Solid-State Chemistry
`of Drugs
`
`~ECOND EDITION
`
`Stephen R. Byrn
`Ralph R. Pfeiffer
`Joseph G. Stowell
`
`SSCI, Inc. ¯ West. L.afayette, Indiana
`www.ssc~qnc.com
`
`Lupin Ex. 1034 (Page 1 of 17)
`
`

`

`SSCI, Inc., 3065 Kent Avenue, West Lafayette, Indiana 47906-1076
`www.ssci-inc.com
`
`Second Edition © 1999 SSCI, Inc. Published 1999. All Rights Reserved.
`
`Printed in the United States of America
`Printing History: 03 02 01 00 99 5 4 3 2 1
`
`Neither this book nor any part may be reproduced or transmitted in any form or by any means, elec-
`tronic or mechanical, including photocopying, mierofilming, and recording, or by any information
`storage and retrieval system, without pe~nission in writing from the publisher.
`
`The citation of trade names or names of manufacturers in this book is not to be construed as an en-
`dorsement or as approval by SSCI, Inc. of the commercial products or services referenced herein; nor
`should the mere referedce herein to any drawing, specification, chemical process, or other data be re-
`garded as a license or as a conveyance of any right or permission to the holder, reader, or any other
`person or corporation, to manufacture, reproduce, use, or sell any patented invention or copyrighted
`work that may in any way be related thereto. Registered names, trademarks, etc., used in this book,
`even without specific indication thereof, are not to be considered unprotected by law.
`
`Library of Congress Cataloging-in-Publication Data
`
`Byrn, Stephen R.
`Solid-State Chemistry of Drugs / Stephen R. Byrn, Ralph R. Pfeiffer, Joseph G.
`Stowell--2nd ed.
`xvii, 576 p. :ill.; 24 cm.
`Includes bibliographical references and index.
`ISBN 0-967-06710-3
`ISBN 0-967-06711-1 (paperback student edition only)
`1. Pharmaceutical Chemistry. 2. Solid state chemistry. 3. Chemistry, Pharma-
`ceutical. I. Title.
`QV744.B995s 1999
`615’.19
`
`The publisher offers a discount paperback edition of this book to registered students
`only. Quantity discounts on the hardover edition are also vailable.
`
`Printed on acid-free paper.
`
`Cover illustration: The figures are space-filling representations of prednisolone 21~
`tert-butylacetate crystal packing diagrams. On the top is Form IV illustrating the
`densely packed crystal lattice, On the bottom is Form V showing the oxygen-
`accessible tunnels produced by desolvation.
`
`Lupin Ex. 1034 (Page 2 of 17)
`
`

`

`Hydrates and Solvates
`
`~"he occurrence of hydrated or solvated crystal forms (see Table 11.1), crystals
`- ~ in which solvent molecules occupy regular positions in the crystal structure, is
`-~. widespread but by no means universal among drug substances, Some classes
`of drugs (e.g., steroids, antibiotics, and sulfonamides) are particularly prone to form
`solvates, but this impression may be partly related to the considerable attention these
`drugs have received. In her classic book on thermomicroscopy, Kuhnert-Brandstfitter
`
`Table 11,1 Partial List of Drugs Discovered Prior to 1971 that Form Solvates
`
`Dr~g Reference
`
`Ampicillin
`Cephalofidine
`Chloramphenicol
`Cholesterol
`Cortisone acetate
`Eluprednisolone
`Erythromycin
`Estradiol
`Fluorohydrocortisone acetate
`Gramicidin
`Griseofulvin
`Hydrocortisone 2 l-acetate
`Hydrocortisone 21-tert-butylacetate
`Nitrofurmethone
`Prednisolone 21-tert-butylacetate
`Succinylsulfathiazole
`Sulfabenzamide
`Sulfaguanidine
`Sulfameter
`Sulfanilamide
`
`(Kuhnert-B randstiitter, 1971)
`
`Austin et al., 1965
`Chapman et al., 1968; Pfeiffer et al., 1970
`Himuro et al., 1971
`Shefter and Higuchi, 1967
`Carless et al., 1966
`Haleblian et al., 197i
`Rose, 1955
`Kuhnert-Brandstatter and Gasser, 1971
`Shefter and Higuchi, 1967
`Olsen and Szabo, 1959
`Sekiguchi et al., 1968
`Shell, 1955
`Biles, 1963
`Borka et al., 1972
`Biles, 1963
`Shefter and Higuehi, 1967
`Yang and Guillory, 1972
`Yang and Guillory, 1972
`Moustafa et aL, 1971
`Lin, 1972
`
`Lupin Ex. 1034 (Page 3 of 17)
`
`

