cis-Selective cyclopropanations using chiral 5,5-diaryl bis(oxazoline) catalysts

Article abstract of DOI:10.1016/S0040-4039(00)01232-6

A series of 5,5-diaryl bis(oxazolines) have been prepared and used as ligands in the copper catalysed asymmetric cyclopropanation of styrene. Unusually, for bis(oxazolines), the diastereoselectivity of the process favoured the cis cyclopropanes with diastereomeric ratios of up to 65:35 being realised. The cis selectivity of the process was rationalised in terms of repulsion between the alkene substituent and the pro-R 5-aryl group of the bis(oxazoline). (C) 2000 Published by Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4039(00)01232-6

Tetrahedron Letters 41 (2000) 7135±7138
cis-Selective cyclopropanations using chiral 5,5-diaryl
bis(oxazoline) catalysts
a,y
b
a,
Karen Alexander, Stuart Cook and Colin L. Gibson *
aDepartment of Pure & Applied Chemistry, University of Strathclyde, 295 Cathedral Street, Glasgow G1 1XL, UK
bHickson & Welch Ltd, Wheldon Road, Castleford, West Yorks WF10 2JJ, UK
Received 18 July 2000; accepted 21 July 2000
Abstract
A series of 5,5-diaryl bis(oxazolines) have been prepared and used as ligands in the copper catalysed
asymmetric cyclopropanation of styrene. Unusually, for bis(oxazolines), the diastereoselectivity of the
process favoured the cis cyclopropanes with diastereomeric ratios of up to 65:35 being realised. The cis
selectivity of the process was rationalised in terms of repulsion between the alkene substituent and the
pro-R 5-aryl group of the bis(oxazoline). # 2000 Published by Elsevier Science Ltd.
Over the last decade a wide range of chiral bis(oxazoline)±metal complexes have been used to
1
catalyse an impressive array of asymmetric transformations. These processes include asymmetric
cyclopropanation and aziridination of alkenes, Diels±Alder and cycloadditions, Mukaiyama
aldol reactions, allylic substitutions, oxidations and hydrosilylation reactions. One of the most
studied procedures has been the bis(oxazoline)±copper and bis(oxazoline)±ruthenium catalysed
2
asymmetric cyclopropanation of alkenes. In general, the procedure is highly diastereoselective
for the formation of the trans cyclopropanes (trans:cis ratios ca. 68:32 to 100:0) and exceptionally
high ee's have been realised (up to 99% ee). Access to the cis cyclopropanes has been more elusive
but very recently there has been growing interest in the development of cis diasteroselective
processes. These methods have involved alkene cyclopropanations with diazoacetate catalysed by
3
chiral complexes including copper(I) bis(oxazolines) derived from tartrate, metal(II)±salen
4
5
6
complexes, dirhodium(II) carboximides and dirhodium(II) carboxylates.
We have previously shown that 5,5-diaryl substituents in chiral oxazolidinone auxiliaries can
7
result in bene®cial diastereoselectivity and chemical eciency. As part of a programme aimed at
generating cis cyclopropanes, we decided to investigate such 5,5-diaryl substitution in bis(oxazoline)
catalysed cyclopropanations. At the outset, we were aware that Singh and co-workers had
*
y
Corresponding author. E-mail: c.l.gibson@strath.ac.uk
nee Gillon.
Â
0040-4039/00/$ - see front matter # 2000 Published by Elsevier Science Ltd.
PII: S0040-4039(00)01232-6

