with a background of 40. This indicated that the DBPO contained less than 0.3 t-butyl peroxide. Yields of Products. The absolute yield of t-butyl alcohol was estimated from total oxygen evolution. However, the ratio of alcohol to peroxide, which is important to our radiochemical results, is obtained by measuring the area under the g.1.c. curves obtained by injecting 2 PI. of either the original product containing unreacted DBPO or product which had been vacuum distilled away from the DBPO onto an 8 ft. X in. column as described above. By comparison with synthetic mixtures they gave weight ratios of 10.5:l and 11.2:1, respectively. Added DBPO did not affect the results. This indicates about 21 moles of alcohol per mole of di-t-butyl peroxide, representing an average chainlength of 10.5. Thus t-BuOH/2(tB~0)~ = 10.5, 02/DBP0 used = 10.5, and average 2d(O2)/d(CO~)= 10.5. The composition of products can therefore be calculated from the DBPO which was used. Alcohol = 21 X DBPO used = 32 mmoles, di-t-butyl peroxide = DBPO used = 1.54 mmoles, t-butyl hydroperoxide = (initial 2BuOOH) - (t-BuOH) = 8.0 mmoles. Calculation of Activities. Of the 10.700 g. total weight used in run 9, 0.094 g. of DBPO remained behind, and 0.138 g. of COZ and 0.515 g. of 0 2 were evolved, leaving 9.935 g. of volatile products of which 9.136 g. were used for dilution with alcohol, peroxide, and hydroperoxide. The activity in the alcohol product based on a chain length of 10.5 is therefore
+
(52.6 mmole 32.4) (0.92 mmole) X (32.4)(0.92 mmole) 18,800 c.p.m./mmole diluted = 52,000 c.p.m./mmole
+
(32.4)(0.92/10.5 26.92 mmole) (32.4)(0.92/10.5) 2803 = 56,050 c.p.m./mmole This represents (56,050)(1) 21.0)(52,000)
(56,050
+
100 = 4.8% of DB'PO activity Decomposition of Cumene Hydroperoxide. A solution of 4.00 ml. of cumene hydroperoxide (2.79 h4) purified by the method of KharaschZ0in 6.00 ml. of chlorobenzene was decomposed at 45.0' in the kinetic apparatus described above. As initiator 0.125 g. (0.055 h4) of radioactive DBPO was used. The DBPO gave a net count of 5230/mg. The average chain length was 7.8 by kinetic measurement and 8.0 over-all (see Table 11). The final solution was cooled to -8OO to condense any volatile materials, and 1.00 ml. (0.793 g.) of di-t-butyl peroxide was added. About 6 ml. of liquid was distilled at 25 O under high vacuum and a small amount of lithium aluminum hydride was added to the distillate and then redistilled. The peroxide was separated by g.1.c. fractionation as described above, shown to contain negligible t-butyl alcohol, and counted. For 84.4 mg. the activity was 4765 c.p.m. This represents a 6.8% yield of di-t-butyl peroxide from DBPO. Acknowledgment. We wish to thank the Academic Senate Committee on Research of the University of California for a grant, and the National Institutes of Health for a Fellowship (A. F.). We are also grateful to Dr. Paul D. Bartlett, Dr. Charles Perrin, and especially to Dr. Richard Hiatt for helpful advice. (20) M. S. Kharasch, A. Fono, and W. Nudenberg, J. Org. Chem., 16, 113 (1951).
The activity in the peroxide is
Mechanisms of Autoxidation. Terminating Radicals in Cumene Autoxidation T. G. Traylor and Carol A. Russell Contribution f r o m the Department of Chemistry, Revelle College, University of California, San Diego, La Jolla, California. Received March 12,1965 The effect of t-butyl hydroperoxide on the rate of cumene autoxidation has been studied. Although t-butyl hydroperoxide does not initiate cumene autoxidation, it accelerates the autoxidation initiated by a,a-azobisisobutyronitrile (AIBN). Furthermore, the rate of autoxidation depends directly on the ratio of cumene hydroperoxide to t-butyl hydroperoxide. These and other data, combined with previous studies, strongly indicate the foIIowing termination mechanism f o r autoxidation of cumene and other tertiary hydrocarbons. The CHa CHa
I I
2 c6HH,-c-oo
CHa
3698
*
+2 C $ I S
?+ I
-0
CHa
*
0 2
CHa C$Iad-O.
I
CHa
+C$I&
+ CH1.
CHa
+ 02 +CHaOO. + CHsOO. +CHzO + (ROOH or ROH) + O2 CHs.
ROO.
deuterium isotope eflect found by Boozer, et al., is aIso explained by this scheme.
Introduction The autoxidations of hydrocarbons at moderate temperatures and oxygen pressures display kinetic
Journal of the American Chemical Society / 87:16 / August 20, I965
behavior requiring that these chain processes terminate by a bimolecular reaction of peroxy radicals. 1i
+nonradical products
2R00.
(1)
The nature of this termination reaction has been the subject of several investigations, 3-7 and several mechanisms have been proposed. In the case of secondary peroxy radicals, which terminate 100-500 times faster than do tertiary peroxy the a-hydrogen has been implicated in the termination step by its isotope effect.4b This termination has been written as a direct one-step reaction. 2RZCHOO.
