Pressure
`Pressure
`
`_ «:30gas r52:mx.3w»
`
`;,
`353%EZEOEEEQEE.__..EEgSm
`2252.35...5
`
`Avmom‘_9.#3
`
`

`

`ISBN: 0-8247-0237-9
`
`This book is printed on acid-free paper.
`
`Headquarters
`Marcel Dekker, Inc.
`270 Madison Avenue, New York, NY 10016
`tel: 212-696-9000; fax: 212-685-4540
`
`Eastern Hemisphere Distribution
`Marcel Dekker AG
`Hutgasse 4, Posffach 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 5 4 3 2 1
`
`PRINTED IN THE UNITED STATES OF AMERICA
`
`Lupin Ex. 1032 (Page 2 of 46)
`
`

`

`This materiai may be p,~otected by Copyright law (Title 17 U S. Code)
`
`5
`
`Generation of Polymorphs,
`Hydrates, Solvates, and Amorphous
`Solids
`
`J. Keith Guillory
`
`The University of Iowa
`Iowa City, Iowa
`
`METHODS EMPLOYED TO OBTAIN UNIQUE
`POLYMORPHIC FORMS
`A. Sublimation
`B. Crystallization from a Single Solvent
`C. Evaporation from a Binary Mixture of Solvents
`D. Vapor Diffusion
`E. Thermal Treatment
`F. Crystallization from the Melt
`G. Rapidly Changing Solution pH to Precipitate Acidic or
`Basic Substances
`H. Thermal Desolvation of Crystalline Solvates
`I. Growth in the Presence of Additives
`J. Grinding
`
`II.
`
`METHODS EMPLOYED TO OBTAIN HYDRATE FORMS
`
`III.
`
`METHODS EMPLOYED TO OBTAIN SOLVATE FORMS
`
`184
`186
`188
`194
`195
`195
`197
`
`198
`199
`201
`202
`
`202
`
`205
`
`183
`
`Lupin Ex. 1032 (Page 3 of 46)
`
`

`

`184
`
`Guillory
`
`IV. METHODS EMPLOYED TO OBTAIN AMORPHOUS
`MATERIALS
`A. Solidification of the Melt
`B. Reduction of Particle Size
`C. Spray-Drying
`D. Lyophilization
`E. Removal of Solvent from a Solvate or Hydrate
`F. Precipitation of Acids or Bases by Change in pH
`G. Miscellaneous Methods
`
`V. SUMMARY
`
`REFERENCES
`
`208
`209
`210
`213
`213
`215
`217
`218
`
`219
`
`219
`
`I. METHODS EMPLOYED TO OBTAIN UNIQUE
`POLYMORPHIC FORMS
`
`Organic medicinal agents that can exist in two or more solid phases
`often can provide some distinct advantages in particular applications.
`The metastable solid may be preferred in those instances where absorp-
`tion of the drug is dissolution rate dependent. The stable phase may
`be less susceptible to chemical decomposition and may be the only
`form that can be used in suspension formulations. Often a metastable
`polymorph can be used in capsules or for tableting, and the thermody-
`namically stable form for suspensions. Factors related to processing,
`such as powder flow characteristics, compressibility, filterability, or hy-
`groscopicity, may dictate the use of one polymorph in preference to
`another. In other cases, a particular form may be selected because of
`the high reproducibility associated with its isolation in the synthetic
`procedure.
`It is essential to ascertain whether the crystalline material that
`results from a synthetic procedure is thermodynamically stable before
`conducting pivotal trials, since a more stable form may be obtained
`subsequently, and it may be impossible to produce the metastable form
`in future syntheses. Conversion from one polymorph to another can
`occur during processing or upon storage. An additional incentive for
`
`Lupin Ex. 1032 (Page 4 of 46)
`
`

