PETER RIESZAND EDWIN J. HART
858
different molecules. In the alcohol series the rates are in the order expected for abstraction of the carbon hydrogen from methanol and t-butyl alcohol, the secondary hydrogen from ethanol and the tertiary hydrogen from isopropyl alcohol. The rates suggest that in ether the secondary hydrogen is removed whilst in ethyl acetate the primary hydrogen of the ethyl group is involved. Probable solvent effects on the quinone reactions prevent any conclusions being drawn from the reaction rates in cyclohexane and acetic acid. Primary Yields.-The results in Table I indicate a tendency for total radical yields to be higher when G(H) is high. Thus the compounds fall into two groups, the primary and secondary alcohols for which G(H) > 2 and G(R) > 6, and the others for which G(H) < 0.5 and G(R) is 4.5 t o 3.1. This tendency may result from the greater ease with which hydrogen atoms escape solvent cage reactions. It seems odd that G(H) for cyclohexanol is so small. One would expect a value similar to the G(H) for either cyclohexane or isopropyl alcohol as is the case for Gm(H2) and G(R). It is possible that traces of unsaturated materials are responsible for the anomaly, but further fractional distillation left G(H2) unchanged (Fig. 9). For the other alcohols G(H) increases in the order MeOH, EtOH, iso-PrOH. This variation is in the opposite sense to that expected from the mass
VOl. 63
spectra of the vapors. Here the relative abundance of the ion corresponding to the production of H atoms (Le., parent ion less one) decreases appreciably along the series. This suggests that ionization processes are not a predominant primary source of H atoms and that possibly dissociative electron capture of the type ROH+e+RO-+H
or charge neutralization are more important. However, it is not immediately obvious how these mechanisms account for the much lower G(H) for butanol. One possibility is that the slightly higher 0-H bond energy compared with the other alcohols' makes the electron capture process less efficient. The values of G(CHa) for the alcohols (except methanol which may involve a special mechanism for CHa production' ) fall in line with the abundance of the corresponding ion (ie., parent ion less 15) in the mass spectra. This would be consistent with CHs as a primary product of the ionization processes. We gratefully acknowledge the financial support of the U.K. Atomic Energy Authority (Research Group Harwell) and the Department of Scientific and Industrial Research.
c
. I
( 7 ) P. Gray, Trans. Faraday Xoc., 62, 344 (1956).
ABSOLUTE RATE CONSTANTS FOR H ATOM REACTIONS I N AQUEOUS SOLUTIONS1 BYPETER R I E S ZAND ~ EDWIN J. HART Argonne National Laboratory, Lemont, Illinois Received February 9, 1069
+
+
In order to determine the absolute rate constant of the reaction H FeIII = Fer1 H+, we have measured the relative ~) then calculated k ( ~ ~+~ 1 1 by 1 ) assuming that ~ ( H + D , ) ,known from gas phase studies, rate constant ratio k ( ~ F+~ I I I ) / ~ ( H + Dand remains unchanged in solution. The initial H D yields from the yirradiation of water containing dissolved deuterium gas, ferric sulfate and ferrous sulfate have been measured at various FeIII/D2 ratios. From these studies we have calculated the rate constant ratio ~ ( H + F ~ I I I ) / ~ ( H + D = * ) 120 f 30 in 0.01 N H2SO4. Since the rate constant of the gas reaction, H -4- DI = HD D, equals 0.4 x 106 1. mole-' sec.-1, it follows tha: the rate constant of the reaction H Fer11 = FeII H + in 0.01 N H~SOC is (0.48 I 0.1) x 107 1. mole-1 sec.-l at 25 Using this result, we have calculated a number of absolute rate Constants for H atom reactions.
