ENCYCLOPEDIA OF
`PHARMACEUTICAL
`TECHNOLOGY
`
`Editors
`JAM ES SWARBRICK
`P~ofessor and Chairman of Pharmaceutics
`School of Pharmacy
`University of North Carolina at Chapel Hill
`Chapel Hill, North Carolina
`
`JAMES C. BOYLAN
`Director
`Pharmaceutical Research & Development
`Hospital Products Division
`Abbott Laboratories
`Abbott Park, Illinois
`
`VOLUME 7
`
`GENETIC ENGINEERING
`TO HYDROGELS
`
`MARCEL DEKKER, INC..
`
`NEW YORK ¯ BASEL ° HONG KONG
`
`Lupin Ex. 1053 (Page 1 of 52)
`
`

`

`Library of Congress Cataloging in Pnblication Data
`Main entry under title:
`
`Encyclopedia of Pharmaceutical Technology.
`editors: James Swarbrick, James C. Boylan.
`
`Includes index.
`1. Pharmaceutical technology--Dictionaries. I. Swarbrick, James, 1934-
`II. Boylan, James C., 1943-.
`[DNLM: 1. Chemistry, Pharmaceutical-encyclopedias. 2. Drugs--
`encyclopedias. 3. Technology, Pharmaceutical-encyclopedias. QV 13 E565].
`RS192,E53 1988 615’.1’0321-dc19
`
`COPYRIGttT © 1993 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.
`
`MARCEL DEKKER, INC.
`.270 Madison Avenue, New York, New York 10016
`
`LIBRARY OF CONGRESS CATALOG CARD NUMBER: 88-25664
`ISBN: 0-8247-2806-8
`
`Current printing (last digit):
`10987654321
`
`PRINTED IN THE UNITED STATES OF AMERICA
`
`Lupin Ex. 1053 (Page 2 of 52)
`
`

`

`Hydrates
`
`Introduction
`
`During the development of a new drug, from the synthetic process to dosage form
`design, the interaction of water with the solid form is of great concern since it
`affects bulk properties and chemical stability. A major consideration in prefor-
`mulation studies is to define the ability of the drug powder to take up water and
`to characterize the state of this water, because this information is relevant in
`developing strategies for the process and storage of dosage forms.
`It is important to know the mechanism of water uptake, where the water is
`located, and the nature of the water molecule environment. The water vapor-solid
`interaction may result in surface adsorption (physisorption or chemisorption) or
`incorporation into the bulk (interstitial voids or part of the crystal lattice). This
`article describes the latter, that is, solids that form solvates and in the case of
`water, hydrates. The water molecule in these crystals is frequently involved in
`hydrogen bonds and contributes to the coherence of the crystal structure.
`Most.of the work in the pharmaceutical literature involving hydrates has focused
`on their characterization [1-4], vapor-solid interactions [5], and hydration-dehy-
`dration kinetics in the solid state [6]. Some work [%9] has addressed the conse-
`quences of the hydrat~ choice on the thermodynamic properties of the solid and
`on pharmaceutical properties such as dissolution and bioavaitability. To date, little
`emphasis has been placed on liquid-solid interactions and the phase changes that
`may result during crystallization.
`It is essential to define the transformations between solid phases in the devel-
`opment of a dosage form, since the presence of a metastable phase during processing.
`or in the final product could result in a system that is in continuous evolution and
`instability in drug release. Erratic bioavailability of theophylline [10] and carba-
`mazepine .[11] from solid dosage forms has been reported to be the result of a
`phase change caused by the formation of a hydrate during dissolution. Crystalli-
`zation in tablets that have been in contact with water has been observed for caffeine
`[12], theophylline [13], carbamazepine [14], lactose [15], and mannitol [16].
`In the case of a solid that exists in various hydrated forms, possible transfor-
`mations associated with exposure to water are of interest in both the liquid and
`the vapor states, for instance, during solubility measurements, wet granulation
`processes, dissolution studies, and accelerated stability tests.
`Questions that may need to be answered are:
`
`Does the solid exist as a hydrate? How is the water incorporated?
`What is the range of relative humidities and temperatures in which the desired
`hydrate is thermodynamically stable?
`
`393
`
`Lupin Ex. 1053 (Page 3 of 52)
`
`

