Rhodium-catalyzed enantioselective cyclopropanation of electron-deficient alkenes

Article abstract of DOI:10.1039/c3sc50425e

The rhodium-catalyzed reaction of electron-deficient alkenes with substituted aryldiazoacetates and vinyldiazoacetates results in highly stereoselective cyclopropanations. With adamantylglycine derived catalyst Rh₂(S-TCPTAD)₄, high asymmetric induction (up to 98% ee) can be obtained with a range of substrates. Computational studies suggest that the reaction is facilitated by weak interaction between the carbenoid and the substrate carbonyl but subsequently proceeds via different pathways depending on the nature of the carbonyl. Acrylates and acrylamides result in the formation of cyclopropanation products while the use of unsaturated aldehydes and ketones results in the formation of epoxides.

Full text of DOI:10.1039/c3sc50425e

View Article Online
View Journal
Chemical Science
EDGE ARTICLE
Rhodium-catalyzed enantioselective cyclopropanation
of electron-deficient alkenes†
Cite this: DOI: 10.1039/c3sc50425e
a
a
b
b
*
Hengbin Wang, David M. Guptill, Adrian Varela-Alvarez, Djamaladdin G. Musaev
a
*
and Huw M. L. Davies
The rhodium-catalyzed reaction of electron-deficient alkenes with substituted aryldiazoacetates and
vinyldiazoacetates results in highly stereoselective cyclopropanations. With adamantylglycine derived
catalyst Rh2(S-TCPTAD)4, high asymmetric induction (up to 98% ee) can be obtained with a range of
substrates. Computational studies suggest that the reaction is facilitated by weak interaction between
the carbenoid and the substrate carbonyl but subsequently proceeds via different pathways depending
on the nature of the carbonyl. Acrylates and acrylamides result in the formation of cyclopropanation
products while the use of unsaturated aldehydes and ketones results in the formation of epoxides.
Received 13th February 2013
Accepted 12th April 2013
DOI: 10.1039/c3sc50425e
www.rsc.org/chemicalscience
direct cyclopropanations of electron-decient alkenes have
been achieved using cobalt porphyrin, palladium acetate, and
Introduction
ruthenium salen as catalysts.9 These reactions, however, are
likely to be mechanistically distinct from the rhodium- and
copper-catalyzed reactions. For example, the cobalt porphyrin-
catalyzed reactions have been shown to proceed via radical
intermediates9a and the palladium-catalyzed reaction has been
proposed to be a [2 + 2] cycloaddition followed by a reductive
elimination.9d On the other hand, cyclopropanation reactions of
electron-decient alkenes using rhodium-catalyzed decompo-
sition of diazo compounds are scarce.10 In this paper, we
describe the rst highly enantioselective cyclopropanation of
acrylate derivatives with rhodium carbene intermediates
derived from aryl and vinyldiazoacetates. The scope of this
reaction is described and a computational analysis is presented
to explain why highly electrophilic carbenoids are able to
cyclopropanate electron-decient alkenes.
Cyclopropanes are a common subunit of natural products and
bioactive compounds.1 Additionally, they serve as building
blocks in complex molecule synthesis, as the opening of the
highly strained three-membered ring can lead to a number of
synthetically useful intermediates.2,3 One of the most effective
strategies for the synthesis of cyclopropanes is the metal-cata-
lyzed reaction of diazo compounds with alkenes.4 The devel-
opment of a wide variety of chiral catalysts has enabled
cyclopropanation reactions to be highly enantioselective.
Among them, chiral complexes of rhodium (e.g. Fig. 1) and
copper are the most widely used catalysts, but their application
in cyclopropanation reactions has been largely conned to
electron-rich and electron-neutral alkenes.4,5 The metal-cata-
lyzed cyclopropanation of electron-decient alkenes is consid-
ered to be much more challenging due to the electrophilic
nature of metal-bound carbenes.4,6,7 One approach has been to
decompose diazo compounds in the presence of a dialkyl
sulde, which generates a sulfur-ylide capable of cyclo-
propanating electron-decient olens.8 Recently, effective
aDepartment of Chemistry, Emory University, 1515 Dickey Drive, Atlanta, GA 30322,
USA. E-mail: hmdavie@emory.edu; Fax: +1 404-727-7766; Tel: +1 404-727-6839
bCherry L. Emerson Center for Scientic Computation, Emory University, 1521 Dickey
Drive, Atlanta, Georgia, USA. E-mail: dmusaev@emory.edu
† Electronic supplementary information (ESI) available: (1) Experimental data for
the reported reactions and a CIF le for the X-ray crystallographic data for 7l, and
15, (2) completed ref. 20, (3) Fig. SI-3 including calculated structures, important
˚
geometry parameters (in A) and relative energies of the epoxidation of (a)
acrylate, (b) acrylamide and methyl vinyl ketone, and (4) Cartesian coordinates
all reported reactants, intermediates, transition states and products of the
reaction of vinyldiazoacetate with methyl acrylate (R
¼
OMe),
N,N-dimethylacrylamide (R ¼ NMe2), and methyl vinyl ketone (R ¼ Me). CCDC
908862 and 908861. For ESI and crystallographic data in CIF or other electronic
format see DOI: 10.1039/c3sc50425e
Fig. 1 Chiral rhodium catalysts.
