Anti-Markovnikov Hydrofunctionalization of Olefins Mediated by Rhodium-Porphyrin Complexes

Article abstract of DOI:10.1002/anie.200351941

A rationally designed mechanistic approach to anti-Markovnikov olefin hydrofunctionalization and its application to the synthesis of heterocycles are described. Porphyrin-rhodium complexes have been shown to exhibit remarkable reactivity and selectivity for each step of the proposed catalytic cycle (see scheme). A critical step of this reaction sequence is a new, facile, and remarkably general carbon-heteroatom bond-forming reductive elimination.

Full text of DOI:10.1002/anie.200351941

Communications
Alkene Functionalization
Anti-Markovnikov Hydrofunctionalization of
Olefins Mediated by Rhodium–Porphyrin
Complexes**
Melanie S. Sanford and John T. Groves*
The development of catalysts for the anti-Markovnikov
addition of H-Nu (H-Nu = H-NR2, H-OR) to olefins could
provide an atom-economical approach to amines, ethers, and
heterocyclic compounds [Eq. (1)].[1] These products are of
Scheme 1. Rhodium–porphyrin-mediated olefin hydrofunctionalization.
The first step of the process involves a reversible olefin
insertion into (TPP)Rh-H to produce a s-alkyl rhodium
complex (Scheme 1, step a).[9] As summarized in Table 1, the
desired reaction proceeded efficiently (within 1 h at 258C) to
Table 1: Olefin insertion into (TPP)Rh-H.[a]
significant utility to both the chemical and pharmaceutical
industries, and as a result the development of new catalysts for
this type of transformation has been identified as one of the
“top ten challenges for catalysis”.[2] However, despite consid-
erable effort in this area over the past two decades,[3–5] general
and selective processes and catalysts for such transformations
have remained elusive. This report describes the rational
design of a new mechanistic pathway, which offers a
potentially general solution to anti-Markovnikov olefin
hydrofunctionalization.
Entry
Substrate
Product
Yield [%][b]
1
2
88
74
3
91
4
5
76
83
We envisaged a three-step catalytic cycle for intramolec-
ular anti-Markovnikov olefin hydrofunctionalization involv-
[a] Conditions: 258C; 12 h; [Rh]~0.01m; [olefin]~0.10m. In all cases the
yield of the reaction was >90% as measured by 1H NMR spectroscopy.
[b] Yield of isolated product.
À
ing: a) olefin insertion into a [Rh] H bond, b) intramolecular
functionalization of the resulting s-alkyl adduct, and c) pro-
tonation of the resulting [RhI]À species to regenerate the
starting [Rh]-H complex (Scheme 1). We anticipated that the
regioselectivity of the reaction would be dictated by the
olefin-insertion step, which should favor the less-substituted
metal s-alkyl species[6] and thereby selectively provide the
anti-Markovnikov product. We report herein that (TPP)Rh-
H (TPP = tetraphenylporphyrin) efficiently mediates each
step of this postulated catalytic cycle to afford heterocyclic
products with extremely high (> 97%) anti-Markovnikov
regioselectivity. The functionalization step serves as a rare
and unusually general example of direct C(sp3)–heteroatom
bond-forming reductive elimination.[7,8]
afford a variety of stable rhodium–alkyl products.[10] This
transformation was highly regioselective, and in all cases the
less-substituted metal–alkyl species was the only insertion
1
product observed by H NMR spectroscopy.[10] Furthermore,
yields remained excellent in the presence of a diverse array of
functional groups, including alcohols, aldehydes, carboxylic
acids, sulfonamides, and nitroalkanes (Table 1, entries 1–5).
This extraordinary functional-group tolerance is particularly
notable because the insertion mechanism involves reactive
RhII radicals as intermediates.[9,11]
The second step involves reaction of the s-alkyl–rhodium
species with an internal nucleophile (Nu; Scheme 1, step b).
À
C Nu bond-forming reductive elimination reactions are
extremely rare, particularly when the a carbon atom is sp3
hybridized and/or b hydrogen atoms are present.[6,7,12] We
note several previous examples of this transformation that
involve methyl–rhodium–porphyrin complexes (e.g., Nu =
PPh3[12a] or RhI[12b]).