`

`Chapter I1 Hydrates and Solvates
`
`(1971) summarized in tabular form some of these examples (see Table 11.2). Never-
`theless, many drugs, including some members of the aforementioned classes, even af-
`ter intensive investigation, are found to always crystallize without solvent inclusion
`(e.g., aspirin and ibuprofen).
`In a classic study, Kuhnert-Brandstatter. (1971) characterized the behavior of hy-
`drates of pharmaceuticals known at that time using thermomicroscopy (see Table 10.2).
`Many of the hydrates listed in this table show unusual behavior that may be caused by
`dehydration prior to melting. Many of these crystals are reported to become opaque,
`and appear dark when viewed by transmitted light (see Chapter 14). Some of these hy-
`drates crack and "jump" during dehydration. This behavior is characteristic of rapid
`solid-state reactions that produce gaseous products.
`Another aspect of solvate formation is that virtually any laboratory solvent can be
`involved; Table 11.3 lists solvents in solvates reported in the crystal structure literature
`on organic compounds which, naturally, includes many crystalline drugs. In some
`solvates, two or even three different solvents occupy their own positions in the .struc-
`ture. Furthermore, a compound may form solvates with a given solvent in different
`ratios, 2:1, 1:1, etc., and in rare cases, a fixed ratio in polymorphic forms.
`Because prediction of crystal structures is not yet generally possible, we must be
`content with examining the crystal structures of compounds after the fact in looking for
`explanations of why solvates do or do not form. On doing so, however, we are left
`with only vague impressions, to wit:
`¯ Certain molecular shapes and features favor the formation of crys-
`tals without solvent. These structures tend to be stabilized by an
`efficient packing that also utilizes intermolecular hydrogen bonding
`and other bonding capacity to a maximum extent. The slightest
`molecular differences may conceivably interfere with this coopera-
`tive effect. As a result, solvate formation within a series of related
`compounds tends to lack a discenfible pattern---each compound has
`a unique response to solvate formation.
`¯ Including specific solvent molecules can stabilize a crystal structure
`by improving either the packing or the intermolecular bonding, es-
`peciaIly hydrogen bonding. Some of the solvents listed in Table
`I 1.3 are nonhydrogen-bonding solvents and thus must serve only
`in a space-occupying capacity. The hydrogen-bonding capacity of
`included solvent molecules is usually fully exploited, although
`some structures are known where such bonding capacity is not ex-
`ercised at all.
`¯ Lower temperatures favor formation of solvates and also higher
`stoichiometric amounts of a given solvent. This is probably due to
`the increased strength of hydrogen bonding at lower temperatures.
`¯ The ease with which solvent is lost varies widely among solvates.
`At one extreme some retain solvent at temperatures well above the
`boiling point of the solvent, at the other extreme, others lose sol-
`vent readily at room temperature. In the latter case, the formation
`of a solvate may be overlooked unless special precautions are taken
`to preserve the composition of the crystals.
`
`Lupin Ex. 1034 (Page 4 of 17)
`
`

