Electrochemical formation and dimerization of α-substituted benzyl radicals. Steric effects on dimerization

Full text of DOI:10.1021/jo951577f

J . Org. Chem. 1996, 61, 3191-3194
3191
tions to this rule have emerged. The reduction potentials
of alkyl radicals lie in the order I° < II° < III°, i.e.,
primary radicals are the easiest to reduce and tertiary
are the hardest.1,7 Alkyl iodides are easier to reduce than
bromides, which in turn are easier to reduce than
chlorides.1 Considering this and the dependence of
reduction potential on the degree of substitution in the
radical, one can readily understand the fact that, for tert-
butyl iodide E1 is positive of E2, resulting in two well-
resolved voltammetric waves.8 Secondary alkyl iodides
give rise to long-lived radical intermediates,9,10a but
primary iodides are reduced to carbanions.10 Tertiary
bromides represent a borderline situation: they exhibit
two barely resolved one-electron waves in some solvents11
and a single two-electron wave in others.8 As long as
the cathode does not consist of a material such as
mercury which can react with radicals,12 one may use
the relative yield of radical coupling product RR as a good
approximate measure of the lifetime (before reduction)
of the radicals R• produced by electrochemical reduction
of a given alkyl halide R-Hal. Such dimers are formed
relatively rarely, and some apparent dimer-forming reac-
tions actually correspond to SN2 displacement on the
alkyl halide by the carbanion intermediate.9
Electr och em ica l F or m a tion a n d
Dim er iza tion of r-Su bstitu ted Ben zyl
Ra d ica ls. Ster ic Effects on Dim er iza tion
Albert J . Fry,*,† J ohn M. Porter, and Peter F. Fry
Department of Chemistry, Wesleyan University,
Middletown, Connecticut 06459
Received August 29, 1995X
The mechanism of electrochemical cleavage of the
carbon-halogen bond is well established.1 Overall, the
reduction consumes two electrons, which are transferred
in two steps (Scheme 1). In the first (and potential-
Sch em e 1
E1
R-Hal + e- 98 R• + X-
E2
R• + e- 98 R-
R- + H+ f RH
In this context, benzylic halides represent an interest-
ing situation. Benzyl chlorides undergo the usual two-
electron reduction to a benzylic carbanion,13 but benzyl
bromides 1d -f exhibit two closely-spaced voltammetric
waves in dimethylformamide (DMF), and the products
change from radical type (RR) to carbanion type (RH)
over electrolysis potentials of just a few tenths of a volt.7,14
The R-chlorobenzyl radical (2a ) is easier to reduce than
benzal chloride (C6H5CHCl2), so it cannot be produced
by direct electrochemical reduction of benzal chloride.15
determining) step, one electron is transferred and the
carbon-halogen bond is broken to produce a free radical.
A second electron transfer then converts the radical to a
carbanion. Generally, the carbanion abstracts a proton
from the medium to produce the dehalogenated species
RH, but in aprotic media the carbanion may be trapped
by a variety of added electrophiles such as silyl halides,2
activated alkenes,3 and carbon dioxide and carbonyl
compounds.4 It is much more difficult to trap the
corresponding radicals in these reductions. As a result
of a great deal of experimentation in a number of
laboratories, the reasons for this difficulty have gradually
emerged. It turns out that for most alkyl halides, E1,
the potential necessary to produce the radical R•, is
negative of E2, the reduction potential of the radical, so
that R• is reduced very quickly as it is formed. Experi-
mentally, this means that most alkyl halides exhibit a
single two-electron voltammetric wave.1a By balancing
the rate of electron transfer to the radical against an
intramolecular isomerization process of known rate,5,6 it
can be estimated that the radical intermediates in most
alkyl halide electrolyses have lifetimes less than 1 µs. In
general, therefore, it will be especially difficult to trap
such radicals intermolecularly. However, a few excep-
We recently showed, however, that 2a can be prepared
by indirect electrolysis, that is, by electrocatalytic reduc-
tion of benzal chloride by electrochemically generated
cobalt(I) bis(salicylideneethylenediamine).16 Under these
conditions, radical 2a undergoes efficient coupling to 1,2-
(7) (a) Fry, A. J .; Powers, T. A. J . Org. Chem. 1987, 52, 2498. (b)
Yamasaki, R. B.; Tarle, M.; Casanova, J . J . Org. Chem. 1979, 44, 4519.
