WINES OF SURFACE FRASES ON CHARCOAL AND COPPER Thank for fh. Dogroo of Ph. D. MiCHEGAN STATE COLLEGE Robert Earle Vandor Vonnon 1954 Ida—‘51:» 4.7V. LIBRARY Michigan State University 0“.— «ILL Shanta: State WDIES 01" SURFACE PHASES ON CHARCOAL AND COPPER By Robert Earle Vander Vennen A THESIS Submitted to tho School of (ll-“unto Studies of Richigan State College of Agriculture and Applied Science in partial mltillment of the requirunents for the degree of MOR 0F PHILOSOPHI Department or Chemiltry 195i: W?! m: of t for g; M”, 5 ‘3’ iez/é ACKNMEDOIEET The writer wishes ta express his sincere appreciation to Professor 1!. '1‘. Rogers for his guidance and helpfulness throughout the course at this work; and to the Office of lift]. Research for s grant subsidising this research. W W M an “ I ”Ill ambition Bhicll Lab. - Won: St]. W 0! Studiel ”it? mbjec 51M lubjec' I “minim ml hm, OCtObeJI Under-glam] “’qu 3%: lo 3. Mme mumition Whoa I Hub” 0: 5m m Kipp; Pu: VITA Robert Earle Vendor Vennen candidate for the degree of Doctor of PhilosOphy Find. sminstion, September 9, 1951;, 10:00 A. 11., Room 128, Kedzie Ohmic-l Leborstory Dissertations Studies of Surface Pheees on Charcoal end Copper Outline of Studio” llsJor subject .. Physical Chemistry Minor subjects - Physics , Hethamtics Bioptphicd. Item” Born, October 15, 1928 , Orsnd Rapids, Hichigen Undergraduate Studies, Calvin College, 19146-50 Graduate Studies, Michigan State College, l9SO-5h ll. 8. Degree, 1951, Physical Chemistry, Dissertations A Magnetochemicel Investigation of the Adsorption of Permegnetio Salts Experience. Q'sduete Research Assistant, Michigan Stste College 1933-5 Huber of American Chemical Society, Society of the 8191 Xi, Phi ‘Ippl Phi, Sign: Pi Signs 111 LMbhub “Minnie I Iflthmctio W. tuneup ”Mun: ABM A stw hes been node of the nature of the surfece of ectiveted shreoel, of the em in which electrolytes ere edeorbed by chercoel, ad of the reeotion et roo- tenpereture between engen end netellio ”PP”. Very little precise infer-stun is ordleble shout the structure of the eel-fees of sotiveted chncoel. A review of the 11th obs“ thet tb dolinent role in determining the snrfece properties of M is pleyed by oxygen chenieorbed on the orbon surfece. By seiteble ehenges in the conditions of satiation of eherooel the nature of the eaten endee en the serfeee chenges ad the edserptien properties of ti. shereeel else chenge. The structure of chercoel eetiveted st mo‘c. and the MM of onions by this shereonl fro- equeons solu- tions see the chief interest of the present investigation. Ash-free chrooels were prepered fro- pure eager end ectinted elder eentrdled conditions. Anelysee for tie elements present were perfernedmdthe morptionofiodinebytheohreeelsseeneesured under stud-’4 conditions. the “reeds were further ehcecterieed by their bilitiee te edeerb eoids ad base. The nonetie suseeptibili- ties of e ”or of shereoels ”tinted under different conditions were deteninedareoIt-pereturendelseetlffl. ndvs‘l. Itves foundpeeeiblebythenepetieneesurenentstodeterninethe-ountof iv muss sass-bed physicslly on the chsrcosl surfece, since loleoulsr cm is Wm. Ohercosl free frm own see prepercd end it nsfsendtohsnoredi-sgnstiethncvgen-conteiningshereeslend its susceptibility see independent of telpersture . Other-stubs! studyingtheneterecfthesurfece csrbcnerides sf chcsosl included direct chelieel emotion of the cherccsl surfeoe with sedi- netqericdste , which ceidsticn yielded crbon diedds. 2b Mes-ed shecrpticn spectre of severel chucesls prepsrsd es nulls were OHM is . stteIpt to chtein inforlstinn shout the structure efssrfecesndes. Studiesseressdeefthsreesmnetrccuteepere- muemoauunmmmemmomm o“. Insetshcfthenshre cfthesdsorhedphsee shenelectrclytes were sdeerhed es ohsrecsl, the osmotic susceptibilities of eelts of iron, sehslt, ad mm sdsorbed on shu-ccsl eere detox-lined. flu mastic susceptibilities of these selts were found to be the one in the edsoehedstetesstheyu-einthecrystellinsfcrs. Magnetism-ermine etlesteqeretures, these seltein tb sdeorbedsteteserefeundtocbsythsOurie-Veisslssfa-psre- septic substance. It res concluded free these results thst the selte hrs hldonthe shsrccsl surfsce hyhonds betvssnnetsl on! own stou, ad s sechsni- for the sdeerption process use postulstsd. “the surfsoe slidetiee cf amulet“! setellic copper see studied et reee W end a 100%. by the netbd of mm “.mbm- ties. It see fend thst the resotien could be encounted for unusually ”We madly-dz: ltd-side: w. 192°I Mm «1m ”Wald", _..-___ ————-——_— by the fernstion cf oupreus oxide, cupric oxide, end by the presence of pivsieelly edserbed oxygen. from independent detersinstions of the tetel .eunt of own sdsorbed end fro- nagnetic neesurenente st room w, 191‘!" end 7ft. the mounts cf eech of the constituents present were eelsslsted. The grss susceptibilities of cuproue oxide "cutie oxides-e else deter-iced etthe three tomes-stuns. TABLE OF CONTMS he. mm: a mn‘gm 6W0...OOOIOOOOCOOCOOOOOIOOOCO0.....0... mwotOOOOOOOOOOO0.0.0.0...00.00000...OOOOOIOOOOOOOO0.0. mm“ BakWOOOOOOOOODOOOOOOOOOOOOOOOO.0.00.00.00.00... mun WM utMOOOOOOOQOOOOOOOOOOOOOOIOOOOCOOOOOOCO mania“ m ”tints-on 0: cmomssesssessssecseseeeees W1.“ room“...OOOOOOOOOOOOOOOOOOCIOOOOOOOOOIOOOOOOOOO mien “atMOOOOOOOOOOOOOOOOOOOOOQOOOOOIOOOOOOOOIOIOOO. “1H”. WOOOOOOOOOOOOOOOO00......0.00.00.00.00... c.t1°n “$6.00....OOOOOOOOOOCOOOOOOOQOOOOOOIOOOIOI.O... man WuOOOOOOOOOOOOOOOOOOOOOOOOOCOOOOOOOOIIOOOOO... uh And”..000.0.0..COO-OOOOOOOOOOOOO.OOOOOOOOOOOOOOOOOOO hummus-1 Tumouu.. ”w.“- md cmm.m000000000000.0.0....0.0.0.0000... lee-respereture Memento.............................. Amman Studiu.oOOOIOOOOOOOOOOOOOOOO.OOOOOOOOOOOOOOOOOOOOOOO “am My“. .‘ awecesseeeeeeeeeeseeeceeeseeeeee Ammtion or 1m...0....C...OOOOCOOOOOOOOOOIOOOOOOOI0.... Adsorption of Acids, Reese, end leutrel 8slte................ Adsorption of Acids end Bases by Cher-cosh Activeted vith M1. or mm.OOCOOOOOODOOCOOOOOOOO0.0.0.0....00...... cum-.1 Studies cf the Adsorption of Psrsenpetic Selts on cmeO0.0000......o...00....OOOOOOOOOOOOOOOOOOOOOQCO.. caplets Che-icel Anelyses of liltrstes................... Enhenge of Pore-spastic Ions for Sodium Ions on Ohercoel. Ravel of Pee-enegnetic Ions free Choreoel................ ”tic ”wai‘h. or m cmmeeeeesseseeeeeseseees lepetie Susceptibilities of Pure Chsrecels st Room rm.m.OOOCOOOOOOOOOOOOO0.0000COIOOOOO0.00.00.000.00. lepstic Susceptibilities of Pure Gherccsls st Lee rmu.‘m..O................C.O...........‘O............ Prepcetion of Dense-free abroad.......................... Studies of the leture of Bur-fees Oxides......................... omen mm.CO....C..................................... M” o: W Ammo M‘cosesceeceseecseseeeeee 1.0“” mm butiomeesseesseeseeeeeseeeeeeseseeseessee 93 Wiueecseeseeessesseceseessecesseeseeseseceseseeeeeeeseece 1m WWWWTBWUMEPWIO M’COOOOOOOOOQOOOOOOIO.OCOOOOOOCIOOOQOOOOOOOOCOOOOOOOIOOOOOO.O 106 WIWMS «mm Mun” Mal Sm- m to: sunset loom: Huh-o] hum: Madam Ishmin 1 Nation Mention We}: hut-mu Inn: on I: “one"... M°NOH| New not: 0 occurs . Continued Pegs A lemma. ROD! C THE SURFACE OXIDATIOI CI COPPER........ 126 Ifiv.h°‘1.nesceseseecseeesceesseeeceseseecesesecseseeeseeesees 126 mu.“ messseeeeseeeseeseeescsssecessesceescscesessece 12" lethods for Determining the mm». of the Oxide File"... 127 “M.OOOCOOOO0.000.000.0000OOOOOOOIIOOOOOOOOOOOOOOO 121 mm MMMOOOOOOOOOOOOOOOCOOOOOOOOOIOOOOOO.0... 12? nmm1OOOIOOOOOOOOOOOOOOOOOOOOOOOOGOOOOOCOOOOOOOOOOO 12? mm. “mmOOQOOOOODOOOOOOOOOOOOOCOOOOOOOOOQOOCO'OO 128 c”““‘ .‘r m ““000...OOOOOOOOOOOOOOOOQ0.0.0.0.00... 1'28 “m- “ mmo‘.’IOOOOOOOOOO.IOOOOOOOOOOOOOOOOOOOOOOO 1” “‘1mun .t h‘mi. OWOOIDOOOOOOOOOOODOOOOOOOOOOOOOIO 132 “new“ 0! Own on Copper WI... 131 locustoche-ietry of Comer-Own Smell................... 1 Wm“ mom.0000000OOOOOOOCOOOOOOOOOOOO...0.0.0.0... 1 M“ “ W'00000000000000000.00.0.OOOOOOOOOOOOOOOOO. 6 ”30.00.000.00.0....OOOOOOOOOOOOOO...OOOOOIOOOOOOOOOOCOOO0.... 15’ MOOIOOOOOOOOOOCDOOOOO00.00.00.000...OOOOCOOOOOOOOOOOIOOOOOO 156 ”WHCOOOOOOOOOOIOOOOOOO.000......OCOOOOOOOOOOOOOI.00.0.00... 1” L13! estates mu 1. We of nemur.. n. rupee-stirs Oredient in the lurrell Furnscsuuununu In. W km 1- th- mm;- MOI-.....mu IV. Det- fer Gained“ of W... V. Chap in Weight of Susceptibility Tube I “hen s nepotio M‘ u ‘M‘dCOOOOOO......OOOOOOOOOOOOOOOOO0.00.0.0... n. on... in mum of lusceptibility Tube 0 when emu. Md ” ‘pm‘dOOQOOO0......000...........OOUOICOOOOOOOO m. Incl-stare W in the lefrigereet lsth............. un. founders tedisnt Men the Susceptibility Tube ad m m huOOOOOOOOOOOOOO000.000.00.000.0000.00.00.00. n. rum-em uni-moat new: the Beluptibility nun“... 1'. Mel Ame-es or Ohm-tie.. n. Iodine Mention II I W of fie-.................. m. Adsorption of Iodine by Chercoel.............u.........o m1, Adsorptien rupee-ties ef ce-srcisl Ohercoeluuuu..." :11. who "overth- ef Met-sod chum-1....umm. ".mtieedicidendlueenchtceslutiweted st 0.000.000.0000...OOOOIOOIOOOCOOOOOOIC00.000.000.00. m. Aterpties of Acid end lose hyOhereoels Astieetsd with m ” “0.000000000000000...0000...00.00.0000... XVII. Wench st Gobelt cum-1e- lclntisn with Mic). cdeOOOOOOOOOOIIO....00......00......OOOOOOOOOOOO... I'm. Inilih'stiee cf chelt cum-u. Solution eith Specielly M11“ GMOOOOOOOOOOCOCCOOOOOOO.....COOOOOOOOO.... Peat 18 19 20 32 35 36 kl ha ha us h? 1.9 51 53 SS 58 60 62 m Cfmocsntinned Page 211. Eleilibretisn ef Oebelt chloride mad lichel Chloride Solu- tias with Low-leqersture Aotivsted Ohsrccsl.............. 63 11. Elk... of Per-ensue lane for sodium lose on Chen-cosh. 6h m. logistic Msptibilities of Pure Cherccsls st Boon '“ramsceeseesseceeeeceeeeeesceeseeseeseesseeeeseeeses 6? n11. lunatic Susceptibilitise of Pure Ghee-cede es Nations of ’WCOOCOOOOOOOOCOOOOOOO.....OOOOOOOOOOOOO00......O. 69 mn. Peroentege cf Koleculsr omen Adserbed .‘.’ Ohsrecel........ 80 m1. Oxygen flycicnlly and Ohmicelly ideas-bed on Ohsrcoel (due) 81 m. mpmmty of Ongen-Iree Ghersoel as nu. Wise cf the mm Absorption of own-......" 9h mm“.- W cup-tum of am one» me me choc-seals .00....OOOOOOCOOOOOOI.........OOOOOOOO......OOCOOO 97 nun. “cry of the hcpsrtiee of levers]. Chen-eons............. lOl nn. Or- Susceptibilitiee sf Obereeel and of Iron (:1) Sulfste W .n emu-0000000000.00.000.0000.0000000000000000 no nx. Q'- Dusseptibility of own end of Iron (In) lelfste W .‘ cm.0.0.0.........C...’.......O.......C... m nn. ur- em-pmmty of Oobelt menu. We on Ohsrccel 111 nm. on. Inceptibility cf Ilengneee (:1) cm». Adscrbed on OMOOOOOOOOOO.00.0.0.0.........OOOQOOOOOOCOOOOOOOO.... n6 nun. le'etis Susceptibilities of Adscrbed Iron 8e1ts........... 111 nm. “Neoptihilitycfcwrieonde esehnctisnst ,m“m0000000000.....OICOOQOCOOOOOOCOOO.......OCOOOOO. 137 m.h3useeptibilityefccpper esehnticncffield W a 'WOOCOOCOC.OCO......OOCOCOOOOOOOOOOOC ”‘6 mu. lepetie SusceptibilityefOupreusOndese etunctionef mm d ’mmeeseesecseeeeeceeeeeeesseesse “1 nannies-c mm. lanai: of Field mm. lunatic use as rmns . Continued Page muz. llspetie Incesptibility of Oupric Oxide es s Function “ “.14 Wk ”4 r.m.r‘bm'.eeeeseesseeecseeeeeeees 1118 mun. Hemetic Susceptibilities of lertielly Oridised Copper eelCOCOOOseesOeeesseeoesescseseeeeeseeoeeesseeess 150 1. Meter 2. in: fig} 3. meta-ice i. Mum: 5. tutu A been ) hue ‘0'.“ re chm '0 ~- MC for choc a. a... m “1’ Ohm ’c t. m in- the. 1.13: cm FICIIRES Hem Peso 1. ”MCCOCOOOOOOOOOOOOOOsssssseQOOIestOOOOIIOOOOOOO0.00. 11 t. m mm bm...000OOOOOOOOOIOOOOOOO0.00.0000...0.0.. 29 3. Electricel control circuit for oped-sting the electronemst” 30 3. cahmn 0‘ .1”WOOOOOOOOOCOOOOOOOOOOO.0.0.0.0... 33 5. retus fer hepatic susceptime deter-insticns........ 37 A Assenbly for lowotelpereture mastic neesurcnents Mocptibility tube 6 6. Q'- susceptibility es s function of reciprocal tenpersture :” .mn casesseeseesseocsees—seesscessececseseeseeeeece 71 1. kn susceptibility es s function of reciprocel masters tor om c‘loseceeeseeeeessecssseceesesseseceseeeeeseees 72 8. dr- susceptibility es s function of reciprocel tempereture ‘" .mn 0.11....OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO... 73~ 9. ar- suseeptibility es s function of reciprccel tenpersture t" .M 6.12.0.0.........OOIOOOOO...OOIOOOOIOOOOOOOOOOO 7h 10. Gr- sesseptibility es s function of reciprocsl temperature m Ohm-1 12' 2m. mOIUOOOOOOOOOOOOOOO00.000.000.00... 7s 11. be. susceptibility es s fmotion of reciprocel tenpereture ‘” .m& 13’ "u. mOOOOOOO0.0...00.000000000000000... 16 1!. b. seseeptibility es s function of reeiprcesl meters I” own 1h. rm. mOOIOOOOOOOOOOOOOOOOOCOOOOOOOOOOOOO 77 13. Apperetus for the properetion of omen-free chercosl....... 8h 11:. left-seed absent-tee band- el ehu'eoel 0-12.................. 92 15. leeipueel susceptibility es s function of tenpererture for ”m‘ (11) cm“ W on cwMQOOOOOOOOOOOOOOOOOO 11, 16. Reciprccel susceptibility es s fmtion cf tenpereture fer scheme» (11) sulfate adsorbed on ehmoel................. 115 HIGHW- xii-dined the (III) in. help-0011 in (In) Marconi 1m (III) 10. burned in (n) n. Minced in (II) ”0 VIII If" 1.13! or HWRR . continued Page 11. Reciprocsl susceptibility es s function of tempereture for iron (In) sulfste, snupls A, edsorbed on chsrcosl......... 1.19 18. leciproeel susceptibility es s function of tempersture for iron (m) sulfate, couple 3, adsorbed on charcoel......... 120 19. Reciproeel susceptibility es o function of tenporeturo for iron (:11) chloride sdsorbed on chercoel 121 . Recipreoel susceptibility es s function of teqursture for ire: (II) chloride, eqle D, sdsorbed on cherccel......... 122 21 e1 susceptibility es s function of tempersture for . Beoipree iron (II) chloride, euple E, edcorbed on cheroosl......... 123 22. 'm W“OOQOOOOOOOOOOCOOOOCOOIOOOOOOOOOOOOOOOOOOOQ0.0. 139 Then 1! pn is mind so I Wu. 1h 1m theories in animated When SURFACE OI" ACTIVATED CHARCOALS Introduction There is probably no aspect of the adsorption by charcoal that has received so much attention as that dealing with the adsorption of electrolytes. liany points of View have been expressed and new difo forent theories have been advanced, but as yet there is no canprehen- sire explanation to account for the behavior of charcoal toward electro- lytes. This is no doubt due to the extreme complezdty cf the charcoal ‘ surface and to the vastly diverse properties it imports to the charcoal. ls funduentel study of electrolyte adsorption by charcoal can be pursued sithout searching into the nature of the surface of the char- coal at hand. And the study of the surface of charcoal is necessarily no less complicated then the study of adsorption itself. It is generally agreed that the dominant role in determining the surface properties of charcoal is played by oxygen chomiocrbed on the carbon surface. It is amazing that although it was postulated alrescb in 1913 that oxygen and carbon at the charcoal surface reacted to fem surface complexes ,‘ very little precise information is available today about tb structures or physical properties of these oxides , even though these have been the subject of new investigations. Il'here is general moment that there are two distinctly different types of charcoals, one of uhich is formed by activation in air or ongen at n- moan h00°6., and the other at higher temperatures, around 800°C. The low- tupereture ohucosl adsorbs prinarily bases from aqueous solution and the high-temperature charcoal adsorbs primarily acids . The purpose of the present investigation, as contained in Parts I and II of this paper, “was to amino by several methods the surface properties of charcoals activated at approximately hoo°c . uith particu- lar euphesis upon the adsorption of inorganic cations by these oharcoels and the nature of the adsorbed phase. The methods used to investigate the nature of the charcoal surface were: the extent of adsorption of acid and base, the nagnetio susceptibilities of pure charcoelo, direct chenicel oxidation of the charcoal surface, infrared spectra of the charcoals, end isotope exchange studies. The nature of the adsorbed phase ad of the forces which bind the cations to the charcoal surface tore studied by determining the nagnetic susceptibilities of adsorbed per-agnotio cations . Historical Beckgound Ohn'coel has often been seemed to be completely max-phone carbon, but that this is not completely true has been shown by way investi- gstions by a umber of wrkers."