[*] Dr. M. S. Sanford, Prof. J. T. Groves
Department of Chemistry, Princeton University
Princeton, NJ 08544 (USA)
Fax: (+1)626-564-9297
E-mail: jtgroves@princeton.edu
Initial investigations focused on the intramolecular cycli-
zation of (TPP)Rh(CH2)4OH to produce THF (Table 2,
entries 1-3). The pendant alcohol was not sufficiently
nucleophilic to cleave the Rh–alkyl bond under neutral
[**] This research was supported by the National Science Foundation
(CHE 9814301). M.S.S. also thanks the NIH for a postdoctoral
fellowship (GM64158-02).
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
À
conditions. However, the addition of base led to rapid C O
588
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/anie.200351941
Angew. Chem. Int. Ed. 2004, 43, 588 –590

Angewandte
Chemie
Table 2: Carbon–oxygen bond-forming reductive elimination.[a]
Table 3: Formation of O and N heterocycles and nitropentane.[a]
Entry
1[b]
Substrate
Product
Yield [%]
73
Entry Substrate
Solvent
C6D6
Base
Yield [%]
A
B
1[b]
2[b]
3[c]
KOtBu
29
60
78
71
40
22
2[b]
>95
[D6]DMSO KOtBu
[D6]DMSO NMe4OH
3[b]
>95
4[c]
[D6]DMSO NMe4OH
70
30
4[c]
69
53
5[b,d]
C6D6
KOtBu
>97
<3
6[c,e]
7[c]
[D6]DMSO NMe4OH
[D6]DMSO NMe4OH
<3
<3
>97
>97
5[b,d]
6[c]
7[c]
8[c]
83
92
[a] Reactions were quantitative by 1H NMR spectroscopy and were
characterized by integration of the ratio of the two organic products.
Unless noted, Rh–alkyl substrates were prepared by insertion of the
corresponding olefin into (TPP)Rh-H. [b] Conditions: [Rh]=8.1 mm;
[KOtBu]=16 mm; [[18]crown-6]=16 mm. [c] Conditions: [Rh]=8.1 mm;
[NMe4OH]=24 mm. [d] Substrate prepared according to reference [13].
[e] Substrate prepared by alkylation of (TPP)Rh with Br(CH2)3OH.
>95
1
[a] Reactions were monitored by H NMR spectroscopy and proceeded
quantitatively to give a mixture of the cyclized product and the
elimination product (olefin). Yields were determined by integration of
the 1H NMR spectrum of the crude reaction mixture. Rh–alkyl substrates
were prepared by insertion of the corresponding olefin into (TPP)Rh-H.
[b] Conditions: 258C; 1 h; [Rh]=8.1 mm; [NMe4OH]=24 mm. [c] Con-
ditions: 708C; 12 h; [Rh]=8.1 mm; [NMe4OH]=24 mm. [d] The elimi-
nation product was allylphenol.
bond-forming reductive elimination. THF was produced in
78% yield under optimized conditions ([D6]DMSO,
NMe4OH (3 equiv); Table 2, entry 3). The only other organic
product identified was 3-buten-1-ol, which is readily recycled
under the hydrofunctionalization conditions. The Rh–por-
phyrin complex was converted into [(TPP)RhI]À quantita-
tively. The cyclization of b-hydroxyalkyl–RhIII derivatives to
produce epoxides also proceeded in high yield (> 97%;
Table 2, entry 5).[13] However, attempts to prepare four- and
six-membered oxygen heterocycles by using this methodology
were unsuccessful. Instead the corresponding olefins were the
only organic products formed (Table 2, entries 6 and 7).[14]
Several pieces of evidence suggest that these unusual
À
À
À
applied to the formation of the C O, C N, and C C bonds of
a diverse range of cyclic compounds. Once again, the only
organic products observed were the desired heterocycle (or
carbocycle in the case of Table 3, entry 4) and the corre-
sponding olefin. The Rh complex was quantitatively con-
verted into [(TPP)RhI]À. For example, a-, b-, and g-substi-
tuted Rh–alkyl intermediates with pendant alcohol groups
underwent cyclization to produce the corresponding tetrahy-
drofurans within 1 h at 258C (Table 3, entries 1–3). A phenol-
substituted rhodium complex underwent facile cyclization at
room temperature to give 2,3-dihydro-2-methylbenzofuran
(Table 3, entry 5). The modest yield (53%) observed for this
system is probably due to the SN2 mechanism, which requires
nucleophilic attack at a sterically hindered tertiary carbon
À
reductive eliminations, which involve the formation of a C O
bond, proceed by an SN2 mechanism. First, the furan/olefin
product ratio increased dramatically upon changing the
solvent from C6D6 to [D6]DMSO (Table 2, entries 1 and 2).
