A multicomponent redox system accounts for the first Nozaki-Hiyama-Kishi reactions catalytic in chromium

Full text of DOI:10.1021/ja9537538

J. Am. Chem. Soc. 1996, 118, 2533-2534
2533
Scheme 1
A Multicomponent Redox System Accounts for the
First Nozaki-Hiyama-Kishi Reactions Catalytic in
Chromium
Alois Fu¨rstner* and Nongyuan Shi
Max-Planck-Institut fu¨r Kohlenforschung
Kaiser-Wilhelm-Platz 1, D-45470 Mu¨lheim/Ruhr, Germany
Scheme 2
ReceiVed NoVember 7, 1995
The addition of organochromium compounds to aldehydes
originally described by Nozaki and Hiyama et al. has evolved
into a very powerful method for C-C bond formation (Scheme
1).1,2 This transformation is highly chemoselective and displays
an exceptional compatibility with an array of functional groups
in both reaction partners. The nucleophiles are readily available
by oxidative insertion of Cr(2+) into a wide range of (func-
tionalized) substrates such as allyl, propargyl, aryl, or alkenyl
halides, alkenyl triflates, allyl phosphates etc.3 Kishi’s finding
that nickel salts exhibit a catalytic effect on the formation of
the C-Cr bond has greatly improved the reliability of this
process,4 which has ever since been frequently used as the key
step in most impressive total syntheses of natural products of
utmost complexity.5,6
Two major drawbacks, however, are pending. First, stereo-
chemical control over the newly formed chiral center in
Nozaki-Hiyama-Kishi reactions is still in its infancy.7 Second,
as Cr(2+) is a one-electron donor, 2 mol of this reducing agent
per mol of halide are required to form the organochromium
intermediate; in practice, a huge excess (usually 200-1600 mol
%) is mandatory. The toxicity of chromium (and nickel) salts,
however, obviously makes this versatile process inadequate for
applications to pharmaceutical chemistry as well as for any
large-scale synthesis. Herein we address this latter shortcoming
and report the first examples of Nozaki-Hiyama-Kishi reac-
tions catalytic in chromium.
Our approach relies on the following concept: CrCl2 can be
prepared by reduction of CrCl3 with different reducing agents.1,8
The low-valent salt inserts into the C-X bond of the substrate
with formation of 1 mol of CrX3 and the organochromium
nucleophile A, which then reacts with the aldehyde giving rise
to a chromium alkoxide B. The high stability of the O-Cr-
(3+) bond formed drives the conversion but makes catalysis a
difficult task. Nevertheless B might react with a chlorosilane
in view of the pronounced oxophilicity of silicon. Since this
would lead to the O-silylated product C and regenerate the
second mole of CrX3, a catalytic cycle might emerge (Scheme
2).
The stoichiometric reducing agent chosen must efficiently
reduce CrX3 but has to be inert toward the reducible substrates;
moreover, its salts should be significantly less toxic than those
of chromium. Our endeavors in low-valent titanium chemistry
suggested the use of the combination of Zn dust and TMSCl
for such a purpose.9 Although some turnovers could be reached,
this particular combination turned out to be impractical. Eno-
lizable aldehydes gave poor results since they were partly
transformed into the corresponding silyl enol ethers (Table 1,
entry 3) due to the Lewis acidity of the zinc halides accumulat-
ing during the course of the reaction.
(1) (a) Okude, Y.; Hirano, S.; Hiyama, T.; Nozaki, H. J. Am. Chem.
Soc. 1977, 99, 3179-3181. (b) Takai, K.; Kimura, K.; Kuroda, T.; Hiyama,
T.; Nozaki, H. Tetrahedron Lett. 1983, 5281-5284. (c) Hiyama, T.; Kimura,
K.; Nozaki, H. Tetrahedron Lett. 1981, 1037-1040. (d) Takai, K.; Kuroda,
T.; Nakatsukasa, S.; Oshima, K.; Nozaki, H. Tetrahedron Lett. 1985, 5585-
5588.