This journal is ª The Royal Society of Chemistry 2013
Chem. Sci.

View Article Online
Chemical Science
Edge Article
to 1 equiv. with a slight drop of the reaction yield and enan-
tioselectivity (entry 10).
Results and discussion
Previously, we had observed that additives such as methyl
benzoate can have a benecial effect on rhodium-catalyzed
carbenoid transformations.11 During studies to further explore
the inuence of ester additives, we happened to nd that ethyl
acrylate can react with rhodium carbene intermediates to form
cyclopropane products. Further study showed that the Rh2(S-
DOSP)4-catalyzed reaction of methyl p-tolyldiazoacetate 1 with
ethyl acrylate 2 produced cyclopropane 3 in 59% yield with
>97 : 3 dr and 77% ee (Table 1, entry 1). Intrigued by this result,
we conducted a systematic study with a series of chiral dirho-
dium catalysts (Fig. 1). As shown in Table 1, the optimum
catalyst was found to be Rh2(S-TCPTAD)4. Though Rh2(PTAD)4
(ref. 5e–g) has been used in cyclopropanation reactions, catalyst
Rh2(TCPTAD)4, which we had previously developed for asym-
metric C–H amination reactions,12 has not been applied to
carbenoid reactions. With this catalyst, in reuxing pentane,
the cyclopropane 3 was formed in 71% yield as a single dia-
stereomer and with 84% ee (entry 8). Generally, the dirhodiu-
m(II) tetrachlorophthalimido-protected amino acid catalysts
provided markedly improved levels of enantio-induction over
the phthalimido-protected analogs (see entries 2 vs. 8 and 3 vs.
6). For example, in the presence of Rh2(S-PTTL)4, the cyclopro-
pane 3 was generated with only 27% ee. However, using Rh2(S-
TCPTTL)4 the level of enantioselectivity increased to 74% ee.13
The same trend was observed with the bulkier ada-
mantylglycine-derived catalysts, where we obtained 35% ee with
Rh2(S-PTAD)4, and 84% ee with Rh2(S-TCPTAD)4. In the pres-
ence of Rh2(S-TCPTAD)4, the amount of acrylate can be reduced
The effect of diazo ester size on the enantioselectivity of the
Rh2(S-TCPTAD)4-catalyzed reaction of ethyl acrylate was exam-
ined for phenyldiazoacetates 4a–d (Table 2). The highest level of
enantioselectivity was observed with tert-butyl phenyl-
diazoacetate 4d (entry 4). Under these conditions, the cyclo-
propane 5d was formed in 78% yield and with 91% ee.
