must receive periodic maintenance so that performance
does not change. Care must also be taken that no part of the process is
changed. A process change. improvement or otherwise, must be revalidated.
Constant review of all specifications, product and process, is also
necessary to enable the specifications to be "tightened1I or "broadened" to
enhance the total validated system.
Improved Tablet Production System Design
ACKNOWLEDCMENT
Special acknowledgment to Mr. Ken Main t P.E . t Division Electrical Bngineer,
Aluminum Company of America, Alcoa, Tennessee, for his help and
insight in the development of the computer process control section.
APPENDIX: LIST OF MANUFACTURERS
73
Mixer/G ranulator
Drydispenser
Jaygo, Inc.
199 Seventh Avenue
Hawthorne, NJ 07506
Diosna
Dierks and Sohne
45 Osnabruck
Sandbachstrasse 1
West Germany
Lodige Mixer/Granulator
MGT Vertical Mixer/Granulator
Littleford Brothers, Inc.
15 Empire Drive
Florence, Kentucky 41042
Fielder Mixer/Granulator
Raymond Automation Co.
508 Westport Avenue
Norwalk, Connecticut 06856
Gral Mixer /Granulator
Maines Collette, Inc.
P.O. Box 818
Wheeling, IL 60090
Fluid-Bed Cranulator/Dryer
Aeromatic
198 Route 206 South
Somerville, NJ 08876
Glatt
Glatt Air Techniques, Inc.
260 West Broadway
New York. NY 10013
Freund
Vector Corp.
675 44th Street
Marion, Iowa 52302
DOUble-Cone and Twin-Shell
Blenders with liquid feed and
vacuum drying capabilities
Double Cone Rota-Cone
Paul D, Abbe
146 Center Avenue
Little Falls, NJ 07424
Twin Shell Processor
The Patterson-Kelley ce., Inc.
Process Equipment Division
East Stroudsburg, PA 18301
Nauta Mixer
Day Mixing Co.
Cincinnati, Ohio 45212
L/D Series Blender
Gemco, Inc.
301 Smalley Avenue
Middlesex t NJ 08846
Specialized Granulation Equipment
Roto Granulator
Key International , Inc.
480 Route 9
Englishtown, NJ 07726
Marumerizer
Luwa Corporation
4433 Chesapeake Drive
P.O. Box 16348
Charlotte, NC 28216
74
REFERENCES
Peck, Anderson, and Banker
1. Sigman, D. A., Drug Cos. lrui, , 125( 5) : 84 (1979).
2. Beckley, J. N. and Bathgate, T. A., Drug Cos. Ind., 125( 4) : 54
(1979) .
3. Schuesaler, O. P., Pharmaceutical Plant Design in Today's Regulatory
Climate. Paper presented before the Pharmaceutical Processing Symposium,
American Institute of Chemical Engineers, 88th National Meeting.
Philadelphia, June 10, 1980.
4. Morrissey. J. T., Renovation of a Pharmaceutical Facility. Paper presented
before the Pharmaceutical Processing Symposium, American
Institute of Chemical Engineers, 88th National Meeting, Philadelphia,
June 10, 1980.
5. Snyder, H., Automated Productfon Procedures in Drug Manufacturing.
Paper presented before the First Annual Conference on QUality Programs
and Government Regulations, Arlington, Virginia, April 12,
1976, Automated Process Controls and GMPs Workshop.
6. Greene, J. H., Productfon Control Systems and Decisions. Irwin,
Homewood. Illinois, 1965, p. 257.
7. Dickinson, J., Pharm. Technol., 2(2):41 (1978).
8. Snyder, H. T., Battista, J. V., and Michelson, J., Development and
Technology of a Semi-Automated Continuous Extrusion Process for the
Preparatfon of a Tablet Granulation. Paper presented before the Industrial
Pharmacy Section, 113th Annual Meeting, American Pharmaceutical
Association, Dallas, April 15 - 29, 1966.
9. Kamman. J. P., Pharmaceutical Plant Design for GMP and Process
Efficiency. Paper presented before the Pharmaceutical Processing
Symposium, Annual Meeting, American Institute of Chemical Engineers,
New York, November 17, 1977.
10. Vosnek, K. J. and Forbes, R. A., Diosna-Granulating with Wet
Particle Size Control. Paper presented before the 69th Annual Meeting
of the American Institute of Chemical Engineers, Chicago,
December 2, 1976.
11. Thurn, U., Soliva, M., and Speiser, P., Drugs Made in Germany, 14:
12 (1971).
12. Ormos, Z. and Hung, J., Ind. Chern., 1:207 (1972).
13. Ormos, Z., Patald , K., Osukas, B., and Hung, J., Ind. Chern.. 1:
307 (1973).
14. Davies, W. L. and Gloor, Vi. T., J. Pharm• sa., 80: 1869 (1971).
15. Roufller , M., Gurny, R•• and Doelker, E., Pharm. Acta Technol.,
21(2) (1975).
16. Shevlin. E. J., Pharm. Technol., 2( 3) : 56 - 59 (1978).
17. Schwartz, J. B., Drug Dev. Ind. Phar•• 14(14): 2071- 2090 (1988).
18. Korsch PH-BODe Automated Tablet Production System Bulletin, Korsch
Tableting, Ine , , Somerville, New Jersey, 1989.
19. Kristensen, H. G. and Schaefer. T., Drug Dev; Ind. Phar., 13(4,5):
803 -872 (1987).
20. Hunter, B. M. and Ganderton, D.• J. Pharm. Pharmacol., 25(Suppl.):
71P - 78P (1983).
21. Lindberg, N. 0., Leander. L., Wenngren. L., Heigesen, H•• and
Reenstierna, B., Acta Pharm. Suecica, 11:603-620 (1974).
Improved Tablet Production System Design 75
22. Leuenberger, H., Bier, H. P., and Sucker, H. B., Pharm. Technol.,
3( 6) :61- 68 (1979).
23. Leuenberger, Hot Pnarm, Acta Helv., 57:72-82 (1982).
24. Ritala, M., Jungensen, 0., Holm. P., Schaefer, T., and Kristensen,
H. Go; Drug Dev. Ind. Phar., 12(11-13) :1685-1700 (1986).
25. Compu-Coat II Bulletin, Thomas Engineering, Inc•• Hoffman Estates,
Illinois, 1986.
26. Jean Y. LeFloc'h, Automation and Networling of Coating Systems,
Thomas Engineering, Inc; , Hoffman Estates, Illinois, April 1987.
27. Ritala , M., Holm. P., Schaefer, T., and Kristensen, H. G., Drug
Dev, Ind. Phar., 14(8): 1041-1060 (1988).
28. Travers, D. N., Rogerson, A. G•• and Jones, T. M., J. Pharm.
Pharmacol" 27(Suppl.): 3P (1975).
29. Lindberg, N. 0., Leander, L., and Reenstierna, B., Drug Dev; Ind.
Pharm., 8: 775 -782 (1982).
30. Gharrta , S. R., Srinivas, R., and Rhodes, C. T., Drug Dev, Ind.
Pharm" 10:305-311 (1984).
31. Lindberg, N. O. and Jonsson, C., Drug uev, Ind. Pharm., 9:
959 - 970 (1983).
32. Timko, R. J., Barrett, J. S., McHugh, P. A., Chen, S. T., and
Rosenberg, X. X., Drug Dev, Ind. Pharm., 13(3):405-435 (1987).
33. Kay, D. and Record, P. C., Mfg. Chemist, 48(9):45-46 (1978).
34. Staniforth, J. N. and Quincey, S. M., Int. J. Pharm.• 32:
177 -185 (1986).
35. Staniforth, J. Nq Walker, S., and Flanders, P q Int. J. Pharm.,
32: 277 - 280 (1986).
36. Spring, M. S., Drug oe«. Ind. Pharm., 9:1507-1512 (1983).
37. Fry, W. C., Stagner, W. C., and Wichman, K. C., J. Pharm. sci.,
73: 420 - 421 (1984).
38. Fry, W. C., Stanger, W. C., Wu, P. P., and Wichman, K. C.,
Computer-Interfaced Capacitive Sensor for Monitoring the Granulation
Process. I: Granulation Monitor Design and Application (C. E.
Capes, ed.), Proceedings of the 4th International Symposium on
Agglomeration, June 2 - 5, Toronto. Ontario, Canada, Book Crafters,
Chelsea, Michigan, 1985, pp , 497 - 501.
39. Fry, W. C., Wu, P. P., Wichman, K. C., and Stanger, W. C.,
Granulation Monitoring. In Proceedings of Pharm , Tech. Conference
'85. Aster Publishing Corp., Springfield, Oregon, 1985, pp.
358-367.
40. Fry, W. C., Wu, P. P., Stanger, W. C., Wichman, K. C., and
Anderson, N. R., Pharm. Technol., 11(10): 30 (1987).
41. Alleva, D. S. and Schwartz, J. B., Drug Dev. Ind. Phar., 12(4)~
471- 487 (1986).
42. Holm, P., Drug Dev; Ind. Phar., 13(9-11):1675-1701 (1987).
43. Pemon, J. P. and Schwartz, J. B., Drug Dev, Ind. Phar., 13(1):
1-14 (1987).
44. Chowhan, Z. T., Ptiarm, Technol., 12(2):26-44 (1988).
45. El-Gindy, N. A., Samaha, M. W., and EI-Maradny, H. A., Drug Dev,
Ind. Phar., 14( 7): 977-1005 (1988).
46. Korsch, W., Pharm. Technol., 5( 4) : 62 (1981).
47. Lachman, L. and Cooper, J., J. Pharm. Sci. I 52: 490 (1963).
76 Peck, Anderson, and Banker
48. Berfield, D., Pharm. Technol., 1(40):53 (1977).
49. Krause, G. and Iorio. T., J. Pharm. sa., 57:1223 (1968).
50. Wurster. D., U.S. Patent 2,648,609.1953.
51. Wurster, D., U. S. Patent 2,799,241, 1957.
52. Wurster, D., J. Am. Pharm. Assoc., Sci. ea., 48:451 (1959).
53. Bulletin 303-873, Graco, Inc., 1977.
54. Skrokov, M. R. (ed.). In Mini- and Microcomputer Control in
Industrial Processes. Van Nostrand Reinhold. New York, 1980,
pp. 1-58.
55. Noyce. R. N•• Sci. Am•• 237(3):63 (1977).
56. Terman, L. H., Sci. Am., 237(3):163 (1977).
57. Hunter, R. P. Automated Process Control Systems Concepts and
Hardware. Prentice-Hall, Englewood Cliffs, New Jersey. 1978.
pp. 15-110. 305-335.
58. Foster. D. Automation in Practice. McGraw-Hill, London, 1968,
pp. 48-54.
59. Leitner, M. and Cocheo, M. S. In Mini- and Microcomputer Control
in Industrial Processes (M. R. Skrokov, ed.). Van Nostrand
Reinhold, New York, 1980, pp. 128-163.
60. Thor. M. G. In Mini- and Microcomputer Control tn Industrial
Processes (M. R. Skrokov, edv) , Van Nostrand Reinhold, New York,
1980, pp. 212- 219.
61. Copeland, J. R. Microcomputers and Programmable Controllers for
Process Control. Center for Professional Advancement, East
Brunswick, New Jersey. 1979.
62. Renner, K. M., Pharm. Technol•• 9(12):24 (1985).
63. Gopal, C., Pharm. Technol•• 13(4):20 (1989).
64. Grosswirth, M., Output. I( 1) : 33 (1980).
65. Byers. T. E., Manufacturing Process Validation: A Systematic Approach
to Quality Assurance. Paper presented at the Management
Conference for the Pharmaceutical Industry, Purdue University, West
Lafayette, Indiana, September 13-15. 1978.
2
Coating of Pharmaceutical
Solid-Dosage Forms
Stuart C. Porter and Charles H. Bruno
Coloreon, West Point, Pennsylvania
I. INTRODUCTION
The application of coatings to the surface of pharmaceutical solid-dosage
forms. especially tablets. has been practiced for over 150 years.
Although such a process is often applied to a dosage form that is
functionally complete I and thus may cause US to reflect on the need for
incurring the additional expense. it is evident that the continued use of
coating processes in pharmaceutical production remains very popular.
Such popularity relates to the many benefits obtained when a dosage form
is coated, which include:
Improved esthetic qualities of the product
Masking of unpleasant taste and odor
Enabling the product to be more easily swallowed by the patient
Facilitating handling. particularly in high-speed filling Ipackaging
lines
Improving product stability
Modifying drug-release characteristics
Over the course of time. coating processes have developed from the
art of earlier years to those that are more technologically advanced and
controlled such that compliance with good manufacturing practices (GMPs)
is facilitated. The design of new equipment, the development of new
coating materials. and the recognition of the impact of applied coatings
on subsequent release of drug from the dosage form have all contributed
to improved products.
77
78 Porter and Bruno
Changes that have occurred in coating processes reflect a desire to;
Consistently obtain a finished product of high and reproducible
quality
Achieve processes in which the economics are maximized, particularly
with respect to process times and equipment utilization
While the methods for applying coatings to solid-dosage forms are
varied. and some are described elsewhere in this book, this chapter will
focus on the processes of sugar coating and film coating. and will discuss
these processes with respect to;
Raw materials
Application techniques
Potential problems
Available coating equipment
II. SUCAR COATINC
A. Introduction
The process of su gar coating, which has its origms in the confectionery
industry. is perhaps one of the oldest pharmaceutical processes still in
existence.
Although in recent years modernization of the process with respect to
panning equipment and automation has taken place, sugar coating is still
considered to be more of an art rather than a science.
While methods (and materials) for coatings date back over 1000 years
(early Islam makes reference to pill coatings based on mucillage of
psyllium seeds), the current pharmaceutical process of sugar coating
originated in the middle of the nineteenth century when sugar as a raw
material became plentiful, and the forerunner of modern panning equipment
was invented.
Although the tendency is to produce pharmaceutical coating pans from
stainless steel, early pans were made from copper because drying was effected
by means of an externally applied heat source. Current thinking,
even with conventional pans, is to dry the coated tablets with a supply
of heated air and to extract the moisture and dust-laden air from the
vicinity of the pan.
Although the sugar-coating process has experienced declining popUlarity
in the United States, it is still retained by many companies worldwide,
since many advantages can be realized, including:
Raw materials are inexpensive and readily available
Raw materials are widely accepted with few regulatory problems (with
the exception of perhaps colors)
Inexpensive, simple equipment can be used
Sugar-coated products are esthetically pleasing and have wide consumer
acceptability
The process is generally not as critical (as film coating) and recovery
(or rework) procedures are more readily accomplished
Coating of Pharmaceutical Solid-Dosage Forms
However, in spite of the relative simplicity of the sugar-coating
process, it does have some potential shortcomings, for example:
79
The size and weight of the finished product results in increased packaging
and shipping costs
The brittleness of the coatings renders the coated tablets susceptible
to potential damage if mishandled
The achievement of high esthetic quality often requires the services of
highly skilled coating operators
The final gloss is achieved by a polishing step which can make imprinting
difficult
The inherent complexity (from the standpoint of the variety of procedures
and formulations used) of the process makes automation
more difficult
In spite of these difficulties, many companies have made excellent use
of modern process technology (including automation), so that the requirements
of current GMPs, including documentation, are readily achieved with
a high degree of reproducibility in the quality and performance of the
finished product.
B. Raw Materials Used in Sugar Coating
As expected, the major ingredient used is sugar (sucrose), although this
may be substituted by other materials (such as sorbitol) for low caloriel
diabetic products (typically in the candy industry). Sugar-coating formulations
are for the most part aqueous.
The sugar-coating process consists of various steps, each designed to
achieve a particular function. Consequently, a variety of additives may
be incorporated into each type of formulation. Examples of such additives
are:
Fillers (calcium carbonate, talc, titanium dioxide)
Colorants (dyes, aluminum lakes, iron oxides, titanium dioxide)
Film formers (acacia, gelatin, cellulose derivatives)
Antiadhesives (talc)
Flavors
Surfactants (as wetting agents and dispersion aids)
Although most of the coating formulations used in the sugar-coating
process are applied as liquids, some (e.g., dusting powders) are applied
dry.
A typical sugar-coating process encompasses five stages:
1. Sealing
2. Subeoatdng
3. Grossing
4. Color coating
5. Polishing
While each of these stages is varied, the common feature throughout
is that the process requires repeated applications of coating liquid, each
80 Porter and Bruno
application followed by a period during which the tablets are allowed to
tumble freely to allow complete distribution of the coating materials. and
finally. a drying period when moisture is removed from the coating prior
to the next application.
Sealing
Most of the coating formulations used in the sugar-coating process are
aqueous. whereas tablet cores are typically porous. highly absorbent. and
formulated to disintegrate rapidly when they make contact with water.
Consequently. if these cores are not appropriately protected at the outset.
ultimate product stability (both physical and chemical) can be seriously
compromised. The purpose of sealing is to offer this initial protection.
and to prevent some tablet core ingredients from migrating into the coating.
and ultimately spoiling the appearance of the final product.
Sealing is accomplished by the application of a polymer-based coating
(either by ladle or spray techniques) to the surfaces of the tablet cores.
Examples of polymers that might be used include shellac, zein. hydroxypropyl
methylcellulose (HPMC), polyvinyl acetate phthalate (PVAP), and
cellulose acetate phthalate (CAP). These are typically dissolved (at a
15-30% w/w concentration) in an appropriate organic solvent, preferably
one of the denatured ethanol products.
While use of shellac has been universal, this polymer can cause problems.
One problem results from the fact that shellac can polymerize on
storage. causing the solUbility characteristics of the coating to change.
This problem can either be minimized by incorporating PVP into the shellac
formulation [1] or by using one of the other. more stable polymers (such
as PVAP).
When using any of the water-insolUble polymers as the basis for a sealcoat
formulation. it is important to apply only the minimum quantity of coating
needed to give the appropriate protection; otherwise drug-release
characteristics may well be affected.
When the seal coat is applied by a ladle technique, detackifiers, such
as talc, are often used to minimize the risk of "twinning" or clumping.
Overzealous use of talc should be avoided. however, otherwise it might be
difficult for the subsequent sugar coat to bond to the surface of the seal
coat.
Finally. if the final product is to have enteric properties, this result
is usually achieved by using one of the enteric polymers (such as PVAP
or CAP) as the basis for the seal coat and ensuring that sufficient coating
material is applied.
Subcoating
SUbcoating is the first major step of the sugar-coating process and provides
the means for rounding off the tablet edges and building up the
core weight. It also provides the foundation for the remainder of the
sugar-coating process. with any weakness in the final sugar coat often
being attributable to weaknesses in the subcoat.
In order to facilitate this buildup. subcoating formulations almost always
contain high levels of fillers such as talc. calcium carbonate. calcium
sulfate, kaolin, and titanium dioxide. In addition, auxiliary film
formers such as acacia, gelatin, or one of the cellulose derivatives may
Coating of Pharmaceutical Solid-Dosage Forms 81
-
Coat~ng
Coating
II'
Figure 1 Schematic of examples of acceptable tablet-core shapes for
sugar coating.
also be included in order to improve the structural integrity of the
coating.
It is important during subcoattng to get effective coverage of the
coating material over the tablet corners and on the edges if a quality result
is to be achieved. To this end. selection of appropriate tablet
shapes is important. Certainly. tablet shapes which minimize the corners
(such as tablets compacted on deep concave punches or dual radius
punches). as shown in Figure 1, can aid in effective coverage. Additionally,
it is necessary to minimize tablet edge thickness. otherwise
twinning will be more prevalent and incomplete edge coverage (by the
coating) is likely to occur (Fig. 2).
Core
can causeJ!}
Incomplete
Edge Coverage
~ can cause
Twinning
Figure 2 Schematic example of poor tablet core shape for sugar coating.
(The twinning problem illustrated is more prevalent when using capsuleshaped
tablets .)
82 Porter and Bruno
Table 1 Examples of Formulations Used in the Lamination Subeoating
Process
Binder solutions
Dusting powders
Gelatin
Gum acacia
Sucrose
Distilled water
Calcium carbonate
Titanium dioxide
Talc (asbestos free)
Sucrose (powdered)
Gum acacia
I
3.3% w/w
8.7
55.3
32.7
40.0% w/w
5.0
25.0
28.0
2.0
II
6.0% w/w
8.0
45.0
41.0
1.0
61.0
38.0
Two main approaches to the process of subcoating are often practiced,
depending on whether a lamination technique or a suspension subcoat
formulation is used. Each has its distinct features and advantages.
LAMINATION PROCESS. The lamination process is perhaps the older
of the two techniques used, and involves alternate applications of binder
solutions and dusting powder until the required level of coating is achieved.
While materials and formulations for binder solutions and dusting powders
are varied. some typical formulations are shown in Table 1.
When using the lamination technique it is important to ensure that a
careful balance is achieved between the relative amounts of binder solution
and dusting powders used. Underutilization of dusting powders increases
the risk of sticking and twinning, whereas overdusting can create tablets
that have brittle coatings.
While achievement of quality results with the lamination process typically
requires employment of skilled operators. there is no doubt that this
type of process can permit rapid buildup of the coating.
On the downside, the lamination process can be messy, more difficult
to use by less-skilled operators, and more difficult to automate.
SUSPENSION SUBCOAT INC PROCESS. In simple terms, suspension
subcoating formulations result from combining the binder and powder
formulations used in the more traditional lamination process. Examples of
a typical formulation are shown in Table 2.
Use of the suspension subcoatfng approach reduces the complexity of
the process, allowing it to be used effectively by less-experienced operators.
and ultimately facilitates automation of the process.
Coating of Pharmaceutical Solid-Dosage Forms
Table 2 Examples of Suspension Subcoating Formulations
I II
Sucrose 40.0% w/w 58.25% w/w
Calcium carbonate 20.0 18.45
Talc (asbestos free) 12.0
Titanium dioxide 1.0 1.00
Gum acacia 2.0
Gelatin (120 bloom) 0.01
Distilled water 25.0 22.29
83
Grossing (or Smoothing)
In order to manufacture a quality sugar-coated product, it is imperative
that the surface of the coating be smooth and free from irregularities
prior to application of the color coat.
While the requisite smoothness may be achieved during the application
of the subeoat , it is not unusual to find that further smoothing (prior to
color coating) is necessary. Depending on the degree of smoothing required,
the smoothing coating may simply consist of a 70% sucrose syrup,
often containing titanium dioxide as an opacifier/whitening agent, and possibly
tinted with other colorants to provide a good base for subsequent
application of the color coat.
If a substantial amount of smoothing is required, as in the case in
which the subcoat tablets have a pitted surface, other additives (such as
talc, calcium carbonate, corn starch) may be used in low concentrations to
hasten the smoothing process.
Color Coating
Many would agree that color coating is one of the most important steps in
the sugar-coating process because of the immediate visual impact that is
associated with overall quantv ,
Use of appropriate colorants, which are dissolved or dispersed in the
coating syrup, allows the desired color to be achieved. Two basic approaches
to coloring sugar-coating syrups exist, each giving rise to differing
coating techniques. These two approaches involve the use of
either water-soluble dyes or water-insoluble pigments.
Prior to the 1950s, soluble dyes were used extensively to achieve the
desired color. This technique was handled by an experienced eoater ,
Who had acquired his skill over many years of work experience. Much of
the color coating required 2 or 3 days, and unless handled properly. resulted
in tablets that were nonuniform in color or mottled, since the soluble
dye can migrate to the surface during drying. Additionally, color
84 Porter and Bruno
reproducibility from batch-to-batch was not predictable. and light sensitivity
with subsequent fading was also a problem when using dyes.
The use of insoluble, certified lakes has virtually replaced the soluble
dye in pharmaceutical tablet coating. Lakes have several advantages;
namely. color migration on drying is eliminated. since lakes are insoluble.
light stability is improved, mottled tablets are a rare occurrence. and coating
time is SUbstantially shortened. While lakes are insoluble. they are
not totally opaque. Consequently, coloring properties can be optimized by
combining lakes with opacifiers such as titanium dioxide. The most efficient
process for color coating involves the use of predtspersed , opacified
lake suspensions. By varying the ratios of lake and opacifier , various
shades can be produced.
DYE-COATING PROCESS. The features of a typical sugar-coating
process that utilizes water-soluble dyes as colorants include:
Sequential application of coating syrups containing specific dye concentrations
(typically, as coating progresses, dye concentrations in
the syrup may be increased until the target color is achieved)
Addition of a quantity of colored syrup (at each stage) that is sufficient
to just wet the total tablet surface, followed by gentle drying
to achieve requisite smoothness and prevent color migration
Employment of relatively low concentrations of colorant (necessary to
achieve final color uniformity). resulting in a requirement to make
anywhere up to 50 separate color syrup applications (particularly
for dark colors)
There is no doubt that in the hands of a skilled operator the quality
of sugar-coated tablets that employ the dye-coating method are difficult to
match (this is particularly true from the standpoint of "cleanliness." depth.
and "brflltance" of the final color).
However. such a process is not without its difficulties, namely:
Color migration problems (resulting from either underdrying or too
rapid a drying) are commonplace.
Color variability. across the surface of individual tablets, Which occurs
as the result of unevenness of the subeoat layer and transparency
of the color coat.
Tablet-to-tablet color variability which may result because the transparent
coloring system has not been uniformly distributed.
Batch-to-batch color variation which is likely to occur because of
variability in the total quantity of color applications made, or as a
result of small differences in amount of colorant weighed out for
each batch (water-soluble dyes produce very intense colors and a
little goes a long way).
The process is time consuming (because of the slow drying required
and the need to make so many individual color applications).
PIGMENT-COATING PROCESS. Pigments have demonstrable advantages
over water-soluble dyes. two important ones being:
1. Lack of solubility in aqueous media (which eliminates color migration
on drying)
2. Superior light stability
Coating of Pharmaceutical Solid-Dosage Forms 85
However, because pigments are discrete, insoluble particles, careful
attention must be paid to the pigment-dispersion process. Hence, the popularity
of commercially available pigment-dispersion concentrates.
Some of the major characteristics of sugar-coating formulations and
processes when pigment colorant systems are used include:
Use of a single-color concentration throughout the color-coating
process, thus making it easier to achieve the target end color (in
order to obtain a different color, it is necessary to vary the ratios
of the lake pigments with respect to the opacifier, titanium
dioxide)
Achievement of batch color uniformity after only a few applications of
colored syrup (often color development is complete after eight to
10 applications, and the remaining five to seven applications are
simply used to smooth off the tablet surface)
Reduced drying times resulting from the fact that the insoluble colorants
do not migrate on drying, and thus can be dried more
rapidly
Overall shortened color-coating process as a result of reduced number
of color applications and shortened drying times
One should, however, be aware of what some might construe to be disadvantages
with pigment coloring systems and the associated coating
process:
Colorants derived from pigments (especially when lakes are used in
combination with titanium dioxide) are generally not as bright or
clean-looking as those obtained with soluble colorants.
If the pigment color-coating process is rushed, it is relatively easy to
produce rough tablets that are difficult to polish.
There is a need to ensure that pigments are effectively dispersed in
the coating syrup (certainly, pigm ent color concentrates eliminate
this problem), otherwise color "specking" might be a problem.
Since most pigment coloring systems contain lakes (which are typically
acidic), it is inadvisable to keep coating systems hot for any
length of time once the color has been added; otherwise excessive
amounts of invert sugar will be formed.
With the exception of the first of these problems, all the others can
easily be avoided, and thus advantages of the pigment coating process
tend to prevail, making it the process of choice.
Summarizing these advantages I they are:
Greater ability to get a uniform color on the surface of each tablet
Greater batch-to-batch color uniformity
Significant reduction in thickness of the color coat
Significant reduction in processing time
Polishing (Glossing)
Since freshly color-coated tablets are typically dull (i.e. I they have a
matte surface finish). it is necessary to polish them in some way to
achieve the gloss that is typical of finished sugar-coated tablets.
86 Porter and Bruno
While methods to achieve a desirable gloss tend to vary considerably.
it is generally recommended that tablets should be trayed overnight (prior
to polishing) to ensure that they are sufficiently dry. Excessively high
moisture levels in tablets submitted for polishing will:
Make achievement of a good gloss difficult
Increase the risk of "blooming" and "sweating" over longer periods of
time
Glossing or polishing can be carried out in various types of equipment
(e. g.. canvas- or wax-lined pans), including that used for applying the
sugar coating itself (which is more typical in automated processes).
Polishing systems that may be used include:
Organic-solvent-based solutions of waxes (beeswax, carnauba wax,
candelilla wax)
Aleoholic slurries of waxes
Finely powdered mixtures of dry waxes
Pharmaceutical glazes (typically alcohol solutions of various forms of
shellac, often containing additional waxes)
Printing
If sugar-coated tablets are to be further identified with a product name,
dosage strength, or company name or logo, this has to be accomplished by
means of a printing process.
Typically, such printing involves the application of a pharmaceutical
branding ink to the coated tablet surface by means of a printing process
known as offset rotogravure.
Sugar-coated tablets may be printed either before or after polishing.
with each approach having its advantages and disadvantages. Printing
prior to polishing enables the ink to adhere more strongly to the tablet
surface, but any legend may subsequently be removed by either friction or
as a result of contact with organic solvents during the polishing process.
Printing after polishing avoids the problem of print rub-off during polishing,
but branding inks do not always adhere well to the waxed tablet surface.
Adhesion of printing inks can be enhanced by application (prior to
printing) of a modified shellac, preprint base solution.
c. Application Techniques
Application of sugar coatings to pharmaceutical tablets has long been considered
one that requires a significant amount of skill on the part of the
operator. While this philosophy has a lot of truth in it, and While it is
certainly difficult for untrained operators to achieve qUality results, the
employment of special techniques (such as the use of suspension subcoat
formulations and coatings colored with proprietary pigment dispersions)
makes quality results achievable even for less-skilled operators.
While many different types of coating formulation (Sec. II.B) will be
applied during the coating process, similarities in application exist for each
of them.
The basic application procedure in each case involves three steps in
sequence:
Coating of Pharmaceutical Solid-Dosage Forms
1. Application of an appropriate volume (sufficient to completely
cover the surface of every tablet in the batch) of coating liquid
to a cascading bed of tablets
2. Distribution of the coating liquid uniformly across the surface of
each tablet in the batch
3. Drying of the coating liquid once uniform distribution is achieved
87
Specific details of actual procedures adopted may vary from companyto-
company. However, the ultimate goal in each case is to ensure that
the coating is uniformly distributed throughout the batch. Although this
goal may be facilitated by the manner in which the coating liquid is applied,
and by manual stirring of the wet tablets to help eliminate "dead
spots" (regions in the tablet bed that are difficult to reach with the coating
liquid), the main mechanism for distribution of the coating liquid relates
to the shearing action that occurs as tablets cascade over one
another.
For the greatest period of time, sugar-coating liquids were applied
manually by allowing premeasured quantities of coating liquid to be poured
across the moving tablet bed. In recent times, there has been a greater
reliance on mechanical dosing techniques, involving the use of spray guns
or dosing "spaeges" (Fig. 3).
One of the major misconceptions concerning the use of mechanical
dosing techniques, particularly spray guns, is that they can exert a
major influence on uniformity of distribution of the sugar-coating liquid.
Again, it is important to emphasize that the main factor controlling distribution
of the coating liquid relates to contact between the cascading tablets,
and transfer of liquid from one tablet to another as the result of this contact.
ThUS, particularly When using spray guns, it is not necessary to
finely atomize the coating liquid in order to ensure effective distribution
of that liquid. Indeed, excessive atomization can cause "fogging!! where
much of the coating liquid can end up on the walls of the pan rather than
on the tablets. Consequently, many advocates for the use of spray guns
simply allow the liquid to stream from the nozzle. For this reason, use
of a device similar to that shown in Figure 3 can be equally effective and
less expensive than using spray guns.
Summarizing, since coating uniformity is achieved as the result of
tablet-to-tablet transfer of liquid coating material, it is not necessary for
each tablet to pass through the zone of application (which is a necessity
in the film-coating process). Factors which influence coating uniformity in
the sugar-coating process are that:
The coating material remains fluid until it is spread across the surface
of every tablet in the batch.
Sufficient volume of coating liquid is applied to ensure that every
tablet in the batch is capable of being wetted (thus liquid volumes
may have to be changed as the process progresses in order to
reflect changes in tablet size and drying conditions).
The coating pan exhibits good mixing characteristics, particularly so
that dead spots are avoided (many coating pans of conventional
design, i.e., the traditional pear-shaped design, may have to be
modified by inclusion of mixing baffles, otherwise mixing may have
to be augmented by manual stirring of the tablets by the operator).
88 Porter and Bruno
Figure 3 Figure showing a dosing sparge for sugar coating.
Coating of Pharmaceutical Solid-Dosage Forms
D. Problems in Sugar Coating
89
In any coating process, a variety of problems may arise. Often such problems
may be related to formulation issues that have been compounded by
those associated with processing.
Problems with Tablet Core Robustness
The attritional effects of any coating process on tablet cores is well
understood. Consequently. tablet cores must be sufficiently robust to
resist the stress to which they will be exposed during coating.
With this in mind; particular attention must be paid to important tablet
physical properties such as hardness (diametral crushing strength), friability,
and lamination tendency. Failure to address these issues is likely
to result in a situation in which tablet fragmentation occurs during the
coating process.
Tablet fragmentation is not only a problem from the standpoint that
the broken tablets will obviously not be saleable (and thus would have to
be inspected out). but additionally. the broken fragments may typically become
"glued" (because of the adhesive nature of the coating fluids) to the
surface of undamaged tablets (Fig. 4); thus spoiling a significant portion
of the batch.
Quality Problems with Finished Tablets
CHIPPING OF COATINGS. Sugar coatings are inherently brittle and
thus prone to chipping if mishandled. Addition of small quantities of
polymers (such as cellulosics, polyvinyl pyrrolidone, acacia, or gelatin) to
one or more of the various coating formulations often helps to improve
structural integrity. and thus reduces chipping problems.
Excessive use of insoluble f'lllers and pigments tends to increase the
brittleness of sugar coatings, and thus should be avoided where possible.
CRACKING OF THE COATING. Tablet cores that expand, either during
or after coating, are likely to cause the coating to crack (Fig. 5).
Such expansion may result from moisture absorption by the tablet core.
or may be caused by stress-relaxation of the core after compaction (a
phenomenon Which is known to occur, for example, with ibuprofen).
Moisture sorption can be minimized by appropriate use of a seal coat,
whereas expansion due to postcompaction stress relaxation can be resolved
by extending the time between the compaction event and commencement of
sugar coating.
NONDRYING COATINGS. Inability to dry sugar coatings properly,
especially those based on sucrose, is often an indicator that excessive
levels (greater than 5%) of invert sugar is present. Inversion of sucrose
is exacerbated by keeping sucrose syrups at elevated temperatures under
acidic conditions for extended periods of time. Such conditions occur
when sugar-coating solutions containing aluminum lakes are kept hot for
too long; or such sugar-coating formulations are constantly being reheated
to redissolve sugar that is beginning to crystallize out.
TWINNING (OR BUILDUP OF MULTIPLES). By their very nature,
sugar-coating formulations are very sticky, particularly as they begin to
90 Porter and Bruno
Figure 4 Figure showing how broken tablets can ruin a whole batch of
product in sugar coating.
Coating of Pharmaceutical Solid-Dosage Forms
Figure 5 Sugar-coated tablets with cracked coating.
91
dry, and allow adjacent tablets to stick together. Buildup of multiples
really becomes a problem when the tablets being coated have flat surfaces
(as shown in Fig. 2) which can easily come into contact with one another.
This can be particularly troublesome with high-dose, capsule-shaped
tablets that have high edge walls. Appropriate choice in tablet punch design
can be effectively used to minimize the problem.
UNEVEN COLOR. Because it has a major impact on final tablet appearance,
the color-coating stage of the sugar-coating process is critical
to ultimate tablet quality.
Uneven distribution of color. particularly with the darker colors, is
often viaually apparent. and thus a major cause of batch rejection. Many
factors may contribute to this type of problem, including:
Poor distribution of coating liquids during application. This may be
caused by poor mixing of tablets in the coating process, or failure
to add sufficient liquid to coat completely the surface of every
tablet in the batch.
Color migration of water-soluble dyes while the coating is drying.
Unevenness of the surface of the subcoat when using dye-colored
coatings. This unevenness causes a variation in thickness of the
transparent color layer that is perceived as different color
intensities.
"Washing back" of pigment-colored color coatings. While pigments do
not migrate on drying, if excessive quantities of coating liquid are
applied during the coloring process, there is a tendency for the
previously applied (and dried) color layers to be redissolved and
distributed nonuniformly; thus giving rise to nonuniform appearance.
This problem is. particularly noticeable for formulations predominantly
colored with aluminum lakes where the level of opacifying
pigments (such as titanium dioxide) is low (La .• dark colors).
Excessive drying between color applications. This can cause erosion
of the color layer and contributes to unevenness in the color coat.
92 Porter and Bruno
"BLOOMING" AND llSWEATING." Residual moisture (in f"mished sugar-.
coated tablets) can often be a problem. Over a period of time ~ this
moisture can diffuse out and affect the qUality of the product. Moderate
levels of is reached, below which there is a critical
cessation of molecular motion on the local scale. Under these temperature
conditions, the polymer exhibits many of the properties of inorganic
glasses, including toughness, hardness, stiffness, and brittleness. For
this reason, the glass-transition temperature is often described as one below
which a polymer is brittle, and above which it is flexible. This definition
is, at times, a little simplistic, and so a better definition of glasstransition
temperature is that temperature above which there is an increase
in the temperature coefficient of expansion [201.
Because the glass-transition temperatures of many of the polymers used
in film coating are in excess of the temperature conditions experienced in
the typical coating process, it is often necessary to modify the properties
of the polymer. This modification allows the final coating to better withstand
the conditions to which it will be subjected in the typical coating
process. An appropriate modification involves the process of plasticization.
Plasticizers reduce the glass-transition temperature of amorphous
polymers and impart flexibility. The basic requirements to be met by a
plasticizer are permanence and compatibility. Permanence dictates that the
plasticizer has a low vapor pressure and low diffusion rate within the
polymeric film, a requirement that favors high molecular weight plasticizers.
Compatibility, on the other hand, demands that the plasticizer be miscible
104 Porter and Bruno
with the polymer and exhibit similar intermolecular forces to those present
within the polymer.
It is not uncommon to see reference made to two types of plasticization.
The first is internal plasticization, and refers to the situation in which
chemical changes are made within the structure of the polymer itself, as
with copolymers (exemplified by many of the acrylic polymers systems used
in film coating). The second, termed external plasticization, occurs when
an external additive (the plasticizer) is combined in admixture with the
polymer.
Effective plasticization is critical when using aqueous polymeric dispersions
in order to ensure that sufficient free volume exists at normal
processing temperatures to facilitate coalescence of the polymeric particles
into a continuous film.
Since the plasticizer by its interaction with the polymer affects the
intermolecular bonding between polymer chains, it is only to be expected
that this additive will change the properties of the coating (Table 8). In
these cases, although well-defined quantitative effects can be demonstrated,
the magnitude of the effect is very much dependent on the compatibility,
or degree of interaction, of the plasticizer with the polymer. Methods for
determining the interaction between film-coating polymers and appropriate
plasticizers have been described by Sakellariou et ale [29] and Entwistle
and Rowe [22]. Plasticizers. by their very nature. are not universal,
since selection is very much determined by which polymer is being used.
COLORANTS. Colorants are included in many film-coating formulations
to improve the appearance and visual identification of the coated product.
Certain types of colorant (as will be discussed later in this section) can
provide other physical benefits.
As described under sugar coating (Sec. II). various types of approved
colorants exist that can be used in film coating. While a detailed review
of colorants that can be used in pharmaceutical dosage forms has been
Table 8 Effects of Plasticizers on the Properties of Film Coatings
Property Effect of increasing plasticizer concentration
l. Tensile strength
2. Elastic modulu s
3. Film adhesion
4. Solution viscosity
5. Film permeability
6. Glass-transition
temperature
Decreased [27]
Decreased
May be increased, but results often variable
[ 27]
Increased. and magnitude of effect dependent
on molecular weight of plasticizer [11]
Can be increased [27] or decreased [28], depending
on chemical nature of plasticizer
Decreased. but magnitude of effect dependent
on compatibility with polymer [29]
Coating of Pharmaceutical Solid-Dosage Forms 105
given elsewhere [30]. it is appropriate to briefly review here those types
that may be used in film coating. Such colorants include:
Water-soluble dyes (e.g•• FD&C Yellow #5 and FD&C Blue #2)
Aluminum lakes (of FD&C water-soluble dyes)
Other lakes (e.g•• D&C Red #6)
Inorganic pigments (e.g •• titanium dioxide, iron oxides. calcium sulfate,
calcium carbonate)
"Natural" colorants (e. g •• riboflavin, tumeric oleoresin. carmine 40)
Unlike sugar coating. film coating had its origins as a nonaqueous
process. Consequently, with few exceptions. most formulators preferred
to use pigments as colorants in their film-coating formulations. While the
more recent introduction of aqueous technology has opened the door to
using water-soluble colorants, use of pigments has persisted owing to the
advantages shown by this form of colorant. including:
Ability to increase solids content of coating solution without dramatically
affecting viscosity [11] (particularly advantageous in aqueous
film coating)
Ability to improve the moisture barrier properties of film coatings [27]
Pigments consist. however. of discrete particles, and thus a substantial
effort is required to ensure that they are well dispersed in the coating
liquid. Such dispersion will be inadequate if dry pigment is simply added
to the coating liquid with the aid of a low-shear stirrer (e. g •• a Lightnin'
mixer). Inadequate pigment dispersion leads to coating defects. as discussed
by Porter [31] and Rowe [32J. Particle size of the pigment within
the film (a parameter often related to efficiency in the dispersion process)
will also affect perceived color [32] and may influence the surface roughness
of the coating [33].
Since colorants are mostly added to a film-coating formulation for their
visual effects. it is important to understand the physical behavior of
colorants in a polymer film. As pigments are used almost exclusively as
colorants in film coatings. further reference in this section to colorants
will be restricted to pigments.
When coloring a film coating not only is it important to create a given
visual effect. but also to ensure that the appearance is as uniform as possible
throughout the particular batch of coated product. and consistent
from batch-to-batch.
Such uniformity is facilitated when the colorant chosen is able to effectively
mask the appearance of the substrate without requiring the use
of excessive quantities of colorant (which could increase the risk of
physical defects in the coated product) or applying excessive quantities of
coating (which impacts total cost of the process). This ability to mask
the substrate is often described in terms of hiding power or opacity of the
colored coating. These characteristics of the coating are often related to
contrast ratio. a term defined as the ratio of the Y tristmulus value for a
111m measured over a black background (i.e •• Yb) to the similar value
measured over a white background (Yw). Thus
Yb
Contrast ratio =Yw x 100
106 Porter and Bruno
Colored films having contrast ratios close to 100 exhibit excellent
hiding power, whereas those having values close to zero are almost transparent
(and thus have poor hiding power). Since contrast ratio is also
affected by film thickness, determinations (and comparisons) must always
be made on films of equal thickness.
Rowe [35] has listed the contrast ratios for coatings containing
various colorants. From his comparison, it is evident that the better results
are obtained for films containing certain inorganic pigments (e. g. ,
titanium dioxide and iron oxides). or those that absorb the higher wavelengths
of visible light (e. g., FDIC Blue #2). These results can be explained
by the theories of light, which predict that hiding power will be
influenced by:
Light reflected at the polymer/pigment interface, which is influenced
by differences in the respective refractive indices of the polymer
and pigment [35]
Quantity (and wavelength) of light absorbed by the colorant
Understanding these basic principles may help the formulator optimize
a particular coating formulation, especially with respect to the concentration
of colorant required, and the quantity of coating needed to develop a
uniform appearance. Some optimal results in this respect have been described
by Porter and Saraceni [15].
While it is important to understand the physical behavior of colorants,
one should not lose sight of the impact that pigments can have on other
physical characteristics of the coating. Some of these effects are listed in
Table 9.
SOLVENTS. While choice of an appropriate solvent deserves careful
attention, the rise in importance of aqueous film coating has virtually eliminated
solvent selection from the formulation process. Nonetheless, certain
types of film coatings and film-coating processes require that SOme organic
Table 9 Effects of Pigments on the Properties of Film Coatings
Property Effect of increasing pigment concentration
Decreased (effect may be minimized by effective
pigment dispersion)
Increased [36]
Little effect [27]
Increased I but not substantially [11]
Decreased [27], unless critical pigment volume
concentration is exceeded
Increased, but magnitude of effect dependent
on refractive index of pigment, and Ught
absorbed by pigment [34]
1. Tensile strength
2. Elastic modulus
3. Film adhesion
4. Solution viscosity
5. Film permeability
6. Hiding power
Coating of Pharmaceutical Solid-Dosage Forms 107
Table 10 Common Solvents Used in Film Coating
Class Examples
1. Water
2. Alcohols Methanol
Ethanol
Isopropanol
3. Esters Ethyl acetate
Ethyl lactate
4. Ketones Acetone
5. Chlorinated hydrocarbons Methylene chloride
1: 1: 1 Trichloroethane
solvents still be used. Thus, to a limited extent, selection of an appropriate
solvent may still be necessary.
A list of some of the more common solvents that have been used in
film coating is shown in Table 10.
When selecting a particular solvent or solvent blend there are several
factors that must be considered. The first prerequisite is the ability to
form a solution with the polymer of choice. In this respect, it is often
difficult to determine whether true solutions are formed or whether they
are mainly macromolecular "dispersions." Banker [37] stated that optimal
polymer solution will yield the maximum polymer chain extension, producing
films having the greatest cohesive strength and thus the best mechanical
properties.
One method for determining interactions between the polymer and solvent.
and aiding in selecting the most suitable solvent for a given polymer,
is to use the solubility parameter approach. This approach is based on
the theoretical treatment of the familiar free energy equation as proposed
by Hildebrand and Scott [38] and is expressed in this way:
L\H = V m m
= overall heat of mixing
=total volume of the mixture
= energy of vaporization of either component 1 or 2
ep = volume fraction of either component 1 or 2
L\H
m
V
m
L\E
where
108 Porter and Bruno
The expression (~E)IV is often termed the cohesive energy density,
written 02} where 0 is the solubility parameter and is equivalent to
([flE]/V)1 2 in the above equation. If 0 1 = O2, then flHm = O. Thus, in
the free-energy equation
flF =~H - T~S
m
where ~F is the free-energy change, T is absolute temperature, and ~S is
the entropy of mixing. The free energy is now dependent on the mixing
entropy. As there is a large increase in the entropy when a polymer dissolves,
the set of circumstances thus far described ensures that there will
be miscibility between the solvent and polymer.
Kent and Rowe [39} used the solubility parameter approach in evaluating
the solubility of ethylcellulose in various solvents for film coating.
They determined the intrinsic viscosities of several grades of ethylcellulose
in a range of solvents of known solubility parameters utilizing the equations
derived by Rubin and Wagner [40}. They graphically evaluated the
effect of solvent solu bility parameter on intrinsic viscosity for a range of
sol vents classified either as (1) poorly hydrogen bonded. (2) moderately
hydrogen bonded, or (3) strongly hydrogen bonded, and determined both
the best class of solvent to use and the optimum solvent solubility parameter.
ThUS, this approach can be used to determine the best solvent for
the polymer from a thermodynamic standpoint. The technique can also be
used for optimizing solvent blends, the individual components of which may
or may not themselves be thermodynamically good solvents.
Before dtssolution can take place the solvent must penetrate the
polymer mass. Once this has occurred, a swollen gel will form which rapidly
disintegrates to form a solution. The rate of solution is facilitated by
small solvent molecules which diffuse rapidly into the polymer mass. Unfortunately.
thermodynamically good solvents are not always kinetically
good. ones. and vice versa; thus, a compromise may be necessary.
An additional function of the solvent system is to ensure a controlled
deposition of the polymer onto the surface of the substrate. If a good coherent
and adherent film coat is to be obtained, the volatility of the solvent
system is an important factor. In the final formulation, the selected
solvent system usually represents a compromise between thermodynamic,
kinetic, and volatility factors and results in a solvent blend being used.
After all formulation factors have been resolved, attention must be
paid to the changes in solvent ratios that can occur during the application
process. This will cause no problem if all components of the blend, or the
least volatile component, are good solvents for the polymer. However, if
this is not the case, the solvent -polymer thermodynamic balance changes
as evaporation progresses. The polymer can thus be precipitated before a
cohesive film is formed. Alternatively, the solubility of the polymer in the
remaining solvent may not be sufficient to ensure that the optimum film
properties will be obtained. In this situation, it is essential to use a
constant-boiling or azeotropic solvent mixture whose composition does not
change on evaporation.
B. Conventional Film Coatings
The greatest area of application for film coatings is that where the coating
is mainly designed to improve product appearance, perhaps improve
Coating of Pharmaceutical Solid-Dosage Forms 109
stability and ease of ingestion of the dosage form, but not alter drugrelease
characteristics from that dosage form.
From this description, it is apparent that esthetics are of paramount
importance, and consequently are likely to influence selection of the raw
materials to be used in the formulation. This selection is often based on
factors that affect the mechanical properties (such as tensile strength,
elasticity, and adhesion) of the coating, allow the smoothest, glossiest
coatings to be obtained, and produce coatings that readily dissolve in the
human gastrointestinal tract.
Conventional film coating is also the area where aqueous technology
has gained the highest acceptance. Thus, with few exceptions, most ingredients
are selected for their solubility in water (this is especially true
for polymers).
Polymers
Common polymers used in conventional film coating are listed in Table 11.
The most popular class of polymers used in conventional film coating
are cellulosics, many of which have good organic-solvent and aqueous
solubility, thus facilitating the transition to aqueous film coating. Of
these cellulosics, ethylcellulose is not water soluble, and was originally
used as a film modifier in admixtures with hydroxypropyl methylcellulose
(HPMC) in organic-solvent-based formulations. To a limited extent, this
blending process has continued in aqueous film coating. In this case,
aqueous solutions of HPMC are mixed with aqueous dispersions of ethyIcellulose.
However, except when one needs to produce special barrier
Table 11 Polymers Used in Conventional Film-Coating
Formulations
Class
1. Cellulosics
2. Vinyls
3. Glycols
4. Acrylics
Examples
Hydroxypropyl methylcellulose
Hydroxypropylcellulose
Hydroxyethylcellulose
Methylhydroxyethylcellulose
Methylcellulose
Ethylcellulose
Sodium carboxymethylcellulose
Polyvinyl pyrrolidone
Polyethylene glycols
Dimethylaminoethyl methacrylatemethylacrylate
acid ester copolymer
Ethylacrylate-methylmethacrylate
copolymer
110
20
Porter and Bruno
O+-----.---,---,.----r----,
o 10 20 30 40 50
% (w/w) of EthylcelluJose in Polymer Mix
(a)
1.0
,,-..
o a.
o 0.8 ........
..J o
~
c
c:J
::>
a:
Cl
Iz
woa:
w
c,
TIME (hours)
Figure 28 Effect of scale of coating process (Wurster) on release of chlorpheniramine
from nonpareils coated with aqueous ethyl cellulose dispersion
(Burelease 10% by weight of coating applied).
Inlet air humidity
Nozzle location
Atomizing air pressure/volume
Spray rate
Coating liquid solids content
Metha [59] has described in detail the importance of many of these
variables and how they relate to the scale-up process. As a word of caution.
careful consideration (in the optimization process) should be given
to the scale on Which the scale-up process is based. Many times, initial
studies are conducted on small (0.5-1.0 kg) laboratory scales; unfortunately
this may not be an appropriate basis for predicting scale-up
factors. By way of example, the data shown in Figure 28 are indicative
of how feasibRity studies conducted in a 1-kg capacity Wurster may not
be entirely predictive of what is likely to happen on scale-up. In this
example, a more appropriate starting point is likely to be the 5-kg pilot
scale. where results obtained more closely match those obtained on the
50-kg scale.
Commercial Equipment
As discussed earlier, most fluid-bed equipment used for fHm coating was
based on the Wurster design, with many pharmaceutical companies using
custom-built equipment under a licensing agreement with WARF. However,
in the last 10- 15 years, several manu facturers of commercial fluidbed
equipment have adapted their designs to fuliill the needs of the
Coating of Pharmaceutical Solid-Dosage Forms 145
film-coating process. The trend has been to provide equipment designed
on the modular concept, where a basic processing unit is intended to
accommodate a variety of processing inserts. These inserts can facilitate
the use of fluid-bed processes for drying, granulating, spheronizing, and
coating.
While many similarities exist for equipment supplied by the various
vendors, opportunities for differentiation exists with:
Clamping systems (i.e., the mechanism for clamping the various removable
sections together). which are typically either compressed
air or hydraulic
Explosion protection, where some manufacturers use 2-bar construction
plus explosion-relief venting, whereas others rely on 10-bar
construction
Filter-bag assemblies, which can be based on either a split-filter
shaking or pulsed-air blow-back systems
Heating units, Where typically either conventional steam or electric
can be provided. but where there is a growing preference for
"face and by-pass" systems that permit more precise control over
processing temperatures while facilitating rapid changes (in temperature)
to be made When required
The trend is to provide standard equipment with appropriate options
so that designs can be more easily customized to meet end-user requirements.
Specialized designs, with taller expansion chambers that facilitate
appropriate deceleration when coating small particles, are becoming common.
As with other types of coating equipment, opportunities for providing completely
automated processes now exist.
GLATT FLUID-BED COATING EQUIPMENT. Originally, the Glatt Air
Techniques' designs were based on the WSG processing unit. which is
capable of accepting a variety of inserts. More recently. however, in
order to better address the situation in which the material to be coated
is of small particle size (ranging from approximately 100 um- 2 mm), the
emphasis has shifted toward the GPCG processing unit. Inserts can be
provided to permit drying. granulating (or top-spray coating), Wurster
coating (bottom spray). and rotor granulationllayering/coating (tangential
spray).
An example of a 5- to lO-kg pilot GPCG unit is shown in Figure 29.
Other common features of the Glatt equipment are:
Filter systems based on the split filter. shaking principle
2-Bar construction with explosion relief (although 10-bar construction,
at an added cost. may be available)
Microprocessor-based control panels when non-explosion-proof operating
environments are permissible
Hydraulic clamping systems
AEROMATIC FLUID-BED COATING EQUIPMENT. Aeromatic fluid-bed
coating equipment is again designed to accommodate a variety of modular
inserts. A typical laboratory scale unit, capable of processing up to
3- 5 kg of material. is shown in Figure 30.
Inserts that can be utilized with this manufacturer's equipment
include:
146 Porter and Bruno
Figure 29 Figure of a Glatt GPCG-5 fluid-bed coating unit. (Courtesy
of Glatt Air Techniquest Ramsey, New Jersey.)
Dryer (with options for batch or continuous processing)
Spray granulator (and top-spray coating)
Aero-coater (Wurster, bottom-spray coating)
Ultra coater (bottom/tangential spray for tablet coating)
Some additional features of Aeromatic equipment are:
Sequential blow-back filter system
Clamping system based on compressed air
Standard 10-bar construction ,
Unlike some of the competitive equipment , the Wurster process (Aerocoater)
does not require a completely separate unit, but rather is created
by adding the inner Wurster chamber (insert) to the product bowl of
the batch granulator unit. This obviously provides some opportunity
for economizing on cost.
Porter and Bruno
Figure 30 Figure of an Aeromatic multi fluid-bed coating unit.
eCourtesy of Aeromatic. Inc., Columbia, New Jersey.)
147
VECTOR-FREUND FLU ID-BED COATING EQUIPMENT. The design of
the Vector-Freund fluid-bed coating equipment was originally based on
that of the Floweoater , which features spray nozzles located (and angled
downward) in the sidewalls of the product container. This design permits
the same equipment to be used for fluid-bed drying, coating, and
granulating.
148 Porter and Bruno
Figure 31 Figure of a Vector-Freund Flowcoater multi fluid-bed coating
unit. (Courtesy of Vector Corporation. Marion. Iowa.)
Coating of Pharmaceutical Solid-Dosage Forms
More recently, the Flowcoater multi unit (Fig. 31) has been introduced,
which features separate inserts for:
Drying
Granulating (and top-spray coating)
Wurster coating
Spheronizing/layering/coating (a tangential spray, rotor unit that is
based on the operating principle of the CF granulator)
149
HUTTLIN KUGELCOATER. In terms of fluid-bed fllm coating, the
Kugelcoater is rather unique, both in basic design and for the fact that
it does not consist of a central processor that accepts multiple inserts.
A simplified schematic of the Kugelcoater is represented in Figure 32,
whereas a photograph of the equipment is shown in Figure 33.
The product container of the Kugelcoater is spherical. Inlet fluidizing
air is introduced via a tube that passes down the center of the product
container. The design permits this air to enter the product container at
the bottom. A series of spray nozzles are also located at the bottom of
the product container in such a way that fluidizing air creates a "balloon"
effect to keep the product being coated away from the spray nozzles.
Exhaust
Al.r
Product
Redirectl.on
Screen
Inlet Al.r
Static Coarse Fl.lter
.---__ Product Contal.ner
"Balloon" Al.r Pocket
Spray Gun
;{}
\lPneumatlc Product Loadl.nq
and Unloadl.nq
Figure 32 Schematic diagram of a Kugelcoater.
150 Porter and Bruno
Figure 33 Fig'Ure of a Hiittlin Kugeleoater ,
national, Englishtown, New Jersey.)
(Courtesy of Key Inter-
The product fluidization pattern is constrained by a product redirection
screen that restricts product entry into the filter system.
This unique design. encompassing the inclusion of multiple spray nozzles
intended to maximize uniformity of distribution of the coating, also
permits pneumatic loading and unloading of product.
Various sizes of the Kugelcoater are available, ranging from lab models
with product container capacities of 2 L to those having capacities of
700 L.
c. Application Equipment
In most modern pan-coating operations, one major emphasis is to provide a
means of mechanically applying coating liquids, and thus avoid manual
techniques such as ladling.