`

`Generation of Polymorphs
`
`185
`
`isolating and identifying polymorphs that provides certain advantages
`is the availability of subsidiary patents for desirable polymorphic
`forms, or for retaining a competitive edge through unpublished knowl-
`- edge. In 1990 Byrn and Pfeiffer found more than 350 patents on crystal
`forms granted on the basis of an advantage in terms of stability, formu-
`lation, solubility, bioavailability, ease of purification, preparation or
`synthesis, hygroscopicity, recovery, or prevention of precipitation [1].
`One question that is likely to arise during the registration process
`is "What assurance can be provided that no other crystalline forms of
`this compound exist?" It is incumbent on the manufacturer of a new
`drug substance to show that due diligence has been employed to isolate
`.and characterize the various solid-state forms of a new chemical entity.
`This may seem to be a daunting task, particularly in light of the widely
`quoted statement by Walter C. McCrone [2] that "Those who study
`polymorphism ate rapidly reaching the conclusion that all compounds,
`organic and inorganic, can crystallize in different crystal forms or poly-
`morphs. In fact, the more diligently any system is studied the larger
`the number of. polymorphs discovered." On the other hand, one can
`take comfoi~ from the fact that some important pharmaceuticals have
`been in use for many years and have, at least until now, exhibited only
`one stable form. Indeed, it seems to this author that there must be partic-
`ular bonding arrangements of some molecules that are so favorable
`energetically as to make alternate arrangements unstable or nonisolat-
`able.
`
`In the future, computer programs, using force-field optimization
`should be perfected to the point where it will be possible to predict,
`with confidence, that a particular crystalline packing arrangement is
`the most stable that is likely to be found. These programs also may
`make it possible to predict how many alternate arrangements having
`somewhat higher energy can potentially be isolated [3,4]. Until that
`time, the developmental scientist is handicapped in attempting to pre-
`dict how many solid forms of a drug are likely to be found. The situa-
`tion is further complicated by the phenor6enon of "disappearing poly-
`morphs" [5], or metastable crystal forms that seem to disappear in
`favor of more stable ones.
`Some polymorphs can be detected, but not isolated. Hot stage
`microscopy has been used extensively to study polymorphic transfor-
`
`Lupin Ex. 1032 (Page 5 of 46)
`
`

`

`186
`
`Guillory
`
`mations. The microscopist can detect numerous polymorphic transfor-
`mations, but the individual polymorphs often prove to be so unstable
`that they cannot be isolated by the usual methods. An excellent example
`of this is the work of Griel3er and Burger on etofylline [6]. These au-
`thors identified five polymorphic forms by thermomicroscopy, but only
`stable Modification I could be obtained by recrystallization, even when
`seed crystals from the hot stage were used. Similarly, Kuhnert-Brand-
`statter, Burger, and V611enklee [7] described six polymorphic forms of
`piracetam, only three of which could be obtained by solvent crystalliza-
`tion. All the others were found only by crystallization from the melt.
`What, then, is a careful investigator to do?
`In this chapter, the various methods used to isolate polymorphs,
`hydrates, and solvates will be described. As Bernstein [8] has observed,
`"The conditions under which different polymorphs are obtained exclu-
`sively or together also can provide very useful information about the
`relative stability of different phases and the methods and techniques
`that might be necessary to obtain similar structures of different chemi-
`cal systems." In this context, it is hoped that the following information
`will prove useful in devising a "screening" protocol for the preparation
`of the various solid state forms of pharmaceuticals. While one cannot
`be absolutely certain that no additional forms will be identified in the
`future, this approach should provide some assurance that "due dili-
`gence" has been exercised to isolate and identify crystalline forms that
`are likely to arise during the normal course of drug development and
`storage.
`
`A. Sublimation.
`On heating, approximately two-thirds of all organic compounds are
`converted partially from the solid to the gaseous state and back to solid,
`i.e., they sublime [9]. While strictly speaking the term sublimation re-
`fers only to the phase change from solid to vapor without the interven-
`tion of the liquid phase, it is often found that crystals are formed on
`cooler surfaces in close proximity to the melt of organic compounds
`when no crystals were formed at temperatures below the melting point.
`The most comprehensive information concerning sublimation tempera-
`tures of compounds of pharmaceutical interest can be found in tables
`
`Lupin Ex. 1032 (Page 6 of 46)
`
`

