zeitler_2007_int j pharm (2)

International Journal of Pharmaceutics 334 (2007) 7884

Drug hydrate systems and dehydrationterahertz pulsed spectr

ka Raltoge CBx 913bridgersityty of Tnish Uke, Uni

er 20r 2006

Abstract

Terahertzlactose, carbthe far infrarto the anhydvery charact 2006 ElseKeywords: Te

1. Introdu

It has becan existforms mayanhydrousthis has timplicationfactors mayexposure to2005b). Suproductiondrug with e

CorresponCentre, P.O. BTel.: +358 9 1fax: +358 9 1

E-mail ad

0378-5173/$doi:10.1016/jpulsed spectroscopy was used to distinguish between different hydrate systems. In the example of four pharmaceutical materialsamazepine, piroxicam and theophylline it was demonstrated that all different hydrate and anhydrate forms exhibit distinct spectra ined. Furthermore the dehydration of theophylline monohydrate was characterised in situ. Here, a phase transition from the monohydraterous form was observed, followed by evaporation of the hydrate water in a second step. The rotational spectrum of water vapour iseristic in the far infrared and can easily be discerned from the terahertz spectrum of the solid state form.vier B.V. All rights reserved.rahertz pulsed spectroscopy (TPS); Hydrate forms; Theophylline monohydrate; Dehydration; Phase transition; Principal components analysis

ction

en estimated that one in every three drug compoundsas a hydrated form (Threlfall, 1995). Hydratedhave different physicochemical properties to theircounterparts, and for pharmaceutical development

herapeutic, manufacturing, legal and commercials (Brittain, 1999; Bernstein, 2002). A number oftrigger and/or promote hydrate formation, includingwater and temperature changes (Airaksinen et al.,

ch conditions frequently occur during dosage formprocesses. For example, before or during mixing axcipients for tablet production it is common practice

ding author at: Drug Discovery and Development Technologyox 56, FI-00014 University of Helsinki, Finland.91 59674;91 59144.dress: [email protected] (C.J. Strachan).

to perform a wet granulation step where the powder blend isgranulated into larger particles by aqueous polymer solutionto achieve better flow properties. The aggregates must thenbe dried, using a process such as fluid bed drying. Duringsuch processes the drug is very likely to form different hydrateand/or anhydrate forms (Morris et al., 2001; Airaksinen et al.,2005a,b; Aaltonen et al., 2006). Hydrate formation in tabletsstored in ambient conditions, especially in dosage forms thatcontain excipients with loosely bound water, such as starch orcellulose, may also occur. Therapeutic failure has been attributedto this phenomenon highlighting the importance to understandthe hydration behaviour of drug materials (Meyer et al., 1992).

Hydrate forms of drugs have been classified into threedifferent categories according to the distribution of waterwithin their crystal lattice: channel, isolated-site and ion-boundhydrates (Morris, 1999). These classes are postulated to exhibitdifferent solid state transformation mechanisms, though suchmechanisms and the consequences thereof are poorly understoodat present. To control the solid state properties of materials

see front matter 2006 Elsevier B.V. All rights reserved..ijpharm.2006.10.027J. Axel Zeitler a,b,c, Karin Kogermann d,e, JukPhilip F. Taday a, Michael Pepper a,c, Jaakko A

a TeraView Ltd., St Johns Innovation Park, Cambridb School of Pharmacy, University of Otago, P.O. Bo

c Cavendish Laboratory, University of Cambridge, Camd Division of Pharmaceutical Technology, Univ

e The Department of Pharmacy, Universif Department of Pharmaceutics and Analytical Chemistry, The Da

Copenhagen, Denmarg Drug Discovery and Development Technology Centr

Received 25 May 2006; received in revised form 8 OctobAvailable online 21 Octobeprocesses studied byoscopyantanen f , Thomas Rades b,

nen d,g, Clare J. Strachan d,g4 0WS, United Kingdom, Dunedin, New Zealande CB3 0HE, United Kingdomof Helsinki, Finlandartu, Estonianiversity of Pharmaceutical Sciences,

versity of Helsinki, Finland06; accepted 17 October 2006

J.A. Zeitler et al. / International Journal of Pharmaceutics 334 (2007) 7884 79

during dosage form production, it is imperative to obtain amolecular level understanding of dehydration and hydrationprocesses.

