Combined experimental and computational studies of heterobimetallic Bi-Rh paddlewheel carboxylates as catalysts for metal carbenoid transformations

Article abstract of DOI:10.1021/jo900998s

(Chemical Equation Presented) The catalytic activity of heterobimetallic Bi-Rh paddlewheel carboxylate complexes has been evaluated for the first time in the context of metal carbenoid chemistry. The Bi-Rh carboxylate complexes were found to effectively catalyze both cyclopropanation reactions and C-H insertions as well as reactions involving ylide intermediates with similar selectivity profiles to analogous dirhodium complexes. The heterometallic complex BiRh(O₂CCF₃)₃(O₂CCH₃) was found to be approximately 1600 times less reactive than its homometallic analogue Rh₂(O₂CCF₃)₃(O ₂CCH₃) toward the decomposition of methyl phenyldiazoacetate. The observed difference in reactivity is in good agreement with a computational model system where axial coordination to the second rhodium active site is considered for the dirhodium catalyst.

Full text of DOI:10.1021/jo900998s

pubs.acs.org/joc
Combined Experimental and Computational Studies of Heterobimetallic
Bi-Rh Paddlewheel Carboxylates as Catalysts for Metal Carbenoid
Transformations
Jorn Hansen,† Bo Li,‡ Evgeny Dikarev,‡ Jochen Autschbach,§ and Huw M. L. Davies*,†
†Department of Chemistry, Emory University, 440 Atwood Hall, 1515 Dickey Drive, Atlanta, Georgia 30322,
‡Department of Chemistry, University at Albany, State University of New York, Albany, New York 12222,
and §Department of Chemistry, University at Buffalo, State University of New York, Buffalo,
New York 14260-3000
hmdavie@emory.edu
Received May 14, 2009
The catalytic activity of heterobimetallic Bi-Rh paddlewheel carboxylate complexes has been
evaluated for the first time in the context of metal carbenoid chemistry. The Bi-Rh carboxylate
complexes were found to effectively catalyze both cyclopropanation reactions and C-H insertions as
well as reactions involving ylide intermediates with similar selectivity profiles to analogous
dirhodium complexes. The heterometallic complex BiRh(O2CCF3)3(O2CCH3) was found to be
approximately 1600 times less reactive than its homometallic analogue Rh2(O2CCF3)3(O2CCH3)
toward the decomposition of methyl phenyldiazoacetate. The observed difference in reactivity is in
good agreement with a computational model system where axial coordination to the second rhodium
active site is considered for the dirhodium catalyst.
Introduction
entities are structurally characterized by a bimetallic core with a
Rh-Rh single bond, bridged by four μ2-carboxylate, carbox-
amidate, or phosphonate ligands. A large number of complexes
have been synthesized in which the ligand structure has been
diversified to a great extent. Many of these have been studied
and used in catalysis.1 The dirhodium catalysts have been
remarkably successful in a range of transformations, particu-
larly in metal nitrenoid and carbenoid chemistry1d involving
aziridinations, cyclopropanation/propenation reactions,1h,o,2
C-H insertions,1d,3 and ylide transformations,1i,k,n,4 but also
Catalysis by dirhodium paddlewheel complexes has been an
area of immense interest for the past three decades. These
*To whom correspondence should be addressed.
(1) (a) Boyar, E. B.; Robinson, S. D. Coord. Chem. Rev. 1983, 50, 109–
208. (b) Davies, H. M. L. Eur. J. Org. Chem. 1999, 2459–2469. (c) Davies, H.
M. L.; Beckwith, R. E. J. Chem. Rev. 2003, 103, 2861–2903. (d) Davies, H. M.
L.; Manning, J. R. Nature 2008, 451, 417–424. (e) Doyle, M. P. Chem. Rev.
1986, 86, 919–939. (f) Doyle, M. P. Recl. Trav. Chim. Pays-Bas 1991, 110,
305–316. (g) Doyle, M. P. Aldrichim. Acta 1996, 29, 3–11. (h) Doyle, M. P. In
Modern Rhodium-Catalyzed Organic Reactions; Evans, P. A., Ed.; Wiley-VCH
Verlag GmbH & Co. KGaA: Weinheim, Germany, 2005; p 473. (i) Doyle, M. P. J.
Org. Chem. 2006, 71, 9253–9260. (j) Doyle, M. P.; Forbes, D. C. Chem. Rev.
1998, 98, 911–935. (k) Doyle, M. P.; McKervey, M. A.; Ye, T. Modern Catalytic
Methods for Organic Synthesis with Diazo Compounds: From Cyclopropanes
to Ylides; Wiley-interscience: New York, 1998. (l) Felthouse, T. R. Prog. Inorg.
Chem. 1982, 29, 73–166. (m) Hansen, J.; Davies, H. M. L. Coord. Chem. Rev.
2008, 252, 545–555. (n) Hodgson, D. M.; Stupple, P. A.; Pierard, F. Y. T. M.;
Labande, A. H.; Johnstone, C. Chem.;Eur. J. 2001, 7, 4465–4476. (o) Lebel, H.;
Marcoux, J. F.; Molinaro, C.; Charette, A. B. Chem. Rev. 2003, 103, 977–1050.
(p) Ren, T. Coord. Chem. Rev. 1998, 175, 43–58. (q) Timmons, D. J.; Doyle, M. P.
J. Organomet. Chem. 2001, 617, 98–104. (r) Ye, T.; McKervey, M. A. Chem.
Rev. 1994, 94, 1091–1160.
(2) Davies, H. M. L.; Antoulinakis, E. G. Org. React. 2001, 57, 1–326.
(3) (a) Dauban, P.; Dodd, R. H. Amino Group Chem. 2008, 55–92.
