www.informahealthcare.comTelephone House, 69-77 Paul Street, London EC2A 4LQ, UK 52 Vanderbilt Avenue, New York, NY 10017, USAT hi rdE di T i on About the bookFreezedrying,orlyophilization,isawellestablishedtechnologyusedinthepreservationofnumerous pharmaceutical and biological products. This highly effective dehydration method involves the removal of water from frozen materials via the direct sublimation of ice. In recent years, this process has met with many changes,ashavetheregulationsthatimpactlyophilizationpractices.ThisneweditionofFreezeDrying/ LyophilizationofPharmaceuticalandBiologicalProductsaddressesthesechangeswithupdatedandnew chaptersonemergingdevelopmentsinlyophilizationtechnology,research,andindustryprocedures. Providing both a scientifc and industrial perspective, this comprehensive text is a valuable resource for all those who use freeze drying technology. Written by leading experts in the feld, the third edition: Includesfundamentalbackgroundonwaterandiceproperties,phasetransitionoffrozenliquids,andthebasicsoffreeze-dryingtechniques Analyzestheimportanceofproductpropertieslooksatthechallengesstemmingfromtheincreasingsensitivity of treated products Extensivelycoversformulationissueslooksatstabilizingadditivesandoverallformulationcharacterization Emphasizesthegrowingsignifcanceofproductcontainerandclosureinteractions Describessomenewup-frontdevelopmentssuchasthecouplingoflyophilizationwithirradiation Reviewsnewregulationsandbiologicalstandards IntroducesissuesregardingmanufacturingautomationprocessesspecifcallylooksatProcessAnalyticalTechnology(PAT)andQualitybyDesign(QbD) About the editorsLOUISREYisaScientifcAdvisorbasedinLausanne,Switzerland.HehasbeenaRegentsLecturerandVisitingProfessor at the University of California, Los Angeles, and at the University of Alaska, Fairbanks. The author,coauthor, or editor of 16 books and 130 papers and notes in eight languages, Louis Rey has presented 250lecturesworldwide.HehasworkedasaresearchscientistattheCentreNationaldelaRechercheScientifque,Paris,France,asaprofessorofphysical-chemicalbiologyattheUniversityofDijon,France,andasHeadofCorporateResearchandDevelopmentforNestl,Vevey,Switzerland.Today,Dr.ReyrunstheLyophilizationLaboratoryofAerialinIllkirch,France,andistheirScientifcAdvisor.HeisaFellowoftheNewYorkAcademyofSciencesandtheAcademyofSavoy,andhasmanagedtheInternationalCoursesonLyophilization.Dr.ReyreceivedhisM.A.(1952)andPh.D.(1958)degreesfromtheSorbonne,Paris,France.JOAN C. MAY has been the Director of the Laboratory of Analytical Chemistry for the Center for BiologicsEvaluationandResearch(CBER)oftheU.S.FoodandDrugAdministration,Rockville,Maryland.Theauthororcoauthorofmanyarticlesandthecoeditorofseveralbooksonfreeze-dryingprocesses,sheisamemberoftheAmericanChemicalSocietyandtheNorthAmericanThermalAnalysisSociety.Dr.MayreceivedherB.S.degree(1965)fromNazarethCollege,Rochester,NewYork,herM.S.degree(1968)fromtheUniversityofWisconsin-Madison,andherPh.D.degree(1971)inanalyticalchemistryfromtheUniversityofNotreDame,NotreDame,Indiana. Postdoctoral appointments included research at New York States Roswell Park Memorial Institute(1972)andtheNationalInstituteofStandardsandTechnology(1973)inGaithersburg,Maryland.DRUGS AND THE PHARMACEUTICAL SCIENCESVOLUME 206Freeze drying/ Lyophilization of Pharmaceutical and Biological Products Third Edition206Freeze Drying/Lyophilization of Pharmaceutical and Biological ProductsEdited byLouis Rey and Joan C. MayFreeze Drying/ Lyophilization of Pharmaceutical and Biological ProductsRey MayThird EdiTion[ vi j ay] D: / i nf or ma_Publ i shi ng/ REY_H100116_2400039/ z_pr oduct i on/z_3B2_3D_files/978-1-4398-2575-4_CH0000_O.3d][5/5/010/12:37:50][114]Freeze Drying/Lyophilization ofPharmaceutical andBiological Products[ vi j ay] D: / i nf or ma_Publ i shi ng/ REY_H100116_2400039/ z_pr oduct i on/z_3B2_3D_files/978-1-4398-2575-4_CH0000_O.3d][5/5/010/12:37:50][114]DRUGSANDTHEPHARMACEUTICALSCIENCESASeriesofTextbooksandMonographsExecutiveEditorJamesSwarbrickPharmaceuTech,Inc.Pinehurst,NorthCarolinaAdvisoryBoardRecentTitlesinSeriesGenericDrugProductDevelopment:International RegulatoryRequirementsforBioequivalence,editedbyIsadoreKanferandLeonShargelProteinsandPeptides:Pharmacokinetic,Pharmacodynamic,andMetabolicOutcomes,editedbyRandall J.MrsnyandAnnDaughertyPharmaceutical Statistics: Practical and Clinical Applications, Fifth Edition, SanfordBolton and Charles BonGenericDrugProductDevelopment:SpecialtyDosageForms,editedbyLeonShargel andIsadoreKanferActive Pharmaceutical Ingredients: Development, Manufacturing, and Regulation,SecondEdition,editedbyStanleyH.NusimFreezeDrying/LyophilizationofPharmaceutical andBiological Products,ThirdEdition,editedbyLouisReyandJoanC.MayLarryL.AugsburgerUniversityofMarylandBaltimore,MarylandRobertGurnyUniversitedeGeneveGeneve,SwitzerlandAjazHussainSandozPrinceton,NewJerseyKinamParkPurdueUniversityWestLafayette,IndianaHarryG.BrittainCenterforPharmaceuticalPhysics,Milford,NewJerseyAnthonyJ.HickeyUniversityofNorthCarolinaSchoolofPharmacyChapelHill,NorthCarolinaVincentH.L.LeeUSFDACenterforDrugEvaluationandResearchLosAngeles,CaliforniaStephenG.SchulmanUniversityofFloridaGainesville,FloridaJenniferB.DressmanUniversityofFrankfurtInstituteofPharmaceuticalTechnology,Frankfurt,GermanyJeffreyA.HughesUniversityofFloridaCollegeofPharmacyGainesville,FloridaJosephW.PolliGlaxoSmithKlineResearchTriangleParkNorthCarolinaJeromeP.SkellyAlexandria,VirginiaYuichiSugiyamaUniversityofTokyo,Tokyo,JapanGeoffreyT.TuckerUniversityofSheffieldRoyalHallamshireHospitalSheffield,UnitedKingdomElizabethM.ToppUniversityofKansas,Lawrence,KansasPeterYorkUniversityofBradford,SchoolofPharmacyBradford,UnitedKingdom[ vi j ay] D: / i nf or ma_Publ i shi ng/ REY_H100116_2400039/ z_pr oduct i on/z_3B2_3D_files/978-1-4398-2575-4_CH0000_O.3d][5/5/010/13:9:3][114]Freeze Drying/Lyophilization ofPharmaceutical andBiological ProductsThird EditionEdited byLouis ReyAERIAL, Illkirch, FranceCabinet d'Etudes, Lausanne, SwitzerlandJoan C. MayUnited States Food and Drug AdministrationRockville, Maryland, U.S.A.[ vi j ay] D: / i nf or ma_Publ i shi ng/ REY_H100116_2400039/ z_pr oduct i on/z_3B2_3D_files/978-1-4398-2575-4_CH0000_O.3d][5/5/010/12:37:50][114]Thiseditionpublishedin2010byInformaHealthcare, TelephoneHouse, 69-77Paul Street,LondonEC2A4LQ,UK.Simultaneously published in the USA by Informa Healthcare, 52 Vanderbilt Avenue, 7th Floor,NewYork,NY10017,USA.Informa Healthcare is a trading division of Informa UK Ltd. Registered Office: 3741 MortimerStreet,LondonW1T3JH,UK.RegisteredinEnglandandWalesnumber1072954.#2010InformaHealthcare,exceptasotherwiseindicatedNoclaimtooriginalU.S.GovernmentworksReprinted material is quoted with permission. Although every effort has been made toensurethat all owners of copyright material have been acknowledged in this publication, we would beglad to acknowledge in subsequent reprints or editions any omissions brought to our attention.All rights reserved. No part of this publication may be reproduced, stored in a retrieval system,or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording,orotherwise,unlesswiththepriorwrittenpermissionofthepublisherorinaccordancewiththe provisions of the Copyright, Designs and Patents Act 1988 or under the terms of any licencepermittinglimitedcopyingissuedbytheCopyright LicensingAgency, 90TottenhamCourtRoad, LondonW1P0LP, UK, ortheCopyrightClearanceCenter, Inc., 222RosewoodDrive,Danvers,MA01923,USA(http://www.copyright.com/ortelephone978-750-8400).Product or corporate names may be trademarks or registered trademarks, and are used only foridentificationandexplanationwithoutintenttoinfringe.Thisbookcontainsinformationfromreputablesourcesandalthoughreasonableeffortshavebeen made to publish accurate information,the publisher makes no warranties (either expressorimplied)astotheaccuracyorfitnessforaparticularpurposeoftheinformationoradvicecontained herein. The publisher wishes to make it clear that any views or opinions expressed inthisbookbyindividualauthorsorcontributorsaretheirpersonalviewsandopinionsanddonot necessarilyreflect the views/opinions of the publisher. Any informationor guidancecontainedinthisbookisintendedforusesolelybymedical professionalsstrictlyasasup-plement tothemedical professionals ownjudgement, knowledgeof thepatients medicalhistory, relevant manufacturers instructions andthe appropriate best practice guidelines.Because of the rapidadvances inmedical science, anyinformationor advice ondosages,procedures, or diagnoses shouldbe independently verified. This book does not indicatewhether a particular treatment is appropriate or suitable for a particular individual. Ultimatelyit isthesoleresponsibilityof themedical professional tomakehisorherownprofessionaljudgements, so as appropriately to advise andtreat patients. Save for death orpersonalinjurycausedbythepublishersnegligenceandtothefullest extent otherwisepermittedbylaw,neither the publisher nor any person engaged or employed by the publisher shall be responsibleorliableforanyloss, injuryordamagecausedtoanypersonorpropertyarisinginanywayfromtheuseofthisbook.ACIPrecordforthisbookisavailablefromtheBritishLibrary.LibraryofCongressCataloging-in-PublicationDataavailableonapplicationISBN-13:9781439825754Ordersmaybesentto:InformaHealthcare,SheepenPlace,Colchester,EssexCO33LP,UKTelephone:+44(0)2070175540Email:CSDhealthcarebooks@informa.comWebsite:http://informahealthcarebooks.com/Forcorporatesalespleasecontact:CorporateBooksIHC@informa.comForforeignrightspleasecontact:RightsIHC@informa.comForreprintpermissionspleasecontact:PermissionsIHC@informa.comTypesetbyMPSLimited,NewDelhi,IndiaPrintedandboundintheUnitedKingdom[ vi j ay] D: / i nf or ma_Publ i shi ng/ REY_H100116_2400039/ z_pr oduct i on/z_3B2_3D_files/978-1-4398-2575-4_CH0000_O.3d][5/5/010/12:37:50][114]ForewordPreservation of perishable products has always been a major concern forhumanity. Manyprocesses have beenusedinthe past, particularlynaturalrefrigeration, thenartificial refrigerationstartinginthe19thcentury. Freeze-drying was developed in the 20th century (1906), not only for biological materialand pharmaceuticals but also for certain foods, and not just instant coffee that isnowwidelyproduced.Prof Louis Rey was a key pioneer of thistechnology, through both publicandprivateresearchprojects, andtodayisstill oneof themost eminent spe-cialists in the field. The International Institute of Refrigeration (IIR) is honored tohavehadProfReyashonorarypresidentofitscommissiondevotedtocryobi-ologyandcryomedicine,andhonoraryheadofitsSectionCwitheverbroaderscopeencompassingbiologyandfoodtechnology.When I became director of the IIRin 2004, through my studies andexperience I was familiar only with freeze-drying of foodstuffs. I was extremelyhappy when I was contacted by Prof Rey and AERIAL at a later date concerninga scientific IIR conference on freeze-drying of pharmaceuticals: New Ventures inFreeze-drying. This milestone event was held in Strasbourg, France, onNovember79, 