`

`Generation of Polymorphs
`
`187
`
`in the textbook of Kuhnert-Brandstiitter [9]. While the information in
`these tables is designed primarily for the microscopic examination of
`compounds, it is also possible to utilize it to determine which com-
`pounds might be susceptible to the application of techniques (such as
`vacuum sublimation) that can be carried out on larger scales and at
`lower temperatures.
`The sublimation temperature and the distance of the collecting
`surface from the material undergoing sublimation have a great influ-
`ence on the form and size of the crystals produced. The occurrence of
`polymorphic modifications depends on the temperature of sublimation.
`In general, it may be assumed that unstable crystals form preferentially
`at lower temperatures, while at higher temperatures stable forms are
`to be expected. Nevertheless, mixtures consisting of several modifica-
`tions are frequently found together. This is the case for barbital and
`for estradiol benzoate. It should be obvious that the sublimation tech-
`nique is applicable only to those compounds that are thermally stable.
`A simple test can be used to determine if a material sublimes. A
`small quantity (10-20 rag) of the solid is placed in a petri dish that is
`covered with an inverted watch glass. The petri dish is heated gently
`on a hot plate and the watch glass is observed to determine if crystals
`are growing on it. According to McCrone [2], one of the best methods
`for obtaining a good sublimate is to spread the material thinly over a
`portion of a half-slide, cover with a large cover glass, and heat slowly
`using a Kofler block. When the sublimate is well formed, the cover
`glass is removed to a clean slide for examination. It is also possible
`to form good crystals by sublimation from one microscope slide to a
`second held above it, with the upper slide also being heated so that its
`temperature is only slightly below that of the lower slide. Cooling of
`the cover slip by placing drops of various low-boiling solvents on the
`top surface will cause condensation of the more unstable forms, the
`lower temperatures leading to the most unstable forms. On a larger
`scale, a glass cold finger or a commercial sublimator can be employed..
`Once crystals of various modifications have been obtained, they can be
`used as seeds for the solution phase crystallization of larger quantities.
`Form I of 9,10-anthraquinone-2-carboxylic acid was obtained as
`needle-like crystals upon sublimation at temperatures exceeding 250°C
`[10]. Fokkens et al. have used sublimation to purify theophylline for
`
`Lupin Ex. 1032 (Page 7 of 46)
`
`

`

`188
`
`Guillory
`
`vapor pressure studies [11]. Sakiyama and Imamura found that stable
`phases of both 1,3-dimethyluracil and malonamide could be prepared
`by vacuum sublimation [12].
`
`B. Crystallization from a Single Solvent
`Slow solvent evaporation is a valuable method for producing crystals.
`Solutions of the material being crystallized, preferably saturated or
`nearly so, are filtered to remove most nuclei and then left undisturbed
`for a reasonable period of time. The rate of evaporation is adjusted by
`covering the solution with aluminum foil or Parafilm® containing a few
`small holes. For a solvent to be useful for recrystallization purposes, the
`solubility of the solute should be on the order of 5-200 mg/mL at room
`temperature. If the solubility exceeds 200 mg/mL, the viscosity of the
`solution will be high, and a glassy product is likely to be obtained. A
`useful preliminary test can be performed on 25-50 mg of sample, add-
`ing a few (5-10) drops of solvent. If all the solid dissolves, the solvent
`will not be useful for recrystallization purposes. Similarly, highly vis-
`cous solvents, and those having low vapor pressures (such as glycerol
`or dimethylsulfoxide) are not usually conducive to efficient crystalliza-
`tion, filtration, and washing operations. The solvents selected for re-
`crystallization should include any with which the compound will come
`into contact during synthesis, purification, and processing, as well as
`solvents having a. range of boiling points and polarities. Examples of
`solvents routinely used for such work are listed in Table 1 together
`with their boiling points.
`The process of solution mediated transformation can be consid-
`ered the result of two separate events, (a) dissolution of the initial
`phase, and (b) nucleation/growth of the final, stable phase. If crystals
`do not grow as expected from a saturated solution, the interior of the
`vessel can be scratched with a glass rod to induce crystallization by
`distributing nuclei throughout the solution. Alternatively, crystalliza-
`tion may be promoted by adding nuclei, such as seed crystals of the
`same material. For example, Suzuki showed that the (x-form of inosine
`could be obtained by crystallization from water, whereas isolation of
`the [3-form required that seeds of the [~-form be used [13].
`If two polymorphs differ in their melting point by 25-50°C, for
`
`Lupin Ex. 1032 (Page 8 of 46)
`
`