The scope of this reaction was investigated with respect to
both the aryldiazoacetates and alkenes (Table 3). Carbenoids
with electron rich aryl groups performed well giving the cyclo-
propane products 7a–d in good yields (61–91%) and with
excellent levels of enantioselectivity (88–94% ee). Both naphthyl
and halo-substituted aryl groups were tolerated giving the
cyclopropanes 7e and 7f in excellent yields and levels of enan-
tioselectivity. The reaction of aryldiazoacetate 6h, containing a
strong electron-withdrawing group, resulted in decreased yield
of the cyclopropane product 7h (22%) relative to the more
electron-rich aryl groups, as well as decreased dr (92 : 8), though
the product was isolated in 91% ee. A variety of acrylate esters,
including the bulky t-Bu and Ph esters gave cyclopropanation
products in good yields (74–91%) and with excellent levels of
enantioselectivity (90–96% ee). Of particular note is that acryl-
amides are also efficiently cyclopropanated to give the products
(7n and 7o) with excellent levels of enantioselectivity (92–94%
ee), albeit in slightly lower yields. Addition of a substituent at
the a position of the acrylate caused a decrease in both dia-
stereo- and enantioselectivity giving cyclopropane 7p in 55%
yield with 84 : 16 dr and 77% ee. The absolute conguration of
product 7l was determined by X-ray crystallography14 and the
structures of other products were assigned by analogy.15
Vinyldiazoacetates were also tested in the Rh2(S-TCPTAD)4-
catalyzed cyclopropanation of acrylates (Table 4). The levels of
enantioselectivity with vinyldiazoacetates were generally higher
(91–98% ee) than those obtained with aryldiazoacetates. The
aryl groups of styryldiazoacetates could be substituted with both
electron rich and electron poor substituents, cyclopropanes
being formed in excellent yields (75–89%) and with excellent
levels of enantioselectivity (95–98% ee). A slight drop of
Table 1 Optimization of the chiral catalyst
Temperature
(ꢀC)
Yieldb
(%)
eec
(%)
Table 2 Effect of ester group
Entry
Catalyst
dra
1
2
3
4
5
6
7
8
Rh2(S-DOSP)4
Rh2(S-PTAD)4
Rh2(S-PPTL)4
Rh2(S-NTTL)4
Rh2(S-BPTV)4
Rh2(S-TCPTTL)4
Rh2(S-TCPTV)4
Rh2(S-TCPTAD)4
Rh2(S-TCPTAD)4
Rh2(S-TCPTAD)4
Rh2(S-TCPTAD)4
Rh2(S-TCPTAD)4
36
36
36
36
36
36
36
36
36
36
23
40
>97 : 3
>97 : 3
>97 : 3
>97 : 3
>97 : 3
>97 : 3
>97 : 3
>97 : 3
>97 : 3
>97 : 3
>97 : 3
92 : 8
59
70
68
69
65
71
62
71
65d
60e
22
81
77
35
27
24
5
74
65
84
79
73
82
71
Entry
R
Product
dra
97 : 3
>97 : 3
>97 : 3
>97 : 3
yieldb (%)
eec (%)
9
1
2
3
4
Me
Et
n-Bu
t-Bu
5a
5b
5c
5d
83
78
84
78
86
85
81
91
10
11
12f
a
Determined by 1H-NMR analysis of the crude reaction mixture.
Isolated yield. Determined by chiral HPLC. 2.0 equiv. alkene
b
c
d
a
Determined by 1H-NMR analysis of the crude reaction mixture.
Isolated yield. c Determined by chiral HPLC.
used. e 1.0 equiv. alkene used. f Dichloromethane was used as solvent.
b
Chem. Sci.
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Edge Article
Chemical Science
Table 3 Scope of aryldiazoacetates and alkenesa,b,c
Table 4 Scope of vinyldiazoacetatesa,b,c
a
dr was determined by 1H-NMR analysis of the crude reaction mixture.
Yields are isolated yield. ee was determined by chiral HPLC.
Reaction was conducted with 0.2 mol% Rh2(S-TCPTAD)4.
b
d
c
with 97% ee (Scheme 2). The formation of 15 is postulated to
begin by reaction of 6f with the nitrile group of 14 to form an
electron-rich oxazole-substituted alkene, which then preferen-
tially reacts with another molecule of 6f to give the observed
cyclopropane product. The absolute conguration of cyclopro-
pane 15 was assigned by X-ray crystallography.14
a
dr was determined by 1H-NMR analysis of the crude reaction mixture.
Yields are isolated yield. ee was determined by chiral HPLC. The
b
c
d
Computational analysis
diazo compound was dissolved in pentane–dichloromethane
(v/v ¼ 10/1).
As rhodium bound carbenes are considered to be highly elec-
trophilic, it was intriguing that the Rh2(TCPTAD)4 catalyzed
cyclopropanation of electron-decient olens worked so effi-
ciently. Palladium- and ruthenium-catalyzed cyclopropanations
have been proposed to occur via a [2 + 2] cycloaddition, and
cobalt-catalyzed cyclopropanations have been demonstrated to
enantioselectivity (91% ee) was observed with o-substituted
styryldiazoacetate 8c. In the case of methoxy-substituted diazo
8b, due to difficulties separating the cyclopropane product 9b
from the rhodium catalyst, the reaction was conducted at
0.2 mol% catalyst loading. Although we have not conducted a
detailed study of the lower limit of catalyst needed for these
reactions, this experiment shows that at least several hundred
catalyst turnovers are feasible in these reactions.
A different outcome was observed when the reaction was
extended to a,b-unsaturated aldehydes and ketones (Scheme 1).
The reactions of acrolein 10 and methyl vinyl ketone 11 with
diazo compound 4d provided epoxides 12 and 13, respectively.
Under these conditions, the formation of cyclopropanes was not
observed. As observed in previous studies of rhodium-catalyzed
epoxidation with diazo compounds,16 there was almost no
asymmetric induction for the Rh2(S-TCPTAD)4-catalyzed epoxi-
dation in either case.
