April, 1949
KINETICS OF
THE
DECOMPOSITION OF POTASSIUM PERSULFATE IN METHANOL 1419
TABLE V COMPARISON OF KINETICCONSTANTS OF SOME~CHLOROETHYLAMINES C1CH2CH2NRR'IN 2 :1 ACETONE-WATER SOLUTIONS AT 25" R
Principal A ~ K A ~ reaction
R'
ki
Constants in 2 :1 acetone-water (time in minutes) k-1 ka kw
15 nol is the average of three runs yielding constants of 0.610, 0.640 and 0.647. In a study of the polymerization of allyl alcohol 10 in aqueous solution made in this Laboratory in 1944,18the decomposition of the initiating persulfate was observed to deviate from first-order kinetics. Because of the polymerization occurring, the 0; nature, functionality, and concentration of the al0 5 10 15 20 25 30 35 cohol was changing during these experiments ; howMinutes. ever, we have now found that in the one experiment Fig. 2.-Decomposition of persulfate (0.01375 M ) a t with a high concentration of allyl alcohol (6.308 in presence of 0.488 Mmethanol, plotted as 79.8" and pH 5.02 M during the run) the decomposition of persulfate was clearly of the 3/2 order, while in the a 3/2 order reaction. The arrow marks 90% reaction. other experiments the kinetics more closely approximates this order than i t does first order. In bution from such a recombination would make the the presence of the high concentration of allyl al- observed first-order rates vary with initial concencohol a t 55' the 3/2 order rate constant was 0.33 tration of persulfate. Furthermore, if sulfate ion radicals and hydroxyl radicals combined they liters1I2 min.-l. would form Caro's acid, which Palmelo found to be absent under these experimental conditions. TABLE I The inclusion of Equation 3 in the above scheme RATECONSTANTS AND FORMALDEHYDE FORMATION IN THE does not imply the exclusion of a similar disproporDECOMPOSITION OF 0.015 M POTASSIUM PERSULFATE AT tionation between a hydroxyl radical and a sulfate 79.8" * 0.05' AT gH 8 Per cent. ion radical. I t seems, however, that with water Initial formaldehyde, as the solvent Equation 2 represents the most methanol based on concn., ks/, liters'/z ks/2 peroxide probable fate for the sulfate ion radical. I t will be moles/l. moles min. d/tCHaOHI decomposed noted that although the sequence Equations 1-3 1.952 *. .. 100.5 involves free radicals, initiation, termination and 1.464 1.04 0.86 99.1 transfer, there is no chain propagation reaction in .80 92.9 0.976 0.79 the sense of attack on the starting material by rad0.488 0.63 .90 82.8 icals. 0.195 .. .. 72.9 By comparison of the over-all rates of decomposition of persulfate with and without methanol Discussion At no point in the investigation has any evi- present we may conclude that in 1 M methanol dence appeared that the decomposition of persul- 96% of the persulfate ions disappear by being atfate ion can be induced by any radicals normally tacked by a free radical other than those present in involved in its aqueous decomposition. Therefore the water reaction. This new free radical must the simplest interpretation of the mechanism is by also be one which is present a t a steady-state concentration proportional to the square root of the some such scheme as the following concentration of persulfate ions. These requireki ments are met by the following scheme.
y
t
5t
-1
&OB-
Soil
+ HpO
20H.
+ 2so4'
(la)
kz
+ HSOa- 4- OH.
kia
ka
+ Hz0 + ' / 2 0 z
(3)
Although there are two kinds of free radicals involved in this chain, admitting of three possible chain-terminating steps, yet the order of the reaction tells us that the sulfate ion radicals SO4& do not recombine with one another appreciably, for if they did so the reaction would be of half order with respect to persulfate ion, and even a small contri(18) M. E. Fleischer, unpublished experiments.
s2os- ---f
(2) SzOs=
2S04'
(14
+ CHIOH +HSO4- f SO,' + CHzOH' kib
kd
(lb)
+
f CHIOH + HSOdCHzOH' (4) ks CHzOH' f Stoa= + HSOI- f Sod' f CHz0 (5) ke 2CHzOH' 4CHsOH f CHpO (6)
SO,-
It would of course be possible that Reaction 2 should occur here as well as in the absence of methanol, and the methanol be attacked by the result-
1422
PAULD. BARTLETTAND JOHN D. COTMAN, JR.
