Three-electron S(N)2 reactions of arylcyclopropane cation radicals. 1. Mechanism

Article abstract of DOI:10.1021/ja9631110

The mechanism of photosensitized nucleophilic substitution reactions on arylcyclopropanes was investigated. Stereochemical experiments with methanol, water, and cyanide as nucleophiles showed that the reactions occurred stereospecifically with complete inversion of configuration at the carbon atom undergoing substitution. Independent generation of the arylcyclopropane cation radicals by nanosecond transient methods showed that they reacted rapidly with nucleophiles with kinetics that were first-order in both the cation radical and the nucleophiles. Through a combination of transient kinetics and steady-state Stern-Volmer quenching experiments, the reaction of the phenylcyclopropane cation radical with methanol was kinetically correlated with the formation of the substitution product. The reaction of phenylcyclopropane cation radical with a series of alcohols as nucleophiles showed small steric effects.

Full text of DOI:10.1021/ja9631110

J. Am. Chem. Soc. 1997, 119, 987-993
987
Three-Electron SN2 Reactions of Arylcyclopropane Cation
Radicals. 1. Mechanism1
J. P. Dinnocenzo,* T. R. Simpson, H. Zuilhof, W. P. Todd, and T. Heinrich
Contribution from the Department of Chemistry, UniVersity of Rochester,
Rochester, New York 14627-0216
ReceiVed September 4, 1996X
Abstract: The mechanism of photosensitized nucleophilic substitution reactions on arylcyclopropanes was investigated.
Stereochemical experiments with methanol, water, and cyanide as nucleophiles showed that the reactions occurred
stereospecifically with complete inversion of configuration at the carbon atom undergoing substitution. Independent
generation of the arylcyclopropane cation radicals by nanosecond transient methods showed that they reacted rapidly
with nucleophiles with kinetics that were first-order in both the cation radical and the nucleophiles. Through a
combination of transient kinetics and steady-state Stern-Volmer quenching experiments, the reaction of the
phenylcyclopropane cation radical with methanol was kinetically correlated with the formation of the substitution
product. The reaction of phenylcyclopropane cation radical with a series of alcohols as nucleophiles showed small
steric effects.
Introduction
The activation of carbon-carbon bonds toward reactions that
occur both stereospecifically and in high yield is a major
challenge in organic chemistry. A particularly well-documented
case is formed by the class of SN2 reactions, which constitute
an essential part of organic synthesis2 as well as a testing ground
for theoretical models.3 Classic four-electron SN2 reactions
suffer, however, from several limitations with respect to the
nature of both the substrate and the nucleophile. For example,
loss of stereospecificity often occurs with increasing steric
demands of either the substrate or the nucleophile by transition
to reactions occurring by an SN1 mechanism. This motivates
research into analogous reactions that do not suffer from this
drawback. One possibility is provided by three-electron SN2
reactions between open-shell substrates (LG•R)+ and closed-
shell nucleophiles (:Nu), schematically represented by eq 1. As
we show in this and the accompanying paper in J. Am. Chem.
Soc. 1997, 119, 994, the reactions of arylcyclopropane cation
radicals with nucleophiles are examples of such reactions. They
proceed with a complete inversion of configuration in the
substitution step and have greatly diminished steric effectsstwo
features that make them attractive for synthetic applications.
Rao and Hixson proposed that methanol-addition product 2
is formed by the mechanism shown in Figure 1. The mechanism
starts with photoinduced electron transfer from 1 to the singlet
excited state of 1,4-dicyanobenzene (DCB) to give 1•+ and
DCB•-. Methanol is then proposed to react with 1•+ by
nucleophilic substitution. Following proton loss from oxygen,
the benzylic radical intermediate is reduced by DCB•- to give
a benzyl anion, which is protonated by methanol to give 2. The
mechanism for formation of 3 will be discussed later.
In a subsequent study of the photooxidation of 1, Mizuno
and co-workers5 observed formation of methanol-adduct 2 (40%)
as well as 1,6-dimethoxy-3,4-diphenylhexane (45%) with 1-cy-
anonaphthalene (1-CN) as the electron acceptor. The latter
product is thought to be formed by coupling of intermediate
benzyl radicals. Mizuno et al. also demonstrated that cyanide,
water, and several other alcohols could be used as nucleophiles.
The observation that PhCHD(CH2)2OCH3 is formed in the
reaction of 1 with CH3OD provided evidence that the mecha-
nism for formation of 2 involves donation of a proton rather
than a hydrogen atom from methanol.
The pioneering work of Rao and Hixson4 on the photo-
sensitized addition of nucleophiles to arylcyclopropanes lead
us to explore whether the reactions involved examples of three-
electron SN2 reactions. They discovered that the 1,4-dicyano-
benzene-photosensitized addition of methanol to phenylcyclo-
propane (1) gave two major products, 2 and 3.
From the standpoint of this work, the most important step in
Figure 1 is the reaction of methanol with 1•+. Rao and Hixson4
proposed that the reaction occurs from a ring-closed cation
radical intermediate because trans-1-methyl-2-phenylcyclopro-
pane did not trans f cis isomerize under the reaction conditions.
While this observation is consistent with the proposed mech-
anism, the unimolecular ring-opening of the cyclopropane cation
radicals followed by rapid nucleophilic capture of methanol (i.e.,
an SN1 mechanism) could not be entirely excluded. In later
work on the photooxidation of bridged bicyclobutanes, Gassman
and co-workers observed products formed from nucleophilic
substitution with inversion of configuration.6a-d Similar results
were reported by Arnold and Du for the photooxidation of
X Abstract published in AdVance ACS Abstracts, January 15, 1997.
(1) Taken in part from the doctoral thesis of T.R.S., University of
Rochester, 1995.
(2) See, e.g.: (a) Carey, F. A.; Sundberg, R. J. AdVanced Organic
Chemistry; Plenum Press: New York, 1990; Part B, Chapter 3. (b) March,
J. AdVanced Organic Chemistry Wiley-Interscience: New York, 1992;
Chapter 10.
(3) Shaik, S. S.; Schlegel, H. B.; Wolfe, S. Theoretical Aspects of
Physical Organic Chemistry. The SN2 Mechanism; Wiley-Interscience: New
York, 1992.
(4) Rao, V. R.; Hixson, S. S. J. Am. Chem. Soc. 1979, 101, 6458.
(5) Mizuno, K.; Ogawa, J.; Otsuji, Y. Chem. Lett. 1981, 741.
S0002-7863(96)03111-3 CCC: $14.00 © 1997 American Chemical Society

988 J. Am. Chem. Soc., Vol. 119, No. 5, 1997
Dinnocenzo et al.