Amongpowder difincreasinglThere aretechniquesinter- andtool to una moleculanon-destruvery rapidtechniquesformationprocesses.been demothe investimaterials a2004, 2005polymorphIn recent tha better inteinformationlattice (Day

The purthe terahertand to evalucharacteris

2. Materia

2.1. Mater

Lactose-D-glucopScientificdibenz[b,f](PXM, 4-zine-3-carbfrom HawMN, USApurine-2,6-Germany).

2.2. Sampl

The Lanhydrate fH2O in a sal., 2002).

The suppolymorphdihydrate f373 K. CBfor 2 h atprepared b

aqueous solution. The solution was cooled slowly to roomtemperature. After filtration the crystals were dried at ambientconditions and the obtained hydrate forms were stored at 54%

M mecrystoredmon

mo

ion oo thhe manhaturrefeCamor tPN0FNOP01

HEOationon thlable

-ray

ay pecor

8 Ae codiff

rahe

pleem,uentaherwas

TPSith

cmaned fcedwer

theaturwindof thasur

. Spceddeh

d temother techniques such as thermal analysis and X-rayfraction (XRPD), spectroscopic methods have beeny used to characterise hydrate/anhydrate systems.a number of advantages for using spectroscopicto analyse these systems. The spectra directly reflectintramolecular properties, and therefore provide aderstand hydration and dehydration processes atr level. In addition, spectroscopic techniques are

ctive and, with the exception of solid state NMR,compared to conventional methods. Some of thehave the potential to study hydrate or anhydrateand dehydration in real time during productionTerahertz pulsed spectroscopy (TPS) has recentlynstrated to be an interesting novel technique for

gation of solid state properties in pharmaceuticalnd products (Taday et al., 2003; Strachan et al.,; Zeitler et al., 2006). It has been used to monitoric phase transitions in real time (Zeitler et al., 2005).eoretical studies first approaches were presented forrpretation of the spectra which predominantly reflecton the intermolecular vibrations in the crystallineet al., 2006; Allis et al., 2006).

pose of the study is to investigate the differences inz pulsed spectra of anhydrous and hydrated materialsate the potential of variable temperature TPS for the

ation of dehydration processes.

ls and methods

ials

-monohydrate (L-H2O, 4-O-D-galactopyranosyl-yranose monohydrate) was obtained from Fisher

(Loughborough, UK). Carbamazepine (CBZ, 5H-azepine-5-carboxamide, USP grade) and piroxicamhydroxy-2-methyl-N-(2-pyridyl)-2H-1,2-benzothia-oxamide-1,1-dioxide, USP grade) were purchasedkins, Inc. Pharmaceutical Group (Minneapolis,). Theophylline (TP, 3,7-dihydro-1,3-dimethyl-1H-dione) was obtained from BASF (Ludwigshafen,

e preparation

-H2O form was used as received. The unstable -orm of lactose (LH) was prepared by heating L-tainless steel Petri dish at 383 K for 6 h (Garnier et

plied form of CBZ was polymorph III. To ensureic purity form III was obtained by dehydration of theorm under reduced pressure (72 mBar) for 24 h atZ form I was prepared by heating the raw material443 K (Lefebvre et al., 1986). The dihydrate wasy recrystallization from a hot (353 K) saturated

RH.PX

I) by rwere s

of theThe

tallizatprior tform t

Alltemper

The(CSD,UK) fCBMZI), FECIDYAand Tinformbasedis avai

2.3. X

X-rwere r

AXS DFor threticalused.

2.4. Te

SamInduchsubseqthe terpelletsusing aUK) w2130(Strachwas us

referenspectra

Fortemperquartzfingerthe meto 90 Kreferen

Theelevateonohydrate was prepared from the raw material (formtallization as described for CBZ dihydrate. Samplesat 54% RH. Pure form I was obtained by dehydration

ohydrate (72 mBar, 24 h, 373 K).nohydrate form of TP was also produced by recrys-f the raw anhydrous material and stored at 75% RH

e measurements. For the samples of the anhydrateonohydrate was dehydrated (72 mBar, 24 h, 373 K).ydrous forms were stored over silica gel at roome.rence codes in the Cambridge Structural Databasebridge Crystallographic Data Centre, Cambridge,

he respective forms are LACTOS10 (L-H2O),1 (CBZ form III), CBMZPN11 (CBZ formT (CBZ dihydrate), BIYSEH (PXM form I),(PXM monohydrate), BAPLOT01 (TP anhydrate)

