Catalytic Asymmetric C-H Activation of Alkanes and Tetrahydrofuran

Article abstract of DOI:10.1021/ja994136c

Rhodium carbenoids derived from methyl aryldiazoacetates are capable of effective catalytic asymmetric C-H activation of a range of alkanes and tetrahydrofuran by a C-H insertion mechanism. Dirhodium tetrakis(S-(N-dodecylbenzenesulfonyl)prolinate) (Rh₂(S-DOSP)₄) catalyzed decomposition of methyl aryldiazoacetates in the presence of alkanes results in intermolecular C-H insertions with good control of regioselectivity, diastereoselectivity, and enantioselectivity. The carbenoids derived from methyl aryldiazoacetates are considerably more chemoselective than carbenoids derived from diazoacetates. They strongly favor C-H insertions into secondary and tertiary sites. Formation of side products due to carbene dimerization is not a major problem with rhodium carbenoids derived from aryldiazoacetates.

Full text of DOI:10.1021/ja994136c

J. Am. Chem. Soc. 2000, 122, 3063-3070
3063
Catalytic Asymmetric C-H Activation of Alkanes and
Tetrahydrofuran
Huw M. L. Davies,* Tore Hansen, and Melvyn Rowen Churchill
Contribution from the Department of Chemistry, State UniVersity of New York at Buffalo,
Buffalo, New York 14260-3000
ReceiVed NoVember 29, 1999
Abstract: Rhodium carbenoids derived from methyl aryldiazoacetates are capable of effective catalytic
asymmetric C-H activation of a range of alkanes and tetrahydrofuran by a C-H insertion mechanism.
Dirhodium tetrakis(S-(N-dodecylbenzenesulfonyl)prolinate) (Rh2(S-DOSP)4) catalyzed decomposition of methyl
aryldiazoacetates in the presence of alkanes results in intermolecular C-H insertions with good control of
regioselectivity, diastereoselectivity, and enantioselectivity. The carbenoids derived from methyl aryldiazoacetates
are considerably more chemoselective than carbenoids derived from diazoacetates. They strongly favor C-H
insertions into secondary and tertiary sites. Formation of side products due to carbene dimerization is not a
major problem with rhodium carbenoids derived from aryldiazoacetates.
One of the most challenging reactions in organic synthesis
is the functionalization of nonactivated C-H bonds. Not only
does such a reaction require an extremely reactive reagent, but
also, regiochemical problems would need to be overcome. Even
though there are several industrial processes for functionalization
of alkanes,1 general laboratory processes for alkane C-H
activation that lead directly to carbon-carbon bond formation
are limited.2 Over the last twenty years, there has been a massive
effort to achieve selective C-H activation of alkanes by the
use of organometallic chemistry.2b,3 One of the central trans-
formations that has been used for this chemistry is the oxidative
addition of highly reactive metal complexes into C-H bonds
(Scheme 1).4 The ultimate goal of this chemistry would be to
develop a catalytic and asymmetric C-H activation process,5
but unfortunately, this is an extremely difficult proposition
because regeneration of the highly active organometallic species
to complete the catalytic cycle is very difficult.
Scheme 1
Scheme 2
In this paper, a method will be described to achieve
asymmetric intermolecular C-H activation of alkanes by using
an alternative catalytic cycle (Scheme 2).6 In this scheme a
(1) (a) Mimoun, H. In ComprehensiVe Coordination Chemistry; Wilkin-
son, G., Gillard, R. D., McCleverty, J. A., Eds.; Pergamon: New York,
1987; Vol. 6. (b) SelectiVe Hydrocarbon ActiVation; Davies, J. A., Watson,
P. L., Greenberg, A., Liebman, J. F., Eds.; VCH Publishers: New York,
1990. (c) ActiVation and Functionalization of Alkanes; Hill, C. L., Ed.; John
Wiley & Sons: New York, 1989.
(2) (a) Guari, Y.; Sabo-Etienne, S.; Chaudret, B. Eur. J. Inorg. Chem.
1999, 1047. (b) Dyker, G. Angew. Chem., Int. Ed. Engl. 1989, 28, 1698.
(c) Sakakura, T.; Sodeyama, T.; Sasaki, K.; Wada, K.; Tanaka, M. J. Am.
Chem. Soc. 1990, 112, 7221. (d) Tanaka, M.; Sakakura, T.; Tokunaga, Y.;
Sodeyama, T. Chem. Lett. 1987, 2373. (e) Jones, W. D.; Hessell, E. T.
Organometallics 1990, 9, 718. (f) Fujiwara, Y.; Takai, K.; Taniguchi, Y.
Synlett 1996, 591. (g) Lin, M.; Sen, A. J. Chem. Soc., Chem. Commun.
1992, 892. (h) Ryu, I.; Sonoda, N. Angew. Chem., Int. Ed. Engl. 1996, 35,
1050. (i) Hill, C. L. Synlett 1995, 127. (j) Zheng, Z.; Hill, C. L. J. Chem.