(2) Reviews: (a) Saccomano, N. A. In ComprehesiVe Organic Synthesis;
Trost, B. M., Fleming, I., Eds.; Pergamon: Oxford, 1991; Vol. 1, pp 173-
209. (b) Cintas, P. Synthesis 1992, 248-257. (c) Hodgson, D. M. J.
Organomet. Chem. 1994, 476, 1-5.
(3) For the use of substrates other than halides, see the following. (a)
Alkenyltriflates: Takai, K.; Tagashira, M.; Kuroda, T.; Oshima, K.; Utimoto,
K.; Nozaki, H. J. Am. Chem. Soc. 1986, 108, 6048-6050. (b) Allylic
phosphates: Nowotny, S.; Tucker, C. E.; Jubert, C.; Knochel, P. J. Org.
Chem. 1995, 60, 2762-2772.
(4) (a) Jin, H.; Uenishi, J.; Christ, W. J.; Kishi, Y. J. Am. Chem. Soc.
1986, 108, 5644-5646. (b) Kishi, Y. Pure Appl. Chem. 1992, 64, 343-
350. This nickel effect has been independently reported by Nozaki et al.,
cf. ref 3a.
A likely substitute for zinc is manganese: it is very cheap
and exhibits an appropriate reduction potential,10 and its salts
are nontoxic and rather weak Lewis acids. Moreover com-
mercial manganese powder does not insert into C-X bonds
except for very reactive allylic halides.11 This was confirmed
by a series of control experiments with iodobenzene and octanal
as the substrates (Table 1); with either Mn/chlorosilane or Mn/
chlorosilane/NiCl2 cat. no product was formed (GC yield <3%).
In contrast, a combination of catalytic amounts of CrCl2 (≈15
mol %, doped with NiCl2), Mn powder, and a chlorosilane
cleanly converted these substrates into a mixture of 1a and its
silyl ether 1b.12 A workup with aqueous Bu4NF gave alcohol
(5) Stereoselective reactions of crotylchromium species have received
particular attention, cf.: (a) Mulzer, J.; Kattner, L.; Strecker, A. R.; Schro¨der,
C.; Buschmann, J.; Lehmann, C.; Luger, P. J. Am. Chem. Soc. 1991, 113,
4218-4229. (b) Jubert, C.; Nowotny, S.; Kornemann, D.; Antes, I.; Tucker,
C. E.; Knochel, P. J. Org. Chem. 1992, 57, 6384-6386. (c) Takai, K.;
Nitta, K.; Utimoto, K. Tetrahedron Lett. 1988, 5263-5266. (d) Buse, C.
T.; Heathcock, C. H. Tetrahedron Lett. 1978, 1685-1688. (e) Lewis, M.
D.; Kishi, Y. Tetrahedron Lett. 1982, 2343-2346.
(6) See: (a) Nicolaou, K. C.; Theodorakis, E. A.; Rutjes, F. P. J. T.;
Tiebes, J.; Sato, M.; Untersteller, E.; Xiao, X. Y. J. Am. Chem. Soc. 1995,
117, 1171-1172. (b) Kishi, Y. Chemtracts: Org. Chem. 1988, 1, 253-
265. (c) Aicher, T. D.; Buszek, K. R.; Fang, F. G.; Forsyth, C. J.; Jung, S.
H.; Kishi, Y.; Matelich, M. C.; Scola, P. M.; Spero, D. M.; Yoon, S. K. J.
Am. Chem. Soc. 1992, 114, 3162-3164. (d) Rowley, M.; Tsukamoto, M.;
Kishi, Y. J. Am. Chem. Soc. 1989, 111, 2735-2737. (e) Schreiber, S. L.;
Meyers, H. V. J. Am. Chem. Soc. 1988, 110, 5198-5200. (f) Chen, S. H.;
Horvath, R. F.; Joglar, J.; Fisher, M. J.; Danishefsky, S. J. J. Org. Chem.
1991, 56, 5834-5845. (g) White, J. D.; Jensen, M. S. J. Am. Chem. Soc.
1995, 117, 6224-6233. (h) Lu, Y. F.; Hartwig, C. W.; Fallis, A. G. J. Org.
