https://theses.gla.ac.uk/

Theses Digitisation: https://www.gla.ac.uk/myglasgow/research/enlighten/theses/digitisation/ This is a digitised version of the original print thesis.

Copyright and moral rights for this work are retained by the author

A copy can be downloaded for personal non-commercial research or study, without prior permission or charge

This work cannot be reproduced or quoted extensively from without first obtaining permission in writing from the author

The content must not be changed in any way or sold commercially in any format or medium without the formal permission of the author

When referring to this work, full bibliographic details including the author, title, awarding institution and date of the thesis must be given

Enlighten: Theses https://theses.gla.ac.uk/ [email protected] YHACIH SWDim OM WK 80IÆD STATE DwmBiQM m mmoGEM bobbed s o u m

à The al 8 «bm itted f o r the degree of

D o c t o r ©f philoeophf . .. of. -1

llBlTOrait^r of Glasgow

L^OT Roxburgh Me Ghi©o BoB&a AoEoOoBoT*

December 1965 ProQuest Number: 10647004

All rights reserved

INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted.

In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion. uesL

ProQuest 10647004

Published by ProQuest LLO (2017). Copyright of the Dissertation is held by the Author.

All rights reserved. This work is protected against unauthorized copying under Title 17, United States Code Microform Edition © ProQuest LLO.

ProQuest LLO. 789 East Eisenhower Parkway P.Q. Box 1346 Ann Arbor, Ml 4 8 1 0 6 - 1346 abstract

UbbelohdP hps propoeed m protom transfer mechpmism to explgalm oloctrioal conductivity atudio© on p series of crystplllnG organic acids rnd it haa been suggested thrt conductivity increaaes v;ith the degree of co-operrtlve hydro^ion bonding in the crystfilo EKperlmenta have been carried out to teat thi© hypothesis by studying the conductivity and. the rate of proton diffusion^ using tritium bb tracer^ in the same single crystpl© of two organic acids exhibiting different degrees of co­ operative hydrogen bonding? benzoic ecid which exists a cyclic dimer in the solid stete and acetic acid which he© linear chains of co-operative hydrogen bonds extending through the crystalo The acids were purified by distillation^ sublimation and zone refining techniques pnd single crystal© were grown from the melt in a Bridgman ov©n„ Diffusion wrs studied by e sectioning technique using tritium labelled benzoic acetic aeide to obtain the hydrogen diffusion rates \nû carbon-14 labelled - benzoic acid wrs used to mersure the diffusion coefficient of the bulk molecule « The electrical conductivity of benzoic acid vmm ^found to be very low in single crystalSo bein^^ less than 10^^^ ohm*"^ cm%^ 7C^ below the meltim^^ points contrf ry to measurements by IJbbelohde but rgreeing with more recent measurements by Eleyo A proton conductivity was founds however^ in oxrdie acid dihydrate a CO—op errtiv ely hydro^,en bonded pcidg which was comparable th?t found In other hydrogen bonded systems<> A tritium diffusion was observed in benzole aeld whieh was considerably faster then the bulk diffusion mû which could not be explained by a proton diffusion mechMisMo This tritium diffusion was found to vary with the water content in the diffusion cell but a lower limit was obtained which satisfied the Arrhenius equation 4>2o0 ® ® 0»5 _o„4 ' I - HT Diffusion etuiie© in p-terphenyl doped benzoic acid cryatalSp deutero benzoic acid and polycrystallin© benzoic acid have helped to elucidate the meehanism which is believed to be the diffusion of water molecules trapped in the crystals through interstitial dofeetao The carbon-14 bulk diffusion in benzoic acid was found to obey the. following Arrhenius equation « r “ 1 D ® Cl 0 8 ) 3S 1 0 ^^ . ©KP J and can b# explained Im terms of a relaxed vacancy diffusion mechanism o Conductivity and diffusion étudiés in acetic acid were both found to be greatly affected by the presence of moisture and no definite conclusion could be reached regarding proton conductivity in this système AOmOYLUmQMEW£S.

The author wiehe© to h is sincere thaKilcs to his supervisor;, Jo Sherwood g for Ma guidance ^ help and oMOuregement throughout this period of research,

The axrthor aim wishes to thank MrSc. Ho Fergie for typing the manuscript and his for tracing all diagramso

Thanks are also duo to the Scientific Research

Council for the award of a maintainano© grant during this course ©f studyo CHAPT'ER I PÜRIP1CATI0W AMD CRÏSTAL GROWTH

Xo Introduction <=oc»c=>«=>*=>cae=>ca< 3ato«s.fi^<=5.«»o=>«cjamt=i 3.,^ 2o Purification of Benzoic Acid il

a ^ D is tin c tio n CQO

b) %©nO Refining cac=,c 3 C3 a C ;3 c=,c 3 C3 <=,oc=,BaK=,c=,oc=, 2Q 5o Design of Bonzolc Add Cry stale Growing Otom « - - 27 4o Growth of Doped Bonzoia âcid Crystals - -«-«.=>-35

Purity of Bonzosic Add — — — — — « — « — — — — 6q Growth of Acetic Add Crystals — — — — — — — 4 CHAPTER II BÏFFUSSOM ABB COBDÜGTIVITY ŒPKBIMfflTS

1 0 Introduction to JDiffueion — « — — — — « — — — — 49

2 o Préparation of HadioactiTO C=> C=> CT» G> <3 c=a to

3 o Depoeitiom %st©M© W C7 C3 to C3 cs ca to to ito to to ca c=> cr?

4o Bi«4«3Sx^^ion Ann©al Oo^^-itions — — — — « — (64

<^0 Th® ^©c ti osning TocSîUiqu© co — «3c»e=> — c=>c=.c=> — w — — —67

6 0 Determination of Radioactivity In the Crystal Section© 77 7o Preparation of Bi^xusion Samples — — — — — — — — — ©3

Ih© D iffus j>on Xvxporimonts — — — esta — — co — eata, — 0 ^ 4

9o Ih© E3tchang0 Experiments — c^ — iacacsta — cacaacs, ^ 0 5 lOoConductivity Studi© C3 ca cr> cl ;» C:) O CD c cr,> tt> CHAPTER I I I THE RESULTSs AND THEIR INTERPRETATION PAGE

8 Tritium and Carbon 14 Diffusion in Benzoic Acid Single C rystal# 91 b Tritium and Carton 14 Diffusion in PolycrystalliM Benzoic Aeid, vX04 8 Tritium techarjg© between Benzoic Acidy/water Vapour 109

â Tritium Diffusion in Acetic ac M ------111

© The Conductivity Studies — « — — — — — — — — — f Tabulated Results — — — — — — — — — — — — — — 117 cmprm IV Discussion op the resoms a Pro tom Transport im Hydrogen Bonded Solid 1 Previous ÇVork — — — — — — — — — — — — — — — — — H©

H Gomduo tivity Results — — — coca — — c=,<=,c=,c»(=, 123 h D iffusion im Bomzoio AoM

1 ^uw&&ary of R esults — — — — = — —— — p* — « — — c=. 120

H i-9®l©©ular D^ffu®io m — — — — 130 m Gicaim Boundary DÜ fusion — — — — — — — — 135 IV TritiuFfl D iffusion — — — — — — —— — — — — — 137 ^ Difiusiom 5i,m Aoets.© Ao%d — — — — — — — — — — — — — 144 ^ Oomolusiom — — — — — — — — — 145

^ i

1 a Unit cell of benzoic acido b Unit cGl'i of acetic acid* 2 Principle of single crystal growtho 3 Distillation of benzoic acid into growing tabe^ 4 Automa.tio zone refixicr D&Oo 1^ 5 Automatic son© refiner too 2o . 6 Benzoic acid c r i s t a l growing oven.^Oo !« I Temperature g rad ie n ts in c ry s ta l groxving qwu Bo 8 Crystal growing oven 2o 9 Qrysfcal growing v@aeel@o ... 10 'QoVo ©pèctra of p-Tezrphenyl and benzoic aoMo' II Purification and crystal.growth of deuterobenzoio acidp 12 Freezing point apparatus^ 13 Water determination in benzoic acido 14 Tritiation of benzole acMc 15 Vacuum sublimation apparatuso 16 Trltiation of acatie acldo XT Beï.zoic aeid evaparatoTo 18 A cetic acid deposlti on aysternn 19 Diffusion cells„ 20 Thermistor controller,circuito 21 .Fine adjustment chucks 22 Sucli ©ifl apparatuso 23 calibration of benzole acid in cycohaxaneo 24 a Effect of scintillator volxme on tritlxm counting efficiency ’b Effect of benzoic acid on tritium counting efficiencyo 25 Oonciuetivity cells^ 26 Graph of VS for tritium diffusion plane in benzoic acid (large vessels)o 21 Arrhenius plots for tritium anci diffusion in benzoic

aod single crystals « ...... 28 a^bo urapn Xog^^A’VS ^Ihror tritium aiffuMon Jj (uOl) plan© :ln benzoic acid small' veaselSo 29 Graph VS for tritium diffusion JT( OOl) plan© in benzoic a.crid -under' dry conditioner, 30 Greÿh log^qA VS T^'for tritium diffusion jf (001) plan© in d©xxterobenzoilo acid and p^^TerptonyX doped benzoic acido 31 Graph log^^A VS %^for trltim i diffusion J j (001) plane in benzoic and deuterbbbnzoie. acid simultayxeouely<, . 32 Graph VS g-g^for tritlWa diffusion JT( 001) plane in benzole aeld froBi TH0 vapour^ .

33 a'ptoo Graph logn VS for diffusion (001) plan© in benzoic acMa . ' 34 Graph of’ V8 %^for 0^'^ di.ffusion parallel to {001} plan© i n ’b©ns6ïre acido - *î A 33 Graph of Xogi^â VS for G* ' 'diffusion in benzole' aold. compacte, '36 Graph of logVS % for tritium diffusion in benzoic acid compacts» 31 Arrhenius plots for grain boundary diffusion in benzoic acici

compacta 0 ' . ■ - ■ . 36 Graph V8 f€ for ©échange reaction between triti&ted benzoic acid and water vapour« ' ' 39 Arrhenius plot for tritium exchange HgO/benzüic acicVo ' • 40 Graph log^gA VS or tritium diffusion if (100) plan© in acetic acid» 41 AoCo conductit>’ity In benzoic acid compacts^ 42 Time variation of AoO» oonducti’^lty in acetic acid single cry a tale 43 a^^bo AaCo and Bo Co conductivity' In oxalic acid, d'ttj^dr’at©^ 44 Schematic xriew of hydrogen -bonded chains ih cry stals^ INTRODUCTION lo

of the physical rncl chemical properties of solids haso imtil recentlyj, been principally confinée to those materials of potential mouatrial importance e^ga aiotelsp semiconducting end ionic crystals^ This work has lead to a 1 fairly well deirelopeo theory of the aolii state in theoe systemaV j\ recent upsurge in intereet In organic mû molecular solids hps evolved^ hoiveTC5% from the discovery of interesting p % aemicondiicting and photoconducting properties in these systems Electrical rnd optical properties have beexa studied in b l©rge xiumber of organic solicla rangixïg from well characteriaed aromatic solids like anthracene and naphthalene" to more complicetedg 6 blologlcplly Interesting protein-like structureso Ix?> all systems it has been foimd that properties are frequently very dependent on the de^x'ee of p u rific a tio n and hence the cry st a l l in© pex’fection of the solidso The imperfections in the crystâlSf, whether natural imperfections due to impurity molecules or non equilibrium line defects like dislocations^.com greatly affect the property being studied and give rise to extrinsic properties which may completely mask the intrinsic property being investigated^ Organic solids differs from metallic and ionic solids in that their binding forces are primarily ?an der Wsiala interactions and the molecules crystalliee in the system which gives the ■greatest packing density except where other factors such as hydrogen bonding have an effecto There have been relatively few studies on organic solids other than ©lectrical/optical. properties m û i n order to uMeratasid th© basic molQeular r reaotiona ooeurrixsg I n thea© aoXida am ©xtonaiw study tB reaotiona ooeurrixsg I n ttea© aoXida am ©xtonaiw study tB required ©B the properties of both ®p©rf©ot® crystals uml oryata.la hevimg a Immn and controlled defect structure « A cl BBS of organic solids which hm important chemical p ro p e rtie s and has not imdorgon© ©Ktexisive study is th a t im which the formatiom of imtermolecular hydrogen bonds controls the structural arrangement of the molecules in the crystals Hydrogen bonded organic crystals can -be divided into two

prixxcipl© categories g co-operative snd nem c@-opemtiv©p although Bom© structures may contain a mixture of botho These have been reviewed by ïjbbelahd© and GMIacher? Co-operative hydrogen bonding is exhibited by crystals which contain infinite chains @f hydrogen bonds extendtog through the cx-ystalo Typical examples of such a structure are the alcohols whose configuration can be represented Bchematically as

where H is a hydrooarboxn groupo Such co-operative essembllea can be further aub-dividecs tot© chains^, sheets and three dimensional networks dependImg on the spatial arrangement of the hydrogen■bonds o 0 0 -opera.tive aystems as their name suggeste oompased of closed cyclic stru c tu re s

cjcfc dimers as exhibited by almost all earboKylio acids« In e ll casesp h o w e v e rth e c@mfi^,uratlom is achieved which leaves the system in its lowest potexxtial erne%gy oonfiguraticmo Interest has been aroused ie the possible role of co­

operative hydrogexx bonded eystema bb pro tonic conductors o It ha© been suggested that hydrogen bonded hydration structure© are present in protein membranes which are favourable to pro tonic c ont .ml systems axid that fast proton trmsport may 1 A play an important role in biological react!im klnetiaso*’^'^ FeCli)-^---Fe(lll) o2cldatlon has been observed .In Iron salts dlaaolved in ice and a mechanism involving proton troAiaport along hydrogen bonded chaixxs has been suggeeteCil * 1 Some1 experimental studies have been made to teat the validit^^ of a proton transp0,rt moehtarlBm in co-operatively hydro^^an bonded solids» 12 In 1951 Kakluchi et alo demonstrated that current fl#w through cetyl alcohol occurred almost entirely by proton charge transporte They did th is by observing the sp e ctra of hydrog;en discharged at the cathode after passing a current through the solid alcohol» SmytM*^1 ha© also proposed © proton transfer meeheaism associated with malecular rototion to explain high cenduetlvlty and clialeotric lose in the higher alcohols^ a- tetrstexy^'-'^Vd hexadecylo The effect of the degree of co-opemtlon on coKxductivity has been etudieo by Pollock mû lîbbelohd©^"'^' in a serie© of orgeuTiic acids ax'xd they found that as the degree of co-operation increased the conductivity increased mid the activation energy decroasedj the conductivity being greater in acids containing water of crystalllaationo They interpreted their results a© Bubstantiating a proton conduction mechanism In hydrogen bonded so3J.dao A comprehensive conductimetrie study has been made by Eley 15 end h ie CO-workers' on sy a terns containing the (G^O — H - M) hydrogen bonded unit ranging from simple molecules like glycine and 051 amide to polyamides and naturally occurring proteins» Only in polyamide© was the conductivity due to proton migration and in all othezc système the coxiduetivity was 10*^ times lower and electroxxic in origine Mey also found that the presence of adsorbed moiature on protein© greatly affect the conductivity^^ This has been shown to give proton conduction in keratln^'^ but IB electronic conduction Im heemoglobtoo OoMucti'vity studies in ice whlchp though not an organic BOlldg is equivalent to an alcohol E-OH with has been exhaustively atudied^^^ and proton transport verified ,21 by this and ©leetrolysis meaaurememta' o 5 c

With the exception of Ice all the above studies showing proton conductivity wer© made on polycrystallina compacta» It hsa frequently been showng howeverg that compacts often give misleading résulta as intergranular end surface effects may occur due to the presence of adsorbed iapuritiea ©og» oxygen and vmtero"PP I t would be preferable^ therefore^ to examine thi© proposed proton conduction in single crystals of materials in which the crystallographic orientation im known and the chemical impurity concentration and non-equilibrium defect structure can be reduced to a minimumo As no intrinsic proton conductivity has yet been conclusively proved in hydrogen bonded organic crystalm it was thought that a useful contribution could be made to organic solid state theory by attempting such an Investigation in single crystals of a series of molecules esdiibitlng different types of co­ operative hydrogen bondingo An independent method of verifying charge transport mechanisms in ionic crystals is to measure both the bulk con­ ductivity and the diffusion coefficient of the ion which is believed to be the charge carrier in the same single crystal over a range of température» If the same mechanism is responsible for both phenomena they will be related through the well known ox Bernst - Einstein equation»

crr> J) m 6

which correlates the specific ^conductivity 9 ^% at temparaturè T with the measured diffusion coofficisntg Do B im the total, numbar of ions of the conducting species per unit volume g e the electronic charge and k Balt‘zraann“e coMtanto The Bei'nst-Elmeteim equation has been experiaontally verified for ionic BOlida by the work of Mepotherg Crook© and pA Maurer who studied DoCc conductivity and Ba"‘ diffusion im the Berne aingle crystals of sodium chlorideo They found that the Bemst-^Einateixi equation imm obeyed exactly in the intrinsic temperature region but not im the e^ttrinsie regiomo This proved that in the intrtosic region the conductivity mam due solely to the migration of aodium iomso This type of measurement has not been attempted in hydrogen bonded organic solide which are protom comductora and it naa thought that such a study might yield interesting Information about the mature of the charge carrier© in the©© syatemso llbbeloMe'^s study of organic aeida"^^ gave specific conductivities of 10^^ to cral for co-operatively bonded acids at temperatures well below their Melting pototSo If this is due to proton conduction a3.one then the diffueian coefficient of the proton calculated from the Beimst - Einstein equation is to om?

which im measureable by tracer techniques0 Before embarkImg on such a studyg however^ i t is a e se n tia l

th a t a 0 ultE>ble isotope of hydrogen is available fo r a d iffu sio n study and should aatiafy th© following requirements 1 . lo The isotope must be detectable tn tra c e r a®oimtSo

2o I t must9 if radioactiveg have a half-life of several times the total diffusion anneal timeo !io It shom'ld not eoxistitute a safety hazard» 2 Of the two ïmmm iaotopea of hydrogen g deuterium^ @ 1© nearest to hydrogen in properties but la non-rad 1 oac11 v@ m û muat be determined by physical methods which are not yet aeasitlTC enough for the experiments envisaged though an aotivotioB analysis method ha© recently been reported In which one mgm» ©amples of deuterlim cm be accurately determined?'^ The second isotope tritium g does ziot occur naturally but im easily prepared by a Li nuclear transforaetion» It is radioactive and dislntegra.tes with the emission of a weak p g 0 ^ 018 MeV and has a h a lf - lif e of 12»3 years mailing i t a suitable isotope for a tracer diffusion studyo An important point arises^ however^ over the isotope effect» In normal diffusion studies Oogo in metallurgy^ this is negligible but the mass difference between tritium and hydrogen ia large 3:1 hence mu appreciable isotope effect may be observed If the proton diffuaea alone » Studying tritium diffusion in a normal and a completely ieuterated hydrogen bonded system and comparing the values obtainedg however g may help to elucidate OP the mechanism o Isotop© studies in have aho?/n that the activa’tion energy of conduction is almost unaltered by eufo- a t 1tuting deuterium for hydrogen» The choice of the m&t suitable systems to study was made with refem xce to the fo3J,owing; points» 1» The systems muat differentiate hrntimm the different types of hydrogen bosadingg co-operative and nan co-operative aa that an assessment cao be made of r-ho effect of co­ operation on the proton transport prepartlea» 2o They should be thermally ©table » 3o They should be obtrîinable in a high degree of purity»

4 o They should have similar properties and crystal ©tructmre so that a direct oomparieon can be made between them » 5o They should not undergo a phase transition between room temperature and the melting poiato

6 0 They should bo mechanical.ly ©table mû should not shatter on application of a shear stress or rapid coolingo

7 o Suitable laotopicality labelled molecules must be available^ also in a high degree of purity» l%- wa© decided that the mono-^carboxylic acids would be most suitable syBtem»' All normal moxio-oarboxylic acids crystall* ime BB cy clic dimer© with the escceptlon of the two in itisâl membersg formic and acetic acidg which have linear chain© of hydrogoxx bonds esstendlng thresh the cry at al» Two acids were chosen for study each ©schibiting one type of hydrogen bonding» Benzoic acid was chosen aa the exemple of a non co­ operative hydrogen banded system consisting of a@n Interacting cyclic dimerso The advantages ©f studying benzoic acid were:

\-51)\z\

2(87

.481 24. °__ L

FIG, io„, UNIT CELL BENZOIC ACID VIEWED ALO NG b o AXIS.

b

UNIT CELL ACETIC ACID. lo It :la obtainable in a high degree of puritj^'o 2o It is theraally at able f.*cl has no solid state traMitio.no 3« It 0rysta3„Xises in the mono ciinio Bjstem^ space group v;lth fcmr moleculee per unit collp Im. t’llth 0 a 5»52Îlp b 5 .1 # , s 21.9Ë aM fê 9 4o It hasi a well defined cleavage plm a (001) - #lch facilitates crfstal ©rlontatlono 1 â 3o Tritium and C*" la b elle d molecules are re a d ilj aval3.abl©o Acetic acid was chosen as the e.i&aisiple of a co^^operatlve hydrogen bonded system and had the folloi^lng advantageso lo It :1s thermal3.y stable îîith no knoim transition below the

m®lting p 0 in t «

(L^j. o Itm crystal striactur© has been characterised by Jones and Templeton?® It ia orthorhombioy space group Pna 2^ with four moleoülea per unit ceilp having a 13o32lj, b 4oOSt m d ?o77Ë<, Fw 0& 3o fritiiM and labelled moleculœ are easily preparedo Disadvantages in using acetic acid are its low melting point of 3.6o7^C and the fact that it is slightly hygroscopic « It was thoughts howeverp that these points could be overcome by suitable expérimental techniqueo Both these acids are of almilar atrength^ benzoic m iû pKa ^ 4h"c4 3 acetic acid pKa - 4«^T ^ and solidify to give systems containing hydrogen bonds of approximately the same lengthp benzoic 2o64a$ acetic 2o61Ët, The crystal struetusvi of 10 o both is approximately the as benzole acid is a].most orthorhoMblCf, & a 97^ and the major difference between these csystala lajj^ype of hydrogen bonding hence it is to be expected that any difference in the proton transport properties will be directly attributable to this factOFo In both these solide the bulk molecule is large and can b© easily labelled with C'^'^ hence a comparison can be made between diffusion of the bulk molecule and that of the protoho This will show whether the proton diffuses with the bulk molecule or not<, Very little is known of molecular diffusion and this study may help to elucidate bulk diffusion mechonisma in organic solids*. As the defect structure of solids has been shown to have a marked effect effect on the properties it was thought that additional information could be obtained from a study of the above system by a) Introducing a substitutional chemical im purity and h) introducing gross non equilibrium defect© by studying conductivity and diffusion in high purity poly- crystalline compactso The object of this theaiSp therefore^ was threefoldo lo To purify and grow large single crystals of hydrogen bonded organic acidso 2o To measureg in these crystpls^ the conductivity and proton diffusion coefficients^ using tritium as tracerj, to confirm or disprove the hypothesis that intrinsic 1 1 0

proton conduction can occur in eo-opormtive hydrogen bcmdod so lid s end elu cid ate the mechpnlsBo 3o To examine the effect of introduced imperfections im the above system®o ji'l 1 9

I lo Iwll-ODUCnOiM •the defect structure of solids has been shown to exert a marked e ffe c t on many so lid s ta te phenomena p i t is important when embarking on a study of ?. particular crystalline solid to prepare single crystals with only an equilibrium defect structure « That is to say that non equilibrium defects such as dislocations^ grain boundaries and point defects introduced by the incorporation of impurities should be reduced to such a level that they have no effect on the intrinsic property being studiedo It is not always possible to attain this equilibrium due^ to the technical difficulties involved and the method of growing such perfect crystals ia still very much an arto There are three principal ways of growing crystala of which all techniques are simple or sophiaticsted adaptations» (a) Growth from aolutioBo (b) Growth from the vapour phase ? (c) Growth from the melto Method (a) ia the most generally used Indus tria l (method and has been used to grow very large single crystals'e«g» a 43 Ibo crystal of ammonium dlhydregen phosphate was grown 32 the Bell Telepiion© L aboratories over a four month period « 13

This method frequently gives flawless crystal© but suffer© from the disadvantage that crystal growth Im very ©I 0 W9 about 1 mm/day being an average rate 3 and th a t the c ry s ta ls may grow in b particular crystal habit which ia experi*- mentally unsuitable*> Crystal habit varies with the solvent used and can often foe modified to force the crystal to grow in a preferred crystalXographic direction^ Hsfolt modifier© are frequently dye© which adsorb on a partiaul.ar crystal plane preventing further growth thereono As the modifier is usually incorporated in the crystal this may affect the intrinsic property being studiedo This method may also introduce microscopic occlusions of solvent which are equally detrimental and hence i© not considered particu3.arly suitable for growing high purity crystals with a uniform defect stru ctu reo This method^ however^ is frequently the only suitable method of growing single crystals which either decompose on heating or undergo a phase transition between room temperature and the melting

Growth of c ry s ta ls from the vapour phms© ia also useful where decomposition on heating or a phaa© change occu rs0 Thia method^howeverp usually requires a fairly high vapour pressure at the growth temperature although 14 sûffla organic crystals haw been grora at vapour presBurea of a fen hundred miorona Growth r&tea can varj ooneiderably and can be as high m X-2 om/da^o Vapour growth gives high purity^ flawless crystals but difficulty la ©iiperieMed in growing oryatsla larger than a few mmo on edgOp predominately due to a large heat of orystalliaati coupled with poor heat dissipation caused by the low thermal conductivity of most organic aolidSo In this method of growth it is also difficult to control the habit of the c ry sta l formedo I4ethod (0% growth from the melt g wm considered the best method of growing crystals of sufficient purityg sis© and imlforsity in the present Investigation« Growth from the melt ia e sse n tia lly a simple one step operation in which a melt of pure material, ie cooled through the freezing transition at a rate alow enough to allow equilibrium single crystal growtho There are two p rin c ip le v a ria tio n s of th is méthode a) The Kyropoulom method*' % y h) The Bridgman 3tockbarger methodo In the Kyropeuloa technique a seed crystal ie slowly withdrawn from a mielt of pure materialo Surface tension draws the melt in contact with the seed above the level of o! o o 0, o 'fcTis (fcenRzyKfKuraTJifz'K: 0 0 0 0 e wT' d 0 *” j Tm 2 , 1 J ' Â

r . y. ? .. PRINCIPLEa,lM-ri»y»H**«e»*ï**5*yes9.<>>ftat** 5Ttâcir»i OF3i»Te 7i*TWihioE«-GeFWt^»»*rS SINGLE5'Ç:^"5par'»*irT%: CRYSTAL o the melt where it then crystallise®* The rete of withdrawal of the seed la adjusted t© give equilibrium freezing mû a single crystal is obtained* This methodp however^ is only suitable for ms te rials w;hioh have a low vapour at the melting point and has been auacassfully used t@ grow large alkali halide single crystals* Unfortunately the organle acid® t® be grown both have fairly high vapour pressures at the melting g of the order ©f several mmo Hgg md sublime hence this method was not considered particularly suitableo The principle ©f the Bridgman Stoekbargerg ©r moving ve©s@l techniquep is illustrated in fig g. * A vessel A containing the melt is slowly lowered from a high temperature regiong T > through the melting point isothermalg T^g in to a low tem perature region g T < In order to grow a single crystal a seed crystal must be formed at the lower end of the growing tube from which the melt will crystallise into one single crystal boule « The formation of this preferential seed crystal which controls the whole growth process has tazec the ingenuity of crystal growers In the pnst and several standmid crystal growing vessels have been designecl'^'g all of which start with polyerystallin© material formed by spontaneous nucléation on supercooling the melto The crux of the problem is to preferentia3,ly select one of 16 r these crystals and this is uaua3*ly accompliehed by either having a .constriction in the growing tub© which allows only one crystal through which then controls the subsequent single crystal growthg or by forcing the crystal to grow through a bent capillary* This latter procedure may also fijiontrol the orientation of the single crystal formed as the growth plane ©f the crystal formed as it travels up the capillary -can be altered by the angle ©f band through which i t is grown^» In th is way the orientation can be c^ltercs-d

The Bridgman Stoekbarger method was chosen to grow crystals of benzole and acetic acid for the following reasons lo The method is relatively simpleo 2o The erystalSg which aublineg can be contained in a seeled vessel which els© prevents external contamination during growtho 3o Large uniform single crystals can be grown in a few dayso 4o It should be possible to grow crystals with different crystallographic orientation* Before groving the crystals the starting materials were purified by the following procedures* ’, g,. Purification ©f Bengoic Acid*

Purification of organic materials can be accomplished by a variety of unit operations of which solvent extraction^ dietillationg sublimation and recryetallisation are th© most comsiono In the present study benzoic acid was required in as pure a form as possible for the reasons given in the previous section* The preparation of high purity benzole acid for a standard substance for calorimetry and aeidimetry has been studied by Schwab & Wichera ' „ They found that purity

greater than 9 9 «9 8 ^ could be obtained by several methodsg

recrystallisation from water or benzeneg fractional freezing or hydrolysis of benzoyl chloride* Purity of 99o99S^ was obtained only by recryatalliaing from benzene eight timea ©r by fractionally freezing twice in a manner almllar t@ the Bridgman technique for grov:ing single crystalso The simplest of these methods was fractional freezing and this grew was the method chosen to^single crystals in this investigation* The starting material used waa Analar reagent grad© benzoic acid supplied by Messrs* Limited* The specificationa of this material guaranteed greater than

9 9 o9 ^ purity with known impurities having the following

mimimum concentrations,, 6 pop^m* metalliCg 0 * 0 2 5 ^ non volatile^

—

FlGj,3. DISTIU.ATION OF BENZOIC AClP INTO GROWlNS TUBE, 1 8 ,

0o02^ ehl®riB@ ana the remainder probably adsorbed oxygen and water^ In a tritium diffusion study water is likely to be th© most detrimental impurity but im this ease i® ©aaily removed by ©imply evacuating the fused acid o'* Per the initial purification of benzoic acid prior to ©ingle crystal growthp zone refining appeared to be th© most suitable techniqueo As no zone refiner waa available it was decided t© construct one, During this period benzoic acid was purified by simple distillation under reduced pressure using the following technique,

JC t (a BiatillEtiûfâo

for th© Initial experiments], growth of benzole acid

©ingle c r y s ta l© the apparatus shown in figo 3 was used,

0 S Approximately 200 gma® of m alar benzoic acid was placed In flask A and evacuated for 1^2 hours at lO"" cm*, Hgo Tap T|^ was closed and a few «« of dry nitrogen introducedo The acid wae carefully melted and degassed by opening tap to the vacuumo This procedure removed meet of the water and it was swept out of the acid by the dissolved air

bubbling ©uto then closed and 1 0 qMo Ego pressure ©f nitrogen introducedo The system was then sealed at and the acid gently distilled over Into th© crystal growing

vesselB the f ir s t 2 0 ml « of distillate being trapped in limb B using a liquid nitrogen trap which served t@ remove any last, traces ©f watero The subsequent 60=^80 mlo @f distillate sufficed to fill the crystal growing tube to within 1'^ of the seal Th© distillation wea stopped at

this point and the growlmg tube sealed at 8 2 » Th© glass at Sg was then fashioned into a hook from which th© vessel was suspended in the crystal growing oven o Th© benzoic acid in the growing tub© waa al.waya erystal clear in th© liquid state end crystallised to give a pur© whit© cry stall in© solido The benzoic acid left in th© distillation tubOj, howeverÎJ always had a lig h t brown colouration « Thi^ indicated that either some degree of purification had taken place or th© acid was decomposingo Experiments by Schwab 4*-P and Wichers ' in which they held 3 samples of high purity © benzole acid in sealed capsule© for 72 hours at 200 Go under one atmosphere of a ir and oxygen and under vacuum^ gave

0o047g O 0 O4 7 and 0o025^ impurity respectively□ Thi© indicated that the rat© of decomposition was very alow and m s probably du© to oxidation hence it seemed likely that purification was occurring during distillationo A further indication that this brom colouration was a natural Impurity vwm later obtained by zone refining during.which a similar brown material was quickly rejectedo 20.

