Intramolecular hydrosilylation and silicon-assisted cross-coupling: An efficient route to trisubstituted homoallylic alcohols

Article abstract of DOI:10.1021/ol006769y

(equation presented) Alkylidenesilacyclopentanes (formed by intramolecular hydrosilylation of homopropargyl alcohols) are efficiently coupled with aryl or alkenyl halides in the presence of tetrabutylammonium fluoride and a palladium(0) catalyst. Yields of cross-coupling were generally high, and the reaction is compatible with a wide range of functional groups. The overall transformation achieves the conversion of homopropargyl alcohols to trisubstituted homoallylic alcohols in a highly stereoselective fashion.

Full text of DOI:10.1021/ol006769y

ORGANIC
LETTERS
2
001
Vol. 3, No. 1
1-64
Intramolecular Hydrosilylation and
Silicon-Assisted Cross-Coupling:
An Efficient Route to Trisubstituted
Homoallylic Alcohols
6
Scott E. Denmark* and Weitao Pan
2
6
45 Roger Adams Laboratory, Department of Chemistry, UniVersity of Illinois,
00 South Mathews AVenue, Urbana, Illinois 61801
denmark@scs.uiuc.edu
Received October 24, 2000
ABSTRACT
Alkylidenesilacyclopentanes (formed by intramolecular hydrosilylation of homopropargyl alcohols) are efficiently coupled with aryl or alkenyl
halides in the presence of tetrabutylammonium fluoride and a palladium(0) catalyst. Yields of cross-coupling were generally high, and the
reaction is compatible with a wide range of functional groups. The overall transformation achieves the conversion of homopropargyl alcohols
to trisubstituted homoallylic alcohols in a highly stereoselective fashion.
The invention and development of efficient and stereospecific
carbon-carbon bond-forming reactions remains among the
central missions of synthetic organic chemistry. One of the
most powerful reactions to emerge in the past quarter century
is the transition metal-catalyzed cross-coupling between
donors in cross-coupling reactions, thus eliminating some
of the problems associated with the aforementioned meth-
ods: toxicity, handling of reagents, and lack of functional
group compatibility. Disclosures from these laboratories have
5a,b,d
5c
established that silacyclobutanes
and simple silanols or
5f
(
(
primarily) main-group organometallic species and organo-
pseudo)halides. Witness the impact of the Stille coupling
silyl hydrides are extremely reactive in cross-coupling with
aryl and alkenyl halides. Preliminary mechanistic investiga-
5
e
of organostannanes, the Suzuki coupling of organoboranes,
or the Negishi coupling of organozinc reagents and the case
is easily made for the stature of these transformations in the
armamentarium of organic synthesis.1
tions reveal that the basic structural requirement for facile
cross-coupling is the presence of an oxygen function on the
silicon (alcohol, ether, disiloxane). Accordingly, we recog-
nized the opportunities for installing a fresh carbon-silicon
bond via intramolecular hydrosilylation of a pendant silyl
We have recently set out to elevate the implementation
of organosilicon reagents in palladium-catalyzed cross-
6
ether. This process would provide an unambiguous route
2
coupling to a similar rank. The pioneering work of Hiyama,
to geometrically defined vinylic silanes necessarily bearing
3
4
Ito, and DeShong established that organosilicon compounds
with heteroatom(s) attached at the silicon can be used as the
(
3) (a) Tamao, K.; Kobayashi, K.; Ito, Y. Tetrahedron Lett. 1989, 30,
6
8
051. (b) Suginome, M.; Kinugasa, K.; Ito, Y. Tetrahedron Lett. 1994, 35,
635.
(
1) (a) Diederich, F.; Stang, P. J., Eds. Metal-Catalyzed Cross-Coupling
(4) (a) Mowery, M. E.; DeShong, P. J. Org. Chem. 1999, 64, 1684. (b)
Mowery, M. E.; DeShong, P. J. Org. Chem. 1999, 64, 3266.
(5) (a) Denmark, S. E.; Choi, J. Y. J. Am. Chem. Soc. 1999, 121, 5821.
(b) Denmark, S. E.; Wu, Z. Org. Lett. 1999, 1, 1495. (c) Denmark, S. E.;
Wehrli, D. Org. Lett. 2000, 2, 565. (d) Denmark, S. E.; Wang, Z. Synthesis
2000, 999. (e) Denmark, S. E.; Wehrli, D.; Choi, J. Y. Org. Lett. 2000, 2,
2491 (f) Denmark, S. E.; Neuville, L. Org. Lett. 2000, 2, 3221.
Reactions; Wiley-VCH: Weinheim, 1998. (b) Tsuji, J. Palladium Reagents
and Catalysts. InnoVations in Organic Synthesis; Wiley: Chichester, U.K.,
1
995.
(2) (a) Hatanaka, Y.; Hiyama, T. Synlett 1991, 845. (b) Hiyama, T. In
Metal-Catalyzed Cross-Coupling Reactions; Diederich, F., Stang, P. J., Eds.;
Wiley-VCH: Weinheim, 1998; Chapter 10.
1
0.1021/ol006769y CCC: $20.00 © 2001 American Chemical Society
Published on Web 12/14/2000