‘ The way patterns are diffuse and shes a graphite-like structure corresponding to a two-dimensional goph- its lattice. nae has been confirmed by electron diffraction studies .‘ Theo charcoal can be said to be amorphous because there exists no thres- dinensionsl network, but it must be borne in mind that charcoal is not entirely devoid of all pattern or regular arrangement of atus. .2. Thou-ban: l mum” m. 80 It at Mm! at til 1H! this! mi} “mm! The) Walt to 391. M charco' MM 11th Iir Man of c Wis, Wing tin-m1, Pea: the carbon stalls in charcoal are held together by very strong onion bonds, as witnessed by its extremely high decomposition temper- stirs. So far I these bonds are strong, the bonds or free valencies ‘left over“ at the edge or surface of charcoal must also be strong, and it is these residual forces that are responsible for the chenisorption . These forces at the surface are so strong that it is very of engen. difficult to get a charcoal surface entirely free from oxygen. lost charcoals used in adsorption experiments have been in some contact Iith air. It is know from several investigations that slow cabustion of carbon takes place even at room temperature: In addition to this, timing the process of activation to impart adsorptive preperties to charcoal, reaction sith oxygen is inevitable. The importance of the existence of surface oxides in the adsorption of electrolytes on char- seal use first mentioned by Schilow and reomutovf” These sorkers later. postulated the existence of two basic surface oxides and one acidic oxide in order to account for the adsorption of electrolytes. the theory of Bchilos and coosorkers will be discussed later. It should be remarked that vith the present incomplete state of Was of the nature of the surface of charcoal, characterisation of ' a given charcoal can only be node by giving a detailed account of the aethod used for its aanufactm'e, the source of the raw natorial, and the results of adsorption experiments. Untold confusion has resulted fru reports of the properties of very poorly defined, and often very impure oharcosls. Er); mos m, chucoal ah. om; to t W m7 co man "do it em til use than in t1 shed lith acid W931 antherL k W 01 the} “I ifl‘ned 1 ml “mu. Althe- Early investigators of electrolyte adsorption on charcoal used chiefly charcoal of animal origin containing from five to fifty percent ash. Owing to the use of such contaminated products the investigators obtained very contradictory results. Using charcoals with high ash content made it impossible to ascertain with any degree of definiteness whether the adsorption of ions depended on the carbon or on the inorganic impurities in the material. Even though the charcoals were very often sashed sith acid much of the ash could not be removed, and it was often doubtful shether all the acid had been washed off. A major advance in the study of charcoal adsorption was made by Bartell and liner" when they devised a nethodto prepare ash-free charcoal from impure starting laterial. Although this method of purification is not sidely used today became charcoals can be made more easily fran pure starting material, their sork first showed the supreme importance of using ash-free char- coals in studies of electrolyte adsorption. Killer and codmrkers were the first to characterise the adsorption of electrolytes on pure charcoals activated at about 9oo°c. They found that, of inorganic electrolytes, these chm-coals adsorbed acids and all anions except the hydroxyl ion, but did not adsorb alkali hydronides nor alkali metal ions. Inorganic electrolytes sore adsorbed hydrolytically, that is, only one of the ions was adsorbed. In 1929 Kruyt and de Kadtn discovered that charcoals activated at a considerably lower teapot-stun than that used by Miller had practically the reverse properties. Alkali ludrosidss and inorganic cations were adsorbed, acids and anions were at curbed, 1:. Hit II mm he theori. mt! discussed 1111s I the re a M taper-It in: Is enth- Wde In all “1" War We of th W; 3150 the mt Idlorbed, end in generel the hydrolytic adsorption of inorganic eelte wee revereed in eign. Sue theoriee of the edeorption of electrolytee by charcoel will not be diecneeed. The eerlioet theory of eignitioence was offered by Killer ee the remit of hie etudiee or edeorption on cherooele ectiveted et high tapereturee."1u He coneidered that the edeorption of electro«- lytee wee entirely electrolytic, that, for example, e solution of eodim chloride ice elightly lydrolyzed end the chloride ion adsorbed cc mole- culer hydrochloric ecid. Evidence for this was the fact that upon edecrption or the chloride tro- eodiun chloride the eolution became less ecidic) eleo that smell emounte of edeorbed eoid were not eble to invert eeee edged- et 60°c. The greeteet difficulty with thie theory liee in the interpretetion of cone of hiller'e own experimentel work; he obeemd that fro arsenic codin- eelte both cation end enion were edeorbed, thereee he hed eleo ehovn thet codin- hydroxide one never edecrbed by hie chercoele. But undoubtedly be we correct in eteting thet cation end union ere not elweye adsorbed to the cane extent. Soon etter the tint report on the prepertiee of ohercoele ectiveted et lee tenperetme, Buretein end Imam“ deecribed experimente with cherccele outgeeeed et lOOO°C. on which no adsorption of electrolytee could be cheerved. Schiller end Tact-Inter. round thet the edecrption or ecid depended on the ougen procure in the eyeten, end thet below two -. preeeure the edeerpticn of ecid repidly deoreeeed. The letter enthore watered the expleneticn thet oxidee were forced cn the eurtece ct the ehereoel in the preeenee of oxygen. The edeorption of hydrochloric eoid -s- ‘ Mmted M of fin char when 0! en mention of h; Mme at an mod. These 0: / --A A..- ‘o a. / \ —& wee interpreted u e neutralization of the elkeline oxidee on the cur- tece or the chercoel. In e later peper Schilow 31 _e_l_._.° eeeumed the exieteme of en ecid oerhon oxide which wee euppoeed to bring ebcut edecrption ct hydroxidee on lov-tempereture chercoele. The poeeible exietenoe of en inner eelt of these two kinds of cxidee wee eleo postu- leted. Theee oxidee ere repreeented as follow" J}... .2... .9... .2. I , 40 -c\ -c\ -c-o -c\ 40 ,0 [o ,o -c\ -c -c\ -c-o -c\\ ,0 1 lo 0 -c\ -c\ -c\ -C-0 ,0 ,o ,0 40 -c\\ -c -c\ -c-o -c\ o I ,o -c *0 Oxidee A end B would be expected to be elkeline end 0 ecidio) D ie I type of inner oxide which would be mphoterio. The exietonce of then eurtece oxidel ie etill hypothetical. A more complete theory regu'ding the edeorption of electrolytes on ell types of chercoele wee leter preeented by mm end do xedt” along the eeme linee ee thet or Ichilcv end hie co-workere . Although the oxide theory hee net with objectione tron verioue euertere, it hee obteined generel eupport during recent yeere. Perhepe the chief obetecle to ite eccoptence ie the hypotheeie regarding the enietence or elkeline cerbon czidee, eince euhetencee of e correspond- in; neture end etehility ere not known in orgenic chenietry. however, -6- the oxide theory scene very satisfactory in its broad outlines for chu-ccels ectiveted st low tampereturoe . I’m-kin end hie coworkers“ used the classical electrochemical theory es s eterting point end seemed thet chercoele suspended in solutions of electrolytes behaved ee reversible gee electrodes. The cherge on the ehercoel wee thought to depend upon the concentration of the ions in solution and the nature end pressure of the gas which was in equilibrium with the system. A charcoal in content tith hydrogen use believed to be negetively charged due to ionizetion of edeorbed hydrogen etcue. The nsgetive charge on the oerbon em‘fece wee seemed to ettrset ceticne from the electrolyte, thus ceusing edeorpticn. The presence of oqgen on the chercoel- wee believed to give e positive cher- eeel due to icnieeticn of lvdroxyl ions. This theory hes been subject to new criticisms end res modified in nearly every paper published by Frankie end hie co-eorkere.“'" Gteenberg" eede e very extensive investigation of the edeorption of electrolytes, chiefly on high-temperature charcoal. Be eede the interesting discovery thet the edeorption of strong electrolytes on mun-meters chccosl could be inhibited by the presence of organic non-electrolytes, end thet electrolytes edscrbed in the sheence of non- electrolytes could be removed Iran the chercoel by edding strongly Idlerbed non-electrolytes. For lov-tenpereture chercoel. the presence of ensue Inn-electrolytes hed no effect whatever on the edeorption e: electrolytes, end it see not possible to desorb the electrolytes from 10"‘t-finture chercoel. This he considered es evidence in fever of '4’ u oxide 1.}. chested t1; cationic tbs In: the flotation 2 be the at ton on higl anion: ere '7 mar; ti 1'1“; Inio: W that 1 '9 "a day We the Dc Wilmer: "med | “perm” the oxide theory of edsorpticn for love-temperature chercoele , end he suggested thet edeorpticn on these chercoele could be interpreted es s cetienio chenge. [for the high-temperature chercoele Steenberg concluded that the edecrption forces were the same for electrolytes ee for non-electrolytes . Ira this etrting point he developed e theory of electrolyte edeorp- ticn on highotmpereture chercoele based on the eesimption thet ell ceticne emept hydrogen on these chercoele were cepeble only of second- ery edeorpticn, thet is , that they existed only in the electricel double 1W! enione, on the other bend, were prinerily edsorbed. He postu- leted thet the strongest edsorptive forces on high-tapereture chercoele were ven der Weele' forces, end demcnetreted thet it was possible to epply the Boom theory to e solution of electrolytes in content with ugh-tempereture chercoel. 0n the basis of the Donnen theory Steenberg presented e theory of hydrolytio edsorption of electrolytes on high- tuperetnre chercoel. Since low-tempcreture chm-cosle were the main interest in the present investigation this theory will not be discussed. A eignificent espect of Steenberlg'e work is the feet thet his theory of the edsorption of electrolytes does not require ”V cosmiptione es to the netnre of the surfece of the chercoel nor es to the nsture of the edserption forces. Vileon end Belle“ heve nede important studies on the reletion of envenomed oxygen te the edeorpticn of ecid end beee. They found that Ifluidised outgeeeed chercoel edsorbed apprecisble mounts of said but no bees, but thet efter oxidation et hoo°c. the edsorption of base was -8- 2" my naked mall the: mud the hymen u; adsorbed ox; IN, I I; boa rem till 0! or nluibu M “Iggy; very marked end that of acid was reduced to very small preportions. - Thee-eel treatment of the charcoal in a vacuum or under nitrogen deo creased the adsorption of base and increased that of acid to an extent depen‘ient upon the tanperetm-e and the time of heating. The amount of adsorbed oxygen removed from: the charcoal by this treatment was deter- lined, as was the decrease in adsorptive capacity of the charcoal. Upon reeoval of the first part of the adsorbed oxygen the loss of two state of oxygen corresponded to a decrease in adsorption of one mole- cule of base. This led the authors to conclude that tee types of euro face oxygen couple: were involved in the sorption of base, the ratio of etus of oxygen to molecules of base being two to one in the one case and four to one in the other. Since they found that the adsorption of iodine in very sinner to the adsorption of acid, the acid was prob- ably physically adsorbed, and ovgen chanisorbed on charcoal did not fore basic complexes instrumental in adsorbing the acid. The studies of King have been very revealing as to the points of difference and similarity between lee-temperature and high-taspereture aha-coals.” Iron studies of chercoele activated at a series of difa ferent temperatures King concluded that chercoele activated in the region 800 to 850°C. had the following properties: eaxhun acid ad- sorption, maxim:- phyeicel adsorption, minimum base adsorption, maximal pl of aqueous suspension, medium electrophoretic mobility of particles lespended in water, sexism: efficiency as oxidizing catalyst, and linen- efficiency in the decomposition of hydrogen peroxide. Charcoels minted :1 s ml: to "period 1: Mull I flag on ab] potion of c mention.I N In sq. mm. Ch decode I: Moth a Mod; 11 Semi hum of 1' hm a Wh' 1 activated at hOOOG. have in nest cases the reverse prepertiee. King see able to prepare only chercoele uhich sore negatively charged when suspended in eater, although some authors have claimed that positive shl'coals sore formed by ionisation of )wdron'l ions from the chrooalfihneu King was able to find definite correlations between the pH of aqueous sus- pension of chercoele and both temperature of activation and acid and base adsorption.“ Beteeen the activation temperatures of 200 and 800°C . the pH of an «new suspension rose linearly with the temperature of aeti- vatien. Charcoals activated at hoo°c. gave a pl! of about five, and chercoele activated at 800°C. gave a pH slightlygreater thm eight. Charcoals inch an ice pH values were best able to adsorb base, and chercoele mioh gave high pl! values vere best able to adsorb acid. leveral studies of adsorption and the adsorbed state have been made by new of nagnetic susceptibility measurements, but few of these stud- ies have dealt Iith the adsorption of electrolytes. Bhatnagar and co. worm“ investigated the adsorption of salts of iron, nickel, cobalt, and manganese on abused and silica gel. They reported that these salts, all noreally par-nagnetic, became diuagnetic when adsorbed on charcoal, and in fact the charcoal itself appeared to become more die-ensue. with silica gel the susceptibilities were not far from additive , that is , the salts had the seas susceptibility in the adsorbed state as in solution. A recent study by [oboeev and coworkers“ on the paruagnetisn Of edsorbed layers is very interesting. They observed that ehen the Montage of the surface covered by adsorbed salts was decreased the -10. planet. e um: deride h mm 1 the arm $.12! the j ta of th. e cine-co: 0! near-be 0.66:! the ' I M of MI ga pea-napeti-issreasedtylcp-mts,ccrrespeedin¢ineeweases tenanstieeueetsefteuefsehaagnetens. whenecbalt(m shleridehemeteeaeedsestedeusilieegel,theeapeti-efthe Winn-usesad‘tethstefthepureseltfiaenonepereentof the surface was severed; tea the surface coverage area decreased to oMthepr-epeti-eftheadscrbedaaltieereasedtoSJtieea thatoftheperesalt. l‘er nee-1m) nitrate Wm adsorbed eaehareealatsu-faeeeevereaesefaeleastlélthssueeeptibilitiee efadeerbedaedpereaalteceecusl,butabentheserfaeeeevera¢eeas 0.0“themtiaseftheaeseebeducheleaslfl3tieeeaeneet esthstefthepereeelt. Thesaeeeptibilitwefplatis-aieerbeden uninspleesfeeeitobeJGpOOtheeasp-eetasthatefbulk pile“. Ireeaesenedeeehareealusperuaastieretherthan Wemuatdnutelqersefirosesacarrierareatceie, aeteryetelline,ianetere. m'aaa-piaaeaaaaaraa-uaaaequunme dthdiletieevasehareeteristieefahesbedlqu-saeieaeastfeund hmheeehuiodaiatwesefthesuesebstaeees. Imamtic eraledoetbythefastthatseeeeptihilitiesserein- mean-lemma. ‘ mummwmtmmoum mallards-ensue silver,ee1ythsfreeateeefi:ichispu-n.apetic. Thenseeptibiliayefsilverdispersedenaniaeetacpperteasnegetive (ii-anstic)eelyeerelative1ydeeselayers,batbee-einereasis¢ly Mtiveuthiaoreasiudilutdee. Milieeentrasttethefindings a! Soleod a wwwu dim cont: a large m (fort by t hithl. m he m We p ”I term “We dc “Min of M 5| regs " ”PMs-i am. he Sela»: efdeluedaaibsllasnebtriedtodetectpar-agneticsilverand eeppwbydispsrsiagtheeenetalsenaluinegel. Aeupleefsepported silveresntsimn‘usilmeaaexninsdandthesithorerepcrtedthat no large treatise ef the silver beans para-auntie. Very eareml effertsbyfieseautherstodetsetpr-egneticcoppereerealsenet fruitful. The uni-a investigators” concluded ma their studies that the verylargeparuaansfi-ofdilutslayereintheadscrbedetetseesdus toadsereessefthsrndoueesofspin,thstis,eriestationofthe aagaatieateaaaoatbaaarudiy enhanced. Theycellsd this ans-aim- Wefdiluteaumsdlayers'supsrpernagnetiu'. Thisscrk leatheremdedntheeeereservetienuntilcsnfiraingreesmhes new,aineethserrersefeeasureeentarelargemnuall “snatched. filmed“ has carried out extensive investigauens sf pumpetic salts dispersed so inert supports each as nemesis», rutile, and sluice. Thessstedieshauledtoceulusionsefpeatintsrestbothfruths pedetefviesef satanytic reastieesandeftlestreetureefthes-aup- pertedsalt. .Ithasheenpesdhlsteeelseletsthseemticeeuentof mwmapetioieaiediunimaw,udmotoam_ estieetesofthestatsefamgetionandthemberefaecnstieneighé bersefeaehetu. Ioetarie and Ierthier”!” investigated the eagnstic properties of mwumormnu k mueuzozez mutiny? was» ‘ >P_J_O_PmU83w W UF.