Competitive cationic pathways and the asymmetric synthesis of aryl-substituted cyclopropanes

Article abstract of DOI:10.1021/ol063026p

Equation presented 1,2-Disubstituted cyclopropanes were synthesized in a nonracemic fashion via activation of the corresponding homoallylic alcohols in excellent yields. A series of substituted phenyl rings showed higher enantiospecificity for the cyclization as the electron-withdrawing ability of the group increased. The results offer strong support for the existence of competing cation mechanisms.

Full text of DOI:10.1021/ol063026p

ORGANIC
LETTERS
2007
Vol. 9, No. 8
1425-1428
Competitive Cationic Pathways and the
Asymmetric Synthesis of
Aryl-Substituted Cyclopropanes
Bruce J. Melancon, Nicholas R. Perl,† and Richard E. Taylor*
Department of Chemistry and Biochemistry, UniVersity of Notre Dame,
251 Nieuwland Science Hall, Notre Dame, Indiana 46556-5670
taylor.61@nd.edu
Received December 14, 2006
ABSTRACT
1,2-Disubstituted cyclopropanes were synthesized in a nonracemic fashion via activation of the corresponding homoallylic alcohols in excellent
yields. A series of substituted phenyl rings showed higher enantiospecificity for the cyclization as the electron-withdrawing ability of the
group increased. The results offer strong support for the existence of competing cation mechanisms.
New methods for the generation of cyclopropane motifs are
ever present in the literature due to interesting structure and
reactivity and their frequent appearance in biologically active
compounds.1 One particular class of reactions, Michael-
initiated ring closures (MIRCs),2 is a highly efficient and
stereoselective route to cyclopropanes. MIRC substrates
include phosphorus, sulfur, and tellurium ylides which
generate cyclopropanes through an anionic process. Con-
versely, the realm of cationic cyclizations is less widely
explored.3 Several years ago, Suzuki and co-workers4
demonstrated that homoallylic triflates can be trapped to
access diastereomerically pure trans-1,2-disubstituted cyclo-
propanes. Contemporaneously, we investigated the use of
allylsilane homoallylic alcohols as a source of vinylcyclo-
propanes through an intermediate â-silylcyclopropylcarbinyl
cation.
We have demonstrated the ability of this method to
produce diastereomerically and enantiomerically pure 1,2-
disubstituted and 1,2,3-trisubstituted cyclopropanes from
readily available homoallylic alcohols in excellent yields.
The high diastereoselectivity of this process is rationalized
by a preference for transition state 1 (TS 1) (Scheme 1).5
Scheme 1. Cationic Cyclopropanation
† Undergraduate research participant, University of Notre Dame.
(1) de Meijere, A.; Editor for special thematic issue on Cyclopropanes
and Related Rings. Chem. ReV. 2003, 103, 931.
(2) For a review of stereoselective cyclopropanation reactions, see: Lebel,
H.; Marcoux, J. F.; Molinaro, C.; Charette, A. B. Chem. ReV. 2003, 103,
977 and references therein.
(3) For a recent review of cation approaches to cyclopropanes, see:
Taylor, R. E.; Engelhardt, F. C.; Schmitt, M. J. Tetrahedron 2003, 59, 5623.
(4) Nagasawa, T.; Handa, Y.; Onoguchi, Y.; Ohba, S.; Suzuki, K. Synlett
1995, 739.
Moreover, our previous work in this area, focusing on alkyl
and heteroatom-substituted alkyl substrates (R ) alkyl),
(5) Taylor, R. E.; Engelhardt, F. C.; Schmitt, M. J.; Yuan, H. J. Am.
Chem. Soc. 2001, 123, 2964.
10.1021/ol063026p CCC: $37.00
© 2007 American Chemical Society
Published on Web 03/16/2007