PH01 (TP monohydrate). For the crystallographicof LH the structure solved by Platteau et al. (2005)

e powder pattern was used as no single crystal datafor this modification.

powder diffraction

owder diffraction (XRPD) patterns of all formsded by using a thetatheta diffractometer (Brukerdvance, Bruker AXS GmbH, Karlsruhe, Germany).mparison of the recorded patterns to their theo-raction patterns reference data from the CSD was

rtz pulsed spectroscopy

material (Table 1) was mixed with polyethylene (PE,Volketwil, Switzerland, particle size < 10 m) andly compressed at a load of 2 tonnes into a pellet fortz spectra at room temperature. The thickness of thebetween 3.2 mm and 3.4 mm. Spectra were recordedspectra1000 V spectrometer (TeraView, Cambridge,

a spectral resolution of 1 cm1 over the range of1 by co-adding 1800 scans as reported previouslyet al., 2005). Blackman-Harris 3-term apodizationor the Fourier transformation. Sample spectra wereagainst the spectrum of a PE pellet and absorbancee calculated.acquisition of terahertz spectra at 90 K a variable

e cell (Specac, Orphington, UK) equipped with z-cutows was used. The sample was attached to the cold

e temperature cell and the cell was evacuated duringements. Liquid nitrogen was used to cool the sampleectra were recorded by co-adding 1800 scans andagainst a spectrum of PE at 90 K.ydration of TP monohydrate was investigated atperatures. TP monohydrate (15 mg) was mixed with

80 J.A. Zeitler et al. / International Journal of Pharmaceutics 334 (2007) 7884

Table 1Amount of sample material used to prepare the pellets for the measurements at293 K and 90 K after mixing with 360 mg PE

293 K 90 K

LH 15 15L-H2O 15 15CBZ I 15 20CBZ III 15 20CBZ dihydrate 30 20PXM I 30 15PXM monohydrate 15 15TP anhydrate 15 15TP monohydrate 15 15

poly(tetrafluoroethylene) (PTFE, Aldrich, UK) and compressedinto a pellet. In situ spectra of TP during heating were acquiredusing a variable temperature cell (Specac, Orphington, UK) asdescribed previously (Zeitler et al., 2005). The sample pellet washeated at 5 K min1 and terahertz spectra were acquired every30 s by co-adding 900 scans per spectrum. Absorbance spectrawere calculated using a reference spectrum of PTFE. The samplechamber was purged with dry nitrogen gas throughout all mea-surements. Spectral acquisition and processing was performedusing TPI Spectra software (TeraView, Cambridge, UK). Thespectral refractive indices of each material at room temperaturewere calculated directly using the measured terahertz electricfield and the sample thickness.

2.5. Multivariate analysis

Multivariate analysis of the spectral data acquired for thedehydration of TP was performed by principal componentsanalysis on centred spectral data between 5 cm1 and 110 cm1(PCA, SIMCA-P, Version 10.5, Umetrics AB, Umea, Sweden).In PCA the data is decomposed into a new set of variables calledprincipal components (PCs). The PCs are composed of two setsof data: the scores (t) and loadings (p) with the scores referringto spectral variation and the loadings representing the spectralcontribution to each PC. The analysis can then be restricted tothe first couple of PCs that describe the highest variance in thespectra (Eriksson et al., 1999).

3. Results and discussion

3.1. Terahertz spectra of different hydrate forms

The crystal structures of all forms were verified by theirXRPD patterns compared to reference powder patterns (datanot shown). All forms show distinct differences in their terahertzspectra that allow the hydrate and anhydrate forms to be readilydistinguished by their room temperature spectra (Fig. 1 andTable 2). The measured refractive indices of the materials werefound to be between 1.6 and 2.3 at room temperature.

Upon cooling the samples to 90 K an increase in intensityof all spectral features can be observed. The peak widths

Fig. 1. TerahCBZ forms I,ertz absorption spectra and frequency dependent refractive indices of the different hIII, and dihydrate, (C) PXM form I and monohydrate, and (D) TP anhydrate and moydrate and anhydrate forms at 293 K. (A) LH and L-H2O, (B)nohydrate.


Fig. 2. Terahe ) LH and L-H2O, (B) CBZ forms I, III, and dihydrate, (C) PXM formI and monohy

decrease, tand featureTable 2).