2007. Thankstothisconference, Iwasabletograspthelively,promisingnatureoffreeze-dryingtechnologyusedforbiologicalproductsandpharmaceuticals. Thepreservationoftheseproductsinvolveshugestakes: thequalityrequiredbytheprocessesusedandtheindustrial productionof suchproductsrequiregreat skill andpreciseknowledgeof theimpact of thepro-cesses on the components of the products handled. Freeze-drying, used alone orin conjunction with other processes such as ionization, enables the quality to bemaintainedatahighlevel. Itisinthebiologicalproductsandpharmaceuticalsectorsthatawiderangeofrefrigerationapplicationsarenotonlyneededbutarealsofast-evolving.This is why it is important to monitor progress in these different researchfields on a regular basis. Prof Rey and Dr. Mays remarkable expertise networkshave made it possible to gather contributions from all over the world to gain thegreatestinsightpossibleintothestateoftheart.Ihopeyouwillenjoythisbookandwillbecomeevenmoreinterestedinthistechnology.DidierCoulombDirector,InternationalInstituteofRefrigerationInstitutInternationalduFroid,Paris,Francev[ vi j ay] D: / i nf or ma_Publ i shi ng/ REY_H100116_2400039/ z_pr oduct i on/z_3B2_3D_files/978-1-4398-2575-4_CH0000_O.3d][5/5/010/12:37:50][114]This page intentionally left blank[ vi j ay] D: / i nf or ma_Publ i shi ng/ REY_H100116_2400039/ z_pr oduct i on/z_3B2_3D_files/978-1-4398-2575-4_CH0000_O.3d][5/5/010/12:37:50][114]PrefaceDrying an aqueous system from the frozen state by direct sublimation of ice hasbeenknownformorethana100years. However, foraverylongtimeit hasremained confidential and restricted to a handful of specialists trying to stabilizedelicatebiochemicalsintheirlaboratory. Then, duringwartime, it becameamajor process to provide human plasma to the Medical Corps on battlefields. Afewyearslater, freeze-drying, nownamedlyophilization, experiencedarock-etingdevelopmentwiththepreparationofantibioticsandmanydelicatephar-maceuticalsandbiological products. Todayitisawell-establishedtechnologywithlargeapplicationsundersteadyexpansion.Despitethisverywideuse, itremains, anyway, achallengingissuebothbecauseof theincreasingsensitivityof thetreatedproducts, mainlyproteins,andbecauseof themoreandmorestringent regulationsandconstraintsthatimpact the industrial process. In 1999, Joan May and myself already gathered inafirstbooksomeofthemainissueslinkedtoourbasicunderstandingofthefreeze-dryingof pharmaceuticalsandbiological products. Fiveyearslater, in2004, the situation had again evolved and it looked sensible to publish a secondedition, revised and enlarged. That situation prevails once more today and thisiswhyInformaHealthcarefeltthatitwouldbeappropriatetolaunchathirdedition of our book addressing both the basic science aspects and their industrialoutput. Wehavetakenupthat request withdeterminationknowingthat wewouldhavetheprivilegetogatheralargepanelofrenownedexperts.Inthat prospect we see that manyfundamental issues are still inthelimelight, suchas water andice abnormal properties, phase transitioninfrozenliquids, andmoreparticularlyglassbehaviour. Onanotherhand, moreand more emphasis is given today to the product/container/closure interactionand this implies more knowledge on the properties of glass and elastomers andontheirpotentialleachingoutofundesirablecomponentsintotheactivesub-stances. The structure of the dry plug is also an important factor in the long-timestorageandfinalreconstitutionabilityofthefreeze-driedproduct. Quiteobvi-ously the nature and amount of additives, bulking agents, and residual moistureare keyissues for thepreservation oflarge molecules as wellas for livingcells,henceformulationremainsofmajorconcern.Last but not least, the industrial operations are more and more under closescrutinyandrequirecontinuousmonitoringandcontrol. Most manufacturingprocessesbeingnowfullyautomated,developmentssuchasprocessanalyticaltechnology(PAT)orqualitybydesign(QbD)arebecomingleadingissues.To cover such a diversified field of competence was not an easy task, andJoanMayandIhavebeenquitefortunatetobeabletogrouptogether, inthisthirdedition, a great number of distinguishedspecialists whoare bringinggracefullytheirknowledgeandexperience.Altogetherithasbeenarewardingexercise for the editors, and we are pretty sure that it will equally be the case forourreaders.vii[ vi j ay] D: / i nf or ma_Publ i shi ng/ REY_H100116_2400039/ z_pr oduct i on/z_3B2_3D_files/978-1-4398-2575-4_CH0000_O.3d][5/5/010/12:37:50][114]Needless to say that all these efforts would have been unsuccessfulwithout theactivesupport of ourpublisher, andit isoursincerepleasuretoextend, here, ourdeepappreciationtoSandraBebermanwhohasbeen, since1999,thedrivingforceofthatenterprise.LouisReyJoanC.Mayviii Preface[ vi j ay] D: / i nf or ma_Publ i shi ng/ REY_H100116_2400039/ z_pr oduct i on/z_3B2_3D_files/978-1-4398-2575-4_CH0000_O.3d][5/5/010/12:37:50][114]ContentsForeword DidierCoulomb....vPreface....viiContributors....xi1. GlimpsesintotheRealmofFreeze-Drying:ClassicalIssuesandNewVentures 1LouisRey2. StructuralandDynamicPropertiesofBulkandConfinedWater 29Marie-ClaireBellissent-FunelandJoseTeixeira3. FreezingandAnnealingPhenomenainLyophilization 52JimA.Searles4. PhaseSeparationofFreeze-DriedAmorphousSolids:TheOccurrenceandDetectionofMultipleAmorphousPhasesinPharmaceuticalSystems 82AdoraM.PadillaandMichaelJ.Pikal5. TheUseofMicroscopy,ThermalAnalysis,andImpedanceMeasurementstoEstablishCriticalFormulationParametersforFreeze-DryingCycleDevelopment 112KevinR.WardandPaulMatejtschuk6. TheRelevanceofThermalPropertiesforImprovingFormulationandCycleDevelopment:ApplicationtoFreeze-DryingofProteins 136StephaniePassot,IoanCristianTrelea,Miche`leMarin,andFernandaFonseca7. Freezing-andDrying-InducedPerturbationsofProteinStructureandMechanismsofProteinProtectionbyStabilizingAdditives 167JohnF.Carpenter,Ken-ichiIzutsu,andTheodoreW.Randolph8. MechanismsofProteinStabilizationDuringFreeze-DryingStorage:TheRelativeImportanceofThermodynamicStabilizationandGlassyStateRelaxationDynamics 198MichaelJ.Pikal9. FormulationCharacterization 233D.Q.Wangix[gaj endra]D:/i nforma_Publ i shi ng/REY_H100116_2400039/z_producti on/z_3B2_3D_files/978-1-4398-2575-4_CH0000_O.3d][10/5/010/20:33:53][114]10. PracticalAspectsofFreeze-DryingofPharmaceuticalandBiologicalProductsUsingNonaqueousCosolventSystems 254DirkL.Teagarden,WeiWang,andDavidS.Baker11. RegulatoryControlofFreeze-DriedProducts:ImportanceandEvaluationofResidualMoisture 288JoanC.May12. Freeze-DryingofBiologicalStandards 317PaulMatejtschuk,MichelleStanley,andPaulJefferson13. MolecularMobilityofFreeze-DriedFormulationsasDeterminedbyNMRRelaxationTimes,andItsEffectonStorageStability 354SumieYoshiokaandYukioAso14. ScanningElectronMicroscopy:APowerfulToolforImagingFreeze-DriedMaterial 372TaunoJalanti15. PharmaceuticalPackagingforLyophilizationApplications 383ClaudiaDietrich,FlorianMaurer,HolgerRoehl,andWolfgangFrie16. ClosureandContainerConsiderationsinLyophilization 396FranL.DeGrazio17. ExtractablesandLeachablesasContainerClosureConsiderationsinLyophilization 413DianePaskietandFranL.DeGrazio18. OptimizingtheThroughputofFreeze-DryersWithinaConstrainedDesignSpace 425JimA.Searles19. MonitoringandControlofIndustrialFreeze-DryingOperations:TheChallengeofImplementingQuality-by-Design(QbD) 441MiquelGalan20. ProcessAnalyticalTechnologyinIndustrialFreeze-Drying 460AntonelloA.Barresi,DavideFissore,andDanieleL.Marchisio21. PracticalConsiderationsforFreeze-DryinginDualChamberPackageSystems 494DirkL.Teagarden,StanleyM.Speaker,SusanW.H.Martin,andThomasOsterberg22. LyophilizationandIrradiationasanIntegratedProcess 527DalalAoude-WernerandFlorentKuntzAppendix:Table,SaturationVaporPressureOfIce....541Index....545x Contents[ achary] D: / i nf orma_Publ i shi ng/ REY_H100116_2400039/ z_product i on/z_3B2_3D_files/978-1-4398-2575-4_CH0000_O.3d][7/5/010/12:30:47][114]ContributorsDalalAoude-Werner Aerial-crt,Illkirch,FranceYukioAso NationalInstituteofHealthSciences,Tokyo,JapanDavid S. Baker Pharmaceutical Sciences, QLT Plug Delivery Inc., Menlo Park,California,U.S.A.AntonelloA.Barresi DipartimentodiScienzadeiMaterialieIngegneriaChimica,PolitecnicodiTorino,Torino,ItalyMarie-Claire Bellissent-Funel Laboratoire Leon-Brillouin (CEA-CNRS), CEASaclay,Gif-sur-Yvette,FranceJohnF.Carpenter UniversityofColoradoHealthSciencesCenter,Denver,Colorado,U.S.A.Fran L. DeGrazio West Pharmaceutical Services, Inc., Lionville, Pennsylvania,U.S.A.ClaudiaDietrich SCHOTTformavitrumag,St.Gallen,SwitzerlandDavideFissore DipartimentodiScienzadeiMaterialieIngegneriaChimica,PolitecnicodiTorino,Torino,ItalyFernandaFonseca UMR782GenieetMicrobiologiedesProcedesAlimentaires,AgroParisTech,INRA,Thiverval-Grignon,FranceWolfgangFrie Ludwig-Maximilians-Universitaet,Munich,GermanyMiquelGalan TelstarTechnologies,S.L.,Terrassa,SpainKen-ichiIzutsu UniversityofColoradoHealthSciencesCenter,Denver,Colorado,U.S.A.TaunoJalanti MicroscanServiceSA,Chavannes-pre`s-Renens,SwitzerlandPaul Jefferson National Institute for Biological Standards and Control, HealthProtectionAgency,PottersBar,U.K.FlorentKuntz Aerial-crt,Illkirch,Francexi[ achary] D: / i nf orma_Publ i shi ng/ REY_H100116_2400039/ z_product i on/z_3B2_3D_files/978-1-4398-2575-4_CH0000_O.3d][7/5/010/12:30:47][114]DanieleL.Marchisio DipartimentodiScienzadeiMaterialieIngegneriaChimica,PolitecnicodiTorino,Torino,ItalyMiche`leMarin UMR782GenieetMicrobiologiedesProcedesAlimentaires,AgroParisTech,INRA,Thiverval-Grignon,FranceSusanW.H.Martin PfizerBiotherapeuticsPharmaceuticalSciences,Pfizer,Inc.,Chesterfield,Missouri,U.S.A.PaulMatejtschuk NationalInstituteforBiologicalStandardsandControl,HealthProtectionAgency,PottersBar,U.K.FlorianMaurer SCHOTTAG,Mainz,GermanyJoan C. May Center for Biologics Evaluation and Research, United States FoodandDrugAdministration,Rockville,Maryland,U.S.A.ThomasO sterberg PfizerProductandProcessDevelopment,GlobalManufacturing,PfizerHealthAB,Strangnas,SwedenAdoraM.Padilla KBIBiopharma,Inc.,Durham,NorthCarolina,U.S.A.DianePaskiet WestPharmaceuticalServices,Inc.,Lionville,Pennsylvania,U.S.A.Stephanie Passot UMR 782 Genie et Microbiologie des Procedes Alimentaires,AgroParisTech,INRA,Thiverval-Grignon,FranceMichaelJ.Pikal SchoolofPharmacy,UniversityofConnecticut,Storrs,Connecticut,U.S.A.TheodoreW.Randolph UniversityofColorado,Boulder,Colorado,U.S.A.LouisRey Aerial-crt,Illkirch,FranceandCabinetdEtudes,Lausanne,SwitzerlandHolgerRoehl SCHOTTAG,Mainz,GermanyJimA.Searles Aktiv-DryLLC,Boulder,Colorado,U.S.A.StanleyM.Speaker PfizerBiotherapeuticsPharmaceuticalSciences,Pfizer,Inc.,Chesterfield,Missouri,U.S.A.MichelleStanley NationalInstituteforBiologicalStandardsandControl,HealthProtectionAgency,PottersBar,U.K.Dirk L. Teagarden Pfizer Biotherapeutics Pharmaceutical Sciences, Pfizer, Inc.,Chesterfield,Missouri,U.S.A.xii Contributors[ vi j ay] D: / i nf or ma_Publ i shi ng/ REY_H100116_2400039/ z_pr oduct i on/z_3B2_3D_files/978-1-4398-2575-4_CH0000_O.3d][5/5/010/12:37:50][114]JoseTeixeira LaboratoireLeon-Brillouin(CEA-CNRS),CEASaclay,Gif-sur-Yvette,FranceIoanCristianTrelea UMR782GenieetMicrobiologiedesProcedesAlimentaires,AgroParisTech,INRA,Thiverval-Grignon,FranceD.Q.Wang