Highly enantioselective synthesis of 1,2,3-substituted cyclopropanes by using α-iodo- and α-chloromethylzinc carbenoids

Article abstract of DOI:10.1002/chem.201202528

Herein, we report the enantio- and diastereoselective formation of trans-iodo- and trans-chlorocyclopropanes from α-iodo- and α-chlorozinc carbenoids by using a dioxaborolane-derived chiral ligand. The synthetically useful iodocyclopropane building blocks were derivatized by an electrophilic trapping of the corresponding cyclopropyl lithium species or a Negishi coupling to give access to a variety of enantioenriched 1,2,3-substituted cyclopropanes. The synthetic utility of this method was demonstrated by the formal synthesis of an HIV-1 protease inhibitor. In addition, the related stereoselective bromocyclopropanation was also investigated. New insights about the relative electrophilicity of haloiodomethylzinc carbenoids are also presented. Propping up the enantioselectivity: The enantio- and diastereoselective formation of trans-iodo- and trans-chlorocyclopropanes from α-iodo- and α-chlorozinc carbenoids by using a dioxaborolane-derived chiral ligand is reported. The iodocyclopropanes were derivatized into a variety of 1,2,3-substituted cyclopropanes and new insights into the relative electrophilicity of α-haloiodomethylzinc carbenoids are presented (see scheme). Copyright

Full text of DOI:10.1002/chem.201202528

FULL PAPER
DOI: 10.1002/chem.201202528
Highly Enantioselective Synthesis of 1,2,3-Substituted Cyclopropanes by
Using a-Iodo- and a-Chloromethylzinc Carbenoids
Louis-Philippe B. Beaulieu,[a] Lucie E. Zimmer,[a] Alexandre Gagnon,[a, b] and
Andrꢀ B. Charette*[a]
Abstract: Herein, we report the enan-
tio- and diastereoselective formation of
trans-iodo- and trans-chlorocyclopro-
panes from a-iodo- and a-chlorozinc
carbenoids by using a dioxaborolane-
derived chiral ligand. The synthetically
clopropyl lithium species or a Negishi
coupling to give access to a variety of
enantioenriched 1,2,3-substituted cyclo-
propanes. The synthetic utility of this
method was demonstrated by the
formal synthesis of an HIV-1 protease
inhibitor. In addition, the related ster-
eoselective
bromocyclopropanation
Keywords: asymmetric synthesis ·
carbenoids · cyclopropanes · halo-
gen scrambling · Simmons–Smith
reactions · zinc
was also investigated. New insights
about the relative electrophilicity of
haloiodomethylzinc carbenoids are also
presented.
useful
iodocyclopropane
building
blocks were derivatized by an electro-
philic trapping of the corresponding cy-
Introduction
developed by using both unsubstituted[9,10] and substituted[11]
iodoalkylzinc reagents. The high levels of enantio- and dia-
stereoselectivity observed in the preparation of 1,2,3-sub-
stitued cyclopropanes by using alkyl-substituted zinc carbe-
noids and allylic alcohols in the presence of a dioxaboro-
lane-derived chiral ligand clearly illustrate the potential of
substituted zinc carbenoids in synthesis.[8d]
Despite its efficiency, the scope of this reaction is limited
due to the instability of iodoalkylzinc carbenoids and the
large number of molar equivalents of the 1,1-diiodoalkane
precursors that are generally needed to achieve high conver-
sions. A more divergent approach would be expected to rely
on the use of an a-functionalized zinc carbenoid
(RZnCHIX), in which the X substituent could be easily de-
rivatized to obtain structurally diverse, enantioenriched
1,2,3-substituted cyclopropanes. Ideally, this species would
be readily accessible from inexpensive reagents. In this con-
text, the development of a halocyclopropanation methodolo-
gy is quite appealing.
The cyclopropane unit continues to generate interest due to
its unique bonding properties, its rigidity, and its numerous
applications in diverse chemical transformations.[1] This
motif is present in several bioactive natural products as well
as synthetic drugs, which has prompted the development of
methodologies for their synthesis.[2] The Simmons–Smith cy-
clopropanation reaction[3] remains one of the most impor-
tant methods for the formation of cyclopropane derivatives
from alkenes.[1e,4] Following seminal reports[5] of a method
that featured a zinc–copper couple and diiodomethane as re-
agents, many variations that increase the efficiency of the re-
action, its reproducibility, and its solvent compatibility have
been reported. For instance, Furukawa et al. reported[6]
a
major breakthrough when they substituted the zinc–copper
couple with diethylzinc to form zinc carbenoids by an alkyl-
exchange reaction with diiodomethane.[7] Subsequently, nu-
merous substituted carbenoids were developed by using this
strategy with other gem-diiodomethyl motifs (RCHI2) and
were employed for the formation of the corresponding cy-
clopropanes.[8] Along with these advances, several enantiose-
lective cyclopropanation reactions of allylic alcohols were
Since the pioneering efforts of Hashimoto and Miyano,[12]
there has been very limited progress in the field of halocy-
clopropanation reactions that use a-iodomethylzinc carbe-
noids. A major drawback of these reactions is their poor dia-
stereoselectivity, which typically ranges from 2:1 to 1:2
(Scheme 1). It appears that the two possible transition states
involving the substituted carbenoid are too close in energy
to provide synthetically useful diastereomeric ratios when
unfunctionalized alkenes are used. Furthermore, the prepa-
ration of this reagent can be complicated by a second alkyl-
exchange reaction between the a-iodozinc carbenoid 1 and
an organozinc reagent (either 1, EtZnI, or Et2Zn), thus lead-
ing to a gem-dizinc carbenoid 2 (Scheme 2). This would lead
to a lower yield of the desired iodo-substituted product.[13]
Although the first versatile and highly diastereoselective
halocyclopropanation reaction of chiral allylic alcohols has
[a] L.-P. B. Beaulieu, Dr. L. E. Zimmer, Prof. A. Gagnon,
Prof. A. B. Charette
Centre in Green Chemistry and Catalysis
Department of Chemistry, Universitꢀ de Montrꢀal
P.O. Box 6128, Station Downtown, Montrꢀal, H3C 3J7 (Canada)
Fax : (+1)514-343-7586
E-mail: andre.charette@umontreal.ca
[b] Prof. A. Gagnon
Department of Chemistry, Universitꢀ du Quꢀbec ꢁ Montrꢀal
P.O. Box 8888, Succ. Centre-Ville, Montrꢀal, H3C 3P8 (Canada)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201202528.
