Probing the mechanism of the anomalous intramolecular C-H insertion reaction of rhodium carbenoids by analysis of kinetic isotope effects

Article abstract of DOI:10.1016/S0040-4039(01)01173-X

The mechanism of the anomalous intramolecular insertion reaction of a rhodium carbenoid into an ethereal C-H bond has been explored using deuterium labelled substrates. Comparison of primary kinetic isotope effects and analysis of ratios of diastereoisome

Full text of DOI:10.1016/S0040-4039(01)01173-X

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 6187–6190
Probing the mechanism of the anomalous intramolecular C–H
insertion reaction of rhodium carbenoids by analysis
of kinetic isotope effects
J. Stephen Clark,* Yung-Sing Wong and Robert J. Townsend
School of Chemistry, University of Nottingham, University Park, Nottingham NG7 2RD, UK
Received 6 June 2001; accepted 28 June 2001
Abstract—The mechanism of the anomalous intramolecular insertion reaction of a rhodium carbenoid into an ethereal C–H bond
has been explored using deuterium labelled substrates. Comparison of primary kinetic isotope effects and analysis of ratios of
diastereoisomeric deuterated products suggests the reactions leading to the anomalous and conventional C–H insertion products
are mechanistically distinct and do not share a common rate-determining step. © 2001 Elsevier Science Ltd. All rights reserved.
We recently reported that anomalous products are
obtained during the preparation of 3(2H)-furanones by
intramolecular C–H insertion of rhodium carbenoids
into the a-CꢀH bonds of ethers.1 For example, the
unusual acetal 4 and the conventional C–H insertion
product 5 were both obtained when the diazoketone 1
was exposed to a variety of rhodium(II) catalysts
(Scheme 1). The formation of the anomalous product 4
suggests that a zwitterionic intermediate (3) is generated
by hydride migration to the carbenoid. In principle, the
ionic intermediate 3 could also serve as a precursor of
the conventional C–H insertion product 5 allowing the
reaction to circumvent the generally postulated transi-
tion states, which involve H-migration to rhodium with
concomitant CꢀC bond formation.2
At the time we first reported the isolation of anomalous
products from intramolecular insertion reactions of
rhodium carbenoids into the a-CꢀH bonds of ethers, we
believed that compounds of this type had not been
isolated from carbenoid C–H reactions.1 However, we
subsequently discovered that Mander and Owen had
isolated an analogous product from a copper carbenoid
C–H insertion reaction.3,4 Following our report, White
and Hrnciar isolated several anomalous products from
intramolecular rhodium carbenoid C–H insertion reac-
tions which appear to be mechanistically related to
those described by Mander and by us.5 It is clear from
these results that more than one reaction pathway is
operative during the insertion of a metal carbenoid into
an ethereal CꢀH bond. In order to gain a better mech-
RhLn
O
H
H
O
hydride migration
N2
O
RhLn
O
O
δ+
O
O
Rh2L4
H
H
3
4
5
?
O
RhLn
H
RhLn
H
O
O
O
1
2
or
direct C−H
insertion
O
O
O
H
H
Scheme 1.
Keywords: rhodium carbenoid; C–H insertion; deuterium; kinetic isotope effects.
* Corresponding author. Tel.: +44 (0)115 9513542; fax: +44 (0)115 9513564; e-mail: j.s.clark@nottingham.ac.uk
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)01173-X

