Oct. 20, 1964
REACTIONS OF PEROXIDE RADICALS
total rate, and the most likely explanation'j of this discrepancy is t h a t some unknown complication occurs in the rate of thermal initiation a t low rates of polymerization. Figure 3 also shows the same relationship using molecular weights calculated by the OliveSchulz equation, and these data are discussed below. Data at 80".-Table I11 gives data for the polymerization of styrene by butyl peroxide a t 80". Unfortunately, dioxane was used for these higher temperature runs before the peroxide-benzene-styrene system was chosen as the standard system. The transfer constant calculated on runs using greater than 0.9 -If peroxide is 29 X A graph analogous to Fig. 4 gives C = 25 X lop4. -4 graph of the data in the form and using of eq. 2 gives a line with slope 1.53 X 62 = 260 a t XO', a value of k d J ' can be obtained as 6.1 X
set.-'. TABLE I11 THEPOLYMERIZATIOX O F STYREXE BY BUTYLPEROXIDE AT
80""
[?I I O ~ ' F 104/Fd c x 1 0 4 ~ 0.47 4.54 0.655 11.1 11.9 38.5 0.93 6.44 0 494 16.4 17.8 32.2 1.42 7.72 0.417 20.7 22.5 28.6 1.89 8.89 0.377 23.8 25.9 24.9 2.36 10.87 0.300 32.5 35 9 31.2 2.84 11 43 0.283 35.2 39.1 28.8 a Styrene molarity held constant a t 4.3 using dioxane as solvent. 6 At 80", R P , ~is~ negligible. , e Gregg and M a y o relations used. See ref. 9 and 10. OlivC-Schulz relations used. See ref. 11. (1)
R P X 106'
Activation Energies.--The values of k d f a t 60 and 80' of 3 . 3 X l W 9 and 6 1 X 10F8 give an activation energy for the dissociation reaction of 33.3 kcal./mole. A value between 35 and 37 kcal./mole would be expected from studies of other alkyl peroxides.16 Using the data in dioxane as solvent, the values of C a t 60 and a t 80' are 7.6 X l W 4 and 29 X giving an apparent activation energy for the transfer reaction of 16 kcal.,/mole. If Ep is taken as 7 . 3 kcal.,/mole, then Et,-is 23 kcal./mole. (16) (a) P. L. H a n s t and J . G. Calvert, J . P h y s . Chem., 63, 104 (1959); (b) L. B a t t a n d S . W. Benson, J . Chem. P h y s . , 36, 895 (1962).
~COYTRIBUTION FROM
THE
4237
OlivE-Schulz Molecular Weights.--Henrici-Olive, Olive, and Schulz have developed a method for calculating molecular weights which attempts to take into account the change in molecular weight distribution which accompanies chain transfer. Their method involves calculation of a viscosity-average degree of polymerization and correcting this to a number-average degree of polymerization using a function which depends on the rate of polymerization and on 62. Tables I and I11 give the degree of polymerization calculated by both methods. =It low peroxide concentration, where the amount of termination by transfer is small, the two methods agree very closely. However, as expected,l 1 the agreement becomes poorer a t higher peroxide concentrations. Figure 1 shows this graphically ; the solid circles are the peroxide-dioxane-styrene data calculated using the Mayo equations, and the squares are for the Olive-Schulz equations. As Olive has pointed out, the Olive-Schulz equations lead to slightly higher transfer constants. This is best seen in Fig. 3 , where the open circles give the results of calculations using the Olive-Schulz calculations, and the instead of transfer constant is obtained as 10 X
8 x 10-4.