Scheme 1 Reactions of unsaturated aldehydes and ketones.
Under the standard reaction conditions, the reaction of
acrylonitrile 14 with diazo compound 6f generated an inter-
esting oxazole-substituted cyclopropane 15 in 96% yield and Scheme 2 Reaction of acrylonitrile.
This journal is ª The Royal Society of Chemistry 2013
Chem. Sci.

View Article Online
Chemical Science
Edge Article
proceed through a radical addition–substitution pathway.9,17
Cyclopropanation by rhodium carbene intermediates, however,
has been proposed to be a direct cyclopropanation of the alkene
in a concerted asynchronous manner.18 For the rhodium-cata-
lyzed reactions of acrylate derivatives, the highly electron-de-
cient rhodium carbene intermediates would be expected to
react preferentially with the more nucleophilic oxygen of an
unsaturated carbonyl system to form a rhodium bound
carbonyl ylide (Scheme 3).19 Therefore, it became of interest to
determine how and why a system that would be expected to
form a carbonyl ylide actually resulted in a highly enantiose-
lective cyclopropanation.
For this reason we conducted detailed DFT studies20 on three
model systems: (a) vinyldiazoacetate with methyl acrylate, (b)
vinyldiazoacetate with N,N-dimethylacrylamide and (c) vinyl-
diazoacetate with methyl vinyl ketone (Scheme 4).
For each of these systems we investigated the mechanisms of
cyclopropane (CP) formation, ylide (YL) formation and epoxide
(EP) formation of these reactions by using dirhodium tetra-
formate as a model catalyst (Scheme 5). The studies showed that
the reaction of carbenoid CARB with SUB is initiated by weak
interaction between the carbenoid and the substrate carbonyl
(resulting in a weakly-bound pre-reaction complex PC, which is
more stable for the N,N-dimethylacrylamide than methyl acry-
late or methyl vinyl ketone), but proceeds via different pathways
from this point depending on the nature of the unsaturated
carbonyl compound.
Scheme 3 Mechanistic question associated with the cyclopropanation.
Fig. 2 Potential energy surfaces (scaled to DG values) of the cyclopropanation,
ylide formation and epoxidation pathways for the reaction of (a) vinyl-
diazoacetate and methyl acrylate, (b) vinyldiazoacetate with N,N-dimethylacry-
lamide, (c) vinyldiazoacetate with methyl vinyl ketone.
Scheme 4 Model reactions for computational analysis.
Potential energy surfaces (PES) of the cyclopropanation, ylide
formation and epoxidation pathways for the reaction of vinyl-
diazoacetate with (a) methyl acrylate, (b) N,N-dimethylacryla-
mide and (c) methyl vinyl ketone are shown in Fig. 2a–c,
respectively.21 For the methyl acrylate case, complex PC or reac-
tants, i.e. CARB + SUB, does not proceed to form the ylide because
the kinetically and thermodynamically preferred pathway
involves direct reaction of the alkene to form the cyclopropane.
As seen in Fig. 2a, the calculated energy barrier for the cyclo-
propane formation is DH ¼ 5.4/DG ¼ 4.9 (DGsol ¼ 4.4) kcal molꢁ1
smaller than that for the ylide formation. Epoxidation requires
even higher energy barrier and is not feasible.
Scheme 5 Schematic presentation of the cyclopropanation, ylide formation and
epoxidation pathways of the reaction of rhodium carbene intermediates with (a)
methyl acrylate, (b) N,N-dimethylacrylamide and (c) methyl vinyl ketone.
Chem. Sci.
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Edge Article
Chemical Science
Computation predicts a signicantly different mechanism for
the reaction of N,N-dimethylacrylamide compared to that of
methyl acrylate. As seeninFig. 2b, replacement of the OMe group in
methyl acrylate by the NMe2 amido group facilitates ylide formation
by stabilizing the ylide complex (as well as the pre-reaction
complex) and making the energy barrier required for the ylide
formation 0.9/1.2 (1.2) kcal molꢁ1 smaller than that for the
cyclopropanation. Indeed, as seen in Fig. 2a and b, the calculated
energy of the reaction CARB + SUB / YL is ꢁ12.3/3.0 (3.2) and
ꢁ13.5/2.1 (2.2) kcal molꢁ1 for methyl acrylate and N,N-dime-
thylacrylamide, respectively. This trend in the energy is consis-
Scheme 6 Possible alternative cyclopropanation mechanism.