found that the reaction proceeds with complete stereospeci-
ficity with inversion. In this manuscript, we report an
extension of our methodology to the synthesis of aryl-
substituted cyclopropanes (R ) aryl). In contrast to our
previous observation, the intermediate benzyl sulfonates may
be prone to ionization and, thus, lead to products with
reduced enantiomeric purity.
Table 1. Aryl-Substituted Cyclopropane Formationa
The general synthetic route for the formation of aryl-
substituted cyclopropanes is illustrated by the phenyl-
substituted case in Scheme 2. Benzaldehyde (3d) was first
Scheme 2. General Sequence to Cyclopropylmethyl Alcohols
a Starting er’s for alcohols 5a-g were 95:5 for enantiomers shown.
bStarting er for bromo-substituted aryl alcohol 5e was 98:2. cRatios
determined by NMR integration of the corresponding aldehyde proton in
intermediate 7. The cis product was found to have been formed with a
similar er. dProduct er determinations were made on their corresponding
cyclopropylmethyl alcohols.
subjected to Brown’s asymmetric allylation conditions,6 to
furnish homoallylic alcohol 4d with 95:5 er. Olefin cross
metathesis5,7 with allyltrimethylsilane yielded allylsilane
homoallylic alcohol 5d in 66:34 selectivity for the E
geometry. We have previously demonstrated that olefin
geometry has little effect on the cyclization, and thus the
mixture of olefin isomers was used directly.
The resultant alcohol 5d was subjected to various condi-
tions for the key cyclization step. After optimization, vinyl
cyclopropane 6d was isolated in 85% yield as a 9:1 mixture
of trans/cis isomers. The enantiomeric ratio of 6d was
determined by a two-step conversion to the cyclopropylm-
ethyl alcohol 8d. 19F NMR analysis of the corresponding
Mosher esters of 8d and comparison to the racemic series
were performed as well.8 With regards to the key cyclization,
our standard conditions of triflic anhydride and 2,6-lutidine
provided the activated sulfonate intermediates necessary for
cyclization. However, much to our surprise, vinyl cyclopro-
panes were isolated in reduced yield. Under these conditions,
a significant amount of silyl-protected starting material was
isolated which effectively shut down triflate formation and
prevented cyclization. Thus, it appears that cyclization is
significantly faster than triflation of 5d. However, mesyl
anhydride was an efficient alternative and vinylcyclopropane
6d was isolated in high yields with no observable silylation
byproducts.
The series of aryl-substituted aldehydes was subjected to
the identical reaction sequence to gain insight on the
stereochemical outcome of the reaction. The homoallylic
alcohols 4a-d and 4f,g were synthesized using Brown’s
asymmetric allylation conditions,6 and homoallylic alcohol
4e was prepared via Soderquist’s method with higher
enantioselectivity.9 Homoallylic alcohols 4b and 4c were
synthesized from 4f (see Supporting Information). Alcohols
4a-g were further elaborated via cross metathesis with
allyltrimethylsilane yielding allylsilane homoallylic alcohols
5a-g (75-90%) in good yield and moderate E/Z selectivity.
Regardless of substitution, the cyclization proved to be
extremely efficient. Moreover, the study of aryl-substituted
substrates revealed two unique aspects of reactivity not
previously observed in the aliphatic system.
(6) Racherla, U. S.; Brown, H. C. J. Org. Chem. 1991, 56, 401.
(7) Scholl, M.; Ding, S.; Lee, C. W.; Grubbs, R. H. Org. Lett. 1999, 1,
953.
(8) (a) Dale, J. A.; Mosher, H. S. J. Am. Chem. Soc. 1973, 95, 512. (b)
Sullivan, G. R.; Dale, J. A.; Mosher, H. S. J. Org. Chem. 1973, 38, 2143.
(c) Ohtani, I.; Kusumi, T.; Kashman, Y.; Kakisawa, H. J. Am. Chem. Soc.
1991, 113, 4092.
(9) Burgos, C. H.; Canales, E.; Soderquist, J. A. J. Am. Chem. Soc. 2005,
127, 8044.
1426
Org. Lett., Vol. 9, No. 8, 2007

