Three-electron S(N)2 reactions of arylcyclopropane cation radicals. 2. Steric and electronic effects of substitution

Article abstract of DOI:10.1021/ja963112s

The nucleophilic substitution reactions on substituted arylcyclopropane cation radicals were studied by a combination of methods including product studies, time-resolved laser flash photolysis, kinetic isotope effects, and quantum chemical calculations. The reactions were found to proceed stereospecifically with inversion of configuration, with high regioselectivity for nucleophilic attack at the more substituted carbon atom, and with very small steric effects. Electronic effects on the nucleophilic substitution regiochemistry and the rate constants were found to be substantial for substituents on the cyclopropane moiety and on the aryl ring.

Full text of DOI:10.1021/ja963112s

994
J. Am. Chem. Soc. 1997, 119, 994-1004
Three-Electron SN2 Reactions of Arylcyclopropane Cation
Radicals. 2. Steric and Electronic Effects of Substitution1
J. P. Dinnocenzo,* H. Zuilhof, D. R. Lieberman, T. R. Simpson, and
M. W. McKechney
Contribution from the Department of Chemistry, UniVersity of Rochester,
Rochester, New York 14627-0216
ReceiVed September 4, 1996X
Abstract: The nucleophilic substitution reactions on substituted arylcyclopropane cation radicals were studied by a
combination of methods including product studies, time-resolved laser flash photolysis, kinetic isotope effects, and
quantum chemical calculations. The reactions were found to proceed stereospecifically with inversion of configuration,
with high regioselectivity for nucleophilic attack at the more substituted carbon atom, and with very small steric
effects. Electronic effects on the nucleophilic substitution regiochemistry and the rate constants were found to be
substantial for substituents on the cyclopropane moiety and on the aryl ring.
Introduction
selected arylcyclopropane cation radicals and their transition
states for nucleophilic substitution by methanol.
Recent studies on the reactions of nucleophiles with arylcy-
clopropane cation radicals have shown some interesting differ-
ences in the structure-reactivity relationships of three-electron
vs. four-electron nucleophilic substitutions. For example, the
1-cyanonaphthalene(1-CN)-photosensitized reaction of (S)-1-
methyl-2,2-diphenyl[1-2H]cyclopropane (1) with methanol pro-
vides optically active (R)-ether 2 as the only product (eq 1).2
The reaction is proposed to proceed through the intermediacy
of 1•+, which undergoes selective, nucleophilic attack of
methanol at Câ rather than Cγ, i.e., at the more highly substituted
carbon atom. This unusual regioselectivity was first observed
by Rao and Hixson for the photosensitized oxidation of trans-
1-methyl-2-phenylcyclopropane with methanol3 and contrasts
strongly with that observed in classical four-electron SN2
reactions, where substitution generally occurs preferentially at
the least hindered carbon atom.
Finally, electronic effects of aryl substitution on the reac-
tivities of para-substituted phenylcyclopropane cation radicals
(5) with nucleophiles were measured by time-resolved nano-
second laser flash photolysis (eq 2).
Results
A. Syntheses. Cis- and trans-1-methyl-2-phenylcyclopro-
pane were prepared from the reaction of trans-4-phenyl-3-buten-
2-one with hydrazine followed by thermal decomposition of the
resulting 2-pyrazoline.4 1,1-Dimethyl-2-phenylcyclopropane
was prepared via the reaction of benzyl chloride/lithium
tetramethylpiperidide with isobutylene.5 The 1,1-diphenyl-2-
alkylcyclopropanes and 1,1-dimethyl-2,2-diphenylcyclopropane
were synthesized by photolysis of diphenyldiazomethane in the
presence of the appropriate alkene.6
cis-3-Methyl-2,2-diphenylcyclopropaneacetonitrile (6) was
prepared in four steps from 1,1-diphenylpropene as shown in
Scheme 1. Reaction of 1,1-diphenylpropene with ethyl diazo-
acetate under rhodium catalysis gave a diastereoisomeric mixture
of cis- and trans-3-methyl-2,2-diphenylcyclopropanecarboxylate
esters. After chromatographic separation of the diastereoiso-
mers, the cis-3-methyl-2,2-diphenylcyclopropanecarboxylate
In this paper, we present a systematic study of substituent
effects on three-electron SN2 reactions of arylcyclopropane
cation radicals. The effects of alkyl substituents on the
cyclopropyl moiety were investigated using monophenyl- and
diphenylalkylcyclopropanes 3 and 4 to probe how the number,
the steric size, and the electronic properties of substituents affect
the reaction regiochemistry. Time-resolved nanosecond laser
flash photolysis was used to measure the substituent effects on
the rate constants for reaction of the cyclopropane cation radicals
with nucleophiles. In addition, secondary kinetic isotope effects
were measured to probe the distribution of positive charge in
the substitution transition states. These experimental studies
were complemented by quantum chemical calculations on
X Abstract published in AdVance ACS Abstracts, January 15, 1997.
(1) (a) Taken in part from the doctoral thesis of T.R.S., University of
Rochester, 1995. (b) A preliminary account of this work has been
published: Dinnocenzo, J. P.; Lieberman, D. R.; Simpson, T. R. J. Am.
Chem. Soc. 1993, 115, 366.
(2) Dinnocenzo, J. P.; Simpson, T. R.; Zuilhof, H.; Todd, W. P.; Heinrich,
T. Preceding paper in this issue (J. Am. Chem. Soc. 1997, 119, 987.
(3) Rao, V. R.; Hixson, S. S. J. Am. Chem. Soc. 1979, 101, 6458.
(4) (a) Casey, C. P.; Polichnowski, S. W.; Shusterman, A. J.; Jones, C.
R. J. Am. Chem. Soc. 1979, 101, 7282. (b) In our hands the reaction yielded
between 0 and 26%. In general, we have found other procedures to be
somewhat more reproducible: Denmark, S. E.; Christenson, B. L.; Coe,
D. M.; O’Connor, S. P. Tetrahedron Lett. 1995, 36, 2215.
(5) Olofson, R. A.; Dougherty, C. M. J. Am. Chem. Soc. 1973, 95, 581.
(6) Baron, W. J.; Hendrick, M. E.; Jones, M., Jr. J. Am. Chem. Soc.
1973, 95, 6286.
S0002-7863(96)03112-5 CCC: $14.00 © 1997 American Chemical Society

SN2 Reactions of Arylcyclopropane Cation Radicals. 2
J. Am. Chem. Soc., Vol. 119, No. 5, 1997 995
Scheme 1a
Scheme 2
a (a) N2dCHCO2Et, Rh2(O2C(CH2)4CH3)4 (b) chrom. sep. (c) LiAlH4.
(d) (i) p-TsCl, (ii) n-Bu4NCN.
ester was reduced to the alcohol, reacted with tosyl chloride,
and then treated with cyanide to give 6. trans-3-Methyl-2,2-
diphenylcyclopropaneacetonitrile (7) was prepared analogously
from the trans-3-methyl-2,2-diphenylcyclopropanecarboxylate
ester.
The para-substituted phenylcyclopropanes were synthesized
by modified Simmons-Smith cyclopropanation7 of the corre-
sponding para-substituted styrenes (see Experimental Section).
B. Photooxidations. Photooxidations of cyclopropanes
6-13 in methanol or acetonitrile were performed in rigorously
degassed solutions containing 1-cyanonaphthalene (1-CN) as a
photosensitizer. Reported yields refer to reactions run on a small
scale and were determined by gas chromatography (GC).8
Reactions were run on a larger scale for preparative isolation
of the photoproducts. The results are presented in Scheme 2.
Photooxidation of cyclopropane 6 and 7 gave products of
nucleophilic attack by methanol with complete inversion of
configuration. Cyclopropane 6 gave as its sole detectable
product the syn substitution product 6a, while 7 gave only the
corresponding anti product 7a. These results show that nu-
cleophilic attack by methanol is not only stereospecific but also
regioselective, in agreement with earlier observations.1-3 This
was also observed for cyclopropanes 8-11 which react with
methanol selectively at the more hindered carbon atom, giving
the methanol-addition products 8a-11a. For 12, the photo-
oxidation in methanol gave in addition to the major nucleophilic
substitution product 12a, the tetralin-derivatives 12b-c, and
4-methoxy-3,4-dimethyl-1,1-diphenylpentane (12d). To deter-
mine the regioselectivity of methanol substitution on 12, the
product of methanol substitution at Cγ was independently
synthesized (Vide infra). GC analysis showed that the reaction
was >99.9% regioselective for substitution at Câ vs.Cγ. In dry
acetonitrile, photooxidation of 12 gave tetralin-derivatives
12b-c and cyclic imine 12e. Cyclopropane 13 provided the
substitution product 13a in methanol. As done for 12, the
product of methanol substitution at Cγ of 13 was independently
prepared. GC analysis revealed that the photooxidation reaction
was >99.99% regioselective. Finally, photooxidation of 13 in
dry acetonitrile gave cyclic imine 13e.
reduced with sodium in the presence of Fe(III)acetylacetonate9
to give syn-ether 6a′ and anti-ether 7a′, respectively. These
1
products were then correlated by H NMR and GC to com-
pounds independently synthesized as shown in Scheme 3.
Reaction of cis-2-epoxybutane with diphenylmethyllithium
followed by methylation of the resulting alcohol gave syn-ether
6a′. Analogously, trans-2-epoxybutane provided anti-ether 7a′.
The structure of photoproduct 8a was determined by inde-
pendent synthesis via the base-catalysed methylation of R,R-
dimethylbenzenepropanol. The structures of the remaining
photooxidation products were deduced by spectroscopic methods
(see Experimental Section).
C. Independent Synthesis of Photoproducts. The stereo-
chemistries of the products from the photooxidations of cyclo-
propanes 6 and 7 in methanol were determined by chemical
correlation. First, methanol-addition products 6a and 7a were
(7) (a) S ) MeO, Me: Furukawa, J.; Kawabata, N.; Nishimura, J.
Tetrahedron 1968, 24, 53. (b) S ) F, Cl: Denmark, S. E.; Edwards, J. P.
J. Org. Chem. 1991, 56, 6974. (c) S ) C(dO)Me: Levina, R. Ya.;
Gembitskii, P. A.; Kostin, V. N.; Shostakovskii, S. M.; Treshchova, E. G.
Zh. Obshch. Khim. 1963, 33, 365. (d) S ) OC(dO)Me: Horrom, B. W.;
Mazdiyasni, H. OPPI Briefs 1992, 6, 696.
In order to determine the regioselectivities for methanol
substitution at Câ vs. Cγ in the photooxidations of cyclopropanes
12 and 13, it was necessary to independently prepare the
(8) See preceding paper in this issue (J. Am. Chem. Soc. 1997, 119, 987)
for general description of the reaction procedure.
