Palvmmmhism in
`Pharmaceutical
`
`Salitls
`
`edited by
`Harry G. Brittain
`Discovery Laboratories, Inc.
`Milford, New Jersey
`
`MARCEZL
`
`( MA1{CEL DEKKER, INC.
`
`DEKKEH
`
`NEW YORK - BASEL
`
`Page 1 0‘ 33
`
`Grunenthal GmbH Exhibit 2041
`
`Rosellini V. Grunenthal GmbH
`
`IPR20l6—00471
`
`

`
`
`
`ISBN: 0-8247-0237-9
`
`This book is printed on acid-free paper.
`
`Headquarters
`Marcel Dekkcr, Inc.
`270 Madison Avenue, New York, NY 10016
`tel: 2126969000; fax: 212-685-4540
`
`Eastern Hemisphere Distribution
`Manccl Dckkcr AG
`
`Hutgasse 4, Postfach 812. CH—4001 Basel, Switzerland
`tel: 44-61-261-8482; fax: 44-61-261-8896
`
`World Wide Web
`
`http://www.dekker.com
`
`The publisher offers discounts on this book when ordered in bulk quantities. For more
`information, write to Special Sales/Professional Marketing at the headquarters address
`above.
`
`Copyright © 1999 by Marcel Dekker, Inc. All Rights Reserved.
`
`Neither this book nor any part may be reproduced or transmitted in any form or by
`any means, electronic or mechanical, including photocopying, microfilming, and re-
`cording, or by any information storage and retrieval system, without permission in
`writing from the publisher.
`
`Current printing (last digit):
`10
`9
`8
`7
`6
`S
`4
`3
`
`2
`
`l
`
`PRINTED IN THE UNITED STATES OF AMERICA
`
`
`
`Page 2 of 33
`
`

`
`V
`
`i
`
`This material may be protected by Copyright law (Title 17 U.S. Code)
`
`Effects of Pharmaceutical
`Processing on Drug Polymorphs
`and Solvates
`
`Harry G. Britlain
`
`Discovery Laboratories, inc.
`Milford, New jersey
`
`Eugene F. Fiese
`
`Pfizer Central Research
`
`Groton, Connecticut
`
`
`
`..«._..r«.:y._r—..-.—.._.».4-_-—-«.-
`
`INTRODUCTION
`
`PRODUCTION AND STORAGE OF BULK DRUG
`SUBSTANCE
`
`EFFECTS OF PARTICLE SIZE REDUCTION
`
`EFFECTS DUE TO GRANULA'I‘ION
`
`EFFECTS DUE TO DRYING
`
`A. Changes in Crystalline Form Accompanying the Spray-
`Dxying Process
`
`332
`
`333
`
`334
`
`339
`
`341
`
`343
`
`331
`
`Page 3 of 33
`
`

`
`332
`
`Brittaln and Flese
`
`B. Changes in the Crystalline State of Lyophilizcd
`Products
`
`VI. EFFECTS DUE TO COMPRESSION
`
`A. Changes in Crystal Form Effected by Compaction
`B. Effects on Tablet Properties Associated with the Use of
`Different Crystal Forms
`
`VII. SUMMARY
`
`REFERENCES
`
`I.
`
`INTRODUCTION
`
`345
`
`348
`
`348
`
`353
`
`356
`
`358
`
`In the previous chapters, the structural origin, energetics, and thermo-
`dynamics of polymorphs and solvates have been largely described for
`pure chemical entities. In most of the studies reported, the compounds
`were intentionally converted among various polymorphic forms for the
`purpose of study. In the present chapter, we will discuss the uninten-
`tional conversion of polymorphs and the desolvation of hydrates upon
`exposure to the energetics of pharmaceutical processing. Environments
`as harsh as 80°C and 100% RH for up to 6 h are not unusual during the
`routine manufacture of dosage forms. As previously noted, the various
`crystalline polymoiphs frequently differ in their heats of fusion by as
`little as 1 kcal/mol, with the transition temperature being well below
`the boiling point of water. In the ease of hydrates, removal of water
`from the crystal lattice requires more energy but is very much depen-
`dent on the temperature and humidity history of the sample.
`In this chapter we will discuss the effects of pharmaceutical pro-
`cessing upon the crystalline state of polymorphic and solvate systems.
`Given the degree of attention lavished on drug substances that is re-
`quired by solid-state pharmaceutical [1] and regulatory [2] concerns,
`it is only logical that an equivalent amount of attention he paid to pro-
`cessing issues. A variety of phase conversions are possible upon expo-
`sure to the energetic steps of bulk material storage, drying, milling,
`wet granulation, oven drying, and compaction. In this setting, an envi-
`ronment as harsh as 80°C and 100% RH for up to 12 h is not unusual,
`
`Page 4 of 33
`
`