+
+
Introduction For a better understanding of aqueous radiation chemistry, a knowledge of the absolute rate constants of H atom reactions with a variety of in-
+
.
absolute value of the rate constant for a single suitable reaction. Since the rate constant for the Dz = H D D is known from gas reaction H phase studies,D our procedure was to measure organic and organic substances is desirable. the relative rate constant ratio, k ( +~Fe"')/k(H + D2), Such rate constants are of theoretical interest in and then to calculate k ( +~ FeIII) by assuming the development of the radical diffusion modelav4 that k ( +~ D ~ Iis the same in solution as in the gas and in testing its ability to predict the effect of phase. When this work was started, it was hoped that the competition of ferric ions and deuterium solutes on the molecular Hzyield. In order to calculate absolute rate constants from molecules for hydrogen atoms could be studied by the large number of published relative rate con- measuring the initial HD yield from yirradiated stants for H atom reaction~,~-sone requires the (5) J. H. Baxendale and D. H. Smithies, 2. physik. Chem., 6, 242 (1) Based on work performed under the auspices of the U. S. Atomic Energy Commission. (2) Radiation Branch, National Cancer Institute, N. I. H., Bethesda 14, Md. (3) D. A. Flandem and H. Fricke, J. Chem. Phyr., 28, 1126 (1958). (4) H.A. Schwarz, J . A m . Chem. Soc., 77, 4960 (1955).
+
+
(1956). (6) J. N. Baxendale and G. Hughes, ibid., 14, 306 (1958). (7) J. N. Baxendale and G. Hughes, ibid., 14, 323 (1958). ( 8 ) W. G. Rothchild and A. 0. Allen, Rad. Res., 8 , 101 (1968). (9) C. Boato, G. Careri, A. Cimino, E. Molinari and G. G. Volpi, J . Chem. Phya., 24, 783 (1956). '
I
Julie, 2959
ABSOLUTE RATECONSTANTS FOR H ATOMREACTIONS IN AQUEOUS SOLUTIONS
solutions containing dissolved deuterium gas at various FeIII/D2 ratios.
K
cW \ I
Triply distilled water10 was used in all experiments. In W Stock solutions with known FelI/FeIII ratios were prepared 6I 50 by addition of standardized hydrogen peroxide to ferrous 0 sulfate solutions in 0.10 N HzS04. The total iron content a of the ferrous sulfate solutions was determined by oxida;tion Y to ferric sulfate with excess HzOz. Ferric ion concentrations were determined by absorption spectroscopy at 3020 A. Solutions used in irradiation experiments were made up 0 from the stock solutions just prior to irradiation and were 0.010 N in HISOn. The pH's of these solutions were measured on a Beckman pH meter and found to be 2.12 =k 0.02. Deuterium gas supplied by the Stuart Oxygen Co. and assaying >99.5% in deuterium atom content was used without further purification except to pass it through a liquid nitrogen trap. The procedures to degas the solutions, dissolve the deuterium gas and fill the 100-ml. syringes have been described previously.lo All solutions prepared by the above procedure contained no gas phase and were irradiated with Cow -prays in a specially designed chamber." The dosage rates for each syringe used in these experiments were measured for the volumes employed by use of the ferrous sulfate dosimeter, and a value of 15.6 FeI"/100 e.v. was used to convert chemical yield to e.v./l. 0 For each experiment two or three syringes were used. 0 After irradiation 10 ml. of solution was added to 10 ml. of 1.6 N HzSOnand the optical density at 3020 A. determined Fig. l.--Fe"I in cells of 5- or l-cm. path length. Approximately 15-20 ml. of solution was introduced into a Van Slyke apparatus, the gas extracted and its volume and pressure measured. Two such gas samples at each dosage were combined for mass spectrometric determination of Hzr H D and DI. NZ and 02 were determined as a check on the absence of dissolved air. The initial concentration of deuterium was measured in each syringe by extracting the gas from an aliquot of the Rolution removed prior to irradiation. Mass spectrometric analyses were carried out on a Type 21-620 Consolidated Electrodynamics machine. Precision is estimated as &3% for points other than blank determinations of HD present initially in Dz where it is &6qib. The percentage compositions of Hz, H D and D2 obtained by the mass-spectrometer, to ether with the micromoles of total gas measured by the $an Slyke, were used to calculate the micromoles of HD and HI produced and of DI consumed.