(9) Tamelen, E. E. van; Rudler, H.; Bjorkland, C. J. Am. Chem. Soc.
1971, 93, 7113.

996 J. Am. Chem. Soc., Vol. 119, No. 5, 1997
Dinnocenzo et al.
Scheme 3
Scheme 4a
a (a) RR′CdCH2, hν. (b) (i) Li/But-PhPh-But; (ii) MeI. (14, R ) R′
) Me; 15, R ) But, R′ ) H).
products that would be formed by substitution at Cγ since they
could not be detected in the crude reaction mixtures by 1H NMR
analyses. Ether 14, which would be formed by Cγ substitution
on 13•+, was prepared by the Paterno-Bu¨chi reaction of
benzophenone with isobutylene10 followed by reductive cleavage
of the oxetane and methylation of the resulting alcohol (Scheme
4). Ether 15, the product of Cγ substitution on 12•+, was
prepared in an analogous fashion starting from benzophenone
and 3,3-dimethyl-1-butene.
D. Kinetics of Nucleophilic Substitution. Hammett Study.
A Hammett study was performed to study the electronic
characteristics of the transition states for the SN2 reaction of
arylcyclopropane cation radicals with methanol and pyridine
as nucleophiles. The cation radicals were produced by photo-
induced electron transfer using N-methylquinolinium hexafluo-
rophosphate (NMQ) as a photooxidant and toluene as a
cosensitizer (eq 3).11 The UV-vis spectra of the cation radicals
were measured by picosecond laser flash photolysis (Figure 1).
The reactivity of the cyclopropane cation radicals with nucleo-
philes (eq 2) in 1,2-dichloroethane (1,2-DCE) was studied using
nanosecond laser transient absorption spectroscopy. In the
presence of nucleophiles the decay of the cation radical signals
followed pseudo-first-order kinetics. The pseudo-first-order rate
constants for disappearance of the cation radicals were deter-
mined by fitting the decay profiles at 540 nm with monoexpo-
nential functions. In all cases, plots of the rate constants vs.
nucleophile concentration showed good linearity. The slopes
Figure 1. UV-vis spectra of the cation radicals of para-substituted
phenylcyclopropanes in 1,2-dichloroethane at 23 °C.
Table 1. Rate Constants (in M-1 s-1) for the Reaction of
Para-Substituted Phenylcyclopropane Cation Radicals with Methanol
(kMeOH) and Pyridine (kpyr) in 1,2-Dichloroethane at 23 °C
substituent
σ+
kMeOH
kpyr
MeO
Me
OC(dO)Me
F
H
Cl
-0.78
-0.31
-0.18
-0.07
0.0
+0.11
+0.41
e 1.0 × 105
3.0 (( 0.5) × 106
6.1 ((0.7) × 108
5.2 ((0.2) × 108
3.2 ((0.4) × 109
3.9 ((0.1) × 109
1.7 ((0.1) × 109
4.0 ((0.4) × 109
1.2 ((0.3) × 106
1.7 ((0.1) × 106
9.7 ((0.1) × 106
1.5 ((0.1) × 107
8.4 ((0.2) × 106
2.9 ((0.3) × 107
C(dO)Me
of the plots gave the bimolecular rate constants, kMeOH and kpyr
,
for reactions with methanol and pyridine, respectively. The
results are presented in Table 1. Hammett plots of these data
vs σ+ and Arnold’s σ•R substituent constants are shown in
Figure 2.12
Effect of Alkyl Substituents on the Cyclopropyl Moiety.
The cation radicals of the arylcyclopropanes 8-13 and 16-18
(10) Arnold, D. R.; Hinman, R. L.; Glick, A. H. Tetrahedron Lett. 1964,
5, 1425.
Figure 2. Hammett plots: log kNu for the reaction of arylcyclopropane
cation radicals with nucleophiles (squares ) pyridine; circles )
methanol) in 1,2-dichloroethane at 23 °C, versus σ+ and σR• .
(11) (a) Todd, W. P.; Dinnocenzo, J. P.; Farid, S.; Goodman, J. L.; Gould,
I. R. J. Am. Chem. Soc. 1991, 113, 3601. (b) Dockery, K. P.; Dinnocenzo,
J. P.; Farid, S.; Goodman, J. L.; Gould, I. R.; Todd, W. P. J. Am. Chem.
Soc. 1997, 119. In press.
were generated by photoinduced electron transfer as described
above. All of the cation radicals gave strong absorption peaks
in the visible region which were recorded by either picosecond
(12) (a) σ+: Hansch, C.; Leo, A.; Taft, R. W. Chem. ReV. 1991, 91,
165. (b) σ•R: Wayner, D. D. M.; Arnold, D. R. Can J. Chem. 1984, 62,
1164. (c) σ+ (COMe) is a calculated value (see ref 12a).

SN2 Reactions of Arylcyclopropane Cation Radicals. 2
J. Am. Chem. Soc., Vol. 119, No. 5, 1997 997
Table 2. Rate Constants (in M-1 s-1) for Phenyl- and
Diphenylcyclopropane Cation Radicals 8•+-13•+ and 16•+-19•+
with Methanol in Dichloromethane at 23 °C
compd
R1
R2
R3
kMeOH
19•+
9•+
Ph
Ph
Ph
Ph
Ph
Ph
H
H
H
H
H
H
H
H
H
H
Me
H
Me
H
Me
1.7 × 107 a
1.5 × 108
8.3 × 107
3.0 × 107
4.8 × 106
3.2 × 108
1.0 × 107 a
3.1 × 107
3.1 × 108
1.5 × 108
Figure 3. B3LYP/6-31G* optimized structures (bond lengths in Å)
for the cation radicals of phenylcyclopropane (16•+; left) and cis-1-
methyl-2-phenylcyclopropane (18•+; right).
Me
Et
10•+
11•+
12•+
13•+
16•+
17•+
18•+
8•+
Pri
But
Me
H
Table 3. Selected Bond Lengths (Å) and Group Charges (Q) for
B3LYP/6-31G(d) Calculated Alkylarylcyclopropane Cation Radicals
H
Me
Me
compd r(CR-Câ) r(CR-Cγ) r(Câ-Cγ) r(CR-CPh) Q(Ph) Q(Câ) Q(Cγ)
16•+
17•+
18•+
8•+
1.591
1.723
1.789
1.943
1.591
1.523
1.511
1.450
1.450
1.463
1.469
1.483
1.431
1.428
1.426
1.422
0.681 0.187 0.187
0.592 0.290 0.158
0.570 0.310 0.155
0.491 0.406 0.121
a Statistically corrected (kobsd/2).
or nanosecond transient absorption methods.2 The kinetics of
their reactions with methanol were determined by monitoring
the decay profiles of the cation radicals at the respective λmax
by using nanosecond laser flash methods as described above.
The bimolecular rate constants (kMeOH) are presented in Table
2.
with the 6-31G(d) basis set.15 The advantages of this method
for calculating cation radical structures and energies have
recently been discussed.16 Selected bond lengths and group
charges for the cation radicals are shown in Table 3.
Based on the data in Table 3, it is clear that phenylcyclo-
propane cation radical is structurally distinct from the three
alkyl-substituted cation radicals. Following the notation of
Hudson et al.,17 16•+ is a 2L1N isomer (two lengthened C-C
bonds, one normal C-C bond) while 17•+, 18•+, and 8•+ are
1L2N isomers (one lengthened C-C bond, two normal C-C
bonds). Several geometric features are most prominently
different in these species. First, it is seen that the addition of
methyl groups at Câ results in a significant increase in r(CR-
Câ) and that the incremental increase is roughly additive with
each methyl group. Concurrent with this are decreases in r(CR-
Cγ) and r(CR-Ph) and increases in r(Câ-Cγ). Second, as shown
in Figure 3, the orientation of the phenyl ring with respect to
the cyclopropane ring changes significantly with alkyl substitu-
tion. In 16•+, the phenyl group bisects the cyclopropane ring,
as expected for a 2L1N structure. For all three alkyl-substituted
cation radicals, the orientation is similar to that shown for 18•+,
where the plane of the phenyl ring is nearly completely aligned
with the CR-Cγ bond. Thus in 16•+, two cyclopropane C-C
σ bonds profit from partial overlap with the π-orbitals on the
phenyl ring, while for the alkyl-substituted cation radicals the
π-orbitals overlap maximally with the long CR-Câ bond. As
shown in Table 3, the geometric changes due to alkyl-
substitution result in significant redistribution of the positive
charge. For example, the phenyl group charge, Q(Ph), in 16•+
is 0.68. This decreases upon 2,2-dimethyl substitution to only
0.49 in 8•+. Simultaneously, there are large increases in Q(Câ)
and small decreases in Q(Cγ) upon alkyl-substitution at Câ.
Calculations of the vibrational frequencies confirmed that all
of the calculated cation radical structures were energy minima.
The calculated expectation values of S2 ranged from 0.762-
0.765, in good agreement with that expected for a pure doublet
state (0.750). That the cation radicals are predicted to have
structures with bonded cyclopropane rings is consistent with
the stereochemical and kinetic evidence obtained from experi-
ment. Confirmation is also provided by approximate estimates
E. Kinetic Isotope Effects. The 1-cyanonaphthalene-
photosensitized reactions of phenylcyclopropane (16) and trans-
1-methyl-2-phenylcyclopropane (17) with n-butanol (n-C4H9OH)
and nonadeuteriobutanol (n-C4D9OH) were studied to determine
â-secondary kinetic isotope effects (KIEs). These KIEs provide
information on the buildup of positive charge on the nucleophile
in the transition states.13 For both 16 and 17 the sole products
are those derived from butanol substitution (Scheme 5). For
17, n-butanol substitution occurs exclusively at Câ, which is
the same substitution regiochemistry observed in the photosen-
sitized reaction of 17 with methanol. The products of reaction
of 16 and 17 with n-C4H9OH vs. n-C4D9OH (16a-b and 17a-
b) were baseline separable by capillary GC, so that the KIEs
could be determined from the product ratios.14 The measured
KIEs for 16 and 17 at 20.0 ( 0.5 °C were 0.9798 ( 0.0033
and 0.9419 ( 0.0032, respectively. An analogous experiment
was attempted with 1,1-dimethyl-2-phenylcyclopropane (8). In
this case, however, several minor products were formed in
addition to the major substitution product. Since such a result
can complicate an accurate determination of the KIE in several
ways, this experiment was not pursued further.
F. Quantum Chemical Calculations: Arylcyclopropane
Cation Radicals. The data in Scheme 2 show that alkyl groups
exert a powerful directing effect on the regiochemistry of
reaction of substituted arylcyclopropane cation radicals with
nucleophiles. To assess how alkyl substituents at Câ affect the
reactant cation radical structures, quantum chemical calculations
were performed on phenylcyclopropane cation radical (16•+),
trans-1-methyl- (17•+), cis-1-methyl- (18•+), and 1,1-dimethyl-
2-phenylcyclopropane (8•+) cation radical. All calculations were
performed with the density functional hybrid-method B3LYP
(15) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (b) Stephens, P.