`

`394
`
`Hydrates
`
`¯ How will the hydrated or anhydrous solid form be prepared? By crystallization
`from liquid solution or by vapor sorption-desorption?
`Is a metastable solid phase desired? How do interactions at the solid-vapor and
`solid-liquid interfaces affect the transformation rate?
`
`¯
`
`In this article, the thermodynamic, crystallographic, and kinetic aspects of phase
`transformations associated with hydrates are discussed, as well as the methods used
`for their characterization.
`
`Hydrogen Bonding Mechanism in Hydrates
`
`The ability of water to form hydrogen bonds and hydrogen-bonding networks gives
`it a unique behavior with respect to colligative properties such as boiling and melting
`points. Similarly, hydrogen bonding between water molecules and drug molecules
`in the solid state dictates its role in the structure of all classes of crystalline hydrates.
`Water is, of course, hydrogen bonded whenever physically possible. This may take
`the form of hydrogen bonding to other water molecules, functional groups on other
`molecules, or anions. Hydrogen bonding to other water molecules is common both
`in the crystal lattice and in interstitial cavities or channels. Hydrogen bonding to
`other moieties and anions in crystalline hydrates is primarily within the lattice. In
`addition, the lone-pair electrons on the water oxygen may be associated with
`metallic cations present in many salts. This interaction is largely electrostatic in
`nature for the metal cations common to pharmaceutics (Na, Ca, K, Mg). These
`metals lack the ability (d-orbitals of proper energy) to form coordinate covalent
`or coordination bonds that some transition metals may form with oxygen. It is
`often stated that Mg has a coordination number of 4. However, this is a result of
`packing (or geometric) restrictions, arising when fitting water molecules around
`the cation in response to the electrostatic attraction. Since these "bonds’° are
`electrostatic, they are not properly described quantum mechanically by a molecular
`orbital, but rather by classical electrostatics [18]. These bonds are often stronger
`than hydrogen bonds with less directional dependence. A typical water hydrogen
`bond is on the order of 19 kJ/mol (4.5 kcal/mol), whereas a sodium-oxygen lone-
`pair electrostatic interaction can be four to five times stronger [18]. These bonds
`also exert their influence through hydrOgen bonds in the form of cooperative effects.
`The specific characteristics of the hydrogen bond are discussed here in the formalism
`of Falkand Knop [17].
`The ubiquitous hydrogen bonding of water is largely a result of the fact that it
`is both a hydrogen-bond donor and acceptor. It may participate in as many as four
`hydrogen bonds, one from each hydrogen and one for each lone pair on the oxygen.
`Classification schemes based solely on the type of coordination of the water oxygen
`have been proposed [17]. As each bond is formed, it makes the other sites more
`attractive as partners for additional bonds. Hydrogen-bond acceptors must be elec-
`tronegative and include oxygen from other water molecules, oxygen and nitrogen
`from other functional groups, and chlorine. Hydrogen-bond donors include protons
`on nitrogen, oxygen, and sulfur, usually found on water, alcohols, amines, etc.
`
`Lupin Ex. 1053 (Page 4 of 52)
`
`