7
136
observed very poor enantioselectivities (< 40% ee) in the asymmetric trans-selective cyclopropanation
8
9
of styrene using 5,5-disubstituted pybox catalysts. More recently, Singh and co-workers have
used 5,5-disubstituted pybox catalysts in enantioselective allylic oxidations while Andrus et al.10
have utilised 5,5-dimethyl bis(oxazolines) in the same processes. Originally, Corey et al. employed
5
,5-dimethyl bis(oxazolines) in the asymmetric Diels±Alder reaction.11
The required starting tertiary amino alcohols 1a±f were prepared by the reaction of excess
7,12
Grignard reagent with the appropriate amino ester hydrochloride salt.
Using the methodology
of Denmark et al.13 the tertiary amino alcohols 1a±f were converted into the diamides 2a±f by
treatment with dimethylmalonyl chloride and triethylamine (Scheme 1). The subsequent trans-
formation of diamides 2a±f into bis(oxazolines) 3a±f initially proved problematic. Using the
1
1
method of Corey for the cyclisation of tertiary alcohol diamides with methanesulfonic acid and
azeotropic removal of water gave the mono cyclised mono dehydrated product. The use of
13
thionyl chloride in toluene at re¯ux resulted in the dehydrated product being formed.
Gratifyingly, treatment of diamides 2a±c and 2e,f with thionyl chloride at room temperature gave
the appropriate bis(oxazolines) 3a±c,e,f. However, similar treatment of diamide 2c gave a
complex mixture of products. Fortunately, diamide 2c could be smoothly converted into the
bis(oxazoline) 3c with zinc chloride in dichloroethane at re¯ux.
i
i
t
Scheme 1. (a) R=Pr , Ar=4-tolyl; (b) R=Pr , Ar=2-naphthyl; (c) R=Bu , Ar=4-tolyl; (d) R=Ph, Ar=4-tolyl; (e)
R=CH Ph, Ar=4-tolyl; (f) R=CH Ph, Ar=2-naphthyl
2
2
The ecacy of 1:1 mixtures of bis(oxazolines) 3 and appropriate copper salts to catalyse the
asymmetric cyclopropanation of styrene 4 with diazoesters 5 was investigated (Scheme 2, Table 1).
Initially, a wide range of reaction parameters were investigated for the isopropyl 5,5-ditolyl
bis(oxazoline) 3a (entries 1±7, Table 1). In the cases using ethyl diazoacetate as the carbene source
there is a preference (up to 65:35) for the cis cyclopropane 6a over the trans cyclopropane 7a
(
entries 1±5, Table 1). The use of tert-butyl diazoacetate 5b as the carbene source results in a
diminishing of the cis selectivity (ca 1:1) but an increase in the enantioselectivity of the process
entries 6 and 7).
(
Scheme 2.
Clearly, the structure of the bis(oxazoline) ligand in¯uences the diastereoselectivity of the
cyclopropanation. This is in contrast to the observations of Pfalz et al. for the semicorrin
catalysed asymmetric cyclopropanations where the trans diastereoselectivity was attributed solely
to interactions between the alkene and the carbene ester group.14 However, we believe that the

7
137
Table 1
Asymmetric cyclopropanation of styrene 4 using 1 mol% of bis(oxazolines) 3
14
general mechanistic protocol is similar to that delineated by Pfalz et al. and extended by Singh
9
and co-workers. Here the enantioselectivity can be understood from the attack of the Re face of
the electrophilic carbenoid by the alkene. This avoids steric repulsion at the transition state
between the carbene ester group and the C-4 substituent at the stereogenic centre. However, to
understand the unusual cis diastereoselectivity we carried out a molecular mechanics based
modelling study.1 This modelling suggested that the carbene is distorted from being orthogonal
to the bis(oxazoline) ring and this further favours alkene attack at the Re face of the carbene.
Additionally, there is potentially an unfavourable steric interaction between the styrene phenyl
group and the pro-R aryl group at C-5 for ul addition which leads to the trans cyclopropane 7.
This adverse steric interaction is avoided for the lk addition and results in the preferential
formation of the cis cyclopropane 6. However, this adverse steric interaction is counter balanced
by a steric repulsion between the styrene phenyl and the ester alkyl group on the carbene in the
favourable lk addition. Thus, the use of tert-butyl diazoacetate 5b results in an increase in the
proportion of the trans isomer 7.
5
2
In contrast to previously reported bis(oxazoline) catalysis with copper(II) salts, the use of
bis(oxazoline) 3a with copper(II) tri¯ate did not require heat or the addition of phenylhydrazine
to initiate the carbene insertion reaction (entries 1±3 and 6). However, in the case of bis(oxazoline)
3
a, the best selectivities were obtained using copper(I) tri¯ate when the reaction was carried out at
ꢀ
0
C. Under these conditions the use of ethyl diazoacetate 5a a€orded a 63:37 mixture of cis
cyclopropane 6a and trans cyclopropane 7a in ee's of 63% and 64%, respectively (entry 5).
Similar conditions with tert-butyl diazoacetate 5b a€orded the corresponding cis 6b and trans 7b
products in a 1:1 ratio with ee's of 73 and 81%, respectively (entry 7).
All the initial reactions with bis(oxazoline) 3a were conducted using 1 mol% of the catalyst
system. However, changing the substrate:catalyst ratio to 1000:1 does not appear to compromise
the stereochemical outcome of the reaction, but results in a lower yield (entry 2).
The in¯uence of the steric requirements at C-4 and C-5 in the enantioselective cyclopropanation
of styrene 4 using copper(I) tri¯ate and ethyl diazoacetate 5a was investigated. These results
indicated that increasing the steric requirements of the 5,5-substituents resulted in little change in