(b) Davies, H. M. L.; Hansen, T.; Churchill, M. R. J. Am. Chem. Soc.
2000, 122, 3063–3070. (c) Espino, C. G.; Fiori, K. W.; Kim, M.; Du Bois, J.
J. Am. Chem. Soc. 2004, 126, 15378–15379. (d) Flanigan, D. L.; Yoon, C. H.;
Jung, K. W. Tetrahedron Lett. 2004, 46, 143–146. (e) Godula, K.; Sames, D.
Science 2006, 312, 67–72.
(4) (a) Doyle, M. P.; Forbes, D. C.; Vasbinder, M. M.; Peterson, C. S. J.
Am. Chem. Soc. 1998, 120, 7653–7654. (b) Doyle, M. P.; Hu, W. H. Synlett
2001, 1364–1370.
6564 J. Org. Chem. 2009, 74, 6564–6571
Published on Web 07/28/2009
DOI: 10.1021/jo900998s
r
2009 American Chemical Society

Hansen et al.
JOCArticle
in Lewis acid-catalyzed processes.1i,5 The ability of such
complexes to effectively catalyze a variety of reactions at
low catalyst loadings has demonstrated their synthetic
potential, particularly in the context of chiral catalysis where
numerous examples of applications in syntheses of natural
products and biologically relevant molecules have been
reported.1d,3e,6 The ease at which simple modifications of
the bridging groups on the catalyst brings about dramatic
changes in reactivity makes these dirhodium complexes
preferable to mononuclear copper or palladium counter-
parts.
The development of dirhodium paddlewheel catalysts has
primarily focused on modifications of the electronic
and/or steric properties of the ligands. Within chiral
catalysis, the most successful classes are the N-arylsulfonyl-
prolinates,1b,m,7 phthalimide protected amino acid derivati-
ves,1m,8 carboxamidates,1i,k,m,9 ortho-metalated arylpho-
sphines,1m,10 and binaphthoyl phosphate complexes
(Figure 1).1m,n,11 More recently, alternative modifications
to the dirhodium scaffold have been explored in catalysis,
such as mixed valence dirhodium(II,III) species,12 complexes
with axially coordinated N-heterocyclic carbenes (NHCs),13
and also mixed ligand (heteroleptic) systems.14 Computa-
tional and kinetic studies of dirhodium carboxylate-cata-
lyzed carbenoid reactions have indicated that carbene
binding occurs only at one of the two rhodium active sites
at a time.15 This is supported by the observation that
dirhodium complexes immobilized via coordination of
FIGURE 1. Various dirhodium catalysts.
polymer-bound pyridine to one active site still display selec-
tivity and reactivity profiles comparable to those in homo-
geneous reactions.16 Dikarev and co-workers reported the
syntheses and structural characterization of a new family of
paddlewheel carboxylates in which one of the rhodium
atoms has been replaced by bismuth-a significantly less
expensive metal than rhodium.17 A preparative solid-state
technique was developed, in which the ability of bismuth(II)
trifluoroacetate to act as a metalloligand toward transition
metal fragments was utilized to prepare BiRh(O2CCF3)4 II,
a heterobimetallic, homoleptic carboxylate complex.17b,c
Further developments of the technique were utilized in
one-step syntheses of mixed-ligand carboxylates cis-BiRh-
(O2CCF3)2(O2CtBu)2 I, BiRh(O2CCF3)3(O2CCH3) III, and
the chiral complex BiRh(O2CCF3)3(O2C(S-(+)-iBu))
(Figure 2). The complexes have typical paddlewheel struc-
tural features, in which the bimetallic core has a bismuth-
rhodium single bond and acts as an anchor for four
μ2-ligands. The structures were shown to remain intact in
solution and displayed Lewis acidity toward basic donors at
the transition metal end only.17b,c
(5) (a) Doyle, M. P.; Phillips, I. M.; Hu, W. H. J. Am. Chem. Soc. 2001,
123, 5366–5367. (b) Doyle, M. P.; Valenzuela, M.; Huang, P. Proc. Natl.
Acad. Sci. U.S.A. 2004, 101, 5391–5395.
(6) (a) Davies, H. M. L. Angew. Chem., Int. Ed. 2006, 45, 6422–6425.
(b) Davies, H. M. L.; Gregg, T. M. Tetrahedron Lett. 2002, 43, 4951–4953.
(c) Davies, H. M. L.; Hansen, T.; Hopper, D. W.; Panaro, S. A. J. Am. Chem.
Soc. 1999, 121, 6509–6510. (d) Davies, H. M. L.; Stafford, D. G.; Hansen, T.
Org. Lett. 1999, 1, 233–236. (e) Doyle, M. P.; Hu, W.; Valenzuela, M. V. J.
Org. Chem. 2002, 67, 2954–2959. (f) Doyle, M. P.; Protopopova, M. N.;
Zhou, Q.-L.; Bode, J. W.; Simonsen, S. H.; Lynch, V. J. Org. Chem. 1995, 60,
6654–5. (g) Hinman, A.; Du Bois, J. J. Am. Chem. Soc. 2003, 125, 11510–
11511. (h) Kurosawa, W.; Kan, T.; Fukuyama, T. J. Am. Chem. Soc. 2003,
125, 8112–8113. (i) Taber, D. F.; Malcolm, S. C. J. Org. Chem. 2001, 66, 944–
953. (j) Wee, A. G. H. J. Org. Chem. 2001, 66, 8513–8517. (k) Yoon, C. H.;
Nagle, A.; Chen, C.; Gandhi, D.; Jung, K. W. Org. Lett. 2003, 5, 2259–2262.
(7) (a) Kennedy, M.; McKervey, M. A.; Maguire, A. R.; Roos, G. H. P. J.