(10) E. J. Hart, S. Gordon and D. A. Hutchinson, J . Am. Chem. Boc., TS, 6165 (1953). (11) R. A. Blomgren, E. J. Hart and L. 8. Markheim, Rev. Sci. InetT., a4, 298 (1953).
I'
IO 0
Experimental
Results and Discussion In some preliminary experiments solutions of either ferric sulfate or ferrous sulfate (205 pm./l.) in 0.01 N H2S9 containing about 670 pm./l. of dissolved deuterium gas were irradiated. A steady-state FeIII/FelI ratio is rapidly approached in both cases. The results for ferrous sulfate are shown in Fig. 1. It is evident from these experiments that precise initial H D yields ( * 3y0 for a given dose) cannot be determined for solutions containing only ferric ions without a substantial concentration of ferrous ions or vice versa. Hence subsequent experiments were carried out at steadystate FeIII/Fe11 ratios. The initial G(HD) and G(H2) values of experiments carried out at approximately steady-state FelI1/Fe'I ratios in 0.01 N H2S04containing about 600pm./l. of dissolved D2 for various Fe"1/D2 ratios are shown in Table I. In every experiment reported here the formation of HD and D2 is a linear function of dose for low conversion. Some typical results (experiment 3 of Table I) are shown in Fig. 2. For experiments in which
I
I
859
y
10
20
30
e v / l x 10.20.
and H D yields from 7-irradiatej FeS04 solutions containing dissolved D?.
l i
10
0 0
10
ev/t
x
10-20
20
30
Fig. 2.-Initial H D and Hzyields from yirradiated solutions containing dissolved Dz at the steady state FeIII/Fell ratio.
the initial FeIII/FeII ratio is not exactly equal to the steady-state ratio, the H D yield does not extrapolate to the H D content of the Dz gas at zero dose. Curvature becomes noticeable on such plots when the concentra.tion of the products, hydrogen and hydrogen deuteride, is about 10% of the total dissolved gas. In all experiments the total amount of gas was found to be constant; that is, the amount of deuterium consumed was always equal to the hydrogen and hydrogen deuteride produced. This observation can be understood by a consideration of the
PETER RIESZAND EDWIN J. HART
860
Vol. 63
TABLE I G(HD) AND G(H1) FOR AQUEOUSSOLUTIONS CONTAINING Dz, Fe'"
AND
Fe"
IN
0.01 N H2S04
DI
(FeIlI) +
DBav.,
Expt.
pM/I.
610 460 675 630 610 620 600
1
2 3 4 5
6 7
FeIII av., &/I.
Fer1 av., &/I.
GYHD)
11.6 31.5 26.7 11.5 101 64.5
22.4 36.2 41.0 22.1 101 70.2 22.5
1.38 0.54 0.72 1.33 0.218 0.380 1.47
11.1 I
dHn)r
I
I
(?(HI)
0.90 .76 .89 .85 .72 .80 .95
B
0.46 .18 .31 .46 ,086 .13 .51
I
+, I / '1
HD yields plotted for determination of
Fig. 3.-Initial
k ( +~F ~ I I J , / ~ ( +H na). The line is drawn with an intercept of one.