`
`Effects of Pharmaceutlcal Processing
`
`333
`
`and the mobility of water among the various components must be con-
`sidered.
`
`ll. PRODUCTION AND [STORAGE OF BULK DRUG
`SUBSTANCE
`
`The first processing opportunity to effect a change in polymorphic form
`or solvate nature is with the final crystallization step in the synthesis
`of the bulk drug substance. Crystallization is thought to occur by first
`forming hydrogen-bonded aggregates in the solution state, followed by
`the buildup of molecules to produce a crystal nucleus. A number of
`parameters are known to affect the crystallization process, including
`solvent composition and polarity, drug concentration and degree of su-
`persaturation, temperature and cooling rate during the crystallization
`process, presence of seed crystals and/or nucleation sites, additives that
`influence crystal habit or add strain to the crystal lattice, agitation, pH,
`and the presence of a salt-forming molecule. It is evident that ample
`opportunity exists for the appearance of a polymorphic change when
`a process is scaled up, or moved to a new site, or run by a new operator.
`Since the discovery chemist would have been able to make gram
`quantities of a quasi-crystalline drug substance, logic holds that the
`process chemist should be able to make kilograms of the same sub-
`stance in a GMP manufacturing setting. While Mother Nature and equi-
`librium may have been fooled at the bench-top (where reaction steps
`are short and yield is improved through the use of anti solvents), longer
`processing times and improvements in purity usually mean that thermo-
`dynamic equilibrium will be achieved for the first time at the scale-up
`stage. On the other hand, the need for high yield frequently is the chief
`motivating force early in a development program, so even the first
`scale-up phase often fails to produce the thermodynamically preferred
`polymorph. Eventually, either at the first scale-up site or when the pro-
`cess is moved to a new site, the thermodynamically preferred form will
`appear. This observation has been attributed to Gay-Lussac, who noted
`that unstable forms are frequently obtained first, and that these subse-
`quently transform to stable forms.
`.
`Eventually, either at the first scale-up site or when the process is
`moved to a new site, the thermodynamically preferred form will appear.
`
`Page 5 of 33
`
`

`
`334
`
`Brittain and Fiese
`
`When this change occurs late in development, retesting of the substance
`is rcquired, to correct the analytical profile of the compound, as well
`as expensive clinical or toxicology testing. In the competitive environ-
`ment of today, time is money, and delaying an NDA can be as costly
`as the expense of additional clinical or toxicological testing. It is the
`appearance of a thermodynamically preferred polymorph late in the
`development cycle that often requires the use of seed crystals during
`’ processing. One occasionally runs into legends of “whiffle dust,” or
`seed crystals circulating in the air-handling system, which lead to the
`occurrence of the disappearing polym.orphs discussed earlier in Chap-
`ter 1. Experienced process chemists who have tried to generate an un-
`stable polymorph for an analytical standard generally support the ten-
`ets of this legend and contribute to its dissemination.
`
`Ill. EFFECTS OF PARTICLE SIZE REDUCTION
`
`The last processing step in the production of bulk drug substances usu-
`ally involves milling to reduce the particle size distribution of the mate-
`rial. This is ordinarily conducted using the mildest conditions possible
`to render a sample homogeneous, or through the use of more rigorous
`milling to reduce the primary particle size in an effort to improve for-
`mulation homogeneity or dissolution and bioavailability. In the latter
`process, a substantial amount of energy is used to process the substance,
`which can result either in a polymorphic conversion or in the generation
`of an amorphous substance. The formation of an amorphous material
`is highly undesirable, since it will often be hygroscopic and more water
`soluble than any of the crystalline forms. The amorphous material must
`also be considered as being a thermodynamically metastable state of
`high energy, and a variety of pathways exist that can result in a back-
`conversion to a crystalline form of the material. This will certainly alter
`the dissolution characteristics and possibly even the bioavailability of
`the drug. In any case, the use of any rigorous milling process requires
`careful analysis for consequent changes in crystallinity or crystal form.
`Hancock and Zografi have reviewed the characteristics of the
`amorphous state and methods whereby this form of a given material
`may be obtained [4], and this information is also covered in the final
`chapter of this book. Amorphous materials are disordered, have the
`
`. 3I
`
`Page 6 of 33
`
`