Chem. Soc.,Chem. Commun. 1990, 361–362. (b) Davies, H. M. L.; Bruzinski,
P. R.; Lake, D. H.; Kong, N.; Fall, M. J. J. Am. Chem. Soc. 1996, 118, 6897–
6907.
(8) (a) Liang, C.; Collet, F.; Robert-Peillard, F.; Mueller, P.; Dodd, R. H.;
Dauban, P. J. Am. Chem. Soc. 2008, 130, 343–350. (b) Anada, M.; Kitagaki,
S.; Hashimoto, S. Heterocycles 2000, 52, 875–883. (c) Hashimoto, S.;
Watanabe, N.; Ikegami, S. Tetrahedron Lett. 1990, 31, 5173–5174. (d)
Hashimoto, S.-i.; Watanabe, N.; Ikegami, S. Synlett 1994, 353–355.
(9) Zalatan, D. N.; Du Bois, J. J. Am. Chem. Soc. 2008, 130, 9220–9221.
(10) (a) Barberis, M.; Perez-Prieto, J.; Herbst, K.; Lahuerta, P. Organo-
metallics 2002, 21, 1667–1673. (b) Estevan, F.; Herbst, K.; Lahuerta, P.;
Barberis, M.; Perez-Prieto, J. Organometallics 2001, 20, 950–957.
(11) McCarthy, N.; McKervey, M. A.; Ye, T.; McCann, M.; Murphy, E.;
Doyle, M. P. Tetrahedron Lett. 1992, 33, 5983–5986.
(12) (a) Catino, A. J.; Nichols, J. M.; Forslund, R. E.; Doyle, M. P. Org.
Lett. 2005, 7, 2787–2790. (b) Catino, A. J.; Nichols, J. M.; Choi, H.;
Gottipamula, S.; Doyle, M. P. Org. Lett. 2005, 7, 5167–5170.
(13) (a) Trindade, A. F.; Gois, P. M. P.; Veiros, L. F.; Andre, V.; Duarte,
M. T.; Afonso, C. A. M.; Caddick, S.; Cloke, F. G. N. J. Org. Chem. 2008, 73,
4076–4086. (b) Gois, P. M. P.; Trindade, A. F.; Veiros, L. F.; Andre, V.;
Duarte, M. T.; Afonso, C. A. M.; Caddick, S.; Cloke, F. G. N. Angew. Chem.,
Int. Ed. 2007, 46, 5750–5753.
(14) (a) Lou, Y.; Horikawa, M.; Kloster, R. A.; Hawryluk, N. A.; Corey,
E. J. J. Am. Chem. Soc. 2004, 126, 8916–8918. (b) Lou, Y.; Remarchuk, T. P.;
Corey, E. J. J. Am. Chem. Soc. 2005, 127, 14223–14230.
(15) (a) Nakamura, E.; Yoshikai, N.; Yamanaka, M. J. Am. Chem. Soc.
2002, 124, 7181–7192. (b) Pirrung, M. C.; Liu, H.; Morehead, A. T. Jr. J. Am.
Chem. Soc. 2002, 124, 1014–1023.
The modification introduced in Bi-Rh systems offers the
possibility of tuning the reactivity of the bimetallic core
itself.17c This is a novel approach to reactivity control in
such systems. More importantly, the availability of the
heterobimetallic carboxylates suggests a unique opportunity
to probe the role of the effective dirhodium framework
through comparative studies of the catalytic performance
(16) (a) Davies, H. M. L.; Walji, A. M. Org. Lett. 2003, 5, 479–482. (b)
Davies, H. M. L.; Walji, A. M. Org. Lett. 2005, 7, 2941–2944. (c) Nagashima,
T.; Davies, H. M. L. Org. Lett. 2002, 4, 1989–1992.
(17) (a) Durivage, J. C.; Gruhn, N. E.; Li, B.; Dikarev, E. V.; Lichten-
berger, D. L. J. Cluster Sci. 2008, 19, 275–294. (b) Dikarev, E. V.; Gray, T.
G.; Li, B. Angew. Chem., Int. Ed. 2005, 44, 1721–1724. (c) Dikarev, E. V.; Li,
B.; Zhang, H. T. J. Am. Chem. Soc. 2006, 128, 2814–2815.
J. Org. Chem. Vol. 74, No. 17, 2009 6565

Hansen et al.
JOCArticle
TABLE 1. Scope of Carbenoid Structure in Cyclopropanation of
Styrene
entry compd
R =
Ph-
Ph-
Ph-
H-
R0 =
catalyst de (%)a yield (%)b
1
2
3
4
5
a
a
a
b
c
-CH3
I
94
>95
74
94
96
47
-CH3
II
Rh2(TFA)4 >95
-CH3
-CH2-CH3
I
I
21
N/A
MeO2C- -CH3
N/R
aMeasured by 1H NMR of crude reaction mixture. bIsolated yields.
FIGURE 2. Bi-Rh carboxylate complexes.17
TABLE 2. Reaction between 6a and Furan
of the two systems. Such an investigation would reveal
whether the dirhodium framework is a necessity to achieve
efficient, catalytic reactions as well as the high levels of
selectivity that are often observed. In this paper, we report
a combined experimental and computational study to ex-
plore catalysis by heterobimetallic Bi-Rh carboxylates in
the context of metal carbenoid chemistry. We will demon-
strate that such species can act as catalysts for carbenoid-
mediated transformations and, furthermore, show how they
compare with analogous dirhodium catalysts in terms of
reactivity and selectivity. The computational studies provide
an evaluation of how the Bi-Rh core affects the mechanism
in terms of catalytic activity and reactivity.
entry
catalyst
product ratio (8:9)a
yield (%)b
1
2
3
I
70:30
47:53
44:56
80
96
95
II
Rh2(TFA)4
aMeasured by 1H NMR of crude reaction mixture. bOverall isolated
yields.