+R,C=O
+ RzCHOH + 0,
amount of methyl radicals. Thus, if the chain in cumene autoxidation could be carried by t-butylperoxy instead of by cumylperoxy radicals, or if cumyloxy radicals could be rapidly trapped in another way, any termination of methylperoxy radicals could be prevented, resulting in an increase in the rate of oxidation. The possibility of changing the chain-carrying radical has been realized recently by Thomas and Tolman,'O who indicated that reaction 10 has a rate constant of about 12 1. mole-' sec.-l. The following sequence is a very likely one. At equal concentrations of cumene
(2)
Since tertiary peroxy radicals cannot undergo this mechanism, other mechanisms have been proposed.6,6a 2R00.
-c
[
0-0
/
\
.O-R
R-0..
+2RO.
2R00.
2RO.
I
I
+
ROOR
+
0 2
+ ROOR
CH3 CsHs-c-0.
]-
I
I
fast
f
0 2 --f
I
C~Hs-c00*
(8)
I
CH3
CH3 CHn
(3)
(4) (5)
CH3
I
+ C Q H ~ C ~+OCH3.
(6)
CH3 CH:,.
C6Hs-C.
CHa
0,
CH3
CH3
+
0 2
+ CHaOO.
C6H6-(&O0. (7)
Blanchards has related the amount of acetophenone produced during autoxidation of cumene to the chain length and has concluded that, although a certain fraction of cumylperoxy radicals decomposes by reaction 4, the termination step is either reaction 3 or reaction 5. Bartlett and TraylorBa confirmed the occurrence of reaction 4 using oxygen tracers but suggested that, reaction 5 being a poor process in cumene, termination might involve methylperoxy radicals produced by the sequence of reactions 4, 6 , and 7. The purpose of the present work is to determine which of these sequences actually constitutes termination of cumene autoxidation. Since the rate of hydrogen abstraction by ROO. is little dependent on the structure of R,4c it is reasonable to assume that interaction of ROO * by reaction 3 or 4 followed by reaction 5 would be similarly independent of the structure of R. In this event, the rate of autoxidation of cumene is expected to be almost independent of the structure of R in ROO-. On the other hand, the rate of autoxidation of cumene should depend on the structure of R in ROO. if the rapid termination reaction 2 is preceded by the slower reaction 6 . For example, cumyloxy radicals are known to undergo reaction 6 in cumene536 to give methyl radicals. In contrast to this behavior, 2-butoxy radicals exclusively abstract hydrogen in cumene* or in hydroperoxide^,^^^ producing a negligible (1) E. J. Bowen and E. L. Tietz, J. Chem. SOC.,234 (1930). (2) C. Walling, "Free Radicals in Solution," John Wiley and Sons, Inc., New York, N. Y., 1957, pp. 418-427, 442-447. (3) G. L. Bolland, Quart. Rev. (London), 3, 1 (1949). (4) (a) G. A. Russe11,J. Am. Chem. SOC., 78, 1047(1956); (b) ibid., 79, 3871 (1957); (c) G. A. Russell and R. C. Williamson, Jr., ibid., 86, 2364 (1964). (5) H. S. Blanchard, ibid., 81, 4548 (1959). (6) (a) P. D. Bartlett and T. G. Traylor, ibid., 85, 2407 (1963); (b) T. G. Traylor, ibid., 85, 2411 (1963); (c) P. D. Bartlett and T. G.
Traylor, unpublished results. (7) (a) R. Hiatt, J. Clipsham, and T. Visser, (1964); (b) R. Hiatt, private communication,
Can. J. Chem., 42, 2754
I,
ktr = 12 1. mole-' 8ec.-1
+ f-BuOOH
>
CH.3 CH3 t-BuOO.
+ C&-&OOH I
(10)
CH3 CHJ
t-BuOO.
CHI
+ CJIb-CHI +CJI5-L. + t-BuOOH kp
I
I
CHa
Zt-BuOO.
+2t-BuO.
(11)
CH3
+ Oz
(12)
and t-butyl hydroperoxide, cumylperoxy radical abstracts hydrogen from t-butyl hydroperoxide about 24 times faster than from cumene, and therefore the chain would be carried and terminated by t-butylperoxy radicals. For the same reasons, any effect t-butyl hydroperoxide has on autoxidation rate should be decreased by adding excess cumene hydroperoxide. Thus, if termination is occurring through methyl radicals, the addition of a small amount of t-butyl hydroperoxide should give an accelerated initial rate which decreases with time. This and other effects have been observed and are discussed below. Results We have designed a continuous recording gasometer for measuring oxygen absorption or evolution. This apparatus allows accurate instantaneous oxygen absorption rates to be measured even though the rate is changing with time. Furthermore, we obtain a continuous recording of total oxygen absorbed and from this can calculate the concentration of cumene hydroperoxide as a function of time. (The oxygen absorbed is nearly equal to cumene hydroperoxide formed at the long chain lengths occurring in this study.) (8) P. D. Bartlett, E. P. Benzing, and R. E. Pincock, J. Am. Chem. SOC.,82, 1762(1960).