Soc., Chem. Commun. 1998, 2467. (k) Brown, S. H.; Crabtree, R. H. J.
Am. Chem. Soc. 1989, 111, 1, 2935. (l) Xiang, J.; Evarts, J.; Rivkin, A.;
Curran, D. P.; Fuchs, P. L. Tetrahedron Lett. 1998, 39, 4163. (m) Xiang,
J.; Jiang, W. L.; Gong, J. C.; Fuchs, P. L. J. Am. Chem. Soc. 1997, 119,
4123.
highly reactive metal carbenoid intermediate will be the species
that will cause the C-H activation by insertion of a carbenoid
into the C-H bond. The carbenoid will be generated from a
reasonably stable metal complex that can catalyze the decom-
position of a diazo compound. The driving force for the
generation of the highly reactive metal-carbenoid intermediate
is the formation of N2 as a byproduct. Once the carbenoid
intermediate has undergone the C-H insertion into the alkane,
(3) Shilov, A. E.; Shul’pin, G. B. Chem. ReV. 1997, 97, 2879.
(4) (a) Arndtsen, B. A.; Bergman, R. G.; Mobley, T. A.; Peterson, T. H.
Acc. Chem. Res. 1995, 28, 154. (b) Waltz, K. M.; Hartwig, J. F. Science
1997, 277, 211.
(6) Doyle, M. P.; McKervey, M. A.; Ye, T. In Modern Catalytic Methods
for Organic Synthesis with Diazo Compounds; Wiley-Interscience: New
York, 1998.
(5) Mobley, T. A.; Bergman, R. G. J. Am. Chem. Soc. 1998, 120, 3253.
10.1021/ja994136c CCC: $19.00 © 2000 American Chemical Society
Published on Web 03/17/2000

3064 J. Am. Chem. Soc., Vol. 122, No. 13, 2000
DaVies et al.
the original metal complex will be formed and the catalytic cycle
can continue. Even though general reviews on C-H activation
of hydrocarbons1,2a,b,3 have largely ignored carbenoid C-H
insertions,6 this process has the potential to be a practical
solution to catalytic C-H activation, as long as the chemose-
lectivity can be controlled.
The metal-catalyzed decomposition of diazo compounds is a
well-established method for the generation of reactive metal-
carbenoid intermediates.6 C-H activation by means of an
intramolecular C-H insertion is a powerful synthetic method.7
Excellent control of diastereoselectivity and enantioselectivity
is possible with this chemistry. In contrast, intermolecular C-H
insertion is not generally considered to be a synthetically useful
process8 even though numerous examples of the reaction have
been reported over the years.9 The reason for this lack of
synthetic utility is that metal carbenoids are prone to dimer
formation10 and in the absence of an efficient trapping agent,
the dimer formation is the dominant reaction process. Further-
more, the regioselectivity of the reported C-H insertions was
not sufficient to be of general practical utility.9
enantioselective.15 Furthermore, we discovered that this type
of carbenoid is much more chemoselective and less prone to
dimer formation than carbenoids derived from alkyl diazo-
acetates.16 This led us to consider that this class of carbenoids
would be able to undergo effective intermolecular C-H inser-
tion. In combination with Rh2(S-DOSP)4, this chemistry could
lead to a practical catalytic asymmetric C-H activation process.
In our preliminary communication of this work, we reported
that methyl aryldiazoacetates underwent C-H activation into
cycloalkanes heated under reflux in 60-89% ee (eq 1)17 and
into tetrahydrofuran in 52-76% ee (eq 2).17 Recently, we have
demonstrated that C-H insertions of aryldiazoacetates R to
nitrogen (eq 3)18 and at allylic positions R to oxygen (eq 4)19
Over the last 15 years, we11,12, and others13 have explored
the chemistry of a different class of metal-carbenoid intermedi-
ates which contain electron-withdrawing and electron-donating
(vinyl or phenyl) groups (1). This class of metal-carbenoid
intermediates has very different characteristics from the tradi-
tional carbenoids derived from alkyl diazoacetates. For example,
its cyclopropanation chemistry is highly diastereoselective11b,14
and by using Rh2(S-DOSP)4 (2) catalysis, its reactions are highly
(7) (a) Taber, D. F.; Song, Y. Tetrahedron Lett. 1995, 36, 2587. (b) Taber,
D. F.; You, K. K. J. Am. Chem. Soc. 1995, 117, 5757. (c) Taber, D. F.;
You, K. K.; Rheingold, A. L. J. Am. Chem. Soc. 1996, 118, 547. (d) Taber,
D. F.; Song, Y. J. Org. Chem. 1996, 61, 6706. (e) Taber, D. F.; Malcolm,
S. C. J. Org. Chem. 1998, 63, 3717. (f) Taber, D. F.; Stiriba, S. E. Chem.
A Eur. J. 1998, 4, 990. (g) Doyle, M. P.; Kalinin, A. V.; Ene, D. G. J. Am.