J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J. J. Phys. Chem. 1994,
98, 11623.
(16) Zuilhof, H.; Dinnocenzo, J. P.; Reddy, C.; Shaik, S. J. Phys. Chem.
1996, 100, 15774.
(17) Hudson, C. E.; Giam, C. S.; McAdoo, D. J. J. Org. Chem. 1993,
58, 2017.
(13) (a) Sunko, D. E.; Hehre, W. J. Prog. Phys. Org. Chem. 1983, 14,
205. (b) Lee, I. Chem. Soc. ReV. 1995, 24, 223.
(14) For other examples of this technique to determine kinetic isotope
effects, see: (a) De Vaal, P.; Lodder, G.; Cornelisse, J. J. Phys. Org. Chem.
1992, 5, 581. (b) Zuilhof, H.; Van Gelderen, F. A.; Cornelisse, J., Lodder,
G. Submitted for publication.

998 J. Am. Chem. Soc., Vol. 119, No. 5, 1997
Dinnocenzo et al.
Scheme 5
spin and charge in the ring-opened cation radical is different
than in the other three cation radicals. This is the principal
reason that 8•+ has a significantly lower BDE(CR). The other
three cation radicals have similar BDE(CR) values due to
approximately offsetting effects of alkyl-substitution on Eox(N)
and BDE(N).
G. Quantum Chemical Calculations: Transition States
for Nucleophilic Substitution. Quantum chemical calculations
were performed to evaluate the effect of alkyl substitution on
the structures and energies of the transition states for nucleo-
philic substitution by methanol. Due to the high computational
demands of these calculations, semiempirical (AM1 and PM3)
methods were employed.21 In all cases, force constant calcula-
tions were performed to ensure that the optimized geometries
were indeed transition states for the nucleophilic substitution
reactions. Selected geometric and electronic features of the
transition states for reaction of phenylcyclopropane cation radical
as well as cation radicals with alkyl substituents at Câ are listed
in Table 5.
In view of the high regioselectivity observed for the nucleo-
philic reaction arylcyclopropane cation radicals at the more
substituted carbon atom, transition state energies were also
calculated for reaction with methanol at the less substituted
carbon atom in order to obtain the differential activation
enthalpies, ∆∆Hq. The calculated values are shown in Table
6. These data will be discussed in detail below.
Scheme 6
Table 4. Oxidation Potentials (V vs. SCE) and Bond Dissociation
Energies (kcal/mol) for Neutral 16, 17, 18, and 8 as well as Their
Cation Radicals
compd
Eox(N)
BDE(N)
Eox(BR)
BDE(CR)
16
17
18
8
1.94
1.73
1.71
1.65
46
43
42
41
0.37
0.37
0.37
0.09
10
11
11
5
Discussion
of the bond dissociation energies of the cation radicals as
determined from the thermodynamic cycle shown in Scheme
6.1b,18 The results of these calculations as well as the associated
thermodynamic data are given in Table 4.19,20 It includes the
oxidation potentials of the cyclopropanes, Eox(N), the estimated
bond dissociation energies of the neutral cyclopropanes, BDE-
(N), and the oxidation potentials of the 1,3-biradicals produced
by homolytic cleavage of the CR-Câ bond, Eox(BR). For 16•+,
17•+, and 18•+, Eox(BR) was approximated by the oxidation
potential reported for the 1-phenylethyl radical (0.37 V).20d For
8•+, Eox(BR) was taken as the oxidation potential for the tert-
butyl radical (0.09 V).20d Thus for 8•+, the apportionment of
Product Studies. Previous steady-state and transient kinetics
experiments have shown that the photooxidation of arylcyclo-
propanes in the presence of nucleophiles such as methanol are
consistent with the reaction mechanism shown in Figure 4.2 The
key feature of this mechanism in the context of the present
discussion is the SN2 reaction between the cyclopropane cation
radicals and methanol.
We begin by discussing the results from photooxidation of
cyclopropanes 8-13. All of the cyclopropanes undergo clean
photooxidation in methanol to give products from nucleophilic
substitution at Câ, except for 12, which additionally gives several
rearrangement products. These will be discussed later. Based
on the product studies and the reaction mechanism, we conclude
that the cation radicals of cyclopropanes 8-13 show a strong
preference for nucleophilic substitution at the more hindered
carbon atom, i.e., at Câ rather than Cγ. No substitution is
observed at CR, which would be expected if the reaction
occurred by an SN1 mechanism.2,19 The regiochemical selec-
tivities determined in the cases of 12 (>99.99%) and 13
(>99.9%) show that the energetic preferences must be signifi-
cant, since neopentyl and tertiary substitutions at Câ are favored
over primary substitutions at Cγ. The results can be rationalized
by assuming that the alkyl groups at Câ stabilize positive charge
in the substitution transition states and that this stabilization
overwhelms the opposing steric effects. This hypothesis is also
consistent with results from the photooxidations of cyclopro-
panes 6 and 7. Here, the substituents at Câ and Cγ are
comparable sterically but not electronically.12a In each of these
cases nucleophilic substitution is observed only at the carbon
atom bearing the more electron-donating methyl group. The
fact that the nucleophilic substitutions also proceed with
complete inversion of configuration provides additional evidence
for an SN2 mechanism.
(18) Wayner, D. D. M.; Parker, V. D. Acc. Chem. Res. 1993, 26, 287.
(19) (a) Oxidation potentials of the neutrals, Eox(N), were obtained from
the equilibrium constants for electron transfer between reference aromatic
compounds and the cyclopropanes as studied by picosecond laser flash
transient absorption spectroscopy (Lingenfelter, T. G.; Simpson, T. R.;
Dinnocenzo, J. P. Manuscript in preparation). (b) Bond dissociation energies
of the neutrals, BDE(N), were estimated based on group additivity
relationships. For 16, BDE(N) was obtained by taking the C-C BDE of
cyclopropane (59 kcal/mol)20a and correcting it for the effect of phenyl
substitution by adding the difference between the C-H BDEs for Ph-
CHMe-H (85 kcal/mol)20b and MeCH2-H (98 kcal/mol).20b For 17, BDE-
(N) was obtained by taking the C-C BDE of 16 (46 kcal/mol) and correcting
it for the effect of methyl substitution by adding the difference between
the C-H BDEs for Me2CH-H (95 kcal/mol)20b and MeCH2-H (98 kcal/
mol).20b For 18, BDE(N) was obtained by taking the C-C BDE of 17 (43
kcal/mol) and correcting it for the effect of having the substituents in a cis
configuration by adding the difference between the heats of formation for
trans-1,2-dimethylcyclopropane (-1 kcal/mol)20c and cis-1,2-dimethylcy-
clopropane (0 kcal/mol).20c For 8, BDE(N) was obtained by taking the C-C
BDE of 16 (46 kcal/mol) and correcting it for the effect of geminal dimethyl
substitution by adding the difference between the C-H BDEs for Me3C-H
(93 kcal/mol)20b and MeCH2-H (98 kcal/mol).20b (c) The oxidation potential
of the biradical, Eox(BR), was estimated by taking the lower oxidation
potential of the two radical sites. E1/2 for the 1-phenylethyl radical (0.37
V)20d was used for 16, 17, and 18. E1/2 for the tert-butyl radical (0.09 V)20d
was used for 8.
(20) (a) Doering, W. v. E. Proc. Natl. Acad. Sci. U.S.A. 1981, 78, 5279.
(b) McMillen, D. F.; Golden, D. M. Annu. ReV. Phys. Chem. 1982, 33,
493. (c) Lias, S. G.; Bartmess, J. E.; Liebman, J. F.; Holmes, J. L.; Levin,
R. D.; Mallard, W. G. In Gas-Phase Ion and Neutral Thermochemistry;
Lide, D. R. J., Ed.; J. Phys. Chem. Ref. Data 1988, 17, Supplement no. 1.
(d) Wayner, D. D. M.; McPhee, D. J.; Griller, D. J. Am. Chem. Soc. 1988,
110, 132. (e) Wayner, D. D. M., personal communication.
Photooxidation of cyclopropane 12 in methanol provides, in
addition to the nucleophilic substitution product 12a, products
(21) (a) AM1: Dewar, M. J. S.; Zoebisch, E. G.; Healy, E. F.; Stewart.
J. J. P. J. Am. Chem. Soc. 1985, 107, 3902. (b) PM3: Stewart, J. J. P. J.
Comput. Chem. 1989, 10, 209, 221.

SN2 Reactions of Arylcyclopropane Cation Radicals. 2
J. Am. Chem. Soc., Vol. 119, No. 5, 1997 999
Table 5. Selected Transition State Data for the Reaction of Several 1-Alkyl-2-phenylcyclopropane Cation Radicals with Methanol
R,R′
H,H
Me,H
H,Me
But,H
Me,Me
H,H
Me,H
H,Me
But,H
Me,Me
method
rq(CR-Câ) (Å)
rq(Câ-O) (Å)
θ
Qq(Ph)
Qq(Câ)
Qq(HOMe)
∆Hq (kcal/mol)
AM1
AM1
AM1
AM1
AM1
PM3
PM3
PM3
PM3
PM3
2.24
2.34
2.34
2.35
2.43
2.22
2.32
2.32
2.34
2.42
2.10
1.97
1.98
2.01
1.88
1.95
1.89
1.87
1.93
1.83
139.9
137.6
130.5
132.9
129.3
143.0
140.7
135.2
136.0
133.4
0.29
0.22
0.21
0.21
0.11
0.27
0.24
0.21
0.22
0.18
0.39
0.50
0.46
0.52
0.54
0.36
0.45
0.41
0.45
0.47
0.12
0.18
0.17
0.16
0.23
0.21
0.24
0.26
0.22
0.29
3.0
4.2
4.1
3.2
8.6
10.0
11.2
9.7
7.5
14.0
Table 6. Calculated ∆∆Hq (∆Hq(Cγ) - ∆Hq(Câ), kcal/mol) for
Reaction of 1-Alkyl-2-phenyl- or 1,1-Dialkyl-2-phenylcyclopropane
Radical Cations with Methanol at Cγ vs. Câ
substituents
AM1
PM3
cis-Me
5.8
10.5
6.3
5.8
9.0
5.0
trans-Me
trans-But
2,2-Me2
15.5
13.3
Figure 6. Proposed mechanism for the reaction of diphenylcyclopro-
pane cation radicals with CH3CN.
of 12 undergoes nucleophilic attack by methanol at Câ in
competition with neopentyl rearrangement that is concerted with
ring-opening of the cation radical.