6188
J. S. Clark et al. / Tetrahedron Letters 42 (2001) 6187–6190
anistic understanding of the anomalous C–H insertion
reaction and identify factors which might influence the
reaction, we prepared deuterium labelled derivatives of
the diazoketone 1 and investigated the product distribu-
tion obtained upon rhodium carbenoid generation.
and 10a obtained from the rhodium(II) trifluoroacetate
mediated reaction of the diazoketone 6a was similar but
not identical to the ratio of conventional C–H insertion
products (7a and 8a) obtained from the same reaction.
In contrast, the ratio of isomeric acetals 9b and 10b
obtained from the rhodium(II) trifluoroacetate medi-
ated reaction of the diazoketone 6b was much higher
than the ratio of the ketones 7b and 8b obtained from
the same reaction.
The simple diazoketone 1 was selected as the precursor
because we had found that a significant amount of the
anomalous C–H insertion product 4 could be obtained
by treating it with an appropriate rhodium catalyst. In
1
In subsequent studies, cyclisation reactions of the
deuterated diazoketone 6c were analysed and the deu-
terium isotope effect in both the conventional and
anomalous C–H insertion reactions were calculated
(Scheme 3). Previous work by Wang and Adams had
shown that a kinetic isotope effect (kH/kD) of 1.2–2 was
likely for the conventional intramolecular insertion
addition, there were no overlapping signals in the H
NMR spectrum of the anomalous C–H insertion
product 4 or in that of the conventional C–H insertion
product 5 (Fig. 1). This greatly simplified peak assign-
ment and meant that accurate integration of signals for
individual protons could be performed.
H
δ 4.29
δ 2.23
H
O
H
δ 3.86
H
δ 2.46
O
H
δ 4.04
δ 3.82
δ 3.94
H
H
H
δ 4.34
δ 4.50
H
δ 4.96
O
O
H
4
5
1
Figure 1. The H NMR assignments for the anomalous and conventional C–H insertion products.
In preliminary studies, the deuterium labelled substrates
6a and 6b were prepared and treated with a sub-stoi-
chiometric amount of either rhodium(II) acetate or
rhodium(II) trifluoroacetate in dichloromethane at
room temperature (Scheme 2). The resulting intra-
molecular insertion reactions delivered mixtures of the
ketones 7 and 8 and the acetals 9 and 10. The
ketone and acetal products were separated by chro-
matography and the ratio of products within each
reaction of a rhodium carbenoid into a deuterated
ether,6 and Sulikowski and Lee had obtained kinetic
isotope effects of similar magnitude during
a
intramolecular insertion of rhodium and copper car-
benoids into deuterated tertiary amines.7 The diazoke-
tone 6c was treated with rhodium(II) acetate or
rhodium(II) trifluoroacetate in CH2Cl2 at room temper-
ature. 1H NMR analysis of the conventional C–H
insertion products obtained from the rhodium(II) ace-
tate mediated reaction indicated a 65:35 ratio of the
diastereoisomers 7b and 8b and a kinetic isotope effect
(11/[7b+8b]) of 1.2. Unfortunately, it was not possible
to isolate sufficient quantities of the products (9b, 10b
and 12) arising from the anomalous C–H insertion to
compare the kinetic isotope effect with that of the
conventional rhodium(II) acetate mediated reaction.
1
reaction manifold was determined using H NMR spec-
troscopy. Analysis of the spectra of the mixtures of
conventional C–H insertion products 7 and 8 revealed
that the ketone 7 predominated in all cases and that
there was a modest but significant dependence of
product ratio on the catalyst used. In the case of the
anomalous products, the ratio of isomeric acetals 9a
X
Y
N2
X
Y
O
O
Y
X
Y
X
O
X
O
O
Rh2L4
Y
Y
+
+
CH2Cl2, rt
O
O
O
O
O
Y
Y
Y
Y
6
7
8
9
10
7:8
7/8
9:10
9/10
Rh2(O2CCH3)4
Rh2(O2CCF3)4
a X = H, Y = D
b X = D, Y = H
77:23
71:29
3.3
2.4
75:25
3.0
75:25a
68:32b
3.0a
2.1b
Rh2(O2CCH3)4
Rh2(O2CCF3)4
88:12b
7.1b
a Diazoketone 6b had 96% deuterium incorporation (ratio corrected)
b Diazoketone 6b had 91% deuterium incorporation (ratio corrected)
Scheme 2.