This method of purification and tube filling was used successfully until the superior technique of zone refining had been developedo

Zone Hefining

Since its introduction by Pfann in 1952 zone reflningg or zone melting aa i t is sometimes calledp has rapidly become one of the most powerful methods of obtaining materials of super purity lo#o impurity concentration© parts per million or lesso The theoretical and practical development took place simultaneously and was the gin© ‘5? np,n of the semiconductor Industry^^o The technique j general application to materials which do mot decompi melting and thus in many cases can be used to prepare organic crystals with eontaminamta well below the level of analytieaî. detectability, In principle zone refining is a aimple adaptation of the normal freezing method of purifieationo Its power lies in the speed with which a large number of unit freezing ope rations can be made couplea with the method of rejection

0 f the impurityo The factor controlling the separation of an impurity between the liquid and solid phase is 2 1

equilibrium diatribution coefficient» fe®3

being the mole fraction o f impurity in the solid mû liquid respectivelyo ïïhen the impurity is more soluble in the liquid phase thèn &o < ! ana the solid becomes pro-- greaaivej^y purer» when the impurity is more soluble in th© 'Solid phase then and th© liquid became® purero fh© theoretical, aafeiumption @f constant distribution coefficient is only applicable in the oaae of very dilute solutions and tea can be obtained by extrapolation of phase diagram® to zero concentrationo In normal, zone refining operation© equilibrium freezing conditions are not attained and a related paraiaeter is used called th® effective distribution coefficient» te o The practical process of zone refining cone1st® of slowly passing, a molten zone of constant length» I» through a length» It» of the solid to be purifiedo If the initial concentration throughout is them th© Impurity distribution after on© pass is given f ' ^ k X "I ^/Gg ® 1 ■“ ( 1 '= k) @%p L “H ’J whereG/G. is th® relative impurity concentration fromthe

top of th© sample (x ^ 0)o The benefit of zonerefining is 22n that several molten zones can be passed through the sample at the Berne time providing that a length of solid exists between them* The degree of purification using this method is limited» hov.ever» and Braun^’^’ has calculated the number of passes beyond which no further purification w ill be obtmlmeéo 0ol'Cte40o3 he calculated the maximum effective number #f sea» n» as

- L 2 r%; 2

I t la possible t@ further purify Ih im material by extracting the pure fraction of the charge and re-Introduclmg it into a longer» narrower zone refining tube thu© giving a larger value of /& and higher purity, for a further discussion of th© many facets of son© refining the stm%dard work OH the subject is by Pfmmsf'" , The applicability of zone refining to th© purification of benzoic acid warn demonstrated by who increased th© purity of a sample of benzoic acid from 9 9 , 9 ^ to more than 9 0 0 9 9 9 ^ by two equilibrium freezing operations, This indicated that the distribution coefficient of normal impîlH'ties in benzoic acid is 4 1 end that, relatively îmi mon© parses m m required to give high purity material. It

Kl. - KS “ Ra = IS S. S P K IN G j. R SCREWED WHEEL. V."VÂKIAC" AÜTOTKASMSFOKMEI?. y RADIUS OF WHEEL,

■vWVW V S

y . / z .

ELECTRIC MOTO^, Fia, i. F ia . z. s c h e m a t ic v ie w . RICIPROCATIWI^ MECHANISM.

FIG. 4 AUTOMATIC ZONE REFINER. 2 3 .. has bI bo beem ehowB that does mot form solid solution© with ben%olG aelà^'" p hence th is method should ale© remove the final traces @f rmtero

Demlga of Zoa© Befimereo

Two simple sono m i l n i n g , aeohaniBme were evolved during th is period of regiearoho Th© f i r s t mechamisH 1© Bhmm in fig 4- « It Im extremely simple aifxd consisted eseemtially of a eorewad pulley wheelp to which a length of copper string was hragecU fh© other end of the string was fastened to a fixed spring and heatere were attached at suitable equally spaced intervalso The actual, mechanism is ehovm Im fig 4 After one revolution of the. wheel th<% string screwed off suddenly and the heaters w©re returned to their initial position by the tension in th© springo This gave a vertical displacement ©quÆ to th© circumference of the wheel mmd thus the heaters were separated by this distance to glv# the required overlap of the zones» Th© pulley wheel u#@d was 2 o5 am» dla» and the speed of the drive motor was

^/S r»pohoj) giving a son© refining speed of 1 ' cmo/hr» which wma quite suitable for refining organic aoliteo Th© heaters used were wound with 33 ohm/yd » chrome ^ wire m glass formers insulated with asbestos paper and the heating current was controlled through a “Ya'Wiac®

&utotrenef0 rmer» several sizes of heaters we re constructed for use with tubes of varying diameter ranging from one to three ce» Attempts were made to zone refine benzoic ecid using two he altera but it ws^ found that the glass tube containing the acid always fractured due to the large volume expansion 40 on meltingî, $/Sm Md even two msu thick walled glass tubing could not withstand the pressure created» 3ome success was obtainec by using Teflon tubing to contain the acido This tubings which is suitable for use at temper-- aturea over 2 0 0 ^0 »g expanded sufficiently to contain the melto Unfortunately the heaters frequently ©tuck to the tubing and melted it at the point of conteoto It was decided;, therefore^ that the zone refining must be carried out in glass tubing in which the acid could be melted^ degassed and sealed imder vacuum* Thia meant that only one heater could be used as it had to start at a free surface for every pass to avoid fracture due to the volume change on meltingo I t was found that even with one heater the tube occasionally fractured if it was exposed to

PULLEY WHEE.L.

VAR I AC.* AUTo Transformer

^ "TEFLON

MOTOR VS

s e c t i o n a a '

FI&.5, AUTOMATIC ZONE REFINER. :i: i draughts or suàâeE -yarlatioEB tu ambient temperature which altered the mone leBgth^ hence it wa8 placed m a draught free pesitiono Thia zom refiner was not particnlarlj suitable for a single heater unit aa th is meant that the heater must travel the fall length of the charge and for a length of 30 cUo K 2o5 cliaos. giving a reasonable chax'g© of 200 the wheel diao required was 10 cBo and the motor speed "^/30 r«p«ho At thia time an Improved design of refiner was evolved flgo S p in which the speed and path length could be easily variedo This refiner wa® again of simple design

0 jM th© principle ©f operation is shown In figoSbo A Sangyamo Weston motor^ ""/B ropohop drove the lower gear from which severml teeth heel boon removedo This gear drove m seooni gear wheel attached to a spindle of variable dimmetero A cord was att.aahe

Tm gearing used was in the ratio 2 si so that the spindle turned twice during every aona paaSo In this wav a total 26 n pass length of 30 em» was obtained» The gome refining speed wsia controlled by varying the electrical input to the motor through a "Simmeratat“ eontrollero Im this way gome refining speed© of 4 * 5 cmoA^ra and under were obtained» In some refining bengoie acid th© procedure adopted was to use a heater giving a 2^3 cm» molten gome » The total length of charge was 35 cm» and th© tuba diameter 2 » 5 cm»

3 or 4 faet passes were initial.ly made at 4o5 cm»/hr» which served to remove the major impurity aa a brown band of solid which was quickly rejected » This wee then followed by a mlnlmma of 20 passe© at 2oO CTo/hro Within two ©r three passes th# acid always started to crystallise as lasv,# aii:%ular pieces^ a sign of high purity» 25 passes wer© always given to ©neur© complet© removal of im purities» Two methods of filling the gone refining tubes were iisedo Initially the gone refining tub© was broken open and the top inch and lower third rejected » The remainder was then placed in a large tube glass blown on to the ©nd #f a crystal grov-ing tube and connected to the vacuum line » The crystals we.i© then evacuated for hFIf an hour before aomitting a l i t t l e nitr©o©o and melting into the growing tub©» The melr was agsdia degassed before admitting 1 0 cm» AjitrO(^©n ana sealed » At a la te r etege th© crystal gi owing tub© was attached directly to the gone re finding 2‘

tube as lu fig» S! » The upper inch of crystal, was alway B melted.into the top of the refining tube before the subsequent material v;es melted into the growing tube »

Xi-5o Design of Benzole Acid Crystal Growing Ovens»

The basic principle in constructing an oven suitable for growing crystal© by the Brldgman-ytockbarger method ia to have two well lagged fumaceso one thermos tatted above the melting point of crystal ano the other thermostatted below itp separated by a baffle in the middle of which the melting point ieothermaX should lie» The major factors controlling the growth of large single crystal© by this method are»

0 , 0 The thermal stability of the oven » bo The temperature gradient ^t the freezing transition» Co The ret© of lowering the crystal through thie

tra n sit 1 ono do The development of & suitably oriented seed crystal» The most common oven design is to have two electrically heated wlre wound furnaces» This is probably because of their widely variable temperature range and the fact that initially this method was used to grow high melting point crystals of matalSg ionic salts end semiconducting materials »

SLOW SPEED MOTO^,

/Z3.

ÊUNVIC THERMOSTAT TT T 5 I, AWDEEMAW 40W EULB, STIRRER

"COSK INSULATOR.

€

g )

FIQ.G, BENZOIC ACID CRYSTAL CROWlMCi OVEN, 28o O A© bengolc acid has n fairly low melting pointy 122 Co aaa bb wire wound furnaces do not give a uniform temperature distribution throughout their length it was decided to construct a crystal growing oven in which a much more uniform and stable temperature could be maintainoa above and below the baffleo To accomplish this thermostatted baths o f liquid were used to obtain the two temperatures requiredo This type o f oven has been constructed by Hood

5) „ , and Sherwood and found satisfactory for growing a low melting point organic ®plastic° crystalo The design of this ©van is illuatratad in flgo é o

The lov.er bath was encloaeo in a 1 ° 2£ 1 ° % 1 "6 '® deep plywood boza A large metal drum fitte d neatly in this box and jT^a© insulated with etraWo The drum was with with a layer of paraffin to suppress evaporation^ and thermostatted using a toluene regulator to c o n t r o l a

50Qw heating elemento Th© lid o f this box waa a 1 “ x 1 " x I’’® thick plywood boara to which a piece o f cork of similar dimensions was gluedo This gave a 1*® thick insulation layer of sufficient strength to support the upper foatho A two -litre beaker which effectively formée, the lower bath was boltec. to th is lid and wejs also fille d with we te r having a paraffin layer on top * The upper vessel consisted of a copper can 5^^" x 10" diao irkth a 2" dia^ hole drilled off«centre in th© base^ through which a tight fittin g pyrex tube was inserted and -aealed with ®Araldit 0 ® epoxy realm * Holes wex’e blown in the lower section of the tube which projected into the two litx ' 8 be alter to allow the lower bath t© attain rapid thermal. @qullibriUE?u The upper bath was fille d with glycerol and had a silicone o i l layer on top to reduce evaporationo A background heater was wound romiü the copper ean^ insulated with asbestosg using 33 ohm/ycU chrome ® wire and lagged with dia« asbestos ropoo It was contr©l.led using a

?ariac to give a bath temperature 2 0 ^ 3 0 ^0 o below that 'requiredo The final temperature control was obtained using a Sunvi© T8 l bim etallic atrip regulator which controlled a 40w Osram bulb giving a temperature stability of t OolO 0^ at 140®Go

The o[ 2>f5 PmU(fn rate of lowering at which ©quilibriu® freezing conditions are still maintained ia a factor which depends an th© materialo Metal single crystals can b© groxm at rates @f several cm^/hro *" whereas anthracene and naphthalene require growth rates less than 2 mmo/%ro The rat© ia aleo related to th© temperature gradient at the freezing point aa the heat of crystallisation must b@ diesipated through the'crystal buü this Increaaea with increasing temperature gradients In the case of benzoic 3 0 . add th© determined by trial and ©rror^ It %mB found that a lowering rat© of 2 oO mmo/hro was satisfactory with a temperature gradient of 20 C^/oeIo The growing tubes wer©p therefor©@ lowered using a SangaTO Weston e lectric motor rated at /I 2 ropohc for a 3 0 ©«poSo aoC« supplyo The 50 CopoSo Ifeiboratory supply was found to increase the proportionally to j-‘/B1 ropch^ The spindle diao was

Go3 2 oiBo and this gave a lowering rate of lo 7 mm very fincp strong cotton thread was used to auepend the growing tub© « It was found that th is occasionally broke in th© hot section of th© oven asid was replaced by a short length of copper “string® which was tied to th© thread above th© hot ©actiono This lowering mechanism was Independently ciampeo from th@ rest of the even to minimise the effect of any vibration caused by the stirrero The problem of the be&>t temperature gradient to use is one in which there are two achools of thoughto A ©harp temperature gradient i© considered to give better and quicker dissipation of the heat of crystalliaation ami increaaes the positional stability of the melting point iaothemalo In favour of a shallow temperature gradient is the fact that the crystal can anneal aa it is formed and will not be subjected to the stresses which occur on rapid coolingo These stresses and the subsequent non-©quilibrium defects they introduce ean^ however^ be annealed out by

TEMPERATURE 6 RApliWT IM CRY&TAL ë^©W)Wë) OVEN N®. S.

1 % L 3 MM: #gM&@OC A&OE

Tm#R » © 8 i 6^ r #

© C

&P §p &APFC ®

zc%S-Rm3sr.a=7&zGzrz;%:rzA=izwJ5=5==&=:T=%%%=)y;?i35æ35%K=æ=%%;%o» f§ï, 0 ^ 8 ''à '^■’4 -fA,

T@MP@^#TWRg m A U m t m AêiVSiC. A O a CRvSmiL

6 #.@WW# #V@M

zpia i^,___

"H t KMP (Œ#im(&(ieik%ir m t

® mU^fi C i©

O

\S3 “H0 jijijisis^sactsïïïSD C3

1 ^4:V; : 8 % srzT;;:TK:3:;% ::'^c^"«'==:::% a;3=^.ïa;'Kz=^.3T:=2% A=y===2&::z% =r;zzza33$Œ ü&3=a^ ^icncîësœiïîÿKSîSjœr^œrrîXz::^ %! o reheç-tlBg the crystal to just below the melting point for a suitable time period followed by gradual controlled eoolingo In this investigation the crystals were grown with a sharp temperature'gradiento The temperature profile -aoroBS the baffle la shown in figo T « The upper bath was ■controlled Bt 20 0^ above the melting point and the lG%er bath at 60®Co» 6o below th© melting poisito The temperature gradient ©btalA'ied ecroaa the baffle waa 20 G^/cmo and was found to be benzoic acid crjatalBo The upper bath temperature waa lim ited t© a. mazclmum of l^O^Co as Schwab #%d W icher^ found that a alow ozidatlve dégradailorn occurred at 2 0 0 ^0 o and that a reversible dehydration to benzoic anhydride and water occurred in th© 1 iquici phase o Th© extent of this déhydration was small 9 howeverg and only occurred when th@ acid was very dryo It was found to return to its initiai value when th© material, passed through th© freezing transiticMo When this crystal growing oven was operating properly good crystal boules were obtained which usually consisted of two or three large single crystal pieces growing with the cleavage plane vertical. ®r at an angle of 60 ^ to the horizontala Onfortarately the ©yatom could not always b© relied upon and several factors which Influenced its

IS o

FIG,:8 . THE CI^YST/AL 6 /?0 m N 6 '.:.0 V £N . 32.

a. Th© ©tirrer in th© high temperature bath had a tendeaey to ©tick ocoa©ai©nally &nû wh©xi thi© oceurred tho crystal vwB restartedo bo Th© ©tirrere alao ©ceaslonally caused ©xcesalv© vibration uhioh affected th© crystal gr@?vth processo Co The cry a ta], could not b@ observed during the grov/th process and the crystallisation had to be carried out completely before it was known whether or not the seeding had been successfulo do When the string broke or the crystal slipped off the hooks, and th is happened several tlmesj, the oven had to be completely stripped down to remove the tubOo A period occurred during which for a combination of the above reasons unsatisfactory crystals were obtained. At this point it was decided to change the crystal growing oven in an attempt to obtain better crystalso The new oven used was one constructed by Dr. Jo Mo Sherwood and successfully used to grow anthracene crystal©^ ffi.pto 217^0o This oTOHj, figo # was an a ll glass oven mad© in Pyreji which enabled the crystal to be observed at all times during the growth precesso A detailed S3 description of the m®n has been given elsewhere and only, th© principle features will be described. Two “Kichrome'^

G. 4 tYSTAL Ê® heater â & B of resiataiAcea 80^ and 130./ure8peetiTOlj xmm ^ound am pyresc glaa© tubes and ivere adjusted to temperatures 10 - 20 0® belo^f those mqulredp (upper heater 130^0 low r heater 80^0« ) wing rheostats in-aerie© with the line supplyo k heater Dp resistmace 100Ap was wound ©n a conoentrie pyrez tube 2*6" diao mid the ©ven was raised to th© required tamiperature end thermostated by aontîfolling the supply to this heater through a ®Sun¥ic® oircuit f 102/5#o whioh used a 8unvle TSl normally closed thermostat as senBing elemento The detailed operation of this circuit is given elsewhere Aa with the previous oven the temperature gradient was adjusted to 20 G^/omo at th© interface but the lower heater was controlled at 95^0^ to avoid the larger temperature drop in the other @v©no The rate of lowering with this oven was 1 mmo/hro and good quality cry stales were obtained when high purity bensoio acid was usedo

The Crystal Growing vessels «

Th© first attempt at crystal growing was mad© in th© vessel shown in figo 9 o A poly crystal line mass resulted ^ however^ containing some singular pieces which were observed by examining piece# of the crystal bouJ.© between 34,

crossed polaroida where four poeitiens of complete ©Etlnotlom were obtained« This indicated the neceasity for a better seeding arrangement and the second cryatal was grown in the vessel shown in îig B h im which a short piece of bent capillary was used t© seed th© eryatalo This crystal was considerably better than the first and gave several singular pi©c©s which had the (001) plan© growing verticallyo àm there were s t i l l too many crystals formed the TOsael shown in flgQc. was adopted for aJ,i subsequent growing vesmelao This veesely which has been used previously S i to grow anthracene crystals g gave aa ln#ulatlmg layer of th© material in the outer vessel which helped to stabilise th© melting point isotharaal in the Inner vessel and was found to Improve the quail% ©f th© crystal obtained eomsider^ ebly o It was net p©asibl©p however^ to obtain a moaoerystal and usually two or three large single crystal pieces were obtained which grew the whole length ©f the tube and gave excellent cleavage and extinction between crossed polaroidSo figc^d was yet another variation ©f vessel C which ‘vap used in th® later stages of crystal growingo Attempts were made to grow crystals with horizontal eleavssg© in the tube by m alterirag th© angle of the capillary « It was foumlp however^ th a t th is had no effe c ts c ry s ta ls were obtained with cleavage plane vertical, and at an angle of 30® and 43® to 33

the vertical irrespective of the angle at which the tube Yjas bento The fai-ct that the cleavage plane could be groi-m at an angleg howeverg Indicated that the growth rate waB Blmi enough for equilibriu© freezing conditione otherwise a vertical cleavage would be obtained along the direction of fastest growths In all the erysta 3.8 grown th© cleavage plane obtained was the ( 0 0 1 ) When crystale were withdrawn from the ©ven and allowed to cool naturally they always fractured due to the thermal strain induced^ When th© erystala were withdrawn from th© oven they were, therefore^ placed in a large^vmter filled dewar flask at the same temperature aa the lower bath^

stoppered anc^alloweci to oool naturally ^

4% PrsparatioB of Dopwé Benzoic Acid Crystalso

Attempts wer© made to alter the defect structure ©Î bensoic acid by introducing known impurities Into the latticeo It was thought that the introduction of cations into the lattice would introduce defect sites in the acidic hydrogen lattice which would increase the conductivity and tritium diffusiono llfobelohde ' ZTleû to incorporate sodium ions in benzoic acid but found no effect on the conductivityo It was thought that lithium^ being emalXcn% wee more l;Ucal;y to enter the lf»ttice subatltutienallyo ân attempt^ therefore^ was made to grow a crystal doped with 1 g1000 parts by weight of lithium benzoateo The crystal obtained was completely polyerysialllM iadicating that th© lithium did not enter t.he lattice sul)sxitutiona3.1y and no further work was attempted on this ays toother crystal wbb grown with p-^terphenyl bm impurity « p-^terph©nyl is sim ilar in shape to a benzoic acid dimer but slif^htly shorter and it was thought thsît it might enter the lattice aubstitutionally and cause a loosening of the lattice, around these aubstitutionaX impurity sites with ite subsequent effect of increasing the rate of self diffuslono This crystal was grown in the normal way using zone relined benzoic acid to which 0oI05 gmso of scintillatlorn grade p--terphexiyl (supplied oy riOaerBo A^iuclear toterpriees LtcU) had been addeo^giving an impurity concentration of ls 6 0 0 o This re the r high initial eoncenc.xation was used as it was not known to what ezctent the p-terphenyl wouIq be incorporatedo When the crystal vmm removed i t was foimo to have a horizontal cleavage a phenomena which had not been observed with any other cry stale The crystal was trMsparent with a small ring of brown material rejeoteo at the topo Th© crystal was not^ however^ completely singular as only partial

F U Y S P f f & T R A

O)

m w

rJtowziiîSWat®?* ZWWRWktW** >wàfto ■.. O 37. extinction obserfea between crossed polaioldSo

The distribution o f p-^terphenyl throu^jh the crystal wne obtained by measuring the bcVo absorption on a Perkin Elmer 137 U oVo ©p®€trophotomete3To The p-terphenyl concentration warn obtained from a calibration graph of p'-terphenyl In benzole aeicU P-terphenyl ©schlbits a broad absorption in the UoVo at ^ max 275 with € max 3500 in hexane BolntiOBo Unfortunately benzoic aeid also absorbs at 275 but gives a much sharper baoci and the p-terphenyl can b© determined from the overlap^ see figc0© o A calxb^ration graph wae conatrucT.eo using p->t0rphenyl/b©nzoic $cid. solutions in hexmx® 0 The value© obtained are tabulate below for a benzoic PC Id concentration of lo3 B^gsi/mlo taking the absorbance at

p-Terphenyl absorption at 300 mj

SAawPLE ABSOHBAÈCE COÈCEAiTHATIOB ml o 0 4o7 % 10” 2 OoQl 2ol5 X 10”^

3 0o25 2.5 21 10” 3

4 0o44 4o9 K 10-3

5 0o55 6o9 K 10"3

6 Oc65 8 .9 K 10” 3 Pour samples of p->t©rph@nyl doped benzoic acid were taken at various positions along th© crystal£, their U«?^ absorption

VAC,

BI4.

CRYSTAL GROWING TUBE.

ZONE REFINING . TUBE. SOOmt.

P(6,||, PREPARATION AND PURIFICATION OF DEÜTERO-BEN ZOIC ACID. spectra üieasureci aM the eoBcexitratlüa ca3.eulat©d from the calibration graphe The results are tabulated belo % 3

SâMPLE POSITION TOÏGH1 üü^OBBAIûCK p-Terphenyl (Bas®ao) agm/25ml . eonoentratioB mgm/ S fflo mgfâp Boâe

1 0 IT. 74 0.02 2„8 % 1 0 “ ^

2 2.5 16.04 0.04 6o3 % 1 0 ”^

4 . 3 5 . 5 14.77 0 . 0 9 .15 oO %

4 8.0 14.15 0.95 3 5 0 % 1 0 “^

The ma j oritur ©f the p^terphenyl there Pore g rejected althemgh a small amomt appeara to have bmmi inoorporatedo

Bouterofeen^oio Acido Im order to study the isotope effect on tritium diffusion iii benzole aoid a crystal, of benssoie acid-d^ was grown end the deuteration was carried out by hydrolysing bengioi© anhydride with deuterated water following th© procedure used by g-f Robertson & IIfobel@hd©„ 120 gmso of reagexit grade beMoic anhydride (Messrso BoDoHo Ltdo) Bopt^ 39o5^Co was purified by 10 paasea on a gome refin©?o Th© upper 100 gmOo ©f this purified material^ Bopto

4 2 - 4 3 ^ 0 oj) mm melted into the robe flask Im fig Ho 20 ml* of SSolp C tCJ. iW) ^ a three fold access^ was introduced

and 4 drops of freshly dijstilled thionyl chloride added t© eataljse the hydrolysiso The tube was aealod at B-j and opened to air through a Ga Gig tube to prevent entry of normal water vapour into the apperaluso The mixture was feflused for 8 hrs, using an isomantlaj to allow complet© hydrolysiso The exoom DgO <3aB removed on evacuation and trapped out using a liquid nitro^.0 n trapo The bensoia acid was then sublimed over^under vacuum,into the 2^ long tube above the gone réfining tubeo The vacuum pump was turned o ffg a l i t t l e dry a ir added and the sublimed acid Bielted down into the son© refining tub®o This procedure was repeated until the gone refining tube \ibjb full g approximately 80 gms acid 9 over 0 period of 4 hours o The materiçil in the gone refining tub© was degassed twice in the melt before sealing at Sg ujader a pressure of 1 0 cm« dry'^'aîro The material, had a very pal© yellow colouration at thia stagCo Th© acid was son© refined 9 50 passes at 2 cm/hro Large transparent crystals were obtained with a mustard coloured Impurity being rejectedo The upper inch of material was melted into the top of the tub© and the subséquent 6 0 - 7 0 gmso melted into the crystal growing tube and sealed at The crystal was grown In the all glass oven at a lowering rate of 1 mmo/hTo The crystal obtained was perfectly transparent^ gave a well defined cleavage plan© at 30^ to the vertical and gave good extinction between crossed poloroidSo Th© crystal was annealed at 120o5^Co for one weeko Th© melting point of

la crystal was lI 9 c3 ^Goj, 3 C^ below that of normal b@ngolc 40 n acid and In good agreement with the value ef Hobertson & lIfob©lohd©«

Annealing of OrystalSo All crystals prepared from .‘sone refined material were annealed for one week at 121^^0o These crystals were designated

ZRlp ZH2 etc a with the exception of th© p-Terphenyl doped cryatalp p-Ter^ and th© deutero bensolc acldg DoBoAo The non ?sone refined crystals were desl^^neted kOB ZoKo

Parity of Bensoic Aeldo Although th© method uaecî to prepare the benzole acid should have ansured a purity 9 9o99 mole ©n attempt mm made to datermino the impurity by 2 -

ao Measuring the freezing range 0 bo Measuring the water eontentn Go Estimation of metallic impurities by motivation analysis

8 . 0 The l?m®zing Hang© The freezing point of benzoic acid for us© as a standard in thermometry has been determined by Schwab & fichera in an exhaustive aeries ©f axperimenteo They found the triple point of pure benzoic acid to be 1 2 2 « 3 6 2 ± OoOOS^Co and 0«03,3® higher under a pressure of 1 atmosphere of dry air» It muB impossible to duplicate their experiments aa they used 400 gme ©f acid in eaeh fixed point cello Such a large amount wae net

RECIPROCATING, MECHANISM. AMDEIRMAN STIRRER

STIRRER, C3,UIDE-

TO WHEATSTONE BRIDGE

TO VACUUM,

TEFLON qUIDE BLOCKS, j CORK SUPPORT, ___

EVACUATED JACKET NtCHROME WOUND HEATER^

STAINLESS STEEL STIRRER,

THER MISTO R r—-,^

DEWAR FLASK.:

P((q, 1 2 , FREEZING POINT APPARATUS, 41c available aa ©ach crystal gromi oîaly weighed 60 gmac àn attempt wae made to construct an apparatus in %'hich the freezing rang® of ©ach crystal used in a diffusion ® 3cperim©nt could be measuredo Th® fin al form of th is apparatus ia shown In figo 0&. a It consisted of a small sample holder^ sample volume Oo5 ml0 3 attached by a B14 com© and socket to a long glass tube in which two tight fitting teflon inserts acted as guides f?nd supports for a Stmtel. F15 thermistor and a platinum wire atirrero Th© stirrer was adjusted to give a vertical displacernvmt of 1 mslng am eccentric drivec This tube warn surrounded by a glm^m jacket which cou3.cl b© evacuated a È small

chrome ^ wound heater controlled through a 2 amp ¥ariac was Med to melt the sample and the whole assembly was thermoistatted in a glycerol filled Dewar flasko The procedure used to oeteimime the freezing point wa® to maintain the outer bath at 75^Co ana alJ„ow the aBsembly to equilibrateo Th© crystal was then melted using th© the s tir r e r started and th© temperature obtained by meaaurlng; th© thermistor resisstanc© on a Wheatstone bridge o The outer jacket was then evacuated t© minimis© th® rate of heat e^tractlomo The variation of temperature with time waf notedo The thermietor was calibrated against a platinum reaiatanc© thermometer and the temperature we.a accurate to ± OoOl^Co

Attempt ,9 were made to ©easur© the freezing point of sample©

©f several single? crystal boules 0 Th© freezing points obtained^ 42,-.

were always lower than e^xpeeted and I t was observod that when the sample was remelted and redetemimed the freezing point became progressively lowero Oogo a aample of boule

ZBo4 gave melting point© 122<,;;0^Co 3 X22o20^^Cop 122ol7^0o It was comclurled that this progressive decrease in freesiimg point was due to absorption of molature in the melt as bomsîoic acid Im signifleaBtij hygroacopic im tha meXto (Schwab & Wichers found a OoX5 C^ decrease in malting point on ©guilibratlon with water vapour at a partial pressure of IB m\o Hg) I t was considered that the effect of th© stirrer would Increase© the rate of absorptiomo On consideration of this fact and the difficulty of obtainlag accurate and reproducible results due to the melt Bupercoollmg to a variable extent in each cleter-^-- mlmatiomp thia method of purity deteraMmtlon waa abandoned in favour of the following more accurate da te rmlnatlom o bo Determination of Water in Benzoic iicido The normal method of water determination in acid© msimg 58 the Kerl Fisher reagent, was considered but was rejected a© the amount of water expected was less than Û0I mgm per 10 gmSo Gas-llguld chromatography was considered and a special column of Cartewax 400 on Teflon 6 was preparedo As the benzoic acid was solid it had to be introduced im eolutiom and mo water peairlould be detected sis it was masked by the broad solvent hamdo The Biethod used to determine the water com tent of benzoic acid was that used by Schwab and to determine the v o latile impurity in high purity bemsoic acid which has boen MR AN! ÆÜU

iff

ME I# W #:R IXTERMIMAT 4 o ahov/n to be almost entirely imtQVo The apperEitus is shami in tigo 113 o Crystals of bensoic aeid '^em placed im A aad degassed fo r i: hour to remove any adsorbed molatur@o Schv/ab & #lcher,$ h@,v© demonstrated that the water absorbed by three samples @f beMoic acid at relative humidity at 23^0« was Oq0004s> 0d0004c OoOOOO mole p on exposure for 45 dayso The acid sublimed in A under continuous vacuum and the volatile Impurity trapped la the “IP tub© with liquid os^ygeno This tube was calibrated between the two atopcocksp 120^1 mlog and included a Pir&mi gouge head (ISdwarda & QQo model GHl) calibrated from Ool micron to 1 mm.o Hgc When ©11 the acid had been sublimed taps g Tp closed and the tube brought to room temperature using a water batho The volatil© Impurityp assumed to be water^, ©xerted its vapour preesur© and provided it was less than the saturation vapour pressure at room temperature the amount of water could be determined knowing the pressure and volum# of the enclosed ayetesio Blank experiments were performed to determine the pressure increase caused by degassing the sample holder during sublimâtloBo The pressure in the aystern after degassing was always lo5 - 2o0 micronso Two blank detem inations increased the pressure t© 20 micronso The volatile impurity in three crystals was determiiaedg a non gone refined cryatal^ a zone refined crystal and the deutero benzoic acid crystal « In all cases the finaJ, pressure was less then the saturation vapour _ 44 pressure O I t was however j, that on standing th© preasur© slowly decreased due t o re-adsorption on the vmllB of the TOsaelo The impurity contexit vjbb therefor© oa3,oulated from the maximum proaaur© observed^ The weight of orystalp pressure inoreaee ami maximum calculated impurity oomGemtratlorn are tabulated below«

CKÏSTAX. WTo SAMPLI MAXIMUM MOLE !/i> ÎMPOHÏÏÏ 6 MS. PHESS5ÜRE ASSüœB TO BB mCREHSSi WATER HICROËS^

iio^ i . , e . 3=060 120 0 .0 0 3 4 Z.Ro'l 5 .1 0 8 90 0 ,0 0 1 5 •DoBoAo 4 .7 9 4 110 0 .0 0 2 0

Oo Metmllie ïmpuritie^Q â3.th©ugh the known m etallic im purities were lew^ F© 0o0002^o Pb 0o0004^ alk ali metal im purities were imlmown and &# these were determined by activation ana3.ysiSo Samples of nine ©ingle crystals and one o f Amalar Benzoic acid^ ranging from Oo3 to Oo5 gmBop were activated for om© hour in a neutron flm o f 10'"^" nopo em" (at the Scottish Hesearoh Beaetor Centro^ East Kilbride) o Each mample was eubsequently ©Ba,lys@d on a hsbm Multiohaunel Pulse Height Analyser aM a ©peetrum o f ©ach obtained om 126 ohaumela in the range Ool 2o5 Me Vo In a ll caeea the |* peak at lo3? M@V was obtained along with it© assooisted background ©nû photopeak at the lower md of the raug©o A small peals waa obtained in the anal a r acid spectrum at lo6 - lo7 Me? presumably due to 01' produced by the reaction In the single crystala^ however^ thie peak was umi existant or only just detectable indicating that the chlorine concentration wma extremely lowo The absolute concentration of Ba was calculated using the following Information^ The total counting efficiency f o r Ma ¥'a uml^g a 3" X 3" Mai crystal is 28p. with a peak to t o t a l rati© o f Oc37o The sample wee taken to be 3 muu from the crystal 1 p surface * The saturation activity for a 10*” neutron flux was é>! ê)@ 590 mc/gffio Mao The h a lf-life of Ma was 15o0 hra,, Using the above information all the samples were found to lie within the rang© l->5 pop^mo Hence i t is assumed that a ll other TOtallio im purities are as small or small©ro

X, é> Growth of Acetic Acid Crystal,Bo As in the case of benzoic acid the Bridgman technique was wml to grow single crystals of acetic acido The growing oven usedp however y had to be re-designed because of the low melting point o f acetic acid (I6o7^0o)

Crystal Growing Oveno The oven used was sim ilar t© the type designed by Hood and Sherwood for their experimental growth of cyclohexane single Grystal»eo In principle this oven was identicaJ. to tha,t ahov.n in figo 6 o the only difference being that the lower m@tal drum was replaced by a 15 li tr e Dewar vesselo The lower temperature bath vmm obtained by packing the Dewar with 46' an aeGtonQ/’‘l)rih0ld“ Blnrrj which maintained ;;. constant temperature of -78^Go for several û &jb mû then alowly started to risoo The effect of this slow rise in the lower vessel wiqs only to slow down the rate of crystal growth m the position of melting point isothermal was low©redo The water in the lower 2 litre beaker was replaced by acetone and the glycerol in the upper bath was replaced by watero M© background heater wab required for the upper bath as it was thermostatt©d at 30^Go The temperature gradient through the Malting point Isothermal was feeasured using a calibrated thermie ter and shmm in figoT The igradimit used waa lÛC^/oâu The crystal growing vessels used were those ahoivn in figoA c. & Û and the lowering rat© used was lo 7 mmo/houro

Purification and Crystal Growthe The sta rtin g material, used was “analar® reagent gracie acetic aoid (Mesarso B^DoH« Ltdo) mopt„ above IGcO^Co (Lito 16 o 604^^0 o ) In the first attempt at crystal growth the acid wae simply dietilledy the fraction^; bop to 1 1 7 *^G« being collected in the growing tube ^ The liquid was degassed three times œ a vacuum line before s©aj,ing under a alight pressure of dry nitrogen^. The crjstaL obtained cons:Leted of many polycrystalo oriented with a vertical cleavage plane miê. Indicated that swingle crystals could be obtained by methodo 4îo

An attempt was made to purify th e acid by zom refining in a deep freeze at -20®0o u^ing the zon® refiner in figo 4" The large expansion on melting fractured glass tubing and teflon tubing was used to contain the acido The a c id was p u r i f i e d slowly by this method as the melting point was raised from X6o2 to 16o55^Go a fte r 30 passeso This method was r e j e c t e d in favour of the following purification p r o c e d u r e recommended by Wibergo As analar acetic acid may contain som e acetaldohyde and other d i s a b l e components Ik litres was reflux with 2p chromic acid ©xiû distilled^ 1 litre of th e middle f r a c t i o n g bopto 115o5 - 116o5^Cop being retained o Thia p introduced som e water a s an impurity but thia was rm distillation after treatment with acetyl b o r a te which reacted with the water present to give acetic acid and boric acid which wa® non volatileo Very p u re acetic acid was o b ta in e d by thia methodo Th© acetyl b orat® was prepared by heating 1 part Boric acid/5 parts acetic anhydride for one hour a t 60^Go On cooling the acetyl borate crystallised out and was filtered off an washed with acetic acid to remove e x c e s s a£ihydrid©o Th© chrciiic distilled acid and the acetyl borate were stored separately until requiredo The crystal® obtained using thie material w©r© of good qua3,ity but s t i l l p ro d u ced several single crystal pieces in th® bauleo The cleavage plane was a-lwaye obtained verticallyo The crystals gave excellent extinction between crossed p o la r® Id : and melted at I6o7^G« An attempt vsbb made to meaaure th© water content o f these crystals by gas liquid ohroraatographj o The column uaed waa 10^î> PoBoGo 400/Teflon 6 on a G riffin and George Mark ÏIB 0«L«Co with a kertharomater deteotoro Due to tli© ineensitivity of th© machina^ however^ th© water content could only be determined to be lees than ls500« the erystalB were withdrawn from the growing oven they were broken open inside a polythene bag whioh im© sequently placed to a deep freeze at =20'^Gc The crystals were then removed andppiaeeâ in stoppered glas^si Jare and stored in the deep freeze until requiredo All hMdling of the eryetala wae performed in this imito C r y s t B1 31 rn e I u re The acetic &cld cry stills g?ve one well defined p easy olervage plf'ne which from theoptical and cleavage studies of Steinmeiz and the zreray structural analysis o f Jones and Templeton concluded to toe the (100) planec Unfortunately this is pan llel to the hydrogen bonded chainso As diffusion studies %ere to be carried out down the hydrogen bonded chaîne attempts were made to cut the crystals perpendicular to the cleavage plane», The cry at els g however^ were so hard in this direction thr-t it was very difficult in th® direction of growth and almost impossible at right angles to i'U The crystals always split along the cleavage plana when this was attemptedo As it was impossible to diffuse in the required direction the diffusion and conductivity were stuaied perpendicular to the elejvrge plana « CHAFTER II EIEEÜSiÜE ANE GONEUCTIVIXY STUEIIS 49,

Ch. H - THE EÏFPÜSIOM EK.PEXÏKiEMTS .