¢U_¢O o \s w d» 83% m noun 9: .oooou . \\\ \\\ _l.-.|\r.|_lel won... 20.92.5500 mtmmEo ulna O... I. -8h- had been reached. The systee was sealed off atpoint A with a hand torch. The combustion tube was removed.frcn the furnace, and‘by Judioious tipping of the combustion tube the charcoal was brought over to the susceptibility tube. ‘Uhsn sufficient charcoal had accumulated in the susceptibility tube, the latter was sealed off from the rest of the apparatus at point 8. than, in a helium at-oaphere, the sale standard taper Joint was quickly removed and replaced‘by the plug or stopper of the susceptibility tube. The values obtained for the sag- nstie susceptibility of this charcoal at several magnetic field strengths and three temperatures are recorded inflrable 117. There appears to be a slight trace of iron present, as seen by the mall drift of susceptibility values frat low to higher field strengths. this sight account for the very slightly lower susceptibility values found.et low temperatures. TABLE»XIV SIS‘JEPIIBILII! G onset-ram CHARCOAL W Tempdeieture Field Btrength 2 92's H __ 17-6 Oars ted- ulcep b ty w I x 10‘ L780 4.31. 4.25 4.28 7.360 4.36 -1 .33 -1 .3 9 ,060 -l .3? ol .3 6 cl . 10,1870 .1 e38 "'1 e37 .1 e32 Average 4-1 .36 -1 .31: -1 .32 -85- ‘fheee susceptibility values are appreciably more negative than values obtained for any chercoele, activated or unactivated, which con- tained ongon. Corriez“ reported that sugar charcoals heat treated at 1200°c. hsd susceptibilities of -o.91 and -o.98 x 10". When still higher temperatures rere need, the susceptibilities becme increasingly sore diuagnetic, approaching but not attaining to the value of graphite (41.1 x 10"). l—rq patterne of these charcoals showed beamings of a graphite-like structure at high temperatures. Heyer“ reported that the gran susceptibility of gas carbon was ol.h x 10.. at rooe temperature. This neterial was probably an inactive e fore of morphous carbon containing very little oxygen, and is probably caparable to charcoal obtained by removing oxygen. Studies of the lsture of surface Oxides Chemical Methods issuing that the surface of activated chercoele contains appreci- able uounte of carbon oxides, it eight be possible by suitable oxidation to relieve sons of these oxides from the surface, obtaining in aqueous ' solution lee soleeular weight cupounds of carbon, hydrogen, and cam. It was reported by nag" that mp1;- washing certain activated chercoele with ester moves from the ohmoal surface a substance identified as oxalic acid. It has been observed in the course of the present investi-- gation that certain ash-free charcoals ehen washed with 2 l sulfuric or whence-1o acid produced a substance ehich could be slowly oxidised by eerie uonim sulfate solution. -86- A suitable oxidant for organic oxidations of this type is sodim netaperiodate . The great advantage to the study of periodate oxidation of organic matter consists in its ability to promote a specific type of noleoular cleavage. The cleavage of organic molecules by periodic acid eddation involves the breaking of carbon to carbon linkages of carbon atas vhich carry Murcsyl or carborvl substituent groups. Thus in the ease of the oxidation of a silple compound such as ettvlene glycol the periodate supplies one steel of oxygen in being reduced to the iodate and the urban to carbon bond is broken with the degradation of ethylene glyeel to tea eclecules of forsaldehyde. On the other hand, periodate under the ace eonditione is without action on such an equally simple orgmie soleouls as ethyl alcohol. In general, periodate oxidizes 1,2 glyeols, ac ohydroxy aldehydcs, cc ahydroq ketcnes, 1,2 diketones, and ac owdrosy acids. Olefins, secondary alcohols, 1,3 glycola, batches, and aldehydes are not affected. Became the reactions are not instane tsneous and because degradation products such as fonaldehyde and formic acid are reducing in character they are studied at room temperature or lover in some cases ." The study of the oxidation of charcoal edth periodate eas carried out as follow. A quantity of charcoal was activated in air at hoo°c., and 211.8 g. of this was nixed with 100 ll. of 0.1 H sodiue netaperiodate. ‘l‘hs mixture eas stirred vigorously for eight hours after wish it res filtered through a sintered glass filter. the filtrate eae found to have a pH of four. One hour after filtration the periodate eas destroyed by .07- the addition of 10 g. of sodium sulfite and the solution was diluted to 200 I1. A test for formaldemrde in the filtrate was carried out according to the following metbd suggested by Professor J. C. Speck." Ten ml. aliquots of the filtrate were transferred with a pipette to separate volumetric flasks equipped with glass stoppers. To each was added 0.5 ll. of 10% chronotropic acid and five ml. of 1h 11 sulfuric acid. The flasks were put into a boiling water bath, the stoppers inserted, and the heating continued for 30 minutes. Then the flasks were cooled to 25°C. and the solutions diluted to 50 ml. with distilled water. Excess sulfur dioxide was rmoved by bubbling air saturated with water vapor through the solution for 15 minutes at the rate of one liter per minute. The color of the solution at this point was a pale yellow rather than the cherry-red color indicative of the presence of foualdehyde. A blank was run using the sons procedure, and it had the same color as the filtrate solution. This test is sensitive to about one nicrogrm of formaldehyde per nl. , and apparently formaldehyde was not present above this concentration. The presence of formic acid in the filtrate was tested for by a sethod similar to that used for formaldehyde .'° Three drops of the filtrate solution was nixed with three drops of 2 N tydrochloric acid and nagnesimn powder was added until there was no further evolution of gas. At this point the liquid was poured off, and to it were added three ml. of 121 sulfuric acid and a little chromotmpic acid. This was heated for 10 ninutes at 60°C. Again there was no positive test -88- but only a pale yellow solution. The limit of identification of this test is one microgr- of formic acid per ml. To five al. of the filtrate from the periodate caddation reaction was added enough hydrochloric acid to give a pH of about two. Then bariul chloride solution was added and a voluminous precipitate of bariu- sulfate was obtained. The precipitate was filtered, the fil- trate nade alkaline (pH about 12) with sodim hydroxide and more barium chloride was added. A white precipitate was formed which dissolved to give a clear solution when the solution was sade acidic, and re-precipi- tated when the solution was again eade alkaline. The precipitate was filtered, washed once with distilled water, and dissolved in dilute perchloric acid. One drOp of 0.005 M potassim permanganate solution was added to this solution and the purple color persisted. When this solution was heated nearly to boiling the color disappeared, but upon the addition of another drop of the permanganate the purple color per- sisted. Thus there was apparently no oxalate present and the precipitate was sost likely carbonate . A spectrOphotonetric stuw was also made of the reaction products of the oxidation of charcoal by periodate. About 1.2 g. of sodiun setaperiedate was dissolved in 50 al. of distilled water, and 10 al. of this was set aside as s blank. The other to ml. of the approximately 0.1 H periodate solution was aimed with 8.1;2 g. of charcoal which had been activated at hOOOC. in air. To a 1.37 g. sample of the same cher- coal was added 50 :1. of distilled water. These mixtures were each stirred continuously for eight hours, then filtered, and the filtrates .89- saved. The spectra were examined with a spectrophotometer* using match- ed Cores (glass) cells with a. one on. path length. The percentage transmission of the periodate-charcoal filtrate was measured against the pure periodate as a reference solution, and the water filtrate was eeasured against pm'e distilled water. The wave length range 800 to 320 nilliniorcns was covered, but there was no absorption from either solution in this range. This indicates that no organic molecules of an complexity were removed from the charcoal by the periodate treat. sent. However small solecules like onlic acid and pyruvio acid absorb at shorter wave lengths than those used and ttnrefore could not be detected in this measurement. , In conclusion, under the conditions used in this stuck there was found in evidence for the presence of formaldelwde, formic acid, oxalic acid, or of solecules which have absorption bands in the region 800 to 320 nillinicrons. However, evidence was found for the presence of car- bonate. It is poesiblo that the periodate oxidised some of the surface asteriel of the charcoal all the way to carbon dioxide. Studies of Infrared Absorption Spectra An attempt was sade to obtain infrared absorption spectra of several charcoal samples in the hope that such spectra, if obtained, would pro- duce eons infornation about the nature of surface oxides. The samples 'Bsoknsn Instrument 00., South Pasadena, Cal., Model DU. -90- were prepared for infrared examination by grinding them in an agate mortar with a suitable dispersing agent and then pressing the null obtained between sodium chloride windows. The dispersing agents were Fluorolube 5* and 311301;“ all of the chercoele were examined in Fluore- lube and some were examined in Bujol also. The spectra were obtained between two and seven microns for Fluorolube and between two and four- teen microns for NuJol. The absorption spectra of both Fluorolube and me). have been published by Hiller and Wilkins.” The instrument used was a Perkin-miner recording infrared spectrome- terf" The compensating been of the spectrometer contained empty sodium chloride windows similar to the windows used to contain the samples. One of thal’luorolube mulls was exanimd with a microccOpe having a calibrated eyepiece, and it was found that most of the charcoal particles were one to two microns in diameter. The chmods examined were chiefly those for which complete ele- mental analyses had been nade (see Table I). In Table XXVI are given the descriptions of these chercoele and their methods of preparation for infrared analysis. One seaple of charcoal 0-12 was prepared by mulling the charcoal in carbon tetrachloride .m The null was spread out on a sodiu- chloride window and winn the carbon tetrachloride evaporated “enam- mectrocheaical 00., liagara Falls, n.1,, Lot number h-lé-Sl. ”a. a. Squibb and Sons, New Iork, n. I. “Perkin-Elmer Carp., wax—walk, Conn., Model 21. ““Hsrck and 00., Reagent grade. -91- .wauo Hmoonmeo no means noapmnomnw vmacnueH .JH ennMfim AnGOQOfiEV newsma m>e3 05 mg 0;. me be m.m em 3 q _ _ q q _ ll om noteetwsueag queoaeg a fine dry powder remained which was examined between sodium chloride window. The only charcoal samples whose spectra showed absorption bands that could not be assigned to Fluorolube or Nujol were the samples or C-lZ. This charcoal both in the Fluorolube null and as the dry powder showed two absorption bands adjacent to each other at 5.6 to 6.0 microns and 6.0 to 6.5 microns. These are shown in Figure 114. The lower wave- length band is in the region of carbonyl absorption and the other band lay be due to ethylenic carbon-to-carbon bonds or to adsorbed water.“ These absorption bands are too broad to make fruitful any attempt to ascertain more precisely what kind of bonds might cause the absorption. Possibly the widths of the bands are indications that the absorbing groups are in quite different enviromnents. Isotope Whange Reactions It was thought that some information about the nature of the sur- face carbon oxides might be gained by equilibrating certain types of chercoele with water containing 0“ and staining the chercoele for the presence of the heavy isotOpe of oxygen. There might be an exchange reaction bet‘deen oxygen in the water and oxygen adsorbed on the surface of the charcoal, or possibly water might react with the carbon of the charcoal to produce surface oxide. The nethod of study was to equilibrate a known weight of charcoal with a Imam weight of the water, then to dry the charcoal and evacuate -93- TABLE nu W10! 0? THE INFRARED ABSORPTIOH Cl CHARCOAIS Charcoal lumber Col c-e 0-6 0-7 0-9 (3-12 Description of Charcoal Activated in ammonia, con- tained 0.711 nitrOgen, 3.10‘ Garcon. Activated in monia, contained 0.6% nitrogen, b.05i engen. Activated in ammonia , contained 2.17% nitrogen, 0.681 oxygen. Activated in sulfur contained 2.931 sulfur, 1.55fi oxygen Unactivated charcoal, contained 8.095 oxygen. Activated in air at hocflc., contdned 25.11% oxygen. Ongen-free charcoal. Obucoal with adsorbed Pe.(50‘) . Dispersing Medium Plucrolube Fluorolube lujol Fluorclube In: ol fluorolube Fluorolube Fluorelube , dry powder Fluorolube Fluorolube -9u- it at a high temperature to collect carbon monoxide and carbon dioxide. The carbon caddes were collected in Pyrex sample bottles and sent out for mass spectroscopic analysis 3‘ About 0.1 to 1.2 g. of charcoal was mined with one or two grams of water containing 1.5% 3.0“.“ For a reaction vessel the bottom end of a Pyrex test tube, with a volume of about three ml. was used. The char- coal was weiglmd into this vessel, the water added and the vessel weighed again. a small cork was used for a stopper and the charcoal and water were antated to mix them thoroughly. After an equilibration time of eight to sixteen hours the water was poured off and the charcoal was dried at 120°C. for one hour. It has been found that drying a char- coal at 115°C. for as long as three hours produces no change in the “ Thedry charcoal was put into a platinum amount of surface complexes. boat and evacuated at 200%. to a pressure of about 5 x 10“ an. The stopcock nearest the vacuum p‘tlllp was then shut off (see Figure 22) and the teapcrature of the charcoal was raised to iooo°c. and maintained at that tequsrature for five to eight hours. The platinum boat and com- busticn we. had been balced out previously at iooo°c . for five hours. When both carbon dioxide and carbon monoxide were collected from the same sample of charcoal, the method of collecting as samples was as fellows. Two gas sample bottles were attached to the manifold of ‘Consclidated mac-in; Corp., rumba, on. “Start Ongen Co., San Francisco, Cal., by pcmission of the Iso- topes Division, 0. 8. Atomic Energy Commission. -95- the vacuum system. Each seaple bottle consisted of a bulb of about 150 ml. capacity, a two no. stepcock, and a 10/30 female standard taper Joint. First, one of the sample bottles was immersed in liquid nitrogen to trap out carbon dioxide. Then the stopcock on this bottle was closed and the othersample bottle immersed in liquid nitrogen to increase the mount of the non-condensible carbon monoxide in the bottle, and the st0pcock of this bottle closed. The stopcock of the first sample bottle was then opened to the vacuum to remove the carbon monoxide; the vapor pressure of carbon dioxide at this teqaerature is extremely low. The stopcock of this sample bottle was closed, the liquid nitrogen traps were removed, and the snple bottles removed from the vacuum system. For two chercoele the count of carbon dioxide was so small that only carbon monoxide was collected. The pressure of gas collected was in the range of 100 to 150 an. at room teweratm'e. ~ Tlree different charcoals were investigated. The first was a char- ecal activated a hoo°c . previously identified as one. From it were collected Gas Snple 1 containing carbon dioxide and Gas ample 2 con- taining carbon monoxide. The second charcoal sample contained iron (III) sulfue to the extent of 0.091 g. of salt per grma of charcoal. This ehmcel yielded Gas Sample 3, which was carbon monoxide. The third ehrecal s-ple was activated at aoo°c., and has been designated pro- viously as 04-10. From this charcoal was taken Gas Smaple h consisting of carbon monoxide. The results of the isotope analyses are presented in Table mn. I -96. TABLE mu OXIDE! ISO‘I'OPE CQIPOSITIONS W CARBON OXIDES REMOVED FRO! CHARCOALS Gas supls Gas 925tmdard 21: Sample 93: hirichment Em'icMent lunber on on on S 1 00. 