ANALYTICAL CHEMISTRY
334 is used, it is possible to obtain caking as defined above tvithout any liquid addition. This caking tendency of very fine catalyst can be eliminated by addition of coarse, nonporous material such as sand, so that the method probably can be applied to very fine porous powders if necessary. Experience with these materials is limited, however. Tests can be carried out on coarse niatcrial, either ground or unground. Pores larger than the particale size will be eliminated by grinding, so that tests on the grind might give values lower than the total porosity prior to grinding.
Table IV.
Density Data on Calcined Silica-Alumina Catalyst Pore Volume, Cc./Gram 0.675
Method TVater titration
’T$$~~~~~~$&(fheck
Density, Gram/Cc. 0.916
0.870
0.913
Skeletal Densitya, Gram/Cc. 2.45
2.44
A value of 2.46 grams per cc. was obtained by the conventional method i 1 ) 11-ing isopropyl alcohol a t 30O C.
PARTICLE .4ND SKELETAL DENSITY DETERMINATION
I k a u s e particle density (density of particles n-hen pores are filled !&h fluid lnediunl) determines elutriatioll and sedimeIltatioll behavior, it is important in regard t o stack losses from c,omnierrial fluid catalyst units, as !vel1 as particle size measurement. At the present time, it is usually determined from pore volume and skeletal density (density of solid if nonporous), using the following relation for density in a mcdilim of negligible dvn$ity such as air.
mix is measured. T o eliminate entrapped air, the bottle should be subjected to a swirling motion or brief evacuation prior to cOn1Plete filling. Particle density can then be calculated as folloa-s:
d, =
sample weight final \\-t.-tare wt.- sample a t . flask v0l.- _ _ _ ~ water density a t temperature vol. added to pore vol. end point
+
~
Skelrtal density is given by: where d, d,
I,‘,
= = =
particle density, gram er cc. skeletal density, gram)&. pore volume, cc. per gram
However, both d, and V p may be a function of temperature, pretreatment, and the like, which affect water sorption and shrinkage so that particle densities calculated from pore volumes and skeletal density obtained under other conditions may be in e i ~ o r . A direct measure of particle density under conditions of interest seems preferable. For example, in determining particle size disparticle densities calculated tribution by the Roller method (8), from pore volume and skeletal densities obtained under other conditions cannot be vorrectly applied. To accomplish the desired objective, the sample under oonditions of interest-e.g., material from a Roller test, which nornially has a high water content because of use of n-et air to eliniiriate electrostatic effects-is first tested for pore volume as desc-iilJcd above. More water is then added to form a slurry and the sample transferred to a volumetric flask or pycnometer of k n o ~ v nvolume arid tare (100 ml. is suitable for a 25-gram sumplc). The bottle i p then filled and acvurately weighed and the trmpei,ature of the
d,
=
sample \\-eight flask vol. - final n-t.-tare wt.- sample wt. \+ ater density a t temperature
The skeletal density data obtained (Table IV) were in good agreement with data obtained in the conventional manner ( I ) by the Stamford Analytical Laboratory, using isopropyl alcohol a8 the medium. LITERATURE CITED (1) Am. SOC.Testing Xaterials, Standards, Part 4, 5 2 , 196 (1952). (2) Ashley, K. D., Innes, W. B., Iizd. Eng. Chem. 44, 2857 (1952). (3) Holmes, J., E m m e t t , P. H., J . P h y s . & Colloid Chenl. 5 1 , 1262 (1947). (4) Iniies, W. B., ABAL.CHEM.2 3 , 759 (1951). (5) Innes, W‘. B., .ishley, K. D., Pioc. -4m. Petroleum Inst. 2 7 , 111, 9 (1947). 16) Ries. H. E.. J . Ain. Chein. Soc. 67. 1249 (1945). i7) Ritter, H . ‘L., Drake, L. C., I ~ DE N . G . CHEM.,ANAL.ED. 17, 787 (1945). ( 8 ) Roller, P. S., Proc. Am. Soc. Testiiig Materials 3 2 , 507 (1932).
