1"cm " - WI ! ‘ L I“? WWII" * ”I ’l m—s {I \JN I coco ‘ \ ? HTHS .5". SEES": {34" WEE“ CWSZfiN‘? SGEEE‘TSSS; GXIEA'ESGN ('2? $25,355: EN PQTASSEUg-‘vfi HYQROKEGE SGLUTEO‘N E‘hcsés felt the» @9539“ a? 53.5. S. Méfifiifikk YEA‘EE Ufii"§§$§?¥ é‘aamaré A; Vamcéasr Egg? E96552 iTHr-rsvs L [B R A R Y Michigan State Univcr ty A SW 0! THE WSW? PO'IENIIAL QXIDA'HG OP SILVER IN POTASSIUM HYIROXIDB 9011mm By W A. Vanda- Lugt AfliESIS Submitted to th- comm of Selma and Arts a: Michigan 5m. University of Agriculture and Appllod Schne- in Partial fulfillment of the ”quiz-cunts fur the dcgr-o at man OF SCIENCE Depart-ant. of (mulch-y 1960 WWW“! the writer viehee to cups-en hie eincere appreciation to 2r. 1‘. Diane at Celvin College end h. J. me for their guidance end help “momma! mums-mu marinara-1M {er happening this project. ABSTRACT The {oration ot 19,0 end Ago was studied by news of the constent potentiel amidstion of e silver electrode i 8313:3331... hydroxide solu- tions. the renge or cummtiom used ves from 8 percent to h2 percent end in eeeh solution, volteges spanning the range 0.05 to 0.1; volts shew the A910 reversible potential were used. The constent potentiel eethod see utilised in en ettenpt to clerify the kinetics end eecheniees or the electrolytic {oration or the oxides of silver. The efficiencies of the electrode processes under cousihretion Iere celculeted by determining the weight geined by the electrode end then coepering this to the quantity of current used. It ees round thet the oxide toreetion processes were appreciately loo percent efficient under ell auditions except those st which men as evolved. has re- lotion between the hydronids ion concentretion end the potentiel required for oxygen evolution was elso noted. The extent of the conversion or A9 to 11920 was studied under e ur- iety or coMitiom in en ettenpt to deter-ins the ochre of the oxide swine end its effect on the fornetion of 190. The results indioete thet it is not possible to reelin 100 percent cmversion of silver to AgpdetflumsmoeotefileotMponmelectrode decreeses the initiel rete of Ago torntion. meshepeotthecurrent-tiee curves obteinedeteechconditienoi' voltege end potessiue hydroxide omoentretion ves given speciel ettention. Mmsverethmoonperedtotlntinoreticeilydetereinedcurves for the motions being midsred. The coepnrieons shoved thet there 2 use egreseent between the theoreticel end uperisentel curves et 0.05 volts only. At ell higher potentials the presence of commotion: ow- rents ceused the reection rete to deviete from the theoretioei. On the besis ol‘ the inforletion geined from these studies some euggestionsereeedeesto thenetureortheelectrodeprooesees in- volved in the electrolytic toreetion of A930 end Ago. IABLB OF “THIS nmmnnwOOOQOOOOOOOOO... mmOIOOOIIOCOOOOOOO RESUL‘IS. .QOCOOOOOIUOO'COOO. Efficiency of, the Electrode m...” . GOM1OROIAQWAQIOeeeeeeee Potentiel for the Aglo-Ago hunter-each Anelorsis of the Current vs. Hilliampere Hour Curves O O 'mdnnimutuumeimpr Agpmlgo. . WOOOIOOOitijibpttotOilOOO APPENDIX WCES eeeeeeee'eeeeeeeeeeeee 0 more» 10 15 16 21 23 25 lemma: A General Review of Silver and Its Oxides Silver is e umber or the Group l-b eetels which ere chenoterieed lay their low chenicel ectivity. ‘ihe generel outer electron configure- tion is (nondmns1 end this pereite the renovel of eore then one else- tron since the energy differences between the ‘(n-l)d end ns electrons ere not very lerge. The cherecterietic oxidetion state of copper is two: for silver it is one end for gold it is three. iiwever, in eddi- tion to the show eost steble oxidetion stetes, it hes been shown thet ooepomds corresponding to ell three oxidetion stetes heve been cherec» terieed i‘or eech of the coinege eetels. Thus it night be expected a... there would be three oxides or silver corresponding to the eepiricel {armies 19,0, Ago end A9303. , . A review of the litereture indicetes thet there is definite evi- dence for the ulstence of the eononlent end bivelsnt oxides of silver. However, thereseens tobesoeequestion ebonttheveliditaretthe interestien used to show the existence of en acids or silver conteining trivelent silver. In the peregrephs thet follow the einrecteristics, eethode of preperetion, end the evidence for the velencs stete or eech oi‘ the oxides of silver will be indiceted. A910 is e oovelent compound which crystalline in e fees-centered cubic lettics. It is cherecterieed by its eeee oi‘ ther-l reduction end its low eolubilitar. lt exhibits both elkeline and ecidic properties; its equeous WIN! ebsorbs cerbon dioxide end its solubility in 2 elheline solution increeses with increesing hydronorl ion concentretion(l). lgao one be prepered by both diuicel end electrolytic eethods. The chenioel sethod involves the precipitetion of 19,0 by treeteent of nono- velent silver ion with eliteli eetel hydroxide solutions. ‘ It cen be propered electrolytieelly by enodic treetnent of silver in elhelins solutions. The OW oxidation potentiel velue given by Letieer tor the Ag—Agzo couple in elkeline solution is «0.31:1; volts. l‘egreetdeelodworhhesbeenreportedonlgo. lheclessioworh