Chem. Soc. 1996, 118, 8837. (h) Doyle, M. P.; Dyatkin, A. B. J. Org. Chem.
1995, 60, 3035. (i) Doyle, M. P.; Zhou, Q.-L.; Raab, C. E.; Roos, G. H. P.
Tetrahedron Lett. 1995, 36, 4745. (j) Doyle, M. P.; Protopopova, M. N.;
Poulter, C. D.; Rogers, D. H. J. Am. Chem. Soc. 1995, 117, 7281. (k) Doyle,
M. P.; Protopopova, M. N.; Zhou, Q.-L.; Bode, J. W.; Simonsen, S. H.;
Lynch, V. J. Org. Chem. 1995, 60, 6654. (l) Doyle, M. P.; Van Oeveren,
A.; Westrum, L. J.; Protopopova, M. N.; Clayton, T. W., Jr. J. Am. Chem.
Soc. 1991, 113, 8982. (m) Doyle, M. P.; Zhou, Q.-L.; Dyatkin, A. B.;
Ruppar, D. A. Tetrahedron Lett. 1995, 36, 7579.
(8) (a) Reference 6, p 115 (b) Ye, T.; McKervey, M. A. Chem. ReV.
1994, 94, 1091. (c) Spero, D. M.; Adams, J. Tetrahedron Lett. 1992, 33,
1143.
(9) For representative examples of intermolecular C-H insertions, see:
(a) Scott, L. T.; DeCicco, G. J. J. Am. Chem. Soc. 1974, 96, 322. (b)
Ambramovitch, R. A.; Roy, J. J. Chem. Soc., Chem. Commun. 1965, 542.
(c) Adams, J.; Poupart, M.-A.; Greainer, L.; Schaller, C.; Quimet, N.;
Frenette, R. Tetrahedron Lett. 1989, 30, 1749. (d) Demonceau, A.; Noels,
A. F.; Hubert, A. J.; Teyssie, P. J. Chem. Soc., Chem. Commun. 1981, 688.
(e) Demonceau, A.; Noels, A. F.; Hubert, A. J.; Teyssie, P. Bull. Soc. Chim.
Belg. 1984, 93, 945. (f) Demonceau, A.; Noels, A. F.; Hubert, A. J.; Teyssie,
P. J. Mol. Catal. 1988, 49, L13. (g) Callott, H. J.; Metz, F. Tetrahedron
Lett. 1982, 23, 4321. (h) Callott, H. J.; Metz, F. NouV. J. Chim. 1985, 9,
167. (i) Domenceau, A.; Noels, A. F.; Costa, J. L.; Hubert, A. J. Mol. Catal.
1990, 58, 21.
(10) (a) Doyle, M. P.; van Leusen, D.; Tamblyn, W. H. Synthesis 1981,
787. (b) Wulfman, D. S.; Pearce, B. W.; McDaniel, R. S., Jr. Tetrahedron
1976, 32, 1251.
(11) For recent examples, see: (a) Davies, H. M. L.; Kong, N.; Churchill,
M. R. J. Org. Chem. 1998, 63, 6586. (b) Davies, H. M. L.; Rusiniak, L.
Tetrahedron Lett. 1998, 39, 8811. (c) Davies, H. M. L.; Ahmed, G.; Calvo,
R. L.; Churchill, M. R.; Churchill, D. G. J. Org. Chem. 1998, 63, 2641. (d)
Davies, H. M. L.; Hodges, L. M.; Thornley, C. T. Tetrahedron Lett. 1998,
39, 2707. (e) Davies, H. M. L.; Doan, B. D. J. Org. Chem. 1998, 63, 657.
(f) Davies, H. M. L.; Stafford, D. G.; Doan, B. D.; Houser, J. H. J. Am.
Chem. Soc. 1998, 120, 3326.
(12) For general reviews, see: (a) Davies, H. M. L. Tetrahedron 1993,
49, 5203. (b) Davies, H. M. L. Aldrichim. Acta 1997, 30, 107. (c) Davies,
H. M. L. Curr. Org. Chem. 1998, 2, 463. (d) Davies, H. M. L. In AdVances
in Cycloaddition; Haramata, M. E., Ed.; JAI Press Inc.: Greenwich, CT,
1999; Vol. 5. pp 119-164. (e) Davies, H. M. L. Eur. J. Org. Chem. 1999,
2459, 9.
(13) (a) Landais, Y.; Planchenault, D. Tetrahedron Lett. 1994, 35, 4565.
(b) Bulugahapitiya, P.; Landais, Y.; Parra-Rapado, L.; Planchenault, D.;
Weber, V. J. Org. Chem. 1997, 62, 1630. (c) Yoshikawa, K.; Achiwa, K.