We have previously demonstrated that donor/acceptor car-
benoids are much more stabilized than the conventional
acceptor carbenoid derived from ethyl diazoacetate.18 The
increased donating capabilityofdonor/acceptorcarbenes creates
a stronger carbene–transition metal interaction, thus strength-
ening the rhodium carbene bond. Since, upon ylide formation,
the nascent carbonyl-carbene bond competes with the Rh–car-
bene bond, we may expect more favorable ylide formation for the
conventional acceptor carbenoids than for donor/acceptor car-
benoids. Consequently, we might also expect complementary
reactivity in these systems. To demonstrate this, we studied the
reaction of ethyl diazoacetate 16 with N,N-dimethylacrylamide 17
(Scheme 7). In this case, no cyclopropanation was observed, and
instead ketoester 18 was unexpectedly formed in 39% isolated
yield. The most likelymechanism for the formation of 18 is initial
formation of the ylide 19 and then cyclization to the epoxide 20.
Amineinduced ring openingof the epoxidewouldgenerate a new
ylide 21, which then undergoes a hydride transfer to form 18.
Such a mechanism would suggest that the initial ylide is
continuing through to product and the lack of cyclopropane
formation is suggestive that the initial ylide formation step is not
reversible, unlike the reaction of vinyldiazoacetate with N,N-
dimethylacrylamide. Together with the computational results,
thisexperimentprovidesstrongevidenceforthepresenceofylide
intermediates in these cyclopropanation reactions.
1
1
˚
tent with the calculated C –O bond distances, 1.479 and 1.456 A,
in the corresponding ylide complexes. In the case of N,N-dime-
thylacrylamide, the slight endothermicity of the ylide formation
and the larger energy barrier required for the forward reaction
(i.e. epoxide formation), make the ylide formation a reversible
process. Since the energy barrier for cyclopropanation is only a
few kcal molꢁ1 more than ylide formation, and since both
transformations start from the same reactants (pre-reaction
complex), the reaction of vinyldiazoacetate with N,N-dimethyla-
crylamide ultimately undergoes cyclopropanation.
In the case of the methyl vinyl ketone, the pre-reaction and
ylide complexes are generated in an analogous fashion to N,N-
dimethylacrylamide (Fig. 2c). However, in this case the ylide
undergoes a favorable cyclization to the epoxide. As seen in Fig. 2c,
the calculated energy barrier for epoxide formation from the
ylide complex is 0.9/0.8 (1.5) kcal molꢁ1 smaller than the barrier
for the reverse reaction. The ylide formation is therefore not
reversible as it was for N,N-dimethylacrylamide. Thus, donor/
acceptor carbenoids do not undergo cyclopropanation of methyl
vinyl ketone mainly due to facile ylide formation followed by epox-
idation. One should mention that the calculated C1–O1 bond
˚
distance, 1.421 A, in the ylide complex of methyl vinyl ketone is
˚
0.058 and 0.035 A shorter than those found in ylide complexes of
Thus, the above-presented computational studies illustrate
the subtle difference in the reactions of donor/acceptor carbe-
noids with methyl acrylate, N,N-dimethylacrylamide, and methyl
vinyl ketone. In each case a weakly bound pre-reaction complex is
formed in which the carbonyl interacts with the carbenoid
(which is thermodynamically stable relative to reactants only at
the DH level). However, beyond PC formation, the reactions
diverge (either from this weakly bound pre-reaction complex or
from the reactants themselves). For the acrylate case,
methyl acrylate and N,N-dimethylacrylamide, respectively.
As shown in Fig. 2, as well as data presented in the ESI,† the
cyclopropanation transition state TS_CP for all studied
substrates reveals a concerted-asynchronous process. Natural
bond analyses show that the interaction of the carbonyl ofmethyl
acrylate with the carbenoid (i.e. formation of pre-reaction
complex PC) makes the alkene even more electron decient. The
pre-reaction complex would therefore have to make the carbe-
noid more susceptible for reacting with electron decient
alkenes. Indeed, the calculated natural charges of the C1 atom
are +0.04 and +0.06 e in CARB and PC, respectively, which would
suggest that metal back-bonding is less signicant in the pre-
reaction complex compared to the free carbene. Since the
cyclopropanation of methyl acrylate by vinyldiazoacetate is
found tobe a facile process, this suggests that the {Rh2}–C1 sigma
bond may be involved in a concerted non-synchronous cyclo-
propanation by nucleophilic attack on the alkene (Scheme 6).
This is opposite to the normal cyclopropanation of electron rich
alkenes, in which it is believed the reaction is initiated by an
electrophilic attack by the carbenoid on the alkene.18 More
extensive computational studies will be necessary to understand
fully this unconventional reactivity of Rh–carbene
intermediates.