`

`Table 11.2 Thermomicroscopic Behavior of Dt~g Hydrates
`
`Drug
`
`Mp (°C) ¯
`
`Remarks
`
`Introduction 235
`
`Apomotphine hydroehloride
`Atropine sulfate
`Bnacine
`Brucine sulfate
`Chloroquine sulfate
`Citric acid
`Cocaine nitrate
`Codeine
`Codeine hydrochloride
`Codeine phosphate
`Cyclophosphamide
`Dihydrocodeinone bitartrate
`Dipropylbarbituric acid
`Emetine hydrochloride
`L-Ephedrine
`Heroine hydrochloride
`Histidine monohydrochloride
`Hydrocortisone hemisuccinate
`Hyoscyamine hydrochlo~’ide
`Lidocaine hydrochloride
`Me,’captopurine
`Mescaline sulfate
`Methicillin sodium
`Morphine
`Morphine hydrochloride
`Morphine sulfate
`Ouabain
`Oxycodone hydrochloride
`Phloroglucinol
`Pyrogallol
`Quercetin
`Quinidine hydrochlofide
`
`Quinidine sulfate
`Quinine bisulfate
`Quinine hydrobromide
`Raffinose pentahydrate
`Reserpine hydrochloride
`a-Rhamnose
`Succinylsulfathiazole
`- Sulphaguanidine
`Terpin hydrate
`Theophylline
`L-Thyroxine sodium
`
`(Kuhnert-Brandstatter, ~t971)
`
`220-260
`190-193
`170
`130-165
`209-213
`152-155
`55-59
`156
`260-275
`225-240
`40-47
`115-130
`148
`205-215
`38-40
`218-232
`155-176
`198-205
`152-155
`65-78
`300-325
`230-250
`182-186
`245-255
`285-310
`230-240
`178-I84
`245-260
`218-220
`133
`300-320
`262-265
`205-210
`155-160
`145-152
`132-135
`125-225
`70-95
`190-193
`187-191
`105.5
`274
`195-202
`
`From 220 °C, turbidity and carbonization
`Substance partially dehydrates
`Crystals are t~u’ety cleat"
`Gradual loss of water
`
`Loss of H20 and turbidity at 60-70 °C
`Needles of decomposition product appear
`Loss of water with turbidity
`
`Melts as hydrate
`
`Commercial product partly dehydrated
`
`Turbidity of crystals from 115-120 °C
`
`From 85 °C, loss of water with turbidity
`Turbidity with loss of water during heating
`
`From 160 °C, turbidity of crystals
`From 125 °C, water escapes with turbidity
`Loss of birefringence
`At 115-140 °C, loss of water with turbidity
`From 80 °C, loss of water with turbidity
`
`Loses.H~O from 90 °C
`From 70 °C, loss of water with turbidity
`At 55-90 °C, turbidity with loss of water
`Turbidity
`
`From 90 °C, loss of water
`
`Turbidity at 60 °C
`Water evolved at 90 °C
`Transforms to anhydrous
`From 180 °C, turbidity of crystals
`
`From 90 °C loss of water
`Transforms to anhydrous
`At 70-80 °C, loss of water
`From 70 °C, effervescence with jumping
`
`Lupin Ex. 1034 (Page 5 of 17)
`
`

`

`Chapter 11 Hydrates and Solvates
`
`Table 11.3 Solvents that Form Solvates with Drugs and Organic Compounds
`
`water
`methanol, ethanol, 1-propanol, isopropanol, 1-butanol, sec~butanol, isobutanol, tert-butanol
`acetone, methyl ethyl ketone
`acetonitrile
`diethyl ether, tetrahydmfuran, dioxane
`acetic acid, butyric acid, phosphoric acids
`hexane, cyelohexane
`benzene, toluene, xylene
`ethyl acetate
`ethylene glycol
`dichloromethane, chloroform, carbon teta’achloride, 1,2-dichloroethane
`N-methytformamide and N,.N-dimethylformamide, N-methylacetamide
`pyridine
`dimethylsulfoxide
`
`Most drug crystals that fall into the category of solvates are, for obvious reasons,
`hydrates-exceptions being ethanol or freon solvates. The organically solvated struc-
`tures, nevertheless, also deserve our attention for several reasons:
`¯ They are often the penultimate solid form of the drug (which should
`therefore be monitored carefully in the interests of good control ),
`¯ They are often specifically chosen for recovery or purification.
`¯ They may have a morphology conducive to good filtration or other
`bulk processes.
`¯ They may be the only crystalline form available for X-ray structure
`determination of a new molecular species.
`¯ They may be useful in their desolvated form as a drug product due
`to superior dissolution properties.
`¯ They may be patentable for any of the above reasons, thus pro-
`longing the manufacturing exclusivity of the drug.
`
`11.1 HYDRATES
`
`The water molecule, because of its small size, is particularly suited to fill structural
`voids. The multidirectional hydrogen bonding capability of water is also ideal for
`linking a majority of drug molecules into stable crystal structures. Figure 11.1 shows
`how a water molecule in an ice crystal forms hydrogen bonds in four directions by act-
`ing twice as an hydrogen acceptor and twice as an hydrogen donor to neighboring
`molecules with identical bonding.
`In hydrated crystal structures,~ we find that water molecules bind not only to other
`water molecules but also to any available functional groups like carbonyls, amines, al-
`cohols and many others that pan accept or donate an active hydrogen atom to form hy-
`drogen bonds. As a result, ~he total hydrogen bonding of water in crystal .hydrates is
`almost always one of the most important forces holding the structure together)
`Owing to their high solubility and biocompatibility, sodium salts are contmonly the
`derivative chosen for acidic drug products. Because the formation of crystal hydrates
`
`Lupin Ex. 1034 (Page 6 of 17)
`
`