Reactions of strained-ring cation radicals with CH3CN to form
cyclic imines such as 12e have been previously reported in
several cases.24 The reaction of CH3CN with 12•+ differs from
these only in that we suggest it proceeds by an SN2 mechanism
(Figure 6). This mechanistic proposal is based on the regio-
chemistry of nucleophilic substitution, which is analogous to
that observed for the reaction of 12•+ with methanol. As
described above, an SN1 mechanism leads to the prediction that
nucleophilic substitution should occur at CR, which is not
observed. The distonic cation radical produced from nucleo-
philic attack by CH3CN can form the imine product in several
ways. It can undergo ring closure as shown in Figure 6,
followed by reduction with either the sensitizer anion radical
or 12. The latter process would lead to a chain reaction which
has previously been shown to be feasible.24a Alternatively, the
distonic cation radical could be reduced by the sensitizer anion
radical to produce a biradical which undergoes ring closure to
12e. Our present data do not distinguish between these
possibilities. A cyclic imine is also formed from the photo-
oxidation of cyclopropane 13 in CH3CN. In this case, the
reaction regiochemistry can be explained by either an SN2 or
an SN1 mechanism. An SN1 mechanism cannot be ruled out
here because ring-opening of 13•+ is expected to give a distonic
cation radical with the positive charge at Câ,19c and thus
nucleophilic capture would be expected to occur there, as
observed.
Figure 4. Mechanism for the photooxidation of arylcyclopropanes in
the presence of methanol.
Figure 5. Proposed mechanism for formation of 12b-d.
12b-c and 12d. As shown in Figure 5, the latter products can
be rationalized by an alkyl-migration mechanism analogous to
that proposed for the solvolysis of neopentyl derivatives22 and
for rearrangements of other strained ring cation radicals.23
Although an alternative SN1-type mechanism in which 12•+
undergoes fragmentation of the CR-Câ bond, followed by either
nucleophilic capture or neopentyl rearrangment, cannot be ruled
out, the observed products argue against it. If unimolecular
fragmentation of the CR-Câ bond occurred, then it should give
a distonic cation radical in which the radical is localized at Câ
and the positive charge at CR.19c This distonic cation radical
should lead to nucleophilic capture at CR, which is not observed.
For this reason we postulate that the ring-closed cation radical
Kinetics. Based on the reaction regioselectivities described
above, the three-electron SN2 reactions of arylcyclopropane
cation radicals 8•+-13•+ appear to be dominated by electronic
rather than steric effects. The steric effects that are present can
be estimated by comparing the second-order rate constants for
reaction of 9•+-12•+ with methanol. As seen from the data in
Table 2, the rate constants decrease along this series. We
attribute this to a steric effect. It is worth noting, however,
(22) (a) Dauben, W. G.; Chitwood, J. L. J. Am. Chem. Soc. 1968, 90,
6876. (b) Ando, T.; Morisaki, H. Tetrahedron Lett. 1979, 20, 121. (c) Shiner,
V. J., Jr.; Reib, R. C. Tetrahedron Lett. 1979, 20, 121. (d) Shiner, V. J.,
Jr.; Tai, J. J. J. Am. Chem. Soc. 1981, 103, 436. (e) Yamataka, H.; Ando,
T. J. Am. Chem. Soc. 1982, 104, 1808. (e) Yamataka, Y.; Ando, T.; Nagase,
S.; Hanamura, M.; Morokuma, K. J. Org. Chem. 1984, 49, 631.
(23) (a) Adam W.; Walter, H.; Chen, G.-F.; Williams, F. J. Am. Chem.
Soc. 1992, 114, 3007. (b) Weng, H.; Sheik, Q.; Roth, H. D. J. Am. Chem.
Soc. 1995, 117, 10655.
(24) (a) Zona, T. A.; Goodman, J. L. J. Am. Chem. Soc. 1993, 115, 4925.
(b) Arnold, D. R.; Du, X. Can. J. Chem. 1994, 72, 403. (c) ref. 23b.

1000 J. Am. Chem. Soc., Vol. 119, No. 5, 1997
Dinnocenzo et al.
that the effects are much smaller than those observed in
comparable four-electron SN2 reactions. For example, the Me/
But ratio in the three-electron SN2 reaction is only 31, whereas
in typical four-electron substitutions it is ca. 105!25 Based on
the small steric effects measured for the three-electron substitu-
tions, it seems plausible that electronic factors can overwhelm
the steric effects, thus explaining nucleophilic attack at the more
substituted carbon atom. Possible reasons for the small steric
effects will be discussed below.
was assessed by fixing the plane of the phenyl ring such that it
was colinear with the CR-Cγ bond. As mentioned above, this
structural feature is common to all of the alkyl-substituted
phenylcyclopropane cation radicals, which have 1L2N struc-
tures. When optimized with this one constraint, the resulting
phenylcyclopropane cation radical structure had one long
cyclopropane bond: r(CR-Câ) ) 1.671 Å. The other cyclo-
propane bond lengths (r(CR-Cγ) ) 1.526 Å; r(Câ-Cγ) ) 1.430
Å) were found to be comparable to those in the alkyl-substituted
phenylcyclopropane cation radicals. Although the 1L2N struc-
ture for phenylcyclopropane cation radical is not predicted to
be a minimum, its energy is only 1.0 kcal/mol above the 2L1N
minimum. Thus it is clear that the potential energy surface for
distortion of the phenylcyclopropane cation toward a long-bond
structure is quite soft. Further rotation of the phenyl group
causes a much steeper increase in the energy of the cation
radical, however. For example, rotation by 90° leads to the
transition state for rotational isomerization which is 12.0 kcal/
mol higher in energy than the ground state, bisected structure.
Transition State Structures and Energies. Selected struc-
tural and electronic parameters for the AM1 and PM3 calculated
transition states for backside nucleophilic attack of methanol
on phenylcyclopropane cation radical are shown in Table 5.
Several common trends are seen in the data. First, both methods
predict r(CR-Câ) to increase significantly on going from the
reactants to the transition state. Second, both methods predict
the CR-Câ-O angle (θ) to be ca. 140°sfar from the 180° angle
nominally found in most conventional four-electron SN2 reac-
tions. Third, nucleophilic attack by methanol results in a
substantial increase in Q(Câ) from reactants to transition state
(from 0.19 to ca. 0.4). The increase in Q(Câ) comes largely at
the expense of Q(Ph), which decreases from 0.68 to ca. 0.28.
The large shift in positive charge away from the phenyl group
provides an understanding of why the rate constants for
nucleophilic sustitution on para-substituted phenylcyclopropane
cation radicals correlate reasonably well with Hammett σ+
substituent constants and why F is positive. Electron-donating
substituents make it more difficult to transfer positive charge
out of the phenyl group toward both Câ and the nucleophile
upon substitution, which results in increased reaction barriers.
It is also interesting to compare rate constants for reaction of
methanol with the cation radicals of 1,1-diphenyl- (19), 1,1-
diphenyl-2-methyl- (9), and 1,1-dimethyl-2,2-diphenylcyclo-
propane (13), which reveal the effects of mono- and dimethyl-
substitution at Câ. As shown by the data in Table 2, the
substitution rate constants increase with increasing alkyl sub-
stitution, a trend opposite that found in typical four-electron
SN2 reactions. Although the origin of this trend is more difficult
to evaluate because both electronic and steric factors change, it
is clear that steric effects do not dominate these nucleophilic
substitution reactions. We considered the possibility that the
increase in rate constants observed with increasing alkyl
substitution at Câ might be due to an internal steric effect,
namely release of strain upon nucleophilic substitution between
the phenyl groups at CR and the cis-alkyl group(s) at Câ. To
test this hypothesis, the rate constants for reaction of the
analogous monophenylcyclopropane cation radicals (16•+, 17•+,
18•+, 8•+) with methanol were measured. As described above,
8
•+ reacts with methanol exclusively at Câ; 17•+ and 18•+ have
been previously shown to do so as well.3 The rate data in Table
2 show that addition of a single methyl group at Câ leads to an
increase in the rate constant for reaction with methanol. The
effect is slightly larger for a cis-methyl group than a trans-methyl
group. Addition of a second methyl group leads to either a
small increase or decrease in rate constant relative to one methyl
group depending on whether comparison is made to the trans-
or cis-methyl derivatives. Based on these data we conclude
that there may be a small internal steric component to the rate
increase observed in the diphenylcyclopropane cation radical
series. A clear interpretation of the data is complicated,
however, by the prediction that the cation radical structures
change significantly upon alkyl substitution (see below).
As shown in Table 5, both computational methods predict
that alkyl substitution results in reasonably smooth changes in
all of the transition state parameters except the activation
energies. For example, rq(Câ-O) is predicted to continually
decrease upon increasing alkyl-substitution. This change is
understandably accompanied by increases in both Qq(Câ) and
Qq(HOMe). These trends suggest that the transition states
generally become “later” with increasing alkyl-substitution. The
increase in rq(CR-Câ) with increasing alkyl-substitution seems
to be in accord with this conclusion. This measure of reaction
progress must be analyzed cautiously, however, because the
reactant r(CR-Câ) distances also increase significantly upon
alkyl-substitution (see Table 3). In fact, the ∆r(CR-Câ)
variations for the ground state cation radicals exceed ∆rq(CR-
Câ) for the transition states. Other ground state cation radical
parameters are also considerably affected by alkyl-substitution
[e.g., Q(Câ) and Q(Ph)]. For this reason, the estimation of
progress along the reaction coordinate cannot be based solely
on trends in the transition state properties.
Cation Radical Structures. The most striking observation
regarding the calculated structures for the arylcyclopropane
cation radicals are the large structural changes that accompany
alkyl-substitution at Câ (Table 3). Most notably, r(CR-Câ)
increases significantly upon alkyl-sustitution at Câ. This stands
in marked contrast to the corresponding neutral cyclopropanes
whose structures are comparatively insensitive to substitution.
The likely origin of the structural changes can be deduced from
the group charges listed in Table 3. It is seen that alkyl-
substitution leads to a significant increase in Q(Câ) and
corresponding decreases in Q(Ph) and Q(Cγ). It is reasonable
to conclude from these data that the structural changes are driven
by the ability of the alkyl groups to stabilize positive charge in
the cation radicals.
The data in Table 3 also reveal that phenylcyclopropane cation
radical is predicted to have a structure with two lengthened
cyclopropane C-C bonds, whereas the alkyl-substituted deriva-
tives have one lengthened bond. The calculated 2L1N structure
for phenylcyclopropane cation radical is somewhat unusual for
cyclopropane cation radicals which generally prefer 1L2N
structures.17,26 For this reason, it seemed of interest to determine
to what degree a 2L1N structure was preferred in this case. This
The increases in Qq(Câ) and Qq(HOMe) with increasing alkyl-
substitution show that substituents significantly polarize the
transition states. These predictions are in agreement with the
â-secondary kinetic isotope effects measured for the reaction
(25) Streitwieser, A. SolVolytic Displacement Reactions; McGraw-Hill:
New York, 1962; p 13.