Results and Discussion
Catalytic Activity: Cyclopropanation Reactions. Catalytic
activities of Bi-Rh complexes I and II were evaluated in the
cyclopropanation reaction between styrene and methyl phe-
nyldiazoacetate (6a, Table 1). The optimal conditions were
determined to be 2 mol % catalyst under reflux when using
dichloromethane as solvent. Full conversion was observed
and the cyclopropane was formed in excellent yield (94%)
and high diastereoselectivity (>95% de) with complex II
as catalyst (Table 1, entry 2). Rh2(O2CCF3)4 performed
similarly in this reaction with 96% yield and >95% de
(entry 3). Complex I gave somewhat lower yield (74%), but
still very high diastereoselectivity (entry 1).
The effect of carbenoid structure was investigated with
complex I. The acceptor carbenoid derived from ethyl dia-
zoacetate (entry 4) produced the corresponding cyclopro-
pane in 47% yield as a mixture of diastereomers (21% de).
The diastereoselectivity is similar to previously reported
values for this reaction when catalyzed by dirhodium cata-
lysts.18 The acceptor-acceptor carbenoid derived from
methyl diazomalonate was presumably not formed under
the reaction conditions, since starting materials were reco-
vered, even after 2 days at 40 °C (entry 5). Overall, the best
result was obtained with methyl phenyldiazoacetate (6a) as
carbenoid precursor and this diazo compound was therefore
used for further studies. The success of the reactions shows
that Bi-Rh complexes are capable of forming the carbenoid
species; however, it was qualitatively observed that the
Bi-Rh-catalyzed reactions appeared to proceed slower than
those catalyzed by the dirhodium counterpart. The lack of
other observable byproducts indicates that the intermediate
displays similar selectivity to dirhodium carbenoids.
The decomposition of methyl phenyldiazoacetate (6a) in
neat furan catalyzed by dirhodium catalysts affords pro-
ducts of monocyclopropanation (8) as well as an alkyla-
tion product (9) where the furan ring has been opened.19
The latter is believed to be formed via a mechanism in which
the stabilization of charge build-up by the furan ring inter-
rupts the cyclopropanation and forms a zwitterionic inter-
mediate capable of undergoing the furan-unraveling reaction
competitively.19b By increasing the electrophilic character of
the catalyst, more of the ring-opened product is normally
observed.19a When the reaction was catalyzed by the rela-
tively electron-rich bismuth-rhodium complex I, the two
products 8 and 9 were formed in a 70:30 ratio in 80% yield
(Table 2, entry 1). The more electron-deficient complex II
afforded an approximately 47:53 ratio of the two products in
an excellent 96% overall yield (entry 2). Rh2(O2CCF3)4
performed essentially identical with the latter (entry 3). For
these reactions, bismuth-rhodium carbenoids again appear
to display similar chemoselectivities to their dirhodium
counterparts.
(19) (a) Davies, H. M. L.; Hedley, S. J. Chem. Soc. Rev. 2007, 36, 1109–
1119. (b) Hedley, S. J.; Ventura, D. L.; Dominiak, P. M.; Nygren, C. L.;
Davies, H. M. L. J. Org. Chem. 2006, 71, 5349–5356.
(18) Doyle, M. P.; Griffin, J. H.; Bagheri, V.; Dorow, R. L. Organome-
tallics 1984, 3, 53–61.
6566 J. Org. Chem. Vol. 74, No. 17, 2009

Hansen et al.
JOCArticle
FIGURE 4. Hammet plot of cyclopropanation catalyzed by Rh2-
(O2CCF3)3(O2CCH3) and III.
SCHEME 1. Competition Studies
FIGURE 3. Disappearance of NdN stretch frequency of diazo
compound as function of time for (a) Rh2(O2CCF3)3(O2CCH3)
(0.2 mol %, over 45 s) and (b) complex III (2 mol %, over 62 min).
The catalyst was injected at t0=10 s.
Reactivity. To investigate the difference in reactivity be-
tween the two catalyst systems toward decomposition of
phenyldiazoacetate, cyclopropanation reactions between
methyl phenyldiazoacetate and styrene, catalyzed by com-
plex III and Rh2(O2CCF3)3(O2CCH3), respectively, were
conducted with all the reagents present in the flask. After
injection of the catalyst into a solution of the reagents in
dichloromethane, the reaction progress was monitored by
ReactIR, following the disappearance of the diazo NdN
stretch frequency at ∼2100 cm-1. The reaction progress is
shown in Figure 3a for Rh2(O2CCF3)3(O2CCH3) and in
Figure 3b for III. The reaction catalyzed by Rh2(O2CCF3)3-
(O2CCH3) went to completion after ∼16 s at 0.2 mol %
catalyst loading (∼500 turnovers). The reaction catalyzed by
2 mol % III went to completion after about 45 min (∼50
turnovers). The average turnover frequencies for the two
SCHEME 2. C-H Activation of 1,4-Cyclohexadiene
obtained from the literature21 for both catalyst systems
(Figure 4). Rh2(O2CCF3)3(O2CCH3) appears to give a
slightly more chemoselective carbenoid (F=-1.20 ( 0.02)
than III (F=-1.00 ( 0.03).
C-H Insertion. 1,4-Cyclohexadiene is a highly activated
system for C-H insertion due to hyperconjugative interac-
tions between the methylene group and the two pendant
double bonds.6d,22 The reaction proceeds cleanly with both
Rh2(O2CCF3)4 and bismuth-rhodium complex II under the
standard reaction conditions (Scheme 2). An 84% yield of 12
was isolated when the reaction was catalyzed by complex II,
comparable to its dirhodium analogue (74% yield). This
reaction demonstrates that the heterobimetallic complexes
are capable of effectively catalyzing C-H insertion pro-
cesses.
catalysts are TOFRhRh ≈ 31 s-1 and TOFBiRh ≈ 0.019 s-1
,
respectively. The dirhodium complex is therefore about
∼1600 times more reactive than the Bi-Rh counterpart
toward diazo decomposition. Even though the rate of the
Bi-Rh-catalyzed reaction is significantly slower than that
catalyzed by dirhodium, the former can still be a high-
yielding process, because under standard reaction conditions
the addition of the diazo compound is the rate-limiting process.