(9) A. Factor, C. A. Russell, and T. G. Traylor, ibid., 87, 3692 (1965). (10) J. R. Thomas and C. A. Tolman, ibid., 84, 2079 (1962).
Traytor, Russell f Terminating Radicals in Cumene Autoxidafion
3699
Table I.
Changes in Rates of Autoxidation of Cumene with Time in the Presence of &ButylHydroperoxide in Chlorobenzene at 60'0 Concn. of Run
Concn. of t-BuOOH, M
2 3 4 4 4 4 4 4 14 14 14 14 14 14 20 20 20 20 20 24 24 24 24 24 24
0 0 0.500 0.500 0.500 0.500 0.500 0.500 0.250 0.250 0.250 0.250 0.250 0.250 2.25 2.25 2.25 2.25 2.25 0.100 0.100 0. 100 0 . 100 0,100 0.100
of Time, sec.
CeHsCM~OOH,b M
...
...
... ...
0 360 1260 3000 4500 5400 0 240 840 1680 3000 4440 0 780 1380 2040 2580 0 900 1560 2280 3840 5040
0.0 0.0124 0.0540 0.128 0.175 0.203 0 0.011 0.0467 0.0752 0.126 0.173 0 0.047 0.085 0.125 0.159 0.0379 0.088 0.110 0.143 0.191 0.226
a [Cumene] = 3.59 M , [AIBN] = 0.100 M . by cumene concentration.
(11) These conditions correspond to the terminal oxidation rates discussed by Tobolsky, Metz, and Mesrobian.12 However, their kinetic
Journal of the American Chemical Society
1 87:16
lo6 X rate O2 absorption, mole 1.-1 set.-'
106 X ratedl cumene sec.-
1.53 1.45
0.425 0.403
4.75 4.33 3.50 2.88 2.98
1.32 1.21 0.97 0.80 0.82
4.77 4.70 4.26 3.62 3.12
1.33 1.31 1.19 1.01 0.87
6.06 6.24 6.24 6.16
1.69 1.74 1.74 1.72
...
...
4.06 3.79 3.35 3.01 2.75
1.13 1.06 0.932 0.838 0.764
1 .OO 1.OO 0.0 0.024 0.0976 0,204 0.259 0.288 0 0.043 0.158 0.231 0.335 0.430 0.0 0.02 0.037 0.053 0.066 0.275 0.469 0.529 0.590 0.656 0.694
Including that produced by autoxidation.
The effect of added t-butyl hydroperoxide on the rates of cumene autoxidation can be seen in Table I and Figure 1. The plot of rate (d[Oz]/dt) against time in Figure 1 shows the usual zero-order reaction in the absence of t-butyl hydroperoxide (D) or in the presence of large concentrations of t-butyl hydroperoxide (A). Furthermore, a large quantity of tbutyl hydroperoxide causes the zero-order rate to increase by a factor of 4.0, even though no reaction occurs until the initiator, AIBN, is added. When a small amount of t-butyl hydroperoxide is used, the rate is initially high but rapidly decreases (B). It is also demonstrated in Figure 1 that addition of cumene hydroperoxide to solutions containing t-butyl hydroperoxide decreases the rate acceleration (C). The rate of cumene autoxidation as a function of tbutyl hydroperoxide concentration was determined by measuring initial oxidation rates at concentrations of tbutyl hydroperoxide from 0 to 10 M (pure t-butyl hydroperoxide). The data are shown in Table I1 and plotted in Figure 2. Although the rate acceleration for small concentrations of t-butyl hydroperoxide is difficult to obtain because of the rather rapid rate decrease with time (see Figure l), reasonable estimates are available by extrapolation to zero time as shown in Figure 1. It is immediately apparent from Figure 2 that small concentrations (0.1 M ) of t-butyl hydroperoxide bring about a fourfold increase in autoxidation rate, but that this does not increase further with more t-butyl hydroperoxide even up to 80% hydroperoxide. At 9 M t-butyl hydroperoxide (90 % concentration) oxygen evolution competes with oxidation, and the rate of oxygen absorption decreases.
3700
Mole fraction of C6H6CMe2OOHc
...
...
c
Of total ROOH.
...
Observed rate divided
Table II. Effect of t-Butyl Hydroperoxide on Rates of Autoxidation of Cumene in Chlorobenzene at 6O.Ooa
Run
Concn. of cumene, M
2b 14 4 5 6 20 7 8 25 9 10 11 17 12 16
3.59 3.59 3.59 3.59 3.59 3.59 3.59 3.59 3.59 3.59 2.87e 2. 156 1.795" 0.7186 0.3596
106 x Concn. of d[Oz]/dt, (106 X d[OJ dr)/cumene, f-BuOOH, mole M 1.-' set.-' set.-' 0 0.250 0.500 1.00 1.50 2.25 3.00 4.00 4.50 5.00 6.00 7.00 7.50 9.00 9.50
1.527 4.79 4.79 5.58 5.92 6. 17d 5.98 5.94 6.14 6.00 4.72 4.02 2.95 0.96