Desymmetrisation of cyclopentadienylsilane by asymmetric cyclopropanation

Article abstract of DOI:10.1002/ejoc.200390157

Desymmetrisation of silylcyclopenta-2,4-diene was carried out by an asymmetric copper(I)-mediated cyclopropanation. An in-depth investigation with various ligands led to the discovery that the PyBox ligands 10a-c were the most efficient ligands for this transformation and led to the cyclopropane 7a with an ee of up to 72%. Further studies aimed at providing insights into the origin of this enantiocontrol are also provided. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2003.

Full text of DOI:10.1002/ejoc.200390157

FULL PAPER
Desymmetrisation of Cyclopentadienylsilane by Asymmetric Cyclopropanation
[a]
[a]
[a]
[a]
´
Florent Allais, Remy Angelaud, Boris Camuzat-Dedenis, Karine Julienne, and
Yannick Landais*[a]
Keywords: Dienes / Silicon / Cyclopropanation / Carbenoids / Copper
Desymmetrisation of silylcyclopenta-2,4-diene was carried
out by an asymmetric copper( )-mediated cyclopropanation.
An in-depth investigation with various ligands led to the dis-
with an ee of up to 72%. Further studies aimed at providing
insights into the origin of this enantiocontrol are also pro-
vided.
I
covery that the PyBox ligands 10a−c were the most efficient ( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
ligands for this transformation and led to the cyclopropane 7a
Germany, 2003)
Introduction
We recently reported that desymmetrisation of silylcy-
clohexa-2,5-dienes of type 1a (n ϭ 1, Scheme 1) through
Sharpless dihydroxylation gave access to diols 2a (n ϭ 1) in
high yield, with complete diastereocontrol and reasonable
enantiocontrol.[1] Cyclopropanation of the remaining
double bond was then shown to provide stereoselectively
the corresponding cyclopropylmethylsilanes 3, which
could be opened, with concomitant desilylation, under elec-
trophilic[2a] (RЈ ϭ H) or nucleophilic[2b] (RЈ ϭ CO2Et) con-
ditions, leading to trisubstituted cyclohexenes which were
converted later into a range of sugar mimics (Scheme 1).
When trying to extend this strategy to the silylcyclopenta-
2,4-diene analogues 1b (n ϭ 0), decomposition of the start-
ing material occurred during dihydroxylation and no trace
of the desired diol 2b could be isolated.[2a] At the time, this
plagued our efforts to access homologous cyclopentitols
and carbo-furanose, thus limiting the scope of our meth-
Scheme 1
Results and Discussion
The limited scope of the strategy above and the known
odology. The preparation of enantiopure carba-C-furanos- utility of cyclopropylmethylsilanes[2a] as synthons for or-
ides was, however, carried out using the reverse strategy in- ganic synthesis prompted us to start an investigation on the
stead, i.e. cyclopropanation of silylcyclopenta-2,4-diene 1b desymmetrisation of silylcyclopenta-2,4-diene 6 through
then dihydroxylation of the remaining allylsilane moiety to copper-catalysed cyclopropanation (Scheme 2, Table 1).
give the fully functionalised cyclopentane 3b.[2] This ap- Based on earlier observations that Schiff bases catalysed
proach eventually led to the synthesis of optically pure cyclopropanation reactions efficiently,[3] our preliminary
carba-C-furanoside 4 in good overall yield, following the studies were carried out with the simple C2-symmetric
coupling of the carba-furanose moiety with a lithio-glycos- Schiff bases 8aϪb (Scheme 3). As shown in Table 1
ide.
(Scheme 2), these led to the expected mono-cyclopropan-
ated product 7a along with a minor isomer 7b (not shown)
but with no enantioselectivity (entries 1Ϫ2, Table 1). Sim-
ilar results were observed with bis-oxazolines 9aϪb,[4]
which are known to be highly efficient ligands for cyclopro-
panations (entries 3Ϫ4). Finally, we were pleased to find
that Nishiyama’s PyBox 10a[5] provided the desired cyclo-
propanes 7aϪb with good diastereocontrol in reasonable
[a]
University Bordeaux-I, Laboratoire de Chimie Organique et
´
Organometallique,
´
351, Cours de la Liberation, 33405 Talence Cedex, France
Fax: (internat.) ϩ33-(0)55Ϫ684Ϫ6286
E-mail: y.landais@lcoo.u-bordeaux.fr
Supporting information for this article is available on the
WWW under http://www.eurjoc.org or from the author.