Electronic Effects. To probe the properties of dirhodium
and bismuth-rhodium carbenoids, competition studies were
carried out in order to quantify the chemoselectivity profiles.
The cyclopropanation of styrene vs. para-substituted styr-
enes gives a measure of the level of benzylic charge build-up
in the cyclopropanation transition state via the Hammet
reaction constant (F) obtained for a series of para substitu-
ents (Scheme 1).20 Good correlations were found when
log(kR/kH) was plotted against σ+ substituent constants
Due to the effectiveness of complex II in catalyzing C-H
insertions, the catalyst was also tested with a substrate where
C-H insertion occurs in competition with cyclopropanation
chemistry. One such substrate is cyclohexene, where both
reaction channels are important with most dirhodium
(21) Hansch, C.; Leo, A.; Taft, R. W. Chem. Rev. 1991, 91, 165–195.
(22) Muller, P.; Tohill, S. Tetrahedron 2000, 56, 1725–1731.
(20) Davies, H. M. L.; Panaro, S. A. Tetrahedron 2000, 56, 4871–4880.
J. Org. Chem. Vol. 74, No. 17, 2009 6567

Hansen et al.
JOCArticle
TABLE 3. C-H Insertion vs. Cyclopropanation
TABLE 4. Vinylogous vs. Carbenoid Reactivity
product ratio
(16:17)a
Z/E-ratio
(17)a
yield (%)b
entry
catalyst
product ratio (13:14)a
yield (%)b
entry
catalyst
1
2
3
II
79:21
84:16
75:25
52
57
1
2
3
Rh2(OAc)4
Rh2(TFA)3(OAc)
III
63:37
26:74
50:50
2.7:1
1.3:1
1.5:1
79
82
80
Rh2(TFA)4
Rh2(OAc)4
50c
aMeasured by 1H NMR of crude reaction mixture. bOverall isolated
aMeasured by 1H NMR of crude reaction mixture. bOverall isolated
yield.
yield. cFrom ref 22.
catalyst systems (Table 3).22 The decomposition of 6a in the
presence of cyclohexene and 2 mol % of complex II afforded
a 79:21 mixture of C-H insertion product 13 versus cyclo-
propane 14 in an overall 52% yield. This is almost identical
with the results obtained with Rh2(O2CCF3)4, which gave a
57% overall yield. For comparison, Rh2(O2CCH3)4 has been
reported to give a 75:25 ratio of the two products in an
overall 50% yield.22 These results further support that the
selectivity of the Bi-Rh carbenoid is very similar to that of
analogous dirhodium catalysts.
SCHEME 3. Mechanism for Catalytic Cyclopropanation by
Dirhodium Complexes1k,15,25
Vinylogous Reactivity. Transient metal vinylcarbenes can
display electrophilic reactivity at the vinylogous position in
addition to at the carbenoid carbon.23 A test reaction for
these competing pathways is the reaction between unsubsti-
tuted vinyldiazoacetate 15 and cyclopentadiene (Table 4)
where product 16 is derived from a tandem cyclopropana-
tion-Cope rearrangement (carbenoid reactivity) and 17
from vinylogous reactivity.23b,e This mode of reactivity has
been observed to increase with the electron-withdrawing
capability of the catalyst, solvent polarity, and size of the
ester group.23b The reaction of 15 with cyclopentadiene
catalyzed by rhodium-bismuth complex III gave a 50:50
ratio of the products in overall good yield (80%). The
analogous reaction with Rh2(O2CCF3)3(O2CCH3) gave a
26:74 ratio of 16 to 17 in overall 82% yield. For comparison,
Rh2(O2CCH3)4 gave a 63: 37 ratio in 79% yield. These
results indicate that the rhodium-bismuth carbenoid
behaves as a slightly more electron-rich system since it dis-
plays less vinylogous reactivity. While the overall dif-
ference is not large, as evident from the aforementioned
Hammet studies, the occurrence of vinylogous reactivity is
particularly sensitive to the structure of the vinylcarbenoid
complex.23b
density functional calculations. Computational details can
be found in the Computational Methods section. The calcu-
lated reaction pathway involving phenyldiazoacetate and
styrene catalyzed by Rh2(O2CH)4 was recently reported.24
The currently accepted mechanism for reactions of
diazo compounds with dirhodium complexes was con-
sidered.1k,15,25 This involves complexation of the diazo-
acetate to the bimetallic catalyst, rate limiting nitrogen
extrusion to form the intermediate metal carbenoid complex,
and finally trapping by an alkene to form a cyclopropane
product (Scheme 3).1k,15,25 Alternative mechanisms have
been suggested in the literature, involving carboxylate ligand
dissociation from the dirhodium core, but were not con-
sidered here since these are not yet widely accepted.14,26
Figure 5 shows the reaction pathway calculated for the
cyclopropanation reaction of methyl phenyldiazoacetate
with styrene catalyzed by both dirhodium formate24 (red
pathway) and bismuth-rhodium formate (blue pathway).