Previous work with aliphatic-substituted systems, ours as
well as Suzuki’s, reported exclusive formation of trans-
cyclopropanes. In contrast, aryl-substituted cyclization sub-
strates provided a consistent ratio of trans-/cis-cyclopropane
products (9:1). This observation suggests that competing
transition states (TS1 and TS2, Scheme 1) are closer in
energy for aryl-substituted substrates. A number of qualitative
differences could account for the greater stability of TS2 in
aryl-substituted substrates including greater dipole minimiza-
tion or simple steric differences.
was a good correlation between the enantiospecificity of the
cyclization and σ+ values11 for the aryl substituent (Table 2).
As illustrated in Tables 1 and 2, electron-rich aromatic
systems showed a loss of enantioselectivity upon cyclization.
As a synthetic solution to this problem, we envisioned
p-bromo-aryl-substituted cyclopropanes 9 (Scheme 4) as a
Scheme 4. Aryl Bromide Cross Coupling
The enantiomeric purity of the cyclopropane products,
reported in Table 1, illustrated a clear trend between aryl
group and reaction stereospecificity. Although aryl groups
substituted with electron-withdrawing substituents cyclized
with complete stereospecificity (Table 1, entries 6 and 7),
an appreciable loss in enantiomeric purity was observed in
electron-rich systems (Table 1, entries 1-3).
Two well-described mechanistic pathways for cationic
processes are relevant here: neighboring group displacement
(NGD, k∆) and ionization (kc) (Scheme 3). The NGD pathway
a Starting er for 9 was 95:5 for enantiomer shown.
Scheme 3. Competitive Cationic Pathways
common precursor to enantiomerically pure cyclopropanes
with electron-rich aryl substituents. Couplings of aryl
bromides with Grignard reagents or dialkylamines mediated
by nickel or palladium are frequent in the literature, such as
those described by Kumada12 and Hartwig.13
TBS ether 9 was cleanly generated from alcohol 8e.14
Coupling with methylmagnesium bromide in the presence
of Ni(II) salts, followed by deprotection of the TBS group,
afforded alcohol 8a in 95% yield. The enantiomeric ratio of
the product was consistent with the starting cyclopropane 9.
Thus, substitution of aryl bromide 9 circumvents the loss of
enantioselectivity observed in the cyclization of 5a.
A logical extension of this process would allow the
preparation of electron-rich aryl cyclopropanes not likely
accessible through cationic cyclization processes. For ex-
ample, diethylamino-substituted aryl cyclopropane 10 was
procured through a similar two-step sequence from 9. Pd-
(0)-catalyzed aryl amination in the presence of S-(-)-BINAP
cleanly generated the desired coupling product. After TBAF
deprotection and Mosher’s ester analysis, the enantiomeric
ratio of 10 (93:7) was similar to the starting material 9 (95:
5). On the basis of σ+ values for an amino substituent (-1.3),
one would expect essentially complete loss of stereoselec-
tivity and/or an inefficient cyclization.
involves direct displacement of the intermediate sulfonate
by the allylsilane through a stereospecific inversion. Alter-
natively, ionization provides an achiral benzylic cation and
the formation of a racemic product. Thus, relative enantio-
meric purity between the starting material 5a-g and the
cyclopropane products 6a-g provides an estimate of the
relative rates of each mechanistic pathway.10 Moreover, there
Table 2. Competitive Cationic Pathways
+
R group
∆ee
k∆/kc
σp
In summary, we have expanded on our first-generation
methodology for the formation of cyclopropanes through
CH3
-32
-10
-10
-5
-6
0
76:24
89:11
89:11
94:6
93:7
100:0
100:0
-0.31
-
-
0.00
0.23
0.49
0.46
CH2OBn
CH2OAc
H
(10) Detailed kinetic analysis also supports this correlation. Creary, X.;
O’Donnell, B. D.; Vervaeke, M. J. Org. Chem., 2007, 9, in press.
(11) Hansch, C.; Leo, A.; Taft, R. W. Chem. ReV. 1991, 91, 165.
(12) Kumada, M.; Tamao, K.; Sumitani, K. Organic Syntheses; Wiley:
New York, 1988; Coll. Vol. VI, p 407.
Br
CO2Me
a
CF3
0
(13) (a) Hartwig, J. F. Angew. Chem., Int. Ed. 1998, 37, 2046. (b)
Christensen, H.; Kiil, S.; Dam-Johansen, K.; Nielsoen, O.; Sommer, M. B.
Org. Process Res. DeV. 2006, 10, 762.
a σm value used for the m-CF3 substituent.
(14) See Supporting Information for the preparation of 8e.
Org. Lett., Vol. 9, No. 8, 2007
1427

cationic processes. The method is concise and provides
access to vinyl cyclopropanes from readily available, non-
racemic homoallylic alcohols. The stereoselective generation
of aryl-substituted cyclopropanes with good diastereo- and
enantioselectivity was demonstrated for most substrates.
Marked loss of enantioselectivity was observed in the
cyclization of electron-rich aryl-substituted precursors. How-
ever, the same substrates were generated with enhanced
enantioselectivity through the use of transition-metal coupling
methods. This chemistry is amenable to the synthesis of
complex cyclopropane targets including natural products,
natural product analogues, and medicinal agents. Results to
this end will be reported separately.
Acknowledgment. Support of this work by the National
Science Foundation (CHE-0517389) and Merck Research Labs
is gratefully acknowledged. B.J.M. is grateful for support from
the University of Notre Dame through a Wolf Fellowship.
Supporting Information Available: Full experimental
and characterization data for all new compounds. This material
is available free of charge via the Internet at http://pubs.acs.org.
OL063026P
1428
Org. Lett., Vol. 9, No. 8, 2007