`

`Generation of Polymorphs
`
`189
`
`Table 1 Solvents Often Used in the
`Preparation of Polymorphs
`
`Solvent
`
`Dimethylformamide
`Acetic acid
`Water
`1-Propanol
`2-Propanol
`Acetonitrile
`2-Butanone
`Ethyl acetate
`Ethanol
`Isopropyl ether
`Hexane
`Methanol
`Acetone
`Methylene chloride
`Diethyl ether ’
`
`Boiling point
`(°C)
`
`153
`118
`100
`97
`83
`82
`80
`77
`78
`68
`69
`65
`57
`40
`35
`
`monotropic polymorphs thelower melting, more soluble, form will be
`difficult to crystallize. The smaller the difference between the two melt-
`ing points, the more easily unstable or metastable forms can be ob-
`tained.
`A commonly used crystallization method involves controlled tem-
`perature change. Slow cooling of a hot, saturated solution can be effec-
`tive in producing crystals if the compound is more soluble at higher
`temperatures; alternatively, slow warming can be applied if the com-
`pound is less soluble at higher temperatures. Sometimes it is preferable
`to heat the solution to boiling, filter to remove excess solute, then
`quench cool using an ice bath or even a dry ice-acetone bath. High
`boiling solvents can be useful to produce metastable polymorphs.
`McCrone [2] describes the use of high boiling solvents such as benzyl
`alcohol or nitrobenzene for recrystallization on a hot stage. Behme et
`al. [14] showed that when buspirone hydrochloride is crystallized
`above 95°C the higher melting form is obtained; below 95°C the lower
`
`Lupin Ex. 1032 (Page 9 of 46)
`
`

`

`190
`
`Guillory
`
`melting form is obtained. Thus the lower melting polymorph could be
`converted to the higher melting polymorph by recrystallizing from xy-
`lene (boiling point 137-140°C).
`To understand how temperature influences the composition of
`crystals that form, it is useful to examine typical solubility-temperature
`diagrams for substances exhibiting monotropic and enantiotropic be-
`havior [15]. In Fig. la, Form II, having the lower solubility, is more
`stable than Form I. These two noninterchangeable polymorphs are mo-
`notropic over the entire temperature range shown. For indomethacin,
`such a relationship exists between Forms I and II, and between Forms
`II and III.
`In Fig. lb, Form II is stable at temperatures below the transition
`temperature T,, and Form I is stable above T,. At the transition tempera-
`ture the two forms have the same solubility, and reversible transforma-
`tion between enantiotropic Forms I and II can be achieved by tempera-
`ture manipulation. The relative solubility of two polymorphs is a
`
`(a)
`
`Z Form ITi Tt
`
`"EMPERATURE
`
`TEMPERATURE
`
`(c) Z
`
`~ Form I
`
`Form II
`
`TEMPERATURE
`
`Z
`O
`C)
`
`Fig. 1 Solubility curves exhibiting (a) monotropy, (b) enantiotropy, and (c)
`enantiotropy with metastable phases. (Reprinted with permission of the copy-
`right holder [15].)
`
`Lupin Ex. 1032 (Page 10 of 46)
`
`