(8) Hoffmann, A. K.; Hodgson, W. G.; Maricle, D. L.; J ura, W. H. J .
Am. Chem. Soc. 1964, 86, 631.
(9) (a) Mubarak, M. S.; Peters, D. G. J . Org. Chem. 1982, 47, 3397.
(b) Peters, D. G.; Willett, B. C. J . Electroanal. Chem. 1981, 123, 291.
(10) (a) Cleary, J . A.; Mubarak, M. S.; Vieira, K. L.; Anderson, M.
R.; Peters, D. G. J . Electroanal. Chem. 1986, 198, 107. (b) La Perriere,
D. M.; Carroll, W. F., J r.; Willett, B. C.; Torp, E. C.; Peters, D. G. J .
Am. Chem. Soc. 1979, 101, 7561.
(11) (a) Vieira, K. L.; Peters, D. G. J . Org. Chem. 1986, 51, 1231.
(b) Fry, A. J .; Krieger, R. L. J . Org. Chem. 1976, 41, 54.
(12) (a) Webb, J . L.; Mann, C. K.; Walborsky, H. M. J . Am. Chem.
Soc. 1970, 92, 2042. (b) Casanova, J .; Rogers, H. R. J . Am. Chem. Soc.
1974, 96, 1942.
(13) Marple. L. W.; Hummelstedt, L. E. I.; Rogers, L. B. J . Electro-
chem. Soc. 1960, 107, 437.
† Correspondence may be adressed to: Dr. Albert J . Fry, Department
of Chemistry, Wesleyan University, Middletown, CT 06459. Phone:
(203) 685-2622. Fax: (203) 685-2211. E-mail: afry@wesleyan.edu.
X Abstract published in Advance ACS Abstracts, April 1, 1996.
(1) (a) Fry, A. J . Synthetic Organic Electrochemistry, 2nd ed.;
Wiley: New York, 1989; Chapter 5. (b) Peters, D. In Organic
Electrochemistry, 3rd ed.; Lund, H., Baizer, M. M., Eds.; Dekker: New
York, 1991; pp 361-400.
(2) (a) Yoshida, J .-I.; Muraki, K.; Funahashi, H.; Kawabata, N. J .
Org. Chem. 1986, 51, 3996. (b) Fry, A. J .; Touster, J . J . Org. Chem.
1989, 54, 4829.
(3) Baizer, M. M.; Chruma, J . L. J . Org. Chem. 1972, 37, 1951.
(4) (a) Silvestri, G.; Gambino, S.; Filardo, G.; Greco, G.; Gulotta, A.
Tetrahedron Lett. 1984, 25, 4307. (b) Sibille, S.; d’Incan, E.; Leport,
L.; Pe´richon, J . Tetrahedron Lett. 1986, 27, 3129. (c) d’Incan, E.; Sibille,
S.; Pe´richon, J .; Moingeon, M. O.; Chaussard, J . Tetrahedron Lett. 1986,
27, 4175.
(5) (a) Fry, A. J .; Mitnick, M. A. J . Am. Chem. Soc. 1969, 91, 6207.
(b) Similar results are obtained in the electrochemical reduction of
6-bromo-1-hexene, where the product consists of roughly equal amounts
of methylcyclohexane and 1-hexene.6
(14) Yamasaki, R. B.; Tarle, M.; Casanova, J . J . Org. Chem. 1979,
44, 4519.