0.002088 0.003783 0.001695 13.0 2 00 .002015 .00292h .000909 7 .0 a 00 .002055 .003271 .001216 9.1: 00 .002020 .0027h8 .000728 5.6 Colman three of Table 1111: gives the noreal isotope ratios for the gases analysed, column four gives the isotope ratios for the gas samples which were collected fron the chercoele, and column five is the difference between columns three and four. If a gas ample were not enriched at all in the heavy isotOpe of oxygen its isot0pe ratio would be that given in column threw. If it were enriched 100%, that is, if the water were instrumental in producing all the gases, its Ou/O1,‘ ratio would be approximately 0.0150, which was the fraction of 0“ in the water; this assures that the water in contact with the charcoal contains appreciably acre oxygen than does the charcoal. To find the extent to which oxygen from tin water entered into the gases analysed it is necessary to sub- tract the isotope ratio for the normal gas from the isotOpo ratio which would have been obtained had the gases contained only ongen from the water, and to divide this into the value actually obtained for the iso- tope mick-ant. This value, shown in column six, is simply the per- centage of oxygen in th carbon monoxide or carbon dioxide gas which had its origin in the water. .97- It is evident that an appreciable amount of owgen from the water entered into the carbon dioxide and carbon monoxide which were collected free the chercoele. On the basis of the drying operations it is con- sidered unlikely that any appreciable mount of water remained adsorbed on the chercoele which might have reacted with the charcoal at higher temperatures. The only alternative is that the water reacted with the charcoal at temperatures not exceeding 200°C" the temperature at which the final removal of water took place. This reaction could be one of two typesl first, the oxygen in water could exchange for oxygen on the charcoal surface leaving the systen essentially intact after the exchange; or second, the water could react with the charcoal in scans sanmr to form a surface oxide. Which of these two processes actually took place cannot be definitely determined from the date. The second alternative seem favored on the basis of recent work by Smith 33 3;.“ These investigators studied the reactions of water with chercoele in the temperature range 25 to 200°c. After a suitable time of contact of water with the sample, the charcoal was baked out Just below the softening point of Pyrex and the gases evolved were collected and analysed. The authors concluded that the reaction of water and car- ben gave 8, and a carbon-ongen complex. There seemed in fact to be two types of complex, onefthat decomposed at roan temperature to give carbon dioxide, and another more stable complex that decomposed at high temerature with the evolution of carbon monoxide. Since the recovery of hydrogen was far below the stoichiometric amount the authors suggested .93. that the stable complex nay contain Ivdreawl groups. Charcoal which contained ongen was much more reactive with water than charcoal treated with lydrogen at 100006.. to remove oxygen. Tb present work shows that the carbon dioxide sample collected contained the largest concentration of oxygen from the water, but that the ocbon monoxide asaple collected from the sac charcoal contained only about one-half as much ongen fro. the water. This indicates that tab types of surface complex say be formed by water, one which is pro-- duced sore readily and appears as carbon dioxide when the charcoal is band out, and another which is not produced so readily by water and appears as carbon nonexide. Since the charcoal which furnished Gas despise l and 2 contained bout 25‘ oxygen the entire charcoal surface was nest lilcsly completely covered with oxygen, and it is difficult to see in this case how water could react with the carbon of the charcoal. Undoubtedly for this charcoal at least the decomposition of water occurred through the nediun of the surface oxides. Sample h, which was charcoal activated at 800°C., evolved carbon monoxide containing the least percentage of oxygen free the water , indicating that reaction between water and this charcoal to fern a surface sample: does not occur so readily as in the case of low-temperature activated chucoals. The presence of appreciable ascents of iron (:11) sulfate on charcoal (ca-pic 3,) append to have had acne catalytic effect on the reaction of water with charcoal. Discussion Tb nature of the surface of activated charcoal has been studied by severalnethods, both chemical and physical. An attenIpt is sade in this section to correlate the results of the various methods, to relate the results to the findings of other observers, and in some cases to postulate csuses for the results in toms of the properties of the our-- face constituents. a summary of the properties of several charcoals is given in Table mm. Studies of the adsorption of acids, bases, and salts by specially prepared chercoele have resulted in no new information on this subject, but hsve been the means for characterizing the adsorption properties of the charcoal: and for confirming the observations reported in the litera- ture.“ Charcoals from which most of the surface owgen was rennved were found to have a small cqaacity to adsorb acids, but bases and neutral salts were not adsorbed at all. Chin-coals activated at approxi- nately 1.00%. were found to have a strong affinity for bases and to ‘ adsorb cations fro- inorganie salt solutions, but these charcoals adsorbed only saall ascunts of acids and of anions. Unlike nest substances charcoal appeu's to have no unique nemetic susceptibility, and in recent years it has become apparent that the Iagnotia susceptibility of a chn‘coal depends upon its past history, and in particular upon its onrgen content. A complicating factor in the study of the susceptibility of charcoal is the near impossibility of freeing the chu'coal suples from all ferromagnetic and perenagnetic inurities. " .100- 84.o. In... In! «ca 3.3 he 5 cos and End. .....1. III 0.9.. can an 5.. o8: fit... Simon ...I. ll. 9%. 23 ha 5 coco 3.6 «30.. Ill I... don 8.» ecu->383 To 3%? III. III 0.3 2.0 coca .t 3.838 ouo ..cl . 8.4.6 80.0 «.8 mm.“ .538 To 1.! 8o.o 084 0.: 8.0 one! one It. 1!... In... 3: 94 :3! won II. ~86 98... 0.2 84 38a! .10 ...l. «8.0 456 0.4... 8.0 38-2 moo ...... 086 24. o o. 2 mo. 4 «Hole «-0 II Red 846 «.2 SA anal: To hommwhnmoam .mmhwdux .umwuwcwux “whoop-mun“ u Ina confluence 839.83 Bacon £380 Echo coda->33 icon-no ago Eu 8 ”Magda In .3. a gag. ~101— It is apparent fro: the present study that the magnetic suscepti- bility of charcoal depends upon the amount of oxygen adsorbed by the charcoal surface, or really upon both the amounts adsorbed physically and also chenioally in the for- of surface oxides. Oxygen adsorbed plvsically is responsible for the variation of the susceptibility of charcoal. with temperature. rho mount of ongen adsorbed chenioally in general deteriines the susceptibility of the charcoal at rcon temerature. Charcoal fru fro- engen is stroneg di-agnotic, char- coal activated at 1.00%. is the acct weakly dimagnetic, and charcoal activated at 800%. possesses a dianagnetinc intmediate between these two. For ecsparison" the gran susceptibility at room temperature of graphite is. ~35 x lo“ and of diamond on x 10". ‘l'he fact that oxygen-free charcoal has a susceptibility of 4.36 a 10" but that char- seal with only 0.11: ongcn hes a susceptibility of -o.9h a 10" suggests that the charcoal itself may become less dinagnetic when the first amounts of oaygen are adsorbed on its surface. The elsnental analyses and ash determinations of the chmoals used in the present stucv persitted calculation of the amunts of oqgen present in the charcoals. The methods of activation and extents of activation of the chercoele were revealed in their ongen contents. . Outgaesed charcoal had a very low oxygen content, charcoal activsted at 600%. had a somewhat higher cvgen content, but the charcoal activated at 3300.6. could always be distinguished ‘by its very high oxygen content. Ithe adsorption of salts on charcoal, the analyses of both ceticne and anions, and the measurements of changes of pH produced results that ~102- are difficult to interpret. It was observed that neither charcoal nor adsorbste solution appeared to be electrically neutral after adsorption took place. Perhaps some electrically charged constituent of the ch0- ooal surface passed into solution but was not detected by the analysis. this is probably related to King's observation” that charcoals were always negative wbn placed in contact with distilled water, even though their aqueous suspensions night be either acidic or basic. A negative charcoal would be able to become neutralized by adsorbing cations. The fact that adsorption of cations by charcoals activated at 1.00%. increased after treatment of the chad-coals with eodiun murcxide is very interesting. It is possible that the reescn for this can be found in the presence of acidic surface oxides. Bchilow's acidic oxide is very such like the anlvdride of a carbozylic acid, and perhaps t1- action of the sodiul quroxide solution on that oxide is in the nature of n alkaline hydrolysis: ,IO 40 —c\ —c ’0 t HIGH -—-> do ...c\\ [We _c\\ 0 In the presence of distilled water lvdrolysis of the surface oxide pro. seeds to a limited extent and provides acidic hydrogen atoms which can be replaced by cations. But in an alkaline ncdiun the reaction proceeds uh sore readily, analogous to the enhanced hydrolysis of acid .103. snlvdrides in an alkaline medium," and the sodium ions can subsequently be replaced preferentially by the divalent or trivalent metal ion. Treatment of the charcoal surface with sodiun netaperiodate in an attmt to obtain the products of a selective oxidation led to results that were inconclusive. The only prochhct obtained was carbonate ion. The fact that the solutions were acidic at all times and that a fairly large amount of carbonate ion was found points to the fact that this was the product of the oxidation rather than simply the result of dissoluo tion of su‘bon dioxide from the atmsphere. Apparently the conditions of oxidation were too drastic to pemit the recovery of any intermediate oxidstion products such as formic acid or formaldehyde. The results of the infrared analyses of charcoals were also too meager to offer my conclusive ideas about the nature of the charcoal surface oxides. The fact that only the chm-coal with 25‘ ongen present showed szv absorption bande indicates that this method say not be very sensitive to small ancunts of surface oxides. however the absorption of infrared radiation in the carbolvl stretching region scene quite definite. Unfortunately the band is too broad to give infomtion on what sort of carbonyl mp or groups night be present. Obtaining the infrared spectra of charcoals say become a satisfactory way to ermine chercoele if the methods of preparing the sanples for infrared examination are improved. It is difficult to relate the adsorption of iodine by chamosls to the results of other studies, but there does seem to be the greatest «oth- iodinc absorption when the cvgen content of the charcoals is greatest. For the charcoals aotivsted in mania there is no clear relation between the absorption of iodine and that of acid, although such a re- lationship has been reported for air-activated chercoele.“ -105- HAQIEI‘OCHEHICAL STUDIES OF THE IATURE OF ADSORBE PARMAWIO SALTS In any study of the nature of adsorption it is important to know the kind of forces that bind the adsorbed species to the adsorbent. A nagnetochemical study of adsorption is often useful for this because the magnetic properties of the adsorbed substance depend upon the forces which bind it to the adsorbent. The results obtained by this method are particularly striking vlnn the adsorbed substance is para-agnetic and saw become dismagnetio upon adsorption. Earlier studies have been made"o" of the adsorption of salts of nickel, nanganese, cobalt, and iron on commercial sugar charcoal and on silica gel. Fm magnetic susceptibility determinations of the charcoal- adsorbate systems it was found that the susceptibilities of these ad- sorbed salts were essentially the same as the susceptibilities of the pure salts. Anomalous results were obtained, hoverer, in the case of iron (III) salts. In the present stuck salts of manganese, cobalt, md iron were adsorbed on specially prepared charcoals, and the magnetic susceptibili- ties of the resulting charcoal-sdsorbate systems determined at low temperatures as sell as at room temperature. the procedure was essen- tially as follows. To a known weight of charcoal was added 10 .00 ml. of an appromately 0.5 l solution of the paramagnetic salt. The mixture was equilibrated for 18 to 21: hours , then filtered through a sintered .106- glass filter. The charcoal was washed quickly with distilled water and the washings added to the filtrate 3 this was then analyzed for the nount of pneumatic cation not adsorbed. The mount of salt adsorbed was calculated from the difference between the amount of salt in the filtrate and the smunt present in 10.00 ml. of the standard solution. The charcoal with mm was dried at 110%. for two hours before m susceptibility was determined. for susceptibility no asurements at low temperatm-es the charcoal samples were always in an atmosphere of helium. When iron salts were used the solutions and chercoele were kept in an inert atmosphere as such as possible to prevent oxidation or reduction. Daring equilibration of charcoal and salt solution the nixture was kept in a steppcrsd flash from which the air had been replaced by nitrogen. the charcoal was dried by heating it in a test tube in which nitrogen was continuously circulated. rh- iron sulfate solutions were stabilised by we; then about ‘ tso normal in sulfuric acid, and the iron chloride solutions contained enough morochleric acid to naha than two normal. Before the suscepti- bilities cf the pure charcoals used for reference were determined they are treated with acid of the sus concentration as was present in the sea-responding salt solution. These charcoals were also dried in an a‘tlnsphsre of nits-oxen so that their susceptibilities would in every wq correspond to the true susceptibilities of the chmoals with no adsorbate. The gran susceptibility of the adsorbed salt could be calculated from the additivity relation .101- xmmm - cxc » (1’Pcuaalt (5) where ,c - weight fraction of charcoal. The temperature dependence of magnetic susceptibility of most para- ugnatio substances is expressed by the Curie~Weiss law I e C (6) T723 ‘ where 0 e Curie cement A - noleculsr field constant. This law is a generalisation of Curie's law I e G , '1' (7) A convenient way to express the susceptibility of a paramagnetic substance is by use of 'effoctivs Bohr magneton nunbers', defined by the relation ‘ A,“ - 2.839 W (a) where 13 - susceptibility par sole. If the sanitude of A is not known but it is desired to get an estimate of the sine of the magneton mber, the following mprozinate relation is useful: A... -2.839 m (9) The advantages to the use of the Bohr nagzaeton number are that unlike the susceptibility itself the nagneton number is independent of tempera- ture, and also that the magneton war is related to tie number of .108- unpaired electrons in the substance by the simple expression fl eff " m (10) where n - number of unpaired electrons. This relation is based on the asstmxption that the permagcetic suscepti— bility is due to spin only, that is, that there is no significant orbital contribution. Duplicate experiments were carried out on the adsorption of iron (11) sulfate on comm-cm charcoal, and magnetic susceptibility de- tor-inations were made at room temperature. Since 10.00 :11. aliquots of the standard solution were equivalent to 53.71; 3 0.03 ml. of 0.09577 3 ceric moniun sulfate solution, this solution was 0.511;? M. Charcoal samples of 7.3M: g. and 7.030 g., denoted as samples I and II, respectively, were equilibrated with 10.00 ml. of the stmderd iron (II) sulfate solution, and the filtrates were equivalent to 16.21; and 50.13 :11. of cerate, respectively. Thus senpls 1 contained 6.20 mg. of iron per gram of charcoal and sample 11 contained 2.75 mg. of iron per pas of charcoal. The gran susceptibilities of the pure charcoal and of charcoal containing adsorbed iron (I!) sulfate are given in Table 1m. . For divalent iron the number of 11an electrons per atom is four and A,” is 13.90. Calculation of effective Bohr nagneton numbers for samples I and II give, respectively, 5.1? and 5.85. Selwood" gives the range of experimental Bohr negneton numbers for iron (11) salts as 5.0 h 5.5. ~109- TABLE m1 cam snacmmnnms or CHARCOAL m or mos (II) MATE ADSORBED on omen. Current Pure Charcoal Sample I Sauple {I Amperes 1210‘ 1:10! 