`

`Generation of Polymorphs
`
`191
`
`convenient measure of their relative free energies. The polymorph hav-
`ing the lower solubility is the more thermodynamically stable form, i.e.,
`the form with the lower free energy at the temperature of the solubility
`measurement. At room temperature, carbamazepine Form I (m.p.
`189°C) is more soluble than is Form III (m.p. 174°C), so the form with
`the higher melting point is more soluble. The polymorphs are enantio-
`tropic with respect to each other [16].
`There are situations in which kinetic factors can for a time over-
`ride thermodynamic considerations. Figure lc depicts the intervention
`of metastable phases (the broken line extensions to the two solubility
`curves). If a solution of composition and temperature represented by
`point X (supersaturated with respect to both I and II) is allowed to
`crystallize, it would not be unusual if the metastable Form I crystallized
`out first even though the temperature would suggest that Form II would
`be the more stable (i.e., less soluble) form. This is an extension of
`Ostwald’ s law of stages [ 17], which states that"when leaving an unsta-
`ble state, a system does not seek out the most stable state, rather the
`nearest metastable state which can be reached with loss of free en-
`ergy." This form then transforms to the next most soluble form through
`a process of dissolution and crystallization. Crystallization of Form I
`when Form II is more stable would be expected if Form I had the faster
`nucleation and/or crystal growth rate. However, if the crystals of Form
`I were kept in contact with the mother liquor, transformation could
`occur as the more soluble Form I crystals dissolve and the less soluble
`Form II crystals nucleate and grow. For crystals that exhibit this type
`of behavior, it is important to isolate the metastable crystals from the
`solvent by rapid filtration so that phase transformation will not occur.
`In the general case, if there are any other polymorphic forms with
`solubiIities below that of Form II, the above-described process will
`continue between each successive pair of forms until the system finally
`contains only the most stable (the least soluble) form. The implication
`of this hypothesis is that, by controlling supersaturation and by harvest-
`ing crystals at an appropriate time, it should be possible to isolate the
`different polymorphic forms. Furthermore, the theory predicts that at
`equilibrium the product of any crystallization experiment must be the
`stable form, regardless of the solvent system. It is apparent, however,
`
`Lupin Ex. 1032 (Page 11 of 46)
`
`

`

`192
`
`Guillory
`
`from the literature that for some solutes it is the choice of solvent rather
`than the effects of supersaturation that determines the form that crystal-
`lizes [181.
`Crystallization of mannitol as a single solute was found to be
`influenced by both the initial mannitol concentration and by the rate
`of freezing [19]. In the range of 2.5% to 15%, the g-polymorph is fa-
`vored by higher concentrations, whereas the ~-polymorph is favored
`at lower concentrations. At constant mannitol concentration (10%), the
`c~-polymorph is favored by a slow freezing rate, whereas the ~5-poly-
`morph is favored by a fast freezing rate.
`Kaneko et al. [20] observed that both the cooling rate and the
`initial concentration of stearic acid in n-hexane solutions influenced
`the proportion of polymorphs A, B, C, and E that could be isolated.
`Garti et al. [21] reported that for stearic acid polymorphs crystallized
`from various organic solvents, a correlation was observed between the
`polymorph isolated and the extent of solvent-solute interaction.
`The reason for using.crystallization solvents having varying po-
`larities is that molecules in solution often tend to form different types
`of hydrogen-bonded aggregates, and that these aggregate precursors
`are related to the crystal structures that develop in the supersaturated
`solution [22]. Crystal structure analysis of acetanilide shows that a hy-
`drogen-bonded chain of molecules is aligned along the needle .axis of
`the crystals. This pattern is characteristic of secondary amides that crys-
`tallize in a trans conformation so that the carbonyl acceptor group and
`the -NH hydrogen bond donor are anti to one another. The morphology
`of acetanilide crystals can be controlled by choosing solvents that pro-
`mote or inhibit the formation of this hydrogen-bond chain. Hydropho-
`bic solvents such as benzene and carbon tetrachloride will not partici-
`pate in hydrogen-bond formation, so they will induce the formation of
`rapidly growing chains of hydrogen-bonded amides. Crystals grown
`by evaporation methods from benzene or carbon tetrachloride are long
`needles. Solvents that are proton donors or proton acceptors inhibit
`chain formation by competing with amide molecules for hydrogen-
`bonding sites. Thus acetone inhibits chain growth at the -NH end, and
`methanol inhibits chain growth at the carbonyl end of the chain. Both
`solvents encourage the formation of rod-like acetanilide crystals, while
`
`Lupin Ex. 1032 (Page 12 of 46)
`
`