At this stage, however, the results do not appear to be worth the effort. The Olive-Schulz calculations require knowledge of S2 in order to calculate the degree of polymerization, and the correct value of a2 remains the largest imponderable of any method. For our purposes, it seems better to calculate the degree of polymerization in the same way for each peroxide and to regard the data as accurate only within the series of similar compounds. K'e hope to have a better idea of the absolute. accuracy of the transfer constants when we complete our measurements of transfer constants for these peroxides using the radioactive tracer method. Acknowledgment.--This work was supported in part b y Contract A T (11-1)-11(9 (at Purdue) and by A T (40-1)-3180 (at LSU) from the A4tomicEnergy Commission and by Public Health Service Research grants from the National Institutes of Health a t both Purdue and L.S.U. Grateful acknowledgment is made to the donors of these funds.
DEPARTMEYTS OF CHEMISTRY UF LOUISIAYA STATEUZIVERSIIY, BATONROLGE,LOUISIAYA, AND P U R D C E UNIVERSITY, LAFAYETTE, I N D I A Y A ]
Reactions of Radicals.
XI.
Ethyl Peroxide, Isopropyl Peroxide, and sec-Butyl Peroxide
BY WILLIAMA. PRYOR,'" D. M. HOSTON, T. R . FISKE, T. L. P I C K E R K NANI) G,~~ E. C I U F F A R I N ~ ~ RECEIVED J A N U A R Y 2 5 , 1964 T h e dissociation constants and chain transfer constants are given for ethyl peroxide, .rec-butyl peroxide, and isopropyl peroxide a t 60 and 80' in styrene solutions. These data are compared with cumparable data previously reported for propyl peroxide, butyl peroxide, and t-butyl peroxide. A11 these dialkyl peroxides have set.-', activation energies for dissociation of 31---37kcal. /mule, and transfer values of kdj' a t 60' of about constants in the range of 3 t o 9 X 1 O P a t 60". I t is proposed that the predominant tnechanisrn for transfer in this system is hydrogen abstraction. Preparation of the di:rlkyl peroxides and 1i;tz:irds i n handlirig thern arc discussed, The transfer constant of l-butyl ether is also reported.
This paper reports data on the three peroxides named in the title. N o previous kinetic studies have been reported on either isopropyl peroxide or sec-butyl peroxide' Data On the unimolecular decomposition
of ethyl peroxide have been reported in the gas phase by three groups of worker^,^--^ but no previous liquid phase results have been published.
( I ) (a) D e p a r t m e n t of Chemistry, Louisiana S t a t e University, Baton
(19381. ( 3 ) K. E. Kebbert and K . J. Laidler, J . ( ' h e m P i i y s , 20, 074 (1952). (4) K Moriya, Rev. P h y s . Chem J a p a x ( H o r i b a V c ~ l , , ,1 4 3 (1946).
Rouge, L a . , ( b ) National Science Foundation Undergraduate Kesrarch Paiticipant during 1962--1963 a t P u r d u e University; (c)Post doctoral Fellow 1962-1964.
(2) E. J. Harris a n d A . C. Egerton, Proc.. R o y . .So(.