(15) Fry A. J .; Sirisoma, U. N. J . Org. Chem. 1993, 58, 4919; 1994,
58, 2914.
(16) (a) Fry, A. J .; Sirisoma, U. N.; Lee, A. S. Tetrahedron Lett. 1993,
34, 809. (b) Fry, A. J .; Singh, A. H. J . Org. Chem. 1994, 59, 8172.
(6) Reed, R. G. Ph.D. thesis, Wesleyan University, 1971.
S0022-3263(95)01577-5 CCC: $12.00 © 1996 American Chemical Society

3192 J . Org. Chem., Vol. 61, No. 9, 1996
Notes
Ta ble 1. Red u ction P oten tia ls of r-Su bstitu ted Ben zyl
Electrochemical reduction of chloride 1a at -1.2 V
afforded a 52:48 mixture of the meso and dl diastereo-
isomers of 1,2-diphenyl-1,2-dichloroethane (3a ) in 93%
yield, together with a small amount of trans-stilbene (2-
3%). Coulometry indicated the consumption of 1.25
electrons/molecule of 1a . Controlled-potential electrolysis
of either diastereomer of 3a is reported to afford only
trans-stilbene.20
Br om id es, C6H5CH(X)Br a
b
compd
X
-Ep
1a
1b
1c
1d
1e
Cl
Br
SiMe3
H
1.53
1.52
1.52
1.48
1.55
Me
a
Voltammograms measured as 1.5 mM solutions in acetonitrile/
Controlled potential electrochemical reduction of gemi-
nal dibromide 1b in acetonitrile containing TBAHFP at
-1.2 V consumed 1.25 faradays/mol of 1b and resulted
in the formation of a mixture of stereoisomeric stilbene
dibromides (3b) and R-bromostilbenes (8) in 69% (meso:
dl ) 5:1) and 31% yield, respectively, together with a
trace of trans-stilbene. Formation of 8 is probably due
to the presence of adventitious water in the solvent and/
or supporting electrolyte. We have repeatedly observed
in this laboratory that electrolyses in wet solvents form
hydroxide ion (electrolysis of water), which in this experi-
ment would convert 3b into 8. Formation of 3b requires
1 faraday/mol of 1b and formation of 8 requires 2 faraday/
mol of 1b (the additional 1 faraday is required to produce
1 equiv of hydroxide). To produce the observed products,
1.31 faradays/mol should have been consumed, in good
agreement with the observed current consumption. The
methyne protons of the major and minor diastereomers
0.1 Bu4NPF6 vs the Ag/0.1 M AgNO3 reference electrode.
dichloro-1,2-diphenylethane (3a ) (which then undergoes
a second electrocatalytic conversion to a mixture of
stereoisomeric stilbenes).15,16 We were interested in
preparing radicals such as 2a by direct electrolysis, which
would be experimentally more convenient. Our previous
work on the electrochemical reduction of (R-bromo-
neopentyl)benzene7a suggested an approach which might
be adaptable to the synthesis of substituted benzyl
radicals. As mentioned above, E1 and E2 are very similar
for benzyl bromide. It seemed that it ought to be possible
to produce substituted benzyl radicals without their
undergoing further reduction to carbanions by carrying
out direct electrochemical reduction of the corresponding
bromides at relatively positive potentials. A secondary
advantage of this mode of radical synthesis over the
electrocatalytic route should be the fact that at such
potentials a wide variety of radical traps could be added
to the electrolysis medium without the concern that they
might themselves undergo electrochemical reduction.