`

`Generation of Polymorphs
`
`193
`
`mixtures of benzene and acetone give hybrid crystals that are rod-
`shaped, with fine needles growing on the ends [23].
`Some solvents favor the crystallization of a particular form or
`forms because they selectively adsorb to certain faces of some poly-
`morphs, thereby either inhibiting their nucleation or retarding their
`growth to the advantage of others. Among the factors affecting the
`types of crystal formed are (a) the solvent composition or polarity, (b)
`the concentration or degree of supersaturation, (c) the temperature, in-
`cluding cooling rate and the cooling profile, (d) additives, (e) the pres-
`ence of seeds, (f) pH, especially for salt crystallization, and (g) agitation
`[22].
`
`Marrnez-Oh~rriz et al. [24] found that Form III of diflunisal is
`obtained from poIar solvents, whereas Forms I and IV are obtained
`from nonpolar solvents. Likewise, Wu et al. [25] observed that when
`moricizine hydrochloride is recrystallized from relatively polar sol-
`vents (ethanol, acetone, and acetonitrile), Form I is obtained, whereas
`nonpolar solvents (methylene chloride or methylene chloride/ethyl ace-
`tate) yield Form II.
`In determining what solvents to use for crystallization, one should
`be careful to select those likely to be encountered during formulation
`and processing. Typically these are water, methanol, ethanol, propanol,
`isopropanol, acetone, acetonitrile, ethyl acetate, and hexane. Matsuda
`employed 27 organic solvents to prepare two polymorphs and six sol-
`vates of piretanide [26].
`According to McCrone [27], in a poor solvent the rate of transfor-
`mation of a metastable to a more stable polymorph is slower. Hence
`a metastable form once crystallized can be isolated and dried before it
`is converted to a more stable phase by solution phase mediated transfor-
`mation. In some systems the metastable form is extremely unstable and
`may be prepared only with more extreme supercooling. This is usually
`performed on a very small scale with high boiling liquids so that a
`saturated solution at a high temperature that is suddenly cooled to room
`temperature will achieve a high degree of supersaturation [28].
`There are many examples in the literature of the use of single
`solvents as crystallization screens. Slow crystallization from acetone,
`acetonitrile, alcohols, or mixtures of solvents yields the Form A of
`
`Lupin Ex. 1032 (Page 13 of 46)
`
`