`

`11.1 Hydrates ~,37
`
`is common for sodium salts of all drug classes, a look at why this is so prevalent may
`be helpful to our discussion of water in crystals.
`The sodium ion in most crystal structures is coordinated to six ligands in an octa-
`hedral array. In sodium chlor..ide, for example, six chloride ions surround each sodium
`ion. In monensin, a crown ether with extraordinary affinity for sodium, the coordi-
`nated ligands are bidentate, forming five-member rings with the sodium (see Figure
`11.2). In sodium salts of drugs and other organic compounds, this strong coordination
`tendency of the sodium ion is satisfied by linkages to any suitable ligand to be found in
`the structure, such as carboxylate, alcohol, carbonyl, amide, sulfonate, and many oth-
`ers, particularly water. While hydrogen bonding between the various ligands tends to
`distort the regularity of the coordination octahedra around the sodium ion, the trend to
`octahedral coordination is almost universal. ¯
`The high affinity of the sodium ion for water is perhaps most striking when we
`find that some sodium ions in acid structures actually coordinate to water molecules and
`the charged oxygen atoms of the acid are relegated to a position more distant from the
`sodium ion, yet always hydrogen bonded to an intervening water molecule. Analysis
`of the charge distribution in some of these structures shows the charges to be quite dif-
`fuse rather than being concentrated at single atomic positions.
`
`Figure 11.1 Stereoview of the hydrogen bonding in ice Form Ih. Only the the oxygen atoms am
`shown (Peterson and Levy, 1957),
`
`Figure 11.2 Stereoview of monensin sodium monohydmte showing the coordination of the sodium
`ion (only the non-hydrogen atoms, expept for the two hydrogens involved in in-
`tramolecular hydrogen-bonds to the carboxylate group, ate shown; key: ¯ C, ¯ O,
`@ Na, @ H20, ® H) (Barrans etal., 1982).
`
`Lupin Ex. 1034 (Page 7 of 17)
`
`

`

`238 Chapter 11 Hydrates and Solvates
`
`The coordination octahedra around sodium ions, in addition to accommodating dif-
`ferent ligandg, also show a remarkable flexibility in how they propagate throughout a
`crystal. Thus, neighboring octahedra can be independent or share points, edges, or
`sides in any combination, usually leading to highly polar chains or sheets that traverse
`the entire crystal. Many properties of the crystal are dependent on structural features of
`this kind. Figure 11.3 is a view of a typical structure of a sodium salt hydrate.
`The counterpart to sodium salts ,are amine hydrochlorides, again commonly en-
`countered among drug products because they are soluble and biocompatible. Of the
`roughly 130 hydrochloride salts listed in the Physicians’ Desk Reference (Greenberg,
`1994), about ten are reported to exist as hydrates. Hydrated forms of amine hydrochlo-
`rides do not, however, fall into the orderly patterns shown by the sodium salts.. Nev-
`ertheless, a few generalizations can be made with regard to the role of water in hydrated
`hydrochloride crystals:
`
`The charged ions--the ammonium ion and the chloride ion--are
`most often in proximity and hydrogen bonded to one or more water
`molecules.
`These water molecules tend to be extensively involved with any
`other hydrogen bond donors or acceptors in the structure.
`
`Figure 11.4 shows the location of the water in fenethazine hydrochloride monohydrate.
`In summary, we have shown that water plays many roles critical to drug crystal
`structures:
`
`¯ water occupies vacancies in packing
`° water hydrogen bonds to functional groups and to other water
`molecules
`° water coordinates sodium.and other cations
`
`Figure 11.3 Stereoview of the sodium acetate tfihydrate unit cell showing the octahedml ma’angement
`of the oxygen ligands around the sodium atoms (symmetry related waters have been
`added to complete the octahedra; Key: O Na, ¯ C, O O, ¢ H20 (Cameron et al,,
`1976).
`
`Lupin Ex. 1034 (Page 8 of 17)
`
`