an oxygen function. We report herein the successful realiza-
tion of this strategy for the preparation of stereodefined
homoallylic alcohols.7
To test the feasibility of the overall transformation, we
chose to examine each stage independently and therefore
selected a substrate that would allow isolation and charac-
terization of both intermediates. The cyclic siloxane, 2a,
derived from 3-pentyn-1-ol, was chosen for ease of handling
common functional groups tested (ester, ketone, nitro,
alcohol, nitrile, ether). For all aryl iodides examined, the
reaction proved to be mild and high yielding except in the
case of 2-nitroiodobenzene (entry 4) which was very slow
and gave a substantial amount of nitrobenzene as a byprod-
uct. Noteworthy features of this process are that (1) electron-
withdrawing or -donating groups exhibit similar reactivity,
(2) ortho substituents on the aryl iodide do not affect the
reactivity significantly, (3) the reaction tolerates diverse
functional groups such as ester, nitro, cyano, ether, and even
free hydroxy group, and (4) the reactions of all halides were
stereospecific, with the exception of 4-nitroiodobenzene and
1-iodonaphthalene, which gave a small amount of the
geometrical isomer.
(
low molecular weight) and the diisopropylsilyl group was
5f
chosen to facilitate purification (stability of the silyl ethers).
Thus, silylation of 3-pentyn-1-ol with diisopropylchlorosilane
provided the silyl ether 1a in good yield, Scheme 1.
Intramolecular hydrosilylation of 1a was effected with a
catalytic amount of Speier’s catalyst to afford 2a cleanly and
in high yield.
Table 1. Palladium-Catalyzed Cross-Coupling of 2 with Aryl
Iodidesa
Scheme 1
b
entry
R
time, h/temperature, °C product yield, %
1
2
3
4
5
6
7
8
9
H
2-Me
6.66/rt
6.83/rt
10.0/35
23/35
6.0/rt
6.50/rt
6.0/rt
6.25/rt
46/45
16/45
3a
3b
3c
3d
3e
3f
3g
3h
3i
88
74
74
56
81
72
70
67c
70
86
2-MeO
2-NO2
3-HOCH2
4-CH3O
4-MeCO
4-NO2
Orienting experiments were carried out using the standard
conditions established for the silanol couplings with a simple
aryl iodide. Thus, siloxane, 2a, was dissolved in 2.0 equiv
of a 1.0 M solution of TBAF in THF, followed by the
2
addition of iodobenzene and 5 mol % of Pd(dba) . Gratify-
5
c
4-CN
4-COOEt
1
0
3j
ingly, the siloxane did undergo the coupling process,
however, at a significantly reduced reaction rate compared
to the related silanols. Moreover, the reaction mixture was
contaminated with a substantial amount of biphenyl (the
product of self-coupling of iodobenzene), and thus the yield
of the cross-coupling product was attenuated. The addition
of various ligands5b or decreasing the amount of Pd(0) did
not meaningfully improve the results. Fortunately, we found
that adding the iodide in portions satisfactorily suppressed
the formation of biphenyl and correspondingly improved the
yield of the desired coupling product. The portionwise
addition of the iodide proved to be effective in reducing the
amount of homocoupling byproduct in most cases. For a few,
very slow reacting substrates, even this expedient was not
helpful (vide infra).
a Reaction conditions: 1.1 equiv of 2a, 2.0 equiv of TBAF, and 5 mol
of Pd(dba)2 were employed for 1.0 equiv of iodide in THF at rt. The
%
iodide was added in portions as specified (see Supporting Information).
b
Yields of analytically pure materials. c Isomeric ratio 95.2/4.8 by capillary
GC analysis.
This variant of the coupling reaction is not limited to
benzene derivatives. For instance, 1-iodonaphthalene, 1-bromo-
4-tert-butyl-1-cyclohexene (an unactivated vinyl bromide),
and 3-iodopyridine reacted with 2a to give the expected
products 3k, 3l, and 3m, respectively, in reasonable to good
yield (Figure 1).
The optimization of this process next turned to the
investigation of the importance of the silicon substituents.
It was of interest to see whether and how the size of the
substituents on the silicon would affect the rate and selectivity
of the reaction. Additional benefits such as improved mass
efficiency and ease of byproduct removal could be realized
with a smaller group in place of an isopropyl group. We
thus focused our attention on the corresponding dimethyl-
siloxane. However, the synthesis of the siloxanes was
problematic. The intramolecular hydrosilylation of dimeth-
ylsilyl ethers with chloroplatinic acid resulted in polymeric
materials. Though the platinum(0)-1,3-divinyl-1,1,3,3-tet-
With a reproducible procedure in hand, we next explored
the scope of the reaction with regard to the nature and
position of substituents on the aromatic ring. The results
compiled in Table 1 reveal good compatibility with all
(6) For an excellent review of hydrosilylation, see: Ojima, I.; Li, Z.;
Zhu, J. In The chemistry of organic silicon compounds; Rappoport, Z.,
Apeloig, Y., Eds.; John Wiley & Sons: Great Britain, 1998; Vol. 2; pp
1
687-1792.
(
7) For a single previous example of this concept see ref 3a.
62
Org. Lett., Vol. 3, No. 1, 2001