Chem. 1993, 58, 4202-4204. (i) Still, C. W.; Mobilio, D. J. Org. Chem.
1983, 48, 4785-4786. (j) Eckhardt, M.; Bru¨ckner, R.; Suffert, J. Tetrahedron
Lett. 1995, 5167-5170. (k) Dancy, I.; Skrydstrup, T.; Cre´visy, C.; Beau,
J.-M. J. Chem. Soc., Chem. Commun. 1995, 799-800. (l) Brandstetter, T.;
Maier, M. E. Tetrahedron 1994, 50, 1435-1448. (m) Paquette, L. A.; Astles,
P. C. J. Org. Chem. 1993, 58, 165-169. (n) MacMillan, D. W. C.; Overman,
L. E. J. Am. Chem. Soc. 1995, 117, 10391-10392.
(8) Wuts, P. G. M.; Callen, G. R. Synth. Commun. 1986, 16, 1833-
1837.
(9) The combination of TiCl3 cat., Zn, and R3SiCl accounts for the first
carbonyl coupling reactions catalytic in titanium, cf.: Fu¨rstner, A.; Hupperts,
A. J. Am. Chem. Soc. 1995, 117, 4468-4475.
(10) Electrochemical redox potentials: Cr3+ + e- h Cr2+ (-0.41 V);
Zn2+ + 2e- h Zn (-0.76 V); Mn2+ + 2e- h Mn (-1.03 V). Handbook
of Chemistry and Physics, 62nd ed.; CRC Press: Boca Raton, FL, 1982.
(11) (a) Cahiez, G.; Chavant, P. Y. Tetrahedron Lett. 1989, 7373-7376.
(b) Hiyama, T.; Sawahata, M.; Obayashi, M. Chem. Lett. 1983, 1237-
1238.
(7) Chen, C.; Tagami, K.; Kishi, Y. J. Org. Chem. 1995, 60, 5386-
5387.
(12) Complete silylation could not be reached. This seems to be a limiting
factor for the number of turnovers.
0002-7863/96/1518-2533$12.00/0 © 1996 American Chemical Society

2534 J. Am. Chem. Soc., Vol. 118, No. 10, 1996
Communications to the Editor
Table 1. Control Experiments for Cr(2+)-Mediated Reaction of
Iodobenzene with Octanal
Table 2. Nozaki-Hiyama-Kishi Reactions Catalytic in
Chromiuma
ClMe2Si(CH2)3CN Me3SiCl
yield (%)b yield (%)b
entry
RX
aldehyde
product
CrCl2
isolated
T (°C) yield (%)b
entry (mol %)a
additives
1
2
3
4
5
6
7
8
400
400
30
15
15
9
20
50
70
50
50
50
50
50
78c
65
Zn, TMSCl
Mn, TMSCl
Mn, ClMe2Si(CH2)3CN
Mn, ClMe2Si(CH2)3CN
Mn, ClMe2Si(CH2)3CN
30-40d
67
72
58
0
0
0e
Mn, ClMe2Si(CH2)3CN, NiCl2 (cat.)
0e
a Unless stated otherwise the reactions were carried out in DME/
DMF (20/3) using CrCl2 doped with NiCl2 (≈15%). b After desilylation
of the admixed 1b. c In pure DMF as solvent. d Formation of 1-(tri-
methylsilyloxy)-1-octene as side reaction, see text. e GC yield <3%.
1a in 67-72% yield, which favorably compares to the 65-
78% obtained under conventional conditions using 400 mol %
of CrCl2 as the reagent.
The experimental parameters of the catalytic process have
been carefully optmized. Reactions of aryl and alkenyl halides
(triflates) proceed properly with ≈15 mol % CrCl2, while allylic
halides require only ≈7 mol % of this catalyst. Except for
allylic substrates, the reactions were carried out at 50 °C ((3
deg) in DME/DMF (20/3)13 since the conversion was too slow
at lower temperatures. The addition of catalytic amounts of
LiI to TMSCl as well as replacing TMSCl by TMSBr,
(EtO)3SiCl, ClMe2SiCH2CH2SiMe2Cl, or ClMe2Si(CH2)3CN (cf.
