H. CESSER, J. T . MULLHAUPT AND J. E. GRIFFITHS
4534
length is similar and the reactions corresponding to reactions 8, 9 and 11 are 13, 14 and 15, respectively.
+
CHzCO hv +CHzCO* CHzCO* +2CHzCO CHzCO
+
(12) (13) (14) (15)
Vol. 79
However, a longer lifetime would be expected for (CH3)2=C=C=O* because of the extra degrees of freedom in this molecule.
Conclusion The results reported in this paper indicate a similarity in the photochemical behavior of the three The ratio k?13/k14 for ketene is 4.7 X l o 4 liters/ ketenes investigated so far. The primary process is mole.6 The close agreement of the two ratios: wave length dependent and similar in ketene and dimethyl ketene, but secondary reactions are difk]3/k14 and ks,/kg, is indicative of the similarity of the primary processes in ketene and dimethyl ke- ferent. Ketene, itself, is unique in that methylene tene. An internal conversion of CH&O* to the cannot isomerize to a stable molecule as can ethyliground state is important in ketene’b and appears t o dene and isopropylidene. The rate of isomerization of isopropylidene is probably faster than the rate be important in dimethyl ketene. The value of kll/ of isomerization of ethylidene since six hydrogens kg compares well with the value of k16/k14 which is are available to shift; however, there is no evidence of the order of 26 at room temperature. to verify this hypothesis. A calculation of the mean life of (CH&=C=C= Acknowledgment.-The authors are indebted t o O* from the integrated absorption coefficient gives hlr. R. Vanselow for his assistance and helpful sug7 = 2 X 10-6 second, which is longer than the mean life for CH&O* which is 0.3 X second.6 gestions.
+
CHzCO* CH* CO CHzCO* +CH,CO
Los ANGELES, CALIF.
(15) G. B. Porter, THISJOURNAL, 79, 827 (1957).
[CONTRIBUTION FROM
THE
DEPARTMENT OF CHEMISTRY, UNIVERSITY OF ROCHESTER, AND THE DEPARTMENT OF CHEMISTRY, OF MANITOBA] UNIVERSITY
The Photolysis of Trimethylaminel BY H. GESSER,J. T.MULLHAUPT AND J. E. GRIFFITHS RECEIVED APRIL 10, 1957 The photochemical decomposition of trimethylamine was investigated over the temperature range of -78 to 175’. The products were hydrogen, methane, ethane and a “liquid.” At high temperatures the methane and ethane are primarily formed by the abstraction and recombination reactions of methyl radicals and the hydrogen is formed by a molecular reaction. At low temperatures both methane and hydrogen are probably formed by molecular elimination reactions.
Introduction The photochemical decomposition of the amines have not been investigated in great detail. The photolysis of methylamine2 seems complicated by the possibility of more than one primary process. The mechanism of the photolytic decomposition of dimethylamine and trimethylamine is even more uncertain. Bamford3 had shown that the products of the photolysis of trimethylamine a t 100’ were hydrogen, methane, ethane and polymer. The mechanism he proposed to account for these products is
+
+
(CH8)sN hv +(CHdtN CHI 2(CHa)zN +(CHJzNH CH&CHz (CH3)tNH
+
+ hv +(CH3)zN + H
+
(CHdzNH H --+ (CHJzN CH1+ (CHs)zN (CHJzNH H 4- CHI +CHI
+
+ Hz + CH,
(1)
(2) (3) (4) (5) (6)
2CHs +CzHs (7) CHsNCHZ+CHa-N=CH2 +Polymer (8)
This mechanism must be revised in view of the com(1) This work was supported in part by 8 Naval Research Contract S.Navy and by the N a t ~ o n a lResearch Council of Canada. (2) For review see W. A. Noyes, Jr., and P. A. Leighton, “The Photochemistry of Gases,” Reinhold Publ. Corp., New York, N. Y., 1941. (3) C. H.Bamford, J . Ckcm. SOC.,17 (1939). of the U.
TABLE r
System A; Reaction time, 4.5 hr. [TMA] [C,Hla] Temp., molecules/cm.a X 10-1’ OK.