Scheme 7 Reaction of ethyl diazoacetate with N,N-dimethylacrylamide.
This journal is ª The Royal Society of Chemistry 2013
Chem. Sci.

View Article Online
Chemical Science
Edge Article
¨
cyclopropanation is preferred over ylide formation. With the
acrylamide, the ylide is preferentially formed but the formation
of this intermediate is reversible, and so, a cyclopropane is ulti-
mately formed. However, with methyl vinyl ketone, ylide
formation is not reversible due to the rapid epoxide formation.
The computational studies not only help to rationalize the
rhodium-catalyzed reactions of donor/acceptor carbenoids with
electron-decient alkenes, but also give novel insights into car-
benoid chemistry in general. Methyl benzoate has been shown to
have a dramatic inuence on the turnover capacity of dirhodium
tetracarboxylates.11 From these computational studies, it would
be a reasonable hypothesis that this is due tointeraction between
the carbenoid and the ester carbonyl, which could protect the
rhodium carbene intermediates from self-destruction. Donor/
acceptor carbenoids are capable of a vast array of intermolecular
C–H functionalizations by means of C–H insertion and this can
be conducted in the presence of a range of heteroatoms and
carbonyl functionality.6 The observation that amide ylide
formation is reversible is suggestive that the range of functional
group tolerance may be due to the reversible nature of carbenoid/
ylide formation. A further interesting possibility is that the ester
carbonyl interaction may present the opportunity to use esters as
directing groups in carbenoid C–H functionalization.
103, 1051–1070; (c) J. Salaun and M. S. Baird, Curr. Med.
Chem., 1995, 2, 511–542; (d) W. A. Donaldson, Tetrahedron,
2001, 57, 8589–8627.
2 For reviews: (a) C. A. Carson and M. A. Kerr, Chem. Soc. Rev.,
2009, 38, 3051–3060; (b) M. J. Campbell, J. S. Johnson,
A. T. Parsons, P. D. Pohlhaus and S. D. Sanders, J. Org.
Chem., 2010, 75, 6317–6325; (c) H.-U. Reissig and
R. Zimmer, Chem. Rev., 2003, 103, 1151–1196.
3 For recent examples: (a) Y. Bai, W. Tao, J. Ren and Z. Wang,
Angew. Chem., Int. Ed., 2012, 51, 4112–4116; (b) H. Wang,
J. R. Denton and H. M. L. Davies, Org. Lett., 2011, 13, 4316–
4319; (c) G. Zhang, X. Huang, G. Li and L. Zhang, J. Am.
Chem. Soc., 2008, 130, 1814–1815; (d) J. Zhang and
H.-G. Schmalz, Angew. Chem., Int. Ed., 2006, 45, 6704–6707.
4 For reviews: (a) H. Lebel, J.-F. Marcoux, C. Molinaro and
A. B. Charette, Chem. Rev., 2003, 103, 977–1050; (b)
M. P. Doyle and D. C. Forbes, Chem. Rev., 1998, 98, 911–
935; (c) H. M. L. Davies and E. G. Antoulinakis, Org. React.,
2001, 57, 1–326.
5 For recent examples: (a) D. T. Boruta, O. Dmitrenko,
G. P. A. Yap and J. M. Fox, Chem. Sci., 2012, 3, 1589–1593;
(b) C. Qin, V. Boyarskikh, J. H. Hansen, K. I. Hardcastle,
D. G. Musaev and H. M. L. Davies, J. Am. Chem. Soc., 2011,
133, 19198–19204; (c) D. Marcoux, S. Azzi and
A. B. Charette, J. Am. Chem. Soc., 2009, 131, 6970–6972; (d)
V. N. G. Lindsay, W. Lin and A. B. Charette, J. Am. Chem.
Soc., 2009, 131, 16383–16385; (e) J. R. Denton and
H. M. L. Davies, Org. Lett., 2009, 11, 787–790; (f)
J. R. Denton, K. Cheng and H. M. L. Davies, Chem.
Commun., 2008, 1238–1240; (g) J. R. Denton, D. Sukumaran
and H. M. L. Davies, Org. Lett., 2007, 9, 2625–2628.
6 (a) H. M. L. Davies and D. Morton, Chem. Soc. Rev., 2011, 40,
1857–1869; (b) H. M. L. Davies and J. R. Manning, Nature,
2008, 451, 417–424; (c) H. M. L. Davies and
R. E. J. Beckwith, Chem. Rev., 2003, 103, 2861–2903; (d)
M. P. Doyle, R. L. Dorow and W. H. Tamblyn, J. Org. Chem.,
1982, 47, 4059–4068.