`
`Effects of Pharmaceutical Processing
`
`335
`
`highest free energy content, have the highest water solubility, and are
`usually hygroscopic. The effect of amorphous materials or amorphous
`regions within pharmaceutical solids is to convert the normal solid
`properties of high elasticity and brittleness to varying degrees of visco-
`elasticity that allow the materials to flow under the mechanical stress of
`milling or tableting. Upon heating, amorphous materials pass through a
`glass transition temperature and convert to a crystalline state. The effect
`of adsorbed moisture from the atmosphere or exeipients is to act as a
`plasticizer, lowering the glass transition temperature, increasing molec-
`ular mobility in the solid, and allowing crystallization to occur at a
`lower temperature. Thus any rigorous milling of a drug substance re-
`quires careful analysis and monitoring for subsequent changes in crys-
`tallinity.
`Much of the literature on the milling of polymorphs involves long
`grinding or ball-milling times, which impart unusual thermodynamic
`stress» to the material and possibly lead to confusion about the real ef-
`fects of milling on polymorphism. In practice, drugs are typically
`milled in a Bantam Mill or a Fitzpatrick Mill (alone or with excipients)
`where the exposure time of the material to stress is very short. One
`would predict that the thermodynamic impact would be very small, and
`therefore one might expect to observe little or no change in polymer-
`phism arising from effects of industrial-scale milling.
`Nevertheless, one can glean an understanding as to the effect of
`milling on polymorphism from ball-rnilling studies, such as these con-
`ducted by Miyamae et al. on (E)-6-(3,4-dimethoxy~phenyl)-1-ethyl-4-
`mesitylimino-3—methyl-3,4-dihydro—2(1H)-pyrimidinone [5].
`In this
`work, the title compound was milled for up to 60 min and then stored
`at ambient conditions for up to 2 months. It was found that the unstable M
`Form A (melting point 118°C) was converted to a noncrystalline solid
`as a result of the milling process but then crystallized upon storage to
`the stable Form B (melting point 141°C).
`Fostedil exists in two polymorphic forms, characterized by melt-
`ing points of 95.3°C (Form 1) and 96.4°C (Form 11), a free energy
`difference of only 71.8 cal/mol at 37°C, and distinctly different infrared
`absorption and x-ray powder diffraction patterns [6]. Solubility studies
`suggested that Form 1 was more stable than Form II. Mechanical mix-
`ing in an automated mortar showed that complete conversion of Form
`II to Form I occurred in 2 h. Milling fostedil in an industrial fluid energy
`
`;
`
`Page 7 of 33
`
`