1
of 3b appear at δ 5.50 and 5.30, respectively, in the H
NMR spectrum. The fact that the major product from
the radical dimerization is meso-3b was established by
addition of Br2 to trans-stilbene to afford the authentic
meso diastereomer, which exhibited its methyne reso-
nance at δ 5.50. It should be pointed out that electro-
chemical reduction of 1b consistently resulted in decay
of the electrolysis current to quite low levels (decay of
the initial current from 100 to 5 mA) before all of the
starting material had been consumed; in the experiment
described above, about 35% of the starting dibromide
remained unreacted. We attribute this to a partial
passivation phenomenon, which we previously observed
with this dibromide.2b An electrolysis carried out for a
long time at very low current did result in total consump-
tion of 1b but resulted in a mixture consisting of 71% of
a 5.4:1 mixture of trans- and cis-stilbenes, 15% of a
mixture of stereoisomers of 8, and 14% of a mixture of
the stereoisomers of 3b. Apparently, 3b is slowly con-
verted into stilbenes under these conditions (see Discus-
sion). Regardless of the side reactions which consume
3b in these experiments, it is notable that electrolysis of
1b results in 100% conversion to dimeric products.
Electrochemical reduction of 1c at -1.2 V at 0 °C in
acetonitrile containing 0.1 M LiClO4 consumed 0.98
electrons per mole of 1c. The product consisted of a
mixture of four isomeric compounds of molecular formula
C20H30Si2 in 94% crude yield in the ratio 39:33:19:9,
Resu lts a n d Discu ssion
Syn th esis of Su bstitu ted Ben zyl Br om id es. Bro-
mination of benzyl chloride by N-bromosuccinimide in
refluxing benzene afforded R-bromobenzyl chloride (ben-
zal chlorobromide, 1a ). Similar bromination of benzyl-
trimethylsilane (6a ) afforded (R-bromobenzyl)trimethyl-
silane (1c). Bromides 1b, 1d , and 1e were commercial
samples.
Volta m m etr y. The reduction potentials of the various
R-substituted benzylic bromides 1 were measured by
linear sweep voltammetry at a glassy carbon electrode
in acetonitrile containing 0.1 M tetrabutylammonium
hexafluorophosphate (TBAHFP). Potentials were mea-
sured relative to a Ag/0.1 M AgNO3 reference electrode
(Table 1).17 All of the benzylic bromides exhibited a
sharp, highly symmetrical voltammetric peak at ca. -1.5
V (see supporting information). The shape of the peaks
is highly suggestive of adsorption of the halides on the
carbon surface. This conclusion is reinforced by the
presence in the voltammograms of a small voltammetric
prewave at ca. -1.2 V and the appearance of the waves
at platinum- and mercury-coated platinum electrodes,
where the substrates gave rise to broad, poorly-defined
voltammograms and the prewave was considerably larger.
(R-Chlorobenzyl)trimethylsilane (6b) exhibited a volta-
mmetric peak potential at -2.10 V. The peak potential
of benzal chloride appeared at -2.20 V.
together with a trace of a fifth substance of mass C30H44
-
Si3 identified only by gas chromatography-mass spec-
trometry (GC-MS). The first two of these substances
proved to be meso- and dl-1,2-bis(trimethylsilyl)-1,2-
diphenylethane (3c), respectively.21-23 After multiple
flash chromatographic separations, the purified yields of
meso- and dl-3c were 21 and 18%, respectively. Although
Electr och em ica l Red u ction . Each halide was sub-
jected to controlled-potential electrochemical reduction
in acetonitrile. Acetonitrile was used because it is a
poorer hydrogen atom donor toward electrochemically
generated benzyl radicals than is dimethylformamide,
which has often been used in the past for electrochemical
experiments in this laboratory.18,19
(18) Save´ant, J .-M. Bull. Soc. Chim. Fr. 1988, 225.
(19) The C-H bond heterolytic bond dissociation energy of CH3CN
is 93-95 kcal/mol, only a few kcal/mol less than that of ethane (98
kcal/mol): Kanabus-Kaminska, J . M.; Gilbert, B. C.; Griller, D. J . Am.
Chem. Soc. 1989, 111, 3311.
(17) Fry, A. J .; Touster, J . J . Org. Chem. 1986, 51, 3905.