Chem. Pharm. Bull. 1996, 43, 2048. (d) Moye-Sherman, D.; Welch, M.
B.; Reibenspies, J.; Burgess, K. J. Chem. Soc., Chem. Commun. 1998, 2377.
(e) Kende, A. S.; Smalley, T. L.; Huang, H. J. Am. Chem. Soc. 1999, 121,
7431. (f) Doyle, M. P.; Zhou, Q.-L.; Charnsangavej, C.; Longoria, M. A.;
McKervey, M. A.; Garcia, C. F. Tetrahedron Lett. 1996, 37, 4129. (g) Axten,
J. M.; Ivy, R.; Krim, L.; Winkler, J. D. J. Am. Chem. Soc. 1999, 121, 6511.
(14) Davies, H. M. L.; Clark, T. J.; Church, L. A. Tetrahedron Lett.
1989, 30, 5057.
(15) Davies, H. M. L.; Bruzinski, P.; Hutcheson, D. K.; Kong, N.; Fall,
M. J. J. Am. Chem. Soc. 1996, 118, 6897.

C-H ActiVation of Alkanes and Tetrahydrofuran
Table 1. Asymmetric C-H Activation of Cycloalkanes
J. Am. Chem. Soc., Vol. 122, No. 13, 2000 3065
by comparison of the optical rotation of the corresponding
carboxylic acid with the literature value.21 The absolute con-
figuration of the other products is tentatively assigned on the
assumption that the same sense of asymmetric induction occurs
in all of the reactions with cycloalkanes.
A C-H insertion process would be expected to display a
considerable kinetic isotope effect.9h,i This was confirmed by
running the reaction of aryldiazoacetates 3a, 3c, and 3d with a
1:1 mixture of cyclohexane and d12-cyclohexane (eq 6). From
the product distribution it was determined that the kinetic isotope
effect was between 2.0 and 2.1.
diazo
Ar
n
product
yield, %
ee, %
3a
3b
3c
3d
3e
3f
3g
3h
3i
C6H5
2
2
2
2
2
2
2
2
2
2
1
1
4a
4b
4c
4d
4e
4f
4g
4h
4j
80
64
76
23a
81
47
78
63
72
62
72
70
95
95
94
88
90
94
94
93
90
95
96
95
p-BrC6H4
p-ClC6H4
p-(MeO)C6H4
o-ClC6H4
m-ClC6H4
p-CF3C6H4
p-MeC6H4
o-BrC6H4
m-BrC6H4
C6H5
3j
3a
3c
4k
5a
5c
p-ClC6H4
a 3 mol % of catalyst was used.
are extremely efficient and useful transformations. In this paper
we describe a systematic study exploring the scope and regio-
chemistry of C-H activation with various alkanes, which, of
course, are very challenging substrates for this type of chemistry.
In the preliminary communication,17 the reactions of aryl-
diazoacetates with cycloalkanes were carried out under reflux
conditions because the yields dropped considerably when the
reaction temperatures were lowered. On reexamination of the
reaction of methyl phenyldiazoacetate (3a) with cyclohexane,
we have found that if the solvent is degassed,20 the reactions
can be carried out at temperatures as low as 10 °C without an
appreciable drop in yield (eq 5). In these reactions, cyclohexane
In the asymmetric cyclopropanation chemistry of alkyl
vinyldiazoacetates with Rh2(S-DOSP)4, the enantioselectivity
drops considerably on increasing the size of the alkyl group
from methyl to tert-butyl.15 This effect is a distinctive feature
of the vinyldiazoacetate/rhodium prolinate system, because in
the diazoacetate system, the highest enantioselectivity of cy-
clopropanation is generally obtained when a bulky ester group
is used. An exploration of the effect of ester size on the C-H
activation revealed that a parallel trend existed between the C-H
activation and the asymmetric cyclopropanation. A slight drop
in enantioselectivity occurred in the C-H activation on changing
from the methyl ester 3a to isopropyl ester 6 while with the
tert-butyl ester 7, the enantioselectivity for C-H insertion was
only 20% ee.
is used as solvent and the aryldiazoacetate is added dropwise
to Rh2(S-DOSP)4 over 90 min. Under these conditions, the C-H
activation product 4a is formed in 95% ee. The reaction is
applicable to a range of aryldiazoacetates 3a-j, and in most
cases the C-H insertion products 4a-j are produced in >90%
ee (Table 1). Even an electron rich aromatic ring can be
incorporated into the aryl diazoacetate (3d) although the yield
of the C-H insertion product (4d) is considerably lower in this
case (23% yield) than what was obtained when higher reaction
temperatures were used (85% yield).17 The reaction of cyclo-
pentane with aryldiazoacetates 3a and 3c similarly results in
the formation of the C-H insertion products 5a and 5c in >90%
ee. The absolute configuration of 4a was determined to be (R)
The standard reactions described above were carried out with
1 mol % of Rh2(S-DOSP)4, but lower equivalents of catalyst
can be used as shown in eq 8. On using 0.1% of catalyst at 24
(16) Davies, H. M. L.; Hodges, L. M.; Matasi, J. J.; Hansen, T.; Stafford,
D. G. Tetrahedron Lett. 1998, 39, 4417.
(17) Davies, H. M. L.; Hansen, T. J. Am. Chem. Soc. 1997, 119, 9075.