Eur. J. Org. Chem. 2003, 1069Ϫ1073  2003 WILEY-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim 1434Ϫ193X/03/0306Ϫ1069 $ 20.00ϩ.50/0
1069

F. Allais, R. Angelaud, B. Camuzat-Dedenis, K. Julienne, Y. Landais
FULL PAPER
yield[6] (over two steps) with 68% enantiomeric excess for
both the major and the minor isomers (entry 5). Cyclopro-
panation using other PyBox-type ligands (10bϪc) also led
to the same behaviour, although with a lower enantioselec-
tivity (entries 6Ϫ7, Table 1).[7] It is noteworthy that no trace
of bis-cyclopropanation product could be detected during
these processes. The involvement of a kinetic amplification
which would raise the level of enantioselectivity by con-
sumption of the minor enantiomer may thus be ruled out.[8]
To the best of our knowledge, this is the first example of a
desymmetrisation using an asymmetric cyclopropanation.[9]
Both diastereomers were readily separated by chromato-
1
graphy and, after extensive H and 13C NMR experiments
(DEPT, COSY, HMQC, NOESY and HMBC), their rela-
tive configurations were determined unambiguously. As ob-
served in the racemic series,[2a] the major isomer 7a was
shown to possess the C4ϪC5-trans configuration, demon-
strating that the carbenoid species approaches anti relative
to the silicon group. More surprisingly, NOE experiments
indicated that the minor isomer 7b had the C4-C5-cis ste-
reochemistry (Figure 1) showing that, to a certain extent,
the electrophilic carbenoid reagent (vide infra) can also ap-
proach syn relative to the bulky silicon group. This result is
particularly surprising considering the wealth of data
indicating anti stereospecificity for electrophilic processes
occurring with chiral allylsilanes.[10]
Scheme 3
Figure 1. NOE effects in 7b (minor)
tate ligand sequesters the metal to form a stable copper
complex in solution. Tridentate ligands such as pyridine-
bis-benzimidazoles and PyBox are known to form stable
helicate complexes with copper and silver in solution.[12]
The structure of these complexes has also been determined
by X-ray crystallography.[12b] Unfortunately, we were unable
to crystallise a 1:1 CuI:10a complex. However, we found
that adding a slight excess of CuIOTf to the 1:1 CuI:10a
mixture restored the catalytic activity of the system. As CuI-
PyBox:10a ratio as low as 1:0.2 could be used without any
noticeable decrease in yield or enantioselectivity (entries
1Ϫ3, Table 2), suggesting that copper aggregates are in-
volved in the process. These aggregates then dissociate when
CuI is in excess, providing the catalytically active species
(probably monomeric) involved in the stereochemistry-de-
termining step. In order to get further insights on the nature
and relative stability of the copper complexes involved in
our cyclopropanation, we examined the possible occurrence
of a nonlinear effect.[13] As depicted in Figure 2 (entries
6Ϫ9, Table 2), we observed a linear relationship between
the ee of the product and the ee of 10a (Figure 2). It is
noteworthy that the diastereocontrol remained the same
when varying the ee of the ligand. We also changed the
metal involved in the reaction and performed the same ex-
Scheme 2
Table 1. Asymmetric cyclopropanation of 6 (Scheme 2); study on
the nature of the ligand L*
[b]
[b]
Entry
L*
8a
8b
9a
9b
10a
10b
10c
dr[a]
eemaj.
eemin.
Yield (%)[c]
1
2
3
4
5
6
7
95:5
95:5
94:6
95:5
82:18
98:2
96:4
2
2
4
0
68
6
Ϫ
Ϫ
Ϫ
Ϫ
68
10
36
20
26
28
47
48
45
41
48
[a]
1
[b]
Estimated from the H NMR spectrum of the crude mixture.
Measured by HPLC (hexane/iPrOH, 95:5), Chiralcel OD-H.
[c]
Overall yield starting from 5 (two steps) after purification by chro-
matography.