BayerScheringPharma,Beijing,ChinaWeiWang PfizerBiotherapeuticsPharmaceuticalSciences,Pfizer,Inc.,Chesterfield,Missouri,U.S.A.KevinR.Ward BiopharmaTechnologyLtd.,Hampshire,U.K.SumieYoshioka UniversityofConnecticut,Storrs,Connecticut,U.S.A.Contributors xiii[ vi j ay] D: / i nf or ma_Publ i shi ng/ REY_H100116_2400039/ z_pr oduct i on/z_3B2_3D_files/978-1-4398-2575-4_CH0000_O.3d][5/5/010/12:37:50][114]This page intentionally left blank[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0001_O.3d][5/5/010/12:33:40][128]1 GlimpsesintotheRealmofFreeze-Drying:ClassicalIssuesandNewVenturesLouisReyAerial-crt,Illkirch,FranceandCabinetdEtudes,Lausanne,SwitzerlandINTRODUCTIONIn their 1906 paper to the Academie des Sciences in Paris, Bordas anddArsonval demonstrated for the first time that it was possible to dry a delicateproduct from the frozen state under moderate vacuum. In that state it would bestable at room temperature for a long time and the authors described, in a set ofsuccessive notes, that this technique could be applied to the preservation of seraand vaccines. Freeze-drying was then officially borne, despite its having been inuse centuries ago by the Inca who dried their frozen meat in the radiant heat ofthesunintherarifiedatmosphereoftheAltiplano.However, we had to wait until 1935 to witness a major development in thefieldwhen EarlW. Flosdorfandhiscoworkerspublishedsomevery importantresearch on what they called, at the time, lyophilization (this name was derivedfrom the term lyophile coming from the Greek l u o B and j i l e i u, which meanslikesthesolvent,describingthegreatabilityofthedryproducttorehydrateagain). Freeze-drying had then received a new name, which has been in currentusesincethen,togetherwithcryodesiccation.Manyauthors, indifferentcomprehensivebooksdedicatedtofreeze-drying,have already described in full detail the scientific history of this method, and we willnot attempt todoit again. Moreover, intheselast 75years, muchresearchandsubstantial development have been devoted to freeze-drying, and it would be of littleuse to list papers that are well known and available to all the specialists concerned.This, indeed, is the very reason why the present book has been designed topresent to the readers essentially new experimental methods and data, as well asrecent developments on our ownbasic understanding of the physical andchemicalmechanismsinvolvedincryodesiccation.Nevertheless,itwouldnotbefairtoskipthenamesofsomeofthegreatpioneers in the field. Earl Flosdorf, Ronald Greaves, and Francois Henaff foughtthe difficult battle of the mass production of freeze-dried human plasma, whichwasusedextensivelyduringWorldWarII. Tothat end, theyengineeredtheappropriate large-scale equipment. Sir Ernst Boris Chain, the Nobel Laureate forpenicillin, introducedfreeze-dryingforthepreparationofantibioticsandsen-sitivebiochemicals.IsidoreGershand,lateron,TokioNeiandFritjofSjostrandproducedremarkablephotographsofbiologicalstructurespreparedbyfreeze-drying for electron microscopy. Charles Merieux, on his side, opened wide newareasfortheindustrial productionofseraandvaccines. Inparallel, hedevel-opedabonebank, afirstmoveinafieldwheretheU.S. NavyMedicalCorpsinvestedheavilyafewyearslaterunderCaptainGeorgesHyatt.At the same time, cryobiology was getting its credentials withmanydevotedandgiftedscientistssuchasBasilLuyet, AlanParkes, AudreySmith,Author contact information: Cabinet dEtudes, Chemindeverdonnet 2, CH1010Lausanne, Switzerland;email: louis.rey@bluewin.ch1[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0001_O.3d][30/4/010/22:28:13][128]Harry Meryman, Christopher Polge, Peter Mazur, andothers. We hadtheprivilege of living this exciting period together with all these people since 1954,andmostofthemwerepresentinLyonin1958whenCharlesMerieuxandIopened the first International Course on Lyophilization, with the sad exceptionofEarlFlosdorfwhohadagreedtodelivertheopeningaddressbutdiedtrag-icallyafewweeksbeforetheconference.Today, 56 years later, we are pleased to see that freeze-drying still holds aremarkableplaceinourmultiplepanelofadvancedtechnologies,morepartic-ularlyinthepharmaceuticalfield.BASICFREEZE-DRYINGLyophilizationisamultistageoperationinwhich,quiteobviously,eachstepiscritical. The main actors of this scenario are all well known and should be understrictcontroltoachieveasuccessfuloperation.Theproduct, that is, theactivesubstance, whichneedstokeepitsprimeproperties.The surrounding medium and its complex cohort of bulking agents, stabilizers,emulsifiers, antioxidants,cryoprotectors, andmoisture-bufferingagents.Theequipment, whichneedstobeflexible, fullyreliable, andgearedtotheultimate goal (mass production of sterile/nonsterile drugs or ingredients,experimentalresearch,technicaldevelopment).Theprocess, whichhastobeadaptedtoindividual casesaccordingtothespecific requirements and low-temperature behavior of the differentproductsundertreatment.Thefinal conditioningandstorageparametersof thefinishedproduct, whichwill vary not only from one substance to another but also in relationship withitsexpectedtherapeuticlifeandmarketingconditions(i.e., vaccinesforremotetropical countries, international biological standards, etc.). Inotherwords, a freeze-dryer is not a conventional balance; it does not perform in thesamewaywithdifferentproducts.Thereisnouniversalrecipeforasuccessfulfreeze-dryingoperation,andtherepetitiveclaimthatthismaterialcannotbefreeze-driedhasnomeaninguntileachsuccessivestepoftheprocesshasbeen duly challenged with the product in a systematic and professional wayand notbytheall-too-common trial-and-error game.TheFreeze-DryingCycleIt is now well established that a freeze-drying operation includes the following:Theadhocpreparationof thematerial (solid, liquid, paste, emulsion) tobeprocessed,takinggreatcarenottoimpedeitsfundamentalproperties.The freezing step during which the material is hardened by low temperatures.During this very critical period all fluids present become solid bodies, eithercrystalline,amorphous,orglass.Mostoften,watergivesrisetoacomplexicenetwork, butitmightalsobeimbeddedinglassystructuresorremainmore or less firmlyboundwithinthe interstitial structures. Solutes doconcentrate and might finally crystallize out. At the same time, thevolumetric expansionof the systemmight induce powerful mechanicalstresses that combine with the osmotic shock given by the increasingconcentrationofinterstitialfluids.2 Rey[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0001_O.3d][30/4/010/22:28:13][128]The sublimation phase or primary drying will follow when the frozen material,placed under vacuum, is progressively heated to deliver enough energy forthe ice to sublimate. During this very critical period a correct balance has tobe adjusted between heat input (heat transfer) and water sublimation (masstransfer)sothatdryingcanproceedwithoutinducingadversereactionsinthe frozen material such as back melting, puffing, or collapse. A continuousand precise adjustment of the operating pressure is then compulsory to linktheheatinput totheevaporativepossibilities ofthefrozenmaterial.Thedesorptionphase or secondarydryingstartswhenicehasbeendistilledaway andthat a higher vacuumallows the progressive extractionofboundwateratabove-zerotemperatures.This,again,isnotaneasytasksinceoverdryingmight beasbadasunderdrying. Foreachproduct anappropriateresidual moisturehastobereachedunder giventempera-turesandpressures.Finalconditioningandstoragebeginswiththeextractionoftheproductfromthe equipment. During this operation great care has to be taken not to losetherefinedqualitiesthathave beenachievedduringtheprecedingsteps.Thus, for vials, stoppering under vacuumor neutral gas within thechamber is of current practice. For products inbulkor inampoules,extractionmightbedoneinatightgaschamberoranisolatorbyremoteoperation. Water, oxygen, light, and contaminants are all importantthreatsandneedtobemonitoredandcontrolled.Ultimatestoragehastobecarriedaccordingtothespecificsensitivitiesoftheproducts (at roomtemperature, 48C, 208C). Againuncontrolledexposurestowatervapor, oxygen(air), light, excessheat, ornonsterileenvironment are major factors to be considered. In that context thecompositionandqualityofthecontaineritself,(typeofglass,elastomersofthestoppers,plastic,ororganicmembranes)havetobeconsidered.At the end, we findthe reconstitutionphase. This canbe done inmanydifferent wayswithwater, balancedsalt solutions, orsolventseithertorestore the concentration of the initial product or to reach a moreconcentratedordilutedproduct.Forsurgicalgraftsorwounddressings,special procedures might be requested. It is also possible to use theproduct as such, in its dry state, in a subsequent solvent extraction processwhenverydilutebiochemicalshavetobeisolatedfromalargehydratedmass,asthisisthecaseformarineinvertebrates.Figure1summarizesthefreeze-dryingcycleandindicatesforeachstepthedifferentlimitsthathavetobetakenintoconsideration. Figure2givesanexampleofatypicalfreeze-dryingcycle.INSIGHTINTOTHEBEHAVIOROFPRODUCTSATLOWTEMPERATUREASensitiveIssue:TheFreezingStepThisinitial operationinthefreeze-dryingprocessis, of course, acritical onesince, iftheproductisimpairedabinitio, thereisobviouslynointeresttogofurther. We will see in the following paragraph what are the governingparameters, thermal, electric, and structural, but before it is worthwhile to makeafewcommentsonthewayfreezingisperformed.GlimpsesintotheRealmofFreeze-Drying 3[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0001_O.3d][30/4/010/22:28:13][128]FIGURE 1 Schematic evolution of the freeze-drying process. Temperatures (upper curve) and watercontent(lowercurve)versustimeareindicated.Abbreviations:Tcs,maximumtemperatureofcom-plete solidification; Tim, minimum temperature of incipient melting; Tim, absolute limit for fast process;Td,maximum allowed temperature for the dryproduct;RMF,final requested residual moisture.FIGURE 2 A typical freeze-drying cycle. Note that the pressure is raised to 0.3 mbar to increasetheheattransferduringsublimationandthenloweredto0.02mbarfordesorption.4 Rey[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0001_O.3d][30/4/010/22:28:13][128]Indilutedsolutions, whichisacurrent caseinfreeze-dryingof pharma-ceuticals, icecandevelopinthecourseofcoolingeitherasawell-definedfrontmoving upward in the liquid from the cold supporting shelf or as a brusk cloud ofindividual germs appearing at the same time in the whole mass of a supercooledfluid. In the first case, and especially when the cooling velocity is low, there mightbe some cryoconcentration of the product, which provokes a solute gradient frombottom to top resulting in an increasing solid concentration in the upper layer. Theresult is, most often, the occurrence of a thin film, rather compact, at the surface ofthe dryplugat the endof the process, whichmight create problems at thereconstitutionstepanddefinitelyimpedesthewatervapor(mass)transfer inthecourse of drying (Fig. 3). In the second