(18) Davies, H. M. L.; Hansen, T.; Hopper, D. J. Am. Chem. Soc., 1999,
121, 6509.
(19) Davies, H. M. L.; Antoulinakis, E. G.; Hansen, T. Org. Lett. 1999,
1, 383.
(20) For a previous example of enhanced yields of intramolecular C-H
insertions after careful degassing, see: Taber, D. F.; Hennessy, M. J.; Louey.
J. P. J. Org. Chem. 1992, 57, 436.
(21) Barth, G.; Voelter, W.; Mosher, H. S.; Bunnenberg, E.; Djerassi,
C. J. Am. Chem. Soc. 1970, 92, 875.

3066 J. Am. Chem. Soc., Vol. 122, No. 13, 2000
Table 2. Asymmetric C-H Activation of Tetrahydrofuran
DaVies et al.
combined diasteromer
ee, %
(major ds)
diazo
3a C6H5
Ar
product yield, %
ratio
10a
10c
10d
10h
10k
67
74
56
60
62
2.8
2.4
3.4
4.0
1.6
97
98
96
97
95
3c
p-ClC6H4
Figure 1. Absolute configuration of 11.
3d p-(MeO)C6H4
3h p-MeC6H4
3k 2-naphthyl
The kinetic isotope effect for the C-H insertion into
tetrahydrofuran was readily determined by competition experi-
ments between tetrahydrofuran and d8-tetrahydrofuran (eq 11).
In each case, the kinetic isotope effect was greater than that for
the reaction with cyclohexane, ranging from 3 to 3.2.
°C, the yield and enantioselectivity (91% ee vs 92% ee) of 4a
is slightly diminished. With 0.01% of catalyst, however, the
reaction conversion is very low and the enantioselectivity of
4a is considerably decreased (70% ee). A similar trend was
observed in the cyclopropanation chemistry catalyzed by Rh2-
(S-DOSP)4.15
The reaction of tetrahydrofuran with aryldiazoacetates has
also been optimized as illustrated in eq 9. After degassing of
Having determined that the C-H activation was an effective
process, the next series of reactions were directed toward
determining the chemoselectivity of the reaction with various
alkanes. From the cyclopropanation chemistry of carbenoids
derived from vinyldiazoacetates, it was expected that a subtle
balance of steric and electronic effects would exist. These
carbenoids preferentially attack electron-rich alkenes in which
charge build up in the transition step of the nonsynchronous
cyclopropanation would be stabilized.13 In a related manner, a
concerted nonsynchronous C-H insertion mechanism would
explain why favorable C-H insertions occur at C-2 of tetrahy-
drofuran. It is also well established, however, that the rhodium
prolinate/vinylcarbenoid system is sterically demanding, as
displayed by its inability to cyclopropanate trans alkenes.13,15
Consequently, in the case of the C-H insertion, it was expected
that the tertiary C-H bonds would be preferentially attacked
on electronic grounds but steric factors might reverse this trend.
The reaction with isobutane was examined to determine the
reactivity difference between primary and tertiary C-H bonds.
Due to its physical properties (bp -12 °C), this is not an ideal
substrate because the reaction needs to be carried out below
the optimum temperatures of 10 °C or greater unless one is
willing to resort to using a pressure reactor. Nevertheless, C-H
insertion into the tertiary C-H bond did occur at -12 °C to
form 13 in 75% ee. The yield of 13, however, was low (20%)
because azine formation was a major competing reaction under
these low-temperature conditions. The absolute configuration
of 13 was determined to be (R) by comparison of its optical
rotation to the literature values.22
the solvents the reaction with tetrahydrofuran can be successfully
accomplished at -50 °C and under these conditions the enantio-
selectivity for the formation of the C-H insertion product 10a
is 90% ee. A further improvement in enantioselectivity was
possible by running the reaction in hexane with just 2 equiv of
tetrahydrofuran. It is well established that the enantioselectivity
for Rh2(S-DOSP)4 catalysis is highly sensitive to solvent effects,
with nonpolar solvents being highly beneficial.15 By using
hexane as solvent the enantioselectivity for the formation of
10a was improved to 97% ee. The observation of high yields
in this reaction is remarkable because it signifies that the carbe-
noid is highly chemoselective, favoring reaction with the THF
even in the presence of a vast excess of hydrocarbon. Under
these optimized conditions, the reaction is applicable to a series
of aryldiazoacetates, and in each instance, the enantioselectivity
for the C-H insertion products 10 is >95% ee (Table 2).