oxidation-reduction balance, since there is no production of oxygen or hydrogen peroxide and no change in the FeIII/FeII ratio. The following reactions take place when aqueous solutions containing D2, FeIIr and FeTI are irradiated Hg0 = Hz, HzO2, H, OH FeII Hz02 = FerrI OH OHFeII OH = FeIrI OHFe'IOHz H = FeIIIOHH) HD Fer10H2 D = FerIrOHFeIII H = Fer1 H + FelI1 D = Fer1 D + H202 H OH HzO Hz02 D = OH HDO Dz H = HD D De + O H = HOD D
+ + + + + + + + +
+ +
+
+ +
+
+ + + + +
+
k~ kZ ka k4 k6 k6
k7 k8 k9
k1o
+
The secondary reactions H H D = H2 D H2 = H D H can be neglected at low and D conversion. Reactions involving two atoms or radicals have been excluded, since the product of their stationary state concentrations is negligible. Let a, b and e be the rates of production (moles 1.-l sec.-l) of H, OH and H202, respectively. Then by applying the stationary state method to the intermediates H, D, OH and Hz02 the following results are obtained
+
+
(HzOz)ss =
C
kdFeI')
+ kdH) + kdD) a
(H)ss = ka(Ferl)
+ k6(Fer1I) + k7(H20z)ss + ks(Dz)
(2)
6.7(HnOn)8s Dn
4.04 10.1
0,0241 0.015 ,079 .012 8.18 ,048 ,007 4.27 ,0232 ,014 28.0 .189 ,003 14.6 ,120 ,005 3.76 ,0237 ,004 b+c (0H)ss = k2(Ferr) kdD2)
X
0.039 ,091 .055 ,038 ,192 ,125 ,028
+
1
30
(FeII)
Co--B i.20
G(HD) =
g(H), g(0H) and g(H20z) are the primary radical and molecular yields and G(HD) is the observed H D yield (molecules/100 e.v.). From equation 5 it follows that a plot of g(H)-y/[G(HD)
- PI should be a straight line with a slope equal to ks/ks and an intercept equal to one. Fortunately, several of the important rate constant ratios which appear in equption 5 have been determined in the course of previous investigations. The ratio has been measured by Allen and Rothchild* in 0.01 N HzS04 and is equal to 7.2 f 0.7. From the work of Baxendale and Hughes,' we can infer that ks/kr = kS/k3. These authors find no isotope effect in the reaction of H or D atoms with FeIII. Since reactions 4 and 3 both involve the rupture of an 0-H bond in the transition state, there should be little change in rate when the reactant is changed from H to D. The ratio OH + F & ) / k ( O H + * E * )has been determined by Dainton and Hardwick, l 2 Hardwick, l 3 and by Rothchild and Allen,a and was found to be 6.7, 8.6 (for 0.1 N HC104) and 5.7 (for 0.01 N H2S04), respectively. A value of 3.5 has been reported14J6 for the ratio OH + H J / ~ ( O H+ D~). (12) F. S. Dainton and T. J. Hardwick, Trans. Faradau Soc., 68, 333 (i957). (13) T.J. Hardwick, Can. J . Chem., 85, 437 (1957). (14) Atomic Energy of Canada, Ltd., Chalk River, Ontario, PRCM-64. (15) H. L. Friedman and A. H. Zeltmann, J . Chem. Phus., '38, 878 (1968).
June, 1959
ABSOLUTE RATECONSTANTS FOR H ATOMREACTIONS IN AQUEOUSSOLUTIONS
Combining the isotope effect of 3.5 with the intermediate value of 6.7 we find k2
&=
k(OH
+
Ferl)
However, it is interesting to note that even if we choose kz/klo = 30, this produces only a 15% decrease in the final result for k5/k9. The ratio k,/k6 was calculated from experiments 4 and 7, in which the y-ray intensity was varied by a factor of 6.8 keeping the ferrous, ferric and D2 concentrations approximately constant. The stationary state H20z. concentrations at the higher intensity (expt. 4, I = 2 X 1019 e.v.1.-* min.-l) was measured by following the post-irradiation increase in the ferric ion concentration and was found to be 1.4 f: O.lpm./l. by extrapolation. If we assume no reaction between H202and H or D atoms, we find from equation 1, (H20z)ss = c/kl(FeIII) = 2.5 pm./l.