Introductlorn to Diffusion

Diffusion in solid© is 9 kinotie phenomena which can be treateo as B macroscopic system without knowledge of the atomic nature of the diffusion prooeaso Diffusion la governed by fundemental lawa which were first applied by Adolf Pick 6g and are kaovn aa Pick “a La?i©o The first of Pick “a laws relates the flux of matter^, J.; across © given plan© to the inatf-ntaneoue concentrrtion (^radient et the plm® mé perpendicular to it I ' . . /. /' \ tj —D W 'G c=.fcT-,=i.c:t,ei.ra«ra

where D'ia the constant of proportionalityp called the diffuaioAi coefficient and has dimension® length'^ aeoT'*^ Diffusion problems can be solved by equation (1 ) only under steady state conoxlxon® where is a constant,) In many diffusion studlesg however^ the concentration gradient varha with time and inis effect requires the introduction of Ficks second lawo o ^ C -p. f ^

assuming constant Where D 1b not conBtant then the second law becomeso 90r The BoXiition to all diffusion problama can ie theory •be found by applying the appropriât© boundary oonôitions ena solving equation ( 2) or ( 3 )« Usually In self diffusion studies the diffusion coefficient i© ecmcentration indepenaent^ BÜÛ QUÛ diffusion Biechemlam predominate©„ Im the ease of organic crystal© the po©sibility of diffusion anisotropy arises due to the low aymmetry of most organic crymtalBg usually raoMclinic or trieliaiCo This will complicate the diffusion kinetics by introducing 3 diffusion eoeffieieats Dxxp Byyg Dz% along the y9 mxx b axes of the crystal o Previous studies on anthracene and fi'-mphthalerae have shown no aignifieaKit variation of diffusion coefficient with crystallographic axe© hence it Im initially assumed that diffusion in these acid© can be treated as isotropic until measurement of the anisotropy have bean madOo There are two common method© of measuring diffusion coefficients in solidso 1) Mon destractive 2) D estructive

■ Im th© non deatr-uctlve method the rate of penetration of the diffusing species la obtained by monitoring its change of concentration outside the solid either by chemical anelysl© in the case of impurity diffusion or in the ease of self diffusion the activ ity of a radioactive la©tope can be measured 51o 6 7 a© in the “surface decrease’ methodo The trouble with these methods is that they do not d ifferen tia te easily between several concurrent diffusion processes and only an overall diffusion coefficient can be obtainedo 2I In the destructive method the solid is sectioned and a concentration p ro file of the diffusing species is ©btainedo The power of thia method lies in the fact that concurrent diffusion processes can be observed9 though not necessarily measured^ and the diffusion coefficient can be obtained to a high degree of accuracy if the experiment is carefully performedo There are two sectioning methods normally used in

1 ) Diffusion in an infinite cylindero 2 ) Diffusion in a semi-infinite cyllndero

Diffusion in m infinite cylinder^ In this method an Infinitely thin deposit of the diffusing species la placed in the middle of two infinitely long cylinder thus

" X This diffusion couple ie then allowed to anneal for a time after which the cylinder is sectioned ©n either aide of the % interface end th© concentration of the diffusing specie abtainea m a function of distance 9 îvom the initial deposit at e;^©o The ©olution to Fiokto I m for these boundary conditions 1#

e(K j t ) B%ip /4 D t e» where 8 is the total emoumt of diffusing specie© present hence imowing the variation of 0 ( 2 ?:) with the diffusion coefficient

ûm he calculated o Biffueion in a semi^infinite cylinder i© variant of the above method in which am lmfimit©ly thim deposit is placed #m (me end of an infinitely long cylindero

and allowed to diffuse as beforeo Im this case the solution to

a & % OlK exp /4 D% catzncao b, ioQo twice the comcemtratiom ©btaimed im the in fin ite cylinder methodo The advamtage© of ueimg the imfim ite cylimderp are th a t a check on th© diffusion coefficient Im given by the symmetry of the comcemtratiom profileoo This method aJ,a© prevents surface d iffu sio n mû evaporation losses o Im général it i© a good method where diffusion coefficients are known to he high for the Kirkendall effect im which the initial boundary is found to movoo This is only found in impajirity (chemical)diffusion md is due to differing diffusion rates of two or more spacieso Diffusion rates in molecular crystal have been shoim to be fa irly Xow^ approximately ^ secT^" neasr tho melting point hence thia method is considered unsuitable as the two aectione of the cylinder would have to be cut and aligned very accuratelyo In order to minimise these errors the diffusion times necessary would become prohibitively large® The semi«infinite cylinder method has several advantageao lo A uniform thin deposit can be placed on th© surface epitaxially by evaporation or from ©olutiono 2 o The crystal can be aligned and sectioned accurately thu© allowing diffusion coefficients as low as secT to be measured in a reasonable length of timeo The minimum time necessary to measure such a alow diffusion rate can be calculated from equation (5)o Assuming a concentration change of 10"^ In a painetmfion of 30 microns with an Initial deposit . approximately one micron thick the time necessary is 600 hre® 1®a® approximately one montho The disadvantage of using this system is possible loss ©f surface activity by evaporation or surface diffusion® These c?n be minimlaedg however^ by pressing active face© together^ having as small a free volume in the diffusion cell as possible

TO VAC UUM

AIR VACUUM

FI&I4 TRITIATION OF BENZOIC ACID. 94 r

Bnû kee^ping than aiiE@aX temperature b.b eosista'^t am poealbleo

I t MBB doo&dedo thereforoo to Wy t l w âlîiUBiou of tr itiim am# Im benggelo ami a e a tie a e li in th is m^o The QEpariaemtal preeedMre required the t©%lmiing etepso lo Pmparatigm of raiioaotivo isotopeso 2o DBBlgm ami eomatruotiom e f a iop@©it@ro 3o Diffueiom AmmemX mmiitiemeo 4 o Wethei #f Seotiomiag th© (srystalo So IteiioohoBicial ©f th© SQotiom»

L.0%9 Preparatlm ®f Eaileaotive Isotoposo ao Tritium laaW llei Bemmel# Aoicl® Thia i© ms of th© eimpleat pr©parati@ms ®f a raiioacitiirelf laliGlloi meleeul©^ beimg prepared hj a aimpi© ©^Eoharng© roaeti@m WtwGm tritiatei water ami th© aeidio ©arbosjlie hjirogem ui th© aoiio Th© aetw ity required wo© oalomlatei' ©m tfiio haai© of a ©urfaee aetlvit^r of 10g000 ©p©o with a deposit thio!m©as of S^3 miorome amd a ©oumtimg effieieme^y of SOjJo fMs rofuired a apeoifio aati'^ity of approEimatoly 3 mioro ouri©© per milligmi am# thia vim th© tritiatiom level aimed ato She appar©tua mmâ is ahoim im figo 0<î^ o Oo35 gm© of toalî^r bem^oie me id waa plaoed im the tritlatlom veaselg ©vaeueted to a vaouum of lO'^^emo Eg mû tap© am# oloa©#o Tap T'l WES opeme# m# th© meroury ^vemtll ^ releas)©# to allow tritimte# water from th© reservoir to b© ûmûmmmi imt# the calibrated tub© using » aoetom^^DrlkoI# ^ siEtmrOo The

s «LASS « ï

(S.

êôPPg^ ‘fyg^

MiAfgfg WiNDINQ, %W' I [_ J H6A?|^ " Cü!s!?|ôu,

%R 35c tritiat©# water had a specific activity of 1 ouri$/S mlo am# approximately 2. omo leagth of calibrate# tub# was filled equivalent te IS mlllleurles ®f tritiiimo TW ''vemW/ miil tap were thorn closed end the trltiatod water diatillod into J. the tritiatiom veeeol where it was left for several deya t# equilibrate o The excess water xmM them removed by imaming th© water im the reservoir and distillim g over the ©xoess with gentle h@ati% of the bem^eic acido After about 1 hour the vont'll am# were closed ^ ogemed t@ oomtimuous vaommm and the remaimimg tra c e s of trltlm te # water were trapped out im liq u id nitrogen ceolo# traps for another houro was them closed am# slowly opened to the atmosphere o The tritiatiom cell waB removed 0 stoppered with Agiosom grease seal and stored In a fridge at «20^0* TOtil requiredo A 50 mgmio sample of th is m iû was vacumi sublimed im a sublimation tub© rnhmm im figo Iffo Th© tube was evacuatod to a pressure ©f l->2 mlcroms % p rio r t® sealimgo The mi chrome wmmd heater had a resiatamc© of 80 ohms ©md was aomtrolleâ by a '*fariac“ autotramsform©ro It was found that a 15 v supply gmv©%temp©rature of approximately 90^0 « at the baa© of the tube miû this was sufficient to give a beautiful, zone @f email well developed erjstale about half way up the tub© with a few ©del cry stals formed at th© enû of the asbestos winding wher© there a sharp temperature dropo The tub© was broken open and each of these Eornes carefully scraped out and placei im stoppered viaISc ApproKimateliy 2 mgmo samples of these gone© and of th© original, tritia t© # Wmgoic acid war© w©igh©à on a BsiorobalgmcGf) dlsBolve# im acimtillator and mad© up t# 23 Bio iu a etamâerê flaMSo One mlo aarapla® of each were tmkem miû commtad mi th© Ecke ©cimtillatiom coumtoro Th© result© obtaime# are shawm im th© follewimg tableo

BEmzOIG ACID WEIGHT COWT BATEp m ACTIVITY ABSOLUTE SPECIFIC mgm cpSo pQ3f mio cp©o per ACTIVITY mc/eigSo o

OBIGIHAL 2 o8 8 5 378 131 oO MAJOR zom 2 o605 340 130o5 O d fS UPPER ZQË-E lo 892 253 135o5 0 o@93

These results coupled with the melting point (122-=-3^Co show th a t th© bemgoie acid 1 © of ©ufficiemt chemical am# radiochemical purity m û the ©rigimaJ„ mmterial wa© used the im ltial tritium diffusion ruma even although the ape activity was only on© third of that imitlally aimed ato It was foundg however^ that this activity warn too low for eomveaiemt use as it required long ceumtimg times t# obtain the required accuracy of l«3?^o A further sample of tritia te c l bemgioio acid vmm prepared from 1 gmo of gome refin ed bemaole acid and Ool mlo of tritlated o Th©©© were sealed off mder a pressure of 2 , 0 oMo a i r im a tub© ccmtaiming a aide limbo The bensoic acid was melted amd refluxed for 15 minso after which the excess tritiatad water warn Gomdeneed into the aide limb using a. liquid nitrogen trapo

VACUUM.

r

f Y ~ TO Hr deposition VAC, apparatus .

mmm

FIS. 16. TRITIATION OP ACETIC ACID. 57 r

Th© benzoic acid was haatea until only a white powder remain©# where upon th© aide limb waa sealed offo The bongaio ©,©14 was them removed md placed im a long tubOo Thie tube warn them Nommée ted to a vacuum lim mû the bemzoie acid sublimed up this tube under a comtimuoue vacuum of 3. microm Hgoe the tritiated water again being trapped by liquid nitrogeno The

benzoic was them removed amd a 1 0 0 mgBo aempl© was re^ublimed under oomtimuou® vaouum» The © peeific aotiviti© ® of the samples were obtained as for the previous preparation with the

exception that the standard solutions were 2 5 0 mlo of Amslar toluemGo The s p e c ific a c t iv it ie s obtained were 7o45 pao/mgBo

and 7 o3 8 |ic/a^mo respectively thus ahowimg radiochemioal purity bo Labelled Bemgoio Acido

Thie m x B obtained d irect from th© Hadiooh©®iorf, G©mtr#o

A T m m h r n i m Ool m illicu ries im 2 o3 8 mgmo from bat#h GM2 1 6 /lo

A m thia vm amoh too active to use directly it warn diluted by di©so

the solutiono The final activity of th© benzoic add v m m

2 o3 7 microouries/kgm« @md this warn used directly« Co Tritium Labelled Acetic Acido

The apparatus ubqû is shown im figo #& « Th© t r i t i a t e d water in re se rv o ir A was frozen using liq u id m'itrogomo AppTOxisately Oo75 mlo of Messrso BoBoHo toalar acetic acid which had Worn fractionally erystaHim d three timea^ mogto Mo53& OoOS^Go mû Ool fitûo of acetyl borate were imtmémceê Into veaeel Q through the side limb B which wbb them mealed ©ffo Vessel 0 vjm then tvmQn down ^ith liq u id OEygem^ evacuated to 1 0 cm<, Hgo #md all taps and mercury wmtil©

? 2 ülmQûo Tap Tg and vesitil m m opened mû th© tritiated water allowed 5 Blnm t@ equilibrate to a vapour pressure of lo 8 cMo Hgo Tap Tg sm clomoêp Tj opened ami the tritiated water mmcendemsed into 0 using a liquid oxygen trap in approximately 2 minute®o Tap Tg warn clomeio Tg opened ami B agmim f il le d to l o i omo Hg and condemned Into Co T% ua# them closed mû Q sealed off under vacuum at Th# mixture

Im 0 mM them boiled gently for a few isimute© to allow the acetyl borate to react with the water presamto ApproXo # of the liquid im C warn them distilled imto Eo Thi© vm ©lowly crystallised using ice water until about # had cryatmlliméo

The aupematamt liquid vmm them ° f la a h e v a p o r a te d imt© 0 by placing a liquid oxygem trap around 0 and mealed at 8 go The liq u id im Sg melted a t 16 « 5 5 lêoTü^Co fhi© liq u id was thorn distilled into F and mlowly orystallisei and ^ f 1 wh^ evaporated twice before sealing at 8 ^o The melting poiat of the liquid im F wee 16oS 16o7^C<> The literature value for the melt lag point of pure acetic acid is i 6 o6 7 *^Co hemce the mmsimum 59

impurity conteat la OoOS^o Th© volume of aeotioa B waa 105 OoOo and knowing the Û

vapour pressure of water vapourg lo 8 omo Hgop the to ta l activity added to Oo75 mlo of Acetic acid was 0o8 millicurie© hence making-the reasonable aaaumptiom of complete equili­ bration thia gave the acetic acid a apeciflc activity of

1 uc/mgmo whichw» ©f sufficiently high activity for diffusion atudiea o

Sc The Deposition Systemso ao Benzoic Acid Depositoro There are two methods of obtaining a thin uniform deposit ouMryatal surface^ deposition from solution ©r^ if the material tm sufficiently volatileg ©vaporatlono Aa bsB^oic acid exhibits

a sufficiently high vapour pressure g 1 tmo a t 9 0 Gog i t waa decided to evaporate radioactive deposits under vacuumo Thia has the following advaiitage over deposition from solutiono

1 ) Even epitaxial deposits can be obtained.o 2) The area being deposited can be uniformly masked and varied w illo b) The eryatal pumped under vacuum for 5 to 10 minute© clean up the surface . c) The radioactive material undergoes a further purification step during the evaporationo The disadvantage of this method is that lOOi^ deposition cannot

ACETONE,/DRI KOLD COOLANT.

MERCURY c o n t a c t .

C R Y S T A L PLATINUM BOAT..

'O VACUUM.

e v a c u a t i o n Ho l e ..

UMSSTEN LEADS.

Fl(q. 1 7 60p b© aehlevod as ivith aolmtlem but f@r a w l l oon^truGteû evmgoratlem m it 50^ efficiency i@ not unreaeonablao Greater efficiencies cam be obtained by cooling the crystal o The system devised for the evaporation of benzoic acid is shorn in flgo It conaiated eeecntlally of a email platinum boat contained im a loO c®o l^do glass tub© a îmi milllmetroB from the top^^hoa© aurfac© to© ground flat© The boat was attached to tungsten leads rjhlch were sealed through the glass at the baa© of this tub©® A brass cover piece ï?lth a centrally drilled halo XoO cm» diao fitted tightly on tog of this tub© and the crystal to deposited mm# placed over the holeo A bras® veeaei ma© then placed over the crystal m é fitted tightly into the coverpieoop pressing lightly upon the crystal, to give good contact « This upper brass vessel contained mercury which acted m a contact liquid for the glass “cold finger“ which was used to cool the crystalo The system was evacuated through the central tube which had a email hole blown in it to giv© aceeea to the outer section of the depoaitor Before this hole was Introduced it was found that the air being evacmated was drawn past the platinum boat so fast that the radioactive material contained in it was blown out-o The thictoea© of deposit required for the solution t@

Pick “s equation to be applicable should be loss than ^ 1 ? hence for a diffusion coefficient of 1 k 1 0 "’^^c®o^aecT^ to be measured in 200 hours the Maximum deposit thidmea© ia B^Suo 6 1 .

The deposit thldmee® m B û Im all miperlmmte wm# I b b b thim

2 ^ 3 aicTOnso M order to give m mon^ epiteisi©! deposit om the crystal BmSmo the evaporation umrst ho carried out fairly quioîây to avoid ohiskor growths a most imdeeimablG offoct whioh hm boom oliaorvGcl om evaporating clepooits onto organic oryetal© wfeon the avogoration rate wao slow enough to allow preferential c rye to ll loat lorn at apeoiflo muwimo aitooo The aecQwmary for good epitmsiol deposition woro found hp OEperimonto It uao found that e herting currant of 3,0 ampo

p 0 0 0 0 6 fo r 4 5 to 60 m@m€m through tho platinum boat gavo a good oepoBito The heatimg tiaa^ howeverg depended « tho vacuum of tho system and the ©moimt of m aterial to he éop@aitoi« The vBOiam achieved varied from deposit t# deposit but was always lee® than 2 mleroms %*o Tho ovmimAimt lino used

c o n siste d of a 2 stage mercury diffusion pump €Oup3 ,od to an oil filled rotary fore pimgo Th© hefting eus rent we?s

obtained using a 6o5Vo 20 amp transformer with m V ariao ooatrolled means input o The current flowing through tho platinum boat was monitored by hf^ving m ^^^ometar iaoorporated im th e e ir e u ito The efficiency of the deposiximi we# meeaured by

QvaporptiBg known weight© of acid onto 1 cmo dia^ glass cover-

©lips and weighing on q miorobalance before rnd after depositioBo This vjas found to give m efficiency of 20± 3po Hence fo r a deposit of thickMSS 2 - 3 micron© om a eurfacs area of OoîB5 cm the weight of acid required to the platinum boat wae 1 - 2 egaso For several diffusion ©^périment© this weight- of material was weighed out to ensure that the pjoper conditions ?iere attainedo In otherSg howevej j, th ie amount of acid could b© judged q u ite easilyo The crystal© were cooled in most of the diffusion 93:periment© using an aceton?/Briko3,d mixture taking the crystal temperature down' t© around to increase the efficiency of depoaitiOBo This procedure wee abandoned to later estperiments when it was realised thpt the cooling was inducing condensation of water vapour onto the crystal- surface during the short period between its removal from the depositor and ©©altog to the diffusion cell which affected the tritium diffusion results quit© seriouslyo The design of the oepositorg however^ was found to be very

successful and was used throughout the eerie© ©f ®3 speriment© without modification for the deposition @f tritiated and 3,abell@d benF,©ic acido

bo Acetic Acid Bepositoro The acetic acid could not be evaporated in the same way m bensoic acid as it is a liquid at room temperature o It “a van©?Ar Û "^2 pressurGp 1 c U o Hgo at 1 7 o2 Co^ is sufficiently high to allow

Ts

j f ) TO VACUUM.

0 ^ 7 4

r" ■"> r TO VACUUM,

X( V TO VACUUM Û A , i K

FKfljS. ACETIC ACID DEPOSITION 5Y&TEM. volume to be filled at room temperature which can be rapidly conûBnmoû on a cooled crystal surfaceo This was the method uaedo The apparatus ia ahown in figo Hf o The eryatal holder aM ooolimg sect lorn similar to that meed Im the Wmmolo aoicl evaporatoro The lower aeetiomme a doming ©yetem in which a known volume filled with the saturation vapour pressure of trifciated acetic acid stored in Ao This dose xim them oomdmmed onto the orystel surface o The actual procedure finally used to carry out a degoaitl was as fQllom^o The se c tio n above Tg was removaable and the crystal placed in position ia a fridge at -20^Go This section was then replaced aad the crystal cooled uming aa acetone/ e 0 B rikold mlscture» Taps Tjj, Tg imm olosedy Tgo Tg opened mad the Imowm volume B evacuated dowa to zero pressure ®a the manometer using a rotary pump« T% m m oloaed^ T-,, opened mû the mercury ventil opemea maû the acetic acid vapour in system allowed to reach ©guilifori»p at 20^0o this nm lo i ©Bo Tj was domed g T^g T^ opened and the upper section puEipei clown through tap T^o Tg vim closed mû Tg ©lowly op one do The mcmometer f e l l rap id ly to mppro3slma%ely Ool ciBo %□ showing that the acetic acid was condensing out mû that the crystal temperature was approsslmately -20^0o from the residual vapour pressureo Tgg were them closed and the upper section returned to the fridge where the crystal was quickly removed

(e.) (fb)

i.ENZOIC ÂCIQ DIFFUSION OELLS.

L

______^ CaViiTAL, t n \\\\\\\\ w \:

FIGj.19. 64 f>

and sealeo in the oiffusion cello The handling ef the crystal in the fridge mM the moat unsatisfactory feature of the nhola operation due to the possible eomtamlnation with water vapour (BoVoPo a t <^20^0oo Oo7T mmo)o This ms mlmimiged by carrying out these transfer operation® mm quickly as poaalbla*

4*, The Diffusion Anne ml Gonditionso

Eo Bensoie Acid Diffusion cellso The diffusion cell# used war© deait^n©# m# that severaJ. cryetale could h% diffused aimultaneoualyj, the erymtal© eould be preseed against each other with glass or mica spacer® and

the cell could be aeeled air tighto The original oello flg< 1 VJBB constructed in brass and had a total internal, volume 15 ccop most of which was taken up by the spring used to pres© the crystal© togethero This cell was found to be satisfactory for temperatures up to lOO^Go above this temperature it was observed that a ©light reaction occurred between the brmae end bensoic acid and th a t evaporation of the crystal, became a serio u s o b j e c t io n to their useo A second set of cells wwm constructed in aluminium figol*f\ In which the free apace was cut down to a mlnlBmmp the total available apace being g cco These were found t© be vastly superior with little @r mo evaporation being observed even at 116®0op 60^ below the melting pointo Both cell© were sealed by screwing both halve© 65, together using teflon tap© t© give a gas tight seal.

Acetic Acid Diffusion Gallo îhQ aimpl© c a ll usad abova y m not suitable for acetic ac-M as tho crystals could mot bo eomvomiomtly hamdlado Tho c e lle ueoci in t h is ease are ahorm in fig o 0^ m û m vB ooastmotod to alustoiUBo Am aootio m iû c r y sta l \mm out to approximately tho correct si^o lo 3 o®<> diao k 0o2 mbo thicko This was thorn hùnûBÛ @m to tho plane surfae© of the eoll using a few drops ©f pure liquid aoetio acid which frose to tho crystal and the cell in tho fridge 0 The square and ©f tho c e l l was clamped in the chuck of the microtome and aligned using the technique described to the next sectiono The crystal was microtomed for several hundred microns m û the sides out off immediately prior to deposition 0 The diffusion cell fitted into the depositor and th© mercury contact to the cold finger fitted into the hole at the top of the cell m ghoim in figo 0# o On removal from the depositor the bottom cover of the cell warn sorQ\7©d on using a teflon tape bqb I o After the diffusion annoal the c e l l vmm remounted and aligned on the microtome in the fridge prior to aectioningo

Thermostat Bathsq Benzoic acid diffusion vms measured in the range 70 116^Co The first bath used was a 5 litre beaker of silicone clip contained in a large Bietal dm® and toaiCLated with ^ U w lm ilK

SIA. .N-

47 K. 22 K. lOOK.

THERMISTOR

3 p0

|i t I2AU7 EN 91.

JK-3WW POTS. DEPENDINÛ ONTÊMR RANGE AND THERMISTOR TYPE,

F i^ .2 0 THERMISTOR CONTROLLER CIRCUIT. I ' The theOToateit M it vjw simply a f i l l e d tolmeme ^regulator t?hidî oomtrolled a 250 watt ImmermlM heater through vaeiam ©witch ¥IOtj3 <, Shia wa® femmd to operate ©atiafactorily with m aaci&racy of Ool^üo over sh o rt period© of time m&d ± OoS^Oo over a period of several dayso At elevated tem perature i t was îom\û that the stirrer eecaeioma3,Iy studs MO ehea th is hmppemed the rum was mbamdomed aa the temperature variatim e with position to the hath m m largeo Where lemg d iffu sio n amnmml tim es xmm Moeasary a secoM amiiealiMg m^m was commtrueted^ . A mi chrome wimitog warn made on am insulated brass foxmer I? diao si long having a c en tral h o l e i i a o This winding wa® insulated with m$bast©a and emc&eed to alum toiM with ©jndamyo miû pieoeSo fhe temperature was maintained by a thermistor

controlled thermostat circuit g f i g o E o a the thermistor being located toe well drilled in the central brass fomer^, The temperature stability ©f this ovem was fmmd to be b e t t e r than ^Oo05O^ uetog a 50^ Bedcmam thermometero The temperature gradient down the oven was measured with a .Fto-«Hh thermocouple and found to fee mumUmt m er m 6 " length at the middle of the ovesno Diffusion celle war© always placed to this region and their temperature measured with a 0 200^Go therraometer which had been standardised against am loFoBo standardised thermometero The acetic acid crystals were annealed in e ith e r a , aaci thermostat feath® aeomrata t# i; 0*0$G^ or a laboratory conatructed low temperature theraoatat deslgaed by Dr« Jo Mo Sherwood for uee dmm to ^20^0o with an accuracy of & 0o05®Co k . fhe S eationing TeohmlqWo The major groblom faced im etudylmg difftiaion by the eamto infinite solid method ia to align the ©ryatal accurately parallel to the initial face amd take umiform aeotioma of the crystal perpendicular to tMe facOo Thera ar© several poamibl# methods of seotlontogo

I n '*f^0 2o Removal ualmg a latheo 3o Microtomingo 4o Bvaporationo 5 o Dissolutiono Grinding has been ©uocessfully used t# measmre very low diffusion coefficients im of the order @f aeco but a high p re c isio n g rin d in g machlm© was specia1..1y constructed mû it ia not a routine methodo Sectioning using a lathe ie the normal metellisrgical practice and ha© beam SB applied to the sectioning of Anthracene o The limiting factor using the lathe^ however^ ia that the minimum section which can be accurately taken ia about ^/lOOO^^ ioOo 25 microns hence the minimum diffusion coefficient which can b© measured 'sjsswaiîsg 6 motive sectioffls sM a fiiffasion œ neal tira© @£

2 0 0 h@ms'r3 i s 8 s. 3,0“”^'^ ea? ©se» ïhis is met eomsiâereâ OTffieism tly l@u Î0X th© preheat etiMy« Seetioaimg by Qvapoa-atioa i® a toehmiguo ohieh ham h@@m reeeatly lîseô t@ %# seetiom io© esy©t©3.© by ©ufoliaimo layam fv@a the Gyymtm. held at «»8D®G» ©mt® a e©13.©0tis»g plat© at '»§0®Oo îhisi aethsio h@w@v05Pp may prefes'em tially Qt«h ©sjrmtals d@om gm in bewcias-iQ© @r dialoeatioB pipes asîdis t© ©bbbto wmifas’aity ®f ©eetl@a thielm ess @md glamarity ef the ©s-ystal faeeo BiffiisiOB ej» ^ fj) coefficient© of 10 œ© aoeT"" Imvo beam qmetcé ugimg th ie metWdo SGctlomimg by dissolution la a oommom m©ta32i?§liioI techiaique end may prove useful i n atW les i n organic materiml but i t is open t o the b b b © objeotloma m evagorationo ¥@ry low diffualoE ooeffioien ta of the order ©f 10"^'^'®©©! meoT^ Iibvq beoEj^by dissolution techniques but im particular to the ©as© of tantm3.UBo The most satisfactory method appeared to be the microtomec The normal aledg© based on rotary m icro to m e ©@ia talce B Q o t i m B of X «=> 15 microns thick with a high degree of moGuraoy © Assuming 6 active sections of 5 mieroma each amd a diffusion time of 200 houra a mimirmm diffusion ©oefflelemt C2f*| ^ 'if Of 3 % oîBo BseT"" ©an be aeasurei© This warn oonsidored auffioiently accurate for the present experimentso %m previous ©experiments #m m olecular crystal© a d iffu sio n c o e ffic ie n t of 2 '■"^1 around ID"'"" cm© mdo^ wa© obtained at temperatures approaching the melting point© Umm diffusion coefficients should be % o % measurable over a range of 10" which wa@ thought mùitSieJxmit tû allow accura-te oaleulatlom of the diffusion parameters © The micretùme used was a Beck Hûtary microtome model

which ?ma calibrated to take 1-15 u see tiens© I n this ml ore tome the crystal ie moved in a vsrticaS, plan© past a Imife- blade fisxed at some slig h t angle to the vertical© f@r

^‘very rotation of the microtome handle the tolf© blade m m m i l towards the cryeta.! by the amount set on the cf-libre^tion dial© The accuracy of the call oration ?jas checkeo over ler^e di-stanceSp hundred of micronSp by Dr© Go M© Hood using a travelling micxoacope eno founc to be t v m o The accuracy of sectioning in the region 5 10 15 was determined by weighing section of a piece of paraffin v m K of toown surface area and density© These weighings were carrieo out cm an OerU ing mlcrobalance„ - " ' o The overall acous^acy of sectioning was foimo to be with average deviations of 5j^o

3%5 3% for 5o 10g and 15 |a slices respectlve3.y o These values were considered sufficiently accurate for the experlmentSo Tha accuracy of the microtome was also esipe riment ally confirmed for benzoic ado using;, the sectioning method of collection described later©

Allgnmexit of the Crystal© AB this WPS the most crucial step in the experiment much thought was given to the best method of alignment© Alignment

SECTION A A.