7 For examples of cyclopropanation using organocatalysis, see:
(a) U. Das, Y.-L. Tsai and W. Lin, Org. Biomol. Chem., 2013,
11, 44–47; (b) L. Gao, G.-S. Hwang and D. H. Ryu, J. Am.
Chem. Soc., 2011, 133, 20708–20711; (c) C. C. C. Johansson,
N. Bremeyer, S. V. Ley, D. R. Owen, S. C. Smith and
M. J. Gaunt, Angew. Chem., Int. Ed., 2006, 45, 6024–6028;
(d) R. K. Kunz and D. W. C. MacMillan, J. Am. Chem. Soc.,
2005, 127, 3240–3241; (e) C. D. Papageorgiou, M. A. Cubillo
de Dios, S. V. Ley and M. J. Gaunt, Angew. Chem., Int. Ed.,
2004, 43, 4641–4644; (f) C. D. Papageorgiou, S. V. Ley and
M. J. Gaunt, Angew. Chem., Int. Ed., 2003, 42, 828–831.
8 (a) V. K. Aggarwal, H. W. Smith, R. V. H. Jones and
R. Fieldhouse, Chem. Commun., 1997, 1785–1786; (b)
V. K. Aggarwal, H. W. Smith, G. Hynd, R. V. H. Jones,
R. Fieldhouse and S. E. Spey, J. Chem. Soc., Perkin Trans. 1,
2000, 3267–3276; (c) V. K. Aggarwal, E. Alonso, G. Fang,
M. Ferrara, G. Hynd and M. Porcelloni, Angew. Chem., Int.
Ed., 2001, 40, 1433–1436.
Conclusion
In summary, these studies demonstrate that aryldiazoacetates
and vinyldiazoacetates are capable of undergoing highly enan-
tioselective cyclopropanations with electron-decient alkenes.
The reaction involves initial formation of a weakly bound pre-
reaction complex between the carbene intermediates and the
carbonyl group of the substrate, but the subsequent reaction is
dependent on the nature of the carbonyl group. Acrylates and
acrylamides result in the formation of cyclopropanation prod-
ucts while the use of unsaturated aldehydes and ketones results
in the formation of epoxides. The optimal catalyst for high
asymmetric induction is shown to be Rh2(S-TCPTAD)4, a catalyst
which had been used previously only for C–H amination. Further
studies are in progress to more thoroughly understand the role of
the pre-reaction complex with carbonyls and to determine
whether this interaction can be used to enhance the selectivity of
other types of rhodium-catalyzed carbenoid reactions.
Acknowledgements
The synthetic studies were supported by the National Institute of
Health (GM-099142-01). The computational studies weresupported
by NSF under the CCI Center for Selective C–H Functionalization,
CHE-1205646. We thank Dr John Bacsa, Emory University, for the X-
ray crystallographic analysis. The authors gratefully acknowledge
NSF MRI-R2 grant (CHE-0958205) and the use of the resources
of the Cherry Emerson Center for Scientic Computation.
Notes and references
1 (a) D. Y. K. Chen, R. H. Pouwer and J. A. Richard, Chem. Soc.
Rev., 2012, 41, 4631–4642; (b) J. Pietruszka, Chem. Rev., 2003,
9 (a) W. I. Dzik, X. Xu, X. P. Zhang, J. N. H. Reek and B. Bruin, J.
Am. Chem. Soc., 2010, 132, 10891–10902; (b) M. Basato,
Chem. Sci.
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Edge Article
Chemical Science
C. Tubaro, A. Biffis, M. Bonato, G. Buscemi, F. Lighezzolo, 18 (a) J. Hansen, J. Autschbach and H. M. L. Davies, J. Org.
P. Lunardi, C. Vianini, F. Benetollo and A. Del Zotto,
Chem.–Eur. J., 2009, 15, 1516–1526; (c) S. Zhu, J. A. Perman
and X. P. Zhang, Angew. Chem., Int. Ed., 2008, 47, 8460–
Chem., 2009, 74, 6555–6563; (b) H. M. L. Davies,
P. R. Bruzinski, D. H. Lake, N. Kong and M. J. Fall, J. Am.
Chem. Soc., 1996, 118, 6897–6907.
8463; (d) S. Chen, J. Ma and J. Wang, Tetrahedron Lett., 19 A. Padwa, Chem. Soc. Rev., 2009, 38, 3072–3081.