`
`336
`
`Brittain and Fiese
`
`mill and a hammer mill resulted in a similar conversion of Form II to
`
`Form 1, presumably through the generation of nuclei on the surface of
`the crystal. The hammer milled material was exposed to less energy,
`which resulted in fewer seed nuclei and therefore less conversion to
`
`Form I even when exposed to elevated temperature and humidity. In
`contrast, since the fluidized energy mill provides more energy, its use
`generated more surface defect nuclei and greater conversion to the
`Form I polymorph. The presence of excipients also was found to exert
`a strong perturbation on the kinetics of the phase transformation. As
`illustrated in Fig. 1, the presence of microcrystalline cellulose retarded
`the conversion of Form 11 into Form I with grinding.
`Chloramphenicol palmitatc Form B was found to transform to the
`less therapeutically desirable Form A during grinding [7]. The transfor-
`mation could be accelerated by the presence of appropriate seed crys-
`tals, and it was also found that the least stable Form C could be progres-
`sively converted from Form B and then to Form A if the grinding times
`were sufficiently long. In another study, it was shown that the tempera-
`ture increases that accompanied the grinding process could accelerate
`
`100
`
`0.0
`
`0.3
`
`0.6
`
`0.9
`
`1.2
`
`1.5
`
`1.8
`
`2.1
`
`Time (hours)
`
`Fig. 1. Effect of grinding time on the Form—II-to-Form-I phase transforma-
`tion composition of pure fostedil (O). and also for fostedil/microcrystalline
`cellulose mixtures having ratios of 3:1 (A), 121 (V), and 123 (O). (The figure
`is adapted from data presented in Ref. 6.)
`
`Page 8 of 33
`
`

`
`Effects of Pharmaceutical Processing
`
`337
`
`3
`
`the phase conversion [8]. In fact, when one grinds chloramphenicol
`stearate for a sufficiently long period of time, Form [[1 first turns into
`Form 1, and ultimately into zm amorphous state [9]. One interesting
`note of this latter study was the finding that the rate of phase conversion
`was accelerated by the presence of microcrystalline cellulose.
`Clearly, the temperature maintained during the milling process
`can influence the degree of any polymorphic transitions. The ot-and 7-
`forms of indomethacin could be converted to an amorphous solid dur-
`ing grinding at 4°C, but at 30°C the ‘y-form converted to the 0t—form,
`which could not be further transformed [10,11]. The authors concluded
`that although the noncrystalline form was stable at 4°C, it evidently
`was unstable with respect to crystallization at 30°C. An illustration of
`the crystal form present in samples of indomethacin as a function of
`grinding time is provided in Fig. 2.
`The transformation behavior of phenylbutazone polymorphs dur-
`ing grinding at 4 and 35°C, and the solid-state stability and dissolution
`behavior of the ground materials, was investigated [12]. The 0t, [3, and
`8 forms were transformed to the new C,-form, which in turn was trans-
`formed to the e-form (which was stable at 4°C). On the other hand,
`
`100
`
`40|
`
`
`
`
`
`Phasecement(5%) 38
`
`012345678910
`
`TIme(hours)
`
`Fig. 2. Effect of grinding time (conducted at 30°C) on the phase composi-
`tion of 7-indomethacin (0), showing also the ot-phase (I) formed. (The figure
`is adapted from data presented in Ref. 11.)
`
`i
`
`Page 9 of 33
`
`

`
`338
`
`
`
`Brlttain and Fiese
`
`during grinding at 35°C, the 5-form was not changed, while the oi-form
`was transformed to the 8-form by way of the C-form. The B-form was
`apparently transformed directly to the 5-form. Once obtained by grind-
`ing, the e-form was transformed to the 8-form under 0% relative humid-
`ity at various storage temperatures after an induction time of a few
`hours.
`
`A common occurrence during grinding is the formation of an
`amorphous, noncrystalline phase. For instance, the grinding of cepha-
`lexin has been found to decrease the degree of crystallinity [13] as well
`as a wide range of physical properties that depend on the crystalline
`content of the material [14]. When the material becomes less crystal-
`line, its stability decreases, and the hydrate forms become easier to
`dehydrate. Similar conclusions were reached with respect
`to the
`strength of the crystal lattice when the effect of grinding on the stability
`of eefixime trihydrate was studied [15]. The degree of interaction be-
`tween the waters of hydration and the cefixime molecules was weak-
`ened by grinding, negatively affecting the solid-state stability.
`Four modifications of cimetidine were prepared as phase-pure
`materials, and each was found to be stable during dry storage [16]. The
`milling process enabled the conversions of Forms B and C to Form A,
`while Form A transformed into Form D only upon nucleation. In all
`cases, the particle size reduction step resulted in substantial formation
`of the amorphous phase. Compression of the various forms into com-
`pacts did not yield any evidence for phase conversion.
`. The pitfalls that can accompany the development of the amor-
`phous form of a drug substance were shown for the capsule formulation
`of
`(3R,4S)-l ,4-bis(4-methoxy-phenyl)-3-(3-phenylpropyl)-2-azetidi-
`none [17]. A variety of spectroscopic techniques were used to study
`the amorphous-to-crystalline phase transition. and it was determined
`that problems encountered with dissolution rates could be related to
`the amount of crystalline material generated in the formulation.
`One general finding of milling studies conducted on hydrate spe-
`cies is that the grinding process serves to lower the dehydration temper-
`atures of ground materials, facilitating the removal of lattice water and
`formation of an amorphous product. This behavior has been noted for
`cephalexin [13,14]. cefixime [15], cyclophosphamide [18], 2-[(2-meth-
`ylirnidazoyl-1-yl)-methyl]-benzo[f]thiochromen-1-one [19], and lacti-
`to] [20], to name a few representative examples.
`
`Page 10 of 33
`
`