No refraat 90 K asthe samplethickness othe absolutbe accurate

3.2. In situ

The dehin situ. Thby a peakdoublet atspectral feain intensity

At 350intensity in93 cm1. Ttransitionsacquisitionthe narrowunder-resotral resolutapodisationwater vapoin Fig. 3).nitrogen thvapour in tfrom the m

hough only two out of the numerous spectral features ofvapour can be observed at this temperature, the doubletTP monohydrate spectrum can no longer be discerned insence of the water vapour signature. A general increaserbance at higher wavenumbers leads to a further decreaseal-to-noise ratio and the spectra therefore become noisierer wavenumbers.rtz absorption spectra at 90 K for the different hydrate and anhydrate forms. (Adrate, and (D) TH anhydrate and monohydrate.

he peak positions blue-shift to higher wavenumberss that overlap at room temperature resolve (Fig. 2 and

ctive indices were calculated for the terahertz signalit was not possible to measure the thickness ofat this temperature. Due to thermal contraction the

f the sample pellets would decrease considerably ande value for the spectral refractive indices would not.

Altwaterin thethe prein absoin signat highdehydration study of theophylline monohydrate

ydration process of TP monohydrate was investigatede monohydrate spectrum at 293 K is dominatedat 55 cm1 with a shoulder at 60 cm1 and a

92 cm1 and 96 cm1. Upon heating the sample thetures red-shift to lower wavenumbers and decrease(Fig. 3).

K two narrow spectral features start to increase inthe spectrum of TP monohydrate at 88 cm1 andhese features correspond to two of the rotationalin the terahertz spectrum of water vapour. With theparameters used for these terahertz measurementslines of the rotational spectrum of water vapour are

lved by orders of magnitude. However, even at a spec-ion of 1 cm1 and after applying BlackmanHarris

during the Fourier transformation, the spectrum ofur is still very distinct and reproducible (see insetAs the sample chamber is carefully purged with dryroughout the measurements, the presence of waterhe spectra indicates the evaporation of hydrate waterolecular crystal.

Fig. 3. Terahertz absorption spectra of TP monohydrate during heating from293 K to 437 K. The spectra are offset in absorbance for clarity. The insetshows the terahertz spectrum of water vapour recorded with the same acquisitionparameters as for the variable temperature measurements.


Table 2Positions of spectral features for the different hydrate forms in the terahertz spectra at 293 K and 90 K

293 K 90 K

LH (50, 53), 65, 76w, 80, 91, 96L-H2O 47, 62, 89CBZ form III , 67, 75w, 85, 96CBZ form I w, 33, 35s, 37w, 47, 55, 59w, 66w, 73, 86, 97CBZ dihydra w, 55w, 63s, 67, 73w, 76, 81w, 86, 91, 98PXM form I w, 58w, 61w, 69, 73, 87, 93PXM monoh 56, 62, 70, 75w, 80, 84TP anhydrate , 41w, 48w, 54, 60, 70s, 82w, 85w, 96TP monohyd s, 58, 64, 69w, 74w, 80, 82s, 94Parenthesis re

An addispectrum adoublet dec53 cm1 instructure toobservation(Suihko etform of TPSuryanarayspectra thapossible thdetectable.behavioursystems (Sbeing obseinto a hydroclosed systThis may eupon dehy360 K leadpeak 100 cmthree waterfurther hea

This obsthree rotatiother peaksonto the Tthe spectratemperaturof the wateat 437 K al

In the c(Fig. 4) thevaporationprocess instep procesin a secondthe evapordiffusion thwas found

Recentlspectra ofmonohydraof the anhy

2006)) and of the monohydrate match the spectra in Fig.other anhydrous form (called form I) was prepared by

g form II to 523 K and keeping it isothermal for 5 h. This iso be the form prepared under similar conditions by Suzuki1989), and is different to the metastable form observeddnis and Suryanarayanan (1997) and Vora et al. (2004).y of the phase transition between the monohydrate andI and between form II and form I was performed but thes were not measured in situ. For each terahertz spectrumthe phase transition a sample of the material was heated