RFXEIVED for review 1Iarcli 2 3 , 1955.
Accepted January 5 , lY3Ci.
Oxidation of Oxalic Acid in Glacial Acetic Acid with CeriumW) 0. N. HINSVARK’
and
K. G. STONE
K e d z i e Chemical Laboratory, M i c h i g a n State University, East Lansing, M i c h .
A solution of ammonium hexanitratocerate in glacial acetic acid may be used as a \ohmetric reagent, provided it is kept in an amber flask and standardized each day. Oxidations must be carried out in the presence of perchloric acid, so that the rate will be reasonably rapid. The oxidation of oxalic acid in glacial acetic acid requires 2 equivalents of cerium(1V) and j ields 2 moles of carbon dioxide per mole of oxalate. Tracer studies with carbon-14 show that both moles of carbon dioxide come from the glacial acetic acid solvent. This discovery was completely unexpected. RIany ox: genated organic molecules do not interfere in the oxidation of oxalic acid.
I
N S P I T E of the fact that nonqueous acidimetry has undergone tremendous development in recent years, nonaqueous redoximetry has received comparatively little attention (9, 1 0 ) . Oxidations of organic substances with cerium(1V) salts in aqueous media are not well understood, but it was thought that an inves. tigation of similar oxidations in nonaqueous media could explain the role of water in the aqueous media. Sodium oxalate has been recommended as a primary standard for cerium(I1’j solutior~s (6, ‘7); therefore, the initial work was on the oxidation of oxalic! acid formed from sodium oxalate in glacial acetic acid solvent. .icetic acid was chosen because it is readily available in pure forni 1
Present address, Girdler Corp., Louisville, Ky.
V O L U M E 28, NO. 3, M A R C H 1 9 5 6
335
and dissolves moRt oxygenated organic moleciilzy. I n :ittempts to study the oxidation of organic substances other than oxalic acid, it was found that the acetic acid solvent participated in the reaction. Solvent participation was also found for the oxalic arid oxidation. This unexpected result makes the oxidation niwhanisni in general lese clear, EXPERIXIENTAL
Acet,ic acid used in t,his work was either reagent grade or analyzed reagent. No purification was required except for the stahility st,udies,for which the acid was dist,illed once from chromium t'rioxide and once from potassium permanganate. All cerium salts were obtained from the G. Frederick Smith Chemical Co. All other chemicals x e r e either C.P. or reagent grade. Acetic. mhydride was redistilled before use. Sodium l-carbon-14:iwtat.e was obtained from the Chem Rad Division, Xurlear Instrument and Chemical Co., Chicago. Cerium(1V) solutions were found t o be sensitive to light. They n-ere kept, therefore, in aniber bottles, and all titrations were done using amber burets. Qualitative tests on gas evolved during ozidat,ions were made with an Orsat-type gas analyzer.
'rnhle 1.