FIQ.21. FINE ADJUSTMENT CHUCK, M.uig a veil lag microBcope wes rejected as insufficiently acüurete, a^d the methoa ohoaen waa baaec m the optical levero The alignment operation x^equlred the constraoiion of a fine adjustment chuck fig ^6 « a© the chuck of the microtome only gave rough positioningo This fine aojuetment chuck vmm held by the microtome chuck and it rotated m% a ball joints tho accurate sotting of the face beln^ attained using tho 3 adjustment A pleoso of paraffin vibb melted on to this chuck mû microtomod to give a smooth facQo à email plan© mirror maoe by vacuum depositing silver om a mieroBcope slide TOs placed on this face anc held im position using a strip of *’3©ll0tap©“ tremaparent adhesive, A ”Sca3,8mp® light source naa positlomed about 3 feet from the microtome amd the light reflected from the mirror onto a plain ^hite paper above the lampo Cross ^/irco were pieced ©m the lamp bulb aad th e ir reflection was focussed on the papero The centre o: the cross iiinB rjae markeu ama th is acted as the refaremce p o in t fo r the plane of the kmife bladeo The chuck %as removed miû the c r y s t a l stuck om using '’Picem©'^ vAich gave a ?rig.ia moumto The m irror vîbb replaced om the face o f the crystal, and remounted om the chucko The face of the crystal rms them brought into the correct by im© of che fime ad jus mem t screws so tmat the reflection ox the cross wires was again at the reference pointo The crystal waa mow aligmed parallel t o the blade but m ot at the ©d-.e- o f the blade « The kmife blade was 1 1 . aojusted to almost the coriect position at th© oryataX face by using an adjusting screw on tho microtomeo This was them locked and sectioning eommeneed until the blade was just serf ping the surface of the crystal o Usually ajE ter a diffusion anneal it was. found that slight variations had occurred on the crystal surface and the first slice was always suspect due to these variations^ slight compression effects and the fact that the position of the crystal surface© was not im oim accuratelyo Allowance was mede for the thickness of the first slice by comparing its weight with the otherso It should be noted that when attempts were mad© to section the crystal with the knife blade vertical the compression effect vmm so great that th© cry at f'l shattered o aperimenting with the angle of the blade showed that at an angle of 15^ to the vertical negligible compression was experienced * At all anglesg howererg in the case of benzoic acid^ it wa© found that on ©Qctloning the acid disintegrated into a powder which had a tendency to “^park^ to© to static chargeo Thi© ©parking occurred only with benzoic acid and not with other organic cry©tale like anthracene ©ni ffiiaphthalene which were found to give a coherent s lic e which could be readily picked off the knife blade using a single edged rasîor blade o This fact is in keeping with the g re a te r hpmmsB oi b©n%olc acid which wfB quite a brittle crystal ©na when © ain^le cryatal was grown it gave a metallic ring when struck o I'ionIngle crystals g however p gave a much 1 2 . more muted soy mi o This problem of alioQ oollootion warn difficult to achieve comp3,et©lyo A method which had previously been up©a to collect ciystals which disintegrate was to place a atrip of aclheaive over the crystal face prior to ©ectioningo Om mlorotoming the section adhere© to the adhesive which is simply removed ©ad eouatoio This method wa© used by Magether Crooks and Maurer t® sectlorn sodium chloride crystalso When th is procedure was attempted the adhesive always stuck to the knife edge as seetlomlmg started buckling the adhesive ©md causing distortion #f th© crystal jhenee it was rejecteio The method uaed was to take a 10 /a ©©ctiano This powdered and spread out to a fen shape over th© Imif© blade with a fairly thick ridge along the knife ©dg©o A 1^" piece of glasa capillary was dipped tot® a vial comtaimtog cyclohezmmeg to which tho solubility of boKisoic acid is only a few mgrn^/mlog ©md capillary attraction filled th© tubOo The end of thi# capillary rms allowed to touch the Imife ©oge mud the cyclohmmmo immediately flowed out wetttog the powdered acidô This ©mall amouat of liquid was sufficient t® blmd the particle© toge'#her and allowed them to fo© removed from the f© blade using a sin gle edged raEor bladec Th© acid was immediately transferred to m piece @f plain .whit© paper 2'" square pno the cyclohexsan© ©Honed to evaporate off over 5-10 mtouteso The section them looked like a white needle ë^bout lomgo It was subsequently transferred to a jf ûTmi glaaa v:lal and prior to weighing m a mlCMbalmee which naighad to 0o002 rngmso The weighing proeeiur© wa© to allow each vial 5-10 mlmutea to reaeh ©quillbriUMg the weight waa reoomWdg the v ia l picked up with tonga mû Inverted over a filter fimuol a ittin g in an empty scintillation vessel<> The vial imm given a ©harp rap to knock all th© acid into the fummel them replaced am th© belamoe am# reweighed while 10 mlo ©f aoimtillator solutiom waa ua©i t# wash the acid into the aototillatiom veaeelo The dlffomnoe im the two ivoighimg© warn taken aa the weight of acid im the aoim^ tillatiom veesel» Dm© to the sparking of the acid it was met possible to eoXlect the section :lm XOO^ yield mû the several opejcatioma also imvo3.ved some loaso The weight per saotiom could vary oomeiderably âepamdihg #m the asaoumt of sparkimgo Am example of m average rum Im mhovm b^lowo fo f- S e e t io a F/t„ DEVIATION saoTiom ' W. 2>Hv?ïâfïÔM 10 i& MGMr. MOM MMW 1 0 11 MOM. m om M iâ i

1 0.615 1 0 1.300 11.3 2 0 . 9 6 6 -17 1 1 . 1.238 , 11.1 3 1.511 12.2 12 0.960 —17.8 4 1.135 => 2.8 1 3 1.173 0.34 5 1.164 - 0.35 14 1.274 9.0 '6 1 . 0 9 0 — 4 . 3 15 1.178 ■ 0.05 7 1.064 “> 8 . 9 16 . 1.194 2.2 a 1 . 1 9 0 1.9 17 1.200 2.7 9 1.207 3.3 18 1.190 1.9

The a'^QS'ag’s w e i g h t per 10 m © © etia® te 1.168 mgia.

Th@ weight ©aleal® ,t©â froB s1 B u r f . & m a r e a @f 1.21 cget. efflo e n â clemaity lo32 gmo/oo I. 06O mgm,

.3ÏBlMMU«»t3eïSSj LOTJcstsewsBï ifcïrr.j3i®5ç^P£tT^‘;:jjîîïi!^

ir 111 jK^KBSWlCSl^^iPHlWw%TCSO The aveTaé^e p of slice collected ^ - 73>bo This ia typical, of the devlatloma miperiemcedo Portanately i t

mot essemtlal to ok^taim the aectlom i n 3.00^t, yiel.d as only the ap ealfic a^ctivity per section was recjuiredo io©o the activity per mllllgramj, to plot the concentration variation with penetration into the crystalo Therefore having shown that the microtome was accurately calibrated the actual penetration

v m s taken from the microtome oalJ-bration and th© weight per section imB only ubbû to deteraine the specific activityj> ©SEcept for the first slice vâwm thictoee© was obtained by comparing it with the average weight per sectiono Thi© aeetlom--

tog method vjbb considered suitable for 10 # aliees where weights of about 2. mgBio were obtained to m accuracy of 2$?

or bette To It vîbb founds ho wave r^ that the diffusion coofficienta to be measured required 5 p> s3.1ces to be taken as the penetration distances became smaller and the weight of th© ©eation foil to Oo4 Oo5 agmo for a 1 mqo emo crystalo Crystals of around .1 aq® cm* were found to be th© optimum #i^e for

Oi'ASuring uniform ©urfaee planarityo Â. more refined method of section collection was developed baaed on a auctioning method described by Labea et a2.o This enabled the complete section to be collected and smaller cry stals to be usedo The apparrtua we© simple and shown to figoelA o Central holes were drilled in 20 -polythene cape for 4 ÛTBM glass vialso Bight angled pieces of Oo5 cmo diao sc 6" Im ^ glrss ocpiilary tubing wero introfiueed to glvo a tight but slldimg fit« Tho p?8t?uôimg edge ivaa S’ouadod off to p F o W o t the e&’y s î t a l from ses’atehijag® AppyoEiaotoly OoS to 3, alt, ®f m ts*a{>js?S.ias 3.;l«jts3.a r/sa fh-ORu in to tliQ vâol by to e e rtia g a ©yffinsG moedlG imto tho poljrthoao oap ohioh v;as aomaotod to a vjatGr pump RhâoSi gfewo a vaeuum of a fOR iaebos Hgo ?ho OEtoîmaS. amd o f #@ csopl.l.l.as'y ®e© 4rat30 bewjteoasrfl® osid fogoogé® aes’oa© the kuifo od^,© es the aootlom oas tekam cm# ohrm th© tiaouiaa uaa high eaoagh avogy parsloî© uas éffa» ia to tha c5Qpi3.3,ai’ÿo ï'h is amablaé the aofe of «m ttlhg to ho ®bo 0 .woâ ths’oughoüJt the ogGMtloa thus ©.usus^lag that a® part of the @m£me oa® belmg ralasod» Shie oeo espeolfally lmpo%"tGmt a t high diffusion tespea'atUE'eo nhere it waa fooaâ that distOFtisas @f tho surfaee ®f a !’©« aiieffoa© eouM @eou,&’<, tïhess th is happened the prooodras*© adopted mbb to tehs the fis’st ssetiors and ohop ®ff mmy pes*t ®f th© faos met beimg @eetiom@d sasiag a shaffp ffasoF blade « Omoe the eseetlom had bao® takea th® eapillas’y m© Kjised a® Im figolfêb am# a fui’ther asaoMat ©f liguia d:rarm through the eapillascy t® v?aah ths’ough aaf adlMDS’tog pag'tloleso ïho aesus’aoy ®f t h i s teohaitp© m& aheekod by talîâag bemmoio mold s e e tio a s i® eyolehezsame, Thee® mtQ made up ts ' 10 mlo Im a ©tanôari flaok mû their VoVo ofootra talma m a

Porkia llraer 3.37 0„Vo Spoots’OphetOEGtes’ &m 1 «sa» oeLlSo ■Sh© W»?„ sp eo tm of beasoie aoid i® oy©3,ah©Kps® f% « 0@

f l B Z ù

@9

Ü

Ù 9■■i» m eu 0

t î O

âi

î^^«rWti**A*ii^^’«=M«H3r3|;ïiJiWlîîJî3;îi3r?LT£ïîtfei^ïTO-vtf3^Sr^ria?jLïv-np«5JiW^fl:,^VJ‘ é cp %s IG

oosiaista ùf 2 hnMm iu the m gim 205 295 mn koomi. as tha B amd C bamda the B bmmd ha'^iEig A mait 229 233 mu vjith lo g # amsi 4ol5 buû the G berna havla^ 2 maxima at max 275 P aBd 282 fi \^ith log@mmi 2o88 mû 2o74 reapoetlvels^c I t mm foimd th a t th© G bamd gave id e al ebsos'baae© peaks in th© region Oo5 to 1q2 absarbaM© urn its for eomeemtrBtlams of Oo5 to 2o0 m^Bio pes' 10 mlo ojoIohexaB© mû a e©libratio/A graphs figoSl^ of peak height at 27!mp avid 282 mp vs acid ooaoemtratiom was ooagtrmeWd ustog astmdard molmtlomB of bon^oio aoici in oyoloh^KMOo The weight per aeotion was th@m reed direotly fro® thim grapho

Th© B 0 ctiom:Wg prooedar© developed was to taka the first sectio n in cyelohezmme to determlm© i t s weight mû th© ©Mfosequent sections to sctotillator aolutiono In th is caa® 10 ralo of scintillator wasaccurately meaaared into a 50 mlo beakero Approximately Oo5 to 1 mlo was drawn from this to trap th© section. Aft©r sectiontog the capillary warn raised to prevent air bubbltog through the aolutiom to mtoimis© mcygem quenching and the rematoder of the acintillator drawn too The special cap was them removed and th© vial stoppered giving a 10 mlo aoluticm ready for acimtillatiom Qmntingo By this method a check could be made on the activity m section tog proceeded o th© sections ceased to b© active several 10 p sections were taken in cycl©hexane t© dotermine the average weight per section and the accuracy of sectioning which was normally found to agree with the theoretical weight to within 5^6o 1 1 .

DETELMlAAïIOh ü? hADIOWjlVITY Ih THK CRYSTAL aKClïOiW.

The radioactive laoiopea used lu this tovewt%ation@ Si tritiu m mv. oerbon - 1 /ig have the folio ring decay achemêss-

He| o'" B te x OoOlB MeVp X2o3 y e a r s «

.5. iC Mail Ool5 6 MeVj) 5j>600 yearso

There la only on© step in each decay scheme resulting in the exïîiteslQn of a weak pi p a r tic le o Both |§> ® have half'-^lives loiig enough to be considered invax^lani durtag the course of any experiment a The low p energies eliminate the use of end window Geiger teller counting for their ôeteciion aM require the more sophisticateé techxiiques of Interna], gas counting; or seln^

tillation counting o The f o r m e r method req u ires the

combustion o f the raaioactlve sample to a suitable gasg GOg or Hpp which can then be counted internally in a Qoli» tube» This method gives high counting efficiencies but la very time consumingo The more rapid method of scintillation counting yields high counting efficiencies with 90> having been recorded for

C ’ and 4 O/ 0 for tritium ^ It was aeeiaedg the ref ore ^ to use aoJ.ntillsttion coimting to detect both iaotopaso The principle of liquid scintillation counting i& as followso Th© radioactive sample is dissolved in a suitable organic solvent contriining a scintillating pho©phorj> Icnown as tno © eintiliator ©olutiofâo bm placed over a photomultiplier tube la a light tight bosu The ^ particle emitted activate© the phosphor to aa excited atat© which decay© with the emlseloa of a photmîp W o This photon interacts with the photocathode of the photomultiplier tub© which emit© aa ©leotroao This electron i© amplifiea through th© dynode chain of the photo^' Bwltlpllar to give a weak electronic pula© wnich is subsequently re-^amplified to operate a scaling unit or. recordero

Counting Assemblieso Two liquid scintillation assemblies ha,ve been usedo Initially the assembling used for and T consisted of the following unitSo ■ An Bek© head unit^ M664Bp having a 2 " water cooled photomultiplierp was followed by a iMUclear Bnterprises non-overloading linear amplifier» iE 5 2 0 2 {, and l^uclear Enterprises differen-cial pula© height analyser» ^E5102o The output from the pulse heij^ht analyser actuated a ï^natran scaling unit» type 1009E» couple© to m automatic timing unit NIO8 0 The atabilisea EoHoTo supply to the head u n it mm supplied by a Dynatron power unit i^l 0 3 o This assembly wa© later replaced by a coeiplete JS^Mclear Enterprise© unit consisting of a h©f?d unit» M5503o having a 1'' water cooled photomultiplier» amplifier hJi;5202y pulse height selector BE5102 ana power supply . The Bynat 79* scaling unit g I009E ai'io timer vms also used with this assembly o

GotmtiDg ConditimSo The best eoimtlmg eonditiona are not neeesaarily those which give the highest ©ffieleney a© this im always asaoeiatod with a high background count rate which will have, a large variation which will prevent accurate determination of low count rateso The conditions required In this inveatlgetion were those which would combine a high efficiency (to reduce the counting time) ami low background with long term stabi3>ity during the time taken to count any particular diffusion run* In each case the best conditions were obtained by varying th© four parameters EoHoT^» amplification factor» pulse height end getewidth around the values recommended by the manufacturero The conditions used are tabulated below*

ECKO BHT AMPLIFICATION P,H. GATE BACSCGROIIMD EFFÏGIEKCÏ VOLTS FACTOR VOLTS VOLTS ep®,, % T 1400 5000 6 30 5.0 17 c2.4 1350 10,000 30 20 0o2 66 inSo Î ^ 750 5,000 6 30 3.0 15 gl4 750 5,000 10 30 Oo5 60 I t was found that th© com ting conditions ^-Itered slig h tly from to day when the counter was switched on but they remained

"lAf &= sescseip'ip r 0R VOLUME OM T i;.m U M c o w N 'n w q BVJ I k K%a I4-©© V Soo& Si '&6a V ”=& r

B

m

E ll

:i;rgc x?TO::jKSti*a:s?accat; 3JLi"krw^ »r/1 s 1 m »

V®LUM@ ©P ^A^MTgLLATbR AD»®© . ML

,ose

*^V3

/S mi Btahle after a 1-2 hour warm up periodo h long term drift in stability waa obaeriyed mBT a period of momtha due to the agelmg of the photomultiplier and consequently al% ht changea had to be made In the counting conditiOMo

S'acter© Affecting the counting Efficiencyo

&0 Ifolume of Scintillator o A depth effect ïms essperienoed due to the photon® being absorbed by the ©cintillatoro This effect is shown in fig< The efficiency was greatest at small irolume® of scintillator but waa very dependent om the volume of acintillatoro Thi# •dependence decreased with in creasin g e c ln tilla to r volumo o The volume of scintillator used in all samples waa 10 mlo and at thia level th© efficiency only varied by OoSÿ&AÆo ' Thie large amount was also necessary to wash th© radioactive sample into the scintillation vialo bo Quenching o I t ha® been shown th a t m^ygen contmining compounds^ especially those containing carbonyl groups^ can quench scintillationso has shown that benseic acid has mo effect on counting efficiency in the concentration rang© êmu to 8^g gm/litr©o The quenching effect of benzoic acid cm th is eyaterrt was checked fo r the ccmcentratiom range of in to r e s tg 0o2 to ê> gms/liirop see fig o 2 4 !^ and a very slow de creme© in efficiency was deserved amounting t@

1 4 to 10 mgmo had no effect on the counting efficiency of r or To

Co The Scintillatoro The counting efficiency of th© system was greatly affected by the scintillator solution usedo One of the moat efficient a c in tiila to r a i© diphenyl ©Easole (PoPoO„) dissolved in tolueneo The wavelength of the emitted photon g ho vj© ver g does not give • maMmum efficiency of photoeatnodic emission o This ia overcome by introducing lo4 ^ dl - 2(5 PhenyloKa^olyl) benzene (PoOo.l*oOoP«) which acta aa a waveshifter to give the wavelength of maximum efficiencyo This was the sclntiliptor used and as large amounts of scintillator were required it was prepared in the laboratory from ^scintillation grade® chemicals ©applied by i^luclear E n terp rises Ltdo The concentration© used were 3oOg/lo PoPoOo ana Odgo/i PoOoPoOoPo This scintillator was found to give the same efficiency'a© the commercial scintillator

UK 213 supplied by Nuclear Enterprise® o

do Dark Adaption and Temperature Variationo It has been observed that if the scintillation vial is exposed to light before counting a residual, phosphoraacencog 82 o or aftercglowg may occur which gives apuriously high counts

I n i t i a l l y 0 Sf.mples werOg theroforop kepz in a light tight bom prior to counting fno it was found thot after two or three minutes in the counter reprooucible counting was obtainedo This period vlso allowed the ©ample to temperature adapt to the temperature of th© water cooled head unlto

Countingo All aaBiples were counted to a statistical accuracy of on© percent or better where possibleo For very low count©g hovmverg th is was not poasibl© and an accuracy @f 3j'o vmB- usually obtainedo The background coimt rate and atendari count rat© was measured periodically during a counting eeaaion to ensure that ho sudden variations in stability occurredo 8 3 .

Preparation of Diffusion Samples

Bonsoio acid ©ingle erystels were readily cleaved along the (OOl) pXenQo Am ezeellemtg smooth eleavage plane was always obtained v;ith few o3.eavaga stepso These steps » however^ prevented the crystal being used directly a® the surfaces were not sufficiently planar. AttMpta were initially made to grind th© crystal face with carborundum and alumina powders and water. This was followed by polishing on a streched pjb.ce of chamoisg uBing cycloh@%ame ©a solvent* The surfaces obtained by this methodg howeverg ware not ©ufflciantly uniform and this method of preparation was abandonad. The method used to prepare th© crystal surface was to eleave the crystal^ cut it to the approxiœt© shape required with a rasor blade ^ align it on the micro tome perpendicular to the cleavage plane and take one micron sections. This resulted in a smooth crystal face ?/ith a perfectly uniform and planar surface with little apparent damage. All crystals used in diffusion studies were prepared In this way including the studies on polyorysta111ne benssoie acid compacts and acetic acid single crystal©. For the studies on benaolc acid parallel to th© (OOl) plane the crystals were cut horizontally from a crystal which had a vertical cleavage plane. Great difficulty was e^periencedijV pieces sufficiently large for ûtftuB%Qu studA© as wm crystal© tended to split along the cleavage plane^ >fi©nc© few studies were made in this direction,

g The Diffusion %p 0 riment@ Benzoic Acid. Tritium diffusion was measured perpendicular to th© (001) plane in single crystals under a variety of conditions in the follow ing ©ysterna a) High purity benzoic acid, t) p«T@rph©nyl doped benzoic acid, c) De (Lite ro benzoic acid.

Th© diffusion of labelled benzoic acid was measured in high purity benzoic acid and deuterobenzoi© acid crystals

perpendicular and parallel to the ( 0 0 1 ) plane. 14 Tritium asid 0 diffusion was measured in polycrystallin© compacts of high purity benzoic acid. Acetic Add.

Tritiuiîî diffusion was measured by th© sectioning technique

in single crystals of acetic add perpendicular to the ( 1 0 0 ) cleevag© plane. The experimental conditions used in the above studies ar© included in th© results. îsehange Expe rimemt b

Xi became apparent that tritium diffusion could be due to fi exchange reaction occurring between the tritiated benzoic acid 3poait and th© small oonocniration of water vapour present in the IffuBion call followed by diffusion of tritiated water into th© rystalo It was decided^ thereforep to tost this ©ssehaAg© Biaotlon 0xp@rimental3,y by monitoring the ra te of tritiu m exchange âtween tritiw labelled benzoic acid crystals and inactive water ripour. Th© apparatus used was simple and consisted of a 250 ml. r.b. Lask with a long neck sealed with a serum cap. Th% flaak was Llowed to equilibrate to the humidity of the room which was mawred with a wet and dry bulb thermometer. A known weight of ire g dry tritiatad benzoic acid was introduced in a small op©n àraiîi glass vial and the run started by placing th© flask in thermostat bath* The neck of the flash p however g was kept > room temperatuf© and one ml. samples of th© vapour were withdrawn syringe through the serum cap at tlm© intervals which depended I the exchange température. The vapour in the syringe was slowly bbled throi^h a scintillation vial containing 10 ml of scintillator trap the tritiated water formed by exchang'^. and this was bsequently counted on the KB5504 tritium counting aBsembly. rmally about 15 samples were taken. This volume represented lass an 3fo of th© total volume of the system g 320 ml. end it was thought Bêo biiat this would not influence th© exchange charaetistrles Blgniflcaqtly The effect of the vapour preaeur© of benzoic acid on the activity of th© vapour was checked by carrying out ®blank® runB In which labelled benzoic acldp apecific activity(g#)uc/mgm^ ma usedo At 38^0 could be detected in the vapour phase >ver a seven hour period* At ll^Q the 0^^ activity rose to a small value within 3 minutes and remained const#nt thereafter /lltMn the accuracy of the %psriment. The temperature range in which this experiment cam he c a rrie d >ut is limited as the acio will sublime to th© cold neck of the tUbe when th© vapour preseura is grer % enough and this will affect he ©xchange kin© tics. A total of B exchange runs were made covering the em perate© rang© 20 « 87

10 Conductivity Studies. In lonle conductor® it ha© boon ®hown that if the mobility of a diffusing particlOg Im Imovm then it© diffusion

©O0 ffielentg Dg can be calculated by memm of th© l)8em©t«-Ein©tein ro3,atioho D ^ ,u k To where k Is Bol*temann®a const* T the absolute temperatureo If the measured conductivitypis du© to one species then It i© related to the mobility thus g O 0^ ^ n z B JÀ where n is th© number of the diffusing species per CoCo ze is the charge on the ion hence the diffusion coefficient and conductivity are related thus*

j g ^

If the seme mechanism is responsible for both conductivity and tracer diffusion then the above equation should hold* If this equation holds over a ran^.© of temperature then the same activation energy should be obtained from Arrhenius plots of diffusion and conductivity. In order to verify whether or not the conductivity of hydrogen bonded organic acids la due to proton migration the conductivity of benzoic acid and acetic acid were measured and their température dependence obtained using both A.C* and D*Go techniqueso A further study was made of the conductivity of oxalic acid dihydrata to compare it with previous experimental, results and determine the effect of water of crystallisation in T-H F L W 5WJkUt.ATôR

6 a

mwewuM

wpwN

CRYSTAL.

a the hydrogen bonded ehelna on the conductlYityo Gomduetlvlty C©ll©o

The conduoti^ity ^ao meeeured œing a simple tw o proba techniqueo The cell ua@d to measure the conductivity o f benmolc acid and oEalie acid dihydrate is ®hom in figo 2 5 9 cell BOo lo This was an b11 hrmm cello 2'® dia® e 3*® longo in which a crystal o r compact was placed between two platinum ©lectrodeso the upper o f which ©crewed down through a teflon insulator to give g o o d electrical contact with the crystalo The cell wae placed in a glaas furnace which was electrically screened and which had b n o n inductively wound nlchrome h@f-,t©r controlled through a ?ariaCo The temperature of the crystal was measured with a Stan tel M5 thermistor which projected through the bottom of the furnace and through a hole drilled in the lower electrode to within one miBo of the crystal surfaceo

In o r d e r to measure the conductivity of acetic acid a different cell design was required due to th© low melting point and hygroscopic n atu re o f the crystal 0 The c e ll i© shown in figo

25 9 cell Bî©o2o The electrodes were stainless steel g the upper being of accurate dimensions 0 o2 5 ®® diao and the lower being m larger plate 1 ®® diao which was kept in good contact with cry a toi by the lower compression springso The cell could b© evacuated and the temperetur© of the crystal measured using a ^Stantel^

1 1 5 thermistor which was locatcul in a centrally drilled hole in the upper electrode * Teflon inaulation was used throughouto Sle©tradeso Two types of electrodes were usedo Graphite electrodes were '■J applied as a colloidal suapenelon in water^ °Aquadag® and the weter evaporated off in an oven at 80®Co Silver electrodes were applied as a suspens!on of colloidal silver in methyl iaobutyl ketone which evaporated o ff rapidly on standings Both these electrodes were found to give satisfactory contacto AoCo Conductivityo Th© AoGo conductivity was measured using a Wayne Kerr Universal Bridge« The conductivity was obtained directly in mhos ;11 the smallest conductivity detectable was 2 k 10 mhOo The specific conductivityg ^ was obtained from the equation r-» where ^ is the thickness of the crystal cross ©action area of the rystalo ^ the measured conductivity DoGc Conductivityo

The D o Q o conductivity was measured using the following circu it

The apparatus used was an Ecko vibrating reed electrometer^ model N6 I 6 B0 A 1 0 5 volt high stability DoGo output from th© electrometer was passed through a potential divider m û a, voltage selected and applied across the standard reaiatoro with key 9 Kp closed « A backing off voltage was applied to the 90a

electrometer from a Py© Precision Decade potentiometer m t û

m i n g the eleetrometer b b a null point detector the voltego E was aoeurately Obtainedo K was then openod and the iroltpge drop over the resietanoe measured using the eloctroaetero Tho unîmown reaiatmcop it>©o the oryatml. reaistanaeg wa© then obtained from the equation

& ^ ^8 - ( ij ” 3- 3 P The largest standard resistor used was 10 ohms and applied voltage of 1 volt was used to minimise polarisation ©ffectso The maximum resistance measurableg therefore was 10 IS ohms assuming that a voltage of 1 m¥ could be accurately

measuredo Preparation of Grystals and Compactso Benzoic acid compacts were prepared in two wayso Initially one inch diao compacts w@r© pressed at 700 poSoio and 4000 poSoi for one minute using analar reagent acido A second batch of compacts were prepared for conductivity and diffusion studies from bensoic acid purified by single crystal growtho This m aterial was powdereo and compressed under vacuum a t 5000 poSoio into diso pellets o These pelleta^ 14 in all g m m annealed at 116®Co for seven days and allowed to cool slowly« Their density was meabured and found to be lo32^ oOl gmocm^p identical to that of single crystalso Pieces of bensoic acid were cut from single crystals both parallel and perpendicular to the {001} planeo These piece© 9Qh were cut to a suitable ©isa n n û ground pleaar on a glasa plate with fia© earbonmdtm end water and aubsequantly poliahod on ©trotched chamois using cyclohoEaa© as solvent a

Oxalic acid compacta m m pressed at 700 paSoio and 4000 poSoio using an alar reagent grade materialo Single cry 53 tala m m grown from aqueaus solution using racryatalliaed analar acido These were out and polished alniilar to th© benaoic acid

Acetic acid crystals were cut in the fridge from mingle crystal, boules using a ra^or blade 0 Due to the cleavage properties thin cryatala were obtained with planar surface® which required no polishing and v m m used directlyo CHAiT'a-; III THE R]?SULTS AIH) THEIR IKTjaJl'HETATION 91.

This ehapfer has iBta the follo^lmg B i n ©ectioaso

ao Tritium amd 0^^' û i f î U B i o n Im benzole meld ©ingl© oryetalSo 1 à bo Tritiugi and G*" diffusion in poljorjata3.1iBO bengeie aoiclo

0 0 Tritium e^^chang© between benzole a,oid and water vapour© do Tritium diffusion in aeetic acid ©ingle crystal©© e© Gonduetivity studies in organic acids© f o Tabulated results ©

Each section has been treated sepa.rately with respect t@ Interpretation of the results though experimental, results for all ©eetioxTis are included in section f in tabulated form© . —

T8T a© Tritium and Diffusion in Benzoic Acid Single_CryBtaJ.ao xteo The experijaent?*!, arrangement used in these diffusion studies was that of an infinitely thin deposit feeing allowed to diffus© through an Infinitely thick crystal perpehdicular to a known plane face© When diffusion occurs by a single mechanism the following solution has been derived fo r the case of constant J) Q n « B «Sacrjc.'SjesscEjcscïi G - f 4 92.

taking logarlthsa to the base 10 of each aide giTOS P

loëiQ c 3-«Sj_o ■• (vm )"^ ■” 2.503 ” 4 at, COWCENTRATIOW oATWE henea a plot o.f the logarithm of the^diffusing species at a pénétration depth % vemum the square of that depth ahoulcl give a straight line of gradient

fY) - 2©303 " 4 Dt Thus I f the anneaJ. tlme^ tt> a t a temperature T ia kmewi the diffusion coefficient at that temperature cam fee calculated

from this gradient© This solution asaumes that there is m lo se from the surface other than that due to bulk diffusion© The results of tritium amd 0^’^ diffusion to bensoie acid are ahom to graphical form to figSoEé'fe34^ from which it can be observed that in most caae^ straight lines of A v@© 1'^ were obtatoedc though to the case of diffusion °tails were observed due to a secondary diffusion mechamismo These graphs indicated that the solution given by equation (1) was the appropriate on© to use©

2o Tritium Diffusion Perpendicular to the (001) Plane© In this system it was discovered that variable results could be obtained depending on the expérimenta], conditions iweûo The roBUlts^ thereforey are tabulated under the conditioBs used©

The toitial tritium diffusion studies mere carried out t o large brass vesaelsg fig 19 « The results of these eicperimente are tabulated in section i and plots of log-^^â vs %' ? are shown Im fig « ^,6 © All these diffualoB rimm were carried out on benzoic acid crystals purified by distillation© The diffUBion coefficients CÊilculated from equation (1) by the method of least mean squareSp see appaMix ly are tabulated baloWo A sample calculation la given in appendix II©

TABLE lo TKIÎÏM DIFFUSION(001) M EOœ HOTBED BEBZOIO ACID (LARGE ?E S m S ) «

1 /m® A ]} -a ÏÏiffusioa Time _ Temp o cffi^aeeT Rîm lOo houffs w o « no 10-^

i m i 18.5 88.8 2.762 23.9 ± 2o6 1 .3 7 9 m e 24.0 93.2 2.729 19.6 2oO 1 .2 9 2 DR9 24 oO 93.5 2.727 24.0 1©0 1.380 DHIO 25.0 104.5 2.651 43.1 loO 1.635 DE13 19.0 108.0 2.623 100 5oO 2.000 DS14 16.0 108.0 2.623 83.0 4oO 1.920 DR1.6 255.5 78.5 2.843 7.50 & Ool5 0.876 DRI17 554.0 93.4 2.721 19.2 3oO 1,284 DB18 2oO 165.7 100.0 2.680 48.0 , ^ 1.681 1 A p lo t of 1os.jqD t <8 /T, 1 iîj® A f ig . S T 9 ®as found te be linear and foil the Arrhenius .equation

2gOOO iiiJSSiLiaOSfi D ^ 175 . mp ■ 163 HT This line was obtained by the method of least mesn squares and the calculation is shown in appendix lïlo Prom these results it appeared that th© diffusion oeefficien was independent of tlme^ concentrcatioB and eryetal perfeoti IN SEMZOIC ACID ( LARÊË VESSELS)

QÎ QZ CJ ^ O) Ô (d) £E^ X o

o

o

cO cu " o , ° / 'S'^ 3 0.1IAI13^ 901 HENlUS PLQTS FOR, TSlT/UM AtslQ C '^ DIFFUSION F I Û .Z 7 IN BENZOIC ACID SINGLE CRYSTALS

oi

CB GK ûj ^

O UL O ^ dfS w N % § . izp. 6 & Li iOL (Q ^ o rai R

o<

5 o O Ô «U

O G

o @ 0 0 c

'Ol'DV ^(OZN^g 'id'W OJo o o «ma CO vSà ^ oJ O (A vS> 04 o Ô Ô g oa — o . 3>3S "izl + a 0 1 901 I - The scatter obtained ?me clue to the inaccuracy of sectioning and the fact that some sublimation of the crystal occurred ïïhiçh prevented long diffusion annealso In an effort to cut down sublimation and avoid the slight tarnishing reaction which occurred at high temperatures in the brass vesselj, diffusion runs were carried out in small aluminium vesselso The results of these experiments are shown in figeogjJ.E^ and the diffusion coefficients obtained tabulated belowo

TJM3LE II TRITIUM DIFFUSlOMjT(001) IJM BENZOIC ACID (SMALL VESSELS

DIFFUSION TIME TEMP. D Log D CRYSTAL 1 //ip® . I A KlIN Ko, h o u r s X 10 5 cm.sec2 ;%1 10 SOURCE 1 ■M2 X 10^^ DH. 56 116 105.5 2.640 lo4 ± 0.7 0.146 ZKc5 DKo57 137 105.5 2.540 lo7 0.3 0.230 ZR.5 DK.58 240 110.8 2.602 0.3 0 o 568 NON Z =K. DR.42b 148 99.5 2.682 2oO ±: 0.4 0.301 ZR .4 DR. 45a 166 99.5 2.682 6oO 0,779 ZR.4 DR.43b 166 99.5 2.682 7 01 i; 0.3 0,852 ZR.4 DR.60a 160 115.0 2.576 4o85 0.4 0.586 WON ZoR., DK.60b 160 115.0 2.576 3o57* % 0.2 0.523 70" Z. R DR. 62 72 110.0 2.6095 6 o 26 :t 0.3 0.798 WOW z . R. DR.63a 160 115.0 2.576 4o05 i: 0.1 0.510 2R.7 DR.55b 160 115.0 2.576 3o30 1 0.1 0.519 ZR.7 DR. 69 134 98.5 2.6905 loOO 1 0.1 0.00 ZR.7 DR.76a 324 110.5 2.606 2o74 & 0.1 0.437 WOW Z. Ko DR.76b 324 110.5 2.606 3o24;% 0.1 0.510 HOW Z, R. DR.84a 222 115.0 2.576 2o93^ 0.15 0.467 ZR.5 DR.84b 220 115.0 2.576 2 o 6 6 & 0.2 0.426 ZR.7 DRo85a 255 103.0 2.658 X o45 0.05 0.163 ZR.4 DR.85b 255 103.0 2.658 lo56 0.07 0.194 ZR.5 “ 2 , . A Log, q A t o X p l o t f o r t h e s e resu lts! is shoun in f ig . Z7

p o i n t s ® on1 l i n e B and A above it. A large exp e r i m s r r t a l ) A v s % TRITIUM DIFFUSION 1 [0 0 I^ PLANE IN &EN 10IC ACID ( s m a l l VESSELS),

X J

V ^ O 0 O % ^ O

jQ) ^ S? 0^ r'~ si) w tO 3 *- fv/à rA ^ ^ ^ 3 K c/ e£ ml ^ af oeiiûûQûanQa

"'( -■ \ >f * ; . LOO 10 A vs TRITIUM DIFFUSION ± (OOlJ PLANE IN BENZOIC ACID (SMALL VESSELS),

O O <2 X

o

s

o X

tso ^ ^ o*x»via cO CV# . . . * . u- o o 'Ul b w /se h f LIIAUDV'^ ®'SOI r m Loq 10 VS X- FOR t r it iu m DIFFUSION 1 PLANE BENZOIC ACiD UNDER DRV CONDITIONS

o c3 X O

Z . d OB ^ Z ç;£3 ^ & 3 e? ^ o ïï£ o

L, I.,-J-1- ^ .—ËSÀiiV.'..—Jl cO . CvJ W Gw /'S'el'O f'AJ.lMl^v') 01 n o i 93 scatter Imltlally observed bat all results were at least ai). scatter was Inltiallj obeerved bat all results were at least ai). order @f magnitude below the diffusion coefficients in the largue veaaelao Initially it was thought that it was due to crystal perfection» The results shown in fable Ilg hQVJQYùr^ proved that this was not the case» It was found that lower diffusion coefficients were obtained when the crystal wea not côoîâd during deposition» This indicated that the diffusion coefficient depended on the water vapour concentration in the diffusion cello When the cryatals ware not cooled during deposition th® diffusion coefficients were found to Ik on line B fig;» £7 A apaclal effort was made to thoroughly dry the crystals and diffusion cells in DRo84g 65 by enclosing them In^PpOr cell in a dry bo% for 24 hours prior to the diffusion anneal » The diffusion coefficients obtaihedjalso fall on line 1 indicating that this represented a lower limit to tritium diffusion in these erystalso Line B obeyed the following ji\rrhen:lus equation

2o0 r 20»000 ± In 300 B ^ Oo5 EXP I - *^0 o 4 L HT The résulta of tritium diffusion experiments in p^terphenyl skoWKi Sù wd doped benzoic acid arej^tabulated below» These were Cf*rried out in aluminium veBs>el0 without cooling the crystalo LOtolo A vs ;c f-OK TRITIUM DIPPUSION J. (OOy PLANE r/q , Dw IN D£UTEI50S£NE0/C A%:ID AMD fa-TERPHENVL DOÇfD ÈiSNZOïc. Acid .

