OPTICALLYACTIVE ~-PHENYLETHYL-~-BUTYL PEROXIDE
June 20, 1952
[CONTRIBUTION FROM
THE
3079
DEPARTMENT OF CHEMISTRY, PURDUE UNIVERSITY]
Optically Active Alkoxy1 Radicals. 11. Preparation and Properties of Optically Active a-Phenylethyl-t-butyl Peroxidel BY NATHAN KORNBLUM AND HAROLD E. DELAMARE~ RECEIVED JANUARY 15, 1952 R’
I
In order to learn whether alkoxyl radicals are able to undergo the rearrangement R-C-0.
R’ -r
I
R-C-0-H
the thermal
I
H decomposition of optically active a-plienylethyl-t-butyl peroxide in thiophenol solution ai-125” was investigated. Optically CHI
I
active a-phenylethanol was obtained in 5570 yield. It is concluded, therefore, that tlic alkoxyl radical CeHb-C-0. is CHI I I H capable of abstracting a hydrogen atom Irom thiophenol without prior rearrailgellielit to CbI16--C-0H. Siiice the optical purity of the peroxide is not known, it can only be said a t present that between 42 and 100% of the alkoxyl yH3 I
radicals C&-c-o*
are converted to a-phenylethanol without undergoing rearrangement.
I
H
The thermal decomposition of alkyl nitrites in the vapor phase has been studied by Steacie and by F. 0. Rice who proposed the mechanism3
+ + + + 2RCHeONO +RCHO 4- RCHzOH + 2 N 0
RCHzONO +RCHtO. NO RCHzO. RCHzONO +RCHiOH RCHONO RCHONO ---f RCHO NO
Recehtly it has been found that the liquid phase pyrolysis of optically active 2-octyl nitrite at 100’ gives %octanone, 2-octanol and nitric oxide and the 2-octanol (80% yield) is optically pure’; this, of course, is completely consistent with the above sequence. If the foregoing mechanism is accepted then it follows that a secondary alkoxyl radical is optically stable and that a rearrangement such as CHI
I CeHia-C-0. I
CH,
I
--+ C~HII--C-OI-I
(1)
€I
does not occur with facility, if at all, in the liquid phase a t In principle, however, such a rearrangement is a real possibility since an 0-H bond is stronger than a C-H bond.6 Also the second free radical of eq. (1)
will be stabilized by resonance with the dipolar CHa
structure6 c~H~~-&-OH;
such resonance is not
i;;e
present in the alkoxyl radical. Of further interest is the fact that an analogous rearrangement takes place readily even a t room temperature.’ Thus the (csH~,)aC-o. +(CeH~,)zc-o-ceH6 question of rearrangements in the sense of eq. 1 cannot be regarded as settled. It is well established that the thermal decomposition of dialkyl peroxides occurs by rupture of the oxygen-oxygen bonds R-GO-R’
+ R’-O.
+R-0.
This suggested a very direct means of producing optically active alkoxyl radicals.which, among other things, would provide a check on conclusions arrived a t from experiments employing alkyl nitrites. Toward this end, the synthesis of optically active CH:
CHs
I C6Hr,C-o-o-c(CH3)~ I H
+ CsHsL-0.
f
I
(c&),C-o~
(1) Paper I in this series: N. Kornblum and E. P. Oliveto, THIS 71, 226 (1949). (2) Allied Chemical and Dye Fellow, 1948-1949. (3) M. Szwarc, Chcm. Reus., 47, 143 (1950). (4) The argument is based on the commonly held assumption that the radical RR‘R”C* is incapable of maintaining optical activity (G. W. Wheland. “Advanced Organic Chemistry,” Ed. 2, J. Wiley and Sons. Inc., New York, N. Y.,1949, p. 713). (6) K. S. Pitzer, THISJOURNAL, 70, 2140 (1948). Although this use of “bond strengths” is an oversimplification, it is noteworthy that for the saturated radical of eq. (1) the difference in “bond strengths” is ca. 11 kcal. per mole; if one employs “bond dissociation energies” the 0 - H bond is stronger than the C-H bond by ca. 23 kcal. per mole. CHI JOURNAL,
I
For the arylated radical CeHr-C-0.
the C-H “bond dissocia-
€! tiorr energy” fulls to ca. 77 kcal. per mole which is to be contrasted with 118 for the 0-Hbond (ref. 3, Tuble AI).
CsHs-C-0-H
1.: CHI
I
C~H~--C--O-H I
H active
+ D.
D
CHI I
CsHs-C-Cb-H
I
f D.
H inactive
(6) Compare P. D. Bartlett and K.Nozaki, ibid. 69,2305 (1947). (7) M.S.Kharasch, A. Fouo and W. Nudenberg, J . Org. Chcm., 15, 770 (1960); h.1. A. Spielman. THISJOURNAL. 67, 1117 (1935). (8) Ref. 3, pp, 149-151; J. H. Raley, I.. hi. Portcr. F, F. K u r t uud W. E. Vaughiln, i b i d . , 73, 17 (1951).
a-phenylethyl-t-butyl peroxide (I) and its thermal decomposition in the presence of a donor of hydrogen atoms (H:D) was undertaken. The optically active a-phenylethoxyl radical (11) can go over to a-phenylethyl alcohol by two paths, only one of which will yield the active alcohol. It is especially noteworthy that in the arylated radical (11) the iiiigration of hydrogen from carbon to oxygen is much more likely to occur than in the strictly aliphatic radical CHs
(49% yield) and t-butyl alcohol (69% yield); lesser amounts of a-phenylethanol and acetone were also isolated.11 It seems very likely that here the acetophenone is produced by a free radical chain sequence
c Ifd
Iiiitiatioi~:I ---t.