1:10 10 4.352 +0.1h8 10.316 12 .362 .739 .Bth 1h .372 .7142 .33b 15 .377 .7147 .326 18 .387 .752 .326 20 .386 .7h2 .322 A similar study was made of iron (III) sulfate adsorbed on cher- coal. To 1.1361; g. of charcoal was added 10.00 :1. of 0.3 1! iron (III) sulfate solution. The sixture was treated in the sale nanner as above. the filtrate was passed through a silver reductor and titrated with 0.09577 I eerate solution, requiring 56.95 al. of cerate. fan I]... of the standard iron (III) solution was equivalent to 58 .99 n1. cerate; thus the standard solution was 0.2825 n. It ... calculated that 1.16 ag. of iron eere adsorbed per gran of charcoal. 'fhe gran suscepti- bilities of the pure charcoal, charcoal with adsorbent, and of the adsorbed salt are given in Table 111. t h the average value calculated for the gran susceptibility of the adsorbed iron (11:) sulfate was 1.6.? x 10". am this the effective lohr sagneton number was found to be h.?h. The 'apin onli' value is 5.92 and the range of values found for iron (III) in pure salts by other experilsntsrs is 5.14 to 6.0. .110- new In cam wSCEH'IBILITI or CHARCOAL AND 6 IRON (III) SULFATE ABSORBED on CHARCOAL W Current Charcoal Charcoal with Adscrbenta Adaorbed Salt Amperea I x 10 I x 10' I x 10' h -°.399 0.7M 4445.0 6 .h91 .671. 1.6.0 lo 5% .661 h? .7 15 .569 .653 1.8.0 a 1.16 ng. iron adsorbed per gran of chancel. It was desirable also to sake magnetic measurements at low tempera- tures to determine whether or not these adsorbed salts obey the Curie luv for paramagnetic substances. Such measurements were made with manganese (II) sulfate, cobalt (II) chloride, and the sulfates and chlorides of iron (II) and iron (III). Cobalt (II) chloride and manganese (II) sulfate were each adsorbed on samples of comereial charcoal by the nethod described above. The cobalt solution was 0.1912 n and 10.00 .1. contained 0.2895 g. of cobalt; and filtrate was found to contain 0.2506 g. of cobalt, so that 0.0389 g. were adsorbed by the 6.665 g. charcoal sample. Il'he gran susceptibili- ties of the charcoal and of the charcoal-adsorbent mixture are given in Table XIII. The susceptibility of the charcoal at roon temperature was -0.h6 x 10.. and it was assumed to be independent of Mature. TABLE m1 W SUSCEPIIBILITI Q COBALT CHLORIDE ABSORBED ON CHARCOAL Temperature Charcoal Charcoal with Adsorbent Adeorbed Cobalt °I I x lo I x lo‘ I x 10’ 196 ‘ +0.96 2h3 1? +2.85 568 *5.8h sg. of cobalt adsorbed per grass of charcoal. These data can be analysed conveniently by the Curie-Weiss equation (6) which can be put into the for! g c r + A (11) If this equation be valid for the substances studied, the plot of reciprocal susceptibility as a function of temperature should be a straight line with the slope scual to C and intercept at the temperature aais equal to (- A ). The susceptibilities of 00th (II) chloride were calculated tron equation (5) and the reciprocal susceptibility was plotted as a function of teaperature as shown in Figure 15. It is evident that the three points do in fact lie on a straight line. The slope of the line was found to be 0.0107, in tens of the gran atomic susceptibility of cobalt, and the value of the intercept (- A) was six degrees. Since free equation (6) C-X(T+A), (12) -ll2- -3) Reciprocal susceptibility (x 10 6.0 ._ .1 I i h.o ._ 2.0 ... L l -50 0 1(1) 200 300 Temperature (OK) Figure 15. Reciprocal susceptibility as a fumtion of temperature for cobalt (II) chloride adsorbed on charcoal. -113- the effective Bohr nagneton number for the adsorbed cobalt could be calculated (equation (8)) from the elope of the line, and was found to be h.87. The value for cobalt (II) chloride given by Stoner“ is 5.01. This there was no ppreciable difference in magnetic properties between adsorbed cobalt (II) chloride and the pure salt. This study was repeated in essentially the same nanner using man- ganese (II) sulfate as adaorbets. The standard manganese solution was 0.130115 1! and the usual procedure was to dilute 10.00 al. of the solu- tion to 100.00 ml. and then titrste h0.00 n1. aliquots with standard 0.015 I potassiua per-anganate solution as described earlier. The h0.00 I1. alienate of standard sangmess (II) solution were equivalent in this tanner to h0.9h :5 0.05 :1. of permanganate, and the filtrate solution was equivalent to 31.55 I 0.10 .1. of permanganate. Thus on the 7.57); g. charcoal sample 12.6 mg. of salt were adsorbed per gram: of charcoal. The gre- auscoptibilities of the charcoal and of the charcoal- adsorbent sixture are given in Table XXIII. The susceptibility of the charcoal at room teaperature was o0.h6 a 10.s and it was assumed to be independent of te-perature. The susceptibility of nanganese in each case could be calculated using the additivity lam expressed by equation (5) and tald.ng the susceptibility of charcoal to be -o.h6 x 10". The reciprocal susceptibility was plotted as a function of temperature and again the three points lay on a straight line, as shown in Figure 16. The value of the intercept was sinus tires degrees, indicating that this salt very nearly obeys Curie's law. The slaps of the line was 0.0025, ea!) Reciprocal susceptibility (x 10 3-0 h— 2.0 _ -50 l 100 0 Temperature ( I) 200 300 Figure 16. Reciprocal susceptibility as a function of temperature for manganese (II) sulfate adsorbed on charcoal . -115- in terns of the gran atonic susceptibility of aanganese. The Bohr magneton number was calculated from ecuaticn (10) to be 6.01;. Stoner's value“ is 5.86. Again the nagnetic preperties of the adsorbed and of the pure salt were found to be essentially the ewe. mu mu cam suecmnmu (r HARNESS (11) name ADSORBED on CHARCOAL #— renpgrature Charcoal Charcoal with Adsorbentit Adsorbed Hangeneee I 1x10 leO' leo' 295 -0.h6 40.78 276 193 +1.“ 1:16 16 44.37 1075 fit.” It. of nanganese adsorbed per gran of charcoal. Inasmch as it had been reported previously" that the ugnetio nonente of iron (II) and iron (III) salts adsorbed on charcoal were snonaleus, a study was sade of these systems which included low tempera- ture nagnetic susceptibility moments. The charcoals used were all activated at h00°0., and the experimental techniques and treatment of data were the sexes as those discussed above. The results are shown in Table IIIIII. Coluln six gives the calculated effective Bohr nagneton Mars for the adsorbed salts, and column seven lists the effective: Bohr nagneton timbers found by other observers from nsasurenents on the pure salts.“ ceiuun eight gives the value of the constant A and eoluen nine the value of the slope . -116- keg: nova-Bel ion 83 a.“ {as 73....er «o is» .89 cannon—i £83338 bunnies-hem an» no '3 ... nooJe and: «5. 59 03.0.. fig «806 n mum mm.m 8a.? 2.4.9. 8n SJ e5...— abd. 8a.? WE. 50.0.. 34.0.. 483 38.0 3 gm $.m 8a.? 34.0. we“ do; as: mine. 5...? m5. 236. 8...? fin? oflbd 3 8a Cum 8%? 43.? are. 3.... enoea 03.9 «and. «.2 48.9.. 0.3.0. 0&9” v 2.8.0 3 and «a... pad. .23.? 8a 2..“ A 8%: 3a.? 09:00 no» $3.9. 93.0.. 3.3 38.0 me. a: 85 and. .23.? new one eaeaveea eeehoa fiem fie» eg u an: a no aloha: ocean . than neonate-en confluence 32%er . ‘1) mg BNH fig 8 mgasga E4: E39: .117- The plots of reciprocal susceptibility as a function of temperature , which test til validity of the Curie-Weiss law for these samples, are given in figures 17 through 21. Within experimental error these plots are all linear, and from the slopes of these lines the Bohr nagneton nubers were calculated. The magnitude of the constant A , the “mole- cular field constant”, is an indication of the nagnetic environment of the pusnbgnetio ions ." For nagnetically concentrated substances where the paranagnetio ions lie closely together A is large in absolute value, but for nagietically dilute substances where there is no nutual inter- action between the nagnetic species A is nearly zero and Curie's law is valid. The values of A given by Stoner“ for iron (III) sulfate, iron (I!) chloride, and iron (III) chloride are 66, .305, and 3, respectively. The fact that the values of A obtained in the present investigation tend to be less than the accepted values for the pure salts nay be due to the greater distance between“ nagnetic neighbors caused by dispersion of the salt upon the charcoal surface. It is clear fro:- the results presented in this section that all the paruagnstie ions whn adsorbed on charcoal retain their paranametisn. Hanganese (n) sulfate and cobalt (n) chloride obey the Curie-Weiss la! very closely and have the same nagnetio properties in the adsorbed state as they have as pure salts. . Similarly both divalent and trivalent iron have essentially the sane paramagnetisn in the adsorbed state as in the pure salts, although in this case the limits of error are appreci- lbly larger. These findings are in conflict with those of Bhatnagar g; 5;.“ who reported that theee salts bee-e diamagnetic upon adsorption .118- e3) Reciprocal susceptibility ( x 10 6.0 #— ] l l -50 O 100 200 300 Temperature (OK) Figure 17. Reciprocal susceptibility as a function of temperature for iron (III) sulfate, sanple A, adsorbed on charcoal. -ll9- -3) Reciprocal susceptibility ( x 10 8.0 - 6.0 -— C h.0 P C 2.0 u- I l 1 -SO 0 1m 200 330 Temperature (OK) Figure 18. Reciprocal susceptibility as a function of temperature for iron (III) sulfate, sanple B, adsorbed on charcoal. -120- Reciprocal susceptibility (x 10-3) h.0 h. 3.0 )— {I 2.0 _. 1.0 .. l l l -50 0 100 200 300 0 Temperature ( I) Figure 19. Reciprocal susceptibility as a function of temperature for iron (III) chloride adsorbed on charcoal. -121- -s Reciprocal susceptibility (x 10 ) 14.0 3.0 2.0 1.0 I l l -50 o 100 200 300 Temperature (OI) Figure 20. Reciprocal susceptibility as a function of temperature for iron (II) chloride, sample D, adsorbed on charcoal. -122- Reciprocal susceptibility ( x 10-3) heo I'- 3.0 h- 2.0%- 1.0-— 1, -§) 0 100 330 Temperature (OI) Figure 21. Reciprocal susceptibility as a function of temperature for iron (II) chloride, sanple E, adsorbed on charcoal . -123- .eel J and that the charcoal even became more diamegnetic when salts were adsorbed. However, tin chercoele used by these workers contained large amounts of paramagnetic impurities, even to tin extent that the char- coals themselves possessed resultant paramagnetisn, so that the validity of their results say be ealled into question. It may be noted that under the conditions of the present investigation there was no appear- wF—“F‘! an. ance of the soocalled "super-paranagnetisa' reported by Ioboaev at 31..." the reason for the very erratic results obtained for the iron (II) .- .. ..__-7_.-_..__ a 1 ‘el'h . ‘ A and iron (:11) salts ie not known. This is particularly strildng in the case ef the iron (II) salts for which the sagastic moments are on L the average even higher than those obtained for the iron (III) salts 3 this is difficult to understand since iron (II) has four unpaired electrons and iron (III) has five. The even complete oxidation of di- valent iron to the trivalent state would not entirely account for the high values obtained for iron(II). It is believed, however, that the iron remained essentially in the divalent state since precautions were taken to minimise its oxidation, and since the iron recovered by desorp- tion from the charcoal surface was nostly divalent (see above). Tb results of this investigation show that it is highly improbable that the adsorbed cations are bound to the charcoal surface by means of covalent bonds to surface carbon atoms since metal to carbon bonds of this type should reduce the magnetic susceptibility of the metal atom very markedly." Furthermore the adsorption forces are much stronger than van der Weals' forces which are operative in physical adsorption as seen by the fact that the cations could not be removed from the .1214- charcoal surface even by leaching with acid. The only reasonable a1- ternstive is that these metal ions are bound to chemisorbed cvgen on the surface of the charcoal. Such bonds would have sufficient ionic character to leave the electron spins of the cation unpaired. This explanation is consistent with the ion embange mechanism proposed by Steenbergu to account for the adsorption of cations on charcoals activated at low temperatures. The charcoal surface may consist of acidic carbon oxides which in the presence of water would foru acidic hydroxyl aromas. The cations could react with these groups liberating hydrogen ions into solution and forming essentially ionic bonds to oxygen atone in the surface. .125- 144444.- -_.._..__._4 _ — up '0 4 <'_-‘.e -s‘i.s A MAGEOGW STUDY OF THE 50me OXIDATION OF COPPER Introduction The corrosion of cepper has been studied very extensively and under a great variety of conditions. This is undoubtedly due to the extreme -4 “x: . 2-4 . '.' usefulness and versatility of capper, both as the pure metal and as an alloying constitmnt. In addition the importance of cuprcus oxide as a semi-conductor has stimulated work on the interaction of ongen with cuprous oxide and the relation between the physical properties of this 1? naterial and the copper-oxygen composition. The copper-copper oxides- cages systeu is interesting from the negnetochenical viewpoint because the sagnetic preperties of the various phases are very different from each other: copper and mm oxide as dienagnetic and the suscepti- bilities are essentially independent of temperature, cupric oxide is pare-agnetic but the susceptibility padually decreases with decreasing temperature, and ongen is a normal paramagnetic substance which obeys Gurie's law. The purpose of this investigation was to examine nametochemically the surface oxidation of copper by pure oxygen at low temperatures (25 to 100°C.) . Pros independent neasurements of the extent of oxida- tion and from magnetic susceptibility deteminations at several tempera-- tures it was considered possible that one night ascertain which oxidation products were formed and also their quantity. Such measurements night .126- show whether or not there was any 'surface par-enametisa' associated with ccppcr or its oxides. Finally it night be possible to detect a change in the magnetic susceptibility of the copper upon the adsorption of hydrogen or oxygen, as Selwood and co-workcrs recent reported for highly-dispersed nickel.” 