The former path was recently reported and is included here
for comparison.24 The calculated structures for the Bi-Rh
pathway are indicated in the figure and are analogous to
the dirhodium structures. Coordination of the diazo com-
pound to the catalyst to form 18 is an exothermic process in
both cases, but less so for the Bi-Rh system (-1.8 kcal/mol
DFT Calculated Reaction Pathways. Reaction pathways
for carbenoid formation and trapping were studied compu-
tationally for model systems. The cyclopropanation reaction
of styrene with methyl phenyldiazoacetate (6a) with both
dirhodium and Bi-Rh model catalysts was investigated in
(23) (a) Yue, Y.; Wang, Y.; Hu, W. Tetrahedron Lett. 2007, 48, 3975–
3977. (b) Davies, H. M. L.; Hu, B.; Saikali, E.; Bruzinski, P. R. J. Org. Chem.
1994, 59, 4535–4541. (c) Davies, H. M. L.; Saikali, E.; Clark, T. J.; Chee, E.
H. Tetrahedron Lett. 1990, 31, 6299–6302. (d) Davies, H. M. L.; Yokota, Y.
Tetrahedron Lett. 2000, 41, 4851–4854. (e) Sevryugina, Y.; Weaver, B.;
Hansen, J.; Thompson, J.; Davies, H. M. L.; Petrukhina, M. A. Organome-
tallics 2008, 27, 1750–1757.
(24) Hansen, J.; Autschbach, J.; Davies, H. M. L. Submitted for publica-
tion.
(25) Yates, P. J. Am. Chem. Soc. 1952, 74, 5376–5381.
(26) (a) Howell, J. A. S. Dalton Trans. 2007, 3798–3803. (b) Howell, J. A.
S. Dalton Trans. 2007, 1104–1114.
6568 J. Org. Chem. Vol. 74, No. 17, 2009

Hansen et al.
JOCArticle
FIGURE 5. Potential energy surfaces (relative to free reactants) at the B3LYP/6-311G(2d,2p)[Rh-RSC+4f][Bi-RLC]//B3LYP/6-31G*[Rh-
RSC+4f][Bi-RLC] level for dirhodium24 and Bi-Rh models.
vs. -7.4 kcal/mol for dirhodium). The loss of nitrogen from
the Bi-Rh complex via TS-I has a potential energy barrier of
19.2 kcal/mol, considerably higher than that for dirhodium
(11.3 kcal/mol). As shown before, the Bi-Rh-catalyzed
reactions were much slower (∼1600 times) than the corre-
sponding dirhodium catalyzed reactions. However, this dif-
ference in reaction rate would correspond to a smaller
difference in barrier heights of the rate-limiting step than
the computed 7.9 kcal/mol difference. One reason why the
difference in the barriers is overestimated could be that the
formate model may not be totally appropriate for the Bi-Rh
system. Other factors discussed below may also change the
difference in barrier heights. One striking difference between
the two systems is that carbenoid formation is endothermic
for Bi-Rh (+2.5 kcal/mol), but quite exothermic for dirho-
dium (-9.1 kcal/mol). Relative to the reactants, the Bi-Rh
carbenoid complex 19 (M=Bi) is less stable by 0.7 kcal/mol,
whereas the dirhodium carbenoid 19 (M=Rh) is stabilized
by -16.5 kcal/mol. Despite this difference in thermodynamic
stability, the kinetic stabilities of the two are not very
different (3.6 kcal/mol barrier for cyclopropanation via
TS-II for the Bi-Rh system vs. 4.5 kcal/mol for dirhodium).
The difference in kinetic stabilities is presumably due to a
reduced π-backbonding component in the Bi-Rh carbenoid
complex.24 The calculations are consistent with the Hammet
study, which showed only minor differences in selectivity
between the two catalyst systems.
In light of the large calculated difference in potential
energy barriers for the rate-determining steps of the two
catalyst systems, which is in poor agreement with our
experimental findings, it is likely that other effects may be
involved that are not reflected in the described computa-
tional models. One such effect is the possibility that the
dirhodium catalysts have an axially coordinated ligand on
the second rhodium during the catalytic cycle.28 No studies
have been reported, to our knowledge, that explicitly de-
scribe the potential influence of such ligands on reactions at
the other metal active site. Pirrung and co-workers could not
demonstrate any evidence for this pathway from kinetic
data, but only a few systems were investigated.15b Since
relatively electron-deficient catalysts were used in the present
study, it is conceivable that the catalytic cycle may proceed
with an axial ligand at the other rhodium center. This ligand
may be solvent, adventitious water, or other coordinating
groups in reactants or products (π-bonds, carbonyl groups).28d
For instance, carbonyl groups in the product, which is
formed rapidly, could be a major contributor to this effect.29
Bi-Rh complexes have been shown to exhibit Lewis acidity
toward basic donor groups at the Rh end only,17c hence, the
system because of the relatively long Rh-C bond length
27
˚
˚
(2.120 A vs. 2.010 A in dirhodium). The presence of a
significant carbenoid trapping barrier for the Bi-Rh carbe-
noid is a consequence of the donor group on the carbenoid
(28) (a) Pirrung, M. C.; Morehead, A. T. J. Am. Chem. Soc. 1996, 118,
8162–8163. (b) Pirrung, M. C.; Morehead, A. T. J. Am. Chem. Soc. 1994, 116,
8991–9000. (c) Cotton, F. A.; Hillard, E. A.; Liu, C. Y.; Murillo, C. A.;
Wang, W. N.; Wang, X. P. Inorg. Chim. Acta 2002, 337, 233–246. (d) Drago,
R. S.; Long, J. R.; Cosmano, R. Inorg. Chem. 1982, 21, 2196–2202.
(29) Petrukhina, M. A.; Andreini, K. W.; Walji, A. M.; Davies, H. M. L.
Dalton Trans. 2003, 4221–4223.
(27) Dikarev, E. V.; Li, B.; Rogachev, A. Y.; Zhang, H.; Petrukhina, M.