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
1
&
ÞÞ
These are not the final page numbers!

Results and Discussion
Iodocyclopropanation reaction: Our preliminary experi-
ments related to the iodocyclopropanation reaction of allylic
alcohols are summarized in Table 1. To maximize conversion
Table 1. Optimization of the iodocyclopropanation with RZnCHI2.[a]
Scheme 1. Cyclopropanation of alkenes with a-diiodomethylzinc carbe-
noids.
CHI3
[equiv]
Additive[b]
SM
Yield
[%][d]
d.r.[e]
N
[%][c]
1[f]
2
3
4
5
4.4
4.4
2.2
2.2
2.2
–
–
–
I2
ꢀ2
ꢀ2
16
5:1
9:1
6:1
–
76 (66)[g]
14
62
–
–
ꢁ98
CF3CH2OH
ꢁ98
–
[a] Unless otherwise noted, the zinc carbenoid was formed by adding
neat Et2Zn to a suspension of CHI3 in CH2Cl2 at room temperature.
[b] A solution of CHI3 in CH2Cl2 was added to a preformed mixture of
Et2Zn and the additive (2.2 equiv) in CH2Cl2. [c] SM is the remaining
starting material as determined by 1H NMR spectroscopy using 1,3,5-tri-
methoxybenzene as an internal standard. [d] The yield of the trans dia-
stereomer was determined by 1H NMR spectroscopy using 1,3,5-trime-
thoxybenzene as an internal standard. [e] The diastereoselectivity was de-
Scheme 2. Stereodivergent methods for the preparation of 1,2,3-substitut-
ed cyclopropane derivatives.
1
termined by H NMR analysis of the crude reaction mixture. [f] The reac-
recently been introduced by Walsh et al.,[8e] an enantioselec-
tive version of this transformation is desirable. We recently
reported our preliminary results concerning two powerful
and complementary methodologies that can provide access
to stereoisomers of 1,2,3-substituted cyclopropane deriva-
tives from stereodefined allylic alcohols by using a-function-
alized zinc carbenoids (Scheme 2).[14]
Both approaches rely on the use of substituted zinc carbe-
noid reagents, which are derived from iodoform and Et2Zn,
and dioxaborolane (3) for the enantio- and diastereoselec-
tive preparation of iodo- and zinc-substituted cyclopropanes.
The complementarity of these reactions resides in the ster-
eochemical outcome of the cyclopropanation. In the iodocy-
tion was performed in the absence of 3. [g] Yield of isolated product is
given in brackets.
to the iodocyclopropane and the diastereomeric ratio (d.r.)
while minimizing the undesired zincocyclopropanation
mediated by the gem-dizinc carbenoid 2, we first studied the
effect of the stoichiometry of CHI3 relative to Et2Zn. A low
yield and modest diastereoselectivity were observed when
cinnamyl alcohol was treated with CHI3/Et2Zn (2:1) under
typical Furukawa–Miyano conditions (Table 1, entry 1). It
appears that under these reaction conditions the zinc alkox-
ide promotes the rapid decomposition of the a-iodozinc car-
benoid.[15] Surprisingly, the incorporation of dioxaborolane
(3) led not only to an improved diastereoselectivity, but also
to a much better yield of the desired trans-iodocyclopropane
5a (Table 1, entry 2). A 1:1 ratio of CHI3 relative to Et2Zn
did not ensure complete conversion (Table 1, entry 3).
Other more-electrophilic carbenoids, such as IZnCHI2 and
À
clopropanation reaction, the newly formed C I bond is in a
trans position to the hydroxymethyl group, whereas a cis re-
À
lationship is observed between the C Zn bond and the hy-
droxymethyl substituent in the reaction that involves the
gem-dizinc carbenoid.
Herein, we describe an extended scope for the enantiose-
lective iodocyclopropanation reaction. We also describe ad-
ditional reactions of these products for derivatization into
diverse 1,2,3-substituted cyclopropane motifs. The synthetic
utility of this methodology is demonstrated by the formal
synthesis of a biologically active peptidomimetic molecule.
We have also developed an unprecedented, highly stereose-
lective chlorocyclopropanation reaction that starts from
chlorodiiodomethane. Finally, new insights into the analo-
gous stereoselective bromocyclopropanation reaction are
also provided.