4‘ ;;-4 . Historical Survey “XS-Z I Methods for Determining the Thickness of the Oxide Film A number of different methods are available for detereining the extent of oxidation and the composition of the oxide film. The uost inportant of these are mesmerized below." (h'svinetric. The increase of weight of the ample is determined, and the count of oxide fouled can be calculated if the couposition of the oxide is known. This is best suited for thick files were inaccur- acies in weighings are not so important as for thin films and mall weight changes. nectron diffraction. This method is useful for exaaination of films 100 A. to 300 A. thick, and the compositions of the files can be detemined. Electron diffraction gives no indication of file: thickness and no inforsstion is obtained about the interiors of thick films. Electrolytic. The time required for cathodic reduction of an oxide film is a measure of the anonmt of oxide present." In the case of copper it is possible to obtain the amount of each oxide, nonovalent or divalent, which is present. -127- W decrease. Iron the decrease in the pressure of the oxidiz- ing gas in a closed system the quantity of material abstracted by the natal can be calculated. This was the method used in the present investi- gation to obtain the total amount of ongen adsorbed. Composition of the Oxide . n One of the basic problems in any study of oxide fomtion is that of determining the composition of the oxide, and for a metal such as copper where two oxides are stable under most conditions this problem K , is doubly important. Innunereble studies of oxide composition have been 2.» made on the caliper-oxygen system over a wide range of temperatures. Allen and Mitchell" stated that as far as they were aware oxygen is not dissociated on copper below 2hO°K. There is no oxide layer formed on copper below 2hOOI. , and this is the critical temperature above which oxidation occurs. up to about 200°C. and above 900°C. the oxide formed is largely cuprous oxide. At intermediate temperatures a mixture of cuprcus and cupric oxides is no arly, always found. Crussn and Riley" found that the essential criterion for deter- sining whether or not cupric oxide would be present as an oxide coupon- out was the thickness of the oxide film. Cuprous oxide film thicknesses of less than too it. contained no cupric oxide, eone of the films in the range 100 A. to 800 A. contained some cupric oxide, but all above 800 A, contained cupric oxide. Factors of secondary importance in that they control file thickness are the temperature of oxidation, the tine of exposure, and the concentration of oxygen in the gas. The above authors ~128- boated copper strips in air at 2ho°c. They reasoned that as the film thickens diffusion through it becomes more and more difficult, and at some point cuprous oxide is expected to begin competing with the capper metal for the attacking oxygen. In the range 150°C. to 350°C. the temperature at which the films are famed has little influence on the formation of cupric oxide in the cuprous oxide films. The work of Hudson" supports these conclusions. Evans has stated" that under reversible conditions the cupric oxide phase should be present from the first and that its failure to arrive is an example of departure from thermodynanic equilibrium. ' When the oxidation products of capper below about 200°C. are ' studied by electron diffraction it is usually reported that the presence of cupric oxide is not detected. The sensitivity of detection of cupric oxide is "tincted to be one or two percent. Preston and Bircumshaw" reported that photogrth of rings at room temperature showed only rings due to cuprous oxide, but that at 100°C. and 183°C. extra lines were occasionally observed which might have indicated the presence of cupric oxide. White and Ger-or" showed that at room telpsrature capper ex- posed to dry owgen at a pressure of 20 me. did not show any lines of cupric oxide. Bound and Richards" studied by electron diffraction the oxidation of thin was of copper in air, and found that at 100°C. the films were completely oxidised to cuproue oxide in Just under an hour. After heat- ing the emples up to 20009. there was no change in the transmission . patterns, but at 250°C. a mum of cuprous oxide and cupric oxide -129- was found. Further heating at 300°C. converted the pattern completely to that of cupric oxide. The composition of the oxide reported by Dankov and Ignatov formed by heating massive capper in air was essentially different from that reported by Bound and Richards, however. The Russian investigators 79 used three mm. electrolytic capper plate, and at 200°C. found diffraction lines of both cupric oxide and cuprous oxide. Heating for 25 minutes under these conditions produced the mammal mount of cupric oxide, and for both shorter and longer heating times many of the cupric oxide lines dropped out altogether. After heating for 150 minutes not a single line of cupric oxide was left. Evidently the outer layer of cupric oxide in the film reached a maximum when heated for 25 minutes. The authors considered that these data were evidence for a mechanism of oxide growth in which the diffusion of capper ions from metal to adsorbed oxygen layer was the slow step. Anderson”0 has emphasized the fact that cuprous oxide is a phase of potentially variable conpoeition, and that the composition of a sur- face oxide on a capper fill may or may not be 01130.net”. This point will be considered further when the mechanism of oxidation is discussed. Mechanism of Oxidation The mechanism by which the oxidation of copper proceeds, as inter- preted from oxidation rate studies, was found to be determined largely by the temperature of oxidation and the thickness of the oxide layer. At temperatures not exceeding a few hundred degrees Centigrade the oxidation follows a logerithnic rate law“ and the slow step in the oxidation is the diffusion of ions across the electrostatic space charge set up within the thin run." it temperatures above 700°C . copper exidisee according to a parabolic rate law” of the form x’ - 2iat where X . film thickness A - constant t '- tine. The rate-determining step in this type of oxidation is the diffusion of capper from the natal to the oxide-oxygen interface.“ At intermediate temperatures the oxidation defies generalisation and it is possible to find theoretical and ezqoerinentsl verification for various laws.” Dell, Stone, and Tiley” have made calorimetric studies of the adsorption of oxygen by capper at 20°C. They found that for the first uptake of oxygen by granulated capper the heat of adsorption was very high, about 110 kcal. per sole. The heat of adsorption rapidly fell to bout 82 keel. per role which is the heat of oxidation of capper to suprous oxide. Later it leveled off to 55 kcal. per nole, which is the heat of adsorption of ongen on euprous ende.“ In between these two clear-cut steps the heat of adsorption gradually decreased from: 82 to SS keel. per mole indicating that in this intermediate region both pro- cesses were taking place. Thus oxidation of the metal was soon arrest- ed and that process»: was replaced by adsorption in the oxide layer. .131- There is considerable evidence that pure water vapor and pure steam have little effect on the oxidation of copper up to 800°C. at atmos- pheric pressure .71 Campbell and Thomas” used an oxygen pressure of 150 mm. and a water vapor pressure of 15 ms. at 100°C. and found that presence of the water vapor retarded the rate of reaction slightly. In similar studies by Preston and Bircumshaw" treatment of cepper with stem for one hour at 100°C. gave cuprous oxide and a second phase which was thought to be cupric oxide. The action of water vapor in ‘ these systens is difficult to explain. Activation of Metallic Copper The activation of copper by repeated oxidation followed with re- duction by hydrogen has been ascribed to an increase in the surface area." Reduction is evidently not simply a reversal of the process of oxidation. As the oxide is reduced by twdrogen, a layer of metallic copper is first formed around the ends. Wdrogen then diffuses through the natal to the oxide, and water is produced at such high pressures that its escape cases cracks and blistering of the surface .“ Repetition of this oxidatisn-reduction cycle gives highly active sur- faces. Reduction wdth carbon monoxide does not produce this effect, but gives a smooth surface of very low activity. Adsorption of Ongen on Copper Oxides Ongen is adsorbed on ouprous oxide at a measurable rate at room temperature, and the process possesses an activation energy. Garner, ~132- Stone, and TileyM found the heat of adsorption at room temperature to be about 55 keel. per sole, which was independent of the thickness of the oxide layer. Disappearance of owgen continued even after 100 hours. Garner, Gray, and Stone reported that at 200°C. the oxidation of cuproua oxide to cupric oxide was appreciable, but not very feet.3‘7 Although st roon temperature the edsorption of omen on cuprous oxide is fairly rapid, the rate of desorption under vacuum is very slow. At room temperature the adsorbed omgen neither evaporates nor does it contribute naterially to the oxidation of cuproua oxide , but the ad- sorbed oxygen is stable on the surface for at least 2h hours. when the authors allowed the edsorbste to react with carbon monoxide and collected the resulting carbon dioxide , they found that only half of the oxygen which was. adsorbed could be removed. ‘fhus about half of the oxygen is 'nobile' and the other half unreactive. Measurements of the semi-conductivity of cuproua oxide shed some light on the interaction of oxygen with this oxide. LeBlsnc and Sachse” found at 20°C. a specific conductance of 10". mho per cm., which changed to 10'. nho per cs. when 0.11 oqgen was present. These observations have since been confirmed by many investigators. The “conductivity of cuproua odds is a minimum for the pure , oxygen-free asterial, but the conductivity increases rapidly with the addition of oxygen to a definite limit. Zhuse end Kurt-chum“ found that the conductivity of cuproua oxide with but 0.6S cupric oxide was very little different from that of pure cupric oxide. .13 3e- Fron the temperature coefficient of conductivity in the range 25 to 250°C. the activation energy for the conductivity process has been determined.“ For a cuproua oxide surface free from adsorbed ovgen the activation energ is S .8 keel per mole 3 on the adsorption of oxygen this value falls to L96 keel per mole. Conductivity was found to be due to positive hols conduction. It is possible that in the presence of oxygen conductivity is a surface phenomenon, while for the pure oxide the conducting ions lie embedded in the material. Garner, Stone, md Tiley“ have reported that no Qpreciable mount of oxygen is adsorbed by cupric oxide. Magnetochesistry of Copper-Omen. System Although the ground state of the capper atom is a doublet 8 state 9" retellic capper is not paramagnetic but diamegnetic. Stoner” has pointed out that for the metals copper, silver, and gold, all of which are dis-apietic, the merioal value of the atomic susceptibility is less for the natal than for the ions. This clearly shows that there is a permanetio contribution from the electrons . Selwood and Dallas" heve reported an interesting attempt to pro- pare setallio copper in the paruagnetic state. Low concentrations of cupric oxide on an inert support had shown characteristics of high dis— persion over the catalyst support. It was thought that by reducing the septic oxide to netallic copper the note]. atoms would be far enough apart that they would exhibit paraegzetisn. The metallic copper proved .131;- to be dianagnetic, however. The conclusion was that the tendency for cohesiw is so strong that even when the surface of the support is lagely unoccupied by copper stone they must still aggregate together quite strongly. . The grew susceptibility of pure copper natal has been detersined by a large number of investigators, and the agreement mug various authors is quite satisfactory. The noun value given in the Inter. national Critical Tables " for 18°C. is ~0.086 x 10". Host values given in the literature range from o0.080 to -0.090 x 10". at room tesperature. The susceptibility of pure copper changes little, if my, with teaperatureg changes in diuagnetisn of five or ten percent are sonstines reported between rcon temperature and 750K.“ Hontgonery“ has found that copper crystals at roon temperature show no sagnetie anisotropy. The mean value of the susceptibility found was ~0.085 x 10... The susceptibility of capper was found to be independent of particle sise in a study by Bhatnagar and coworkers.“ Copper pulverized to 0.2 nicrons had the susceptibility -0.081 x 10'". Earlier Rae” re- ported that as particle sise decreased the susceptibility increased, but these results say have been due to lack of proper precautions for excluding oxygen. Sons interesting studies have revealed that the sagnetic suscepti- bility of bulk netsllio copper is a function of the stress to which the suple hss been subjected. Hutchinson and Room“ found that at -135. room temperature tho susceptibility of mnealed copper was -o.oeo x 10" but after cold-working it became -o.07h x 10". The ssmples had dif- ferent temperature coefficients as well, but at high enough temperatures (about 100 °x.) the susceptibilities has me about equal. Similar re- sults have been reported by others."’" Of the reported values for the magnetic susceptibility of cuproua oxide at roos tonporoturo, that of Bhatnagsr and Hitra,”° -0 .188 x 10“, is probably the nest reliable. Suples prepared by reduction of rehling's solution and by electrolysis give the some results . Other satisfactory values reported are -O.18 x 10" by noon and cohuth,“‘ and ~0st x 10" by mama.“ Due to the difficulty of removing traces of parasagnetio cupric compounds from the emples, many investi- gators have reported that cuproua oxide is paramagnetic.” he de- terIinations of the susceptibility of cuproua oxide below room tempera- ture have cone to the attention of the present writer. Several different values for the semiotic susceptibility of cupric ends at roo- tupu'ature have been reported, and they are only in general agreesent. ChSneveau‘” found +3.6 x 10", Birch‘“ reported +3.01: 1 10" at 10°C., and Ishiwaram' found+3 .20: 10". at 20.300. Ishiwu-a seems to have lads the only measurements on the pure compound at low temperatures, and his results are given in Table nnv.. The temperature dependence of susceptibility for cupric oxide is very unusual. Selwood and co-workers have studied the susceptibility of cupric oxide supported on magnesia, rutile, and alumina.""'°° Pros magnetic -135- TABLE m GRAN SUSCEPI'IBILITY OF CUPRIC OXIDE AS A FUNCTION OF TEMPERATURE Temperature (tram Susceptibility 03 e I X 106 ~20.3 *3.20 . 2 e7 3 am .30 e5 3e07 ~h5.9 2.95 «65.0 2 .77 “811.0 2 e63 -98 .2 2 .55 ~113 .1 2 .h2 -128.h 2.3h -1139 e1 2 e28 .16? e9 2 .17 I ‘ measurements made between room temperature and 90°I. they concluded that the cupric oxide obeys the Curiedweiss lax, and has a.nagnetic ment corresponding to one unpaired electron. noun-1m studied the magnetic properties of oopper oxidised to cuprous oxide at 950°C. to iooooo. ‘When the oxide Ins exposed to air at high temperatures it ens disnagnetic at roan temperature, but became paramagnetic at very low temperatures. When the oxide had been vacuum- tempered at high temperatures it was found to be paramagnetic at room temperature but diamagmtie at low temperatures. Hamel was not able ' to find any correlation betveen the magnetic susceptibility and comi- conductivity of cuprous oxide with varying oxygen content. -137- Fogerimentel Procedures The vacuum system constructed for this investigation is depicted in Figure 22. The distinguishing feature of this vacuum system was that from tin two-stage diffusion punps to the manifold, tubing of 20 m. ameter or greater was used; this construction permits of high pumping speeds. The single exception to this was the right-angle stopcock isolating the manifold from the pumping system; the bore of this stop- cock was 15 me. The HcIeod gmge used was capable of measuring pres- sures ae los'as l leO'. II. ‘Ths pumping system consisted of a two- stege mercury diffusion pumpfl backed by a mechanical pump.M The mania fold of the vecmm system was made from 22 m. Pyrex tubing, and storage flasks were constructed from round-bottom flasks, two of then having volumes of tee liters and the other having a volume of one liter. The nanometer used for measuring pressure ohms was node from seven m. Pyrex tubing and was equipped with a one meter scale. The suscepti- bility tube could be attached to the vacuum syetcn through a standard taper ground joint, and between that Joint and the mmifcld was sealed a coarse sintered glass filter so that the contents of the susceptic— bility tube would not esc ape into the witch! when the tube was being evacuated . a B. 8. Martin and Ccnuamy, Eveneton, Ill., Hodel H-hOlBO-l. 1“‘I'szney Manufacturing Compaq, Boston, 24853., Type CVD—BhBl. .133. E398 =53m> .mm 0.8m?” maZDa ’55:qu >Kaomw2 QSDQ NDDP >P_J_0_FQUOWDw . 