ramethyldisiloxane complex (Pt(DVDS)) gave clean in-
tramolecular hydrosilylation, any attempts to obtain the
siloxane in a pure state or to scale-up the preparation of the
siloxane led to oligomerization.8
in Table 2. Thus, 2-iodoanisole and 4-iodobenzonitrile also
coupled smoothly with 2b (formed in situ from 3-pentyn-
1-ol) using the one-pot procedure (10 mol % of Pd(dba)
2
),
affording good yields of cross-coupling products. In these
cases as well, 5% Pd(dba)
efficiently.
2
failed to promote the reactions
Table 2. One-Pot Arylation of Homopropargylic Alcohols
c
Pd, time, min/
yield, %
entry
1a
R
R′
%
emp, °C product (isomeric ratio)
Me
H
10
40/rt
3a
85
(
97.2/2.8)
75
2a
3a
Me
Me
2-OMe 10
480/35
300/rt
3c
3i
(98.3/1.7)
4-CN
H
10
10
5
74
(
(
(
96.7/3.3)
4b
5b
n-Bu
n-Bu
40/rt
90/rt
4
4
84
97.7/2.3)
85
98.3/1.7)
H
Figure 1. Cross-coupling of 2a with nonbenzenoid halides.
a
Reaction conditions: (1) homopropargylic alcohols (1.3 equiv), TMDS
1.0-1.2 equiv), (2) 0.30-0.37% Pt(DVDS) in THF (1.5 mL/mmol alcohol).
3) TBAF (2.2 equiv), aryl iodide (1.0 equiv). b Reaction conditions: (1)
(
(
homopropargylic alcohol (1.3 equiv), TMDS (1.0 equiv); (2) remove TMDS
in vacuo, (3) 0.30-0.37% Pt(DVDS) in THF (1.5 mL/mmol alcohol), (4)
TBAF (2.2 equiv), aryl iodide (1.0 equiv). c Yields of analytically pure
materials.
A practical solution to this problem was conceived in the
form of a one-pot protocol that would (1) obviate the need
to isolate and purify the delicate silyl ether and siloxane and
(2) improve the overall efficiency of the process. Thus, to
streamline the procedure and minimize the formation of
potentially deleterious byproducts, we employed tetrameth-
yldisilazane (TMDS) as the silylating agent. Treatment of
The capriciousness of the one-pot process is believed to
be rooted in the residual tetramethyldisilazane (TMDS) which
may poison the Pd(dba) . We were unable to remove the
2
excesses of this reagent owing to the volatility of the silyl
ether 1b. To test this hypothesis, we made recourse to a
higher molecular weight homopropargylic alcohol, 3-octyn-
3
-pentyn-1-ol (1b) with 0.77 equiv of TMDS (neat) followed
by dissolution in THF and closure under the action of Pt-
DVDS) produced the siloxane 2b efficiently by GC/MS
(
analysis (Scheme 2). Unfortunately, the subsequent cross-
coupling reactions were not reproducible and often stalled.
Varying the amount of fluoride, reaction concentration, and
palladium source resulted in no significant improvement.
Fortunately, it was found that by increasing the loading of
1
-ol. When this alcohol was employed, the excess TMDS
was easily removed by evacuation upon the completion of
silylation. Intramolecular hydrosilylation with Pt(DVDS)
proceeded normally, but now, the cross-coupling reaction
could be effectively promoted with only 5 mol % of
Pd(dba)
2
to 10%, the complete consumption of iodobenzene
Pd(dba)
h. If 10 mol % of Pd(dba)
complete within 40 min, affording the desired product in
2
to give the expected products in 85% yield within
was achieved within 40 min, giving the desired product in
2
2
was used, the reaction was
85% yield. The results for this and other iodides are collected
9
8
5% yield.
Comparison of the results in Tables 1 and 2 clearly shows
Scheme 2
the advantages of the one-pot procedure: (1) ease of
experimental protocol, (2) aryl iodide need not be added
portionwise to avoid homocoupling, (3) intermediate silyl
ethers need not be isolated (one step versus three), and (4)
superior overall yields. However, in one important category
(
8) We ascribe this to the strain of the silacyclopentane ring and the low
energy barrier to silyl ether exchange.
9) Without removing the TMDS, the reaction gave a 63% yield of the
(
product (72% conversion) over 44 h on a small scale, with 10 mol % catalyst
loading.
Org. Lett., Vol. 3, No. 1, 2001
63