Solubilit>-of Cerium(1V) Salts in Acetic Acid Salt
Normality 0 0 0 0
005 0On 01
06
A Fisher Elecdropode (0.025 pa. per s( ale division) equipped with 2-cm., 18-gage platinum wire active electrodes IT as used for ainperometric end points. A Sargent potentiometer was used for potentiometric measurements. A magnetic stirrer was used during titration; provision was made for introducing nitrogen into solutions Tvhere necessary. Counting in the radioisotope studies was done with a Suclear scaling unit, Jlodel 163, in conjunction with a Tracerlab windowless flow counter, SC16, fed with Matheson Geiger flow gas (helium-isobutane). Solutions of cerium( I V ) salts were prepared by saturating gldria1 acetic acid at 60" C. with the dried, finely powdered salt, allowing the solution to cool in the dark to room temperature, and filtering through a sintered-glass filter of medium porosity. A substance approximating ceric hydroxide was prepared by neutralizing a solution of cerium( I V ) nitrate with dilute ammonium hydroxide and washing the precipitate thoroughly with water, then air drying. T h e solid thus obtained could well be a basic nitrate. The silver-silver chloride electrode used in this work \vas prepared in the usual way (2). Organic chemicals studied as interferences n-ere C.P. quality and mere not purified further. Titrations were carried out by weighing portions of sodium oxalate or pipetting aliquots of a glacial acetic acid solution of sodium oxalate into 150-ml. beakers, then adding 50 ml. of glacial acetic acid and enough i o % perchloric acid t o make the finid mixture 1 S in perchloric acid. The resulting solution v a s titrated M ith ammonium hexanitratocerate as described below. For cases where the carbon dioxide was t o be examined, the titration was carried out in a conical flask attached to a watcr condenser, n-hich led t o a n absorption bulb containing standard O.2,V barium hydroxide. T h e sample solution was deaerated with carbon diouide-free nitrogen. T h e carbon dioxide formcd was mashed by passing through water and was s m p t into tlie barium hydroxide solution with the same gas The excess base was titrated with standard arid t o determine the carbon dioxide atmorbed. Solvent participation studies were carried out using sodium I-carbon-14-acetate in the following way. A 0.1-me. quantity oi tagged sodium acetate (about 8 mg.) was dissolved in 20 ml. of glacial acetic acid. Exactly 1 ml. of this solution was added to exactly 50 ml. of solution containing the sodium oxalate to be oxidized and t o exactly 50 ml. of cerium(1V) solution, so that dilution of the carbon-14 would not occur on mixing. T h e iequired amount of cerium(1V) solution was added t o the oxidation solution and the carbon dioxide caught as previously described. T h e barium carbonate was collected on a sintered porcelain crucible, dried, and weighed. A SO-mg. portion was transferred to a 2-rm. aluminum counting dish, and the activity was counted in the flow counter. A w i g h e d portion of the solvent containing carbon-14 was burned to carbon dioxide in a combustion train
arid the act,ivity determined in the same manner. By these procedures the counts per minute per milligram of barium carbonate from both the solvent and the oxidation kvere det,ermined. I n order t o prove positively that oxalic arid did not exchange wit,h acetic acid in t,he presence of perchloric acid, a mixture containing cerium(III), tagged acetic acid, and sodium oxalate dissolved in 20 ml. of glacial acetic acid {vas allowed to stand at rooin temperature for 4 days. At, t,his t h e the volume was reduced to a few milliliters by vacuum distillation at room temperature. The residue was t,aken up in water, the cerium(II1) ivas precipitated by neutralization with dilute ammonium hydroxide, and the oxalat,e was isolated as calcium oxalate. After mishing and drying the calcium oxalat,e s h o w d no higher rount t,han t,hr linckground count for that day. SOLUBILITY O F CERIUII(1V) SALTS IN ACETIC ACII)
\Then this work was started, no inforrnation n-as availal,le o n the solubility of cerium(II-) salts in g1aci:il acetic acid. Saturated solutions of ceric sulfate! ammonium trisulfatoceratc, re*,ie hydroxide, and ammonium hexanitratoceratc were prepared. The first three were analyzed by adding 2 ml. of 70% perchloric acid t o 50-ml. portions and titrating ivith standard iron(I1) perchlorate in glacial acetir acid (3). Iron(I1) perchlorate can not be used in the presence of nitrate, hon-ever, and the solution of ammonium hexanitratocerate was analyzed b>- titration of sodium oxalate as described beloiv. The approximat e solubilities found are listed in Table I. Exact solubilitics have little meaning because solutions are not stable. I n a n attempt to improve the solubility of cwic hydroxide, the dry solid was moistened with mineral acids including sulfuric, perchloric, and nitric. I n all cases the solubility in acetic arid was increased, but the resulting solution completely decomposed in a few hours, as shonm by the loss of color and oxidizing power. From these experiments it \vas concluded that only ammoniuni hrxanitratocerate was worth further invwtigation.