5.sz— m<0 nZDawmou 0... ENE 02_>mo CMPU’OZ‘Z own”... are w mind-E mo<10km -139- The volume of tin manifold was determined in the following manner. The volume of the one liter storage flask was determined before it was attached to the manifold by weighing the amount of water required to fill it. That weight of water was 1093 g. which at 23°C. occupied a volume of 1095 cc. This flask was ttnn sealed to the manifold, the whole systu was evacuated, and helium introduced to the nsnifold but not to the flask. The pressure of heliln was read on the nanometer and then the stopcock to the storage flask was opened and the pressure again read. Iron the ideal gas law the volume of the manifold enclosed entirely by stopcocks was found to be 1614 cc. In a similar manner the volume of the susceptibility tube filled with ample plus the tubing connecting the susceptibility tube to its stopcock was found to be seven cc. The small correction for the change in volume of the vacuum systen due to the rise md fall of the nercury colum in the nanometer was made where necessary. Finely divided not allic copper was prepared by reduction of a copper sulfate solution with mdrasine , according to the method of Garner, .22. 2..“ In 300 ml. of deninerslised water was dissolved us .9 g. of copper sulfate ponbemdrataf and to this was added an ludrasine hydrate" until the evolution of hydrogen ceased. A mudchr-brown copper suspension was formed. The copper natal was separated fron the liquid by “Hallinckrodt Chemical Works, Analytical Reagent grade. "hathieson Chemical Co. , technical grade. -lhO- eentrifugation aid the product was washed with denineralized water six times. The copper was then washed once with absolute alcohol and fil- tered through a sintered glass filter. It was finally put into a weighing bottle and exposed to the drying action of anhydrous calcium chloride in a desiccator for one day. The weight of product obtained was about ten grams. Complete elimination of all oxidation products on the surface of the capper was insured by tln following treatment. The capper was put into the susceptibility tube and the tube and ample were attached to the vacuun system. The couple was heated by a heater consisting of hichrcne heating wire wound at half-inch intervals around a Pyrex tube 0.6 inches in diameter and seven inches long; the heating wire use covered with asbestos paper and then wound with heat~resistant tape. The copper smile was initially heated under vacumn to a temperature of 200°C. For he hours the pressure was maintained at 10" m. and then for a period of about three weeks the smxple was alternately exposed to hydrogen at atmospheric pressure and then evacuated to a very low pres- sure. At the end of this the the ample was no longer able to take up hydrogen and the reduction was complete. The reduced copper was then evacuated at a pressure of 10"“ mm. for to hours. The sample, which had been at 200°C. during all these Operations, was then cooled. to room temperature and helium was admitted as the supporting atmosphere. Preliminary magnetic susceptibility measurements on this product indicated the presence of a ferronamtic contuninant. The capper was -151- then removed from the susceptibility tube and washed thoroughly with 10% turdrochloric acid, then several times with denineralized water , and finally was dried 15 minutes at 110%. Thia copper was put into the susceptibility tube and the reduction treatment again given the sample until it was no longer able to take up hydrogen. It was then evacuated to 10" n. and isolated from the ‘nanifold of the vacuum systole by a stopcock. Oxygen which was taken free a cylinder and passed over anim- dreus calciun sulfate was then introduced into the unifold until at- sospheric pressure was reached. The oxygen pressure was read with the sane-etc and the stepcock to the susceptibility tube opened to allow contact between oxygen and the ssnple. After a suitable tins of contact, usually 2h hours or longer, the cages pressure was again read from the nanometer. The aseunt of ovgen adsorbed by the sample was calculated from the change in pressure after correction was node for the expansion of oxygen iron 16‘: cc. to 171 cc. lo attempt was ads to control the temperature precisely since room temperature did not very nore than one or two degrees Centipede during the course of an adsorption determina- tion. When the cages had been in contact with the sanple for a suit- able length of tine the mph was emuated to 10" or 10" a. and heliun was introduced as the supporting atnosphere preliminary to the susceptibility nemrenenta . . Guprcua oxide was prepared by the following method.ma Tnnty grass of copper sulfate pentahydrate,’ and to g. of c. P. grade sodiua e . Hallinckrodt Chemical Works , Analytical Reagent grade . -1142- s potassiul tarts-ate, were dissolved in 1100 ll. of demineralised water. Then G. P. sodius Ruth-oxide solution was added until the pH of the copper solution was approsinately nine . To the clear solution was added 25 g. of dextrose and the solution was heated to boiling. It was kept near t1. boiling point for 30 ninutea during which tine it was stirred continuously. A finely-grained, bright red precipitate of cuprous oxide settled out. The supernatant liquid was decanted from the precipitate and the cuproua oxide was washed several tines with dominerslized water. It was then transferred to a weighing bottle and dried in a vacuum desiccator overnight. The asgnetic susceptibility of the mrous oxide sasple so prepared was—0.162 x 10.. naasured at roon tespsrstura, OOlpII‘Od with the litera- turs vain-“W of 4.188 x 10". Cuprous oxide has the ability to adsorb paramagnetic oxygen tenacioualy, and probably that was why this sample ‘ was less dianagnetic than would be expected for the pure asterial. The work of Garner, 9; 3;." showed that this adsorbed oxygen is stable to evacuation at rooa temperature but that it can be removed by baking in s vscuun at 160.0. This procedure was tried on several cuprous oxide samples, but in every case the sample changed in color from bright red to black or dark blue indicating that some cuproua oxide had been oxidised to tin black cupric oxide. An alternative seth for removing the adsorbed ongen hon cuproua oxide”. was. to treat the sample with carbon monoxide; the reaction, ‘I J. 1. Baker Chemical co., c. I». grade. -11.}. which involves oxidation of, carbon monoxide to carbon dioxide, was said to be completed in about 30 minutes at roon temperature. Treatment of a cuprous oxide sample with carbon monoxide at room temperature did not cause an appreciable change in the magnetic susceptibility of the sample, however, so treatment at 100°C. was tried. The ample was evacuated to 10.. an. at room temperature, carbon monoxide from a cylinder* was intro- duced until a pressure of 200 an. was reacted, and then the temperature was raised to 100°C. The reaction was allowed to proceed for three hours, then the carbon dioxide was renoved by evacuation and the sample [...4 ‘Fwwfl-WF— “*1 A. V -..' - . . .- . A "1. . was cooled to roan temperature . The magnetic susceptibility of this sample is recorded in the following section. A copper analysis of this sample was perforaed by the standard icdoeetric procedure 3" the average of three deterainations showed that the cuprous oxide ample contained 85.16 : 0.0“ copper. The theoretical ceppor assay for cuproua odds is 88.821. The low value found for the copper analysis is aoat likely due to tb presence of hydrated water. ‘ Cupric oxide was prepared by boiling a solution of cupric sulfate which had been lads alkaline. In 500 ll. of delineraliaed water was dissolved 25 g. of copper auliate pentahydrato. To this was addedtwo grams of sodiua fluoride to prevent the precipitation or occlusion of any traces of forroaapetic iron present. Cuprio hydroxide was pre- sipit ated by adding sodiu- hydroxide solution until the copper solution was distinctly alkaline. The solution was then boiled and stirred *Hatheson cm, Inc., Joliet, 111. -m- continuously for 30 minutes until the precipitate turned black and began to coagulate. The cupric oxide was filtered by centrifugation and it was washed five tines with denimralised water. It was then collected on a watch glass, dried for 2.5 hours at 135°C” and ground in an agate nortar. Pralininary detorlinations of the nagnetic susceptibility of this ‘4. hr ‘4‘ '1 saaple slnwed it to have a susceptibility of +3 .75 a 10", compared with the value of +3.20 3 10" given by Ishiwara.‘“ mobbed-m found a“: yen-(M‘s! ‘12. sis (.1 that during its preparation cupric oxide was able to adsorb large quan- tities of gases, particularly nitrogen and oqgen, but that a satis- i factory way to renove these gases was to heat the salple to 300°C. for one hour. Therefore this seaple of cupric oxide was placed in a covered silica crucible and heated to 300°C. for one hour in a amffle furnace .' The sample was then returned to the susceptibility tube and heated at 100°C. for 2.5 hours at a pressure of approainately l a 10'. an. After the aametic susceptibility of this sample was detersined approxi- aately 0.11 g. anplea were weighed and dissolved in dilute sulfuric acid. These were analysed iodonetri‘cally in the same calmer as were the samples of cuprous! oxide and the average of three analyses gave "18.53 : 0.09% capper. The theoretical copper assay for cupric oxide is 79.891 and the in van. obtained was lost likely due to water of hydration or incompletely decomposed cupric muroxide . The procedures for sating nagnetio susceptibility neaaureaenta were the sane as those described in Part I of this paper. “an. n numb 0o., Chicago, 111., 110491 a 10.1. Results and Discussion The aagnstic susceptibilities of metallic capper sanples which lid been repeatedly oxidised and reduced were always essentially the same regardless of how often this cycle was repeated. The nagnotic suscepti- bilities of several aanples of penulated metallic copper as functions of nan-us field strength and detersinsd at three different tenperatures are given in Table an. The values at ths highest and strength are TABLE mm m SUSCEPIIBILIT! w COPPER AS A FUNCTION 08' FIELD STRI'BCI'H AND TWERATURE Tompgrsture 60 gégld Strength 1; X. Oersteds 9 Oersteds 10. 70 Oersteds 1.: ..m 06 _ 298 -0.096 -0.087 43.085 .083 .082 .080 .0814 .081 .080 .085 .081: .082 .070 .075 .075 .070 .075 .077 A'”m e 0 Coco, 192 -0.077 ~0.079 43.081 75 43.076 -0.075 ~0.077 .066 .071; .073 .079 .082 082 Average 3.577 3 0.003 the nest reliable because the changes in weight were greatest and because the effect of ferromagnetic impurities on the susceptibilities was the least. -1h6- The magnetic susceptibility values obtained for cuprous oxide are shown in Table mu. It is possible that some or all of the temperature dependence of susceptibility for cuproua oxide may be due to adsorbed muse-n. mam xxm -.. i: manure suscmzsmr cr cameos onus AS A FUNCTION or HELD smears AND immense Z 3: '5 Field Strength ' 0 Temperature Oersteds 227 a. A 1920:. '1fo Susceptibility 5:1 x x 106 ' “~- 7,360 " ‘ -o.191 -0.18h -0.166 9,060 .o.19s -0.18? -0.169 10.h70 -O.l96 -o .188 .0.168 The aagnetic susceptibility of cupric oxide is shown in Table nun. The values obtained at room temperature are in accord with those of Ishiwaram' (see Table XIII!) but at low temperatures Ishiwara's values are sonewhat lower . The aagnetic susceptibilities of several angles of copper which had been oxidised to various extents were determined at two or tlu'ee temperatures. The resulting mixtures contained in general aetallic copper, cuproua oxide, cupric oxide, and oxygen. The gran susceptibili- ties of the sutures could therefore be expressed by the am of the susceptibilities of each constituent wultiplied by the weight fraction of that constituent present, that is, TABLEmVII HACKEIIC SU33H’TIBEITI a CUPRIO OXIDE AS A I'UHUI‘ICN W FIELD STRENG‘H AND TWERATURE W Field Strength 0 Temperature 0 Oersteds Joe x. 195%. j6 x. Susceptibility I x 10° 13,780 *3 .21 +2 .85 *2 .21; 7 ,360 +3 .21; +2 .81 +2 .32 9,060 +3.25 +2.81 +2.30 10 .1170 +3 .22 e2 .82 +2 .29 Average +3 .23 e2 .82 e2 .29 xmixture ' P01! 10:: " l,tlu.0 I01130 + Pf5u0 xCuO ’ PO. 10s (13) where P - weight fraction. If the susceptibility of each component is known then there are spparohtiy four unknown factors in equation (13), namely, the weight fraction of each component. Actually there are only three unknown factors since the sum of the weight fractions lust equal unity. Since one of the unknown weight fractions can be calculated fron equation (13) 1: all the others are known and the susceptibility of the mixture is known, two additional neuurenents amat be aade on the system to solve the equntion completely. These measurements were of the total amount of oqgen adsorbed and of the magnetic susceptibility of the mixtures at 7’1. The weight fractions of each component in the oxidized copper system were calculated by a method of successive approximations . From the susceptibility values of the mixture at two temperatures the amounts of omgen and cupric oxide present were calculated, assuming that the - weight fraction of copper was unity and that the contribution of cuprous oxide to the susceptibility of the mixture was negligible. Then by sub- tracting from the total amount of ongen present the amount calculated to be present as adsorbed oxygen and as cupric oxide tbs amount of " 'ouprous oxide was calculated; the sum of the weight fractions of these A components was subtracted from unity to give the weight fraction of copper. Then taking into account the contribution of copper and cuprous - a s. ..g \nt-*w no runes).(-. LX‘? ‘. r _ .s oxide to the susceptibility of the mixture the aacunts of adsorbed oqgen ..--..r md cupric oxide were reocalculated. Then the uounts of copper md cuprous oxide were again calculated, and if different from those obtained earlier the cycle of calculations was again repeated. The magnetic sus- ceptibility of ongen as a function of temperature is given by Selwood." The results of these measurements are presented in Table mun. The last column of Table mm is the susceptibility of the mixture calcu- lated from the weight fractions and susceptibilities of the cosponents (equation (13)). 1 Samples 1 through 5 of Table mun represent the oxidation of a fresh copper surface. for all the other angles the starting material was copper which had already reacted with a knots: amount of oxygen; these samples were first evacuated and than equilibrated with oxygen to allow the adsorption of additional amounts of oxygen. The oxidation of all these samples except numbers 10 and 11 was carried out at room temperature, but the two sup1ss were enema st 100°c. Capper was first .119- «no.0. «8.0. E. «8. 0. $40 «30.0 ...: 8.0 8. « 80.0. 0% d" 20.0. «8.0. E. 30.0. «d...0 008.0 3: 8.0 8.” 48.0. 0% 0a $0.0. $0.0. E. $0.0. 2.0.0 «80.0 «... «0.0 $0.0 $0.0. 03 m $0.0. $0.0. E. 80.0. $0.0 38.0 m... n06 3.0 80.0. 0%. 0 A30.0. 30.0. E. 80.0.. 30.0 080.0 0.0 as 3.0 000.0. 0% .. «00.0. «00.0. E. 8.0. «8.0 080.0 H. « 00.0 3.0 «8.0. 0% 0 80.0. 80.0. m« 08.0. 20.0. 03 «8.0. «8.0 080.0 «A 00.0 8.0 «8.0. 0%. m $0.0. 08.0. ml. $0.0. «.806 38.0 0.0 84 80.0 08.0. 0% a «8.0. «no.0. ma. “3.0. 30.0. 03 $0.0. 000.0 m«8.0 0 0 84 340 $0.0. 08 n $0.0. $0.0. E «00.0. 08.0. 03 «8.0. 88.0 38.0 0 n 84 80.0 «8.0. 08 « «00.0. «00.0.. m0. $0.0. $0.0. 00H 80.0. 00.0 28.0 m« H SJ «8.0 «00.0. 