case, when nucleation starts up all at oncethroughouttheliquidthestructureofthefrozen massismorehomogenous,andthis may lead to a more finely porous dry cake, but if the degree of supercooling isimportant and the ice development is pretty fast it may result also in the rupture ofthe vial. It is thus advisable in that particular case of freezing to measure ahead thecrystallizationvelocityofthesolutioninfunctionofthedegreeof supercooling(Fig.4)toget a bettercontroloftheprocess.Another concern is that supercooling might not be regular from one shelf tothe other one and even all over one single shelf. As a result nucleation may occur atdifferent temperatures and give different structures among the treated vials. Then itbecomesadvisabletocontrol thestructureofall productstoequallycontrol thenucleation step and make it simultaneous for the whole batch. The use of acousticof ultrasound stimulators seems to give anelegant solutiontothat problem.Anotherinterestingissueisthatifthefreezingisdelayed,foroperationalreasons, andtheliquidstandsfortoolongatimeincontact withtheinsidesurface of the vials, there might be an important interaction between the solutionandthe glass walls. Activesubstances might thenbeadsorbed, ofteninanirreversibleway, orelseglasschemicalsmight leakoutintotheliquid. Theseproblems will be addressed further in this book but they have to be taken care ofwhenhandling solutions in whichthe active substances are invery smallquantity(geneticallyengineeredbiochemicals)orassmall individual particles(bacteriaandvirusvaccines).Finally, the freezingprocess might beverydelicatetocarrywhentheinitial liquidis anunstable emulsionanda fast processing, makinguse ofcryogenics,mightbethendesirable.AnInterestingDevelopment:SoftIceTechnologyEverybody is familiar with the concept of sherbets and soft ice-creams. They arebasicallyfrozenplastic pastes, whichpresent ahighviscosityandrelativelymoderatenegativetemperatures. Theyarenotfreeflowingbutcanbestirredandmixedunder moderate mechanical strength andcan incorporate solidparticles like chocolate crumbs, fruit, nuts, andothers without losing theirstructure. Their rheological properties are rather complex and highly dependentontemperature.Theyaremostgenerallycommercializedassuchbutmightbealso the first step of a more elaborate process. For instance, in the freeze-dryingofcoffee,westartwiththeproductionofaviscousconcentratedcoffeeextract,whichisturnedintoaplasticicyfoambyinjectionof carbondioxideunderpressureinarotativescrapedsurfaceheatexchanger. There,theliquidiscon-vertedintoasherbet-likefoamypaste, whichis spreadover ametallic beltGlimpsesintotheRealmofFreeze-Drying 5[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0001_O.3d][30/4/010/22:28:13][128]FIGURE 3 (A) Scanning EM of the upper surface of a freeze-dried pellet showing a membrane-likeskinwithelongatedventsfortheescapeofwatervapor.(B)ScanningEMoftheinsidemiddlepart of afreeze-driedplugshowingtheextremelyporousmaterial resultingfromthesublimationofice-crystals.Abbreviation:EM,electronmicrograph(Microscan).6 Rey[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0001_O.3d][30/4/010/22:28:13][128]conveyorandfrozenhardinacoldroom. Thenitisbroken, groundorsliced,sifted,andfinallyfedintothefreeze-dryingplant.Inthepreparationofmanypharmaceuticalsweknowthatsomecompo-nentsintheformulationarenot compatibletogether. Forinstance, if wemixacetyl salicylic acidwithsodiumbicarbonate insolution, thereisan immediatereactionandavigorousreleaseofcarbondioxide.Nevertheless, itmightbeofinterest to freeze-dry them together to have instant sparkling aspirin, but this isnot possible, andthe only way around thisis tocompact theproductstogetherinthedrystate. Theresultingtabletisnotverystableandittakesarelativelylongtimetogetbackintosolution.Conversely,ifsuchreactive substancesaremixedtogetheratlowtempera-ture after incorporation in a soft ice they will not react and the resulting paste canbemoldedinappropriateshapesandhardenedbyfurthercooling.Thematerialcanthen be freeze-dried withoutdifficulties.Inthatstate,lyophilized aspirin, forinstance, isperfectlystableandwhenwaterisaddedbackit reconstitutesasasparklingfluidinafewsecondsbecauseofits highporosity.Thesofticetechnologyisthusaveryprecioustooltopreparecomplexproducts, mostoftenfororalroute. Purechemicals, drugs, vitamins, andmin-eral saltscanbesuccessfullyfreeze-driedinthat wayinratherelaboratefor-mulations since it is possible to mix together sherbets-lines issued fromdifferentsolutionsandevenaddtothewholeothersolidingredientsasfinelydispersedpowders. Thekeytosuccessinthatprocessisagoodcontrolofthetemperatureoftheicypaste,whichisgenerallypreparedinacylindricaldou-ble-wallheatexchangerwithacontinuousscrapedsurfacemaintainedattem-peratures between 48Cand 208Cdependingonthenatureof thetreatedproducts. Whenthesoft icemixtureisdulycompletedit canbemoldedbyconventional equipment inplasticblistersandfrozenhardinablast tunnel.FIGURE4 Crystallizationvelocityofasolutionversusitsdegreeofsupercooling.GlimpsesintotheRealmofFreeze-Drying 7[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0001_O.3d][30/4/010/22:28:13][128]Thenthismaterial canenterthefreeze-dryer. Altogether, itisarathersimpleonlinegalenicprocess.ThermalandElectricPropertiesA fundamental paradigm of freeze-drying is to understand that, in almost everycase, there is no direct correlation between the structure of a frozen product anditstemperaturesinceallstructuralfeaturesdependessentiallyonthethermalhistory of the material. In otherwords, theknowledge of the temperature of afrozensolutionisnotenoughtoallowtheoperatortoknowitsstructuresincethislatterdepends,essentially, onthewaythistemperaturehasbeenreached,thatis,onthefreezingcycle.For instance, wedidshowin1960that anaqueoussolutionof sodiumchlorideat 258Ccouldbeeitherasponge-likeicenetworksoakedwithhighlyconcentratedfluidsifthesystemhasbeencooledprogressivelyfrom 208Cto 258Coratotallyfrozensolidwherealltheinterstitialfluidshavecrystallizedaseutectics if the system has been cooled, first to 408C and then rewarmedto 258CActually, whenwater separates as pure ice, as is the case for dilutedsolutions, there might be a considerable degree of supercooling in the remaininginterstitial fluids. It isthencompulsorytogotomuchlowertemperaturestorupture thesemetastablestates andprovoke theirseparationas solidphases.This, indeed, has a great significance because it is precisely within thosehypertonicconcentratedfluidsthattheactivesubstancessoak,whethertheyarevirusparticles, bacteria, ordelicateproteins, andtheretheycanundergoseriousalterationsinthisaggressiveenvironment. Thisisthereasonwhyweadvisedtocool theproduct at sufficientlylowlevelstoreachwhat wehavecalled Tcs, the maximum temperature of complete solidification. However, whenfrozen,thematerialwillonlystarttomeltwhenitreacheseithereutectictem-perature or what we called Tim, the minimum temperature of incipient melting.Differential thermal analysis (DTA) and differential scanning calorimetry(DSC) are useful techniques to proceed to this determination. They can be veryadvantageouslycoupledwithlowfrequency(LF)electricmeasurementssincetheimpedanceofthefrozensystemdropsinaspectacularwaywhenmeltingoccurs (Figs. 5and6). Inother terms, ahighelectric impedance is alwaysrelatedtoastateof utter rigidity. Moreover, theelectric measurements aremore reliable thanDTA/DSCalone. Indeed, whenwe are dealing withacomplexsystemmoreparticularlywhenit containshighmolecular weightcompoundsthe material hardens progressively during freezing and, often, inthecourseofrewarmingshowsincipientmeltingonlyatrelativelyhightem-peratures. Until that point the DTA curve remains silent. Unfortunately, thisis not always a sign of absolute stability since, quite often, a softening of thestructureappearsmuchearlier. It isourexperiencethat asharpdecreaseinelectric impedance is a clear warningfor anoperator whoshouldtryandmaintaintheproductduringprimarydryingattemperaturesbelowthislimit.Figure7demonstratesthisphenomenoninthecaseofU.S. standardforper-tussis vaccine lot 9, and Figure 8 shows the same behavior for the Saizen mass10 mg from Serono. In both cases it is quite obvious that sublimation has to be8 Rey[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0001_O.3d][30/4/010/22:28:13][128]done at temperatures much lower than those, which could be derived from theDTAcurvesalone.The evolution of electric properties is, thus, at least for us, a major parameterinthedesignandfollow-upofafreeze-dryingcycle.Itcanevenbeusedforanautomatic control of the whole operation, as we proposed in the past.FIGURE 5 Differential thermal analysis (DTA) and impedance (1000 Hz Z sin j) of a 2 p. 100solutionofCl Nainwaterduringcontrolledfreezing.FIGURE 6 Differential thermal analysis (DTA) and impedance (1000 Hz Z sin j) of a 2 p. 100solutionofCl Nainwaterduringcontrolledrewarming.GlimpsesintotheRealmofFreeze-Drying 9[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0001_O.3d][30/4/010/22:28:13][128]GlassandVitreousTransformationsIt is well known that certain solutions/systems do not crystallize when they arecooleddown, especiallywhenthisis doneinarapidprocess(quenchingincryogenic fluids has longbeeninuse byelectronmicroscopyspecialists toprepare their samples). Cooling results, in that particular case, in the formationof a hard material that has the properties of a glass. This vitreous body can proveto be stable, and when rewarmed it softens progressively and goes back to liquidFIGURE 7 Differential thermal analysis (DTA) and impedance (1000 Hz Z sin j) of U.S. FDAstandardpertussisvaccinelot9.FIGURE8 Differential thermal analysis(DTA)andimpedance(1000Hz Zsin j)ofSeronoSaizenmass10mg.10 Rey[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0001_O.3d][30/4/010/22:28:13][128]again. However, inthecourseof rewarmingit generallyundergoes amajorstructuralchangeknownasthevitreoustransformation. Duringthisprocesstheglass, stillinthesolidphase, evolvesfromasolid-likestate(withlowspecificheat andlowspecificvolume) toaliquid-likestatewithasharpincreaseinspecific heat and specific volume. This occurs at a well-known temperature, thetemperature of vitreous transformation, generally quoted as Tg. Atypicalexample, pure glycerol, is shown in Figure 9. This phenomenon is fullyreversible.Insomeothercasestheglass, whenrewarmed, becomeshighlyunstable,andwhenthe vitreous transformationis completed(or sometimes duringthistransformationitself)itcrystallizesoutabruptlyshowingamarkedexothermicpeak. This, to the contrary, is an irreversible process called devitrification (Fig. 10).Alloftheglassmightthendisappearoritmightdevitrifyonlypartially. Ifthematerial isthencooledagainandrewarmedasecondtime, theinitialvitreoustransformation disappears or is substantially reduced (and sometimes shifted to ahigher temperature) and the exothermic peak no longer exists. During this cycle,whichwecalledthermal treatmentin1960, themetastablesystembecameunstableandhasbeenannealed.Figure11showsthistypeofevolutioninaglycerol-water-Cl Na