(26) Krogh-Jespersen, K.; Roth, H. D. J. Am. Chem. Soc. 1992, 114,
8388, and references therein.

SN2 Reactions of Arylcyclopropane Cation Radicals. 2
J. Am. Chem. Soc., Vol. 119, No. 5, 1997 1001
of phenylcyclopropane cation radical (16•+) and trans-1-methyl-
2-phenylcyclopropane cation radical (17•+) with n-BuOH vs d9-
n-BuOH. Inverse isotope effects are observed in both experi-
ments, consistent with charge buildup on the nucleophile in the
transition states. Reaction of 17•+ shows a more inverse isotope
effect, consistent with the greater Qq(Nu) predicted in this case.
As mentioned above, the CR-Câ-O angle (θ) for nucleo-
philic attack of methanol on phenylcyclopropane cation radical
is significantly less than 180°. The reaction of cyclopropane
cation radical (C3H6•+) with nucleophiles is also calculated to
have a nucleophilic attack angle of <180°.27 The origin of the
deviation from a colinear displacement can be traced to the
electronic mechanism of nucleophilic substitution. Using a
valence bond model, the reaction barriers for three-electron SN2
reactions on σ-cation radicals have been previously analyzed
in terms of the crossing of two states that are related by single
electron transfer from the nucleophile to the σ* orbital of the
bond undergoing nucleophilic attack.28 The avoided crossing
interaction of these states is proportional to the overlap between
the nucleophile lone pair orbital and σ*. The CR-Câ-Nu angle
which maximizes this overlap, and thereby minimizes the
activation barrier, is expected to occur for θ <180° for the
reaction of cyclopropane cation radicals with nucleophiles, as
schematically illustrated below.
Figure 7. Transition states for nucleophilic substitution by methanol
at Cγ on the cation radicals of cis-1-methyl-2-phenylcyclopropane (left)
and trans-1-methyl-2-phenylcyclopropane (right)
predicted to have little effect on ∆Hq, consistent with the
experimental rate data (Table 2). However, addition of a second
methyl group is predicted to lead to a large increase in ∆Hq
and thus a large decrease in rate constant. This is not observed
experimentally. Further work will be required to understand
the discrepancy.
Table 6 lists the calculated differential activation enthalpies
for reaction of alkyl-substituted phenylcyclopropane cation
radicals with methanol at Câ vs. Cγ. Here both computational
methods give consistent results and good agreement with
experiment. In all cases, a high regioselectivity for reaction at
Câ is predicted and observed. An unanticipated result from these
calculations is the prediction that the energetic preference for
reaction of cis-1-methyl-2-phenylcyclopropane cation radical at
Câ vs. Cγ is significantly greater than for the isomeric trans-
cation radical. The likely origin of this effect can be seen by
comparing the transition state structures for substitution at Cγ
for the cis- and trans-cation radicals (Figure 7). Note that
conjugation of the CR-Cγ bond undergoing nucleophilic cleav-
age with the phenyl group leads to an eclipsing interaction
between one of the ortho-phenyl hydrogens and the methyl
group in the cis-isomer which is absent in the trans-isomer. As
a result, substitution at Cγ in the cis-isomer is disfavored both
sterically and electronically. Unfortunately, the prediction
cannot be tested in these particular cases because both experi-
mental regioselectivities are larger than can be accurately
measured. Nonetheless, this may provide a new structural
element which could be useful for controlling the regioselectivity
of other nucleophilic substitutions.
As shown in Table 5, both rq(Câ-O) and θ are calculated to
decrease with increasing alkyl-substitution at the carbon atom
undergoing nucleophilic displacement. The fact that θ decreases
on going from trans-1-methyl-2-phenylcyclopropane cation
radical to trans-1-tert-butyl-2-phenylcyclopropane cation radical,
while Qq(Câ) and Qq(HOMe) remain relatively unaffected,
suggests that θ is at least partially controlled by steric factors.
Consistent with this interpretation, methanol substitution on 1,1-
dimethyl-2-phenylcyclopropane cation radical shows the small-
est θ. It should be noted that the steric consequences of the
trends in rq(Câ-O) and θ oppose each other. The decrease in
rq(Câ-O) with alkyl-substitution increases the steric interactions
between the nucleophile and the substituents attached to Câ. In
contrast, the simultaneous decrease in θ decreases the steric
interactions. Consequently, based on the calculations, it is
difficult to predict the direction or magnitudes of the steric
effects from the structural changes in the transition states. Thus
it is not surprising that the calculated activation enthalpies for
the corresponding nucleophilic substitution show no clear trend.
For example, addition of a single alkyl group at Câ is predicted
to have a modest effect on ∆Hq, regardless of whether the
substituent is cis or trans, large or small. However, addition of
a second alkyl group is predicted to raise ∆Hq significantly.
The calculated ∆Hq values listed in Table 5 show a significant
method dependence. ∆Hq calculated for reaction of phenylcy-
clopropane cation radical with methanol using PM3 is 7.0 kcal/
mol higher than that calculated using AM1. A similar trend is
seen in the remaining ∆Hq values. It is clear that the ∆Hq values
calculated with PM3 are too large to be consistent with the
experimental rate constants. Despite the absolute differences
between the AM1 and PM3 data, the predicted trends in ∆Hq
with alkyl-substitution at Câ are reasonably similar. The results
are best described as in modest agreement with experiment,
however. For example, addition of one methyl group at Câ is
Conclusions
The nucleophilic substitution reactions of arylcyclopropane
cation radicals with alkyl-substitutents on the cyclopropane
moiety proceed in good yields by an SN2 mechanism on the
ring-closed cation radicals. Substitution takes place at the most
highly substituted carbon atom with a high degree of regiose-
lectivity, even when the carbon is tertiary or neopentyl. Steric
effects measured for alkyl substituents attached to the carbon
atom undergoing substitution are very small. The preference
for nucleophilic attack at the more substituted site is ascribed
to the electron-donating ability of the alkyl-substituents which
stabilize positive charge in the substitution transition states. This
electronic effect is supported by experimental data (isotope
effects and competitive substituent effects) and by computational
results.
Experimental Section
The general techniques and apparatus are described in the experi-
mental section of the preceding paper (J. Am. Chem. Soc. 1997, 119,
987).2
Ethyl cis- and trans-3-Methyl-2,2-Diphenylcyclopropanecarbox-
ylate. Ethyl diazoacetate (1.20 mL, 11.4 mmol) was added over 6 h
by means of a syringe pump to solution containing 1,1-diphenylpropene
(25.72 mg , 132 mmol),29 rhodium hexanoate dimer (38.28 mg, 57
(27) Shaik, S.; Reddy, A. C.; Ioffe, A.; Dinnocenzo, J. P.; Danovich,
D.; Cho, J. K. J. Am. Chem. Soc. 1995, 117, 3205.
(28) Shaik, S. S.; Dinnocenzo, J. P. J. Org. Chem. 1990, 55, 3434.

1002 J. Am. Chem. Soc., Vol. 119, No. 5, 1997
Dinnocenzo et al.
µmol),30 hexane (30 mL), and diethyl ether (10 mL). After 2 h, the
resulting solution was concentrated. Column chromatography using
hexane as an eluent (until the excess diene was eluted), followed by
hexane:ethyl acetate (98:2) gave a mixture of the title compounds as a
colorless oil (2.76 g, 86%). Anal. Calcd for C19H20O2: C, 81.40; H,
7.19. Found: C, 81.00; H, 7.17. The isomers were separated by
medium pressure liquid chromatography.
Ethyl cis-3-Methyl-2,2-diphenylcyclopropanecarboxylate. (1.44
g, 45%): 1H NMR (CDCl3): δ 7.32-7.10 (m, 9.74 H), 4.16-4.03
(m, 2.03 H), 2.40 (d, J ) 8.9, 0.98 H), 2.12-2.03 (m, 1.05 H), 1.45
(d, J ) 6.6, 3.14 H), 1.21 (t, J ) 7.1, 3.07 H).
After separation of the layers, the aqueous layer was extracted with
pentane (3 × 30 mL). The combined organic layers were washed with
brine, dried, and concentrated to give a yellow oil which gave after
column chromatography (97:3 hexane:ethyl acetate) a white solid (339
mg, 36%). 1H NMR (CDCl3): δ 7.40-7.09 (m, 10.12 H), 2.46 (dd, J
) 17.3, 1.00 H), 2.21 (dd, J ) 17.3, 7.3, 1.04 H), 1.92-1.84 (m, 0.97
H), 1.80-1.72 (m, 0.99 H), 1.18 (d, J ) 6.6, 2.87 H). 13C NMR
(CDCl3): δ 146.72, 137.20, 131.09, 128.79, 128.15, 126.77, 125.81,
119.24, 37.38, 24.49, 22.98, 15.22, 10.80. Anal. Calcd for C18H17N1:
C, 87.41; H, 6.93. Found: C, 87.21; H, 6.88.
trans-3-Methyl-2,2-diphenylcyclopropaneacetonitrile (7). This
material was prepared using the same procedure as above except trans-
3-methyl-2,2-diphenylcyclopropanemethanol was used instead. Puri-
fication of the crude product by column chromatography (90:10 hexane:
ethyl acetate) gave a white solid (0.27 g, 52%). 1H NMR (CDCl3): δ
7.34-7.18 (m, 10.23 H), 2.23-2.01 (m, 2.08 H), 1.71-1.60 (m, 1.91
H), 0.97 (d, J ) 6.0, 2.77 H). 13C NMR (CDCl3): δ 141.64, 141.55,
129.86, 128.72, 128.39, 126.88, 126.51, 119.35, 41.25, 25.79, 22.80,
18.56, 14.89. Anal. Calcd for C18H17N1: C, 87.41; H, 6.93. Found:
C, 87.11; H, 6.99.
(2R*,3R*)-3-Methyl-4,4-diphenyl-2-butanol. Under an argon at-
mosphere, lithium (154 mg, 22 mmol) was added to a solution of
biphenyl (1.54 g, 10 mmol) and 1,1,2,2-tetraphenylethane (3.35 g, 10
mmol)29 in tetrahydrofuran (85 mL). After 2 h, the reaction mixture
was refluxed for 45 min and then cooled to -10 °C. trans-2,3-
Epoxybutane (1.21 mL, 14 mmol) in tetrahydrofuran (15 mL) was added
dropwise over 45 min. The reaction mixture was warmed to room
temperature for 2.5 h and then several drops of water were carefully
added. Extractive workup followed by column chromatography
(starting with methylene chloride followed by ether) gave a clear
colorless oil (2.29 g, 69%). 1H NMR (CDCl3): δ 7.43-7.11 (m, 9.99
H), 3.87 (d, J ) 11.5, 0.93 H), 3.79-3.72 (m, 1.01 H), 2.38-2.27 (m,
0.95 H), 1.25 (brs, 1.09 H), 1.17 (d, J ) 6.5, 3.06 H), 0.84 (d, J ) 6.7,
2.97 H).
(2S*,3R*)-3-Methyl-4,4-diphenyl-2-butanol. This material was
prepared using the same procedure as above except cis-2,3-epoxybutane
was used. Column chromatography (85:15 methylene chloride:ethyl
aceate) gave a clear colorless oil (97%). 1H NMR (CDCl3): δ 7.32-
7.12 (m, 10.41 H), 3.79-3.71 (m, 0.94 H), 3.58 (d, J ) 11.1, 1.02 H),
2.72-2.60 (m, 0.99 H), 1.41 (broad s, 1.01 H), 1.08 (d, J ) 6.3, 2.77
H), 0.82 (d, J ) 6.7, 2.77 H).