`

`194
`
`Guillory
`
`fosinopril sodium, but rapid drying of a solution of this compound
`yields Form B, sometimes contaminated with a small amount of Form
`A [29]. A rotary evaporator can be used to maintain a solution at the
`appropriate temperature as solvent is being removed.
`Form I of dehydroepiandrosterone was obtained by recrystalliza-
`tion from warm ethyl acetate, acetone, acetonitrile, or 2-propanol. Form
`II was obtained by rapid evaporation, using a vacuum from solutions
`in dioxane, tetrahydrofuran, or chloroform (which are higher boiling,
`less polar solvents) [30].
`
`C. Evaporation from a Binary Mixture of Solvents
`
`If single-solvent solutions do not yield the desired phase, mixtures of
`solvents can be tried. Multicomponent solvent evaporation methods de-
`pend on the difference in the solubility of the solute in various solvents.
`In this approach, a second solvent in which the solute is sparingly solu-
`ble is added to a saturated, solution of the compound in a good solvent.
`Often a solvent system is selected in which the solute is more soluble
`in the component with the higher vapor pressure. As the solution evapo-
`rates, the volume of the solution is reduced and, because the solvents
`evaporate at different rates, the composition of the solvent mixture
`changes.
`Occasionally, crystals are obtained by heating the solid in one
`solvent and then pouring the solution into another solvent or over
`cracked ice. Otsuka et al. [31] obtained phenobarbital Form B by add-
`ing dropwise a saturated solution of the compound in methanol to water
`at room temperature. Form E was obtained by the same technique, but
`by using a saturated solution of phenobarbital in dioxane.
`Kitamura et al. have shown that the fraction of Form A of L-
`histidine decreases quickly when the volume fraction of ethanol in an
`ethanol-water solvent system increases above 0.2, and that pure Form
`B is obtained at a 0.4 volume fraction of ethanol [32]. The transforma-
`tion rate for conversion of Form B to Form A decreases with ethanol
`concentration. The authors postulated that the concentration of the con-
`former that corresponds to Form A decreases more with ethanol con-
`centration than that of Form B, and so the growth rate of Form A will
`also decrease.
`
`Lupin Ex. 1032 (Page 14 of 46)
`
`

`

`Generation of Polymorphs
`
`195
`
`An example of precipitation in the presence of a second solvent
`is seen in the case of indomethacin. The 7-crystal form of indomethacin
`can be obtained by recrystallization from ethyl ether at room tempera-
`ture, but the (x-form is prepared by dissolution in methanol and precipi-
`tation with water at room temperature [33]. Precipitation can also result
`from the addition of a less polar solvent. Form II of midodrine hydro-
`chloride, metastable with respect to Form I, can be prepared by precipi-
`tation from a methanolic solution by means of a less polar solvent such
`as ethyl acetate or dichloromethane [34].
`In Fig. 2, three crystalline modifications of thalidomide are illus-
`trated. These were obtained by solvent recrystallization techniques and
`differ both in crystal habit and in crystal structure. Two of the forms
`were obtained from a single solvent, and one from a binary mixture.
`
`D. Vapor Diffusion
`
`In the vapor diffusion method, a solution of the solute in a good solvent
`is placed in a small, open container that is then stored in a larger vessel
`containing a small amount of a miscible, volatile nonsolvent. The larger
`vessel (often a desiccator) is then tightly closed. As solvent equilibrium
`is approached, the nonsolvent diffuses through the vapor phase into the
`solution, and saturation or supersaturation is achieved. The solubility
`of the compound in a precipitant used in a two-solvent crystallization
`method such as vapor diffusion should be as low as possible (much
`less than 1 mg/mL), and the precipitant (the solvent in which the com-
`pound is poorly soluble) should be miscible with the solvent and the
`saturated solution. The most frequent application of this technique is
`in the preparation of single crystals for crystallographic analysis. An
`illustration of the technique is provided in Fig. 3 [35].
`
`E. Thermal Treatment
`
`Frequently when using differential scanning calorimetry as an analysis
`technique, one can observe an endothermic peak corresponding to a
`phase transition, followed by a second endothermic peak corresponding
`to melting. Sometimes there is an exothermic peak between the two
`endotherms, representing a crystallization step. In these cases it is often
`
`Lupin Ex. 1032 (Page 15 of 46)
`
`