Figure 1. Rao-Hixson mechanism for the photooxidation of 1.
tricyclene6e and by Weng and Roth for the photooxidation of
quadricyclane.6f Most recently, the stereochemical outcome of
methanol addition to (1R,5R)-(+)-sabinene under photooxidation
conditions has been reported by Roth and co-workers.7 The
products of reaction with methanol with sabinene were found
to be optically active, although the stereochemical configurations
and optical purities were not determined. Nonetheless, the
results show that the reaction is partly, if not completely,
stereospecific.
Figure 2. Stereochemical results for the photooxidations of 4-7.
In this report, we present results on the mechanism of the
photooxidation of 1 and several alkyl-substituted analogues (4-
7).8 By a combination of transient pulsed laser and steady-
state experiments, direct evidence is provided that free, ring-
closed cyclopropane cation radical are generated as reaction
intermediates and that they rapidly react with added nucleophiles
to give substitution products in high yields. In addition,
stereochemical experiments with 4-7 show that the cation
radicals react with nucleophiles in a completely stereospecific
manner with inversion of configuration at the atom undergoing
nucleophilic substitution.
Figure 3. Synthetic scheme for the preparation of 8a-c and 9a-c.
was used with 7. The adducts of methanol are designated as
8a-10a, those of water as 8b and 9b, and those of (formally)
hydrogen cyanide as 8c and 9c. As shown in Figure 2,
cyclopropane 4 gave the anti isomers 8a-c as the sole detectable
substitution products. In contrast, cyclopropanes 5 and 6 gave
only the corresponding syn products 9a-c. These results require
stereospecific nucleophilic substitutions with inversion of con-
figuration. Taking into account the detection limits of the GC
methods used, all of the reactions were found to be >99.7%
stereospecific. Similarly, photooxidation of 7 (22.8 ( 0.3%
optically pure) in the presence of methanol gave (R)-ether 10a10
(22.4 ( 0.7% optically pure). This result requires addition of
methanol with 98 ( 4% inversion in the substitution step.
Results
A. Synthesis of 4-7. Compounds 4-6 were prepared by
the cyclopropanation method of Olofson and Dougherty9 using
benzyl chloride/lithium tetramethylpiperidide and either trans-
2-butene or cis-2-butene, yielding r-1,t-2-dimethyl-t-3-phenyl-
cyclopropane (4) and a mixture of r-1,c-2-dimethyl-c-3-
phenylcyclopropane (5) and r-1,c-2-dimethyl-t-3-phenyl-
cyclopropane (6), respectively. The mixture of 5 and 6 was
used in some of the experiments, while pure samples of pure 5
and 6 were obtained by preparative gas chromatography for
other experiments. Optically active (S)-1-methyl-2,2-diphenyl-
[1-2H]cyclopropane (7) was received as a gift from Prof. H. M.
Walborsky.10
C. Independent Synthesis of the Photoproducts. The
absolute configuration and optical purity of ether 10a derived
from photooxidation of cyclopropane 7 was established by
comparison with literature data.10 The stereochemical configu-
rations of the products derived from photooxidation of cyclo-
propanes 4-6 were determined by independent synthesis as
shown in Figure 3. The syntheses start with the reaction of
benzylmagnesium bromide with either trans- or cis-2-epoxy-
butane11 which gave anti and syn alcohols 8b and 9b,
respectively. These alcohols were transformed into the corre-
sponding anti and syn methoxy ethers 8a and 9a by alkylation
with methyl iodide. Butyronitriles 8c and 9c were prepared
from alcohols 8b and 9b by tosylation followed by an SN2
reaction with cyanide. Due to inversion of configuration in the
second step, 8b gives 9c, and 9b gives 8c.
B. Photooxidations of 4-7. Stereochemical Outcome of
Substitutions. Preparative photooxidations of cyclopropanes
4-7 with several nucleophiles were performed in degassed
acetonitrile using 1-CN as sensitizer. Compounds 4-6 were
reacted with methanol, water, and cyanide, while only methanol
(6) (a) Gassman, P. G.; Olson, K. D.; Walter, L.; Yamaguchi, R. J. Am.
Chem. Soc. 1981, 103, 4977. (b) Gassman, P. G.; Olson, K. D. J. Am. Chem.
Soc. 1981, 103, 4977. (c) Gassman, P. G.; Hay, B. A. J. Am. Chem. Soc.
1986, 108, 4227. (d) Gassman, P. G. In Photoinduced Electron Transfer;
Fox, M. A., Chanon, M., Eds.; Elsevier: Amsterdam, 1988; Part C, Chapter
4.2, p 70. (e) Arnold, D. R.; Du, X. Can. J. Chem. 1994, 72, 403. (f) Weng,
H.; Roth, H. D. J. Org. Chem. 1995, 60, 4136.
(7) Weng, H; Sethuraman, V.; Roth, H. D. J. Am. Chem. Soc. 1994,
116, 7021.
(8) Preliminary results have been previously reported: Dinnocenzo, J.
P.; Todd, W. P.; Simpson, T. R.; Gould, I. R. J. Am. Chem. Soc. 1990,
112, 2464.
(9) Olofson, R. A.; Dougherty, C. M. J. Am. Chem. Soc. 1973, 95, 582.
(10) Walborsky, H. M.; Murari, M. P. Can J. Chem. 1984, 62, 2464.
(11) Closs, G. L.; Goh, S. H. J. Org. Chem. 1974, 39, 1717.

SN2 Reactions of Arylcyclopropane Cation Radicals. 1
J. Am. Chem. Soc., Vol. 119, No. 5, 1997 989
Table 1. Rate Constants (×10-7 M-1 s-1) for the Reaction of
Phenylcyclopropane Cation Radical (1•+) with Alcohols in
Acetonitrile (ACN), 1,2-Dichloroethane (DCE), and
Dichloromethane (DCM) at 23 °C
alcohol
k(alcohol)ACN k(alcohol)DCE k(alcohol)DCM
methanol
1.6 ( 0.1a
1.4 ( 0.1a
1.10 ( 0.08
1.43 ( 0.09
0.73 ( 0.05
1.52 ( 0.09
1.46 ( 0.09
1.19 ( 0.06
1.56 ( 0.08
0.95 ( 0.04
1.99 ( 0.08
ethanol
isopropyl alcohol
butanol
tert-butyl alcohol
1.10 ( 0.06
a Data collected up to 0.15 M of the alcohol.
line with a slope (kq) of 1.9 × 1010 M-1 s-1. The transient
species also reacted rapidly with tris-p-tolylamine. Disappear-
ance of the transient was accompanied by appearance of a new
species with λmax ) 670 nm, which agreed well with that
reported for the tris-p-tolyl amine cation radical.18 Finally,
kinetic measurements showed that the transient reacted with
methanol in a reaction that was first-order in each; the second-
Figure 4. UV-vis spectra from the photosensitized (solid line) and
γ-radiolysis (dotted line) oxidation of 1.
