1580 (11) R. Breslow, private communication. (12) J. C. Traeger and F. P. Losing, in preparation.
F. P. Losing,* J. C. Traeger NRCC No. 14552, Division of Chemistry National Research Council of Canada Ottawa, Ontario, K1 A OR6 Canada Received November 26, 1974
Radical Production from the Interaction of Closed Shell Molecules. 11. The Reaction of Organic Sulfides with tert-Butyl Peroxybenzoates' Sir: In recent years it has become clear that a number of the reactions in which nucleophiles react with substrates to give substitution products actually proceed by a chain mechanism involving radicals and radical ions.2 In this context, the reactions of nucleophiles (donors) with peroxidic substrates provide a lode of rich variety for study; these reactions can involve either an s N 2 reaction (eq 1) or an electron transfer (ET) process (eq 2). Many peroxide-nucleo-
D + XOOY
D
+
XOOY
SN2
ET
[ W X
-OY]
t
--
- d[ TBP]
ultimate products
[De'
+ .OX +
-OY]
nism," most of the critical reviews of the DMA-BPO system ignore ET and continue to formulate the DMA-BPO reaction as an s N 2 process." Clearly, E T reactions of peroxides are not at all easy to establish, and further studies are needed. We previously reported on the reaction of sulfides with benzoyl peroxide, a process which does not produce radic a l ~We . ~ now ~ describe the superficially similar reaction of sulfides with tert-butyl peroxybenzoates (TBP's) which does lead to radicals and for which we propose an E T mechanism.12 This system is of interest in comparison both with the BPO system and also with ortho-mercapto-substituted TBP, studied by Martin, et al.l3,l4 The rate of decomposition of TBP in sulfide solutions can be written as in eq I where k H is the rate constant for homolysis and k s and k's are the second-order and pseudofirst-order rate constants for the sulfide-TBP reaction. The fraction of the bimolecular, assisted process that leads to radicals is calculated by comparing the rates of peroxide and scavenger disappearance. (The TBP disappearance was monitored using the peroxide infrared band at 1758 cm-I; ~ t y r e n e 3or~ galvinoxyl in Iirnitingl5 or excessi6 concentration were used as scavengers.) dt
(1)
=
k,[TBP]
+
k,[R,S][TBP] =
-+
Table I gives the data; the last three columns give the per cent radical production. Despite the lack of precision in these small value^,^^,^^ it is clear that the interaction of sulphile reactions which involve an SN2 mechanism are fides with TBP derivatives does produce radicals. Conservaknown,3 but very few E T reactions of peroxides have been tively, averaging all the data in the last three columns, 2.3 proposed. f 1.5% of the bimolecular interactions produce radicals." Recently E T reactions have been postulated for the interaction of diphenylhydroxylamine with benzoyl peroxide It is important to establish that the radicals we observe do not arise from homolysis of an intermediate produced in (BPO)4aand alkyllithium and Grignard reagents with alkyl an s N 2 process. It is clear that species such as BzOS+Me2 peroxides.4b In these systems, the leakage of radicals from a or B Z O S ( M ~ ) = C H ~which , ~ ~ can be formed from TBP seemingly nonradical reaction has been interpreted as an Me2S by s N 2 processes, do not produce radicals since the important clue that an ET mechanism is involved. BPO-Me2S system does not form radical^.^^,^^ Homolysis Distinguishing between s N 2 and ET paths is subtle and of t-BuOS+Me2 was excluded by showing that can be very difficult. For example, in 1950 Homer5 postulaMeOS+Mez, synthesized independently,22initiates the poted that the rapid radical production which results from the lymerization of styrene too slowly to be responsible for radireaction of dimethylaniline (DMA) with BPO results from cal production in the Me2S-BPO system. an ET mechanism. In 1957 in his book, Walling6 dismissed this possibility and proposed that the DMA-BPO reaction The products (gas chromatographic analysis) from rcacis a normal s N 2 process, and that radicals arise from the tion of 1 M dimethyl sulfide with TBP at 80° in CC14 conhomolysis of the ammonium salt, BzO-+NMelPh, which is taining 