on the higher oxides of silver use done by Noyes, 35 L1, Their studies or the oxidation or silver nitrete in nitric ecid by oeone indicete thet the oxidetion stete of silver in the block comound produced is bipositive. In edditlon, the feet that on electrochemicel reduction tee Modem or electricity ere obteined per greo eton of silver would seem to be conclusive evidence tor the divelent stete oi‘ silver in this coepomd. The eegnetio dot: on Ago .1... not give definite evidence for the velenoe etete of silver. The diveient stete of silver would be diluted to exhibit mantle properties but voiding end lieeernovshi(2) ‘ report thet solid Ago is not peremgnetic. However, in nitric ecid solution it is pereugnetic. This permgnetiss her not been observed in elheline solution. The romle or the oxide indicetes e lel stoiche ioeetry but there is oviducs thet this my not be neoessery. DirkseU) tomd thet prolmged enodio treeteent of silver in potessiuo hydroxide solution produced oxides whose compositions veried from A90, to £90,... Detenineticn of the aidetion potentiels of these electrodes shoved thet cm W to the potentiels for the divalent mm of 3 silver. Letieer's velue {or the oxidation potentiel of the AgZO-Ago toque in slimline solution is .057 volts. A90 is less steble then mo end in the presence or silver it will slowly decompose to fore the lore stebls AgzOUi). Ago cen elso be prepared by chenicel end electro- cheeioel methods. The chenicel methods involve the oxidation of Ag or silver nitrete by oeondS). pmgmotesw), and persuli‘etes(7). Ago cen be propered electrolyticelly by the oxidetion or silver nitrete bettnen pletimn electrodee(8) or by the enodic treetieent of silver in elkeline solution(9). The evidence for the existence of the trivelent acids of silver is less conclusive. Jirse end Jelinek(10) reported the {oration of Agzoa by «tidieing silver (I) oxide with em. However, the mound wee very unsteble end decomosed repidly in weter. Hickling end Taylor(ll) maest- thet the winery product of the acideticn of Ag.0 is A9303, which subsequently deomiposes to fare A90. However, it hes been suggested thet there is no conclusive evidence for the trivelent oxide of silver(l2). All eveileble dete can be setistectorily expleined by entering thet the highest aidetion stete of silver in ell of these processes is e posi- tive two. Our work is concerned with the electrocheeicel properetion of the oxides of silver in potusitn ivdroxide solution. the nmlent end divelent crideticn stetee of silver ere epporent when e silver electrode is treeted enodicelly in elkslins solution. At potentiels below 0.35 volts shove the Ag-dgzo potentiel Agzo is for-ed end et higher poten- tiels the product is AgO. However, there ere sole problems connected h vith the enodic treetesnt or e silver electrode in potessiue hydroxide solution. 'ihe neture end mechanics of the electrode processes ere still open to question. Also, results have been obtainedCl) which indicete thet the conversion of silver to silver(1) oxide in potessiue hock-oxide solution is never couplets. The reletion between the cmcentretion of mm: ion end the efficiency of the process is meant obscure. It is the purpose of our work to etteapt e cleriticetion of some of the problsls cited above end possibly make suggestions es to the netm‘e end eechmisns of the motions involved. Host or. the writ on the fomtim or the oxides of silver hes been done using constant eta-rent techniques. However, the electrocheeioel kinetics of reections ere Motions or overvoltege tether then mt densitvibh Therefore, the use of constant potentiel methods cen re- veel some espects od’ reection himtics which cennot he obteined through the use of constent current esthods. At content potentiel e perticuler reection on be chosen for study, vhsrees et constent current the re- setion thet is kinetioelly cost fevored may teke piece. ‘ihis is per- timlerlytrueviurethegrcvthotenatidetileeeyrmdertheelectrode pessive. the rete of en electrodtesicel reection carried out st coast-ht potentiel is dependent on e teacher or rectors, such es, concentretion polerisetion, the electricel resistence of the systee, the overvoltege, end the retes or lineer diffusion of the ions involved. Harever, in lost electrocheeioel reections this is somewhat sieplit‘ied, beceuse the lieiting rector is either the overvoltege or the rete of diffusion. Before e abstenee can reset et en electrode surfece it met overcome 5 en energr barrier. This energy barrier in en electrochemicel process is the overvoltage. The magnitude of the overvoltege can be evelueted tor deteeeining the potentiel et which the desired motion viii tehe piece end cowering this with the equilibrium potentiel, celculeted iron the Hornet equation. is the potentlel is further incl-sued, the theoreticel rete oi‘ the reection will increase very repidly. But this theoretical rete oen be etteined only if the rete oi‘ trensport of ions to the electrode surfece is equel to the damned. This is normellv not the case end thus et increasing potentiels the retee or eost electro— chemicel reections ere