system.Itisworthwhilementioningthatinbothcasesthevitreoustransformationinitiatesamarkeddecreaseof theelectricimpedance,thoughatfirstsightthesystemremainsacompactsolid.Themagnitudeandtemperatureoftheseeventsareofcoursedependenton the system under investigation, as can be seen in Figure 12, which comparesD2O/glycerolandH2O/glycerolmixtures.Thistypeofbehaviorisnotuncommoninpharmaceutical preparations.Indeed, it is often required that a thermal treatment be applied to the material toFIGURE9 Differential thermalanalysisandimpedance(1000Hz)ofpureglycerol.GlimpsesintotheRealmofFreeze-Drying 11[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0001_O.3d][30/4/010/22:28:13][128]FIGURE10 Theoretical diagramof thelow-temperaturebehavior of asystemsusceptibletopresentglassformation.Specificheathasbeentakenasthemarker.FIGURE 11 Low-temperature behavior of a 50:50 glycerol-H2O system containing 10 per 1000 ClNa. RW1: recordingsof differential thermal analysisandZsinjduringthefirst rewarmingfrom1968C to 958C. When the devitrification peak is completed (958C), the sample is immediatelycooled backto 1968C.RW2:Secondanalysis from 1968Cto 208Cafterthermaltreatment.12 Rey[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0001_O.3d][30/4/010/22:28:13][128]secure a steady and successful freeze-drying operation. In the example depictedin Figure 13A and B, it was discovered that a refined biochemical (recombinantglycoprotein-Mol-2 fromthe Center for Molecular Immunology inHavana,Cuba)neededtobestabilizedbythermaltreatmenttoraisethesofteningtem-perature (as witnessed by the evolution of the electric impedance) from328C to198Candallowaneasyfreeze-drying.Inotherinstancesasthiswillbedevelopedlaterinthisbookbyseveralauthors(Randolph,Searles,Carpenter,andPikal)itappearsessentialtosafe-guardtheglassystate, whichisindeedanabsoluteprerequisitetosecurethetertiarystructure of active proteins. Here againa precise knowledge of theboundariesandsensitivitiesofthisglassisdefinitelyneeded.Finally,letusmentiontheinterestingcaseofthesucrose/watersystems,which after drying gives a very sensitive porous cake susceptible to undergoingadverse transformations at above-zero temperatures and in the dry state because oftheexistenceof aratherhighTg(578C). It is, thus, compulsorytokeeptheseproductsattemperatureslowenoughtopreventtheshrinkageofthepelletbywhathasbeencalledarubberbehavior.BULKING AGENTS: A MATTER OF CONCERN IN SUBLIMATIONThis is, indeed, the heart of the lyophilization process: the step during which thefrozenproduct is driedfroma solidphaseby direct sublimationof ice.During thatphase thedryingboundary sinks intothefrozenmaterial from toptobottomorelsefromoutsidetoinsideleavinganupper layerorsurroundingshelloffineporousmaterial. Theresultingdrymatrixisthuscarvedfromthefrozenmassundervacuum. Toachievethisoperationthisfrozenmasshastokeep a structure when ice disappears and water vapor moves away. To that end,FIGURE 12 Differentialthermal analysis of50:50/100 mixtures ofglycerolandD2O/H2O.Bothnormal andheavywater-glycerol mixturesdemonstratethesamebehaviorbut thetransitiontemperaturesforD2Oareshiftedupwardby4.58C.GlimpsesintotheRealmofFreeze-Drying 13[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0001_O.3d][4/5/010/9:22:50][128]FIGURE13 (A, B) Recombinant glycoprotein-Mol-2. 1416RW: Inapreliminaryanalysistwoslight exothermicphenomenawereobservedbyDTAat 25.58Cand 16.48C, followedbyaprogressive melting starting around88C. However, the electrical measurements alreadyshowedamarkeddecreaseinrigiditystartingat 328C. 1417RW1: Thermal treatment wasappliedbycoolingthesolutiondownfrom 208Cto 508C, thenrewarmingto 218C. Thethermal curveisalmost identical tothe1416RWandthefirst exothermicpeakshowsup. Thesystemisthencooledbackfrom 218Cto 508Candrewarmedagainto 188C. 1417RW2shows, during the second rewarming, that the first exothermic phenomenon has disappeared. Asa consequence, the Z sin j is by far higher 1417RW2 and rigidity good until 208C. To improve,once more, the low-temperature stability of the system, it is cooled back a third time from 188Cto 508Candthenrewarmedfor analysis. 1417RW3showsthat, inthecourseof thethirdrewarming, all accidents have disappeared on the DTA curve, which only indicates progressivemelting from 88C upward. Moreover, the stability of the frozen structure has been substantiallyimproved sincenowZsin jremains almost constant during thewholeperiod andonlystarts todropat 198Cinsteadof 328Cwithout thermal treatment. It istheneasytofreeze-drythesolutionat 208Corslightlybelow.Abbreviation:DTA,differential thermal analysis.14[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0001_O.3d][30/4/010/22:28:13][128]bulking agents have to be added to build that structure if the initial solution istoo diluted. Lactose, sucrose, mannitol, PVP, dextrose, methyl cellulose, glycine,andmanyothershavebeenusedtofulfillthistask.NewCarriersIn somecases, however,the introductioninto theformulationof theseforeigncompoundsmightbeundesirableand,ideally,itwouldbebyfarbettertodowithout themandhandle the active substance alone. Aninterestingoptionmight then be to enclose the liquid to be dried into a porous matrix where it willbekeptinthecourseofdryingandthuspreventedtoflyawaywiththewatervaporstream. Porouspolymers, sinteredmetals, ceramics, porousglass, inor-ganic textiles, multilamellar pads, even small silica or glass beads, etc., could, atfirst glance, beresortedprovidedtheydemonstratesimultaneouslyacertainnumberofproperties.Theycanadsorbtheliquidsolutioneasilyasahydrophilicmaterial andwithstandfreezinganddryingwithoutmechanicalrupture.Theydonotinteractwithanyelementoftheformulation.Theyarecleananddeprivedofresidues,particles,orcontaminants.They hold enough liquid per unit volume, which means that they present atleasta30%to50%porosity.Theycanbe shapedas well-definedgeometrical units: disks, rods, andspherestobeincorporatedintothevial.Their pore structure is thin enough to hold the solution but wide enough toletthe watervapor escapefrom thefrozen liquid, which means an open-porestructure,aspongewithinterconnectinginterstitialchannels.Theyreleasetheadsorbedactiveproductwhentheyarefloodedwiththereconstitutionfluid.To our own experience, fulfilling these requirements is not an easy task and,actually, very few products are susceptible to provide this complete set of prop-erties. While most of the structural issues can be solved, the most difficult, by far,remains the latter one: The carrier should release the totality or at least a majorknown amount of the active substance at the time of reconstitution. To that end,developments have been made to exhaust the carrier by percolating through itthe dissolution fluid under pressure, as would be done with a conventional onlinefilter. Special syringes have been manufactured where the original solution to bedriedis pumpedthroughthe carrier, thenallowedtodrythere, andfinallyextracted by the reconstitution fluid using the same type of mechanism.Today, theinert carrierissueisstill underdevelopment, andit ismorethan likely that the enormous amount of research, which is currently devoted tonew materials, whether glass and glass derivatives, polymers, fibers, etc., mighthelp ustofindnewsupportswith almostnil adsorptionpropertiesandstillanopenporousstructure.OPERATINGPRESSURENumerousbooksandpapershavebeenwrittenonthephysicsofprimaryandsecondarydrying. However, wewouldliketostressonesinglepointthathasnot always been clearly understoodthe centralrole of the operating pressure.GlimpsesintotheRealmofFreeze-Drying 15[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0001_O.3d][30/4/010/22:28:13][128]Indeed, aswehaveshowninFigures1and2, freeze-dryingdeservesanaccurate andcontinuous control of the pressure inthe chamber duringthewholeoperation.With the exception of vacuum freezing (or snap-freezing generally used inthe food industry), the initial cooling is always done at atmosphericpressure,sometimesinaseparatecabinet.Primary drying, to the contrary, is performed under vacuum. Whereas in theyoung days of the technique it was felt by most specialists that the higher thevacuumthe better the process could be. It was shown by Neumann and Oetjenet al. that throttlingthe water vapor flowbetweenthe chamber andthecondenser was increasing the speed and efficiency of the operations. Later on,RieutordandIpatentedtheairinjectionprocess,whichcouldbeappliedtoany equipment whether the condenser coils were placed in a separate chamberor in the drying cabinet itself. Since then, the air bleed has been used almostuniversally sincea substantial risein the pressure(0.10.5 mbar according tothe Timof the treatedproduct) provedtoincrease considerablythe heattransfer to the sublimation interface essentially by gas conduction-convection.As a consequence, the temperature of the heaters couldbe substantiallyreduced, which prevented melting of the still-frozen core and/or scorching ofthe already existing dry layer. Monitoring and control of this pressure is still anessential part of the process, and they can be geared to the intrinsic propertiesof the product such as its temperature or, better, its electric impedance. At anyrate, during this whole sublimation period, the vacuum level is the master keyto the heat transfer and can help to rescue a product that is becoming too hotandstartingtosoftendangerously. Indeed, insuchasituation, pullingthevacuumdown immediately works as a real thermal switch with aninstantaneous result, whereas cooling the shelves requires a longer time.Secondarydryingisgenerallycarriedout under higher vacuumwhentheproduct has reached an above-zero temperature (or its electric impedancehas reachedthe upper limit). Indeed, it has beenshownbydifferentauthors that isothermal desorptionwas faster andallowedlower finalresidualmoistureswhenthepressurelayinthelevelof102mbarbutthathighervacuums(103mbar)didnotdrasticallyimprovetheoperation.Final conditioninghasalwaysbeen muchdiscussedand,inthelastdecades,vacuum and dry neutral gas both got their supporters. Today, stopperingthe vials under a slightly reduced pressure of dry nitrogen gas looks to bethefavoriteoption. Someexperimentsthat wedidinthepast andthatremainedunpublishedpushustothinkthatstopperingunderdry argoncould give better results for long-term storage, as is the case, for instance,forinternationalbiologicalstandards.SOMECHALLENGINGWAYSTOINVESTIGATETHEFINALDRYPRODUCTThe dry freeze-dried cake undoubtedly has a very peculiar structure since, as wehavealreadyseenbefore, it has beencarvedunder vacuumfromasolidmatrix(Fig.3).Attheendofthisprocessandwhenitisstillundervacuumitsinternalsurfaceisquitecleanandveryreactive. Itcanbeeasilyunderstoodthatits firstcontactwitha foreignelement,suchas thegas used torupturethevacuum, is determinant. This is, in fact, why the choice of this gas as well as the16 Rey[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0001_O.3d][30/4/010/22:28:13][128]procedure to introduce it are critical. In that context the structure of the dry cakeandthepropertiesofitsinternalsurfaceconstituteamajorelement.Another