<3

é dü) Q ^ ST vO Ê^' oi G£ lof (L^' m O O £H Ü a a Ci.I z f NOo J g g «a § II w ^ & hr tLbJ) hr O-

w G w /'9^0 CAHAIIOV) '^'901 TABLK I I I . TKITIim UÎFFÜ3I0B JI(OOl) l'â P-TERPHENYL DOPED BEinZQIC ACII (3MALL VESSELS) o

1 o D i-'IPPUSlOÏ) , 2 = =1 Log3_oD TIME TEMP, /T°A CittoSec.) iriUi'S i<0„ hours V 0 1 9 X 10^ 2C 10 4- 12 l)Ko67a 256 100 o 5 2.676 3 .1 5 * 0.>1 0.500 DRo67b 256 100.5 2.676 2.50 ± Oc.1- 0.400 DH o 66 169 90.0 2.7535 1.50 & Oc,22 0.177 DH oBOc 250 115.0 2.576 4.01 S 0(,05 0.604 DH.806 250 115.0 2.576 2«6o * 0.,12 0.415 DR„85c 255 103.0 2.658 2.26 a Oc,1 0.355

A Log,„D ■vs ^/T plo t of these re s u lts i s shovm ;In fig .2 7 5 ips'wll line D The Arrhenluîs eqiif^tion ?me obtained by the least meaxi squares method to y ie ld

4" 75 7»450y ± Ci,g 2,200 D 31 XO^■8 EIP “”4 o 4 «n> RT .J

A series of diffusion runs were made on deuterobenzoic acid to determine the isotope effect on diffusiono The results o are tabulated in section f and ra plots in The calculated diffusion coefficients are tabulated belowo i: ]

TABLE IV TRITIUM BIFTUSlOB (001) XU MUTEHOBBWZOIC ACID.

a,

DIPPUSIOM TÎMK TTMP. ■ /ï'^A sra.ssc. VE33KÎ, RUR j o . h o iv m ^C. 'it 10"'4 X 10 ^ USED, ’”X DR.M./|? 47 120 100.0 2.679 2.812.81 * * 0.1 0. 1 0.449 8MAM, DR. 12 112 115.5 2.573 11.0 1.05 SK.fila 67.5 115.0 2.576 6o9 ± 4 0.856 ” DK,6lb 67,5 115.0 2.576 5t>9 j: lo 5 0.771 DR.72a 204 110.5 2.605 37.6 1: Oc7 1.575 LARGE DR.72b 204 110,5 2.506 36 oO ±: lo il­ 1.556 " DR.80a 250 115.0 2.576 3on±- O' l s 0.502 SMALI, DR.aOb 250 115.0 2.576 3o33^ 0 M 0.528 " 3. A p lo t of q D VB / T %B i a c l .uded in i'ig m .

Th© r'Kk&'XyslB ©f th©s© reaultî-g is C 0 mp.licat©d by the fa.et that in DHo47g 52ÿ BUÛ 61 the c ry sta le were ©ooled during depoailion thue explaining the large variation obaerveclo Thia meant that an Arrhenius ezpremBiotx could not be obtained<> DRo61 g 72 and 80 illustrate the fact that oryBt&Ll© armealed in the ssi^e cell under the aame conditional give almost identical diffusion coefficients though the actual va3.U0s vary widelyo The Isotope effect mi tritium diffusion waa obtained by carrying out diffusion runs on noma3> benzoic acid and

deute:<-ob0 ?;ASoic acid in the same cell « The results are tabulated bel©Wo 36.

ÏABÏJ'i; ? ÎHIÏÎIIM DIFPUüIOkM BEMZOIC ACID AMD DEliïBBOBEHZOlfi ACID SîMULîAh EOOSI.Ï .

DIPPUSÎOi CRYSTAL D AVERAGE Mo. ? B&fW ..^12

DR.7§a 37.6 DR.72b ÏÏ.B.A» 36.0 1.21 DHo73a R o A. 41.0 DR.73b BoA.

DR.80m D oB. A o 3,17 DR.80b D.B.A. 3.33 1.015 DH.BOc B. A o 4.01 DK.eOd B.A

DR.82a o A DR.02b .D o B o A o 15o0 0o75

To determine th© effect of th© size ©f veseel used in the diffusion eimeal a diffusion run was in a large vesselo The cryatels were not odoled an deposition and they vmre wrapped in aluminiums foil to suppress aublimatim The results are tabulated im section f and fig., o fho diffusion ooaffiGienta obtained are tabulated below» TABLE VI TRITIUM DIFEUSIOM iTiOOl) Ifl ZOMS RKEIiED BEBZOXC AGID (LARGE VESSEL;„ 1 DIVFUSÏOM /T"A -.1 .3 eia.? see.'* RDM Mo. %j o j'i. A A DR.73a 110.5 2 . 5 4 1 ,0 1 1 .2 3.2.63.2 ‘7*?4 204 o5 2.6095 4 8 .0 1 1 .7 11.688 V 'à (' /A OR. TRITIUM DIFFUSION A L ( o o \ ) PLAN a c id and DEUTEfROBENZOIC ACiD.

o ; g o ! A e/ ti MM p Z G O û û et 3 O ^

'w B w /'S -qb f^A J JAUP v} 99,-

The of DR.72 plao carrlao out In thlB eell is tabulate# 1 ill Tehle IVc A plot of ^D VB /f of. thaa© la ahoma Im fig^EXi? pointe Thaa© dlffumiem ooefficieiita lie almoet on the uppex* linoo Th© lower mlues are oonalde/md due to the f^ot that the orjatala were not cooled on d©poaltion and subsequently contain lesa water Tapourn

An 0 Kperlment wae performed to observe the tritium diffusion rate In beiraolo aeld cryatale in the prasenee of TRO Tapouro lifiaetiTO c ry s ta ls were prepared in th© umual wayo They wore placed in a glass tub©^, evacuated g and a little THO vapour introduced^ A small amount of dry nitrogen was introduced and th© tube sealed under reduced pressurco The crystals were amnealed and sectMmed in the'usual way c The results are tabulated Is se ctio n f and shown to figoSE o The d iffu sio n coefficients obtatoed are tabulated belawo

TABLE ? ï î TRITIUM DIFFUSl O i X (001) IS BSSZOÏC ACID lU GOITÀCT WITH THÏÏÎATBD WATER» (GLASS VESSEL)»

DÎFFüSIOf TIMS /T®A 2 -2 LQiSiq D CRYSTAL RUM Io„ hours (.4 O s 10^ ,5Ï 10^^ 4» 12 SOURCE DRc82a 250 115 oD 2.576 10.9^ 0 .4 1.036 ZR.7 DR»82b 250 115 oO 2.576 15.0& 0.5 1.176 D oS o A o iq. 3'<£.

L 0 6 10 A VS ^ FOR TRITIUM DIFFUSION J. ( 0 0 1) PLANE IN BENZOIC ACID FROM THO VAPOUR,

rOCVJ m g oi

CM

CJ

'LU / 'sW'O (AllAllOV) 01 9 0 1 Thaae Imdleated that the ÎHO G.iseha% 0 d oaaily th the bemsolc aol 8 eml diffama cl te a sim ilar mij to a l l other trltimm diffasioa msüBo fhe diffUBi» ooefficlesit© obtained m m 3-4 times greater than those obtained nMer dzqr ceBfiitionBo

p iffu s ic i.1 of leus^olo Acid - ^ (001) lllî-ne in Ban^elo A old The results of this stud^/ are tabulated te section . f and vs ii plots in In ell or se tiw dlatinet limes were obtained inclicf'ting that diffusion m& oocurrteg by t%o different raeehfniamso The slow initial d iffu sio n warn aonaidered to he clue to bulk diffusion of the bensaole acid moleeule and the fast prooess due to diffusion Qt bemsoic acid moleoulea down dislocations or other structurel defectt> The bulk diffusion coefficient vim obtained by assuming thar. i.he diffusion profile represented the superposition of two Independent prooegsaa The p lo t of log A vs for the bulk precees was ©btàteed by pro jeetteg the ’’fa st‘d diffus lorn Itee back to @ amd subtracting the contribution due to It

1 row the to ta l actrvitywtrao The diffusion coeffioienta were calculated as for tritium diffusion and are tcteulated belowo UWV.J ly rwrv kw uirrvaiwN -L. ; ru^N t IN BENZOIC AGO,

O

w m a a Ù

\A) £ui I S '4'3 ( a 1 !AI1'3V ") 9 0 1 l iQ. LOaioAVS c'*^ DIFFUSION 1 (OOl) PLANE BENZOIC a c i d .

J} O ; @ Q e» @

ij M c8 r- 6^ 3 ^ {f^ 6^ iT* r** m.

o

w Svj/s'd'o (AiiAiiov) 0 1 ^on ao*

T.ABLB VI II BSSüZOÏG AC IB - ,14 0 BULK DIFFUSlOM j «M BE.N20IC ACÏDo

■"/TA p DIFFUS mw TI.W.E TEMP. @!iîfsee*.^ Bog'll q D CRISTAL RU|' too hsura ■’Co s 10^ SOURCEV X 10^"^ 4" 1 4 DR „ 22b 403 105.2 2.6425 20 <,Q ^ 10 oO lo740 mB ZoRc DRo23a 329 116.5 2.566 55.>0 5oO loT40 mn Z oH a DR.23b 3.1 6 . 5 2.566 69 io848 mn ZoHo ÏÏH.24 54? 110.5 2.605 13 t 5 1 . 1 1 4 mi^ ZoEo DRo32 3.72 110.0 2.610 5 0 * lo700 EoHo5 DRo7?b 236 110.5 2 . 6 0 6 3 2 ± 2 1 . 9 0 2 SoEo4 DRoTle 236 110.5 27 Oo5 lo4M B„B 0 A 0 DR. 78$!. 496 103.5 2.655 8,>6 ± loO On 936 ZoE 0 4 0 DR.78b 4 9 6 103.5 2,655 11.>0 fa Oo3 JL. 0 W ^ mn Z oH'i DR.79b 202 115.0 2.576 66,,0 è lo9 1 fj820 ZoR o5

P A pl©t of % % ia ahom in pelmta # amd T o Theae diffuai©.^ ©©efficiente are exceptionally fàiaû approach the lim it of the Bcctioiaing methodo Dlffuaiom rm a ï)Ho22 ^ 32 were c a rrie d ©ut iiBing the original sectioning technique (see p« amdar© leas accurate thmi which mmm carried out ueing the suction technique the accuracy of the experiment the bulk diffusion independent of crystal, source and the data excluding DRo22g 32^ Vi\9.mÎQimû to fit the Arrhenius 2 -^7o6 % 10' B lo8 % 1012 '1.79 / C5.-^ BenEoic Acid « 0^^' Diffusion (001) Benzoic Acido fhe result of two attempts to measure diffusion parallel ,earag© plane are tabulated in section t and log.^A to 102-

21^ plats shomi I n fig . 3+ . Th© profile obtalMd was sim ilar to that for diffusioxi perpenâieislar to th© cleavage plane and diffusion coefficients were calculated by the same procedure and tabulated below. 14 TABLE IX G DÎEFÜSÎOM 11 (001) BœOÎC ACID. 1 /«i© A 3) DIFFUSION TlMl TEMP. 5 L©g,jvD CRYSTAL ROM ÈΩ. hours ®Go % 10^ +14 SOURCE 33 182 109.8 2.S11 45 1.65 2.R.3 34 162 100.5 2.678 25 1.40 ZR.3

Dm® to the email -©ryatals uaodp aurfase area Oo5 ©ii y and the aubsequent short diffusion anneal these results are mot eomaidered partieularly aeourate© They serve to imdicatejj however^ that diffusion in this direction is very low mû approaches that obtained perpendicular to the (001) planec

The '’F e e t’’ D iffueion Process in B©ng;oie Acido This process was found to vary with the crystal usedo The diffusion coefficientajericiilated aasuming that the profile was linear and could b© calculated as for the bulk diffusion^ gave

1 C* *S T O 1 values in the range 1 x 10'"^ to 5 % 10^' cmo 0 ee% but there wa© no pattern evident In a log. va 1 /T ploto It was sumed that the diffusion waa due to a diaXoeatlon or other structurel crystal defect mechanism bM no fu rth e r vioxk was attempted on this systemo L O f i l î o A V S % 0 “'' Il ( ^ 0 0 1) P L A N E H é s. 5 4 BE

OS Ç Z § 2

O o '%' O o CvQ ô o ô o ô 103.

Surface D iffus ion in Benjsoic Aciüo In aevara,! tr ltlim d iffu sio n the baelc facje of the crystal was removed by scraping it off aith a raser blade after the diffusion annealo The section imm weighed ami countc In all oases it was found that within the experimental error the activity of the back face was th© same e.® the front face indicating that there was a rapid surface oqulllbrlmm of tritiated bensoio acid caused either by e cmrfac© diffusion or vapour exchange processo The 1 A diffuaiMp howeverp gave a different results In this case the ©urf©.c® activity wa© monitored using a portable GoMo eountero The surface activities obtained.are tabulated belowo

TABLE X

BOTUSIOM AWÈiEAL COUMT SATE GOUSSIT RATS RUK isl0„ TIME PROMT SURPACK BACK SURFACE hoiirso C O p O ® Û Cop o ® o 77a 236 350 220 77b «0 360 220 77e T9 500 220 11Û. 99 500 220 78a 496 500 250 78b TV 500 3.00 79 202 500 300 These results show that surface m igration of benzole acid - is much slower than trltiated benzoic acid and must go by a d iffe re n t mechanlamo GMÜ.N BOlJïlBARy DIPFUSIOM OF TRÎÎIUM AMD G'^'^ IM FOÎAT.RXSTALUMB BaVXOiG AGÏB,, The l ' i attempt to solve the problem grain boundary diffusion was made by Elsher» *" His solution was approximate but simple to apply to experimental resultSo i4- The exact solution was published by Whipple but was in a form that was oifiicult to apply to experlEiental results and hence until recently the Fisher solution was the one commonly used o A comparison of the two solutions has been maae oy Le Claire who has published in graphical form some numeilcal evaluations of the Vu'iij.iple solutiono Tnis enables estimates to be made of the error involved in using the Fisher solutiono The matnemaLical moael usea in tne analysis is to represent the greln oouuoary as an isotropic slab of material of width B within which diffusion occurs accoroing to Fick°a lawsp with a coefficient much greater thmx D» the coefficient of bulk diffusion on either aie® of the boundaryo With the boundary conditions

C-(Xpyj,t) s C o ; t > o ; y = o where y is the direction of the graiUg z the direction at right angles to the graiUg Co the constant surface concentrtlion rnd t the tiaie of diffusion Fieh@r"a solution is

^/Co“ y p ) erfc & ^ (t) 105. nhsre vÿ j, ^ pnd ^ nr© the reduced almeasloD leaa quanlrtiee.

,n r< o' 4 % g .„ 7 ® CStÿfe i /* " *D iDW%. 5 ) ' whare

T, ^J œ . 1 / Sias , . . = \ t.)

In the Fisher solution the iBst term is a conatp Tf hence s plot of In c va y should be linear mû a vain© of 1 c F’a easily obtainedo -

Le Cl el re has evaluated c) ( 9 } as a fimction of ■for variauB values of p and presented it in graphical ÎQ'm thus allowing theWhipple diffusion coefficient to be evaluated«

The résulté of the diffusion of Benzoic Acid and tritiatod benzole acid in polycryetallIne compacts are shown in the foim of (specific activity} vso î the penetration Into the crystal in flgeS%#&raapec11vely. Acooîélng to the Fisher analyaie these should be linoar ezcept foj the high point at the surfaceo The curves are mEssentially W I i t ; L /HCID COMPACTS,

o % 5>~ O 0 X - f o E3 <1 sill ÇÎ JQ -Û « w W W ^ (U N(V 0^ £!f Û£ ûi ^ ca a o o o o û

XO

1 ,.... >1., I I o O "4" 15 U) UJ g vu I %el3 r ^ ll AIJ.3V) ®‘?)0“l L 0 6 A vs, X TRITIUM DIFFUSION IN FlCq.36

BENZOIC ACID COMPACTS

a O CN

O s J o v3o 0 o >

^ ^ ^ ^ «20 O OQ . , . ^ S2f iÉ 2 2 Û Cl Û ÏL, Oi

'W ' ^ U I/çfifo (Al/Auov) "^'901 1 0 6 , linear v.ith some Hailing*' occurring at deep penetrstioneo This cur'^atu.re is in agreement vâth that predicted by the Whipple soIutioHo The grain bo unde ly diffusion coefficients i?e re calculated from the Fisher equation„ 1 ■6-- . s . fi K , Æ 8 X Go @%p w i I I D*'’/ \J) % } J assuming erfc 5 ^ 3., using the linear part of the curve to obtain d Inc, The values dy of jD used were calculated from the bulk diffusion ©quation

C©G0 Poll©!) obtaineo in this investigation, The value of c to be used was the only unknown quantityo Other grain boundary studies have sho?m that d is of the % r n oraer ©f 2 3 lattice spacing® hence a va3*ue @f & 251 was assumedo The g rad ie n ts were calculated by the le a s t mean square method and the diffusion coefficients obtained tabulated in table II ûïffimlon ©©'^efficient's ©btaiy^9# by |S t© " from which th© valm ©f

;sxiVirriviTi:?C'{?-’'tSy I^AI graph, Ths PH UTOd In ealctilatloma are ahcwm in tabl^ Kilo '"ha calculated valuer of the Whipple diffusion coefficients are also included in table XI^ 10? o

TABLTi XI C14 BOURDAKY DIFFUSlOR ÏR BENZOIC ACID. 3, 1 _ îdT?- ah er LogjQ l?hippl@ DOgjg â c=>j|, ©fiioseco D^ lîhip'pla ç " - X i F ish es* 3.0 „l4 ° 3: 10"° “5*10 H so DRo 26a 91.0 2.746 38.4â2 1.59 102 2.008 26b 91.0 2.Î46 55 A 4 1.75 147 2,167 27a 103.5 oVmg 260 »2Q ' 2.42 415 2.618 103 .5 -g:r5^ 260 a 20 2,42 415 2.618 28a 78.6 ê;843 5.760.2 0,756 12.9 1.110 28b 78.6 2.843 7.4® 0.5 0.780 16.7 1.223 1.47 66 1:820 "’•5 2.805 ^ TRITIUM 101.0 2i6î3 243 & 60 2.386 350 2.545 71 78.5 2.843 20 & 2.5 1.301 34 1.532 86 96.1 2.708 115 & 10 2,061 198 2.297

TABLE XII CALCÜLATIOfil DP LBOLAIRE PARAMETER ’A"

DIFFUSION (Dt)^ RUN 5 ^ 4 f P A X 10 em. 5ï 3.0 et0. DR. 26 “ 3.63 68 107 385 9.5 2.01 21 8.15 100 123 4000 1,95 1.10 28 2.19 56 256 1800 6.05 1.30 29 2.60 §6 370 3500 6.25 1.31 66 3.88 . 120 310 7830 3.50 1.20 71 2.21 100 453 5650 6.40 1.30 86 8,26 120 146 554 6.20 1.31 Plots of lot-inu D Vd Vrp .1 fo; both sets of diffusion e ffic ie n ts ere shown in fig u re 37^ Both Fisher and Whipple nes are approximately linear and H it the Arrhenius equation : p exx) r I where the individual p^i^rnmetors are given by o i— ' h i '"‘I 3 following equations^ " Diffusion

37„100 ± 2^200 F isher ,■5 -«-3.5) X 10 I2CP RT =7.0

Whipple (i„i ) X 10^^ KXP 3%500 2,500 0.97 HT Ltium D iffusion

Whipple (loB ® ) 3c 10^. EXP 27JL00 ^ 700 “lo 0 HT

These results show that within the experimental aecuracy ,tium and d iffu se by the same rnechanismo The decrease in activation energy obtain@d on applying the Claire analysis is In agreement with that predicted/ 4 A combination of the tritium and 0“1 Whipple data fitted following Arrhenius equation, 31;^0 ± 2,000 1,0 'f 13 \ . .11 «tt?Mo»*ttCTc ------— " is. HT r I ^ / AKRHENtUS PLOT FOR GRAIN SOUNOARy

diffusion in &ENZOIC AOD COMPACTS, C\ CKp“ O--' !?■ 0 r D

Ci o o Ô

w o o c

vD

'QDV 3I62N39 Id § o ■v^ oo CO Ô CJ % C\J CJ 'T «X o O ? WO Ol-f- *^‘001 0 1 M

Co Réaction Between Xritiated Benaoie Acid and .Viter VapouE It I? as foundjthat when tritia ted foenaoio aeid and Inactive water vapour were in contact an exchange reaction took place and the water vapour became radioactive^ The* reaction was found to obey neither first nor second order Kinetics but instead gave llnee r plots of radioactive concentration in the vapour vs, scuare root of exchange timCo This indicated diffusion controlled kinetics similmr to that obtained by 'Kitknecht in inorganic hydroxides with the exception that no initial rapid surface e^chang© was obaer^ The exchange equation must^ therefore^ have the following form^ Ht A . (Dt)^ where Ht ^ the vapour phase activity at time;fcp is equilibrium activity of the vapourj, A is a constant depending on the aurfac© area and geometry of the crystals^ D tne d^ffusion

The results of the exchange experiments are tabulated in section ERl bo ERlOa Plots of Rt vSo ^ are shown in fig 38^ Although there la some scatter due to inaccuracy in the samplihg method (poB5) a linear relation is apparent indice.ting that equation (3) is the correct solution^ The gradient^m^ of these slopes is given m ss . A ,

H to ACTIVITY OF VAPOUR E s Vs FOR EXCHANGE REACTION BETWEEN TRlTlATED BENZOIC ACID AND WATER VAPOUR.,

RUN N®, SYMBOL jf . C3 £R, I A £R , 2 -m - X ER. 3 'Q V ER. 4 0 8 a E R; (o B ER, 5 4 ER, 9 0 ER. 10 Ü (3 V X 0 Q 0 B

i % © 0 © 0 V A

I

A

X---—L J______L J______L 10 II 12 G 14- 15 i(o n IS i9 1 1 0

Values €if m were obtained by th© snethod of least mean aouares» Values of EtoiiSTO calculated fxom the volume of the cell g water vapour prelawo and Weight of sample„ Values of *m vjero Ho-0 caloulatod and, ar© tab u lated baloWo TAB&B Kill mmTIVB BlWmiOK GOBliPieiEMTS OBTAIBKD i?EOM EIGHAMGE ixpimMfflTao ^oHo TEMPo j4mA WToTBAo H Vapour Total Boo m * m Oo mgm

1 22qO 3o3a© 6o40 8,43 47o4 543 0o9?6 ic79 Oo253 2 53o0 36059 5o50 9o85 ^ 40l? # 5 2cIM 4c45 0,549 5 65«o 2,947 5,00 10,8 ' 37,0 439 3,59 8,25 0,917 1 3®oO 3 , 2 2 3 4o90 11.0 36.3 428 1.20 2.80 0.448 Î l l M 2 „ B 0 5 5.30 P.83 40.? 536 6.22 11.20 1.050 5 ?S = 3 2o8S0 5.70 9.3 42.2 5? 9 11.13 19.1 1.282 5 43oO 3.162 5.03 10. S 30.7 543 1.97 3.62 0.358 LO 60o2 3.000 4o9o 11.0 36.3 343 4.00 7.35 0.866

A p lo t of l o ^ i Q ff." (®4Di) i s shown :I n fig u re 3i). This

(lot is CSSiantially llm oar in %@ rang 1® 20 " 70°0 h®ne© the pa “OCÔSS ^an be represented by the followlY% Arrhenius e^uationo 14 nOOO ± 7 0 0 1 RT Am thee© are relative diffusion coefficients a pr@=essponen11 al actor cannot fee calculated. This activation energy represents a a%lmum value as the increasing diffusion coefficient at high temperat re is due to the superposition of another exchange process which^ rom th© linear plots also obtained^ indicate tkt this ia due to nother diffusion meehanis®. This may fee the bulk diffusion of BRBolc acid m©lecul©So o o ARRHENIUS PLOT LÔ^iq Û VS '/ t FOR TRITIUM

EXCHANGE HaÛ / SENZÛIC AùD.

O ©s ep r> OJ •îj" . r/1 ^ ( 3/, a , w °' 901 X X . '

Trltlmi diffusion perpendicular to the (100) plane in acetic aeido This was found to b© very difficult to handle as %11 operations had to be carried out in a deep frees© at « 20^0o fh© diffusion experiments were carried out in an identical manner to the bensoic acW systdm<, A total of 19 diffusion runs were attempted over the temperature range -5-10 to - 15^G of which only 4 gave reasonabl© resultSo The other eleven runs were lost due to the poor alignraant or very uneven surface after the diffusion annealp the crystals splitting when the sides were cut off prior to sectioning and lack of activityo Due to difficulty in obtaining the weight of each section it was assumed that microtome was accurate and the specific activity was obtained as counts per second per 10 micron section^ Th© results of the acceptable diffusion runs are shown in figure

4o as plots of log^^QA^ VS % * There is eemsidarably scatter in the experimental points but th© diffud. on profiles are obvious curves in all cases ^ except run 4p indicatmg that equation ( d ) is not the coriect solution to this problem^ if ^ as with diffusion in benzoic acid^ these curves represent the superpositi©^'^ of two difxusion process then an estimate of the bulk diffusion Coefficient can be m%de from the Initial slope of th© curve^ TRITIUM DIFFUSION + (ioo) PLANE IN ACETIC ACID, Fl(%. f 0

OJ (\i

co

O i------I à____ I____ I____ i____ L_ -r 2^ oo -^rr\

31 maximum toIuo of tMîSbulk diffusion coefficient^ are tabulated below mBLE mv TRITIUM DIFPU8I0M IN ACETIC ACID.

ROT . TSMP, TIC«S MA1ÎMUM BULIC oC houips DIFFUSION COEFFICIENT 1 Oo0±„O5 25,0 lo3 ^ 10' 10 Z »13„§i .Og 25,0 le X 10 ‘11 5 -5o2 ± »05 48,0 5 X lO' -11 4 0.0 & ,05 48,0 1 X 10

Inspection of these results shows that there is no apparent correlation between diffusion coefficient and temperature. It is belie ved that the variation in these rest^lts Is due io variable amounts of water vapour being absorbed on the crystals >rlor to th© diffusion anneal. It has been shown that th© conductivity of acetic acid is very dependent on adsorbed water p and alao that dlffuiM. on in bermoic acid is very dependent n the w ater vapour in the d iffu sio n c e ll. Hence i t was concluded hat a similar phenomena ivae occurring in this system. Due to the ifflcultio© involved in Improving the diffusion technique this ystem was abandonedo The intrinsic diffusion coefficient of ritium in acetic acid £ |00 ) plane was concluded to be leasthan 10' 11 a t 0% o GoKiductivil?/ Studies; lo Conductivity in bamoie aeid The reeuitapf.â«Co conductivity measurements on. analar bensioic acid compacte are shown in figure 41. These results w#re obtained on the same compact using a heating rate of poSC^/min. In the first rim^ run. tra n s ie n t e ffe c ts were, observed in the region 20 - §0^0 which were not evident in the oubsequent runs. The Cf^nductivity was found to deerease. with . each run probably - due to a eintering reaction taking place which io indicated by the curvature of the Arrhenius -plots above 100^0. In the temperature range 40 - 100^0 runs 2 and 5 had the sam© activation energy and run 3 can b© represented by the following Arrhenius equation.

C» % éo o T 3^1©

Attempts to measure.AoOo conductivity in compacts of purity benmolc ado (p @9 ) ware unsuccessful a® only transient effects were observed on the ve%y lowest .scale of the Instrumant.

C » 1l 1 «ai This indicated that the bulk-^ccndwtiviiy warn lass than 10^^ ohm”

Attempts to measureBoOo conductivity in single crystals and polycrystals was also unsuccessful*-which indicated th? t the bulk seaduetivlty was less than below th® meXtiag point. These results suggest that the conductivity in the analar Compacts are*-due to* an-im purity-of which- th e most - like4^- te w-ate-r m m ©tterapta were #md@ to dry this material prior to pelleting. \.c. COMDUCTWlTY BENZOIC acid COMPACTA

o O «

o

o Gw (T^ ‘ p , - ^ j P l - AilAli-soQNOD 2o Conductivity In ace ti e

A3 the conductivity could not be m©asu%*ed * dov/n the hydrogen bonded chains {p4S ) an atxempt was made- to measure it at right angles to them down th© (100) planeo The conductIvity of oingle crystals of acetic acid was measured at 0^0 with PgOg at thm baee #f the oondim©ell figure ?S 1!h© AoOo conductivity.?ms found to vary wl# time and th© iPf@rliî0 ntEl r©s®lta #\r© sfc\ofm graphically in figure 42 o The crystals were placed in the- oonduetivity cell at =>20^0 in a fridge and then inserted in the, thermostat batho • In run GHl the co nductivi-ty rose to 6 sr mho then decreased with- time aa shown to 2 % mho after 2 hours a • The temperature was allowed -to warm up slowly t© IG^G over a furtte 2 hour period but th© conductivity remained unaltered^ . This was the lirMt ©f detection of th©-instrument heno©’itrm© eçnolûded that tho) bulk mnductivl%r was le ss than 2 it 10“^^ ohm as th© 8Gll constant was approscimately onoo In CH2 the cell was not dried thoroughly prior to inserting th© crystal and some water droplets were observed on the iifie of th© ©ell; in this case- the PgOg was not sufficient to absWè all the moisture ?*nd a rapid Izaereas© in «^inductivity was observed due to adsorption. . if water on th© aryetal* On pumping th© crystal to remove the water ^ rapid decrease in - was observed and the conductivity igain f©13, t© a very low valueo TIME m iA T lO N OF A.O, COWOUCTlVlTY IN ACETIC ACID SINGLE CRYSTALS AT 0®C, FlG,4 2 .