Ethyl trans-3-Methyl-2,2-diphenylcyclopropanecarboxylate. (0.77
g, 24%): 1H NMR (CDCl3): δ 7.35-7.09 (m, 10.04 H), 3.93-3.85
(m, 1.93 H), 2.43 (m, 0.97 H), 2.28 (d, J ) 5.6 Hz, 0.97 H), 1.01-
0.96 (m, 6.09 H).
cis-3-Methyl-2,2-diphenylcyclopropanemethanol. A solution of
ethyl cis-3-methyl-2,2-diphenylcyclopropanecarboxylate (2.27 g, 8.08
mmol) in tetrahydrofuran (3 mL) was added dropwise over 1 h to a
stirred suspension of lithium aluminum hydride (271 mg, 7.14 mmol)
and tetrahydrofuran (30 mL). After refluxing for 2.5 h, water (5 mL)
was carefully added. The mixture was transferred into a separatory
funnel along with diethyl ether (50 mL) and a 70 mL of a 5% HCl
solution. The layers were separated and the aqueous phase was
extracted with ether (3 × 70 mL). The combined organic layer was
washed with water, dried, concentrated, and distilled (140-145 °C,
0.03 mm Hg) to give a light yellow oil (1.76 g, 92%). 1H NMR
(CDCl3): δ 7.36-7.06 (m, 9.85 H), 3.92-3.87 (m, 1.03 H), 3.55 (dd,
J ) 10.2, 5.4, 1.03 H), 1.84-1.70 (m, 2.19 H), 1.33 (brs, 0.99 H),
1.16 (d, J ) 6.3, 2.91 H). Anal. Calcd for C17H18O1: C, 85.67; H,
7.61. Found: C, 85.41; H, 7.61.
trans-3-Methyl-2,2-diphenylcyclopropanemethanol. This material
was prepared using the same procedure as above except ethyl trans-
3-methyl-2,2-diphenylcyclopropanecarboxylate (1.98 g, 7.07 mmol) was
used. The resulting product was distilled (140-145 °C, 0.03 mmHg)
to give a clear light yellow oil (1.44 g, 85%). 1H NMR (CDCl3): δ
7.35-7.13 (m, 10.04 H), 3.51 (dd, J ) 11.3, 6.0, 0.97 H), 3.30 (dd, J
) 11.3, 7.9, 0.99 H), 1.74-1.62 (m, 2.03 H), 1.31 (broad s, 1.00 H),
0.93 (d, J ) 5.9, 2.96 H). Anal. Calcd for C17H18O1: C, 85.67; H,
7.61. Found: C, 85.54; H, 7.64.
cis-3-Methyl-2,2-diphenylcyclopropaneacetonitrile (6). p-Tolu-
enesulfonyl chloride (0.778 g, 4.1 mmol) was added in several portions
to a solution of cis-3-methyl-2,2-diphenylcyclopropanemethanol (0.896
g, 3.8 mmol) and pyridine (9.0 mL) which was maintained at -10 °C.
After 15 min, the solution was warmed to 4° for 12 h. The reaction
mixture was then poured into a separatory funnel containing ice-cold
methylene chloride (100 mL). The organic layer was washed succes-
sively with 20 mL portions of cold water, cold 5% HCl, cold 5% sodium
bicarbonate, and cold brine. It was then dried and concentrated to give
0.33 g of a light yellow oil. This material was immediately placed in
a flask containing a solution of tetra-n-butylammonium cyanide (3.66
g, 13.6 mmol) in DMSO (15 mL). After 2 h the solution was poured
into a separatory funnel with pentane (100 mL) and water (50 mL).
(2R*,3R*)-3-Methoxy-2-methyl-1,1-diphenylbutane (7a′). (2R*,
3R*)-3-Methyl-4,4-diphenyl-2-butanol (406 mg, 1.7 mmol) in tetrahy-
drofuran (2 mL) was added dropwise over 20 min by means of a syringe
pump to a mixture of sodium hydride (51 mg, 2.1 mmol), methyl iodide
(156 µL, 355 mg, 2.5 mmol), and tetrahydrofuran (6 mL) maintained
at 46 °C. After 30 min, additional methyl iodide (50 µL, 114 mg, 0.8
mmol) was added. After 2.5 h, the reaction mixture was cooled to
room temperature, and several drops of water were carefully added.
Extractive workup followed by column chromatography (95:5 hexane:
ethyl acetate) gave a clear colorless oil which formed a white solid on
standing (345 mg, 80%). 1H NMR (CDCl3): δ 7.34-7.11 (m, 10.33
H), 3.56 (d, J ) 11.4, 0.98 H), 3.25 (s, 2.86 H) 3.24-3.16 (m, 0.97
H), 2.88-2.76 (m, 1.01 H), 1.01 (d, J ) 6.3, 2.90 H), 0.78 (d, J ) 6.7,
2.95 H). Anal. Calcd for C18H22O: C, 84.99; H, 8.72. Found: C,
84.78; H, 8.68.
(29) Simes, B. E.; Rickborn, B.; Fluornoy, J. M. Berlman, I. B. J. Org.
Chem. 1988, 53, 4613.
(30) Johnson, S. A.; Hunt, H. R.; Neumann, H. M. Inorg. Chem. 1963,
2, 960.
(31) (a) Miller, J. B. J. Org. Chem. 1959, 24, 560. (b) Smith, L. I.;
Howard, K. L. Organic Synthesis Wiley: New York, 1959; Collect Vol. 3,
p 351.
(32) Walborsky, H. M.; Murari, M. P. Can. J. Chem. 1984, 62, 2464.
(33) (a) Pretch, E.; Clerc, T.; Seibl, J.; Simon, W. Tables of Spectral
Data for Structure Determination of Organic Compounds; Springer-
Verlag: Berlin, 1983; p I-190. (b) Socrates, G. Infrared Characteristic
Group Frequencies, Wiley-Interscience: Chichester, 1980; p 55. (c) Meyers,
A. I.; Ritter, J. J. J. Org. Chem. 1958, 23, 1918.
(34) MOPAC93; Dr. J. J. P. Stewart & Fujitsu Ltd., Tokyo, 1993.
(35) (a) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.;
Johnson, B. G.; Robb, M. A.; Cheeseman, J. R.; Keith, T. A.; Petterson, G.
A.; Montgomery, J. A.; Raghavari, K.; Al-Laham, M. A.; Zakrzewski, V.
G.; Ortiz, J. V.; Foresman, J. B.; Cioslowski, J.; Stefanov, B. B.;
Nanayakkara, A.; Challacombe, M.; Peng, C. Y.; Ayala, P. Y.; Chen, W.;
Wong, M. W.; Andres, J. L.; Repogle, E. S.; Gomperts, R.; Martin, R. L.;
Fox, D. J.; Binkley, J. S.; Defrees, D. J.; Baker, J.; Stewart, J. J. P.; Pople,
J. A. Gaussian 94, Versions C.3 & D.1; Gaussian, Inc.: Pittsburgh, PA,
1995.
(2S*,3R*)-3-Methoxy-2-methyl-1,1-diphenylbutane (6a′). This
material was prepared using the same procedure as above except
(2S*,3R*)-3-methyl-4,4-diphenyl-2-butanol was used. Chromatography
(98:2 hexane:ethyl acetate) gave a clear colorless oil which formed a
white solid on standing in the freezer (60%). 1H NMR (CDCl3): δ
7.33-7.10 (m, 10.43 H), 3.94 (d, J ) 11.5, 0.91 H), 3.14-3.07 (m,
3.77 H) 2.30-2.21 (m, 0.93 H), 1.09 (d, J ) 6.3, 3.01 H), 0.83 (d, J
) 6.8, 2.95 H). Anal. Calcd for C18H22O: C, 84.99; H, 8.72. Found:
C, 85.02; H, 8.71.
Reduction of Products from the 1-Cyanonaphthalene-Sensitized
Photooxidations of cis-3-Methyl-2,2-diphenylcyclopropaneacetoni-
trile (6) and trans-3-Methyl-2,2-diphenylcyclopropaneacetonitrile
(7) in Methanol. Reductive decyanation9 of the product obtained from
the photooxidation of cis-3-methyl-2,2-diphenylcyclopropaneacetonitrile
gave a product which showed an identical GC rentention time and 1H
NMR spectrum to independently synthesized 6a′, prepared as described

SN2 Reactions of Arylcyclopropane Cation Radicals. 2
J. Am. Chem. Soc., Vol. 119, No. 5, 1997 1003
above. Analogously, application of the reaction to the product obtained
from photooxidation of trans-3-methyl-2,2-diphenylcyclopropane-
acetonitrile gave a product which showed an identical GC rentention
time and 1H NMR spectrum to independently synthesized 7a′, prepared
as described above.
3-Methoxy-3-methyl-1-phenylbutane (8a). A solution of R,R-
dimethylbenzenepropanol (0.289 g, 1.76 mmol) in tetrahydrofuran (2
mL) was added over 5 min to a suspension of sodium hydride (54.5
mg, 2.27 mmol), methyl iodide (160 µL, 2.57 mmol), and tetrahydro-
furan (6 mL) maintained at 40 °C. After 30 min, additional methyl
iodide (100 µL, 1.61 mmol) was added. After 1.5 h, extractive workup
gave a light yellow oil which, after chromatography (90:10 hexane:
ethyl acetate), gave a colorless liquid (0.285 g, 94% yield). 1H NMR
(CDCl3): δ 7.30-7.20 (m, 4.59 H), 3.26 (s, 2.87 H), 2.69-2.64 (m,
2.03 H), 1.83-1.77 (m, 2.01 H), 1.24 (s, 6.50 H).
tive reactions which contained no internal standard. 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
and irradiated with nominal 300 nm light bulbs. The reaction mixture
was subsequently analyzed by GC, in which the retention times were
compared with those of independently synthesized materials where
available. All reported yields are based on GC analysis of the analytical
reactions which were run in duplicate. A portion (typically 3-6 mg)
of the major products were isolated by preparative GC (column: 6′ ×
3/8′′ 17% XF-1150 on Anachrom Q, except when noted otherwise) and
1
analyzed by H NMR, GC, and other techniques as described.