limited by the retes of ionic diffusion. EXPERIMBTAL ‘ihe spperetus shown in Figure 1 (see Appendix) ves designed to cherge the silver electrodes et constent potentiel. Resistencs A is e precision resistor from vhich the desired potentiel cen be epplied toeneleetrode. lnthisveyelectrodefl canbensi'nteinedetecon- eteht poteetiei ehove electrodes it end on sieeteede a is the moped-i- sentel electrode eodeespreperedbypressing eoisteggooeepletinue screen end reducing it thereslly to silver. Ileotredss c end 6' ere motively lerge electrodes prepered by pressing 19.0 on e silver screen. ‘ihementflouthroughthsosll isnsuuredhyteppingfrtn resisteeoe D end feeding into e recording potentioester. The voltege difference betsesn the eocpeeieehtei electrode end the mo electrodes is assured by tepping free resisteeos E and feeding into e recording potentisester. Althsughthe ebovedsscribedproosss is eetuelwepreeess cer- ried out et censtsnt cell potentiel, it is in effect it constent elec- trode potentiel process, for by eshing the Agzo electrode surfece very lsrge inocuperiscnvith the gross Misceereeoftheeocperieentel electrode the can-rent density on the reference electrodes will elueys beveryseell. 'ihus thelgfi sleetrodeuillbsepsretingveryclose to its reversiblepotentiel. lnsuvonssheveessi-sdthstthe veltsgss assured ms the pistes of the cell represent the potentiel difference bet-sen the expert-Intel electrode end the ‘230 reversible peteetiei. 7 ‘ihe procedure which use employed to study the cons’tent potentiel oxidetion of silver cen best be divided into two perts. The first pert of our procedure was designed to show the effect of e prior cherge on the voltege-current cherecteristics of A910 and Ago formtion in verious cowentretims of potassium hydroxide. “to socce- plish this the silver electrode was treeted emdiooiiy et 0.05 volts for forty-eight hours or until the current use core. The electrode was then soaked in distilled ate: for et nest. eight hours to veeh out the electrolyte, dried in e streets of purified nitrogen, pieced in e desiccetor for et leest eight hours end then weighed. The increese in weight ues “suited to be due to cmygen. The electrode, which now oenteined sons “,0, use pieced on cherge et.0.35 volts for forty-eight hours end the gain in weight ves egein noted. ‘ihis ves repeeted et 0.1.0 volts shove the hop potentiel. It should be noted thet vs devieted from the stove outlined sequence of volteges in eight percent potessiue hydroxide solution beceuse so little eiivei- ves converted to 1930 et 0.05 volts. 'lhesecondpertofourprocedurevesdesignedtostWtheeffect of medially cherging e silver electrode et 0.15, 0.50, end 0.35 volts respectively, in vericus concentretims of potessiue hydroxide. The eatperieentel electrode ves thereelly reduced to silver betsesn sech veltegs step. has gein of avgen end the efficiency of the process vers egein noted st seat of. these velteges. ‘ihe composition of the product on the upsrissntel electrode ves deter-iced by visuel inspection, by ohteihihg it-rey ditri-ectioe petterns with e Horelco x-rxy spectrometer, and by conpering its potentiel to that of the [£92049 electrodes . RESULTS Efficiency of the Electrode Processes The efficiency of the process was determined by assuming that the increase in weight emerienced by the silvcrelectrode one due to oxygen and compering this to the number of nilliempere hours of elec- tricity consumed in the process. It was else around that the solu- bility of the .1910 would not produce an significant loss of oxide from the electrode em‘feoe. _‘ihese results ere given in Mle I (see Ap. mix). lhe date show thet et 0.05 and 0.15 volts shove the A910 potential the enodic mtidetion of silver was approotiutely 100 percent efficient in mmtretions of potassium hydroxide between eight percent end forty percent. lhis would indicate that under these conditions the reection by which silver is converted to A910 is the only reection of em con- sequence. this reection is generelly considered to proceed eccording to the following equetiom 2A9 on.“ 19,0 +1130 #- 2? In the forestion of A930 end Ago et 0.30 end 0.5 volts sons ex- perieenteldifficulv ves mtered end the results ere not es eeey tointerpret. Hm, thereeppeers tobesproncunceddrop incur- rent efficiency in forty percent potessim hydroxide. At 0.1.0 volts the ohenge in meat efficiency with mm ion cenomtretion is very pronounced. In these ms, the electrode hed been enodieed et 0.05 volts end 0.35 volts respectively, for M hours 10 prior to the application of the 0.1.0 volt potential. However, the . prior foreation of the oxides of silver on the electrode surface is may not relevant to the efficiency of the process. Figure 5-) «a Mk I (see Appendix) shoes that treatment of a previously anodized electrode at 0.35 volts resulted in the formation of Ago with 100 per- osnt mt efficiency. Figure 2 (see Appendix) contains a plot of current efficiency vs. potassium ivdroocide concentration and shown that the went efficiency falls off very rapidly with increasing lvdroxide ion concentration. The decrease in wait efficiency is due