D. Direct Spectroscopic Evidence for Involvement of
Cation Radicals. Transient absorption spectroscopy was used
to measure the UV-vis spectrum of the species obtained from
the photooxidation of phenylcyclopropane (1). The transient
was generated by irradiation of a dioxygen-saturated, acetonitrile
solution containing 1 (0.08 M), N-methylquinolinium hexafluo-
rophosphate (NMQ+, 10-4 M), and toluene (2.0 M) with a laser
pulse (355 nm, 25 ps, 800 µJ). In this experiment, the first
order rate constant (km) was 1.6 × 107 M-1 s-1
.
E. Kinetics for Reaction of 1•+ and 4-6•+ with Alcohols.
The kinetics for the reaction of 1•+ with a series of alcohols
were studied at 23 °C in acetonitrile (ACN), 1,2-dichloroethane
(DCE), and dichloromethane (DCM). Plots of the decay rate
constants vs [alcohol] were linear, and the second-order reaction
rate constants were obtained from the slopes of the plots. The
results are given in Table 1. The cation radicals of 4-6 were
similarly generated and have broad, visible absorption spectra
similar to 1•+ with maxima at 563, 560, and 551 nm,
respectively. The rate constants for reaction of 4-6•+ with
singlet excited state of NMQ+ (E* ) 2.7 V vs SCE)12 is
red
quenched by toluene (Eox ) 2.3 V)13 to produce the N-
methylquinolyl radical (NMQ•) and free toluene cation radical
with a high quantum yield.14 The toluene cation radical
subsequently reacts with 1 (Eox ) 1.9 V)15 by electron transfer.
This co-sensitization method with the positively charged
sensitizer NMQ+ produces a much higher yield of cation radicals
than by directly using 1 in high concentration.16 The photolysis
solution is saturated with dioxygen to remove NMQ•. Dioxygen
reacts with NMQ• to give products that are optically transparent
in the spectral window of interest (400-700 nm). The transient
UV-vis spectrum obtained from 1 under these conditions has a
maximum at 540 nm and is shown in Figure 4 (solid line). The
lifetime of the transient is unaffected by dioxygen and, in the
absence of added quenchers, decays mainly by second-order
kinetics, presumably due to return electron transfer.
methanol in DCM are 3.7, 1.9, and 4.4 × 107 M-1 s-1
,
respectively.
Although the bimolecular kinetics for the reactions of 4-6•+
with methanol are consistent with an SN2 mechanism, they do
not exclude the possibility that the cation radicals undergo rapid,
reversible ring-opening followed by relatively slow capture of
methanol. To test this hypothesis, the photooxidations were
performed in the absence of nucleophile to determine if
isomerization of the cyclopropanes was competitive with
formation of the methanol adducts. In practice, photooxidation
of a mixture of 5 and 6 in CH3CN showed < 2% isomerization
to 4 after 60 h of irradiation. For comparison, preparative
experiments in the presence of 2 M methanol were complete
within 2 h.
Ionization of 1 by γ-radiolysis in a glass matrix of CFCl3
and BrCF2CF2Br at 77 K gave a species with a UV-vis spectrum
that is virtually identical to that obtained by transient absorption
spectroscopy (Figure 4, dotted line).17 We assign these spectra
to the phenylcyclopropane cation radical. Support for this
assignment is provided by the chemical trapping experiments
described below.
Time-resolved nanosecond transient absorption spectroscopy
was used to measure the rate of decay of the transient species
in acetonitrile in the presence of different concentrations of good
electron donors, such as 1,2,4,5-tetramethoxybenzene (TMB).
The cation radical was generated in the same manner as
described above except the excitation source was a nanosecond
laser (ca. 15 ns, 340 nm, 4-6 mJ). A plot of the first-order
decay rate constants for the transient vs [TMB] yielded a straight
F. Stern-Volmer Quenching Experiment. Although the
transient kinetic experiments described above show that phen-
ylcyclopropane cation radical reacts with methanol, they do not
rigorously prove that this reaction is related to the formation of
methanol-adduct 2 in the steady-state photolysis experiment.
A Stern-Volmer quenching experiment was performed to
determine if these two reactions are mechanistically linked.
Addition of increasing amounts of TMB to an acetonitrile
solution containing 1, 1-CN, and 0.45 M MeOH decreased the
quantum yield for formation of 2. A plot of φ0/φ vs [TMB]
was linear and provided a Stern-Volmer slope (kq/km[MeOH])
of 2.2 × 103 M-1. Multiplying the slope by the methanol
concentration gives an estimate for kq/km of 1.0 × 103, which
is in good agreement with the value obtained from transient
kinetics, 1.2 × 103.
(12) Yoon, U. C.; Quillen, S. L.; Mariano, P. S.; Swanson, R.; Stavinoka,
J. L.; Bay, E. J. Am. Chem. Soc. 1983, 105, 1204.
(13) Howell, J. O.; Goncalves, J. M.; Amatore, C.; Klasinc, L.;
Wightman, R. M.; Kochi, J. K. J. Am. Chem. Soc. 1984, 106, 3968.
(14) Dockery, K. P.; Dinnocenzo, J. P.; Farid, S.; Goodman, J. L.; Gould,
I. R. J. Am. Chem. Soc. 1997, 119. In press.
Discussion
(15) Lingenfelter, T. G.; Simpson, T. R., Dinnocenzo, J. P. Manuscript
in preparation.
(16) Todd, W. P.; Dinnocenzo, J. P.; Farid, S.; Goodman, J. L.; Gould,
I. R. J. Am. Chem. Soc. 1991, 113, 3601.
(17) The radiolytic oxidation of 1 at 77 K in a Freon matrix was
performed according to previously published procedures (Bally, T. In
Radical Ionic Systems, Properties of Condensed Phase; Lund, A., Shiotani,
M., Eds.; Kluwer Academic Publishers: Dordrecht, 1991).
Reaction Mechanism. We start by discussing the transient
kinetics experiments described above in the context of the
mechanism originally proposed by Rao and Hixson for the
photooxidation of phenylcyclopropane in the presence of
(18) Gould, I. R.; Ege, D.; Moser, J. E.; Farid, S. J. Am. Chem. Soc.
1990, 112, 4290.

990 J. Am. Chem. Soc., Vol. 119, No. 5, 1997
Dinnocenzo et al.
methanol (Figure 1).4 Regarding the nature of the reactive
cyclopropane intermediate, the strong similarity of the UV-vis
spectra obtained from oxidation of 1 by photoinduced electron
transfer and by γ-radiolysis (Figure 4) demonstrates that the
same species is produced in both experiments. The transient
kinetics experiments show that this species reacts with the good
electron donors TMB and tris-p-tolyl amine with diffusion-
controlled rate constants. Furthermore, the reaction with tris-
p-tolyl amine leads to the formation of the tris-p-tolyl aminium
cation radical. These facts are all consistent with assignment
of the transient species to 1•+. Finally, the agreement between
the rate constant ratios determined for reaction of the species
with TMB vs methanol by time-resolved and steady-state
experiments provides a direct experimental link between the
reaction of 1•+ with methanol and the formation of 2 in the
preparative photolysis experiment.