`

`Hydrates
`
`395
`
`Free water (vapor) has an OH bond length of 0.0957 nm (0.957 ~) and an
`HOH angle of 104.52°. As soon as the molecule starts interacting with other
`molecules through hydrogen bonds, coordination, or other electrostatic bonds, the
`molecule is distorted from its free conformation, The OH bond length usually
`increases up to 0.01 nm for an exceptionally strong hydrogen bond, but is more
`typically on the order of 0.001 to 0.002 nm for organic hydrates with hydrogen
`bond lengths of 0.27 to 0.29 nm (O--O distance). Depending on the hybridization
`of the water oxygen as spa (trigonal coordination) or sp3 (tetrahedral), the HOH
`angle is more typically 109.5. and 120°, respectively.
`The limits of length of a hydrogen bond are bound on the lower end by the
`van der Waals radii of the two atoms, and on the upper end arbitrarily by the
`length c~f the weakest hydrogen bond observed. This can be seen more quantitatively
`by expressing the hydrogen bond distances shown in Fig. 1 in terms of the van der
`Waal radii, as in Eq. (I).
`
`R(H--Y) < r(H) + r(Y)- 0,02 nm
`
`(1)
`
`Since r(H) = 0.12 nm, R(H--Y) < r(Y) - 0.1 nm where 0.02 nm is the combined
`experimental and statistical uncertainty. The compounds studies by Falk and Knop
`[17] were inorganic and small organic hydrates.. Of the 129 compounds studied, only
`one failed this criterion. Hydrogen bond lengths are often given as the distance
`between O and Y (e.g., oxygen-oxygen or oxygen-chlorine). This is because in
`x-ray diffraction studies it is often difficult or impossible to accurately locate the
`hydrogen atoms due to their inherently low scattering and their relatively high
`mobility. (Because of the large cross-sections, hydrogens are often located by
`neutron diffraction studies.) Under these circumstances, crystallographers report
`
`H
`
`r(O-H)
`
`O
`
`R(H
`
`RIO--Y) Y
`
`FIG. 1.
`
`Formalism used in the discussion of hydrogen-bond strength and length [17].
`
`Lupin Ex. 1053 (Page 5 of 52)
`
`

`

`396
`
`Hydrates
`
`the O-:Y distance, as shown in Eq. (2), they know to be reliable. The geo.metric
`constraints may also be applied to such data.
`
`Substituting in Eq. (1) and setting r(OH) = 0.098 nm, gives Eq. (3).
`
`R(O--Y) _< R(H--Y) + r(OH)
`
`(2)
`
`R(O--Y) -< r(Y) + 0.198 nm
`
`(3)
`As the electrostatic bond strength of the donor to the water’s oxygen (X)
`increases, the length of the H--Y bond decreases (the bond strengthens). This
`cooperative effect is also seen as the number of hydrogen bonds per water molecule
`increases. Hydrogen bonds prefer to be linear (with the angle between the OH
`and Y approaching 180°), but may adopt a range of angles at the expense of the
`bond strength [17,19].
`
`Classification. of Hydrates
`
`Crystalline hydrates may b’e classified by structure or energetics [I7]. In the struc-
`tural classification scheme presented here, the hydrates are divided into three
`classes which are discernabte by the common analytical techniques available. Clas-
`sification by structure is the most common and useful approach. A good classifi-
`cation system should direct the preformulation scientist to the characteristics of
`the particular class and help in identifying a new sample. The classification scheme
`given here also reflects the distinct behavior of each class with respect to the
`characterization techniques.
`Class 1, isolated lattice sites, represents the structure where water molecules
`are isolated from direct contact with other water molecules by intervening drug
`molecules.
`In Class 2, water forming lattice channels, the water molecules included in the
`lattice lie next to other water molecules of adjoining unit cells along an axis of the
`lattice, forming "channels" through the crystal. The empty channels are actually
`a conceptual construct. A corresponding low density structure with empty channels
`would not be expected to be physically stable.
`Class 3, metal-coordinated water, addresses the effect of the metal-water in-
`teraction on the structure of crystalline hydrates. This interaction can be quite
`strong (e.g., Mg(H20)6 which dehydrates at very high temperature) compared to
`the other "bonding" in a molecular crystal [18]. Drugs with solubility, dissolution,
`or handling problems are most often recrystallized as salts of sodium, calcium,
`magnesium, or potassium, and are often hygroscopic to some degree [20].
`
`Methods
`
`Methods suitable for the characterization of hydrates may be classified as energetic
`(thermal and spectroscopic) or structural [1~,17]. Irrespective of the method or
`classification, each method must yield information on the structure, composition,
`or energy of association in the hydrate.
`
`Lupin Ex. 1053 (Page 6 of 52)
`
`