Table 11.
Effect of Acid on Potential of Cerium(tI1)Cerium(1V) Couple in Acetic Acid
rlcid
s of Acid
HCIOI
0,23
0.50
i n
i.5
H2SOI
2.5 2.5
E us.
Ag-AgC1, Volts 0 99 1 00 1 02 1 04
1 oti
0 81
DETECTION OF END POINTS
Three possible methods for the detection of the end point in titrations of oxalic arid with acetir acid solutions of a.mnionium hexanitratocerate were considered-namely, indicatory, potentionieti,ic, and arnperometric with tivo active electrodes. The indicator method was soon abandoned when it was found that the oxidation of oxalic acid, even in 1S perchloric acid, was fast hut not instantaneous. Therefore, the end point was usually missed, because no indication of the vicinity of the end point could be observed. I n order to use the potentiometric method for end-point detection, information was required concerning suitable electrodes for use in 1 S perchloric acid in glacial acetic acid, potentials for cerium systems in acetic acid, and rates of attainment of stable potentials. T o study potential changes, a solut'ion of sodium oxalate in 1 S perchloric acid was oxidized with ammonium hexanitratocerate and a n equal amount added in excess so that approximately equivalent amounts of cerium( 111)and cerium( IV) were present. When the process was followed using platinum and calomel electrodes, large breaks in potential were found in the vicinity of the equivalence point, but individual values were not
336
ANALYTICAL CHEMISTRY
Table 111.
a
39.0
Sensitivity of 4mperometric End Point"
Potential,
2 "El
LIV.
A M I . ' 111.
200 225 250 275 300
8 13 8.76
~ i i g .XaL!?Oa
1'aI)lc I\
.
9.76 10.5 10.3 titrated wit11 20.D7 nil. 0.029O.V Ce(I\')
Reproducihilit> of -4mperometric End Points
Ce(IV) S o h .
S of Ce(1V) S o h . 0.0261, 0.0259, 0.0260 0.0426, 0.0426, 0.0425 0.00913, 0.00914
reproducible and consideritble variation with t,ime was observed. This system might be used for end-point detection, but it nould be of little value in measuring the cerium(II1)-cerium(1V) formal redox potential. T h e messurements were then repeated using platinum or silver--silver chloride electrodes with different acid concentrations. At least three conclusions may be drawn from the data (Table 11). First, the values have only a relative meaning, because the valuc for the silver-silver chloride clcctrode in 1.Y perchloric acid in glacial acetic acid is unknown. Second, the solution contains a multitude of species, and what i,s being measured is not clear. Third, there is a definite anion effect, because the addition of sulfuric acid t o a perchloric acid solut,ion immediately lowers the potential. .4s a result, these values must he looked upon as only preliminarj-; further work is in progress t o clarify the situation. .4n unusual observation in the course of the work was t h a t flakes of silver chloride fell from the surface of the electrode as it stood in acetic acid. T h e end point may be detected using silver-silver chloride electrodes if necessary, but the flaking is an objection. T h e amperometric end point using two active platinum electrodes for titrations with cerium(1V) in aqueous medium has been noted (8). There was no reason t o believe that the method ~vouldnot work in nonaqueous media. h portion of sodium oxalate was dissolved in 1.Y perchloric acid in acetic acid itnd an equivalent amount of cerium(1V) solution TTas added plus 1 drop more. Cleaned platinum electrodes were inserted, and the sensitivity, microamperes per milliliter, was determined with different potentials applied (Table 111). At the same time it v a s found that, the sensitivity as well as the rate of oxidation was greater in 1S acid than in 0 . 