Figure 5. Proposed mechanism for the formation of products from
the photosensitized oxidation of 1 in the presence of methanol.
The stereochemical probes of the photosensitized reactions
of arylcyclopropanes 4-7 with nucleophiles show that the
reactions are stereospecific and proceed with inversion of
configuration at the site of nucleophilic substitution. These data
are consistent with ring-closed arylcyclopropane cation radical
intermediates undergoing nucleophilic displacement. An alter-
native hypothesis that might explain the stereochemical and
kinetic results would be reversible opening of the cyclopropane
ring in the cation radicals followed by rapid nucleophilic capture
before bond rotation, i.e., an SN1-like mechanism. If this were
occurring, then one would expect to see rapid isomerization of
the starting cyclopropane in the absence of nucleophile. This
was not observed and consequently a mechanism involving
nucleophilic capture of ring-opened cation radical intermediates
is ruled out. The fact that the cation radicals do not rapidly
isomerize in the absence of nucleophiles is consistent with
thermodynamic cycle and quantum chemical calculations.19
These calculations show that the CR-Câ/γ cyclopropane bonds
in the cation radicals are substantially weakened upon one-
electron oxidation, but still have significant bond strengths
(ca. 10 kcal/mol for phenylcyclopropane cation radical).
of the methyl group at Câ to stabilize positive charge in the
substitution transition state. This explanation is consistent with
the regiochemistry of related arylcyclopropane cation radical
substitutions and quantum chemical calculations.19
The lack of benzylic substitution on the ring-closed phenyl-
cyclopropane cation radical can be rationalized by the relative
stabilities of the expected primary substitution products, shown
below. Nucleophiles presumably attack 1•+ at Câ/γ rather than
CR, because the former pathway provides the more stable
distonic cation radical, which benefits from benzylic stabilization
of the radical center.
Following nucleophilic substitution, the reaction products can
be explained by the three-way partitioning of benzyl radical 12
as shown in Figure 5. This leads to a matter of synthetic interest.
All of the preparative experiments described here differ from
previous experiments4,5 in that dioxygen was removed by
rigorous freeze-pump-thaw degassing and that 1-cyanonaph-
thalene (1-CN) was uniformly employed as photosensitizer.
These modifications significantly improve the reaction yields
by suppressing radical-coupling products. In addition to being
synthetically useful, the observations are mechanistically reveal-
ing. Removal of dioxygen by degassing most likely alters the
partitioning of 12 in the following way. At low conversion of
starting material, the sensitizer anion radical (S•-) and 12 are
formed in equal amounts. To the extent that 12 dimerizes,
however, the steady-state concentration of S•- will increase.
After a brief reaction time, the steady state concentration of
S•- will be much greater than of 12. As a result, reaction of
12 with S•- will subsequently dominate.21 If dioxygen is
present, however, S•- will react with dioxygen by electron
transfer which will prevent the condition [S•-] . [12] from
being achieved.
The regiochemistries of substitution provide a second argu-
ment against reaction via ring-opened intermediates. For
example, ring-opening of 4•+, 5•+, or 6•+ would be expected to
give distonic cation radical 10•+ rather than 11•+.20 If 10•+ were
formed then one would expect nucleophilic addition to CR,
which is not observed. The same argument also applies to the
regiochemistry of substitution on 1•+; here the argument is even
more persuasive since formation of a primary carbocation would
be required to explain the observed regioselectivity.
The regiochemistry of reaction of 7•+ with methanol is
interesting because nucleophilic substitution occurs at Câ rather
than Cγ, i.e., at the more hindered carbon atom. The stereo-
chemical outcome of the reaction shows that the substitution is
best rationalized by an SN2 mechanism. The substitution
regiochemistry is, therefore, clearly not determined by steric
effects. We instead propose that it is governed by the ability
The use of 1-CN as a sensitizer instead of DCB is thought to
favor reduction of 12 by the sensitizer anion radical rather than
coupling. Since the reduction potential of 1-CN is -1.98 V22
and the reduction potential of DCB is -1.60 V,22 it is likely
that the 1-CN•- is more efficient at reducing 12 than DCB•-.
Assuming that the reduction potential of 12 is similar to that of
1-phenylethyl radical (-1.60 V),20 then the electron transfer
reaction from DCB•- to 12 is thermoneutral, while the electron
(19) Dinnocenzo, J. P.; Zuilhof, H.; Simpson, T. R.; Lieberman, D. R.;
McKechney, M. W. J. Am. Chem. Soc. 1997, 119, 994.
(20) This can be deduced from the relative oxidation potentials of the
1-phenylethyl radical and the 2-propyl radical, which are 0.37 and 0.47 V,
respectively: Wayner, D. D. M.; McPhee, D. J.; Griller, D. J. Am. Chem.
Soc. 1988, 110, 132.
(21) (a) Fischer, H. J. Am. Chem. Soc. 1986, 108, 3925. (b) Wagner, P.
J.; Thomas, M. J.; Puchalski, A. E. J. Am. Chem. Soc. 1986, 108, 7739.
(22) Kavarnos, G. J.; Turro, N. J. Chem. ReV. 1986, 86, 401.

SN2 Reactions of Arylcyclopropane Cation Radicals. 1
J. Am. Chem. Soc., Vol. 119, No. 5, 1997 991
transfer from 1-CN•- to 12 has a driving force of ca. 0.38 V or
8.8 kcal/mol.
alcohol with R-aryl vinyl carbocations32 (kHOMe/kHOBu ) 9-21)
t
and benzhydryl carbocations33 (kHOMe/kHOBu ) 8-21) are larger
t
than the three-electron SN2 reactions described here. Similarly,
the selective nucleophilic attack of methanol on 7•+ at Câ rather
than Cγ suggests that steric effects are diminished for these three-
electron SN2 reactions. This conclusion is also consistent with
the small steric effects observed for alkyl-substituted phenyl-
cyclopropane cation radicals discussed in the accompanying
paper in J. Am. Chem. Soc. 1997, 119, 994.19
Spectroscopy and Kinetics. It is interesting to compare the
observed spectrum of 1•+ (λmax ) 540 nm) with the cation
radical of cumene, which has an absorption maximum at 440
nm.23 The remarkable red-shift of the spectrum of 1•+ suggests
extensive delocalization of spin and charge into the cyclopropane
ring. This is in accordance with expectations for a ring-closed
structure and is difficult to rationalize by a ring-opened cation
radical. In the latter case, the UV-vis spectrum might be
expected to resemble that of either benzyl radical (λmax ) 316
nm)24 or benzyl cation (λmax ) 305 and 445 nm).25 Extensive
delocalization in 1•+ is also consistent with CIDNP effects
observed in the photooxidation of a variety of arylcyclopro-
panes26 as well as with quantum chemical calculations.19
Conclusions
Direct evidence is provided that photooxidation of aryl-
cyclopropanes in the presence of nucleophiles proceeds via a
bona fide three-electron SN2 mechanism. This mechanism is
supported by transient absorption experiments which demon-
strate the intermediacy of arylcyclopropane cation radicals and
by transient kinetic methods which show that the cation radicals
rapidly react with nucleophiles with second-order kinetics.