(20) Lund, H.; Hobolth, E. Acta Chem. Scand. 1976, 30B, 895.

Notes
J . Org. Chem., Vol. 61, No. 9, 1996 3193
not high, the 39% combined purified yield of the diaster-
eomers of 3c is substantially better than from the
previous literature procedure for preparation of this
substance.23 The remaining products are a mixture of
[p- and o-(trimethylsilyl)benzhydryl]trimethylsilanes, 4
and 5, respectively. The two compounds are separable
of dimer 3c are electroinactive and are stable to the
electrolysis conditions. The reduction potential of dichlo-
ride 3a is substantially negative (≈1 V) of the potential
(-1.2 V) at which the electrolysis was carried out.20 For
this reason, the few percent of stilbene produced in the
electrolysis could not possibly be produced by direct
electrochemical reduction of 3a at -1.2 V. It is probably
formed by SN2 attack of electrochemically-generated
bromide ion on 3a to afford the vicinal bromochloride 7,
which is reducible to stilbene at the electrolysis po-
tential.27-29 The meso and dl diastereomers of dibromide
3b, the initial products of reductive coupling of benzal
bromide (1b), are readily reduced to stilbene at the
electrolysis potential.29
by capillary GC-MS, but not by flash chromatography.
Their structures could be assigned by the 1H NMR
spectrum of the roughly 2:1 mixture and by their mass
spectral cracking patterns. Two other experiments of
mechanistic significance should be mentioned here. Ben-
zyltrimethylsilane (6a ) was produced in low yield in
addition to meso- and dl-3c, 4, and 5 when the electrolysis
of 1c was carried out at -1.3 V. Electrochemical reduc-
tion of (R-chlorobenzyl)trimethylsilane (6b) at -1.7 V
produced only 6a .
The formation of substantial amounts of head-to-tail
coupling products in the electrochemical reduction of 1c
warrants comment. The electrochemical reduction of
1-bromo-1-phenylethane (1e) affords a mixture of dimers
analogous to those produced from 1c, but in different
proportions. Head-to-tail dimers constitute less than 3%
of the products.7b The proportion of head-to-tail coupling
of radicals of the type C6H5CHR• rises to 28% with R )
SiMe3 (present work) and 44% with R ) t-Bu.7a,30 There
is a rough correlation between the conformational A-
value31 and the proportion of head-to-tail product as the
R group increases in size through methyl (A ) 1.7),
trimethylsilyl (A ) 2.5), and tert-butyl (A g 4.7).31 There
is also a second variable to consider: the meso:dl ratio
of head-to-head dimers. Greene pointed out the fact that
in principle this ratio should be unity because there is
no activation energy for dimerization of radicals and
therefore the transition states cannot respond to steric
factors.32 However, it is now known that this stricture
can break down when R is very bulky, since head-to-head
coupling brings the two groups into close proximity.
There ought to be a greater than additive33 steric effect
on dimerization; for example, the steric energy of cis-1,3-
diaxial dimethylcyclohexane is more than twice the
A-value of methyl.34 This effect is apparently sufficiently
large for the coupling of sterically hindered radicals to
result in an activation energy for radical dimerization.35
Nonunity meso:dl ratios have been observed in cases
The voltammetric wave in the vicinity of -1.5 V vs Ag/
0.1 M AgNO3 exhibited by each of the benzyl bromides
1a -c corresponds to reductive cleavage of the carbon-
bromine bond to afford the corresponding R-substituted
benzyl radical (2).1 As we showed previously with
R-bromoneopentyl bromide,7a the dimeric products pro-
duced in the electrolysis must be produced by dimeriza-
tion of the intermediate radicals. The waves for the
substituted halides 1a -c are not significantly different
from that of benzyl bromide (1d ), probably because of the
modest stabilizing effect of halogen24 and silicon25 on
carbon-centered radicals. It is critically important to
produce these radicals at quite positive potentials in
order to avoid reduction to the corresponding carbanions.