[8e]
CF3CH2OZnCHI2 (Table 1, entries 4 and 5, respectively),
did not lead to any formation of the desired product.
To gain further insight into the nature of the carbenoid in-
volved in the iodocyclopropanation, we quenched the pre-
formed zinc carbenoid with D2O. A ratio of CHI3 relative to
CHDI2 of 1 was observed upon forming the carbenoid at
room temperature and with a 2:1 stoichiometric ratio of
CHI3 relative to Et2Zn.This indicates that, under these con-
ditions, diethylzinc undergoes a single alkyl exchange with
iodoform to generate EtZnCHI2 (Hashimotoꢃs reagent).
The optimal reaction conditions (Table 1, entry 2) were
selected to evaluate the scope of this reaction (Table 2).
&
2
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Highly Enantioselective Synthesis of 1,2,3-Substituted Cyclopropanes
FULL PAPER
Table 2. Scope of the iodocyclopropanation reaction.
Gratifyingly, the iodocyclopropanation reaction proved to
be highly enantioselective and displayed good to excellent
diastereoselectivities with most of the allylic alcohols tested.
Synthetically useful yields of the isolated trans diastereomer
were obtained from the reaction of a variety of allylic alco-
hols.[16] Both electron-rich and -poor (E)-3-aryl-2-propen-1-
ols (Table 1, entries 1–9), some of which contain bulky aryl
substituents (Table 1, entries 2 and 3), were well tolerated.
Both primary and secondary (E)-3-alkyl-2-propen-1-ols
(Table 1, entries 11–16) underwent the iodocyclopropanation
reaction smoothly, although higher concentrations and
longer reaction times were needed to ensure complete con-
version. 2,2,3-Trisubstituted allylic alcohols (Table 1, en-
tries 17 and 18) gave products in excellent diastereo- and
enantioselectivities. The reaction proved to be chemoselec-
tive towards allylic alcohols (Table 1, entry 10) and was
compatible with functionalities such as nitroarenes (Table 1,
entry 5), chloroarenes (Table 1, entries 3 and 6), alkyl chlor-
ides (Table 1, entry 13), and tert-butyldimethylsilyl ethers
(Table 1, entry 12). The iodocyclopropanation of (Z)-2-
hexen-1-ol illustrated the marked influence of 3; a complete
inversion of diastereoselectivity was observed when the re-
action was performed in the absence of this ligand (Table 1,
entry 15). In all instances, the diastereomers were readily
separated by flash column chromatography, and the cis
isomer was less polar than the trans in all cases. It is worth
noting that the iodocyclopropanation of 4a could be per-
formed on a 10 mmol scale to provide 5a in 64% yield with
a d.r. of 9:1 and 96% ee.
The proposed transition-state model for the enantioselec-
tive iodocyclopropanation of cinnamyl alcohol in the pres-
ence of dioxaborolane 3 is depicted in Figure 1. The treat-
ment of cinnamyl alcohol with diethylzinc results in the for-
mation of a zinc alkoxide, which reacts with the dioxaboro-
lane to produce an ate complex. We believe that the bulkier
butyl substituent on the boron atom of this complex adopts
the sterically less-congested pseudoequatorial position,
whereas the allylic alkoxide assumes the electronically more
favorable pseudoaxial position. The resulting ate complex is
believed to coordinate to the zinc carbenoid in a bidentate
fashion through the allylic alkoxide and one of the Lewis
Product
Method[a] Yield d.r.[c]
[%][b]
ee
[%][d]
1
5a
5b
5c
5d
5e
5 f
A
B
66
73
81
70
64
69
9:1 98
2
18:1 98
12:1 98
7:1 96
7:1 97
9:1 98
3
A
A
B
4
5[e]
6
A
7
5g
5h
5i
A
A
A
B
B
55
70
74
53
66
7:1 98
11:1 98
6:1 96
5:1 98
5:1 96
8
9
10
5j
11
12
5k
5l
B
B
63
59
5:1 96[f]
6:1 95
13
5m
14
5n
5o
5p
B
B
B
67
42
46
4:1 95
4:1 99
8:1 90
15[g]
16
17
18
5q
5r
B
65
68
16:1 91
A
ꢁ20:1 93
[a] Method A: 0.1m in CH2Cl2, RT, 15 h. Method B: 0.2m in CH2Cl2, RT,
1
24 h. [b] Yield of isolated trans diastereomer. [c] Determined by H NMR
analysis of the crude reaction. [d] Determined by GC or SFC on a chiral
stationary phase. [e] Yield after dihydroxylation of the remaining traces
of the starting allylic alcohol. [f] Determined after removal of the pro-
À
tecting group. [g] The product where the C I bond is in a cis relationship
with the hydroxymethyl group was formed exclusively when the reaction
was conducted in the absence of 3. Mes=mesityl, TBS=tert-butyldime-
thylsilyl, Cy=cyclohexyl.
Figure 1. Representation of the proposed transition-state model for the
enantioselective iodocyclopropanation reaction of cinnamyl alcohol.
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
3
&
ÞÞ
These are not the final page numbers!