`

`Hydrates
`
`397
`
`Polarized Light Microscopy (PLM) and Hot Stage
`
`Optical microscopy is one of the most familiar and commonly available analytical
`tools, although underutilized in formulation research. Coupling microscopy with a
`hot stage provides a rapid method for determining crystallinity, crystal habit, and
`solvation of a compound. This may often be accomplished with far less than 1 mg
`of compound, which is of obvious importance when the supply of a new chemical
`entity is limited.
`
`Crystallinity
`
`The first property of a new compound to be determined, is whether it is crystalline
`or amorphous. This becomes all the more important if moisture is associated with
`it. Some compounds are obtained as large crystals. More often the crystal habit
`is small needles that may, at first inspection, be indistinguishable from amorphous
`material. The presence or absence of crystallinity may be determined by observing
`a few particles of the compound between crossed polarizers at moderate magni-
`fication.
`Crystals are distinguished from amorphous material by "repetition of their
`constituent atoms (molecules) in a three-dimensional array" [21]. This orderly
`arrangement imparts a certain symmetry to the crystal with respect to rotation
`about axes, reflection through planes, and inversion through a point Of the con,
`stituent molecules. These symmetry "elements" also dictate the directional de-
`pendence of many physical and thermodynamic properties of the crystal, including
`the way light behaves when passing through the crystal [22].
`Of particular interest in PLM is that each of the six (or seven [21]) possible
`crystal systems has at least one axis of symmetry, called the optical axis. Each
`unique optical axis has a different refractive index. In a polarizing light microscope,
`light is polarized both before encountering the sample by the polarizer and after
`passing through the sample by a similar filter called the analyzer. When the polarizer
`and analyzer are 90° out of phase, an empty field of view is black because the only
`light passing through is polarized in the plane perpendicular to the analyzer. Sim-
`ilarly, when a crystal is oriented with an optical axis parallel to the direction of
`light propagation, the crystal disappears. This is called the "extinction position";
`it represents no rotation (change in the plane .of vibration) of the polarized light.
`Amorphous material depolarizes polarized light and has a grayish appearance under
`crossed polarizers. Crystal belonging to the cubic system have many equivalent
`optical axes. They are termed "optically isotropic" and disappear at any orientation
`when observed through crossed polarizers. Most crystalline drugs are anisotropic
`and disappear at 90° intervals when rotated [6,21].
`Anisotropic crystals exhibit another feature called "birefringence," which is one
`of the least complicated-tests for crystallinity in an unknown sample. It is also one
`of the most difficult phenomena to explain on the molecular level. Birefringence
`results from the directional dependance of the refractive index in an anisotropic
`crystal. When white polarized light is incident on an anisotropic crystal at any
`position other than extinction, the plane of polarization of its component colors
`(or wavelengths) are each rotated by a differing amount (actually becoming cir-
`cularly polarized). The amount of rotation exhibits a complicated dependance on
`
`Lupin Ex. 1053 (Page 7 of 52)
`
`