at least in part, to the evolution of oxygen at the anode and our results indicate that at potentials of 0.30 volts and above, the re- action is favored by increasing potassium hydroxide concentration. This indicates that the potential for oxygen evolution at the Jig-£910 and Agzo-Ago couples decreases with increasing hydroxide in concentra- tion. Reference to the curves labeled '0' on Figures is, S, and 6 (see Appendix) shoes that the want densities were appmimtely the case in all omeentrations of potassium Wide. lhus the meat density factor is largely eliminated in the consideration of the differences in the potential for oiqygen evolution at the various concentrations of electrolyte. Conversion of Ag to A920 Results reported previously by Dirkse (3) indicated that -at potentials below that at which Ago forms it was impossible to convert all of the silver on the active electrode to A930. Inspection of Table I (see Appendix) reveals that our-writ verifies this contention. The 11 best conversion (approximately 80%) was obtained in 28 percent potassiu hydroxide at potentials of 0.15 and 0.30 volts. Under these conditions the mt-nillianpere hour curves (Figure 5, d and on see Appendix) shovahighcmentdensityattheanodefor some tineandthena rather abrupt drop to zero current. 'ihese ms also show that the total mmber of nillismpere hours of electricity consuned at this high current density is approximately the same at both potentials. ‘ihis offers further substantiation for the idea that the rapid drop in curd- rent dmity occurs when all of the surface silver atoms have reacted to for: A930. me rapid change in current density could be accounted for in two we. The first suggestion takes into accomt the high electrical re- sistance of A920. ‘ihe specific electron conductance of A910 is reported to be 10““ alas“ (it) and thus the formation of the oxide filn on the "face of the electrode would increase the electrical resistance of the systasandbring aboutadecrease incurrentflov. Harever, the in- creased resistance would only decrease the rate of the reaction and thus c-plsts ccnversim of A9 to AgzO should be achieved after a sufficiently long anodic name. Our results show that this does not occur. 'ihs other possibility is that the A910 layer presents a barrier to the dif- fusion of hydroxide ion to the unrescted silver. When the formation ' of the oxide film on the surface is complete the reaction ceases since there are no sore inrdrootide ions available for reaction at the site of the free silver. It should be noted that, in this neclnnism, there is no electron flow across the A910 film. The electron transfer takes 12 p13“ at the 311m - hydroxide ion marina and than an rename. of an mm mt be duo to an difficulty a: mum man-m4. ion through tho mud. lust. Th. data in Tabb 11 91V. tuition-.1 Widow. on tho (“than of “,0. (5.: mix), mu. data m obtained using an m m!- mul «animus aux-that! tuna, with tho much that thc volt- Iac [nan-nu var. lunar. An chatted. at «141m at 0.05 volts for 2141;01:0de tau-cm inmlghtwu mud. flamma- m-u nun-od- uu m arm-god .: 0.10 volu for mom- 2!. m: I!!! thalamus again «mama. TM! Wm mud» 0.05 wit. 111W. up to 0.35 volts. Innpuum of mm 11 (no Ap- ”11):th maximummuonortg “1930th um medic trutnmt at 0.10 volts. Hm, the parent of Ag «worm to 19,0 mm with an W40 tan manna. mm 7 (m Ap- ptndix) m a plot of the met“: conductivity at punish: 1mm:- Mc solutions superman“ on u plot or the parent of sum marked to 19,0 at 0.10 volts VI. flu petunia- Wlh Manual. This on» indium that flu aunt of 1930 tor-tum in I Mica of tho mama mung: a! flu 01.3mm. use, inspection of mm 11 (000W) m1. memmmuumumcmm is «mum by mum medic tut-mt for 2%; hour pct-1m at ouch 01‘ th. mum halal 0.35 V01“. Sines ths “nativity at tho che- trolyto, and not tho nativity of thc Waldo in, is involvad, thin held on: to indicate that the with?" at the 01.0erth much an dupth of tin oxide :11- on the sum theta-ado. It do“ not an pouclhhtocaphln mammmisumaMndem 13 in lower ooncentratione of lid-i, becauee if it were kinetically poeeible to ouidiee the eilver thie would have occurred. at least to acne extent, during the 96 houre or anodic treatment. A consideration or the voltage dietribution acroee the cell before and after the formation of the oxide layer at firet euggeeted a poeeible explamtion tor the above reeulte. When the voltage ie first applied acre” the cell the eilver electrode ottere little or no reeietame and therefore the full potential gradient is applied aoroee the elec— trolyte. Oxidation me at the silver electrode and an oxide file ie W. Since the oxide file has a high electrical reeietence thie torcee a redietribution of the potential acroee the cell. A cer- tain tractim oi' the cell potential then We a voltage gradient acroee the oxide layer and the eignitioant tact ie that the aagnitude or thie voltage gradient inoreaue ae the epeciflo conductivity of the electrolyte increaeee. If we now new that the oontimed oxidation of eilver involves the diffueion of hydroxide icne through the oxide layer, m:- the influence or the voltage gradient, it night he nag- geeted that. in eolutione of high electrical resistance, the difference in potmtial m not be great enmgh to overcome the reéietance of the file to the trmeport of hydroxide ion to the ten-noted eilver. nu. emanation ie at leaet ooneietent with the oheeflation that the ex- tent of the reaction ie a Motion of the epecitio condutivity rather than the activity of the hydroxide ion. Thie explanation, while tenpt- log, ie imorrect, however; because an increaee in cell potential would inoreaee the voltage gradient acroee the oxide layer, multing in the toe-uticn or aore Agp. Our reeulte ehou that thie doae not occur. 