`

`11.2 Conditions Under which Hydrates May Form 239
`
`Figure 11.4 Stereoview of the coordination of water in fenethazine hydrochloride monohydrate.
`O H, ¯ C, @ N, ® O, Q S, @ CI (Obata et aL, 1985).
`
`Key:
`
`11.2 CONDITIONS UNDER WHICH HYDRATES MAY FORM
`
`When we see the. manifold ways in which water can be bound in various crystal-
`structures we should expect and indeed find that each hydrate structure has its own
`characteristic binding energy for the water molecules in it. Thus,(the mere presence of
`water in a system is not sufficient reason to expect hydrate form~ttion, rather, it is the
`activity of water that determines whether a given hydrate structure forms.’) We have al-
`ready pointed out that lsome, compounds do not seem to form hydrate( even though
`they~are soluble in water.
`~he most obvious that favors the formation of crystal hydrates is of
`course when an aqueous solution of a substance is evaporated, cooled, or otherwise
`altered to reduce the solubility of the substance. Supersaturation followed by ,nuclea.
`lion will result in the formation of hydrate crystals provided that that form exi.stst Fig-
`ure 11.5 shows an example of how different crystal forms may be obtained’when
`evaporation of an aqueous solution is conducted at different temperatures. The solubil-
`ity profile in Figure 11.5 shows relationships that typify drug systems capable of
`forming several structures with different degrees of hydration.
`Figure 11.6 shows typical solubility relationships that govern formation of hy-
`drates in mixtures of water and an organic solvent. This diagram would naturally
`change with temperature, with the higher hydrates becoming unstable as the tempera-
`
`evaporation
`
`Figure 11.5 Crystal forms produced when evaporations are performed at different temperatures.
`
`Concentration
`
`Lupin Ex. 1034 (Page 9 of 17)
`
`

`

`240 Chapter 11 Hydrates and Solvates
`
`ture increases. One factor that many researchers fail to consider is the activity of water
`in mixed solvents. Figure 11.7 shows the relationship between the activities and the
`activity coefficients of water and ethanol in solutions of these miscible solvents, (Re-
`call that ~ = 7’rr;, where t~i is the activity of component i, ~ is the activity coefficient for
`component i, and xl is the mole fraction of component i in the solution.)
`
`99
`
`9"7
`
`95
`
`93
`
`Percent Ethanol in Water
`
`Figure 11.6 Hydrates of cefazolin sodimn formed in water-ethanol mixtures.
`
`4.0
`
`3.5
`
`3.0
`
`1.0
`
`0,5
`
`0 0,1 0.2 0.8 0.4 0,5 0,6 0.7 0.8 0,9 1.0
`Hg3
`
`XEtOH
`
`EtOH
`
`Figure 11.7 Activity (o~) and activity coefficient (7) plot for the ethanol-water system at 20 °C (Data
`from the International Critical Tables; Washburn et al., 1928).
`
`Lupin Ex. 1034 (Page 10 of 17)
`
`