Furthermore, the reactions show complete stereospecificity and
proceed with inversion of configuration at the atom undergoing
substitution.
The rate constants for reaction of the cation radicals of 1 and
4-6 with alcohols are very high for SN2 reactions, but all
significantly below that expected for a diffusion-controlled
reaction. The linearity of the plots of the first-order decay rate
constants vs. [MeOH] indicate a 1:1 stoichiometry of the
quenching process, in line with the observed products and
postulated mechanism. Since the pseudo-first-order rate con-
stants for nucleophilic capture at the nucleophile concentrations
used in the preparative experiments are substantially lower than
the rate constants for separation of ion radical pairs,27 it follows
that the cation radicals undergo nucleophilic substitution after
diffusional separation of the geminate ion radical pairs, i.e., they
react as “free” cation radicals.
Experimental Section
General Techniques. All 1H-NMR spectra were recorded at 300.1
MHz, using a General Electric/Nicolet QE-300 spectrometer. 13C-NMR
spectra were recorded at 75.5 MHz using a General Electric/Nicolet
QE-300 spectrometer and calibrated using internal chloroform. Ultra-
violet and visible absorption spectra were recorded using a Hewlett
Packard 8452A diode array spectrophotometer. Low-resolution electron
impact mass spectra were obtained using a Hewlett-Packard 5890A
gas chromatograph equipped with a J&W Scientific Durawax 3, 30 m
× 0.32 mm column with a 0.25 µm film thickness, and a Hewlett-
Packard 5970 mass selective detector. High-resolution mass spectra
were obtained using a VG-7035 mass spectrometer with perfluoro-
kerosene as a standard. Elemental analyses were performed by
Quantitative Technologies Inc., P.O. Box 470, Salem Industrial Park,
Bldg. 5, Rt. 22E, Whitehouse, NJ 08888. Analytical gas chromatog-
raphy was performed on a Hewlett Packard 5890 Series II chromato-
graph equipped with a J&W Scientific Durawax 3, 30 m × 0.32 mm
column with a 0.25 µm film thickness, a flame ionization detector, a
Hewlett Packard 7673A Controller, and a personal computer. Prepara-
tive gas chromatographic separations were performed on a Varian-
Aerograph 1720-1 using helium as the carrier gas and a thermal
conductivity detector. Inert atmosphere manipulations were conducted
under nitrogen in a Vacuum Atmospheres HE-43-2 DRI-LAB glovebox
equipped with a Vacuum Atmospheres HE-493 DRI-TRAIN. Degassed
samples were prepared using three or more freeze-pump-thaw cycles
on a manifold connected to a mercury diffusion pump (1 × 10-6 mmHg)
in conjunction with a Sargent Welch Model No. 1405 oil vacuum pump.
Transient kinetic and absorption experiments were performed on
nanosecond and picosecond apparatuses described previously.34
Materials. Reagents were from commercial sources except when
noted otherwise. Phenylcyclopropane was vacuum-distilled before use.
1-Cyanonaphthalene was sublimed before use. All syntheses were
performed under a nitrogen atmosphere except when noted otherwise.
Column chromatography was performed using silica gel (230-400
mesh; EM Science) and several different eluents (Vide infra) and
invariably followed by concentration by rotary evaporation under
reduced pressure. Whenever intermediary drying of organic layers was
necessary, anhydrous sodium sulfate or magnesium sulfate were used,
except when noted otherwise. Reagent grade solvents were used for
Schepp and Johnston have recently determined rate constants
for the reaction of methanol with substituted styrene cation
radicals to be in the range of 104-108 M-1 s-1, depending on
substituent pattern.28 It is interesting to note that methanol reacts
with the trans-â-methylstyrene cation radical ca. 20 times slower
than with styrene cation radical. In contrast, the methyl-
substituted phenylcyclopropane cation radicals 4-6•+ react as
fast or faster than 1•+, pointing to a different nature of substituent
effects on the SN2 reactions.19,29
The data in Table 1 show that the rate constants of
nucleophilic substitution are rather insensitive to both the solvent
polarity and the steric bulk of the nucleophile. The small solvent
polarity effects are consistent with that expected for reaction
between a charged species and a neutral one.30 The reactivity
differences between methanol and tert-butyl alcohol (kHOMe
/
t
kHOBu ) 1.5-2.1) are smaller than analogous rate data reported
for four-electron SN2 reactions. For example, the relative rate
constants for reaction of 1-bromobutane with methoxide and
tert-butoxide are ≈20:1.31 It is also worth noting that the
relative rate constants for reaction of methanol and tert-butyl
(23) Kelsall, B. J.; Andrews, L. J. Phys. Chem. 1984, 88, 5893.
(24) Arbour, C.; Atkinson, G. H. Chem. Phys. Lett. 1989, 159, 520.
(25) McClelland, R. A.; Chan, C.; Cozens, F.; Modro, A.; Steenken, S.
Angew. Chem., Int. Ed. Engl. 1991, 30, 1337.
(26) (a) Roth, H. D.; Schilling, M. L. M. J. Am. Chem. Soc. 1980, 102,
7958. (b) Roth, H. D.; Schilling, M. L. M. J. Am. Chem. Soc. 1981, 103,
7210. (c) Roth, H. D.; Schilling, M. L. M.; Hutton, R. H.; Truesdale, E. A.
J. Am. Chem. Soc. 1983, 105, 153. (d) Roth, H. D. Croat. Chim. Acta 1984,
57, 1165. (e) Roth, H. D.; Schilling, M. L. M.; Schilling, F. C. J. Am. Chem.
Soc. 1985, 107, 4152. (f) Roth, H. D. Top. Curr. Chem. 1992, 163, 133.
(27) (a) Ojima, S.; Miyasaka, H.; Mataga, N. J. Phys. Chem. 1990, 94,
7534. (b) Yabe, T.; Kochi, J. K. J. Am. Chem. Soc. 1992, 114, 4491. (c)
Arnold, B. R.; Noukakis, D.; Farid, S.; Goodman, J. L.; Gould, I. R. J. Am.