A. Organometallics 2008, 27, 3728–3735.
J. Org. Chem. Vol. 74, No. 17, 2009 6569

Hansen et al.
JOCArticle
FIGURE 7. Potential energy surfaces (relative to free reactants)
at the B3LYP/6-311G(2d,2p)[Rh-RSC+4f][Bi-RLC]//B3LYP/
6-31G*[Rh-RSC+4f][Bi-RLC] level for trifluoroacetate bridged
complexes.
FIGURE 6. Influence of axial coordination on reaction pathway
compared to the coordination-free system.24 Potential energy sur-
faces (relative to free reactants) at the B3LYP/6-311G(2d,2p)-
[Rh-RSC+4f]//B3LYP/6-31G*[Rh-RSC+4f] level.
role of a coordinating ligand on dirhodium for reasons of com-
putational cost. The reaction pathways remain qualitatively
the same as before, but significant potential energy differ-
ences can be observed. Coordination of the diazo compound
to the catalyst complexes has become much more exothermic
compared to the formate model systems (-5.4 kcal/mol for
Bi-Rh and -13.3 kcal/mol for dirhodium). Since the intro-
duction of the trifluoroacetate ligands rendered the diazo
coordination process much more exothermic, it is likely that,
at least very electron-deficient dirhodium complexes such as
Rh2(O2CCF3)4 undergo the catalytic cycle with axially co-
ordinating groups at the second rhodium center, as discussed
above. For BiRh(O2CCF3)4, the nitrogen extrusion barrier is
only 1.6 kcal/mol lower than that for the formate system
(17.6 kcal/mol vs. the previously calculated 19.2 kcal/mol).
This barrier decreased by 2.1 kcal/mol for the dirhodium
system by introducing the trifluoroacetate ligands, a similar
reduction to the former. The difference between rate-limiting
barriers for the two systems remains about the same. The
thermodynamic stabilities of the carbenoid complexes have
changed significantly. For Bi-Rh, the carbenoid formation
is now slightly exothermic (-0.4 kcal/mol), whereas it is
much more exothermic for dirhodium (-11.4 kcal/mol). In
accordance with more electrophilic systems, the cyclopropa-
nation barriers are somewhat lower compared to those
calculated in the former models, but the trend remains the
same (Bi-Rh has a slightly lower barrier in the product-
determining step).
Charge Distribution. Figure 8 shows the calculated charges
on the two metal atoms in dirhodium formate (Rh1 = blue,
Rh2 = red) and Bi-Rh formate (Rh3 = violet, Bi = orange)
along the reaction pathyway for the cyclopropanation reac-
tion. For the rhodium formate-catalyzed pathway, coordi-
nation of the diazoacetate leads to charge reduction for
Rh2 (red line) by about -0.2 units. This is in accordance
with observations by Nakamura and co-workers in a
computational study of diazomethane and methyl diazoace-
tate.15a The drop in charge signifies reduction of Rh2 and
shows that it can act as an electron sink during the catalytic
cycle.15a For Bi-Rh formate, both Rh and Bi become only
partially reduced during the coordination event. However,
effect is absent in this system. To study the effects of this
phenomenon for dirhodium, the reaction pathway was
recalculated for the dirhodium tetraformate system with an
acetone molecule coordinated to the second rhodium as a
model system. A comparison of the two pathways is given in
Figure 6, with the catalyst structure indicated. For this
catalyst system, the coordination of the diazo compound is
less exothermic (-3.1 kcal/mol vs. -7.4 kcal/mol without
axial ligand). A very large effect is seen in the nitrogen
extrusion barrier (18 f TS-I), which is now increased to
16.0 kcal/mol, 4.7 kcal/mol higher than previously calcu-
lated. The now much smaller difference in this barrier
between dirhodium and Bi-Rh systems (3.2 kcal/mol) is
more consistent with the observed decomposition rate dif-
ference. An estimate of the difference in activation energies
for the two catalysts can be made from the kinetic data
reported above (for Rh2(O2CCF3)3(O2CCH3) and complex
III) by applying the Arrhenius equation. A reactivity differ-
ence of ∼1600 corresponds to roughly 4 kcal/mol difference
in activation barriers-in good agreement with the calcula-
tions considering that the computational model affords a
number of approximations, and considering the error bars of
density functional calculations.
In the coordinated dirhodium system, the carbenoid com-
plex formation is now less exothermic, only -3.1 kcal/mol.
Another important observation is that the potential energy
barrier for the product-determining step (19 f TS-II) is
virtually unchanged by the presence of the coordinating
ligand (4.6 kcal/mol vs. 4.5 kcal/mol). This suggests that
the selectivity of such reactions is not much modified by the
axial ligand, in agreement with previously reported studies
with immobilized dirhodium complexes.16
The significance of the electron-withdrawing nature of the
catalyst ligands was also investigated. Currently, only sys-
tems with strongly electron-withdrawing ligands are avail-
able for the Bi-Rh core (trifluoroacetate), and have
therefore been used in the experimental part of this
study.17b,c The reaction pathway was calculated by using
Rh2(O2CCF3)4 and BiRh(O2CCF3)4 as catalysts (Figure 7).
The calculations were conducted without reinvestigating the
6570 J. Org. Chem. Vol. 74, No. 17, 2009

Hansen et al.
JOCArticle
catalytic process similar to what has previously been de-
scribed for dirhodium species.