In the original communication, the relative stereochemistry
of 10 was tentatively assigned as (2S*,RR*)17 but the absolute
configuration was not determined. Proof of the absolute con-
figuration of 10k as (2S,RR) was obtained by conversion of
10k to crystalline 11 (eq 10), whose structure was confirmed
by X-ray crystallography (Figure 1).

C-H ActiVation of Alkanes and Tetrahydrofuran
J. Am. Chem. Soc., Vol. 122, No. 13, 2000 3067
To explore further the selectivity between primary and tertiary
C-H bonds, the reaction of 3a with 2,3-dimethylbutane was
examined. Under refluxing conditions (58 °C), using 1% of cata-
lyst, this reaction resulted in the formation of the tertiary C-H
insertion product 14 in 24% yield and 46% ee. The low yield
in this reaction is probably due to the steric difficulty in reaction
of the carbenoid at the tertiary C-H bond in 2,3-dimethylbutane,
which is relatively crowded. Even though the phenyldiazoacetate
was added dropwise over 1 h, carbene dimer 15 and azine 16
were formed (64:36 ratio) as major side products.
azoacetates.19,24 Further confirmation of the remarkable attributes
of Rh2(S-DOSP)4 was observed by evaluating a series of cata-
lysts in the reaction of 3a with 2-methylbutane. None of the
traditional catalysts that were examined gave greater than 30%
yield of C-H insertion product 17a. Electron-rich catalysts are
not particularly effective at intermolecular C-H insertion into
alkanes,9d while of all the standard catalysts, only Rh2(S-DOSP)4
is fully soluble in 2-methylbutane under ambient conditions.
The lack of C-H insertion into the methylene group of
2-methylbutane is indicative of a steric influence because the
reaction with cyclohexane demonstrates that secondary sites are
susceptible to the C-H insertion. To compare the rate of
reaction between secondary and tertiary sites, competition
experiments were carried out using a mixture of cyclohexane
and 2-methylbutane. Reaction of phenyldiazoacetate 3a at 24
°C with a 1:4.6 mixture of cyclohexane and 2-methylbutane
gave a 3.2:1 mixture of 4a and 17a. After considering the
statistical factor, the tertiary C-H bond in 2-methylbutane is
of similar reactivity to the C-H bond in cyclohexane. The same
reaction with 4-bromphenyldiazoacetate 3b gave a 2.6:1 mixture
of 4b and 17b. The more electrophilic carbenoid that would be
derived from 4b appears to be more selective for a tertiary C-H
bond than the carbenoid derived from 4a.
As it is generally considered that the formation of both
carbene dimer and azine is due to the reaction of the carbenoid
with excess diazo compound,23 a study was undertaken to try
to improve the reaction by altering the amount of catalyst that
was used. When 5% of catalyst was used, the yield of 14
remained about the same but the enantioselectivity was improved
to 66% ee. Curiously, the ratio of 15 and 16 also changed (75:
25). An even more dramatic change occurred when 0.2% of
catalyst was used. Only a 4% yield of 14 in 13% ee was
obtained, while the ratio of 15 to 16 was changed to 32:68.
The observation that 15 is preferentially formed with high
catalyst loading while 16 is formed with low catalyst loading
is not consistent with the general view that both carbene dimer
and azine are derived from reaction of a carbenoid with the
diazo compound.23 Instead, a more reasonable explanation is
that only the azine 16 is formed from reaction of the carbenoid
with the aryldiazoacetate, while the dimer 15 is formed by
reaction between two rhodium carbenoid intermediates.
To explore the reactivity profile between secondary and
tertiary sites, the reaction between 3a and 2-methylbutane was
examined. Decomposition of 3a with Rh2(S-DOSP)4 at 24 °C
in 2-methylbutane resulted in the formation of 17a in 60% yield
and 68% ee. The higher yield obtained with 2-methybutane
compared to 2,3-dimethylbutane is presumably because the
tertiary C-H bond in 2-methybutane is less sterically encum-
bered than the tertiary C-H bonds in 2,3-dimethylbutane. No
evidence (<5%) for C-H insertion into the secondary C-H
bond was observed. The regiochemistry is very different from
the published results on the rhodium-catalyzed reaction of ethyl
diazoacetate with 2-methylbutane, where a mixture of primary,
secondary, and tertiary C-H insertion products were formed.9d
We have previously communicated that Rh2(S-DOSP)4 appears
to be ideally suited for C-H insertions of aryl- and vinyldi-
The absence of reactivity at the secondary C-H site in
2-methylbutane is a promising indication that chemoselectivity
between secondary C-H bonds in different steric environments
should be feasible. This concept was tested in the reaction of
aryldiazoacetates with 2-methylpentane. Rh2(S-DOSP)4 cata-
lyzed decomposition of 3a in the presence of 2-methylpentane
(22) Bright, D. A.; Mathisen, D. E.; Zieger, H. E. J. Org. Chem. 1982,
47, 3521.