`
`Effects of Pharmaceutical Processing
`
`339
`
`
`
`
`
`oz-FormContent(96)
`
`100
`
`'6’.‘33
`
`0
`
`0
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`B
`
`9
`
`10
`
`Time (hours)
`
`Flg. 3. Changes in ot—form content for the ot-monohydrate (I), ot-anhydrate
`(C), and [3-anhydrate (O) modifications of lactose with grinding time. (The
`figure is adapted from data presented in Ref. 22.)
`
`Although most studies have been concerned with the state of the
`drug entity in either bulk or dosage form, milling can also affect the
`crystalline state of excipients. For example, isomerization was found
`to take place during the grinding of the ot-monohydrate, on-anhydratc,
`and [3-anhydratc forms of lactose [21]. The crystalline materials were
`all transformed into amorphous phases upon grinding, but these materi-
`als quickly re-sorbed water to approach the equilibrium composition
`ordinarily obtained in aqueous solutions. This behavior has been illus-
`trated in Fig. 3. In a subsequent study [22], these workers determined
`the degree of crystallinity during various stages of the grinding process
`and concluded that while the isomerization rate of the ot-monohydrate
`phase depended on the crystallinity, the rates for the two anhydrate
`phases depended on the content of absorbed water.
`
`IV. EFFECTS DUE TO GRANULATION
`
`When the requirements of a given formulation require a granulation
`step, this will represent an opportunity for a change in crystal form of
`
`
`
`Page 11 of 33
`
`

`
`
`
`340
`
`Brittaln and Flese
`
`the drug to take place. The problem will be most acute when a wet
`granulation step is used, since a potential transforming solvent is added
`to the blend of drug and excipients. The mixture is physically processed
`in the wet stage until homogeneity is obtained and then dried at a fairly
`significant temperature. The rationale for including a granulation step
`is to improve How and blend homogeneity, permitting the high-speed
`compression of tablets. The granulation process is used to maintain the
`distribution of the drug throughout the formulation even during free
`flow, thus blend homogeneity.
`Clearly, one can consider a wet granulation to be equivalent to a
`suspension of the drug entity in a mixture of solvent and excipients.
`Since the usual solvent is water, one can encounter a variety of inter-
`conversions between anhydrates and hydrates, or between hydrates and
`hydrates, which are mediated by the presence of the solvent. It is
`equally clear that one should not expect to be able to wet-granulate the
`metastable phase of a particular compound if that metastable phase is
`capable of transforming into a more stable form. A discussion of so]-
`vent-mediated phase transformations has been given in an earlier chap-
`ter and need not be repeated here.
`It has been noted that chlorpromazine hydrochloride Form II ex-
`hibits severe lamination and capping when compressed, and that wet
`granulation with ethanol and water significantly improves the tableting
`characteristics [23]. The reason for this process advantage was shown
`to entail a phase change of the initial Fonn 1] to the more stable Form
`I during the granulation step. It was concluded that the improvements
`in tabletability and tablet strength that followed the use of wet granula-
`tion were due to changes in lattice structure that facilitated interparticu-
`late bonding on compaction.
`The effects of granulation solvent on the bulk properties of gran-
`ules prepared using different polymorphic forms of carbamazepine
`have been studied [24]. It was found that although the anhydrate phase
`of the drug transformed to the dihydrate phase in the presence of 50%
`aqueous ethanol, the phase transformation did not take place when ei-
`ther pure water or pure ethanol was used as the granulating solution.
`This unusual finding was attributed to the fact that the solubility of the
`drug in 50% ethanol was 37 times higher than that in pure water. Gran-
`ules processed using 50% ethanol were harder than those processed
`with pure solvents and led to the production of superior tablets. This
`
`Page 12 of 33
`
`