0% a OOH N R OOH H s “w comb—.0 decoy can!” lea-“ugh on: alumna go BE gm .3 gas 035 ggh reduced with hydrogen at 100%., sample 10 represents the oxidation of the clean capper surface, and sample 11 represents an addition of the oxide layer started by sample 10. For each of the partially oxidised copper unples there proved to be a unique solution to equation (13), indicating that all the factors contributing to the magnetic susceptibility have been accounted for. In those cases where susceptibility neasureaenta were made at 196°l. it was possible to calculate what the susceptibility of the nizture should be, and agreement was within the experimental error. Since the nametic susceptibilities of the aixtures can be accounted for in a satisfactory manner froa the susceptibilities of the known constituents of the mix- ture, there is no apparent surface nagzetism nor change in the suscepti- bility of the copper upon the worption of oxygen. The treatment of the nagnetia susceptibility data presented here stfl'l to be a satisfactory wq to analyse the products resulting free the oudation of capper without recourse to chemical methods. Particu- larly interesting is the sensitivity of this nethod to small nounts of adsorbed ougen. This would seen to be of interest in the study of cuprous oxide senioconductors where small amounts of oxygen are very important . The accuracy of the determination of the components of the mixture is estinated to be about 105, so that this netted is most useful in low cementration ranges. the limit to the accuracy is in the aagnetic susceptibility deterunations which are accurate to about three percent. The sources of error are chiefly the lack of uniform packing of the -151- samples in the susceptibility tube and the inability to get completely unifora and reproducible samples . The composition of the oxidized film on the surface of capper, as shown in Table man, is interesting. It seems that certain amounts of ongen and of cupric ends were built up on the surface rather readily at roan temperature, but as the oxidation proceeded the growth of the oxide layer was due mostly to cuprous oxide wkdle the amounts of adsorbed oqgen and cupric oxide did not increase very such. In fact the anount of cupric oxide present hardly changed at all. However, comparison of samples 9 and 11, which were oxidized at room temperature and at 10000., respectively, shows that while oxidation at 100°C . . approximately doubled the amount of cuprous oxide present, the amount of cupric oxide and of adsorbed oxygen increased approximately four-fold. Thus oxidation at the higher temperature appears to have favored the fixation of ovgen as cupric oxide and as adsorbed engen. -153- 5mg A study has been aade of the nature of the surface of activated charcoal, of the tamer in which electrolytes are adsorbed by charcoal, and of the reaction at recs temperature between ovgen and metallic ”PP”. The structure of the surface of activated charcoal is still largely A _ ’ um..- ant-“Ia I '. unlmown, even though chucoal' has a great natty well-known uses as an adsorbent. A review of the literature showed that the capacity of a charcoal to adsorb electrolytes from aqueous solution can be varied by ‘ if suitable changes in the temperature of activation of the charcoal. Charcoal activated in air at 800°C. has the greatest affinity for acids sod mods, while charcoal activated at 1.00%. has the sen-us ability to adsorb bases and cations. The latter type of charcoal has been of chief interest in the present investigation. Ashpfree chercoele were prepared from pure sugar and activated under controlled conditions. Analyses for the elements present were performed and the adsorption of iodine by the chercoele was leasured under standard conditions. The charcoale were furtlnr characterised by their abilities to adsorb acids and bases. The magnetic susceptibilities of a number of charcoale activated under different conditions were determined both at recs temperature and also at 192°:. and 75°x. It was found possible by the magnetic measurements to determine the moment of oxygen adsorbed physically on the charcoal surface, since molecular oxygen is 453- paramagnetic. The dependence of the magnetic susceptibility of charcoal on temperature was attributed to this physically adsorbed oxygen, while the magnitude of the susceptibility at room temperature was believed to be determined largely by the amount and kind of surface carbon oxides. Charcoal free from ovgen was prepared and it was found to be more dia- magnetic than ovgen—containing charcoal and its susceptibility was independent of tenperature. 03“”! u.”A f1 . . ‘ I ‘ x -Al Other methods of studying the nature of the surface carbon caddes of charcoal included direct chemical oxidation of the charcoal surface I with sodimn netaperiodate, which oxidation yielded carbon dioxide. The infrared absorption spectra of several chercoele prepared as nulls were obtained, but in the region from two to seven microns only one charcoal, which contained 25% omen, showed any absorption bands; those were in the regions of carbonyl and of etIVIenic carbon-to-carbon absorption. Studies were made of the reaction at room temperature of charcoal with water enriched in the omen isotOpe o“, The water reacted with the chercoele to fern surface oxides which were removed from the charcoals at high temperatures as carbon dioxide and carbon monoxide. In a stub of the nature of the adsorbed phase when electrolytes are adsorbed on charcoal, tin aagnetic susceptibilities were determined of salts of iron, cobalt, and manganese adsorbed on charcoal. The mag- nctic susceptibilities of these salts were found to be the same in the adsorbed state as they are in the pure crystals. Pro. magnetic measure- nents at low temperatures, these salts in the adsorbed state were found to obey the Curie-Vein law for pneumatic substances. It was concluded -1514... fron these results that the salts were held on the charcoal surface by bonds between natal and oxygen atoms, not directly by bonds with carbon atom. An adsorption process was postulated in terms of an ion exchange reaction in which hydrogen ions are removed from chmoal to the aqueous solution. The surface oxidation of granulated metallic capper was studied at roan temperature and at 100°C. by the nath of nagnatie susceptibilities. It was found that the reaction could be accmnted for magnetically by the formation of cuprous oxide , cupric oxide, and by the presence of plvsically adsorbed oxygen. tron independent determination of the total amount of oxygen adsorbed and from thematic neuureuents at roan tenperatm'e, 192°:., and 7ft. the aaounts of each of the constituents present was-calculated. The gr- susceptibilities of cuprous odds nd cupric ends were also determined at the three temperatures. 4.55- APPENDIX flu, elitative Anslnig of Charcoal Ash Two samples of comercial sugar charcoal (Riser and Amend) were ignited to determine ash content. Heights of charcoal taken were 1.563? and 1.1029 s., aid the ash contents were 0.0678 and 0.0652 g. giving, respectively, It.” and 1:36} ash. . the ash t"... difficultly soluble in cold water and the solution had a basic reaction to litms. It was almost completely soluble in both a-. 's a‘ dilute nitric and dilute Indrochlorio acids. lo precipitate was formed with main chloride solution, nor with hydrogen sulfide in either acidieorbasio solution. Tbisindisatedtheabsancesfallionsof Groups I, II, and III of the qualitative analysis scheme. A precipitate was forum! with “on!“ carbonate in basic solution. 2 a—po..- :-_F.“hlr_fv.:_~f_q i Darius and strontiu- ware found to be absent, and nagnesiu and calcius present, using the chrome-sulfate-exalate procedure for Group IV cations. Ilene tests were used which showed the presence of sodiun and the absence of potassiun. A precipitate of chloride was not for-ed with silver nitrate, but a voluuneus precipitate of sulfate was foreed when bariun chloride solution was added. Special teats rei- iron (11) and iron (:11) were nude with amoniun thiocyanate, potassiu- ferricyanide, and potassiu- ferrocysnide . There w- no evidence of any trace of iron present, although the presence of a small mount of fem-agnetic inpurity was indicated by osmotic ‘ measurenents . ~156- REERMCES 1. r. r. n. Rhead and a. v. wheeler, J. Chen. Soc., 2161, (1913). 2. r. Debye and r. Scherrer, rm. 2., g. 291 (1917). 3. u. Hoffnan and D. Wilson, 2. nek'trechen., 93.. 50). (1936). h. r. frendelenburg, latureissenschaftsn, 3;, 17} (1933). s. a. r. stricuand-conotahie, Trans. Faraday sec.. 35. 1071. (1938). 6. a. n. Lowry and o. 11. Matt, J. An. choa. 30s., 5;, 11.08 (1920). 7. a. achiiew and a. fscmtow, z. physik. Chen... 5133, Id (1929). a. s. schiiow and x. Tsolnutew, £15.. M, 233 (1930). 9. l. achilow‘ h. achatunowekaja, and I. 'rseb-utow, aid" 11112, 211 1930 . 10. r. E. Bartel]. and E. J. Miller, J. An. chee. scc., 14,, 1866 (1922). 11. s. 3. arm and o. s. dexadt, rancid-2., g1. 1.1. (1929). 12. a. J. Killer, J. A... on... sea.. 99, 1150 (192).). 13. E. J. Miller, J. 2m. comic ch... 29 2967 (1932). 11. a. amt-1- and A. Prairie, 2. pro-1k. cm" 9311,. 219 (1929). 15. s. a. Iruyt and c. s. deladt, renew-2., 3;, 21.9 (1931). I 16. i. Frunkin, rabid-2.. 2. 123 (1930). 17. a. lucidnshy, a. Burstein, and A. W, Acta lesioochil. 0. R. 3. 8.. _1_2_, 795 (191.0). 18. B. Steenberg 'Adsorptioa ad Inhange of lens on Activated chm-cod; Ahquist and Wiksells, opp-m. 19th. 19. J. n. wiiaen and r. a. man, J. colloid aei.. 5, 550 (1950). 20. A. use. J. on... 3a., 996 (1938). .157- 21. H. R.(!rws; and 1‘. but, free. Acad. sci. dusted-den, 18., $70, 193 e 22. 0. a. 'erstraete, Iatuurw. fijdschr., 9.1.3.: 107 (1936). 23. 11. Bakh, Acta Physicochil. u. a. s. 8., _1_1._, 1163 (191.1). 21.. A. King, J. one... 800., 889 (1935). 25. s. s. Bhatnagsr x. a. Ratlmr, and r. L. Iapur, Indian J. Phys” 3. 53 (1925). 26. l. I. Iobosev, V. I. Evdoldnov, I. A. Zubovieh, and A. I. Hel'tsev, Zhur. Pia. this... 33. 13349 (1952)) G. A., 21, 5711511. 27. r. w. Selwood and a. 8. Dallas, J. 111:. Chen. 80s., 19, 2115 (191.8). 28. P. V. Selwood, Bull. soc. chin. France, 1189 (19119)- 29. A. Boutaric and r. Berthier, J. chin. prom, 32. 729 (19112). 30. 1. Boutaric and r. Berthier, Conpt. rand.. 335. 62 (191.2). 31. 14. Goldsaith, J. Chen. Plus” E.» 523 (1950). 32. R. Juzamns Lubbe, and L. Beinlein, z. norg. Chem, 328, 105 19 9 . 33. a. Jusa, a. hash-1a, and 1!. Hahn. Anus. one... 51. 357 (1938). 31.. a. 4m. mm. ch»... 62.. 251. (191.8). 35. R. Jusa, and l. fentschert, 2. anorg. Ghee" 332., 165 (1950). 36. L. 3. Bennett and J. 3. Warts. 3. PM. 01's., 2,, 23110-9”). 37. O. Oocrty, Pres. Intern. Conn. Pare and Appl. Chen. llth. Congr., London. 357 (1938). 39. G. 0mm. ibid.. 332, 7115 (1950). 1.0. a. x. smtain and s. 3.1111)». thur. no. In... 33, 113 (19119); as Asp k2, M3761). 1.1. w. a. l-ewis and A. a. hot-er, Ind. Eng. Obel., 56,, 81.9 (1951:). ~158- 112. P. n.(van 3m- Lay and J. r. Wibaut, nee. trav. chin., 21, 111.3 1932 . 1.3. r. r. Bertha and J. a. Walton, J. rm. on»... 1,1. 329 (191.3). 1.1;. W. W. Scott, “Standard Hetlnds of Chemical Analysis ," Fifth Ed. , D. Van lostrand, New Iork, I. 1., 1939. 1.5. 11. c. Chirneide, l. J. 01111:, and o. a. c. Proffitt, Analyst, 1g. 35h (19117). 1.6. J. £6.18“. and I. Iarplus, Ind. M. 011..., Anal. Ed., _1_8-, 191 19 . 117. P. w. Selwood, 'Kagietochenistry,‘ Interscienoe Publishers, Inc., '9' tort. '. 1., 19h.)- hB. R. E. VanderVennon, Unpublished H. S. thesis, Michigan state College, 1951. hge Ae m, Jo ch'. 8°09, 1,489 (1937)e 50. W. F. Hillebrand and o. E. F. Lundell, aApplied Inorganic Analysis) Join Wiley and Sons, Inc., flew fork, N. 1., 1929, page 150. 51. 1!. H. Willard and I. E. Furnan, "momentary "uantitative Analysis ,“ Third Ed., D. Van llostrand Go., Inc., New fork, I. 1., 19110, DIED 273. 52. r. 11. Gas, J. Indian Chen. soc.. .23.. 38 (191.11). 53. r. w. sawed. aha. am. .19.. 70 (191.6). Sh. handbook of Gheeietry Id musics, 35th 3d,, Chemical Rubber Publishing 00., Cleveland, Ohio, 1953. 55. 1'. Corr-1a. Capt. rend" :91. M6 (1935). $6. 8. Hey-r. Ann. ply-13., 99,, 330 (1899). 57. A. ling, 3. Chan. '00., 18112 (1933). 58. e. r. saith, 'Psriodie Acid and Iodic mid," Third 3d,, the 0. r. unith 0heaica1 co., Coin-m, Ohio, 1950. 59. Professor 3. O. Speck, privste eouunioation. 60. F. reigl, 'Qualitative Analyses by Spot rests) third Ed., Elsevier Publishing 0a., low York, a. 1., 191.6, page 397. 61. r. 1. Mine:- and 0. a. Wilkins, Anal. 0hen., 3.1,. 1291 (1952). 62. H. (ll-In, 'm'ganic Ghenistry " Volume III John Wiley and Bone, Inc., see Ice-k, s. 1.. 191.0, chapter . 63. a. B. Anderson and P. n. Matt, J. Flo's. colloid Chen" 2;, 1308 (19117). 61.. n. L‘s-.153, 0. Pierce, and c. 1). Joel, J. Phys. chain... 18.. 298 19 . 65. s. Heller and r. r. reung, J. All. on... see., 19, 1.155 (191.8). 66. n. L. Shriner and R. 0. Neon, 'fhe Systematic Identification of Organic Col-pounds ' 1hird 9d,, John Wiley and Sons, Inc., now rot-1:, s. 1., 191.8. 67. )4. 15‘??? and R. E. Vanda-Venus; J. Am. Chem. Soc., 12, 1715 9 3 e 68. E. c. Stoner, magnetisa and Hatter,‘ letm and Go., Ltd., Mn, 1931‘s 69. L. Pulling, 'Iature of the Obnical Bond ' Second Ed" Cornell University Press, Ithaca, l. 1., l9 0. 7o. 2. w. Selwood 2. 2. Phillips and s. Adler J. Am. Chen. see. 16, 228211951). . ' ' ' 71. 2. r. Tyleoote, J. Inst. Hotels, 1.5;. 259 (1950). 72. c. c. Grusan and a. A. may. J. ipp1. 21”., 11., 631 (19110). 73. J. A. Allen and J. U. litehell, Discussions trolley 8011., g, 361, (1950). 71.. c. i. nuiaen. 21:11. he... 11, 96 (1931:). 7S. 0. 3. Eva. bass. linemen-a. 800., 92, 335 (19113). 76. o. 11. Preston and 1.. 7.. siren-haw, run. he... 32, 706 (1935). 17. 1. 3.84113; and h. I. Garner, Trans. neotroehea. Soc“ 9;, 30$ 19 a 78. n. song-$3311 n. A. Richards, Free. Plus. see. (London). a, 256 ~160- 79. P. D. Denkov end D. V. Ignetov, Invest. liked. In): S. 8. 8. R. 00861. this. Ink, 231: (19119); c. 1., 1.3, 981.83. 80. J. 8. Anderson, Means-16m Feredq 800., 8, 362 (1950). 81. 1'. x. Rhodin, J. Am. Chen. 868., lg, 5102 (1950). 82. s. Cebrsre end I. r. 30%, napu. prop. ”we” 13, 163 (191.9). 83. R. )1. Doll, 1. i. ”one, end P. r. 1110:, hens. Isl-eds: 800., £12, 195 (1953). 81.. J. Bsrdaen, n. w. mun, end 7:, Shockley, J. Chen. Hum. 23,. 711. (191.6). 85. W. E. Gsrner, I. 8. Stone, and P. I. Tiley, Proo. Roy.800. (London), 1211, 872 (1952). E. Csnpbell end 0. I. Tholee, Trans. Electrochon. 800., 2;, 623 (19h7). 87. w. E. Gernor end I. a. Stone, Ieture, $28.: 909 (191.6). 88. C. E. Reneloy, J. In“. Hotels, 65, 1117 (1939). 89. W. E. Gamer, '1'. «I. Qty 300., _8_, 21.6, (1950). 90. x. 1.68m end a. Seohse, Ann. Plveik, _1_1_, 727 (1931). 91. 7. Zhnse end I. 7. Kurt-chm, mm. 2. 86:362., 2, 1.53 (1932), 0. 1.. 31, 2360. 92. a. E. White, “Introdustion to Atomic Spootre,‘ nearer-3111 Book 00., New Ian, I. 1., 1931;, page 82. 93. Inna-notional Critioel Tebles, liberal-Hill Book 00., 1110., New Ion-k, s. 1., 71, 1930. pes- 35):. 9h. 1'. s. Entohineon end J. Rookie, letnre, 152, 537 (191.7). 95. c. 0. 11011173011017, mm. 301., 36, 1.98 (1930). 96. s. s. Bha‘bnsger u. x. 7cm, and n. Anew-Esq, “11618.2" 19,, 9 (1937). 97. s. a. 12:0, pm. mm Am. 361., g, 21.9 (1935). 86. W end I. 8. 300m, Dieoussions Fendey .161- 98. r. x. m. m r. 9. cut-at, mm. an... 39, 151.7 (1931). 99. I. E. nan-u, 1m... 31, 6311 (1931). 100. s. 3.(Ihets)1s¢er end I. 0. litre, Current 801. (Indie), ,1_, 31.3 1933 . 101. w. 11-- u 11. 80mm), 2. an. 011..., 233, 101. (1931). 102. n. 1.. than, J. 801. m. lsseeroh (Indie), 9;: 1. (191.7). 103. 0. Chane-1, J. phru. rum, 8». h, 9, 163 (1910). 1011. I. 811-611, 3213.. Ber. 6, z, 137 (1928). 105. !. Ishim, Science Ropes. 261nm Imp. 111.19., 3, 309 (1911.). 106. P. messy“ end L. 1.. tools, Discueions vadq 800., 8, 222 19 . 107. r. nan-.1, Am. rm. 39. 1:67 (1937). 108. W. E. Corner 2. J. (rev ad P. 8. bone Pm. 307. 800. 8121, 301 (19119). ' ' ' 109. r. w. W, J. n. 011-. 800., 39,, 701 (1898). .162. MAY1q'55 JAN 1 5 ’5]; WHY Um mm? 118111133 Thesis c.2 Vandcr Vennen, R. 3. Studios of surface phases of charcoal and copper. Ph.D. 1954 ! Thesi I CW8??? no" 0:7 Vander Venncn, B. E. Studies of surface phases of charcoal and copper. Ph.D. 1954 [high-oping. Name. ISSUED TO 1 -. ~7~1~~~ ‘7' 7.”? Au? be..'..'.;‘1.k.,‘5 a"? 11.1. M71117111111111]matilljfljlilflllflilflflyiflffli‘lWI“