5 5 acid. From thesc results nn :tpplied potential of 275 m v . was selected for all future work. T o check the reproduciliility of the amperometric end point, three different cerium(1V) solutions were used t o titrate aliquots of a, solution of sodium oxalate in acetic acid, made 1.1- in perchloric acid. T h e results are summarized in Tahle IV. STABILITY OF ACETIC ACID SOLUTIONS OF AMMOSIURI HEXANITRATOCERATE
Because thc slow oxidation of acetic acid by cerium(1V) in aqueous media has been reported (1, 4, 6 ) , it was necessary t o determine the stability of acetic acid solutions of ainmonirim hexanitratocerate. dqueous solutions of cerium(1V) are sensitive t o light ( 6 ) . -4solution of ammonium hexanitratocerate in acetic acid with no perchloric acid present was divided int6 two parts. One part n-as placed in a clear, glass-stoppered borosilicate flask, the other in an amber, glass-stoppered borosilicate flask. Both Lvere stored on the bench top a t room temperature (23" t o 26" C.) with no additional protection from light. Portions were removed periodically from each flask and used t o titrate aliquots of a known eolution of sodium oxalate in acetic acid. T h e results are sum-
marized in Tslile V. The times noted are within i O . 5 houi,; therefore, the results rcprepent crude rate data. Obviously, cerium(1V) solutions in acetic arid must be kept in dark bottles and titrations performed using amber burets t o minimize derompo.:ition by light. It is also apparent that frequent stuntlnrdizntion is neccwary. The effect of perchloric acid on the stability of ncetic ticid solutions of rerium(1V) n-as also determined. For this study an amount of perchloric acid t o yield the desired concentrat.ion \ v a ~ added to the cerium(1T') solution in a n amber flask. Aliquots were removed periodically and added to excess sodium oxalate in 1 S perchloric arid, and the excess mas titrated rvith another standardized cerium(1T') solution. From the results shown in Table 1'1, it can be seen that, the oxidizing por-er of ccriuni(1V) .solutions in acetic acid in the presence of perchloric acid decrease3 very rapidly. Therefore, under t,hese conditions exrws techniques are not applicable and only small amounts of cerium(1V) may be added during direct titrations. Thcse limitations restrict the use of acetic acid solutions of ceriuni(1V) conaiderahly. ~
~
Table Y. Light Sensitivity of icetic Acid Solutions of Cerium(1V) 1 itlie l)a) s
Y of Ce(IY) S o h . ____ Clear flash
Ambei-kasK
0 0260 0 0241 0 0230 0 0185 0 0134
0 1 2
3
0 0 0 0 0
0260 0255 0253 0248 0221
Table VI. Stability of Acetic Acid Solutions of Cerium(1V) Containing Perchloric Acid TIme, Minutes
0 . 3 ,V
HClOa
S of Ce(1V) Solution 0 75N HClOa
1 OY HClOi
POSSIBLE REDUCTANTS IN ACETIC ACID
one point in this work it appeared desirable t o have :t reducing agent available t o remove cerium( IV) quickly by reduction t o cerium(II1). T h e various substances tried included arsenious oxide, sodium arsenite, stannous chloride, ferrous PUIfate, Mohr's salt, Oesper's salt (ferrous ethylenediaminesulfate tetrahydrate), and ferrous chloride. ,411 these were eithei only sparingly soluble in acetic acid or underwent a reaction with the solvent. Sodium nitrite \vas soluble and readily oxidized, but a solution of sodium nitrite in acetic acid evolved a colorless gas after only a few minutes. Iron(I1) perchlorate was found to be suitable under certain conditions (8). S o completely satisfactory reducing agent is known for use in acetic acid. OXIDATION OF OXALIC ACID