RUN SYMBOL.

Q (tN THIS ' THE TIM

•070

So 100 i$o COMDUCrfVifY OXALIC ACID DlHYORATE COMPACTS

RUNN2. SYM BOL.

O

00^1 '(3d55 \ *0054- u.w, WW|\iu;uv-M V uI /nwlLJ U I n TU NMI c.,

RUN N2, SYMBOL, A 0 X T4 20 Q TA

-£ : I •2 .

t'O i

i ■1

'£

'4 -S •2 I l 'L f « 7!a bulBtoâ He su it This seetloa includes the results of all successful diffusion raias attempted in benzoto aeido The runs are numbered and tabulated in the order that th©^ igera performed and rai-ge from BH7 to DH 6 6 0 These résulta are followed bj the results of th© essohange reactio n between t r i t i a t e d oenBoio aoid and water vapour and are numbered ERl to ERlOo ISlomenolatur© - The nom©ïAolature the d iffu sio n re s u lts ia ae folloi^/So Total activity of the section^ counts per seoondo A Specific activity of the sectionj, counts par second per mgm

:îi' Penetration depth to the end of each sectionj> mieronso • -IS Somre of the pénétration to the middle of each sect!on^ P 6 cm" 2c 10 & Penetration to the midole of each sectiong microns^ BïPFïJSiOa .tUÜM JSTF .16.,2(> hours at 80.,e i: 0 .l‘^C

3etio n log A K X2 V 1 10 0.25 2 17.85 1.513 20 2.25 3 7.09 0.913 30 6.25 4 4 ^ 65 0«661 40 12,20 5 1.4S 0..184 50 20.30 6 0 . 60 30n2U

BXFFlioIOM RlhN :DH8 24«G hours at 93.»2 ± 0

8ect?.ou HA .L O ë ' A .21 1 10 0 .2 5 2 4.84 0.782 20 2.25 3 3.26 0.62:5 30 6.25 4 1.59 0.367 40 12,20 5 0.30 -0.380 50 20,20

JDIOTUSIOh UUh B R 9 24.0 hours at 93.5 i- Oo2^ü deotloxi 11 Log A 0.062^1 2 221,0 2.533 1 00 Of.. 3 262 «0 .42 ' â. 75.46 X o V * i t'i DIFFUSlOîl RUS :OK10 23oO hours at 104.2 ± 0.2°C

Se G t i on P's Log A Au 1 36 o 2 2.070 5 0,062 2 61 c 8 2 oOX9 15 1.000 3 68o 6 1.940 25 4 .0 0 4 40,0 1.724 35 9 .0 0 5 35o3 1.548 45 16 0 00 6 1.340 55 2 5. 00 7 10.3 1.038 ■55 3 6 .0 0 8 3 J.0 0 a 498 75 4 9 .0 0

MFFUSIOh KüiNi DKI3

19-0 hours at 108 ±_ 1 ,0®C.

Section «s Log A X X" 1 30.0 1 « 386 10 0.25 2 27.4 1.490 20 2.25 3 23.15 1.45 30 6.25 4 31/7 1,48 40 12,20 5 22.8 1,366 50 20.30 6 21 o 0 1.315 60 30o20 7 11.7 1 c066 70 42.30 8 ô e 4 3 0.735 80 56.20 9 2.40 0,394 90 72.30 10 1.26 0.104 100 90.20 DÎ'FF’UdIOt't RUM DR14 16.0 hours a t 108 ,0 ± 0

S ectio n Log A X 1 750,0 2.630 20 loOO 2 IW. » 5 2.455 30 6,25 3 101,5 2,272 40 12,20 4 59»6 2,140 50 20,30 5 55.8 1.924 50 30,20 6 35«7 1,586 70 42,30 7 6 ,2 1,007 80 56,20 8 4 0 42 0,898 SO 72,30 9 5.08 0,722 100 90.20

UIFFÜSIOM RUM DHX6

2 3 5 , 5 hours at 78,5 ^ 0 ,2 ° c.

Sectlorn % Log A K x2

I 1 0 , 6 0 1 ,287 10 0 ,2 5 2 13,50 1,318 20 2.25 3 19,80 1 ,338 3 0 6 , 2 5 4 1 4 , 6 5 1 ,227 40 12,20 5 8,50 1 ,089 5 0 2 0 , 3 0 6 5 . 6 0 0 ,839 60 30,20 7 3 , 3 0 0 0 62 70 4 2 , 3 0 8 2,10 0 .41 80 5 5 , 2 0 9 1,20 0 ,134 9 0 7 2 , 3 0 10 0 , 5 0 •,0 ,210 100 90,20 ®)i) li'f’iiJS'KDSÎ wm 2334 m k'QM 1% 9'%- - 3 & @■aAc.

Rgj eps. Log A % mles’ons i r ^ X 10" i cm;P 1 11.17 1,356 10 0,25 2 13.47 1,813 20 2.25 3 8.14 1,370 30 6,25 4 8,57 1,344 40 12,2 5 11,87 1,330 50 20,3 6 5.67 1 , 3 0 2 60 30.2 7 9,26 1,250 70 42,3 a 7,00 1,148 80 56o3 9 7,32 l«06l 50 72,3 10 7.55 1,232 100 90»3 11 5,73 0,911 110 110 o 2

mmusioM mh b ris 165^7 hours a t 100s0 ± 0,2°C,

S ection Log A % A

1 9,46 1 , 1 2 9 10 0,25 2 10,11 1,056 20 2,25 3 11,13 1,000 30 6 , 2 5 4 4,05 0,964 40 12,2 5 4,13 0,908 5 0 20,3 6 4,61 0 , 9 3 .0 60 30,2 7 3,92 0,878 70 42.3 8 4.05 0,849 80 5 6 , 3 S 4 ,06 0,812 9 0 72,2 xo 3 , 5 6 0 , 7 2 3 100 90,2 12 2,67 0,383 120 1 3 2 , 14 1.74 0 , 4 3 2 1 4 0 182 16 0,87 0 , 1 3 7 3 .6 0 2 4 0 riFFUdlO# KÜW 2 5a 529 flours a t 115.5 & 0.05®C.

S ectio n «B A X 1 192 2.402 6 0.09 2 260 2.476 14 1.00 3 238 2.239 24 3.51 4 9.92 1 .105 34 8.40 5 2.02 0.217 44 1 5 .2 6 0.75 -0 .2 7 54 24.0 7 0 .4 5 —0.40 64 34,8 8 0.45 —0.47 74 47.6 9 0.28 —0. 6*7 84 62.5 10 0.40 —0.50 94 79.2 11 0,46 -0 .5 0 104 98.0

DIFFUSlow HUM DK22b 405 hours a t 105.2 6 0.2®C

S ectio n Log A 3C î2 1 291 2.562 10 0.25 2 68.0 1.926 20 2.25 3 2.36 0.344 30 6.25 4 3.27 0.520 40 1 2 .2 5 0.88 -0 .1 5 50 20,3 5 0.53 —0.46 60 3 0 . 2 7 0.57 —0.40 70 4 2 .3 DFPFUSÏON HUW T,)HS3h

l'SB at 116.5 0.05 rSectlOB Log A s

1 2 3 8 2.567 4 0 .04 2 1 5 6 2.431 B 0 .36 O'? 3 155 2,263 14 1 0 4 128 1 . 9 8 6 24 3 L, 6 1 5 29.5 1.323 34 8 o40 6 8.72 0.764 44 15 o2 7 3.90 0.454 54 2 4 0 0 8 3.63 0.347 6 4 34 0 8 9 3.04 0.331 74 47 0 6 10 3.36 0.334 84 62 o5 11 3.37 0.302 94 79 «2 12 2.60 0.205 104 98 .0

DIÏi'.FOSÎOW HUH ;0R24 54? hour© a t 110.5 Æ 0.2®C

S ectio n Log A X A

:i 376 2.710 & 0 . 1 6 2 6 9 , 0 1.868 18 1.69 3 7 . 3 3 0.874 28 5.30 4 3 . 9 4 0 . 6 5 9 38 10.90 9 2.79 0,484 48 18,5 6 4.70 0 . 6 6 9 58 28.1 7 3.85 0.595 68 39.7 8 3 , 3 3 0.468 78 53.2 9 2 . 6 4 0.404 88 6 9 , 0 10 2 . 6 0 0.322 98 8 6 0 6 11 2.78 0 . 4 1 2 108 106 13 2 . 3 5 0.312 128 151

15 2,74 0.414 1 4 0 204 DIFFUSION

hours at »0 a- 0 . 05®c

S e c tio n Log A X 1 3^90 3.53*? 5 2 S»80 2.%b 15. 3 350 2 ,476 25 4 152 2.14 0 35 5 b3o5 1,831 45 o 43.3 1. 5ol 55 7 25.5 l.< 7 8 65 8 13.7 1. C2o 75 9 10.8 O.Bsl 85 10 7.23 0.732 95 13, 5.14 0,575 105 12 3.90 0,467 115

DIFFUS I O.u HUD DK26b

Sect.ion Log A %

1 4260 3.376 7,5 2 600 2.796 20 3 274 2.434 30 4 150.5 2.123 40 5 105,5 1.874 50 6 45.6 1.572 60 7 32,0 1.545 70 8 17.2 1.139 80 9 14.4 1,043 90 10 7.66 0,785 100 11 7.52 0.782 110 12 4.24 0.490 120 13 4.44 0.537 130 19 2.70 0.272 150 DIFFUSION KUW DH27a l&o5 houH's at 103»5 0»05®G

S ectio n Log A 3 5 1 3570 3.553 5 2 862 2.991 15 3 478 2.619 25 4 178 2,316 35 5 147 2.116 45 ô 9 0 .8 1.853 55 7 4 1 .8 1.670 65 8 35.9 1,470 75 9 24.8 1.344 85 10 18.4 1.151 95 11 11.8 1.010 105 12 8.24 0.892 115

DIFFUSION RUN DH276

S ectio n Lag A z 1 1550 3.534 2 .5 2 1500 3,227 7 .5 3 830 2.910 15 4 622 2.642 25 5 283.5 2.402 35 6 162,0 2.118 45 7 93.9 1.936 59 8 59.2 1.746 65 9 40.9 1.559 75 10 25.6 1.143 85 11 23.8 1.293 95 12 15.25 1.116 105 13 12.5 1.052 115 14 10.05 0.916 125 15 7 .5 0.810 135 DlPPUSIOffl KÜD DR28a 67 hour® a t 78.6 -t 0.05®i

S ectio n Log A S 1 1407 3.214 4 2 642 2.721 13 3 23.9 2.257 23 4 92.5 1.854 33 5 44» 5 1.506 43 6 3.4.5 1 . 1 5 2 53 7 9.25 0.824 6 3 8 4 o 04 0.55 73 9 2.86 0 . 4 0 83 10 2.3.6 0.20 93 11 1.77 0,11 1 0 3 12 1,45 —0 . 2 4 113

DIFFUSION RUa DR28b

S ectio n Leg A 2

1 1515 3 . 2 9 2 4 2 1000 2 . 9 5 6 1 5 3 286 2.428 2 3

4 138 2 . 0 3 3 3 3 5 67.9 1,713. 4 3 6 27.4 1 , 3 5 6 5 3 7 .17.1 1 . 1 5 4 63 6 1 1 .3 0.918 73 9 6.40 0,663 85

11 2 . 6 4 0 . 2 9 4 1 0 3

13 2.13 0 . 1 9 3 1 2 3 DIFFUSION RUN DH29 0p 42 h ours a t 84.3 - 0.05 U 0

S ectio n % Log A s 1 2635 3.255 5 2 742 2.787 1 5 3 420 2.542 25 4 247.5 2.300 35 5 152 2,081 45 6 8 1 .3 1.886 55 7 38.0 1.605 6 5 8 3 2 . 8 1.450 75 9 2 4 . 0 1.256 85 1 0 13.55 1.041 95 1 1 7 .0 8 0.87 105 1 2 4.40 0.64 115

DIFFUSION HUM DK32

1 7 2 hours a t 110.0 4 0 . 05

S ectio n Log A 3S

1 2517 3.799 5 0.0625 2 1619 3.526 1 0 0 . 5 6 2

3 815 3.109 2 0 2 . 2 5 4 104.5 2.074 3 0 6.25 5 34.2 1 . 6 2 0 40 1 2 . 2 6 1 4 .7 50 1 15.7 1 . 2 0 0 6 0 §8 : 1 8 1 1 . 0 1.052 70 42.3 9 1 4 . 8 1.183 80 5 6 . 3 1 0 7 .8 1.014 9 0 72.2 1 1 1 5 .8 1.342- 1 0 0 90.2 12-16 2 0 « 9 0 . 8 8 8 150 1 5 6 17-21 9 .2 0 . 4 2 4 2 0 0 306 raPFDSÏOW SUM DR33 182 h o u rs a t 109.8 i 0..5®C.

S ectio n Log A s 1 ) 2 ) 72 .7 2.313 10 0.25 3 15.50 1.537 20 2.25 4 2.10 0.558 30 6.25 5 0.29 —0.084 40 12.2 6 0.21 -0 .3 2 50 20.5 7 0 .3 5 —0 « 17 6 0 30.2 8 0.18 -0 .2 5 70 42.3 5 0 .5 0 0.075 80 56.3 10 0.26 -0 .1 0 90 72.2 11 0.45 0.018 100 90,1 12 0.25 -0 .3 0 13 0,18 -0 .3 2

DÏFFÜSÎOW HTO DH34 162 hours a t 100 <,5®C.

S ectio n Log A 25 1 332.3 2.996 5 0.0625 2 109.0 2.594 10 0.56 3 27.5, 1.983 15 1.56 4 1 .26 0.704 20 3.06 5 0.46 0.05 25 5.06 7 0.12 -0 .0 7 35 10.6 DïjmrsiOD ROW DR36

4* A 116 hours a t 105.5 *==' Vy 0 1®C.

Section Log A X

1 17 .7 3 1 . 4 6 2 5 0 . 0 6 2 5 2 29.88 1 . 4 9 0 15 1 . 0 0 3 0 .5 4 0.814 2 5 4 . 0 0 4 0 .1 —1.0 0 35 . 9 . 0 0

B.ÎFFÜSÏON HUM DR37 Hr- A 137 hours a t 105.5 «=• Vo1®C,

S ectio n Log A E 1 260.5 2.369 10 0.25 2 9 5 .0 2.020 20 2.25 3 4 3 .8 1 . 6 5 0 3 0 6.25 4 14 0 5 1,136 40 12,2 5 0 50 20.3

BIfmSXOM RUN DH38 240 hours a t 110,8 i o . 05®C.

<=3p S ectio n Log A 2£ 2t

1 8 .3 1 . 3 9 3 5 0.0625 3 20,9 1,398 25 4.00 5 8 .4 0 . 9 2 5 45 16.0 7 3 .5 0.408 6 5 36.0 9 0 « 6 4 . 0 DIFFUSION RUN BK42b 148 hours at 99» 5 t: 0.05®C

S ectio n Leg A % «a M 27» 3 1.403 10 0.25 3 1 6 .3 1.236 20 2.25 4 7 .4 0.040 30 6.25 5 1»9 0.279 40 12.2 6 0 50 20.3

DIFFUSION RÜR BR43a 166 h o u rs a t 99.5 ±0.05® G .

S ectio n Leg A s 1 122 1 .9 7 8 10 0.25 2 195 2.105 20 2.25 3 9 6 ,9 1.950 30 6.25 4 4 4 .4 1.587 40 12.2 5 1 9 .3 1.258 50 20.3 6 8 .4 0.682 60 30.2 7 5.54 0.570 70 42.3 8 2.80 0.406 80 56.3 10 2.10 0.318 100 9 0 .2 12 1 ,4 0.123 120 132 14 0 .6 —0 D 28 140 182 DIPFOaiOw BUD DH43b

ition Log A 2£ ■=”2% 1 54,2 1.755 10 0.25 2 35.5 1.567 20 2.25 3 17.5 1.354 30 12.2 4 1 8 .7 1.306 40 20.3 5 1 6 .2 1.174 50 30.2 6 8.72 0.883 60 42.3 7 3 .8 0 0.560 70 56.3 8 1 «66 0.133 80 72.2 9 1.67 0.125 90 90.2

DIFFUSIOft HUM DR47 120 hours a t 100oO 6 0.05®C

S ectio n Log A % 1 1 a 4^2 10 0.25 2 13c4 lo232 20 2.25 3 6o0 Go 781 30 5.25 4 2 .3 0o266 40 12.2 5 Oo6 ""Oo 34 50 20.3 DIFFUSIOa itUW DR52 112 hours at 115.5 0«05®0

S ectio n Log A E 1 1 58.0 2,031 10 0,25 2 32,2 1.991 15 1.56 3 35.9 2.037 20 3.06 4 56.4 2.091 30 6,25 5 45,2 1.048 40 12.2 6 52,0 1.936 50 20.3 7 66,1 1,904 60 30.2 a 21,3 1.622 70 42.3 10 8 ,0 0.938 §0 7 2 .2 12 1 ,2 0.135 100 110 14 0 120 132 DIFTOÎOtl RW 160o0 houra a t 115.0®C.

S ectio n 'a Log A. % 1 riiï 10 0.25 2 58o0 1,954 20 2.25 5 28oB 1.585 30 6.25 4 26o5 ■ 1,549 40 12.2 5 Bo40 1.050 50 20.3 6 4o45 0.784 60 30.2 7 le 70 0.356 70 42.3 8 lo55 0.316 80 5 6 , 4 9 2.70 0.256 100 81

DIpi^USÎON RUM DR6üb

S ectio n Lsg A s

1 cr> Cia 20 1.00

2 1 7 7 . 5 2.249 3 0 6 . 2 5 3 1 0 9 . 5 2.040 4 0 12.2 4 1 8 ,3 . 1.263 5 0 20.3 5 5 . 5 0.814 6 0 3 0 . 2 6 2.1 0 . 3 2 2 70 42.3 RUM DRôla 67.5 hours at 119.0 ± 0.09°G

S ectio n XïOg A }&

1 C13 10 0,25 2 15 Io56 3 24.9 1,494 20 3 o0 6 4 1 1 ,0 1 ,1 3 8 25 5o06 5 1 2 ,9 1.207 30 7o56 6 6,70 0 ,9 2 3 35 10 «6 1 2,35 0,468 40 14ol B 3 ,7 0.365 50 20o3 9 2 ,9 0 .2 5 8 60 50o2 10 3 ,0 0.274 70 42c 3 14 0 ,7 —0.40 110 90c2

DIFFUSlOM RUM DR6lb

S e c tio n Log A z

1 6 4 ,6 1 , 6 6 4 15 Oo56 2 1 2 ,3 1 ,5 7 8 20 3 o06 3 1 1 .4 1 ,3 8 0 2 5 5o06 4 5 .6 0 .6 9 5 35 9oO 3 1 .9 0 0 ,4 8 5 45 X60O 6 rca 25o0 7 1,90 0 . 2 4 7 65 36 oO a 1,2 0 ■■ ■ 0,077 ■ ■ •75 49o0- 9 64 a 0 10 0 95 81 oO DIFFUSlow HUM DR62 7 2 ,0 hours a t 110,0 t 0 „05®C,

S ectio n «S Log A X 1 10 0,25 2 13,60 1,435 15 1 ,5 6 3 8,80 1,246 20 3,06 4 1 4 ,7 1,167 30 6,25 5 6 ,3 0,800 40 12,2 6 1 ,4 0,146 50 20,3 7 1 ,2 0,079 60 30,2 8 0 70 4 2 ,3

DIFFUSION RUW 63a 160 hours a t 1 1 5 ,0®C,

S ectio n «S Log A s ■ 1 an 20 1 ,0 0 2 325 2,813 2 5 5,06 3 247-5 2,657 3 0 7,56 4 249,5 2,699 35 10,6 5 287,5 2,459 45 1 6 ,0 6 116,3 2 0 O6 4 55 2 5 , 0 7 20,0 1 , 3 0 1 6 5 36,0 8 5,1 0,700 75 4 9 , 0 9 2,1 0 ,3 2 95 6 4 , 0 DIFFUSIOM RUÜi DR63b

S ectio n Log A X 1 1 15 0 ,5 6 2 357 2,775 20 3,06 3 258 2,654 25 5,06 4 93,4 2,192 30 7,56 5 9 3 ,2 2,192 35 1 0 ,6 6 7 6 ,3 2,104 40 14,1 7 7 5 ,4 1,798 50 20,3 8 13,9 1,064 60 30,2 9 5 ,0 0,620 70 4 2 ,3 10 0 ,5 —0,4 80 56,3

DIFFUSION HUM DH66a 42 ,0 hours a t 101,0 t 0 ,

S ectio n Log A % 1 (.100) (2 ,5 8 ) 5 2 15 3 34,2 1,534 25 4 28,4 1,454 35 5 23.1 1,364 45 6 19,6 0,992 60 7 9 ,7 0,987 75 8 9 ,2 0,624 85 9 1 ,0 0 ,0 0 95 10' 3 ,3 0,520 105 11 0 „ ' 115 VlTTOHKTnM HUM DWfiTa DIFFUSION HUN DH67a 256 hour© a t 1 0 0 ,9®( 3,

S ectio n A s 1 10 0 ,2 5 2 15 1 ,5 6 3 130,0 2,453 20 3,06 4 127,7 2,425 25 5,06 5 71,0 2,170 30 7 ,5 6 6 60,4 2,100 35 10,6 7 51,5 2,031 40 14,1 8 25.9 1,732 45 18,1 9 36,0 1,575 55 25,0 10 12,4 1,111 63 56,0 11 6,1© 0,804 73 4 9 .0 12 5,20 0,524 85 6 4 , 0 13 4 ,0 0 0,320 110 100,0 17 0 165 272

DIFTOSIOÎs! RUN DR67b

Seetâosî Ii0g A %

1 Cxa 10 0 , 2 5 2 15 1 ,5 6 3 1 3 2 , 3 2 , 3 5 2 20 3 , 0 6

4 122,2 2,317 2 5 5,06 5 161,0 2,435 3 0 7 ,5 6 6 145,9 2,393 35 10,6 7 50.0 1,929 40 1 4 ,1 8 30,0 1,706 45 18,1 9 30,5 1,416 55 25,0

10 9 , 7 2 0 , 9 2 0 6 5 3 6 , 0 11 3,05 0,420 75 4 9 , 0 12 0,80 —0,16 85 6 4 ,0 DIFFUSION HUM DR68 3.69 oO houra a t 90,0®G,

S ectio n Log A 3S 1 5 0,0625 2 4 ,5 1,216 10 0 ,5 6 3 3 ,8 1,133 15 1 ,5 6 4 1 ,■? 0,784 20 3,06 5 3. ,5 0,730 25 5,06 6 0 SO 7 ,5 6

DIFFUSION RUN DRôgb .0 hour© 8 t 98,5 t 0,2®C,

S ectio n Lag A % 1 5 0,0625 g 10 Go 56 3 55,4 2,044 15 io56 4 37,5 1,876 20 3o06 5 13,17 1,421 25 5o06 6 3,44 0,838 30 7o56 0,50 0 ,0 0 35 10o6 0 40 14 ol MFFUSÎON RUN DR71

6 8 oO h o u r s a t 78»5®C

S e e tia n Lag A s 1 5 ,0 2 343 0 6 2,494 1 5 , 0 3 219,6 2,300 25,0 4 77,0 1,846 35,0 5 38,4 1,543 4 5 , 0 6 50.7 1 , 4 4 6 5 5 , 0 7 19,0 1,238 6 5 , 0 8 8,75 0 , 9 0 0 7 5 ,0 9 8,25 0,876 85,0 3.0 7,60 0 , 8 4 0 9 5 ,0 13. 7,55 0,536 110 12 6,00 0,417 130 13 1 4 5 14 155 15 2,05 - 0 , 3 0 170 DIFFUSION ROT DR72a

2 0 4 h@ur® a t 3.10,5

S ectio n «B Lag A %

1 10 0 , 2 5 2 61,8 2,128 15 1,56 3 57,5 2,097 20 3,06 4 5 5 , 9 2 , 0 7 5 25 5,06 5 5 9 , 6 2,113 30 7,56 5) 7} 99,4 2 , 0 3 3 40 12,2 8 5 4 , 9 2,077 45 10,1 9 57,4 2 , 0 9 7 50 22,6 10 48,4 2,022 55 2 7 , 6 11) 12) 158.1 1,961 75 42,3

7 9 , 4 1,936 85 6 4 , 0 14 5 9 , 7 1,813 95 81,0 17 4 7 ,8 1,716 1 2 5 1 4 4 20 16,3 1,248 ' 155 225 23 7 , 7 0 0 , 9 2 3 185 289 26 2,40 0,417 2 1 5 441 2 9 1 ,2 0 0,116 245 576 35 0 DIFFUSION Küi BR?2b

S e ctie n L@g A s ' 1 143 2.489 15 0,56 2 94,5 1.830 25 4 ,0 0 3 115.9 1,890 35 9 .0 0 4 92 0 6 1,797 45 16,0 5 84.1 1 .7 5 8 55 25,0 6 99,7 1,829 65 36,0 Î 92,3 1,795 75 49,0 8 69.9 1,675 85 64,0 9 9 3 .9 1.603 95 81,0 3.0 86,9 1,769 105 100 13, 81,7 1.742 115 121 12 61,0 1,614 125 144 14 18,9 1,106 145 196 15 1 6 ,8 1,055 155 225 18 10,2 0,838 185 324 20 4 ,1 0 0,443 205 400 22 2,30 0,190 223 505 24 0 .3 -0 ,7 0 245 600

DIFFUSION RUN DR75a I h o u rs a t ll0^3^Qo

S ectio n Leg A % K 1 .15 0.56 2 37,8 1,934 20 3,06 3 53.9 1.786 30 6,25 4 68,3 1.910 40 3.2,2 5 29,8 1,83.1 45 1 8 ,1 6 30.0 1,034 50 22.6 7 64,0 1,883 60 30,2 8 fix» Ç O iîrt 1,860 80 49.0 9 45,2 1,7.11 90 72,2 10 49,9 1,754 100 90,1 3.1 20,0 1,356 130 156 12 8 ,3 0 0,975 160 240 13 4,40 0,700 190 342 14 1 ,8 0 0,310 220 15 1 1 ,0 0,055 250 600 BimviBiQË Rim

S ectio n B iîûA A.

2 80.4 8.25 83.4 6.25 4 44.2 10,6 5 ■^o9 40 14,1 S 5,2 50 20 o 1 65.1 o &O è 53.3 ?0 4 2 ,3 9 65.4 56.2 10 53.1 72.2 II 39,0 1,677 90,1 18 44.2 1,733 p 13 o 140 ]U33 14 12.2 '‘M2 170 15 Î lA. ïl 6 t* i «V Ki 16 Cl9 o y r ^ i S

DIFFUSION Hlia BH76a 0, 3k-., a t ^,5

le tio a Rg A 1 ‘ 1 0.25 g 54, 3 2ol06 15 1 ,5 6 3 36. 2 lo934 4 53, l 2o095 5,01 5 2 5 , Û le 7 5 9 30 o ... 6 24, 8 lo766 35 10,6 7 87, 8 I 0 8 I 3 40 14 ol 8 24, 3 lo762 45 10,1 9 22, 3 le 420 55 25.0 10 8, 10 0o979 65 5 6 ,0 0 11 0 37 0 0 4 4 6 75 49.0 ÏS Qo64 Oa ÔO 85 6 4 . 0 15 *•■^0 0 6 81.0 DIFFUSION RUN m 76b

Ssetiosa Log à s 1 “^ 1 10 0 ,2 5 2 87 00 2 , 3 3 1 15 1 ,5 5 3 47,9 2,068 20 3,06 4 42o5 2,016 5 ,0 6 5 45,1 2,042 50 7 ,5 6 6 37,7 1 , 9 6 4 35 1 0 ,6 7 30 ol 1,866 40 1 4 ,1 0 23 0 6 1,760 45 1 8 ,1 9 24,0 1,472 55 ■ 2 5 , 0 10 1 3 ,0 1,200 65 3 6 ,0 1 .1 . 8,85 1 , 0 3 3 75 49,0 12 l o t s 0 , 3 3 64,0 .13 95 8 1 ,0 14 0 ,6 0 “©0 I 4 105 100

DIFFUSION RUN DR77Î)

2 3 6 ho u rs a t u o ,5 ®e.

S@etloa Log A s

1 1 5 0 , 5 6 2 2,000 20 3 , 0 6 3 372:7 3,027 25 5,06 4 36,2 2 , 0 1 4 30 7 ,5 6 S 4,10 1 , 0 6 9 35 1 0 ,6 6 3 ,7 0 1,024 40 1 4 ,1 7 1 ,2 6 0,557 18,1 8 0 , 6 3 0,256 50 22,6 9 1 ,8 8 0,430 60 3 0 , 2 10 1,00 0,155 70 4 2 , 3 11 0,90 0,1.1 80 56,2 12 90 7 2 , 2 13 0,80 0,058 100 90,1 18 O0 6 5 —0 , 3 150 210 DIFFUSION HON DR77©

S ectio n Log È % 1 (10,000) (4 ,0 ) 10 0,25 2 814.7 3,464 15 1,56 3 129,7 2.666 20 3,06 4 15.3 1,736 25 5,06 5 1.15 0.614 30 7,56 6 1 .6 3 0.766 35 10,6 7 2.45 0,942 40 14,1 8 0,50 0.254 45 18,1 9 0.35 0,097 50 22,6

DIFFUSION RUN DR?8a 496 hours a t 103o5®C

S ectio n Log A s

1 4000 3.616 10 0 . 2 5 2 31,2 2,963 15 1.56 3 5.94 1 0 2 4 2 20 3,06 4 2.50 0,857 2 5 5,06 5 1.32 0 , 5 9 0 3 0 7 ,5 6 6 1,34 0,596 35 10,6 7 1,07 0.498 40 14,1 8 0,71 0,320 45 18,1 9 0,79 0,356 50 22,6 10 0,67 0 , 2 9 5 55 27,6 11 0,49 0,158 60 33,0 &.T.pî?£3sioai mm m is h

2 qcüok S B ?. & Log k z I 11,540 4,223 10 0 2 790 3,380 15 1 3 172,5 2,719 20 3o06 4 49,9 2,180 25 5 a06 5 28,9 1,943 30 7 o56 6 1 8 ,8 1,756 35 10 o6 Y 1 4 ,4 1,540 40 14 ol 8 12,44 1,577 45 18 ol 9 3,70 1,050 50 22 o6 10 1 4,00 1,083 55 27 d 6 11 2,32 65 36 oO 12 2,44 ■ 0,568 75 49 oQ 13 1 ,4 5 0,342 85 64 oO 14 3,10 0,973 95 81 nO 15 1 ,1 2 0,230 135 ;

° DIFFUSION RÜM :0R79 202 hours a t 115. 0 ±. 0,05®C„

S e ctio n Lag A ■ z ï'2 1 (5000) (4 ,0 ) 15 0 o56 2 141,5 2,754 20 3o06 3 59,0 2,374 25 5 oOS 4- 1 6 ,8 1,828 30 7 5 2,53 1,00 35 10 o6 6 0 ,5 3 0,327 40 X4ol 7 0,36 0,158 45 18 ol 8 0 ,2 5 0,00 50 22 o6 9 0 .2 - 0 ,1 55 27 o6 10 0 ol —0,4 60 33 aO MPPiJSIOM HUM DH60e

1 2 5 0 ho u rs 5at 1 1 5 , 0 i o , 0 5 ®ct t>

S ection Log A X

.1 1 5 0,56 2 50,7 3,207 20 3,06 3 4 7 ,2 2,176 2 5 5,06 4 2 8 ,7 1 , 9 6 0 3 0 7 , 5 5 5 35,0 2,046 35 10,6 6 2 7 .3 1.93B 4 0 14,1 7 1 4 ,4 1,660 45 18,1 8 1 3 , 0 1,616 5 0 22,6 9 3,40 1,034 27,6 10 6 ,0 0 1,280 60 33.0 11 2 .4 8 ' 0 , 0 9 6 70 42 ,3 12 0,45 0,154 00 56,3 14 0 ,2 - 0 , 5 100 9 0 , 2

BIPPUSIOW RUM DRaOb

S ectio n Log A % 1 0,25 2 .5 2,095 2,25 c5 2,090 5,06 78,8 2,197 30 7,56 7 9 ,3 35 10,6 6 9 ,0 l l 45 ( o g 1,427 55 tic, O 7,10 0,852 65 ^woO 2,30 0,362 49 oQ 1 ,9 0,280 64 oO 11 •0,15 MFFUSÏOM HUî) DH8 O0

Sectioxî %

1 1 0 0.25 2 275o8 2.441 2 0 2.25 3 § 6 0 8 2.287 2 5 5 , 0 6 4 1 5 0 .8 2,480 3 0 7,56 5 83.1 2.221 35 10.6 6 1 2 7 . 3 2.105 45 1 6 .0 1 8 5 .Q 1 , 9 3 4 55 2 5 . 0 8 41.2 1.616 65 36.0 9 10.55 1.024 75 4 9 . 0 10 6.60 0.820 85 64.0 11 1 .7 0 0 . 2 3 0 95 81.0 18 0 .3 “O. 5 105 100.0

DÏPFÜSÎO&I RÜW DRSOâ

S eetio îi Lag A % 1 10 0.25 2 74,6 2.174 1 5 1.56 3 7 2 .8 2.162 20 3 . 0 6 4 4 0 .8 1 . 9 1 2 25 5.06 5 26.1 1.718 3 0 7.56 6 2 9 . 2 1 . 4 6 6 4 0 ■ 12 ,2 7 21,0 1 . 3 2 2 50 20.3 0 5.30 0.724 60 30,2 9 2 .0 0 0 . 3 0 1 70 42.3 10 1 .1 0 0 . 0 4 0 80 56.3 11 1 .2 0 0.080 9 0 72.2 12 0 ,7 “ 0.25 100 9 0 . 1 13 0 , 3 “ 0 ,5 2 120 1 3 2 DIPFUSIOm saw BKS2a 250 houses a t 115.0 & 0 .0 5 'ÿpV a

Sect!OH :',s Xi0g A a 1 20 1 .00 2 s a , 3 25 5 .06 3 50.4 2.527 30 7 .56 4 65.9 2.643 35 10 .6 5 72.1 2.682 40 14 oX 6 66.1 2.644 45 18 .1 1 7 0 .3 2.671 50 22 .6 8 9 2 .S 2.489 So 30 .2 9 2 9 .8 1 .5 9 8 TO 42 .3 10 41.6 2.142 80 56 .3 11 31.3 2.020 90 72 .2 12 1 5 .2 1.705 100 90 .1 13 9.10 1.482 no 110 14 4.30 1.156 120 132 15 2.10 0.846 130 BÎFFB'aîOM EOH BR82I)

S e e tio a X S'^ 1 10 0.25 2 131 2.538 15 1 . 5 6 ' 3 139 2.563 20 3.06 4 178 2 0 6 7 2 2 5 5.06 5 427 2.750 35 9 . 0 0 6 407 2 . 7 2 9 45 1 6 .0 ■ 7 260 55 2 5 . 0 8 , 261.7 2.537 6 5 36.0 5 151 2 . 2SB 75 49.0 10 9 9 .7 2,114 85 6 4 . 0 11 9 7 .2 2 . 1 0 7 95 81.0 12 50.7 1 .825 1 0 5 100 13 43,1 1.755 115 121 14 1 6 .1 1.326 125 144 15 1 5 .0 1.295 1 3 5 169 16 1 1 .0 1.160 145 1 9 6 17 5 .3 0 0 .8 4 A 155 225 18 5.60 0.868 165 256 20 3.20 0.624 185 3 2 4 23 0 .6 “ 0 .1 215 4 4 1 BIFFUSSOH HBW DR84a 222 hsura a t i;i5oO®C

S o etlo a I,eg A 21 I 388,5 2.290 10 0.25 2 06 oO 1.955 15 1 .5 6 3 107,0 2.029 20 3.06 4 7 1 ,8 1,857 25 5.06 5 33,2 1.522 30 7.56 6 3 3 ,3 1.523 35 10.6 T 2 9 ,8 1.475 40 14,1 a 1 8 .0 1.274 45 18.1 9 3.1.4 1.057 50 22 0 6 10 2.50 0.3 9 8 55 27,6 11 3.10 0 .1 9 65 36,0 12 3 .1 ,0.19 75 49 oO 13 0 .8 “0 .4 RR 64oO 14 0 95 81,0

DIFFUSION RUN

S ectio n «a s S'S 1 413.3 2.616 5 0.0625 2 117.3 2.070 10 0.56 3 8 3 .5 1.922 15 1.56 .4 131.9 2.012 20 3.06 ; 57.4 1.759 25 5.06 è 32.0 1.716 30 7.56 7 60.0 1.4 7 8 40 1 2 .2 8 23.0 1.061 50 20.3 9 4 .2 0 0.322 30.2 10 1 .2 0 0 .0 8 70 42.3 MFFÎIfSïOà! EU.N M 84e

8setâ©a % ;cic?g A E S'* 1 147 gol68 10 0.E5 S 90. S SO g.29 3 ?0oS X Q S5Û 30 6.25 4 40.4 c.) 0 W 40 la . 2 5 1 5 .8 1.19S 50 20.3 6 7.70 0.887 60 30.2 1 2.30 0.362 70 42.5 8 0,5 “0.30 80 56.5

DÏ5POSIOM WM mm&. 255 hem: ira at îfJ3.c re .