important factor is the level of the residual moisture in the pelletanditsevolution,ifany,duringstorage.Finally, it is definitely of interest to know how this internal water is boundtothestructure.Theseare three recurrentproblems infreeze-drying, amongmany others.Weshallproposesomeselectedexperimentalmeansfortheirinvestigation.InternalSurface:BETApplyingthe Brunner, Emmett, andTeller (BET) methodto pharmaceuticals,whetheringredientsorfinishedproducts, isof commonuseinpowdersforthedetermination of their specific area and pore sizes. However, generally, the operatorhasareasonableamount of substanceinhands, andmost oftenthismaterial isrelatively resistant to water vapor and can be manipulated without too many risksin the openspace of a laboratoryor, if needed, withina conventional glovebox.The situation is quite different with sensitive freeze-dried specimenscontainedinsealedvials or ampoules, whichpresent botha veryreducedweightand, ofcourse, averylowdensityaswellasahighhygroscopicity. Inthat case, the prolonged contactwith moist air can provoke a real collapse ofthe internal structure, which cannot be restored to its original state by prolongedpumping. Then a very precisemethodology hasto be followedto get a reliablemeasurement.Wewouldliketoexplainhowweproceedtothatend.The sample to be checked (5100 mg of dry product in a sealed vial or ampoule)is quickly opened with a diamond saw and the bottom part, containing theplug, placed in a special stainless steel BET cell that is immediately capped,weighted, and connected to the manifold of the pumping device. Vacuum isthen pulled overthesample at room temperature.Two to three days later vacuum is broken with dry nitrogen gas and the BETcellconnectedtoanautomatic,computerized analyzer(Den-Ar-Mat1000).Vacuumis pulled again and repetitively checked until the leak rate (combinedwiththe gas release fromthe product) falls to the order of 23 104mbar/sec.The BET cell is then flooded with helium for the determination of the samplevolume,whichallowsthedeterminationofitsrealdensityandporosity,sincewealreadyknowbothitsweightanditsapparentvolume.TheBETcellisagainpumpedandimmersedinliquidnitrogen.Whentemperatureandpressurearestable, knownamountsofheliumareintroduced in the measuring chamber; thanks to special gas-tightmicrosyringes(7mLfor thesmallest one) toperformthecalibrationofthecavitybeforeadsorption.The BETcell ispumped againandthen themeasurementcan start.Knownamounts of the adsorption gas are then introduced in the specimenchamber and each time we wait for a steady equilibrium. The adsorptionisothermis,thus,constructedprogressivelythroughouttheBETrange.When saturation is reached, the BET cell is slowly pumped down, stepwise,toensurecontrolleddesorptiontomeasureporesizeanddistribution.Forabsolutesurfacesrangingfrom50to5m2, themeasuringgastobeused is nitrogen. For absolute surfaces of 5 to 0.5 m2, the measuring gas is argon.For absolute surfaces below this level we use krypton. In the latter case, the onlyGlimpsesintotheRealmofFreeze-Drying 17[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0001_O.3d][30/4/010/22:28:13][128]possible determination is that of the specific area since, with krypton being solidat liquidN2temperature, it isnot feasibletoperformacompleteadsorption-desorptioncyclefortheevaluationofporedistributionandsizes. Forobviousreasons, kryptonmeasurement isboundtobecometheroutineonefor mostfreeze-driedpharmaceuticals.Figure 14, for instance, represents the krypton adsorption curve for a sampleof Gonal F75 UI fromSerono, and we can see that it gives a pretty accurate readingdespite the fact that the absolute surface measured is only 0.07 m2.It is beyond the scope of this chapter to indulge in more details about thistechnique, but it might be of interest to note thatit allowsthe evaluation of theinfluence, ontheproductstructure, ofdifferentfactorssuchastheinitialcon-centrationofthestartingfluid, therateoffreezing(macroormicrocrystallinestructures),thepertinenceofthedryingcycle,andofthefinalconditioning.Sincemanyfunctionalpropertiesofthedrycakederivefromitsinternalsurfaceandporositysuchas solubility, oxygen, andwater sensitivity, thesemeasurements might be of great interest, especially for the assessment ofreliabilityofanewprocessinviewofitspotentialvalidation.Some people will ask why we did not use penetration by mercuryfor themeasurement of pores. Our answer is that for delicate, flexible products that canbe easilycrushedor simplydistortedintheir morphologybyoutside con-straints, we found that the application on the cake of pressures ranging to a fewthousandbarscouldbeconsideredirrelevant.EquilibriumWaterVaporMeasurementinaSealedVialFor the reasons that we have just described, the measurement of residualmoisture within an isolated vial or ampoule is a difficult undertaking. Joan May,in this book, also deals with this topic, and she explains how the FDA developedFIGURE14 Determinationof thespecificareaof freeze-driedSeronoGonal F75UI byBET(sampleweight0.0325g).18 Rey[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0001_O.3d][30/4/010/22:28:13][128]very accurate standardized techniques to execute these determinations bychemical Karl Fischer titration or thermo gravimetric techniques. The onlydrawbackisthat,inbothcases,thematerialisdestroyed.It is, in fact, because of that specific point that our approach has followed adifferentroute.Useanonintrusivemethodleavingtheproductintact.Beable,underthoseconditions,todorepetitivemeasurementsonthesamevialinthecourseoftimetoassessitsevolutionduringstorage.Figure15showstheprincipleof themethodwedevelopedjointlywithJ. Mosnier(deceased). Theselectedvialorampouleisplacedinathermostaticmetallicblockwheretemperaturecanbemaintainedconstant from 108Cto608C. Ononesideof theblockametallicfinger,havingat itsendatinypolished stainless steel cylindrical mirror (2 6 mm), is pushed in close contactwith the outside wall of the glass vessel and the connection is made optically andthermallytight,thankstoatraceofsilicongrease.Ontheothersideisplacedanear-infrared diode that beams light throughout the vial on the mirror, which, inturn,reflectsthatlightonaphototransistorplacedinasymmetricalposition.Whenthetemperatureofthemirrorisequaltoorhigherthanthatofthemetallic block (hence the temperature of the sample) the level of reflected light issteady. Tostart measurement the temperature of the mirror is progressivelydecreased, thanks to combined Peltier and cold nitrogen gas. Because of the goodthermal contact on a very limited surface (12 mm2) the internal wall of the flask isequallycooledonasmall isolatedspot. When, finally, thetemperatureof themirror (and the temperature of the spot on the internal surface) reaches the valuecorresponding to the saturated water vapor pressure within the flask, some dew(or ice) is deposited on the internal wall that immediately becomes diffusive. Thereflectedlightdropssharply.Themeasurementisdone.Themirrorcanthusbewarmedagain, dewevaporates, lightisrestored,andthesamplecanbeplacedback into the storage cabinet for further determination. Figure 16 gives anFIGURE15 Operatingprincipleof thedevicemadetomeasuretheequilibriumwater vaporpressureinsideasealedvial.GlimpsesintotheRealmofFreeze-Drying 19[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0001_O.3d][30/4/010/22:28:13][128]example of this determination for a sample of Novartis calcitonin dry substance ina sealedampoule measuredat +258C. At the corresponding temperature of15.88Cthewatervaporpressureinside theampouleis1.53 mbar.Thesensitivityof thismethodisveryhighsincemeasurementsarestillpossibleforcondensationtemperaturesof 608C, which, inavial of 10cm3,means thatless than 0.1 mgof wateris present in the overhead spaceabove thepellet.Overthelast15yearssystematicdeterminationshavebeendoneincol-laboration withDr.May fromFDA andwithseveraldrug companieson a vastnumber of products. Most of them have been followed during storage (at 208C,48C, or room temperature) for more than two years. Dr May reports on some ofherfindingslaterinthisbook.Without going into too much detail, let us mention some of the key issuesofthisparticularwork.Insealedglassampoulesandforthesameproduct wefoundsubstantialdifferences in the equilibrium water vapor pressures according to batchesand, within a given batch, according to the position of the ampoule on theshelves of the freeze-dryer. The so-calledwall, door, condenser, andcornereffectscouldthenbeanalyzedandamended.Insealedglassampouleswealsowitnessedanevolutioninthecourseofstorage showing that often the remaining water was restructured in thecourseoftimegoingfromarathermobileformtoamorefirmlyboundform,displayingalowerwatervaporequilibriumpressure(Fig.17).In stoppered vials we could quite easily follow the transfer of moisture fromthe stopper toward the plug, which acted as a getter pump until a newequilibriumwas reached(Fig. 17). We couldestablish, then, a directcorrelation between the water vapor pressure measurements and the KarlFischertitration(Fig.18).FIGURE16 Novartiscalcitonindrysubstanceinampoules.20 Rey[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0001_O.3d][30/4/010/22:28:13][128]Sometimes our method helped the pharmaceutical technicians to modify theformula of the excipients and stabilize the water content. They could alsobetterunderstandtheas-yet-unknownbehaviorofcertainadditiveslikeheavypolyalcohols,glycerol,ortensioactiveagents.FIGURE 17 Evolution of the condensation temperatures of freeze-dried Cibacalcin and Revasc(Novartis)inthecourseofstorageat208C.FIGURE18 Evolutionwithtimeofthecondensationtemperatureofthevial gasphaseandoftheKarl Fisherresidualmoisture(measuredonanalternatesample)ofafreeze-driedvaccine.GlimpsesintotheRealmofFreeze-Drying 21[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0001_O.3d][30/4/010/22:28:13][128]Today we are convinced that there are numerous applications for this newanalytical tool, especially when controls have to be performed on rare, high-cost,or infectious material for which a nonintrusive remote technique is compulsoryor that repetitive measurements needtobe done onthe same sample overprolongedperiods.Low-TemperatureThermoluminescenceofDryMaterialThermo luminescence analysis has been applied to solids for quite a long time inarcheologicalresearchandradiationdosimetry. Inbothcasesthesamplesthathavebeenactivatedbynatural and/orman-maderadiationsourcespresentacertain number of energy traps that can be emptied by heating. For the datationofceramics, theheatingcycleisextremelyfastandbringsthesampletomorethan 3008C in less than a couple of minutes. The magnitude of the light emissionunder well-standardized conditions is directly linked to the age of the material,whichcanbe determinedover several centuries withanaccuracyof afewpercentagepoints.For radiationdosimetry, withinor aroundnuclear plants as well asinopengreenfieldsandintheenvironmentatlarge,theanalystsmakeuse,gen-erally, of so-called thermoluminescent solids that are susceptible to