A. B. Charette et al.
basic amide moieties. The most suitable conformation for
CHI delivery is that in which the allylic chain is in its most
stable conformation. It is generally accepted that over the
course of the Simmons–Smith reaction the s*CÀ orbital
I
must be synperiplanar to the pCÀ orbital.[17,18]
C
Of the two possible reactive conformers that meet this
prerequisite, the conformer that contains an iodine atom
trans to the sterically encumbered dioxaborolane–alkoxide
complex leads to the major diastereomer. Recent density
functional theory (DFT) studies of analogous cyclopropana-
tion reactions conducted by Yu et al.[19] have corroborated
this model by minimization of torsional strain, 1,3-allylic
strain, and the ring strain that is generated in the intramo-
lecular cyclopropanation transition state.
Figure 2. Representation of the proposed transition-state model for the
enantioselective iodocyclopropanation reaction of (Z)-cinnamyl alcohol.
We observed one exception to the proposed model in the
iodocyclopropanation of (Z)-3-phenylprop-2-en-1-ol (4s;
Scheme 3).[20] In this particular case, the diastereoselectivity
diastereoselectivity due to the weaker chloro–arene interac-
tion than the corresponding iodo–arene interaction.[23] The
chlorocyclopropanation of
a (Z)-cinnamyl alcohol did
indeed lead to a 1:1 diastereomeric ratio of products (see
Table 6, entry 7 below).
To illustrate the versatility with which 2,3-disubstituted io-
docyclopropanes can be derivatized into several highly func-
tionalized 1,2,3-substituted cyclopropanes, the unprotected
or the O-benzyl-protected iodocyclopropanation products
were submitted to a lithium–halogen exchange reaction fol-
lowed by the addition of
a variety of electrophiles
(Table 3).[24] In all cases, the lithium–halogen exchange pro-
ceeded cleanly with complete retention of configuration
when tBuLi was used. The electrophiles were also cleanly
intercepted by the cyclopropyllithium reagents with com-
plete retention of configuration (Table 3, entries 1–6). Inter-
estingly, the reaction of an unprotected alcohol was success-
ful when an additional equivalent of tBuLi was employed
(Table 3, entries 7–9). A mixed cuprate could also be pre-
pared from the iodocyclopropylmethanol 5a and was subse-
quently treated with allyl bromide to provide the corre-
sponding allylated derivative in good yield (Scheme 4).[8f]
The benzyl-protected iodocyclopropylmethanol 7 could be
used in a Negishi cross-coupling reaction[25] with aryl iodides
(Table 4).[26] An extensive screening of phosphine-based li-
gands indicated that several were effective in promoting the
reaction.[27] Tris(ortho-tolyl)phosphine was chosen to elabo-
rate the scope of the reaction. Electron-rich (Table 4,
entry 5) and electron-deficient (Table 4, entries 1, 7, and 8)
aryl iodides were well tolerated in this reaction. Sterically
encumbered aryl iodides (Table 4, entries 4 and 6) also
proved to be competent coupling partners.
Having developed a powerful and versatile methodology
for the synthesis of enantioenriched 2,3-disubstituted trans-
iodocyclopropanes, we embarked on establishing its synthet-
ic utility by using it to elaborate the cyclopropane subunit of
the HIV-1 protease inhibitor 11.[28] The inhibitorꢃs cyclopro-
pane core (13) was accessed very efficiently from (E)-crotyl
alcohol by using our stereoselective iodocyclopropanation
methodology (Scheme 5). Lithium–halogen exchange fol-
lowed by quenching with carbon dioxide resulted in the for-
mation of the corresponding carboxylic acid, which was pro-
Scheme 3. Contrasting diastereoselectivities in the enantioselective iodo-
cyclopropanation of (E)- and (Z)-cinnamyl alcohol.
was reversed and the cis diastereomer was favored. This se-
lectivity contrasts with that of the iodocyclopropanation of
(E)-cinnamyl alcohol (4a) and, more strikingly, with that of
(Z)-2-hexen-1-ol (4o), which both result in the preferential
formation of the trans diastereomer. This observation led us
to propose that the reversal of diastereoselectivity is due to
a
stabilizing halogen-bonding interaction between the
HOMO of the aryl substituent (in 4a–j and 4r) and the a*
orbital of the carbon–iodine bond of the electrophilic zinc
carbenoid (Figures 1 and 2).[21,22] This electrostatic stabiliza-
tion could also account for the increase in diastereoselectivi-
ty when more electron-rich aryl substituents are present on
the starting allylic alcohol. This interaction is important be-
cause a remarkably higher trans diastereoselectivity was ob-
served in the cyclopropanation of the bulkier mesityl-substi-
tuted allylic alcohol (for example, see Table 2, entry 1 vs. 2).
If the arene–halide interaction is the main stereodirecting
effect for (Z)-cinnamyl alcohol, it is expected that the relat-
ed chlorocyclopropanation would occur with a much lower
&
4
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Highly Enantioselective Synthesis of 1,2,3-Substituted Cyclopropanes
FULL PAPER
Table 3. Functionalization of iodocyclopropanes by a sequential Li–I ex-
change and electrophilic quench.
Table 4. Negishi cross-coupling of iodocyclopropane derivatives.
Product
Electrophile
Yield
[%][a]
Product
Yield
[%][a]
1
2
3
4
5
6
7
8
9
8a
8b
8c
8d
8e
8 f
8g
8h
8i
PhC(O)Ph
ClCO2Me
MeI
90
70
97
85
75
85
65
84
88
1
2
3
4
5
6
7
8
10a
10b
10c
10d
10e
10 f
10g
10h
65
80
85
84
73
59
62
51
Bu3SnCl
CO2
ClC(O)NMe2
PhC(O)Ph
Bu3SnCl
CO2
[a] Yield of isolated product as a single diastereomer. dba=dibenzylide-
neacetone, Tol=tolyl, Naphth=naphthyl.