11; In fact, the electrode resiete further oxidation at all potentials up to the potential at which 59,0 is oxidised to A90. It were then, that the aide rue prothwed at 0.1 volts, regardieas a m thicimess, prevents the diffusion of Wide ion to the eurface of the active eiiver. If this is the case, the only reaction possibh ie the conver- sion of A910 to A90, at the 119.0 - potaseiua hydroxide interface. may, the utmt or silver-to-eilver (I) oxide convex-lion an be dependent on a timber of factors, such as, specific connectivity, the voltage at which the initial aide tile is forced. and the electrical resistance of 4920. Inspection of Table I (eee Appendix) reveals that. in 8 percent and 20 percut xou, after the test electrode had been amdically treated to for: A910, the application cd‘ e pctential of 0.35 volte reeulted in a less or weight, this occurred despite the fact that oxidation was taking place at the electrode as shown by the current flow, Figures 3-»: and 13-h (see Appendix). In 28 percent Kai this loes did not occur. In feet over the voltage span of 0.05 to 0.35 volts the current ei'i'iciw was 100 percent. This mt. lossei‘onvgenhadbemohservedbetoremeei‘elt that a saratentic study of the oxidation of silver in ROI-i solution as required to discover the reaction or process «sociable. Table 11 (see Appendix) oontaine a emery or the results obtained in this study. medateshcethattlnre isadei'inite lossofaeighti‘roathe electrode ae it is treated at increasineg acre positive potentials, uptothepointahereAgObegine toforn.iiovevar; reductionoi‘the electrode and much of its weight to the weight of the original 15 electrode showed that some of the silver had been lost. It was fomd that the anomt of silver lost in each series exceeded the loss eno- periencsd by the electrode during anodic treatment. A graph of the graas of oxygen gained per gram of silver vs. voltage on Figure 8 (see Appendix) indicates that the loss in weight is independent of the KG! caicentration, quantity of oxide on the electrode, and the voltage. lime it appears rather certain that a competing reaction could not be involved. Evidently there is a certain amount of cracking and flaking of! of the oxide material during the processing of the electrode and this accomts for the appamt loss of oxygen. The loss of mo could also eaqaose some free silver to the electrolyte and this would accent for the fact that some current was observed to flow during the tile tint the anode appeared to be losing oxygen. It would also follow that tin ctu'rmt efficiencies calculated for this process can only he approa- irate. Potential for the Agzo-Ago lbanstorntion Reference tol‘ablel (seeAppendix) shows tintAgO is notpaocuoec in 8 percent potassiua ivdroaide solution until the electrode was held at 0.10 volts above the A930 potential. In higher ccoomtrations Ago was termed at 0.35 volts. Since the sturdard reversible potential of Ago is about 0.26 volts above the Agzo potential, it is evident that the potential required for the AgzO-AQO transforntion is apprentiaateiy 0.1 volts above the stmdard potential and that this value decreases with increasing hydroxide ion concentration. This is the expected train! it we assun the following reaction: lg,o+2ai’ I mgowpna‘, mdexpresslihterasoftheflernstewation, a....5.-2=g22 tggggyrfit.‘ Since Ago and A910 are solids we w consider their activities to be mityandit is thenapparehtthatastheactivityofthehydratide ion imeaeesaodthewateractivitydeoreaeee, mumunnuw 3'. Justification for the assmption that hydroxide ion and not water is theactive species cahbe obtainedfroarigures 3-6, eaodf (see Appendix). a comparison of the mt densities m that the rate of the reaction increases with increasing tordrontide ion concntratim. If water were the active species one would expect the opposite effect. It‘ will be shown later in this paper that the rate mtrolling factor (inthepoteotialrangecousideredhere) is thespeedatehichthe active aaterial is brought to the etc-face of the electrode. The higher viscosity and lower water activity in the more concentrated potassiu Wide solutions would loner the rate at which water aolecules would diffuee to the electrode surface. lhus it seem alike]: that eater could be direothr involved in this process. Analysis of the Chrrent VS. Hilliaopere lour (harves Hhen the potential of the working electrode was held at 0.05 volts shove the «,0 potential (ourve "a'I in Fig. 3 to 6; see Appendix) the meat slowly decreased with tiae. However, when the silver electrode was oxidiDd at higher potentials ('d' in Pig. 3 to 6) the current re- aaimdmtantroraemalhmmumdmppaarapmytoam 17 low value. 