Vol. 71
ant hydroxyl radicals instead of directly by sulfate tion 6 in preference to a direct union of the two hyion radicals. I n the related problem of the emul- droxymethylene radicals to yield ethylene glycol. sion polymerization of allyl acetate, l 1 although The high yield of formaldehyde is scarcely concluthe concentration of the organic compound in the sive on this point, since the chains are obviously aqueous phase was considerably less than here, it long enough to limit the occurrence of chain-terwas definitely shown that a t least 75% of the mination products t o a few per cent. of the total. polyallyl acetate chains were initiated by sulfate Any of these indicated variations could be introion radicals rather than by hydroxyl radicals. l8 duced into the proposed reaction scheme without Allyl acetate, therefore, competes very success- vitiating its kinetic consequences. It should be noted that the evidence for the fully with water for the sulfate ion radicals, and i t absence of chain decomposition of persulfate in seems probable that methanol does likewise. The sequence of steps la-4-5-6 leads to the water in the absence of organic compounds is of a negative character and is not altogether complete. kinetic equation I n the decomposition of benzoyl peroxide in ether solutions the kinetics, as here, is that of a first order reaction, yet induced decomposition is readily It is reasonable in view of the strong acceleration suspected because the reaction is so much faster by methanol that the second term of this expres- than in other solvents, and the presence of free sion should be more important than the first, and radicals other than those of oxygen is shown by the such a mechanism would therefore predict the ob- uptake of oxygen during the reaction with the served 3/2 order kinetics. The sequence of steps formation of new peroxides. I n the case of persullb-4-5-6 leads t o the slightly different equation fate decomposing in water solution there can be no ’ free radicals other than those of oxygen, and oxygen-sensitivity is therefore not to be expected even if a long chain reaction were taking place. There is a possible mechanism fsr this reaction This equation, again with neglect of the first term which includes induced decomposition of persulon the right, predicts not only the 3/2 order of the fate ions by hydroxyl radicals, and which still reaction but also the linear dependence of the rate predicts decomposition of the first order. Such a constant on the square root of methanol concen- mechanism includes adding to Equations l a and 2 tration (Table I). Equations (7) and (8) are de- the following ks rived on the assumption that every sulfate ion OH’ 5208- ---+ HSOasod’ +l/zOz (9) radical reacts with methanol to produce a hykio droxymethylene radical. As the concentration of OH’ + Sod’ + HSOd‘/zOz (10) methanol is reduced to the point where this is no The induced decomposition of persulfate by hylonger true, Equations 1-6 predict that an increasing fraction of the sulfate ion radicals will react droxyl radicals will be of the first order now only if with water (or possibly with the accumulating Equation 10 represents the mode of chain terminaformaldehyde), leading to shorter chains and tion rather than Equation 3. This possibility is smaller total amounts of formaldehyde, as ob- mentioned only for completeness; as far as our evidence goes, persulfate decomposition in water served. It is assumed .that the removal of hydrogen may well be an example of the case, unusual to date, in which the spontaneous, unimolecular defrom methanol occurs from carbon, leading to a composition is uncomplicated by any induced hydroxymethylene radical rather than from oxy- process involving the free radicals of the chain. gen to yield a methoxyl radical, since the relative Summary normal energies of the C-H and 0-H bonds (87.3 and 110.2 kcal. mole, respectively)20 make the The decomposition of potassium persulfate a t former process seem much the more probable. 79.8’ in aqueous solution at fiH 8 is strictly of the Reaction 5 , as depicted, may well be an over-all first order, with no evidence of induced decompoprocess which actually occurs in two steps sition by free radicals analogous to that observed CHzOH’ -t S~OS= d sod’ + HOCHZOSOI- (5a) in the decomposition of diacyl peroxides in organic solvents. The presence of methanol or allyl alcoHOCHzOSOs- +CHzO HSOI(5b) hol accelerates the reaction (as much as 25-fold in Very little is known as yet concerning the mode of attack of free radicals upon peroxides in the in- the case of methanol) and the alcohols change the reaction to one of 3/2 order. In the reaction with duced decomposition. Finally, there is no definite evidence for Equa- methanol formaldehyde is produced, its amount being equivalent to the persulfate decomposed (19) Of the polymer molecules, 75% contained sulfate end-groups. when the initial methanol is 1.5 M,and becoming The fraction of the kinetic chains initiated by sulfate ion radicals less a t lower methanol concentrations. Mechamust be at least as high as this and may be higher because of chain transfer leading t o some polymer molecules without any end-group nisms are presented to account for these observafrom the initiator (ref. 11). tions. (20) L, Pauling, ‘Wature of the Chemical Bond,” Cornel1 Uni-
+
+
+
+
versity Press, Ithaca, N. Y., 2nd edition, 1940, p. 53.
CAMBRIDGE, MASSACHUSETTS RECEIVED OCTOBER 11,1948