1.61 1.62 1.40 1.39 1.22 1.22 1.09 1.09
1.44 1.28
1.22 1.09
299 298 347 349 398 400 448 445
RCHC Rcma
Rnn
molecules/cm.~/ sec. X 10-12
0.58 0.50 1.21 1.04 1.99 2.45 3.42 3.62
1.51 1.46 1.52 1.56 1.79 1.43 0.88 0.29
ki/ki 112 (mol./cm.’/ sec.) 1 1 2 X 10”
0.96 0.72 1.05 0.98 1.55 1.20 1.50 1.21
3.9 7.4 12.2 33.4
paratively low value of 8.8 k ~ a lfor . ~ the activation CHI
+ (CHs)sN -+- CH4 + (CHs)zNCHz
(9)
energy of the abstraction reaction in comparison to 7.2 k ~ a lfor . ~reaction 5. It was believed that the mechanism could be tested by a study of the product yields in the presence of a hydrocarbon and a t low temperatures where reactions requiring high activation energies would become unimportant. Cyclopentane was chosen as the hydrocarbon because its vapors did not interfere with the analysis of the products and because it has a low value for the activation energies of abstraction by hydrogen atoms, 7.7 k ~ a land . ~ methyl radicals, 8.5 k ~ a l . ~ (4) A. F. Trotman-Dickenson and E. W. R.Steacie, J . Chem. P h y s . , 19, 329 (1951). ( 5 ) H. I. Schiff and E. W. R. Steacie, Can. J . Chcm., 29, 1 (1951).
Sept. 20, 1957
PHOTOLYSIS OF TRIMETHYLAMINE Experimental
The experiments were performed on two different apparatuses. System A (Rochester) consisted of a reaction system similar to that used previously6 and was free of stopcocks, mercury cut-offs being used to isolate parts of the vacuum apparatus. The light source was a Hanovia S-100 Alpine Burner. The parallel beam was collimated by one quartz lens and a stop, and was made to fill the quartz cell. The cell (vol~ ) 2.5 cm. in diameter ume, 50 ~ m . was and 10 cm. long and was enclosed in a brass block. System B (Manitoba) was similar to that of McElcheran and Steacie7 and consisted of a mercuryfree vacuum line in which reactants were stored and the photolysis done; and an analysis line free of stopcocks utilizing mercury float valves. The two units were separated by two stopcocks and a small oil diffusion pump. The light source was a Hanovia S-500 medium pressure mercury arc. The light beam was collimated as above. The cell (volume, 130 cmS8)was 2.6 cm. in diameter and 25 cm. long. The cell was designed for low temperature studies and is shown in Fig. 1. The volume of the cell and connecting tubing was 260 cm.8. The low temperatures were obtained in all cases by acetone-Dry Ice mixtures. The pressures of reactants were measured bv -~ a silicon oil manometer in conjunction SCALE: cm, = with a bourdon gage which was used as a null point indicator. ternThe analvses were uerformed bv utilFig. perature cell: A, izing a trap for solid nitrogen t o sepauartz-Pyrex seals; arate the methane and hydrogen from vacu;m jacket; the ethane, and a Le Roy still for fractionations. A copper-copper oxide sides; furnace maintained a t 215’ was used to D, coolant level; analyze the hydrogen. The trimethylE, cemented quartz amine was obtained from Eastman white label 2570 aqueous solution. It window. was dried by passage through potassium hydroxide pellets and by, repeated distillations from - 7 8 O and degassed a t -140 . The cyclopentane was Phillips reagent grade and was used after bulb-to-bulb fractionation and degassing. The intensity of incident light was varied by means of neutral density filters of aluminum evaporated on quartz. Degassed trimethylamine was distilled into the cell through a trap maintained a t -80”. After an experiment it was observed that some gases were produced which were not volatile a t -75”. These “liquid” products could not be identified but were estimated in System A to be from 10% to 50% of the hydrogen formed.