ertain time, allowed to cool and then a pellet was preparedhe heat treated material. With this approach it was possiblecribe the overall phase transition from the monohydrateanhydrate (form II). However, the second step of thiss involving the evaporation of the hydrate water (see Fig.ld not be detected. The observation of this step requires thes to be monitored in situ as the event of water evaporationient and intrinsically tied to the temperature of the sample.hough in the work by Upadhya et al. (2006) the powdermonohydrate is heated rather than the compressed pellets no indication of a metastable form of TP as describeddnis and Suryanarayanan (1997). It is possible that theable form, once formed during the dehydration, convertsnto the stable anhydrate during the preparation of thepellet. To confirm the formation of the metastable formdehydration it would therefore be necessary to acquirespectra of a sample of TP monohydrate in powder formthan in a pellet.30w, 36w, 47w, 54w, 62, 76w, 87w, 94w, 104w 33, 41,18, 40, 46, 60, 85, 96, 108 18, 41,28w, 41, 60, 68w, 80w, 88s, 91, 105s, 109, 122 31w, 4531, 44w, 52, 60w, (85, 93), (111, 116, 122) 25w, 29

te 39w, 51s, 62, 71, 80w, 101 39w, 4632, 40w, 50w, 63s, 68, 81s, 89, 106 36w, 44

ydrate 39, 46, 52w, 65, 75, 96, 125 41, 49,25w, 32, 43w, 54, 65w,s, 90 28w, 36

rate 28w, 55, 59s, 77s, (91, 96) 30w, 52fer to unresolved spectral features, w denotes a weak feature and s a shoulder.

tional broad peak starts to emerge at 32 cm1 in thet 355 K. At the same time the intensity of the initialreases drastically and the peak at 55 cm1 shifts to

dicating a phase transition of the monohydrate crystalthe crystal structure of the anhydrous form. Thiscorresponds very well to studies by thermal analysisal., 1997). An intermediate metastable anhydrousupon dehydration has been reported (Phadnis and

anan, 1997). There are no features in the terahertzt suggest that this metastable form occurred. It isat this form is occurring at concentrations that are notHowever, it has also been noted that the dehydrationof TP is considerably different in open and closeduihko et al., 1997), with the reported metastable formrved under open conditions. Compression of the TPphobic matrix of PTFE can be considered closer to a

em, since the water cannot freely escape the sample.xplain the apparent absence of the metastable formdration in the terahertz spectra. Further heating tos to an additional peak at 25 cm1. Another sharp

1 due to water vapour appears. The intensity of thevapour features remains weak but constant during

ting.ervation changes at 375378 K. The intensity of the

onal spectral features increase strongly and nearly allof the water vapour spectrum appear superimposed

P anhydrate spectrum at 378 K. For the next 120 sl signature of water vapour is very dominant until thee has reached 388 K. From then onwards the intensityr vapour signal decreases rapidly and in the spectruml water vapour features disappear.ontour plot of the variable temperature experimente peak shifts, phase transition and hydrate water

et al. (1. Theheatinlikely tet al. (by PhaA studform Isampleduringfor a cusing tto desto theproces3) couprocesis transEven tof TPthere iby Phametastback isampleduringin siturathercan be followed in more detail. The dehydrationthis sample (a compressed pellet) is clearly a twos, with a phase transition and the water evaporationstep. The time lag between the phase transition and

ation step is likely to result from the water vapourrough the PTFE matrix. The dehydration experimentto be reproducible.y, Upadhya et al. (2006) have reported the terahertz

two different anhydrous modifications and thete of TP. The spectra that have been presented for onedrous forms (called form II in the work by Upadhya

Fig. 4. Contour plot of the terahertz spectra during dehydration of TPmonohydrate.


Fig. 5. PCA p nents(p[1] and p[2]

PCA waexperiment92.7% ofand secondchange untincrease insmaller inc5B) are posnegativelythe changethe monohy

With fuHowever,decreases bfurther decof p[2] (Figa strong reoration stepclearly sep

The furis associatanhydrate s

4. Conclus

Anhydrcal materiarecorded aperature arhow variabdration inboth intermtional tranthe phasethe evapormakes TPSmechanism

Acknowled

The autSmithKlinetemperatur

R aademke ton fu

nce

n, J.,itorinraman

s.en, S

tanen,-stateS Phaen, Santanlid do

.G., Pertz195n, J.,tallogH.G.York., Ze

influeulatio447, L.,avariaa.