activationoveralargespanofdoses.Espagnanetal.developedarefinedmethodtothatendusinglithiumfluoridetablets, whicharefirsttotallydeactivatedat4008Cand then exposed to the radiation field at ambient temperature. Reading is donebyrecordingtheemissionpeaksduringtheir progressiveextractioninthecourse of a constant-speed (18C/sec) heating, which generally shows upbetween2008Cand4008C.Attheonset ofourresearchwork, wetriedtooperateinthesamewaywith freeze-dried productsthathad received a strong g-irradiation (4060 kGy)at room temperature. The results were disappointing because we were limited toratherlowtemperatures(below1008C)and, evenso, whenthematerialcouldstandhightemperatureslikefreeze-driedsilica. Thisiswhyweattemptedtoapply to these products the same methodology that we used previously for ourinvestigationonwaterandsolutions.Thedrysamples areirradiatedat 1968Canddeactivatedbyagentlerewarming (1.58C/min) to roomtemperature. Our preliminary results lookpromising(Fig.19).Figure 20, for instance, shows the behavior of a freeze-driedplug ofmannitol (residual moisture1.5%). Wecanseethat, incomparisonwiththeinitial solution (at 10% mannitol), the thermoluminescence of the dry material issome 20 times stronger and does not follow the same pattern. The lower peak (at1528C)isverynarrowwhereasthehigherone(near 1188C)issubstantiallydepressed. Thisisnot surprisingif weassumethat thefirst peakisdirectlyconnected to the water molecule itself while the second one seems to be gearedtothethree-dimensionalnetworkofthewatermoleculeswithinthecrystallinelattice.Additionalstudiesdoneonfreeze-driedsilicagels(microbeadsgelsfromRhodia, Salindres) give some support to this idea. In Figure 21, for instance, wecanseethat whenweshift fromalargelyhydratedmaterial (likethefreshstarting material at around 40% residual moisture) to the low residual moisture22 Rey[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0001_O.3d][30/4/010/22:28:13][128]FIGURE19 Low-temperaturethermoluminescencesetup.Apreviouslyirradiatedfrozensam-ple placed in a liquid N2 cryostat emits light during controlled rewarming. The corresponding glowisrecordedbyaphotomultiplierversustemperature.FIGURE20 Thermoluminescenceof a10p. 100mannitol-H2Osystemafter g-irradiationat1968C(30kGy).GlimpsesintotheRealmofFreeze-Drying 23[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0001_O.3d][30/4/010/22:28:13][128]freeze-dried gels, the 1208C/1008C emission is first almost completely erasedandthendisappears (for 2%residual moisture). At the same time the low-temperaturepeakincreasesandsharpens.We thinkthat low-temperature thermoluminescence appliedtofreeze-driedsolids is susceptibletogiveus newinformationonthewater trapswithin the solidmaterial,andthatwith a better knowledge ofthe temperature,magnitude, andshapeoftheemissionpeakswemightbeabletobeamsomelight onthe role of residual moisture infreeze-driedproducts andonthemechanismsbindingwatermoleculestotheirsupportingmatrix.UNCONVENTIONALDEVELOPMENTS:NEWPROSPECTSWaterIsNottheOnlyAvailableSolventUntil now, almost all lyophilization operations have been carried with aqueoussolutions, but we know since a long time that water is not the only solvent thatcansublimatefromthefrozenstate.Mineral solvents: We didshow, decennia ago, that ammonia andcarbondioxide,whicharemostinterestingsolvents,mightdistillawayfromthefrozenstate. Thecaseofcarbondioxideisevenaratherremarkableonesinceitsublimatesathighspeed, atatmosphericpressure, andatalowtemperature of 78.88C. Moreover, it is a nontoxic chemical and leaves noanalytically measurable traces in the final dry product. Knowing itsexceptionalsolventpropertiesinthesuper-criticstateitmightbecomeavery valuable extraction and supporting mediumfor many unstablechemicalsandbiologicalsthatneedtobestabilizedasdryforms.NH3is more difficult to handle but it might offer equally interestingopenings for thepreparationof transient compounds manufacturedinFIGURE21 Thermoluminescence of fresh and freeze-dried SiO2gels (microbeads) afterg-irradiationat 1968C(30kGy).24 Rey[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0001_O.3d][30/4/010/22:28:13][128]liquidammoniaandunabletobebrought toroomtemperatureunlesstheyarefreeze-driedfromtheirfrozensolutionat 1208C.Organicsolvents:Manyorganicsolventsaresusceptibletobeeliminatedbysublimation fromtheir frozen solution, for example, dioxane, diethylamine, chloroform, cyclohexane, dimethylsulfoxide, and benzene. Weshowed that they could be used for the preparation of delicatebiochemicals (like phospholipids). In that case, however, the mainproblemistheresidualamountofsolventthatcanbefoundinthefinaldry product andthis, of course, is a redundant issue for the wholepharmaceutical industry. For that reason the use of pure organic solventsinfreeze-dryinghasnotreallybeendeveloped. However, theirapplica-tionaspartialconstituentsinmixedaqueoussolutionshasgivenrisetomost interesting new processes. This is the so-called cosolvent issue that isdealt with in details further in this book by Dirk Teagarden, who has beenpioneeringinthatfieldforalongtime. Itisexpectedthatthisapproachwillfindmoreandmorenewapplicationsinthefuture.ContinuousOperationsFormorethan65yearsfreeze-dryinghasbeenalmostexclusivelydonewithinthe pharmaceutical andbiological industries as a batch-type operation. Thestartingliquidsolutionsaredistributedinvialsorfilledintrays,frozen,dried,and then handled as one single batch. To the contrary, the food industry, whichhadtofacelargeproductions, hasprogressivelyshiftedfrombatchtosemi-continuous and even purely continuous operations. The results are a substantialincreaseinproductivityandamorehomogeneousproduct.Tomyfeeling, it is ample time for the pharmaceutical andbiologicalindustries to take up that challenge and enter the processanalytical technology(PAT)constraintswithbettermonitoringandcontroloftheiroperations.Indeed, any chemical engineer dreams of a continuous process because it iseasy to control and gives manufactured products a standard equal quality. In foodareasfreeze-dryinghasnotescapedthistrendand, asearlyasinthe60s, semi-continuous to continuous equipments have been designed and built. Leybold wasamongtheveryfirsttodoit,andOetjendevelopedthecontinuousqualitycon-trolled(CQC)processintoanindustrialrealityformilkproducts. Theoperationwas, indeed, sequenced into several phases. The frozen products, most often undergranular form, wereloadedontrays placedonaspecial carrier, hangedtoamonorail that traveled all along the freeze-drying tunnel between heating plates insuccessive steps through vacuum locks closed by sliding gates. The total cycle timefromtheentrancelocktotheoutletwasoftheorderofseveral hours.An alternativetothis systemwas introduced byAtlas who pioneered theso-calledConradsysteminwhichtheloadingofthefrozengoodswasdonetraybytraythroughasmallsidelock.Notonlycoffeeandmilk,butalsoveg-etables,fishfillets,andmeathavebeensuccessfullytreatedinthisway.Quiteobviouslythistechnologyworkedbutitwasstillasemicontinuousprocessandfoodindustrywaseagertodevelopafullyautomatedcontinuousoperationforoneof itsleadingproductsontheinternational market, that is,instant coffee. I had the opportunity to live this development very closely, and Icantellyouthatithasbeenverydifficultsincemoststepsoftheinstantcoffeeprocessinghadtoberevisited.GlimpsesintotheRealmofFreeze-Drying 25[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0001_O.3d][30/4/010/22:28:13][128]Thus, itisnotsurprisingthatitisonlyinthemid-1970sthatafullycon-tinuouslinewasintroduced. Inoneof themost advanceddesignsgranulatedfrozenextract isfedcontinuouslyintothedryer througharotatinglockanddepositedona20to30mlongvibratingtrayonwhichittravelsasafluidizedbed. It moves, indeed, onits ownwater vapor cushionall alongtheheatedsurface that provides the energy for sublimation. During that long transport, thegranules are guided by vertical ribs into parallel channels, and great care is takento prevent attrition from mechanical shocks between the granules themselves andwith the surface and walls of the tray. Though limited, attrition nevertheless doesoccur andgenerates alot of dustthat has tobestoppedbyalongsemi-cylindrical screen deployed all over the vibrating tray. In that way, the fines arenotcarriedawaywiththevaporstreamtowardthecondenserandthepumps.Underthosecircumstances,itcanbeeasilyunderstoodthatacontinuousfreeze-drying plant is a highly sophisticated piece of equipment that is designedand put together by the food industries themselves, who keep the wholedevelopment asstrictlyconfidential andassembledifferent componentspur-chasedfrommultipleunconnectedmanufacturers.Theefficiencyofavibratedtrayfreeze-dryerisenormousandthedryingtimesdropbymorethanoneorderofmagnitude.Wearespeakingintermsofminutesinsteadofhours, andthroughputoftensoftonsperdayisnolongerunrealistic. Moreover, the process is intellectually clean and the end product isconstantinquality.This is thereasonwhyI havebeen, for averylongtime, dreamingtointroducethattechnologytotheprocessingofbiologicalsandpharmaceuticals.Indeed,despitetheelaborate design proposed by equipmentmanufacturersandthecarethatthedrugcompaniestakeoftheirfreeze-dryingoperations, wearestill facingtherecurrent problemof potential heterogeneitybetweenvialsandampoules within a same single batch. The spatial distribution of 10,000 to 100,000vials in a multishelf freeze-drying cabinet remains a problem. Some sit close to thedoor, someright inthemiddle, othersnearthecondenser, someintheuppershelves,othersdown,and,letusbeserious,theydonotdryinthesameway.Iknowthatdrugmanufacturersclaimthatall theproductsthattheyreleasearewithinvery closegivenstandards andIknow,also, that theregulatoryagencieskeepa closeeyeon that issue andthat they aremoreandmore stringenton thevalidation tests. It remains, anyhow, that the homogeneity of a single batch is stillamatterofdeepconcern. Canwesolvethatproblem?Idobelieveso. Averysimple idea is then to switch to a semicontinuous or continuous process as this isdone in the food industry. Let us forget about the holy paradigm that a vial thathas been initially loaded with a given volume of solution has to remain so until itisfullyprocessed, cappedwithitsinsidefreeze-driedcake, andplacedintoitscommercial retail box. Let us try to consider a process (Fig. 22) in which the initialsolutionisdistributedasindividual dropletsfrozenintospherical granulesofgivensize, continuouslyfedintoavacuumchamber, andspreadonaheatedconveyor. On that tray, most probably a vibrated tray, they glide self-suspendedontheirwatervaporcushionasathin, regular, andfluidizedcloudatawell-controlled operating pressure fit for sublimation, say a hundred microbar. At theendofthetraytheyreachatransferlockthatdischargesthemonanothercon-veyorplacedinasecondchamberfit fordesorptionandsecondarydryingatpressuresofsometensofmicrobar. Finally,theyenteranultimatelockandaredischargedintoareceivingbinunderdryneutralatmosphere.Theefficiencyof26 Rey[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0001_O.3d][30/4/010/22:28:13][128]suchaprocessistremendous, andifwehavegranulesof 1to3mmsize, thewhole freeze-dryingoperationmighttakelessthan30minutes.Now, at the end of the road, we have a population of granules, all dried inthe same conditions, all equal in quality and size that can be numbered, and fedintodrysterilevialsassomanydistributionmachinescando. Theresult isabatchof identical vials containingeach, say100 1freeze-driedgranulesinstead of more or less regular cakes, painfully manufactured over several days.Moreover, the product elegance is maximized and, in that dispersed form, itsreconstitutionisamatterofseconds.Thisisnot wishful thinkingsincewedohavethebasicknowledge, thepractical know-howandamultiyearexperienceofthistypeofoperation. Itisjust achangeof mind, anewapproachthat deviatesfromhistorical practice.Unfortunately,weknowthatthisissometimesmoredifficulttopass.FIGURE22 Engineeringdesignofapharmaceutical continuousfreeze-dryingplant.Thesys-temdisplaysfoursections.