J. S. Clark et al. / Tetrahedron Letters 42 (2001) 6187–6190
6189
D
H
O
O
O
H
D
+
+
O
O
O
H
H
D
7b
30
33
8b
16
17
11
N2
H D
Rh2(O2CCH3)4
Rh2(O2CCF3)4
54
50
kH/kD = 1.2
kH/kD = 1.0
O
Rh2L4,
CH2Cl2, rt
O
D
H
H
H
6c
D
H
O
O
O
+
+
O
O
O
H
H
D
9b
10b
12
Rh2(O2CCF3)4
4
39
57
kH/kD = 1.3
Scheme 3.
However, both sets of products could be isolated from
the rhodium(II) trifluoroacetate catalysed reaction. In
this case, the diastereoisomers 7b and 8b were obtained
in a 66:34 ratio and the kinetic isotope effect (11/[7b+
8b]) was 1.0. In contrast, the isomeric anomalous prod-
ucts 9b and 10b were obtained as a 9:91 mixture and
the kinetic isotope effect (12/[9b+10b]) was 1.3.
finding which is indicative of the non-linear transition
state required for intramolecular H-transfer.10
In summary, our results demonstrate that the conven-
tional and anomalous C–H insertion products 5 and 4
are not produced by a common rate-determining step
and that the ionic intermediate 3 is unlikely to be an
intermediate prior to formation of the acetal 4. Further
studies to elucidate the mechanism of the reaction
leading to anomalous C–H insertion products are in
progress.
We also explored the temperature and solvent depen-
dence of the reaction. The ratios of products obtained
from the rhodium(II) trifluoroacetate mediated reaction
in dichloromethane were identical at 0°C, room temper-
ature or reflux.8 When the same reaction was performed
in 1,2-dichloroethane the ratio of products differed
from those obtained in dichloromethane, but there was
little variation in the relative ratios of products within
each reaction manifold between 0 and 83°C.9
Acknowledgements
We are grateful to the EPSRC for financial support
(Grant no. GR/L62856).
Several important findings emerge from our results.
Firstly, there is a modest but significant difference
between the kinetic isotope effects for the conventional
and anomalous C–H insertion reactions of diazoketone
6c mediated by rhodium(II) trifluoroacetate. This
finding rules out a common rate determining CꢀH bond
cleavage step for the conventional and anomalous C–H
insertion reactions (Scheme 1). Secondly, the high level
of diastereocontrol (9b:10b, 9:91) obtained from the
anomalous C–H insertion reaction contrasts with the
modest level of diastereocontrol (7b:8b, 66:34) obtained
during the conventional C–H insertion reaction; a simi-
lar though less pronounced effect was observed during
the cyclisation of the diazoketones 6a and 6b. This
result demonstrates that the stereochemistry defining
steps differ in the conventional and anomalous C–H
insertion reactions. It also precludes the intermediacy of
an equilibrating or freely rotating rhodium enolate 3
generated by hydride transfer prior to CꢀC bond for-
mation since this would produce a 1:1 mixture of the
labelled acetals 9 and 10 (Scheme 1). Finally, the kinetic
isotope effects for both C–H insertion reactions are
generally independent of the reaction temperature, a
References
1. Clark, J. S.; Dossetter, A. G.; Russell, C. A.; Whitting-
ham, W. G. J. Org. Chem. 1997, 62, 4910.
2. (a) Taber, D. F.; Ruckle, Jr., R. E. J. Am. Chem. Soc.
1986, 108, 7686; (b) Doyle, M. P.; Westrum, L. J.;
Wolthuis, W. N. E.; See, M. M.; Boone, W. P.; Bagheri,
V.; Pearson, M. M. J. Am. Chem. Soc. 1993, 115, 958; (c)
Taber, D. F.; Malcolm, S. C. J. Org. Chem. 1998, 63,
3717.
3. Mander, L. N.; Owen, D. J. Tetrahedron Lett. 1996, 37,
723.
4. The principal author would like to express his sincere
apologies to Professor Mander for an oversight which
meant that this important paper (Ref. 3) was not cited in
our original publication (Ref. 1).
5. White, J. D.; Hrnciar, P. J. Org. Chem. 1999, 64,
7271.
6. Wang, P.; Adams, J. J. Am. Chem. Soc. 1994, 116,
3296.

6190
J. S. Clark et al. / Tetrahedron Letters 42 (2001) 6187–6190
7. Sulikowski, G. A.; Lee, S. Tetrahedron Lett. 1999, 40,
7b:8b:11 (30:12:58), 9b:10b:12 (2:42:56); reflux 7b:8b:11
(29:14:57), 9b:10b:12 (5:40:55).
8035.
8. Product ratios in dichloromethane: 0°C 7b:8b:11
(33:16:51), 9b:10b:12 (5:39:56); reflux 7b:8b:11 (34:15:51);
9b:10b:12 (5:39:56).
9. Product ratios in 1,2-dichloroethane: 0°C 7b:8b:11
(31:13:56), 9b:10b:12 (3:42:55); room temperature
10. (a) Kwart, H. Acc. Chem. Res. 1982, 15, 401; (b) Anhede,
,
B.; Bergman, N.-A. J. Am. Chem. Soc. 1984, 106,
7634; (c) Dakin, L. A.; Ong, P. C.; Panek, J. S.; Staples,
R. J.; Stavropoulos, P. Organometallics 2000, 19,
2896.
.