1
4835
(where R represents “rate of formation”) is the possible quenching effect of cyclopentane on an excited trimethylamine molecule. This would seem to be important at all temperatures. TABLE I1 System B rr. Rela. Mi1 tive mole/ incident cc. X Temu.. Time, inOK. sec. 10-18 tensity
0.20 1.5 0.69 0.27 0.11 1.4 0.77 0.92 0.84 0.83
193 213 210 212 211 213 213 211 211 212
1800 1800 1800 1800 900 1800 1560 1800 3600 5400
100 100 100 100 100 100 100 48 26 4.3
RCHC RCZHO Rnn molecules/cc./ sec. X 10-12
1.19 1.78 1.77 1.67 1.72 1.67 1.48 0.81 0.74 0.20
3.45 3.96 4.27 4.15 4.00 3.69 3.65 1.84 1.86 0.41
R s . RHS-I- &s Rcai: Rear
2.27 1.00 1.88 1.08 2.63 0.9;‘ 2.99 0.89 2.99 0.86 1.85 1.04 2.37 0.95 0.98 1.03 1.20 0.96 0.26 0.89
1.9 1.2 1.5 1.8 1.8 1.1 1.6 1.2 1.6 1.3
If i t is assumed that methane and ethane are formed only by reactions 9 and 7, respectively, then i t is possible t o evaluate E3 - ’/2E7 by an Arrhenius plot of the relation R C H ~- = (Rct~ti) ‘1%[TMA 1
ks/k7‘/2
(where [TMA] represents the concentration of trimethylamine). Such a plot is shown in :Fig. 2 where comparison is made with the results of Trotman-Dickenson and S t e a ~ i e . The ~ value of
i, cp
Results and Discussion Preliminary experiments on the photolysis of trimethylamine, in which decomposition was carried to less than 1%) showed that the products were methane, hydrogen and ethane and no trace of polymer. Thus, unless dimethylamine shows very strong mercury line absorption, the reaction sequence postulated by Bamford3 for hydrogen production must be erroneous. The rates of formation of the products are given in Tables I and 11. The results a t high temperature show that in all cases the rate of hydrogen formation decreases when cyclopentane is added and would seem to indicate the absence of hydrogen atoms in the reaction system. Another possible explanation for this observed decrease in RH^ (6) H. Gesser. THIS JOURNAL. 77, 213% (1955). (7) D . E . McElcheran and E. W. R Steacie, Can. J . C h e m . , in press.
:
2.4 2.8 3.2 10s/T. Fig. Z.-Arrhenius plot of k 9 / k 7 1 / 2 . Open circles are from the work of Trotman-Dickenson and Steacie. Filled circles represent the present work. 1.6
2.0
E9 (ETis assumed to be zero) for the two highest temperatures is 7.1 kcal., as compared to Steacie’s value of 8.8 kcal. This discrepancy is probably due to an additional step for methane formation at the lower temperature. The change in slope of the Arrhenius line rules out the alternate explanation,
namely, an additional step for ethane formation a t the high temperature. The disagreement between the intercept values may be due to strong absorption for trimethylamine (in comparison to acetone) and the consequent smaller effective volume. This also accounts for the small effects observed at low temperatures when the concentration of trimethylamine and the incident intensity were variqd. Trimethylamine begins to absorb a t about 2370 A.8 with an extinction c?efficient of about 0.13 X 106 mole a t 2500 A: which increases to 1 X lp6 cm.2/mole a t 2330 A. and 4 X lo6 a t 2000 A. These high extinction coefficients imply transinittancy values much less than 0.05% in agreement with the above conclusion. In the presence of cyclopeiitane, methane can also be formed by the reaction CH3 Cyclo-CsHio + CHI -t C>CIO-C&I~ (10) It is possible to calculate4 a value of El0 (positive for the two highest temperatures only) of 14.3 kcal. This value is about 6 kcal. too high and signifies an incomplete mechanism. Ai reaction sequence utilizing an excited intermediate in the primary process can account for the results
querice is consistent with the results a t low tciiiperature where a decrease in coilcentration of triI.~ methylamine from 0.27 X lo1? ~ o ~ . / c ~toI 0.11 X lo1* mol./cm.3 is accoinpanied by a slight increase in RcH,. At low temperature, the results (Table 11) show that R c ~ H , / ( X C H ~ R H ~is) approximately equal to unity (0.97 O.OG), and that RH,IRcH,has an average value of 1.5 with a mean deviation of i 0.24. It is reasonable to assume that reactioii 9 does not occur at the low temperature. Hence if all methyl radicals react t o form ethane, then thc methane and hydrogen are formed by- the disI)roportionation of the dimethylamine radical. I t is therefore proposed that a t low temperatures, reactions 1I , 12, 13 and 7 are followed by the reaction