While there are similarities in requirements (for mechanical application
of coating liquids) between the sugar-coating and film-coating processes,
Coating of Pharmaceutical Solid-Dosage Forms 151
certain differences exist that almost certainly influence selection of appropriate
equipment.
Sugar-coating liquids typically have solids contents in excess of 70%
(w/w), and consequently can be extremely viscous. Contrast this with
the fact that film-coating liquids rarely contain greater than 20% nonvolatile
materials (Which typically fall in the range of 5 -15%).
Uniform distribution of sugar-coating liquids is achieved by the cascading
action (of the product being coated) that takes place within the pan
and causes transfer of liquid from one piece to another. This transfer
should occur before any substantial drying occurs. Thus, the application
equipment needs only to be very simple in design and allow the coating
liquid to stream onto the surface of the cascading product.
Conversely, film-coating liquids (almost without exception) need to be
applied in a finely atomized state and uniformly distributed across the surface
of the product being coated.
In the context of this chapter. mechanical application refers to any
method of introducing a coating liquid across the surface in an appropriate
manner other than hand ladling.
Application Sparges
Application sparges were discussed earlier in this chapter (and illustrated
in Fig. 3) and refer to simple equipment that allow the coating liquid to
stream across and onto the surface of the cascading bed of the product
being coated.
The simplicity of this type of equipment makes it extremely suitable
for use in the sugar-coating process.
Airless Spray Equipment
Airless spray equipment generally consists of a particular design of spray
gun that contains a nozzle with an extremely small orifice (typically
200-400 um or 0.009-0.015"). The coating liquid is delivered under significant
pressure (often in the range of 3.5 - 20.0 MPa or 500 - 3000 psi)
and velocity, and literally explodes into tiny droplets as it emerges from
the nozzle. Because of the high shear generated at the nozzle, a nozzle
insert or tip made of tungsten carbide is used. Nozzle tips are available
in a wide range of configurations that permit cone spray or flat (or ovalized)
spray patterns to be generated.
Liquid flow rates through airless spray equipment are typically high.
Stearn [62] described how such liquid flow rates. based on examination of
fundamental flow theories, are:
Directly proportional to the cross-sectional area of the nozzle orifice
Directly proportional to the square root of hydraulic pressure (of
coating liquid)
Inversely proportional to density of the coating liquid
Since a minimum hydraulic pressure is required for optimal atomization
of the coating liquid. and there is a practical lower limit on orifice size
(before clogging becomes a serious problem), airless spray equipment is
mainly restricted to applications where high-volume delivery of coating
liquids is required. Such applications include production scale. organicsol
vent-based film coating or similar scale sugar coating.
152 Porter and Bruno
Air-Spray Equipment
Air-spray application equipment is generally less expensive than the corresponding
airless equipment, and is capable of delivering coating liquids at
low to relatively high application rates. The functional components of an
air-spray gun are the fluid cap (through which the liquid emerges) and
the air cap (through which compressed air is delivered to create the
driving force for atomization). Typically, the air cap fits over the fluid
cap (or nozzle) and forms an annulus that allows compressed air to impinge
on the stream of coating liquid emerging from the fluid nozzle. This impingement
causes the coating liquid to be broken up into tiny droplets.
Droplet size and size distribution are typically controlled by atomizing
air pressure and volume. In most pan-coating operations, in order to
allow the coating liquid to be spread out, the air cap also has "wing tips"
that permit compressed air to be directed laterally onto the atomized spray
so that the spray pattern is ovalfzed, In order to avoid increased risk of
spray drying and turbulence. care should be taken not to try and ovalize
excessively the spray patterns. For this reason, air-spray guns typically
are unable to cover as much of the surface of the product bed as airless
guns, and thus more air-spray guns must be used for a particular application
(e.g., in a typical 48" Accela-Cota setup, whereas two airless guns
may provide adequate coverage, three or more air-spray guns may be
required to provide equivalent coverage).
Air-spray equipment is typically suited to small-scale coating operations
and all those involving the use of aqueous formulations (where the
atomizing air effectively augments the drying capabilities of the airhandling
system). Air-spray equipment should be used with great care
for organic-solvent-based film coating processes, since the atomizing air
can cause excessive spray drying.
Ultrasonic-Spray Equipment
Recently, attempts have been made to utilize ultrasonic nozzles as a means
of effectively atomizing and applying coating liquids. Such nozzles are
commonly used in oil burners and as part of the lithophotographic process
used in the semiconductor industry. and to apply specialized coatings onto
the surfaces of medical devices.
The major advantages of such nozzles are that they are economical to
operate (although initial capital cost can be high). produce atomized
liquids with uniform size distributions and low velocities, and eliminate
many of the problems associated with overspray and fogging.
While various designs of ultrasonic nozzles are available, those used
for coatings typically have relatively large orifices and conical tips. The
operating principle of an ultrasonic nozzle is similar to that of a speaker
in audio equipment. In this case, ceramic piezoelectric transducers convert
electrical energy into mechanical energy, which is transmitted to a
titanium horn that forms the atomizing tip. Coating liquid that is introduced
onto the atomizing surface absorbs some of the vibrational energy.
causing a wave motion to be set up in the liquid. If the vibrational
amplitude is carefully controlled, the liquid wID break free from the surface
of the nozzle as a fine mist (with median droplet sizes in the range
of 20-50 um) .
So far, the use of ultrasonic nozzles with pharmaceutical coating
liquids has met with t.ery limited success. Most polymeric-coating
Coatin.g of Pharmaceutical Solid-Dosage Forms 153
solutions appear to present a problem with respect to achieving effective
and-controllable atomization. This limitation seems to reflect the complex
rheological and cohesive properties of these solutions. Such nozzles might
be used more effectively where the coating liquid is a dispersion (e. g. ,
many of the formulations used in sugar coating or the aqueous latex systems
used in film coating).
Berger [63] has given a more complete description of ultrasonic nozzles,
their applications, and limitations.
D. Metering/Delivery Equipment
This discussion of metering/delivery equipment refers to that equipment
Which enables coating liquid to be delivered from a bulk holding tank to
the application equipment described in the previous section (Sec. IV.C).
Pressurized Con.tainers
Pressurized containers (or pressure pots, as they are commonly called)
were once commonly used with many organic-solvent-based film-coating
systems and have recently been reintroduced as components of some
specialized automated delivery systems.
Basically, such a system consists of a special liquid holding tank that
can be pressurized by compressed air. Once pressurized, the liquid is
forced from the tank through feed lines to the spray nozzle. Fluid delivery
rates with such equipment will be dependent on air pressure.
liquid viscosity and density, length and internal diameter of the feed lines,
and design of the spray nozzles used.
All other things being equal (for a given system and coating liquid),
flow rate is controlled via a pressure regulator that determines pressure
within the tank.
Precision with such a system may be inadequate when attempting to
deliver low-viscosity aqueous-coating liquids (such as latex dispersions).
Peristaltic Pumps
Peristaltic pumps (or "tubing" pumps) are commonly used in many coating
operations. The operating principle of these pumps is based on the ability
to create liquid flow in a tube by squeezing and stretching that tube.
Most common peristaltic pumps have a pump head that allows a piece of
tubing to be clamped in a U - shaped fashion against a rotating mechanical
device (such as "fingers" or rollers) that squeezes the tube and creates
motion of the liquid within the tube. To function effectively, the material
from which the tubing is made should have sufficient resiliency that it
quickly recovers its original shape after squeezing. For this reason. preferred
tubing is made from silicone rubber.
Peristaltic pumps have two major advantages:
1. They are simple and inexpensive (to purchase and operate).
2. They are sanitary (since the liquid being transferred does not
come in direct contact with the mechanical pump head). and cleanup
is easUy accomplished by passing a cleaning liquid through the
tubing, and tubing that is difficult to clean can be replaced at
little cost.
154 Porter and Bruno
There are some major drawbacks with such equipment, including:
They are not positive displacement pumps.
Linearity, and often accuracy, decreases as the pump speed is increased
(particularly at the high end of the speed range).
Accuracy decreases as the tubing wears and fatigues.
Their effectiveness is limited by liquid viscosity.
Gear Pumps
Gear pumps are becoming extremely popular in aqueous film coating. They
are positive displacement pumps that rely on two counterrotating gears to
draw liquid into and through the pump housing. Because the gears come
into contact with the coating liquid, they must be made of noncorrodible
materials. Gear pumps can transfer liquids with a wide range of viscosities,
but rely on the inherent "lubrdctty" of the coating liquid to minimize wear
in the gear mechanism.
Some limitations of gear pumps are:
Cost
Requirement for more complicated clean-up procedures
Likelihood of excessive wear of mechanism when pumping highly pigmented.
low-viscosity coating liquids
Likelihood of inducing coagulation of latex-coating systems as a result
of shear developed as liquid passes through pump head
While many gear pump s rely on a direct mechanical drive. some recent
models have an indirect magnetic drive.
Piston Pumps
Piston pumps. which are also positive displacement pumps, rely on a reciprocating
piston (typically powered by a compressed air motor) to transfer
the coating liquid. Such a pump can easily handle very viscous liquids
and so is well suited to sugar-coating and most film-coating operations.
Those piston pumps driven by air motors usually have a particular
pressure ratio rating (such as 30; 1, 10; 1, or 5: 1). This rating refers to
amount by Which pressure supplied by the air operating the pumps is increased
when transferred to the liquid being pumped. For example J 10
psi air pressure applied to a 30:1 ratio pump will generate 300 psi of liquid
line pressure. For this reason, such pumps are commonly used as the delivery
mechanism in airless spray systems.
More recently, electrically driven piston pumps, which have variable
speeds and stroke lengths and generate very little line pressure, have
been introduced as components of air-spray systems used in aqueous film
coating. An example of such a pumping system is the VOlumetric pump
PD 1000/S supplied with the Pellegrini-OS coating system.
Piston pumps, like gear pumps, are relatively expensive and not as
easy to clean as tubing pumps. In addition, the shearing action as the
pump reciprocates may induce coagulation of latex-coating systems. This
is particularly true when any significant line pressure is generated.
Coating of Pharmaceutical Solid-Dosage Forms
V. AUTOMATED COATING PROCEDURES
A. Jntroduction
155
In the early days of tablet coating, the addition of coating liquids, determination
of distribution/mixing times. and application of drying and
exhaust air were all manual operations. Decisions as to when heated air
should be applied, coating liquid should be added, and how long to allow
for uniform distribution of the coating liquid were all judgmental factors on
the part of the operator.
In order to achieve a high level of reproducibility in the process and
simplify the documentation process, it is apparent that the above situation
has many limitations.
To effectively automate a process, it is necessary to remove the requirements
for direct (and repeated) involvement by the operator. The
process must be well developed and precisely defined (when a poorly defined
process is automated it will simply produce consistently unacceptable
results time after time). In particular, it is imperative to avoid using any
critical steps that can only be achieved on such a limited basis as to
render the whole process impractical.
The benefits to be derived from a well-automated process are;
Independence from operator judgment
Achievement of consistent product quality
Achievement of a fully optimized process with greater potential for
minimizing processing times
Production of hard-copy documentation
It is critical, however, that an automated process possesses a fail-safe
mechanism so that process shutdown takes place in the event of process
failure. Such a procedure will reduce the liklihood that a batch of product
is ruined.
The automation of a coating process typically involves two basic
activities:
1. Timing and sequencing (more appropriate for sugar coating where
a sequence of events must occur at specific time intervals)
2. Measuring and controlling appropriate processing parameters
The second of these two activities can be the more difficult, particularly
since appropriate precision in measuring devices may be difficult to
achieve. Measuring and control devices may be classified as;
Electromechanical
Pneumatic
Electronic
There is no doubt that electronic devices are preferred in order to
deal with the complexities of today's coating processes. Electronic devices,
however, are subject to drift, and thus must constantly be zecalibrated
to ensure accuracy and precision.
156 Porter and Bruno
Figure 3lJ. Figure of a totally automated film-coating process. (Courtesy
of Meltech. East Hanover. New Jersey.)
A description of some pan-coating control systems has been given by
Thomas [64], and an example of a modern fully automated tablet coating
process is shown in Figure 34.
B. Automation of Sugar Coating
In basic terms, sugar coating can be considered. from the automation
standpoint. to be a complex. noncritical process. Complexity stems from
the significant number of timing and sequencing functions that are required.
Noncriticality (from a relative standpoint) occurs because many of the coating
parameters need not be controlled to extreme accuracies. For example,
accurate control of airflow and temperature is not critical as long as specific
objectives (such as achieving a particular state of product dryness
before the next addition of coating liquid occurs) are met.
Early attempts at automation of the sugar-coating process involved
achieving the appropriate sequencing and timing by means of timers.
Such an approach is very simple but has limitations. namely:
If constant, predictable drying conditions are not achieved (such as
use of temperature and humidity-controlled drying air). then seasonal
variations in conditions may result in predetermined sequencing
and timing to be totally inappropriate for a specific set
of conditions.
As coating build-up occurs, there may be a need to adjust the volumes
of coating liquid applied.
Coating of Pharmaceutical Solid-Dosage Forms 157
Over the last 25 years, various descriptions of automated procedures
have appeared in the pharmaceutical literature [65 - 67] •
An early commercially available system, the Drdamat , has been described
by Rose [68].
A more recently introduced commercial, fully-automated, sugar-coating
process is the Sandomatic system, which was first described by Melliger
and Goss [69].
The important independent process variables that need to be monitored
and controlled in a typical sugar-coating process are:
Drying air volume (or flow)
Drying air temperature
Drying air humidity
Pan speed
Quantity of coating liquid applied (at each step)
C. Automation of Film Coating
In comparison to sugar coating, film coating is a simpler, but more critical
process to automate. Little sequencing is required, but important process
conditions must be monitored and controlled to a high degree of accuracy.
Important independent variables that need to be measured and controlled
in a film -coating process inc1u de:
Drying air volume
Drying air temperature
Drying air humidity
Pan speed
Spray rate (more precisely, the delivery rate from each spray gun in
a multiple-gun setup)
Atomizing air pressure/volume
Pattern air pressure/volume
In addition, it may be necessary to control various dependent variables,
particularly in the aqueous process where the margin for error can be
much less than in other types of coating process. Some of the important
dependent variables that need to be monitored are:
Tablet bed temperature
Exhaust air temperature
Exhaust air humidity
Tablet bed moisture
Since these variables cannot be controlled directly, it will be necessary
to control the interaction between those independent variables that influence
a particular dependent variable. For example, exhaust air temperature
may be kept under control by maintaining an appropriate balance between:
Inlet air volum e
Inlet air temperature
Inlet air humidity
Spray rate
158 Porter and Bruno
In the aqueous film-coating process, the importance of taking into account
the moisture content of the inlet drying air cannot be overemphasized.
The simplest way of dealing with this issue is to keep inlet air
humidity constant by use of a dehumidificat .on process. Such an approach
can, however. prove to be extremely expensive, since the volumes of air
that need to be conditioned are typically very large (in an aqueous filmcoating
process, e.g•• a typical 48" Accela-Cota coating a batch of tablets
in 90 min wID require approximately 5000 m3 of conditioned air).
An alternative is to compensate for variation in humidity of inlet air
by adju sting one or more of:
Air volume
Inlet air temperature
Spray rate
Lachman and Cooper [70] have described an early attempt to automate
the film-coating process. More recently, the automation principles of the
Sandomatlc process have been applied to film coating. Also, virtually
every supplier of pan-coating equipment can provide automated equipment,
and selection may well depend on the preferences, and individual needs,
of the pharmaceutical manufacturer.
REFERENCES
1. Signorino, C. A., U.S. Patent 3,738,952 (1973).
2. Rowe. R. C .• Int. J. Pharm., 43:155 (1988).
3. McGinlty, J. W., Aqueous Polymeric Coatings for Pharmaceutical
Dosage Forms. Marcel Dekker, New York, 1988.
4. Skultety, P. F., Rivera, D., Dunleavy, J., and Lin, C. T. Drug
Dev. Ind. Pharm., 14(5):617 (1988).
5. Reiland, T. L. and Eber, A. C •• Drug Dev; Ind. Pharm., 13(3):
231 (1986).
6. Thoennes, C. J. and McCurdy, V. E., contributed paper at PT
Section of 3rd Annual Meeting of AAPS, Orlando, Florida, Oct. 1988.
7. Ebey, G. C.. Pharm. Tech.. 11( 4) : 40 (1987).
8. Rowe, R. C., J. Pharm. Pharmacal., 29: 723 (1977).
9. Fisher, D. G. and Rowe, R. C., J. Pharm, Pharmacol., 28:886
(1976) •
10. Rowe, R. C., J. Pharm. Pharmacal., 30:669 (1978).
11. Aulton, M. E., Twitchell, A. M., and Hogan, J. E.. Proceedings of
AGPI Conference, Paris (1986).
12. Rowe, R. C., J. Pharm. Pharmacal., 32:851 (1980).
13. Rowe, R. C. and Forse, S. F., Acta Pharm. Tech., 28( 3) : 207
(1982) •
14. Twitchel, A. M•• Hogan, J. E., and Aulton, M. E., J. Pharm.
Pharmacal., 398: 128P (1987).
15. Porter, S. C. and Saraceni, K., Pharm. Tech., 12(9):78 (1988).
16. Kara, M. A. K., Leaver, T. M•• and Rowe, R. C., J. Pharm.
Pharmacal., 34:469 (1982).
17. Rowe, R. C., Pharm, Int., 6(9):225 (1985).
Coating of Pharmaceutical Solid-Dosage Forms 159
18. Wicks, Z. W., Jr., FUm Formation. Series on Coatings Technology,
Federation of Societies for Coating Technology, Philadelphia (1986).
19. Burrett , H., Offic. Dig., 34(445):131 (1962).
20. Williams, M. L., Landel, R. F., and Ferry, J. D., J. Am. Chem.
Soc., 77:3701 (1955).
21. Wicks, Z. W., Jr., J. Coatings Technol., 58(743):23 (1986).
22. Entwistle, C. A. and Rowe, R. C., J. Pharm • Pharmacol., 31: 269
(1979) •
23. Rowe, R. C., J. Pharm. Pharmacal., 33: 423 (1981).
24. Bindschaedler, C., Gurney, R., and Doelker, E., Labo-PharmaProbl.
Tech., 31(331):389 (1983).
25. Bradford, E. B. and Vanderhoff, J. W., J. Macromol. Chem., 1(2):
335 (1966).
26. Gross, H. M. and Endicott, C. J., D&CI, 86( 2) : 170 (1960).
27. Porter, S. C., Pharm • Tech., 4( 3) : 66 (1980).
28. Crawford, R. R. and Esmerian, O. K., J. Pharm. set., 60(2):312
(1971) .
29. Sakellariou, P., Rowe, R. C., and White, E. F. To, Int. J. Pharm.,
31: 55 (1986).
30. Woznicki, E. J. and Schoneker, D. R., Coloring agents, In Encyclopedia
of Pharmaceutical Technology. Marcel Dekker, New York, in press.
31. Porter, S. C., Int. J. Pharm, Tech. Prod. Mfr., 3( 1) : 21 (1982).
32. Rowe, R. C., Pharm. Acta. Helv .• 60(5/6):157 (1985).
33. Rowe, R. C., J. Pharm. Pharmacol., 33;51 (1980).
34. Rowe, R. C., J. Pharm. Pharmacol., 36: 569 (1983).
35. Rowe, R. C. and Forse, S. F., J. Piuirm, Pharmacal., 35:205
(1983) •
36. Rowe, R. C., Int. J. Pharm., 14:355 (1983).
37. Banker, G. S., J. Pharm, Sci., 55;81 (1966).
38. HUdebrand, J. and Scott, R., The Solubility of Non-Electrolytes,
3rd Ed. Reinhold, New York, 1949.
39. Kent, D. J. and Rowe, R. C., J. Pharm. Pharmacal., 31(5):269
(1979) •
40. Rudin, A. and Wagner, R. A., J. Appl. Polym. Sc•• , 19:3361
(1975) •
41. Johnson, K., Hathaway, R. D., and Franz, R. M., Presented at the
Contributed Paper Session of the 3rd Annual Meeting of AAPS,
Orlando, Florida, 1988.
42. Porter, S. C., In Aqueous Polymeric Coatings for Pharmaceutical
Dosage Forms (J. W. McGinity, ed.), Marcel Dekker, New York,
1988, p , 317.
43. Ozturk, S. S., Palsson, B. 0., Donohoe, B., and Dressman, J. B.,
Pharm. se«, 5( 9) : 550 (1988).
44. Edgar, B., Bogentaft, C., and Lagerstrom, P. 0., Biopharm Drug
Disp., 5:251 (1984).
45. Theeuwes, F., J. Pharm. sa., 64(12):1987 (1975).
46. Lindholm, T. and JuIsin, M., Pharm. tna., 44(9):937 (1982).
47. Jarnbhekar, S. S., Breen, P. J., and Rojanasakul , Y., Drug Dev.
Ind. Pharm., 13( 15) : 2789 (1987).
48. Rowe, R. C., J. Pharm. Pharmacal., 35:112 (1983).
49. Rowe, R. C., J. Pharm. Pharmacol., 32: 851 (1980).
160 Porter and Bruno
50. Breech. J. A., Lucisano, L. J., and Franz, R. M., J. Pharm.
Pharmacol.. 40: 282 (1988).
51. Rowe. R. C. and Forse, S. F., Acta. Pharm. Technol.. 28(3):207
(1982).
52. Kim. S., Mankad. A •• and Sheen, P. Drug Dev. Ind. Pharm., 12(6):
801 (1986).
53. Down. G. R. B., J. Pharm. Pharmacal., 34(4):281 (1982).
54. Shah, B. B., Contractor, A. M.. and Auslander. D. E., Pharm.
Res., 5(10):S-253 (1988).
55. Wurster. D. E., U.S. Patent 2.648,609 (1953).
56. Hall. H. S. and Hinkes, T. M., Presented at the Symposium on
Microencapsulation: Processes and Applications. American Chemical
Society. Chicago, Aug. 27-31, 1973.
57. Glatt Technically Speaking, 2( 1) : 1 (1989).
58. Jones. D. M., Pharm. Technol., 9(4):50 (1985).
59. Mehta, A. M., Pharm. Technol., 12( 2) : 46 (1988).
60. Olsen. K. W. and Mehta. A. M., Int. J. Pharm. Tech. Prod. Mfr••
6( 4) : 18 (1985).
61. Johansson, M. E., Ringberg, A., and Nicklasson, M., J. Microencapsul••
4(3):217 (1987).
62. Stern. P. W., J. Pharm; Sci•• 63(7):1171 (1974).
63. Berger, H. L., Machine Design, July 1988.
64. Thomas. R., Pharmaceut. Eng., 16:Aug. -Oct. (1981).
65. Rieckmann, P., Drugs Made in Germany, 6:162 (1963).
66. Heyd, A. and Kanig, J. L., J. Pharm. sa., 59(8):1171 (1970).
67. Fox, D. C., Buckpitt, A. E., Laramie, M. V., and Miserany, M. E.,
Presented at A.I.Ch.E. Meeting, New York, Nov. 17, 1977.
68. Rose, F., D&C.I•• 44:Nov. (1971).
69. Melliger, G. and Goss, L•• Presented at A.Ph.A. Annual Meeting,
New York. April 1977.
70. Lachman. L. and Cooper. J., J. Pharm. Sci; , 52( 5) : 490 (1963).
3
Particle-Coating Methods
Dale E. Wurster
University of Iowa. Iowa City, Iowa
I. INTRODUCTiON
There are many important pharmaceutical reasons for applying coatings to
both pure chemicals used as medicinal agents and to physical -chemical
systems employed as dosage forms [l-4J. Similar coatings are also widely
used in other industries to coat particulate solids. One of the most widespread
pharmaceutical uses today is for the control of drug release from
various types of oral or parenteral dosage forms designed to have either
an enteric, a timed, or a sustained-release effect. Thus, particle coatings
can be formulated so that the escape of the drug from the dosage
form occurs primarily via a diffusional process. The various physical
parameters influencing diffusion can then be modified in order to control
the rate of drug release. In other systems, the dissolution kinetics of
the applied coat serve to control the availability of the contained drug.
In still other cases, drug release can be controlled by chemical reactions
involving the coating material or by simple attritional effects on the coat.
No matter what mechanism of controlled drug release one wishes to use, it
is apparent that this can usually be achieved by the proper formulation of
coating materials and the manner in which they are applied.
A second type of effect achieved by coating procedures is the enhancement
of the chemical stability of drugs. It is possible to formulate
coating materials to guard against the penetration of substances which will
cause specific degradation reactions. Thus, coatings can be designed to
resist the penetration of water vapor when the product contains drugs
which are susceptible to hydrolytic decomposition. For those drugs which
degrade by oxidative reactions, coatings can be used which inhibit the
transport of oxygen to the lab:ile compound. In cases where detrimental
161
162 Wurster
interactions result from materials contained within the physical system rather
than from external influences, the incompatible materials can be separated
by suitable coatings.
Of course, coating procedures have long been used to occlude drugs
with an unpleasant taste from the taste receptors, or to impart pleasing
colors to a product to make it more attractive visually. Certain other
physical parameters, such as particle density, surface characteristics, or
geometry, can also be altered by the use of coatings to obtain processing
advantages. For example, coatings can sufficiently change the geometry
and surface of small particles so that the flow properties of a powder are
enhanced.
It is the primary intent of this chapter to present a discussion of
those processes which are useful in particle coating to meet the above
needs. While the theoretical mechanisms involved in controlling such factors
as drug release and chemical stability are of great interest to all who
work with these systems, a discussion of these mechanisms is not within
the scope of this chapter. Primary emphasis is, therefore, directed toward
the processes which are useful for the coating of solid particles. However,
methods which are mainly useful for the microencapsulation of liquids are
also briefly discussed. This is because these processes can usually be
modified so that, in some manner, particulate solids can also be coated.
II. WURSTER PROCESS
A. Introduction
The commonly called air-suspension particle-coating method, also known as
the Wurster process,* has been utilized for many years by the pharmaceutical,
chemical, food, agricultural, and other industries. The rapidity
of the operation, the ability to control variables, the uniformity of the coat
produced, and the fact that it can be used to coat particles varying greatly
in size. shape, and density are some of the main advantages of the
process. Further, the process does not restrict either the kind of coating
materials used or the solvents employed in the coating fluids. Even chemical
reactions designed to produce polymer coatings from applied monomer
solutions can be easily conducted in the column.
Because it is possible to control the variables of the process the method
is sufficiently flexible to make it extremely versatile. Thus, the
process can be utilized to coat medicinal tablets [5 -13] or other large particles,
to microencapsulate fine particles [14], to control drug release from
coated particles [15,16], and to prepare compressed tablet granulations
[17,18] or the like. By the proper adjustment of process conditions, it is
possible to either coat solid particles Or to cause particles to aggregate.
It is also possible to control the density, geometry, friability, and other
physical parameters. of the treated particles.
There is little doubt that the Wurster process provides the most rapid,
efficient, and cost-effective method of producing film-coated compressed
"'Wisconsin Alumni Research Foundation, Madison. Wisconsin.
Particle-Coating Methods
tablets. Similar advantages are gained in microencapsulation and compressed-
tablet granulation procedures.
B. General Description
163
The Wurster process* can be appropriately be described as an upwardmoving,
highly-expanded pneumatically transported bed of particles coupled
with a downward-moving, more condensed, fluidized bed of particles on the
periphery of a vertical column. The two beds are separated by the tubular
central partition. Figure 1 is a schematic drawing of the equipment. As
indicated above. the particles to be coated are pneumatically conveyed upwardly
through the central tube, (b). of the coating chamber. As the
particles pass the atomizer, (d). they are wetted by the coating fluid and
then Immediately subjected to drying conditions created by the heated COnveying
air moving upwardly in the column. The partially coated solid particles
move downwardly in a near-weightless condition along the periphery
of the column. (c). where further drying occurs. When the solid particles
reach the lower end of the column they are directed back into the
upwardly-moving bed and the entire process is repeated. The air pump,
(g). and heater (f), provide the heated support air for the process, and
the distribution plate (e) directs the proper volume of air to the central
and peripheral regions of the column.
The proper adjustment of the airflow, the temperature, and the fluid
application rate are all critical to the successful operation of the process.
Obviously. the drying kinetics are influenced by the airflow rate and the
temperature of the air. These kinetics in turn dictate the fluid application
rate. Since drying is a cooling process. the temperature of the particle
surface is lower than either the inlet or exhaust air temperatures.
This permits the coating of heat-sensitive materials, since process conditions
can be adjusted so that the exhaust temperature will not exceed the
temperature which the product can tolerate. Thus, it is readily apparent
that the drying kinetics can be enhanced by increasing either the airflow
rate or the air temperature while maintaining the fluid application rate
constant.
C. Control of Airflow
The proper airflow rate is most important for the successful operation of
the Wurster process. It is apparent from the general description of the
process that the two-component fluidized bed employed in this process
*Before using this or other fluidized bed processes. it is important to become
familiar with the basic principles that govern fluid-bed systems.
The following textbooks are included in a long list of textbooks that are
very useful for this purpose. (1) Cheremisinoff, N. P., Hydrodynamics of
Gas -Solids Fluidization. Gulf Publishing Co•• Houston, 1984; (2) Davidson,
J. F., Fluidization, Academic Press, New York, 1971; and (3) Howard,
J. R., Fluidized Beds, Applied Science Publishers Ltd., London, 1983.
164
---+--- Settling chamber (a)
.....+----- Central tube (b)
....---- Peripheral Tube (c)
Atomizer (d)
~§E~~"""--- Distribution plate (e)
Heater (1)
Figure 1 Schematic drawing of air-suspension coating apparatus.
Wurster
differs considerably from a conventional fluidized bed. Since the solid particles
undergo pneumatic transport up the central portion of the column
and descend in the peripheral region, different air velocities are required
in the two regions. Although an air distribution plate is positioned at the
base of the column (see Fig. 1). an air distribution zone also exists just
above the plate. It is in this zone in the loaded column that the air undergoes
further distribution to the directional flow and velocity conditions
which prevail above the plate and which differ markedly from the conditions
imposed by the plate alone [19]. Nevertheless, it is extremely important
that the plate is properly designed for the particular production
process for Which the method is being used.
Minimum Fluidization Velocity
When particulate solids are placed on a perforated plate in a vertical column
and a gas, usually air, is passed upwardly through the bed of particles
the bed at lower air velocities will initially remain fixed. However,
as the air velocity is increased the particulate bed expands, the individual
Particle-Coating Methods 165
particles begin to move in the air stream. and an increase in the void
space of the bed is observed. The transition of the bed from the fixed
to the fluidized state occurs at the point where the pressure drop across
the bed is equal to the weight of the bed per unit of cross-sectional area.
This is often referred to as incipient fluidization, and the air velocity required
to produce this state is known as the minimum fluidizing velocity.
Thus. the pressure drop across the bed at incipient fluidization is given
by [20.21]
(1)
Where &Pb := pressure drop across bed
Emf =bed voidage at condition of minimum fluidization
p =density of solid particles s
o =density of fluidizing gas
g
Hmf = height of bed at minimum fluidization
g =gravitational constant
For uniform. small spherical particles in the approximate range of
20-100 um and densities up to 1.4. the equation derived by Kozeny [2023]
and tested by Carman over the range of bed voidages of O. 26 - 0.89
[24.25] shows that the relationship between the pressure drop and the
minimum fluidizing velocity at incipient fluidization can be written in the
following manner
3
E
V:= mf
mf 5(1 _ E )2
mf
(2)
Here Vmf = gas velocity at minimum fluidization
S = specific surface area of solid particles
J.l =the viscosity of the gas
From equations (1) and (2). it follows that
3 e
V = mf
mf 5(1 - Emf)
(3)
If it is assumed that the bed is composed of uniform spherical particles.
S = 6/d, where d is the particle diameter and correspondingly. the bed
voidage, Emf' is considered to be 0.4 then [20)
0.4
3
V := mf 5(1 - 0.4)
( 4)
166 Wurster
Another simple and convenient method for predicting the approximate
minimum fluidization velocity for solid particles in the 50- to 500-jJm size
range takes advantage of the mathematical expression recommended by
Woodcock [21,26)
( 5)
Here Vmf = minimum fluidization velocity in mls
p =the solid particle density in kg/ m3
p
d =mean particle diameter in meters
v
When the bed is composed of larger particles it has been suggested by
some [20] that more general type equations such as those given by Ergun
[27] and Black [28] be employed to arrive at the pressure drop across
the bed.
Effects of Increasing Air Velocity
The influences on the bed created by increasing air velocities are shown
in the idealized graph [29] presented in Figure 2. As stated previously,
the pressure drop across the bed increases with an increase in the velocity
of the air directed through the bed. Initially. the bed remains fixed, but as
the air velocity is further increased the bed expands somewhat and the
0..
0.7 .... ;:r
0
~
CI.l
0.6 1-
:§
,
0.5 .~
"0
?:
u
0.4 '"c,
'" OJ
III
~
0.3 '0
::E
0.2
0.1
100 90 80 70
Wet bulb
temperature
of
0.021 ,
?: :c
'E
" J:
"-I '" ~
!~
TEMPERATURE (Degree F)
Figure 6. Humidity chart.
~
~
~a~
~
180 160 140 120
Temperature IDegree F)
100 80 60
Alcohol-drying chart for coating solution. Figure 7
....
~
PERCENTAGE SATURATION I
15% 10% 5%
005I I i ./ / / • I. 1 2
5
Safe operating
limit
I I I2.0 ~
0041 I I . I - { --,., 1 I I I '" if ... ... E
't
~~
o I I Im--.:-='\l: ~~ .r:. '- - «I) o ;;; ~ > ~.c 003 - - - - I 15 0-- ~-=- -~ ~ ~'~~'.".'~I"'" I =1 i~
u
;;;
'0
DOlt I ~~'* -=---t==:= - - - -10 E
001 - - - - - - - - - - - - ~ - t ~, J 05
Particle -Coa ting Metho ds 173
is maintained, then attrition of the particles may occur. In actual practice,
then, a humidity condition above 85% for some materials can lead to aggregation,
whereas below 45% attrition may occur.
Drying charts can be constructed and used for organic solvents or solvent
mixtures in a manner similar to the above use of the humidity chart.
In Figure 7, a drying chart for ethyl alcohol solutions of coating materials
is shown. The required calculations for chart construction are conventional
and need not be treated here.* As in the case of humidity charts,
these charts ignore the influence of dissolved substances on the solvent removal
rate.
The determination of the fluid application rates using solvent-drying
charts such as shown in Figure 4 differs slightly from that of a humidity
chart. This, of course, is because the atmosphere employed as the support
air does not normally contain a significant amount of the organic solvent.
From the alcohol-drying chart shown in Figure 5, it can be observed
that if the air is 140°F and if an outlet temperature of 98°F is desired,
then 0.024 lb of alcohol/1b Qf dry air can be used. From the righthand
scale it can be observed that under the above conditions, 1.13 ml of
the coating solution per cubic foot per minute of air can be employed.
Then if the support air is 100 efm , the fluid application rate for the coating
solution is 113 ml min-1•
Since alcohol is an inflammable solvent, the process conditions can be
set so that a wide margin of safety exists. The lower inflammability limit
for alcohol is O. 0716 1b of alcohol vapor/lb of dry air. In Figure 4, the
safe operating limit is set at two-thirds of this value or 0.0477 lb of alcohol/
1b of dry air. The maximum inlet temperature is also indicated as is
the process limit. By setting operating conditions within these parameters,
a wide margin of safety is achieved [30].
E. Drying Properties of the Process
When particles are wetted by a solvent it can be assumed that normally the
solvent penetrates into the particle and also that a pool of solvent resides
on the surface. When the wetted particle is subjected to a drying process,
the solvent on the surface is readily removed at a constant rate or a steady
state exists. The solvent which has penetrated the particle subsequently
returns to the surface by diffusion and capillarity, and thus a nonsteady
state exists for the removal of the solvent residing within the particle.
Figure 8 shows a typical drying curve for the solvent loss from a particle
as a function of time.
In the Wurster process, a thin layer of the coating fluid is atomized
onto the surface of the solid particles. Because of the elevated temperatures
and the large volume of air passing over the particles, the solvent
is rapidly removed and drying occurs primarily under the steady-state
condition.
*The interested reader is referred to Perry, R. H. and Chilton, C. H.,
Chemical Engineers' Handbook, 5th Ed., McGraw-Hill Book Company,
New York, 1973, Chap. 20; and Pilot Plant Operation, Wurster Coating and
Granulating Process, Wisconsin Alumni Research Foundation, Madison,
Wisconsin, Sec. VI.
174 Wurster
Time
Figure 8 Water loss as a function of time.
Falling rate period
Constant rate period
Drving
rate
(W) Free moisture
Figure 9 Drying rate versus free moisture.
Particle-Coating Methods 175
In drying processes, the nonsteady state or falling rate and the steady
state or constant rate of solvent removal is often depicted as shown in
Figure 9, in which the drying rate is plotted against the free solvent content.
The following relationship is also often employed to determine the
constant drying rate in conventional drying processes [32J.
dw
Cit =
h A (t - t )
t a s =k A (P _ P )
A. gsa (7)
= wet surface temperature
= mass transfer coefficient in Ib/(hr)(ft2)(atm)
=vapor pressure of H2
0 at t s
=the partial pressure of H
2
0 in air
where dw dt =drying rate in Ibs , H20 /h
h
t = total heat transfer coefficient in Btu /(hr)(ft2)(OF)
A = area in ft 2
A =latent heat of evaporation at t s
t =air temperature a
t s
k g
Ps
Pa
To obtain the falling rate in drying processes, we can use [32]
( dw) = K (W - W )
dt F e
(8)
Here W = average moisture content at time, t (dry basis)
W =average moisture content (in equilibrium with external eondieditions)
(dW) =(dW) =K(W - W )
dt C W=W dt F W-W c e
'c J C
where Wc = critical moisture content, and F =falling rate and C =the
constant rate, or
(9)
K =
(dw)/(dt)C
W - w c e
(10)
( 11)
Figure 10 shows the difference in the rate at which water is removed
from a product in the air-suspension method as compared to more traditional
drying methods. As indicated previously. the process can he monitored
by folloWing the exhaust temperature. Also J as shown in Figure 11,
176
... s'" ... ~ iii
~~
"0
G>
>oE
e...
J!l s
----,I • Air suspension drying
I
\,\
\
1----- \
\
\
\
Elapsed drying time ....
Wurster
Figure 10 Comparison of air-suspension drying with traditional drying
methods.
Inlet air temperature
r
Dried product
/ removed
Product temperature
Elapsed drying time ....
Figure 11 Air-su spension drying.
Particle-Coating Methods 177
.,
:; ...
~.,
Q.
E.,
IInlet
air temperatu re
Product
..--temperature
Elapsed drying time ..
Figure 12 Traditional drying process.
the exhaust temperature remains relatively constant during the steady-state
drying which occurs during the coating process. When atomization is
stopped. the temperature rapidly rises and the dried product can be immediately
removed without exposing it unnecessarily to a high temperature
for a long time. If during the actual coating process the exhaust temperature
falls below that set for the process. this suggests that the fluid
application rate is too great; the particles may become too wet. and aggregation
will occur. Conversely. if the exhaust temperature rises, then the
column condition is drier than desired and particle attrition may occur.
The equipment can be instrumented to monitor such changes and to make
corrective changes. Figure 12 shows the drying curve of a wetted product
by more traditional methods. In the case of substances which form hydrates.
the slight temperature changes can actually be utilized to follow
the conversion of one hydrate form to another (Fig. 13).
F. Uniformity of Coat
One of the main advantages of the air-suspension coating process is the
uniformity of coat that can be achieved on both small and large particles.
If the coating fluid is applied at a constant rate, both the weight of the
particles and the cube of the radius for spherical particles should increase
linearly with time. Figures 14 and 15 show this to be the case [5]. Even
a very thin. 1-lJm thick coat applied to a small crystal is extremely uniform.
as is shown in Figure 16.
178
Inlet air temperature
Wurster
Exhaust air temperature
Drying tirn«
Figure 13 Air-suspension drying of a hydrate.
o 60 120
Time (min)
180
Figure 14 Weight of particles as a function of process time (constant
atomization rate).
o 60 120
Tmu- {Ill Ill)
180
Figure 15 Particle size of spheres as a function of process time (constant
atomization rate).
179
180 Wurster
Figure 16 A t-um thick coat on a single crystal.
G. Particle Aggregates, Granulations, and
Particle Build-Up
As previously stated, it is possible to prepare particle aggregates of different
densities and geometries depending upon column conditions. For
example, one can prepare aggregates which are comparable to those made
for a compressed tablet granulation by the wet method. By utilizing materials
which are tacky when wet, a column condition approaching saturation
of the support air can be employed. In the case of water solutions
containing such materials as sugars and gums, a humidity condition in the
column exceeding 85% relative humidity (RH) will usually result in particle
aggregation. Irregular particles prepared by this process are shown in
Figure 17. If a friable, low-density spherical particle is desired. this
Particle-Coating Methods
Figure 17 Irregular aggregate of particles (1500 um) ,
181
182 Wurster
Figure 18 Porous low-density spherical aggregate of particles (1000 um) ,
Particle-Coating Methods 183
can also be accomplished. In this case a high-humidity column condition
together with the atomization of the binder solution as large liquid particles
is required. By utilizing a very small particle size of the solid material
(less than 15 um) and a large droplet size of the atomized solution.
the small solid particles are allowed to impinge upon and be entrapped by
the large liquid particles. Upon the subsequent removal of the solvent. a
porous. low-density particle results. This type of particle is shown in
Figure 18. The fragility of this type particle is evident by the presence
of the fractured spheres in the figure.
Spherical particles of near-theoretical density can be prepared by coating
particles with a slurry which contains both suspended solids and a
binder. In this case, column conditions are adjusted so that a coating
process rather than an aggregation process results (Fig. 19). If water is
Figure 19 High-density, slurry-coated particles (2000 um) .
184 Wurster
the solvent, a 65% humidity condition in the column can be used as a
starting point.
Granulations of effervescent materials are normally quite difficult to
prepare by conventional methods because of the close control which is required
over the reaction rate, the drying rate, and the residual water.
The air-suspension process is well suited for the preparation of these
granulations [33] because of the ease with which the process conditions
can be controlled and monitored.
In air-suspension coating procedures. the lower particle size limit is
approximately 50 um , but smaller particles have been coated. There is no
practical upper limit for larger particles. In fact, very large particles
are efficiently coated with great ease. When particle aggregates (tablet
granulations, etc.) are prepared. very fine particles of only a few micrometers
can be aggregated into larger particles without difficulty. Of course,
the particle aggregates can subsequently be coated, if desired, without removing
them from the column.
Lastly. because the air-suspension process operates as a closed system
with respect to the working area, people are protected from exposure to
dangerous chemicals and solvents. Similarly, the loss of solids can be prevented
and solvents can be reclaimed from the exhaust so that contamination
of the environment does not occur.
III. CENTRIFUGATION
The centrifugation method" is a unique modification of the simple extrusion
technique of producing microcapsules [34-40]. As shown in Figure 20.
the simple extrusion method utnizes a device consisting of two concentric
tubes containing aligned fluid nozzles. The liquid material to be coated is
extruded through the nozzle of the inner tube into the coating fluid contained
in the outer tube. Initially. the fluid extrudes as a rod surrounded
by the coating fluid, but the rod ultimately breaks up into droplets which
are then immersed in the coating fluid. As the extruded droplets pass
through the nozzle orifice of the outer tube. the coating fluid forms a surface
coat which encases the extruded particle. Spherically shaped particles
are formed by the surface tension of the liquid. By suitable means
the formed coat is converted to a more rigid structure. Hardening baths
are usually employed for this purpose. These baths are designed according
to the coating material used. Thus, in some cases the bath contains a
chemical agent which will react with the coating material to yield an insoluble
compound. A common example of this is the use of calcium chloride
in the bath to convert sodium alginate in the coat to the insoluble calcium
alginate [39]. In other cases, the bath may simply be a nonsolvent for
the coating material, whereas in still other cases only a decrease in temperature
is needed to harden the coat.
This process is primarily employed to prepare microcapsules of liquids.
Thus, the coating fluid is immiscible with the liquid to be encapsulated.
As previously stated. however, the main concern of this chapter is a discussion
of methods by which discrete solids can be coated. When this
"'Southwest Research Institute, San Antonio. Texas.
186 Wurster
method is used to coat solid drugs it is convenient to prepare a slurry of
the solid material. The alurry is then encapsulated in the same manner as
described above for a liquid. Also, in some cases solids, such as fats,
resins, waxes, or like materials, can be extruded and coated while in the
molten state [37]. Obviously, other particulate solids can also be included
in the molten solid. Substances in the molten state can also be used as
coating materials.
The initially employed gravity flow and simple extrusion devices were
found to have some general processing problems. Thus, in gravity flow
systems it is apparent that certain limitations are imposed upon the size of
the capsule produced. Also, with simple extrusion devices capsule uniformity
is a problem. To overcome such problems and yet maintain a high output
of product, devices utilizing centrifugal force and having multiorifice
extrusion heads, as shown in Figure 21, were designed. It is important to
note here that the capsule size can be controlled by varying such factors
as the orifice size, the rotational speed J and the flow rate of the fluid to
be encapsulated. As would be expected, with a uniform orifice size and a
constant rotational speed, an increase in the fluid flow rate results in an
increase in the capsule size. Conversely, with a uniform orifice size and
a constant fluid flow rate, an increase in the rotational speed yields smaller
Material to be
coated
I
I f
=J-=- I --
- -
.....__ Coating
fluid
Coated
particle ----.
Figure 21 Schematic drawing showing one nozzle unit of a multiorifice
centrifugal extrusion head.
Particle-Coating Methods 187
capsules. The centrifugation method is capable of producing microcapsules
in the 100- 200 um range.
A still further modification [39) of this method, the so-called submerged
nozzle, utilizes three concentric tubes in which the outermost tube
contains a carrier fluid. As the fluid to be encapsulated is extruded from
the innermost tube. it is surrounded by the coating fluid in the center
tube and encapsulation is effected in a manner similar to that described
above. However, in this case the carrier fluid in the outermost tube carries
the microcapsules away from the nozzle. Naturally. the carrier fluid
must not be miscible with the coating fluid. This modification of the
process is stated to be most useful when the fluid coat is fragile and ruptures
upon impact with the hardening bath used with the centifugal devices.
The size of the microcapsules produced in this system is again a
function of the nozzle size and fluid flow rate; however, the carrier fluid
flow rate is also important in this regard. Thus, an increase in the carrier
fluid flow rate, other factors remaining constant, results in a decrease
in the capsule size.
IV. SPRAY DRYINC
The application of the spray-drying process to the production of microcapsules
has been found to be useful for a variety of materials and can be
Water-immiscible liquid
Ol'
water- insoluble solid particles
Aqueous coating
solution
Prepare an O/W emulsion or an
aqueous suspension of the solid
I Spray-dry o/»: emulsion
or
aqueous suspension
J
Product
Frae-flowtng dry powder of
encapsulated Iiquid or coated solid
Figure 22 Flow chart for spray-dry process of coating liquid or solid
particles.
188 Wurster
Fluid
Exhaust
Centrifugal
collector
Blower - .....,
t
Atomizer
...
--,..,..
/ I
/'
'.(
I " , -
\ /' - ......,
".......
I --
\
\ ....... -,
)... J
\ -.-
\
\,I
+
Filtered----...- ...
heated__•
inlet air__II
DrV
product
Figure 23 Schematic diagram of spray-drying apparatus.
employed to encapsulate both liquids and solids [1,41- 46]. In the case of
liquids. the fluid to be coated is first emulsified. The film-forming coating
material is contained in the continuous phase of the emulsion and the
fluid to be coated becomes the dispersed phase. The emulsion is subjected
to the spray-drying process and the solvent constituting the continuous
phase is removed. The film former is thus deposited upon the surface of
the dispersed particles. The resulting product is a free-flowing powderlike
material containing the encapsulated liquid. Essentially, the same
procedure is followed when this method is used to produce coated solids.
Here, however, a suspension of the solid particles in a solution containing
material is first prepared. Solvent removal during the spray-drying
process deposits the coating material on the surface of the solid particles.
Particle-Coating Methods 189
The spray-dry process usually produces coated aggregates rather than
coated single particles. A simplified flow chart for the overall process is
shown in Figure 22, and a schematic drawing of the spray-drying equipment
is presented in Figure 23. Since emulsion systems are employed in
the interfacial polymerization microencapsulation method (see page 194),
the spray-drying process may also be utilized to recover the microcapsules
prepared by this technique. The principles governing the spray-drying
process are well known and need not be treated here.
V. AQUEOUS-PHASE SEPARATION, COACERVATION
In the coacervation coating method* [1,47 - 61], an aqueous colloidal solution
of a hydrophilic polymer having film-forming properties is employed.
The material to be coated may be either a liquid or a solid. When liquids
are microencapsulated an oil-in-water-type emulsion is prepared, with the
liquid to be coated occupying the dispersed phase and the aqueous colloidal
solution the continuous phase. When solids are coated, a suspension
of the particulate solid in the aqueous colloidal solution is prepared. In
either case, by suitable means, phase separation is produced with a layer
of the coacervate derived from the hydrophflic colloid forming around the
dispersed particles. Finally, the coacervate layer is treated to cause it
to become more firm. The process can be one of simple coacervation involving
only a single colloidal solute, or complex coacervation utnizing two
or more colloidal solutes can be employed.
The steps involved in the coating process are presented in a general
manner in Figure 24.
In simple coacervation, the single colloid contained in the continuous
phase can be caused to deposit on the surface of the dispersed particles
as a result of such physical influences as salting-out effects or the addition
of another solvent. Such influences, of course, result in the dehydration
of the hydrophfiic colloid. Gelatin is commonly employed as the
hydrophilic colloid here, since the above effects are readily observed by
the addition of a highly soluble salt such as sodium sulfate or ammonium
sulfate, or by the addition of alcohol. A flow chart of the simple coacervation
process is shown in Figure 25. The process allows the use of other
hydrophfiic film-forming colloids as well as other dehydrating and filmhardening
agents.
A. Particle-Coating Methods
Complex coacervation, in contrast to simple coacervation, requires at least
two hydrophilic colloids in the continuous phase of the fluid system. This
type of procedure allows the application of the well-known phenomenon
whereby two oppositely charged colloids discharge with each other to produce
a coacervate, as is shown in the following general-type reaction:
[Colloid 1]+ + [colloid 2]- ---- [colloid 1]+[colloid 2]- ~
*National Cash Register Company, Dayton, Ohio.
190 Wurster
+---......1-Dispersed phase
-oooo1l-Aqueous continuous phase -0 - 0 - containing fllm·former
_ _ (gelatin)
Phase separation
(3 phasel
Coacervate _ ..._0.0-~0 ~:: ~
o 0 ° -0-0
o ° 0 0
Hardened -II---~Io....o'
film
Continuous
phase
dispersed ----It-,,,,,,,"
phase
Coacervate
film
\1~~-=_,._Dispersed
phase
OIWemulsion
(2 phase)
Coacervate film
on particles
Dispersed
phase-lt-....
Coated particles
Figure 24 Schematic diagram of steps involved in coacervation coating.
Unless the coacervate remains highly hydrated it will separate from solution.
Thus, when such a weak union between the charged colloids occurs
fine droplets initially appear which subsequently coalesce to produce a
continuous phase. Figure 26 is a flow chart for the complex coacervation
process.
The process is a batch method and can be conducted on an experimental
basis with ordinary laboratory equipment. Pilot plant and production-
scale operations require only such simple equipment as suitable tanks,
ffitration, or other separation equipment and means to control such process
conditions as pH and temperature.
The particle size of the dispersed phase (i.e •• the solid particles or
liquid droplets to be coated) will govern to a large degree the size of the
coated particles produced. From this it is evident that coated particles,
varying widely in size, can be prepared. The literature indicates this
size range to be 5- 5000 '11m in diameter [2]. The thickness of the coat is,
of course, a function of the amount of coating material on the surface of
the particle. In normal commercial coating procedures the weight of the
Particle-Coating Methods
Water-immiscible liquid
or
water-insoluble solid particles
Aqueous coat in/;
solution
(gelatin 10% w/wj
191
Prepare 1 liter of an o/w emulsion (20% dispe rsed phase)
or
aqueous suspension of solid particles
j
Add slowly" ith mixing a 20% "Iw ~a2S04 solution
to produce coacervation (approx, 400 mil;
maintam temperature at 50'C above this point
j
Gel the colloid by pouring coacervate mixture into approx, 7-fold
its volume of 7a;, wlw :\a2S04 solution. At this point
and below maintain the temperature at 19°C or le-,s
j
Filter and wash coacervate with cold water
to remove salt
j
Treat Ii lte red mater-ial wtth J7% Ior rnaldehvde
(approx , 2 liters) to harden coacervate
j
F iller and" as h part ic les \\ ith cold \\ atl' r
to remove hardentna .urent
j
Dn to remove remaining solvent
Figure 25 Flow chart for simple coacervation process.
192
Water-immiscible liquid
or
water-insoluble soUd particles
Aqueous coating solution
(acacia 10% w/w) adjust to
pH 6.5 with 10% NaOH
Wurster
Prepare 1 liter of O/W type emulsion
(30% dispersed phase) or aqueous suspension of
the solid particles
I Add 700 ml of 10% gelatin (isoelectric point pH 8.0)
solution with mixing
I Add warm water until a coacervate is produced. The collotd
concentration is reduced to approximately 3% (the less water the
finer the particles). Adjust pH to 4.0 to 4.5 by gradual addition of
10% acetic acid. Maintam temperature at 50·C for above steps
I Pour coacervate mixture into 2 times its volume of cold
water. Maintain temperature at S·C or below in this
and subsequent step I Treat coacervate with 37% formaldehyde solution. Adjust
pH to 9.0 with 10% NaOH (a macromolecular electrolyte such
as CMC has been used in this step in place of NaOH
to inhibit aggregation
I Remove aggregate of encapsulated material and wash with water I Dry and comminute aggregated material
Figure 26 Flow chart for complex coacervation process.
Particle-Coating Methods 193
coat varies between 3 and 30% of the particle weight. but coatings of
1-70% have been applied.
The flexibility of this coating process is further emphasized by the
wide variety of coating materials that can be utilized. A partial list of
suitable coating materials which can be employed. depending on whether the
aqueous-phase separation or nonaqueous-phase separation method is used,
follows:
Gelatin
Gelatin /acacia
Gelatin/acacia/vinylmethylether maleic anhydride
Gelatin I acaciaI ethylenemaleic anhydride
Carboxymethylce1lulose
Propylhydroxycellulose
Polyvinylalcohol
Cellulose acetate phthalate
Bthylcellulose
EthylenevinyIacetate
Nitrocellulose
Shellac
Wax
While the coacervation method is capable of coating a large number of
materials. it is readily apparent that the material to be coated must be
stable to the process conditions. Particular consideration must therefore
be given to the influence of temperature, pH, solvent system, etc., on
product stability.
VI. NONAQUEOUS-PHASE SEPARATION
Phase separation [2.62,63] can be induced in organic solvent systems as
well as aqueous systems. Again, the suspended particle may be either a
solid or an immiscible liquid. Thus, for dispersed liquids, emulsions of
the W10 type are employed in contrast to the more conventional coacervation
method in which 0 IW-type emulsions are used. Obviously. watersoluble
solids can also be included either by suspending them in the organic
solvent or by dissolving them in the internal phase of the W10-type
emulsion.
The polymeric coating material is dissolved in the continuous organic
solvent phase. Phase separation and the coating of the solid or liquid
dispersed particles is accomplished by adding a solvent which is miscible
with the continuous organic solvent phase but which is a nonsolvent for
the polymeric coating material. Triangular phase diagrams are a most useful
tool when dealing with phase separation in either nonaqueous [62] or
aqueous [64] systems.
Hardening of the polymer coat on the particles is usually accomplished
either by the addition of more of the nonsolvent or by separating the
coated particles from the system. washing them with a nonsolvent liquid,
and drying.
194 Wurster
VII. INTERFACIAL POLYMERIZATION
The interfacial polycondensation or polymerization technique [63 - 65] has
been applied to the microencapsulation of various liquids. However. as
with the other processes discussed in this chapter. particulate solids can
be suspended in the liquid SO that the microencapsulated particle contains
both the solid and the liquid.
The process consists of bringing two reactants together at the interface
of the dispersed and continuous phases in emulsion systems [66].
This is usually accomplished by emulsifying the liquid containing the first
reactant (dispersed phase) into the continuous phase, which is initially devoid
of the second reactant. Additional continuous phase containing the
second reactant is then added. This interfacial polymerization reaction
produces a continuous fnm of the formed polymer around the dispersed
phase. A diagram of the interfacial process is shown in Figure 27. The
recovery of the microcapsules from the continuous phase can be accomplished
by spray drying, flash evaporation, filtration, Or other separation
techniques [67]. A flow chart of the process is shown in Figure 28.
The various polymer-coating materials which have been utilized to prepare
microcapsules by the described process include polyamides (nylon).
polyurethanes. polysulfonamides, polyesters, polycarbonates, and polysulfonates.
The particle size of the product created by this method varies
in accordance with the particle diameter of the dispersed phase. Therefore,
particles varying greatly in size. from approximately 3 to 2000 um in
diameter, can be prepared.
Water-soluble monomer
.1IllI....-- Interfacial
reaction
Oil soluble 0 42 L
Di = initial dose = 500 mg or 5,000.000 J.lg
It is desired to formulate this drug into a sustained-release product releasing
drug over a 12-h period such that the serum level is maintained at
10 llg mI- l • Step 1: Using equation 2 from the text
Rate in = rate out = k 0 = CtkelVd r
-1 -1 =10 ].lg ml x 0.2 h x 42,000 ml
-1 =84,000 ].lg h
Sustained Drug Release from Tablets and Particles 205
o Step 2: After calculating the zero-order release rate constant. k r • calculate
the total dose per tablet. Using equation 3 from the text
W = D. + k ~
1 r
-1 = 500,000 llg + (84,000 g h x 12 h)
= 1.508.000 ug'
=1.5 g per tablet
Since a large tablet will be generated. the formulator may wish to prepare
O.75-gm tablets so that the patient takes two tablets every 12 h.