2008, 49, 6781–6783; (e) Y. Chen, J. V. Ruppel and 20 The calculations were conducted at the M06L/
X. P. Zhang, J. Am. Chem. Soc., 2007, 129, 12074–12075; (f)
J. A. Miller, W. Jin and S. T. Nguyen, Angew. Chem., Int. Ed.,
2002, 41, 2953–2956.
{(SDD+4f)Rh+[6-311+G(d,p)]} level of theory using the
Gaussian_09 program package (M. J. Frisch, et al.,
Gaussian 09, Revision A02, Gaussian, Inc., Wallingford, CT,
2009. For competed reference see ESI†). The optimized
structures of the reactants, intermediates, transition states
and products of this reaction were calculated in the gas
phase without any symmetry constrains. Solvent effects (n-
pentane with 3 ¼ 1.8371 was used as a solvent) were
incorporated by performing single point energy
calculations on the gas-phase optimized geometries using
the PCM polarizable conductor calculation model. For
M06L method see: Y. Zhao and D. G. Truhlar, J. Chem.
Phys. 2006, 125, 194101–194118. For SDD basis sets and
associated effective core potential see: (a) M. Kaupp,
P. V. R. Schleyer, H. Stoll and H. Preuss, J. Chem. Phys.,
10 (a) V. N. G. Lindsay, D. Fiset, P. J. Gritsch, S. Azzi and
A. B. Charette, J. Am. Chem. Soc., 2013, 135, 1463–1470; (b)
S. J. Gharpure, M. K. Shukla and U. Vijayasree, Org. Lett.,
2009, 11, 5466–5469; (c) H. Xu, W. Zhang, D. Shu,
J. B. Werness and W. Tang, Angew. Chem., Int. Ed., 2008,
47, 8933–8936; (d) B. Leroy, Tetrahedron Lett., 2005, 46,
7563–7566.
11 H. M. L. Davies and C. Venkataramani, Org. Lett., 2003, 5,
1403–1406.
12 R. P. Reddy and H. M. L. Davies, Org. Lett., 2006, 8, 5013–
5016.
13 A similar improvement in enantioselectivity for Rh2(S-
TCPTTL)4 over Rh2(S-PTTL)4 has been observed in the
cyclopropanation of electron rich alkenes by acceptor/
acceptor carbenoids. It was reasoned that the halogen-
bonding keeps the catalyst in a rigidied conformation.
See ref. 5d.
14 The crystal structures of 7l and 15 have been deposited at the
Cambridge Crystallographic Data Centre under deposition
numbers CCDC 908862 and 908861 respectively.
15 Similar types of products have been obtained in racemic
¨
1991, 94, 1360–1366; (b) A. Bergner, M. Dolg, W. Kuchle,
H. Stoll and H. Preuss, Mol. Phys., 1993, 80, 1431–1441; (c)
M. Dolg, H. Stoll, H. Preuss and R. M. Pitzer, J. Phys.
Chem., 1993, 97, 5852–5859. See also: Z. Li, V. Boyarskikh,
J. H. Hansen, J. Autschbach, D. G. Musaev and
H. M. L. Davies, J. Am. Chem. Soc, 2012, 134, 15497–15504.
For PCM method see: (d) J. Tomasi and M. Persico, Chem.
Rev., 1994, 94, 2027–2094; (e) R. Cammi and J. Tomasi, J.
Comput. Chem., 1995, 16, 1449–1458.
form
by
the
palladium-catalyzed
reaction
of 21 The calculated energies presented below were referenced to
the reactants, (HCO2)4Rh2C[(CH]CH2)(CO2Me)] (CARB)
plus substrate (SUB), and presented as DH/DG(DGsol),
where DH and DG are the gas-phase enthalpy and Gibbs
free energy, respectively. DGsol is calculated as DGs + [DG ꢁ
DE], where DGs is the PCM calculated free energy in
solution (using n-pentane as solvent), and DE is the gas-
phase electronic energy. Full geometry parameters of all
calculated and reported reactants, intermediates,
transition states and products are given in the ESI.†
aryldiazoacetates with acrylates (ref. 9d). In this original
paper, the relative conguration was incorrectly assigned
as the
catalyzed reactions preferentially form the E isomer (see
ESI† for greater details).
Z isomer. Both the palladium and rhodium-
16 (a) H. M. L. Davies and J. DeMeese, Tetrahedron Lett., 2001,
42, 6803–6805; (b) M. P. Doyle, W. Hu and D. J. Timmons,
Org. Lett., 2001, 3, 933–935.
17 B. F. Straub, J. Am. Chem. Soc., 2002, 124, 14195–14201.
This journal is ª The Royal Society of Chemistry 2013
Chem. Sci.