This is evident from the reduction of 6b, which produced
only benzyltrimethylsilane (6a ) even when electrolysis
was carried out at -1.7 V (at the foot of its voltammetric
wave). Thus, even at the most positive potential at which
the electrochemical cleavage of the carbon-chlorine bond
of 6b can be effected, the intermediate radical 2c is still
quickly reduced (the reduction potentials of benzyl
radicals are about -1.6 to -1.8 V).7,26 Similarly, reduc-
tion of benzal chloride (reduction potential ) -2.2 V)
affords carbanion-derived products.16b Fortunately, it is
possible to carry out the preparative scale electrochemical
reduction of the benzyl halides 1a -c at -1.2 V, well
positive of the voltammetric peak potential and the
reduction potential of the radical. The actual electroac-
tive species is probably the adsorbed halide, whose
reduction potential is presumably somewhat positive of
that of the peak potential.
(27) There is ample precedent for SN2 attack on starting material
in electrolyses of alkyl halides; see, e.g., ref 10b and: Fry, A. J .; Fry,
P. F. J . Org. Chem. 1993, 58, 3496.
(28) Fawell, P.; Avraamides, J .; Hefter, G. Aust. J . Chem. 1991, 44,
791.
(29) Dibromide 3b is reduced at -0.62 V vs Ag/0.1 M AgNO3: Inesi,
A.; Rampazzo, L. J . Electroanal. Chem. 1974, 54, 289.
(30) Powers, T. A. B. A. Thesis, Wesleyan University, 1986.
(31) Eliel, E. L.; Wilen, S. H. Stereochemistry of Organic Compounds;
Wiley: New York, 1994; pp 695-697.
The three dimers 3a -c are each different with respect
to further cathodic electrochemistry. The stereoisomers
(32) Greene, F. D.; Berwick, M. A.; Stowell, J . C. J . Am. Chem. Soc.
1970, 92, 867.
(21) Huang, H. H. Austr. J . Chem. 1976, 29, 2415.
(33) Steric interaction energies rise sharply as the distance between
the two interacting groups decreases, though there is no general
agreement on how to represent the functional dependence of the
repulsive van der Waals’ interaction on distance. Some investigators
model such interactions using an exponential function, while others
prefer an r-12 function: Burkert, U.; Allinger, N. L. Molecular
Mechanics; American Chemical Society: Washington, D.C., 1982.
(34) (a) Reference 31, pp 701-702. (b) The experimental cis-1,3-
diaxial dimethyl interaction energy of 3.7 kcal would undoubtedly be
far greater (.100 kcal) except for the distortions which the cyclohexane
ring undergoes to minimize the steric repulsion between the two methyl
groups: Allinger, N. L.; Miller, M. A. J . Am. Chem. Soc. 1961, 83, 2145.
(22) Hauser and Hance23 reported the synthesis of 3c in 1952
without discussion of its stereochemistry; their reported mp of 150-
152 °C corresponds to the meso diastereomer.
(23) Hauser, C. R.; Hance, C. R. J . Am. Chem. Soc. 1952, 74, 5091.
(24) Cain, E. N.; Solly, R. K. J . Am. Chem. Soc. 1972, 94, 3830.
(25) Doncaster, A. M.; Walsh, R. J . Chem. Soc., Faraday Trans. 1,
1976, 72, 2908.
(26) (a) Sim, B. A.; Milne, P. H.; Griller, D.; Wayner, D. D. M. J .
Am. Chem. Soc. 1990, 112, 6635; (b) Hapiot, P.; Konovalov, V. V.;
Save´ant J . Am. Chem. Soc. 1995, 117, 1428; (c) Wayner, D. D. M.;
Sim, B. A.; Dannenberg, J . J . J . Org. Chem. 1991, 56, 4853.