(23) For a general discussion of carbene dimerization, see ref 6, pp 624-
627.
(24) Davies, H. M. L.; Stafford, D. G.; Hansen, T. Org. Lett. 1999, 1,
233.

3068 J. Am. Chem. Soc., Vol. 122, No. 13, 2000
DaVies et al.
Figure 2. Absolute configuration of 20.
under refluxing conditions (62 °C) gave rise to the tertiary C-H
insertion product 18a and the secondary C-H insertion product
19a in a 58:42 ratio and 75% combined yield. Remarkably, 19a
was formed in a 4:1 diastereomeric ratio and 86% ee, while no
evidence (<5%) of the regioisomeric secondary C-H insertion
1
product was observed in the H NMR of the crude mixture. A
Figure 3. Relative rates of reaction of 3a with various substrates.
similar reaction with the bromophenyldiazoacetate 3b gave a
77:23 ratio of 18b and 19b in 50% yield. The secondary C-H
insertion product 19b was again formed in a 4:1 diastereomeric
ratio and 86% ee. The change in ratio is once again indicative
that the more electrophilic carbenoid that would be derived from
3b is more selective for a tertiary C-H site than the carbenoid
derived from 3a. The absolute configuration of 19b was
determined to be (2R,3S) by X-ray crystallographic analysis of
the pyrrolidine amide derivative 20 of 19b as shown in Figure
2.
insertion product 21a in 90% ee. Similarly, the reaction between
adamantane and 3b gave a 70% yield of 21b in 96% ee.
The absolute configuration of 4a, 10a, 13, and 19b was
determined unequivocally. The absolute configuration of all the
other products is tentatively assigned on the assumption that
the re face of the rhodium carbenoid is attacked in each case.
Supporting evidence for this tentative stereochemical assignment
was obtained from the CD spectra of 4a, 5a, 13, 14, 17a, 18a,
18b, 19a, 19b, 21a, and 21b, all of which show a strongly
negative absorption at 220 nm.21
One of the most distinctive features of the C-H insertion
chemistry of aryldiazoacetates is the remarkable chemoselec-
tivity that is possible. To quantify this effect competition
experiments between various substrates were examined and the
results are summarized in Figure 3. In most instances the
competition experiments were run using a 1:1 mixture of
substrates, but in the cases where one of the substrates was
considerably more reactive that the other, the ratio of substrates
was altered accordingly. From these data, it is clear that these
C-H insertions display remarkable regioselectivity. Reaction
with THF and N-BOC-pyrrolidine is more favorable than
reaction with cyclohexane by factors of 1700 and 2400,
respectively. C-H activation of THF is about 10 times less
favorable than Si-H insertion or cyclopropanation of styrene.
On the basis of the reactivity profile described above, 2,2-
dimethylbutane was considered to be a promising inert solvent
for the C-H insertion chemistry because it has a reasonable
boiling point and the C-H bonds in 2,2-dimethylbutane are at
either primary or sterically crowded secondary sites. A test
reaction with 2,2-dimethylbutane under the standard reaction
conditions revealed that no appreciable reaction between the
carbenoid and 2,2-dimethylbutane had occurred. This meant that
2,2-dimethylbutane is an appropriate inert solvent for reactions
in which solid hydrocarbons such as adamantane are used as
substrates. Reaction of 3a with 2 equiv of adamantane in 2,2-
dimethylbutane gave rise to 67% yield of the tertiary C-H

C-H ActiVation of Alkanes and Tetrahydrofuran
J. Am. Chem. Soc., Vol. 122, No. 13, 2000 3069
C-H sites and for the 2-position of tetrahydrofuran is consistent
with this mechanism because these are the sites that are best
suited to stabilize the positive charge buildup in the transition
state. The actual trajectory of approach of the alkane is not
known, but to generate the correct absolute stereochemistry, the
alkane would need to approach over the ester group of the
carbenoid. The increased reactivity of bromophenyldiazoacetate
3b compared to phenyldiazoacetate 3a for reaction at a tertiary
C-H site rather than a secondary C-H site is also consistent
with this model, because charge build up in the transition state
of 3b would be enhanced over 3a. The generally higher
enantioselectivity of C-H insertions at secondary sites compared
to tertiary sites would presumably be because a close approach
of the alkane to the catalyst is less likely for reaction at a
crowded tertiary C-H insertion site compared to a secondary
site.
Figure 4. Model for asymmetric cyclopropanation.