S.,tallisan, Alots of TP monohydrate during heating. (A) Scores for first and second compo).

s used to further investigate the variable temperature(Jrgensen et al., 2006). The first PC represented

the spectral variation, the second, 5.7%. The firstscores (t[1] and t[2], respectively, Fig. 5A) do not

il the sample reaches 343 K. At this point a larget[1] can be observed until 365 K, along with a

rease in t[2]. The loadings of the first PC (p[1], Fig.itively correlated to the TP anhydrate spectrum andcorrelated to the TP monohydrate spectrum. Thus,in t[1] is associated with the solid state change fromdrate to the anhydrate crystal structure.

rther heating, little change can be observed in t[1].t[2] sharply increases at 362 K and subsequentlyack to its original value at 386 K, followed by a

rease until heating is terminated at 437 K. Inspection. 5B) reveals that the loadings for the second PC havesemblance to the water vapour spectrum. The evap-

of the hydrate water from the sample pellet can bearated from the phase transition through t[1] and t[2].ther decrease in t[2] after the water vapour signaled with peak broadening and red shifting in thepectrum until the experiment is terminated.

ions

ous and hydrated forms of crystalline pharmaceuti-ls can easily be distinguished using TPS. Spectrat 90 K indicate that several features at room tem-e multiple overlapping modes. It was demonstratedle temperature TPS can be used to monitor dehy-situ. As terahertz spectra represent information on

CJS, Jthe Acalso liTurune

Refere

Aaltonemon

andpres

AiraksinRansolidAAP

AiraksinE., Rof so

Allis, Dterah1951

BernsteiCrys

Brittain,New

Day, G.Mthecalc110,

ErikssonMegUme

Garnier,crys

Jrgense

olecular interactions in solid samples and rota-

sitions in the gas phase it is possible to observetransition during dehydration and in a second stepation of the hydrate water from the sample. This

a valuable technique for the study of dehydrations.

gements

hors would like to acknowledge Ian Lynch, Glaxo-, Stevenage, UK for the generous loan of the variable

e cell used for the low temperature experiments.

Antikainein evaluat

Lefebvre, C.,J.C., 1986compressi

Meyer, M.C.,1992. Thclinical fa

Morris, K.R.,(Ed.), PolYork, USA

Morris, K.R.,approacheduring ma(t[1] and t[2]). (B) Loadings of the first and second components

nd JA are grateful for the financial support fromy of Finland (decision number 107343). JA wouldacknowledge the Finish Cultural Foundation (Elli

nd).

s

Kogermann, K., Strachan, C.J., Rantanen, J., 2006. In-situg of solid state changes during fluid bed drying using near infraredspectroscopy. Chem. Eng. Sci. doi:10.1016/j.ces.2006.08.061; in

., Karjalainen, M., Kivikero, N., Westermarck, S., Shevchenko, A.,J., Yliruusi, J., 2005a. Excipient selection can significantly affectphase transformation in formulation during wet granulation.

rm. Sci. Tech. 6 (2), article 41, http://www.aapspharmscitech.org/.., Karjalainen, M., Shevchenko, A., Westermarck, S., Leppanen,en, J., Yliruusi, J., 2005b. Role of water in the physical stabilitysage formulations. J. Pharm. Sci. 94, 21472165.

rokhorova, D.A., Korter, T.M., 2006. Solid-state modeling of thespectrum of the high explosive HMX. J. Phys. Chem. A 110,9.2002. Polymorphism in Molecular Crystals. IUCr Monographs onraphy. Oxford University Press, Oxford.(Ed.), 1999. Polymorphism in Pharmaceutical Solids. M. Dekker,.itler, J.A., Jones, W., Rades, T., Taday, P.F., 2006. Understandingnce of polymorphism on phonon spectra: lattice dynamicsns and terahertz spectroscopy of carbamazepine. J. Phys. Chem. B456.Johansson, E., Kettaneh-Wold, N., Wold, S., 1999. Multi- andte Data Analysis; Principles and Applications. Umetrics AB,