`
`Effects of Pharmaceutical Processing
`
`341
`
`work demonstrates the extreme need to evaluate the robustness of the
`
`granulation step during process development.
`The crystal characteristics of a drug substance can undergo a se-
`quence of reactions during wet granulation and subsequent processing.
`When subjected to a wet granulation step where 22% w/w water was
`added to dry solids, and the resulting mass dried at 55°C, an amorphous
`form of (S)-4-[[[ l.-(4-fluorophenyl)-3-(1-methylethyl)-1H-indol-2-ylJ-
`ethynyl]-hydroxyphosphinyl]-3]-hydroxybutanoic acid, disodium salt,
`was obtained [25]. However, this amorphous gradually became crystal.-
`line upon exposure to humidity conditions between 33 and 75% RH.
`Depending on the exact value of the applied humidity, different hy-
`drates of the drug substance could be obtained in the formulation.
`
`V. EFFECTS DUE TO DRYING
`
`Whenever a wet suspension containing a drug substance is dried, the
`possibility exists that a change in crystal state will take place. It has
`been amply shown in earlier chapters that anhydrates or amorphates
`can be produced by the simple desolvalion of a solvate species, and
`such reactions are always possible during any drying step. The produc-
`tion of anhydrate phases by simple drying is one example of a reaction
`that can accompany the drying process, and this reaction type has been
`fully discussed in an earlier chapter.
`In the simplest type of drying procedure, moist material from the
`wet granulation step is placed in coated trays and the trays are placed
`in drying cabinets with a circulating air current and thermostatic heat
`control. Historically, the tray drying method was the most widely used T
`method, but more recently, fluid-bed drying is now widely used. In
`drying tablet granulations by fluidization, the material is suspended and
`agitated in a warm air stream. The chief advantages of the fluid-bed
`approach are the speed of drying and the degree of control that can be
`exerted over the process. The wet granulations are not dried to zero
`moisture; for a variety of reasons one seeks to maintain a residual
`amount of moisture in the granulation. It is evident from these simple
`considerations that the combination of solvent and drying conditions
`provides a suitable environment for the generation of new polymorphs
`or solvates when such conversion routes are available.
`
`I
`I
`
`Page 13 of 33
`
`