0.2 M styrene (to eliminate induced decomposition) the first intermediate in the ionic displacement p r o c e s ~ . ~ .are ~ 90% t-C4H90H, 96% C H ~ S C H ~ O C O C ~ H S Horner later partially recanted9 and suggested that both ( BOMS),3e,'3,24six unidentified compounds in less than 1% ET and an s N 2 reaction (followed by homolysis) are re- yield, and no DMSO, PhCOlH, or tert-butyl benzoate. sponsible for the radical production observed. Despite the Without styrene the products are 90% z - C ~ H ~ O40-50% H, fact that research published by several groups in recent ChH5C02H, and 50-60% BOMS. years appears to support Homer's original ET mechaEquations 2a, 2b, and 3 present a generalized mechanism s i m i l a r o r identical products
(2)
+
Table I. The Reaction of Dialkvl and Arvl Alkvl Sulfides with Substituted tert-Butvl Peroxvbenzoates 1). Table I1 presents these data. Some of the data are not very precise, but they are adequate to distinguish inverse from normal isotope effects, which is all that is necessary. The use of isotope effects for identifying s N 2 reactions is well known. Isotope effects are smaller for s N 2 than S N I reactions because of higher vibrational frequencies for deuterium in both substrate and nucleophile a t the transition state for s N 2 relative to S ~ 1 . These l ~ higher frequencies can be rationalized as being due to hybridization, steric, hyperconjugative, or inductive effects." Computer studies correctly predict inverse or near unity isotope effects for N deuterated substrates in s N 2 reactions and normal isotope effects for Sxl reactions,18 and inverse isotope effects are predicted and observed for s N 2 reactions in which the n u cleophile is deuterated in the @-position.'9*20$21 In contrast, vibrational frequencies for @-hydrogens in the donor are loosened a t the transition state in ET reactions, and normal isotope effects are observed. Weakened bonding (i.e., loosened vibrational frequencies) in the transition state produces normal isotope effects; therefore, normal secondary isotope effects, k H / k D > 1, will be observed in E T reactions whenever the ionized electron is lost from an orbital which has bonding character a t the @ C-D bond. (Inverse isotope effects will be observed when ionization occurs from an orbital which was antibonding character a t
Table I. Reactions of Peroxides with Nucleophiles (Donors) No.
Peroxide
1 2 3 4 5 6 7 8 9
BPOe BPO BPO BPO BPO TBPl TBP
Donor
PhNMe2 PhzNOH Me2S ArCH=CHAr Me2C=CMe2 Me2S Ph3P o-MeSTBP 2-MeS-3-t-BuOOCOTBP o-Ph%C--CHTBP o-Ph2C==CHBPO
10
11
Pb
Accelerationa
peroxide
3X 6X 5x 4 x 1x
lo4 (40") lo5 (40') 104(400) 103 (45017 102 (45") 17 (80') 2 x 103 (go3) 5 x 103 (80")" 1 X lo5 (80')n 42 (80")" 387 (70")l'
+1.6! +0.8
+I . 2 +1.3 +1.2
PC
donor
?
radicald
Ref
18h
6a 2h 6e 6f 6g bh 6i 6j 61 6m 6m
-2.70
100
-1.31 -1 .O'
0 IO1 0 2 0 50 50 80
-1.773" -1.30
+0.7
-1.8
11
The acceleration of the rate of peroxide disappearance in a 1 .O M solution of the nucleophile relative to the rate in the same solvent without nucleophile. Hammett equation p when substituents are in the Ar group of the peroxide. c Substituents in the Ar group of the nucleophile. Per cent of the total reaction that produces scavengeable radicals. e Benzoyl peroxide. J Reference 6b. Reference 6c. Reference 6d. This work; nucleophile is Ar substituted ArSCH3. ' For reaction of m,ni'-Br2BP0 with trr~iis-p-p'-(MeO)*stilbene.With u-. tert-Butyl peroxybenzoate. m For reaction with fert-butyl p-chloroperoxybenzoate. Relative to unsubstituted peroxide. * Reference 6k. Q
71
Table 11. @-Deuterium Isotope Effects for Reactions of Nucleophiles (Donors) with Substrates ~~~
~~
~
~~
~~
~
Temperature Substrate
Nucleophile
CH~OTS CH~OTS t-BuOOH BPOb BPO C102
TBPd 3,5-(NOz)zTBP a
Per molecule, b Benzoyl peroxide.
c
( "C)
Mechanism
ks/kDa
Ref
51.29 51.29 80 40 24.8 25.2 80 80
sN2 sN2 sN2 sN2 ET ET ET ET
0 883 i 0.008 0.952 + 0.002 0.93 i 0.03 0.88 i 0.05 1,53c 1.3 I .08 i 0.06 1.06 i 0 . 0 3
19a 19a 6h 6h 2h 2g 6h 6h
May be due in part to a contribution from a primary isotope effect. tert-Butyl peroxybenzoate.
Journal of the American Chemical Society
/
97.6
/
March 19, 1975