With these results in hand we then investigated the mech- periments with RhCl3.[5c] Surprisingly, the reaction did not
anism of the reaction and the origin of the enantioselectiv- afford the desired cyclopropane. Finally, negligible temper-
ity in more detail. We first varied the CuI/10a ratio and ature effects were observed on the enantioselectivity (72%
found that when PyBox 10a was slightly in excess or in an vs. 68%) and yield of the process (entries 3Ϫ5, Table 2).
equimolar amount with the copper() salt, not only was
We also varied the number of chelating groups (N) on
cyclopropanation not observed but ethyl diazoacetate was the ligands in order to obtain additional information about
not even decomposed.[11] This may indicate that the triden- the implication of the different chelating sites of PyBox.[14]
1070
Eur. J. Org. Chem. 2003, 1069Ϫ1073

Desymmetrisation of Cyclopentadienylsilane by Asymmetric Cyclopropanation
FULL PAPER
Table 2. Asymmetric cyclopropanation of 6 (Scheme 2)
Entry Ligand[a] T (°C) dr[b] eemaj.
eemin.
Yield (%)[d]
[c]
[c]
1
2
3
4
5
6
7
8
9
(S)-10a[e] 20
(S)-10a[f] 20
84:16 68
83:17 66
82:18 68
83:17 68
85:15 72
84:16 72
84:16 52
85:15 32
85:15 20
77
74
68
84
82
72
52
32
22
4
48
48
48
48
50
47
42
43
45
23
46
(S)-10a
(S)-10a
(S)-10a
(S)-10a
20
26
11
20
Scheme 4
allylsilane 7a produced the desired epoxide, which was not
isolated but transformed directly into the alcohol 12[15]
through an acid-catalysed Peterson reaction. Swern oxida-
tion of the alcohol finally furnished the known cyclopro-
pane 13[16] with a specific rotation of Ϫ135° (ref.[16] Ϫ255°),
indicating that 7a has the absolute configuration
(1R,4S,5R,6R) shown below.
(S)-(70%) 20
(S)-(50%) 20
(S)-(30%) 20
10
11
11a
11b
20
20
91:9
6
89:11 16
2
[a] Unless indicated otherwise the CuI:ligand ratio is 1:0.5. [b] Estim-
ated from the 1H NMR spectrum of the crude mixture. [c] Measured
by HPLC (Hexane/iPrOH, 95:5), Chiralcel OD-H. [d] Overall yield
over two steps after purification by chromatography.[e] CuI:ligand
[f]
ratio of 1:0.8. CuI:ligand ratio of 1:0.2.
Scheme 5
With the absolute configuration of 7a in hand and con-
sidering the information collected previously, we next at-
tempted to explain the origin of the enantioselectivity. Re-
cent calculations by Noorby et al.[17] have shown that the
stereochemistry-determining step of copper-catalysed cyclo-
propanation proceeds through a concerted but asynchron-
ous addition of an electrophilic copper-carbenoid species[18]
onto the alkene. An open transition state was thus proposed
Figure 2. Variation of the ee of 7a as a function of the ee of li-
gand 10a
Ligands 11a and 11b were thus designed as bidentate ana- for the cyclopropanation of styrene, where bonding between
logues (Scheme 4). In the former, the trinuclear skeleton the less-substituted end of the olefin and the carbenoid
was maintained with a nonchelating phenyl replacing pyrid- centre occurred very early on, leaving a partial positive
ine and in the latter, one oxazoline was removed. As charge at the benzylic position. The cyclopropane was then
summarised in Table 2 (entries 10Ϫ11), PheBox (11a) led to formed upon ring closure at a later stage, although still in
low enantioselectivity but also to a poorer conversion than a concerted manner. In our case, such a scenario could also
10a. PyOx (11b) led to cyclopropanes 7aϪb with yield and operate with the development of a partial positive charge
diastereoselectivity similar to those obtained with 10a but β to the silicon group, stabilised through hyperconjugation
again with poor enantiocontrol. This demonstrates that the (silicon β-effect).[19] The diene would thus approach the li-
three chelating sites of PyBox are required to obtain high gand-copper-carbenoid complex as shown in Figure 3. The
levels of enantioselectivity. These observations and the ab- enantioselectivity of the process is thought to be governed