([.ondon,,
A 166, 1
4238
PRYOR, HCSTON,FISKE, PICKERING, AND CIUFFARIN
Secondary alkyl peroxides are difficult to obtain in good yield. Primary alkyl peroxides, including ethyl peroxide, can be obtained in about 50% yield using the appropriate alkyl methanesulfonate and the synthesis developed b y Mosher, et nl.: Tertiary peroxides can be obtained in 80 to 100% yields directly from the alcohol using the Milas method.6 In the case of secondary peroxides, however, the XIilas method fails and the Mosher method gives very poor yields. In their original report,sathe Mosher group obtained a 17% yield for sec-butyl peroxide. However, in our laboratory, this procedure has not given yields of purified peroxide of greater than 1 lye. Similarly, isopropyl peroxide has not been prepared in our laboratory in yields of better than 1Oyousing this method. The cause of the poor yields is k n o w n : both the alkyl methanesulfonate and the peroxide product are unstable in the strong basic solutions used in the Mosher method, and decomposition of both the starting material and the product competes with the accumulation of peroxide. Improved syntheses are described in the Experimental section in which the peroxides are removed from the reaction mixture as they are formed Experimental Preparation of sec.-Butyl Methanesu1fonate.--A slightly improved yield of the sulfonate can be obtained by the following procedure: 296 g. of sec-butyl alcohol and 458 g. of methanesulfonyl chloride are held between 5 and 8 " while 632 g. of pyridine is added over 4 t o 6 hr. Stirring is continued for 1 hr. Then 1.4 1. of 6 A' HCI is added, keeping the temperature below 30". Distillation [b.p. 58" (0.5 n i m . ) ] gives yields of 78 to 817;. Preparation of sec-Butyl Peroxide.--In a 1.51. flask are placed 228 g. of scc-butyl methanesulfonate. 75 g. of 30% hydrogen peroxide, and 200 i d . of rncthanol. The temperature is kept below 10" while 83 g. of KOH in an equal weight of water is added over 30 min. A pressure of 500 to 5.50 m m . is applied and the flask is heated to 60" and methanol allowed to distil into a series of traps. Additional methanol is added a s the reaction proceeds t o keep the mixture homogeneous. After heating for 2 hr. 50 ml. more of hydrogen peroxide is added, and the reaction is allowed to run for another 4 t o 6 hr. The trap contents are then extracted with Tetralin, washed s.ith 5'; KOH and water, and dried over sodium sulfate. The resulting solution is bright orange, but the color disappears when distillation begins. The peroxide is distilled through a short Yigreux colunin a t G 1 - ~ i 3 ° ( 5 0 m m . ) . Theyieldis 1 9 . 1 g . ( l i r ~ ) . Purity of sec-Butyl Peroxide.--The n.1n.r. ( n e a t ) consists of a triplet centered a t 0.90 p.p.1". (relative intensity 3.0) due to the y-methyl group, a doublet at 1.13 p.p.m. (intensity 3.0) due t o the 8-methyl group, a complex quintet a t 1.4,5 p.p.m. (intensity 2.0) due t o the methylene group, and a sextet a t 3.92 p . p , m . (intensity 1.0) due to the e-hydrogen atom. The infrared spectrum is an excellent diagnostic of purity. The peroxide has peaks at 3.4, 6.9, 7.4, 7.6, 7.7 (weak), 8 . 0 (weak), 8 . 9 ( b r o a d ) , 0 . 0 , 9 . 8 ,10.1, 10.2, 10.4, 11.2, 12.1,and 1 2 . 8 ~ . Preparation of Isopropyl Peroxide.-This peroxide has not been well characterized previously.' Isopropyl inethanesulfonate8 is coilveiiiently prepared on a 3.&ITlok scale in 80'; yield. The peroxide is then best prepared as follows. Isopropyl methanesulfonate (139.4 g , ) is placed in a 1-1. three-necked flask equipped with stirrer, reflux condenser, and dropping funnel. The flask is heated to 51-52" and held there throughout the (.5! (a) F . Welch. 11. I t . Williams. and H S. hlosher. J A ? n . ('huiii. S o c . , 77, 551 (19,55), (hr u' A P r y o r a n d I ) 51 H u i t < , n , J . O Y R .C h ? n i , 2 9 , 515 (1964). 16) S . .4 hlilas a n d I).
Xr Surgenor, .I A n i Chini . i o c , 68, 20.5 ( l Y 4 6 ) . ( 7 ) G . I t . hichlillan has reported t h e synthesis a n d photolysis r,f isopropyl peroxide (Lbiti., 83, :3018 ( I S H I ) , and private communications] T h e only physical constant which Mchiillan was able t o $obtain i b t h e tmiliny ]mint a t atmospheric pressure ( c n . 91"). He h a s summarized t h e literature on this peroxide !see his note R J . Although t h e peroxide has been listed in second^ a r y wurces. these a r e either incorrect citations of t h e original literature or were u n i u p p ~ r t r dt>y d a t a . Xiichlillan's i b t h e firit hotin fill? preparation ( 8 ) H I