[a] Isolated yield of product as a single diastereomer. Bn=benzyl.
Scheme 4. Allylation through the cyclopropylcuprate.
tected as the methyl ester. Oxidation of the primary alcohol
to the carboxylic acid, followed by peptide coupling with
benzylamine and saponification of the methyl ester, led to
the isolation of an advanced intermediate (18) towards the
HIV-1 protease inhibitor in 33% overall yield over six steps,
which compares favorably to the reported eight-step route
developed by Martin and collaborators, which has an overall
yield of 11%.[28a]
Scheme 5. Formal synthesis of a peptidomimetic HIV-1 protease inhibi-
tor. a) EtZnCHI2, 3, CH2Cl2, RT, 63%, d.r. 5:1, 94% ee; b) i) tBuLi; ii)
CO2, Et2O, À788C to RT, 65%; c) SOCl2, MeOH, 08C to RT, 89%;
d) RuCl3·xH2O, NaIO4, CCl4/H2O/MeCN, 93%; e) EDC·HCl, BnNH2,
DMAP, CH2Cl2, RT, quant.; f) LiOH·H2O, THF, MeOH, RT, 96%.
Chloro- and bromocyclopropanation reactions: To increase
the scope of the enantioselective cyclopropanation reaction,
we were interested in investigating the analogous stereose-
lective chloro- and bromocyclopropanation reactions, which
have very few literature precedents.[8f] This reaction would
also provide further evidence to support the models present-
ed in Figures 1 and 2 because the halogen-bonding interac-
tions may not be as strong with these two halides. The a-
chloro and a-bromozinc carbenoids could be generated
from the readily accessed haloforms ClCHI2 and BrCHI2,
respectively.[29] The chlorocyclopropanation performed by
using 4.4 equivalents of ClCHI2 at room temperature led to
an unsatisfactory mixture of chlorocyclopropane and iodocy-
clopropane products and a low conversion (Table 5, entry 1).
These observations suggested that the zinc carbenoid may
be unstable under these reaction conditions. Lowering the
reaction temperature led to interesting ratios of 19a/5–6a
and complete conversion (Table 5, entries 2–4). Under the
optimal conditions, chlorocyclopropane 19a was isolated in
good yield and with a high ee (Table 5, entry 4). Decreasing
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
5
&
ÞÞ
These are not the final page numbers!

A. B. Charette et al.
Table 5. Optimization of the chlorocyclopropanation.
Table 6. Scope of the chlorocyclopropanation.
Product
Method,[a]
d.r.[c]
ee
4a
[equiv]
T
[8C]
Ratio
SM
Yield [%][b]
[%][d]
G
19/5–6a[a]
[%][b]
1
2
3
4
5
6
4.4
4.4
4.4
4.4
3.3
2.2
25
3:1
12:1
ꢁ20:1
ꢁ20:1[c,d]
ꢁ20:1
–
60
53
10
1
2
3
19a
19h
19g
A, 78
ꢁ20:1
ꢁ20:1
10:1
93
97
85
À20
À40
À78 to À40
À78 to À40
À78 to À40
ꢀ2
B, 73
B, 73
B, 55
B, 73
B, 55
A, 46
19
ꢁ95
[a] Determined by 1H NMR analysis of the crude reaction mixture. The
d.r. for the formation of 19a was ꢁ20:1. [b] SM is the remaining starting
material as determined by 1H NMR analysis by integrating the signals
from the chlorocyclopropane and cinnamyl alcohol. [c] Diastereomerical-
ly pure 19a was isolated in 78% yield. [d] A 93% ee was determined by
SFC on chiral stationary phase.
4
5
6
7
19b
19k
19n
19s
ꢁ20:1
ꢁ20:1
ꢁ20:1
1:1
95
90[e]
92[e]
98[f]
the amount of ClCHI2 relative to diethylzinc resulted in
lower conversions (Table 5, entries 5 and 6). Interestingly,
conducting the reaction in the absence of the dioxaborolane
3 (Scheme 6) led to a much lower diastereoselectivity, a
result which highlights the role of the dioxaborolane ligand
for both the enantio- and diastereoselectivity of the reac-
tion.
[a] Method A: 0.1m in CH2Cl2, À78 to À408C, 15 h. Method B: 0.2m in
CH2Cl2, À78 to À408C, 24 h. [b] Isolated yield of diastereomerically pure
19. [c] Determined by 1H NMR analysis of the crude reaction mixture.
[d] Determined by SFC on a chiral stationary phase. [e] Determined by
SFC analysis of the benzoylated product on a chiral stationary phase.
[f] Determined by 19F NMR analysis of the corresponding Mosherꢃs ester.
Scheme 6. Decreased diastereoselectivity observed for the chlorocyclo-
propanation of 4a in the absence of 3.
Scheme 7. Attempted bromocyclopropanation reaction.
The enantioselective chlorocyclopropanation reaction is
compatible with cinnamyl alcohol derivatives that contain
electronically different substituents (Table 6, entries 1–4).