sence of a nonlinear effect suggest that the catalytically act- by steric interactions developing between substituents on
ive species in the stereochemistry-determining step may in- the carbene (carboxylate) and those on the chiral ligand
volve only one chiral ligand (vide infra). If copper aggreg- (iPr groups), during carbenoid pyramidalisation, as pro-
ates are present in solution, they probably dissociate with posed earlier by Pfaltz[4a] and confirmed by Noorby’s calcu-
low energy of activation.[13]
lations.[17] The preferential attack of the carbenoid species
The absolute configuration of 7a was determined using at the pro-(S) double bond (anti relative to the SiR3 group,
the sequence summarised in Scheme 5. Epoxidation of the Figure 3) would thus prevent steric interactions developing
Eur. J. Org. Chem. 2003, 1069Ϫ1073
1071

F. Allais, R. Angelaud, B. Camuzat-Dedenis, K. Julienne, Y. Landais
FULL PAPER
´
determined at the Centre Regional de Mesures Physiques de l’Ouest
between the iPr group at C-1 and the carboxylate. Calcula-
tions also suggested that bonding between Cu and the car-
benoid centre was best represented as CuϪCϩ, correspond-
ing to the lone pair of a singlet carbene coordinated to a
CuI cation.[17] As already mentioned, CuI cations are effi-
ciently coordinated by tridentate ligands. The three nitro-
gens present in PyBox (10a) would thus be required to
stabilise the monomeric carbenoid species through coor-
(Rennes). Optical rotations were measured on a PerkinϪElmer po-
larimeter using a 10 cm cell. CH2Cl2, Et3N were distilled from
CaH2. THF was distilled from over sodium/benzophenone. Li-
gands 10aϪc were purchased from Aldrich (99%) and used in with-
out further purification.
(Cyclopenta-2,4-dien-1-yl)dimethylphenylsilane (6):[2a] A 2.5  solu-
tion of nBuLi in hexanes (14.4 mL, 36 mmol) was added, at Ϫ78
dination to CuI. Finally, the formation of the minor isomer °C, to a solution of freshly distilled cyclopentadiene (2 g, 30 mmol)
in anhydrous THF (90 mL). The solution was stirred at Ϫ78 °C for
45 minutes, then a solution of PhMe2SiCl (5 mL, 30 mmol) in THF
(10 mL) was added. The reaction mixture was stirred at Ϫ78 °C
for 1.5 h, then quenched with a saturated solution of NH4Cl and
allowed to warm to room temperature. The organic layer was de-
canted and the aqueous layer extracted with diethyl ether. The com-
bined extracts were washed with brine, dried over MgSO4 and con-
centrated in vacuo to give 6 as a yellow oil (5.4 g, 92%). The crude
product was used in the next step without further purification.
7b, with relatively high enantioselectivities (Table 1Ϫ2), may
be rationalised using a similar transition state model. This
would result from an approach of the carbenoid species, syn
relative to the silicon group. Such an approach would not
suffer from too much steric crowding because the bonding
between the carbenoid centre and the allylsilane is thought
to occur first at the γ-position relative to the silicon bond
(vide supra). Although this has not been firmly established,
using this model the pro-(R) enantiotopic double bond of
6 would thus be cyclopropanated to form 7b (Figure 1).
Ethyl 4-(Dimethylphenylsilyl)bicyclo[3.1.0]hex-2-ene-6-carboxylate
(7a and 7b): Two drops of ethyl diazoacetate were added to a dark
red solution of copper() trifluoromethanesulfonate-benzene com-
plex (28 mg, 0.1 mmol) and PyBox (10a; 15 mg, 0.05 mmol) in dry
CH2Cl2 (3 mL) to initiate the reaction. Then, a solution of 6
(200 mg, 1 mmol) and ethyl diazoacetate (140 µL, 1.3 mmol) in
CH2Cl2 (10 mL) was added over a period of 3.5 h and the dark red
mixture turned successively orange and then brown. At the end of
the addition, the solvent was removed under reduced pressure and
the remaining brown oil was purified by chromatography through
silica gel (petroleum ether/CH2Cl2, 60:40) affording 7aϪb as a col-
ourless oil (138 mg, 48%). The two diastereomers were then separ-
ated by chromatography over silica gel (2% ethyl acetate/petro-
leum ether).