'lhis difference in behavior suggests that different rate controlling factors are involved. Itwaepointedout in the introductim tlnttwooftheratecon— trolling factors which must be considered in a constant potential pro- cess are overvoltage and rate of linear diffusion. If we seems that the electrolytic oxidation of silver proceeds in accordance with the following equations Agoon“ -—--.~. Vaigzo+venpuf , then following the treataem. of blahay (15) we can indicate the de- pendence of meat on overvoltage by - if i - -C e if where med-l Ag Kistherateconstant,distheareaoftheelectrode,risthel'araday, Af , 4 and e M represents the fraction of molecules in the activated state. A,P,andcm-arecmstantso “if .R‘l' I l i KCAQ , and substituting which, at 25' becomes logi-logc‘g‘t 75%; + logK'. The plot obtained in Figure 9 (see Appendix) was derived by aestning that the electrode surface had a total capacity of 100 nillianpere hours and this was considered to bo CA9. Thus, after 10 ailliampere hours of 18 charge the «attenuation of Ag was mu to 0.9, and after 20 ailli~ upere hours cm was 0.8, etc. The plot of i vs. nillialpere hours shows that the rate increases with potential and at constant overpoten~ tial the rate decreases with ties. The high current rate predicted by the ctrrent~overvoltage relation- ship is seldom achieved at high overvoltage values because a point is soonrsaclxedwhere the rate ofdiffusionoftheelectrolyte to the elec- trode surface liaits the rate of the reaction. Under these conditions a concentration gradient is established when the electrode and the. body of the solution, the concentration of the reacting species being esro at the electrode surface. Delahay (16) has shown that the liaiting current in a constant potential process controlled by the rate of dif- fusion is expressed;hy l 'l/ltl/l 1 . i -nFAD./zc° vherenis thenmberofelectrons, P is the Faraday, D, is the diffusion coefficient, and c0 is comma-a- l tion. The quantity n! fl 0’ is a constant in a given potassium s57! hydroxide solution. Simpmying, 1 .. Kt'Ji/z A graph of this equation (Figure 10, see Appendix) shows that. the cur- mt will decrease with time. The above considerations would lead us to suspect that the race-- tion is rate controlled by the overvoltage at 0.05 volts and by the rate of diffusion at higher potentials. However, mination of marvel 19 'a' and u! in Figures 3 to 6 (see Appendix) indicates that w could be rate controlled by either faotor while «H mm neither of the theoretical ms. At this point it is necessary to note that in the derivation of the equation for current inthe‘diffusion controlled process it use assumed that the concentration gradient remains constant. This would ‘ not held tn. if the electrode vere held at a constant potential for a long period of tins, because the difference in density hetveen the electrolyte near the electrode surface and the body of the solution would soon cause convection currents. Laitenen and liolthoff (17) studied the effect of emotion curmts on the shape of curt-mt vs. voltage curves in a diffusion controlled constant potential process. M found that large deviations froa the theoretical were woduoed by eonveetiu meats and that the amt of deviation is depuxdent on the eriutation of the electrode in the solution. If the electrede is positioned so that diffusion can take place only in a vertical direction, and if the solution of lowest density is always at the top, convection is eliainated end the theoretical we've is obtained. However, if dif- fusion is peraitted fro: the side, contraction takes place, the concen- trationgradimtis largelyreeoved, andagreatermnnerofionsare am available for reaction at the electrode surface than would he provided by the process of linear diffusion. This results in a high constant current for a period of tine with a rapid drop in current can aost of the active anterial has reacted. In this sort our electrode ass mounted vertically in the potassiun Wide solution and the oxidation was carried out over a period of ac Zhhoursornore. Itseelsvaryprobablp thenthatthercactimcsr. ried out at potentials greater than 0.05 volts vars controlled by the rate of. transport or mm ion to the electrede smface. For the reaction at 0.05 volts it is not: quite obvious that the liaiting eta-mt is established by the ovemltage. ‘ihe fact that the comection currents which are established, have little or no effect on therateofthereactien, indicates thattherate is independentofthe woe ion concentration. ' nushapeofthe'cwmt-tise curves inthercantpotassiua invdroxide, Flame 6, e and f (see Appendix), differs aarkedly froa those observed in lower concentrations. The initial mrent density use high but it dropped rapidly to low value in a very short tine._ Analysis of the experimental electrode revealed that “.0 us foraed at 0.30 volts and A90 at 0.35 volts, but the «tent of oxide forsation use very low in each case. Since previous work indicated that the aesunt ofsgzoandAgOpa-oduced athigh mattdensityvas relatedte the porosityortotalsta‘faceareaoftheelectrode,vesuspectedthatthe high tssperatu'e of theraal reaction av have been. responsible for a decrease in porosity.» To check this. an A910 electrode was rmced at soo'c.'ud placed on charge at 0.35 volts. a high current density vss lintained for h.