`

`398
`
`Hydrates
`
`the wavelength of the light and the refractive index, thickness, and orientation, of
`the crystal. The rotation due to the crystal is responsible for the transmission of
`light through the analyzer since some components of the light are no longer per-
`pendicular to the slit. This is also responsible for the colors characteristic of bi-
`refringence. An often spectacular range of colors in and about the fringes of the
`mass is the result of birefringence through crystals of varying thickness. Birefrin-
`gence charts such as the Michel-Levy chart are used to assess the degree of crys-
`tallinity [21]. The presence of birefringence in a solid is, therefore, a key charac-
`teristic for crystallinity when present.
`
`Desolvation on the Hot Stage
`
`The hot stage is a very accurate heater which is constructed to allow observation
`of a sample during controlled heating. A simple test for the presence of water in
`a drug is to immerse a few crystals in mineral oil and heat. If the crystals dehydrate,
`droplets appear as escaping bubbles. Many hdyrates show this behavior at ap-
`proximately 100°C, although some compounds dehydrate at much higher temper-
`atures [23]. If no droplets appear, it is possible that the drug is a solvate with a
`solvent soluble in the mineral oil. This can be checked by repeating the process
`using silicon .oil or some other high boiling liquid immiscible with the suspected
`solvent. This technique is sensitive to the heating rate and should be performed
`very carefully. A heating rate from 1 to 5°C/min is a reasonable range for initial
`studies. Fusion Methods in Chemical Microscopy [24] is a very good reference.
`Another hot-stage related phenomenon well documented by Byrn [6] is the
`dehydration of a particular class of hydrates called "channel hydrates." The concept
`is that there are channels in the crystal structure in which water (or other solvents)
`may reside. Upon heating, the crystal may show the dehydration as a "darkening"
`along the. axis of the channel. Hot-stage dehydration under crossed polars also
`shows, by its degree of biretringence, whether the drug remains crystalline after
`dehydration. This behavior is discussed later in terms of the crystal structures of
`ampicillin.
`
`Differential Scanning Calorimetry (DSC) and Thermogravimetric
`Analysis (TGA)
`
`These techniques are simple in concept and provide valuable qualitative and quan-
`titative, information concerning the existence and type of hydrate and solvate
`present,
`The DSC experiment entails heating (or coolingi a reference pan and a pan
`containi.ng the smnple at identical rates of temperature change. The signal measured
`is usually the difference in the amount of energy it takes to maintain the equal
`rate. If the sample absorbs heat during a phase transition (melting, vaporization),
`it takes more energy to maintain the equality; this is called an endothermic event.
`If heat is released by the sample due to crystallization or degradation, it takes less
`energy to attain the same temperature increase as the reference pan; this is called
`an exothermic event. Dehydration is an endothermic process because it takes energy
`to disrupt the association between the water and its host or between water mole-
`cules. However, DSC must be. coupled with TGA to distinguish between a phase
`transition and a desolvation or degradation.
`
`Lupin Ex. 1053 (Page 8 of 52)
`
`

`

`Hydrates
`
`399
`
`The shapes and temperatures of the DSC events of hydrates are very sensitive
`to the experimental conditions. The type of sample pan, heating rate, and purge
`gas all affect the data [23,25]. The sample pan may b.e open, hermetically sealed,
`or partially sealed. In an open pan, water may be lost by advection as well as
`dehydration. A hermetically sealed pan may increase the apparent temperature of
`dehydration because of the increased pressure in the head space. A crimped pan,
`or a hermetically sealed pan with a pinhole (<1 mm) punctured lid, is usualls~ the
`best choice although care must be taken not to "mask" events [26]. This arrange-
`ment allows pressure equilibration without excessive advective loss. Figure 2 shows
`the effect of the sample-pan type on the dehydration of theophylline [23]. A high
`heating rate may not"allow enough time for closely spaced events to occur inde-
`pendently. Too low a heating rate may result in a curve showing continuous water
`
`4
`
`I I
`200
`100
`
`0
`
`300
`
`1
`400
`
`500
`
`Temperature (°O)
`
`FIG. 2.
`
`Effect of sample-pan type and environment on DSC curves for sodium theophylline mono-
`laydrate recorded at a heating rate of 5°C/min. Samples were placed el) on an open aluminum
`pan with purging of.nitrogen, (2) on an open aluminum pan without purging of nitrogen, (3)
`in a pan closed by crimping, and (4) in a hermetically sealed pan. (Reprinted with permission
`from Ref. 23.)
`
`Lupin Ex. 1053 (Page 9 of 52)
`
`