Figure 3. Transition state model for the asymmetric cyclopropan-
ation of 6
Major Diastereomer (7a): ee 68%. (Rf ϭ 0.45; EtOAc/petroleum
ether, 95:5). IR (KBr): ν˜max ϭ 3069 cmϪ1, 3050, 2979, 2957, 2903,
2871, 1732, 1722, 1717 (CϭO), 1428, 1398, 1380, 1283, 1263, 1250,
1161, 1115, 1047, 943. 1H NMR (CDCl3): δ ϭ 7.51Ϫ7.57 (m, 2
H, Harom), 7.34Ϫ7.41 (m, 3 H, Harom), 5.83Ϫ5.89 (m, 1 H, 2-H),
5.46Ϫ5.51 (m, 1 H, 3-H), 4.10 (q, J ϭ 7.2 Hz, 2 H, CH2),
2.40Ϫ2.47 (m, 1 H, 1-H), 2.33Ϫ2.37 (m, 1 H, 4-H), 2.07Ϫ2.12 (m,
1 H, 5-H), 1.24 (t, J ϭ 7.2 Hz, 3 H, CH3), 0.88Ϫ0.91 (m, 1 H, 6-
H), 0.31 (s, 3 H, SiCH3), 0.29 (s, 3 H, SiCH3) ppm. 13C NMR
(CDCl3): δ ϭ 174.4 (CϭO), 137.1 (Carom), 133.9 (2 ϫ CHarom),
131.6 (C-3), 129.6 (C-2), 129.2 (CHarom), 127.8 (2 ϫ CHarom), 60.3
(CH2), 39.7 (C-4), 35.1 (C-1), 29.1 (C-6), 27.5 (C-5), 14.3 (CH3),
Ϫ5.2 (SiCH3), Ϫ5.5 (SiCH3) ppm. MS (CIϩ, NH3): m/z ϭ 304 [M
ϩ NH4]ϩ, 287 [M ϩ 1]ϩ, 209, 152 [M ϩ 1ϪPhMe2Si]ϩ. MS (EI):
m/z ϭ 135 (100) [PhMe2Si]ϩ. HRMS (EI) C17H22O2Si: calcd.
286.13891; found: 286.13800. [α]2D0 ϭ ϩ147 (c ϭ 0.27, MeOH).
Minor Diastereomer (7b): ee 68%. (Rf ϭ 0.40; EtOAc/petroleum
ether, 95:5). IR (KBr): ν˜max ϭ 3068 cmϪ1, 3050, 2979, 2956, 2901,
2871, 1738Ϫ1733 (CϭO), 1427, 1404, 1382, 1248, 1186, 1140, 1115,
1029, 951. 1H NMR (CDCl3): δ ϭ 7.52Ϫ7.57 (m, 2 H, Harom),
7.33Ϫ7.39 (m, 3 H, Harom), 5.63Ϫ5.69 (m, 1 H, 3-H), 5.54Ϫ5.60
(m, 1 H, 2-H), 4.03 (q, J ϭ 7.2 Hz, 2 H, CH2), 2.68Ϫ2.71 (m, 1 H,
Conclusion
In conclusion, we have developed an enantioselective ap-
proach towards cyclopropanes (i.e. 7), having a useful allyl-
silane moiety, which are valuable intermediates for organic
synthesis. As shown in earlier reports, synthons such as 7
can be elaborated further to cyclopentitols[2a] and carba-C-
disaccharides.[2b] Our approach also provides a more
straightforward access to recently discovered excitatory
amino acids.[16] Moreover, some mechanistic insights have
been obtained which might be useful for the design of more
efficient ligands for the cyclopropanation of cyclic as well
as acyclic (Z)-olefins.