-Insection1,theliquidsarefrozeninregulardroplets,forinstance,asarainfallingagainstacountercurrentofcoldairatatmosphericpressure.-In section 2, the frozen granules of a given standard size are spread throughout a vacuum lockon a heated vibrating tray where they perform their primary drying at a moderate vacuum near0.1mbar.-In section 3, subsequent to a new transfer through a vacuum lock, the already ice-free granulesare deposited on a conveyor under a controlled infrared or microwave heating at a rather lowpressureinthe0.01to0.001mbarrange.Theretheyaccomplishthesecondarydrying.-Insection4, afinal lockallowsthefreeze-driedmaterial tobebrought backtoatmosphericpressure byinjection ofadry,neutralgassuch asnitrogen orargonandtheyaredistributedbythenumberintopreviouslysterilizedvialsthatarefinallycapped.GlimpsesintotheRealmofFreeze-Drying 27[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0001_O.3d][30/4/010/22:28:13][128]Some people will claim that instant coffee manufacturers do not care aboutsterility, that individual freezing of small droplets is a difficult undertaking, andthat, maybe, mechanically, the granules will not prove resistant to the process. Iknow quite well that nothing yet has been completely solved but I equally knowthat we do have industrial solutions to those specific issues (frozenbeadsgenerators, isolators, automatic remotehandlinginsterileenvironment, etc.)and I do not have the slightest doubt that they can be successfully challenged if,atthesametime, somedevelopmentsaremadeontheformulationside. ThisnewapproachwouldequallyfitbetterwiththenewPATrequirements.More thana hundredyears after its discovery, lyophilization(freeze-drying or cryodesiccation) remains a challenging issue that requires an educatedmixof classical, basic multidisciplinarydisciplines, andnewrocketingtech-nologies. This book is the actual demonstration that it is not a freeze-driedknowledgebutacontinuouslyevolvingvividexperience.FURTHERREADINGSBordasF,dArsonvalM.CRAcadSciParis1906;142:1058and1079;143:567.FlosdorfEW, MuddS. Procedureandapparatusforpreservationinlyophileformofserum and other biological substances. J Immunol 1935; 29:389. (See also Flosdorf EW.FreezingandDrying.NewYork:Reinhold,1949:1280.)HauduroyP. Histoiredelatechniquedelyophilisation. In: ReyLR, ed. TraitedeLyo-philisation.Paris:Hermann,1960:316.ReyLR. Thermal analysis of eutectics infreezingsolutions. AnnNYAcadSci 1960;85:513534.ReyLR, ed. Fundamental aspects of lyophilization. In: ResearchandDevelopment inFreeze-Drying.Paris:Hermann,1964:2343.ReyLR. Automatic regulationof thefreeze-dryingof complexsystems. Biodynamica1961;8:241260.Rey LR. Glimpses into the fundamental aspects of freeze-drying. In: Cabasso VJ, RegamyRD,eds.Freeze-DryingofBiologicalProducts,BiologicalStandardizationSeries,.Vol.36.Basel:S.Karger,1977:1927.Grossweiner LI, Matheson MS. Fluorescence and thermoluminescence of ice. J Chem Phys1954;22:15141526.ReyLR.Lowtemperaturethermoluminescence.Nature1998;391:418.ReyLR. Thermoluminescence of ultra-highdilutions of lithiumchloride andsodiumchloride.PhysicaA2003;323:6774.ReyLR. Thermoluminescenceof deuteratedamorphousandcrystallineices. RadiationPhysChem2005;72:587594.NeumannK. LesProblemesdemesureetdereglageenlyophilisation. In: ReyLR, ed.TraitedeLyophiiisation.Paris:Hermann,1960:1411.OetjenGW, EhlersW, HackenbergU, etal. Temperaturemeasurementsandcontrol offreeze-dryingprocess. In: Freeze-dryingofFoods. Washington, D.C.: National Acad-emyofSciencesandNationalResearchCouncil,1962:2542.RieutordL.Internationalpatents,1961.EspagnanM, Plume P, MarcellinG. Dosimetrie des doses elevees par Thermolumi-nescence des pics profonds du Fluorure de Lithium: Fli: Mg, Ti. Radioprotection 1991;26:5164.28 Rey[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0002_O.3d][30/4/010/14:4:49][2951]2 StructuralandDynamicPropertiesofBulkandConfinedWaterMarie-ClaireBellissent-FunelandJoseTeixeiraLaboratoireLeon-Brillouin(CEA-CNRS),CEASaclay,Gif-sur-Yvette,FranceINTRODUCTIONThe structural anddynamic properties of bulk water are nowmostlywellunderstoodinsome ranges of temperatures andpressures. Inparticular, inmanyinvestigationsusingdifferenttechniques, suchasX-raydiffraction, neu-tronscattering, nuclearmagneticresonance(NMR), differential scanningcalo-rimetry(DSC), moleculardynamics(MD), andMonteCarlo(MC), simulationshave been performed in the deeply supercooled regime (110), and in a situationwheretheeffectsduetothehydrogenbondingaredominant.However, in many technologically important situations, water is not in itsbulkform, but insteadattachedtosome substrates or fillingsmall cavities.Common examples are water in porous media, such as rock or sand stones, andwaterinbiologicalmaterialasintheinteriorofcellsorattachedtosurfacesofbiologicalmacromoleculesandmembranes.Thisiswhatwedefinehereastheconfinedortheinterfacialwater.Water in confined space has attracted a considerable interest in the recentyears. It iscommonlybelievedthat thestructureanddynamicsof water aremodifiedbythe presence of solidsurfaces, bothbya change of hydrogenbondingandbymodificationofthemolecularmotion, whichdependsonthedistanceofwatermoleculesfromthesurface.Understandingofthemodificationfrombulkliquidwaterbehaviorwhenwaterisintroducedintoporesof porousmediaorconfinedinthevicinityofmetallicsurfacesisimportant totechnological problems, suchasoil recoveryfrom natural reservoirs, mining, heterogeneous catalysis, corrosion inhibition andnumerousotherelectrochemical processes. Waterinporousmaterials, suchasVycor glass, silica gel, and zeolites, has been actively under investigation becauseof itsrelevanceincatalytic and separationprocesses. Inparticular, thestructureof water near layer-like clay minerals (1112), condensed on hydroxylated oxidesurface (13), confinedinvarious types of porous silica (1422) or incarbonpowder(23)hasbeenstudiedbyneutronand/orX-raydiffraction.Inthefieldof biology, theeffects of hydrationonequilibriumproteinstructureanddynamicsarefundamentaltotherelationshipbetweenstructureandbiologicalfunction(2430).Inparticular,theassessmentofperturbationofliquid water structure and dynamics by hydrophilic and hydrophobic molecularsurfacesisfundamental tothequantitativeunderstandingofthestabilityandenzymaticactivityofglobularproteinsandfunctionsofmembranes.Examplesof structures that impose spatial restriction on water molecules include polymergels, micelles, vesicles andmicroemulsions. Inthelast threecases, sincethehydrophobic effect is the primary cause for the self-organization of thesestructures, obviously the configuration of water molecules near the hydrophilic-hydrophobicinterfacesisofconsiderablerelevance.29[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_files/978-1-4398-2575-4_CH0002_O.3d][30/4/010/14:4:49][2951]The microscopic structure of bulk and confined water is currently studiedby using X-ray or/and neutron diffraction techniques which are complementarytechniques. Thesediffractiontechniquesallowtoaccesstotheintermolecularpair correlation function g(r) (31) of a system which is the probability density offinding another atom lying in another molecule at a distance r from any atom. InX-raymeasurements, g(r)isthepaircorrelationfunctionofthemolecularcen-ters,toagoodapproximationequaltotheoxygen-oxygencorrelationfunction.In neutron measurements, g(r) is the weighed sum of the three partial functionsrelative,respectively,totheoxygen-oxygenpairs,oxygen-deuteriumpairsanddeuterium-deuterium pairs. In particular, it is heavily dominated by deuterium-deuteriumandoxygen-deuteriumpartialcorrelationfunctions.Thestructuralanddynamicpropertiesofwatermaybeaffectedbybothpurely geometrical confinement and/or interaction forces at the interface.Therefore, a detailed description of these properties must take into account, thenature of the substrate and its affinity to form bonds with water molecules, andthehydrationlevel ornumberofwaterlayers. Todiscriminatebetweentheseeffects, reliable model systems exhibitinghydrophilic or hydrophobic inter-actions with water are required. This looks the appropriate strategy to bedevelopedtoaccesstosomeunderstandingofthebehaviorofwaterclosetoabiologicalmacromolecule,aspresentedinthefollowingsections.Inthepastfewyears, computersimulationsandtheoreticaltreatmentsofthestructureanddynamicsof waterindifferent kindsof environmentshavebeen undertaken (3240). Some important results are now available. For instance,moleculardynamicssimulationsindicatethatthewaterdensityincreasesupto1.5g/cm3inthefirst fewangstroms of theshell aroundaproteinandgiveinformationconcerningthepaircorrelationfunctionsandorientationsofwatermolecules(41). Instead, it hasbeenshownexperimentallythat athinlayerofwatervaporisformedbetweenliquidwaterandahydrophobicsurface(42).The purpose of this chapter is to account for the more recent developmentsabout the structure and the dynamics of bulk and confined water as a function oftemperature. Examplesrelativetointeractionsofwatermoleculeswithmodelsystemsaswellaswithbiologicalmacromoleculeswillbepresented.THESTRUCTUREANDDYNAMICSOFLIQUIDWATER:ASHORTREVIEWInspiteofanenormousamount ofexperimentsperformedwithliquidwaterunder different external conditions (19), many of its properties remain not fullyunderstood. Themainreasonisthecomplexityoftheintermolecularpotentialresulting from the formation of intermolecular hydrogen bonds. Such bonds arestrongly directional andtheir study imposes the considerationof quantumeffects.Mostoftherecenttheoreticaldevelopmentshavebeenachievedbycom-puter simulations of the molecular dynamics using several sophisticated,effectivepotentials(3,10).Suchpotentialsarewrittenadhoctosimulateatthebest both the microscopic structure and the thermodynamic and transportproperties. Ageneralproblemisthat, eitherthepotentialimposesatoomuchstrongandice-likestructuretoreproducethethermodynamicproperties, oritreproduceswellthepaircorrelationfunctiong(r)andthen,theso-calledwateranomalies are not well reproduced. Many progresses have been achieved30 Bellissent-Funel andTeixeira[vijay][6 X9 Tight][D:/informa_Publishing/REY_H100116_2400039/z_production/z_3B2_3D_file