`

`11.3 Stability of Hydrates at Different Relative Humidities 241
`
`3.2
`
`Percent Water Content of Ethyl Acetate Solution
`
`Figure 11.8
`
`Crystallization of hydrated crystal forms from so-called water-immiscible solvents.
`(Note that the water activity increases from zero to one over the range of water content
`from 0% to 3,2% water,)
`
`So-called water-immiscible solvent mixtures sometimes present problems in con-
`trolling crystal form because the finite, but very low, water solubility means that the
`water activity in that solvent can vary from zero to one with only slight changes in wa-
`ter concentration. Therefore,(when substances that are capable of forming multiple hy-
`drates are dissolved in water-i’rnmiscible solvents, different hydrates can be encountered
`if the water content of the system is not rigorously controlled~(see Figure 11.8). This
`/
`can prove difficult in large scale operations.
`At this point it should be obvious that good solubility data are vital to understand-
`ing the crystallization behavior of a compound, particularly if multiple forms are en-
`countered.
`
`11.3 FORMATION OF HYDRATES IN AIR: STABILITY OF HYDRATES AT DIFFERENT
`RELATIVE HUMIDITIES
`
`One of the most important characteristics to be determined for any drug is how it re-
`sponds to changes in relative humidity, This knowledge is essential to providing
`proper conditions for the handling and storage of any solid drug product in order to
`avoid possible recrystallization into a new and perhaps undesired form. For example,
`plaster or cement is purchased as a free-flowing powder, but after the addition of water,
`the material undergoes a recrystallization to the hydrated forms. The quantitative as-
`pects of changes in a drug with respect to relative humidity are also important to the
`proper execution of any mass-based analysis.
`The best method to describe the moisture-uptake or moisture-loss characteristics of
`a compound is to prepare a water content versus relative humidi,ty (RH) diagram based
`on data obtained after equilibration of the solid at various RH1 This can be accom-
`plished rather .simply by obtaining a single Karl Fischer meast~rement or other water
`analysis of the substance and following changes by gravimetric monitoring until con-
`stant weight is obtained for a multiple of samples stored in various RH chambers, An-
`other method is to subject a single sample to gravimetric analysis using a recording mi-
`crobalance connected to a system with controllable atmosphere.
`Griesser and Burger (1995) have developed a semimiero hygrostat (see Figure
`11.9) which is a modification of a hygrostat reported by Schepky (1982). The semimi-
`
`Lupin Ex. 1034 (Page 11 of 17)
`
`

`

`242 Chapter I1 Hydrates and Solvates
`
`stopper with hooks
`
`19/26
`
`adapter
`
`34/24
`
`250 mL
`amber bottle
`
`sample cradle
`
`saturated
`solution
`
`E
`E
`
`!~
`
`70mm
`
`,~__-!
`
`Figure 11.9
`
`Semimiero hygrostat for gravimetdc sorpton-desorption moisture studies (Gdesser a~d
`Burger, 1995).
`
`cro hygrostat consists of an amber wide-mouth reagent bottle with a 34/24 standard-
`tappered ground-glass joint, a 34/23 male to 19/26 female standard-tappered ground-
`glass adapter, and a modified 19/26 standard-tapper ground-glass holiow stopper. The
`amber bottle contains a saturated salt solution to control the relative humidity (Nyqvist,
`1983). The stopper has a long hook on the bottom (inserted through the adapter to hold
`the glass sample basket directly above the saturated salt solution) and a short hook on
`the top (attached to a bottom-weighing balance when weighing). The sample to be
`studied (approximately 300 rag) is distributed evenly in the sample basket and then the
`semimicro hygrostat is assembled and sealed. At various time periods, the hygrostat is
`positioned below the bottom-weighing balance and the stopper-sample assembly is
`slighly separated from the adapter and suspended from a wire loop attached to the bal-
`ance arm of the balance to determine any change in the sample weight. After recording
`the data, the stopper-sample assembly is detached from the loop and the hygrostat re-
`sealed. In this manner, weighings can be made without disturbing the equilibrium of
`the system. Several automated systems have been developed to measure moisture up-
`take (e.g., Surface Measurement Systems, 1998; VTI Corp., 1998).
`Figure 11.. l0 shows the kind of behavior that might be found for a compound that
`can exist as an anhydrate, monohydrate, or dihydrate at room temperature. The step-
`wise changes in moisture content, of course, are verifiable by other methods covered in
`Chapter 2.
`Note that Gibbs’ phase rule considerations dictate the depicted behavior: when
`there are two condensed phases present, vapor pressure cannot vary; when there is
`
`Lupin Ex. 1034 (Page 12 of 17)
`
`