Experimental Section
General Remarks: NMR spectra were performed in CDCl3 solu-
tions with chloroform as internal reference (1H: δ ϭ 7.26; 13C: δ ϭ
77.0 ppm). Structural assignments were based on DEPT, COSY,
HMQC, HMBC and NOESY experiments. 1D H and 13C NMR 4-H), 2.37Ϫ2.47 (m, 1 H, 6-H), 1.84Ϫ1.91 (m, 1 H, 5-H),
1
spectra were recorded on a Bruker AC 200 spectrometer at 200 and
50.3 MHz respectively. 2D NMR experiments were recorded on a
Bruker ARX 400 spectrometer (400 MHz). HPLC were performed
on a Waters 600 equipped with a 996 photodiode array detector
and a Chiralcel OD column. IR spectra of thin layers of samples
between KBr plates were obtained on a Bruker instrument. LRMS
1.61Ϫ1.69 (m, 1 H, 1-H), 1.20 (t, J ϭ 7.2 Hz, 3 H, CH3), 0.30 (s,
3 H, SiCH3), 0.29 (s, 3 H, SiCH3) ppm. 13C NMR (CDCl3): δ ϭ
169.5 (CϭO), 137.3 (Carom), 133.9 (2 ϫ CHarom), 133.7 (C-3), 129.1
(CHarom), 127.7 (2 ϫ CHarom), 124.2 (C-2), 59.7 (CH2), 35.1 (C-4),
32.0 (C-6), 23.3 (C-5), 20.7 (C-1), 14.3 (CH3), Ϫ4.9 (SiCH3), Ϫ5.6
(SiCH3) ppm. MS (CIϩ, NH3): m/z ϭ 304 [M ϩ NH4]ϩ, 287 [M
data were recorded on a HP mass spectrometer. HRMS data were ϩ 1]ϩ, 209, 152 [M ϩ 1ϪPhMe2Si]ϩ. MS (EI): m/z ϭ 135
1072 Eur. J. Org. Chem. 2003, 1069Ϫ1073

Desymmetrisation of Cyclopentadienylsilane by Asymmetric Cyclopropanation
([PhMe2Si]ϩ, 100). HRMS (EI) C17H22O2Si: calcd. 286.13891;
FULL PAPER
[3c]
Ito; T. Katsuki, Synlett 1993, 638Ϫ640.
T. Uchida; R. Irie;
T. Katsuki, Synlett 1999, 1793Ϫ1795. [3d] T. Uchida; R. Irie; T.
found: 286.13800.
[3e]
Katsuki, Synlett 1999, 1163Ϫ1165.
Catal. 2002, 344, 131Ϫ147.
T. Katsuki, Adv. Synth.
Ethyl 4-Hydroxybicyclo[3.1.0]hex-2-ene-6-carboxylate (12): A 70%
suspension of mCPBA (300 mg, 1.2 mmol) was added, at 0 °C, to a
suspension of silane 7a (290 mg, 1.0 mmol) and NaHCO3 (100 mg,
1.2 mmol) in CH2Cl2 (10 mL). The reaction was stirred at 0 °C for
1 h then allowed to warm to room temperature. After 6 h at room
temperature, the mixture was diluted with diethyl ether, washed
with a 20% solution of NaOH, dried over MgSO4 and concentrated
in vacuo. The resulting crude mixture was then diluted with CH2Cl2
and amberlite IR-120 (200 mg) was added to the solution. After
12 h at room temperature the mixture was filtered, concentrated
in vacuo and the resulting white solid was purified by silica gel
chromatography (petroleum ether/EtOAc, 90:10 Ǟ 80:20) to give
12 as a colourless oil (70 mg, 41%). 1H NMR (CDCl3): δ ϭ
6.20Ϫ6.18 (m, 1 H, ϭCH), 5.70Ϫ5.67 (m, 1 H, ϭCH), 4.60Ϫ4.50
(m, 1 H, CHOH), 4.12 (q, J ϭ 7 Hz, 2 H, OCH2CH3), 2.53Ϫ2.33
(m, 3 H), 1.39Ϫ1.13 (m, 3 H) ppm. Other spectroscopic data were
in good agreement with those reported in the literature.[16]
[4] [4a]
H. Fritschi; U. Letenegger; A. Pfaltz, Helv. Chim. Acta
[4b]
1988, 71, 1553Ϫ1565.
D. A. Evans; K. A. Woerpel; M. M.
Hinman; M. M. Foul, J. Am. Chem. Soc. 1991, 113, 726Ϫ728.