Saetaosî % Leg A K 2 8 2 .6 2.157 10 0.25 2 .1.45.1 2.155 15 1.56 3 88.9 1.950 ■20 5.06 4 69.6 ,1.843 25 5 0 06 5 55.1 1.742 30 7.56 6 1 5 .4 1.188 35 10.6 6.5 0.815 49 14.1 a ■9.716 45 10.1 3.0Q 0.477 50 22.6 i .40 “0 0,154 611 30.2 0 .8 '“•'0 0 4 70 4 8 .3 DIFFUSION RUN

SGetiQSl A '^8 s

3. 5 1 1 2,204 1 5 0,56 2 68.8 1,654 20 5,06 3 46o2 3 . 0 6 6 5 25 5,06 4 32,2 1,508 3 0 7,56 5 38,6 1,589 10,6 6 14,1 1,149 40 14,1 f 6,90 45 18,1 8 3,30 0 , 5 1 9 5 0 22,6 9 1,20 55 27,6 3.0 0,3 60 33,0

d iffu sio n nuk meg®

L®S A 5S 1 2,018 20 1,00 2 78,5 1,861 2 5 5,06 3 42,3 1,627 3 0 7 , 5 6 4 31,2 1,494 35 10,6 5 3,7,0 1 , 2 3 1 40 14,1 6 20,3 1 , 3 0 a 45 18,3. 1 6 0 6 0 0,820 50 22,6 8 5,70 0,756 55 27,6 9 SO 33,0 10 1,00 0,00 65 3 9 , 1 11 0,4 “ 0,22 70 4 5 , 6 :oj.ïFiysïow :«ua Dites SOcO h©US’S a t 96,1 t 0,05®C

SestloH :a L©S A 1 1 61? 5 2 112 2,236 15 3 121,7 2,085 25 A 88.1 3., 9 4 6 33 5 49,5 1,695 45 6 34,2 55 1 20,7 1,316 65 8 27,4 1,136 80 9 18 ,0 0,955 100 10 1 6 ,8 0,925 120 11 6 ,6 0,519 .140 12 9 ,1 0,658 ISO 13 6 ,8 0,532 ISO SCÏÎAHGE RON BM2 ïsffipQî/'atuî.'e 2 2 ,0 ± 3.o0®B(. Tempemtere 53,0 't. l„ 0* wf-i o Wto ‘i’S'itiatea Bsasoie Wto I J i o & o So50 SgBo AaWi 6i>40 mga. Ooimt B ate SamBle Bate «13.88 1 3olê 0,2 1 & O 3oQ 2 4.00 .1,0 2 2 .8 g 3 .3 3 4,80 1,3 3 3,61 4 o i 4 5.58 1,9 4 4f O 6,;i« 3 6.25 1,8 Sc'’'* lo3 $ 6,70 2,9 1 Sol 10,9 7 7,42 . ;lol '? 8 oO 1 8 ,9 8 133 3,4 8 8./IÎ 3Jo6 9 0,65 4,3 9 •eg.. IB , 4 10 9,22 5,56 2.0 3.1 o 40 11 9,74 6,6 11 12,12 25.2 .12 «,!>'Î A vij O&tAF: ^ 6.35 12 18 , 23, g 13 10,72 7,7 14 11,19 7,65 15 12,64 9,3 16 13,22 3.0,6 17 14,32 18 16,50 3,0,9 19 17,74 15.6 ao 18,4® 14,9 ZL 4V.» ^ O g ef 16,3. as 20.65 18,0 as 24,3 19.5 24 39,0 54.6 IXCHâMOK HUN BR3 îaCHAHGE RWN BM ïem^5©ratiaK’0 6 6 ,0 & 0 ,5 G« 38,0 & 0,5®G, ï3to SoBoAo 5,00 raga. fto fcBoA, 4 ,9 0 àMBo

Gmmt Bat.0 Oanfât Safe® fûiiîB Iî„, 608/ml, mams H e§m/&lç

1 1,41 0,8 1 2335 1 3 2. 8,24 2,0 2 3,16 2 ,0 3 3,16 5,3 3 2 3 4 3,87 7,1 4 4,46 3 3 5 4,47 àoè 5 5,10 3 3 ê 5,00 10.3 6 5,91 4.2 1 5,47 11,5 7 6,70 5.2 S 6,32 3.4,05 8 8.06 5.9 9 7,06 19.1 9 8.09 8.2 3.0 7,75 20,6 :w 10,00 9 ,7 13. . 8,36 26,4 11 10,96 1 1 .4 12 8,94 29,5 12 11,61 1 1 ,a 13 9 ,4 8 S3 3 13 12,45 3.4 o 4 u 3.4,55 ;i5o6 15 15,17 1 4 3 3.6 17,60 î% S 11 21,10 Zâol lS0HàS5G£ a w acIÎAM G l SÜW EH6 0. o V V o ' aw0S*afcUTC 7 1 ,0 10o05® e , Wt, G*‘‘'^BoA, 1 o35 Kgcio é Geimt Heto Si® s e a s u ffa M © aisi® 4 ,4 2 8 ,9 17.0 4 4 0 4à 16.0 5oM 24.0 ^ lé i 26.0 o,<%" 2 8 ,2 p: 33,0 7,41 33.7 11 idO ol 2 3o 45.8 EXGHAME- BUB ER7 EiGHAmE nm ER8 T@mp©ratur© 71.o o i‘o„5®c. Temperature 76o5- Wto o A o 1.*90 mgMo Wto ToB o A o 5o70 mgmo Sample Count Hat© Sample Count Rate mia© R_ opa/m l 0 mlmm Rgg opa/mlo 1 .5 10 oO 1 lo41 lo 7 2 10 SoO 2 2o24 4 o6 3 15 7 .5 3 3ol6 17.0 4 20 7oO 4 3o87 24.4 5 25 7o2 5 4.46 c=> 6 30 6oO 6 5oOO 43o2 7 40 6o4 7 5 o46 42 o 4 5o91 49.7 9 6o32 45.7 54.7 n 63,4 12 % 7,41 59.7 7,75 71.9 8,06 70,0 11 aCHAWGE RUN SR9 EXCHANGE RUN BRIO •4' Temperature 43.0--Oo5 G Tempe ra tu re 60 oü - 0 o 5 G o Ho To E o A o 5,05 mgm. Wto T,B,A, 4,90 mgm.

Sample C am t Rate Sample é 0® m t Hat© Blnm R ep e/m l , mime. R op®/siI,

1 2,24 2 .7 1 2,24 0 ,6 2 3clS 3 ,5 2 3,16 1 ,0 3 3.87 5 ,5 3 3,87 4 ,3 4- 4,47 4 .6 4 5,00 1 2 ,6 S 5 ,0 0 5 ,9 5 5,47 1 5 ,9 6 5 ,4 8 6 ,7 6 5,91 13,1 7 . 6,32 7 ,6 7 6,32 1 5 ,5 8 7 ,4 0 9 ,8 8 6,70 2 0 .3 9- 8,35 1 1 ,5 9 7 .4 1 1 6 ,2 10 9 ,8 2 7 .7 10 8 ,0 6 16 ,9 11 10,0 11,1 11 8 ,9 4 23,9 12 10,71 1 4 ,4 12 9.21 22.9 13 n ,40 1 4 ,8 13 9 ,4 8 27,0 14 12.07 23,9 14 10,00 31,9 15 13,22 25.7 15 10.50 31,9 16 13,80 25.7 CHAP'J ':x IV DISCUSSION ON THN HNSULT3 118 a* Proton TremCer In Bonded SoliüSo io Previous Worko E¥id©neo for proton transport In oo«-op@ralively hydrogen bonded orgenlo solida ha© been obtained from a® Diffusion ©tudieso bo Conductivity studieSo

' 8 0 Diffusion ©tudieso There have^ until thia workp been no direct diffusion measûrementa in hydrogen bonded organic cryatala The neFrest approach has been exchange studies in hydrogen bonded inorganic crystalso PÊ In 155S Wei and Bernstein observed proton diffusion occurring in boehmltOp alumina monohydrate o They studied the exchange of DpO 18 vapour T?ith polyeryatallime boehmit© from &1^130^Qo and found that after an initial surface exchange a diffusion controlled reaction took place between the deuteri mi mià the crystal which had the following Arrhenius dependenceo

r ..“9 r™M4âfîâJ,jao'^ \ u x " 5 E 10 ^ ««P L RT where is the highest posaible diffusion coefficient as the 1 0 surface area of the crystals was not accurately knowno 0 *" d iffu s io n was found to ;fc> 6 " very much slower than deuterium indicating that a fast proton mechanism was operaiingo They also found that proton mobility in beyeritOj^ ^ ^Al(OH)^£> was an order of magnitude greater than thiSo A s im ila r stu d y was mao© by Feitkn© cht ©t tr itiu m diffusion from watei vapour into axidFeOOtl and th#ir 119

te ra te d aiiia3.,ogu©e and the smm diffusion controlled

Qxchang© k in e tic s were o b t a i n e o ©a In b o e h m i t e mû s a t i s f i e d th© fo llo w in g ArrhemluB equ@tionso

M(0H)2 Do 2 % 1 0 “ ^ ©Ep j jji|

J Several interesting points arise from these studieSo Proton diffusion la coBsiderably faster than metal loma at the same low température© All these eystemm have in their structure a

c@*^qp©rative system of hydrogen bondB© All give very sëîëiII pre-©KpoMntial factorsp indicating a negative entropy of

actl-vatlom which Bernstein suggests may be a function o f a p ro to n ic mechaniam© The a c tiv a tio n e n e rg ie s o f th ese p ro c e sse s are all of the order of the activation energies found in

hydrogen bonded solids asshibiting proton conductivity 1 ^ ) The only eo-^-operatiTOly hydrogen bonded system in which a direct diffusion study has been attempted is ice© The first measurement xmm made in I95B by Kuhn and Thurkauf who observed the diffusion rate of deuterium and oxygen-’lB in polycrystaXline ice ana found %he a erne diffusion coefficient for each loO^O.S x *^’10 2 1.0*"’ ' cm©aecT indicating that at thia temperature diffusion waa du© to migration of an entire water molecule© This measurement was made a t one tem perature only and no A rrhenius eq u atio n was obtaineo for the processo It was thoughtposaible^ however^ that thia bulk diffusion could be due to a premeltiag phenomena am the mmaemrement was made so near the melting pointo This would mask a slower proton oonduetion aechemismo During the cours© of this Investigation two other studle® were made on ice by 3)engel and Riehl 'and Itsgaki in which tritium diffusion was measured In mingle crystal^of ice aa a fimction of temperature and the following Arrhenius equations o b tain ed r=* D ® 2 exp RT Dengel and Rlehl

D 2 0 8 X 1 0 ^ ïtEgalei 1Q Hiehl has also measured the diffusion of@" in ice anc tritium % diffusion in ice crystals doped with HF and found that in all oases the activation energy was the aam©o These results appear to con#.rB^ that 'm'"the cas© of ice the proton diffusion associated with the aieasured proton conductivity is masked by a rapid diffusion of neutral water molecules in the lattice© The mechanism of this bulk diffusion process has been examined theoretically and vacancy end # In terstitial mechanisms ® proposed© The most likely mechanism Êp*?? is that proposed by Oaaager end uum Ib of a ®free^ inter­ stitia l machaniam in which a water molecule travels through several Interstitial positions before occupying a vacant lattice s i t 0 o Granicher has euggeatedj, howevero that as the activation ©nergy for tritium diffusion agrees with that for D©G© conductivityjj hydrogen diffusion may proceed by the same mechanism in both cases© 1H

Conductivity Studies© Comductlvity studies Iiotq bem made in a large number of o rg an ic mii inorganic hydrogen bmûBû aolida and mechaniama based cm proton migration along eo^operative chain© ©f hydrogen mi ^af%302&)SScg%&4 X 14^ P o llo ck mû ÜbbolûMe Made am imtereetimg ©tudy on a m icim of organic acidSg acetylemodicarboxylic acid mû its dihydrateg oxalic acid and it® dlhydrate^ be&molo acid and furoie acidg and found that the conductivity imereaaed and activation ©mergy decreased in that order with the hydrates exhibiting greater conductivity than the parent acid© They concluded that the conductivity was pro tonic amdg aa the order above waa one of decraaaimg co^=>@p©rati@mf, that the conductivity depended on the degree of co-^operatiTO hydrogem bomdimg preaem to They did mot 9 howeverg verify the mature of the oherge carriers and their ©tudie© were made em polycrystallime compact® well below the melting pointo The immt exhaustive series of conductivity atuclie© Im [jff hydrogen bonded solids hm beam made by Kiey and hi© c@-work©re/'* cm synthetic and naturally occurring solid© comtaimimg the 0^" hydregem bomdo Im all case© a defimite conductivity was found im the dry state which in general was found to be electronic im mature with energy gaps detmrmimod from the temperature variation ®f eomductamee in the region §'=.#? which ia of the order of band gap calculated for repeat units ©f the 12 type ffi— 0^0**^^ and Eley suggested that these are intrinsie semiconâuatùTs due to electron mobility in the GO^^WH ayetern© In the ease of polyamide^^“ it was obaerved that at loo temperatures the conductivity was loo and electronic with a band gap o'f 2c5@Vo At high tersiperature^ hooever^ the conduct­ ivity became protonle with coMuctWitiea 10"^ time© greater than the electronic conductors anû a much smaller activation energy of conduct long loi«^I©3©¥ was obtained© They concluded that thii was due to the rotation ©f the amide group and that aelf'=' ionisation was the rat© determining step of the conduction mechanisM.■ o '4 Bley has also found that the conductivity of these systems can be greatly influenced by adsorbed water'^and h© has shown that in the case of haemoglobin 7^ adsorbed water increase© the electronic conductivityo It has also been shown that proton ÛW conductivity can be obtained on adsorption of water and it i been suggested that the water acts as a plasticiser which @ facilitates the reorientation of the hydrogen bonded chain* The model substance of a co-operatively hydrogen bonded lattice is Ice and a proton conduction mechanism has been proved beyond dw bt in which th© rat© determining step appear© to be the ionisation of the water molecule<, For a static conductivity to b© observed in a co-operatively hydrogen bonded ©olid it is necessary for each hydrogen bonded chain to re­ orient to its original state before another charge can pass 123 along ito In the cas© of ice neutral orientational defects known bb Bjerrmm defects in which one hydrogen bond is doubly occupied mû another unoooupied have been proposed t o e x p la in this reorientation which mnmt occur in all hydrogen bonded solids which are pa^oton conductors® The conductivity of moth©'r hydrogen bonded cryatalg MBonira dihydrogen phosphate has recently been reported by Murphy in which proton conductivity has been shown to ©xiat and i t vim proposed that the mechanism vm analogous to conduetl in iC8o il Conductivity Resultso

The c o n d u c tiv ity o f benzoic acid wa© found to b© very lowo The conductivity ©f anelar benzole acid compact© was found t© obey the following Arrhenius ©quation in the temperature rang©

4 0 “ 1 0 0 ® C . 0"' a 1 % 3 7 , 5 0 0 RT In high purity compacts and ©ingle crystals both perpendicular and parallel t o the (001) plane the conductivity was not measurable and must b© less than 2 % ohi^ ci^ within 70^ of the melting pointo Thi© agrees with the value recently found 1 El taoTI 1 by Bley € t e l o f 1 x 10” ^ ohm cm two d eg rees below the melting pointo

Th© conductivity of acetic acid perpendicular t o the (lOC,) plane was found to vary with the amount of a d s o r b e d w ater a n d f e l l t o an unmeasurable valu© on drying® The bulk conductivityg 12

ft therefore g was concluded to fo© than 10 '1 0 ohm" cm o'" a t o®e, ■ flw coM uctivity &f ©xalic acid dihyârat© bath mb compacts ami ©ingle crystals gave linear Arrhenius plote having the earn© activation energy 21-24 K oals/mol© but which varied in absolut© 3 cemductivlty by m order of 10® A comparisoB of coMiactivity atudiea in organic acid© is shown im ta b le o^VI

GOBJDUCTIFIW IM ORGAMIG ACIDS® rYP33 BIAMPDB SliteUK - —1 —1 oh®o eso 3YCLÎC mmm bkwzoig ACID eam gaets 1 .8 3S 10“°" UBBBLOHDl Analas’ ceiagasts < 2 % 10”-^^ pup® eompaete < 2 a; THIS WOIEC JECOOI) < 2 s 10“^^ THIS ITOHK “ ” ;U.(0Q1) < 2 % THIS WORI sitagl© eP 3?®'c@J. ‘ 1 2S 10%^^ I^J ^D'ïÇ* (at 120 C.) MROÎC ACID e0®pa@t 1.8 z 10°* PIfALIG ACID K s jn “13 KOIIDO AND ODA y. “ BoaSQi OXALIC ACID h aias ■ ,2H«0

eoinpaets 23,5 2 .2 K 10“ ® 21=»24 3 .5 3Î THIS WORK gJtogle epystal 22.0 1.5 s 10“^^ THIS WOHK ssalie Aeiâ “ ©empaet 40.5 1 .1 X irn-S ubbelohdk ' 125

TYPE EXAMPLE E. 0” (50®C) REFEKENCE A Kcal/ffîole ohmT cmo ACETYLENE DÎCARBOXYLÎC ACID <7 12 o 5 7ol X 10^ ÜBBSLOHDE ACETIC ACID Single Crystal 1 (100) - < 10*"^^ THIS WORK

Several conclusions can b© drawn from the above table® lo The conductivity in crystals having cyclic dimers la very much lower than in crystal© which have co-operative chains of hydrogen bonds® 2o The conduetivity varies considerably with the degree of compaction and much lower conductivities are obtained in single crystal©® 3o The effect of water on the crystals is to inereaee the conductivity® 4o The conductivities are so low in single crystal© that the proton diffusion coefficient as calculated from the Nernst ™ Einstein equation would not be measurable by a tracer diffusion study using a sectioning technique® The low conductivity obtainec in this work for analar benzole acid does not compare favourably with Ubbelohde**© values® It is possible that his result is due to transient effects which can occur at the low temperature ( s e e e.go Fig^v > run 1) at which ha was w orking^20-50^0®,and to the f a c t th a t he was working near the insulation lim it of hi© cell® tufnol® 10^^ ohms cm®

V.» ,jfÆ. I?l ft.F. o ' ’tût, 4'i %9 V»'' @%AL& AUB) j / " \ . W N V ^ ^ m ^ 9 t ’"V .& VN r" %V .. / e ' " e * " %

!

' . o A ’f4i ©3HVOP^©e#.M „ iuf . 'V\- & m"" 'V> ^ PMeSPNATm Ka'^ "ip / <

. x - V \ ^ '

, A‘'* 3, 44». S^ÊHIWATOÊ VmW ©F Ê.Ô» ©Pi^ATsVi I The conductivity found in this study is believed due to tl preaenc© of ©.teorbed mo la tu re in the anaiar aold as no attempt xmm made to dry it® It may be significant that the activation energy o f th ia procea© 7o6Kcal©/m@l© la approximate3„y th e ©©me m the activation energy of mobility of water diffusing in benzoic aold single crystals (see p #0 )« The conductivity of oxalic acid dlhydrat© i© interesting in that the same activation energy i© always obtained® This result oan b@ compared with conduetivity studies in other hydrogen bonded system© which proton conduction ha© 0, GEIBtàh % Rl? 4 'A % ®A ohm em ohm^oBV Kesl/saol© Keal/®®! OXÆÏC AG ID 2HgO 6.35:10' 1.52:1012 ûM DOmMIDES 4 E AMMONXWÎ DÏHÏDH06W >10 BHOSPHAÎS 6 20.4 1 3Î lo ' O' MURPHI “ XG,1 o % 11.0 22 BIGBN ^ 5ÎÎ m 2ol2Sl0‘ 1 s 10' ■Î GRASiïGHIB 3. EXTRAPOLATED YAlMMBo A sohemmtie view of the hydrogen bonded chain© ooeurring in the above eryetale im- mhmm in fig«44o Only in ice are adjacent hydrogen bond© eonneeted through the same ato®o Murphy hm© analysed hi© result© for ammonium dihydrogen phosphate md ha© explained them in term© of ionic conduction in 12Î

sol&â dieleo-fcrica «h©r© th@ aotlvatlem Qaeygy eoasists & î tüo termso ®A ® ^ * % % wh©TO le the ©nergy fo r diaaociation into ion pair© and the activation energy of mobility o Muxphy obto.ineâ f # r ammonium éihydrogem phosphate by studying toped crystal© and obteinei a value of 10o5 Koal/mol© heme© 19^8 Kcals/aolOo A % similar value has beem obtained for 1©©% 22 Kcal/mole hemoe it does not mreaaommble to accord a aimilax^ va3,ue to oxalic acid dihydrate which has the mmm Q-41 — 0 Mnd® Em the oa« of polyaraMea a higher value ha© been proposed

^ 30 Koala/mol©® A large varietiom existe between the activation energy of iee tm û th e o th e r oryatWLm which im due to the 3,@w@x” activBtiom exwrgy of migration in icQo This i© due to the hydrogen homûm im io© being in jm stm poaltlem w herew in the other cryetals they are separated by a three at@a system which subsequently require© a much greater energy for re- ©rientatioxi of the ehaim© into a suitable orientation for comduotiomo

àn Imtereetimg observation m a d e by Murphy who f o m i |^î‘î0S|<;»bca.fe th a t after overheating th e crym tals of mmouium dihydrogem in c re a se In the conduetivity of pure erystal© by a fmetox^ of 100 could be obtained which gave the seme aetivatiom ©norgy® Shi© ©oulcl only be explaim ed im term© of © tru e tu ra l ©hang© which

Imereaaed the equilibrium eomeemtreitlom of imtriaai© i o n © ami i t was suggested that imtexmal surfaces m re formed® A s im ila r explanation may h o l d f o r the variation observed In oxalic acid t©o IV bo BiffusiM Aeldq io SuBimary o f R é s u lta o Molecular aelf-dlffuaiom mB observed in benzole acM single erjetal© by the sectioning technique using labelled benzoic acM as tracées Tm diffusion meohanlama were oWervedpafaat p ro c e ss BMÛ a alow proceaso The slow process was believed due t@ bulk diffusion and- wa® found to satisfy the following Arrhenius equation perpendicular to the (001) planeo 12 % 10 HT The “fast‘d process varied with the crystal used and is believed due to non^aquilibriTO structural defects in the crystal©o Tritium diffusion in ©ingle crystals of benzoic a c i d 001) wa© found to be much greater than diffusion amd'^mry with th© w ater vapour c o n te n t o f the d iffu s io n -collo Dmder dry conditions a lower lim it for diffusion wa© obsorved which @li£ th e fo llo w in g Arrhemimm eqmationo 2oamÆi&:Qo .0 S’,S’s 0'v «»0o4 HT i ^ ) %thlm ©Eperimental. error tritium \mm Is® found to diffuee at the same rate in single crystals of benzoic acid ^ d^o Tritium diffusion in benzoic acid ©ingle cry©tala doped with p« terphonyl %md©r dry conditions were found to obey the following A rrhenius

4o7 * 2 â é M A ^ i 8 S 4o 4 129

Tritium and diffusion in polycry©tallIme bansoic acid compact© were found to be the same wlthim esperimental accuracy and ©atiafiec the following Arrhenius equation 1 .,A 1 B s 1 X IQ^-’- . a & ia a ü a .Q S (iM ©OT RT J

Ml e%ohm % 0 reaction waa observed between polyorystalline tritiatad benzoic acid and inactive water vapour which had diffusion controlled kinetioso The rat® eomtrollimg prooe©© was foimd to obey the Arrhenius equation

in the temperature range 20«^' ^ bo lo molecular Diffusiorio 1 /i Diffusion ©f 0 “ labelled benzol© acid was much ©lower than that of tritimm and appeared to have tm d is tin e t diffusion mechanisms operating simultaneously whereas with tritiim only one diffusion process was observed□ Similar diffusion profiles have been obtained in other diffusion studies in erystailin© solidsf, and it has been ©oneluded that th© slew process i® due to bulk or XatLie© diffusion and the process due t® structural diffuaiofâc. This latter process may vary considerably in rat© aa the type of strap©tural defect present in any crystal, will depend on the impurities present aM t&0 uon ©cjuilibri» defects ii it reduced during crystal growths Th® large variation found in the ®faat^ diffusion process agrees

with ü reaao ning » Th© bulk diffusion process is vers?' slow compared to previous "î?û^iœ V studies on molecular ®©3,Ms in whieh bulk diffusion ooef;ri©i©nts at th© melting point w©r© found to be approximately telO cm'“’a©c ‘ «Ï g ^ "5[ compared with 2o3 ^ 10*^ cm eecr*" in this study a Several

atudies have recently been mad© on anthraeen© -and imphthalmn©

and . mu#; lower va3.ues have been obtained whi#. are of the order

found in this stm%o â oompariaon ©an b© made between th© ârrh©nius diffusion parameter in teœoic acid and other molecular crystal® anil

thes© are shown in table JD M l® XWI SSMF BÎPFÜSÏOM MOLBCUMB ORYS'UIS

PLASTIC CRYSTALS ©BS^g©©"* Steal/mol® Steal/mole RsYsranee cî5rtjrsîps>ir*%p:s;î=T::r,

OHGAMO^OHlAmice Anthracene I 42o4 23o3 ■ Sherwood & Thomsono XI 16 White III 22 Labes ÊOE. Haphthalen© 42oî 1? m i t e Benzoic l=8sl0-*-^ o 0 2 1 ,6 T h is Work,

ertieî5"5-^î5iTi=tT 10 0,79 Oo 46 Cr©mer<

j?’rom th is ta b le I t be m ^n thm.tg in general^ diffusion le charaotariW ed by ^îarge pra-^expontial factors and activation energies which are of the order of twice the latent heat of sublimation^ In the case of organic crystals there are two largo discrepancies in this statmentg pivalic acid amâ anthracan© III, In pivali© acid ^ however ^ it is bm lie red that moleœlar self diffusion warn not being measured (p %4-B) o In anthracen© III it is beliered by the author that this result ®as due to the quality ®f the ©rystals used as th® difî'iîsîioia profile s ofetaiasd mare aiailar to thcs© found in palycrysjtalline . log

Larg© pre*^©5sp0n©Kit.ial factors ar© asa®©iat@d wit# larg#

^ntm p'lm of activation t o T th© d iffu sion pmceoa which t n turn is Indicative of a eo-operative phamomena in which more than

om moleoul© Is involved in th@ diffusioja prooaaSo From the atomic theory of d iffu sio n the pre-^expontiaX factor^ Do,

„ 800 cam ho ©xpresaed as B@ sal 0?' "f 0ssp ^ (ij) where % ia a constant depending on the d iffu sion mechanism, a

is the jump distance in the atomic (or moleoular) processp l u & frequency which is approximately the mean vibrational frequency

of m i atom about its ©gmllibrium aifeOp B is the gam constant aM is activation entropy fo^ the diffusion process. à valu© of therefor©, be ©aloulated for benzoin

aoMp IfaM 1? are known. If is a small constant not far from unity, a the jump distance in the Wmzolo acid lattice down th'a (OOl) plane is §,47Êg is normally taken a© th© Deby© frequency of the lattice which ©an be obtaiïiad from the D©by© temperature ®p D from th© equation

©_ era3 c2S=j;?r==:-j» . k . ? "^ih” Th® B@hy@ fmquQia^j &f Wnwle a©î4 is 116%

, 3 , 2Î 1=385l10“‘*-®2£ 116 4 s; 6 .6 2 % lO"^^ - 1.88x10''-^ S8@“ ^ Substituting these *v©lues in ©quption (81 ) end using th© value of In equation ( 6> ) 1 S X ^12 t**»'*#'C*i Ttw !Tïr:fCi»IfS»lÿa«=ÏE-Jaf«!taRI»ElîItt-£t» I f 1 .6 2 « i a > - ^ ^ (5.47 » 10®) =

5 .5 X 1 0 ^ ^ - 2.303 ^ 2 z 14.52

Thia large activation entropy la larger than expected from a siaipl© vacMoy mechanism if compared with systema in which UË, auch mechanlem have been well established and indicates a mechanlam in which aeveraX molecules are involved. The theoretical energy of formation of a vacancy in a iuoleculrr solid is approximately equal to the latent heat of 803 sublimation, Expérimental evidence has been obtained from ©pecific heat measurements in solid argon which indicate that

- lo4 21 H^o This suggests that a vp3.ue of approsrimately twice the l a t e n t heat of aubllM f-^tion,) la not an imreaeonabla value for a vacancy process a© ^ AH^ where 1# the enthalpy of migration. This valm i© also of th# order found in metal vacMcy diffusion atudias, In metol©^ however^ there is little relaxation of th# aurrounaing Irttice on vacancy form ation om to their electronic structure wnereaa in aolecu3.ar c r y s t a l s due t o th e ir malM Van der Waale binding a rela x atio n o f up to has been auggeeted thim giving a email region- around the vacancy in which the molecule© ere mobile ©no tend in th# lim it totmro© liquid like behaviour, Thi© type of defect would be expected to have a large entropy change asBOeiated with the relaxation of the ©urrounding molecule©, Mlgmtion could then take place by a molecule melting into this cluster &.b another freeze© out with the lower lim it of the activation energy of migration being of th© order of the latent heat ©f fusion g This mechaniam has been proposed by j83achtrieb and Handler &Bé) to explain ©elf diffusion in alkali matais and by Hood mxû m Sherwood for self diffusion in oyolohemne. It was suggested that th© number of moleculee agsooiated with each relaxed vacancy or ® relation % was given by the equation

c«> Ü L. ^ esara;rii52Sj % where t im th© melting point Ef the entropy of formation of a io©o vacancy ioQo n entropy of fusion.