Conversely, it is possible to reduce the size of the tablet by recalculating
for a shorter sustaining time. For example. suppose 6 rather than 12 h of
sustaining time is used.
-1 W= 500,000].lg + (84,000 g h x 6 h)
= 1.004,000 ug
= 1.0 gm per tablet
Suppose the formulator finds that the particular sustaining mechanism is
better approximated by first-order rather than zero-order release. From
the earlier calculations. the formulator knows that for 12 h of release the
tablet should contain about 1.5 gm and the zero-order release rate is
84.000 1Jg h -1. From these data it is possible to approximate the firstorder
rate constant k using equation 4
r
1 kelCtVd
k =::-::"-~r
W - D
i
-1 -1 = 0.2 h x 10 llg ml x 42.000 ml
1,508,000 - 500,000 llg
B. Drug Properties Considerations
There are a number of physical-chemical and derived biological properties
of the drug that either preclude placement of the drug in a sustainedrelease
system or have an adverse influence on product design and performance.
Some of these considerations are listed in Table 2. With almost
all of these properties we refer to them as restrictive factors. making
formulation of a sustained-release system difficult, but not impossible.
Thus, by changing the type of sustaining mechanism, the dose. or the
Table 2 Drug Properties Adversely Influencing a Sustained-Release Dosage Form
Property Explanation
~
~
Physical-chemical properties Dose size
Aqueous solubility
Partition coefficient
Drug stability
If an oral product has a dose size greater than 0.5 gm, it is
a poor candidate for a sustamed-release system, since addition
of the sustaining dose and possibly the sustaining mechanism
Will, in most cases, generate a substantial volume product that
will be unacceptably large.
Extremes in aqueous solubility are undesirable in the preparation
of a sustained-release product. For drugs with low water
solubility, they will be difficult to incorporate into a sustainedrelease
mechanism. The lower limit on solubility for such product
has been reported [30] to me 0.1 mg/ml. Drugs with
great water solubility are equally difficult to incorporate into
a sustained-release system [311. pH-Dependent solubiltty ,
particularly in the physiological pH range, would be another
problem because of the variation in pH throughout the GI
tract and hence variation in dissolution rate [31].
Drugs that are very lipid soluble or very water soluble, i.e.,
extremes in partition coefficient, will demonstrate either low
flux into the tissues or rapid flux followed by accumulation in
the tissues. Both cases are undesirable for a sustainedrelease
system [32, 33] •
Since most oral sustained-release systems, by necessity fare
designed to release their contents over much of the length of
the GI tract, drugs which are unstable in the environment of
()
;::T'
I:l
~
I:l
;::l
Q.
~.... ;::l
Cf.l
C
;::l
Biological properties Absorption
Distribution
Metabolism
Duration of action
Therapeutic
the intestine might be difficult to formulate into prolonged
release systems [34]. Interestingly, placement of a labile
drug in a sustained-release dosage form often improves the
bioavailability picture [31].
Drugs that are slowly absorbed or absorbed with a variable
absorption rate are poor candidates for a sustained-release
system. For oral dosage forms, the lower limit on the absorption
rate constant is in the range of 0.25 h- 1 [31] (assuming
a GI transit time of 10-12 h)
Drugs with high apparent volumes of distribution, which in
turn influences the rate of elimination for the drug. are poor
candidates [31].
Sustained-release systems for drugs which are extensively
metabolized is possible as long as the rate of metabolism is
not too great nor the metabolism variable with GI transit or
other routes.
The biological half-life and hence the duration of action of a
drug obviously plays a major role in considering a drug for
sustained-release systems. Drugs with short half-lives and
high doses impose a constraint because of the dose size
needed and those with long half-lives are inherently sustained
[16,35].
Drugs with a narrow therapeutic range require precise control
over the blood levels of drug, placing a constraint on sustained-
release dosage forms.
~
OJ
E' -::l
(Il
1:4
o2
~
~
(Il
(i)
~
OJ
(Il
~
::I
§:
(i) .... OJ
s
1:4
;§'
'"'$ ....
0"
(i)
OJ
~
C
"'1
208 Chang and Robinson
route of administration, it might be possible to generate a sustained-release
system. Frequently, a seeminly undesirable property of a drug or
dosage form can be overcome or minimized by placement of the drug in a
sustained-release system. For example, low drug bioavanability due to instability
may sometimes be overcome by placement in a sustained-release
system [311.
III. FABRICATION OF SUSTAINED-RELEASE
PRODUCTS
Having established the desirable concentration versus time proflle as depicted
in Figure 1, we can now ask two related questions: What approaches
can be taken to achieve this type of profile, and how should the sustained-
release product be constructed? Our concern, therefore, is to
examine the potential mechanisms available and to describe the general
nature of the dosage form construction.
A. Repeat-Action Release and Continuous
Release
To maintain the drug level at a constant desired value, we can employ
frequent dosings of drug to generate a series of peaks and valleys in the
blood level proflle, whose mean value lies on the plateau of the ideal case.
This approach is shown in Figure 2. The success of this approach depends
on the frequency of the multidoses because. obviously, the more
frequent the dose the smaller the peaks and valleys and the closer the
extreme drug level will adhere to the plateau value. This is the approach
taken with the Spansules, where four dosage units were employed in each
Spansule. One dose unit provided drug in a nonsustained form to establish
the initial blood level of drug and the other three doses were intended
to release drug at 2-, 4-, and 6-h intervals. Depending on the
drug properties, other intervals can be used, as can a greater number of
repetitive doses, although more than four becomes impractical from a manufacturing
standpoint.
An alternate approach is to employ a continuous release of drug. With
this method, a nonsustained portion of the dosage form is needed to rapidly
establish the therapeutic level of drug in the blood, and then by
some suitable mechanism, drug is continuou sly provided in a zero- or
first-order fashion. In this regard it is appropriate to comment on the
drug concentration versus time profile for a system releasing drug via
first-order kinetics. Contrary to the case where drug is released via
zero-order kinetics, as depicted in Figure 1, first-order release produces
a more or less bell-shaped proflle. Whether the bell shape is symmetrical
or skewed and whether it is narrow or wide depends on the drug and
the sustained-release system employed; that is, the kinetics of release.
Examples of this type of profile will be discussed and demonstrated later
in the chapter.
B. Mechanisms of Sustained Release
A zero-order release of drug is needed for the dosage form, which means
that the rate of drug release is independent of drug concentration
Sustained Drug Release from Tablets and Particles 209
Toxic level
Minimum Effective \
level \
\
\
\
\
\
\
\
....
\...
E
:l
C
.a1--4-----------------------"'""':~- s
~
88
C>
2 o
Time )
Figure 2 Repetitive release approach to sustained-release. The dotted
line represents the ideal sustained-release profile.
de =k 0 (5)
dt r
or expressed in amounts
(6)
At times it is not possible to generate a constant-release product and a
slow first-order release of drug is employed. A slow first-order release
will approximate a zero-order release as long as only a fraction of drug
release is followed [291; that is. less than one half-life is followed.
To attain a zero-order release rate, we have several mechanisms and
dosage form modifications that we can employ. We will restrict our coverage
of potential mechanisms to those that can be employed in the coating
approach to sustained release.
210 Chang and Robinson
Diffusion
A number of sustained-release products are based on diffusion of drug.
The follOWing discussion. although somewhat naive, will bring into perspective
those properties that should be considered in the diffusion
approach.
Fick's first law of diffusion states that drug diffuses in the direction
of decreasing concentration across a membrane where J is the flux of the
drug in amount/area-time.
dC J =-Ddx
where D is the diffusion coefficient in area/time, C is the concentration,
and x is the distance. Assuming steady state, equation 7 can be integrated
to give
6C
J = -D T
or expressed in more common form when a water-insoluble membrane is
employed
elM ADK 6C
dt = i
( 7)
(8)
( 9)
where A is area. D is diffusion coefficient, K is the partition coefficient of
drug into the membrane, i is the diffusional pathlength (thickness of coat
in the ideal case), and !l C is the concentration gradient across the
membrane.
In order to have a constant rate of release, the right-hand portions
of equations 8 and 9 must be maintained constant. In other words, the
area of diffusion, diffusional path length, concentration increment, partition
coefficient. and diffusion coefficient must be invariant. Usually, one
or more of the above parameters will change in oral sustained-release
dosage forms giving rise to non-zero-order release.
The more common diffusional approaches for sustained drug release
are shown in Schemes 3 and 4. In most cases, the drug must partition
into a polymeric membrane of some sort and then diffuse through the membrane
to reach the biological milieu. When the tablet or microcapsule contains
excess drug or suspension. a constant activity of drug will be maintained
until the excess has been removed, giving rise to constant drug
release. In Scheme 3 the polymer is water insoluble. and the important
parameter is solubility of drug in the membrane. since this gives rise to
the driving force for diffusion. In Scheme 4 either the polymer is partially
soluble in water or a mixture of water-soluble and water-insoluble
polymers is used. The water-soluble polymer then dissolves out of the
film. giving rise to small channels through which the drug can diffuse.
The small channels would presumably give a constant diffusional path
length. and hence maintain constant conditions as described earlier. Although
diffusion through the channels should be much more rapid than
diffusion through the membrane noted in Scheme 3, it is possible to have
a situation whereby membrane diffusion, being quite rapid in this ease,
Sustained Drug Release from Tablets and Particles
Membrane Reservoir
211
Scheme 3 Diffusion control of drug release by a water-Insoluble polymer.
is within an order of magnitude of pore diffusion. In this event, both
types of diffusion, membrane and pore, will provide contributions to the
overall diffusion rate and the equations would have to be modified to account
for these combined effects.
Dissolution
In this case, the drug is embedded (coated) in a polymeric material and
the dissolution rate of the polymer dictates the release rate of drug. The
Membrane Reservoir
b~~J.~
a ::. :.\)
D ' .
\)~ . ' .... : .
~. "O~
c:::J Pores produced by soluble portion
of polymer membrane
Scheme 4 Diffusion control of drug release by a partially water-soluble
polymer.
212
Membrane Reservoir
Chang and Robinson
Scheme 5 Dissolution control of drug release via thickness and dissolution
rate of the membrane barrier coat.
Scheme 6 Dissolution control of drug release via polymer core erosion or
polymer-coating erosion. (Note: Although equations are based on spherical
tablets, erosion of conventional flattened tablets is similar but there is
a combination of various R terms where R depends on the axis being
measured. See Ref. 38. It is assumed that diffusion of water, metabolites,
or biologically active components is not rate limiting.
Sustained Drug Release from Tablets and Particles 213
drug release rate, if governed by erosion or dissolution, can be expressed
as
dM =A dx fCC)
dt dt (10)
where (dx)/(dt) is the erosion rate, fCC) is the concentration profile in
the matrix, and A is area. A constant erosion rate can produce zeroorder
release kinetics, provided the drug is dispersed uniformly in the
matrix and area is maintained constant [36,37). Oftentimes, swelling of
the system or a significant change in area produces non-zero-order release.
The common forms of dissolution control are shown in Schemes 5 and 6.
In Scheme 5 we have a barrier coat across a microcapsule or nonpareil
seed containing drug, and the release of drug is dictated by the dissolution
rate and thickness of the barrier coat. Varying the coating thickness,
or layering concentric spheres of coating material and drug reservoir material,
gives rise to different release times, producing the repeat action
dosage form. Once the polymer has dissolved, all of the drug contained
in the capsule or seed is available for dissolution and absorption. In
Scheme 6 the drug is either embedded in a polymer or coated with a watersoluble
polmyer, which in turn is compressed into a slowly dissolving tablet.
The release rate is controlled by the dissolution rate of the polymer
or tablet.
Osmosis
Placement of a semipermeable membrane around a tablet, particle, or drug
solution, which allows creation of an osmotic pressure difference between
the inside and outside of the tablet and hence "pumps" drug solution out
of the tablet through a small orifice in the coat, can be used as a sustained-
release mechanism. The key component of the system is the ability
of a drug solution to attract water through a semipermeable membrane by
osmosis. Since the drug solution is contained within a fairly rigid system,
drug solution can be pumped out of the tablet or particle at a controlled
constant rate if a small hole is created in the coating surface and a constant
activity of drug, that is, excess drug, is maintained. Controlling
the rate of water imbibition thus controls the rate of drug delivery. This
can be seen in the following expression (12)
dV = k A (6 1l' - 6P)
dt t
(11)
where dVI dt is the flow rate of water, k , A, and t are the membrane
permeability, area, and thickness, 6 11" is the osmotic pressure difference,
and liP is the hydrostatic pressure difference. Keeping the hydrostatic
pressure small relative to the osmotic pressure, equation 11 reduces to
dV A -= k - (61l') dt .-
(12)
By maintaining the right-hand side of equation 12 constant, a zero-order
release system will result.
214 Chang and Robinson
COATING METHODS. Coating is a versatile process which imparts
various useful properties to the product. Modifications of drug-release
patterns such as enteric release, repeat-action release, and sustained
release are the most important pharmaceutical applications of coating. Although
many coating techniques, for various purposes, have been detailed
in previous chapters, we have elected to discuss the coating methods for
the purpose of modified release.
Pharmaceutical coatings have been classified into four basic categories:
sugar coating, microencapsulation, film coating, and compression coating.
Sugar coating of compressed tablets and granules is regarded as the oldest
process, involving the multistage build up of sugar layers through deposition
from an aqueous-coating solution and coating powder. Sugar coating
falls outside the scope of this chapter owing to the inability to control
drug release through a highly water-soluble sugar barrier.
A. Microencapsulation
Microencapsulation is a process in which tiny particles or droplets are surrounded
by a uniform coating (so-called microcapsule) or held in a matrix
of polymer (so-called microsphere). A number of microencapsulation techniques,
including aqueous phase separation, three-phase dispersion, organic
phase separation, and interfacial polymerization, have been used to
encapsulate pharmaceuticals and to retard the liberation of drug from
microcapsules.
Aqueous phase separation methods to prepare microcapsules Include the
simple coacervation of hydrophilic colloids with ethanol or sodium sulfate as
dehydrating agents. In addition, one can employ a complex coacervation
of two dispersed hydrophilic colloids of opposite electric charges with subsequent
pH change. Aqueous phase separation was patented as an encapsulation
process by Green in 1955 [39]. and has been used commercially
since that time for numerous applications.
The three-phase dispersion method involves dispersion of the materials
to be coated in a nonsolvent liquid phase containing colloidal or filmforming
materials such as gelatin. This two-phase system is then dispersed
in a third phase by emulsification. through spraying or other
means. and the coating material is gelled, usually by cooling. The resulting
gelled droplets are then either separated and dried or partially dehydrated
with an aliphatic alcohol and then separated and dried. Variations
on this process have been used commercially to encapsulate oilsoluble
vitamins such as vitamin A. One of the earliest applications of
this method was reported in a British patent in 1938 [40].
Generally, aqueous phase separation and three-phase dispersion methods
are used to encapsulate water-insoluble or poorly water-soluble materials
which are usually poor drug candidates to incorporate into a sustained-
release system. Gelatin is the most commonly used microencapsulating
agent for the two processes described above. Even for poorly
water-soluble drugs, gelatin is usually not able to achieve the desired
dissolution profile owing to the hydrophilic properties of gelatin. The use
of formalin in treatment to cross-link the gelatin and/or dual coating of
gelatin microcapsules may be necessary to improve the dissolution profile.
In addition, these encapsulation processes would be precluded for heatsensitive
materials and substances with a stability and/or solubility problem
at the pH of coacervation.
SUstained Drug Release from Tablets and Particles 215
Nonaqueous phase separation involves dispersion of core material in
an organic continuous phase in which the wall-forming polymer has been
dissolved. Phase separation is induced by the addition of a nonsolvent,
incompatible polymer, inorganic salt, or by altering temperature of the
system. The organic phase separation method can be employed in the
manufacture of microcapsules using various water-insoluble polymers as
coating materials to enclose water-soluble drugs and to slow the drugrelease
rate.
Encapsulation via interfacial polymerization was pioneered by Chang
[ 41, 42] in work designed to produce artificial cells. This method involves
the reaction of various monomers, such as hexamethylene diamine
and sebacoyl chloride. at an interface between two immiscible liquid
phases to form a film of polymer that encapsulates the dispersed phase.
Capsules formed by interfacial polymerization usually have relatively thin
semipermeable membranes highly suited for artificial cell studies or applications
requiring permeable walls. This method may give rise to questions
about toxicity of the unreacted monomer, the polymer fragments,
and other constituents in the process, instability of the drug in the reaction
medium during the polymerization period, fragility of the microcapsules,
and high permeability of the coating to low molecular weight
species [43]. Various other polymerization procedures, tneluding bulk,
suspension, emulsion, and micellar polymerization, have received considerable
academic interest to entrap active materials in polymer matrices.
However, inherent problems of polymerization procedures such as impurities
in the system, limited drug solubility in the monomer, excessive
drug degradation caused by reaction with the monomer or initiator. and
possible entrapment of drug in polymers may prevent the pharmaceutical
industry from actively engaging in this approach to control drug release.
In general, microencapsulation techniques using liquid as a process
medium are rather complicated and difficult to control. Several process
difficulties such as hardening of the capsule shell, isolating the microcapsules
from the manufacturing vehicle, and drying the microcapsules to
form free-flowing powder should be solved in order to ascertain batch to
batch uniformity. Equipment required for microencapsulation by these
methods is relatively simple: It consists mainly of jacketed tanks with
variable speed agitation. The process can be carried out on a production
scale with good reliability. reproducibility. and control. However,
process control, product quality control, and scale-up problems appear to
be the limiting factors influencing general acceptance by the pharmaceutical
industry.
The most common mechanical microencapsulation process is the spraydrying
technique, which consists of rapid evaporation of the solvent from
the droplets. Spray-drying techniques may produce monodispersed freeflOWing
particles which can be directly compressed into tablets, filled into
capsules, and suspended in water. However, the microcapsules obtained
by spray drying tend to be very porous because of rapid volatilization
of the solvent. Spray-congealing techniques accomplish coating solidification
by thermal congealing of the molten coating materials such as hydrogenated
castor oil, cetyl alcohol, monoglyceride and diglyceride , etc.
Spray-congealed coatings are less porous but require coating materials
that melt at moderate temperature.
Successful attempts using spray-drying and congealing techniques to
control the release of sulfa drugs have been reported [44 - 47] .
216 Chang and Robinson
B. Film Coating
Film coating involves the deposition of a uniform film onto the surface of
the substrate, such as compressed tablets, granules, nonpareil pellets,
and capsules. Intermittent manual application or continuous spraying of
coating solution onto a mechanically tumbled or fluidized bed of substrates
allow the coating to be built up to the desired thickness. Because of the
capability of depositing a variety of coating materials onto solid cores,
this process has been widely used to make modified-release beads and
tablets in the pharmaceutical industry.
Properly designed film coating can be applied to pharmaceutical products
to achieve performance requirements such as rapidly dissolving coatings,
sustained- or controlled-release coatings, and enteric coatings. The
polymer(s) used in coating formulations is the predominant factor for the
properties of the film coat. Water-soluble fflm formers such as methyl cellulose,
hydroxypropyl methylcellulose, hydroxypropylcellulose, polyethylene
glycol, polyvinyl pyrrolidone, etc , , form a rapidly dissolving barrier.
Enteric materials such as cellulose acetate phthlate, polyvinyl acetate
phthalate, methacrylic acid ester copolymers, etc., form acid-resistant
films. The hydrophobic water-insoluble polymers such as ethyl cellulose,
cellulose acetate. cellulose triacetate, cellulose acetate butyrate, and methacrylic
acid ester copolymers are used to extend the release of drug over a
long period of time. Depending upon the physicochemical properties of the
drug and the substrate formulation, several coating approaches have been
employed to regUlate drug release.
Partitioning Membrane
Partitioning membranes, continuous hydrophobic polymeric films which remain
intact throughout the gastrointestinal tract, can be applied onto the
coating substrate by using a single polymer or a combination of waterinsoluble
polymers. Since drug molecules cross the membrane by both a
partition and a diffusion process, solubility of the drug in the polymeric
material is a prerequisite to permeation. The polymeric material should be
carefully selected to have the desired permeability to the drug and water
in order to achieve the desired release profile, In addition to thickness
of membranes, the permeability of the film can also be adjusted by mixing
two water-insoluble polymers in any desired proportion.
Dialysis Membrane
Frequently, the partitioning membrane is too effective to regulate drug release.
In other words, the drug within the coating would be released
very slowly or be released not at all for a long period of time. The inclusion
of hydrophilic additives within the coating along with hydrophobic
polymer(s) creates pores when the additives are dissolved by water, which
guarantees the penetration of water and elimination of drug entrapment.
When the drug molecule leaves the membrane by diffusing through pores
filled with dissolution media (dialysis mechanism) the size of the drug molecules
and solubility of the drug in a dissolution medium are important
factors in transport. Some water-soluble additives such as sodium chloride,
lactose and sucrose have poor sohrbilfty in organic solvents and may
be micronized and suspended in a solvent-based coating system. Watersoluble
polymers such as methyl cellulose. polyvinyl pyrrolidone, and
Sustained Drug Release from Tablets and Particles 217
polyethylene glycol are commonly mixed with hydrophobic polymers to regulate
drug release owing to their excellent film-forming properties and
solubility in organic solvents. In addition to film thickness, the ratio of
soluble components to insoluble polymer in the coating influences the release
rate. Porous membranes may also be prepared by incomplete coating
of hydrophobic polymers. However, strict process control is necessary to
ascertain the reprodUcibility owing to sensitivity to the coating weight.
Fat Wax Barrier
Mixtures of waxes (bees wax, carnauba wax, etc.) with glycerol monopalmitate,
cetyl alcohol, and myristyl alcohol can be applied onto the substrate
to form a barrier by hot-melt coating. Hot-melt coating is the most
economical process owing to elimination of solvent cost and the inexpensiveness
of the coating materials. However, a higher level of coating, compared
to polymeric film coating. is normally required to retard the liberation
of the drug.
Incorporation of Enteric Materials into
the Formulation
In general, pH-independent dissolution is the ideal attribute of a controlledrelease
dosage form. However, most drugs are either weak acids or weak
bases; their release from delivery systems is pH dependent. If the drug
has a higher solubility in acidic than in basic media. enteric material may
be incorporated into the rate-controlling barrier or core matrix to minimize
the effect of pH-dependent solubility. In another approach. physiological
acceptable buffering agents can be added to the core formulation to maintain
the fluid inside the rate-controlling membrane to a suitable constant
pH. thereby rendering a pH-independent drug release [48]. Enteric material
also can be incorporated into rate-controlling membranes or core
matrices to create pH-dependent release of the drug. The dosage form
with pH-dependent dissolution characteristics may be beneficial in some
cases to improve the extent of absorption by dumping the dose in time
and preventing the unabsorbed dose being entrapped in the stool. Sustained-
release preparations overcoated with enteric material can be utilized
as an intestinal delivery system with sustained-release properties. Entericcoated
dosage forms can be overcoated with a drug layer to form a repeataction
preparation.
CORE PREPARAT ION. Very few drug particles possess adequate
physicochemical properties for the usual coating process. These properties
include (1) suitable tensile, compaction, shear, impaction, and attrition
strengths to avoid destruction during the coating process; (2) approximately
spherical shape to obtain good flow and rolling properties in the
coating equipment; (3) suitable size and size distribution; and (4) suitable
density to avoid escape of the drug particles during the coating process.
Certainly. different types of coating equipment have their own capabilities
to handle the cores. leading to different requirements of the cores for
various equipment. For the pan coating process, a relatively large particle
size (larger than 500 urn) and a spherical shape are generally considered
necessary to provide excellent rolling in the coating pan and to avoid
the agglomeration and/or aggregation owing to inefficiency of drying and
long contact time among the cores. Although fluidized -bed coating
218 Chang and Robinson
systems expanded coating capabilities dramatically. suitable strengths and
weight of the cores are needed to avoid excessive attrition of drug particles
and suction of drug particles into the filter during the coating
process. The major advantages of using pure drug particles as coating
substrates are elimination of the core-making process and a less bulky final
product. Potassium chloride and acetyl salicylic acid crystals are typical
examples which have been satisfactorily coated in a coating pan or a
fluidized-bed coating system.
1. Compaction process. Apparently. tablets are the most common and
easiest dosage form to coat. Tablets with excellent friability. hardness,
and edge thickness are preferred for coating. However, sustained-release
film-coated tablets may prematurely dose dump due to accidental rupture
of the coating film. The use of a multiple-unit instead of a single-unit
dosage form is a pharmaceutical trend because of the presumed reduction
of the inherently large inter- and intrasubject variation linked to gastrointestinal
transit time. Tablets of small dimensions have been successfully
prepared from polyvinyl alcohol and subsequently cross-linked at the surface
to form a quasimembrane-controlled system for a multiple-unit dosage
form [49]. In order to achieve high output in large-scale manufacturing,
this approach to the preparation of cores may face formulation difficulties
and tablet tooling problems; i.e •• multitipped punches. The limited size
flexibility of the tableting method to manufacture cores for a multiple-unit
device is another disadvantage.
Other methods with production capacity. such as slugging, ehllsonator ,
and Hutt Compactor, can be used to produce granules in the compaction
mode. However, irregular-shaped granules are commonly observed and
extensive sieving is necessary to remove the fines and oversized granules.
2. Surface-layering process. Another common approach to core production
involves the use of substrates and enlargement of the substrates
by a surface-layering technique. Thus, nonpareil seeds of various sizes
or sugar crystals are used as the substrate. Application of an active
substance onto an inert substrate can be carried out by uniform coating
in a rotating coating pan in the presence of a suitable adhesive. In detail.
the substrate is uniformly wetted by manual spreading of a binding
solution. followed by attachment of the active substance to the surface of
the substrate. Commonly used adhesives include solutions of polyvinyl
pyrrolidone polyethylene glycols, cellulose ethers, natural gums, shellac,
zein, gelatin, and sugar syrup. Suitable binder(s) and solvent systems
for the binder( s) must be found in order to have smooth production of
cores without excessive agglomeration of the cores or separation of the
drug particles. Enteric binders. such as cellulose acetate phthalate and
shellac, may impart pH-dependent dissolution properties to the final product.
Separating agents, such as talcum and magnesium stearate, may be
used to eliminate or reduce tackiness of the adhesives. Trituration technique,
to blend the potent drug with the auxiliary agents, is usually required
to obtain a uniform distribution of the drug onto the substrate
surface.
The powder layering process requires a great deal of repetition, and
is thus time consuming. Moreover. undesired agglomeration or aggregation
and adhesion to the wall of coating equipment can occur. To avoid
the labor-intensive powder layering process in a coating pan. a centrifugal-
type fluidized bed with a powder feed device (CF-Granulator) has
Sustained Drug Release from Tablets and Particles 219
been used to produce high-quality pellets. The stirring chamber of the
CF-Granulator consists of a fixed specially curved wall stator and a directed
rotating plate rotor. Fluidization air, through a gap slit between
the stator and rotor, prevents the substrates from falling. The substrates
are whirled up along the wall of the stator owing to centrifugal
force of the rotor and to the upper part of the wall due to the fluidization
air. SUbsequently, they drop due to gravitational force. During the
spiral stirring operation, layering powder is metered to the fluidized bed
from a powder feed unit. A binder solution is sprayed from a spraying
gun to cause binding of the powder to the substrate surface.
It is also possible to dissolve or to suspend the drug in the binder
solution and to apply this liquid uniformly to the surface of the substrate
using a coating pan or fluidized-bed system. However, there are several
difficulties with this approach, including the tendency to clog the nozzle
with the slurry. A large quantity of solvent may be needed to dissolve
or suspend the active substance. In addition, drug loss to the air stream
and possible adverse aggregation of the cores can occur. Another inherent
disadvantage of the layering process is the possible formation of
unduly large pellets owing to the use of nonpareil seeds as a substrate.
Small-sized substrates, such as sugar crystals, can be used to eliminate
bulkiness of the pellets. The finer the substrate, the finer the drug particle
should be and the more difficult the process.
3. Agglomeration process. Alternatively, the drug particles in the
powder bed can grow by wet agglomeration. The extent of granule growth
depends on the amount of granulating solution, the type of binder, the
force of agitation, and heat applied. The conventional wet granulation
method to prepare the substrate for coating can give rise to problems
such as irregular-shaped particles with a coarse and porous surface, soft
and pliable particles, and a broad granular size distribution. Application
of relatively large amounts of coating material may be required owing to
capillary suction of the coating fluid to pores. The porous structure and
irregular shape of the granules may lead to an unpredictable sustainedrelease
coating. Additional powder layering to round off the granules
into a sphere or to smooth surfaces or perhaps to increase the strength
of the pellets may be important. However, Kohnle et al , [50] have successfully
produced microspherules which are suitable for enterlc- and
sustained-release coating through agglomeration of fine drug particles by
using a Twin Shell Blender with an intensifier bar assembly.
The inclined dish granulator or disc pelletizer is well known and high1y
utilized in the fertilizer, iron ore, and detergent industries. It also
has been adopted throughout the pharmaceutical, chemical, food, and
allied industries. The equipment, known as the nodulizer and pelletizer,
are available for continuous production of spherical pellets. The unit
normally consists of a shallow cylindrical dish, motor drive, adjustable
scrapers, spraying system, and powder feed device. As the pan rotates
about an inclined axis, the raw material bed is rolled by centrifugal force
and maintained as a uniform deposit of material onto the base of the pan
by a plow. Scrapers also prevent buildup of materials on the dish surface.
Powdered ingredients must be milled, premixed, and deaerated in
order to have a uniform chute fed to the unit. Powder materials are
continuously metered to the pan at a specific location, normally at a point
three-fourths of a radius unit from the top of the dish. The spray angle
depends to a large extent on positioning of the sprays and their distance
220 Chang and Robinson
from the bed powders; commonly, a 60 spray angle is used. The spray
droplet size should be adjusted according to feed size and desired granule
size. In other words, if small granules are desired, a fine droplet spray
should be used and vice versa. Also, the finer the feed material. the
finer the spray droplet.
There are several theoretical equations that can be used to calculate
the rotational speed of the pan. However, appreciable discrepancy exists
in theory and practice. Usually 20-30 rpm seems a reasonable rotation
speed. Following agglomeration, the finished granules are raked over by
the dish rim, and the rim height can be adjusted to control granule size.
Pronounced size segregation is the principal feature of dish granulation.
This ensures almost perfectly spherical pellets with a narrow particle size
distribution. Important parameters in the dish pelletizer include powder
feet rate, position of powder feed chute, spraying rate, position of the
spraying gun, spray nozzle size, angle of inclination, rotational speed, rim
height, powder bed depth, and pan size. These variables interact to
some extent to produce the final granules.
However, the noncontinuous granulation process in which spraying and
drying stages are alternately repeated has been employed to maintain constant
moisture levels and to produce high-density granules. A high level
of perfection in shape and size of the granules cannot be obtained by
agglomeration mechanism using a conventional fluidized-bed granulator.
Recently, modified fluidized-bed units. such as the Roto-Processor, the
Spir-A-Flow, and the Glatt Rotor Granulator/Coater, ell utilize a rotating
disc at the bottom of a fluidized-bed, replacing the air-distribution plate.
This modification supposedly combines the advantages of the dish granulator
and the fluidized-bed granulator. It has been demonstrated that the
rotary fluidizer-bed granulator can be used to produce spherical granules
with high density by an agglomeration mechanism [51]. In general, the
layering mechanism using nonpareil seeds or sugar crystals as a substrate
to build spherical-shaped pellets is relatively easy compared to the agglomeration
mechanism.
4. Bxtrusion-eptierontzatton process. Spherical pellets can also be
produced using an extrusion-spheronization process. The main processing
steps include (1) dry blending, (2) wet granulation, (3) kneading, (4)
extrusion, (5) spheronization, (6) drying, and (7) screening. A thoroughly
wet granulation containing the drug, diluent. and binder is forced
through a radial or axial extruder with a suitable die design, such as a
perforated die or multiple-hole die, by means of a screw feeder to produce
roughly cylindrical extrudates. The extrudate size and final pellet size
are determined by the size of the die used on the extruder. The pellet
mill, a radial extruder, was initially developed for the agricultural industry
to densify and upgrade the particle size of poultry and animal
feeds. In operation, the preconditioned material is fed continuously in a
controlled fashion to the pelleting chamber. The motor-driven outer perforated
die ring causes the roller( s), Which is mounted inside the die ring,
to turn. The feed, carried by the rotation of the die ring, is compressed
and forced through the holes in the die ring. As pellets are extz-uded , a
knife, or knives, mounted at the exterior of the die ring. cuts the pellets
to length. Application of the pellet mill to pharmaceutical products has
been extremely limited. However, it deserves special mention owing to its
ability to produce pellets with high density and low friability at a high
output.
Sustained Drug Release from Tablets and Particles 221
The extruded granules can be converted into consistently sized
spheres by use of a Japanese device called a marumerizer or an English
version of the device called a spheronizer. This device consists of a stationary
cylinder with a smooth wall and a grooved rotating disc. The
centrifugal and frictional forces, generated by the rough rotating baseplate,
spheronize and densify the extruded granule. A typical time for
the spheronization process would be approximately 5 min per batch, depending
upon the nature of the material. Recently, a unique device called
the Roto-Coil has been designed 8S a continuous spheronizer with no moving
parts. This device consists of a spiral-shaped pipe. The extrudate
is spheronized by passing through the pipe in a rotation movement with
the aid of negative pressure generated by a fluidized-bed system. The
advantages of the extrusion-spheronization process include: (1) production
of the spherical pellets without using seeds, leading to reduction of
the bulk of final product; (2) the ability to regulate size of the pellets
within a narrow particle size distribution; (3) the ability to produce highdensity,
low-friability, spherical pellets; and (4) the ability to achieve
excellent surface characteristics for subsequent coating, leading to an
homogeneous distribution of coating material( s) onto the spherical pellets.
It is also possible to agglomerate the finely divided solids into spherical
matrices from a liquid suspension [52]. In the spherical agglomeration
process, fine pow del'S are dispersed in the liquid. With controlled agitation,
a small amount of a second liquid, which is immiscible with the first
liquid and preferentially wets the solids, is added to induce formation of
dense, highly spherical agglomerates. Another approach to prepare
spherical matrices in the liquid state is the instantaneous crosa-Iinking of
drug-containing droplets of aqueous solution of sodium alginate in a suitable
hardening bath, such as calcium chloride solution. This entrapment
technique in recent years has become the most Widely used method for
immobilizing living cells and enzymes. It enjoys several advantages, such
as a simple production process, a wide size range of beads, and a narrow
size distribution [53]. The spray-drying process has been used to produce
a spherical matrix. However, the spray-dried cores may be porous
and fluffy because of rapid volatilization of the solvent. The spraycongealing
process, also called the prilling process, is well known in the
fertilizer industry for production of urea pellets and ammonium nitrate
pellets. This process can be defined as the process by Which a product
is formed into particles, usually of spherical shape, by spraying a melt of
the product into a chamber of suitable configuration through which cooling
air is passed. The process is fairly restricted to matrix materials
having suitable properties; namely, a high melting point and low heats of
crystallization [54]. These several techniques may have potential applicability
to produce spherical cores containing drug for SUbsequent sustainedrelease
coating, and thus deserve further investigation. The physical
properties of the product, such as pellet size and size distribution, pellet
density, strengths, globocity, pore and pore distribution I ate , , are different
for various methods and may be important for subsequent coating
as well as in product performance. Table 3 summarizes the various possible
methods to core production and some selected pharmaceutical examples.
LATEX/PSEUDOLATEX COATING. The strict air-quality controls
instituted by different federal agencies, spiraling solvent costs. the high
price of solvent recovery system, and potential toxicity as well as to some
Table 3 Possible Methods for Core Manufacture and Some Selected Pharmaceutical Examples t\,)
t\,)
t\,)
Method and example
Crystallization
Compressed tablet
Min-Tablet [49J
Granular
Layering [55J
Layering [56]
Equipment
Oslo crystallizer
Tablet press
Tablet press
Chilsonator, Hutt
Compactor
CF-Granulator
Fluidized-bed
coater
Drug
Potassium chloride
acetyl salicylic acid. etc.
Various
Diprophylline
Various
Theophylline, pseUdoephedrine
hydrochloride. diphenhydramine
hydrochloride
Dexamphetamine sulfate
Carrier binder
N/A
Various
Polyvinyl alcohol
Various
Gelatin, sodium
carboxymethylcellulose.
Kaolinl
Eudragit E-30D.
Povidone
Gelatin
Comments and
variable studied
Very few drug crystals
possess adequate properties
for coating.
Single-unit device. The
accidental rupture of the
film barrier may cause
premature dumping of
drug.
Relatively large matrices
for multiunit device.
Irregular shaped granules.
Extensive screening
and recycling of undesired
granules ,
Evaluation of CF-Granulator
for pelletization.
As cores for hot-melt and
cellulose acetate phthalate
coating to obtain sustained
reslease and enteric
properties.
Layering [57] Coating pan Sodfum salicylate Ethyl cellulose As cores for fluidized-bed
coating to obtain a product
with enteric and sustained-
release properties.
Agglomeration [58] Planetary mixer. Hydrochlorothiazide Microcrystalline
high shear mixer cellulose
Agglomeration [50] P.K. Blender with Aspirin, caffeine, phenyl- Povidone Extensive screening and
intensifier bar ephedrine hydrochloride. recycling of undesired
chlorpheniramine maleate granules.
Agglomeration [59] Dish granulator N/A Maize starch Evaluation and studies of
operating conditions for
the formation of pellets.
Agglomeration [51] Rotary fluidized Butalbital Lactose, corn Comparison of rotary
bed starch. and fluidized-bed granulator
povidone with conventional fluidized-
bed granulator.
Agglomeration [58] Rotary fluidized Hydrochlorothiazide Microcrystalline
bed cellulose
Extrusion-spheroniza- Extruder and N/A N/A General description and
tion [60] marumerizer or discu ssion of extru sion
spheronizer spheronization technology.
Extrusion-spheroniza- Extruder and Dibasic calcium phosphate. Microcrystalline Comparison of extrusiontion
[61] marumerizer or magnesium hydroxide, cellulose spheronization process
spheronizer sulfadiazine. aeatoamlno- with conventional wet
w
phen granulation.
w
~
Method and example
~
~
Table 3 ( Continued)
Equipment Drug Carrier binder
Comments and
variable studied
Extrusian-spheronization
(62]
Extrusian-spheronization
[63]
Extrusion- spheronization
[64]
Extrusion-spheronization
[651
Extrusion- spheronization
[66J
Extruder and
marumerizer or
spheronizer
Extruder and
marumerizer or
spheronizer
Extruder and
marumerizer or
spheronizer
Extruder and
marumerizer or
spheronizer
Extruder and
marumerizer or
spheronizer
N/A
Acetoaminophen
Theophylline. quinidine
bisulfate, chloropheniramine
maleate, hydrochlorthiazide
Acetoaminophen
N/A
Microcrystalline
cellulose. sucrose,
lactose
Microcrystalline
cellulose
Microcrystalline
cellulose, sodium
carboxymethylcellulose
Microcrystalline
cellulose, carboxymethylcellulose
Lactose, dicalcium
phoshydrate
povidone and
microcrystalline
cellulose
Effect of the spheronization
processing variables, dwell
time and speed, on the
final granule properties.
Effects of spheronization
processing variables including
water content, extrusion
speed. screen size,
spheronizer speed and
spheronizer time on tablet
hardness and dissolution
rate.
Effect of different diluents
and drug-diluent ratio on
the final granule properties.
Use of factorial design to
evaluate granulations prepared
by extrusionspheronizatton
.
Elucidation of the factors
that influence migration of
solvent-soluble materials
to the surface of beads
made by extrusion-spheronization.
~
~
til
Extrusion -spheroniza- Extruder and Theophylline, quinidine Microcrystalline Microcrystalline cellulose
tion [58] marumerizer or sulfate, chlorpheniramine cellulose has excellent binding
spheronizer maleate, hydrochloro- properties for granulation
thiazide to be spheronized.
Spherical-agglomera- Liquid agitator Sulfamethoxazole, White beeswax, Parameters affecting the
tion [52] sulfanilamide ethylcellulose size and release behavior
of resultant matrix.
Gellation [53] Dripping device Enzymes and living cells Calcium alginate Widely used method for
and calcium chlor- immobilizing living cells
ide solution as a and enzymes. Possible
hardening agent pharmaceutical application
for pellet making.
Gellation Dripping device Diphenhydramine chloride Gelation, glycerin
and coolant
liquid
Spray congealing Spray congealer N/A Materials with high Variables affecting the
[67,54] melting point and spray congealing process,
low heats of and possible pharmacrystallization
ceutical application for
pellet making.
226 Chang and Robinson
extent explosiveness and danger of these solvents have given pharmaceutical
and food supplement processors considerable incentive to remove
organic solvents from the coating process. The most commonly used methods
to eliminate organic solvents are presented in Table 4. At present
and in the foreseeable future, latex or pseudolatex coatings appear to be
the best choice to eliminate solvent-based coatings for controlled drug release.
Several techniques. including emulsion polymerization. emulsionsolvent
evaporation. phase inversion, and solvent change. can be employed
to prepare suitable latex{pseudolatex dispersion systems. Each method
has advantages and disadvantages based upon ease of preparation. latex
stability. convenience of use, film properties, and economics.
METHODS TO PREPARE LATEX DISPERSIONS
1. Emulsion polymerization. Polymerization in an emulsion state can
be applied to a wide variety of vinyl, acrylic, and diene monomers with
water solubility in the proper range, usually 0.001-1. 000% [89 - 911. The
basic formula used in emulsion polymerization consists of water, monomer,
surfactant, and initiator. The system contains three phases: Water containing
small amounts of dissolved surfactant and monomer. monomer droplets
stabilized by surfactant, and much smaller surfactant micelles saturated
with monomer. The initiator decomposes into free radicals which
react with monomer units at three possible sites for polymerization. The
radicals can react with dissolved monomer in the aqueous phase. or can
diffuse into the monomer droplets or the micelles. Obviously. there is
very little initiation in the aqueous phase owing to low monomer concentration
in the water phase. The diffusion rate of the radicals, which is
directly proportionsl to surface area, is far greater into the micelles, and
thus there is virtually no initiation in the monomer droplets. Free radicals,
from initiator decomposition, begin polymerization in the monomer solubilized
in the micelles. The micelles transform into growing particles. As these
monomer-polymer particles grow, they are stabilized by more surfactant at
the expense of uninitiated micelles, which eventually disappear. The
growing latex particles are continually supplied with monomer by diffusion
through the aqueous phase from monomer droplets. The latter gradually
decreases in quantity as polymerization proceeds, until at a conversion of
about 60%, they disappear completely. All free monomer has then diffused
into the latex particles. The polymerization rate decreases as the
monomer in the particles is depleted by further polymerization. The schematic
process for emulsion polymerization is shown in Scheme 7.
2. Emulsion -solvent evaporation technique. The emulsion -solvent
evaporation technique [92 - 97] also called emulsion hardening, has been
widely used to prepare microspheres for controlled drug release [98 -100] •
The technique involves dispersion of drug in an organic polymer solution,
followed by emulsification of the polymer solution in water. After continuous
stirring, the solvent evaporates and drug-containing rigid polymer
microspheres are formed. The procedure for preparing pseudolatex is
essentially the same as that described above (Scheme 8). The polymer
emulsion, with droplets so small they are below the resolution limit of the
optical microscope. can be accomplished by subjecting the crude emulsion
to a source of energy such as ultrasonic irradiation or by passing the
crude emulsion through a homogenizer or submicron dispenser. The
polymer solvent is normally stripped from the emulsion at elevated temperatures
and pressures to leave a stable pseudolatex. If foaming is not a
problem. the solvent may be removed under reduced pressure.
Sustained Drug Release from Tablets and Particles 227
Purified monomer by extraction
or adsorption
initiator
Water + surfactant and/or
emulsion stabilizer +
Emulsify monomer as internal
phase in water (oil in water
emulsion)
Bubble nitrogen through the
emulsion to remove air and
oxygen
Use heat and agitation to
induce polymerization
True latex
Scheme 8
3. Phase inversion technique. The phase inversion technique [101104]
involves a hot-melt or solvent gelation of the polymer, which is then
compounded with a long-chain fatty acid such as oleic acid, lauric acid,
or linoleic acid using conventional rubber-mixing equipment such as an
extruder. When the mixture is homogeneous a dilute solution of an slkali
is slowly added to the mixture to form a dispersion of water in polymer.
Upon further addition of aqueous alkali under vigorous agitation, a
phase inversion occurs and a polymer in water dispersion is produced
(Scheme 9).
4. Solvent change and self-dispersible technique. An ionic water
insoluble polymer, which may be generated by acid-base treatment or
chemical introduction of functional groups, such as ammonium groups,
phosphonium, Or tertiary sulfonium groups, may be self-dispersible in
water without any need for additional emulsifier [105,106]. Generally,
the polymer is first dissolved in a water-miscible organic solvent or in a
mixed water-miscible organic solvent system. The pseudolatex can then
be obtained by dispersing the polymer solution in deionized water into
~
~
Oc)
Table" Common Methods to Eliminate Organic Solvents in the Coating Process
Method
Compression coating
Aqueous solution
Mixed organic aqueous
system [68]
Alkali salts [69. 70J
Function
Compressible materials
Water-soluble film
formers
Enteric materials
Ent-eric materials
Examples of coating materials
Su gars, hydroxypropylmethylcellulose,
polyvinyl alcohol
Methylcellulose. hydroxypropyl
cellulose, hydroxypropylmethylN
cellulose
Polyvinyl acetate phthalate,
carboxylmethylethylcellulose.
hydroxypropyimethylcellulose
acetate phthalate.
Shellac. hydroxypropylmethylcellulose
phthalate. cellulose
acetate phthalate, cellulose
acetate trimellitate.
Comments
Totally eliminates organic solvents.
It is not well accepted by pharmaceutical
industry owing to complicated
mechanical operation and
formulation problems.
Film formers giving solutions of low
viscosity are the most suitable for
use. Totally eliminates organic solvents.
but unusitable for controlled
drug release.
Parti8lly eliminates organic solvents.
May be suitable for enteric coating,
but not practical for controlled
drug release.
To form gastric fluid resistant coatings,
a volatile neutralizing agent,
ammonium hydroxide or morpholine,
is preferable to neutralize the
enteric materials. Totally eliminates
organic solvents.
t"
t"
to
Hot melts
Aqueous dispersions
of waxes and lipids
[ 71]
Coating emulsions
Latex: dispersions
[73 -80]
Materials with low melting
point
Waxes and lipids
Almost all waterinsoluble
polymers
AImost all waterinsoluble
polymers
Hydrogenated oil. wax, solid
polyethylene glycol
Castor wax. carnauba wax.
Cu tina HR. Hoechst Wax E.
Durkee 07
Cellulose acetate phthalate.
hydroxypropylmethylceI1ulose
Ethylcellulose pseudolatex ,
Eudragit RL/RS pseu dolatex ,
Eudragit E-30D latex
Organic solvents can be eliminated
completely. However. organic solvents
may be needed to thin the
hot melts, in some cases. Heating
devices such as steam jackets or
heating tape is needed for the
spraying system to avoid solidification
of the coating material.
Ladle process may be more practical,
less troublesome.
Totally eliminates organic solvents.
Aqueous dispersions of waxes and
lipids may not be superior to hot
melt coating.
Partially eliminates organic solvents.
Still in their infancy as pharmaceutical
coatings.
Totally eliminates organic solvents.
Latex dispersions usually have low
viscosity and a high solids content.
Latex systems have some applications
in ophthalmic delivery systems
[81]. injectable colloidal delivery
system [82]. and molecular entrapment
techniques for sustainedrelease
dosage forms {83-88].
230
Polymer + water
immiscible solvent
Chang and Robinson
Water + surfactant and/or
stabilizer
Scheme 8
Emulsify polymer-solvent as
internal phase in water (crude
emulsion)
Subject to ultrasonic
irradiation, homogenizer or
subm1cron disperser to
generate a fine emulsion with
sUbmicroscopic droplet size
Eliminate the organic solvent
Pseudolatex
the polymer solution under mild agitation. The organic solvent(s) is subsequently
eliminated from the aquous-organic solution to leave a stable
latex (Scheme 10). The absence of emulsifiers has several interesting
consequences, such as stability to heat and mechanical shear, and dilutability
with organic solvents. Table 5 lists the general features of four
commercially available latex-coating systems for controlled drug release.
Recently, latex/pseudolatex coating has been further expanded to use
cellulose acetate pseudolatex for elementary osmotic pumps [107] and
water-based silicone elastomer dispersion for controlled-release tablet coating
[108,109].
The aforementioned techniques also can be used to prepare latex or
pseudolatex of enteric polymers. Because it is costly to ship aqueous dispersions
and some enteric materials are susceptible to hydrolysis, sprayor
freeze-drying techniques may be used to dry aqueous polymeric
Sustained Drug Release from Tablets and Particles 231
Hot melt or solvent
gelation of polymer
Scheme 9
long chatn fatty acid.
such as oleic acid
Mix thoroughly to form
homogeneous mixture
Add a dilute solution of an
alkali slowly into the mixture
Water in polymer dispersion
f
Further incorporation of
aqueousal ka 11
Polymer 1n water dispersion
pseudolatex
dispersion and form a redispersible aqueous enteric coating system.
Table 6 lists the general features of three dispersible aqueous enter1ccoating
systems.
Basic Considerations in Coating and Accessory Coating Equipment
COATI NG EQUIPMENT. A perforated pan as well as a conventional coating
pan equipped with hot air supply, spray system, pan baffles, and variable
coating pan speed can be used for latex coating. However, the relative
inefficiency of drying and longer contact time in the coating pan may cause
penetration of water into the core and thus discontinuity or irregularity of
the film {1l0]. In the past 10 years, there has been a significant increase in
the use of fluid-bed technology to granulate materials for improved compression
and to coat cores for desired properties such as controlled drug release,
enteric release, appearance, or taste masking. Fluid-bed coating using
232
Polymer with ionic character.
Polymer undergoes acid-base
treatment to generate ionic
character or polymer undergoes
chemical modification to yield
ionic character + watermiscible
organic solvents Water
Chang and RObinson
Mix under agitation to form
latex dispersion in aqueous
organic solvent system
I
Eliminate the organic solvent
(s)
Pseudolatex
Scheme 10
centrifugal-type or Wurster column-type equipment provides ideal conditions.
such as rapid surface evaporation. controllable inlet air temperature. and
short contact time, for latex coating.
ACCESSORY EQUIPMENT
1. Nozzle systems. Characteristics of three different types of nozzle
are listed in Table 7 [111]. The ultrasonic nozzle is a relatively new
and entirely different type of atomizing nozzle that offers several advantages
over conventionsl nozzles. However, application of ultrasonic nozzles
is still in its infancy. Thorough investigations are necessary to dedermine
the feasibility of their pharmaceutical applications. Hydraulic
guns are sometimes used in place of air-atomizing nozzles for large filmcoating
processes. The nozzles tend to clog with latex coating because
the airless system generates a shear which may coagulate the formulated
latex. Pneumatic nozzles have been adapted for fluid-bed systems, and
have been shown to be an acceptable nozzle system for latex coating.
The atomizing air is exposed to the product and, therefore, must be
free of oil and other contaminants.
Sustained Drug Release from Tablets and Particles 233
Table 5 General Features of Four Latex Coating Systems for Controlled
Drug Release
Latex system
Eudraglt E-30D
(Rohm Pharma)
Aqua-Coat
(F .M.C. Corp.)
Surelease
(Colorcon, Ine .)
Budz-agft RS30D
and RL30D
(Rohm Pharma)
Method
Emulsion
polymerization
Emulsion-solvent
evaporation
Phase inversion
Self-dispersible
General features
Poly (ethylacrylate, methylmethacrylate)
latex, 30% w/w solid content.
No plasticizer is required. May
contain residual monomer, initiator.
surfactants , and other chemicals
used in the polymerization process.
Unplasticized ethyl cellulose dispersion.
Contains sodium lauryl sulfate
and cetyl alcohol as stabilizers.
30% w/w solid content.
Fully plasticized ethyl cellulose
dispersion. Contains oleic acid,
dibutyl sebacate, fumed silica, and
ammonia water. 25% w/w solid
content.
Unplasticized poly (ethylaerylate ,
methylmethacrylate. trimethylammonioethylmethacrylate-
chloride)
dispersions. Contains no emulsifiers.
30% w/w solid content.
Table 6 General Features of Three Dispersible Aqueous Enteric Coating
Systems
Coating system
Aquateric
(F .M.C. Corp.)
Coateric
( Colorcon, Inc.)
Budragft L - 100-55
(Rohm Pharma)
Method
Emulsion-solvent
evaporation followed
by spray
dry technique
Mechanical means
to reduce particle
size of
the polymer
Emulsionpolymerization
followed by
spray dry
technique
General features
Redispersible cellulose acetate
phthalate coating system. Contains
polyoxypropylene block copolymer
and acetylated monoglycerides.
Plasticizer is required.
Completely formulated dispersible
polyvinyl acetate phthalate coating
system.
Redispersible poly (ethyl acrylatemethacrylate
acid) coating system.
Contains polyvinylpyrrolidone,
polyoxyethylene sorbitan fatty acid
ester, and polyethylene glycol.
Plasticizer is required.
~
(,,)
"'"
Table 7 Nozzle Characteristics by Type
Ultrasonic nozzles Hydraulic (pressure) Air atomfzing (two fluids)
Principal of operation
Average microdrop
size
Spray velocity variability
flow rate
Minimum achievable
flow rate
Maximum achievable
flow rate
Ornice &ze/cloggabllity
Ultrasonic energy concentrated
on atomizing surface.
causes impinging liquid to
disintegrate into a fog of
microdrops
20-50 Il (depending on frequency)
Low: 0.2-0.4 ms, Infinity
variable from zero flow to
rated capacity
o gph
30-40 gph
Large: up to 3/8" -uneloggable
Pressurized liquid is forced
through orifice. Liquid is
sheared into droplets.
100 - 200 u at 100 psi (higher
pressure reduces size)
High: 10-20 mls ± 10% of
specified rating
0.5 gph
No limit
Very small, usually subject
to clogging
High-pressure air or gas
mixes with liquid in the nozzle:
Air imparts velocity to
liquid which is then ejected
through an orifice
20-100 1.1 (air pressure from
10-100 psi)
High: 50- 20 ml s , Infinitely
variable from 20% of maximum
capacity to maximum
0.3 gph
No limit
Very small, usually subject
to clogging
Sustained Drug Release from Tablets and Particles 235
2. Pumping systems. Peristaltic or gear pumps are used in combination
with nozzles to form a spray system. The ab'fiity of a pumping system
to deliver the coating liquid at the required rate for the duration of the
coating cycle is critical for a uniform coating. A flow integrator can be
used to eliminate pulsation of output flow which may cause uncontrolled
wetting disruption of the spraying process.
3. Humidity control. In order to maximize the drying efficiency of
the fluid-bed machine, it may be necessary to dehumidify the inlet air.
Furthermore, the amount of moisture in the inlet air can significantly influence
batch-to-batch variabfiity of the coating process. Therefore, it is
important to control ambient air humidity in the coating operation.
PROCESS VARIABLES
1. Fluidization air temperature. Film formation from a solvent-based
system is dependent upon the entangling and packing of polymer molecules
as the solvent evaporates. Relatively low fluidization air temperatures
should be used to prevent spray drying of coating materials because of
low heats of vaporization for commonly used solvents. The mechanisms of
film formation from a latex system involves the softening of latex spheres
caused by plasticization and/or temperature, the contact of latex spheres
resulting from loss of water, followed by deformation and coalescence of
the latex spheres owing to capillary force and surface tension of the
polymer to form a continuous film [1121. The temperature of the inlet air
has a dual function; to evaporate water and to soften and coalesce the
latex spheres. Latices, as contrasted to aqueous solutions, have a very
low affinity for water, and therefore relatively low temperatures can be
used to efficiently evaporate water. However, column temperature is critical
for latex softening and coalescence. In order to generate a continuous
film, the column temperature must be higher than the minimum film-formation
temperature. If the temperature is too high, it may cause excessive
drying and softening of the latex film, and hence result in electrostatic
interaction and agglomeration problems. Minimum film-formation temperature
[1131 should be used as a guideline to select the temperature of the
inlet air. The glass-transition temperature [1141 and ffim-softening temperature
[761 can also provide useful information for choosing the column
temperature.
2. Spray rate. The liquid spray rate affects the degree of wetting
and droplet size. At a given atomization air pressure, increasing the
liquid spray rate will result in larger droplets and a higher possibfiity to
overwet the coating substrates. Slowing the spray rate may cause electrostatic
problems owing to low bed humidity, especially at high temperature
settings.
3. Volume of the fluidized air. Since a sluggish or vigorous fluidization
can have detrimental effects on the coating process, such as side-wall
bonding and attrition of core substrates, proper fluidization should be
maintained throughout the coating process.
4. Atomization pressure. Atomization pressure affects the spraying
pattern and droplet size. Excessive high atomization pressure may result
in the loss of coating materials and breakage or attrition of the substrates.
Excessive low atomization pressure may overwet the core and cause sidewall
bonding.
236 Chang and Robinson
FORMULATION VARIABLES
1. The nature of the plasticizer. There are many plasticizers that
are compatible with ethyl cellulose and polymethyl methacrylate and can be
used for plasticization [115]. For the polymer with a relatively high glasstransition
temperature, a plasticizer with a strong affinity to the polymer
must be found in order to form a resistant film. Plasticizers with low
water soIubfiity are generally recommended for controlled drug release.
Dibutyl sebacate, diethyl phthalate, triacetin, triethyl citrate, and acetylated
monoglyceride often give satisfactory results.