[4c]
R. E. Lowenthal; S. Masamune, Tetrahedron Lett. 1991,
32, 7373Ϫ7376.
[5a] H. Nishiyama; Y. Itoh; H. Matsumoto; S.-B. Park; K. Itoh,
[5]
[5b]
J. Am. Chem. Soc. 1994, 116, 2223Ϫ2224.
H. Nishiyama;
Y. Itoh; Y. Sugawara; H. Matsumoto; K. Aoki; K. Itoh, Bull.
Chem. Soc. Jpn. 1995, 68, 1247Ϫ1262. [5c] For the use of related
bis(pyrazolyl)pyridines, see: D. L. Christenson; C. J. Tokar; W.
B. Tolman, Organometallics 1995, 14, 2148Ϫ2150.
[6]
[7]
The moderate overall yield over the two steps is due to the
formation of several isomers in the silylation step (5 Ǟ 6)
which do not react in the second step.
It is noteworthy that attempts to perform this asymmetric
cyclopropanation using Doyle’s chiral Rh2(MEPY)4 led to no
enantioselectivity: M. P. Doyle, M. A. McKervey, T. Ye in
‘‘Modern Catalytic Methods for Organic Synthesis with Diazo
Compounds’’, John Wiley & Sons, New York, 1998.
S. L. Schreiber, T. S. Schreiber, D. B. Smith, J. Am. Chem. Soc.
1987, 109, 1525Ϫ1529.
[8]
Ethyl 4-Oxobicyclo[3.1.0]hex-2-ene-6-carboxylate (13): A solution
of DMSO (65 µL, 0.92 mmol) in CH2Cl2 (1 mL) was added, at
Ϫ60 °C, to a solution of oxalyl chloride (39 µL, 0.46 mmol) in
anhydrous CH2Cl2 (1 mL). After 10 minutes, a solution of the alco-
hol 12 (70 mg, 0.41 mmol) in dry CH2Cl2 (0.5 mL) was added drop-
wise at Ϫ60 °C and the mixture was stirred for 15 minutes. NEt3
(0.3 mL, 2.1 mmol) was then added to the solution. The reaction
mixture was stirred for 5 minutes at Ϫ60 °C, then allowed to warm
to room temperature and water (2 mL) was added. After 10 mi-
nutes, the organic layer was decanted then washed successively with
water, a 1% solution of HCl, a 5% solution of Na2CO3, water, dried
over MgSO4 and the solvents were concentrated in vacuo. The res-
idue was purified by gel chromatography (petroleum ether/EtOAc,
95:5) to give 13 as a white solid (62 mg, 90%). M.p. 97Ϫ99 °C. IR
(KBr): ν˜max ϭ 3070 cmϪ1, 2950, 1715, 1695, 1469, 1265, 1177, 876,
[9] [9a]
T.-L. Ho, ‘‘Symmetry: A Basis for Synthesis Design’’, John
Wiley & Sons, New York, 1995. [9b] M. C. Willis, J. Chem. Soc.,
Perkin Trans. 1 1999, 1765Ϫ1784.
I. Fleming, J. Dunogues, R. Smithers, Org. React. 1989, 37,
57Ϫ575.
[10]
[11]
`
During copper()- and rhodium()-mediated cyclopropana-
tions, decomposition of ethyl diazoacetate produces various
amount of diethyl fumarate and maleate along with the cyclo-
propane. These were absent in our experiment when using a
1:1 10a/CuI ratio.
[12] [12a]
C. Provent; S. Hewage; G. Brand; G. Bernardinelli; L. J.
Charboniere; A. F. Williams, Angew. Chem. Int. Ed. Engl. 1997,
36, 1287Ϫ1289. [12b] S. Rüttimann; C. Piguet; G. Bernardinelli;
B. Bocquet; A. F. Williams, J. Am. Chem. Soc. 1992, 114,
4230Ϫ4237.
[12c]
C. Piguet; G. Bernardinelli; G. Hopfgartner,
1
854, 818. H NMR (CDCl3): δ ϭ 7.61 [dd, J ϭ 5.7 Hz, 2.6 Hz, 1
Chem. Rev. 1997, 97, 2005Ϫ2062.