7
138
the cis:trans ratio but diminished the enantioselectivity (entries 5 and 8). However, for the 4-
benzyl bis(oxazolines) 3e and 3f there is an increase in both the enantioselectivity and the cis
diastereoselectivity on increasing the steric demand at the 5-positions (entries 11 and 12). On the
other hand, increasing the steric requirements at the stereogenic centres resulted in a decrease in
the cis diastereoselectivity (entries 5, 9 and 10)
Acknowledgements
We are grateful to Hickson & Welch for funding (K.A.). We also thank Dr Terry Lowdon
Hickson & Welch) for valuable discussions.
(
References
1
. For a review of the use of chiral bis(oxazoline)±metal complexes in catalytic asymmetric syntheses see: Ghosh,
A. K.; Mathivanan; Cappiello, J. Tetrahedron: Asymmetry 1998, 9, 1±45.
2
. For reviews of asymmetric cyclopropanations see: Doyle, M. P.; Protopopova, M. N. Tetrahedron 1998, 54, 7919±
7
945. Doyle, M. P.; McKervey, M. A. Chem. Commun. 1997, 983±989. Singh, V. K.; DattaGupta, A.; Sekar, G.
Synthesis 1997, 137±150.
3
4
. Boulch, R; Scheurer, A.; Mosset, P.; Saalfrank, R. W. Tetrahedron Lett. 2000, 41, 1023±1026.
. Niimi, T.; Uchida, T.; Irie, R.; Katsuki, T. Tetrahedron Lett. 2000, 41, 3647±3651. Uchida, T.; Irie, R.; Katsuki, T.
Tetrahedron 2000, 56, 3501±3509. Uchida, T.; Irie, R.; Katsuki, T. Synlett. 1999, 1163±1165.
. Doyle, M. P.; Davies, S. B.; Hu, W. Chem. Commun. 2000, 867±868 and references cited therein.
. Ishitani, H.; Achiwa, K. Synlett. 1997, 781±782.
5
6
7
8
9
. Gibson, C. L.; Gillon, K.; Cook, S. Tetrahedron Lett. 1998, 39, 6733±6736.
. DattaGupta, A.; Bhuniya, D.; Singh, V. K. Tetrahedron 1994, 50, 13725±13730.
. DattaGupta, A. D.; Singh, V. K. Tetrahedron Lett. 1996, 37, 2633±2636. Sekar, G.; DattaGupta, A. D.; Singh,
V. K. J. Org. Chem. 1998, 63, 2961±2967.
1
0. Andrus, M. B.; Argade, A. B.; Chen, X.; Pamment, M. G. Tetrahedron Lett. 1995, 36, 2945±2948. Andrus, M. B.;
Chen, X. Tetrahedron 1997, 53, 16229±16240.
1. Corey, E. J.; Imai, N.; Yhang, H.-Y. J. Amer. Chem. Soc. 1991, 113, 728±729.
1
1
1
1
1
2. Itsuno, S.; Nakano, K.; Miyazaki, K.; Masuda, H.; Ito, K. J. Chem. Soc. Perkin Trans. 1 1985, 2039±2044.
3. Denmark, S. E.; Nakajima, N.; Nicaise, O. J.-C.; Faucher, A.; Edwards, J. P. J. Org. Chem. 1995, 60, 4884±4892.
4. Pfaltz, A. In Modern Synthetic Methods; Sche€old, R., Ed. Springer: Berlin-Heidelberg, 1989; Vol. 5, pp. 199±248.
5. Molecular mechanics calculations were carried out using PCModel 5.03, Serena Software, Bloomington, Indiana.