2. The amount of the plasticizer. Experiments should be performed
to determine the most favorable proportion of plasticizer. Low levels of
plasticizer may not overcome the latex sphere's resistance to deformation
and result in incomplete or a discontinuous film. On the other hand, a
high proportion of plasticizer may result in seed agglomeration, sticking,
and poor fluidization problems caused by excessive softening of the polymer
film. The best result generally is obtained with plasticizer concentration
in the range of 15- 30% based upon the polymer.
3. Incorporation of plasticizer. Plasticizer can be incorporated into
the latex system during the preparation process which provides more consistent
plasticization, more effective use of plasticizer, and a relatively
nonseparable plasticized latex system. Plasticizer also can be added to the
latex system under mild agitation. Agitation speed, mixing time. and separation
of plasticizer during the coating process should be considered in
the preparation of the plasticized latex system.
4. The solids content of the latex system. Generally. 8-20% solids
content give the best results [112]. However, higher or lower solids content
can be used to achieve a rapid buildup of film thickness or coating
uniformity, respectively.
5. Additive. Water-soluble chemicals can be incorporated into the
latex film to enhance its dissolution rate. On the other hand, the addition
of hydrophobic powder such as talc. magnesium, stearate, or silica in a
latex-coating system not only alters drug release, but also facilitates
processing by reducing tackiness of the polymer film.
6. Dual Latex Ipseudolatex coating. Most latice or pseudolatices are
stabilized by high surface potential of deflocculated particles arising from
ionic functional groups on the polymer or ionic surfactant or stabilizer.
For example, the positive charge on Eudragit RS30D and RL30D pseudolatex
particles arising from quaternary groups on the polymers and the
negative charge on Aquacoat and Surelease pseudolatex particles originated
from the anionic surfactants such as sodium lauryl sulfate and oleic acid
are the major stabilizing factor. Also, the size of the latex sphere and size
distribution are important factors which affect stability, rheological properties,
and film properties. Small, monodispersed latex particles are
required to have complete coalescence of the latex sphere. Generally. it
is not recommended to use dual latex/pseudolatex coating systems because
of the possible incompatibfiity of two latex/pseudolates systems, the different
glass-transition temperatures of the polymers. and the different sizes
of latex spheres. However, in a British patent, Eudragit E-30D was mixed
with Aquacoat to prevent the coated granules sticking together and to
Sustained Drug Release from Tablets and Particles 237
improve the dissolution-retarding effect [116]. It has been demonstrated
that Eudragit RS pseudolatex can be mixed in any proportion with Eudragit
RL pseudolatex. A wide range of release rates for theophylline could be
obtained by changing the ratios of Eudragit RS pseudolatex and Eudragit
RL pseudolatex , The enhancement of theophylline release caused by increasing
the amount of Eudragit RL pseudolatex is due to its high permeability
to water and theophylline [80].
7. Overcoating. During the curing or storage stage, the pellets
coated with latex or pseudolatex may adhere to one another because of the
softening and tackiness of the film. This could have a detrimental effect
on the dissolution properties of film-coated products. Nevertheless, an
overcoat that is water soluble can solve the problem of tackiness of latex
film without changing the dissolution profile. Immediate mixing of latexcoated
pellets with some separating agents such as talcum. magnesium
stearate. and other dfiuents also can prevent the formation of clumps of
pellets during storage.
C. Compression Coating and Embedment
Specially constructed tablet presses such as Drycota and Prescota are
available for compressing a polymeric or sugar composition onto a drugcontaining
core. This technique has been utilized to create a polyvinyl
alcohol diffusion barrier surrounding a core tablet containing various active
ingredients as described by Conte et al , [117]. In vitro dissolution tests
show that zero-order release kinetics of drug from compression-coated tablets
can be achieved as long as the thermodynamic activity of the drug
within the closure and the barrier characteristics are maintained constant.
Salomon et al, [118] coated potassium chloride tablets with a thin layer of
hydroxypropyl methylcel1ulose by a compression technique for delayedrelease
purposes. Such compression-coated tablets release potassium chloride
at a constant rate. Enteric coating by a double-compression technique
has been reported [119] using a mixture of triethanolamine cellulose acetate
and lactose as coating materials.
Other variations of compression coatings such as inlaid tablets and
layer tablets may have application in preparing a sustained-release tablet
with an immediate-release and a separate slow-release portion. In general,
the release rate from compression-coated tablets may be modified by core
composition and characteristics. thickness of membrane layer, composition
of membrane layer, and geometry of core tablet and final tablet. However,
the expense of compression coating and layer tablet production. complicated
mechanical operation, production problems such as multiple granulations,
improper centration, capping, and limited compressible and permeable barrier
materials limits its adoption as a popular technique to control drug
release. The compression-embedding technique has received increasing
attention to prepare controlled-release matrix tablets, and intensive research
in this area is being conducted by pharmaceutical scientists. There
are three different types of matrix tablets; I, e., hydrophilic matrices.
plastic matrices, and fat-wax matrices. which can be differentiated by the
matrix-bufiding materials.
238 Chang and Robinson
Hydrophaic Matrix Tablet
Utilization of a hydrophilic matrix as a means to control drug release was
disclosed in U. S. Patent 3,065,143. Sodium carboxymethyl cellulose,
methyl cellulose, hydroxypropyl cellulose, hydroxyethyl cellulose, polyethylene
oxide. polyvinyl pyrrolidone, polyvinyl acetate. carboxy polymethylene.
alginic acid, gelatin, and natural gums can be used as matrix
materials. The matrix may be tableted by direct compression of the blend
of active ingredient(s) and certain hydrophilic carriers, or from a wet
granulation containing the drug and hydrophilic matrix material( s) • Several
commercial patented hydrophilic matrix systems are currently in use,
such as the Synchron Technology I 120J and hydrodynamically balanced
system [121J. The hydrophilic matrix requires water to activate the release
mechanism and enjoys several advantages, including ease of manufacture
and excellent uniformity of matrix tablets. Upon immersion in
water, the hydrophilic matrix quickly forms a gel layer around the tablet.
Drug release is controlled by a gel diffusional barrier that is formed and/or
tablet erosion. The effect of formulation and processing variables on
drug-release behavior from compressed hydrophilic matrices has been
studied by a number of investigators [122-134] and can be summarized as
follows:
1. The matrix building material with fast polymer hydration capability
is the best choice to use in a hydrophilic matrix tablet formulation.
An inadequate polymer hydration rate may cause premature diffusion of
the drug and disintegration of the tablet owing to fast penetration of water.
It is particularly true for formulation of water-soluble drugs and excipients.
2. The amount of hydrophilic polymer in tablet formulations was reported
to have a marked influence on the disintegration time and dissolution
of the tablet. The disintegration time was extended as polymer content
increased. The release rate of drug was decreased when the proportion
of polymer was increased but differed quantitatively with different
drugs and different matrix-building materials. Slower hydration polymers
can be used at higher concentration level to accelerate gel formation or
reserved for water-insoluble drug( s) .
3. Generally, reduced particle size of the hydrophilic polymer ensures
rapid hydration and gel formation, leading to a good controlled release.
The impact of polymer particle size on the release rate is formulation
dependent, but may be obscured in some cases. The particle size of
a drug. within a normal size range, may not significantly influence the
drug release from the matrix tablet. Extremes of drug particle size may
affect release rate of the drug.
4. Viscosity characteristics of the polymers are of great importance
in determining the final release properties of the matrix tablet. Generally,
the drug-release rate is slower for a higher viscosity - grade polymer.
5. Commonly, water-soluble excipients in the matrix tablet can increase
drug release. However, addition of water-soluble materials may
achieve a slower rate by increasing viscosity of the gel through interaction
with hydrophilic polymers or by competition with matrix material for
water. When water-insoluble nonswellable excipient(s) or drug(s) is used
in the matrix system stress cracks can occur upon immersion in water because
of the combination of SWelling and nonswelling components on the
tablet surface.
Sustained Drug Release from Tablets and Particles 239
6. For some hydrophfiic matrix building materials, pH may affect the
viscosity of the gel which forms on the tablet surface and its subsequent
rate of hydration. Under acidic conditions. carboxypolymethylene and
sodium carboxymethyl cellulose have little or no retarding effect on the
drug-release rate. Gelatin forms gels of higher viscosity in acidic media
and is more effective in retarding drug release as compared to a basic
media.
7. No conclusions could be drawn as to the effect of compression
force on drug-release behavior owing to the different properties of the
various hydrophilic matrix materials. However. tablet size and shape can
significantly influence drug-release kinetics.
Fat-Wax Matrix Tablet
The drug can be incorporated into fat-wax granulations by spray congealing
in air [135 -13S}. blend congealing in an aqueous media with or without
the aid of surfactants I 139 -142}. and spray-drying techniques [44].
In the bulk congealing method. a suspension of drug and melted fat-wax
is allowed to solidify and is then comminuted for sustained-release granulations
[143]. The mixture of active ingredients. waxy mat erial( s). and
filler( s) also can be converted into granules by compacting with a roller
compactor. heating in a suitable mixer such as a fluidized-bed and steamjacketed
blender, or granulating with a solution of waxy material or other
binders. Fat-wax granulations containing drug obtained from all of the
above processes may be compressed to form tablet cores or directly compressed
into a final tablet form with sustained-release properties.
The drug embedded into a melt of fats and waxes is released by
leaching and/or hydrolysis as well as dissolution of fats under the influence
of enzymes and pH change in the gastrointestinal tract. Enteric
materials such as cellulose acetate phthalate. polyvinyl acetate phthalate,
methacrylate copolymer, zein. and shellac may be used to prepare matrix
tablets with somewhat a similar drug-release mechanism. In general, the
primary constituents of a fat-wax matrix are fatty acids and/or fatty
esters. Fatty acids are more soluble in an alkaline rather than an acidic
medium. Fatty esters are more suscepitble to alkaline catalyzed hydrolysis
than to acid catalyzed hydrolysis. The surface erosion of a fat-wax matrix
depends upon the nature and percent of fat-wax and extenders in the
matrix [136]. Other factors such as drug particle size and drug concentration
affects release of the drug from the matrix system [141]. The addition
of surfactants to the formulation can also influence both the drugrelease
rate and the proportion of total drug that can be incorporated into
a matrix [137 • 142}• Polyethylene. ethylcellulose, and glyceryl esters of
hydrogenated resins have been added to modify the drug-release pattern
[135] •
Plastic Matrix Tablets
Sustained-release tablets based upon an Inert compressed plastic matrix
were first introduced in 1960 and have been used extensively clinically
[144]. Release is usually delayed because the dissolved drug has to diffuse
through a capillary network between the compacted polymer particles.
Commonly used plastic matrix materials are polyvinyl chloride. polyethylene.
vinyl acetate/vinyl chloride copolymer, vinylidene chloride/acrylonitryle
copolymer. acrylate/methyl methacrylate copolymer, ethyl cellulose.
240 Chang and RObinson
cellulose acetate, and polystyrene [145]. Plastic matrix tablets, in which
the active ingredient is embedded in a tablet with coherent and porous
skeletal structure, can be easily prepared by direct compression of drug
with plastic material( s) provided the plastic material can be comminuted or
granulated to desired particle size to facilitate mixing with drug particle.
In order to granulate for compression into tablets the embedding process
may be accomplished by:
1. The solid drug and the plastic powder can be mixed and kneaded
with a solution of the same plastic material or other binding agents
in an organic solvent and then granulated.
2. The drug can be dissolved in the plastic by using an organic solvent
and granulated upon evaporation of the sol vent.
3. Using latex or pseudolatex as granulating fluid to granulate the
drug and plastic masses.
Drug release from the inert plastic matrices was affected by varying
formulation factors such as the matrix material, amount of drug incorporated
in the matrix, drug solUbility in the dissolution media and in the
matrix, matrix additives, and the release media. Since the mechanism of
controlling drug release in the plastic matrix is the pore structure of the
matrix:, any formulation factors affecting the release of a drug from the
matrix may be a consequence of their primary effect on apparent porosities
and tortuosities of the matrices. These release factors can be summarized
as follows:
1. The release rate increases as the solubility of the drug increases,
but there seems to be no direct relationship between the two variables.
2. The release rate increases as the drug concentration increases.
An increase in release rate cannot be explained on the basis of increasing
matrix porosity [146]. Rather it has been attributed to changes in matrix
tortuosity with drug concentration [147] and to decreased diffusional resistance
by shortening the length of the capillary joining any two drug
particles [148].
3. It is possible to modify the release rate by inclusion of hydrophilic
or hydrophobic additives to the matrix. The release of a sparingly soluble
substance can be increased by the addition of physiologically inert but
readily soluble material such as polyethylene glycol, sugars, electrolytes,
and urea [146,149]. The decrease in the release rate on the addition of
hydrophobic substance may be due to decreased wettability of the matrix
[150] •
4. The release rate from plastic matrix tablets could be decreased by
exposure to acetone vapor without changing the release mechanism. The
extent of the reduction was found to be dependent on the amount of
acetone absorbed [147,151]. The tensile strength of the tablets increases
by heating the polymer matrix above the glass-transition temperature.
However, porosity also increases with a marked increase in the release
rate [152,153].
5. The release rate increased as the particle size of the matrix material
increased and as the particle size of the drug decreased.
6. Increasing compaction pressure up to the full consolidation point
tends to decrease the pore formed among the polymer particles, resulting
in a slower drug-release rate [154].
Sustained Drug Release from Tablets and Particles 241
Microcapsules. microspheres, or coated pellets also can be compressed into
tablets or embedded in a drug-containing matrix. Sustained-release tablets
made from individual coated particles may have very different release characteristics
than the original coated particles depending upon whether or not the
tablets disintegrate to expose the majority of the coated particles to the dissolution
environment and whether the coated particles are damaged by the compression
process [155]. Generally. nondisintegrating tablets made from ethyl
cellulose microcapsules followed matrix -release kinetics are much slower than
the uncompressed microcapsules. When compression force was sufficient to prevent
breakup of the tablets, greater compression force had little effect on
rate of dissolution I 156 - 160] • Apparently. nondisintegrating tablets made
from individual coated particles did not provide any advantages over matrix
tablets but a more complicated manufacturing process. A dispersible tablet
containing individual coated particles distributes the drug-containing coated
particles in the gastric content to minimize the high local concentration of
drug and to reduce the inter- and intrasubject variation linked to gastrointestinal
transit time.
However. the sustained-release properties of the coated particles may
be lost due to cracking of the membranes and rupture of the characteristic
microcapsule tails. The amount of damage was found to be related to the
compressibility and particle size of excipients in the tablet formula as well
as to the compression pressure. Large particles led to greater damage
as noted by increased dissolution rates of the disintegrating tablets [161].
It also has been found that a combination of microcrystalline cellulose and
polyethylene glycol provides maximum protection from the damage of potassium
chloride microcapsules by reducing interparticle friction [162].
Matrix: particles have the advantage of being very rugged, less subject to
dose dumping, and more resistant to compression damage than membrane
systems. Little change in the dissolution rate of cellulose acetate butyrate
microspheres containing succinyl sulfathiazole has been reported when tableting
with microcrystalline cellulose and carboxymethyl starch under compression
force between 35 and 350 MPa [163].
IV. SUSTAINED-RELEASE PRODUCTS THROUGH
COATING
The preceding discussion has provided the framework for the design of
sustained-release products. Application of these principles in the pharmaceutical
industry has resulted in varying degrees of success in achieving
consistent, nonvarying blood levels of drug from sustained-release dosage
forms. It is instructive to examine some of these dosage forms more closely,
and to analyze where and why they fail to provide sustained release
and whether the failure can be detected in vivo or in vitro. Of course,
the reader should keep in mind that these are "failures" only in a relative
sense; some forms are very much superior to others, but all do provide
some sort of sustained therapy beyond their nonsustained counterparts.
Whether this sustained blood or tissue level results in a measurable improvement
clinically is often debatable.
It would appear from the earlier sections of this chapter that one can
employ the theoretical calculations on release-rate and dosage-form design
with great precision to formulate a prolonged-action dosage form. In
point of fact, although this has been done in a few studies, the vast
242 Chang and Robinson
majority of published work has employed these principles and calculations
as a working guideline. The formulator obtains the desired release-rate
constant and size of tablet by suitable calculation and attempts to generate
this release pattern by some sustained-release mechanism. such as
coating. If the release rate of drug is dissimilar to the calculated value.
the dosage form is appropriately modified. Since the ideal release pattern,
resulting in a tissue drug concentration -time profile similar to that shown
in Figure 1. is seldom obtained, the formulator will usually be satisfied
with either a bell-shaped profile that has a broad. relatively flat plateau,
or an extension of biological activity. This qualitative approach. which at
times is very empirical, is clearly evident when one reads the literature
dealing with prolonged-action dosage forms. Thus. unless the formulator
has good control over the release mechanism, which is rarely the case. he
or she wfil be working empirically or semiempirically in preparing the
dosage form. Moreover. because of this lack of control over the release
mechanism, a blood drug profile will usually result that shows simple prolongation
rather than the invariant drug level of the sustained type.
In the following discussion, we have arbitrarily classified the approaches
used in coating sustained-release products into various SUbdivisions. A
good argument could be constructed that all of the following subdivisions
are artificial. since all coated sustained-release dosage forms utilize dissolution
and diffusion to varying degrees. However. the dosage forms mentioned
in each section usually appear to have one major mode of providing
sustained action. These modes can derive from the way the material is
produced. as in the case of microcapsules and bead polymerization, or from
the way the body handles the dosage form, as in the case of soluble coated
granules or impermeable films coated in a tablet. The division has been
created solely for the purpose of organizing the literature in the area and
should not be viewed as being rigid or exclusive.
At the end of some of the SUbdivisions, the reader will find a section
entitled case study. These sections are intended to provide sufficient experimental
detail to allow the novice formulator to initiate preparation of a
prolonged-action dosage form. We will thus examine selected published reports
in more detail, providing as much of the experimental methods employed
as is practical. In addition. when available, the appropriate equations
to describe the release of drug from the dosage form and the type
of in vitro and in vivo profile generated will be noted.
A. Sustained Release Utilizing Dissolution
Dissolution methods in sustained release generally refer to coating the individual
particles, or granules, of the drug with varying thicknesses of
coating material so that dissolution of the coat, resulting in release of the
drug contained within, occurs over a long time span owing to the thickness
differences of the coats. These coated particles are then either compressed
directly into tablets (e.g .• Spacetabs), or they can be placed in capsules
(e.g•• the Spansule or Plateau Cap dosage forms), as shown in Scheme 11.
Alternatively. the dissolution may be that of an exterior tablet coating
Where a portion of the drug is placed in the tablet coat and dissolves rapidly
to provide enough drug to quickly reach therapeutic levels. whereas
the sustained-release interior of the tablet utilizes some other method, to
be discussed subsequently. to provide controlled, long-term release of
Sustained Drug Release from Tablets and Particles 243
WAX OR 0
00
DRUG 0° POLYMER •• + 00:0 > 00 c::::=o •• 000 0
0 WAX OR
BLANK DRUG POLYMER NON SUSTAINED SEEDS COATED COATED DOSE
CAPSULES
TABLETS
Scheme 11
medicament. Exterior tablet coats may also have differentially soluble constituents
that dissolve and provide an outer shell to maintain diffusion
path lengths for the drugs contained in the shell. This form of sustained
release is particularly useful for relatively insoluble drugs because it keeps
the dose from disintegrating and being spread out over the gastrointestinal
tract, thus providing some regulation of the dissolution medium and area
for control in maintaining slow release of drug.
Pulsed Dosing
Included in the pulsed dosing category are slowly dissolving coatings such
as the various combinations of carbohydrate sugars and cellulose-based
coatings as well as polyethyleneglycol bases, polymeric bases, and waxbased
coatings. Colbert [1641 and Johnson [165] provided a particularly
complete cross section of patents issued since 1960 based on these digestible
bases. These coating materials are used in preparing sustained-release
dosage forms that follow the approach of various thickness coated granules
or seeds combined with uncoated granules which are dispensed in capsule
or compressed tablet form. Examples are the Spansule [166-175] and
Spacetab [176 -180] formulations. These coatings vary in thickness and
when they COver a drug granule their digestion by fluids in the gastrointestinal
tract results in abrupt release of the medicament at selected
time intervals to provide pulsed dosing for periods up to approximately
12 h.
With digestible coating materials, the important factors a formulator
must take into account are the dissolution rate of the coating material, the
thickness of the coat, and the changed are for disintegration and dissolution
that the increased thickness provides [3 - 5. 10, 18, 181 , 182]. As the
drug-coated granules traverse the gastrointestinal tract, the coating is
slOWly solubilized by the gastrointestinal fluids. Since the granules have
varying thickness coats, one anticipates a staggered release of drug.
Therefore, by combining a large number of mixes of different thickness
coated granules, a horizontal, or very nearly so, blood drug concentration
244 Chang and Robinson
versus time curve for extended periods of time should result. In practice.
it is often difficult to combine a large number of granules having many different
thicknesses of coating. so that the more common approach is to employ
one-quarter of the granules in uncoated form, thus providing for immediate
release and rapid attainment of therapeutic blood levels of drug.
with the remaining three-quarters of the granules being split into three
groups of varying coating thicknesses to provide a sustaining effect,
through pulsed dosing, over the desired time period. Although the approach
of one-quarter uncoated, three-quarters coated is very common.
other combinations, such as one-third coated, two-thirds uncoated. have
also been employed. The ratio is determined by properties of the coated
granules; that is, dissolution rate and derived drug properties such as the
elimination rate constant.
CAPSULES. Since the introduction of the Spansule sustained-release
dosage form in the early 1950s. there have been numerous studies on the
release of active drugs from this type of preparation [172,173,183 -1981 ,
and also on the clinical effectiveness of these preparations in maintaining
therapeutic activity over extended time periods [166 -171,174]. Examples
of the types of drug formulated as coated granules include antihistamines
[166,167], belladonna alkaloids U68,171}, phenothiazines [173,174,185,1871.
combinations of the above [169], antihypertensives [45], cardiac muscle
dilators [199 - 201], anorexigenic agents [175, 184, 187,192, 193], steroid
anti-inflammatories [183], and nonsteroidal anti-inflammatories such as
aspirin [191].
There are several ways to prepare drug-coated beads or granules. A
common procedure is to coat nonpareil seeds with the drug and follow this
with either a slowly dissolving wax or polymer coat of varying thickness.
Conventional pan-coating or air-suspension coating techniques can be employed
for this purpose. Types of coating materials and properties will be
discussed later in this chapter. Coatings such as these can also be accomplished
through microencapsulation, to be discussed shortly, wherein
the drug solution or crystal is encapsulated with a coating substance.
The selection of coating material dictates whether pulsed or sustained drug
release occurs.
An illustration of this approach is the series of papers by Rosen et ale
I 189. 190] in which they describe the release of 14C-labeled Dextroamphetamine
sulfate and 14C-labeled amobarbital from both non-sustainedand
sustained-release dosage forms employing wax-coated granules. As
can be seen in Figure 3. the amount of drug released in each time period
is progressively less as the percentage of wax-fat in the coating increases.
This is the expected behavior, and indicates that sustained release can
be varied quite readily by changing the makeup of the coating material for
the drug granules. Other studies [185,191] along these same lines have
indicated that such behavior of coated granules is the general rule. Note
in Figure 3 that continuous release of drug occurs. This is because a
spectrum of granule size was employed in the study, The dashed line in
Figure 3 indicates the release pattern when only a few granule sizes are
combined.
The coated granules or seeds can be placed into a capsule for administration
to the patient. Actual photographs of the in vivo disintegration
and dispersal of coated sustained-release granules in a capsule were
provided in a series of papers by Feinblatt et al , [200,201]. Using
Sustained Drug Release from Tablets and Particles
C) zz
;;: 20
::ii:
wa::
w
C) 10
ct: 8 ~z
w 6 o
II:
W 4 0...
F
2
I
0 2
245
Figure 3 In vitro release pattern of dextroamphetamine sulfate pellets
pan coated with various amounts of wax-fat coating. (A) 17% coating,
(B) 15% coating, (C) 13% coating, (D) 11% coating, (E) 9% coating,
(F) 7% coating, and (G) selected blend of uncoated pellets and coated pellets.
(From Ref. 189, used with permission.)
roentgenography, the authors showed that after 10-12 h the coated granules
in a sustained-release capsule were well dispersed in the gastrointestinal
tract and the active ingredient was completely dissolved.
In a different study concerned with in vitro measurement of drug release
from sustained-release coated granules in a capsule, Royal [192] described
a modification of the United States Pharmacopeia CUSP) tablet disintegration
apparatus that allowed him to follow the release of drug from
capsule dosage forms. Further modifications [193] allowed him to test different
brands of capsules containing a wide range of granule sizes, and
thus be able to compare the brands as to their efficiency in maintaining
continuous release of medicament for extended time periods. Souder and
Ellenbogen [175] described a method for "splitting" (separating) sustainedrelease
coated granules to ensure an even mix of large and small granules
and to prevent the stratification of such mixes that often occurs when
pooling granules for analysis. They then measured timed release of drug
in a manner differing somewhat from the approach used by Royal. The
results, however, were the same, and indicated that the blending of larger
and smaller coated granules, presumably due to thickness of the sustainedrelease
coating, was effective in maintaining the characteristics of sustained
drug release. There are several approaches that can be employed for
246 Chang and Robinson
dissolution testing of coated pellets. and most investigators employ modifications
of the rotating vial techniques. such as is described in the National
Formulary (NF) XIII. or modification of the dissolution test presented in
uss XIX.
In clinical evaluations of capsule pulsed dose, prolonged-action dosage
form, and sustained-release products. the objectives of the experimental
techniques change from that of ensuring that drug release is indeed occurring
gradually over a long time to that of (1) showing that the sustained-
release dosage form adequately maintains therapeutic blood or tissue
levels for extended times (l72.173,183,185-188,191I, (2) provides relief to
the patient over a long period [166 -169, 171,174], and (3) perhaps reduces
the incidence of side effects due to the "peak" effects of non-sustainedrelease
dosing [166,167,169]. Of course, many of the earlier studies were
conducted before development of the sophisticated methods now employed in
gathering and interpreting blood or tissue drug level data. [166.167, 169,
171,174], and indeed many such studies were rather qualitative in their assessment
of prolonged-action dosage forms. As a result of this lack of
sophistication. the focus was either on monitoring some biological response
or on urinary receovery of active drug or its metabolites. For example,
the reports of Heimlich et al , [172,173] on sustained-release phenylpropanolamine
and trimeprazine and of Sugerman and Rosen [188] on sustainedrelease
chlorpromazine present very detailed and thorough urinary analyses
of drug content and provide, within the limitations set forth by the authors.
a very good analysis of sustained-release characteristics and their worth
with respect to administration of these drugs. While results of urinary
drug analysis alone can give quantitative information on sustained-release
characteristics, it has been pointed out [202] that there are dangers in
this approach and comparisons of blood and urine data do not always coincide.
When possible. both blood and urine samples should be collected and
analyzed.
In the early 1960s, the first really good effort at measuring and ana1yzing
drug levels in the blood began to appear in the literature [183 -186] .
One of the earliest of these is the study by Wagner et al , [183] on the
sustained action of prednisolone in dogs and humans and the comparability
of these data to in vitro findings. The work by Rosen and Swintosky
[187,189,190] contributed to this field by employing radioactive-labeled
drug.
Hollister [1861. in a report on sustained-release meprobamate products,
presented both urine and blood drug level data and pointed out how they
reinforce the interpretation based solely on one or the other.
Of course the nature of the disease state itself can be a major reason
for publtshlng more qualitative results, as was one in some of the earlier
work. Those studies measuring relief of symptoms due to sustained-release
antihistamines [166.167] have few measurable effects that are easily quantitated,
whereas other reports in the area, such as those dealing with ulcer
patients and relief of pain, can at least quantitate volume and acidity of
gastric acid and digestive juices [168]. The literature is speckled with
data that show a sustained effect is operative, but give no indication as to
how well it compares to multiple dosing of ordinary tablets.
The question of the bioavailabUity of sustained-release dosage forms as
compared to conventional systems is an important issue. For most drugs
placed in a sustained-release system, the bioavailabUity is less than in conventional
dosage forms. There are the occasional drugs that are unstable
Sustained Drug Release from Tablets and Particles 24'1
in gastrointestinal fluid or have absorption problems [31] which become
stabilized or show improved absorption when placed in a sustained-release
system. A typical example of reduced bioavailability is the study by
Henning and Nybert [203] on quinidine, in which the rapidly dissolving
tablet was shown to have greater bioavailabUity than either quinidine
Durules or Longacor. Compared to the rapidly dissolving tablet, the
Durules had a bioavailability of 76% and the Longacor 54%.
TABLETS. With the tablet dosage form the concern for the thickness
and area of granular coating remains, the distinguishing feature of this
dosage form being that of tablet disintegration as contrasted to gelatin
capsule dissolution. An added problem here may be the influence of excipients
used to produce a compressed tablet in the disintegration/dissolution
process. Of course, When no fillers or excipients are used. the
coated granules alone are compressed and can fuse together, resulting in
altered dissolution patterns [18], depending on the properties of the coating
material. The role of excipients on the dissolution pattern of compressed
coated granules has not been investigated extensively. One would
not expect substantial effects on the dissolution process of the individual
seeds or granules, but perhaps there would be an influence on fracture or
fusing of seeds. Reports using other sustaining mechanisms show, as one
would expect, that in compression some of the coated particles are fractured.
Green [204]. for example, found that in microencapsulated
sustained-release aspirin Which was subsequently tableted, a small amount
of aspirin was immediately released, suggestive of fractured coats. This
need not be a problem, since the degree of fracture and immediate release
is frequently small and can usually be incorporated into that portion of the
dosage form that provides for immediate blood drug levels. Other studies
have shown that the dissolution pattern was very much influenced by tablet
hardness, so that compression force would be important. These findings
were with specific formuations , but they suggest that these variables must
be considered.
Because the dissolution patterns of sustained-release coated granules
are essentially the same whether the granules are filled into a capsule or
compressed with excipients into a tablet, the dissolution studies described
earlier are applicable here provided extensive fusing of the granules does
not occur in the tableting process.
Here, as in the case of capsules, there has been a wide range of drugs
formulated as sustained-release coated granules and compressed into tablets.
Antispasmodic-sedative combinations have been investigated [176],
as have phenothiazines [172 -180], anticholinesterase agents, and aspirin
[205,206].
Steigmann and coworkers [176] described the clinical evaluation of
Belladenal Spacatabs , a combination of the natural levoratory alkaloids of
belladonna and phenobarbital, in patients with peptic ulcer and other
gastrointestinal disorders. They measured gastric secretion and bowel
motility and found that. for the most part, the results with the Spacetabs
formulation were as good or better than those obtained with conventional
nonsustained forms, and the convenience to the patient of once or twice a
day dosing was recognized.
The treatment of myasthenia gravis with neostigmine bromide formulated
as Mestinon Bromide Timespan was examined by Magee and Westerberg
248 Chang and Robinson
[2051. The need for sustained release is particularly acute here. since
with treatment via non-sustained-release dosage forms, the patient generally
is very weak upon arising in the morning and remains in this state
until the morning tablet dose can be absorbed. A means of maintaining
patient strength and comfort throughout the hours of sleep was needed.
The results of their study show that the value of this sustained-release
dosage form is greatest when taken at bedtime. at times even allowing the
patient to arise with sufficient strength to dress before taking morning
medication. Results of daytime use were variable, and absorption appeared
to be less than optimal in comparison with non-sustained-release
tablets.
Probably the largest body of clinical data has been compiled for the
phenothiazine tranquilizer thioridazine. Mellinger [177], in an early
evaluation of thioridazine formulated as Me1laril Spacetabs, examined
serum concentrations following administration of the drug as a liquid concentrate.
as tablets crushed in a mortar, as intact tablets, and as the
Spacetab sustained-release tablet. He found that the drug persists in
the blood for long periods from all the dosage forms studied, and attributed
its slow excretion to possible slow metabolism. There were no
striking advantages to the sustained-release formulation over that of conventional
tablets. This last finding has been reiterated by other workers
{178] , and as of this time, the Spacetab formulation of thioridazine
is not being marketed.
The clinical evaluation and comparison of sustained-release and nonsustained-
release aspirin tablets was reported by Cass and Frederik [206.
207] • They measured the duration of analgesic relief obtained via a
series of different dosage regimens in patients suffering from a variety
of eoronie illnesses. In all cases, the sustained release form provided
longer and more predictable analgesic relief than any of the other nonsustained-
release tablets tested.
Case Study of Slowly Soluble Wax Coating on Nonpareil Seeds [189].
Prolonged action mechanism. Pulsed dosing in repeat action fashion.
Drug-containing nonpareil seeds were coated with various thicknesses of
digestible waxes which were intended to release drug at various times after
dosing.
Type and method of coating. A kilogram of medicated non-sustainedrelease
pellets was prepared in a 12-in coating pan. The pellets were
composed of 31.2 gm of dextroamphetamine sulfate; 58.8 gm of a 1:1 mixture
of starch, USP, and pOWdered sucrose, USPj and 90 gm of U.S.
No. 16 to 20-mesh sugar pellets. The nonpareil seeds were placed in a
conventional coating pan and wetted with a waterI alcohol! gelatin mixture
consisting of gelatin 10% wlv, hydrochloric acid 0.5% vlv, and water 30%
v lv , and alcohol 70% v I v (90% ethanol, 10% methanol). When the mixture
became tacky the drug diluent was added to the rotating seeds. After a
short period of drying. one-fourth of the seeds were removed (these
represented the uncoated portion). The remaining seeds Were then coated
to varying thicknesses with a wax formula consisting of glyceryl monostearate
11% wlw, glyceryl distearate, 16% wlw, white wax 3% wlw, in
carbon tetrachloride 70% wIw. Six different groups of pellets with approximately
7. 9, 11, 13, 15, and 17% of wax coating were initially prepared
and, through trial and error, a final blend consisting of 25% noncoated
pellets, 55% of the 11% wax-coated seeds, and 20% of the 9% waxcoated
seeds gave a satisfactory in vitro release pattern. Each group of
Sustained Drug Release from Tablets and Particles
Table 8 Blend and Desired in Vitro
Release Patterns, [14C] Dextroamphetamine
Sulfate
%In vitro release
at time interval (h)
249
Blend
Desired
a
0.5
34
39
2
56
62
4.5
79
80
7
92
90
aAverage in vitro pattern of 15 commercial
lots of sustained release dextroamphetamine
[ 190].
pellets was screened through U.S. No. 12 onto U.S. No. 25 standard mesh
sieves to remove lumps and ffnes ,
In vitro test. In vitro dissolution tests in artificial gastrointestinal
fluids were conducted according to the method of Souder and Ellenbogen
[50] • The dissolution pattern for all six coated seeds is shown in Figure
3, and the results on the blend in Table 8. A satisfactory prolongation
is obtained.
In vivo release. The in vivo release study was conducted in humans
and employed the following dosage regimens:
1. I5-mg sustained-release dosage form
2. 15-mg nonsustained dosage form
3. 5-mg nonsustained dosage form
4. 5-mg sustained-release dosage form given at intervals of 0, 4, and
8 h
This particular plan was chosen to determine performance criteria of
1. Whether the sustained-release formulation provided a prompt initial
dose
2. Whether it was similar to 3 times daily drug administration
3. Whether it was dissimilar to an equivalent nonsustained dose
4. Whether equal doses in different dosage forms were equally effective
in making drug available for absorption
5. Whether there was any significant variability among subjects receiving
the various regimens
The study was conducted in a crossover sequence, as shown in Table 9.
Blood and urine samples were collected at various times postdosing,
and the results are shown in Figures 4 and 5. The cumulative urinary
excretion data for human subjects is shown in Table 10. From the table
the following conclusions were drawn:
1. In the first 3 h following administration of the drug, the sustainedrelease,
3 times daily regimen, and the 5-mg single dose were
250
Table 9 Human Study Plan of Various Dosage
Regimens
Subject Initial
a 1 Week later
1 A B
2 B C
3 C D
4 D A
5 A C
6 B D
7 C A
8 D B
9 A D
10 B A
11 C B
12 D C
13 A D
14 B C
15 C B
16 D A
aA, 15-mg sustained release dosage form given
at 0 h; B, I5-mg non-sustained-release dosage
form given at 0 h; C. 5-mg non-sustainedrelease
dosage form given at 0 h; D, 5-mg
non-sustained-release dosage form given tid,
at 0, 4, and 8 h.
Source: From Ref. 189, used with permission.
Chang and Robinson
similar. Thus, the sustained-release formulation did provide a
prompt initial dose.
2. The plasma and urine plots for the sustained and 3 times daily
regimen were similar, whereas the sustained and I5-mg plain capsule
were dissimilar.
3. The sustained-release dosage form made as much drug available
as the other two I5-mg regimens.
4. The variation in plasma and urine data is not greater with the
sustained-release dosage form than with the 3 times daily regimen.
Sustained Drug Release from Tablets and Particles 251
240
•••• I ........ 0.,
..'
_..-
6 8 10 12 24 32 48
POSTDRUG SAMPLING TIME (hours)
.~, _.~.......-.- ,. -.. f ".,,0- • ... • ...·i ..
, ~ -·-1. ~~_
; .............-~
......- _.
..J
UJ > 120 w
..J
~80
~
Q. 40
,....-._,,,,,.------1
~ 200 '
E ' I
~ 160
Figure 4 Adjusted average human plasma levels. Each line represents
eight subjects per regimen in a balanced incomplete block crossover design.
---, lS-mg dextroamphetamine sulfate sustained-release dosage form;
.•.•••• , S-mg dextroamphetamine sulfate capsule, tid; - - - -, 15-mg
dextroamphetamine sulfate capsule; -' - • -, 5-mg dextroamphetamine sulfate
capsule. (From Ref. 189, used with permission.)
0.6
Vl
~ 0.5
- 0.2
a:
<
z 0.' a:
::>
o 3 6 9 12 24 32
END OF POSTDRUG COLLECTION INTERVAL
(hoursl
Figure 5 Adjusted average human urinary excretion rates. Radioactive
counts are expressed as average milligrams of dextroamphetamine sulfate
per collection interval divided by the number of hours in each interval.
---, lS-mg dextroamphetamine SUlfate sustained-release dosage form;
.•••••• , 5-mg dextroamphetamine sulfate capsule, tid; - - - -, lS-mg
dextroamphetamine sulfate capsule; -' -' -, 5-mg dextroamphetamine sulfate
capsule. (From Ref. 189, used with permission.)
252 Chang and Robinson
Table 10 Adjusted Average
a
Urine Recoveries in Humans, mg of
Dextroamphetamine Sulfate Equivalent
Collection interval (h)
Dosage form 0-12 0-24 0-48
5-mg capsule 2.18 3.90 4.89
15-mg sustained-release capsule 4.30b 8.24b
11.78
b
5-mg capsule, 3 times each day 4.32 8.48 12.30
15-mg capsule 6.14 9.31 11.86
~ach figure is the average for eight humans.
bNo figure included in a brace is significantly different from any other
figure included in that brace (p < 0.05).
Source: From Ref. 189, used with permission.
Sustained Dosing
Although the principal factors controlling drug release are very similar in
this case to those noted in the previous section, they do differ in at least
one important aspect; that is, the drug is made available in a continuous
rather than a pulsed fashion. The continuous release of drug is a result
of the drug being impregnated in a slowly dissolving film; as dissolution
occurs, drug becomes available [10,18,208 - 210]. This type of coating is
very similar to the embedding of the drug in an insoluble matrix, which
will be described later in this chapter. The difference lies in the fact
that these products are microencapsulations of drug particles or granules,
whereas the matrix tablets are formulated in a different manner.
MICROENCAPSULATION. Tanaka and coworkers [95] investigated the
effects of formalin treatment on the hardness of gelatin microcapsules of
sulfanilamide and riboflavin. As can be seen in Tables 11A and 1ill, the
treatment of gelatin micropellets containing sulfanilamide by immersion in
10% formalin/isopropanol for 24 h results in a lO-fold increase in time to
release 100% of the drug. These results were mirrored when sulfanilamide
and riboflavin micropellets were administered to dogs and blood levels of
Table l1A Dosages and Contents of SA and RF Micropelletsa
Sample Dosage and contents
1 Gelatin micropellet containing 33.2% SA
3 Micropellet, treated for 24 h, and containing 10.0% SA
aSA and RF refer to sulfanilamide and riboflavin, respectively.
Source: Reproduced with permission of the copyright owner.
Sustained Drug Release from Tablets and Particles
Table 118 Percentage of Accumulative SA Recovered in
the in Vitro Dissolution Testa
Sample 1 Sample 2
Time % Time %
5 min 32.9 5 min 5.9
10 min 59.5 10 min 9.9
15 min 76.5 20 min 30.7
35 min 80.0 30 min 39.6
1 h 89.6· 45 min 53.4
2h 99.5 1 h 61.5
3 h 99.7 2h 82.0
3 h 86.0
5 h 91.6
7 h 94.1
23 h 98.2
30 h 99.6
aSA and RF refer to sulfanilamide and riboflavin. respectively.
Source: Reproduced with permission of the copyright
owner.
253
the drugs versus time were measured. Figures 6 and 7 show that, indeed,
sustained blood levels of sulfanilamide and riboflavin were obtained
when the micropellets were hardened. Nixon et al , [211,213] studied
gelatin coacervate microcapsules of various sulfa drugs and the effect of
various coacervating agents on in vitro release of drug. They found that
hardened microcapsules gave a more prolonged release of drug in both
acid and alkaline pepsin medium. Temperature and pH effects were also
investigated. and from the data it was concluded that dissolution was the
controlling step rather than diffusion of drug through a microcapsular
wall.
The development of the complex form of coacervation as a tool for coating
pharmaceuticals was developed by Phares and Sperandio [214]. The
technique was further investigated and developed by Luzzi and Gerraughty
[217,219] and by Madan et al, [215,216]. They examined the effects of
varying starting pH. starting temperature, ratio of solid to encapsulating
materials, quantity of denaturant, and final pH. Their results indicate
that manipulation of all these variables affects some degree of change in
the microcapsules and the resulting drug-release rate. When the drug to
be encapsulated by the gelatin/acacia system was a waxy solid, such as
stearyl alcohol. the coacervation procedure had to be modified because
microscopic examination showed that, in the case of drug particles smaller
254 Chang and Robinson
--~-.--- .......... .-- . ..--.... -.. ~
co
.E
o
aa
-J
ClJ:
Z
o.~---"":":----~---~---~---~--~
TIME (hours)
Figure 6 Logarithm of SA levels in blood against time after administration
of SA gelatin micropellets to dogs. (0) Unitreated micropellets, (.) micropellets
treated with formalin/isopropanol for 24 h. (From Ref. 213, used
with permission.)
"ij.
.5
oo9
al
~
Zo
i= -c
ex:
Iz
woz8
TIME (hours)
Figure 7 Logarithm of RF concentration in blood after administration of
RF solution and its gelatin micropellets to dogs. ( 0) RF solution, (.)
untreated micropellets, ( ...) micropellets treated with formalin /isopropanol
for 24 h. (From Ref. 213, used with permission.)
Sustained Drug Release from Tablets and Particles 255
than 250 urn, encapsulation was accomplished by several droplets aggregating
and coalescing around the particles rather than by a single droplet as was the
usual case. With drug particles larger than 250 um, they suggested that encapsulation
was the result of a direct interaction of gelatin with acacia on the
surface of the particles. Scanning electron micrographs lend credence to their
suggestion. Merkle and Speiser [218] prepared cellulose acetate phthalate
coacervate microcapsules and evaluated them with respect to optimum coacervation
and encapsulation conditions. Their findings indicate that while the
amount of drug encapsulated had no significant effect on the particle size distribution
of the microcapsules, it did influence the release rate. The suggested
mechanism for this system is drug diffusion through the shells. When the
shells are plasticized in a manner similar to that described earlier. the release
rate control is altered from drug diffusion through the shells to dissolution of
drug in the microcapsules.
More recently, Birrenbach and Speiser [220] employed polymerized micelles
to produce very small particles. termed nanocapsutes , to be distinguished from
microcapsules. Although this approach is suggested for colloidal solutions to
administer antigenic material, it has potential for parenteral drug delivery.
Nylon microcapsules have been examined by McGinity et al , [221], who
described an improved method for making them. and by Lu zzi et al .• who evaluated
the prolonged release properties of nylon microcapsules that had been
either spray dried or vacuum dried. Both methods of drying produced a large
reduction in the dissolution rates of the microcapsules, but since the spraydried
material was free flowing as compared with the vacuum-dried material,
the authors felt that spray drying would result in more uniformity and reproducibility
of release rates. On the other hand. increased release rates can be
obtained by incorporating sucrose into the nylon microcapsules [18]. Interestingly.
and perhaps not surprisingly, Luzzi et al , found when compressing
thesa microcapsulated drugs into tablets that the release rate was inversely
proportional to the tablet hardness.
Bead polymerization as a technique for preparing sustained -release
dosage forms was described by Khanna et al , [222] and further evaluated
by Khanna and Speiser [223]. This technique results in drug being embedded
in the coating so that its release characteristics are similar to those
to be described in the section on the plastic tablet below. By varying the
concentration of the a -methacrylic acid content of the polymer solution, the
coating can be made to release drug over a wide range of pH values and
the release is prolonged for 12-15 h. By combining mixtures of beads with
various concentrations of a -methacrylic acid, the correct sustained-release
pattern can be made. Seager and Baker [224] described a somewhat different
system for making microencapsulated particles in the subsieve size range.
These resulted in drug being embedded in an inner core, whereas the outer
core was a shellac coating. The release pattern showed good sustainedrelease
properties.
Si-Nang et al, examined the diffusion rate of encapsulated drug as a
function of microcapsule size. The influence of the coating on diffusion and
the determination of the coating thickness were presented in terms of complex
mathematical equations. The equations used were of a general nature
and allow a quick estimation of the coating thickness, and thus can be useful
in modifying the microencapsulation procedure to attain desired release
capabilities.
Crosswell and Becker [225) reported a bead polymerization technique
for producing sulfaethylthiadiazole and acetaminophen microencapsulationa.
256 Chang and Robinson
When polystyrene beads were produced in the presence of drug solution,
no sustained release was evident; whereas if the beads were produced
without drug present and allowed to expand via exposure to n-pentane
and subsequent boiling in water, and then the drug solution was allowed
to seep into the deep channels produced in the expanded polystyrene
beads, the release pattern exhibited good sustaining properties.
The extensive use of polymer-drug interactions to produce sustainedrelease
dosage forms was promoted by Willis and Banker [226]. The drug
was made to interact with a cross-linked copolymer such as 1, 12-dihydroxyoctadecone
hemiester of poly(methylvinylether/maleic anhydride) and the
salt that was formed as a result of this exhibited good sustained-release
properties upon dialysis testing in artificial gastric and intestinal fluids of
tablets and granules fashioned from the polymer-drug salt entities.
Banker and various coworkers [227 - 231] further studied molecular
scale drug entrapment as a means of producing sustained-release dosage
forms. Although the technique employed appears to be similar to bead
polymerization, it is not clear whether each drug particle was coated (entrapped)
within the polymer or whether the drug was actually chemically
bonded to the polymer. For purposes of this discussion. the results are
similar to those obtained via bead polymerization processes, so they are
Included here. The authors prepared and tested cationic drugs such as
methapyrilene hydrochloride. chlorpromazine hydrochloride, atropine sulfate,
etc; , for their amenability to the entrapment procedures. The subsequent
increases in sustained-release properties Were tested and
reported.
Further exploration revealed that additives SUch as organic acids
could greatly facilitate drug entrapment by increasing the degree of interaction
between the drug and the polymer, and could provide more control
over the sustained-release characteristics. Other variables, such as
flocculation pH, rate of agitation, use of different polymers , etc., can
exert significant effects on sustained-release properties. Some of the important
variables and their influences are described in Table 12.
Anionic drugs, such as sodium phenobarbital, sodium salicylate,
chloral hydrate, etc; , were also entrapped via coagulated (gelled) polymer
emulsion systems. These polymer-drug products were tested in a manner
similar to that employed for the cationic drugs and the increase in and
reproducibility of the sustained-release products was noted. The authors
make special mention of the fact that their procedures result in a highly
uniform distribution of drug throughout the polymeric system and no drug
segregation, with resultant variability in blood levels of drug after dosing,
was evident upon scale-up processing and blending for incorporation into
sustained-release dosage forms.
All the procedures mentioned thus far in this section are at least
loosely related via their production of microencapsulated drug particles
that employ slowly soluble films or coatings as the encapsulating material.
A good portion of the literature in this area is experimental, usually resulting
in statements that reflect how well these methods might be for
actually producing commercial sustained-release products. As was mentioned
earlier. very few products. aside from aspirin, have been actually
formulated from these microencapsulated particles. and thus there is a
scarcity of clinical evaluations in the literature. In in vitro evaluation of
sustained-release systems, one would like to have linearity in drug release
versus time up to 60-70% or more of drug content. However. it is
Sustained Drug Release {rom Tablets and Particles
Table 12 Summary of Effects Produced by Variables on the Molecular
Scale Drug Entrapment Method of Banker and Coworkers
257
Variable studied
Methapyrilene base and
hydrochloride salt
Entrapment of cationic
drugs by polymeric
flocculation
Addition of a suitable
organic acid to the system
described in Ref.
227
Flocculation pH and rate
of agitation
Entrapment of anionic
drug by polymeric
gelation
Effect on entrapment
No appreciable difference in equilibrium
dialysis release rates between free base
and hydrochloride salt. either as solutions
or as solids. Prolongation of release
of drug from granular and tableted
forms in vitro was shown.
The flocculation of highly concentrated
colloidal polymeric dispersions (latices)
in the presence of the drug in solution
which is to be occluded provides the
entrapment mechanism. Significant increases
in duration of action and reduction
in acute toxicity were demonstrated.
Excellent control of sustained-release
characteristics. Drug entrapped as
the carboxylate salt or in conjunction
with the appropriate dicarboxylic acid.
while demonstrating SUbstantially complete
drug dissolution release in intestinal
fluid. could be maintained in
the entrapped form in aqueous suspensions
during storage periods in
excess of 1 month. Greater binding
of drug was also shown.
Effect of an increase in flocculation pH
was twofold. It increased both the
amount of drug bound and the rate at
which drug was released. Increase in
the stirring rate during flocculation
resulted in an increase in the amount
of drug bound to the polymer.
The phenomenon of gelation of the
polymer emulsions by the addition of
divalent cation (Mg2+) was utilized for
the entrapment of various anionic drug
materials. Good control both in vitro
and in vivo sustained release characteristics
were shown. With increases in
drug concentration. release was rapid
and a poor release pattern resulted.
With diminished drug concentration. too
small an amount of drug was released
initially to provide therapeutic levels.
Smaller size gel particles resulted in
more rapid and complete release of
Ref.
226
227
228
229
229
230
258
Table 12 (Continued)
Chang and Robinson
Variable studied
Uniformity of dlstrdbution
of amine drug in
solid dispersions
Effect on entrapment
drug. This suggests mixing gel particle
sizes to get a particular release
pattern.
Excellent reproducibility of drug content
throughout the entire entrapment
product as demonstrated in both flocculated
(high drug levels) and deflocculated
(low drug levels) systems. Dry
blending was inferior to molecular scale
drug entrapment in distributing small
quantities of drug uniformly.
Ref.
231
common to see studies showing linearity in the sustained effect for only
30-40% of the system. The study by Nixon and Walker [211] using gelation
coacervate showed linearity up to 60% release.
Of the clinical evaluations reported, microencapsulated aspirin has received
the lion's share of attention. Bell et al , [232] pointed out the
need to monitor blood levels of both acetylsalicylic acid (ASA) and its
metabolite. salicylic acid (SA). The need for this is obvious because ASA
is reportedly far more potent as an analgesic than SA, so that modification
of sustained-release products to provide more ASA and protect it from
metabolism to SA become very important. The Bell study compared sustained
-release aspirin versus regular aspirin administered as a single dose
and in divided doses and analyzed the blood levels of ASA provided by
each dosage form or regimen. This resulted in statements to the effect
that sustained-release aspirin gave greater analgesic effects than regular
aspirin because the ASA blood level remains higher for longer periods of
time. When total salicylate blood levels are measured all salicylates presumably
producing and contributing to the anti-inflammatory effect of
aspirin, the sustained-release product still gave release rates that were
too small for sustained-release aspirin. Optimum levels were achieved with
about 3% of the drug in the encapsulated material.
Green [80] investigated sustained-release aspirin tablets formulated
from microencapsulated particles and found the blood level curves of
salicylate to be virtually flat with repeated dosing as opposed to a sawtooth
effect exhibited with repeated dosing of regular aspirin. No conclusion
was drawn from this observation, but the influence is that a flat
blood level curve represents better control over the therapeutic regimen.
Rotstein et al, [233] examined sustained-release aspirin for use in the
management of rheumatoid arthritis and osteoarthritis. In short-term
double-blind crossover studies, the doses employed were large enough that
the patients received relief from both regular and sustained-release aspirin
and observable differences between the two were obscured. However, in
long-term usage studies, all the patients who had been on any type of
previous salicylate therapy preferred the sustained-release aspirin
Sustained Drug Release from Tablets and Particles 259
formulation because it reduced the frequency of dosage and supplied more
medication during the night hours, thus relieving the morning aches and
pains of arthritis. In both studies the incidence of side effects was significantly
lower with sustained-release therapy. These clinical studies
have thus established that microencapsulation of aspirin formulated into
sustained-release dosage forms is an attractive and effective alternative to
nonsustained dosage forms.
IMPREGNATION. There are two general methods of preparing drugimpregnated
particles with wax: (1) congealing and, (2) aqueous dispersion.
In the congealing method, the drug is admixed with the wax material
and either screened or spray congealed. The aqueous dispersion approach
is simply spraying or placing the drug-wax mixture in water and
collecting the resulting particles.
In a series of papers by Becker and associates [225,234 - 239], the
formulation and release characteristics of wax impregnations of sulfaethylthiadiazole
were thoroughly investigated. The authors looked at dispersant
concentration, effects of surfactant addition, effects of different waxes and
modifiers, etc., as to their effect on release rate and proportion of total
drug constituting the prolonged-release fraction. As might be expected.
the size of the microcapsules produced and the physical properties of the
various wax coating materials had profound effects on the release patterns
reported. When the spray-congealed formulations of wax and sulfaethylthiadiazole
were compressed into tablets the release mechanism appeared to
be due to erosion, solubilization, and leaching of the drug from the tablet.
Noone model could describe the release pattern over the 48-h period of
the study. It has been reported, in general terms, that the aqueous dispersion
method gives higher release rates for all waxes tested, presumably
due to increased area and perhaps the physical entrapment of water. Further,
with aspirin as the test drug, the dissolution rate increases in the
order stearic acid > spermaceti > hydrogenated cottonseed oil.
Case Study of Slowly Soluble Wax Microencapsulation [234]. Prolonged
action mechanism. Drug is impregnated in a slowly dissolving wax.
Type and method of coating. Bleached beeswax or glycowax S-932 is
mixed with drug, in a ratio of 1 part drug to 3 parts wax. and heated on
a water bath to 75QC. Typical amounts were 24 gm beeswax and 8 gm
sulfaethylthiadiazole (SETD). In a separate container heated to 80°C was
1 ml sorbitan mono-oleate and 1. 15 ml polysorbate 80 in 400 ml of distilled
water. The aqueous phase was slowly added to the wax mixture with continuous
stirring at a predetermined speed until the mixture cooled to approximately
45QC. Stirring at around 300 - 400 rpm produced particles that
were primarily in the size range of 30-100 mesh. The drug-wax particles
were separated from the aqueous phase by filtration, washed with distilled
water, and dried. Fractionation of the resulting particles into three
mesh sizes 16-20, 30-40, and 50-60 was accomplished with USP sieves,
and only the 50- to 60-mesh particles were retained for in vitro and
in vivo testing.
In vitro test. Dissolution tests were conducted in a modified USP
disintegration apparatus, and the results are shown in Figures 8(a) and
8(b). Drug release during the first 15 min was in direct relation to the
specific surface, since the particles were assumed to have a uniform distribution
of drug on the surface. After the first 15 min period, the rate
of drug release varied as a function of mesh size and appeared to follow
260
e
Chang and Robinson
(a)
400
I
TIME (hours)
2
o
(b)
2 3
TIME (hours)
4
Figure 8 (a) In vitro dissolution rates of SETD from various mesh sizes of
SETD-glycowax particles in 0.1 N HCI. (e) 16-20 mesh, (A) 30-40 mesh,
Ce - C = O. (b) In vitro dissolution rates of SETD from various mesh sizes
of SETD-glycowax particles in alkaline pancreatin solution. (.) 16- 20
mesh, (A) 30-40 mesh, (0) 50-60 mesh. (From Ref. 234, used with
permission.
first-order kinetics. The first-order dissolution appears to be due to the
changing surface area.
In vivo release. Urinary excretion rates of SETD-glycowax, 50-60
mesh size range, were compared to rates for plain SETD in four humans.
The 50-60 mesh size was selected because the in vitro release data were
most similar to the in vitro release of a similar commercial prolongedrelease
SETD product. Figure 9 shows the comparison of the excretion
rates for the plain SETD and the prolonged release form. The SETDglycowax
particles released 50% less drug during the first 3 h , and then
the rates increased so that at the end of 24 h , 71% of the total SETD had
been excreted as compared to 85% for plain SETD. Over 72 h, the
Sustained Drug Release from Tablets and Particles
080
060
261
o
TIME (hours)
12 15
Figure 9 Average urinary excretion rates of free SETD for four humans
receiving a 3.9-gm oral dose of SETD in (a) plain form and (b) SETDglycowax
combination. (e) plain SETD, (0), SETD-glycowax combination.
(From Ref. 113, used with permission.)
in vivo SETD-glycowax release gave 85% excreted versus 81% in the
in vitro experiment.