Propagation:
I
C&16 --C 0. t- ( C I L ) A C---!.),
I
+ R-0.
I
CsH,-C-O-O-C(
i1
+ ROH
1
C36Ns---C--O---(j--C(CH3)R
CHa
I
C6Hs -CEO
?H1 I- --+.
1
I
1%
I t was always contaminated with unreacted broniide and efforts to consume all the bromide by using an excess of the hydroperoxide salt resulted in a sharp decrease in the yield of I. Simple distillation of the peroxide failed to free it of the bromide and rectification was accompanied by slight decomposition of the peroxide. Consequently, a procedure was devised in which the bromide was removed by treating the crude product with thiourea; the resulting salt was then easily separated from the peroxide by washing with water.g This gave I of 99.92 f. 0.04 mole per cent. purity; it has the infrared absorption characteristic of a dialkyl peroxide.I0 Attention was next directed to finding a good hydrogen donor (H:D) for reaction (2)
---+ CHa I
C~H~---.$- o -11 -+. (CHX-0-H
CHI)^
CHJ
a-Phenylethyl-t-butyl peroxide (I) was obtained in 30-3570 yield from a-phenylethyl bromide and the potassium salt of t-butyl hydroperoxide
I $. 2 H : D
--+ CH3
C6H*3-c-o.5
c'sJs ..C R r +. (CH3)3C---(j--(j-K
H
1- r):r) (2)
11
When peroxide I was decomposed in cuniene at 130' a 31% yield of crude 2,3-dimethyl-2,3-diphenylbutane was obtained but none of the expected a-phenylethanol could be isolated. In contrast, a very appreciable amount (ca. 3770) of acetophenone was found. The rather remote possibility that the acetophenone could have arisen from the base-catalyzed decomposition of the p e r ~ x i d e ,the ~ glass furnishing the alkalinity, was disposed of by the finding that peroxide I is readily decomposed when kept a t 125" in a quartz system by itself. The major products are acetophenone (9) In a n early experiment a n a t t e m p t was made t o remove the unreacted a-phenylethyl bromide b y treatment with piperidine; this not only removed the bromide b u t also converted a large portion of the peroxide into acetophenone. This led t o the discovery t h a t peroxide I is subject to a base-catalyzed elimination reaction which produces acetophenone a n d t-butyl alcohol [N.Rornblum a n d H.E. DeLaMare, THIS JOURNAL, IS, 880 (lQ5l)l.It is this sensitivity toward base which causes a drop in the yield of peroxide when a n excess of the potassium salt of t-butyl hydroperoxide is employed. (10) We are indebted t o Dr. E. R. Blout of the Polaroid Corporation for the determination and interpretation of the infrared d a t a (cf. Experimental).
+ (CH3),C--O.
Thiophenol proved to be an excellent medium for minimizing the chain reaction and achieving the reaction described by eq. (2). Upon heating peroxide I in thiophenol a t 125' for ten hours, 55-57% yields of pure a-phenylethanol were obtained; small amounts of acetophenone (9-12%) were simultaneously produced.l28'2a Significantly, the yields of pure diphenyl disulfide in these experiments ranged from 66-71%, Apparently thiophenol is a good enough donor of hydrogen atoms to be able to compete favorably with the peroxide itself, especially when used in large excess. When levorotatory peroxide I, a Z 4-~ 51.0" ( I 1 dm.), was heated in thiophenol the a-phenylethanol produced had a 2 5-~10.6' ( I 1 dm.) ; this was converted to the crystalline hydrogen phthalate ester, [ a I z 3+9.8'. ~ These rotations correspond to an optical purity of 24% for the alcohol and 27% for the hydrogen ~ h t h a 1 a t e . l ~It is, therefore, clear that a t 125' in the liquid phase a sizable fraction of the alkoxy1 radicals I1 are able to abstract hydrogen from thiophenol without (11) Unpublished work by Mr. S. I.. Clark of this department. T h r analogous experiment conducted i n glass also gave a 49% yield ot acetophenone; cf. Experimental. (12) The a-phenylethanol and acetophenone isolated from these experiments account for 64-69% of peroxide I. H a d the a-phenyl-' ethoxyl radicals (11) undergone rearrangement of the carbon skeleton ethyl phenyl ether and (or) benzyl methyl ether would have resulted Since no special effort was made to isolate these ethers t h e possibility remains t h a t they are formed in minor proportions. An analogous rearrangement in which a phenyl group migrates is known for the neophyl radical (W. H. Urry a n d 14. S. Kharasch, THISJOURNAL, 66, 1438 (1944); S. Winstein and T;. H. Seubold, J r . , i b i d . , 69, 2917 (lWij),
(12a) ADDEDIS PRo