(kl=71 1. molesA1sec.-1).12 This indicates that about one-half of the hydrogen peroxide formed in experiment 4 is removed by reaction with H and D atoms. Since the steadystate concentrations of H and D atoms are approximately proportional to the y-ray intensity, we find (H202)expt.4/(H20z)expt.7 = 3.9. From the observed G(HD) values of experiments 4 and 7 we find by substitution into equation 5 that k7/k6 -- 6.7 and kg/kg = 65. The contribution of the term, (k7/k5)[(H202)ss/(D2)],t o the value of z is appreciable only a t low total ferrous plus ferric concentrations and is shown in Table I. The ratio k8/k4 will not differ by more than a factor of 2 or 3 from k7/k3, which is equal to (kT/ kg).(Icg/k3). Hence we find that the term ( k s / k4) [ ( H Z O ~ ) ~ ~ / (] F ~ isI I negligible ) compared to (kdk4) [(FeIII)/(FeII) 1. From the experimental data of Table I and the rate constant ratios in the literature, the quantities g(H)y/(G(HD) - p) and x were calculated and are given in Table I. The primary radical and molecular yields at pH 2.1 are g(H) = 3.3,8g(0H) = 2.6 (by material balance), g(H2) = 0.45 and g(Hz02) = 0.80. A plot of g(H)y/(G(HD) - 8) us. x was found to be linear (Fig. 3). The straight line, drawn by inspection and with an intercept of one, has a slope equal to k ( +.~ FeT")/k(H + D ~ ) = 120 f 30. The rate constant for the reaction, H Dz = HD D, in the gas phasegJ5J6a t 25" is 0.4 X lo5 1. moles-l sec.-l. Assuming that the rate constant, k ( 4-~ D ~ ) ,remains unchanged in solution, we find
+
IC(H + ~ ~ 1 1 1=) (48k12) X lo5 1. moles-' sec.-l at 25" for 0.01 N H2S04. Although it is possible DZ = that the ratio ksoln/kgas for the reaction H HD D a t ordinary temperatures might be as large as two to three," the assumption which we have made seems reasonable, since it has been found1* tliat the conversion of para to orthohydrogen takes place 1.2 times as fast in aqueous solution as in the gas phase when catalyzed by oxygen. The corresponding figure for nitric oxide, a polar molecule which is probably more strongly solvated, is 2.2. This provides evidence that the collision numbers in the two phases are in the same ratio.
+
+
= 23
krOH iDa)
+
(16) J. Hirschfelder, H. Eyring and B. Topley, J . Chem. Phys., 4, 170 (1936).
86 1
TABLE I1 APPROXIMATE RATECONSTANTS FOR R k x 10-5,
+H
AT
25"
1. moles-1
R
see. - 1
0 4 48 6.7 10000 19 1.9c 6.5c 29 900
D 2
F P
Fe" 0 2
PH
Gas
2 . 1 (HzS04) 2 . 1 (HzSOn)
...
HCOOH 3 and 1 CH&OOH 1 (&SO41 CHaCOCHZ 1 (HzSOd) CHaOH 1(HzSOI) HCHO 1 (HzS04) 13oc 1(HzSOI) DCO2Ha EtOH 27OC 1 OLS04) CU2+b 1O6Oc 1 (HzS04) Benzoquinone 57000~ 1 (HZSO,) H+. For D abstraction. *For Cu2+ .+.H = C u + From the data of Baxendale and Smithies, reference 5.
+
A list of approximate rate constants for the reactions of some inorganic and organic compounds with H atoms is given in Table 11. Rothchild and Allen8 have found that the ratio k ( +~ oe)/ k ( +~ FeII') in 0.01 N HzS04 is equal to 208 f 40. Therefore, it follows'that the rate constant k ( +~02) = (10 f: 3) X lo*1. moles-l set.-'. From Hart'sig value for the ratio k ( +~o Z ) / k+( ~HCOOH) = 540 f 80, we find k~ + HCOOH = (19 f 6) X 1051.moles-l sec.-'at pH 3 and 25". This value of k ( +~HCOOH) was used to calculate several of the rate constants of Table I1 from the data of Baxendale and Srnithie~,~who have measured the rates of a number of H atom reactions relative to the rate of the reaction H HCOOH = Hz COOH a t pH 1.