\Then sodium oxalate is dissolved in acetic acid and titrated with animonium hesanitr:ttocerate, the rate of oxidation is $low. If the solution is made 1.V in sulfuric acid, the rate increases a little. If the solution iP mnde 1-V in perchloric acid, the rate becomes fast but not instantaneous. This emphasizes the need for the use of perchloric acid in acetic acid solution. -4t the time the work was started, it \vas assumed that 2 equivalents of oxidizing agent would be consumed per mole of sodium oxalate and t h a t 2 moles of carbon dioxide would be formed se the sole product. .4fter some preliminary experiments with other
V O L U M E 28, N O . 3, M A R C H 1 9 5 6
337
conipound~.,it was clear t h a t the stoichiometry of thc oxalate oxidation had t o be checked. T h e cerium(1V) content' of a .,iinple of ammonium hexanitratocerate was d e t e r m i n d by ;tqueous reduction in the usual way, and a weighed portion \vas clissolved in acetic acid to yield a solution of known concentration, \\-cighrd portions of podium osalate ivere dissolvcd in acetic acid :vhich was 1-V in perchloric acid and titrated n-ith the ccrium(1V) wlution to a n ampcrometric end point. The carhon dioxide cvolvrd wvas absorbed in barium hydroxide, and the esccss barium hydroxide vats titrated with standard acid. The results in T:tble \'TI show t h a t the stoichiometry appears to be normal. 1,ater experiments suggested t h a t the acetic acid solvent part icipated in cerium(1V) oxidations. The work described above \vas repeated \Tit11 the addition of sodium 1-carbon-14-acetate. The cnrhon dioxide formed was isolated as barium carbonate and rounted. The specific activity of the barium carbonate from thr acetic acid solvent was 2.54 (Table VIII). However, only onc carbon atom in the acetic acid was labeled; therefore, the spec~ificactivity of barium carbonate per carboxyl group is 5.08. Hecause the specific activity of the barium carbonate from the oxithtion is 5.45, all the carbon dioside isolated came from carhoxyl group carbon and 2 moles of acetic acid participate per mole of oxalic acid. S o n e of the carbon dioxide isohted can rome froin the osalic acid n-hich has reacted.
Table IX.
Effect of Oxygenated Impurity on Oxidation of Oxalic Acid
Added Impurity Sodium formate
Amount of Impurity 30 nig.
Acetic anhydride
3 0 nil.
Ethyl alcohol
2 . 0 nil.
E t h y l alcohol, plus acetic aiihydridp
1 , 0 nil., 3 0 nil. 1 0 ml., 3 0 ml. 2 0 ml.
JIg. Sa9C'OG Tnkeri Found 74.4 59.3 53 9 6B.5 51.7 367
?4 A 29.3 53.9
38.9
3ti 9
52 G 41 7 42.6 50 mg. (C€I20)3 S 5 . 6 315.7 0 . 5 ml. h-0 end 0 2 ml. 35 1
l s o p r o y y l alcohol
Forninldehyde Iienzaldehyde I r e t on e
2 . 0 mi.
(;lycerol
0 , 5 nil. 0 5 ml..
Glycerol iiliis acetic anhydride
1;thylene glycol Ethylene glycol, plus acetic anhydride
di:crose
3 . 0 ml. 0 5ml. 3 . 0 ml. 0 5 ml. 0 . 5 mi., 3 0 nil. 0 . 5 ml.. 3 . 0 ml. 100 m e .
.io
fitit;
51.0 369
50.4 41 6
42.5 55 9 36 8
point 31 9
50.3 48 .j 48.3 53 7 53.6 KOend point 1
45.4
43
74.4
74 9
tG
N o end point 53.8
54 2
38.1
38.7
KOend point No end point
N o end point No end point
?io end point
'I'ahle VII.
Oxidation of Oxalic Acid in icetic Acid [0.0290.V CelIV) I JIinole COz per Mniole
M g NanCzOa
24 2 10 7 78 6
'I'ahle VIII.
NazCz04 2 01 2 03 0 99
Specific Activity of Barium Carhonate JIg.
San1ple Hackeround .4cetic acid dodiiini oxalate
J l e q . Ce(1V) per hImole.