Furthermore, modification of the nucleophile from a primary
to a secondary alkoxide (Table 2, entries 3 and 4) led to a
significant decrease in the ratio of the cyclic product to the 3-
buten-1-ol. Both results are indicative of competing SN2 and
E2 reactions, whereby the SN2 transition state is favored in
more polar solvents[15] and with less sterically hindered
nucleophiles. Han has also shown that the cyclization of
(TPP)Rh(CDH)CH(CH)3OH (a deuterated analogue of the
substrate in Table 2, entry 5) produces [D1]propylene oxide
with 100% inversion of stereochemistry at the a carbon
atom.[13] This experiment clearly implicates an SN2 mechanism
that proceeds by backside attack at the carbon atom bonded
to Rh (as opposed to precoordination of the alkoxide to the
Rh center).[7] Importantly, the proposed mechanism is con-
sistent with that of a closely related RhIII/RhI alkyl-exchange
reaction, which proceeds through an SN2 pathway.[12b]
À
center. Finally, the C C bonds of cyclopentane (Table 3,
entry 4) and C N bonds of pyrrolidine derivatives (Table 3,
entries 6–8) were also constructed readily by using this
methodology. Although harsher conditions (708C, 12 h)
À
À
were required relative to the analogous C O bond-forming
reactions, the yields and SN2/E2 product ratios remained
excellent.
The final step of the catalytic cycle involves protonation of
[(TPP)RhI]À to regenerate the (TPP)Rh-H catalyst
(Scheme 1, step c). Nelson and Dimagno recently reported
that the pKa value of (TPP)Rh-H is approximately 11;[12a]
therefore, as expected, the addition of excess trifluoroacetic
acid to the crude reaction mixtures resulted in quantitative
regeneration of (TPP)Rh-H.[16] This transformation com-
As summarized in Tables 2 and 3, C(sp3)–Nu reductive
elimination is extremely general in this system and could be
Angew. Chem. Int. Ed. 2004, 43, 588 –590
www.angewandte.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
589

Communications
[9] The mechanism of this reaction (involving [L]RhII· intermedi-
ates) has been studied, but its scope and functional-group
tolerance have not been reported. a) X. F. Fu, L. Basickes, B. B.
Wayland, Chem. Commun. 2003, 520 – 521; b) K. W. Mak, F.
Xue, T. C. W. Mak, K. S. Chan, J. Chem. Soc. Dalton Trans. 1999,
3333 – 3334; c) D. J. Anderson, R. Eisenberg, Organometallics
1996, 15, 1697 – 1706; d) R. S. Paonessa, N. C. Thomas, J.
Halpern, J. Am. Chem. Soc. 1985, 107, 4333 – 4334;
pletes the proposed catalytic cycle and makes it possible to
recycle the expensive rhodium porphyrin.
In summary, we have developed and implemented a new,
rational mechanistic approach to the design of catalysts for
the anti-Markovnikov hydrofunctionalization of olefins. We
have shown that (TPP)Rh-H mediates each step of the
proposed catalytic cycle with high selectivity, and have
demonstrated a new and remarkably general carbon–hetero-
atom bond-forming reductive elimination reaction. The
reactions described herein do not yet constitute a working
catalytic cycle. Preliminary attempts to carry out these
transformations under catalytic conditions (e.g., with several
phenol-based substrates and a variety of weak bases) have
thus far been hampered by the incompatibility of step a with
the polar solvents required for steps b and c. However, the
reactions as performed do allow for the facile recycling of the
valuable porphyrin–Rh-H complex. Additionally, we antici-
pate that the design principles presented herein will serve as a
valuable foundation for the development of a new generation
of regioselective olefin-hydrofunctionalization catalysts. Cur-
rent efforts in our laboratories are aimed at improving these
systems in terms of catalytic turnover through modification of
the solvent system, as well as by steric or electronic
perturbation of the porphyrin ligands. Efforts to circumvent
reaction-medium limitations through the use of solid-sup-
ported Rh catalyst systems are also in progress.