Petit, S., Coquerel, G., 2002. Dehydration mechanism andtion behaviour of lactose. J. Therm. Anal. 68, 489502..C., Miroshnyk, I., Karjalainen, M., Jouppila, K., Siiria, S.,

n, O., Rantanen, J., 2006. Multivariate data analysis as a fast toolion of solid state phenomena. J. Pharm. Sci. 95, 906916.Guyothermann, A.M., Draguetbrughmans, M., Bouche, R., Guyot,. Polymorphic transitions of carbamazepine during grinding andon. Drug Dev. Ind. Pharm. 12, 19131927.Straughn, A.B., Jarvi, E.J., Wood, G.C., Pelsor, F.R., Shah, V.P.,

e bioinequivalence of carbamazepine tablets with a history ofilures. Pharm. Res. 9, 16121617.1999. Structural aspects of hydrates and solvates. In: Brittain, H.ymorphism in Pharmaceutical Solids. Marcel Dekker Inc., New, 126179.Griesser, U.J., Eckhardt, C.J., Stowell, J.G., 2001. Theoretical

s to physical transformations of active pharmaceutical ingredientsnufacturing processes. Adv. Drug Deliv. Rev. 48, 91114.


Phadnis, N.V., Suryanarayanan, R., 1997. Polymorphism in anhydroustheophyllineimplications on the dissolution rate of theophylline tablets. J.Pharm. Sci. 86, 12561263.

Platteau, C., Lefebvre, J., Affouard, F., Willart, J.F., Derollez, P., Mallet, F., 2005.Structure determination of the stable anhydrous phase of alpha-lactose fromX-ray powder diffraction. Acta Crystallogr. Sect. B-Struct. Sci. 61, 185191.

Strachan, C.J., Rades, T., Newnham, D.A., Gordon, K.C., Pepper, M., Taday,P.F., 2004. Using terahertz pulsed spectroscopy to study crystallinity ofpharmaceutical materials. Chem. Phys. Lett. 390, 2024.

Strachan, C.J., Taday, P.F., Newnham, D.A., Gordon, K.C., Zeitler, J.A.,Pepper, M., Rades, T., 2005. Using terahertz pulsed spectroscopy toquantify pharmaceutical polymorphism and crystallinity. J. Pharm. Sci. 94,837846.

Suihko, E., Ketolainen, J., Poso, A., Ahlgren, M., Gynther, J., Paronen, P.,1997. Dehydration of theophylline monohydratea two step process. Int. J.Pharm. 158, 4755.

Suzuki, E., Shimomura, K., Sekiguchi, K., 1989. Thermochemical study oftheophylline and its hydrate. Chem. Pharm. Bull. 37, 493497.

Taday, P.F., Bradley, I.V., Arnone, D.D., Pepper, M., 2003. Using terahertzpulse spectroscopy to study the crystalline structure of a drug: a case study

of the polymorphs of ranitidine hydrochloride. J. Pharm. Sci. 92, 831838.

Threlfall, T.L., 1995. Analysis of organic polymorphsa review. Analyst 120,24352460.

Upadhya, P., Nguyen, K., Shen, Y., Obradovic, J., Fukushige, K., Griffiths,R., Gladden, L., Davies, A., Linfield, E., 2006. Characterization ofcrystalline phase-transformations in theophylline by time-domain terahertzspectroscopy. Spectr. Lett. 39, 215224.

Vora, K.L., Buckton, G., Clapham, D., 2004. The use of dynamic vapour sorptionand near infra-red spectrscopy (DVS-NIR) to study the crystal transitionsof theophylline and the report of a new solid state transition. Eur. J. Pharm.Sci. 22, 97105.

Zeitler, J.A., Newnham, D.A., Taday, P.F., Strachan, C.J., Pepper, M., Gordon,K.C., Rades, T., 2005. Temperature dependent terahertz pulsed spectroscopyof carbamazepine. Thermochim. Acta 436, 7177.

Zeitler, J.A., Newnham, D.A., Taday, P.F., Threlfall, T.L., Lancaster, R.W.,Berg, R.W., Strachan, C.J., Pepper, M., Gordon, K.C., Rades, T., 2006.Characterization of temperature induced phase transitions in the fivepolymorphic forms of sulfathiazole by terahertz pulsed spectroscopy anddifferential scanning calorimetry. J. Pharm. Sci. 95, 24862498.

Drug hydrate systems and dehydration processes studied by terahertz pulsed spectroscopyIntroductionMaterials and methodsMaterialsSample preparationX-ray powder diffractionTerahertz pulsed spectroscopyMultivariate analysis

Results and discussionTerahertz spectra of different hydrate formsIn situ dehydration study of theophylline monohydrate

ConclusionsAcknowledgementsReferences

zeitler_2007_int j pharm (2)

Documents