NazCzOa 2.08 1.96 2.03
BaCOs 8.5.4
.IO 0
Corints/hlin., Coiints/hIin., Gross Net 36.9 127.2 310.4
90.3 273.5
Connts/JIin. per .\In. BaC03 2 54 3.45
titrated t o the amperometiir end point with cerium(1V). From the results in Table I X it is seen that polyols and tartaric arid 1-ield no end points. Perchloric acid plus acetic anhydride in acetic acid should be a n excellent acetylating mixture; therefore, acetic anhydride was added t o the polyols, and some iniproveinent \\as found. T h e addition of acetic anhydride t o ethyl alcohol, however, decreases the accurarv. No experimental evidence is available concerning the cause for this effect. I n general it may he concluded that oxalic acid may he determined in the piewnce of a nuxnher of ouygenatcd niolrculep. 4CKYOW LEDGMEYT
Financial support for thip m-ork ivas generously provided Sational Science Foundation under a grant, T\;SF-G281. \Then solvent participation n-as found, the question of the niechanism wts considered briefly. If the oxidation proceded via free methyl radicals formed by oxidation of the solvent, methanc or ethane would be expected as a side reaction product (,$, 5, 11, 12). T h e gases formed on a large scale oxidntion ~vei'r w e p t on a stream of carbon dioxide into 40% potassium hyilmside solution in a gas buret. No measurable amount of inwluhle gas was found, indicating t h a t hydrocarhon formation w:ts not a n important process. Tests for t'he presence of peroxides in the oxidation mixture were inconclusive. At this time the oxidation mechanism is unknown. Further lvork is in progress to esplain the participation of the acetic acid solvent in the oridation. EFFECTS O F OTHER SUBSTANCES ON OXIDATION OF OXALIC ACID
Because os:tlic acid could be oxidized a t a rapid rate lvith c,cbrium(IV) in acetic acid IThich was 1-1- in perchloric arid, and lwrause test tube experiments shon-ed t h a t many oxygenated organic niolecules TTere oxidized very slowly, i t appeared that oxalic acid might be titrated in the presence of these moleculee. For tjhese csperiments definite quantities of sodium oxalate and inipurit?; were added to 1 S perchloric acid in acetic acid and
111
the'
LITER4TURE CITED
Foster. L. LI..Pavne. J. H.. J . A m . Chem. SOC.63. 223-5 11941). , ~, ( 2 ) Glasstone, S., "Introduction to Electrochemistry," p p ~3 3 - 4 0 , D. C. Heath Co., S e w Tork. 1941. (3) Hinsvark, 0. K . , Stone, K. G., . ~ N . < L . CHEV.27, 371-3 (1955). (4) Kharasch, RI. S., Friedlander, H. S . ,Crey, XV. H., J . Oro. Chem. 16, 533-42 (1951). (5) Raley, J. H., others, J . A m . Chetrt. SOC. 73, 15-17 (1951). (6) Smith, G. F.. "Cerate Oxidimetry," G. Frederick Smith Chemical Co.. Columbus Oliio. 1942. (7) Smith, G. F., Duke, F. R., ISD. Esc,. C H n r . . d x i i . . ED. 15, 120-2 (1943). (8) Stone, K. G.. Scholten. H. G., . ~ N . A L .C H n r . 24, 671-4 (1952). (9) Tomicek, 0.. Heyrovsk?, .\., Collection Czechosioi. Cheni. Comnzuns. 15, 997-1020 (1950). (10) Tomicek, O., Valcha. J., Ihid., 16-17, 113-26 (1951-52). (11) Waters, W. rl.. "Chemistry of Free Radicals," pp. 245-7, Clarendon Press. Oxford, 1948. (12) Waters, W. A . , "Mecanisme de l'oxydation," pp. 101-9, I?. (1)
~~~
~
\
-
-
~
-
I
Stoops, ed., Huitieme Conseil Internationale de CIiimie Solvay, Brussels, 1950. R E C E I V Efor D review October 19, 1955. Accepted Deceniher 19, 1955, Division of Analytical Chemistry, 128th Meeting, h C S , JIinneapolis, Jlinn.. September 1955. Abstracted from a thesis presented t o JIichigan State University by 0. S . Hinsvark in partial fulfillment of requirements for degree of doctor of philosophy, December 1954.