Chem. Soc. 1995, 117, 4399.
(28) Johnston, L. J.; Schepp, N. P. J. Am. Chem. Soc. 1993, 115, 6564.
(29) Dinnocenzo, J. P.; Lieberman, D. R.; Simpson, T. R. J. Am. Chem.
Soc. 1993, 115, 366.
(30) Parker, A. J. Chem. ReV. 1969, 69, 1.
(31) Reeve, W.; Erikson, C. M.; Aluotto, P. F. Can. J. Chem. 1979, 57,
2747.
(32) (a) Kobayashi, S.; Zhu, Q. Q.; Schnabel, W. Z. Naturforsch. 1988,
43b, 825. (b) Verbeek, J.-M.; Stapper, M.; Krijnen, E. S.; Van Loon, J.-D.;
Lodder, G.; Steenken, S. J. Phys. Chem. 1994, 98, 9526.
(33) Bartl, J.; Steenken, S.; Mayr, H. J. Am. Chem. Soc. 1991, 113, 7710.
(34) (a) Chen, L.; Farahat, M. S.; Gaillard, E. R.; Farid, S.; Whitten, D.
G. J. Photochem. Photobiol., A 1996, 95, 21. (b) Arnold, B. R.; Atherton,
S. J.; Farid, S.; Goodman, J. L.; Gould, I. R. Photochem. Photobiol. in
press.

992 J. Am. Chem. Soc., Vol. 119, No. 5, 1997
Dinnocenzo et al.
all extractions and workup procedures. Anhydrous benzene, diethyl
ether, and tetrahydrofuran were obtained by distillation from sodium
benzophenone ketyl under nitrogen. Acetonitrile was distilled under
nitrogen successively from aluminum trichloride, potassium perman-
ganate/lithium carbonate, potassium bisulfite, calcium hydride, phos-
phorous pentoxide and then stored under nitrogen over molecular sieves
(3 Å) prior to use. Deionized water was used for the preparation of
all aqueous solutions. Dichloromethane was refluxed with aluminum
trichloride under nitrogen for 2 h, cooled, and allowed to settle
overnight. After decanting, the dichloromethane was washed with
water, dried over calcium chloride, distilled from phosphorus pentoxide
and stored under nitrogen. Anhydrous dimethylsulfoxide was distilled
from calcium hydride in Vacuo. 2,2,2-Trifluoroethanol was dried over
anhydrous calcium sulfate and sodium bicarbonate and distilled at
atmospheric pressure. Pyridine was distilled from barium oxide.
6.8, 2.93 H). HRMS (EI): Calcd for C12H18O1 (M+): 178.1358.
Found: 178.1346.
(2R*,3R*)-2-Methyl-1-phenyl-3-(p-toluenesulfonyloxy)butane. p-
Toluenesulfonyl chloride (73.39 mg, 0.385 mmol) was added to a
solution of (2R*,3R*)-3-methyl-4-phenyl-2-butanol (48.46 mg, 0.295
mmol) and pyridine (740 µL) at 0 °C. After 26 h, the reaction mixture
was subjected to an extractive workup followed by column chroma-
tography (90:10 hexane:ethyl acetate) to give a white solid (0.033 g,
35%). 1H NMR (CDCl3): δ 7.85 (d, J ) 7.4, 1.71 H), 7.40-7.17 (m,
4.96 H), 7.08 (d, J ) 7.4, 1.86 H), 4.74-4.67 (m, 1.02 H), 2.79 (dd,
J ) 12.9, 4.3, 1.05 H), 2.48 (s, 3.00 H), 2.28 (dd, J ) 12.9, 9.2, 1.07
H), 1.99-1.86 (m, 1.02 H), 1.30 (d, J ) 6.8, 3.19 H), 0.83 (d, J ) 7.4,
3.12 H). The material was immediately used to prepare 9c.
(2R*,3S*)-2-Methyl-1-phenyl-3-(p-toluenesulfonyloxy)butane. This
material was prepared in 44% yield using the same procedure as above
except (2S*,3R*)-3-methyl-4-phenyl-2-butanol (0.049 g, 0.30 mmol)
was used. 1H NMR (CDCl3): δ 7.75 (d, J ) 7.4, 1.79 H), 7.33-7.18
(m, 4.88 H), 7.04 (d, J ) 7.4, 1.90 H), 4.58-4.49 (m, 1.04 H), 2.67
(dd, J ) 12.6, 4.9, 1.02 H), 2.46 (s, 3.00 H), 2.27 (dd, J ) 12.6, 8.6,
1.04 H), 2.05-1.94 (m, 1.02 H), 1.28 (d, J ) 6.5, 3.22 H), 0.84 (d, J
) 6.8, 3.08 H). The material was immediately used to prepare 8c.
(2R*,3R*)-2,3-Dimethyl-4-phenylbutanecarbonitrile (8c). A solu-
tion of (2R*,3S*)-2-methyl-1-phenyl-3-(p-toluenesulfonyloxy)butane
(32.9 mg, 0.103 mmol), potassium cyanide (26.8 mg, 0.412 mmol),
and dimethylsulfoxide (2.5 mL) was stirred at room temperature for
1.7 h followed by heating to 90 °C for 18 h. Extractive workup
followed by column chromatography (95:5 hexane:ethyl acetate) gave
a colorless oil (0.016 g, 89%). 1H NMR (CDCl3): δ 7.38-7.16 (m,
5.65 H), 2.78-2.58 (m, 2.77 H), 1.94-1.83 (m, 0.87 H), 1.31 (d, J )
7.2, 2.93 H), 1.11 (d, J ) 6.7, 2.78 H). HRMS (EI): Calcd for
C12H15N1 (M+): 173.1204. Found: 173.1179.
r-1,t-2-Dimethyl-t-3-phenylcyclopropane (4). Preparation of this
material was adapted from the method of Olofson and Dougherty.9 A
flask equipped with a Dewar condenser was successively charged with
dry ether (20 mL), trans-2-butene (18.5 g, 330 mmol), benzyl chloride
(2.80 mL, 24.33 mmol), and tetramethylpiperidine (800 µL, 4.74 mmol).
The flask was immersed in a -10 °C bath, and methyllithium (1.4 M
in ether, 17.4 mL, 24.36 mmol) was added over 2 h by means of a
syringe pump. The solution was warmed to room temperature for 3 h,
and then 5% HCl (50 mL) was carefully added. The resulting mixture
was transferred into a separatory funnel and extracted with 2 × 30 mL
ether. The combined ethereal extracts were washed successively with
portions of brine, saturated sodium bicarbonate, brine, and dried.