Table 2) did not significantly improve the yields.
The catalytic procedure nicely applies to a set of representa-
tive Nozaki-Hiyama type reactions.14 As can be seen from
Table 2, iodobenzene reacts smoothly with aromatic as well as
with different aliphatic aldehydes. Entry 4 highlights the
chemoselectivity of this process; thus, Cr(2+) exclusively inserts
into the aryl iodide while the alkyl chloride group of 6-chlo-
rohexanal is compatible. Aryl halides containing either potential
donor sites or reducible functional groups such as 2-iodo-
thiophene or ethyl 4-iodobenzoate can also be used. Alkenyl
iodides and, in particular, alkenyltriflates3a turned out to react
smoothly to the corresponding allylic alcohols. However, we
noticed some influence of the electronic properties of the chosen
aldehyde on the reaction path; aliphatic aldehydes, benzalde-
hyde, and analogues bearing electron-donating substituents
reacted properly, while methyl 4-formylbenzoate as the elec-
trophile gave the corresponding pinacol as the major product
(Scheme 3). In this case the ester group obviously facilitates a
single-electron transfer from Cr(2+) to the aldehyde function,
with formation and subsequent dimerization of the ketyl radical
anion. This observation is in good accordance with the
experience gained with other pinacolization methods.15 Allylic
bromides were found to react particularly well in THF at
ambient temperature rather than at 50 °C, requiring only ≈7
mol % CrCl2 which must not be doped with NiCl2 in order to
prevent Wurtz-coupling.16
a Unless stated otherwise, all reactions were carried out with CrCl2
(15 mol %, doped with cat. NiCl2), Mn powder (4.2 mmol), aldehyde
(2.5 mmol), RX (5 mmol), and chlorosilane (6 mmol) in DMF/DME
(20/3) at 50 °C. b Refers to the product obtained after desilylation
(aqueous Bu4NF) of the crude mixture. c Isolated after acetylation of
the crude product. d 4,4′-Bis(ethoxycarbonyl)diphenyl as byproduct
(20% based on I-C6H4-COOEt). e The reaction was performed with
undoped CrCl2 (7 mol %) at ambient temperature in THF. f The
R-methylene-γ-lactone was cyclized upon treatment of the crude product
with catalytic amounts of p-TsOH‚H2O.
Scheme 3
%), which is constantly recycled by the relatively nontoxic
manganese. The limiting factor for the number of turnovers
seems to be the incomplete σ-bond metathesis between the
chromium alkoxide initially formed and the chlorosilane addi-
tive.12 We are now intensively working to refine this crucial
step in order to further reduce the quantity of toxic Cr(2+) and
to further explore the full scope of the new procedure.
Acknowledgment. We thank the Fonds der Chemischen Industrie
for generous financial support.
Supporting Information Available: Details on the purification of
the reagents, a representative experimental procedure, and the listing
of the IR, 1H NMR, 13C NMR, and MS data of all products (5 pages).
This material is contained in many libraries on microfiche, immediately
follows this article in the microfilm version of the journal, can be
ordered from the ACS, and can be downloaded from the Internet; see
any current masthead page for ordering information and Internet access
instructions.
In summary, we have described the first Nozaki-Hiyama-
Kishi reactions using catalytic amounts of CrCl2 (7-15 mol
(13) Pure DMF, although superior in the stoichiometric reaction (Table
1, entry 1), gave inferior yields. The CrCl2 should be colorless; pale-green
batches result in lower yields.
(14) A representative procedure is described in the supporting informa-
tion.
(15) Fu¨rstner, A.; Csuk, R.; Rohrer, C.; Weidmann, H. J. Chem. Soc.,
Perkin Trans. 1 1988, 1729-1734 and literature cited therein.
(16) One might envisage that allylic halides react with Mn without any
assistance of Cr(2+) (cf. ref 11). Control experiments, however, showed
that under the conditions employed the combination Mn/TMSCl without
CrCl2 catalyst leads to very poor conversion (19% in GC after 90 h reaction
time!).
JA9537538