+
+
(17) R. P. Bell, Ann. Rep. Chem. Soc., 86, 82 (1939). (18) L. Farkas and U. Garbatski, Trans. Faraday SOP.,36, 263 (1939). (19) E.J. Hart, J. Am. Chem. Soc., 76, 4312 (1954).
862
P. G. CLAY,G.R. A. JOHNSON
AND
J. WEISS
Vol. 63
THE ACTION OF ‘OCO7-RADIATION O N AQUEOUS SOLUTIONS OF ACETYLENE BY P. G. CUY, G. R. A. JOHNSONAND J. WEISS Department of Chemistry, King’s College, University of Durham, Newcastle on Tyne, England Received February 0, I369
W o ?-irradiation of water saturated with acetylene ( 1 atm.) in the absence of oxygen, yielded Q, C? and Cd-a!dehydes and a white solid polymer. The irradiated solution showed a very strong, broad absorption band m t h a maxlmum at about 200 mp. Aqueous solutions of acetylene-oxygen mixtures, on irradiation yielded glyoxal and hydrogen peroxide, 0.3). The formation of tke radiation products has been studied a t with small amounts of an organic hydroperoxide ( G different ratios of oxygen/acetylene, as a function of p H and in the presence of added ferrous sulfate. Under certain conditions, glyoxal is apparently formed by a chain reaction.
-
Introduction Much work has been done on the radiation chemistry of acetylene, in the gas phase, notably by Lind and his co-workers.’ I n the absence of oxygen, using various radiations, a yellow solid polymer was formed and some aromatic compounds also were produced. I n the case of acetylene irradiated with a-particles, benzene accounted for about 15% of the products.2 In the presence of oxygen, gas phase irradiation of acetylene resulted in the formation of carbon dioxide, carbon monoxide and a liquid polymer.a We have studied the action of W o y-rays on acetylene in aqueous solutions as part of a program investigating the radiation chemistry of aqueous solutions of simple unsaturated compounds. Experimental Preparation of Solutions.-Ordinary distilled water was redistilled from alkaline potassium permanganate and then from dilute sulfuric acid. The pH of this water was about 5.5, solutions of lower H were obtained by adding sulfuric acid and higher values y! the addition of sodium hydroxide. Acetylene was used from a cylinder (British Oxygen Co. Ltd.), after washing with 10% sodium bisulfite solution followed by 20y0 sodium hydroxide solution. After this treatment no acetone could be detected in the solutions p r e pared for radiation. Oxygen (British Oxygen Co. Ltd., medical grade) was used directly from the cylinder. The solutions were prepared and irradiated in cylindrical Pyrex vessels equipped with a side arm and tap. The vessels were deaerated by repeated pumping with a two stage oil pump and shaking. The appropriate gas mixtures were pre ared in a gas buret of conventional design and were admittef to the vessel by the side arm. The vessel and its contents was shaken in contact with the gas to bring about saturation of the solution. For the experiments in the absence of oxygen the acetylene was condensed out in a trap cooled by liquid nitrogen and freed from any non-condensable gases by repeated melting and pumping. The vessel and side arm, previously evacuated, were opened to the container holding the solid acetylene and the latter allowed to warm up. The vessel was agitated during the time when the acetylene pressure was building up in the apparatus and for some time after atmospheric pressure was reached. Irradiation Arrangements .-The samples were irradiated with 6oCo 7-rays from a 500 Curie source of the type d e scribed by Hochanadel and Ghormley.4 The dose rate, measured by the ferrous sulfate dosimeter (10-8 M ferrous sulfate in 0.1 N sulfuric acid) was 0.86 X 10-7 (e.v./N) ml.-1 based on G F . ~= 15.5. (1) Cf. S. C. Lind, “The Chemical Effects of Alpha Particles and Electrons,” 2nd edition, The Chemiaal Catalog Co., New York, N. Y., 1928. (2) C. Roaenblum, THIS JOURNAL, 62, 474 (1948). (3) 8. C. Lind, D. C. Bardwell and J. H. Perry, J . A m . Chem. Soc., I S , 1556 (1926). (4) C. J. Hochanadel and J. A. Ghormley, Rsv. Sci. Instr., 22, 273 (1951).