+
k, , +1 1 2 + s 3(C1I3)*S--~
+
+ +
(CH3)3?; IZV +(CEIJ,Xr (CH3)3N* hl --+ (CH3jjX RI (CH3)3S*--+ (CH3)2?; CH3
-
(11) (12) (13)
+
where M is a molecule or the wall. Such a se(8) E Tannenbauni P h p , 21, 311 (1953)
I ,
E hl Coffin and 4 J Harrison J
Cheiiz
----+C€Jd
-1- Y
ki,
wlicre k , ' k b is 1.2and X and ' I arc the liquid 1)roclucts arid perhaps dimethylaniinc. I t is hoped that a detailed study of the reactions of the dimethylamine radical will further clarify the mcchanism. Acknowledgment.-The authors are indebted t o I l r . George Ensell of the National Research Council of Canada for the construction of the low temperature cell. WINNIPEG, CANADA
,C(lX'l'RIBU'l'l(JS I X U M l'HE
D E P A K ~ X E01'S ~CIIEhfIS~K\',~ ' 0 I . Y T E C I L S I C
I
I N S ' I I l ' U l I< 01; ~ ~ l i O O K l . Y S
Photoreduction of Methylene Blue by Ethylenediaminetetraacetic Acid'"," BY GERALD OSTERAND KEIL ~VOTHERSPOOI~ RECEIVED APRIL1, 1957 hlcthyleiic blue ill tlic presciice of etli~~leiiccliaminetetra~~cetic acid (EDTA) is reduced to the leuco t l y ~O I I irrdiatiiiii nith red light. The rate of photoreduction depends upon PH in the same way a s does the b,tse titration of EDTA. E D T A is consumed in the reaction suggesting that it is osidized although it does not normally fuiiction as a reducing agent. A number of nitrogen-containing chelating agents were tested but only those wth secondary or tertiary nitrogens 1 ) e I ~ : t v ~ d as electron donors in the photochemical reaction. T h e photoreduction involves a lo:ig-lived excited st:ite of the d y e ( IO5 times that of the first electronically excited state) and is retnrded by srnzll amounts of p-phenylenediatnine. T!lc rate of reageneration of the dye by near ultraviolet irradiation of the leuco forin increases with increasing hydrogen ion coiiceritr:ititiil
Introduction Early workers in the development of photographic bleach-out processes noted that thiazine dyes are photoreduced readily in the presence of mild reducing agents such as allylthiourea, anethol, glyoxal2 and ferrous ~ u l f a t e etc. ,~ When a thiazine dye is photoreduced by ferrous ion t o give the leuco dye, the ferric ion thus produced oxidizes the leuco dye in the dark, and for intermediate illu(1) (a) This paper represents a part of the dissertation t o be submitted by Xeil Wotherspnon t o the faciilty of the Gradirate School the Polytechnic Institnte of n r o o k l , ~i ~ n ~partial fufillineiit o f the requirements for the degree of Doctor of Philosophy. ( b ) This researvh was supported by the United Statr.s Air Force thriiiigh the Air Force office of Scientific Research o f tile Air Reiearcli auql 1 ) t : v i ~ l o p n l e n t Command under Contract KO, AF 18(600) 1182. (2) hZ. Aludrovcic. Z.wiss. Photo., 26, 171 (1928). (3) IC. Weber, 2. p i i y s i k . Chei~z, B15, 18 (1931); G. IIolst, ;bid., B169 9 (19341. cjf
mination levels, a steady state is achieved.4 Stannaus chloride and ascorbic acid, on the other hand, do not allow the leuco dye to revert to its colorcd form, but under excitation with near ultraviolet light the dye is r ~ g e n e r a t e d . ~ Nickerson and Merke16 in their studies 011 the photochemistry of riboflavine noted that methylene blue is photoreduced with ethylenediarninetetraacetic acid (EDTA). This is an unexpected result since EDTA does not normally function as a reducing agent. It is the purpose of the present paper to investigate the role of EDTA as Ltn elcc(4) J. U'eiss, T r n i i i . i : n i o ~ / u r S o c . , 32, 1131 (l!l3(i): 35, 18 (19:3:)),, See also 1%;. l < a t ~ i n o w i t ~"hI ,' l ~ i t ~ r s y n t h r Interscience Publishers, I n c . , New I'ork, S . Y . , l!bG. p. 7ii. ( 5 ) 0 Oster and N. TVotherspnon, J . Chenz. Piirs., 22, 157 (1954). ( 6 ) J . I
