The origin of diastereoselectivity in cyclopropanation reactions of the iron carbene complex [(η5-C5H5)(CO)2Fe=CH[(η6-o-CH3OC6H4)Cr(CO)3]](1+)

Article abstract of DOI:10.1016/S0022-328X(00)00731-2

The origin of diastereoselectivity in cyclopropanation reactions of the iron carbene complex [(η⁵-C5H5)(CO)2Fe=CH[(η⁶-o-CH3OC6H4)Cr(CO)3]](1+) has been investigated. The chromium participation has been proved by the upfield shifts of both 1H- and 13C-NMR for C_α-H and C_α, which stabilized the iron carbene and may result in a late transition-state for the cyclopropanation reaction via tricarbonylchromium-complexed iron carbene. Hence, trans selectivity is expected. However, when using aromatic alkenes, a ? stacking effect may exist during the transition-state, resulting in the cis selectivity for cyclopropanation.

Full text of DOI:10.1016/S0022-328X(00)00731-2

www.elsevier.nl/locate/jorganchem
Journal of Organometallic Chemistry 617–618 (2001) 751–754
Communication
The origin of diastereoselectivity in cyclopropanation reactions of
the iron carbene complex
[(h5-C5H5)(CO)2FeꢀCH[(h6-o-CH3OC6H4)Cr(CO)3]]+
Qinwei Wang, M. Mahmun Hossain *
Department of Chemistry, Uni6ersity of Wisconsin-Milwaukee, Milwaukee WI 53201, USA
Received 11 September 2000; accepted 25 September 2000
Abstract
The origin of diastereoselectivity in cyclopropanation reactions of the iron carbene complex [(h5-C5H5)(CO)2FeꢀCH[(h6-o-
1
CH3OC6H4)Cr(CO)3]]+ has been investigated. The chromium participation has been proved by the upfield shifts of both H- and
13C-NMR for CaꢁH and Ca, which stabilized the iron carbene and may result in a late transition-state for the cyclopropanation
reaction via tricarbonylchromium-complexed iron carbene. Hence, trans selectivity is expected. However, when using aromatic
alkenes, a p stacking effect may exist during the transition-state, resulting in the cis selectivity for cyclopropanation. © 2001
Elsevier Science B.V. All rights reserved.
Keywords: Iron carbene; Cyclopropanation; Transition-state; Diastereoselectivity; p Stacking
Due to continuing interest in asymmetric cyclopropa-
nation reactions via chiral iron carbene complexes, we
have employed tricarbonylchromium-complexed iron
This result triggered our interest to explore the origin of
cis/trans selectivity in cyclopropanation of alkenes via
carbene 1. Here, we would like to report our exciting
preliminary findings in the origin of the diastereoselec-
tivity in cyclopropanation reactions with tricar-
bonylchromium-complexed iron carbenes (see Fig. 1).
Cyclopropanes were prepared by the reaction of iron
carbene 1 with corresponding alkenes using the proce-
dure described earlier (Scheme 1) [1]. The resulting
cis/trans ratios for the cyclopropanes 5 are listed in
Table 1. With styrene and styrene-derivatives carbene 1
produced, primarily, cis cyclopropanes (entry 2–5).
With a-methyl styrene carbene 1 also gave high
cis:trans ratio (entry 1), whereas, with buta-1,3-diene
and propene carbene 1 produced mainly trans cyclo-
propanes (entry 6–7).
carbenes
R
and
S
[(h5-C5H5)(CO)2FeꢀCH[(h6-o-
CH3OC6H4)Cr(CO)3]]+ (1) successfully for excellent
asymmetric induction, and also expanded the scope of
these chiral-at-carbene-ligand iron carbenes by reacting
them with a series of styrene derivatives [1]. The car-
bene 1 reacted with styrene derivatives gave mainly
cis-cyclopropanes. This was consistent with the dicar-
bonyl iron carbene complexes (Cp(CO)2FeꢀCHR+)
which are known to exhibit high cis selectivity in car-
bene transfer reactions to alkenes [2]. In contrast, we
recently observed that the cyclopropanation reaction of
carbene 1 with propene gave predominantly trans cy-
clopropane (cis:trans, 1:7), whereas, the uncomplexed
iron carbene 2 gave mostly cis cyclopropane from
propene (cis:trans, 5:1). The reaction of propene with
dicarbonyl iron carbene complexes is a well-docu-
mented reaction and has always provided cyclo-
propanes with high cis selectivity [2], therefore, the
observed trans selectivity with carbene 1 is unusual.
* Corresponding author.
Fig. 1.
0022-328X/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved.
PII: S0022-328X(00)00731-2