`

`400
`
`Hydrates
`
`loss, characteristic of absorbed water when the compound is really a weak crystalline
`hydrate. Depending on the instrument, a heating rate of 5 to 10°C/min is a rea-
`sonable starting point. However, many runs may be needed to accurately assess
`the dehydration behavior if the compound is very sensitive to experimental con-
`ditions. The purge gas (dry nitrogen or argon for better accuracy) helps heat transfer
`and sharpens endOtherms.
`
`Quantitative Information
`Among the most useful quantitative information available from DSC are the melting
`and dehydration temperature ranges,, heats of fusion for the drug, and heats of
`vaporization of water. From TGA, the stoichiometry of a hydrate may be obtained
`as well as the isothermal rate of dehydration.
`The area of a DSC endotherm generated on a properly calibrated instrument
`is a direct measure of the reaction heat (fusion, crystallization, sublimation, va-
`porization, decomposition, etc.) of the sample, as shown in Eqs. (4) and (5).
`
`therefore
`
`(4)
`
`(5)
`
`where H is the enthalpy, M the mass, Cp the heat capacity, T the temperature,
`and t the time.
`The accuracy of this calculation depends on the accuracy of the input sample
`mass because the quantity is calculated as zXH, per unit mass (or mole). If the
`sample has dehydrated prior to melting, the AHf is for the anhydrous material.
`The amount of mass due to the moisture present in the initial material must be
`subtracted to accurately determine the quantity. If only a single melt occurs and
`the TGA curve shows no weight loss, the heat of fusion should be representative
`of the original hydrated species.
`The presence of a large amount of excess or "clustered water" may be probed
`with low-temperature techniques [26]. Evidence is sought for the melting of ice
`while the sample is scanned from well below 0°C through melting. If the heat of
`fusion matches that of bulk water, the water is presumed to be clustered. Great
`care must be taken not to condense moisture upon cooling or to induce a teln-
`perature-dependent crystal transformation. Therefore the test is best carried out
`in a hermetically sealed pan.
`The TGA experiment simply measures the weight change of a sample as a
`function of temperature or time (isothermal dehydration). The instrument is an
`extremely sensitive variable-temperature balance. Once the identity of a solvent
`in a crystal is shown to be water by complimentary methods, the stoichiometry of
`the hydrate may be determined.
`The stoichiometry of water to drug in a hydrate is relatively easy to determine
`with the help of TGA. The implicit assumptions are that the weight loss attributed
`
`Lupin Ex. 1053 (Page 10 of 52)
`
`