`
`Te‘
`
`._'-'"_;;:
`
`1,;
`"
`
`‘=
`
`342
`
`Brlttaln and Fiese
`
`During ordinary manufacturing processes, crystals of the drug
`substance are dried under vacuum or through the circulation of warm
`dry air. Evaluation of wet and dried samples of the bulk substance by
`thermogravimetric analysis and hygroseopicity studies usually yields
`the necessary insight into the limiting parameters of the drying process.
`Since drying times can be relatively long and the environmental tem-
`H perature mild (such as 24 to 48 h at 60°C), partial or total polymorphic
`conversion is possible. The likelihood that phase conversion will take
`place becomes increasingly less as the solvent is removed during dry-
`ing. This area has been thoroughly covered earlier in Chapter 5, so
`only a few illustrative examples will be quoted at this time.
`Shefter et al. used x-ray diffraction analysis to show that over-
`drying ampicillin trihydrate resulted in an unstable amorphous product,
`and that the drying of theophylline hydrate yielded a crystalline anhy-
`drous form [26]. Kitamura et al. dehydrated (with vacuum) cetixime
`trihydrate to obtain a crystalline anhydrous form, which also proved
`to be unstable [27].
`Otsuka et al. studied the effect of humidity and drying on the
`polymorphic transformation rate at 45°C for six forms of phenobarbital
`[28]. Figure 4 shows that the monohydrate (Form C) and the hemihy-
`drate (Form E) slowly convert to the anhydrous Form B under high-
`humidity stress conditions (45°C and 75% RH), while the conversion
`is very fast at low humidity (45°C and 0% RH). This is reasonable,
`since the driving force to form anhydrous material is a dry environment.
`Thepkey point of Figure 4 is the rapid conversion of phenobarbitol in
`less than 12 at 45°C and 0% RH, which represents very reasonable
`processing conditions for an industrial drying oven.
`The sodium salt of 3-(((3-(2-(7-ch1oro-2- quinolinyl)ethenyl)phe-
`nyl)((3-dimethylamino)-3-oxo-propyl)thio)propanoic acid represents
`an unusual case of a hygroscopic drug substance, where the surface-
`active substance (in both crystalline and lyophilized forms) produced
`nonflowing, semisolid masses with exposure to increasing relative hu-
`midity [29]. The substance was shown to form first an amorphous mate-
`rial and then a mesomorphic phase, as the moisture sorption increased
`with exposure to relative humidity.
`The favorite endpoint of the process chemist is to “dry to constant
`weight,” but this laudable goal can lead to the production of a desol-
`
`Page 14 of 33
`
`

`
`Effects of Pharmaceutical Processing
`
`0.4
`
`0.3
`
`0.1
`
`0.0
`
`0
`
`(1-(1-X)"’)’
`
`1
`
`2
`
`3
`
`4
`
`5
`
`Time (days)
`
`Jander plots of one polymorphic transformation of phenobarbital.
`Flg. 4.
`Illustrated are data for the conversion at 45°C of Form C (monohydrate phase)
`to Form B (anhydrate phase). The 0% relative humidity atmosphere points
`are given by the open symbols, and the 75% relative humidity atmosphere
`points are given by the closed symbols. (The figure is adapted from data pre-
`sented in Ref. 28.)
`
`vated product. Without doubt, the overdrying process has produced
`many anhydrous lots of bulk drug substance, which unfortunately are
`eventually found to be unstable.
`Since the effects of simple drying have been more fully discussed
`in detail in Chapter 5 of this book, we will turn to case studies that
`illustrate how the use of two specialized drying methods can lead to
`phase intereonversions.
`
`A. Changes in Crystalline Form Accompanying the
`Spray-Drylng Process
`
`The spray-diying process first requires the formation of a slurry to be
`sprayed, which ca11 be a concentrated solution of the agent to be dried
`or a dispersion of the agent into a suitable nondissolving medium. The
`dispersion is then atomized into droplets, which are exposed to a heated
`atmosphere to effect the drying process. Completion of the process
`yields a dry, freefiowing powder that ordinarily consists of spherical
`
`!
`
`,'
`
`i?
`
`.
`
`Page 15 of 33
`
`