3194 J . Org. Chem., Vol. 61, No. 9, 1996
Notes
the combined solvents were removed by rotary evaporation. The
residue was extracted with pentane, washed twice with distilled
water and saturated brine, and dried over MgSO4, and the
pentane was removed. Analysis by a combination of GC-MS
and NMR spectroscopy (the combination of two spectroscopies
is necessary because the ortho head-to-tail isomer 5 is not
resolvable from meso-3c by GC-MS) showed the product to
consist of a 39:33:19:9 mixture of meso:dl:para:ortho dimers
meso-3c, dl-3c, 4, and 5. The crude product (3.269 g, 94%) was
flash chromatographed over silica gel with pentane as eluent to
afford 0.741 g (21%) of meso-1,2-bis(trimethylsilyl)-1,2-diphen-
ylethane (3c) [mp 150-152 °C (lit.23 150-152 °C); 1H NMR δ
7.1-7.3 (m, 10 H), 2.70 (s, 2 H), -0.42 (18 H); m/e 326 (m+) 309,
253, 223, 180, 179, 73 (100)], followed by 0.624 g (18%) of dl-3c
[1H NMR δ 6.87 (m, 4 H), 7.0-7.1 (m, 6 H), 2.60 (s, 2 H), -0.086
(s, 18 H); m/e 326 (m+), 309, 253, 223, 179, 180, 135, 73 (100)],
followed by a 2:1 inseparable mixture of [p-(trimethylsilyl)-
benzhydryl]trimethylsilane (4) and its ortho isomer 5, 0.513 g
(15%). The mixture of 4 and 5 was submitted for microanalysis.
Anal. Calcd for C20H30Si2: C,73.54; H,9.26. Found: C,73.47;
H, 9.34. The 1H NMR spectra of 4 and 5 could be assigned
because of the unequal amounts of the two isomers in the
mixture; injection of this mixture into the GC-MS spectrometer
after chromatographic separation of the meso isomer permitted
measurement of the mass spectrum of 5. Para-isomer 4: 1H
NMR δ 7.0-7.3 (9 H), 3.46 (s, 1H) and 2.02 (s, 2H), .02 (s, 9 H),
-0.03 (s, 9 H); m/e 326 (m+), 311, 253, 238, 223, 180, 179, 178,
145, 73 (100). 5: 1H NMR δ 7-7.3 (m, 9 H), 3.63 (s, 1 H), 2.17
and 1.89 (AB quartet, J ) 13.9 Hz, 2H), .08 (s, 9 H), -0.02 (s, 9
H); m/e 326 (m+), 238, 223, 179, 178, 165, 73 (100).
where the head of the radical is highly sterically hin-
dered.36 The effect requires very large R groups; even
the relatively bulky trimethylsilyl groups result in a very
modest increase in the meso:dl ratio (1.18 in the present
work) over the value of 1.00 observed for R ) methyl.32
Exp er im en ta l Section
Ben za l Ch lor obr om id e (1a ). Benzyl chloride (25.0 g, 0.2
mol) and N-bromosuccinimide (35.0 g, 0.2 mol) were heated in
200 mL of benzene for 2 h with 0.05 g of azobisisobutyronitrile.
After filtration of the mixture, the resulting solution was distilled
in vacuo (45 °C/0.5 mm) to afford 1a (28.8 g, 70%) (lit.37 bp 31.5
°C/0.15 mm); 1H NMR δ 7.6 (dd, 2 H), 7.35-7.45 (m, 3 H), 6.76
(s, 1 H).