`

`Hydrates
`
`401
`
`to water is distinguishable from that due to other sources, and that the DSC curve
`shows corresponding endotherms. The TGA curve is often displayed as the amount
`of loss as percent of the initial mass. To determine the ratio of water to drug, the
`mass of each component must first be calculated and the number of moles of water
`to drug determined, as shown in Eq. (6).
`
`mole water
`mole drug
`
`(grams water lost)/(18 g/mole)
`(grams sample - grams water lost)/(molecular weight of drug)
`(6)
`
`The stoichiometries determined by TGA may be used with enthalpies of dehydra-
`tion from DSC to assess tl~e strength of the water-drug association relative to bulk
`water. "Nonstoichiometric water" content may occur for several reasons discussed
`later. The isothermal rate of dehydration of drug crystals may also be measured
`by TGA. There are both energetic and geometric components of dehydration
`kinetics. In practice, the goal is to determine the "window" of time at a given
`temperature that a hydrated drug is physically stable. Although this is a quantitative
`procedure, the information is used qualitatively.
`
`Qualitative Information
`The qualitative information obtained with these techniques is related to the struc-
`ture of the solid. The DSC curves are a plot of heat flow (bH/~t) vs. temperature
`(or time), whereas the TGA curves show weight change vs. temperature (or time).
`The shape of the curve corresponding to a dehydration is related to the distri-
`bution of the energies of association between the water molecules and their host
`.and the kinetics of mass transfer in the crystal. If all the water molecules in a crystal
`occupy identical sites, they should have the same energies of association. When
`the sample is supplied with sufficient heat energy to overcome that association, the
`water molecules should be released at the same time. This would correspond to .a
`sharp endotherm in the DSC curve spanning a relatively narrow temperature range.
`The TGA would show sharp loss of weight corresponding-to the tempera, ture range
`observed in the DSC (Fig. 3). If the water exists in varied environments the energies
`of association are distributed over a greater range. This results in a broad endotherm
`or no distinguishable endotherm, in the extreme of loosely adsorbed water in the
`DSC curve, and a continuous weight loss in the TGA curve. In general, the more
`regular the environment and the more crystalline the hydrate, the sharper the
`thermal events.
`The temperature at which dehydration occurs is also related to the hydrate
`structure. Some crystals dehydrate at very low temperatures (25-40°C), whereas
`others retain their water until fusion occurs. Crystalline hydrates may dehydrate
`at low temperatures and still have sharp DSC and TGA events [27]. Typically, the
`higher the dehydration temperature, the sharper the endotherm and weight loss.
`Water in channels may be uniformly and strongly bound and still show the onset
`of dehydration at low temperatures. Channels may also dehydrate more rapidly
`than "isolated site" crystalline hydrates when isothermally treated at moderate
`temperatures. This kinetic effect is a result of the packing of the crystal and the
`statistical reality that a fraction of the water-water or water-drug hydrogen bonds
`
`Lupin Ex. 1053 (Page 11 of 52)
`
`

`

`402
`402
`
`Hydrates
`Hyd rates
`
`0
`0
`0
`
`Doom
`
`.ohmfi
`
`L
`
`
`
`”Eon—.5mo.v>0mm
`
`0
`
`03
`
`80vmSHEwQEmH
`
`
`
`.BmhfibmvofivmflmooH853UmQ
`
`.mm.GE
`
`
`
`(fi/M) Mou leeH
`
`Lupin Ex. 1053 (Page 12 of 52)
`Lupin Ex. 1053 (Page 12 of 52)
`
`
`
`

`

`B
`
`100-
`
`.57~ ~
`457 ~g)
`
`rn
`x
`
`80
`
`0
`
`5O
`
`~o
`
`T~mperBt.ure
`
`~o
`TGA V4.0D DuPon’~ 2_000
`
`FiG. 3b. TGA plo~ for cephradirte dihydrate.
`
`