`
`344
`
`Brittain and Fiese
`
`particles of relatively uniform particle size distribution. For instance,
`the morphology of spray-dried lactose has been contrasted with that
`of the monohydrate and anhydrous phases of lactose, with significant
`differences being reported [30].
`One interesting feature of spray-drying is the possibility of con-
`trolling the final crystal form through manipulation of the drying condi-
`tions. For instance, three different crystalline forms of phenylbutazone
`were prepared from methylene chloride solution by varying the drying
`temperature of the atomized droplets between 30 and 120°C [31]. One
`of the isolated polymorphs could not be obtained by any crystallization
`procedure and could only be produced using extremely slow solvent
`evaporation rates.
`Using phenobartitone and hydroflumethiazide as examples, it has
`been shown that high-energy solids can be produced through the use
`of spray-drying [32]. Commercially available phenobartitone is ob-
`tained as Form II, but with spray-drying the authors were able to obtain
`Form 111 having a large specific surface area. The analogous spray-
`drying of hydroflumethiazide yielded an amorphous solid. The spray
`drying of indomethacin produced a viscous glassy phase, which was
`found to be physically unstable with time [33]. Upon storage, the glassy
`solid converted into a mixture of crystalline Forms I and H. Representa-
`tive x—ray powder diffraction patterns of these materials are shown in
`Fig. 5.
`Spray-drying is often used to encapsulate drug substances into
`excipient matrices, but phase interconversions are still possible in such
`situations. Sulfamethoxazole Form I was microencapsulated with cellu-
`lose acetate phthalate and talc, colloidal silica, or montmorillonite clay
`by a spray-drying technique, and the drug was found to convert to Form
`11 during the process [34]. Increasing the amount of cellulose acetate
`phthalate in the formulation led to increased amorphous drug content,
`but increasing the tale level yielded more polymorphic conversion. In
`a subsequent study, the authors showed that no phase conversion took
`place in formulations containing colloidal silica [35].
`Frequently, the spray drying process yields amorphous materials,
`which undergo crystallization upon storage or exposure to suitable en-
`vironmental conditions. The effect of various additives on the recrystal-
`lization of amorphous spray-dried lactose has been studied [36]. The
`
`Page 16 of 33
`
`

`
`Effects of Pharmaceutical Processing
`
`10
`
`1!
`
`20
`
`25
`
`Scattering Angie (deg. 2-8)
`
`Fig. 5. X-ray powder diffraction patterns for (A) crystalline indotnethacin,
`(B) spray—dried indornethacin, and (C) indomethacin spray—dried with 5%
`polyvinylpyrrolidone and stored for two months. (The figure is adapted from
`data presented in Ref. 32.)
`
`presence of magnesium stearate was found to inhibit the recrystalliza-
`tion process, while layers of microcrystalline cellulose in the spray-
`dried products could result in long lag times prior to recrystallization.
`It was deduced that the onset of crystallization was critically related
`to the mobility of water in the formulations.
`
`B. Changes in the crystalline State of Lyophllized
`Products
`
`Another drying procedure that can yield changes in drug crystal form
`is that of freeze—drying, or lyophilization. In this approach, the material
`
`Page 17 of 33
`
`

`
`
`
`346
`
`Brlttaln and Flese
`
`to be dried is prepared as an aqueous solution or suspension and then
`frozen rapidly and cooled to a temperature below its eutectic point.
`The frozen formulation is then exposed to vacuum and the ice removed
`by sublimation. Since lyophilized products are often dried to moisture
`levels less than 1%, the materials are ordinarily hygroscopic and must
`be protected from adventitious water to prevent any unwanted phase
`transformation steps. In usual practice, the product is produced in its
`amorphous state, which is favorable for its subsequent solubilization
`at the time of its intended use. Consequently, the study of any possible
`moisture—induced crystallization is important to the characterization of
`a lyophilized product.
`The effect of storage conditions on the crystalline nature of lyoph-
`ilized ethacrynate sodium has been reported [37]. Samples of the drug
`substance were dried to various moisture levels and then stored at either
`
`60°C or 30°C/75% RH. It was found that crystalline drug substance
`was obtained after short periods of time as long as sufficient levels of
`water were either present in the initial lyophilized solid or introduced
`by adsorption of humidity. The crystalline form of the drug was identi-
`fied as a monohydrate phase.
`Cefazolin sodium is capable of existing in a number of crystalline
`modifications, but the amorphous state is produced by the lyophiliza-
`tion process [38]. The amorphous form retained its nature at relative
`humidities less than 56% but converted completely to the pentahydrate
`above 75% RH. Interestingly, the hygroscopicity of the amorphous
`formwas essentially similar to that exhibited by the dehydrated mono-
`hydrate or dehydrated on-form. Lamotrigine mesylate was also found to
`exhibit a moisture-dependent amorphous-to-crystalline transition [39].
`The transition was found to be independent of the presence of mannitol,
`whether the amorphous solid was produced by lyophilization or by
`spray-dryin