[10] Products were identified by 1H NMR spectroscopy based on 1H–
Rh coupling and on upfield shifts of the alkyl resonances.[9] The
linear product was formed in situ in > 90% yield, whereas the
branched isomer was not detected; thus, the lower limit of
regioselectivity corresponds to ~ 97%.
[11] a) K. J. Delrossi, X. X. Xhang, B. B. Wayland, J. Organomet.
Chem. 1995, 504, 47 – 56; b) B. B. Wayland, S. J. Ba, A. E. Sherry
J. Am. Chem. Soc. 1991, 113, 5305 – 5311.
[12] a) A. P. Nelson, S. G. Dimagno, J. Am. Chem. Soc. 2000, 122,
8569 – 8570; b) J. P. Collman, J. I. Brauman, A. M. Madonik,
Organometallics 1986, 5, 215 – 218.
[13] H.-Z. Han, Ph.D. Thesis, Princeton University, 1992.
[14] The origin of this effect is currently under investigation.
[15] K. A. Cooper, M. L. Dhar, E. D. Hughes, C. K. Ingold, B. J.
MacNulty, L. I. Woolf, J. Chem. Soc. 1948, 2043 – 2049.
[16] The addition of an excess of trifluoroacetic acid (~ 5 equiv
relative to the base) to the crude cyclization reaction mixtures in
C6D6 (e.g., Table 2, entry 1) or [D7]DMF resulted in quantitative
regeneration of (TPP)Rh-H. The analogous reaction in
[D6]DMSO produced a complex mixture of (TPP)Rh products
owing to the apparent instability of (TPP)Rh-H in this medium.
Received: May 21, 2003
Revised: October 24, 2003 [Z51941]
Keywords: alkenes · hydroamination · porphyrinoids ·
.
reductive elimination · rhodium
[1] B. M. Trost, Science 1991, 254, 1471 – 1477.
[2] J. Haggin, Chem. Eng. News 1993, 71(22), 23.
[3] a) C. M. Jensen, W. C. Trogler, Science 1986, 233, 1069 – 1071;
b) D. Ramprasad, H. J. Yue, J. A. Marsella, Inorg. Chem. 1988,
27, 3151 – 3155.
[4] a) M. Utsunomiya, R. Kuwano, M. Kawatsura, J. F. Hartwig, J.
Am. Chem. Soc. 2003, 125, 5608 – 5609; b) M. Beller, C. Breindl,
M. Eichberger, C. G. Hartung, J. Seayad, O. R. Thiel, A. Tillack,
H. Trauthwein, Synlett, 2002, 10, 1579 – 1594; c) M. Beller, O. R.
Thiel, H. Trauthwein, C. G. Hartung, Chem. Eur. J. 2000, 6,
2513 – 2522; d) M. Beller, H. Trauthwein, M. Eichberger, C.
Breindl, T. E. Muller, Eur. J. Inorg. Chem. 1999, 1121 – 1132;
e) M. Beller, H. Trauthwein, M. Eichberger, C. Breindl, J.
Herwig, T. E. Muller, O. R. Thiel, Chem. Eur. J. 1999, 5, 1306 –
1319.
[5] For an alternative approach, see: A. Seayad, M. Ahmed, H.
Klein, R. Jackstell, T. Gross, M. Beller, Science 2002, 297, 1676 –
1678.
[6] R. H. Crabtree, The Organometallic Chemistry of the Transition
Metals, 2nd ed., Wiley, New York, 2001, chap. 7.
[7] For recent examples of direct carbon–heteroatom bond-forming
reductive elimination under mild conditions, see: a) B. S. Wil-
liams, K. I. Goldberg, J. Am. Chem. Soc. 2001, 123, 2576 – 2587;
b) J. F. Hartwig, Acc. Chem. Res. 1998, 31, 852 – 860; c) S. S.
Stahl, J. A. Labinger, J. E. Bercaw, Angew. Chem. 1998, 110,
2298 – 2311; Angew. Chem. Int. Ed. 1998, 37, 2181 – 2192.
[8] Reductive elimination induced by oxidation is more common;
for a recent example, see: a) B. L. Lin, C. R. Clough, G. L.
Hillhouse, J. Am. Chem. Soc. 2002, 124, 2890 – 2891.
590
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2004, 43, 588 –590