`

`404
`
`Hydrates
`
`can break at any temperature. This point is reexamined in the Fourier transform
`infrared spectroscopy (FTIR) discussion.
`Additional qualitative information is obtained from the DSC and TGA curves
`after dehydration, where in the DSC curve no melting endotherm occurs if the
`anhydrous form of the drug is amorphous. The temperature increase until decom-
`position is observed. Decomposition may show as an exotherm or endotherm, but
`unlike melting is usually accompanied by a weight change in the TGA.
`If dehydration does not precede melting, a DSC exotherm (and no TGA weight
`change) may occur after melting. This represents hydrate melting, followed by a
`recrystallization to the anhydrous form which then melts [28]. Many common
`hydrates have a dehydration endotherm followed by the melting endotherm of the
`anhydrous material. This can be verified by heating the sample in the DSC just
`past the dehydration endotherm and cooling and reheating to the melting endoth-
`erm. This is called thermal cycling. If no dehydration endotherm is present while
`the melting endotherm remains unchanged, the melting point is that of the an-
`hydrous material.
`
`Karl Fischer Titration
`
`The Karl Fischer method (KF) of moisture determination is an oxidation-reduction
`titration in which water is consumed during the oxidation of sulfur dioxide by iodine
`[29]. The reactions may be written as follows:
`
`CsHsNI2 + CsHsN SO2 + CsHsN + H20
`
`~ 2 CsHsN HI + CsHsN SO3
`
`CsHsN SO3 + CH3OH > CsHsN(H)SO4CHa
`The end point occurs when all the water i~s consumed and an excess of the pyridine-
`iodine complex appears. This may be determined by volumetric or electrometric
`titration. Pharmaceutical hydrates typically contain relatively small amounts of
`moisture (1,5%) and are usually analyzed coulometrically. The coulometric titra-
`tion is more sensitive and requires little sample, depending on the moisture content.
`A range of 0.1 to 3 mg total moisture is reasonable for most commercial instruments
`(Brinkman). However, the procedure for handling samples with very Iittle moisture
`(~0.1% w/w) must include precautions to avoid moisture pickup.
`Most samples in pharmaceutical applications are solids, and it is desirable that
`they be soluble in the methanolic Karl Fischer reagents. This is not a necessary
`condition, however, if sufficient time is allowed for the extraction of water. Al-
`ternatively, the sample may be dissolved in an anhydrous solvent prior to analysis.
`Some compounds may cause erroneous results in a Karl Fischer titration by reacting
`with one of the components to form water or shifting the iodide-iodine redox
`couple. Carboxylic acids can esterify with the methanol and liberate water, alde-
`hydes can form acetals and release water, and thiols may be oxidized to disulfides
`by iodine [29,30]. Substituted Karl Fischer reagents are commercially available for
`aldehydes and organic acids (e.g., in the Hydranal series from Riedel-deHaen),
`but the thiol-type problems are more difficult to handle. Oven attachments are
`also available to heat the sample in a stream of dry nitrogen which is bubbled into
`the titration vessel to avoid the problem.
`Coupling the Karl Fischer results with the TGA results shows what portion of
`the TGA weight loss is attributable to moisture and what is due to other volatiles.
`
`Lupin Ex. 1053 (Page 14 of 52)
`
`

`

`Hydrates
`
`405
`
`This shows the importance of linking techniques when dealing with hydrates. In
`the absence of single-crystal x-ray (see below), combining information from various
`techniques is necessary to obtain structural and stoichiometric information.
`
`X-ray Diffraction Techniques: Powder X-ray Diffraction and Single-Crystal
`X-ray Diffraction
`
`The x-ray diffraction (XRD) techniques employed for characterizing hydrates and
`other pharmaceutical solids are used for single crystals and powder. In single-
`crystal XRD the structure of the crystal is elucidated, whereas powder XRD pro-
`vides a fingerprint of the crystal structure for comparison with other samples. The
`x-rays .are diffracted "by the electrons around the atoms. The higher the electron
`density, the more intense the scattering. The relationship between the wavelength
`(X), diffraction angle (~)), and the distance between the periodic planes of diffracting
`atoms (d) is given by Bragg’s law in Eq. (7)
`
`nX = 2d sin 0 (n = 1,2,3,...)
`
`(7)
`
`Single- Crystal X-ray Diffraction
`
`The smallest translationally repeating unit of a crystal is called the unit cell. Knowl-
`edge of the unit cell structure makes it possible to construct "packing diagrams"
`which consist of unit cells translated in three-dimensions symmetry operations (Fig,
`4, see color plate). The structure obtained from single-crystal XRD shows the
`position of water molecules regularly incorporated in the lattice. Single-crystal
`XRD distinguishes between fractional occupancy of sites by water from true frac-
`tional hydrates. In fractional hydrates, for example a hemihydrate, the mole ratio
`of water to drug is 0.5:1. This same approximate ratio may be coincidentally
`observed by TGA or KF in a crystal that prefers to be a monohydrate at only 50%
`occupancy of the water sites.
`The packing diagrams also reflect the nature of the association of water with
`the lattice. The water may be hydrogen bonded in the lattice with