The overall not.lv©!ion entropy is given by where m is th© entropy of migre.tlon hence must be known before can be caleulatedo M V From th© preceding theory however A@> should be very small compared to AS ^

n £S3 A S f

^n the cam© of benzoic acid the entropy of fusioBg AS is

16,4 ©X theref 0 2 "© in tW#-oeae n ^ 6‘=^^ moleeulea, This number Is small comp area with eyolohexame wher© m = 20 and it s©@me unlikely that a liquid like region sxiatB in benzole acid, mo S im ll02" valuae of n were obtained for anthraaen© (n^’5^-^6) and m3 naphthalene 5-6), The minimum activation energy predicted by thl# theory is ■Ki A # A 087 , which for Wmzoic acid ^ 21,8g 4" 4^1g ^ 27,0 Koals/mol# The observed motivation energyg however^ is 44^0 Koala/mol# which suggest# that In such a email "[email protected]" there is mo liquid Ilk© behaviour and that a considerable activation energy of m ig ratio n im mtill requiredo The diffusion nw&hanimm f o r molecular diffusion in benzoic aeidg therefore^, empeazm to be ome of xfacancy diffusion through relaxed vacancies.

piffusiono Both tritium and 1, Æ ‘ diffusion in polyorystallIme benzoic acid were found to give apprmclmately the same Arrhenius ©quation® when analysed for th©Whipple solution to the grain boundary problem by the method of LeClaire (poto4)o This molmtiom vm# ‘i 4 much superior to th© piaher solution and lowered th e C ’*’ diffusion aotivatiom energy by 10# Th© fa c t th a t the same d iffu sio n param©ta ra were obtained B©mna that the aame mechanism im responaible for b o th phenomena, Th© activation energy of grain boundary diffusion ia normally lose then that for bulk diffusion hence from the tritium d iffu sio n BtMûim in single crystal# of foonzoic acM the activation energy should b© less than 20 Kcal/mole, Ae it is very much greater this indicstea that same mechanism does not operate under grain boundary diffusion conditions and hence it is 'concluded that both isotopes migrate by molecular diffualo&i, A comparison with other aysteme in which bulk and grain boundary diffusion have been measured la shown below.

MOLECULE REBBREMOE % L H B BULK G« Bo KCal/raol@ KCal/mole h L Ag 45 21.5 0.48 Zn 23 14 0.61 #6 QÛ 18o5 13 0.71 t6 F© 64 40 0.63 % Benzoic Acid 44oO 31.8 0.72 THIS WBK Dümphthalen© 42.7 29.6 0.69 J©S From these r e s u l t s I t appsara that a o le e u l a ? âiffiœioffl ha® oharacteriatics similar to metal crystals and auggaste that a relaxed vacancy diffusion procès© ie not unreaaonabl© in these molecular crystal©. It i© interesting to note that the activation energy of the grain boundary diffusion process^ 51,8 Kcal/moleg approximates to the value expected from the liquid like "relmxion theory predicted for bulk diffusion where th© minimum activation energy predicted vmm 27 Keel/mole. This suggests that in non plastic crystals this theory may find application in the more highly disordered regions in grain bouMarieso ili ïï-itiuM Diffusion ia Beasoie Aeiü Single GrystalB. In th© initial tsritiuB and G diffusion studies tritium 1 A mm found to diffuse faster than C by a factor of 100 and it was thought that thi© xmm due to a proton tra n s fe r mechamiam. It was soon reallmedp howeverp that thia was impossible as the corresponding conductivity was foimd to be less than 2 x ©too a fmi degrees below the melting point. If this were dm© to proton conduction the corresponding proton diffmslom coefficient obtained by applying th© Memat^Eineteim equation ^ 'U i© < 10 œoSeCo o AB the tritium diffusion coefficienta obtained were of the order of 10"^^^ om§ 0 #c%^ pro tom comdmctivity im this system la m.egliglbl© and another dlffmelom mechamlam must be occmrrlmg, There appear to b© only two poælbl© ûiîînmimi mechamlemmo 1} fh© hydrogen diffmsee by itself by a memtral switchlmg mochamiem, Thia could involve breaking thé four hydrogen bonds of two cyc3,i© âlrmm^ reorlemtimg two bemzoio m M moleoulesg reformatlom of a mom dimer and tramslatloHx of th© hydrogen across th© hydrogen b«io The aotivatiom emergy maeeeaary for auch a prooo 0 ©g however^ may b© prohlbitivaly large, Ê) Mffuslom of mm Impurity molecule im the lattlo# «, Several pieces of expérimental evidemc© were obtained from which a p o ssib le mechamiara mm deduced, The tritium diffusion coefficient was found to vary with th© water content in the diffusioia call, A lower 3,1mlt to the diffusion ©©efficient was obtained which rematoad raalterecl on ûrylng the cell# prior to th© diffusion anneal, When the diffusion cell contained approximately a constant water.vapour content the tritium diffusion was found to ©bay an Arrhenius equation having the same activation energy ae diffusion under dry comditiomao These facts indicated that tritium diffuses by m im purity meehemlam which may be d iffu sio n ©f w ater mol ecu], es in the bemzoic acid lattice, Tritium diffueion waa found to have .approximately the aam# diffusion coefficient im teuterobemzoic acid and normal b©mz@i© acid when carried @ut im the same diffusion cello If the diffusion were due to'a' neutral diffuaiom of hydrogen a© suggested im 1) above them a larg© imotopo effect would Im expected which could r@mg© from 1,41 a©c©rdimg t© classical kinetic theory t© a maximum ®f 3,3 from etatiatical rat© theory according to r§f; Bigelaiaemo Th© fact that this mm mot observed imiioate© that the hydrogen is dAffuslmg m part @f mom© larger speciem which may b© a water molecml©, From'the grain boundary diffusion studies im polycrystmlllm© b©mz@ic acid i t mm found that tritium mû diffused - at the a«© rate, This shows that under these conditions tritium diffuses with th© bWlk moleculeg a further indication that im th e single crystal© m im purity ymohmimy^ i© operating which ia mped'by the imcre#©ed bulk diffusion im grain boimdarioBo îhe rapid ©échange reaction found between tritiated bemsmlc acM g#d inactive water vapour ©howa that in th© diffusion aameal cell in the temperature range usedg 88-115^0g equilibrium between fHO mû tritlated benzole acid will be rapidly attained. This Indicated that diffusion of trltlated water in Mmzoic aoM may occur uMeir the ©oBclltions usedo From the above Information it was concluded that umler th© exporlmeatal eemditioE&e used tritium diffusion to Wmzolc acid oocur# hf diffusion of water moleoule# through the crystal lattice and that the lower Itolt of the diffusion coefficient foimd mB éim to water trapped im the crystal during growth^ fher© are two meahmaismm by which this topmrlty diffusion cam ooourg either by am im torstitial or vaoamoy moohmism. If water diffuse# through vaoamoios im the Immmio m fA lattice thorn the aotivatiom ©morgy of th© prooeas From the experimental atudiea om argom it ha© boom ©hoim tha,t 084% US' AH ^ loi lo4^AH, , Hemoo for bem»!© aoM Â II ^ 24^31 Koul/ 08T BO-le 21,0 ICoall/gaolOo She emergy of migrmtlemo o # f a water molecule will fee ©mal3> but merertholeaa it appear© that the energy required for such a s^eofeamiam is much la rg e r than th a t foiUMl ezEgerimemtellyg 20 Kcal/mol©o llmo® th is rmggmtm th a t mm i n t o r a t i t l a l iiff u s io n meohemimm i© more lik e ly , Evidemoe for am to term titla l meehamlam was obtaimed from the traltlum d iffu sion stu d ies to p^torplumfl doped crystals under "dry" eomditiom© to which a much lower motivation energy warn JLW obtstoedg 7,5 Kcal/mele, was mot expected to introtoc© vacmnciee in the 3,attiee but ûm t® it being smaller than the benzoic aeicl it was thought that it could give rise te im teratitial holee through which water molecules could diffuse. If this cam be regarded a© atoilar t@ the cmee ef eog, m dapei with OdOlg to which a m n equilibrium vacancy cemeemtratiam is introduced g thg mi extrimaio diffuai@m mey be obmrved wi# Lo If this analogy i© vaS emâli,., for water cliffusion im bemzoie mUi ^ 7o5 ICca3./m©l© tS. ^ 28^7,5 - 12o5 ICcal/mol© where AfL ia the energy required to fom the interstitial hole, Estimmteaef the amtropieaof formation and migration may also be calculated ualmg equation ( HO ) for For tritim diffu^lmg lime B fig,27 the activation entropy •L^î A b fer this mechanism earn be calculated if the mame values ©f

^ 0 a and 9 are asamedo AB i s m tg however^ markedly depemdemt values and th© approximate value i© 2 2 A g ^ (5o# .E X loBg X H

_ '4 _ & 8 î^SoO @oU, This value is much lower than that for bulk diffusion miû ia mot imreasomable fo r am t o t e r a t i t i a l meehamiam, For the tritium diffusion im the p==terph@nyl doped crysta3.s ^ «"23,3 OoUo Acoordimg to th© mbove reaaomimg t h is should rep resent 9 A g , This large negative value was imexpeotod but may be quite genuine ae other diffusion studies ®f small Gmleculee im cryatols u£.l K&»

Of larger moleoul®® have yîeléleâ jaegstiv® entropies @f aetivatloa lâô ©og. Hg/Ms, A s ''“ “ 5.3 e.a."'; M j/M a le ite o -4 .8 5 e.mg aM

oaXcMlateû valaea @f A 8., far éiffw io n ©f Hm m û argon thrangh ® m polyvinyl a@etat© s®r® -2o2 sM -7 .8 e.m. re s p s c tiw ly .

The entropy ®f form ation ®f the fi@f©«tp ASj b therefor© is

1 A 3 p U 3 ^ 9oO c=s n CO, o ilil Û Again this ia raasonabXa value being about E'&alf the activation entropy for bulk diffueiono Another factor which imdieated interstitial diffusion wm th© variation #f tritium diffusion with water ecmtent im the cello This phamomema was not reccgaised at the ttoe and a# quantitative data ia availablej, h^acejoaly a qualitative tliaoœaiom om b# mai# Gomcemtration depemdemt d iffœ io m has been observed fo r iEM vapours diffuaimg im polymers' but it warn thought that a more comparable ayaWm was the diffusion ®f Em im EmO which has beam observed by Md M#@re vary aeeerdimg t® the equation m M mm where a is a temp, depemdemt ©ometaut mmd g mn. im the pressure ot Mn vapour above the crystal, This wm sim ila r t the relation expected for imteratitial. diffusion #f & diffusion ©©efficient could be expresead as ,S)o5 ■CSÎT»i Hf L ÎI3ft3 tsJ fch© Qmrg^ @f fomat&en ©Y the intesfstifeial gn tm n sltl© n , ÂG fch© &mrg^ @f fomet&en ©Y the intesfstifeial gn defect g and d^ the atomic jump dlatance. Hence under conditions of constant pressure llineea’ Arrhenius p3,ota would be expected giving s'îonstant activation energy, A similar effect haa als® Wen observed in diffusion of cobalt in OoO as a function of oxygen pressure but in thia case the exponent of the pressure was 8M* found to TOrj slightly with temperature, ' The variation of tritium diffusion with water vapour ©.ppeara to fit a similar o f Mechanismo In this case there are two activation free energies to be explained AGq, andAG^o From the above discussion or r©.ther

^ ^ 20,0 Kcal/mole for the intrinsic diffusion process,ç thereforep ah estimate ofAH^ i> T68^ ) tha enthalpy of formation of the defect which in this ces© la presuBted to be a water molecule on an interstitial defect site can be obtained from the activation energy of the diffusion process under a constant water vapour content^ ire© figure 27 line A,

AHp ^AlineA

^ 21,2 ± 2 - 20o0 ± lo2 1,2 â; 5o2 Kcal/molCo i«©o the enthalpy of formation of theolefect is wry email, A ro% h calc u la tloKi shows th# to b© tru e . The processes Involved in making this defect are a) th© creation of an interstitial hoi©g an ondothermic process ^ for which AlL ss ia.,g KoalB/mola and h) removaJ. of a water moXeouJe from th© va/pour and replacing it in this intoratitial position^ an exothermic process which will b© of the order of the latent heat of condoMation of water^ ^ '^lO Keel/mole, The energy of fos’mation of t>lm d efect p the ref 010p la 12,5 - 10 ^ 2,5 Kcel/mole ^ in ©pproxtoete agreement with th© valu© found exporimentally, It is beloved that th© evidence given above is conclusive proof of an interstitial migration of water in benzole acid crystals, A comparison cam be made wlth other systems in which tritium diffusion has been measured by a sectioning technique and are shown below

MOMOULB E. Bo Mi^ Ba <3 A ^ ' KT%TH J? ^ îtot/wïjSa crf’ssE î&aî/wsajle

BmîieîC Â#@ 80.0 0 . 3 21,0 0.92 9. mua^fsSQtïfe %WA-UO ACID 3.0.0 2.23 10.0. 1.00 12.3 ïe® I « 13.3 2.0 11.3 1.19 15 Ï I 13.1 2.8se10^ 1.39 In all three crystal© a ctee corrélation ©sriste between the parameters of the diffusion equation and it is possible that the same diffusion mechanism %b operating. In tha pi Valle acid diffusion study the acidic hydrogm we.a tritium labelled 0 I t -wb.b belig ved th a t bulk d iffu sio n was being measured as the acid had been shown to hav© a v©ry low conductivity in the solid and thus the tritium would remain firmly bound to bulk 'molecule0 The result obtained^ howeverdid not agree wit] other measurements on plastic crystals (see table XVtll), A possible explanation is to b© found in th© purity of the. crystals used which were known to have 0,04 mole # impurity^ probably water whl& is known to be very difficult to remove. ' He mo a the author suggests that the diffusion mechanism ieon© of interstitial diffusion of water molecules trapped in the latt#© during crystal growth. Experiments are being conducted at present In this Ik boratory to determine the mo.«.@cular diffusion parameters using 0^^ labelled acid. The fact that similar diffusion characteristics were obtained in lc©B in which it has been conclusively proved that th© diffusion occurs by migration o f the bulk molecule^ suggests that in this system diffusion may Indeed occur by the "faee" interstitial qg, mechanism proposed by Cnsager a n d hurmela,

Co Diffusion in Acetic Acid, Due to th© elobvag© properties of th© crystals diffusion could not bo measured along the hydrogen ©ond©d chains and was^ th e re f 0 1 Gp attempted perpendicular to them. Large variations in tritium diffusion rates, wore obtained ^ how ever;) which, were believed- ûim to water adsorbed on the crystals as it had been shown in the previous section that this can greatly Influence tritium diffusion benzoic acid. The lowest diffusion co«-©ffici©nt obtained w

c a H P c=3 1 iOi cm " sec'"' . at o 0 hence this mm t represent an upper lim it for Ik or intrinsic proton diffuaion.q,,__

D u g to th© d lfficu ltto s Involved in handling this system it mu abandoned. X4^ do Conclu©!onSo Iroton transport 1© greatly enhanced In systems having co-operetlvely hydrogen bonded chains due to the low energy rnigrcition pathways they provide. The presence of moieti^e In all types of hydrogen bonded solidhowever^ aûn affect the observed proton migration process. Attempted proton diffusion stud^ s in benzoic acid single crystal.^ using a tracer sectioning technique resulted in the discovery of a concentration dependent diffusion process which was concluded to be diffusion of impurity water molecules in the lattice by an ii*ter- s t i t i a l mechanlmio A lower limit for this process was observed which obeyed the Arrhenius equation* 4*2 D = Oof 20,^ 00 ^ 1 ,3 0 0 ‘^Qo 4 — df .B is belie TOcI due to intrinsic diffusion of water moleculo® trapped In the crystal during growth. Bulk diffusion in benzoic acid obeyed the following Arrhaal

équation ^ ^

« 6ŒîreZ5alltoïsrriRr«ey7 44,000 ^ 4^00 1.Î9 J fi cm which it was concluded that a vacancy mechanism was Operating in which molecular diffusion occurred through relaxed vacancieso

Diffusion and conductivity studiesin acetic acid were found to be very dependent on the water vapour present in the surround g atmosphère. These result© indicate that the physical properties of aystem© containing hydrogem bonds may b© greatly ImflueMced by th# presencepf tracas of nmlBtw© and th a t g re a t car© must b© taken when intrineic propert# m of such system® are being studied.

©o Future Work A logical œtenâ-om ®f this work would be to determine the exact dependenoepf the tritium diffusion coefficient on water vapour by carrying out a series of diffusion ^périment® imdtr controlled water vapour pressureso Due to the slow diffusion rat© this may be more easily performed by a water vapour exchange technique o

Am extension of this study to other hydrogen bonded solW© iu desirable to verify that such an interstitial diffualom of water in a regular close pac&<.©d lattice is not confined to organic acidSr.

Vérification of proton diffusion Dy a tracer technique ohoaM b© attempted in a system which exhibits a sufficiently high conductivity for proton diffusion to be measurable by a tracer soctioning technique, A suitable crystal for such a study might be Ammonium dlhydrogen phosphate^ Rm?BHSMGKS lo l?o S a lts,”ïhe Moteim Theory of Sollfls? MeGraw-HiJ.! M.Y. 1940 S. JoJ. Brophy and J.W. Buttrey (R)S) "Orsasiie Ser.jieonc1uctors« McMillan HoYo 1962 3, i« tollm an am# M, S ilv e r (EDS) "'Symposium on E le c tric a l Conductivity in Organic S o l i d e ^ Duk© U niversity I960 Interscianc© 1961, 4o- iF^6@o PhySo H ot , I960 U9 1226 §0 S^o Riehlp Ref a 3 p, 61, 6o DoBo E l e j o DoG, P a rflttg M,J, Perry and DoH, Tayaum^ Tran©, faraday So®, 1955'49- îo Aola Kitaidgorodskil»^ Organic Oift>em,crystaXlosrahpy""cpoS3 C<^^nBult&r: Bo G,(L Pimentai aM A, L, McClellan, ' ""The h y d r o g e n B o n r O ‘^ B u reau ^ ^

3o AoE, Dbbelohde and ICoJ, Oallaeher, Acta Orjst, 1955 B 71 lOoMo Eigen and Lo Be Meyer, Pro®, Roy, Bo Qo 193B A24? 303 lloHo Ao Morn©, J , Imorg^ -EuclearGhem, 1963 1139

1 2 ,Y0 Kakluchl, H, Komatsu and Bo Kyoya, J* Ghem, Phy©ol95I M 132 13oQal?o Smyth in "Thyale# and Ghem^^try of the O rganic Solid State 0 Labes and W elsm berger, (EDS) ) Vololo pol28 Imterael<^mc@

14odo MoOo Pollock and AoH, Ubbelohde, Trane, Faraday Soo,1956 .§2 112 15oMoMo Cardw and.BoDo ll©y<, BieCo laradaÿ Soc, 1959 2? 115 16, Do Do Iley and RaB, Leslie "’Adsorption o@ Water on Solid Protein^. Advances in Ohemical Physics Wf pa23B (Suohesne ED) InteraciJnoê • - •• • -, ' y 19 6 ^0 o lîoGo King amd Jo A, Medley, J, Ooll, Bel, 41949 4 9

3,00 Do Do Ile y ami D, Bplvey, Hat^. © i 9 6 0 ^ | | 123 ISoJoOo Dacroly, M, Granioter, Oo Jaeeard^ 'Belv, Phys, Acta 1957 30 465 2O0A0 BteiMaTOo iolv, Phys, Ac ta 1937 30 5B1 2 I 0O0 Jaceardg RelVo Phys, A cta 1959 32 89 22oGoPo Dr@t?n ®M Bo A fte rg u t Hef, 2po@9« 23pioBo Liddiard «liandtach Der Physik"' V # Bo flugg© EDo ' Bprl%er-V@rlag Berlin 195? 24oDo Mapether, EoEo 0rooks and B, Mau#orp J, Ghem, Phys, 1950

' 18 1231 / 25o l’o Fabbrio B, Laszarini and ¥« Sanguist, Int, J, Applied Hado 1.1964 15 43? 26o Mo Eigen, L, De Meyer and H, Go Spatz, "'Colloquium on the Physics of loe Crystals” Brlenbach^Zurleh 1962<> 21 o GoÂo Sim, JoMo Eofeerteon and ï, Hp Goodwin Acta Gryst«1955 § 15: 2Bo âoHo Winchell ”Optical Properties of Organic Crystals”, University of Wisconsin Press, 1943, 29o HoBo Jones and D, Templeton, Acta Cryst, 1958 11 48? 30o Ho G« W, Wyctoff ^Crystal Structures” fig, 11¥A, 30a,

31o Fo Ao Krd'ger, ”Th© Chemis tsry of Imperfect Crystals”o Horth- Holland Publishing Go, Ametêrdam 1964o 32, H, Ko Buoklayo ”Oryatal Growth”, p,54 '^'iley 1932, ^ 33p IWd, p,339o 34o Go Bo Gottlieb, J, llTOtroohemo Soe, 1955, 112‘ 903 33o Bo ICyropoulos, Zo Anorg, Ohemo 1926 154 308 36o Po Wo Bridgman^ ProOo Amer, Acad, A rts, Soi, 1925 |0 305 3?o Bo Co Kremar©, J, Opt, S©©, Am, 194? J? 33? 38o Jo Bo Sherwood and So Jo Thomson, J, Soi, I n s t, 1961 J? 242 39o Ko To Bo S cott g So K, Hutchinson and H, Lapage, AoWoHdo Report BOo 0-4/53o 400 Ko Bo Wiberg ^Laboratory Techniques in Organic Chem# try”, McGraw-Hill 1960o 41o Fo Wo Schwab and E, Wichers, J, Research BoBoSo 1940 23 74? 42o Wo Go MmUo Jo Metals 1952 4 74? 43o lo Bo Hannay (ED) ”8©miCDnductors” , Heinhcld Bolo l339o 44o Bo Go Wclf and Ho P, Deutsche Baturwisso 1934 41 423 43o Go Jo Sloan In ”Physies and Chem# try of the Organic Solid 8%t@” ?®lœ© I {BD o Do Fmto M, Mo Labes and A, Weisabergero ) p, Imtermelenee BoYo 1963o 453 lo BmMUo Brito Jo Appl, Phys, 195? 8 457 41 o lo Go Pfann Malting”o Wiley, i,Yo 1958, 48o Ho Jeseupo Quoted in Ref o Ho 49o Wo Do LawaoB miû 8, hielaen "'Preparation of Single Crystals” Butte rvj or th Londo n 195Bo 50o So Ko Hutchinson and H, Lapaga, AoWoHoKo Report Ho, 0-73/54o 31o Go Mo Hood and J , 53, Sherwoodo B rit, J, AppX, Phy©lea,1963 14 21 Ao Jo Goss and So Wointrouh Mature 1951 If? 549q Jo Mo Sh©Moodo Pho Do fhesiso OniTOrei'^ of Durham, 1959o 54o Fo Wo Schwab and Eo Wichers, J , Research Mo B, So 1945 J4 333 33o Lo Wo P ie te tto Pro©, Roy, So©, 1955 A,142 355, 56o Ao lo OillaMo Do Ho Hey, A, Lambert, J, Chem, Soc, 1941 564, 57o Jo Mo lobertaon and A, Ho Ubb©l©hd©o Proo, Hoy, So© (Lends n ) 1959 A170 222 5So Jo Mitchellg Do M, Smith,”Aquam%try^ Interacience 1956, 5®o Fo Wo Schwab, WicWsra, J, Research MoBoSo 1944 35 121 6bo Lo Iy@n®, ”Guide te Activation fnalyela” Van Moatrand l@64o 6l o ”Th© Hadiochemlcal ssanua 1 ” P art 1, p, 18 anrl 56, The had 1 ochemlca 1 Gentre, Amerehamo England, 1962, 62q go Bo Oroutham©! "'wmma Hay Spectrom etry”« o (> o H, Z. teistm lle g s’o 1921 J6 160 ®§o P. Hi. ShswBOBi "w lffusioïî iïi Boliâs**. MeCJaaw-iiill 1963. 66o j . feo 8h©raooâ aM u. J. Ifflait© “ïataraatloifïsi SyraposiMia mi Offgasîifâ c;s*y@tals". Ohieago. amj 19&3.

„ Go Wymonie “Ts^aeaw Radioactifs im M©tallu2*gi® Physlqwi©*’. Dmnodo i'as’is I960 lo Wo '=Diff«Jsioia® Academic Prasso M.Y. I960. 59. ïMdo P. 16. iQo Ao Sravigelskas and E. Kl&'kaMall. îra n s. A.IoMoK 3.94# 1 |1 130 ?lo Wo Kiahm aM M. SMrlmnaf ^ Malv. Ghlm. Acta 1958 £1 938 ?2o HaM&ook ®f Ghsmfe try aM Physio®^ 43rd m itlprn.p.2403«2420 Gh@®i©al R«feb®r Publighing o©. 1961=g„ T3o 00 B@%@1 afflâ Bo Rl@hl, “Oelloquiissaa &n th® Phj®i©s of Io@ (Jrystals'’ Erl©abach <- Zurich 1962. 74, Ho Eo Pawel and T, So Luaidy, J, Chem, Solids 1963 26 957 13o Go Mo Hoodo Pho Do TSiesiSo PpiTOrsity of Glasgow 1955, 7'6o Go Ho Lee, Ho K, Kevorkian, F, So Reuoroft and }L IL Labes Jo Ghem, Phys, 1965 £2 1406 77o Ho lo Ongnade and Ho W, Lmmbo J , Amor, Oh@m« SoOo 1962 74 31B9 7So«^xJSM©nhoff, Anal, Ohem, 1950^22 329 79, Go Go Boll and M, F, Hajes ”Liquid Sointill.ation Counting” Pergammon 19§8o xt>.Davidson and P Feignisono Into Jo Applied Rad"no I, 19,57 2 1 lo Bo Scteam, ”Organic Scintillation Deteetoa”, ' p 86 Elsevier

B3o Do Lo William®, Po Mo May©©, Ho Jo K"Ard©l and Wo Mo Hegei Muceonics 1956 1,4 62 B5o Jo Co PiaLero Jo Applied Phys, 1951 1 74 84o Ho To- Whipple, Phil 1954 45 1225 S5o Ao Do Le G laire ”G©nfer©n@© on D iffusion and Mas® Transport in Solids”, Reading 1962, B rit, J , Applied Phÿaie© 1962 ^ 43 % BSo p. Go ShewMono Ref, 65 p d ll o 87o Wo Peitkneeht, Ao Wytt©nbach, W© Buser, ^Reactivity of (D©.B®©rs, ED) po234. Elsevier 1961 8®o ¥.c Wei and M, Bo Bern®tain© Jo Phya, Ghem, 1939 §3 738 S9o Ko Itagakio Jo Phys, aoc* Japan, I^é4- ij. |o9l 900 Ho Riahlo p riv a te Gommunl©at 1 on , 91o Ho Gmnicher, phys, Kendene, Materia, 1963 1 1 9So Go HasSo PhySo I«ett©r@ 1962 3 126

93 0 Lo Onsager and Lo Ko HmnelSo ProCo Mat, Acad, 8clo 1963 JO 208 94o Do Do Elej and Do lo Spivey, Trans, Faraday S©c<, 1961 57 2280 93o Ho B je rr« o Science A52 113 383 9S= S« J. Murphy» J . Applied, Phys, 1964 21 2609 " D» D» Bley, Ao So Paracett aM M» R, W illis, datura 1963 2 0 0 2 5 5 '0 s. KoMe and T. Cda, B u ll, (3h®m„ ^oc, J&pmn, 19i4 2% 56? 99o A, Sperasip Z, Bleetrochea, 1955 31» Î00» JoMo Sheraood aad 8.J» ^hoaaoa, Trm.B„ Faraday Sec,I960 5 6 I 4 4 3 101, HoGo Jamaglm and JoMa Sherwood ^International Sympoalmm om Organic Crystals” Chicago 1965o 102o GoHc, Leeg HoKo Kevorkian^ PoSo Heucroft and Labea^ Jo CheMoPhyso 1965 §J2 1406 103o DoJo Whiteg Private Communication, 104o GoMo Hood and JoM® Sherwood^ ”Gonf©r@nc© on M olecular Motions and Phase Transition© in Molecular CrystaS.©” Paris June 1965 To be published in Jo Chin Phys, 1 0 #o Ao Berneg Go Boat© and Mo de Pasg II ^uov© Ciment® 1962 2£ 1179 3 .0 5 o l o H o H achtrieb and GoSo Handler g Jo Ghemo Phya, 1 9 6 1 ^ 9 7 4 3 . 0 7 o HoBo Cuddeback m é HoGo Drickamerg J« G h e m o Phya, 1951 1^ 790 1 0 0 0 ICo "Reactionen in und m Festen Steffen" Berlin 1 9 5 5 1 0 9 o K o Cremerp Z o Phyaik Cheno 1938 B29 4 4 5 n o , Hofo 65 Chapter 2, 47 256 l l l o Gofo FfâTOkaTOg HoEo MeCcskey and Gofo King g Jo Research jNoBoS,

112o Do Lasarusg Sclid State Physics 116 Academic Press i 9 6 0

1 1 3 o Eefo 68 Pa: 1 9 6 1 78 1449 Po Plubacherg AoJo L eadbetter and JoAo M@rris®ng ProCoPhySoScCo GoMo Hood and 3oMo Sh©reo@dg Molecular Crystals ^ (1966)

U 6 0 Uolio ^ach trleb miû GoSo Hamdlerg JoChemoPhye, 1955 2 g 1187

1 1 7 o Mo Davies end J d o Jcmes franso Faraday Scco 1954 ^ 1042 3.16o Lo BigeleisOTp Brookhaven Conference Report” Isetcplc Esschang© Reactions and Chemical Kinetics” Brookhaven National Laboratory (Long Islfind) MoYo

3 . 1 9 o Kef 0 6 5 P 0 I 4 7

I 2 O0 HoMo Bprrerg Trana, Faraday SoCo 1942 J8 78 121o Po Me8resQ Jo Amero Chem, Soo, 1954 3*22o Po Meare^o Polymem s S tru ctu re and Bulk Prcpertie© Ch, 12 Van Mostmnd London 1965 « 123o KoAo S©c@® and WoJ® Moore^ Jo Chemo Phys, 1957p 3*24 o RoKo Carter and FoD, Richardson g Jo Metals 1954 6 1244

1 2 5 o Margenemu md Murphy "Mathematic© of Physic© and ChoBietsy" P 0 I 5 9 Van Kostrand Hdo 1956o APPENDIX I

The parameters m and e in the ©quation^ y ^ ÏÏÏK c were calculâted by the method of iemat mean squares,

For a given set of n oo-ordinated p %g ^

The gradient^ ia given by _ y M

(S <=. E. ^ ^ ^ ^ (g. 90% ^ The error in pm^ and the error inpp pOp can be calculated a® follows. The deviation of each point from the L,M,So lin©p dp is d ^ y^ - {m.x^ -3- e) \ The error in m is given by I/H 11 '.'1 pm s Oo6745 "

2> a \%! — ^ The error in c is given by fpc ^ n—E A computer programme was devised to c alcu late cp pm^ pc. This programme was used to ca lcu late both d iffu sio n e o e fiic ie n ts and the parameters of the Arrhenlus enuation and

The computer used was a F e rra n ti ®Siriua®o APPEMDÎX II Oalcülation of the Diffusloa Cosffieieato ïh® soliâtiom of pi ok "s equation for the eoaditioas useâ vim C/C„/A ” . ®sp W# J 3 S<2 lognA B/G_ s •=" -=^2======.===» ^ 1®Sia A #3t 2.305 4Bt Hemce the diffaaioa eoefficiemt eaa‘ be obtalmed from the gradieato "> 0 0 of a p lo t of l®g,AA VB 1

■n =. 1 ______” 2.303“i a i 4 The gradieatj, m, was obtaimed by the method @f least aaara squares &m Appeadis Ï. ©.go DS18 with .u 10 ea?(, Ï = log-j^â (o.p.So/iaa. ) X ¥ XI

0.25 1 . 1 2 9 0.0625 0.2823 2.25 1.056 5.0625 2.3160 S. 25 1.000 39.0625 6.2500 12.20 0.964 148.8400 11.1608 20.30 ■ 0.908 412.0900 10.4324 30.20 0.910 912.0400 27“4820 42.30 Q.B18 1189.2900 31.1394 0.849 3158.4400 47.7138 12.30 0.812 5221.2900 38.1016 90.20 0.123 8136,0400 65.2146 132.00 0.383 11420.4000 50.5560 182.00 0.432 33120.4000 18.6240 240.00 0.131 m 51600,0000 32.8800 S.JÎÏ3 886.45 10,181 121960.1622 431.4189 a a 0.003803 & 0.000111

2.303 % 4 % (0.003803^ 0.000111) % 165©1 % 3600 4 .8 fe .2 K 10"’^^ea?£39eT^ APPEâPÏX III

OmlGMlatlom ©f the Pprmmetera ©f theArrheaiœ Equation f ^ D ^ I - | |

2^303 RT

à pl©t 0 f l@g. va ^/T glvea a straight lime ©f gradient

A amd Imtereept^ © ^ i©g^ 2 o303B The gradient^ lmter©e%)t emd their erram w ra ©btalmed hj the le a s t mean aqaaree ©wputer pr©gra»©o For T ritim Diffmalom to Bemgele àeM Figo27 Ito© A®

.0 & 0.318) E 10^ © a 2.244 0.994

„■ JS^ ^ 2.303 E 1.98 E (4.110& 0.310) e 10'

® 21 (,200& 2000 I©al/aol©