B. Sustained-Release Utilizing Diffusion
This section describes coated sustained-release systems that are distinguished
from those in the previous section primarily by their mechanism
of action (i.e., diffusion). as well as by their means of application and
by the appearance of the final dosage form. Some of the microencapsulated
materials discussed earlier released drug via a diffusion process, but the
bulk of the diffusion systems will be discussed here. In the case of slowly
soluble films and coatings, the reader will recall that these were generally
applied via pan-coating techniques or by some form of polymerization
with the end result being coated pellets or granules or microspherules
that could then be compressed into tablets or placed in a capsule. The
products in this section are, for the most part, press-coated films whereby
a core is fed into the die and the coating material is then pressed onto
it. resulting in a single-coated tablet. or they are whole tablets or particles
that have been coated via air-suspension techniques. As in the previously
described cases, the area and the thickness of the coating are important
parameters. but they take on added importance here because diffusion
of drug, either from the core or from the coating itself, must pass
through the barrier represented by the coating Or film. Insoluble ffims
will thus present a rigid barrier that will act to keep the diffusion path
length of drug from the tablet interior to the absorption site (outside of
tablet) relatively constant. Some of the ffims deserdbed below are also
slowly soluble, but their maintenance of constant path length is still the
most important activity.
262 Chang and Robinson
As might be expected, a lot of the research in this area has to do with
the mechanics of the coating process itself, in addition to the research on
developing new coatings and measuring the efficacy of administering drugs
in this manner. Of course, the patent literature is filled with polymeric
systems that can be used via press-coating or air-suspension techniques
[ 154,165, 240, 241]. and the research literature likewise [242 - 250] • However,
since the press-coating and air-suspension processes are utilized in
the pharmaceutical industry, a body of research has been developed to
overcome some of the manufacturing difficulties presented by these techniques
[251- 254] •
The original work that developed the air-suspension technique has
been reported by Wurster [255,256]. This patented process is now widely
used in the industry, since it is a fast. efficient way of making uniform
coatings on granules and core tablets. The use of the technique for coating
aspirin granules has been described by Coletta and Rubin [257].
Wood and Syarto [258] continued the same line of work involving coating
aspirin with various ratios of ethylcellulose to methylcellulose. As the
methylcellulose dissolved, whereas the ethylcellulose did not, the shell left
behind presumably provided a restraining barrier for keeping the ASA
diffusion path length constant. These authors attempted to correlate the
in vitro dissolution pattern with that obtained in vivo, as shown in Figure
10. As can be seen from the correlation lines, with high-content
ethylcellulose the in vitro/in vivo correlations are good. As the percent
of methylcellulose in the coating is increased, the correlation falls off, and
this points out the fallacy of using in vitro data alone to predict in vivo
performance of these tablet coatings.
Many researchers have, of course, looked at the polymeric materials
used in the coating processes with an eye to developing better, more
durable, and more easily applied coats. Polyvinylpyrrolidone-acetylated
monoglyceride [244], styrenemaleic acid copolymer [245], hydroxypropylcellulose-
polyvinylacetate [259, 260], and many other polymers have been
studied for their value as enteric coatings [243, 248, 250] and for use in
developing sustained-release preparations [246 - 248] • Enteric coating is a
SUbject dealt with elsewhere in this text, but the application of enteric
coats can be accomplished via press-coating and air-suspensions techniques
just as described here.
Donbrow and Friedman [259] reported the release of caffeine and
salicylic acid from cast films of ethylcellulose and the release rates were
found to agree with both the classic first-order equation (log drug retained
in ffim versus time) and with the diffusion-controlled release models
(drug release linearly related to square root of time) as developed by
Higuchi [260]. More stringent mathematical treatment of their data resulted
in the diffusion-controlled release model being most appropriate to
describe their data. Borodkin and Tucker [261,262] also used cast fflms
of drug in hydroxypropylcellu!ose, studying the release of salicylic acid,
pentobarbital. and methapyrfiene. These drugs were released according
to the same diffusion-controlled model described above. Further work to
modify the system resulted in zero-order drug release obtained by laminating
a second mm without drug to the releasing side of the film containing
drug. Thus, the nondrug layer functions as a rate-controlling membrane,
and the drug-containing film serves as a reservoir. In vitro zeroorder
drug release for the three species mentioned was demonstrated
using this technique.
Sustained Drug Release from Tablets and Particles
3.0
2.5
/
263
Ii;
w
Io
a:
!::
>~
>to
W
:E
t=
/ ,
/
TIME IN VIVO [hours)
8 10 12
Composition and Characteristics of Test Delayed-Release Aspirin Products
Used in This Study
Code
1348 134C 152 138
Ratio ethyl to methylcellulose 75/25 25/75 82.5/17.5 10010
Aspirin mesh size -20 -40 -20+40 -20+40
Amount of coating (wt %) 2.7% 4.8% 6% 6%
Tablet disintegration time (sec) 40-60 25-35 3 3
Aspirin content 5 gr 5 gr 5 gr 5 gr
Cornstarch 0.87 gr 0.90 gr 0.96 gr 0.96 gr
Talc 0.13 gr 0.13 gr
Figure 10 Correlation between in vivo absorption and in vitro release
rates for corresponding fractions of total salicylate considered. (.) 152.
(.) 138, ( ... ) 134C. (0) 134B. (From Ref. 258. used with permission.)
264 Chang and Robinson
Since the technique described in this section can be used to coat core
tablets or granules, it is conceivable that many of the clinical studies described
earlier could be applicable in this case. However, a very good
clinical study of the Sinusule sustained-release dosage form has been descrfbed
[263]. This product is somewhat different from the others examined
thus far in that the film applied to the granules actually functions
as a microdialysis membrane. Thus, one need not worry about the acidity
of the gut, the digestive process, nor the contents of the gut. The only
requirement is that there be fluid in the gut. The fluid, primarily water,
passes through the dialysis membrane into the sphere and dissolves the
granule of drug. This drug then diffuses through the intact membrane at
a rate proportional to the permeability of the membrane, the mobility of
the drug molecule, and the concentration of the drug within the hydrated
microdialysis cell. When this product was tested in patients having hay
fever. upper respiratory infection, and miscellaneous respiratory allergies.
good to excellent results with a' minimum of side effects were reported for
a large majority of the test population, This novel sustained-release
dosage form is not marketed anymore.
Case Study of DuffusiDn-Controlled Coating [261]
PROLONGED ACTION MECHANISM. Drugs are dispersed in a watersoluble
polymeric coating which. when SUbjected to an aqueous environment,
allows slow diffusion of drug into the leaching fluid.
TYPE AND METHOD OF FILM FORMATION. Hydroxypropylcellulose,
average molecular weight of 100,000, having viscosity 75-150 cps as a 5%
water solution, and polyvinylacetate, average molecular weight 500,000,
having viscosity 90 -110 cps as an 8.6% benzene solution. were used.
Drugs were pentobarbital, salicylic acid, and methapyrilene. The films were
cast from a solution containing 10% solids (drug plus polymer), using
methylene chloride/methanol mixture (9:1) as the solvent. The polymers
were added as dry powders, as was salicylic acid, While the pentobarbital
and methapyrilene were added from stock methylene chloride solutions.
Films were cast from the solutions at various wet thicknesses (0.64- 2. 54
mm) using a knife on Teflon-coated plate glass. The films were allowed to
air dry at least 48 h before evaluation. The percent drug in the dry film
was calculated from the ratio of drug and polymer weights used.
MECHANISM OF DRUG RELEASE. Release of drug from a matrix can
be either via a first-order process or via a diffusion-controlled process.
Which release is operative can be ascertained by appropriate data treatment.
First-order release would be a linear plot follOWing the normal
first-order equations, whereas a diffusion-controlled process should result
in an S-shaped curve following the equation
Q =In.. (2A - EC ) C t
T S S
where Q =amount of drug released per unit area of tablet exposed to
solvent
D = diffusion coefficient of drug in the permeating fluid
E =porosity of the matrix
(13)
Sustained Drug Release from Tablets and Particles
T :::: tortuosity of the matrix
A :::: concentration of solid drug in the matrix
C :::: solubility of drug in the dissolution medium
s
t =time
The above equation is more commonly expressed as
265
(14) Q = k t 1/ 2
H
where kH:::: ED[(2A - Cs)C s]1/2 for plotting purposes.
IN VITRO TEST. Rectangular films measuring 2.2 x 4.0 em (8.8 cm
2)
were cut using a razor blade with a microscope cover glass as a template.
The film was weighed and its thickness was measured at all four corners
and the center with a micrometer. A thin coating of high-vacuum silicone
lubricant was applied to a 2.54 x 7.62 crn microscope slide and the film
was pressed into the slide, making sure that all edges adhered and no
lubricant touched the exposed surface. The slide was placed at an angle
into a 250-ml beaker in a 37°C-water bath containing 200 m1 of pH 7 buffer
preheated to 37°C. A nonagitated system was used to eliminate turbulence
effect on release rate and to maintain film integrity. Periodic
assay samples (:tl0 min-I) were obtained by removing the slide, stirring
the solution, and pipetting a 5-ml sample, and then reimmersing the slide
with film into the buffer solution. Beakers were covered throughout the
length of the runs (7 h at least and 31 h in the extreme for slow-releasing
films) to prevent evaporation. The samples were assayed by ultraviolet
(UV) spectrophotometry at 240 nm for pentobarbitsl in 0.1 N NH40H, at
312 nm for methapyrilene in 0.1 N Hel, and at 297 nm for salicylic acid in
O.1NNaOH.
Table 13 shows the results of comparing the drug release to firstorder
and square root of time equations. When these data were evaluated
to determine which equation gave best fit, the correlation coefficients for
the best statisticsl lines and the lag times (time intercept extrapolated to
Q = 0) were used as the principal criteria. Although the corrslation coefficients
looked good for either mechanism, when these types of data are
plotted out, as is shown in Figure 11, the curvature in the first-order
mechanism was evident. This indicated that the Q versus t 1/ 2 relation
more readily described the mechanism. Correlation coefficients were generally
greater than 0.995 for this type of plotting and deviations, when
they occurred, were random rather than the result of curvature. Linearity
in release held through 75-80% of drug release when a constant concentration
gradient was operative.
The effect of film thickness on the rate of dru g release for pentobarbital
is shown in Table 14. These results indicate that the release
rate constant, kH, is independent of film thickness. However, film
thickness wID affect the duration of drug release, as is shown in the last
column where t1/2 represents the time, in minutes, for 50% of the drug in
each particular film to be released.
Tables 15-17 show the effects of varying polymer ratio on the release
rate, kH, for the drugs methapyrilene, salicylic acid, and pentobarbital,
respectively. In general, an acceleration of release rate can be obtained
Table 13 Comparison between First-Order and Q Versus t 1/ 2 Treatments of Pentobarbital Release Rate Data
First-order
Q versus t 1/2
Drug
1ag 1ag concentration Hydroxypropylcel1ulosel Number Correlation Correlation
(%)8 polyvinylacetate ratio of runs (min)b coefficient {min)b coefficient
36.4 10:0 4 -3.2 0.996 3.1 0.997
18.2 10:0 4 -14.0 0.957 2.6 0.993
18.2 9:1 2 -9.8 0.986 2.8 0.996
18.2 8:2 2 -6.5 0.986 2.3 0.993
18.2 6:4 2 -84.0 0.982 1.4 0.996
18.2 4:6 4 -142.0 0.983 0.3 0.998
18.2 2:8 4 -176.0 0.981 0.6 0.998
18.2 1:9 4 -176.0 0.982 0.7 0.998
18.2 0:10 3 -214.0 0.979 1.0 0.997
9.1 0:10 4 -199.0 0.981 1.3 0.990
aWeight of drug per weight of dry film.
b
All tIag and correlation coefficient values expressed are mean values.
Source: From Ref. 259, used with permission.
~
OJ
9Q
;::J
tQ
Q
;::J
~
~
S'
~
;::J
Table 111 Effect of Film Thickness on Pentobarbital Release Rate Constant and Half-Life
Wet fUm
Drug thickness Dry film
kH t 1/ 2 concentration Hydroxypropylcel1ulose/ setting thickness Correlation
(%)a polyvinylacetate ratio (mm) (J..lm)b (mg cm- 2 min- 1/ 2) coefficient (min)
36.4 10:0 0.64 44.0 ± 1.1 0.33 0.996 14.8
1.27 56.2 ± 2.1 0.29 0.996 25.2
1. 91 100.2 ± 4.6 0.33 0.998 49.8
2.54 109.3 ± 3.3 0.32 0.998 60.2
18.2 4~6 0.64 61.4 ± 4.9 0.032 0.999 360
1. 27 113.2 ± 9.9 0.034 0.999 1,280
1.91 145.0 ± 8.0 0.034 0.999 2,850
2.54 204.2 ± 3.1 0.037 0.997 3,850
18.2 1:9 0.64 76.8 ± 1. 2 0.0101 0.996 8,520
1.27 118.4 ± 2.6 0.0102 0.999 19.200
1.91 202.8 ± 2.2 0.0101 0.998 50.900
2.54 268.0 ± 2.4 0.0102 0.998 83,600
~eight of drug per weight of dry film.
bThickness =mean :t standard deviation of five measurements.
Source: From Ref. 259, used with permission.
t\)
0)
~
268 Chang and Robinson
Figure 11 Comparison between first-order release treatment and Q versus
t1/2 treatment of data from a film containing 26.3% methapyrilene at a 5:5
ratio of hydroxypropylcellulose/polyvinylacetate. (e) log (QCQ - Q) versus
t , and (0) Q versus t 1/ 2. (From Ref. 259, used with permission.)
by increasing the proportion of hydroxypropylcellulose to polyvinylacetate.
The large variations in the t 1/ 2 values reflect the changes in the polymer
ratio and, in addition, the changes in film thickness.
c. Sustained-Release Utilizing a Combination of
Dissolution and Diffusion
The products to be discussed in this section are those that provide the
sustaining portion of the dose in some sort of relatively insoluble core that
has been impregnated with the drug. This core is almost always coated
and the coat contains that portion of the dose meant for immediate release
upon dissolution of the coat in the stomach. Once this occurs, the gastrointestinal
fluids are free to permeate the core and thus slowly leach out
the drug. Of course, the possibility exists that the core material can
sometimes be dissolved slowly to provide drug, but this is usually not the
the case. The diffusion of drug out of the core is the major mechanism
for providing drug in sustained release form. In these preparations the
area over which diffusion cocurs remains relatively constant (especially if
no core dissolution occurs) and the amount of drug is in excess. The
Sustained Drug Release from Tablets and Particles 269
Table lS Effect of Hydroxypropylcellulose/Polyvinylacetate Ratio on the
Release Rate from Films Containing 26.3% Methapyrilene
HydroxypropylcelluloseI
Dry film
k
H t 1/ 2 polyvinylacetate thickness Correlation
ratio (llm)a, b (mg cm- 2 min- 1/ 2) coefficient (min)
10:1 210 ± 7.4 0.549 0.999 33
9:1 210 ± 6.5 0.593 0.997 28
8:2 216 ± 9.1 0.497 1.000 42
7:3 181 ± 7.8 0.272 0.995 99
6:4 218 ± 4.4 0.217 0.998 226
5:5 208 ± 4.6 0.225 0.999 190
4:6 142 ± 4.5 0.191 0.997 124
3:7 157 ± 3.3 0.094 0.998 630
2:8 136 ± 3.3 0.089 0.999 523
1:9 131 ± 6.0 0.075 0.998 681
0:10 185 ± 0.9 0.074 0.996 1404
aThickness :;: mean ± standard deviation of five measurements.
b All films cast using a wet thickness setting of 2.54 mm,
Source: From Ref. 259. used with permission.
factor that changes, in this case, is the path length term in Fick's first
law. As more drug diffuses out of the core, the permeating gastrointestinal
fluid must travel an increasingly longer and more tortuous path
to get to the remaining drug. The dissolved drug, in turn, has to diffuse
out via the same altered pathways. Thus, a tortuosity factor must
be included in the equation to describe release.
One of the earliest of these core-type products to be described was
Duretter, developed by Sjogren and Fryklof [264] in Sweden. It differs
from other core-type tablets, such as Ciba's Lontab, in that the core is
produced by directly compressing a granulate of the drug and an insoluble
plastic material so that a coherent, porous skeleton of the matrix material
forms around the drug. In core tablets such as Lontab, on the other
hand, the drug is incorporated into the melted matrix material and this is
then spread out to dry. After drying it is granulated, and this granulation
is then compressed into the core tablet [265]. In the final result the
differences in manufacturing are not evident, and the release patterns of
drug from these species are similar in many respects.
The release rate and absorption characteristics of various drug incorporated
into a plastic matrix type of tablet have been extensively studied.
Sjogren and Ostholm [266] studied the release of nitroglycerin, lobeline
hydrochloride, 82Br-Iabeled ammonium bromide, creatinine, potassium
270 Chang and Robinson
Table 16 Effect of HydroxypropylcelIulose/Polyvinylacetate Ratio on the
Release Rate from Films containing 20.0% Salicylic Acid
HydroxypropylcelluloseI
Dry film kH t1/2 polyvinylacetate thickness Correlation
ratio (llm)a,b (mg cm- 2 min- 1/2) coefficient (min)
10:0 189 ± 2.8 0.403 0.997 28
9:1 223 ± 5.6 0.336 0.997 57
8:2 210 ± 0.9 0.328 0.993 53
7:3 229 ± 2.5 0.241 0.998 116
6:4 209 ± 6.6 0.216 0.984 120
5:5 204 ± 7.9 0.175 0.997 175
4:6 145 ± 0.5 0.115 0.997 206
3:7 156 ± 3.4 0.122 0.997 210
2:8 181 ± 7.9 0.112 0.998 338
1:9 222 ± 4.3 0.073 0.998 1190
0:10 172 ± 1.8 0.057 0.996 1040
s..rhickness = mean ± standard deviation of five measurements.
bAll films cast using a wet thickness setting of 2.54 mm,
Source: From Ref. 259. used with permission.
penicillin V. and dihydromorphinone hydrochloride both in vitro and in vivo.
in cats and humans. They noted good correlations between in vitro and
in vivo results. both in blood levels of active drug. and also when monitoring
a particular pharmacological effect produced by the drug, although of
course in vivo determination of release rate was much more difficult to estimate.
In a continuation of this type of study. Sjogren and Ervik [267]
developed an automatic spectrophotometric method for studying continuous
release rates of quinidine bisulfate and ephedrine hydrochloride from
Duretter plastic matrix tablets and obtained highly reproducible results.
The complex interplay between the processes of release and degradations
of substances dispersed in polymeric matrixes has been described in
great detail by Collins and Doglia [268J. Although their discussion is of
general nature, the analogy to drugs and pharmaceutical systems is apparent.
EI-Egakey et al , [269] and Asker et al , [270- 272] have delved
into the in vitro release of drugs from polymeric matrixes and granulations.
In the case of water-soluble drugs, merely granulating the drug with the
melted matrix material will provide suitable sustained release of drug.
whereas with more water-insoluble drugs, a coating providing drug for
immediate release and absorption takes on increasing importance as the
degree of water solubility decreases. Obviously, the limiting case here is
a drug with virtually no water solubility. This drug would not need to
til r::
0.... QS·
Table 17 Effect of HydroxypropylcelluloselPolyvinylacetate Ratio on the Release Rate from Films ~
~
Containing 18.2% Pentobarbital t::l
2
Wet thickness Dry film cq
Hydroxypropylcellulose! setting thickness k
H Correlation t 1/ 2 ::tI
(mg cm- 2 min- 1!2) ~
polyvinylacetate ratio (mm) ( llm)a coefficient (min) irn
~
10:1 1. 27 98 ± 5.4 0.225 0.995 20 a- :1
9:1 2.54 163 ± 3.4 0.224 0.995 57 hi
Qe-
8:2 1. 27 87 ± 6.4 0.178 0.987 25 -~.... rn
6:4 2.54 236 ± 7.1 0.0767 0.997 1.010 Q
;:!
~
4:6 1.91 145 ± 8.0 0.0342 0.999 2.280 'tl
Q
""$ .... ....
2:8 1. 27 94 ± 1.9 0.0161 0.999 3.630 So
~
0
1:9 1. 27 118 ± 2.6 0.0102 0.999 19.200
0:10 1.91 210 ± 15.5 0.0260 0.998 6.960
aThickness =mean + standard deviation of five measurements.
Source: From Ref. 259. used with permission.
~
~
272 Chang and Robinson
be placed in a special sustained-release dosage form. since it is inherently
long acting by nature of its poor aqueous solubility. Of course, the method
of preparation of the granulating materials, the choice of in vitro dissolution
media. and the plastic matrix material chosen will all have an influence
on release of the drug from the sustained-release dosage form.
Water-soluble drugs dispersed in hydrophilic matrixes were studied by
Lapidus and Lordi [273]. Their results indicate that chIorpheniramine
maleate dispersed in methylcellulose is release-rate controlled mostly by
drug diffusivity rather than by polymer dissolution and water permeability.
Thus, even for drugs formulated in a water-soluble matrix, one which
would itself be subject to erosion and absorption in the body. the determining
factor in providing sustained release is still diffusion of the drug
out of the matrix. This important fact was further elaborated on by
Huber et al , [274] for hydrophilic gums versus the matrix material. In
this case. the mechanism of prolonged release was determined to be drug
diffusion from, and eventual attrition of. a gel barrier at the periphery
of the tablet core.
A modification of this type of matrix tablet was described by Javaid
et 81. [275]. They used a lipase-lipid-drug system to provide sustained
release whereby the erosion of the matrix due to the hydrolytic action of
lipase on the substrate was the desirable first step in obtaining release of
the drug for absorption. Accelerators of lipase activity, such as calcium
carbonate or glyceryl monostearate , could be used to tailor make a sustained-
release tablet to provide a desired release profile. Javaid and
Hartman [276] then tested these enzyme-substrata-drug tablets in dogs,
and the action of lipase to control drug release was confirmed. Tablets
containing lipase consistently gave higher and more uniform blood levels of
drug than those without. as is evident in Figure 12. It is apparent that
the enzyme-substrate type of matrix tablet has potential for commercial use
in providing sustained release.
The type of plastic matrix sustained-release dosage form that has been
investigated the most extensively is undoubtedly that of a drug dispersed
in an insoluble. inert matrix [3-5,10,15,18,278-285]. The kinetics of
release and the methods of treating the matrix tablets have been reported
~
UJ 10
i;j
oJ 08 oo
906
OJ
0.5 ie "3 4 5 6 7 8
HOURS
9
Figure 12 Average sulfamethizole blood levels (mg%) in dogs after receiving
tablets containing 100% drug in drug-lipid granules with 5%
glyceryl monostearate. (D) tablets with lipase and (0) tablets without
lipase. Standard errors are plotted around the averages. (From Ref.
276, used with permission.)
Sustained Drug Release from Tablets and Particles 273
by Farhadieh and coworkers [277,279] for drugs dispersed in a methyl
acrylate-methyl methacrylate matrix. As seen in Table 18, treatment of
the tablets by exposure to acetone vapor results in a significant decrease
in the release rate of drug from the matrix. This, coupled with control
of the treatment temperature, allowed production of sustained release tablets
with highly reproducible release rates. An investigation of the problems
that could result from chewing an insoluble matrix was the subject of
a paper by Ritschel [280]. If the drug is incorporated in the pores and
channels of a matrix tablet, potential toxicity could occur if the tablet was
inadvertently chewed by the patient, since this would result in the release
of large quantities of the drug. For drugs with a narrow therapeutic
range, such as nitroglycerin, this could be a problem. When the
drug was dissolved directly into the plastic matrix material. the problem is
alleviated since even with mastication the drug is not free to be absorbed.
Sjuib et al , [281,282] continued the work of Higuchi on the study of
drug release from inert matrixes. The original physical model described
by Higuchi and coworkers [286 - 294] was tested for binary mixtures of
acidic drugs and also for binary mixtures of amphoteric drugs. Analysis
showed that the physical model could describe the experimental data quite
well and also pointed out that the precipitation of the drug in the matrix
during release into alkaline media, such as might be encountered in the
intestinal fluids, has to be considered more important than theories involving
supersaturation of the drug in the .matrix. The mathematics of
both matrix-controlled and partition-controlled drug release mechanisms
were elucidated by Chien et al , [285]. They pointed out that the transition
that occurs between the two processes is dependent on the magnitude
of the solution solubility of the drug and, via a series of equations, indicated
that this term dictates the mechanism and rate of drug release from
the polymer matrix.
Other research in the area of drug release from insoluble matrix tablets
has found that the square root of time relation originally proposed by
Higuchf [260] described the advance of the solvent front into the tablet
[ 283] • Compression force was not a major factor and drug release was,
of course, proportional to the total surface area.
As can be seen from the above cited work, a tremendous amount of
time and effort has been expended to elucidate and quantify the factors
that are important in obtaining controlled-release from plastic matrix-type
tablets. The perturbations introduced by varying diluents, matrix material,
and so on are all important in describing the in vitro release of drug
from these tablets, and presumably are equally important in describing the
in vivo release rates. However, we would be remiss if the clinical evaluation
of these types of dosage forms are not included here, since the literature
is full of examples in which systems showing good in vitro possibilities
were unsuccessful in an actual clinical trial.
The effects of gastric emptying time and intestinal peristaltic activity
on the absorption of aspirin from a sustained-release tablet containing
coated particles in a hydrophilic gel type matrix and conventional aspirin
tablets were described by Levy and Hollister [295]. The lag times in the
absorption profiles reported were undoubtedly due to the time required
for transfer of the dosage form from the stomach to the intestine. The
authors point out the pitfalls of plotting averaged individual absorption
data versus time and how this could lead to erroneous assumptions about
whether a dosage form is a good candidate for sustained release or not.
t\)
Table 18 Effect of Acetone Vapor Pressure on the Release Rate of Drug from 100~mg Sodium ~
Pentobarbital Tablets at Three Different Temperatures
Acetone vapor Acetone
k x 101 t 1/ 2 pressure absorbed
Temperature (mmHg) (mg/tablet) e: L (g cm2 sec- 1/ 2) (hr)
Untreated 0 0.575 9.24 5.38 3.38
tablets
37° 217 16.2 0.567 11.5 4.93 4.03
244 19.0 0.559 13.3 4.64 4.54
267 25.3 0.559 19.7 3.87 6.52
307 30.7 0.535 52.2 2.44 16.4
347 44.9 0.530 154.9 1.43 47.8
34° 217 18.4 0.560 12.2 4.79 4.26
244 22.5 0.51 16.6 4.19 5.57
267 27.9 0.539 43.2 2.67 13.7
307 39.7 0.534 198.2 1. 26 61.5
(")
4.01 6.08
;:3"
31° 217 26.1 0.551 18.6 Q
;:::l
244 29.3 0.540 36.4 2.91 11.5 tQ
Q
;:::l
267 32.8 0.528 86.2 1.93 26.2 Q,
::tl
0 Source: From Ref. 277, used with permission. t:r S·
~
;::I
Sustained Drug Release from Tablets and Particles 275
Individual absorption data that results in a first-order plot can often appear
to be zero-order, indicating good sustained release, when such data
are averaged. Thus, it is necessary to look at the data for each individual
in the test population when assessing sustained-release characteristics.
Nicholson et al. [296] desertbed the blood and urine levels obtained in
humans following ingestion of sustained-release tablets. It was noted that
the sustained-release tablets gave less variation between maximum and minimum
blood level concentrations and more uniform urinary excretion rates
than conventional release tablets. Theophylline aminoisobutanol, administered
in a tablet containing a sustained-release core having a matrix of
hydrophilic gums, a delayed barrier coat on the core, and an outer coat
containing the drug for immediate release was investigated by Kaplan [297].
Both in vitro and in vivo results were obtained and the blood level data
correlated well. Although differences were noted in the urinary data, no
explanation was tendered other than noting that these differences have
been reported previously. The in vivo absorption and excretion of radioactively
tagged [l0-14Clpentylenetetrazol was studied by Ebert et al.
r278) • Human volunteers were given either a single dose of sustainedrelease
insoluble matrix tablets or three divided doses of conventional tablets.
It was shown that the sustained-release form gave absorption and
excretion patterns similar to those obtained in the divided dose case.
As noted earlier in the section on coated granules, the earliest commercially
available sustained-release preparations available for clinical trials
seem to be those providing antihistamines [166,167,298,299]. Such research
has led to clinical trials of sustained-release triethanolamine trinitrate
[300] for use in angina pectoris. The drug was provided in a
plastic matrix that leached out drug over a 7- to B-h span. Patients reported
no undesirable side effects and the frequency and severity of attacks
were diminished in 80% of those tested.
De Ritter [301] used urinary excretion rates to evaluate nicotinic alcohol
tartrate administered to humans via Roche's Roniacol Timespan matrix
tablets. This dosage form uses a coating containing drug to provide
the immediate release portion of the dose and the sustaining portion is
provided by erosion and/or leaching of the insoluble matrix in the gastrointestinal
tract. The results indicate good correlation of in vitro and
in vivo release rates, and the drug is as completely available in this form
as in conventional tablets. In addition, the intense flushing caused by
non-sustained-release tablets is absent with the sustained release form.
The tension-relieving and sedative properties of pentobarbital sodium,
administered via conventional capsules and Abbott's Gradumet plastic
matrix sustained-release form, were clinically evaluated by Cass and
Frederik [302]. Although the grading system employed was qualitative in
nature, the results indicated that the Gradumet form was useful in providing
daytime tranquiJization and it was virtually free of untoward side
effects.
The Gradumet matrix dosage form has also been evaluated for treatment
of iron deficiency anemia via the sustained release of ferrous sulfate
[303,304]. In all cases, the hematocrit and hemoglobin responses were
virtually identical whether the non-sustained-release form or the matrix
tablet was administered. However, the incidence of reported gastrointestinal
upset due to ferrous sulfate, an important problem because more
iron is absorbed by fasting patients while at the same time causing more
gastrointestinal distress, was greatly diminished when the dose was given
276 Chang and Robinson
as a sustained-release tablet. Crosland-Taylor and Keeling [305] formulated
their own sustained-release ferrous sulfate tablets in an inert
polymer matrix and they added [59pe] SO4 as a marker to aid in evaluation.
They point out that the previoualy mentioned hematocrit and hemoglobin
responses are relatively insensitive means for comparing conventional and
sustained-release tablets, and their results indicate variable absorption
from sustained-release forms. They found no real advantage, even in the
area of management of side effects, to the use of sustained-release forms
of ferrous sulfate. It is not clear, then, whether there is any significant
advantage to providing oral hematinics in sustained-release form.
Owing to the many unpleasant side effects of oral potassium supplementation,
such as extremely salty taste and severe gastrointestinal upset,
the past few years have seen a need for a slow-release potassium-providing
product. Slow-K, marketed by Ciba, is the best example of this type of
product. The sugar-coated wax matrix contains 600 mg of potassium chloride
that leaches out gradually over a 4- to 6-h period. Tarpley [306)
studied the patient acceptability of this product as well as its ability to
msintain normal serum potassium levels in clinical trials comparing it to oral
potassium liquid preparations. His findings indicate that both types of
products maintain serum K+ levels, but the sustained-release tablet gives
much less incidence of nausea, abdominal pain. cramps, diarrhea, etc.,
and, of course, since it has little or no taste before being swallowed, it
was much more palatable and acceptable to all the patients involved.
As was noted earlier, aspirin has been tested for delivery via every
new dosage form developed for the past 20-25 years. Here, as in the
case of oral antihistaminics, the simple arithmetic of profit and loss statements
has compelled pharmaceutical firms to develop sustained-release
aspirin preparations [307]. The market is largely due to the vast numbers
of arthritics who need a preparation that will provide aspirin throughout
the night and eliminate morning stiffness, the most common plight of
arthritis sufferers. Many products have been extensively tested in clinical
trials as well.
Wiseman and Federici [308] have developed a sustained-release aspirin
tablet utilizing the matrix principle. By carefully monitoring both in vitro
and in vivo data, the authors were able to fashion a product that gave
constant plasma salicylate concentrations on chronic administration. Wiseman
[309] then extended the tests to the clinic where highly reproducible,
stable plasma salicylate concentrations were attained that overcame the
fluctuations due to multiple dosing of conventional tablets. The stable
levels do not exhibit the peaks in serum salicylate levels shown by conventional
tablets; therefore the incidence of gastrointestinal upset was reduced
and the valleys, or low points, were eliminated, thus providing therapeutic
levels of salicylate throughout the night. Harris and Regalado [310], however,
compared conventional and sustained-release aspirin for their ability
to provide relief to patients in order that the latter might perform simple
tasks requiring phalangeal dexterity. They reported no difference between
the two types of tablets, but a majority of the patients preferred the sustained-
release form. The -inference here is that the sustained-release
product gave relief throughout the night and did not cause the unpleasant
gastrointestinal side effects, thus contributing to the patient's preference.
Also, the convenience of eliminating frequent daytime doses was significant
in the preference for sustained-release aspirin.
Sustained Drug Release from Tablets and Particles 277
These are just a few examples of the benefits obtained from providing
aspirin in a sustained-release matrix tablet and of sustained-release dosage
forms for administering aspirin. in general. The recent flourish of timed
release and double- or triple-strength aspirin tablets to the market attests
to the desirability and. consequently. the profitability of long-acting
aspirin preparations.
Case Study of Combination Dissolution -Diffusion
Coating [280]
PROLONGED ACTION MECHAN ISM. Coated or uncoated plastic matrix
tablets. Solvent dissolves the coat containing the initial dose and drug is
leached out of the plastic core to maintain therapeutic levels. A diagrammatic
structure of the tablet is shown in Figure 13 for the drug combination
nitroglycerin and proxyphylllne.
Complete Depot Dosage Form
Structure of Depot Phase
Initial Phase
Drug Liberation < 5 min
Depot Phase
Slow Drug Liberation 1
Proxyphylline k
r
Nitroglycerin kl
r
0.40 hr- l
0.092 hr- l
~ Pla'ti, Pa,ti,l"
.:..: "..... .: :.. Nit,091y,,"n Df s so l ved in Pl as t ic
-: ,;: :', " Proxyphylline
Drug Release from Depot Phase
Nitroglycerin Diffuses
f rom P1astic tla t r i x
Release: q5~ in 5 hr
57'X in 8 hr
Proxyphylline Dissolves
and Diffuses through Pores
Release: 95t in 5 hr
100'1 in B hr
Figure 13 Diagrammatic structure of the peroral sandwich tablet with
timed release and its release Characteristics. (From Ref. 280. used with
permission. )
278 Chang and Robinson
Table 19 Pharmacokinetics of Proxyphylline in a Timed-Release Tablet into
Which Nitroglycerin Was to Be Incorporated
Biological half-life
Absorption rate constant
Elimination rate constant
Time to reach peak
Therapeutic concentration to maintain
for 12 h
Single dose producing desired blood
level
Liberation constant from depot phase
Equation for plasma concentration
Percent absorbed relative to the
amount ultimately absorbed
Maintenance dose
Initial do se
t 1/ 2 == 4.3 (h)
-1
Ka = 1.3 (h )
K
el = 0.163 (h -1)
T ::= 2.5 (h)
P
B
D ::= 0.8 (mg/100 L)
DB ::= 0.48 (g)
k 1 = 0.4 (h-1)
r
C ::: 11.5·e- 0• I 53t -12.5.e-1. 3t
AT A' 100 == CT + KelTCDT
Percent after 0.25 h == 35.6
0.5 h::: 45.5
1. 0 h::: 74.4
1.5 h == 100.0
1
D
1 ,
::= DB - D •(k T) = O. 177 (g)
M r p
Total dose per tablet
KelB D
W ::: D - D '(k IT ) + =: 0.489 (g)
B M r p k 1
r
Source: From Ref. 280, used with permission.
Sustained Drug Release from Tablets and Particles 279
EXPERIMENTAL DESIGN. The pertinent design variables for this
dosage form are shown in Table 19. Nitroglycerin was dissolved in 1: 1
alcoholI acetone solvent and was then impregnated in plastic granules of
either polyvinylchloride. ethylcellulose, polyamide, polyvinylacetate. or
polyacrylate. The plastic granules must be partly soluble in the alcohol/
acetone solvent. The solvent was evaporated and the plastic mass containing
nitroglycerin with screened to particle sizes of 0.5 -1.0 mm, These
particles were then mixed with proxyphylline and compressed into tablets.
The final two-layered tablet contained 0.2 mg nitroglycerin and 180 mg
proxyphylline in the immediate-release coat and 5 mg of nitroglycerin, in a
solid-solid solution matrix with the plastic material, and 310 mg of proxyphylline
in the matrix: pores, constituted the sustaining section.
IN VITRO TEST. The results of in vitro testing of the tablet are
shown in Figure 14. The solid and open circles indicate that good sustainment
is achieved with this tablet and that if accidental mastication
occurs. as illustrated by the other lines in the figure, some of the sustaining
effect is lost. Proxyphylline was released with an apparent first-order
rate of k r 1 = 0.40 h-! and nitroglycerin had k r
1 =0.092 h- 1. The data
indicate that the proxyphylline, incorporated into the pores of the matrix,
dissolves as soon as the artificial gastrointestinal fluid enters the pores
resulting 95% release within 5 h. Nitroglycerin, because it is dissolved
into the plastic matrix as a solid -solid solution, must diffuse through the
plastic material into the artificial gastrointestinal fluid. Consequently, its
release is much slower, resulting in only about 45% being released within
5 h.
8 7 6 5 :3 4
HOURS
2
____Q /0- a 6 0_0
., ,......., 1).--0 _0---
~C~~
~n~ ~O
M~ o~ u/ ___e
____8
/
0 ~8.
8 __8 • ::;::::;"
o /" ---%---t--e
/'e .:::::.- .:::::::-.__
8/' .~.---.
..........-:.---.~. 8/. ./ .---.--- .../
100
~ 90
0a: 80 t:
> 70 zz
60 0
j:: «a:
w
2:
...J
e
;:)
a:
0
Figure 14 In vitro drug liberation. Proxyphylline: (0) intact tablet,
(A) cut into two parts, (lJ) cut into four parts, (Q) powdered. Nitroglycerin:
(.) intact tablets, (~) cut into two parts, (.) cut into four
parts, (e) powdered, (From Ref. 280, used with permission.)
280 Chang and RObinson
2 3 456
HOURS
7 8 9
Figure 15 Nitroglycerin blood level after peroral administration. ( • )
intact tablet, ( ... ) cut into two parts, (.) cut into four parts. and (0)
powdered and masticated. (From Ref. 280. used with permission.)
Although the in vitro release study showed that in divided tablets the
prolonged activity was essentially lost for proxyphylline. the nitroglycerin
still maintained its sustaining activity.
IN VIVO TEST. Figure 15 shows the results of the in vivo testing
for sustained release of nitroglycerin in these products. These are the
mean curves from three human subjects. As stated by the author, three
SUbjects is too small a population to generate definitive statistics, so that
this study should be repeated with more subjects. The results did indicate
good sustained release of nitroglycerin and served as a useful experimental
guide.
D. Sustained Release Utilizing Osmosis
An example of sustained release through osmosis is tbe so-called osmotic
tablet [11,12.311]. The coating in this case is just a semipermeable membrane
that allows penetration of water. but not drug, to dissolve its contents.
The dissolved drug plus diIuents establishes an osmotic pressure
and forces drug solution to be pumped out of a small hole in the tablet
coating. The rate of this drug pumping can be controlled through core
composition, coating material. and delivery orifice. A pictorial representation
of the osmotic tablet is shown in Figure 16.
The tablet imbibes fluid through its semipermeable membrane at a constant
rate determined by membrane permeability and by osmotic pressure of
the core formulation. With the system at a constant internal volume, the
SUstained Drug Release from Tablets and Particles
SEMI PERMEABLE
MEMBRANE
281
Figure 16 Elementary osmotic pump cross section. (From Ref. 12, used
with permission.)
tablet wID delivery, in any time interval, a volume of saturated solution
equal to the volume of solvent uptake. The delivery rate is constant as
long as an excess of solid is present inside the device, declining parabolically
toward zero once the concentration falls below saturation, as is
shown in Figure 17.
The principles upon which this device is based were presented earlier
in this chapter. A key factor in proper drug delivery is the size of the
40
...:c 30
E
L1I
~
cr: 20
>- cr:
L1I >
:J
~ 10
20
'z (HOURS)
Figure 17 In vitro release rate of potassium chloride from elementary
osmotic pump in water at 37°C. (I) range of experimental data obtained
from five systems, ( ) calculated release rate. (From Ref. 12,
used with permission.)
282
30 - ...
:c
g'
~ 20
~
a:
>-
~ 10
>
:J
wa
Chang and Robinson
5
HOURS
10
Figu re 18 In vitro and in vivo release rate of potassium chloride from
elementary osmotic pumps. ( ) average in vitro rate from systems
of the same batch; and (6). (0). (e) average release rate in one system
jn the GI tract of dogs 1, 2, and 3, plotted at the total time period
each system resided in the dog. (From Ref. 12, used with permission.)
delivery orifice. Two conditions must be met. relative to the size of the
delivery orifice, in order for it to be successful.
1. It must be smaller than a maximum size, Amax, to minimize the
contribution to the delivery rate made by solute diffusion through
the orifice.
2. It must be sufficiently large, above a minimum size, Amin. to minimize
hydrostatic pressure inside the system that would affect the
zero-order release rate.
Too small a hole will depress the delivery rate below that of the desired
constant delivery.
In vivo tests in three dogs are shown in Figure 18. That zero-order
release is maintained for long periods is evident.
It should be quite obvious from the mechanistic description of this
apparatus that drug delivery would be independent of stirring rate and
independent of pH. These are sizable advantages to a sustained release
system and have indeed been shown operable both in vitro and in vivo.
v, COATING MATERIALS
Other sections of this text have dealt with coating substances insofar as
describing their properties, coating technology, quality control, etc.
Some of these coating materials. for conventional coating purposes, can be
employed to produce sustained-release products depending on the thickness
of coat employed as well as addition of fillers to the coating substance.
Table 20 lists some of the more common coating substances and
their properties. This listing is not intended to be comprehensive, but
Table 20 Coating Materials and Their Properties
Type of coating
Barrier coating
(includes microencapsulation)
Most suitable
dosage forme s)
1. Film-coated tablets
2. Film-coated pellets
or granules placed
in gelatin capsules
3. Compressed tablets
containing mixtures
of barrier-coated
particles with
filler particles
4. Compressed tablets
containing only
barrier-coated particles
forming a
matrix
Examples
Various shellacs [18]
Beeswax [18]
Glyceryl monostearate [18]
Nylon [18]
Acrylic resins [18]
Cellulose acetate butyrate
[ 20]
dl-Polylactic acid [20]
1,6-Hexanediamine (20]
Diethylenetriamine [20)
Polyvinylchloride [20]
Sodium carboxymethylcellulose
{2431
Various starches 1244]
Polyvinylpyrrolidone [245]
Acetylated monoglycerides
[245J
Gelatin coacervates [211]
Styrene/maleic acid copolymer
[245]
1.
2.
3.
Probable release
mechanisms
Diffusion and
dialysis
Some disintegration
possible
Also have had pHdependent
dissolution
and some enzymatic
breakdown incorporated
into some films,
but these are, therefore,
poor "barriers"
1.
2.
3.
4.
Properties
Slow or incomplete
release
Coating is subject
to fracture during
compression
Release depends on
solubility of the
drug and pore
structure of the
membrane
Obtain constant release
when water or
GI fluids pass
through barrier to
dissolve drug and
form a saturated
solu tion within the
tablet
Table 20 (Continued)
Most suitable Probable release
Type of coating dosage formes) Examples mechanisms Properties
Barrier coating Gelatin coacervates [881
(includes micro- Styrene/maleic acid 00- encapsulation) polymer [124]
(cont)
Embedment into 1- Compressed granules Glycerol palmitostearate 1. Gradual erosion of 1. Slow or incomplete
a fatty coating into a tablet [18J the coat. aided by release
(similar to em- 2. Compressed granules Beeswax [18] pH and en zymatic 2. Difficult to control
bedding in a placed in a gelatin
hydrolysis of the
release pattern due
matrix: of fatty capsule
Glycowax [181 fatty acid esters to variations in pH
materials) Castor wax [18] 2. Coating may contain and enzyme content
3. Multilayered tablets
portion of the dose of the GI tract
4. Compression-eoated
Carnauba wax {18]
for quick release
tablets Glyceryl monostearate [18] upon hydrolysis with
Stearyl alcohol [18] subsequent slow release
from erosion of
core
Repeat action 1- Sugar coating of an Cellulose acetate phthalate 1. pH-Dependent disso- 1. Variations due to
coatings enteric-coated core [l8] lution and enzymatic changing stomach
tablet (Many of the examples breakdown emptying times
2. Compression coating listed for "Barrier Coating" 2. Outel' coating re- 2. Not a "true" susof
an enteric-coated apply here also) leases first dose tamed form as decore
tablet rapidly in stomach fined in text
fluids. inner entericcoated
core releases
a second dose at
some later time in
intestinal fluids
Coated plastic 1- Multilayer tablets Polyethylene [18] 1. Outer coating con- I. Slow or incomplete
matrix 2. Compression-coated Polyvinylacetate [18] taining active drug release
tablets
dissolves rapidly to
2. Only water-soluble
Polymethacrylate [18] provide drug for im- or fairly water-
Polyvinylchloride [18]
mediate absorption soluble drugs can
Ethylcellulose [18] 2. Above process is be used
followed by leaching
3. Plastic matrix skel- Silicone devices [20] of drug from inert eton is excreted in
Methylmethacrylate [20]
matrix via penetraits
original. shape
tion of 01 fluids into
Ethylacrylate [20] pores of the matrix
in the feces
2-Hydroxyethylmetha-
4. Drug liberation depends
only on solucrylate
[20] bility in 01 fluids,
1.3-Butyleneglycoldimetha- completely indecrylate
[20] pendent of pH,
Ethyleneglycoldimethaenzyme
activity,
concentration. or
crylate I 20]
01 motility
Coated hydro- 1. Multilayer tablets Carboxymethylcellulose 1. Outer coating con- 1. Drug liberation rate
philic matrix
2. Compression-coated U8l taining active drug is dependent on
dissolves rapidly type and amount of
tablets Sodium carboxymethyl- along with rapid dis- gum used
cellulose [18]
solution from the
Hydroxypropylmethylcellu- surface of the 2. High water solubflfty
lose [18] matrix to provide of the drug is abdrug
for immediate
solutely necessary
Methacrylate hydrogels [19]
absorption
Polyethyleneglycols [106]
Table 20 (Continued)
Type of coating
Coated hydrophilic
matrix
(cont)
Most suitable
dosage form(s) Examples
Probable release
mechanisms
2. Above process continues
until a viscous
gelatinous barrier
is formed
around the matrix
surface
3. Once the gelatinous
barrier has formed.
diffusion and dissolution,
via erosion.
occur at a slow. controlled
rate
Properties
3. Release is controlled
by drug diffusivity
more than by gum
dissolution or water
penetrability as long
as the hydrated
gelatinous layer remains
intact
Sustained Drug Release from Tablets and Particles
rather is a starting point for the formulator interested in preparing a
su stained-release product through coating.
VI. SUMMARY
287
During the past several years we have witnessed an increased number of
publtcatlons dealing with polymer coatings as a means of sustained drug
action. Most of these approaches have been aimed at the parenteral Or
specialty areas and as yet many have not been brought to the market
stage, but their potential utility is apparent. Our ability to manipulate
film properties gives us a powerful tool by which to prolong drug delivery
and produce variable drug release rates. It seems reasonable to conclude
that we are limited in attempting to produce variable release rates from
conventional tablets or pellets if we rely strictly on dissolution rates and
shape factors, but the addition of polymer coatings with variable properties
gives us considerable latitude in producing sustained-release products
with varying release rates.
REFERENCES
1. Lipowsky, I., Australian Patent 109,438, Filed November 22, 1938.
2. Williams, A•• Sustained Release Pharmaceuticals. Noyes Development
Corp., Park Ridge, New Jersey, 1969.
3. Eriksen. S., Sustained action dosage forms. In The Theory and
Practice of Industrial Pharmacy, 1st Ed. (L. Lach man, H. A.
Lieberman, and J. L. Kanig, eds.), Lea & Febiger, Philadelphia,
1970.
4. Grief. M. and Eisen. H•• Prolonged action oral medication. Am.
Prof. Pharmacist, 25: 93 (1959).
5. Hanselmann, B. and Voight, R.• Wirkungsverlangerung von
arzneimitteln unter besonderer berucksichtigung der operoralen
depot-arzneiform. Piiarmazie , 26:57 (1971).
6. Lazarus. J. and Cooper. J.. Absorption, testing. and clinical
evaluation of oral prolonged action drugs. J. Pharm. Set. 50: 715
(1961) •
7. Lazarus. J. and Cooper, J., Oral prolonged action medicaments:
Their pharmaceutical control and therapeutic aspects. J. Pharmacol.,
11: 257 (1959).
8. Robinson. M. J •• Making sustained-action oral dosage forms. Drug
Cos. Ind .• 87: 466 (1960).
9. Stempel. E.• Patents for prolonged action dosage forms. 1: General
products. Drug Cos. Ind•• 98:44 (1966); 2: Tablets. Drug
Cos. Ind., 98:36 (1966).
10. Parrott. E. L.• Pharmaceutical Technology: Fundamental Pharmaceutics.
Burgess. Minneapolis, 1970. pp , 100-103.
11. Theeuwes, F. and Higuchi. T., U.S. Patent 3,845.770 (1974) (assigned
to Alza Corp., Palo Alto. California).
12. Theeuwes, F.. Elementary osmotic pump. J. Piuirm, sci., 64: 1987
(1975) •
288 Chang and Robinson
13. Theeuwes J F., Ashida, K., and Higuchi, T., Programmed diffu sional
release rates from encapsulated cosolvent system. J. Pharm. Sci.,
65: 648 (1976).
14. Cooper, J. and Rees, J. E., Tableting research and technology.
J. Pharm. Sci., 61: 1511 (1972).
15. Sjogren, J., Studies on a sustained release principle based on an
inert plastic matrix. Acta Pharm. suecica, 8: 153 (1971).
16. Nelson, E. J Pharmaceuticals for prolonged action. CUn. Pharmacal.
Ther., 4:283 (1963).
17. Edkins, R. P., The modification of the duration of drug action.
J. Pharm. Pharmacol., 11: 54T (1959).
18. Ritschel, W. A., Peroral solid dosage forms with prolonged action.
In Drug Design, Vol. IV (E. J. Ariens, ed.), Academic Press,
New York, 1973, pp. 37-73.
19. Andrade, J. D., Hydrogels for Medical and Related Applications.
ACS Symposium Series 31, American Chemical Society, Washington,
D.C .. 1976.
20. Paul, D. R. and Harris, F. W., Controlled Release Polymeric Formulations.
ACS Symposium Series 33, American Chemical Society.
Washington, D.C., 1976.
21. Robinson, J. R., Controlled release pharmaceutical systems. In
Symposium on Economics and Market Opportunities for Controlled
Release Products. 172nd National Meeting of the American Chemical
Society, San Francisco, August 29-Sept. 3, 1976, pp. 210-226.
22. Hall, H. S. and Hinkes, T. M., Air suspension encapsulation of
moisture-sensitive particles using aqueous systems. In Microencapsulation:
Processes and Applications (J. E. Vandagaer, ed.),
Plenum, New York, 1974.
23. Lehmann, K., Acrylic resin coatings for drugs: Relations between
their chemical structure, properties and application possibilities.
Drugs Made in Germany, 11( 34) : 145 -153 (1968).
24. Lehamnn , K. and Dreher, D., Permeable acrylic resin coatings for
the manufacture of depot preparations of drugs. 2: Coatings of
granules and pellets and preparation of matrix tablets. Drugs Made
in Germany, 12:59 (1969).
25. Hall, H., personal communication.
26. Kwan, K. C., Pharmacokinetic considerations in the design and evaluation
of sustained and controlled release drug delivery systems.
In Sustained and Controlled Release Drug Delivery Systems (J. R.
Robinson, ed.), Marcel Dekker, New York, 1978.
27. Levy, G., Pharmacokinetic aspects of controlled drug delivery systems.
In Temporal Aspects of Therapeutics (J. Urquhart and F. E.
Yates, eds.), Vol. 2 of Alza Conference Series, Plenum, New York,
1973, pp. 107-127.
28. Soeldner, J. S., Chang, K. W., Aisenberg, S., and Hiebert, J. M.,
Progress towards an implantable glucose sensor and an artificial beta
cell. In Temporal Aspects of Therapeutics (J. Urquhart and F. E.
Yates, eda.}, Vol. 2 of Alza Conference Series, Plenum, New York,
1973, pp , 181-207.
29. Robinson. J. R. and Eriksen, S. P., Theoretical formulation of sustained-
release dosage forms. J. Pharm. Sci., 55: 1254 (1966).
Sustained Drug Release from Tablets and Particles 289
30. Fincher, J. B., Particle size of drugs and its relationship to absorption
and activity. J. Pharm. Sci., 57: 1825 (1968).
31. Lee, V. B. and Robinson, J. R., Drug properties influencing the
design of sustained Or controlled release drug delivery systems.
In Sustained and Controlled Release Drug Delivery Systems (J. R.
Robinson, ed.), Marcel Dekker, New York, 1978.
32. Hansch, C. and Dunn, W. J., Linear relationships between lipophilic
character and biological activity of drugs. J. Pharm. Sci•• 61: 1
(1972) •
33. Fujita, T., Iwasa, J., and Hansch, C., A new substituent constant,
derived from partition coefficients. J. Am. Chem. Soc.. 86: 5175
(1964) •
34. Beerman, B., Hellstrom. K., and Rosen, A., Metabolism of propantheline
in man. Clin, Pharmacol. Ther., 13: 212 (1972).
35. Niebergall, P. J., Sugita, E. T., and Schnaare, R. L., Potential
dangers of common drug dosing regimen. Am. J. Hasp. Pharm•• 31:
53 (1974).
36. Brooke, D. and Washkuhn, R. J •• Zero-order drug delivery system:
Theory and preliminary testing. J. Pharm. Sci.. 66: 159 (1977).
37. Lipper, R. A. and Higuchi, W. I., Analysis of theoretical behavior
of a proposed zero-order drug delivery system. J. Pharm. Sci.,
66: 163 (1977).
38. Cobby, J., Mayersohn, M•• and Walker, G. C., Influence of shape
factors on kinetics of drug release from matrix tablets. I: Theoretical.
J. Pharm. sa., 63: 725 (1974).
39. Green. B. K., U.S. Patent 2,712,507 (1955), Chem. Abstr. 50:12479a
(1956) •
40. British Patent 490,001 (1938) through Chem. Abstr•• 33:814 (1939).
41. Chang, T. M. S., Science. 146: 524 (1964).
42. Chang. T. M. S., Science J., 3:63 (1967).
43. Deasy, P. B., Microencapsulation and Related Drug Processes,
Marcel Dekker, New York. 1984, p , 120.
44. Asker, A. F. and Becker, C. H., J. Pharm. Sci•• 55:90-94 (1966).
45. Takenaka, H., Kawashima. Y., and Lin, S. Y., J. Pharm. sci., 69:
1388 - 1392 (1980).
46. Cusimano, A. G. and Becker. C. H., J. Ptuirm, Sci., 5:1104-1112
(1968).
47. Hamid, I. S. and Becker. C. H., J. Pharm. Sci., 59:511-514 (1970).
48. Pederson, A. M. et al., U. S. Patent 4,572,833, Feb. 25 (1986).
49. Colombo. P •• Conte, U., Caramella, C., Gaz zaniga , A., and
LaManna, A•• J. Controlled Release, 1: 283 - 289 (1985).
50. U.S. Patent 3,400.185 (Sept. 3, 1968).
51. Jager, K. F. and Bauer, K. H., Drugs Made in Germany. 15:61-65
(1982) •
52. Kawashima, Y., Ohno , H., and Takenaka, H., J. Pharm. Sci•• 70:
913 -916 (1981).
53. U.S. Patent 4,401,456 (Aug. 30, 1983).
54. Roberts, A. G. and Shah, K. D., cnem. Eng., 12:748-750 (1975).
55. Ghebre-Sellassie, I., Gordon, R. H., Fawzi, M. B., and Nesbitt,
R. U., Drug Dev, Ind. Pnarm.; 11: 1523-1541 (1985).
56. Caldwell, H. C. and Rosen, E., J. Pharm. Set, 53:1387-1391 (1964).
290 Chang and Robinson
57. U.S. Patent 4,083,949 (1978).
58. Handbook of Avicel Spheres, FMC Corp , , Philadelphia, (1982).
59. Wan, L. S. L. and Jeyabalan, T., Chern. Ptuirtn, suu., 33:54495457
(1985).
60. Conine, J. W. and Hadley, H. R., Drug Cosmet, Ind., 106:38-44
(1970) •
61. Jalal, 1. M., Malinowski, H. J., and Smith, W. E., J. Pharm, Sci.,
6: 1466 (1972).
62. Woodruff, C. W. and Nuessle, N. 0., J. Pharm. sct., 61: 787-790
(1972) •
63. Malinowski, H. J. and Smith, W. E., J. Ph arm• sct., 63: 285 - 288
(1974).
64. O'Connor, R. E. and Schwartz, J. B., Drug Dev ; Ind. Pharm., 11:
1837 -1857 (1985).
65. Chien, T. Y. and Nuessle, N. 0., Pharm, Tech., 9(4):42-48
(1985).
66. Malinowski. H. J. and Smith, W. E., J. Pharm. Sci.. 64: 1688 -1692
(1975) •
67. Scott. M. W., Robinson, M. J •• Pauls, J. F •• and Loritz, R. J.,
J. Pharm. Sci., 53:670 (1964).
68. Osterwald, H. P •• Pharmaceut. Res., 2: 14 -18 (1985).
69. Pharmaceu tical Glazes Product Information Bulletin, Somerset, New
Jersey, William Zinsser & Company.
70. Stafford, J. W., Drug Dev; Ind. Pnarm, , 8:513-530 (1982).
71. Bagar-ia , S. C. and Lordi, N. G.• Aqueous Dispersions of Waxes and
Lipids for Pharmaceutical Coating, Paper presented at the AAPS
Second National Meeting, Boston, Massachusetts, (1987).