Steric hindrance on the aromatic ring was also tolerated
(Table 6, entry 2). Finally, the reaction proceeded smoothly
with allylic alcohols containing a primary or secondary
(Table 6, entries 5 and 6, respectively) alkyl substituent. In
all cases, reasonable yields and excellent diastereoselectivi-
ties and enantioselectivities were observed.[30] To the best of
our knowledge, this is the first reported enantioselective
chlorocyclopropanation reaction.
Upon attempting the analogous bromocyclopropanation
under the optimal reaction conditions, we observed com-
plete conversion to a mixture of iodo- and bromocyclopro-
panes. Surprisingly, the iodocyclopropane was the major
component of the mixture (Scheme 7).
terium carbenoid-quenching experiments with BrCHI2 as
the carbenoid precursor (Scheme 8). Upon the addition of
one equivalent of diethylzinc to two equivalents of BrCHI2
and subsequently quenching with deuterated acetic acid, we
observed all possible combinations of deuterated dihalome-
thanes by GC/MS, with complete consumption of the
BrCHI2 starting material. This observation is indicative of a
halogen scrambling that takes place on the zinc carbenoid.
We observed the formation of the same byproducts when
the stoichiometric ratio of BrCHI2 to diethylzinc was
brought to 1:1, which suggested that this exchange could
also be intermolecular in nature and mediated by the pres-
ence of zinc halides. Similar results were obtained upon
quenching the carbenoids generated from diethylzinc and
chlorodiiodomethane.
These observations indicate that several reagents are si-
multaneously present in these reactions. The formation of a
single, major, chloro-substituted cyclopropane product from
At this point, we wanted to shed some light on the mecha-
nism of iodocyclopropane formation so we performed deu-
&
6
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Highly Enantioselective Synthesis of 1,2,3-Substituted Cyclopropanes
FULL PAPER
Conclusion
We have developed conditions for the enantioselective iodo-
cyclopropanation of allylic alcohols from commercially
available reagents. The versatile trans-iodocyclopropylme-
thanols have been functionalized to gain access to a variety
of highly enantioenriched 1,2,3-trisubstituted cyclopropanes,
which would be difficult to access otherwise. The synthetic
utility of the stereoselective iodocyclopropanation reaction
was demonstrated by the formal synthesis of a biologically
active peptidomimetic molecule. In addition, we have inves-
tigated the analogous stereoselective bromo- and chlorocy-
clopropanation reactions. Although the bromocyclopropana-
tion was inefficient due to the competing iodocyclopropana-
tion reaction that stemmed from a halogen scrambling on
the zinc carbenoid, the enantioselective chlorocyclopropana-
tion proceeded in good yield and with excellent stereoselec-
tivity. We demonstrated that the chloro-substituted zinc car-
benoid is more reactive than the corresponding iodo- and
bromo-substituted zinc carbenoids. To the best of our
knowledge, these are the first accounts of enantioselective
chloro- and iodocyclopropanation reactions. These method-
ologies provide an efficient route to numerous, synthetically
versatile, enantioenriched 1,2,3-trisubstituted cyclopropanes.
Scheme 8. Deuterium quenching experiments of the zinc carbenoids gen-
erated from XCHI2 (X=Br and Cl).[31] nd=not determined.
chlorodiiodomethane suggests that the chloroiodomethyl-
zinc carbenoid 22 is more reactive than the corresponding
diiodomethylzinc species 1. Conversely, the bromoiodome-
thylzinc species 21 and the analogous diiodomethylzinc 1
seem to have comparable reactivities (Scheme 9).
Acknowledgements
Scheme 9. a-Dihalomethylzinc carbenoid equilibration.
This work was supported by the Natural Science and Engineering Re-
search Council of Canada (NSERC), the Canada Foundation for Innova-
tion, the Canada Research Chair Program, the Centre in Green Chemis-
try and Catalysis (CGCC) and the Universitꢀ de Montrꢀal. L.P.B.B. and
A.G. are grateful to NSERC for postgraduate fellowships (PGS D and
PGF A, respectively). The authors are grateful to Dr. Alexandre Lemire
for useful discussions.
This was confirmed by conducting a competition experi-
ment in which all three reagents were present (Scheme 10).
When 4a was treated with the reagents derived from dieth-
ylzinc and equimolar amounts of all three halodiiodome-
thanes, the chlorocyclopropane 19a was the major product.
This result unambiguously confirms that the chloroiodome-
thylzinc carbenoid that contains the more electronegative
chloro substituent is more reactive than the other two re-
agents, probably due to an increased electrophilicity.
[1] a) K. B. Wiberg, Acc. Chem. Res. 1996, 29, 229–234; b) K. B. Wiberg
in Houben-Weyl Methods of Organic Chemistry, Vol. E17a (Ed.: A.
de Meijere), Thieme, Stuttgart, 1997, pp. 1–27; c) A. de Meijere,
´
Chem. Rev. 2003, 103, 931–932; d) M. Fedorynski, Chem. Rev. 2003,
103, 1099–1132; e) H. Lebel, J.-F. Marcoux, C. Molinaro, A. B.
Charette, Chem. Rev. 2003, 103, 977–1050; f) H.-U. Reissig, R.