752
Q. Wang, M.M. Hossain / Journal of Organometallic Chemistry 617–618 (2001) 751–754
Scheme 1.
In order to explain the observed cis/trans selectivity
Based on the proposed model, for the less reactive
iron carbene 1 the transition state will be late and trans
selectivity is expected (Fig. 4). Indeed, the reaction of
carbene 1 with propene gave a 1:7 cis:trans ratio,
suggesting a late transition-state. While with the more
reactive uncomplexed carbene 2, the transfer to
propene gave a 5:1 cis:trans ratio via an early transition
state.
in cyclopropanation reactions involving iron carbenes,
a transition-state model was proposed by Brookhart
(Fig. 2) [3]. According to this model, for more reactive
dicarbonyl systems Cp(CO)2FeꢀCHR+, the transition-
state is early as shown in 6. Site 2 would be sterically
less crowded than site 1 for substituents, resulting in cis
selectivity for monosubstituted alkenes [3]. In the late
transition-state model 7, (with less reactive carbene)
substituents will prefer site 1 to site 2, and trans selec-
tivity will be exhibited in cyclopropanation reactions.
In order to comprehend the origin of diastereoselec-
tivity in the cyclopropanation of the carbene 1 employ-
ing the proposed transition-state model, we need to
determine whether the transition-state of carbene trans-
fer is ‘early’ or ‘late’. It has been reported that in the
If the ‘late’ transition state assumption for carbene
transfer reaction with 1 is true, it is surprising to notice
that the carbene transfers to styrene and its derivatives
Table 1
Ratio of cyclopropanes 5 from the reaction of carbene 1 with alkenes
chromium-complexed
benzylic
carbocation
the
chromium does participate with the positive charge at
Ca carbon [4] and this was confirmed by the spectro-
scopic evidence by Olah [5]. Substantial upfield shifts in
13C- and 1H-NMR for Ca and CaꢁH of chromium-com-
plexed benzylic carbocation relative to uncomplexed
benzylic carbocation were attributed to the strong par-
ticipation of chromium [5]. Similar upfield shifts of 13C-
1
and H-NMR for Ca and CaꢁH of carbene 1 relative to
2
clearly demonstrate a strong participation of
chromium with the positive charge of carbene carbon
(structure 1%, Fig. 3) [6]. It is also reported that the
chromium participation stabilizes the benzylic carboca-
tion and makes chromium-bound carbocations less re-
active, and, in some cases, completely inhibits their
reactivities [7]. Similarly, chromium complexation could
also make the carbene 1 more stable and less reactive.
Experimentally, it was observed that the characteristic
purple color of the reaction mixture, assumed to be
carbene 1, persisted for a longer time than that of the
corresponding uncomplexed carbene 2.