Photooxidation of cis-3-Methyl-2,2-diphenylcyclopropaneaceto-
nitrile (6) in Methanol. A degassed solution of 6 (154 mg, 623 µmol)
and 1-cyanonaphthalene (67 mg, 438 µmol) in methanol (12.0 mL)
was irradiated for 73 h. Concentration followed by column chroma-
tography (85:15 hexane:ethyl acetate) gave a light yellow oil (145 mg,
83%), 6a. 1H NMR (CDCl3): δ 7.38-7.16 (m, 10.05 H), 4.12 (d, J )
11.8, 1.00 H), 3.27-3.18 (m, 1.03 H), 3.17 (s, 2.94 H), 2.64-2.57 (m,
0.95 H), 2.46 (dd, J ) 17.1, 7.2, 1.00 H), 2.30 (dd, J ) 17.4, 2.8, 1.01
H), 1.23 (d, J ) 6.2, 3.03 H). 13C NMR (CDCl3): δ 141.66, 141.57,
128.77, 128.67, 127.57, 127.28, 126.79, 126.54, 119.03, 75.65, 55.82,
52.74, 41.51, 14.80, 13.18. Anal. Calcd for C19H21N1O1: C, 81.68;
H, 7.58. Found: C, 81.53; H, 7.37.
Photooxidation of trans-3-Methyl-2,2-diphenylcyclopropane-
acetonitrile (7) in Methanol. The same procedure as above was
followed except 7 was used. Column chromatography (85:15 hexane:
ethyl acetate) gave a light yellow oil (80%), 7a. 1H NMR (CDCl3): δ
7.39-7.16 (m, 10.29 H), 3.90 (d, J ) 12.3, 0.97 H), 3.40-3.33 (m,
0.98 H), 3.26 (s, 2.86 H), 3.06-2.98 (m, 1.01 H), 2.52 (dd, J ) 17.1,
3.1, 1.02 H), 2.21 (dd, J ) 17.1, 6.6, 1.01 H), 1.15 (d, J ) 6.5, 2.86
H). 13C NMR (CDCl3): δ 142.84, 142.35, 129.08, 128.80, 128.03,
127.95, 126.95, 126.67, 119.97, 74.55, 56.41, 53.88, 45.65, 17.27, 14.89.
Anal. Calcd for C19H21N1O1: C, 81.68; H, 7.58. Found: C, 81.32;
H, 7.54.
Photooxidation of 1,1-Dimethyl-2-phenylcyclopropane (8) in
Methanol. A degassed solution of 8 (25.0 mg, 0.163 mmol)6 and
1-cyanonaphthalene (4.3 mg, 28.1 µmol) in methanol (50 µL) was
irradiated for 1.25 h. GC analysis showed the retention time of the
major product (99%) to be identical to that of independently synthesized
3-methoxy-3-methyl-1-phenylbutane (8a). A portion of this product
was isolated by column chromatography (90:10 hexane:ethyl acetate)
and gave a colorless liquid whose 1H NMR was identical to that of 8a.
1H NMR (CDCl3): δ 7.30-7.20 (m, 5.07 H), 3.26 (s, 3.04 H), 2.69-
2.64 (m, 1.97 H), 1.83-1.77 (m, 2.00 H), 1.24 (s, 5.92 H).
Photooxidation of 2-Methyl-1,1-diphenylcyclopropane (9) in
Methanol. Analytical scale: A degassed solution of 9 (10 mg, 48
µmol),6 1-cyanonaphthalene (2.6 mg, 17 µmol), and tetraethylene glycol
dimethyl ether (15 µL, internal standard) in methanol (600 µL) was
irradiated for 28 h. GC analysis showed 99% conversion and formation
of one product (86%; 9a). Preparative scale: A degassed solution of
9 (125 mg, 600 µmol) and 1-cyanonaphthalene (46.5 mg, 310 µmol)
in methanol (9 mL) was irradiated for 39 h. GC analysis showed
formation of 83% of product and 13% remaining cyclopropane.
Purification by column chromatography (neutral alumina; 95:5 hexane:
ether) gave 29 mg (22% yield) of oil which was spectrally identical to
a sample of 3-methoxy-1,1-diphenylbutane (9a).32
2-tert-Butyl-1,1-diphenylcyclopropane (12). Diphenyldiazomethane31
(0.978 g, 5.03 mmol) was reacted with 3,3-dimethyl-1-butene (26.12
g, 310 mmol) following a literature procedure.6 Removal of the
volatiles followed by extraction with pentane gave an orange oil which
after vacuum distillation (0.01 mmHg, 78-82 °C) afforded 0.666 g
(53%) of a clear oil. 1H NMR (CDCl3): δ 7.16-7.51 (m, 10.23 H),
1.68 (m, 0.93 H), 1.58 (m, 0.98 H), 1.11 (dd, J ) 9.4, 4.6, 0.98 H),
0.87 (s, 8.88 H). Anal. Calcd for C19H22: C, 91.14; H, 8.86. Found:
C, 91.30; H, 8.86.
3-Methoxy-2,2-dimethyl-1,1-diphenylpropane (14). Lithium (16.5
mg, 2.38 mmol) was added to a solution of 4,4′-di-tert-butylbiphenyl
(681 mg, 2.56 mmol) in THF (6 mL) at 0 °C. After 3 h, the solution
was cooled to -78 °C, and 3,3-dimethyl-2,2-diphenyloxetane10 (0.250
g, 1.05 mmol) in THF (650 µL) was added dropwise. After 3 h, the
solution was warmed to room temperature. After 10 h, extractive work-
up followed by chromatography (90:10 hexane:ethyl acetate) gave a
light yellow oil shown to be 2,2-dimethyl-3,3-diphenyl-1-propanol
(0.182 g, 69%). 1H NMR (CDCl3): δ 7.47 (d, J ) 7.4, 4 H), 7.16-
7.33 (m, 6 H), 4.07 (s, 1 H), 3.35 (d, J ) 6.0, 2 H), 1.38 (t, J ) 6.0,
0.9 H), 1.03 (s, 6.1 H). A solution of the alcohol (59 mg, 0.25 mmol)
in THF (0.31 mL) was added dropwise over 25 min to a solution of
NaH (0.100 g, 4.2 mmol) and iodomethane (57 mg, 0.42 mmol) in
THF (800 µL) maintained at 45 °C. Extractive work-up followed by
column chromatography (90:10 hexane:ethyl acetate) gave 12 mg of
the title compound 14. 1H NMR (CDCl3): δ 7.47 (d, J ) 7.4, 4H),
7.17-7.35 (m, 6H), 4.17 (s, 1H), 3.27 (s, 2.8 H), 2.95 (s, 1.8H), 1.04
(s, 1H). HRMS (CI) calcd for C18H23O (M + 1): 255.1748. Found:
255.1748.
2-tert-Butyl-3-methoxy-1,1-diphenylpropane (15). Reaction of
benzophenone with 3,3-dimethyl-1-butene following a literature pro-
cedure10 gave 3-tert-butyl-2,2-diphenyloxetane. 1H NMR (CDCl3): δ
7.17-7.58 (m, 10H), 4.65 (dd, J ) 9, 8, 1 H), 4.54 (dd, J ) 9, 8, 1 H),
3.61 (t, J ) 9, 1 H), 0.72 (s, 9 H). Reduction with lithium as described
above gave 2-tert-butyl-3,3-diphenyl-1-propanol. 1H NMR (CDCl3):
δ 7.16-7.41 (m, 10 H), 4.20 (d, J ) 9, 0.9 H), 3.60-3.75 (m, 1.9 H),
2.31-2.38 (m, 0.9 H), 0.92 (s, 9.3 H). Methylation as described above
gave the title compound 15 (7%). 1H NMR (CDCl3): δ 7.07-7.44
(m, 10 H), 4.19 (d, J ) 9, 1 H), 3.14-3.30 (m, 2 H), 2.84 (s, 3 H),
2.26-2.35 (m, 1 H), 0.88 (s, 9 H). HRMS (CI) calcd for C20H27O (M
+ 1): 283.2061. Found: 283.2063.
Trans- and cis-1-methyl-2-phenylcyclopropane (17) and (18) were
synthesized according to the procedure of Casey et al.4a Distillation
of the crude reaction mixture (3.37 g; 75%) afforded a mixture of the
title compounds 17 and 18 as a colorless oil (1.18 g, 26%). The isomers
Photooxidation of 2-Ethyl-1,1-diphenylcyclopropane (10) in
Methanol. Analytical scale: Same conditions as for 9 except irradia-
tion time was 16 h. GC analysis showed >99% conversion and
formation of one product (87%; 10a). Preparative scale: Same
conditions as for 9. GC analysis showed formation of 82% product
and 14% remaining cyclopropane. Column chromatography as de-
scribed for 9, gave 79 mg (58%) of pure material, which was
characterized to be 3-methoxy-1,1-diphenylpentane (10a). 1H NMR
(CDCl3): δ 7.17-7.35 (m, 9.85 H), 4.23 (t, J ) 7.9, 0.98 H), 3.27 (s,
3.00 H), 2.99 (m, 0.98 H), 2.21 (dd, J ) 8.1, 6.0, 2.03 H), 1.57 (dq, J
) 7.4, 5.2, 2.03 H), 0.91 (t, J ) 7.4, 3.14 H). HRMS (EI) Calcd for
C18H22O (M+): 254.1671. Found: 254.1671.
1
were separated by preparative GC (10′ × /4′′ Apiezon L + 2% KOH
on Chromosorb W-AW 80/100).
trans-1-Methyl-2-phenylcyclopropane (17). 1H NMR (CDCl3): δ
7.25-7.03 (m, 5.34 H), 1.61-1.55 (m, 0.92 H), 1.20 (d, J ) 5.8, 2.92
H), 1.11-1.02 (m, 0.92 H), 0.92-0.86 (m, 0.95 H), 0.78-0.72 (m,
0.95 H).
cis-1-Methyl-2-phenylcyclopropane (18). 1H NMR (CDCl3): δ
7.31-7.16 (m, 5.07 H), 2.13-2.06 (m, 0.96 H), 1.20-1.11 (m, 0.95
H), 1.03-0.95 (m, 1.03 H), 0.81 (d, J ) 6.2, 2.95 H), 0.62-0.57 (m,
0.99 H).
Photooxidations. General Procedure. A Pyrex photolysis vessel
was charged with a solution containing the cyclopropane, 1-cyano-
naphthalene, and internal standard in methanol, except for the prepara-
Photooxidation of 2-Isopropyl-1,1-diphenylcyclopropane (11) in
Methanol. Analytical scale: Same conditions as for 10. GC analysis
showed >99% conversion and formation of one product (85%; 11a).