Concentration gave a light yellow oil (2.8 g) which after column
chromatography (hexane) gave a colorless oil (1.23 g, 35%). 1H NMR
(CDCl3): δ 7.33-7.18 (m, 5.17 H), 1.80 (dd, J ) 7.2, 5.9, 0.98 H),
1.24 (d, J ) 5.8, 3.01 H), 1.04-0.86 (m, 4.94 H).
r-1,c-2-Dimethyl-c-3-phenylcyclopropane (5) and r-1,c-2-Di-
methyl-t-3-phenylcyclopropane (6). This mixture of diastereomers
was prepared using the same procedure9 as above except cis-2-butene
was used. Column chromatography (hexane) gave a colorless oil (0.80
g, 25%). 1H NMR (CDCl3): δ 7.34-7.20 (m, 5.07 H), 2.04 (t, J )
8.9, 0.93 H), 1.26-1.15 (m, 2.32 H), 1.00-0.93 (m, 5.68 H).
(2R*,3R*)-3-Methyl-4-phenyl-2-butanol (8b). This material was
prepared following the procedure of Closs.11 Benzylmagnesium chloride
(18.25 mL, 0.90 M in ether, 16.43 mmol) was added dropwise to a
solution of trans-2,3-epoxybutane (1.30 mL, 14.9 mmol) and 20 mL
ether. Extractive workup followed by concentration gave 2.12 g of a
yellow oil which after column chromatography (90:10 hexane:ethyl
acetate) gave a colorless oil (1.03 g, 42%). 1H NMR (CDCl3): δ 7.34-
7.19 (m, 5.18 H), 3.85-3.76 (m, 0.94 H), 2.85 (dd, J ) 13.3, 5.9, 0.99
H) 2.43 (dd, J ) 13.3, 9.0, 0.96 H), 1.88-1.75 (m, 0.94 H), 1.34 (bs,
0.99 H), 1.23 (d, J ) 6.3, 3.00 H), 0.89 (d, J ) 6.9, 3.00 H).
(2S*,3R*)-3-Methyl-4-phenyl-2-butanol (9b). This material was
prepared as described above except cis-2,3-epoxybutane was used.
Column chromatography (90:10 hexane:ethyl acetate) gave a colorless
oil (1.07 g, 44%). 1H NMR (CDCl3): δ 7.33-7.18 (m, 4.85 H), 3.78-
3.68 (m, 0.97 H), 2.91 (dd, J ) 13.4, 4.9, 1.00 H) 2.37 (dd, J ) 13.4,
9.4, 1.01 H), 1.92-1.78 (m, 0.99 H), 1.37 (bs, 1.01 H), 1.24 (d, J )
6.3, 3.15 H), 0.86 (d, J ) 6.8, 3.03 H).
(2R*,3R*)-2-Methoxy-3-methyl-4-phenylbutane (8a). A solution
of 0.9 mL of water, sodium hydroxide (0.90 g, 22 mmol), tetra-n-
butylammonium iodide (0.206 g, 0.56 mmol), methyl iodide (1.40 mL,
22.5 mmol), and (2R*,3R*)-3-methyl-4-phenyl-2-butanol (0.370 g, 2.25
mmol) was refluxed for 95 h. Extractive workup followed by column
chromatography (90:10 hexane:ethyl acetate) gave a colorless oil (0.33
g, 82%). 1H NMR (CDCl3): δ 7.32-7.17 (m, 4.86 H), 3.36 (s, 2.96
H), 3.24 (dq, J ) 6.2, 3.7, 1.02 H) 2.86 (dd, J ) 13.2, 5.6, 1.06 H),
2.35 (dd, J ) 13.2, 9.2, 1.10 H), 1.89 (m, 0.97 H), 1.14 (d, J ) 6.3,
2.96 H), 0.86 (d, J ) 6.8, 3.05 H). HRMS (EI): Calcd for C12H18O1
(M+): 178.1358. Found: 178.1320.
(2S*,3R*)-2,3-Dimethyl-4-phenylbutanecarbonitrile (9c). This
material was prepared in 67% yield using the same procedure as above
except (2R*,3R*)-2-methyl-1-phenyl-3-(p-toluenesulfonyloxy)butane
was used. 1H NMR (CDCl3): δ 7.34-7.16 (m, 4.94 H), 2.89 (dd, J )
13.5, 5.5, 0.96 H), 2.63-2.48 (m, 1.84 H), 2.11-2.00 (m, 1.00 H),
1.34 (d, J ) 7.2, 3.17 H), 1.01 (d, J ) 6.8, 3.10 H). HRMS (EI):
Calcd for C12H15N1 (M+): 173.1204. Found: 173.1173.
Photooxidations. General Procedure. A Pyrex photolysis vessel
was typically charged with a solution containing the cyclopropane,
1-cyanonaphthalene, n-pentyl ether (internal standard), and the nucleo-
phile in 600 µL of acetonitrile. The solution was degassed by three
freeze-pump-thaw cycles and then sealed with a hand torch. The
sealed tube was placed in a Rayonet photoreactor containing nominal
300 nm light bulbs and irradiated for 1-3 h. The mixture was
subsequently analyzed by GC, in which the retention times were
compared with those of independently synthesized materials. All
reported yields are based on GC analysis and corrected for differences
in response factors. A portion (typically 3-6 mg) of the major
3
product(s) was isolated by preparative GC (column: 6' × /8" 17%
XF-1150 on Anachrom Q, except when noted otherwise) and analyzed
1
by H NMR, GC, and other techniques as described.
r-1,t-2-Dimethyl-t-3-phenylcyclopropane (4) in the Presence of
Methanol. A degassed solution of 4 (23.9 mg, 163.4 µmol), 1-cyano-
naphthalene (4.3 mg, 28.1 µmol), n-pentyl ether (25.3 mg, 159.8 µmol),
methanol (50 µL), and acetonitrile (600 µL) was irradiated for 1.25 h.