$ hours andthe camosition of the product tree “09.“. the secedectrode vas then reduced at 900°C. and treated anodioaliy at the sanepotential. 'ihe current densitydroppedtoa lovnlue in 20 aimtes and the product mention use “0.“. The ”iting point of silver is approxieateh' 900°C. so it is quite evident that at high t-peraturesoaeofthesilvarfusesandtlmsreducesthc surfacearea ll of the electrode. ‘ihe rapid decrease in currmt density is undoubtedh' due to the tornatim of the film of nap, which offers high resistance tothei‘lwoflvdroocide ions to the Macs ofthemreactedsilver. Hechanise of the Conversion of Ag,0 to Ago It was pointed out earlier that lgzo has a high electrical re- sistance but this epperenthr is not involved in the control of the re- ection rate in the {creation of A930, since there is no electron flee through the oxide tile dieing the reaction. However, in the conversion of 19,0 to Ago the situation is different. It we assess that the reac- tion 01‘ A910 with hydratids ion is represented by: 13.930420? —-----9 21190 4H1002e‘ , ondii'vealsoassus that hydroxide ions are available only at the surface of the 119,0, then it can be seen that electrons vill bemusported across the A930 later. the resistance or Agzo to electron flow should then limit the rate of the reaction. fiidence for this was observed. When an electrode that contained the minus amt of Ag,0 on its surface was anodized at 0.35 volts, the conversion to Ago occurred (Table II, 22.7% KOH, see Appendix). The eta-rent flow, initially, was very low end grodmlly in- creasedtoaeeximoveraperiodoi‘atevhoms, This is interpreted es evidence that the resistence of A930 limits the electron flow, but as the tomtion of Ago oontimes the thickness of the 119,0 layer de- mandthis allows mmti'loetoincreeseteanxim. lteightalsehenotedthst, while ourwrk csmotproveerdis- prove the existence or “‘03, it does indicate that AgO is produced directly from Agzo and Ag without the formation of A9103 es en 22 intermediate, as suggested by Hickling and Taylor (11). The standard acidaticn potential value given by Latinor for the Ago-A9303 couple in alkaline solution is —0.7h volts and therefore it is not likely that we could have formed any A9203, even at the highest potentials need. “GE lg! Mr “A SWY while the revolts are qulitative in nntm, it appears that several suggestions can be sade concerning the constant potential oxidation of Ag in potassium hydroxide solution. 1. 2. 3. h. The current efficiencies for the reactions which result in the flotation of A920 and Ago are 100 percent efficient at poten- tials below that at which mgen evohition occurs. Evidence has been obtained which indicates that the rate con- trolling reaction in the electrolytic formation of the oxides of silver, in potassium hydroxide solution, involves the reac- tion of Wide ion, and not veter, at the electrode surface. lhe conversion ong to 19,0 is never complete and this is due to the foraation of a film of oxide, which hydroxide ion camot penetrate, on the surface of the silver electrode. It is also apparent that the extent to which Ag is converted to 19,0 is a function of the specific conductivity of the potassium ivdroe- ids solution. It is obvious that 1930; need not be an intermediate in the foraetion of “0. we believe that this work shows that important information on the kinetics and necinnim of electrode processes can be obtained by a study utilising constant potential techniques. Also, further work on the constant potential oxidation of silver is indicated. lhis' investi- gation has shown that more reproducible electrodes should be obtained, 2); in order that a note accurate anslysis of the sin-face area factor my be eade. It would also be necessary to devise and ocperiesntal set-up which would eliminate convection meats in a prolonged electrolysis. APPNDIX 26 A WV: 4) To RECORDER D To ,chonorn 4—4 E __‘_J Figge 1 Diagram of the circuit used for constant potential oxidation. D0 :90“- 4 E 00'- '4» :~: 7» v m 0L . ha 6 50!- -1 E40}- 1} :2 w w a 20- 4L 8 'OF d O l i i 1 M 20 30 40 50 '/.KOH Figure? The relation betaeen current efficiency and hydroxide ion concen- tration for the anodic treatment of an Ag-AgzO electrode at 0.1; volts above the A920 potential. TABLE 1 Summary of Results of Constant Potential Oxidation of Silver VORZQZEEW % current coEvfigtcd W xaon Figure potential efficiency to A920 Product surface 8.25 J—a 0.05 h 70 9 1900.5; 1920 b 0.15 90 57 A00...” A920 “a? A o 9:95“ A ““9 mm “5L“- 3‘99“? A ”1° d 0.35 9h 59 - Moe.“ 591° e 0.1;0 A81; - Agofig A90. 20.7 h—a 0.05V 93 51’ Moe.“ 191° b 0.35 0 h8 “0..“ 19,0 M“: a“: :90: -A— r. 3": : rim d 0.15 100 57 100.... 