Identification and Determination of Products.-(i) Hydrogen peroxide was identified and determined by the titanium sulfate reagent.6 Since the products of irradiation, in the absence of oxygen, absorbed strongly at 405 mM, hydrogen peroxide formed under these conditions was measured by the ferrous thiocyanate method. (ii) Organic hydroperoxides, under normal conditions, do not react with the titanium sulfate reagent but oxidize iodide8 and r e act with ferrous thiocyanate.’ Determination of the total peroxide in the irradiated solutions showed the presence of an organic hydroperoxide. AB its yield was rather low (G N 0.3) the identity of this hydroperoxide has not yet been established. (iii) Formaldehyde and acetaldehyde were identified by paper chromatography of their 2,kdinitrophenylhydrazones according to the method of Gasparic and Vecera.8 Crotonaldehyde was identified on the chromatogram by its 2,4-dinitrophenylhydrazone. Further evidence was provided by elution of the 2,4dinitrophenylhydrazone spot with alcohol and the s ectra of the eluted spot in the neutral and alkaline alcohofc solution were compared with those of an authentic sample of crotonaldehyde 2,Cdinitrophenylhydrazone and shown to be identical. The aldehydes were determined by the method of Johnson and Scholes.e (iv) Glyoxal and glycolqldehyde were identified by the purple-blue color of their 2,4-dinitrophenylhydrazones in strongly alkaline solution. The irradiated solution was treated with the dinitrophen lhydrazine reagent and the resulting red ppt. filtered off, g i e d and dissolved in benzene, addition of ethanolic sodium hydroxide produced the characteristic purple-blue color indicating the presence of either glyoxal or glycolaldehyde. The method of Dechary, et al.,10 was used t o distinguish between these two compounds and t o estimate them. The method depends on the formation of a blue colored derivative when solutions of glyoxal or glycolaldehyde are treated with the 2,3-diaminophenazine reagent. Glyoxal reacts with the reagent in acetic acid solution; glycolaldehyde reacts only under conditions where conversion into glyoxal can occur, Le., in 10 N sulfuric acid. (v) Oxygen was determined by the Winkler method” in the following way: solutions of acetylene and oxygen were made up exactly as for irradiation, the side arm was removed and 15 ml. of etroleum ether placed on the top of the solution. One mE of saturated MnC12 solution was added below the surface of the pet. ether and the flask gently shaken until the solutions were mixed. One ml. of solution of 33% NaOH and 10% KI was added and shaken in a similar fashion and the pt. formed was allowed t o settle for 2 minutes when 5 mf: of concd. HC1 was added. The liberated iodine was titrated against thiosulfate.
Results Irradiations in the Absence of Oxygen.-Irradiation of water saturated with acetylene (1 atrn.) (5) G. M. Eisenberg, Ind. Eng. Chem., Anal. Ed., 16, 327 (1943). (6) C. J. Hochanadel, THIS JOURNAL, 66, 597 (1952). (7) A. C. Egerton, A. J. Everett, G. J. Minkoff, 9. Rudrakanchans and R. C. Salooja, Anal. C h i n . Acta, 10, 423 (1954). (8) J. Gasparic and M. Vecera, Coll. Czech. Chem. Communs., 22. 1426 (1957). (9) G. R. A. Johnson and G. Scholes, Analyst, 77, 937 (1954). (IO) J. M. Dechary, E. Kun and H. C. Pitot, Anal. Chem., 26, 449 (1954). (11) C.f. W. W. Ecott, “Standard Methods of Chemical Analysis," London, 1926, p.-1436.