Computational Methods
All calculations were performed with the Gaussian ’03 soft-
ware package,30 using Density Functional Theory (the three-
parameter hybrid functional B3LYP31) to locate stationary
points on the potential energy surface. The structures were
geometry optimized with a basis set consisting of the 1997
Stuttgart relativistic small-core ECP and corresponding basis
set [Stuttgart RSC 1997 ECP]32 for Rh, augmented with a
4f-function (ζf(Rh) = 1.350),24 the 1997 Stuttgart relativistic
large-core ECP and basis set [Stuttgart RLC 1997 ECP] for Bi,33
and the split valence basis set 6-31G* for all other atoms
(C, H, N, O, and F). This composite basis set is referred to
as 6-31G*[Rh-RSC+4f][Bi-RLC]. In this work, the discussion
is based on single-point potential energies calculated at
the B3LYP/6-311G(2d,2p)[Rh-RSC+4f][Bi-RLC]//B3LYP/6-
31G*[Rh-RSC+4f][Bi-RLC] level. Stationary points were char-
acterized by normal coordinate analysis at the 6-31G*[Rh-RSC
+4f][Bi-RLC] level of theory.34 Transition states were con-
firmed to have a single imaginary vibrational mode whereas
equilibrium structures were confirmed to have no imaginary
vibrational modes. Transition states were further characterized
by intrinsic reaction coordinate (IRC) analysis using default
parameters with subsequent geometry optimization to confirm
that the stationary points were smoothly connected to each
other. Basis set definitions and pseudopotential parameters
were obtained from the EMSL basis set exchange library.35
Calculated structures have been visualized with use of
Mercury.36
FIGURE 8. Charge distribution on heavy atoms as a function of
reaction coordinate for dirhodium and Bi-Rh formate model systems.
upon formation of the carbenoid complex a reduction of
about -0.2 charge units occurs for Bi, very similar to that of
Rh2 in the dirhodium system. This demonstrates that Bi
can also act as a redox pool, facilitating the catalytic
cycle. Further evidence for this comes from the elongation
˚
of the Rh-Bi bond during the reaction from 2.582 A to
2.649 A.
˚
Conclusions
In this paper, we have demonstrated that heterobimetallic
bismuth-rhodium paddlewheel carboxylate complexes can
act as effective catalysts for metal carbenoid transformations
at low catalyst loadings. The catalysts have been shown to
effectively mediate cyclopropanation reactions and C-H
insertion processes as well as reactions involving zwitterionic
intermediates with donor/acceptor diazo compounds. The
heterobimetallic complexes display similar chemo- and dia-
stereoselectivities as their dirhodium analogues.
BiRh(O2CCF3)3(O2CCH3) decomposes methyl phenyl-
diazoacetate about ∼1600 times slower than the correspond-
ing dirhodium complex in the cyclopropanation reaction
with styrene. DFT calculations of the reaction pathways with
model systems are consistent with the observed reactivity
difference only if axial coordination to the second rhodium
atom in the dirhodium catalyst model is taken into account.
Furthermore, the calculations support the observation that
the selectivity profile of Bi-Rh donor/acceptor carbenoids is
similar to that of analogous dirhodium carbenoids. A charge
distribution analysis shows that the rhodium-bismuth core
can act as a bifunctional electron pool, facilitating the
Acknowledgment. Support of this work by the National
Science Foundation (CHE-0750273 for H.M.L.D., CHE-
0619422 for E.V.D., and CHE-0447321 for J.A.) is gratefully
acknowledged. Support was also provided by the Center for
Computational Research at the University at Buffalo. We
also wish to acknowledge the preliminary studies on this
project conducted by Dr. Janelle Thompson.
Supporting Information Available: Complete ref 30, experi-
mental details and characterization data for all new compounds,
and Cartesian coordinates, structures, and a summary of cal-
culated parameters for all new structures reported. This material
is available free of charge via the Internet at http://pubs.acs.org.
(33) Kuechle, W.; Dolg, M.; Stoll, H.; Preuss, H. Mol. Phys. 1991, 74, 1245.
(34) Foresman, J. B.; Frisch, A. Exploring Chemistry with Electronic
Structure Methods; Gaussian, Inc.: Pittsburgh, PA, 1993.
(35) (a) Schuchardt, K. L.; Didier, B. T.; Elsethagen, T.; Sun, L.;
Gurumoorthi, V.; Chase, J.; Li, J.; Windus, T. L. J. Chem. Inf. Model.
2007, 47, 1045–1052. (b) Feller, D. J. Comput. Chem. 1996, 17, 1571–1586.
(36) (a) Bruno, I. J.; Cole, J. C.; Edgington, P. R.; Kessler, M.; Macrae, C.
F.; McCabe, P.; Pearson, J.; Taylor, R. Acta. Crystallogr. B 2002, 58, 389–
397. (b) Taylor, R.; Macrae, C. F. Acta. Crystallogr. B 2001, 57, 815–827. (c)
Macrae, C. F.; Bruno, I. J.; Chisholm, J. A.; Edgington, P. R.; McCabe, P.;
Pidcock, E.; Rodriguez-Monge, L.; Taylor, R.; van de Streek, J.; Wood, P. A.
J. Appl. Crystallogr. 2008, 41, 466–470. (d) Macrae, C. F.; Edgington, P. R.;
McCabe, P.; Pidcock, E.; Shields, G. P.; Taylor, R.; Towler, M.; van de
Streek, J. J. Appl. Crystallogr. 2006, 39, 453–457.
(30) Gaussian 03, Revision C.02; Frisch, M. J.; et al. See the Supporting
Information for the complete reference.
(31) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652. (b) Lee, C.;
Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785–789.
(32) (a) Kaupp, M.; Schleyer, P. v. R.; Stoll, H.; Preuss, H. J. Chem. Phys.
1991, 94, 1360. (b) Dolg, M.; Stoll, H.; Preuss, H.; Pitzer, R. M. J. Phys.
Chem. 1993, 97. (c) Bergner, A.; Dolg, M.; Kuechle, W.; Stoll, H.; Preuss, H.
Mol. Phys. 1993, 80, 1431.
J. Org. Chem. Vol. 74, No. 17, 2009 6571