Preparative scale: A degassed solution of 11 (49.1 mg, 208 µmol) and

1004 J. Am. Chem. Soc., Vol. 119, No. 5, 1997
Dinnocenzo et al.
1-cyanonaphthalene (8.0 mg, 52 µmol) in methanol (1 mL) was
irradiated for 12 h. GC analysis showed formation of 82% product
and 14% remaining cyclopropane. Column chromatography as de-
scribed for 9 gave 29 mg (31%) pure material, which was characterized
to be 3-methoxy-4-methyl-1,1-diphenylpentane (11a). 1H NMR
(CDCl3): δ 7.14-7.34 (m, 9.86 H), 4.22 (dd, J ) 10.9, 4.9, 0.94 H),
3.26 (s, 2.82 H), 2.82 (m, 1.00 H), 2.20 (m, 1.06 H), 2.04 (ddd, J )
13.8, 8.2, 4.9), 1.92 (m, 2.05 H including δ 2.04 peak), 0.88 (2 d, J )
6.7, 6.7, 6.3 H).
Photooxidation of 1,1-Dimethyl-2,2-diphenylcyclopropane (13)
in Acetonitrile. Analytical scale: Same conditions as for 13. GC
analysis showed 100% conversion and formation of one product, 13e
(94%). Preparative scale: A degassed solution of 13 (0.226 g, 1.02
mmol) and 1-cyanonaphthalene (63 mg, 23 µmol) in acetonitrile (15
mL) was irradiated for 83 h. Purification of the crude reaction product
by flash column chromatography (80:20 hexane:benzene) gave 124 mg
of 13e. 1H NMR (CDCl3): δ 7.18-7.34 (m, 10.28 H), 2.61 (s, 2.00
H), 1.85 (s, 2.88 H), 1.28 (s, 5.84 H). IR (CDCl3): 1647 cm-1 (CdN
stretch).33 HRMS (EI) Calcd for C19H21N (M+): 263.1674. Found
263.1635. Calcd for C18H18N (M+ - 15): 248.1439. Found: 248.1471.
Kinetic Isotope Effects. Phenylcyclopropane (0.20 g), 1-cyano-
naphthalene (0.04 g), n-butanol (1.00 g) and 1,1,2,2,3,3,4,4,4-nona-
deuteriobutanol (1.00 g) were dissolved in acetonitrile to make up 50.0
mL solution. Aliquots (5.0 mL) were added to Pyrex tubes which were
purged with argon, sealed, and irradiated in a Rayonet photoreactor at
300 nm for 2 h at 20 °C. Samples of the unirradiated and irradiated
reaction solutions were analyzed by GC. The unirradiated samples
provided the ratio of isotopologic butanol and the irradiated samples
the ratios of the isotopologic products. All experiments were performed
in triplicate, and all samples were analyzed at least five times by GC.
Control experiments showed that the product isotope ratio did not vary
within the reaction time. To determine the isotope effect on the GC
detection, both isotopologic products (Ph(CH2)3OC4H9 and Ph(CH2)3-
OC4D9) were synthesized independently. Phenylcyclopropane (0.30
g), 1-cyanonaphthalene (0.10 g), and n-butanol (1.00 g) were dissolved
in acetonitrile to make up a 7 mL solution. This was irradiated for 21
h under argon (GC: 96% yield of Ph(CH2)3OC4H9). The product was
isolated by preparative TLC (90:10 hexane:ethyl acetate) to give 0.189
g (39%) of Ph(CH2)3OC4H9. 1H NMR (CDCl3): δ 7.28-7.43 (m, 5
H), 3.51 (t, J ) 6.0, 2H+2H), 2.80 (t, J ) 8.0, 2 H), 1.98-2.03 (m, 2
H), 1.66-1.71 (m, 2 H), 1.49-1.54 (m, 2 H), 1.05 (t, J ) 8.0, 3 H).
13C NMR (CDCl3): δ 142.55, 128.74, 126.18, 71.14, 70.33, 32.88,
32.42, 31.86, 19.92, 14.43. Analogously, an independent experiment
with HOC4D9 gave Ph(CH2)3OC4D9 (19%). 1H NMR (CDCl3): δ
7.28-7.43 (m, 5 H), 3.48 (t, J ) 6.0, 2 H), 2.76 (t, J ) 8.0, 2 H),
1.92-2.01 (m, 2 H).
Photooxidation of 2-tert-Butyl-1,1-diphenylcyclopropane (12) in
Methanol. Analytical scale: Same conditions as for 9 except irradia-
tion time was 40 h. GC analysis showed 86% conversion and formation
of four products: 12a (60%), 12b (6.1%), 12c (3.6%) and 12d (3.7%).
GC analysis showed that relative to 12a, <0.1% of the independently
synthesized 15 was present in the crude reaction mixture. Preparative
scale: A degassed solution of 12 (1.80 g, 7.19 mmol) and 1-cyano-
naphthalene (440 mg, 2.87 mmol) in methanol (50 mL) was irradiated
for 91 h. The crude material was purified by flash column chroma-
tography (95:5 hexane:ether) to give 80 mg of pure 12a. Purification
of 12d was achieved by using preparative GC (10% OV-1 on
Chromosorb W/AW). Compounds 12b and 12c were isolated from
the reaction in acetonitrile (Vide infra).
12a: 1H NMR (CDCl3): δ 7.15-7.35 (m, 9.88 H), 4.24 (dd, J )
11.8, 4.1, 0.99 H), 3.35 (s, 2.90 H), 2.68 (dd, J ) 10.2, 1.4, 0.99 H),
2.37 (ddd, J ) 14.0, 11.6, 1.4, 0.99 H), 1.99 (ddd, J ) 14.4, 10.3, 3.9,
0.99 H), 0.90 (s, 9.26 H).
12d: 1H NMR (CDCl3): δ 7.13-7.70 (m, 10.83 H), 4.00 (dd, J )
12.1, 3.4, 1.00 H), 2.96 (s, 2.71 H), 2.44 (t, J ) 11.8, 1.00 H), 1.59
(m, 2.07 H), 1.08 (2 s, 5.56 H), 0.93 (d, J ) 6.7, 2.85 H).
Photooxidation of 2-tert-Butyl-1,1-diphenylcyclopropane (12) in
Acetonitrile. Analytical scale: Same conditions as for 12. GC analysis
showed 100% conversion and formation of three products: 12b (25%),
12c (47%), and 12e (7%). Preparative scale: A degassed solution of
12 (253 mg, 1.02 mmol) and 1-cyanonaphthalene (39.4 mg, 260 µmol)
in acetonitrile (4.5 mL) was irradiated for 37 h. The products were
isolated by flash chromatography followed by preparative GC (6′ ×
1/4′′10% Apiezon L + 2% KOH on Chromosorb W/AW 80/100).
12b: 1H NMR (CDCl3): δ 7.46 (d, J ) 7.7, 0.95 H), 7.18-7.29
(m, 4.43 H), 7.02-7.09 (m, 2.37 H), 6.89 (d, J ) 7.7, 0.95 H), 4.22 (t,
J ) 5.4, 0.95 H), 2.00-2.06 (m, 1.03 H), 1.55-1.91 (m, 2.06 H), 1.40
(s, 3.09 H), 1.22 (s, 3.17 H), 0.94 (d, J ) 6.8, 3.01 H). HRMS (EI)
Calcd for C19H22 (M+): 250.1722. Found, 250.1710.
The isotope effects on the response factors (signal H/signal D) were
subsequently determined to be 1.0124 ( 0.0024 (butanols) and 1.0135
( 0.0024 (products). Since their ratio equals 1 within experimental
error, no additional correction factor was applied. On the basis of this,
no correction factor was applied in the case of 15 also.
Computational Methods. Semiempirical calculations (ROHF) were
performed with MOPAC9334 using the AM1 and PM3 parameters
implemented therein. Transition states were located (with keyword
TS) using the highest point of a reaction coordinate calculations
(variation of C-O with 0.1 Å, from 3.0 to 1.6 Å) to obtain a reasonable
starting geometry for the transition state optimization. B3LYP calcula-
tions were performed with Gaussian 94, using the 6-31G* basis set
implemented therein.35
12c: 1H NMR (CDCl3): δ 7.44 (d, J ) 7.9, 0.95 H), 7.14-7.33
(m, 5.67 H), 6.97 (t, J ) 7.4, 0.95 H), 6.73 (d, J ) 7.8, 0.87 H), 4.08
(dd, J ) 10.9, 6.4, 0.95 H), 1.78-1.92 (m, 3.15 H), 1.40 (s, 3.15 H),
1.23 (s, 3.23 H), 0.94 (d, J ) 6.3, 3.08 H).
12e: 1H NMR (CDCl3): δ 7.15-7.38 (m, 9.67 H), 3.60 (m, 1.09
H), 2.54 (dd, J ) 13.2, 6.4, 1.09 H), 2.38 (dd, J ) 13.2, 9.9, 1.09 H),
1.92 (d, J ) 2.1, 2.92 H), 1.02 (s, 9.12 H). Decoupling at δ 1.9
simplifies δ 3.6 to a dd with J ) 9.8 and 6.4 Hz; these couplings
correspond to those between the methine hydrogen and the two
diastereotopic methylene hydrogens. IR (CDCl3): 1643 cm-1 (CdN
stretch).33 HRMS (EI) Calcd for C21H25N (M+): 291.1987. Found,
291.1963.
Acknowledgment. Research support was provided by the
National Science Foundation (CHE-9312460) and by the donors
of the Petroleum Research Foundation, 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 Netherlands Organization for Scientific Research (NWO)
for a postdoctoral Talent-stipendium.
Photooxidation of 1,1-Diphenylcyclopropane (13) in Methanol.
Analytical scale: Same conditions as for 9 except irradiation time was
27 h. GC analysis showed 100% conversion and formation of one
product, 13a (93%). GC analysis showed that relative to 13a, <0.01%
of independently synthesized 14 was present in the crude reaction
mixture. Preparative scale: A degassed solution of 12 (20.6 mg, 92.8
µmol) and 1-cyanonaphthalene (3.5 mg, 23 µmol) in methanol (750
µL) was irradiated for 13 h. Purification of the crude by preparative
GC (6′ × 1/4′′ 10% Apiezon L + 2% KOH on Chromosorb W/AW
80/100) gave 13a. 1H NMR (CDCl3): δ 7.14-7.37 (m, 10.56 H), 4.17
(t, J ) 6.3, 1.04 H), 3.12 (s, 2.88 H), 2.36 (d, J ) 6.4, 1.92 H), 1.05
(s, 5.60 H).
Supporting Information Available: Gaussian 94 archive
files are available for the B3LYP/6-31G* optimized minima (2
pages). See current masthead page for ordering and Internet
access instructions.
JA963112S