When the C-H insertion occurs at a pro-chiral methylene
group, two stereogenic centers are generated during the C-H
insertion. Although a full understanding of the structural
requirements for the diastereoselection is not yet available, it
is clear that impressive diastereocontrol is possible in certain
systems.18,19 The diastereocontrol that is observed is consistent
with the model shown in Figure 5 where the large group (L)
projects upward from the catalyst, the small group (S) points
toward the carbenoid and slightly toward the catalyst, and the
medium group (M) projects away from the carbenoid but slightly
toward the catalyst. Using this model, the reaction with
2-methypentane would be predicted to form the (2R,3S) isomer
(eq 18), the reaction with N-BOC-pyrrolidine would be predicted
to form the (2S,2′S) isomer (eq 19), while the reaction with
allyl silyl ethers would be expected to form the (2S,3R)
products (eq 20). Further studies will need to be carried out
to determine if this simple predictive model is generally
applicable.
Figure 5. Model for the asymmetric C-H insertion.
Even though selective reactions at tertiary C-H sites are
possible, in general there is a delicate balance in reaction
between secondary and tertiary C-H sites. Tertiary sites are
preferred on electronic grounds but secondary sites are preferred
on steric grounds.
Discussion
From the above experiments it is apparent that the asymmetric
C-H insertion of aryldiazoacetates has many similarities to the
asymmetric cyclopropanation of phenyl- and vinyldiazoacetates.
The most reactive C-H bonds are those in which the carbon
would be able to stabilize a buildup of positive charge during
the transition state, although the rhodium carbenoid is a sterically
demanding reagent and steric effects may dominate over
electronic effects. In both reactions the influence of ester size
and solvent on the asymmetric induction follows a parallel trend.
We have proposed that the asymmetric cyclopropanation occur
by a transition state presented in Figure 4.15 The catalyst behaves
as if it is D2 symmetric with the arylsulfonyl groups aligned
R,â,R,â. The two thickened vertical lines in Figure 4 represent
the blocking effects of the arylsulfonyl groups (due to the
symmetry of the catalysts, the chiral influence on the bottom
face of the catalysts would be the same as the top face). The
cyclopropanation is considered to be concerted but nonsyn-
chronous with the alkene approaching in a side-on mode over
the ester group of the carbenoid.15 Apparently, the major reason
for the success of the phenyl and vinyldiazoacetate cyclopro-
panations is that the trajectory of approach of the alkene to the
carbenoid is much more demanding than is the case with other
carbenoid systems.
The generally accepted mechanism for intramolecular C-H
insertions involves the direct insertion between the carbenoid
and the C-H bond by means of a three-membered transition
state.25 Due to the similar trends between the cyclopropanation
and the C-H insertion of aryldiazoacetates, we propose that
the C-H insertion also occurs in a concerted but nonsynchro-
nous manner with build up of positive charge at the carbon of
the C-H bond (Figure 5). The kinetic isotope effect that was
observed for cyclohexane and tetrahydrofuran is consistent with
such a mechanistic interpretation. The higher value obtained
with THF is indicative that the transition state for cleavage of
the C-H bond is more advanced in THF C-H insertions than
the cyclohexane C-H insertions. The selectivity toward tertiary
One of the intriguing discoveries that was made during these
studies is that in the case of phenyldiazoacetate 3a, when poor
trapping agents are present, carbene dimer formation is preferred
using high catalyst loading while azine formation is preferred
using low catalyst loading. This result is an indication that the
carbene dimer from 3a is likely formed by reaction between
two rhodium carbenoid species, and this suggestion is counter
to the generally held opinion that carbene dimers arise from
reaction of a carbenoid with the diazo compound.23 Furthermore,
(25) For a general discussion on the proposed mechanisms of intramo-
lecular C-H insertions, see ref 6, pp 133-162.

3070 J. Am. Chem. Soc., Vol. 122, No. 13, 2000
DaVies et al.
these observations indicate that the rhodium carbenoid has a
sufficiently long lifetime in the hydrocarbon solvents to allow
two rhodium carbenoids to react together. Further studies are
in progress to determine the full ramifications of this observa-
tion.
In summary, these studies demonstrate that carbenoids derived
from aryldiazoacetates are capable of effective asymmetric C-H
activation of a range of alkanes and tetrahydrofuran. These
carbenoids are considerably more chemoselective than car-
benoids derived from diazoacetates and strongly favor C-H
insertions at secondary and tertiary sites. Even though general
reviews on C-H activation of hydrocarbons1,2a,b,3 have largely
ignored carbenoid C-H insertions, it is clear from this work
that rhodium carbenoids derived from aryldiazoacetates are
practical intermediates for asymmetric intermolecular C-H
activation. Due to the high regioselectivity, diastereoselectivity,
and enantioselectivity of these reactions, we expect this
chemistry to be of considerable value in organic synthesis.
Acknowledgment. Financial support of this work by the
National Science Foundation (CHE 9726124) is gratefully
acknowledged.
Supporting Information Available: Complete experimental
data, X-ray data for 11 and 20, and NMR data for 3l, 4b, 4e,
4g, 4h, 4i, 4j, 8, 10k, 11, 13, 14, 17a, 17b, 18a, 18b, 19a, 19b,
20, 21a and 21b (PDF). This material is available free of charge
via the Internet at http://pubs.acs.org.
JA994136C