`

`11.5
`
`Factors Governing the Formation of Solvates in Mixed Solvents 243
`
`dihydrale
`
`dellquescenoJ
`j"
`
`..................................transition’i£~°~between .....................
`monohydrate to
`~
`dlhydrate
`RHo
`
`monohydrate
`
`the etlcal
`transition between
`anhydrate to
`monohydrate
`
`anhydrate
`
`¯ 0 C 100
`A
`B
`
`% RH
`
`Figure 11.10
`
`Idealized moisture-uptake profile, Starting with a sample of the anhydrous form at 0%
`RH and increasing the percent relative humidity, the sample will transform to the
`monohydrate form at point A, Increasing the percent relative humidity further, the
`sample will transform to the dihydrate at point B. Increasing the percent relative humid-
`ity still further, the sample will begin to deliquesce (RH0) at point C.
`
`only one condensed phase, vapor pressure will vary over a range. (This principle pro-
`vides the constant humidity in environment chambers containing the condensed phase,
`that is, crystals and saturated solution.)
`
`11.4 DELIQUESCENCE AND EFFLORESCENCE
`
`At this time, in the interests of good communication, we should clarify the terms deli-
`quescence (to become liquid from the adsoption of atmospheric water) and efflores-
`cence (to change to a powder from the loss of water of crystallization), because they
`are often used without recognizing the fact they are relative terms, not absolute terms.
`In other words, the conditions under which the respective behavior is observed to
`spontaneously gain or lose moisture must be stated. It is incon’ect to make the blanket
`statement that a substance is, or is not, "deliquescent" or "efflorescent." Thus, Ia sub-
`stance may have been observed to deliquesce or effloresce, but this observation fs valick
`only above or, :respectively, below a certain RH value and at a fixed temperature.j
`These limiting values, of course, vary considerably from one substance to another.
`
`tl.5 FACTORS GOVERNING THE FORMATION OF SOLVATES IN MIXED SOLVENTS
`
`When a solution of a compound in an organic solvent is evaporated, the results are
`analogous to the formation of hydrates depicted in Figure 11.5. Depending on the
`forms available to the given system, the resulting crystals may be unsolvated or sol-
`vated with the relevant solvent, again dependent on temperature.
`
`Lupin Ex. 1034 (Page 13 of 17)
`
`

`

`244 Chapter I1 Hydrates and Solvates
`
`It is common practice in the pharmaceutical industry to use mixtures of solvents for
`the crystallization of a drug. Because many drugs can form multiple solvates, the use
`of mixed solvent solutions can greatly multiply the probability of obtaining a crystal
`solvate.
`Often crystallizing a drug involves the use of a "good" solvent to obtain a fairy
`concentrated solution. A miscible "antisolvent," chosen for its low solubility for the
`given drug, is then added to the solution to induce crystallization by forming a super-
`saturated solution of the mixture. In the most desired case, the solubility of the drug
`decreases smoothly during this process and an unsolvated crystal form is obtained (see
`Figure 11.11). In systems prone to solvate formation, however, the solubility behavior
`of the drug can be strikingly different as the solvent composition varies from one ex-
`treme to the other (Pfeiffer et al., 1970; and others). Rather than a gradual decrease in
`solubility, these authors found not only that there are discontinuities in the solubility
`versus solvent composition curves but also that these discontinuities demarcate the
`boundaries between zones where different solvates are obtained. Moreover, the solu-
`bility maxima can be remarkably higher in the mixed solvents than in either pure sol-
`vent, a finding that can be extremely useful in process design. Figure 11.12, adapted
`from Pfeiffer et al. (1970), gives examples of solubility diagrams that correlate solubil-
`ity behavior with formation of different solvates. Hydrate formation is included in
`these considerations, as are examples of solvates containing two solvents.
`Th, e interpretation of solubility curves like those in Figure 11.12 is that they reveal
`some kind of strong solute-solvent interactions. We can postulate that:
`
`1 The nature and concentration of differently solvated solute species
`that exist at different solvent ratios will change considerably as we
`move from one side of the diagram to the other.
`2 Each alternative crystal form in the diagram will grow best when
`the solvated solution species it favors is at maximum concentration.
`
`Thus, a methanolate crystal might grow well when all of the solute molecules in
`the solution are surrounded by dipole-oriented methanol molecules, and the growth of a
`methanol-hydrate crystal would be favored by a high concentration of solute molecules
`surrounded by a certain ratio of water and methanol molecules, and so forth.
`
`lOO
`
`80
`
`’~ 40
`o
`
`20
`
`o
`v/v% Solvent A 0 20 40 60 80 100
`vtv% Solvent B 100 80 60 40 20 0
`
`Figure 11.11
`
`Solubility of a substance versus concentration of two solvents A and B.
`
`Lupin Ex. 1034 (Page 14 of 17)
`
`