72. Bauer, K. H. and Osterwald, H., Pnarm, Ind., 41: 1203 -1209
(1979) .
73. Pondell, R. E., Drug Dev, Ind. Pharm., 10:191-202 (1984).
74. Banker, G. S. and Peck, G. E., Pharm. Technol., 5(4):55-61
(1981) •
75. Ghebre-Sellassie, I., Banker, G. S., and Peck, G. E., WaterBased
Controlled-Release Drug Delivery Systems, Pharm. Tech.
Conference 1982 Proceedings (Aster PUblishing Corp , , Springfield,
Oregon, 1982), pp. 234-241.
76. Goodhart, F. W., Harris, M. R., and Nesbitt, R. W., Pharm.
Technol. 8( 4) : 64-71 (1984).
77. Lehman, K., Acta Pnarm, Fenn., 81( 2) : 55 -74 (1984).
78. Lehman, K., Acta Pharm. Fenn, , 91( 4) : 225 - 238 (1982).
79. Lehman, K. and Dreher, D., Int. J. Piuirm, Tech. Prod. Mfr., 2( 4):
31- 43 (1984).
80. Chang, R. K., Hsiao, C. H., and Robinson, J. R., Pharm. Technol.,
11(3) :56-68 (1987).
81. Gurny, R., Pharm. Acta Helv., 56(4-5):130-132 (1981).
82. Gurny, R., Peppas , N. A., Harrington, D. D•• and Banker, G. S.,
Drug Dev, Ind. Pharm., 7(1): 1-25 (1981).
83. Willis, C. R. and Banker, G. S., J. Ptuirtn, Sci., 57:1598-1603
(1968) .
84. Goodman, H. and Banker. G. S., J. Pharm. sa., 59:1131-1137
(1970) •
Sustained Drug Release from Tablets and Particles 291
85. Rhodes, C. T., Wai, K., and Banker, G. S., J. Pharm. sa., 59:
1578 -1581 (1970).
86. Rhodes, C. T., Wai, K., and Banker. G. S., J. Pharm. Sci., 59:
1581-1584 (1970).
87. Boylan, J. C. and Banker, G. S., J. Pharm. sci., 62:1177-1183
(1973) •
88. Larson, A. B. and Banker, G. S., J. Pharm. Sci, , 65:838-843
(1976) •
89. Garden, J. L., High Polymers, 29:143-197 (1977).
90. Woods, M. E., Dodge, J. S., Krieger, 1. M., and Pierce, P. Eo;
J. Paint. Technol., 40( 527) : 541- 548 (1968).
91. Hansmann, J., Adhaesian, 20( 10): 272- 275 (1976).
92. Vanderhoff, J. W. et al., U.S. Patent 4,177,177 (Dec. 4, 1979).
93. Sanders, F. L. et aI., U.S. Patent 3,652,676 (Feb. 15, 1972).
94. Burke, O. W., Jr., U.S. Patent 3,652,482 (1972).
95. Miller, A. L. et al., U.S. Patent 3,022,260 (1962).
96. Aelony, D. et al., U.S. Patent 2,899,397 (1959).
97. Burton, G. W. and O'Farrell, C. P., Rubber Chem. Technol., 49(2):
394 (1976).
98. Chang, R. K., Price, J. C., and Whitworth, C. W., Ph arm.
Technol., 10( 10) : 24 - 33 (1986).
99. Chang, R. K., Price, J. C., and Whitworth, C. W., Drug Dev; Ind.
Pharm., 13:1119-1135 (1987).
100. Chang, R. K•• Price, J. C., and Whitworth, C. W., Drug Dev; Ind.
Pharm., 12: 2335, 2380 (1986).
101. Cooper, W.• U.S. Patent 3,009,891 (1961).
102. Date, M. et al , , Jpn. Patent 7306,619 (1973).
103. Warner, G. L. et al., U.S. Patent 4,123,403 (Oct. 31, 1978).
104. Leng, D. E. et al., U.S. Patent 4,502,888 (Mar. 5, 1985).
105. Judd, P., Brit. Patent 1,142,375 (1969).
106. Dieterich, Do; Keberle , W., and Wuest, R., J. Oil Col. Chern.
Assoc., 53(5):363-379 (1969).
107. Gurny, R., Bindschaedler, C., and Doelker, E., Recent Advances
in Latex Technology, Proceedings Intern. Symp. Control. ReI.
Bioact. Mater. 13, 1986 (Controlled Release Society, Ino , , Lincolnshire,
Illinois), pp , 130 -131.
108. Li, L. C. and Peck, G. E., Water Based Silicone Elastomer Dispersions
as Controlled Release Tablet Coating (1) -Preliminary
Evaluation of the SHicone Elastomer Dispersions, Proceedings
Intern. Symp. Control. ReI. Bioact. Mater., 14, 1987 (Controlled
Release Society, Ine , , Lincolnshire, illinois), pp , 148 -149.
109. Li, L. C. and Peck, G. E., Water Based Silfcone Elastomer Dispersions
as Controlled Release Tablet Coating (II) -Formulation
Considerations, Proceedings Intern. Symp. Control. Rel. Bioact.
Mater., 14, 1987 (Controlled Release Society, Ine , , Lincolnshire,
illinois), pp. 150-151.
110. Mehta, A. M. and Jones, D. M., Pharm. Technol., 9(6) :52-60
(1985) .
111. Sono-Tek Ultrasonic Nozzles, Technical Information Bulletin,
Poughkeepsi, New York, Sono-Tek Corporation, 1984.
112. Bindsschaedler , C., Gurny, R., and Doelker, E., Lobo-Ptiarma-:
Probl. Tech., 31/311:389-394 (1983).
292 Chang and Robinson
113. Standard Test Method for Minimum Film Formation Temperature of
Emulsion, American Society for Testing and Materials, Designation
D2354-81, pp. 407 -409 (l981).
114. Standard Test Method for Transition Temperature of Polymer by
Thermal Analysis, American Society for Testing and Materials,
Designation D3418-75, pp. 840 -843 (1976).
115. Modern Plastics Encyclopedia, McGraw-Hill, New York, 1984-1985,
pp , 635-644.
116. D.K. patent Application GB 086725A.
117. Conte, D., Colombo, P., Caramella, C., and LaManna, A., Farmaco
Ed. Pract., 39:67 (1984).
118. Salomon, J. L., Doelker, E., and Buri, P., Pharm. Irui, , 41:799
(1979) •
119. Bludbaugh, F., Zapapas, J., and Sparks, M., J. Am. Piuirm,
Assoc. (Sci. ea.i, 47, 12:857-870 (1958).
120. Decoursin, J. W., Conference Proceedings of the Latest Developments
in Drug Delivery Systems, Aster PUblishing Corp., 1985, pp.
29-32.
121. Sheth, P. R. and Tossounian, J., Drug Dev; Ind. Pharm., 10:
313 - 339 (1984).
122. Huber, H. F., Dale, L. B., and Christensen, G. L., J. Pharm.
sct., 55: 974 (1966).
123. Kaplan, L. L., J. Pharm. sa., 54: 457 (1965).
124. Kornblum, S. S., J. Pnarm, Sci.. 58:125 (1969).
125. Lapidus, H. and Lordi, N. G., J. Pharm. Sci•• 55:840 (1966).
126. Huber, H. F. and Christensen, G. L., J. Pharm. Sci., 57:164
(1968) .
127. Baun, D. C. and Walker. G. C.• Pharm. Acta Helv., 46:94 (1971).
128. Choalis, N. H. and Papadopoulos, H., J. Pharm. Sci., 64: 1033
(1975) .
129. Lapidus, H. and Lordi, N. G., J. Pharm. sct., 57:1272 (1968).
130. Nystrom, C., Mazur, J., and Sjorgren, J., Int. J. Pharm•• 10:
209- 218 (1982).
131. Touiton, E. and Donbrow, M., Int. J. Pharm., 11:131-148 (1982).
132. Daly, P. B .• Davis, S. S •• and Kennerley, J. W., Int. J. Pharm.,
18:201-205 (1984).
133. Burl, B. and Doelker, E., Pharm. Acta Hetv ; , 55(7-8):189-197
(1980) •
134. Handbook on Methocel Cellulose Ester Products, Dow Chemical
Company, Midland, Michigan, 1984.
135. Raghunathan, Y. and Becker, C. H., J. Pharm. sei., 57:1748
(1968) •
136. Cusimano, A. G. and Becker, C. H., J. Pharm. Sci., 57:1104
(1968) •
137. John, P. H. and Becker, C. H., J. Pharm. Sci" 57:584 (1968).
138. Robinson, M. J., Bondi, A., and Swintosky, J. V., J. Am. Pharm,
Assoc. Sci. sa., 47:874 (1958).
139. Kowarski, C. R., Volberger, B., Versanno, J., and Kowanski, A.,
Am. J. Hosp. Pharm., 21: 409 (1964).
140. Yamamoto. R. and Baba, M., Jpn. Patent 1849.1960, through
Chern. Abstr. 54.20100d, 1960.
Sustained Drug Release from Tablets and ParUcles 293
141. Draper, E. B. and Becker, C. H., J. Pharm. Sci, , 55:376 (1966).
142. Robinson, I. C. and Becker, C. H., J. Pharm. Sci., 57:49 (1968).
143. Lazarus, J., Pagliery, M., and Lachman, L., J. Pharm. Sci., 53:
798 (1964).
144. Sjogren, J. and Fryklof, L. E., Pharm, Rev., 59:171 (1960).
145. Rowe, R. C., Manuf. Chem, Aerosol News, 46(3):23 (1975).
146. Desai, S. J., Simonelli, A. P., and Higuchi, W. I., J. Pharm. se«,
54: 1459 (1965).
147. Farhadien, B., Borodkin, S., and BUddenhagen, J. D., J. Pharm.
Sic., 60:209 (1971).
148. Rowe, R. C., Elworthy, P. H., and Ganderton, D., J. Pharm.
Pharmacal., 24(Suppl.): 137P (1972).
149. Endicott, C. J., U.S. Patent 3,087,860 (1963).
150. Singh, P., Desai, S. J., Simonelli, A. P., and Higuchi, W. 1.,
J. Pharm. Sci., 56:1542 (1967).
151. Farhadien, B., Borodkin, S., and BUddenhagen, J. P., J. Pharm.
Sci., 60: 212 (1971).
152. Rowe, R. C., Elworthy, P. H., and Ganderton, D., J. Pharm.
Pharmacal•• 25(Suppl.): 112P (1973).
153. Rowe, R. C., Elworthy, P. H., and Ganderton, D., J. Pharm.
Pharmacal., 25(SuppL): 12P (1973).
154. Desai, S. J.. Singh, P., Simonelli, A. P., and Higuchi, W. I.,
J. Pnarm, sa., 55: 1235 (1966).
155. Chang, R. C. and Price, J. C., J. Biomater, Appl., 3( 1) :80
(1988) •
156. Jalsenjak, I., Nixon, J. R., Senjkovic, R., and Stivic, 1.,
J. Pharm. Pharmacal., 32:678 (1984).
157. Jalsenjak, I., Nicolaidon, C. F., and Nixon, J. R., J. Pharm.
Pharmacol., 29: 169 (1977).
158. Nixon, J. R., Jalsenjak, I., Nicolaidon, C. F., and Harris, M.,
Drug Dev. Ind. Pharm., 4: 117 (1978).
159. Nixon, J. R. and Agyilirah, G. A., J. Pharm, Sci., 73:52 (1984).
160. Chemtob, C., Chaumeil, J. C., and N'Dongo , M., Int. J. Pharm.,
29:83 (1986).
161. Hasagawa, A., Nakagawa, H., and Sugimoto, 1.. Yakugaku Zasshi,
104:889 (1984).
162. Walker, S. E., Ganley, J., and Eaves, T., Tableting Properties
of Potassium Chloride Microcapsules. In F.I.P, Abstracts, 37th
International Congress of Pharmaceutical Congress of Pharmaceutical
Sciences, The Hague, Sept. 1977, p. 31.
163. Abdel Monem Sayed, H. and Price, J. C., Drug Dev. Ind. Pharm.,
12: 577 (1986).
164. Colbert, J. C., Controlled action drug forms. In Chemtcal Technology
Review, No. 24. Noyes Data Corp , , Park Ridge, New
Jersey. 1974.
165. Johnson, J. C•• Tablet manufacture. In Chemical Technology
Review, No. 30, Noyes Data Corp., Park Ridge, New Jersey,
1974.
166. Green, M. A., One year's experience with sustained release antihistamine
medication. Ann. Allergy, 12: 273 (1954).
167. Rogers, H. L., Treatment of allergic conditions with sustained release
chlorpropenypyridamine maleate. Ann. Allergy, 12:266 (1954).
294 Chang and Robinson
168. Berkowitz. D.. The effect of a long-acting preparation (Spansule)
of belladonna alkaloids on gastric secretion of patients with peptide
ulcer. Gastroenterology. 80: 608 (1956).
169. Morrison, S., The use of a sustained release tranquilizer-anticholinergic
combination in patients with disturbed digestive function.
Am. J. Gastroenterol., 29:518 (1958).
170. Grahn. H. V., Antihypertensive effects of reserpine in sustainedrelease
form: A comparative study. J. Am. certatr. Soc., 6:671
(1958).
171. Burness, S. H., Clinical evaluation of a sustained release belladonna
preparation. Am. J. Dig. ou., 22:111 (1955).
172. Heimlich, K. R., MacDonnell, D. R., Flanagan, T. L., and O'Brien,
P. D.. Evaluation of a sustained release form of phenylpropanolamine
hydrochloride by urinary excretion studies. J. Pharm. Sci.,
50:232 (1961).
173. Heimlich, K. R., MacDonnell, D. R., Polk, A., and Flanagan.
T. L., Evaluation of an oral sustained release dosage form of
trimeprazine as measured by urinary excretion. J. Pharm. Sci•• 50:
213 (1961).
174. Vasconcellos, J. and Kurland, A. A., Use of sustained-release
chlorpromazine in the management of hospitalized chronic psychotic
patients. Dis. Nerv; Sys., 19:173 (1958).
175. Souder. J. C. and Ellenbogen, W. C., Laboratory control of dextroamphetamine
sulfate sustained release capsulea, Drug Stand., 26:
77 (1958).
176. Steigmann, F., Kaminski, L•• and Nasatir, S., Clinical-experimental
evaluation of a prolonged-acting antispasmodic-sedative. Am. J.
Dig. Dis., 4:534 (1959).
177. Mellinger. T. J., Serum concentrations of thioridazine after different
oral medication forms. Am. J. Psychiatry. 121: 1119 (1965).
178. Hollister, L. E., Studies of prolonged-aetion medication. II: Two
phenothiazine tranquilizers (thioridazine and chlorpromazine) administered
as coated tablets and prolonged action preparations.
Curro Ther. Res•• 4:471 (1962).
179. Mellinger. T. J.. Mellinger, E. M., and Smith, W. T.. Thioridazine
blood levels in patients receiving different oral forms. CUn.
Pharmacol. Ther., 6:486 (1965).
180. vestee, N. D. and Schiele, B. C•• An evaluation of slow-release
and regular thioridazine and two medication schedules. Curro Ther.
Res., 8:585 (1966).
181. Swintosky, J. V., Design of oral sustained-action dosage forms.
Drug Cosmet. Ind., 87: 464 (1960).
182. Blythe, R. H., The formulation and evaluation of sustained release
products. Drug Stand., 26: 1 (1958).
183. Wagner. J. G., Carpenter. O. S •• and Collins. E. J.. Sustained
action oral medication. I: A quantitative study of prednisolone in
man, in the dog and in vitro. J. Pharmacal. Exp. Ther•• 129:101
(1960) •
184. Nash, J. F. and Crabtree, R. E., Absorption of tritiated d -desoxyephedrine
in sustained-release dosage forms. J. Pharm. Sci•• 50:
134 (1961).
SUstained Drug Release from Tablets and Particles 295
185. Rosen, E. and Swintosky, J. V., Preparation of a 35s labeled trimeprazine
tartrate sustained action product for its evaluation in
man. I. Pharm. Pharmacal•• 12: 237T (1960).
186. Hollister, L. E., Studies of delayed-action medication. I: Meprobamate
administered as compressed tablets and as two delayed-action
capsules. N. Engl. I. uea., 266: 281 (1962).
187. Rosen, E., Tannenbaum, P., Ellison, T., Free, S. M., and Crosley,
A. P., Absorption and excretion of radioactively tagged dextroamphetamine
sulfate from a sustained-release preparation. I.A.M.A.,
194: 145 (1965).
188. Sugerman, A. A. and Rosen, E., Absorption efficiency and excretion
profile of a prolonged-action form of chlorpromazine. Clint
Pharmacal. Ther., 5: 561 (1964).
189. Rosen, E., Ellison, T., Tannenbaum, P.. Free, S. M., and Crosley,
A. P., Comparative study in man and dog of the absorption and excretion
of dextroamphetamine- 14C sulfate in sustained-release and
nonsustained-release dosage forms. I. Pharm. sci., 56: 365 (1967).
190. Rosen, E., Polk, A., Free, S. M., Tannenbaum, P. J., and Crosley,
A. P., Comparative study in man of the absorption and excretion
of amobarbital-14C from sustained-release and nonsustained-release
dosage forms. I. Pharma. Sci•• 56: 1285 (1967).
191. Khali, S. A. H. and Elgamal, S. S., In vitro release of aspirin
from various wax-coated formulations. I. Pharm. Pharmacal., 23:72
(1971) .
192. Royal, J., In vitro method for the determination of the rate of release
of amphetamine sulfate from sustained release medication.
Drug Stand•• 26:41 (1958).
193. Royal, J., A comparison of in vitro rates of release of several
brands of dextroamphetamine sulfate sustained release capsules.
Drug stana«, 27: 1 (1959).
194. Campbell, J. A., Nelson, E., and Chapman, D. G., Criteria for
oral sustained release medication with particular reference to
amphetamine. Can. Med. Assoc. I., 81:15 (1959).
195. Nash, R. A. and Marcus, A. D., An in vitro method for the evaluation
of sustained release products. Drug Stand., 28: 1 (1960).
196. Vliet, E. B., A suggested in vitro procedure for measuring the rate
of drug release from timed release tablets and capsules. Drug
Stand., 27: 97 (1959).
197. Beckett, A. H. and Tucker, G. T., A method for the evaluation of
some oral prolonged-release forms of dextroamphetamine in man,
using urinary excretion data. I. Pharm. Pharmacal. I 18: 72S
(1966) •
198. Campbell, J. A., Evaluation of sustained-action release rates. Drog
Cosmet. Ind., 87: 620 (1960).
199. Ipsen, J., Mathematical relationship of in vitro release rates and
biological availability of controlled-release nitroglycerin (Nitrong).
Curr. Ther. Res., 13:193 (1971).
200. Feinblatt, T. M. and Ferguson, E. A., Timed-disintegration capsules.
An in vivo roentgenographic study. N. Engl. I. Med., 254:940
(1956) •
201. Feinblatt, T. M. and Ferguson, E. A., Timed-disintegration capsules
(Tympcaps): A further study. An in vivo roentgenographic study,
296 Chang and Robinson
blood level study and relief of anginal pain with pentaerythritol
tetranitrate. N. Engl. J. Med., 256:331 (1957).
202. Aming, Y. M. and Nagwekar, J. B., Rationale for apparent differ·
ences in pharmacokinetic aspects of model compounds determined
from blood level data and urinary excretion data in rats. J. Pharm,
sa., 65: 1341 (1976).
203. Henning, R. and Nyberg, G•• Serum quinidine levels after administration
of three different quinidine preparations. Bur, J. Clin.
Pharmacol., 6: 239 (1973).
204. Green, D. M., Tablets of coated aspirin microspherules: A new
dosage form. J. New Drugs, 6: 294 (1966).
205. Magee, K. R. and Westerberg, M. R., Treatment of myasthenia
gravis with prolonged-action mestinon. Neurology, 9: 348 (1959).
206. Cass, L. J. and Frederik, W. S., CliniCal comparison of a sustainedand
a regular-release aspirin. Curro Ther. Res., 7:673 (1965).
207. cass, L. J. and Frederik, W. S., A clinical evaluation of a sustained-
release aspirin. Curl". Ther. Res., 7:683 (1965).
208. Sirine, G., Microencapsulation. Drug Cosmet. Ind., 101:56
(Sept. 1967).
209. Luzzi, L., Microencapsulation. J. Pharm, Sci•• 59: 1367 (1970).
210. Bakan, J. A. and Sloan. F. D., Microencapsulation of drugs. Drug
Cosmet. Ind., 110: 34 (March 1972).
211. Nixon, J. R. and Walker, S. E•• The in vitro evaluation of gelatin
coacervate microcapsules. J. Pharm. Pharmacol., 23:1478 (1971).
212. Matthews, B. R. and Nixon, J. R., Surface characteristics of
gelatin microcapsules by scanning electron microscopy. J. Pharm.
Pharmacal., 26:383 (1974).
213. Tanaka, N., Takino, S., and Utsumi, I., A new oral gelatinized
sustained-release dosage form. J. Piuirm, Sci., 52:664 (1963).
214. Phares. R. E. and Sperandio , G. J., Coating pharmaceuticals by
coacervation. J. Pnarm, sct., 53: 515 (1964).
215. Madan, P. L., Luzzi, L. A., and Price, J. C., Microencapsulation
of a waxy solid: Wall thickness and surface appearance studies.
J. Ph arm. sa., 63: 280 (1974).
216. Madan, P. L., Luzzi, L. A., and Price, J. C., Factors influencing
microencapsulation of a waxy solid by complex coacervation.
J. Pharm. sa., 61:1586 (1972).
217. Luzzi, L. A. and Gerraughty, R. J., Effects of selected variables
on the mtcroencapsulation of solids. J. Pharm• Sci., 56: 634 (1967).
218. Merkle, H. P. and Speiser. P •• Preparation and in vitro evaluation
of cellulose acetate phthalate coacervate microcapsules. J. Pharm.
sa., 62: 1444 (1973).
219. LUZzi, L. A. and Gerraughty, R. J •• Effect of additives and formulation
techniques on controlled release of drugs from microcapsules.
J. Pharm. sct., 56: 1174 (1967).
220. Birrenbach, G. and Speiser. P. P.. Polymerized micelles and their
use as adjuvants in immunology. J. Ptuu-m, set., 65:1763 (1976).
221. McGinity. J. W.. Cobs. A. B.. and Martin, A. N., Improved method
for microencapsulation of soluble pharmaceuticals. J. Pharm. Set.,
64:889 (1975).
Sustained Drug Release from Tablets and Particles 297
222. Khanna, S. C., Jecklin, T., and Speiser, P., Bead polymerization
technique for sustained-release dosage form. J. Pharm. Sci•• 59:
614 (1970).
223. Khanna, S. C. and Speiser, P., In vitro release of chloramphenicol
from polymer beads of a -methacrylic acid and methylmethacrylate.
J. Ptuirm, sct., 59: 1398 (1970).
224. Seager, H. and Baker, P., The preparation of controlled release
particles in the subsieve size range. J. Pharm. Pharmacal., 24:
123P (1972).
225. Croswell, R. W. and Becker, C. H., Suspension polymerization for
preparation of timed-release dosage forms. J. Pharm. sct., 63:440
(1974) •
226. Willis, C. R. and Banker, G. S., Polymer-drug interacted systems
in the physicochemical design of pharmaceutical dosage forms.
I: Drug salts with PVM/MA and with a PVM/MA hemi-ester.
J. Pharm. Sci•• 57: 1598 (1968).
227. Goodman, H. and Banker, G. S., Molecular- scale drug entrapment
as a precise method of controlled drug release. I: Entrapment of
cationic drugs by polymeric flocculation. J. Pharm. Sci•• 59: 1131
(1970) •
228. Rhodes, C. T., Wai, K•• and Banker. G. S., Molecular scale drug
entrapment as a precise method of controlled drug release.
II: Facilitated drug entrapment to polymeric colloidal dispersion.
J. Pharm, sa., 59: 1578 (1970).
229. Rhodes, C. T •• Wait K•• and Banker, G. S., Molecular scale drug
entrapment as a precise method of controlled drug release.
III: In vitro and in vivo studies of drug release. J. Pharm. Set,
59: 1581 (1970).
230. Boyland. J. C. and Banker, G. S., Molecular-scale drug entrapment
as a precise method of controlled drug release. IV: Entrapment of
anionic drugs by polymeric gelation. J. Pharm. Sci•• 62: 1177
(1973) •
231. Larson. A. B. and Banker, G. S.. Attainment of highly uniform
drug dispersions employing molecular scale drug entrapment in
polymeric latices. J. Pharm. sct.; 65: 838 (1976).
232. Bell, S. A•• Berdick, M.• and Holliday, W. M•• Drug blood levels
as indices in evaluation of a sustained-release aspirin. J. New
Drugs, 6: 284 (1966).
233. Rotatein , J., Estrin,!., Cunningham, C., Gilbert, M•• Jordan. A.,
Lamstein • J.. Safrin, M., Wimer, E., and Silson, J., The use of a
sustained-release aspirin preparation in the management of rheumatoid
arthritis and osteoarthritis. J. GUn. Pharmacal.. 7: 97 (1967).
234. Draper. E. B. and Becker. C. H.. Some wax formulations of
sulfaethylthiadiazole produced by aqueous dispersion for prolongedrelease
medication. J. Pharm. ss«, 55: 376 (1966).
235. Robinson,!. C. and Becker, C. H., Sulfaethylthiadiazole (SETD)
release from synthetic wax prolonged-release particles. I: Effect
of dispersant concentration. J. Pharm. Sci •• 57:49 (1968).
236. John, P. M. and Becker, C. H., Surfactant effects on spraycongealed
formulations of sulfaethylthiadiazole-wax. J. Pharm. Sci••
57:584 (1968).
298 Chang and Robinson
237. Cusimano. A. G. and Becker, C. H., Spray-congealed formulations
of sulfaethylthiadiazole (SEl'D) and waxes for prolonged-release
medication. Effect of wax. J. Pharm. sct., 57: 1104 (1968).
238. Raghuathan, Y. and Becker. C. H•• Spray-congealed formulations of
suIfaethylthiadiazole (SETD) and waxes for prolonged-release medication.
Effect of modifiers. J. Pharm. sa., 57: 1748 (1968).
239. Hamid, I. S. and Becker, C. H., Release study of sulfaethylthiadiasole
(SETD) from a tablet dosage form prepared from spraycongealed
formulations of SETD and wax. J. Phorm, Sci., 59: 511
(1970) •
240. Wurster, D. E., U.S. Patent 2,648,609 (August 11, 1953).
241. Wurster, D. E., U.S. Patent 2,799,241 (July 16, 1957).
242. Doerr, D. W., Series, E. Rot and Deardorff, D. L., Tablet coatings:
Cellulosic high polymers. J. Am. Pharm. Assoc., Sci. Ed.,
43: 433 (1954).
243. Gagnon, L. P., DeKay. H. G.. and Lee, C. 0 •• Coating of granules.
Drug Stand., 23:47 (1955).
244. Ahsan , S. S. and Blaug, S. M., A study of tablet coating using
polyvinylpyrrolidone and acetylated monoglyceride. Drug Stand., 26:
29 (1958).
245. Wagner, J. G., Veldkamp, W., and Long, S., Enteric coatings.
IV: In vivo testing of granules and tablets coated with styrenemaleic
acid copolymer. J. Am. Pharm. Assoc., Sci. Ed., 49:128
(1960) •
246. Kleber, J. W., Nash, J. F OJ and Lee. C. C., Synthetic polymers as
potential sustained-release coatings. J. Pharm. Sci., 53: 1519
(1964) .
247. Asker, A. F. and Becker. C. H., Some spray-dried formulations of
sulfaethylthiadiaoole for prolonged-release medication. J. Pharm.
sa., 55: 90 (966).
248. Lappas, L. C. and McKeehan, W•• Polymeric pharmaceutical coating
materials. II: In vivo evaluation as enteric coatings. J. Pharm.
sa., 56: 1257 (1967).
249. Seidler, W. M. K. and Rowe, E. J., Influence of certain factors on
the coating of a medicinal agent on core tablets. J, Pharm. Sci., 51:
1007 (1968).
250. Powell, D. R. and Banker, G. S., Chemical modifications of polymeric
fllm systems in the solid state. I: Anhydride acid conversion.
J. Pharm. Sci•• 58: 1335 (1969).
251. Tsevdos, T. J., Press-coated and multi-layer tablets. Drug Cosmet.
Ind.. 78: 38 (1956).
252. Windheuser. J. and Cooper, J., The pharmaceutics of coating tablets
by compression. J. Am. Pharm. Assoc., Set. Ed., 45:542 (1956).
253. Cooper, J. and Pasquale, D., ThE' present status of compression
coating. Pharm. J., 181: 397 (1958).
254. Lachman. L •• Speiser, P. P., and Sylwestrowicz. H. D., Compressed
coated tablets. I: Measurement and factors influencing
core concentration. J. Pharm. Sci., 52:379 (1963).
255. Wurster, D. E., Air-suspension technique of coating drug particles.
A preliminary report. J. Am. Pharm. Assoc., Sci. Ed •• 48: 451
(1959) •
Sustained Drug Release from Tablets and Particles 299
256. Wurster. D. E., Preparation of compressed tablet granulations by
the air-suspension technique. II. J. Am. Pharm. Assoc., Sci.
sa., 49: 82 (1960).
257. Coletta. V. and Rubin, H., Wurster coated aspirin. I: FUm-coating
techniques. J. Pharm. Sci., 53:953 (1964).
258. Wood, J. H. and Syarto, J., Wurster coated aspirin. II: An
:In vitro and :In vivo correlation of rate from sustained-release
preparations. J. Pharm. sa., 53:877 (1964).
259. Donbrow, M. and Friedman, M., Timed release from polymeric films
containing drugs and kinetics of drug release. J. Pharm. sct., 64:
76 (1975).
260. Higuchi, T., Mechanism of sustained-action medication. Theoretical
analysis of rate of rel~se of solid drugs dispersed in solid matrices.
J. Pharm. sa., 52: 1145 (1963).
261. Borodkin, S. and Tucker, F. E., Drug release from hydroxypropyl
cellulose-polyvinyl acetate fflms , J. Pharm. sct., 63: 1359 (1974).
262. Borodkin, S. and Tucker, F. E., Linear drug release from laminated
hydroxypropyl cellulose-polyvinyl acetate films. J. Pharm. Sci., 64:
1289 (1975).
263. Bercher, P. R., Bevans, D. W., Gormley, J. D., Hubbard, R. E.,
Sullivan, D. D.. Stevenson, C. R.• Thomas, J. B .• Goldman. R.,
and Silson, J. E.. Sinusule: Timed-release therapy in allergic
rhinitis. with human in vivo release rates and clinical statistics.
Curro Ther. Res., 9:379 (1967).
264. Sjogren, J. and Fryklof. L. W., Duretter: A new type of oral
sustained action preparation. Farm. Rev. (Stockh.), 59:171 (1960).
265. Cooper, J •• Lontab repeat action tablets. Drug Cosmet. Ind., 81:
312 (1957).
266. Sjogren, J. and Ostholm, I., Absorption studies with a sustained
release tablet. J. Piiarm, Pharmacol., 13:496 (1961).
267. Sjogren, J. and Ervik. M., A method for release rate determination
from sustained-release tablets. Acta Piuirm, Suecica, 1: 219 (1964).
268. Collins. R. L. and Doglia, S., Theory of controlled release of
biologically active substances. Arch. Environ. Contam. Toxicol., 1:
325 (1973).
269. EI-Egakey, M. A., EI-Khawas, F., EI-Gindy, N. A., and
Abdel-Khalik, M•• Release study of some drugs from polymeric
matrices. Pharmazie, 29: 286 (1974).
270. Asker, A. F., Motawi, A. M., and Abdel-Khalek, M. M., A study
of some factors affecting the in vitro release of drug from prolonged
release granulations . I: Effect of method of preparations.
Piuirmazie, 26: 170 (1971).
271. Asker, A. F., Motawi, A. M., and Abdel-Khalek, M. M., A study
of SOme factors affecting the in vitro release of drug from prolonged
release granulations. 2: Effect of dissolution retardent.
Pharmazie, 26: 213 (1971).
272. Asker, A. F., Motawi, A, M., and Abdel-Khalek, M. M., A study
of some factors affecting the in vitro release of drug from prolonged
release granulations. 3: Effects of particle size, enzymatic
contents of pepsin and pancreatin, bUe and ionic concentration.
Pharmazie, 26: 215 (1971).
300 Chang and Robinson
273. Lapidus, H. and LordiIN. G.. Some factors affecting the release of
a water-soluble drug from a compressed hydrophilic matrix.
J. Ph arm. Sci., 55:840 (1966).
274. Huber, H. E., Dale, L. B•• and Christenson, G. L•• Utilization of
hydrophilic gums for the control of drug release from tablet formulations.
1: Disintegration and dissolution behavior. J. Pharm. Sci.,
55:974 (1966).
275. Javoid, K. A., Fincher, J. H.• and Hartman, C. W., Timed-release
tablets employing lipase-lipid-sulfamethiazole systems prepared by
spray-congealing. J. Pharm. sa., 60: 1709 (1971).
276. Javaid, K. A. and Hartman, C. W., Blood levels of sulfamethizole in
dogs following administration of timed-release tablets employing
lipase-lipid-drug systems. J. Pharm. Sci•• 61:900 (1972).
277. Farhadieh , B" Borodkin, S., and Buddenhagen, J. D•• Drug release
from methyl acrylate-methyl methacrylate copolymer matrix.
II: Control of release rate by exposure to acetone vapor. J. Pnarm,
sa., 60:212 (1971).
278. Elbert. W. R., Morris, R. W., Rowles, S. G., Russell, H. T ••
Born. G. S., and Christian I J. E.. In vivo evaluation of absorption
and excretion of pentylenetetrazol-10- 14C from sustained-release and
nonsu stained -release tablets. J. Pharm. Sci., 59: 1409 (1970).
279. Farhadieh. B., Borodkin , S., and Buddenhagen, J. D•• Drug release
from methyl acrylate-methyl methacrylate copolymer matrix.
1: Kinetics of release. J. Pharm. sct., 60:209 (1971).
280. Ritschel. W. A. I Influence of formulating factors on drug safety of
timed-release nitroglycerin tablets. J. Pharm. sct., 60: 1683 (1971).
281. Sjuib, F., Simonelli, A. P., and Higuchi, W. I., Release rates of
solid drug mixtures dispersed in inert matrices. Ill: Binary
mixture of acid drugs released into alkaline media. J. Pharm. Sci.,
61: 1374 (1973).
282. Sjuib, F., Simonelli, A. P., and Higuchi, W. 1.. Release rates of
solid drug mixtures dispersed in inert matrices. IV: Binary
mixture of amphoteric drugs released into reactive media. J. Pharm.
Sci., 61:1381 (1972).
283. Goodhart, F. W., McCoy. R. H., and Ninger, F. C., Release of a
water-soluble drug from a wax matrix timed-release tablet. J.
Pharm. sei., 63: 1748 (1974).
284. Choulis, N. H. and Papadopoulos I H., Timed-release tablets containing
quinine sulfate. J. Pharm. sct., 64: 1033 (1975).
285. Chien, Y. W., Lambert, H. J., and Lin, T. K., Solution-solubility
dependency of controlled release of drug from polymer matrix:
Mathematical analysis. J. Pharm. sct., 64: 1643 (1975).
286. Desai, S. J., Simonelli, A. P., and Higuchi, W. I •• Investigation of
factors influencing release of solid drug dispersed in inert matrices.
J. Pharm. set», 54: 1459 (1965).
287. Desai, S. J •• Singh, P., Simonelli. A. P., and Higuchi, W. 1.,
Investigation of factors influencing release of solid drug dispersed
in inert matrices. II: Quantitation of procedures. J. Pharm. Sci.,
55; 1224 (1966).
288. Desai, S. J., Singh, P., Simonelli. A. P., and Higuchi, W. 1.,
Investigation of factors influencing release of solid drug dispersed
Sustained Drug Release from Tablets and Particles 301
in inert matrices. III. Quantitative studies involving the polyethylene
plastic matrix. J. Pharm. sct., 55: 1230 (1966).
289. Desai, S. J., Singh, P., Simonelli, A. P., and Higuchi, W. 1.,
Investigation of factors influencing release of solid drug dispersed
in inert matrices. IV: Some studies involving the polyvinyl
chloride matrix. J. Pharm. Sci.. 55: 1235 (1966).
290. Singh. P., Desai. S. J •• Simonelli, A. P •• and Higuchi, W. I.,
Release rates of solid drug mixtures dispersed in inert matrices.
I: Noninteracting drug mixtures. J. Pharm, sct., 56:1542 (1967).
291. Singh, P., Desai, S. J., Simonelli, A. P., and Higuchi, W. I.,
Release rates of solid drug mixtures dispersed in inert matrices.
II: Mutually interacting drug mixtures. J. Piuirm; set., 56: 1548
(1967).
292. Singh, P., Desai, S. J., Simonelli, A. P .• and Higuchi, W. I.,
Role of wetting on the rate of drug release from inert matrices.
J. Pharm, Sci., 57: 217 (1968).
293. Schwartz, J. B., Simonelli, A. P., and Higuchi, W. I.. Drug release
from wax matrices. I: Analysis of data with first-order
kinetics and with the diffusion-controlled model. J. Pharm. Sci.,
57:274 (1968).
294. Schwartz, J. B., Simonelli, A. P., and Higuchi, W. I., Drug
release from wax matrices. II: Application of a mixture theory to
the sulfanilamide-wax system. J. Pharm. sct., 57: 278 (1968).
295. Levy, G. and Hollister, L. E., Dissolution rate limited absorption in
man. Factors influencing drug absorption from prolonged-release
dosage form. J. Pharm. Sci., 54: 1121 (1965).
296. Nicholson, A. E., Tucker, S. J., and Swintosky, J. V., Sulfaethylthiadiazole.
VI: Blood and urine concentrations from sustained
and immediate release tablets. J. Am. Pharm. Aasoc; , Sci. Ed., 49:
40 (1960).
297. Kaplan, L. L.. Determination of in vivo and in vitro release of
theophylline aminoisobutanol in a prolonged-action system. J. Pharm.
Sci•• 54:457 (1965).
298. Laekenbaeher, R. S., Chlortrimeton maleate repeat action tablets in
the treatment of pruritic dermatoses. Ann. Allergy. 10:765 (1952).
299. Bancroft, C. M., Tripelenamine hydrochloride and chlorprophenpyridamine
maleate. A comparison of their efficacies in the treatment
of hay fever. Ann. Allergy, 15: 297 (1957).
300. Fuller, H. L. and Kassel, L. E., Sustained-release triethanolamine
trinitrate biphosphate (Metamine) in angina pectoris. Antibiot. Med.,
3: 322 (1956).
301. DeRitter, E., Evaluation of nicotinic alcohol Timespan tablets in
humans by urinary excretion tests. Drug Stand., 28: 33 (1960).
302. Caas, L. J. and Fredertk , W. S., Clinical evaluation of long-release
and capsule forms of pentobarbital sodium. Curro Ther. Res., 4:
263 (1962).
303. Webster, J. J., Treatment of iron deficiency anemia in patients with
iron intolerance: Clinical evaluation of a controlled-release form of
ferrous sulfate. CUr'r. Ther, Res•• 4: 130 (1962).
304. Morrison, B. 0 •• Tolerance to oral hematinic therapy: Controlledrelease
versus conventional ferrous sulfate. J. Am. Geriatr, Soc., 14:
757 (1966).
302 Chang and RObinson
305. Crosland-Taylor, P. and Keeling. D. H•• A trial of slow-release tablets
of ferrous sulphate. Curro Ther, Res •• 7:244 (1965).
306. Tarpley. E. L•• Controlled-release potassium supplementation. Curro
Ther. Res., 16:734 (1974).
307. 0 'Reagan. T.. Prolonged release aspirin. Drug. Cosmet. Ind., 98:
35 (April 1966).
308. Wiseman, E. H. and Federici, N. J., Development of a sustainedrelease
aspirin tablet. J. Pharm. Sci.. 57: 1535 (1968).
309. Wiseman. E. H.• Plasma salicylate concentrations following chronic
administration of aspirin as conventional and sustained-release tablets.
Curro TheY'. Res •• 11:681 (1969).
310. Harris, R. and Regalado, R. G., A clinical trial of a sustainedrelease
aspirin in rheumatoid arthritis. Ann. Ph)!s. Med•• 9:8
(1967) •
311. Chandraesekaran. S. K•• Benson, H., and Urquhart. J., Methods
to achieve controlled drug delivery: The biomedical engineering
approach, pp, 557-593. In Sustained and Controlled Release Drug
Delivery Systems (J. R. Robinson, ed.), Marcel Dekker, New York,
1978.
5
The Pharmaceutical
Pilot Plant
Charles I. Jarowski
St. John's University, Jamaica, New York
I. INTRODUCTION
A. Primary Functions of the Pharmaceutical
Pilot Plant
The primary responsibility of the pilot plant staff is to ensure that the
newly formulated tablets developed by product development personnel will
prove to be efficiently, economically, and consistently reproducible on a
production scale. Attention must be given to the cost of manufacture,
since low production costs provide a competitive advantage. This is
especially for tablets containing generic drugs. As a consequence, welldesigned
tablet formulas must be processed intelligently to insure that
each unit operation is optimized. Thus, final compression conducted at a
rapid rate is economically advantageous. However, some of this advantage
will be lost if excessive production times are required and undesirable
yield losses accrue during the blending, granulating, drying, and compression
steps. Furthermore, excessive strain On the tablet presses can
lead to the need for both more frequent press repair and resurfacing of
the tablet punch faces.
The manufacturing instructions transferred to the production department
should be clearly written, readily understood, and unambiguous.
The widespread use of word processors makes standardization of such procedures
desirable. Attention should be given to the use of in-house production
equipment, unless sufficient data has been accumulated to support
the economic advantage of purchasing new equipment. Alternate manufacturing
equipment and procedures may also be required for international
companies intending to manufacture pharmaceutical products at several sites.
303
304 Jarowski
The physical properties and specifications for the manufactured tablet
formulation should match those established earlier by the product formulator,
the pilot plant staff. and the quality control department. Therefore,
the tablets manufactured on a production scale should possess the
proper weight. thickness. and content uniformity. Tablet hardness. disintegration,
and dissolution rates in simulated gastrointestinal media must
be consistently attainable to ensure optimal bioavafiability.
Additional responsibilities for pilot plant personnel include the evaluation
of new processing equipment and aiding in finding causes and providing
solutions for problems that occasionally arise during routine tablet
manufacture. The manufacturer of quality tablets at a rate of 10, 000 per
minute attests to the competence of tablet technologists. Such achievements
have been made possible by the interaction of Skilled research,
process development, and production personnel, Manufacturers of processing
equipment have also played key roles by supplying excellent apparatus
for improving unit operations. such as particle size reduction. mixing.
drying. granulating, compressing. and coating.
II. SELECTED FACTORS TO BE CONSIDERED
DURING DEVELOPMENT
The establishment of ideal formulations and procedures for plant manufacture
must take into account a variety of factors.
A. Efficient Flow of Granules
The high speed of tablet manufacture currently attainable requires that
the granule flow from hopper to die cavity be unimpeded. To accomplish
such desired free flow, several parameters need study. First of all, the
particle size distribution of the granules must be accurately defined. The
addition of glidants such as talc, starch. and magnesium stearate may be
indicated for the large-scale manufacture of the tablet formulas developed
in the laboratory.
The efficiency of granule flow can be checked in the pilot plant by
conducting experiments under extended running conditions at high speed.
Tablets produced for at least 3 to 4 h will possess uniform weight and
hardness if granule flow has been efficient. Granule flow may be judged
to be satisfactory and stfll tablet weight and hardness may be unsatisfactory.
Such deficiencies in many instances can be corrected by increasing
the concentration of fines.
B. Ease of Compression, Speed of Manufacture,
Yield of Finished Tablets
Tableting equipment, punches, and dies wID require less maintenance if
tablets are routinely easily compressed. Excessive compressional forces
applied in the manufacture of tablets wID lead to an increased frequency
of maintenance and repair. Tablets that are produced under such conditions
are apt to possess excessive hardness. and are apt to exhibit reduced
dissolution rates and erratic bioavailability.
The Pharmaceutical Pilot Plant 305
A high yield of finished tablets and the speed of their manufacture
are obvious economic advantages that an ideal formulation should ensure.
C. Content Uniformity
The use of finely milled ingredients in tablet manufacture is advantageous.
The oral absorption efficiency of poorly soluble neutral drugs is improved
as their surface area in the intestinal tract is increased. Granulations
prepared with such finely milled material and excipients possessing coarser
particle size may lead to classification problems under high-speed, largescale
tablet manufacture. Particle size distribution and blending time must
be well defined during the product development stage to avoid unacceptable
variation in drug concentration between individual tablets.
D. Advantage of Accurately Defined Variables
Process development pharmacists must conduct well-controlled scale-up experiments
on the tablet formulations developed in the laboratory. Such
variables as drying time, residual moisture, granule particle size distribution,
percentage of fines, and compressional force should be accurately
defined within specified limits. Deleterious effects encountered outside
these specified limits should also be recorded. Such detailed data developed
during the pilot stage will become valuable as tableting problems
arise in future production runs.
Reworking a production batch of tablets not meeting quality control
standards has peen complicated by recent rulings by the U. S. Food and
Drug Administration (FDA). The reason for the poor content uniformity,
the poor dissolution rate, or some discoloration, for example, must be discovered
and explained before reworking of such batches can be considered.
The value of a well-equipped p:ilot plant manned with a skilled staff will.
become increasingly apparent to managers of a pharmaceutical company
whose volume and variety of tablets produced is increasing.
III. TYPES OF ORGANIZATIONAL STRUCTURES
RESPONSIBLE FOR PILOT OPERATIONS
The types of organizational structures found in the drug-manufacturing
industry depends upon a number of factors. In a small company. there
may not be space allocated for pilot development of tablet formulations.
The tablet formulator in the product development department may be responsible
for the ill'st few production batches of tablets. Most l:ikaly he
or she would also be called upon to aid in the solution of problems encountered
in future batches.
In a large organization, a pharmaceutical development team may be responsible
for the transfer of new tablet formulations to production. A
well-equipped space for conducting scale-up runs of new tablets will also
be provided in such an organization. Such a developmental team may report
administratively to the head of pharmaceutical research and development.
In other instances. such a group will be a part of the production
division.
306 Jarowski
The various types of organizational structures that can exist for
pharmaceutical pilot development are mentioned below.
A. Research Pharmacist Responsible for Initial
Scale-Up and Initial Production Runs
Figure 1 represents an example of an organizational structure wherein no
direct responsibility exists administratively for the pilot operation. This
abbreviated structure is concerned only with administrative responsibility
for the research and production of tablet dosage forms. The solid lines
connecting the boxes represent administrative lines of authority and responsibility.
The broken lines represent lines of communication.
In such an organizational structure, direct responsibility rests with the
director of pharmaceutical research and development and the manager of
solid-dosage form research. The tablet formulators, who report administratively
to the manager. are responsible for the scale-up and initial
production runs of their tablet formulations.
There are advantages that can be cited for such an arrangement.
Thus, the tablet formulators who know the most about their new tablet are
directly involved with the manager and section head in the production division.
who will be responsible for the routine manufacture of their tablet.
Communicating such information as stability data, physical properties. and
the reasons for the procedure adopted is best conveyed by the tablet
formulator, who alone is aware of the variations in formulation that were
investigated. Several batches of the preferred tablet formula had to be
Director of pharmaceut1cal
Research
III
I
Manager of Solid
100-
Dosage Form Research
I : '" ....'" ....'"
Pharmaceutlcal
".' ----- Research Staff
Technlcal Dlrector of - Production
•II
I
". Manager of Tablet and
Capsule Production
j
~
Sectlon Head -----
i
•
Operators
Figure 1 Abbreviated organizational diagram of a pharmaceutical company
where no direct responsibility exists administratively for the pilot operation.
The Pnarmaceutical Pilot Plant 307
prepared for accelerated and long-term stability testing. In addition.
many clinical batches may also have been prepared by the tablet formuIator-,
The advantage that accrues to the production division is that individuals
with the most technical information concerning the new tablet are directly
involved in the initial production runs of the new formula. The
tablet formulators from the research division will also gain from such an
arrangement. They in turn gain valuable experience concerning the types
of problems that can arise during the large-scale manufacture of their
tablets. Such experience wD1 serve to improve their competence in developing
future tablet products.
There are disadvantages that can be cited for such a working relationship.
Assigning the research division the responsibility for the first three
to five production batches means a significant commitment of research time.
The research pharmacist may strive to develop formulations and procedures
which are preferred by the production division in order to avoid
complaints. Thus. for example, resistance to a direct-compression or
double-compresskm procedure might be encountered. The heavy investment
in equipment suitable for wet granulation methodology by the production
division will naturally result in their continuing support for such procedures.
A person not thoroughly familiar with the production division's equipment,
materials, ete , , could cause unnecessary expense in urging the
acquisition of new manufacturing equipment. Furthermore, a product development
pharmacist is not as familiar with the plant operations as a pilot
plant person whose job responsibility necessitates knowing the production
division's personnel and facilities.
In a small company -where equipment and personnel are limited -the
tablet formulator can more readily introduce a formula due to the limitations
of facilities and personnel. In a big company. the pilot laboratory is
called upon to bridge the gap between the production and research divisions.
Manufacturing problems can arise with different lots of bulk materials.
Efficient problem solving will be favored if the troubleshooter is
familiar with the production division's facilities and has a good working relationship
with research and production personnel.
The objective of the research pharmacist is to develop practical tablet
formulas that cause minimal manufacturing problems. However. he or she
must not become complacent. A less practical. more expensive procedure
may produce a tablet exhibiting superior bioavaBabnity. Such a formula
modification may be difficult to promote within an organization while patent
protection exists for the drug in the less-efficient tablet. The alert and
capable tablet formulator will not wait for generic competition to initiate
such an improvement in oral absorption through the use of a new manufacturing
procedure.
Formulations that are manufactured very efficiently can still be improved
by using less-expensive excipients. Automated procedures can
significantly reduce manufacturing costs. Such improvements can only be
made after an investment of time and effort. Frequent interruptions in
research effort result when the research pharmacist is called upon to
search for the solution to a problem that has arisen during the manufacture
of a tablet formulation developed by him or her. Such interruptions are
unfortunate because they are distracting and disruptive to long-term
research.
308 Jarowski
B. Pharmaceutical Pilot Plant Controlled
by Pharmaceutical Research
The abbreviated organizational diagram depicting such an administrative
structure is shown in Figure 2. There are a number of advantages in this
type of organizational structure. Thus, newly hired research pharmacists
can include service in the pharmaceutical pilot plant in their orientation
program. The scheduling of scale-up runs is under the research division's
control. Clinical supplies of a new drug in a tablet dosage form should be
controlled by the research division when it is prepared in the pilot plant.
If any scale-up problems are encountered in the preparation of clinical
lots, corrective measures, such as dosage formulation modification, can be
made immediately. Such corrective measures should be made as rapidly as
possible to ensure that the formulation being clinically evaluated will be
the one ultimately marketed and that it will prove to be practical to manufacture.
The pharmaceutical research and development division is dedicated to
developing dosage forms of new drugs and to search for more economical
manufacturing procedures and formulations. Thus, it is expected that in a
pilot plant under research control, new processing equipment, automated
procedures, and alternate excipients will be actively investigated. Ideally,
the pharmaceutical pilot plant facility should be located close to the manufacturing
unit. With such proximity several advantages accrue:
Technical D1rector of
Production
•II
I
Manager of Tablet and
Capsule Product1on
,,I
I
Section Head 1n
Charge of
Tablet Production
I •I!
I Operators I
,
I
I
I
Director of Pharmaceutical Research
and Development ..
• j ,I
I : I
Manager of So11d Manager of
Dosage Form ..~ Pilot
Research Development
: : I I I ! I
Pharmaceutical Pharmacists and
Research Staff . Engineers
Figure 2 Abbreviated organizational diagram of a pharmaceutical company
where direct responsibility for the pilot plant is under the research
division.
The Pharmaceutical Pilot Plant 309
1. Scale-up runs can be readily observed by members of the production
and quality control divisions.
2. Supplies of excipients and drugs. cleared by the quality control
division. can be drawn from the more spacious areas provided to
the production division.
S. Supplies of packaging material will be easily accessible.
4. Access to engineering department personnel is provided for equipment
installation. maintenance, and repair.
5. Subdivision and packaging of clinical supplies will be expidited by
temporarily borrowing personnel from the production division.
The principal disadvantage that can arise from such an organizational
structure is that inadequate time may be provided to solving production
problems when they arise. The fact that such problems arise unpredictably
makes it difficult to justify maintaining an adequately staffed
troubleshooting team in the production division. An experienced pilot
plant development team is technically trained to solve such problems. Its
frequent interaction with both research and production personnel provides
this team with an excellent base of diverse information which is of value in
troubleshooting. All that would seem to be needed to surmount this possible
disadvantage is the assignment of priorities by closely cooperating
heads of production and research.
C. Pharmaceutical Pilot Plant Controlled
by the Production Division
Pilot development of tablet formulations from the research division under
such an organizational structure would be carried out by personnel reporting
administratively to the production division. The abbreviated organizational
diagram depicting this is shown in Figure 3.
In such an organization, the responsibility of the tablet formulator is
to establish the practicality of his or her formula and manufacturing procedure
in pilot plant equipment. Such equipment is most likely to be similar
to that used in the production division. This similarity is advantageous
because a successful scale-up in the pilot plant augers well for
future success in large-scale manufacturing equipment.
The frequency of direct interaction of the tablet formulator with production
personnel in the manufacturing area will be reduced under this
organizational structure. The presence of a pilot plant team will make it
likely that manufacturing problems will be directed to their own pilot plant
personnel. High priority can be assigned because the pilot plant is under
the control of the production division. Thus, any troubleshooting that is
required will be initiated expeditiously. Having competent individuals in
the pilot plant will serve to reduce interruptions on the research effort
by the production division.
On the other hand, initiation of a developmental project that may require
pilot plant equipment not representative of production models is
likely to meet resistance. The pressures of manufacturing a line of products
on schedule is bound to make the pilot group ever sensitive to production
priorities. As a consequence, a sustained commitment on the development
of new manufacturing procedures is likely to be assigned a
lower priority.
310 Jarowski
Section Head in
Charge of
Tablet Production
Technical Director of
Production
Manager of Tablet and
Capsule production
Section Head of
Pilot Development
Director of Pharmaceutical . Research
I
I
II
•·
Manager of Solid Dosage l--
Fot1\l Research
r
•I:
-'
Pharmaceuti ca 1
Research Staff
1----
Figure 3 Abbreviated organizational diagram of a pharmaceutical company
where direct responsibility for the pilot operation is under the production
division.
IV. EDUCATIONAL BACKGROUNDS OF
PILOT PLANT PERSONNEL
Pilot plant staff members concerned with the development of tablets should
understand the functions of the ingredients used, the various unit operations
involved, and the general methods of preparing tablets. They
should possess knowledge of the physicochemical factors which could adversely
affect the bioavailabfiity of the drug component from a tablet
dosage form. In addition, good communication skills, both oral and
written, are essential because these individuals must frequently interact
with members of the pharmaceutical research and production divisions.
An individual with graduate training in industrial pharmacy is well
qualified for such a responsibility. Industrial pharmacy core curricula
and electives usually include the following courses: product formulation
(lecture and laboratory), manufacturing pharmacy (lecture and laboratory),
industrial pharmacy. biopharmaceutics. homogeneous systems, pharmaceutical
engineering. evaluation of pharmaceutical dosage forms, pharmaceutical
materials, regulatory aspects of drug production and distribution,
principles of quality assurance, reaction kinetics, and physical chemistry
of macromolecules.
Tablet production operations are by no means static. The following
technological trends attest to anticipated production complexity: (1) preparing
tablets by continuous processing, (2) aqueous film coating, (3)
direct compression of tablets, and (4) use of fluid-bed technology in
granulating powder blends and in coating tablets. The addition of mechanical
and chemical engineers would be advantageous in the selection of new
The Pharmaceutical Pilot Plant 311
equipment offerings being considered for such operations as well as in the
organization of the most economical sequence of unit operations.
Whatever the technological trend ~ once the goals are defined. the best
combination of trained personnel can be determined for the pilot plant.
V. PILOT PLANT DESIGN FOR TABLET
DEVELOPMENT
The 1979 amendments of Current Good Manufacturing Practices of the Food
and Drug Act are fairly comprehensive with respect to the design, maintenance,
and cleanliness of a production facility used in the manufacture
of drugs and their various dosage forms. One of the paragraphs of the
regulations on Current Good Manufacturing Practices reads as follows:
Buildings shall be maintained in a clean and orderly manner and
shall be of suitable size. construction and location to facilitate
adequate cleaning, maintenance and proper operations in the manufacturing,
processing, packaging, labelling or holdings of a
drug.
Thus, a drug shall be deemed to be adulterated if the methods used for
its manufacture, processing. packaging, or holding are not operated in
conformity with good manufacturing practice. The above regulations apply
to the pilot plant area as well, since it serves as the site for the
preparation of clinical supplies of drugs in various dosage forms.
The design and construction of the pharmaceutical pilot plant for tablet
development should incorporate features necessary to facilitate mamtenance
and cleanliness. If possible, it should be located on the ground
floor to expedite the delivery and shipment of supplies. Extraneous and
microbiological contamination must be guarded against by incorporating the
following features in the pilot plant design:
1. Fluorescent lighting fixtures should be the ceiling flush type.
2. The various operating areas should have floor drains to simplify
cleaning.
3. The area should be air conditioned and humidity controlled (75°F
and 50% relative humidity). The ability to maintain a much lower
relative humidity in a confined area may be desired if special
processing conditions are required, such as the development of
effervescent tablets. In other areas, only clean but not conditioned
air is needed, such as in the site where the fluid-bed
dryer is installed.