Zimmer, Chem. Rev. 2003, 103, 1151–1196; g) Cyclopropanes: A.
de Meijere, S. I. Kozhushkov in Science of Synthesis, Vol. 48 (Ed.:
H. Hiemstra), Thieme, Stuttgart 2009, pp. 477–588; h) A. Galano, J.
Alvarez-Idaboy, A. Vivier-Bunge, Theor. Chim. Acc 2007, 118, 597–
606; i) S. R. Goudreau, A. B. Charette, Angew. Chem. 2010, 122,
496–498; Angew. Chem. Int. Ed. 2010, 49, 486–488.
[2] A. Gagnon, M. Duplessis, L. Fader, Org. Prep. Proced. Int. 2010, 42,
1–69.
[3] a) H. E. Simmons, T. L. Cairns, S. A. Vladuchick, C. M. Hoiness,
Org. React. 1973, 20, 1; b) A. B. Charette, A. Beauchemin, Org.
React. 2001, 58, 1.
[4] A. B. Charette, M.-N. Roy, V. N. G. Lindsay in Stereoselective Syn-
thesis: Reactions of Carbon–Carbon Double Bonds, Vol. 1 (Ed.:
J. G. de Vries), Thieme, Stuttgart, 2011, Chapter 1.14, pp. 731–817.
[5] a) H. E. Simmons, R. D. Smith, J. Am. Chem. Soc. 1958, 80, 5323–
5324; b) H. E. Simmons, R. D. Smith, J. Am. Chem. Soc. 1959, 81,
4256–4264; c) E. P. Blanchard, H. E. Simmons, J. Am. Chem. Soc.
1964, 86, 1337–1347; d) H. E. Simmons, E. P. Blanchard, R. D.
Smith, J. Am. Chem. Soc. 1964, 86, 1347–1356.
Scheme 10. Competition experiment with a mixture of reagents.
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
7
&
ÞÞ
These are not the final page numbers!

A. B. Charette et al.
[6] a) J. Furukawa, N. Kawabata, J. Nishimura, Tetrahedron Lett. 1966,
7, 3353–3354; b) J. Furukawa, N. Kawabata, J. Nishimura, Tetrahe-
dron 1968, 24, 53–58.
[7] For other important variations, see: a) S. E. Denmark, J. P. Edwards,
J. Org. Chem. 1991, 56, 6974–6981; b) J. C. Lorenz, J. Long, Z.
Yang, S. Xue, Y. Xie, Y. Shi, J. Org. Chem. 2004, 69, 327–334;
c) M.-C. Lacasse, C. Poulard, A. B. Charette, J. Am. Chem. Soc.
2005, 127, 12440–12441; d) A. Voituriez, L. E. Zimmer, A. B. Char-
ette, J. Org. Chem. 2010, 75, 1244–1250.
Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/
data_request/cif
[17] C. Zhao, D. Wang, D. L. Phillips, J. Am. Chem. Soc. 2002, 124,
12903–12914.
[18] a) T. K. Dargel, W. Koch, J. Chem. Soc. Perkin Trans. 2 1996, 877–
881; b) A. H. Hoveyda, D. A. Evans, G. C. Fu, Chem. Rev. 1993, 93,
1307–1370.
[19] T. Wang, Y. Liang, Z.-X. Yu, J. Am. Chem. Soc. 2011, 133, 9343–
9353.
[8] a) J. Furukawa, N. Kawabata, J. Nishimura, Tetrahedron Lett. 1968,
9, 3495–3498; b) J. Nishimura, J. Furukawa, J. Chem. Soc. D 1971,
1375–1376; c) S. Miyano, Y. Matsumoto, H. Hashimoto, J. Chem.
Soc. Chem. Commun. 1975, 364–365; d) A. B. Charette, J. Lemay,
Angew. Chem. 1997, 109, 1163–1165; Angew. Chem. Int. Ed. Engl.
1997, 36, 1090–1092; e) H. Y. Kim, A. E. Lurain, P. Garcꢄa-Garcꢄa,
P. J. Carroll, P. J. Walsh, J. Am. Chem. Soc. 2005, 127, 13138–13139;
f) H. Y. Kim, L. Salvi, P. J. Carroll, P. J. Walsh, J. Am. Chem. Soc.
2009, 131, 954–962.
[20] The absolute configurations of 6a and 6s were assigned by compari-
son with known compounds from the literature, see reference [14a].
[21] These p-halogen interactions have been observed by X-ray crystal-
lography and account for 33% of all halogen bonds in protein com-
plexes with halogenated ligands, see: a) H. Matter, M. Nazarꢀ, S.
Gꢅssregen, D. W. Will, H. Schreuder, A. Bauer, M. Urmann, K.
Ritter, M. Wagner, V. Wehner, Angew. Chem. 2009, 121, 2955–2960;
Angew. Chem. Int. Ed. 2009, 48, 2911–2916; b) Y. Lu, Y. Wang, W.
Zhu, Phys. Chem. Chem. Phys. 2010, 12, 4543–4551.
[9] a) A. B. Charette, H. Juteau, J. Am. Chem. Soc. 1994, 116, 2651–
2652; b) A. B. Charette, H. Juteau, H. Lebel, C. Molinaro, J. Am.
Chem. Soc. 1998, 120, 11943–11952.
[22] For a review of important molecular interactions in medicinal chem-
istry, see: C. Bissantz, B. Kuhn, M. Stahl, J. Med. Chem. 2010, 53,
5061–5084.
[10] For examples of catalytic asymmetric Simmons–Smith cyclopropana-
tion of allylic alcohols, see: a) A. B. Charette, C. Brochu, J. Am.