`

`c)
`
`Fig. 2 Three crystalline modifications of thalidomide obtained by solvent
`recrystallization. (A) Form I obtained as bipyramids by slow crystallization
`of thalidomide in 1:1 dimethylformamide:ethanol at room temperature. (B)
`Form II obtained by immersing a saturated solution of thalidomide in acetoni-
`trile in an ice bath. (C) Form III prepared as tabular crystals from a solution
`in boiling 1,4-dioxane, filtered, then allowed to cool to room temperature.
`(Photomicrographs courtesy of Dr. S. A. Botha, the University of Iowa.)
`
`Lupin Ex. 1032 (Page 16 of 46)
`
`

`

`Generation of Polymorphs
`
`Vapor
`containment .-----.~-
`jar
`
`solvent vapor
`
`Cargille Microbeaker
`0.5 mL or 0.1 mL
`
`crystals of drug
`
`197
`
`vapor of
`non-solvent
`
`drug dissolved
`in solvent
`
`non-solvent
`
`Fig. 3 Crystallization by vapor diffusion. (Reproduced with permission of
`the author [35] and the copyright holder, Pfizer, Inc.)
`
`drug solution in solvent
`
`possible to prepare the higher melting polymorph by thermal treatment.
`Thus chlorpropamide Form A is obtained by recrystallization from eth-
`anol solution, but Form C is obtained by heating Form A in an oven
`maintained at 100°C for 3 hours [36]. While the [3-form of tegafur is
`obtained by the evaporation of a saturated methanol solution, the y-
`form is obtained by heating the [3-form at 130°C for one hour [37].
`Form II of caffeine is prepared by recrystallization from distilled water,
`but Form I is prepared by heating Form II at 180°C for 10 hours [38].
`
`F. Crystallization from the Melt
`
`In accordance with Ostwald’s rule [17], the cooling of melts of poly-
`morphic substances often first yields the least stable modification,
`which subsequently rearranges into the stable modification in stages.
`Since the metastable form will have the lower melting point, it follows
`that supercooling is necessary to crystallize it from the melt. After melt-
`ing, the system must be supercooled below the melting point of the
`metastable form, while at the same time the crystallization of the more
`stable form or forms must be prevented. Quench cooling a melt can
`
`Lupin Ex. 1032 (Page 17 of 46)
`
`

`

`198
`
`Guillory
`
`sometimes result in formation of an amorphous solid that on subsequent
`heating undergoes a glass transition followed by crystallization [39].
`On a somewhat larger scale, one can use a vacuum drying pistol
`and a high boiling liquid such as chlorobenzene to achieve the desired
`end. Form II of p-(IR,3S)-3-thioanisoyl-l,2,-2-trimethylcyclopentane
`carboxylic acid was obtained by recrystallization from a 50:50 v/v
`benzene:petroleum ether mixture. Form I then was obtained by melting
`Form II in the vacuum drying pistol [40]. Caffeine Form I is prepared
`by heating Form II at 180°C for 10 hours [38]. Yoshioka et al. [41]
`observed that when the amorphous solidified melt of indomethacin was
`stored at 40°C, it partly crystallized as the thermodynamically stable
`y-form. Yet at 50°C, 60°C, and 70°C, mixtures of the (z- and the y-
`form were obtained. Sulfathiazole Form I is obtained by heating Form
`III crystals (grown from a dilute ammonium hydroxide solution at room
`temperature) at 170°C for 30-40 minutes [421.
`
`G. Rapidly Changing Solution pH to Precipitate
`Acidic or Basic Substances
`
`Many drug substances fall in the category of slightly soluble weak
`acids, or slightly soluble weak bases, whose salt forms are much more
`soluble in water. Upon addition of acid to an aqueous solution of a
`soluble salt of a weak acid, or upon addition of alkali to an aqueous
`solution of a soluble salt of a weak base, crystals often result. These
`crystals may be different from those obtained by solvent crystallization
`of the weak acid or weak base. Nucleation does not necessarily com-
`mence as soon as the reactants are mixed, unless the level of supersatu-
`ration is high, and the mixing stage may be followed b