(r-Br om oben zyl)tr im eth ylsila n e (1c). Benzyltrimethyl-
silane (6a ) (20 g, 0.12 mol) and N-bromosuccinimide (21.7 g, 0.12
mol) were allowed to react in benzene at reflux as in the
preceding synthesis of 1a . After removal of the solvent, analysis
by GC-MS demonstrated the residue to be a 11:46:43 mixture
of 6a , (R-bromobenzyl)trimethylsilane (1c), and (R,R-dibro-
mobenzyl)trimethylsilane (9). Distillation in vacuo afforded 1c23
(8 g, 55%) [m/e 244, 242, 199, 201, 169, 171, 90, 73 (100)] and
923 (9 g, 63%) [m/e 324, 322, 320, 251, 249, 247, 205, 203, 201,
170 (100), 168 (100), 89, 73]. After two distillations, dibromide
9, which had previously been reported as a liquid, crystallized
upon standing, mp 30-31 °C.
Rep r esen ta tive Electr och em ica l Red u ction . Electr o-
ch em ica l Red u ction of 1c. (R-Bromobenzyl)trimethylsilane
(1c) (5.185 g, 21.3 mmol) was dissolved in 50 mL of an 0.1 M
solution of LiClO4 in CH3CN. The solution was placed in a
divided cell38 containing a graphite cloth39 cathode, Ag/0.1 M
AgNO3 reference electrode, N2 inlet and outlet, magnetic stirring
bar, and an anode compartment consisting of a cylindrical tube
terminating in a coarse fritted disk covered with a methylcel-
lulose/BuN4PF6/DMF conducting gel38 and a graphite cloth anode
imbedded in the gel. The anolyte was 0.1 M LiClO4 in CH3CN
containing a small amount of hydrazine as anodic depolarizer.
Electrolysis was carried out at -1.2 V. Over a period of 36 h
the electrolysis current decayed from 25 mA to background level
(5 mA). The cathode had to be shaken at intervals to prevent
passivation by a deposit of crystalline product. The electrolysis
was terminated after a passage of 2026 C; the theoretical current
consumption is 2060 C. The cell was rinsed with pentane, and
Electr och em ica l Red u ction of 1a . Electrolysis of 1a (2.0
g) at -1.2 V afforded 1.14 g (93%) of a mixture shown by GC-
1
MS and H NMR analysis to consist of a 51:47:2 mixture of meso-
3a , dl-3a , and trans-stilbene, respectively.
Electr och em ica l Red u ction of 1b. Electrochemical reduc-
tion of 1b carried out in the usual manner consumed 1.25
faradays/mol of 1b and afforded a mixture that consisted of 35%
of 1b and 65% of a 61:39 mixture of 3b and 8 by GC-MS.
Ack n ow led gm en t. Financial support was supplied
by the National Science Foundation under Grant No.
CHE-9413128. Latorya Hicks purified 9 and obtained
it in crystalline form for the first time in this laboratory.
A summer fellowship for Ms. Hicks (Lane College) was
provided by a National Science Foundation Research
Experience for Undergraduates (NSF-REU) grant to the
Wesleyan University Chemistry Department.
(35) Eichin, K.-H.; McCullough, K. J .; Beckhaus, H.-D.; Ruchardt,
C. Angew. Chem., Int. Ed. Engl. 1978, 17, 934.
(36) Peyman, A.; Beckhaus, H.-D.; Kramer, D.; Peters, K.; von
Schnering, H. G. Chem. Ber. 1991, 124, 1989.
(37) Mark, V. U. S. Patent 3 420 900; Chem. Abstr. 1969, 70, 67765s.
(38) Reference 1, Chapter 10. Electrolysis cells were obtained from
The Electrochemicals Co., 116 Maple Shade Rd., Middletown, CT
06459.
Su p p or tin g In for m a tion Ava ila ble: Voltammograms of
compounds 1a -e (6 pages). This material is contained in
libraries on microfiche, immediately follows this article in the
microfilm version of the journal, and can be ordered from the
ACS; see any current masthead page for ordering information.
(39) Hand, R.; Carpenter, A. K.; O’Brien, C. J .; Nelson, R. F. J .
Electrochem. Soc. 1972, 119, 74.
J O951577F