GC analysis showed the major product (95.7%) to have retention time
identical to that of independently synthesized (2R*,3R*)-2-methoxy-
3-methyl-4-phenylbutane (8a) and the absence of g0.1% of the
(2S*,3R*)-2-methoxy-3-methyl-4-phenylbutane (9a). A portion of the
1
product was isolated by preparative GC; its H NMR spectrum was
identical to that of the independently synthesized 8a.
r-1,c-2-Dimethyl-c-3-phenylcyclopropane (5) and r-1,c-2-Di-
methyl-t-3-phenylcyclopropane (6) in the Presence and Absence of
Methanol. The same procedure was followed as for 4 except a mixture
of 5 and 6 was used. GC analysis showed the major product (83.2%)
to have a retention time identical to that of independently synthesized
(2S*,3R*)-2-methoxy-3-methyl-4-phenylbutane (9a) and the absence
(2S*,3R*)-2-Methoxy-3-methyl-4-phenylbutane (9a). This mate-
rial was prepared in 72% yield using the same procedure as above
except (2S*,3R*)-3-methyl-4-phenyl-2-butanol was used. 1H NMR
(CDCl3): δ 7.32-7.17 (m, 5.48 H), 3.34 (s, 2.74 H), 3.17 (dq, J )
6.1, 5.6, 0.95 H) 2.83 (dd, J ) 13.4, 5.0, 1.04 H), 2.35 (dd, J ) 13.4,
9.3, 1.04 H), 1.97 (m, 0.95 H), 1.14 (d, J ) 6.3, 2.88 H), 0.82 (d, J )
of g0.1% (2R*,3R*)-2-methoxy-3-methyl-4-phenylbutane (8a).
portion of the product was isolated by preparative GC; its H NMR
A
1
spectrum was identical to that of independently synthesized 9a. In a

SN2 Reactions of Arylcyclopropane Cation Radicals. 1
J. Am. Chem. Soc., Vol. 119, No. 5, 1997 993
similar experiment run without added methanol, GC analysis after 61
h of photolysis showed formation of 1.2% of 4.
dimethyl-4-phenylbutanecarbonitrile (8c). A portion of the first product
was isolated by preparative GC; its H NMR spectrum was identical
1
r-1,t-2-Dimethyl-t-3-phenylcyclopropane (4) in the Presence of
Water. A degassed solution of 4 (16.8 mg, 115 µmol), 1-cyanonaph-
thalene (4.9 mg, 31.9 µmol), n-pentyl ether (15.6 mg, 98.3 µmol), water
(40 µL), and acetonitrile (600 µL) was irradiated for 2 h. GC analysis
showed the major product (81.6%) to have a retention time identical
to that of independently synthesized (2R*,3R*)-3-methyl-4-phenyl-2-
butanol (8b) and the absence of g0.2% of (2S*,3R*)-3-methyl-4-
phenyl-2-butanol (9b). A portion of the product was isolated by
preparative GC; its 1H NMR spectrum was identical to that of the
independently synthesized 8b.
r-1,c-2-Dimethyl-c-3-phenylcyclopropane (5) and r-1,c-2-Di-
methyl-t-3-phenylcyclopropane (6) in the Presence of Water. The
same procedure as above was followed except a mixture of 5 and 6
was used. GC analysis showed the major product (74.3%) to have a
retention time the same as that of independently synthesized (2S*,3R*)-
3-methyl-4-phenyl-2-butanol 9b and the absence of g0.2% of (2R*,3R*)-
3-methyl-4-phenyl-2-butanol 8b. A portion of the product was isolated
by preparative GC; its 1H NMR spectrum was found to be identical to
that of the independently synthesized 9b.
to the independently synthesized 9c.
(S)-(+)-1-Methyl-2,2-diphenyl[1-2H]cyclopropane (7) in the Pres-
ence of Methanol. A degassed solution of 7 (80.73 mg, 386 µmol),
1-cyanonaphthalene (19.36 mg, 126 µmol), n-pentyl ether (63.83 mg,
403 µmol), methanol (170 µL), and acetonitrile (2.0 mL) was irradiated
for 3 h. Concentration gave 129.0 mg of a light yellow oil from which
a portion of the product (77.6%) was isolated by preparative GC (6' ×
3/8" 20% SE-30). The product was analyzed by 1H-NMR, and its
spectrum was found to be identical to that reported for 3-deuterio-3-
methoxy-1,1-diphenylbutane.10 1H NMR (CDCl3): δ 7.18-7.34 (m,
9.83 H), 4.21 (dd, J ) 9.8, 6.2, 1.00 H), 3.26 (s, 3.02 H), 2.28 (dd, J
) 13.8, 6.1, 1.05 H), 2.13 (dd, J ) 13.8, 9.9, 1.04 H), 1.15 (s, 3.06 H).
The optical rotation of the starting material [R]Hg25 +34.3 (4)° (c 2.46,
CHCl3) and product [R]Hg25 -4.3 (2)° (c 4.34, CHCl3) were measured
using a polarimeter. The optical purities of the starting material and
the product were compared to literature values10 for (R)-(-)-1-methyl-
25
2,2-diphenylcyclopropane ([R]Hg -150.6° (c 1.18, CHCl3)) and (S)-
25
(+)-2-methoxy-4,4-diphenylbutane ([R]Hg +19.2° (c 2.10, CHCl3)).
The optical purities were thus determined to be 22.8 (3)% and 22.4
(7)%, respectively.
r-1,t-2-Dimethyl-t-3-phenylcyclopropane (4) in the Presence of
Cyanide. A degassed solution of 4 (17.04 mg, 116.5 µmol), 1-cyano-
naphthalene (5.1 mg, 33.2 µmol), n-pentyl ether (14.4 mg, 91.1 µmol),
potassium cyanide (46.8 mg, 719 µmol), water (100 µL), and acetonitrile
(600 µL) was irradiated for 3 h. GC analysis showed the formation of
two major products with retention times identical to those of (2R*,3R*)-
2,3-dimethyl-4-phenylbutanecarbonitrile (8c, 26.5%) and (2R*,3R*)-
3-methyl-4-phenyl-2-butanol (8b, 37.7%), and the absence of g0.3%
of (2S*,3R*)-2,3-dimethyl-4-phenylbutanecarbonitrile (9c). A portion
Acknowledgment. Research support was provided by the
National Science Foundation (CHE-9312460) and by the donors
of the Petroleum Research Fund, administered by the American
Chemical Society. T.R.S is the recipient of an Elon Huntington
Hooker Graduate Fellowship. H. Z. is grateful to the Nether-
lands Organization for Scientific Research (NWO) for a
postdoctoral Talent-stipendium. We thank Professor Thomas
Bally (University of Fribourg) for obtaining the matrix UV-vis
spectrum of phenylcyclopropane cation radical. We also thank
Professor Sason Shaik (Hebrew University) for helpful discus-
sions and Professor Harry M. Walborsky (Florida State Uni-
versity) for a generous gift of 7.
1
of the first product was isolated by preparative GC, and its H NMR
spectrum was identical to that of the independently synthesized 8c.
r-1,c-2-Dimethyl-c-3-phenylcyclopropane (5) and r-1,c-2-Di-
methyl-t-3-phenylcyclopropane (6) in the Presence of Cyanide. The
same procedure as above was followed except a mixture of 5 and 6
was used. GC analysis showed the formation of two major products
with retention times identical to those of (2S*,3R*)-2,3-dimethyl-4-
phenylbutanecarbonitrile (9c, 21.8%) and (2S*,3R*)-3-methyl-4-phenyl-
2-butanol (9b, 23.0%) and the absence of g0.3% of (2R*,3R*)-2,3-
JA9631110