4. High -density concrete floors should be installed in all areas where
there is heavy traffic or materials handling.
5. The walls in the processing and packaging areas should be enamel
cement finish on concrete masonry for ease of maintenance and
cleanliness.
6. Equipment in the pharmaceutical pilot plant should be similar to
that used by the production division in the manufacture of tablets.
The diversity of equipment on hand should enable the developmental
team to conduct unit operations in a variety of ways.
Space should be provided for evaluating new pieces of equipment.
....
Qo
312
Jarowski
1.1 7.1 7.5 13.1 13.2
7.0 13 .0
1.0 i , Drying Area
Granulating
7.9
1.9
Blending
2.8
8.0
8.3 8.5
Coating Cleaning Area
Area 9.0 [l;J I 9.3 ] 2.7
Packaging Area
I 3.1 1
3.0
I 3.1 I
3.2
4.0
Conference Room and
Library
5.0 6.0
Manager's Office sec'ty
Office
r-
U
10.1 10.2
weigh~ 10.0 -
Area I 10.3
Storage Area for Drugs
and Excipients
i i ,o
Quarantine Area
~
12.0
J
I
-I t--------------- 60'
I,-
I
Figure 4 Floor plan for tablet development.
The Pharmaceutical Pilot Plant 313
A typical floor plan is shown in Figure 4; the floor plan reference is
shown in Table 1. A typical floor plan for a small pilot plant is shown in
Figure 5, and its reference is shown in Table 2. The variety of equipment
installed must be narrow because of space limitations. However, efficient
use of such limited space can be made by borrowing from the production
division, as needed, portable tablet presses, mills, and coating
pans.
Table 1 Floor Plan Reference for the Tablet Development Area (Fig. 4)
1. 0 Tablet compression room
1.1 Heavy-duty tablet press for
dry granulation
1.2 Rotary tablet press, 54
stations
1. 3 Rotary tablet press, 16
stations
1. 4 Compression-coating tablet
press
1.5 Single-punch tablet press
1. 6 Tablet deduster
1. 7 Storage cabinet for the
punches and dies
1.8 Bench top with wall cabinets
Ancillary equipment; balances,
friabilator, disintegrationtesting
apparatus, dissolutiontesting
apparatus. hardness
tester
1. 9 Bench top with wall cabinets
Ancillary supplies: tools,
vacuum cleaner
2. 0 Coating area
2.1 Coating pan. 36-in diameter
2.2 Coating pan, 30-in diameter
2.3 Coating pan, 20-in diameter
2. 4 Perforated coating pan.
24-in diameter
2. 5 Polishing pan
2. 6 Fluidized bed coating
apparatus
2. 7 Airless spray equipment
2. 8 Area for preparing syrups •
solutions, and suspensions
for sugar or f"11m coating
2.9 Bench top with wall cabinets
3. 0 Packaging area
3. 1 Work tables
3.2 Storage cabinets containing
packaging supplies
4.0 Conference room and library
5.0 Manager's office
6.0 Secretarial office
7.0 Granulating area
7.1 Comminuting mill
7. 2 Sigma blade mixer
7.3 Planetary-type mixer
7.4 Tornado mill
7.5 Fluidized bed spray granulator
7.6 Comminuting equipment
7. 7 Extructor
7.8 Chilsonator
7.9 Bench top with wall cabinets
8.0 Blending area
8. 1 V shaped blender with intensifier
bar, 1 ft 3 capacity
8.2 V shaped blender with intensifier
bar, 5 ft3 capacity
8. 3 V shaped blender with intensifier
bar, 20 ft3 capacity
8. 4 Ribbon blender
8. 5 Lodige blender
314 Jarowski
Table 1 (Continued)
9.0 Cleaning area
9.1 Sink
9. 2 Ultrasonic cleaner
9. 3 Bench top with wall cabinets
10.0 Weighing area
10.1 Scale
10.2 Scale
10.3 Bench top with wall cabinets
Small scales and balances
positioned on bench top
11.0 Storage area for drugs and
excipients
12. 0 Quarantine area
13. 0 Drying area
13. 1 Tray and truck dryer
13. 2 Fluidized bed dryer
13.3 Freeze dryer
14. 0 Milling area
14. 1 Hammer mID
14.2 Ball mID
14.3 Muller
14.4 Fluid energy mID (micronizer).
2 in diameter
15.0 Spray-drying area
15.1 Spray-dryer
1.1 I 3.1 I
1.2 1.4 3.0 3.2 1.0 Granulating
Tablet Drying - Compression Milling 3.3
1.3 n 03.513.6(
3.4
I 1.5 I
3D'
--- •
I
-'•
u I 4.1 I
4.0 T:2
~
Weighing
Quarantine
storage
j
5.0
6.0
fl Off~ce
30 I ~
U
2.1
2.0
2.2 Coating r--
Area 2.4
2.3
I 2.5 I
•••
...
Figure 5 Floor plan for a small pilot plant for tablet development.
The Pharmaceutical Pilot Plant
Table 2 Floor Plan Reference for the Small Pilot Plant (Fig. 5)
315
1. 0 Tablet compression room
1. 1 Tablet press, single-punch
1.2 Rotary tablet press, 16
stations
1.3 Tablet deduster
1.4 Bench top with wall cabinets
Ancillary equipment: balances ,
friabilator, disintegrationtesting
apparatus, dissolutiontesting
apparatus, hardness
tester
1. 5 Bench top with wall cabinets
Ancillary supplies: tools,
vacuum cleaner
2.0 Coating area
2.1 Polishing pan, 36-in diameter
2.2 Coating pan, 36-in diameter
2.3 Airless spray equipment
2.4 Area for preparing syrups,
SOlutions, and suspensions for
sugar or film coating
2.5 Sink
3.0 Granulating, drying, and
milling area
3.1 Tray and truck dryer
3. 2 Hobart mixer
3. 3 Comminuting mill
3.4 Hammer mill
3. 5 Fluidized bed dryer
3. 6 V-shaped blender with intensifier
bar
4.0 Weighing area
4.1 Bench top with wall cabinets
4.2 Scale
5.0 Quarantine and storage area
6.0 Office
VI. PAPER FLOW BETWEEN RESEARCH, QUALITY
CONTROL, AND PRODUCTION
When the tablet formulation has reached the pilot developmentsl stage it
becomes the responsibility of the tablet formulator to issue clearly written
instructions for its manufacture. In addition, analytical specifications for
the ingredients must be thoroughly delineated. Thus, for example, the
use of micronized drug may be specified. The in-process steps describing
the milling conditions to be followed must be spelled out. In addition,
particle size characterization of the micronized material must be included.
In other instances, the latter characterization would be included as a
quality control responsibility for purchased micronized bulk drug. The
tablet formulator is well aware of the need to use micronized material on
the basis of dissolution studies conducted during the early stages of
formula development. When bulk drug specifications are being established,
such information is related to members of the quality control and production
divisions. A raw material (RM) number is designated for the micronized
bulk drug along with any physical or chemical information deemed
significant. Melting point is not only an important indication of the purity
of a compound, it also serves as an additional checkpoint for the desired
polymorphic form on a crystslline compound. A stringently low trace
metal content may be specified in some instances if improved drug stability
dictates it.
316
Specification Number
RM 1000
Specification
Classification
Release and Purchase
Specification
Effective Date Super8edes
7/11/80 10/10/79
Product Description
ASCorbic Acid, DSP, FCC
Empirical Formula C6Hg0 6
Molecular Weight 176.13
Jarowski
Page
1 of 1
Grade
'l'~let
Label Claim if Applicable
Description
Specific Rotation at 25
0C
Heavy Metals
Lead
Arsenic
Residue on Ignition
Assay
Mesh size
white, practically odorous,
crystalline powder with a
pleasantly tart taste
+20.5 to 21.SoC
Maximum 20 ppm
Maximum 10 ppm
Maximum 3 ppm
Maximum 0.1%
Minimum 99%
100% through a No. 20 u.s. Std
sieve
Maximum 60\ through a No. 80
sieve
Maximum 8% through a No. 200
sieve
Figure 6 Example of a specification sheet for a bulk vitamin.
A typical specification sheet for ascorbic acid is shown in Figure 6.
As there may be several grades or particle sizes of ascorbic acid within
the company. it is important to remember to specify the desired grade (in
this example. RM-1000).
Figures 7 -9 illustrate manufacturing instructions for tablets containing
50 mg of hydrochlorothiazide. On the first sheet (Fig. 7) are listed the
ingredients, the amounts per tablet, and the amounts of each required for
the batch size. At the top of the figure there appears the words "product
code number." Such a designation will help to avoid confusion by the
various departments involved in the manufacture and release of the finished
tablets. There may be hydrochlorothiazide tablet potencies other
than 50 mg also being manufactured in the company. When a new drug is
to be marketed in a tablet dosage form. a new product code number is
assigned and subsequently used.
There are columns headed by the words "compounded by" and "checked
by. II When each ingredient is weighed the individual who made the weighing
puts his or her initials along side the ingredient weight; the observer
Product and Potency Product Code Number 1 Lot Number Page
Hydrochlorthiazide Tablets, 50 mg 1 of 3
Batch Size Date Started Date Completed
5,000,000
Prepared By Production Approval Quality Control
Approval
Numbel Ingredient Grade Grams/Tablet Grarns/Batct !comReunded Checked RM Lot
Bv II 11
1 Hydrochlorthiazide ~.S.P. 0.050 250,000
2 Microcrystalline
Cellulose (PH 101) N.F. 0.050 250,000
3 Dicalcium phosphate, dihydratE N.F. 0.099 495,000
4 Magnesium Stearate N.F. 0.001 5,000
Total 0.200 1,000,000
Figure 7 List of ingredients used in manufacturing hydrochlorothiazide tablets (50 mg) as
they would appear in a set of manufacturing instructions.
"'3
;:l"
tl)
~;
(')
~
.....
5' a
"tl
1=:1
Q...
:g s.....
Co) ....
""I
Product and Potency Product Code Numbe~ Lot Number Page
Hydrochlortniazide Tablets, 50 mg 2 of 3
Steps Manufacturlng Instructions Compounded Checked
By By
1 Pass ltems I, 2 and 3 through a #16 screen using a comminutlng
mlll to break up any lumps.
2 Blend the three lngredients in a V shaped blender for 1 hour.
3 Add the magneslum stearate and contlnue blend~ng for 15 mlnutes.
4 Subdivlde the blended materlal lnto polyethylene-llned drums.
SUbmlt representatlve samples to the Quallty Control divlslon to
determlne lf the blend is homogeneous.
5 Upon receipt of clearance from Quallty Control, tablet the blend
wlth 5/16" standard, round concave puncres on the 54 statlon,
rotary tablet machine.
6 Submlt representatlve samples of the finished tablets to the
Quality Control division for final release.
ee ....
OJ
Figure 8 An example of manufacturing instructions for the prepa.ration of tablets.
~
~
I;Q
fI>' ....
Product and Potency Product Code Number Lot Number Page
Hydrochlorthia~ide Tablets, 50 mg 3 of 3
Tablet Specifications
Weight of Tablet 0.200 g
Weight Range 0.190 - 0.210 g
Potency 50 mg!tablet
Range in Potency 46.25 - 53.75 g !tablet
Hardness Range 7 - 9 (Strong-Cobb units)
Dis~ntegration Time Less than 5 minutes (U S P method)
Disso1utl.on Rate 60% of the labeled potency shall dissolve
withl.n 30 ml.nutes (U S P method)
Color of Tablet Whl.te
Figure 9 An example of a tablet specification form which is attached to the manufacturing
instructions.
~
;:,'
~
~
~
R
~.... .... oa
'l:l
~o....
~
Q;::s ....
C.:l .....
to
320 Jarowski
of the operation also places his or her initials as a double check on the
accuracy of the weighing operation. A similar procedure should be followed
for each step in the compounding operation. Such double checking
is required by the Food and Drug Administration for all new drug application-
type products. However, even if it were not required. such a
double-checking procedure should be followed to ensure that there are no
weighing or compounding errors.
The lot number of each ingredient used must be recorded. Such recordings
become of special importance when one is attempting to trace the
history of a particular batch of tablets. Troubleshooting will be simplified
in some instances by tracing the source of a raw material. A typical problem
that may arise can be due to a particular lot number of a substandard
excipient that could have been inadvertently used.
On the second sheet of the proposed manufacturing instructions (see
Fig. 7) the stepwise procedure to be followed is described. The words
chosen should be readily understood by the variety of individuals who will
be reading these instructions. Thus, quality control personnel considering
the establishment of in-process controls and pilot and production operators
who will be following the procedure for the first time should not be
confused by cryptic phrases containing technical words not easily understood.
The tablet formulator should keep in mind that his or her manufacturing
procedure may also be circulated to foreign manufacturing plants.
Manufacturing instructions that are clearly written are advantageous because
they avoid confusion which may lead to errors.
Figure 9 illustrates a list of specifications for the finished tablets. As
can be seen in Figures 7-9, spaces for the product code and lot number
are provided at the top of the form. The three forms become a part of
the file for the lot number of the batch of tablets prepared. These three
sheets serve as a historical record. Attachments that are included, for
example, would be the clearance data from the quality control division and
a record of the physicians receiving the material for clinical study.
VII. USE OF THE PILOT PLANT SCALE-UP TO SELECT
THE OPTIMAL PROCEDURE FOR THE PREPARATION
OF DEXAMETHASONE TABLETS
A company was planning to expand its tablet-manufacturing facility. The
company's tablet products in greatest demand were potent corticoids and
cardiovascular drugs. These water-insoluble, neutral drugs were to be
used in concentrations ranging from 0.1 to 5.0 mg per tablet.
The presence of 0.1 mg of drug in a tablet weighing 200 mg represents
a drug dilution with tablet excipients of 1: 2000. Maintaining content uniformity
could be a problem in the routine manufacture of such tablets. In
addition to content uniformity, the dissolution rates of such water-insoluble
drugs can be a problem. Such drugs as the corticoids and the cardiac
glycosides, on occasion, have been recalled by the Food and Drug Administration
because of a lack of bioequivalency.
The company's top management requested the pharmaceutical research
and development department to determine which tablet-manufacturing procedure
should be adopted for this line of tablet products. Their 0.25-mg
dexamethasone tablets were considered to be representative of the several
drugs whose production volume were to be expanded.
The Pharmaceutical PIlot Plant 321
One of four tablet-manufacturing procedures were to be considered:
(1) wet granulation; (2) microgranulation , (3) direct compression; and
(4) double compression. Each of these procedures had been used successfully
by the pharmaceutical research department in the development of
the 0.25-mg dexamethasone tablets. The responsibility now rested with
the pilot plant personnel to scale-up each of the four procedures. The
basis for selection of the optimal procedure would take into consideration
the following: (1) simplicity of manufacture; (2) physical properties of
the finished tablets. (3) content uniformity; (4) dissolution rate of the
dexamethasone from the tablets; and (5) cost of manufacture. On the
basis of the procedure selected decisions could be made about new equipment
purchases for the expanded manufacturing facility. The tablet
formulations are listed in Table 3 [1-3].
Dexamethasone was chosen as the drug in this study because it is a
water-insoluble. neutral compound whose bioavailability can be adversely
Table 3 Formulations of Experimental Dexamethasone Granules and
Tabletsa
Ingredients Wet Micro- Direct
(mg/tablet) granulation granulation compression Slugging
Dexamethasone 0.25 0.25 0.25 0.25
Starch. dry 10.00 11.00 15.00 11.00
Acacia. powdered 2.60 3.00 3.00 3.00
Lactose 130.75 130.25 126.75
Lactose. direct 127.75
compresaionv
Starchc
1.00
Acacia. powderedc 0.40
Purified water 18.60 8.00
Starch, dryd
4.00 4.00 4.00
Magnesium stearate 1. 00 1. 00 2.00 1.00
Talcum, powdered 0.50 2.00 3.00
Magnesium
e 1.00 stearate
Total 150.00 150.00 150.00 150.00
aBatch size for each method was 10; 000 tablets.
bDirectly compressible lactose.
"used as starch Iacacia paste.
d
Added to the dry granules.
eAdded for the final. compression.
Source: From Refs. 1-3.
322 Jarowski
affected by a poor tablet formulation. In each procedure the micronized
dexamethasone was mixed with the excipients by using a geometric dilution
procedure in a V-shaped blender of 3 kg capacity. The four procedures
followed are summarized below.
A. Wet Granulation
Homogeneous blends of the dexamethasone and excipients were granulated
with starch/acacia in a planetary mixer. The wet mass was passed through
an oscillating granulator equipped with a No. 8 screen and oven dried
overnight at 45°C. After dry screening through a No. 16 screen, the
granulations were blended with the remaining dry starch, lubricated,
and compressed.
B. Microgranulation
Microgranulates were prepared by a procedure first described by de Jong
[4) • An advantage cited for this procedure is that a blend of finely powdered
drug and excipients can be converted into compressible granules
with only minor changes in particle size. Thus, micro granulation is differentiated
from conventional wet granulation by the absence of large agglomerates.
A relatively small quantity of granulating liquid is used in the
process of forming microgranulates. As a consequence, the blended powder
particles become coated by a thin fUm of binder which reduces interparticular
attraction and yields a free-flowing powder which can be compressed
to form a tablet. Hydrophobic powders coated in such a manner
are rendered less hydrophobic, and thus more readily dispersible in water.
Since such microgranulations are devoid of large agglomerates, powder
blends granulated in this manner exhibit excellent content uniformity.
rapid dissolution. and excellent bioavailability.
In the preparation of dexamethasone microgranulates a homogeneous
blend of micronized drug, powdered acacia, starch, and lactose was placed
in a planetary mixer and uniformly moistened with a minimal quantity of
purified water. The uniform distribution of water is a critical step. Failure
to distribute the water uniformly will mean that portions of the blend
will have been inadequately granulated. As a result excessive capping
will result during the compression stage. The addition of excessive water
during the granulating step will result in the formation of macrogranules.
The uniformly moistened granules were passed through a No. 16 sieve
and tray dried at 45°C overnight. The dried granules were passed through
a No. 16 sieve. Additional starch and magnesium stearate were added to
the sieved microgranules and the blend was compressed into tablets.
C• Double Compression
Slugs were made using a heavy-duty press with t-In diameter punches.
The slugs were granulated through an oscillating granulator fitted with a
No. 16 sieve. Additional lubricant was added to the granules and, after
additional blending, the granules were compressed into tablets.
The Pharmaceutical Pilot Plant
D. Direct Compression
323
The dexamethasone, binder, disintegrant, and anhydrous lactose were
blended by geometric dilution. The homogeneous blend was lubricated and
compressed.
The granules prepared by the three granulation procedures and the
directly compressible blend were compressed on a 16-station tablet machine
in which only 4 stations were used. Flat punches measuring 5/16 in in
diameter were employed. Particle size analyses were performed on samples
of the powder blend before granulation, after granulation, and after compaction
and disintegration by using a nest of sieves and an electromagnetic
sieving machine. The particle size distribution of the granules after
compaction and disintegration was determined by the method of Khan and
Rhodes [5].
These authors prepared compacts by means of a hydraulic press. To
effect disintegration of the compacts without agitation, which might cause
granule fracture, a particularly effective disintegrant was included in the
compacts. In the case of compacts prepared from insoluble materials, such
as dicslcium phosphate, distilled water was used as the disintegration
vehicle; for other systems appropriately saturated solutions were used.
Particle size determinations of the disintegrated compacts were made using
a particle size counter or an air-jet sieve.
The dexamethasone tablets were allowed to disintegrate into particles
and granules by soaking them in absolute ethanol saturated with the drug.
Aqueous media were avoided because of the presence of water-soluble
lactose. The particles and granules were separated by filtration and dried
at 40°C for 16 h. The dried particles and granules were then gently
brushed through a 16-mesh screen. The screened particles and granules
were then subjected to a particle size analysis using a nest of sieves and
an electromagnetic sieving machine.
Representation of Particle Size Distribution Data
In Table 4 are shown the weight percentages and cumulative weight percentages
of particles on each of the sieves for the following: (1) the
dexamethasone powder blend, (2) dried granules prepared by wet granulation
before compaction, and (3) dried granules after compaction and
disintegration.
Inspection of the data in Table 4 reveals that 100% of the particles in
the dexamethasone powder blend have a particle size of less than 251 um,
since all the particles passed through the GO-mesh sieve. Approximately
50% of the material should have diameters of less than 152 um and greater
than 75 um, This follows, since 13% of the material was collected on the
120-mesh sieve and 43% was collected on the 200-mesh sieve. The aperture
for the 100-mesh sieve is 152 um and the aperture for the 200-mesh
sieve is 75 urn,
A method of representing the particle size data to obtain a rough estimate
of the average diameter is to plot the cumulative percentage over or
under a particular size versus particle size. This is shown in Figure 10,
using the cumulative percentage oversize for the data in Table 4.
As an example, the points on the dexamethasone powder blend curve
were generated as follows. All of the particles of the dexamethasone
600
>;
::r
(ll 30
~
~
~
(')
(\)
l:: ....
[
"tl
~
0....
::g s....
Equivalent Sieve Mesh Numbers (ASTM)
80 60 50 40
I I J' I I I I ,~ ~!_ I I I 100 l:J ~I"i 'II"'''' .,~.... ........ _~
o
... .J:: en
~ 50
Q) .e 60 ... ..!!!
:J
E 70
8
Q)
N 80
.~
~
090
Figure 10 Cumulative frequency plot of the data in Table 4. (0) dexamethasone powder blend,
(11) dried granules prepared by wet granulation before compaction, (D) dried granules after compaction
and disintegration. (From Refs. 1-3.) ~
~
326 Jarowski
powder blend passed through the 60-mesh sieve, therefore 0% are greater
than 251 um (sieve aperture diameter). Only 1% of the particles that
passed through the 60-mesh sieve were retained on the SO-mesh sieve.
Hence, only 1% of the particles are greater than 1'18 um, Of the particles
that passed through the SO-mesh sieve, 8% were retained on the 100-mesh
sieve. As a consequence, 9% of the particles have diameters greater than
152 um (1 plus 8%). The remaining points on the dexamethasone powder
blend curve as well as the curves for the dried granules before compaction
and after compaction and disintegration were plotted in a similar
fashion.
The curves shown in Figure 10 can be used to estimate the average
particle size of the particles. This is determined by drawing a line parallel
to the abscissa which meets the ordinate at the 50% point. Perpendicular
lines are drawn at the points where the horizontal line intersects
each curve. The average particle size diameter is read on the abscissa.
A rough estimate of the average diameter of the particles can be obtained
from the sigmoid curves in Figure 10. A perpendicular to the
abscissa is drawn from the point where the 50% point on the ordinate
meets the curves. By this method, the dexamethasone powder blend has
an average diameter of 84 urn, The dried granules obtained after wet
granulation have an average diameter of 339 um, the dried granules obtained
after compaction and disintegration have an average diameter of
237 um,
Extrapolation of data from sigmoidal curves as shown in Figure 10 is
not recommended. A linear relationship is preferred for greater accuracy.
When the weight of particles lying within a certain size range is plotted
against the size range or average particle size a frequency-distribution
curve is obtained. Such plots are valuable because they give a visible
representation of the particle size distribution. Thus, it is apparent from
a frequency-distribution curve which particle size occurs most frequently
within the sample. Hence, it is desirable to determine if the data will
yield a normal frequency-distribution curve.
A normal frequency-distribution curve is symmetrical around the mean.
The standard deviation is an indication of the distribution about the mean.
In a normal frequency-distribution curve, 68% of the particles lie ±1
standard deviation from the mean, 95.5% of the particles lie within the
mean ±2 standard deviations, and 99. 7% lie within the mean ±3 standard
deviations.
Unfortunately, a normal frequency-distribution curve is not commonly
found in pharmaceutical powders. Pharmaceutical powders tend to have an
unsymmetrical, or skewed. distribution. However, when the data are
plotted as frequency versus the logarithm of the particle diameter, then
frequently a typical bell-shaped curve is obtained. A size distribution
fitting thu? pattern is referred to as a log-normal distribution.
A log-normal distribution has several properties of interest. When the
logarithm of the particle diameter is plotted against the cumulative percentage
frequency on a probability scale a linear relationship is observed.
Selected data from Table 4 is plotted in this manner as shown in Figure
1l.
On the basis of the linear relationship of the data shown in Figure 11,
one can determine the geometric mean diameters and the slopes of the
lines. The geometric mean diameter is the logarithm of the particle size
equivalent to 50% of the weight of the sample on the probability scale.
The Pharmaceutical Pilot Plant 327
9
o <:) 0 00 00
"'" LJ'1 \D r- ee "'0
..-l
ooM
Particle Size CJ,tm)
o
o.., o 0 0 00 o 0 0 00
." Ifl \ll r-. CXI
Figure 11 A plot of oversize cumulative weight percentage versus particle
size. (0) dexamethasone powder blend. (6.) dried granules prepared by
wet granulation. (0) dried granules after compaction and disintegration.
(From Refs. 1-3.)
The values for the geometric mean diameters as shown in Figure 11 are
(1) 88 um for the dexamethasone powder blend. (2) 322 um for the dried
granules prepared by wet granulation. and (3) 250 um for the dried granule
granules obtained after compaction and disintegration.
The slopes of the lines are given by the geometric standard deviations.
which are the quotients of anyone of the following ratios:
84% undersize
geometric mean diameter
geometric mean diameter
84% oversize
16% oversize
geometric mean diameter
geometric mean diameter
16% undersize
Knowing the two parameters. slope and geometric mean diameter. one can
determine the particle size distribution at various fractional weights of the
sample from the graph and reconstruct a log-normal distribution curve.
Mention should be made of the origin of the numbers 84 and 16%. In a
328 Jarowski
normal or log-normal distribution curve, 68% of the particles will lie less
than one standard deviation away from the mean or geometric mean diameter.
To put it another way, 34% of the particles will be larger than the
mean or geometric mean (34 + 50 = 84) and 34% of the particles will be
smaller than the mean or geometric mean (50 - 34 = 16).
In Figure 11, the ordinate data are spaced logarithmically. Equally
good linear data are obtained if semilog plots are used; that is, one plots
oversize cumulative weight percentage versus the logarithm of the particle
size. A comparison of the average particle size of granules obtained from
log-probability and semilog plots is shown in Table 5. In the table, the
average particle size data tabulated attest to the close agreement of the
two procedures. Semilog plots were adopted for the presentation of all
particle size data (Figs. 12-15). The average particle diameters determined
from the figures are summarized in Table 6.
Slugging produced the widest particle size distribution and the largest
average particle diameter (Fig. 14). Forty percent of the weight of the
sample had particle diameters greater than 596 um. Furthermore, only
91% of the sample was collected on the nest of sieves, and thus 9% of the
sample possessed particle diameters less than 44 um, The original dexamethasone
powder blend was completely collected on sieves ranging from
80 to 325 mesh. After compaction and disintegration, the granules with
the largest particle diameter amounted to only 7%. Fifteen percent of the
sample after compaction and disintegration passed through the 325-mesh
screen. Such data are indicative of size reduction occurring during the
two compressional stages. Such size reduction was not found to be
deleterious, since the granulations prepared by slugging yielded satisfactory
tablets. If the concentration of particles with diameters less than
44 um had exceeded 25 to 30%, irregular granule flow and capped tablets
cotUd have been obtained.
Table 5 A Comparison of Average Particle Sizes of
Granules Prepared by Three Granulating Methods and
a Direct-Compression Powder Blenda
Average particle size
(um)
Log probability Semilog
Granulating method plot plot
Wet granulation 322 315
Microgranulation 100 100
Slugging 410 400
Direct compression 122 120
aTh e average particle sizes were obtained from log
probab:ility and semilog plots.
The Pharmaceutical PUot Plant 329
Table 6 Average Particle Diameters of Dexamethasone Powder Blends and
Granules Prepared by Wet Granulation, Microgranulation, and Slugginga
Average particular diameter (11m)
Wet Micro- Direct
Processing stage granulation granulation Slugging compression
Powder blend 88 88 88 120
Granules 315 100 400
Disintegrated 235 90 120 120
tablets
aAlso included are dried granules obtained after compaction and disintegration
and tablets prepared by direct compression.
lOo-r_------------------.,
Particle Size l#.tml
Figure 12 Effect of wet granulation on the particle size distribution of a
dexamethasone powder blend before and after compaction and disintegration.
( 0) dexamethasone powder blend, (A) dried granules before compaction,
(D) dried granules after compaction and disintegration. (From
Refs. 1-3.)
330 larowski
o 0 0
o 0 0 0
U"\ \0 r-- aJ
oo
""
oo
,., o 0 00 00
ll'I \D r--a:I 0'10
r-I
o.&--... ......TIr--r-'-,---,r----T-r--r--H
o...
Particle Size (101m)
Figure 13 Effect of microgranulation on the particle size distribution of a
dexamethasone powder blend before and after compaction and disintegration.
(0) dexamethasone powder blend, (A) dried granules before compaction,
(0) dried granules after compaction and disintegration. (From
Refs. 1-3.)
Microgranulation brought about the least change in particle size distribution
(see Fig. 13) • The slopes of the three lines are similar. The
amounts of material passing through the 325 mesh screen were 0% (dexamethasone
powder blend), 2% (before compaction), and 10% (after compaction
and disintegration). Such increases in fine particles are not
significant.
Granules prepared by wet granulation were much larger than the
original powder blend (see Fig. 12). Ninety-five percent of the sample
was collected on 30- to 120-mesh sieves. Only 5% of the sample passed
through the 325-mesh sieve. After compaction and disintegration a significant
size reduction occurred. The 200-mesh sieve collected 21% of
the granules and the 325-mesh sieve retained 10% of the sample. Prior to
compaction and disintegration none of the granules had been retained by
either of these sieves.
Direct compression produced the least change in particle size before
and after compaction and disintegration (see Fig. 15). The slopes of the
two lines are practically identical. Ninety-seven percent of the particles
The Pharmaceutical Pilot Plant 331
100-------------------------
o a 0 00
o 0 0 00
qo "" \0 r--oo
o
<:>
M
oo
"'I
o a 0 00 00
~ Lf> \0 r- 00 "'0
r-l
O-l--.....-...-._ ......-...-z;.........--r--.......-r--r-"""Y'"T
80
10
90
70 !
~ 60
l1.
~
en 50
~
~ 40
~
~
(5 30
l!l
~ 20
o
Particle Size (p:mJ
Figure 14 Effect of slugging on the particle size distribution of a dexamethasone
powder blend before and after compaction and disintegration.
(0) dexamethasone powder blend, (6) granules prepared by slugging,
(0) dried granules after compaction and disintegration. (From Refs. 1-3.)
in the directly compressible blend were collected on 50- to 325-mesh sieves,
whereas 92% of the particles after compaction and disintegration were collected
on such sieves. The fines which were collected in the receiver
(3 and 8%, respectively) are of concern in this instance, since micronized
dexamethasone accounts for only 0.16% of the sample weight. As a consequence,
classification and poor content uniformity might be anticipated.
None of the original dexamethasone powder blend passed through the
325-mesh sieve. Hence, compressed tablets prepared from granules obtained
by the three granulating procedures could be expected to exhibit
differences in content uniformity and weight uniformity.
Dexamethasone Homogeneity in the
Granules and Tablets
The relative homogeneity of dexamethasone in the granules prepared by
the four procedures is shown in Table 7. Microgranulation produced the
best dispersion of dexamethasone. The fine granule size most likely accounts
for the homogeneity of drug distribution in the microgranulate.
332
70
i... c 60 80- .. a.
i. 50
"ij
3:
.~ 40
~
~ 30 <3 ..hi '2 20 ..>
0
10
0
0 0 00 r--m 0 0 0 00 • li'l \0 1"-00
Particle Size (pm)
Jarowski
Figure 15 Effect of direct compression on the particle size distribution
of a dexamethasone powder blend before and after compaction and disintegration.
(0) direct compression powder blend, (0) dried granules
after compaction and disintegration. (From Refs. 1-3.)
Table 7 Homogeneity of Dexamethasone Distribution in Granules
Prepared by Four Different Methods
Average Maximum Minimum Coefficient
content content content of variation
Granulating method (mg) (mg) (mg) (%)
Wet granulation 0.240 0.265 0.222 5.97
Microgranulation 0.251 0.253 0.247 0.99
Direct compression 0.249 0.267 0.238 3.90
Slugging 0.247 0.252 0.243 1.43
Source: From Refs. 1-3.
The Pharmaceutical Pilot Plant 333
On the other hand ~ wet granulation exhibited the poorest homogeneity.
This is attributable to the larger particle size and the wider distribution
of particle size of the granules before and after compaction. The slugging
method was found to be the second best granulating procedure for producing
homogeneous granules. In the slugging method, dexamethasone
was dispersed in powdered lactose in a suitable blender. After the addition
of a portion of the starch and acacia the blend was lubricated with
magnesium stearate and slugged under high pressure. Unlike wet granuIation
, the slugging method did not call for SO many processing steps,
such as wetting ~ wet screening, drying, dry screening~ lUbricating ~ and
compressing. Moreover, bonding of drug excipient had occurred during
the slugging operation. Subsequent granulation did not result in segregation
of dexamethasone because of such bonding. Thus, the homogeneity
of drug distribution in granules is not only dependent on granule size
distribution and/or drug particle size, but also is dependent on the granu1ation
method. The coefficient of variation is also high for the blend prepared
for direct compression. This is attributable to the great difference
in particle size between the active ingredient and the exeipienta. The
dexamethasone was micronized (size range 1-5 llm); the directly compressible
lactose had an average particle diameter of 120 um.
Control charts for the weight variations of dexamethasone tablets are
shown in Figures 16-19. The control chart is a useful measure for
process control. It is based on standard deviation or range (R).
Each of the circled .20ints in Figures 16-19 represent the average
weight of five tablets (X) taken at s-mtn intervals up to 40 min. Thus, a
total of 40 tablets were individually weighed. The highest and lowest
weight of each time interval Jlre also shown in the four figures. The
average for the 40 tablets (jC) is represented by the solid horizontal line.
The horizontal broken lines above and below the solid line represent 3
standard deviations from the mean. For a normal curve distribution of
Time (minutes)
5 10 15 20 25 30 35 40
158
157
156
155
t1' 154
! 153 .,
152 .c:
t1' .... lSI III
~ 150
149
148
147
146
x - A R 2
Figure 16 Control chart for the weight variation of dexamethasone tablets
prepared by the wet granulation method. (From Refs. 1-3. )
334
T1me (minutes)
157
156
155
D'>
E 154
+I 153
~ 152 .J.._--I-~~~:::::~~:::::::~~:::::=G~~~--~,4:""+X
~ 151
150
149
Jarowski
Figure 17 Control chart for the weight variation of dexamethasone tablets
prepared by microgranulation. (From Refs. 1-3.)
weights. this means that 99.73% of the tablets in the batch will weigh
within the range represented by the upper and lower limits; for a skewed
curve distribution. 95% or more of the tablets wID weigh within the upper
and lower limits represented by the horizontal broken lines.
The standard deviation is used as the measure of spread for almost all
industrial frequency distributions. It is the root mean square deviation of
Time (minutes)
162
161
160
159
158
~ 157
tl'
,5 156
+I 155
~
~ 154 &153
152
151
150
149
148
5 10 15 20 25 30 35 40
Figure 18 Control chart for the weight variation of dexamethasone tablets
prepared by slugging. (From Refs. 1-3.)
The Pharmaceutical Pilot Plant 335
40 35
T~me (m~nutes)
15 20 2S 30 10 5 161
160
159
159
157
156
155
lS4'1-_--:I:-------~I11""""---_1't_---~tt_X
153
152
~ 151 !' 150
-4.1 149
-§, 14 9
·rot &147
146
145
144
143
142
141
140
139
Figure 19 Control chart for the weight variation of dexamethasone tablets
prepared by the wet granulation method. (From Refs. 1-3. )
the readings in a series from their average. The sample standard deviation
is obtained by extracting the square root of the sums of the squares
of the series from the average, divided by the number of readings. This
is represented symbolically as follows: j -2 -2 -2 Standard = (Xl - X) + (X2 - X) + (X 3 - X) +
deviation n
... + (X - X)2
n
X :: value of each reading n
X :: average value of the series
n :: number of readings
The standard deviation of the data presented in Figure 16 is calculated
as follows: X, the average weight for the 40 tablets = 152.4 mg
336 Jarowski
Time (min)
5 10 15
(X - X) (X - ib 2
(X - X) (X - JC)2 (X - X) (X - Jb 2
+2.8 7.84 -0.4 0.16 +2.8 7.84
+1.6 2.56 +3.3 10.89 -1.6 2.56
+3.2 10.24 +0.1 0.01 -2.6 6.76
+1.2 1.44 -2.1 4.41 -3.1 9.61
+0.4 0.16 +0.2 0.04 -0.2 0.04
Time (min)
20 25 30
(X - X) (X - Jb
2
(X - ,b (X - X) 2 (X - X) ex - X) 2
-2.4 5.76 +2.8 7.84 -0.2 0.04
-1.8 3.24 -4.5 20.25 +5.1 26.01
-2.3 5.29 -1.2 1.44 -1.0 1.00
-2.6 6.76 +0.6 0.36 -1.0 1.00
-1.5 2.25 +3.6 12.96 +1.0 1.00
Time (min)
35 40
eX - X) (X - 5(2) (X - X) (X - 5(2)
-1.2 1.44 +0.1 0.01
+3.1 9.61 +4.2 17.64
-2.9 8.41 -2.8 7.84
+1.6 2.56 -1.8 3.24
-3.0 9.00 +2.6 6.76
S d d d . t' / 226.27 tan ar evia IOn = 40 = .; 5.657 = ±2.38
Standard deviation of
the sample average
standard deviation of the lot =
In
= 2. 38 = 2.38 =: 1. 06
15 2.24
Three standard deviations of the sample average = 3 x 1.06 =: ±3. 18
The Pharmaceutical Pilot Plant 337
From the calculations shown, the upper line of the control chart (see
Fig. 16) is set at 152.4 = 3.18 = 155.58; the lower line of the control
chart is set at 152.4+ 3.18= 149.22.
Obviously, it would be tedious to gather a series of samples of small
size, determine the values for central tendency and spread for each of
these samples, and then go through the laborious calculations that are involved.
However. statisticians have simplified the calculations by preparing
a table of constants (Table 8) [6].
Through the use of the following equation the control limits can be
cslculated much more easily (R = average of the range values):
Upper broken line (Fig. 16) == X + A2R
Lower broken line (Fig. 16) =X - A2R
Thus. for samples consisting of five units at each time interval the
value for A2 has been calculated to be 0.577 (see Table 8). Substitution
of the factor into the above equations is shown at the bottom of Figure 16.
Note the close agreement of the limit values derived by either method of
calculation,
Table 8 Factors for Computing
Control Limits When Range is Used
as a Measure of Spread
Number of
observations
in sample
(n)
Factor for
control limits
(A
2)
2 1.880
3 1.023
4 0.729
5 0.577
6 0.483
7 0.419
8 0.373
9 0.337
10 0.308
11 0.285
12 0.266
13 0.249
14 0.235
15 0.223
Source: From Refs. 1-3.
338 Jarowski
Examination of Figures 16-19 reveals that the tablets prepared by the
microgranulation procedure were most uniform in weight (Fig. 17). This
conclusion can be drawn from the following considerations: (1) the average
range value. 3. 85. is the smallest for the four granulation procedures;
(2) the average value for the five tablets at each time interval always fell
between the upper and lower limits; and (3) the weight spread between
the 40 tablets was least for the microgranulation procedure (156.6 - 150 =
6.6 mg).
An interesting correlation can be seen between the average range values
R from the four control charts (see Figs. 16-19) and the average
particle size of the granules used in preparing the tablets (see Table 5).
On the basis of the data in Table 5. one can conclude that the lower the
average particle size of the granules the lower the average range value.
Thus. the four granulating procedures rank in the following order with
respect to increase in average particle size and R value: micro granulation •
1001-1, 3.85mg; direct compress., 120 u . 3.9mg; wet gran .. 31511,
5.35 mg : and slugging. 400 u . 6.6 mg. Thus weight variation is a function
of the average particle size of the granules.
Content uniformity was determined for the tablets prepared by the four
manufacturing procedures. Assay data collected on 40 tablets were treated
in a fashion similar to that shown in Figure 16. Control charts for content
variation of dexamethasone tablets are shown in Figures 20- 23. Each
point on these figures represents the average spectrophotometric absorbance
(at 239 nm) for five tablets individually assayed at 5-min intervals up to
40 min.
Microgranulation and slugging can be seen to be the best procedures
for preparing dexamethasone tablets possessing good content uniformity
(see Figs. 21 and 22). Their average range values of 0.017 and 0.016.
respectively, were much lower than those values for the wet granulation
and direct-compression procedures.
Confirmation of these conclusions was derived from assays performed
on 50 individual tablets selected at random from each batch of tablets.
The coefficients of variation (standard deviation/average) were significantly
lower for tablets prepared by the micro granulation and slugging procedures
(Table 9).
Table 9 Coefficient of Content Variation for Dexamethasone Tablets
Prepared by the Four Manufacturing Proceduresa
Number
Manufacturing of Average Standard
procedure assays absorbance deviation
Wet granulation 50 0.430 0.0248
Microgranulation 50 0.452 0.0092
Direct compression 50 0.450 0.0182
Slugging 50 0.455 0.0129
aContents expressed in terms of absorbance.
Source: From Refs. 1-3.
Coefficient
of variation
(%)
5.77
2.04
4.04
2.84
The Pharmaceutical Pilot Plant 339
5 10 15 20 25 30 35 40
41
tl
I::
III . 435 s 430+.-------+------+-----++---t-X
~ 425
+J 420
I::
~ 415
g 410
o
405
400
395
390
385
380
375
370
Figure 20 Control chart for the content variation of dexamethasone tablets
prepared by the wet granulation method. (From Refs. 1-3.)
Time (ml.nutes)
5 10 15 20 25 30 35 40
47
QI 47
l)
I:: 46 III X A}l .Q 46 +
~
0
1/1 45
~ 45 X
+J 44 I:: A2R Q.I 44 -
+J
J::
0 43 u
43
Figure 21 Control chart for the content variation of dexamethasone tablets
prepared by microgranulation. (From Refs. 1-3.)
340
Tlme (mlnuteS)
Jarowski
5 10 15 20 25 30 35 40
QJ
Uc
.3
+ A2R "" 0
III
~ X
.j.J
C
QJ - Ai .j.J c0
u
Figure 22 Control chart for the content variation of dexamethasone tablets
prepared by slugging. (From Refs. 1-3.)
Tlme (minutes)
5 10 15 20 25 30 35 40
48
48
475
47
46
~B 46
~ 45
~ 4SO+----....l~~:::::~-----.....-_t-~~--;_ X
III
~ 445
44
~ 43
, -, "- .- _"--'11" ...-,~-. --_. ..
-.~
, ~
~_." .# ~. ;. "- --.-" <, ~'-"~'
-»
~ - , " TABLET ~ TABLET
_.-. ..... _- -... ~--- .... "". ./"'" -, PRESS ....- _..- _...... ._--.... PRESS "lit"<,, 4- - .....~- ...... -- ... ./
Figure 16 Air handling in compression area.
PACKAGING. Some tablet-filling machines are designed with a selfcontained
vacuum system that returns the air, filtered through an absolute
filter, back to the packaging area. There should also be some provisions
made at the cottoning stations. If the filling machine being used does not
have an air-filtering system, dust pcikups of approximately 300 cfm
should be provided at the hopper station and 50 CFM at the bottle chute.
C. Humidity/Temperature Controls
From the standpoint of both product protection and employee comfort,
careful consideration must be given to humidity and/or temperature controls.
Unless otherwise indicated, conditions of 45% RH and 70°F are
generally adequate for critical manufacturing areas such as compression
and coating. Comfort conditioning should be provided for weighing,
400 Connolly. Beretler, and Coffin-Beacn
granulation, compression. and packaging operations. However. temperature
controls should be such that little or no variation will be caused by
external ambient temperatures. Thus, comfortable working conditions are
achieved and there should be no impact on characteristics of in-process
materials, such as granulation. raw materials. etc. Warehousing operations
should have some adequate type of ventilation, particularly in areas
of high storage. either pallet racks or pallet-to-pallet-type storage. The
ventilation could be provided by large roof fans to circulate air. In addition,
some form of supplemental air heating. such as hot-air blowers.
should be provided fol' cold areas. such as shipping or receiving docks.
D. Water Systems
Although water systems are not covered to any great extent in the CGMPs,
CGMP regulation 211.48 states that the supply of potable water in a plumbing
system must be free of defects that could contribute contamination to
any drug product [6]. Thus, an effective water system is a necessity.
In recent years, various techniques have been developed for producing
the high-quality water required in ever-increasing quantities in pharmaceutical
manufacturing operations. These include ion-exchange treatment.
reverse osmosis. distillation, electrodialysis. and ultrafiltration. In fact.
the United States Pharmacopeia (USP) XXI defines purified water, USP.
as water obtained by these processes or other suitable processes. Unfortunately,
there is no single optimum system for producing high-purity
water. and selection of the final system(s) is dependent on such factors
as raw-water quality. intent of use. flow rate, cost. etc. [1].
In the pharmaceutical industry, the classes of water normally encountered
are well water. potable water. USP (which complies with Public
Health Service Drinking Water Standards), purified water. USP, and specially
purified grades of water. such as water for injection. USP. or FDA
water for cleaning and initial rinse in parenteral areas (as defined in the
CGMPs for large-volume parenterals). For purposes of this chapter. only
the first three classes will be addressed because these are the classifications
encountered in processing for solid- and semisolid-dosage forms.
Well Water
Well water, as the title indicates, is water drawn directly from a well.
The water may not be either chemically or microbiologically pure because it
is untreated. Therefore. the use of well water should be restricted to
nonmanufacturing operations. such as lawn sprinklers. fire protection systerns,
utilities, and the like.
Potable Water
Potable water, USP, is city water or private well water that has usually
been subjected to some form of microbiological treatment. such as chlorination.
to meet the United States Public Health Service Standard with respect
to microbiological purity. Potable water is fit for drinking. and is
generally used in processing operations for cleaning and sanitation purposes.
Periodic monitoring of use points should be conducted to ensure
adequate residual chlorine levels and the absence of microbial contamina
Tablet Production
tion. If necessary t additional chlorine should be added to the water
supply as it enters the plant, or a suitable flushing program should be
implemented to ensure adequate chlorine levels at point of use.
401
Purified Water
Purified watert USP, is usually prepared from water that meets the potable
water standard. Purified water is treated to attain specified levels of
chemical purity and it is the type of water used in most pharmaceutical
processing operations. Bowever, this class of water is not without problems
in that the requirements of no chlorides presents special concerns
from a microbial contamination standpoint.
Purified water. USP. generally is produced by deionization or distillatton
, although reverse osmosis or ultrafiltration systems might be utilized
if the required chemical purity could be achieved. As a starting point in
these processes, water softening or activated carbon filtration frequently
is employed as a pretreatment process to remove calcium and magnesium
ions or chlorine and organic materials. Ion exchange and demineralization
through deionization is a very common method of obtaining the purified
water, USP t used in the pharmaceutical industry.
Water-Treatment Equipment
Deionization equipment should be sized to ensure frequent regeneration and
a recirculating system should be installed on the unit that approaches the
rated flow of the deionization unit. Procedures should be written to ensure
that all water-treatment equipment is properly operated. monitored.
maintained. and sanitized on a regular basis.
WATER FILTRATION. Water filtration generally is approached on the
basis of two major considerations: prefiltering to prevent large particulates
from entering the system and microfiltering to remove bacteria. Prefilters
are generally the replaceable cartridge type with porosities ranging as
high as 25 u, Microfiltering is usually accomplished with a O.2-11 absolute
filters, which will remove most bacteria.
A proliferation of filters within the system should be avoided, since
what might be implemented as a protective measure could readily develop
into a problem wherein retained bacteria are given the opportunity to
multiply. In any system employing filters for the control of bacteria t the
filters and all of the downstream piping must be effectively sanitized on a
regular basis. There should be procedures written for filter inspection,
replacement, integrity testing, and monitoring on a scheduled basis.
SAN ITIZATION PROCEDURES. Sanitization is best accomplished
through several methods. After periods of low usage of water, the system
should be flushed. with a supply of water that has residual chlorine.
Periodic hyperchlorination also is recommended. Effective microbial control
can be maintained by storing water at 80°C. However, this approach
is expensive and presents some hazards to personnel and material. Ultraviolet
radiation may be used. but it has limited application because of the
many factors which can reduce its effectiveness. written instructions
should be developed for sanitization procedures for water-treatment equipment
on a regular, prescribed basis.
402 Connolly. Berstler. and Coffin-Beach
E. Plant Pest Control
The CGMP regulation 211. 56a states that
Any building used in the manufacture, processing, packing, or
holding of a drug product shall be maintained in a clean and sanitary
condition. Any such building shall be free of infestation by
rodents, birds, insects, and other vermin (other than laboratory
animals). Trash and organic waste matter shall be held and disposed
of in a timely and sanitary manner [6].
A pest-control program should be developed that will ensure the integrity
and quality of products produced and comply with existing legislation.
The program should be written to include a general statement of
purpose and the company position. Effectiveness of the program should be
assured by defining the plant individual with overall responsibility for the
program and how the responsibility will be carried out. The training and
experience requirements of the extermination staff should be delineated,
whether they be in-house or SUbcontracted personnel. Assistance in supporting
the program may be gained from other plant personnel by their
pointing out problem areas.
The program should be issued to Une management personnel and
periodically updated in order to keep the program current. The program
should contain a list of approved pesticides to be used in the plant. Individual
sheets should be prepared for each specific item of use [5].
Basic information should be spelled out as follows:
1. Trade name of the pesticide
2. Classification
3. Type of action
4. Chemical name and concentration of active ingredient
5. Effective for:
6. To be used for:
7. Area of usage
8. Mode and frequency of application
9. Toxicities and any specific toxic symptoms, if known
10. Status of government approval
11. Specific restrictions and cautions
The development of sheets as indicated above will serve a twofold purpose.
First, the sheets can be SUbjected to approval by the plant safety
organization to determine if the materials comply with the Occupational
Safety and Health Administration (OSHA) requirements and the requirements
of other state or local agencies. Second, the sheets would also
facilitate compliance with CGMP regulation 211. 56c, which states the
following:
There shall be written procedures for use of suitable rodenticides J
insecticides, fungicides, fumigating agents. and cleaning and
sanitizing agents. Such written procedures shall be designed to
prevent the contamination of equipment, components, drug products
containers, closures, packaging, labeling materials, or drug
products and shall be followed. Rodenticides, insecticides, and
Tablet Production
fungicides shall not be used unless registered and used in accordance
with the Federal Insecticide, Fungicide, and Rodenticide Act
(7 U.S.C. 135) [6].
403
Written records of regularly scheduled inspections and preventive treatments
should be maintained. Emergency or special service! should be documented
specifying the type of problem encountered, the service rendered,
effectiveness of the treatment, and any follow-up that might be required.
The program should also specify when production interruptions might
be necessary, either due to the presence of a specific pest or to avoid
possible contamination during the treatment to exterminate a pest.
All manufacturing areas should be constructed using nonporous materials
on the walls and floors. Any protrusions such as pipes and electrical
boxes should be minimized. Space should be allocated carefully to
provide sufficient rooms for aU operations. There should be adequate
lighting and the areas should be remote from any openings to the outside.
Care should be taken that adequate training in understanding CGMPs be
given to all personnel. Outside contractors must also be trained and understand
CGMPs before embarking on any construction or remodeling efforts
having to do with pharmaceutical manufacturing.
VI. EQUIPMENT SELECTiON
A. Granulation
There are a multitude of equipment manufacturers, each with a specific advantage.
But, more importantly, the initial formulations corning from the
scale-up laboratory should be done in equipment that most closely mimics
the final processing equipment that will be used in production. The success
or failure of a manufacturing process depends on the time and effort
put into the formulation design and equipment selection.
Selection of equipment for dry blending. dry granulation, or wet
granulation is a process that must start at the time the formulation is first
conceptualized. There are numerous types and designs for each processing
technique.
Dry Blending
The first choice for dry mixing is in tumble-type blenders, and here, as
elsewhere in this section, an effort will be made to identify the more commonly
used designs and some of the manufacturers of that equipment.
This information is intended as a guide and is not proported to be a complete
listing. Figure 17 shows some configurations of a double-cone blender,
both straight and offset.
Figure 18 illustrates both the conventional twin-shell or Vee blender,
and the newer cross-flow or short-leg Vee blender. Double-cone and
twin-cone blenders are supplied by Gemeo, J. H. Day Corp •• and Patterson
Kelly Co., among others.
The important considerations to take into account when developing a
formulation for these type mixers are deagglomeration of the raw materials
prior to charging the blender, not exceeding the rated capacities which is
approximately 60% of the total capacity. particle size uniformity, dry and
humid conditions, reducing vibration. and avoiding over mixing.
404 Connolly, Berstler, and Coffin-Beach
(a) (b)
Figure 17 (a) Double-cone blender, (b) slant-cone blender.
Ribbon blenders of the type shown in Figure 19 may also be used for
dry blending. However, care must be exercised here when formulating
tow-dosage products. Dead spots such as discharge ports must be cleared
and the materials recycled. Samples must be taken from multiple locations
when validating this process. Some ribbon blender suppliers are Marion
Mixers, Ine; , J. H. Day Corp., Charles Ross & Son Co., and S. Howes
Co., among others.
(a) (b)
Figure 18 (a) Vee blender, (b) short-leg vee blender.
Tablet Production
Figure 19 Ribbon blender.
405
Dry Granulation
Dry granulation is becoming more popular as equipment improvements continue
to accelerate. Moisture-sensitive actives and some heat-sensitive
actives can now be prepared as efficiently as less sensitive products. A
combination of roll compaction and sizing coupled with improved cleanabllity
of the equipment have encouraged formulators to look at this method more
closely than in the past.
Figure 20 shows a typical design for roll compaction. Again, equipment
manufacturers have realized the importance of being able to precisely
control feed rates, compaction force, and particle size, and have
carefully addressed these parameters when scaling from lab equipment to
production equipment. Some roll compaction equipment suppliers are
Vector Corp., Fit zpatrick Corp., Alexanderwerke, and Bepex, among
others.
Wet Granulation
The wet-granulation technique is benefiting from improved processing
equipment. High-shear granulating equipment is being developed and improved
upon. Equipment manufacturers are studying changeover times,
clean up, end point measurement, ease of discharge. and many more points
to improve quality and productivity. Granulator/dryer combinations are
also becoming more available. Fluid-bed drying, microwave drying J and
vacuum-drying equipment are rapidly replacing the traditional tray
dryers.
B. Compression
Introduction of the high-speed, computer-controlled tablet press has had a
profound impact on the industry. With capacities of over 800,000 doses/hr.
presses no longer need to be dedicated to a single product. Many products
may be manufactured on one machine to utilize excess capacity and to
justify the high costs. Depending on the manufacturing strategies of a
particular company, different criteria would be used to select their equipment.
Large production rate capacities become important as the batch size
406 Connolly. Berstler, and Coffin-Beach
Figure 20 Roll compactor. (Courtesy of Alexanderwerke.)
increases and the number of different products decreases. Cleanability
and change-over time become important as the batch size decreases and
the number of products increases. Some manufacturers claim as little as
4 h for a complete changeover to a different product.
Theory of Computer-Controlled Presses
If every particle and granule were the same size and shape. and if every
die cavity were filled with exactly the same amount, and if every punch
Were exactly the same, etc.. the compression force for each tablet produced
would be identical from die cavity-to-die cavity, from revolution -torevolution,
and from the beginning to the end of a batch. Unfortunately.
the particle size distribution may vary slightly from die cavity-to-die
cavity, from revolution-to-revolutton , and from the beginning to the end
of a batch. Therein lies one of the sources of variation of weight and
hardness from tablet-to-tablet.
Tablet Production 407
Consider two identical die cavities, A and B. A is filled with small
particles, whereas B is filled with larger particles. While both die cavities
are filled, the weight of the powder in A is heavier than the weight of
the powder in B. The compression force will be larger for A than for B
as they pass under the same compression rollers. Obviously, the tablet
resulting from A will be heavier than the tablet resulting from B. The
compression forces can therefore be directly correlated to tablet weights.
The basic assumption used by all computer controlling systems is that
all significant variations in compression forces are resultant from the actial
weight of the material being compressed, hence tablet weight. The
target tablet weight can be correlated to a target compression force that is
determined during the setup operation. This target compression force can
then be used to control the compression operation.
During the prestart operation, a target compression force will be determined.
As the press begins production, it will warm up and its operating
condition will change. As the punches warm, they will elongate
slightly. As the electronic components warm. their signals may vary
slightly. The sum of the press's varying operating conditions is a drift
in the compression force/target weight correlation.
Until recently, it was up t9 ~ the operator to verify that the target compression
force correlated to the target tablet. If this was found not to be
the case, an adjustment needed to be made before compression could continue.
With the Introduction of tablet measurement/feedback systems.
control is taken a step further and the loop is closed. Compression forces
are periodically referenced to actual produced tablet measurements. All
necessary adjustments, including weight and hardness adjustments, are
then performed automatically and the compression force/tablet weight correlation
is adjusted automatically as necessary.
Instrumentation Strategies
There are currently two basic strategies used by the many tablet manufacturers
to computer-control their presses. The first is to set up the
press to produce tablets with the desired weight. hardness, thickness, friability.
etc. The next step is to convert the compression force signal to a
reference value. This reference value corresponds to the target tablet, and
is usually displayed as an average per revolution of each station's actual
compression force. This reference value must be established at the start
up for each batch and is then used to control the tablet weights. When
the average force value drifts from the reference value and reaches an
adjustment value, a signal is sent to the weight-control motor to adjust the
filling depth appropriately. Some manufacturers provide for the duration
of the weight-control signal to be set so that the adjustment brings the
force value back to the reference value. Compression forces of individual
stations are measured, and if their force values are within a set reject
range, the tablet is accepted as a good tablet. If the force value is outside
the reject range, the tablet is rejected as an out-of-spec tablet.