`

`11.5
`
`Factors Governing the Formation of Solvates in Mixed Solvents
`
`245
`
`12
`
`10
`
`1
`
`12o
`
`1oo
`
`2O
`
`o
`vlv% Formamide 0
`vtv% Water 100
`
`20 40 60 80 100
`80 60 40 20
`0
`
`0
`
`v/v% Methanol
`v/v% Water 100
`
`20
`80
`
`40 60 80 100
`60 40 20 0
`
`320
`
`16
`
`14
`
`12
`
`::3
`o~ 6
`
`0
`0
`20 40 60 80 100 vlv% Acetonttrile 0
`20 40 60 80 . 100
`vtv% Aoetonitrile
`v/v% Water 100 80 60 40 20 0 v/v% Water 100 .80 60 40 20 0
`
`Figure 11.12
`
`Solubility diagrams for a drug in mixtures of organic solvents and water (Pfeiffer, et
`aL, 1970).
`
`We again find ourselves dependent on adequate solubility data if we expect to con-
`trol crystallizations in mixed solvents.
`
`11.6 STABILITY OF ORGANIC SOLVATES IN AIR
`
`Like hydrates, organic solvates also respond to changes in the vapor pressure of the
`relevant solvent by losing or gaining weight. From a diagram of solvent vapor pres-
`sure versus solvent content, as in Figure 11.13, we see behavior analogous to the hy-
`drates in the percent relative humidity versus water-content diagram as in Figure 11.10.
`In this case, however, the non-solvated form transforms directly to the disolvate form
`at approximately 15% v/v acetonitrile in triethylene glycol. {Even without absolute val-
`ues, the vapor pressure is the most convincing evidence ofthe true stoichiometry of a
`solvate. Just as all hydrates lose water at zero RH, organic solvates will eventually lose
`
`Lupin Ex. 1034 (Page 15 of 17)
`
`

`

`±46
`
`Hydrates and Solvates
`
`<1)
`
`o-(
`
`I I I I
`20
`40
`60
`80
`
`I
`100
`
`AcetonitrileiTriothylene Glycol (%v/v)
`
`Figure 11.13 Crystal-vapor equilibria for cephalexin-acetonitrile, The abscissa describes the volume
`composition of acetonitdle-tdethylene glycol mixtures used to provide a range of ace-
`tonitrile vapor pressures whose exact magnitude was not determined (Pfeiffer, et al.,
`1970),
`
`solvent when completely exposed to air (i.e., at zero solvent vapor pressure). Like-
`wise, above a certain solvent vapor pressure, the solvated crystals will dissolve.) Once
`more, the limits and rates that pertain to a given case are highly variable.
`
`11.7 SUMMARY
`
`In this chapter we have reviewed hydrates and solvates, the conditions under which
`they form, and their stability. In addition, deliquescence and efflorescence are defined
`and the use of solubility versus composition diagrams to understand crystallizations is
`reviewed. This review provides insight into approaches which can be used to under-
`stand solvates and hydrates.
`
`REFERENCES
`
`Austin, K. W. B., A. C. Marshall, and H. Smith (1965) "Crystalline modifications of ampieillin" Na-
`ture 208 999-1000.
`Ban’ans, Yvette, Marc Alleaume, and Georges Jeminet (1982) "Sodium complex of the ionophore
`monensin B monohydrate" Acta Crystallogr., Sect. B, Struct. Crystallogr. Cryst. Chem. 1138
`1144-1149.
`Biles, John A. (I963) "Some crystalline modifications of the tert-butylacetates of prednisolone and
`hydrocortisone"./. Pharm. Sci. 52 1066-1070.
`Borka, Laszlo, Per R. Kristiansen, and Kjell Backe-Hansen (1972) "Pseudopolymorphism of nitrofur-
`methone" Acta Pharm. Suecica 9 573-580.
`Cameron, T. Stanley, Kh. M. Mannan, and Md. Obaidur Rahman (1976) "The crystal structure of so-
`dium acetate trihydrate" Acta Crystallogr. Sect. B, 1132 87-90.
`
`Lupin Ex. 1034 (Page 16 of 17)
`
`

`

`References 247
`
`Carless, J. E., M. A, Moustafa, and H. D, C. Rapson (1966) "Cortisone acetate crystal forms" J.
`Pharm. Pharmacol, Suppl. 18 190S-197S.
`Chapman, J. H., J. E. Page, A. C. Parker, D. Rodgers, C, J, Sharp, ~d Susan E. Staniforth (1968)
`"Polymorphism of cephaloridine" J. Pharm. Pharmacol. 20