Chem. Soc. 1995, 117, 11367–11368; b) H. Takahashi, M. Yoshioka,
M. Shibasaki, M. Ohno, N. Imai, S. Kobayashi, Tetrahedron 1995, 51,
12013–12026; c) S. E. Denmark, S. P. OꢃConnor, J. Org. Chem. 1997,
62, 584–594; d) A. B. Charette, C. Molinaro, C. Brochu, J. Am.
Chem. Soc. 2001, 123, 12168–12175; e) H. Shitama, T. Katsuki,
Angew. Chem. 2008, 120, 2484–2487; Angew. Chem. Int. Ed. 2008,
47, 2450–2453.
[11] See reference [8d] and S. R. Goudreau, A. B. Charette, J. Am.
Chem. Soc. 2009, 131, 15633–15635.
[12] S. Miyano, H. Hashimoto, Bull. Chem. Soc. Jpn. 1974, 47, 1500–
1503.
[23] The general trend for halogen bond strength is Cl<Br<I, see: E.
Parisini, P. Metrangolo, T. Pilati, G. Resnati, G. Terraneo, Chem.
Soc. Rev. 2011, 40, 2267–2278.
[24] For examples of lithium–iodide exchange on cyclopropane deriva-
tives, see: S. F. Martin, M. P. Dwyer, Tetrahedron Lett. 1998, 39,
1521–1524.
[25] For reviews of the Negishi reaction, see: a) E. Negishi, Bull. Chem.
Soc. Jpn. 2007, 80, 233–257; b) E. Negishi, Angew. Chem. 2011, 123,
6870–6897; Angew. Chem. Int. Ed. 2011, 50, 6738–6764.
[26] For seminal examples of Negishi cross-couplings performed on iodo-
cyclopropanes, see: a) E. Piers, P. D. G. Coish, Synthesis 1995, 47–
55; b) E. Piers, P. D. G. Coish, Synthesis 2001, 251–261; c) refer-
ence [24].
[13] a) A. B. Charette, A. Gagnon, J.-F. Fournier, J. Am. Chem. Soc.
2002, 124, 386–387; b) J.-F. Fournier, S. Mathieu, A. B. Charette, J.
Am. Chem. Soc. 2005, 127, 13140–13141; c) J.-F. Fournier, A. B.
Charette, Eur. J. Org. Chem. 2004, 1401–1404.
[14] a) L. E. Zimmer, A. B. Charette, J. Am. Chem. Soc. 2009, 131,
15624–15626; b) L.-P. B. Beaulieu, L. E. Zimmer, A. B. Charette,
Chem. Eur. J. 2009, 15, 11829–11832.
[15] The significant amount of (3-phenylcyclopropyl)methanol formed in
this reaction may indicate that the gem-dizinc carbenoid 2 is gener-
ated under these reaction conditions.
[16] The absolute configurations of 5e and of the O-3,5-dinitrobenzoyl
derivatives of 5k, 5o, 5p, and 5q were determined by X-ray crystal-
lography and they are all consistent with the model shown in
Figure 1; see the Supporting Information. CCDC-744857 (5e),
CCDC-901922 (5k), CCDC-901921 (5o), CCDC-901923 (5p), and
CCDC-902069 (5q) contain the supplementary crystallographic data
for this paper. These data can be obtained free of charge from The
[27] See the Supporting Information for further details.
[28] a) S. F. Martin, G. O. Dorsey, T. Gane, M. C. Hillier, H. Kessler, M.
Baur, B. Mathꢆ, J. W. Erickson, T. N. Bhat, S. Munshi, S. V. Gulnik,
I. A. Topol, J. Med. Chem. 1998, 41, 1581–1597; b) A. Reichelt, S. F.
Martin, Acc. Chem. Res. 2006, 39, 433–442.
[29] D. B. Li, S.-C. Ng, I. Novak, Tetrahedron 2002, 58, 5923–5926.
[30] The absolute configuration of the O-3,5-dinitrobenzoyl derivative of
19a was determined by X-ray crystallography; see the Supporting
Information. CCDC-901920 (19a) contains the supplementary crys-
tallographic data for this paper. These data can be obtained free of
charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
[31] It was not possible to unambiguously quantify nor identify CHDCl2
since the reaction was run in CH2Cl2.
Received: July 16, 2012
Published online: && &&, 0000
&
8
&
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!

Highly Enantioselective Synthesis of 1,2,3-Substituted Cyclopropanes
FULL PAPER
Asymmetric Synthesis
L.-P. B. Beaulieu, L. E. Zimmer,
A. Gagnon, A. B. Charette* &&&& — &&&&
Highly Enantioselective Synthesis of
1,2,3-Substituted Cyclopropanes by
Using a-Iodo- and a-Chloromethylzinc
Carbenoids
Propping up the enantioselectivity:
The enantio- and diastereoselective
formation of trans-iodo- and trans-
chlorocyclopropanes from a-iodo- and
a-chlorozinc carbenoids by using a
dioxaborolane-derived chiral ligand is
reported. The iodocyclopropanes were
derivatized into a variety of 1,2,3-sub-
stituted cyclopropanes and new
insights into the relative electrophilic-
ity of a-haloiodomethylzinc carbenoids
are presented (see scheme).
Chem. Eur. J. 2012, 00, 0 – 0
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
&
9
&
ÞÞ
These are not the final page numbers!