The second strategy, which most tablet manufacturers are using or
moving toward, is the use of internal calibrations so that actual compression
forces are measured and displayed. The benefits include a simplfied
computer setup and a standard/common base by which batches of the same
product can be compared for compression characteristics and statistical
analysis. The adjustment range settings and the reject range settings
408 Connolly. Berstler, and Coffin-Beach
initiate the same responses as they do in the first strategy. The difference
is that they are set and displayed as actual compression forces.
Rejection Strategies
The rejection mechanisms are also of two basic strategies. pneumatic reject
and mechanical reject. The pneumatic reject relies on a timed burst
of air to direct a bad tablet to a reject chute. The disadvantages with a
dusty product are clogging of the jets and inducing airborne dust particles.
Also, depending on the speed of the press, up to six or more
tablets may be rejected for each bad tablet. The advantages are no
moving parts and the timing is fixed and not varied owing to table speed
changes.
The mechanical reject is usually a gate or arm that swings out to divert
the bad tablet to the reject chute. The timing is critical and must
be set correctly. As the table speed changes, the timing may need to be
changed. Also, tablets of different sizes and shapes will require different
timings and gate settings. Abrasive dusts can interfere with proper rejection
by causing sluggishness. On the other hand, mechanical mechanisms
can be set to reject only one tablet, regardless of speed.
Induced Die Feeder (IDF) Strategies
Another distinguishing feature of high-speed machines is the induced die
feeder (IDF). The IDF is basically a mechanism to ensure adequate and
uniform filling of the die cavities as they pass the feeder at high speeds.
It uses a combination of paddles which pushes or plows the powder over
the dies. Virtually every combination of paddle placement and blade configuration
is represented by some manufacturer. Different IDF configurations
and speeds may induce different compression characteristics of the
same material. Because the IDF agitates the powder, and its speed can
be varied, some additional mixing may be considered to occur in the feeder
for some formulations. This can sometimes become a source of problems
rather than a solution.
Power Delivery Systems
Some manufacturers offer a powder feed controller which maintains a constant
level of powder that feeds the IDF. The effects of the sudden increase
of weight by adding a scoop full of powder to a hopper can be
seen very clearly on computer monitoring/controlling devices that have
graphical displays. These effects are eliminated by the powder feed controller,
and help to create more uniform products by eliminating the
powder level variable.
Future Trends
Most press manufacturers currently have on the market presses and controlling
systems which provide a complete loop for controlling all of the
tablet parameters. Tablet sampling and checking systems take a sample of
tablets at specified intervals, automatically perform weight, hardness, and
thickness measurements, perform statistical calculations, and feedback the
appropriate signals to the press to adjust the weight and/or hardness.
These signals include readjusting the target compression force and the
ol:lo. o
co
~an~:fdcturer Elizabeth Kilian Kikusui Korsch Manesty Stokes
Hata
:-1odel AP-38-SC T300 Libra Pharmdpress e xov apr ss s -15 Stokes 454
336
Station :'-:0. 38 32 36 36 .,I5 .,I5
Output/Hr: I
i ~1ax (:,.1) 160 2-10 216 216 - 222 270
I ~in(Yl) 34.6 9.6 21.6 71.3 5J 96
:vrain Comp 9 80 kX 8 8 6.5 6
Pre Comp 3.5 28 k~ 8 2 - 2
Computerized: I ~lonitoring y Y y Y Y Y
I
Control by y y y y y y
Porce
Control by y Y Y y y :;:
I weight
Figure 21 General comparison of currently available tablet presses.
410 Connolly, Berstler, and Coffin-Beach
window of limits. The target compression force is therefore being constantly
verified against actual tablet parameters.
Some manufacturers currently offer systems where all of a product's
parameters are stored on a disk and the operator loads the information
in a computer. The computer will tell the press where the initial settings
are, produce some tablets for automatic testing. make any necessary adjustments
until the tablets are at target, and then begin production. Figure
21 is a general comparison of some of the tablet presses currently
available and is not all inclusive.
VII. PERSONNEL
A. Training
operators
This group includes the mixmg and granulating operators and the compression
operators. The training of the mixing and granulation operators
should include the following:
1. Cleaning procedures
2. Familiarization with and ability to identify the codes and names of
the raw materials being used in the product being mixed
3. Proper label control and reconciliation
4. Handling, usage, and operation of equipment in area
5. Proper handling of raw materials in each operation
6. Importance of precise mixing times and geometric dilutions
7. Proper labeling and handling of mixed materials
8. Quality assurance and validation procedures
9. Training to observe and look for foreign matter in the raw
materials
10. Product reconciliation
11. Ability to check mathematics required by the formula
The training of the compression operators should include the following:
1. Cleaning procedures
2. Ability to identify and distinguish from quarantine and released
materials
3. Proper label control and reconciliation
4. Handling, usage, and operation of equipment in the area
5. Proper handling and hopper loading of mixed material
6. Proper handling of bulk tablets produced
7. Quality assurance and validation procedures
8. Training to observe and look for foreign matter in the mixed material
and bulk tablets
9. Product reconciliation
10. Ability to check mathematics required by the formula
Mechanics
Mechanics have specialized training in addition to the same training as the
operators. During their setup operation, mechanics operate the machine
Tablet Production 411
and, also during periods where there are no setups to be performed, they
may be asked to operate a machine. The training of mechanics, in addition
to operator training, should therefore include the following:
1. Cleaning procedures for tooling and machinery
2. Breakdown and setup of machinery
3. Complete operation of the machines
4. Proper handling of the tooling
5. Theory behind computer-controlled tablet presses
6. Setup and control of computerized units
7. Maintenance and repair of machinery
8. Troubleshooting of machinery
Supervisors
Supervisors need to have the necessary training and background to lead
and direct their operators and mechanics. They must have an intimate
knowledge and preferably some hands-on experience of the operators' and
mechanics' jobs. In addition, supervisors must also possess the supervisory
skills necessary to maintain the high standards demanded by the
pharmaceutical industry.
In addition to the job-specific responsibilities outlined above, all manufacturing
employees must be versed and trained in CGMPs and in the appropriate
standard operating procedures (SOPs) governing their area.
VIII. ROLE OF MANUFACTURING
A. Marketing Support
Marketing is usually the only organizational link to the customer, and any
feedback on the sales impact of quality and delivery COmes through as
marketing requests. Also, the direction that applied R&D takes should be
driven by the marketing function. Marketing should encourage the development
of products that would generate the highest potential revenues
with the greatest margin. Each product developed or produced should fit
into an overall marketing plan. Without insight to the marketing plan, the
R&D and Production Departments might question the wisdom of their directions,
creating unnecessary, undesirable boundaries between departments.
The importance of marking to an organization is obvious. However,
the importance of accurate, timely information from marketing to various
other departments cannot be stressed enough. The role of manufacturing,
therefore, is to support the marketing function. Successful support might
be defined by improving quality, reducing manufacturing lead times, lowering
costs, and hastening manufacturing response to changing demands.
The role of all other departments in the organization is to support, in one
way or another, manufacturing [7].
B. The Production Plan
A finished goods requirements plan should drive the master production
schedule, which provides a detailed action plan from placing purchase
orders to final packaging schedules.
412 Connolly. Beretler, and Coffin-Beach
The master schedule is the operations statement for production and related
activities. It is a positive commitment to perform certain activities
within the required time span. In the packaging area, the master schedule
states that the Packaging Department will package X amount of a put-up
during week N on day Y. In order to fulfill this commitment, the Quality
Control Department must release the bulk during week N-l. For quality
control to meet its commitment, the Manufacturing Department must have
the bulk produced in time for the former to perform its functions. In
order for the Manufacturing Department to produce the product, approved
materials must be available X number of weeks before the manufacturing is
to be completed. the number of weeks being dependent on the manufacturing
cycle time (including dispensing time). Having approved material
available to manufacturing requires that quality control perform their
analysis in the stated time span for quality control lead time. For this to
happen the material must be received as sCheduled. The Purchasing Department
must, therefore. place the order X weeks before the material is
to be received, where X is the vendor lead time. Purchasing must also
follow up with the blender to assure the material will be received on time.
If the material is rejected, purchasing must issue a replacement order. In
order for purchasing to issue the order on time, production planning must
requisition the material at least 1 week before the order is to be placed.
Given the large number of components and materials involved in pharmaceutical
production operations, the task of manually performing the above
"time-phasing" routine would be virtually impossible. This time phasing of
the production, quality control, and purchasing processes is one of the
functions performed by materials requirement planning (MRP).
C. Material s Requ irement Plan
Materials requirement planning starts with the master schedule. Using the
bill of materials, lead time, capacities, and other product-related database
information, together with information on inventory, open production orders,
and open purchase orders, the materials requirement planning system will
determine the quantity of materials and the date they are needed for each
phase of the production process. The logic used in making the determination
is basically the same as that described under the master SChedule.
The master schedule states which put-ups are scheduled to be packaged
during the weekly time periods. By applying the bills of material for the
put-ups to the quantities to be packaged, the system can determine the
quantities of packaging components and bulk that are needed. Working
within the established lead times, the system will show a demand for the
items in the correct time segment. In turn, the system will plan replacement
orders to maintain specified levels of inventory. The planned order
will be shown as being available in the time frame needed and will be
placed "on order" by considering the purchasing, production, and quality
control lead times. The procedure is carried out for each phase of production
and the release of a planned order at one level generates requirements
at the next level. For example, if finished products were assigned
to 0 level code, items used to produce finished stock would be assigned a
1, products used to produce bulk would be assigned a 2. and so on until
the purchased materials are reached.
Tablet Production 413
The MRP lends itself well to computerization and can perform other
functions, such as alerting management to items with excessive inventory.
It can suggest corrective actions (such as delaying or canceling open
orders) and can predict inventory levels and investments for each time
period in the planning horizon [2].
D. Production Scheduling
All of the elements previously described could be categorized as planning
elements. From the planning phase one must move into the execution
phase, with the transition being supplied by the master schedule.
The development of a production schedule is of prime importance because
this is the basic means for monitoring production activities. Progress
can be Checked and results reported based on the production or planning
period. The production planning schedule forms a basis for decision
making during the production cycle. Production scheduling is one of the
most detailed and demanding tasks in the organization.
Scheduling may be performed in an adequate manner either manually or
mechanically, depending on the size and scope of the organization. Any
scheduling system requires basic input data from the production plan consisting
of What?, When?, and How many? Regardless of the circumstances,
monitoring production activities with the production schedule becomes the
key to successful fulfillment of the production plan. This is the method by
which control is exercised over the production operation. A proper schedule
can optimize production and inventory costs by proper sequencing of
order quantities and time phasing. No matter how small the organization.
the development of a sound production schedule is a tool which cannot be
neglected. The production schedule is the cement which creates the
foundation for an effective production organization that can meet sales
demands.
IX. INDUSTRY OUTLOOK
The advances over the next decade in equipment, design, instrumentation.
and process control techniques will certainly be significant. The nature
of these advancements are difficult to predict. Certainly, computer-integrated
manufacturing will come into its own in the pharmaceutical manufacturing
industry as a whole and most assuredly in tablet production.
Statistical process-control techniques will become widespread throughout
the industry. Numerous data-collection devices and computer systems are
currently available and in use as a means of implementing statistical process
control. Computer-controlled tablet presses and automated material
handling devices are available that virtually remove the operator from the
need to control the operation.
The general state of the industry is that there exists many islands of
automation. The challenge for the future is to first integrate these
islands with a computer network so that process data can be easily collected,
and so that the inventory position and scheduling can be optimized
on a real-time basis. The second challenge is that of validation of these
systems to the complete satisfaction of the individual manufacturers and of
the U.S. Food and Drug Administration.
414 Connolly, Berstler, and Coffin-Beach
The authors stated in the introduction to this chapter that "our-s is an
evolving technology-based science that requires canstant attention. II With
this in mind, the professionals of our industry must stay abreast of current
practices and recognize the emerging trends. Only in this way, will
the industry continue to be profitable and provide an important service in
an increasingly competitive world market.
APPENDIX: LIST OF SUPPLIERS
A C Compacting Presses
North Brunswick, New Jersey
Aeromatic, Inc.
Towaco, New Jersey
Alexanderwerke
Remscheid, West Germany
Bepex Corp.
3 Crossroads of Commerce
Rolling Meadows, illinois
Charles Ross & Son Co.
Hauppauge, New York
Diosna
Osnabruck, West Germany
Elizabeth Hata Int., Inc.
North Huntingdon, Pennsylvania
Fitzpatrick Corp.
Elmhurst, illinois
Fluid Air, Inc.
Napersville, illinois
Gemco (The General Machine Co.
of N.J.)
Middlesex, New Jersey
Glatt Air Techniques, Inc.
Ramsey, New Jersey
Glen Mills, Inc.
Maywood, New Jersey
Gral- Collette
Northbrook, illinois
H. C. Davis Sons Mfg. Co., Inc.
Bonner Springs, Kansas
Hollan d-McKinley
Malvern, Pennsylvania
Indupol Filtration Assoe , , Inc.
Cresskill, New Jersey (Torit)
Inppec , Inc.
Milford, Connecticut (Kilian)
J. H. Day Corp.
Cincinnati, Ohio
Jaygo, Inc.
Mahwah, New Jersey
Kemutec, Inc.
Bristol, Pennsylvania
Korsch Tableting, Inc.
Somerset, New Jersey
Lightnin Mixing Equipment Co.
Rochester, New York
Littleford Bros., Inc.
Florence, Kentucky
Mane sty , Thomas Engineering, Inc.
Hoffman Estates, Illinois
Marion Mixers, Inc.
Marion, Iowa
Micropul
Summit, New Jersey
Millipore Corp.
Bedford, Massachusetts
Mocon (Modern Controls, Inc.)
Minneapolis, Minnesota
Tablet Production
Natoli Engineering
Chesterfield, Missouri
Patterson-Kelly Co.
East Stroudsburg, Pennsylvania
Raymond Automation Co , , Inc.
Norwalk, Connecticut
S. Howes Co •• Inc.
Silver Creek, New York
Scientific Instruments &
Technology Corp.
Piscataway. New Jersey
Stokes-Merrill
Warminster, Pennsylvania
SUGGESTED READINGS
Thomas Engineering. Inc.
Hoffman Estates, Illinois
United Chemical Machinery
Supply, Inc.
Toms River. New Jersey
(Kikusui)
Urschel Laboratories, Inc.
Valparaiso, Indiana
Vector Corp.
Marion, Iowa
415
Alessi. P. Operational MRP vs , Integrated MRP, P 811M with APIC S
News, June 1986.
Anton, C. J. and Malmborg, C. J. The Integration of Inventory
Modeling and MRP Processing: A Case Study. Production and
Inventory Management, 2nd Quarter, 1985.
Ballou, Ronald H. Estimating and auditing aggregate inventory
levels at multiple stocking points. J. Operations Management
1(3): 143-154 (Feb. 1981).
Bryson, W. L. Profit-Oriented Inventory Management. APICS
22nd Annual Conference Proceedings, 1979. pp , 88-91.
The Competitive Status of the U. S. Pharmaceutical Industry.
National Academy Press, Washington, D.C., 1983.
Hadley, G. and Whitin, T. M. Analysis of Inventory Systems.
Prentice-Hall, New York. 1963.
Lotenschtein, S. Just-in-Time in the MRP II Environment, P &1M
Review with APICS News, Feb. 1986.
Mehta, N. How to Handle Safety Stock in an MRP System. Production
and Inventory Management, 3rd Quarter. 1980.
Ott, E. T. Process Quality Control. McGraw-Hill. New York,
1975.
REFERENCES
1. Artiss , D. H. and Klink, A. E. Good Manufacturing Practices
for Water: Methods of ManUfacturing, Testing Requirements
and Intended Use. AIChE 70th Annual Meeting, New York,
Nov. 1977.
2. Berry, W. L., Whybark, D. C., and Vollmann, T. E. Manufacturing
Planning and Control Systems. Irwin. Homewood,
Illinois , 1984.
416 Connolly, Berstler, and Coffin-Beach
3. Boucher. T. O. and Elsayed, E. A. Analysis and Control of Production
Systems. Prentice-Hall, New York, 1985.
4. Guideline on General Principles of Process Validation. U •S. Food
and Drug Administration, Washington. D.C., May 1984.
5. Guidelines for Plant Pest Control Program. PMA. Washington, D.C.,
Jan. 1975.
6. Hitchings. W. S.. IV, Tuckerman, M. M., and Willig, S. H. Good
Manufacturing Practices for Pharmaceuticals. Marcel Dekker, Ine , ,
New York, 1982.
7. Hutt , M. D. and Speh, T. W. Industrial Marketing Management.
Dryden Press. Chicago. 198!.
8. Peterson, R. and Silver, E. A. Decision Systems for Inventory
Management and Production Planning. Wiley, New York, 1985.
9. Pharmaceutical Process Validation. Marcel Dekker. New York, 1984.
10. Tableting Specification Manual, IPT Standard Specifications for
Tableting Tools. Academy of Pharmaceutical Sciences and American
Pharmaceutical Association, 1981.
11. The Theory and Practice of Industrial Pharmacy. Lea & Febiger,
Philadelphia, 1976.
7
The Essentials of
Process Validation
Robert A. Nash
St. John's University, Jamaica, New York
I. INTRODUCTION
The U. S. Food and Drug Administration (FDA) in its most recently proposed
guidelines has offered the following definition for process validation
[1] :
Process validation is a documented program which provides a high
degree of assurance that a specific process (such as the manufacture
of pharmaceutical solid dosage forms) will consistently produce
a product meeting its predetermined specifications and quality
attributes.
According to the FDA. assurance of product quality is derived from
careful (and systemic) attention to a number of (important) factors, including:
selection of quality (components) and materials. adequate product
and process design. and (statistical) control of the process through
in-process and end-product testing.
Thus it is through careful design and validation of both the process
and its control systems that a high degree of confidence can be established
that all individual manufactured units of a given batch or succession
of batches that meet specification will be acceptable.
According to FDA's Current Good Manufacturing Practices (21CFR
211.110)
Control procedures shall be established to monitor output and to
validate performance of the manufacturing processes that may be
responsible for causing variability in the characteristics of inprocess
material and the drug product. Such control procedures
417
418
shall include, but are not limited to the following, where appropriate
[2J:
1. Tablet or capsule weight variation
2. Disintegration time
3. Adequacy of mixing to assure uniformity and homogeneity
4. Dissolution time and rate
5. Clarity, completeness, or pH of solutions
Nash
The first four items listed above are directly related to the manufacture
and validation of solid dosage forms. Items 1 and 3 are normally associated
with variability in the manufacturing process, while items 2 and 4 are
usually influenced by the selection of the ingredients in the product formulation.
With respect to content uniformity and unit potency control (item 3)
adequacy of mixing to assure uniformity and homogeneity is considered to
be a high-priority concern.
Conventional quality control procedures for finished product testing
encompass three basic steps:
1. Establishment of specifications and performance characteristics
2. Selection of appropriate methodology, equipment, and instrumentation
to ensure that testing of the product meets specification
3. Testing of the final drug product, using validated analytical and
test methods in order to insure that finished product meets
specifications.
With the emergence of the pharmaceutical process validation concept, the
following four additional steps have been added
4. Qualification and validation of the processing facility and its
equipment
5. Qualification and validation of the manufactUring process through
appropriate means
6. Auditing, monitoring, sampling, or challenging the key steps in
the process for conformance to specifications
7. Requalification and revalidation when there is a significant change
in either the product or its manufacturing process [3J
II. TOTAL APPROACH TO PHARMACEUTICAL
PROCESS VALIDATION
It has been said that there is no specific basis for requiring a separate
set of process validation guidelines since the essentials of process validation
are embodied within the purpose and scope of the present Current
Good Manufacturing Practices (CGMPs) regulations [21. With this in mind.
the entire CGMP document. from subpart B through subpart M. may be
viewed as being a set of principles applicable to the overall process of
manufacturing, i.e., solid dosage forms or other drug products and thus
may be subjected, SUbpart by subpart, to the application of the principles
of qualification, validation, control, as well as requalification and revalidation.
where appropriate. Although not a specific requirement of current
The Essentials of Process Validation 419
regulations. such as comprehensive validation approach with respect to
each subpart of the CGMP document has been adopted by many drug
firms.
A checklist of validation and control documentation with respect to
CGMPs is provided in Table 1. With the exception of subpart M (sterilization),
the rest of the CGMPs are directly applicable to the manufacture of
solid dosage forms.
Table 1 Checklist of Validation and Control Documentation
Subpart
A
B
C
D
E
F
G
H
I
J
K
L
Section of CGMP
Introduction
Organization and
personnel
Buildings and
facilities
Equipment
Control of raw materials,
in-process
material, product
Production and
process controls
Packaging and
labeling controls
Holding and distribution
Laboratory controls
Records and reports
Returned and salvaged
drug product
Air and water
quality
Validation and control
documentation
Establishment of QA & PV functions
Establishment and facility installation
and qualification [4.5)
Plant and facility installation and
qualification (4,5)
Maintenance and sanitation [6]
Microbial and pest control [7)
Installation and qualification cleaning
methods [8]
Incoming components (9]
Manufacturing non-sterile products
[10]
Process control systems [l1J
(instrumentation and computers)
Depyrogenation, sterile packaging,
filling, and closing [12.13]
Facilities [14]
Analytical methods [15]
Computer systems [16J
Batch reprocessing [17]
Water treatment and steam systems
air, heat, and vacuum handling
[18 - 20]
420
Table 1
Subpart
M
(Continued)
Section of CaMP
Sterilization
Validation and control
documentation
LVPs [21,22]
Autoclaves and process
Parametrics [23 - 25]
Aseptic facilities r26J
Devices [27]
Sterilizing filters [28,29]
Nash
Table 2 Process Validation Matrix or Checklist of Activities to be
Considered
Personnel
(manpower)
(people systems)
Parts
(components. inprocess.
finished
product)
Process
(machines)
(buildings, facilities.
equipment. support
systems)
Procedures
(methods)
(manufacturing and
control. documentation.
records)
The Essentials of Process Validation 421
The CGMPs may also be viewed as consisting of the following four essential
elements:
1. Personnel. The people system and manpower required to carry
out the various tasks within the manufacturing and control
functions.
2. Parts. The raw materials and components used in connection with
the manufacture and packaging of the drug product as well as the
materials used in association with its control.
3. Process. The buildings. facilities. equipment. instrumentation.
and support systems (heat. air. vacuum. water, and lighting) used
in connection with the manufacturing process and its control.
4. Procedure. The paperwork, documentation, and records used in
connection with the manufacturing process and its control.
Thus the four elements of CaMPs listed above may be combined with
the four elements of pharmaceutical process validation (Le .• qualification,
validation. control, and revalidation) to form a 4 x 4 matrix with respect
to all the activities that may be considered in connection with the manufacture
and control of each drug product. An example of such a process
validation matrix provides a simple checklist of activities to be considered
in connection with the general principles of process validation (Table 2).
III. ORGANIZING FOR VALIDATION
The mission of quality assurance in most pharmaceutical companies today
has grown in importance with the advent of process validation. The
process validation concept, which started as a subject noun (validation) in
Table 3 Specific Responsibilities of Each Organizational Structure within
the Scope of Process Validation
Engineering
Development
Manufacturing
Quality
assurance
Source: Ref. 31.
Install, qualify , and certify plant. facilities. equipment
, and support systems.
Design, optimize, and qualify manufacturing process
within design limits, specifications, and /or req uirements.
In other words, the establishment of process
capability information.
Operate and maintain plant, facilities I equipment, support
system s , and the specific manufacturing process
within its design limits, specifications, and/or requirements.
Establish approvable validation protocols and conduct
process validation by monitoring. sampling, testing.
challenging. and/or auditing the specific manufacturing
process for compliance with design limits, specifications,
and lor requirements.
Table 4 Validation Progress Gantt Chart
Qualification stage Validation stage
II:>.
1>.:1
1>.:1
Key elements Design stage Installation Operational Prospective Concurrent
(Batch records and
validation documentation)
Scale-up phase -
(process optimization
and pilot production)
Facilities and equipment
Process and product
Engineering phase .. Manufacturing
~(Validation
/Pro'ocolSl
Developmental phase --(
formula definition
and stability testing)
start-up
QA and manufactUring phase
(full production)
--- _ .. --_.._----- .
Time line for new product introduction
~
Qrn
;::r
The Essentials of Process Validation 423
the late 1970s, has been turned into an action verb (to validate) by the
quality assuance function of many drug companies. Quality Assurance was
initially organized as a logical response to the need to assure that CGMPs
were being complied with. Therefore, it is not surprising that process
validation became the vehicle through which quality assurance now carries
out its commitment to CGMPs [301.
The specifics of how a dedicated group, team, or committee is organized
in order to conduct process validation assignments is beyond the
scope of this chapter. It is clear, however, that the following responsibilities
must be carried out and that the organizational structures best
equipped to handle each assignment is presented in Table 3.
The concept of divided validation responsibilities can be used for the
purpose of constructing a validation progress time chart (Table 4). Such
a chart is capable of examining the logical sequence of key events or milestones
(both parallel and series) that take place during the time course of
new product introduction and is similar to a Gantt chart constructed by
Chapman [32J.
In Table 4, facilities and equipment are the responsibility of Engineering
and Manufacturing, while process and product are the responsibility of
the product and process development function(s). The engineering and
development functions in conjunction with quality assurance come together
to prepare the validation protocols during the qualification stage of product
and process development.
IV. PROCESS VALIDATION -ORDER
OF pRIORITY
Because of resource limitation, it is not always possible to validate an entire
company's product line at once. With the obvious exception that a
company's most profitable products should be given a higher priority, it is
advisable to draw up a list of product categories that are to be validated.
The following order of importance or priority with respect to validation
is suggested to the reader:
Sterile Products and Their Processes
1. Large-volume parenterals (L VPs)
2. Small-volume parenterals (SVPs)
3. Ophthalmics and other sterile products
Nonsteriie Products and Their Processes
4. Low-dose/high-potency tablets and capsules
5. Drugs with stability problems
6. Other tablets and capsules
7. Oral liquids and topicals
V. PILOT-SCALE-UP AND PROCESS VALIDATION
The following operations are normally carried out by the development function
prior to the preparation of the first pilot-production batch. The development
activities are listed as follows:
424 Nash
1. Formulation design I selection, and optimization
2. Preparation of the first pilot-laboratory batch
3. Conduct initial accelerated stability testing
4. If the formulation is deemed stable I preparation of additional
pilot-laboratory batches of the drug product for expanded nonclinical
and/or clinical use
The pilot program is defined as the scale-up operations conducted subsequent
to the product and its process leaving the development laboratory
and prior to its acceptance by the full-production manufacturing unit.
For the pilot program to be successful, elements of process validation
(i.e., product and process qualification studies) must be included and completed
during the developmental or pilot-laboratory phase of the work.
Thus product and process scale- up should proceed in graduated steps
with elements of process validation (such as qualification) incorporated at
each stage of the piloting program [33J.
1. Laboratory Batch. The first step in the scale-up process is the
selection of a suitable preliminary formula for more critical study and testing
based upon certain agreed-upon initial design criteria, requirements
and lor specifications. The work is performed in the development laboratory.
The formula selected is designated as the (lX) laboratory batch.
The size of the (1X) laboratory batch is usually 3 - 5 kg of a solid or semisolid,
3 - 5 liters of a liquid or 3000 to 5000 units of a tablet or capsule.
2. Laboratory -Pilot Batch. After the OX) laboratory batch is determined
to be both physically and chemically stable based upon accelerated.
elevated temperature testing (1. e.. 1 month at 45°C or 3 months at
38°C or 38°C/80% RH). the next step in the scale-up process is the preparation
of the (lOX) laboratory -pilot batch. The (lOX) laboratory -pilot
batch represents the first replicated scale-up of the designated formula.
The size of the laboratory -pilot batch is usually 30 - 50 kg. 30 - 50 liters
or 30,000 to 50.000 units.
It is usually prepared in small-size. pilot equipment within a designated
CGMP approved area of the development laboratory. The number
and actual size of the laboratory -pilot batches may vary in response to
one or more of the following factors:
a. Equipment availability
b. Active drug substance availability
c. Cost of raw materials
d. Inventory requirements for clinical and nonclinical studies
Process qualification or process capability studies are usually started in
this important second stage of the pilot program. Such qualification or
capability studies consist of process ranging, process characterization,
and process optimization as a prerequisite to the more formal validation
program that follows later in the piloting sequence.
3. Pilot Production. The pilot-production phase may be carried out
either as a shared responsibility between the development laboratories and
its appropriate manufacturing counterpart -or as a process demonstration
by a separate, designated pilot-plant or process-development function.
The two organizational piloting options are presented separately in Figure
1. The creation of a separate pilot-plant or process-development unit
The Essentials of Process Validation 425
PRODUCTION PILOT
PLANT _I~I- DEVELOPMENT
LABORATORY
JOINT PILOT OPERATION
Pilot Batch
PRODUCTION DEVELOPMENT
LABORATORY
~ Reque.t ';1
" Pilot Batch /' --------Completion
Report
Figure 1 Main piloting options. (top) Separate pilot plant functionsengineering
concept. (bottom) Joint pilot operation.
has been favored in recent years for it is ideally suited to carry out
process qualification and/or validation assignments in a timely manner.
On the other hand, the joint pilot-operation option provides direct communication
between the development laboratory and pharmaceutical production.
The objective of the pilot-production batch is to scale the product and
process by another order of magnitude OOOx) to, for example, 300-500
kg, 300-500 liters, or 300,000-500,000 dosage-form units (tablets or
capsules) in size. For most drug products this represents a full production
batch in standard production equipment. If required, pharmaceutical
production is capable of scaling the product and process to even larger
batch sizes should the product require expanded production output. If
the batch size changes significantly (say to 500x or 1000 x) additional
validation studies would be required.
Usually large production batch scale-up is undertaken only after product
introduction. Again, the actual size of the pilot-production (lOOx)
batch may vary due to equipment and raw material availability. The need
for additional pilot-production batches ultimately depends upon the successful
completion of a first pilot batch and its process validation program.
Usually three successfully completed pilot-production batches are
required for validation purposes.
In summary, process capability studies start in the development laboratories
and/or during product and process development continue in welldefined
stages until the process is validated in the pilot plant and/or
pharmaceutical production.
An approximate timetable for new product development and its pilot
scale-up program is suggested in Table 5.
426 Nash
Table 5 Approximate Timetable for New Product Development and Pilot
Scale-Up Trials
Event
Formula selection and development
Assay methods development and formula optimization
Stability in standard packaging 3-month read-out (Ix size)
Pilot-laboratory batches (lOx size)
Preparation and release of clinical supplies (lOx size) and
establishment of process qualification
Additional stability testing in approved packaging
6 -8-month read-out (1x size)
3-month read-out (lOx size)
Validation protocols and pilot batch request
Pilot-production batches (lOOx size)
Additional stability testing in approved packaging
9 -12-month read-out (Ix size)
6 -8-month read-out (lOx size)
3-month read-out (lOOx size)
Interim approved technical product manual with approximately
12-months stability (Ix size)
Totals
VI. PROCESS CAPABILITY DESIGN
AND TESTING
Calendar
months
2-4
2-4
3-4
1-3
1-4
3-4
1-3
1-3
3-4
1-3
18-36
Process validation trials are never designed to fail. Process validation
failures> however, are often attributable to an incomplete picture of the
manufacturing process being evaluated. Upon closer examination of the
problem, often failures appear to be directly related to an incomplete
understanding of the process's capability or that the process qualification
trials were not properly defined for the job to be done.
Process capability is defined as studies that are carried out to determine
the critical process parameters or operating variables that influence
process output and the range of numerical data for each of the critical
process parameters that result in acceptable process output.
Thus. the objectives of process capability design and testing may be
listed as follows:
1. To determine the number and relative importance of the critical
parameters in a process that affect the quality of process output
2. To show that the numerical data generated for each critical
parameter are within at least statistical quality control limits (i , e. ,
The Essentials of Process Validation
±3 standard deviations and that there is no drift or assignable
cause of variation in the process data
427
If the capability of a process is properly delineated, the process should
consistently stay within the defined limits of its critical process parameters
and product characteristics [34J.
Process qualification, on the other hand, represents the actual studies
or trials conducted to show that all systems, subsystems, or unit operations
of a manufacturing process perform as intended. Furthermore, that
all critical process parameters operate within their assigned control limits
and that such studies and trials, which form the basis of process capability
design and testing, are verifiable and certifiable through appropriate
documentation. Process qualffleation is often referred to as Operational
or Performance Qualification.
The manufacturing process is briefly defined as the ways and means
used to convert raw materials into a finished product. The ways and
means also include: people, equipment, facilities, and support systems
that are required in order to operate the process in a planned and an
effectively managed way. Therefore, let us assume that all people, equipment,
facilities, and support system s that are required to run the process
qualification trials have been themselves qualified and validated beforehand.
The steps and the sequence of events required in order to perform
process capability design and testing are outlined in Table 6.
Using the basic process for the manufacture of a simple tablet dosage
form, we will attempt here to highlight some of the important elements of
the process capability and qualification sequence.
1. Basic information is obtained from the (Ix) size laboratory
batch.
a. Quantitative formula is scaled to (lOx) size batch and rationale for
inert ingredient selection provided.
b. Critical specifications, test methods, and acceptance criteria for
each raw material used in the formula are provided.
c. List of proposed specifications, test methods, and acceptance criteria
for the finished dosage form are provided.
d. Interim stability report on (Lx) size laboratory batch is provided.
e. Detailed operating instructions for preparing the (lOX) size batch
are provided.
2. Preparation of a simple flow diagram of the process should be provided.
A good flow diagram should show all the unit operations in a
logical sequence, the major pieces of equipment to be used, and the
stages or operations at which the various ingredients are added. The
flow diagram, shown in Figure 2, outlines the sequence of unit operations
used to prepare a typical tablet dosage form by the wet granulation
method. In Figure 3, the enclosed large rectangular modules represent
the various unit operations in the manufacturing process. The arrows
represent transfers of material into and out of each unit operation. Each
large rectangular module or box indicates the particular unit operation,
the major piece of processing equipment employed, and the facility in
which the operation takes place. The sequential arrangement of unit
428 Nash
Table 6 Protocol for Process Capability Design and Testing
Objective of piloting
program
Types of process
Typical processes
Definition of process
Definition of process
output
Definition of test
methods
Analysis of process
Review and analysis of
data
Pilot batch trials
Pilot batch replication
Need for process capability
redefinition
Process capability of
evaluated process
Final report and
recommendations
Process capability design and testing
Batch. intermittent. continuous
Chemical, pharmaceutical, filling, and packaging
Flow diagram. equipment/materials in-process,
finished product
Potency. yield, physical parameters
Methods, equipment. calibration traceability,
precision, accuracy
Definition of process variables, influence
matrix, fractional factorial analysis
Data plot (x -y plots, histogram, control
chart) time sequence, sources of variation
Define stable/extended runs, define sample
and testing, remove sources of variation
Different shifts and days, different materials ,
different facilities and equipment
Data analysis, modification of influence matrix.
reclassification of variables
Stability and variability of process output, conformance
to defined specifications, economic
limits of process
Recommended SOP, limits on process adjustments.
recommended specifications
operations should be analogous to the major sequential steps in the operating
instructions of the manufacturing process.
3. Using the flow chart (Fig. 2) as a guide , a list of process or control
variables are next drawn up for each unit operation or step in the
process. A test parameter or response to be objectively measured is then
assigned to each process variable. The control parameters (Le •• process
variables plus their test parameters) for the manufacture of compressed
tablets by the wet granulation method are shown in Table 7. According to
the data presented in Table 7. there are six unit operations or processing
steps, with from two to five process variables for each unit operation and,
with the exception of tablet compression (finished tablet analysis), one
key test parameter for each of the processing steps. Please note that the
first unit operation, I, e .• the weighing of active and inert ingredients,
was eliminated from Table 7. Weighing operations are a general consideration
in all manufacturing processes. Balances and measuring devices are
normally qualified and validated separately on a routine basis as required
by CGMP's guidelines.
WEIGH ACTIVE &: FILLERS
Drums RmS
U
PREBLEND POWDERS
Vertical MIxer Rm 10
~
GRANULATE BLEND tI.
t.:l
'"
438 Nash
3. The overall design does not balance to zero because there is a
constant +2 for each process variable. Since there are, this is due to the
fact that in our design eight -13s in trial No. 1 and eight +15s in trial
No. 8 were created.
The advantage of the fractional factorial experimental design is that.
using No. 8 qualifleation trial, eight important process variables were
tested at both their lower and upper control limits. In addition. three
processing steps [dry milling. lubricant addition (blending). and tablet
compression] were all shown to be critical with respect to a measured response,
namely tablet hardness. Since tablet hardness also influences
tablet dissolution, this second outcome was also indirectly evaluated by the
experimental design chosen.
Using a larger fractional factorial experimental design, 12 or even 16
process variables could be tested by expanding the qualification trials to
say 12 or 16 pilot runs. But in practice, eight pilot-laboratory batches
is most likely the maximum number that is reasonable to produce for this
purpose. Those unit operations that were found. by fractional factorial
experimental design, to be critical with respect to process capability design
and testing (size reduction, blending. and tablet compression) could
then be subjected to more extensive investigation during the laboratory
stage of product and/or process development.
C. Optimization Techniques
Optimization techniques are used to find either the best possible quantitative
formula for a product or the best possible set of experimental conditions
(input values) that are needed to run the process. Optimization
techniques may be employed in the laboratory state to develop the most
stable, least sensitive formula. or in the qualification and Validation stages
of scale-up in order to develop the most stable. least variable process
within its proven acceptable range(s) of operation, Chapman's so-called
PAR principle [39]
Optimization techniques may be classified as Parametric Statistical
Methods and Nonparametric Search Methods.
Parametric Statistical Methods
Parametric Statistical Methods, usually employed for optimization, are
full-factorial designs [40]. half-factorial designs [41]. simplex design [42].
and Lagrangian multiple-regression analysis [43]. Parametric methods are
best suited for formula optimization in the early stages of product development.
The application of constraint analysis. which was described previously.
is used to simplify the testing protocol and the analysis of experimental
results.
The steps involved in the parametric optimization procedure for pharmaceutical
systems have been fully described by Schwartz [44]. The
optimization technique consists of the following' essential operations:
1. Selection of a suitable experimental design
2. Selection of variables (independent Xs and dependent Ys) to be
tested
The Essentials of Process Validation 439
Table 13 Results of a Three-Components Simplex Design for Tablet
Hardness
Transformed Average
Excipient components proportions tablet
hardness
Run no. Xl X2 X3 Xl X2 X
3
(SCU)
I 55 10 10 1 0 0 6.1
2 10 55 10 0 1 0 7.5
3 10 10 55 0 0 1 5.3
4 32.5 32.5 10 0.5 0.5 0 6.6
5 32.5 10 32.5 0.5 0 0.5 6.4
6 10 32.5 32.5 0 0.5 0.5 6.9
7 25 25 25 0.33 0.33 0.25 7.3
8 32.5 21.25 21. 25 0.5 0.25 0.25 7.2
3. Performance of a set of statistically designed experiments (I.e.,
23 or 32 factorials)
4. Measurement of responses (dependent variables)
5. Development of a predictor, polynomial equation based upon statistical
and regression analysis of the generated experimental data
6. Development of a set of optimized requirements for the formula
based upon mathematical and graphical analysis of the data generated
According to Bolton, one of the most useful methods of defining optimal
regions of formulation characteristics is based upon the application
of simplex matrix design [45]. For example, formulations may be constructed,
using constraint analysis. so that the total amount of excipients
(Xs) to be added to the powder mix is never more than 75 mg of a total
tablet weight of 300 mg. A brief outline of the simplex technique, used in
connection with this example is shown in Table 13.
In the transformations shown in Table 13, the highest excipient concentration.
55 mg, is assigned a value of one and the lowest excipient
concentration, 10 mg, equals 0 is assigned a value of zero. Coefficients
for X10 X2. and X3 in the following polynomial equation are the hardness
values from run Nos. 1, 2, and 3. Simple equations for calculating coefficients
for the following terms: xr X2. Xl X3, X2 X3. and Xl X2 X3
are given in the reference [45].
where
440
Example (run no. 8):
Nash
y =6.1 (0.5) + 7.5 (0.25) + 5.3 (0.25) - 0.8 (0.125) + 2.8 (0.125)
+ 2 (O~0625) + 15 (0.03125)
Y =3.05 + 1.875 + 1.325 - 0.1 + 0.35 + 0.125 + 0.47 =7.1
The best tablet hardness (7.4 value SCU) containing all three excipients
is obtained at Xl =0.25, X2 = 0.5. and X3 = 0.25. where Xl = 18.75 mg.
X2 =37.5 mg. and X3 = 18.75 mg.
Nonparametric Search Methods
Nonparametric Search Methods are relatively simple techniques used to finetune
or optimize a process by varying the critical process parameters that
were found during the qUalification and validation stages of process development.
The procedure is so constrained that no process variable is
ever permitted to exceed its lower or upper control limit. In searching a
given process. it is assumed that there is optimum peak (set of experimental
conditions or inputs) where the process operates most efficiently.
There are two basic search methods that are used for this purpose. The
first method is called evolutionary operation (EVOP) and a second method
is called random evolutionary operation (REVOP). The difference between
the two methods, EVOP and REVOP. is not objectivity but simplicity of the
experimental design chosen.
Process Improvement Through EVOP
The process variables whose perturbation or slight change might lead to
improvement in process performance are usually identified during the qualification
trials where the operational and control limits for the process
have been developed. Next. initial perturbation steps away from the present
operational inputs are selected for each of the critical process variables.
These steps must be sufficiently small so that no input goes beyond the
control limits of the process and no output goes out of product and process
specification. In the traditional box EVOP design [46] a simple two-factor.
two-level design is created about the present condition or input values (see
Fig. 4).
Analysis of the data presented in Figure 4 shows that the path of
steepest ascent to improve tablet hardness is in the direction of run No. 1
to run No.4, which is a change of +3 seu in tablet hardness. A second
two-factor, two-level box is constructed about run No.4 where one corner
of the box is the original condition, run No.1. Following the same procedure,
a series of from 8 to 24 runs, using the path of steepest ascent,
is usually required to ascertain the optimum input values for tablet hardness
and rapid tablet dissolution.
By using a simplex EVOP design [47] where connecting triangles are
created from 2 additional experimental conditions, it is possible to complete
the optimization search in fewer trial runs than Box EVOP (see
Fig. 5).
Process Improvement Through REVOP
Random evolutionary operation (REVOP) is a comparatively little-used
method for process and product optimization [48 J• The technique.
The Essentials of Process Validation 441
~ 3000 ....
ai
El 2500 G2
0~
r:: 2000
0 -;
1ft i 1500
0c
6 9 13 15 18 21
Blend Time (min.) Xs
Figure 4 Optimization by box Evolutionary Operation. Tablet hardness
(SeU shown in parenthesis). Optimum conditions (2000 lbs and 6 min).
developed by F. E. Satterthwaite, employs a random direction of movement
under constraint analysis to discover a probable pathway of ascent
to peak process performance. If the linear direction chosen is not promising.
the direction is reversed and the opposite pathway is chosen in an
effort to improve performance. Movement continues along the new path in
direction previously established, as long as the results are positive.
Movement will then proceed at right angles when progress ceases on the
previously chosen pathway. Peak performance is almost always achieved
in less than 20 trials runs (see Fig. 6).
In summary, process capability studies and qualification trials should
be underaken during the first stage of pilot scale-up, i.e .• with the preparation
of the pilot-laboratory batch (lOx) size). The objective of such
~ 3000 -~
2500 G~
~
~ 2000
1e 1500 rc
0 3 6 9 13 15 18 21
Blend Time (min.) x,
Figure 5 Optimization by simplex evolutionary operation. Tablet hardness
(SeU) shown in parenthesis.
442 Nash
.< 3000 ~(20) out of specs --.. ~(18J ~(l5) @
2500 ~
~
(10) ~ ~
2000 ®-P.
Q)
~
2
198J~ • • 50"C
1.96
1.94
§ 192
lii ...
C 1.90
CD u
c: 8'88
Cl
.3 1.86
184
182
180
178
3 6 9 12
TIme 10Months
Figure 2 Percent of alkaloid remaining' as a function of time for a first -order reaction.
::r::
~
;:s
;:s
~
Stab ility Kinetics
Table 1 Percent of Alkaloida in Tablet Formulation
Time (months)
Temperature
(OC) 3 6 9 12
50 98.92 97.83 96.70 95.72
60 96.16 92.47 88.92 85.51
70 88.41 98.16 69.10 61. 09
aInitial 100%.
463
As log Co is a constant I then a plot of log (drug concentration) against
time will produce a straight line whose slope is equal to -K /2.303.
The constant, K. is the reaction velocity constant or specific reaction
rate, which expresses the fraction of material that reacts in a given unit
of time expressed in reciprocal seconds, minutes, or hours. For example,
let K = 0.01 h -1; then the rate at which the drug is de grading is 1%h -1.
The half-life equation is
t =0.693
1/2 K
and the shelf-life equation is
t = 0.105
90 K
Example: A first-order rate of hydrolysis degradation of an alkaloid at
1 mg per tablet manufactured by aqueous wet granulation technique is
shown in Table 1, which. if graphically plotted I should give a straight
line as shown in Figure 2.
C. Pseudo-First-Order Reaction
(8)
(9)
When the reaction rate depends on the concentration of two reactants or a
bimolecular reaction that is made to act like a first -order reaction. then it
is called pseudo-first -order reaction. for example, when one reactant is
present in greater quantity than the other or when the first is kept at a
constant concentration in relation to the second. Under these circumstances,
one reactant appears to control the rate of reaction even though
two reactants are present because the concentration of the second reactant
does not change significantly during the degradation. An example
of such a reaction is the degradation of sodium hypochlorite tablets in
aqueous solution, which is highly pH dependent and follows a pseudo-firstorder
reaction because the hydroxyl ion concentration is high compared to
the concentration of the carboxylic ion.
464 Hanna
VI. KINETIC STUDIES
Solid-dosage forms that are tablets and capsules constitute a large majority
of pharmaceutical products. However. few kinetic studies and studies on
rates of drug degradation in solid state have been published. Drug solid
degradation in absence of excipients and moisture generally follows a
nucleation chemical reaction rate which may approximate a first-order reaction
rate. In the presence of moisture and excipients, the reaction rate
could be either zero-. first-. or pseudo-first-order. It should be noted
that there is little difference in degradation rates estimated by the zeroor
first-order reactions when less than 15% degradation has occurred.
When the amount of degradation found is in excess of 15% the order of the
reaction which accounts for the degradation should be established.
Stability study at room temperature is the surest method of determining
the actual shelf life of a product. Unfortunately, it is difficult to make an
accurate expiration date prediction until 2 or 3 years of data are generated,
a situation which can be further complicated by the frequent need
for tablet formulation changes that would require additional long shelf-life
study at actual shelf-life conditions.
The principles of chemical kinetics for the evaluation of drug stability
are based on the fact that reaction rates are expected to be proportional to
the number of collisions per unit time. Since this number increases with
the increase of temperature, it is necessary to evaluate the temperature dependency
of the reaction. Experimentally, the reaction rate constant is observed
to have an exponential dependence on temperature as expressed by
the Arrhenius equation:
or
K ::: A (-Ha/RT) (10)
or
log K =log A
lIH
2.303RT (11)
K1 Ha (1 1 )
log K2 = 2. 303R T 2 - T1 (12)
where K = specific rate of degradation
-1 -1
R = gas constant (1.987 cal deg mol )
A =frequency factor (constant)
T ::: absolute temperature (tOe + 273.16°e)
Ha = activation energy of the chemical reaction
A plot of log K against lIT will produce a straight line whose slope is
equal to -lIHa/R (2.303), and is known as an Arrhenius plot (Fig-. 3).
The reSUlting line obtained at higher temperatures is extrapolated to obtain
Stab ility Kinetics
4.0
3.0
1.0
465
o 30
50
335 35
II
lIT X 103
II
25 o
4.0
Figure 3 Arrhenius plot of log K against reciprocal of absolute temperature.
room temperatures and K 25 is generally used to obtain a measure of the
stability of the drug under ordinary shelf conditions at room temperature;
however, any other desired temperature can be equally obtained.
From Figure 3, log K25 is 2 and K25 will be 100 months or 8.3 years
under shelf conditions at room temperature (25°C) if the original points at
60, 70, and 80°C in the Arrhenius plot were calculated from monthly stability
points. Activation energy can also be calculated from an Arrhenius
plot by using the slope of the line which is equivalent to -Ha/2.303R.
From Figure 3, the slope of the line is - 3. 5 x 103. Then
3 Ha
-3.5 x 10 = 2.303 x 1.987
466
and
Ha =(-3.5 x 103)(2.303)(1.987)
= 16.0 kcal/mOI- 1
Hanna
Arrhenius projection enables predictions to be made that are based on
accelerated data in such a way that any change in the relationship between
the degradation at different temperatures will be detected and estimated.
For example, using equation 10 for the prediction of a tablet product expiration
date at shelf storage (25°C) from accelerated stability study at
40°C and activation energy of 10 keal Zmol ;
K =A(-Ha/RT1)
40
and
K = A (-Ha/RT2)
25
then
K 40 A (-Ha/RT1)
K
25
:: A(-Ha/RT 2)
-1
Assuming that Ha = 10 kcal/mol ,then T1 = 273.15 + 40 = 313.15 and
T 2
= 273.15+ 25= 298.15.
K
40
A-10,000/1,987(313.15)
K
25
= A- 10, OOO/ 1. 98 7( 298 . 15)
=2.24465
Shelf stability at 25°C = 12 x 2.245 = 26.94 months. Tablet product
with 10 kcal/mol- 1 activation energy of the chemical reaction and stable at
40°C for 12 months will have an expiration date at shelf life (25°C) of
27 months.
When reactant molecules, e. g., A + B, in a tablet formulation proceeds
to products, e. g., C + D, the energy of the system must change higher
than that of the initial reactant and is defined as the activation energy
(Ha) (Fig. 4).
The usual range for activation energies for tablet formulation decomposition
is about 10-20 kcal/mor 1, except if diffusion or photolysis is rate
determining. Then the rate is about 2 to 3 kcal Zmolr L, which rarely occurs
in tablet degradation. For reactions in which the heat activation energies
range is more than 50 kcal Zmolr l , the rate of degradation is not of any
practical significance at the temperature of shelf-life storage of tablet
formulations. For tablets with activation energy values higher than
Stability Kinetics
30
20
~
"0
E <,
(;j
0
~ H..=10 K cal/mol
>. r Cl
~ I Q)
c:
w t 10
A+B
Reactants - Products
A+B -C+D
Extent of Reaction
Figure q Activation energy.
C+D
467
20 kCal/mol- 1, the error in shelf-life prediction will be on the conservative
side. For example, using equation 10, for a tablet product with 20 kcall
mol- 1 activation energy and stable at 40°C for 3 months:
K 40 A(-Ha/RT1)
K25 =A(-Ha/RT2)
A- 20,000/1. 987(313.15)
= ~~~~~......,..,,....,..,,~---:-::-:-
A- 20.00011. 987( 298.15)
A-32.14250
=~--:-::"---
A-33.75960
= 5.03847
468 Hanna
Shelf stability at 25°C =3 x 5.03847 =15 months. For this tablet
product we will predict an expiration date of 15 months at shelf-life storage.
If in reality the activation energy of the chemical reaction of these tablet
product was on the higher side, that is, 25 kcal/mol- 1 instead of the previously
assumed 20 kcal/mol- l, then using equation 10
K40 A- 25,OOO/I.987(313.15)
K 25 = A- 25,OOO/1.097(298.15)
= 7.54867
and the predicted shelf stability at 25°C will be 3 x 7.54867 = 22.5 months.
which indicates that our original prediction of 15 months expiration date was
on the conservative side. On the other hand. if the real activation energy of
the chemical reaction was on the lower side, that is. 15 kcal/mor- 1 rather
than the originally assumed 20 kcal Imol- 1• then by using equation 10
K40 A-15.000/1. 987(313.15)
--=
K 25 A-15.000/1. 987(298.15)
=3.36296
and the predicted shelf stability at 25°C will be 3 x 3.36296 =10 months.
which indicates that our original prediction of 15 months expiration date was
a risky prediction.
To USe the Arrhenius equation. three stability storage temperatures
are obviously the minimum as the more additional temperatures are used the
more the accuracy of extrapolation is enhanced.
Table 2 gives an example of the application of the Arrhenius equation
to predict the stability of tablet product Z, Lot No. XI containing active
ingredient A.
1. Determine the potency of active A in the tablet product at appropriate
intervals of time when held at three temperatures. 50. 60, and 70°C.
as shown in Table 2. make two plots to determine if the degradation reaction
is zero- or first-order.
2. First. plot potency (y) against time (t) in weeks for the three temperatures
on normal graph paper, as shown in Figure 5. Then plot log
potency (log y) against time (t) in weeks for the three temperatures on
semilogarithmic graph paper as shown in Figure 6.
Stability Kinetics 469
110
100
90
80
70
<:( 60
~
ti
<:(
~ 50
~
8!.
40
30
20
10
o 5 10 15 2030 40 50 60
Weeks (I)
90 120
Figure 5 Zero-order plot for Active A Tablet Z Lot No. X.
3. Draw a straight line of best fit through the five points in both
graphs. Select the plot in which the line most closely fits the determined
points, including the original potency determined at zero time. which should
agree with the potency estimated by extending the degradation lines back
to zero time. In this example, these two conditions are met by the first
plot figure (y versus t or a zero-order reaction).
4. Determine for each of the three temperatures the slope of the
degradation rates (KO) in units of y (in case of first-order reaction Kl
in units of log y) by selecting two points lying on the line to determine
the slopes for each of the three temperatures. as shown in Table 3. The
slope of a line is defined as change in y divided by the change in x,
5. Plot the log values of KO on semilogarithmic graph paper against
the reciprocal of absolute temperature as shown in Figure 7.
470
Table 3 KO Values for Active
A in Tablets Z Lot No. X
Temperature
(OC) KO
50 -0.53
60 -2.5
70 -4.28
Hanna
6. Draw a straight line of best fit through the three points and extrapolate
it to the lower temperatures.
7. Estimate the rate of degradation at the desired temperature. If
this is 25°C, then the KO from the graph is 0.33.
8. Calculate the expiration data (tx) from equation 13.
1
9
8
7
>: 6
Ol sI< '" * * * '"
*** *** '" '" '" '" '" " .. '"
**'" *"'* '" '" >/< '" '" '" '" '"
*"'* "'** '" '" '" '" '" '" * '"
"''''''' "''''''' * * '" '" '" '" '" '"
"'*'" "'*'" '" '" * '" '" * '" '"
*** *** '" * '" '" '" " '" '"
"'** *"'* '" * '" ... * * '" '"
"''''* *** '" * 'I< '" '" '" '" '"
~
~
::t:
~
;:l
;:l
$:)
Stability Kinetics 491
25· 45" 37·
0.1.._ ...._ ..._ ... .._ ...__..
90· SO' 70° 60·
1
T
o
~ ....
Fiqure 11 Accelerated stability plot.
value versus liT are above 75 kcal , but one or more is below 17 kcal or a
minimum of t90 was achieved after at least 52 weeks at 30°C, then the label
storage condition of store at controlled room temperature is used. If t90
value versus liT is lower than 25 kcaI and above 45 kcal , or a minimum of
t90 was achieved after at least 4 weeks at 40°C and 1 week at 50°C. then
the label storage condition of store below 30°C is used. Other label storage
conditions. like protect from freezing or protect from light, should be used
as necessary.
G. Stability During Shipment
Appropriate labeling conditions derived from suitable stability studies must
be described to assure proper protection of tablet products during shipment.
Periods of time that a tablet product may be exposed to extreme
high or low temperature should be determined and appropriate label recommendations
are used accordingly. e. g .• may be adversely affected by exposure
to 50°C more than 3 weeks.
492 Hanna
XIII. COMPUTER APPLICATION
Computer programs are an efficient means of assisting in the design of stability
studies, especially complex models, and the interpretation of data.
A typical computer program can provide an efficient control of sample
storage at given conditions, control inventory. schedule test intervals t
sample at schedule inter-vals , test requirements and analytical procedure
for each sample, determine manpower required for sample testing, review
data, evaluate statistical data, highlight out-of-limits results, maximize research
and development capabilities, interpret data, plot results, estimate
shelf life, prepare reports for regulatory submission or commitments and
product stability trend analysis. Implementation of computer use in stability
programs will allow efficient statistical evaluation of the kinetic data,
handling of complex administrative functions t organization of voluminous
amounts of data, and reduce the potential for human errors. Several stability
computer programs are commercially available from pharmaceutical as
well as specialized computer companies. Example for such computer forms
and stability reports are presented in Figures 12 and 13. Stability data
listed in these two figures are filled according to the instructions per the
coding letter mentioned beneath each requirement as follows:
a = Computer transaction code
b = Product list number
c =Product lot number
d = Assay type: chemical or biological, etc.
e =Assay name
f =Date of assay as per the stability protocol
g = Time of the stability point
h = Number of assay code as per the written stability-indicating assay
monograph
i = Number of monograph code as per the written stability-indicating
assay monograph
j = %RH
k =Temperature in °c
I =Other atability conditions, e. g ., light
m = Date assay was performed
n =Chemist or technician name/or initial who performed the assay
o =Number of tests performed
p =All results of assay performed
q =Average assay results
r =Previous assay results reported
s = Units the assay is to be run by, e. g., average tablet weight
(ATW)
t =Number of samples analyzed
u = Total time for stability study as per the stability protocol
v = Product release limit
w = In-house or action limit
y = Name of supervisor who check
ed and approved the analysis
z =Date of supervisor approval