191° a 0.30 86 53 Moms A930 1 0-35 # f 100 r“ _ Agoo.es_ 19° 28.: S-a * 0.05 87 63 ”0,.” * fiAgzO b 0.35 100 “- ' 1909.1: A9° m1“ a PM- u “7 m M :‘1 .._ Efrain“: d 0.15 9h 79 Moe,» M30 e 0.30 100 76 190.,“ 19,0 t 0.3% w 96‘ - . A909.“ ‘00 at.) d—e 0.05 100 50 A410,,” 19,0 b 0.35 77 "- ‘90s.” ‘90 we __ 3J0 _ A 8“ -~ _ “Sum A90“ d 0.15 100 56 A900.25 A9.20 O 0.30 6h 8 £900.04 A910 Figures 3-6 Girrmt vs. nilliaxnpere hour graphs for the constant potential oxidation of a silver electrode in 8%, 20%, 28%, and i405 potassiua hy- droxide. line comes labeled a, b, and c were obtained by anodio treat- aent at 0.05, 0.35, and 0.1.0 volts above the 1910 reversible potential. the electrode was not reduced between voltage steps. The me on the lower half 01‘ each page (d, e, and 1‘) were obtained by anodic treat-eat of tin electrode at 0.15, 0.50, and 0.55 volts respectively, with the electrode being reduced to silver between each potential increase. tok ..\.m 0' ON MC 393 .. w. Utt<fiu 4-! 0% 0.. ON _ _ O. 0.. on I L r. T l I Q P P > 2.0 00 0* 00 q fi q 1: I C P b P .> 2.0 3 ON O. 9 383dHVl1‘HH Figure 3. 30 J _ L e S and 18. NOW Om OV ON ~ /~ HI rl. u h _ h .>O+.O ”KOO..- c U¢Utt<34=s 8 0+ Ow — _ O0 O. 0? ON O~ SQHHJNVI'I‘HN Figure h. 31 @130 I I mama—1‘. 4.22 OO 0... ON Om OO 0v ON 00. ON OO O? ON 4 fi _ N u u — fl — fi _ _ .l J 1 ll 1 [W L. ‘4 y. _ P _ _ FM F _ e _ _ vF s) “”010 > oneo s) “-00 IOM .\o mm 00 0* ON 00 0? ON 00 0* ON _ _ wit/Ff i LII/q j l I 41 r]. II. n I. .1 TI 1 rl — .v n O p _ _ p _ L _ _ r s 0+6 5 mud > 3.0 $383dHVI11IN Figure 5. mKOOI I Hmumrcsqit OO 2 ON _ .JJI 2 oo 00 0+ ON 3 _ _ 1/. T e _ _ _ F .> mad 10x x New oo 0... ON _ a m IO. _ _ _ $2.0 OO 0... ow | [Z In JO. S 3 HEdHVI 71!“ Figure 6. TABLEII A etc-nary or the results obtained by charging a silver electrode at each or the potnetials listed, for a period‘ot 2h hem-s or until the current flow was rare, no reduction between steps. Volts above ‘ Loss of 1 silver Total weight 01’ mm 1910 potent. to A930 mgen gained silver 8.21; 0.05 18 0.0066 0.10 18 0.0068 0.15 18 0.0066 0.20 15 0.0057 0.25 13 0.00148 0.30 111 0.0052 21.1; 0.05 25 0.0061 0.10 75 0.0201 0.15 73 .5 0.0198 0.20 71 0.0191 0.25 11.5 0.0192 0.30 71.5 0.0192 0.35 76 , 0.0203. 0.00.1.5 22.7 0.05 1.5 0.01“; 0.10 82 0.0262 0.15 80 0.0258 0.20 71 0.0215 0.25 'I? 0.020.. 0030 78a5 0.0250 ' 0.35 590 0.03711 0.0032 - . “WI. 1 1—— - - be 0.05 a 0.0103 0.15 . 61 0.0193 0.20 61 0.0192 0.25 60 0.0191 0.30 59 0.0187 0-35 0.0177 0.0026 w 3h 100 I , , I 1 07 90»- d e 80* . «0.6 ‘3': '( .. <3 ‘70 “ >_ I— e E o 60- r L. Lu U E 3 Lu 50*- .. > g 3 U ‘2... ~04- g, 4( E! (I ‘5 L“ 30»- -+ G .3. :2 U) .J '0 °\. 20?- . IO - ‘ l0 ZC>$4 “aDP'SCD ‘10 50 Figure? The relation between the percentage of silver converted to A920 and the specific conductivity of KOH. Specific conductivity is plotted on the right ordinate and % silver converted to A920 on the left. The upper curve is specific conductivity. 35. O. o 0.08 0.06 Of 5 .0 ’3 4. v 8‘/. KOH a 2'73 KOH A 25%KOH o 40% KOH O ('3 Lu F l GRAHS OF OXYGEN PER GRAN 0F SILVER 0 O N I J O o T L l l 1 1 1 i 1 0.05 mo 0.15 0.90 02 s 0.3 035 VOLTs ABOVE A,~A,20 POI‘ENTIAL Figures 8 A graph of the data of Table II showing the inertness of the partially oxidized electrode between'0.l and 0.3 volts above the Ag-Agzo potential. L (ARBITRARY burrs) i.( HluIANPEREJ) 350 300 250 to 8 I50 § 0‘ O 70 60 50 4O 3O 20 IO I I I ‘r f r I I I b -— 0.15 v —---— OJOv .— —----- 0.05v .1 ’_ -4 _ -—J -...‘ -- e—t h——-.1.-._.I_---L— r---_r_---I—-L.---T....--JI.- IO 20 30 40 50 60 7o 80 90 I00 MILL! A HPERE HOURS Figure 9 Theoretical graph of rate dependence from overvoltage theory. The electrode capacity is 100 milliampere hours. i T [ T I i 1 7 I \ \ --- 42% KOH \ ‘\ ——2s% KOH .1 _ \ \ 2'./o KOH #— —I .— A F. q L 1 1 1 1 l 1 1 O I 2 3 1 5 6 7 8 7 IO TIME (hours) 11.932 10 Theoretical graph of rate vs. time for a constant potential process controlled by the rate of linear diffusion. 36. (1) (2) (3) (h) (S) (6) (7) (3) (9) ( 10) (11) (19' (13) w.) [(15) (16) (17) REWCES 1(i. L.)Johnston, 1“. Cute, A. B. Garrett,J. Am. Chem Soc. 55, 2311 1933 . As Be Welding 18,. 713 (19515. T. P. Dirkse, mm Technical Report on Contract Hour-1682(Ol), 31 Decanter 1956. T. P. Dirkse, First Technical Report on Contract Norm-1632mm, I. A. Rammski, Doklady Akadenii Hank, U.S.S.R., ?. Abfloyes, J. L. Hoard, K. S. Pitmr, J. Am. Chem. Soc. 51, 1221 193 . P. Jirsa, z. anorg. allgea. Chem. 235, 302 (1935). P. 1.. Carmen, mm. Faraday Soc. 3, 566(19311). Se mm. Z. We 8119811. We fig 331 (1901). Jo Ce "hit. a To Pin“, To Pa Dirk“, We Enema 50‘s 29, 1.57 (19115) P. Jirsa, J. Jslinek, Z. anorg. allgu. (he. 12;, 61 (1926). A. Hickling, D. “harlot, Disc. Faraday Soc. lie. 1. 277 (19117). c. a. H. sum, 0. Hargerison, n-m. Faraday Soc. 2;, 925 (1955). a. mum, n. 11. 11am, m. m s... 51, 71 (1955). 11. Le Blane, H. Sackse, 91mm. Zeit. .2, 887-9 (1931). P. Dclahay, mew Instruaental nethode in Electroeheaiehy,’ p. 311. Interscience Publishers Inc., New York (19511). P. Delaney, ”New Instrumental Methods in Electrocheaistsy,‘ p. 51, Intarscience mush." 1110., New York (19511). n. 1. Lsitinen and 1. n. Kolthofi’, J. Am. om. Soc. g1, 33M. (1939). llllllHltl