Q. Wang, M.M. Hossain / Journal of Organometallic Chemistry 617–618 (2001) 751–754
753
Fig. 4.
Fig. 2.
all gave mostly cis cyclopropanes. This raises the ques-
tion, what factor is contributing to the cis selectivity in
cyclopropanation of less reactive chromium-complexed
iron carbene 1? The answer may be the p–p interaction
of two phenyl groups during the transition-state, which
could keep them cis to each other. It is known that the
tricarbonylchromium-complexed arene exhibits p–p in-
teraction with another p system referred as the ‘p
stacking’ effect [8]. Such p–p interaction between the
two arene groups is possible during the cyclopropana-
tion reaction involving carbene 1 and aromatic alkenes.
This interaction could stabilize the transition-state, re-
sulting in cis selectivity in cyclopropanation (Fig. 5). If
our assumption is true, then the reduction of the p
stacking effect by an electron-withdrawing group on
styrene could also reduce the cis selectivity. Indeed, the
cis:trans ratio goes down from 10:1 (styrene) to 6:1
(p-chlorostyrene), to 3:1 (p-trifluoromethylstyrene). The
cis:trans ratios of cyclopropanes made from p-methyl-
styrene (7:1) and from p-trifluoromethylstyrene (3:1)
clearly indicate that the selectivity depends mainly on
the electronic rather than the steric factor. Recently, the
electronic factor was also found to be responsible for
the syn selectivity in the cyclopropanation reactions of
(phosphanyl)(silyl)carbene with styrene and styrene
derivatives [9].
Fig. 5.
its cis partner due to the p stacking effect. The resulting
very high cis:trans ratio (30:1) is consistent with this
prediction. When carbene 1 was reacted with buta-1,3-
diene, a lower cis/trans selectivity was observed (1:3)
than for styrene and its derivatives (entries 1–5). This
may be because the p stacking effect in this case (only
one CꢀC double bond available) is much weaker than
that of arenes. Furthermore, when no p stacking is
possible as in the case of propene, the trans-selectivity
was observed as expected by the late transition-state
model.
In summary, the origin of the diastereoselectivity in
the reaction of carbene 1 with various alkenes has been
investigated. Due to the participation of chromium, the
carbene 1 is quite stable and less reactive, resulting in a
late transition-state in the carbene transfer reaction.
Hence, according to the transition-state-model trans
selectivity with monosubstituted non-aromatic alkenes
is expected. However, with aromatic alkenes a strong
p-stacking effect may exist during the transition-state,
which results in the cis selectivity. Further kinetic and
spectroscopic studies are underway and will be pub-
lished in the near future.
In order to further verify the existence of p stacking
in the transition-state, we have tested the two other
alkenes, a-methyl styrene (entry 1) and buta-1,3-diene
(entry 6). In the reaction of carbene 1 with a-methyl
styrene, it is predictable that the tricarbonyl chromium
arene would obviously select Ph over a CH3 group as
Fig. 3.

754
Q. Wang, M.M. Hossain / Journal of Organometallic Chemistry 617–618 (2001) 751–754
[3] M. Brookhart, Y. Liu, E.W. Goldman, D.A. Timmers, G.D.
Acknowledgements
Williams, J. Am. Chem. Soc. 113 (1991) 927.
[4] (a) S.G. Davies, T.D. McCarthy, in: F.G. Stone, G. Wilkinson
(Eds.), Comprehensive Organometallic Chemistry II, vol. 12,
Pergamon, New York, 1995, p. 1039. (b) M. Acampora, A.
Ceccon, M.D. Farra, G. Giacometti, G. Rigatti, J. Chem. Soc.
Perkin Trans. 2 (1977) 483. (c) P.A. Downton, B.G. Sayer, M.
McGlinchey, J. Organometallics 11 (1992) 3281.
Financial support for this research was provided by
the PRF. We thank Dr. Frank Holger Foersterling for
technical assistance in obtaining the NMR data.
[5] G.A. Olah, S.H. Yu, J. Org. Chem. 41 (1976) 1694.
[6] For carbene 1: 1H-NMR l: 13.7/13.3 (FeꢀCH), 13C-NMR l:
275.8/258.8 (Fe=CH); For uncomplexed carbene 2: 1H-NMR l:
16.3 (FeꢀCH), 13C-NMR l: 320.1 (FeꢀCH).
[7] S. Top, G. Jaouen, J. Org. Chem. 46 (1981) 78.
[8] (a) R.H. Mitchell, T.K. Vinod, G.W. Bushnell, J. Am. Chem. Soc.
112 (1990) 3487. (b) G.B. Jones, B.J. Chapman, J.E. Mathews, J.
Org. Chem. 63 (1998) 2928.
[9] (a) D. Bourissou, O. Guerret, F.P. Gabba¨ı, G. Bertrand, Chem.
Rev. 100 (2000) 39. (b) S. Goumri-Magnet, T. Kato, H. Gor-
nitzka, A. Baceiredo, G. Bertrand, J. Am. Chem. Soc. 122 (2000)
4464.
References
[1] (a) Q. Wang, M.F. Mayer, C. Brennan, F. Yang, M.M. Hossain,
D.S. Grubisha, D. Bennett, Tetrahedron 56 (2000) 4881. (b) Q.
Wang, F. Yang, H. Du, M.M. Hossain, D. Bennett, D.S. Gru-
bisha, Tetrahedron Asymmetry 9 (1998) 3971. (c) R.D. Theys,
M.M. Hossain, Tetrahedron Lett. 36 (1995) 5113.
[2] (a) M. Brookhart, W.B. Studabaker, Chem. Rev. 87 (1987) 411.
(b) R.D. Theys, R.M. Vargas, Q. Wang, M.M. Hossain,
Organometallics 17 (1998) 1333.
.