[13] [13a]
H, C(O)CHϭCH], 5.74 [d, J ϭ 5.7 Hz, 1 H, C(O)CHϭ], 4.15 (q,
J ϭ 7 Hz, 2 H, OCH2CH3), 2.96 (dt, J ϭ 4.8 Hz, 2.6 Hz, 1 H),
2.65Ϫ2.59 (m, 1 H), 2.27 (t, J ϭ 2.6 Hz, 1 H), 1.26 (t, J ϭ 7 Hz, 3
H, OCH2CH3) ppm. 13C NMR (CDCl3): δ ϭ 203.4 (CϭO), 168.0
(COOCH2CH3), 159.7 [C(O)CHϭCH), 129.7 [C(O)CHϭ), 61.4
(OCH2CH3), 45.9, 30.1, 29.0, 14.2 ppm. [α]2D0 ϭ Ϫ 135 (c ϭ 1,
MeOH) [ref.[16]: [α]2D0 ϭ Ϫ 255 (c ϭ 1, MeOH)]. Other spectroscopic
data were in good agreement with those reported in the literat-
ure.[16]
H. B. Kagan; C. Girard, Angew. Chem. Int. Ed. 1998, 37,
2922Ϫ2959. [13b] K. Mikami; M. Terada; T. Korenaga; Y. Mats-
umoto; M. Ueki; R. Angelaud, Angew. Chem. Int. Ed. 2000,
39, 3532Ϫ3556.
[14]
[15]
C.-X. Zhao; M. O. Duffey; S. J. Taylor; J. P. Morken, Org. Lett.
2001, 3, 1829Ϫ1831.
The relative configuration of the allylic alcohol 12 was assumed
to be cis as shown, resulting from an epoxidation which oc-
curred anti relative to the silicon group.[10]
[16] [16a]
[16b]
E. D. Moher, Tetrahedron Lett. 1996, 37, 8637Ϫ8640.
C. Dominguez; J. Ezquerra; L. Prieto; M. Espada; C. Pedregal,
Tetrahedron: Asymmetry 1997, 8, 511Ϫ514.
T. Rasmussen; J. F. Jensen; N. Ostergaard; D. Tanner; T. Zi-
egler; P.-O. Norrby, Chem. Eur. J. 2002, 8, 177Ϫ184.
Various Hammett studies and calculations on the mechanism
of copper-catalysed cyclopropanation have shown that the pu-
tative copper carbenoid species is electrophilic in nature.[17]
[17]
[18]
Acknowledgments
We thank the Region Aquitaine and the CNRS (COST-D12 pro-
gram) for financial support. We are also indebted to Dr. I. Pianet
for NMR experiments.
Similar conclusions were reached with copper-catalysed inser-
[18a]
tion of diazo ester into SiϪH bonds, see:
Y. Landais; L.
[1] [1a]
R. Angelaud, Y. Landais, J. Org. Chem. 1996, 61,
Parra-Rapado; D. Planchenault; V. Weber, Tetrahedron Lett.
[1b]
[18b]
5202Ϫ5203.
1997, 38, 8841Ϫ8844.
R. Angelaud, Y. Landais, Tetrahedron Lett.
1997, 38, 229Ϫ232.
L. A. Dakin; P. C. Ong; J. S. Panek;
[1c]
R. Angelaud, O. Babot, T. Charvat,
R. J. Staples; P. Stavropoulos, Organometallics 2000, 19,
Y. Landais, J. Org. Chem. 1999, 64, 9613Ϫ9624.
2896Ϫ2908.
[2] [2a]
[19]
Y. Landais, L. Parra-Rapado, Eur. J. Org. Chem. 2000, 2,
J. B. Lambert, Tetrahedron 1990, 46, 2677Ϫ2689 and references
[2b]
401Ϫ418.
R. Angelaud, Y. Landais, L. Parra-Rapado, Tet-
cited therein.
rahedron Lett. 1997, 38, 8845Ϫ8848.
Received October 8, 2002
[O02554]
[3] [3a]
[3b]
T. Aratani, Pure Appl. Chem. 1985, 57, 1839Ϫ1844.
K.
Eur. J. Org. Chem. 2003, 1069Ϫ1073
1073