the multistage procedure employing 2a was superior, namely,
stereospecificity. In all cases involving the dimethylsiloxane,
a small amount of the Z-isomer was observed (1.7-2.3%).
All attempts to suppress the formation of this isomer were
unsuccessful. For instance, this minor product was not
eliminated by decreasing the reaction concentration for
hydrosilylation. Changing the hydrosilylation catalyst also
failed to remedy the situation.10
the geometry of the vinylsilane, and (3) as the locus for
directing the formation of a new carbon-carbon bond. In a
very mild and environmentally friendly manner, homopro-
pargyl alcohols were elaborated stereospecifically to trisub-
stituted homoallylic alcohols in good yields. These structures
are often encountered in natural products and also are useful
1
1
synthetic intermediates. It has not been lost on the authors
that the siloxane itself is a very versatile subunit for further
1
2
synthetic manipulation.
The effect of the silicon substituent on the rate and
selectivity of the cross-coupling had been addressed previ-
The ability to introduce silafunctional units in a controlled
and stereoselective fashion by, inter alia, intramolecular
hydrosilylation, silylformylation, and disilylation bodes well
for the manifold applications of this methodology. These
studies will be disclosed in due course.
5c
ously with silanols. In that study the size of the substituent
methyl versus isopropyl) had little effect on the rate of
(
coupling. In the case at hand, the dimethylsiloxanes reacted
faster than the corresponding diisopropylsiloxanes. Interest-
ingly, this fact may also explain why it was not necessary
to add the aryl iodide portionwise, given the competitive rate
of cross-coupling compared to homocoupling of the aryl
Acknowledgment. We are grateful to the National
Science Foundation (NSF CHE 9800387) for generous
financial support.
5
c
iodides. However, as was observed with silanols, the
stereospecificity was dependent on substituent. The dimethyl
derivatives tended to give a small amount of the isomeric
products, whereas the diisopropyl analogues gave the cou-
pling products with near perfect stereochemical fidelity.
Supporting Information Available: Procedures for the
preparation and full characterization of 1a, 2a, 3a-m, and
4
along with representative procedures for coupling reactions.
This material is available free of charge via the Internet at
http://pubs.acs.org.
In conclusion, we have demonstrated the expanded syn-
thetic potential of our newly invented silicon-based cross-
coupling. In this incarnation, the silicon atom has served in
the following capacities: (1) as a temporary protecting group
for the hydroxyl function, (2) as a temporary tether to fix
OL006769Y
(11) See the following and references therein: (a)Takeda, T.; Kabasawa,
Y.; Fujiwara, T. Tetrahedron 1995, 51, 2515. (b) Okuma, K.; Tanaka,
Y.-i; Hirabayashi, S.-i; Shioji, K.; Matsuyama, H. Hereocycles 1997, 45,
1
385. (c) Crousse, B.; Alami, M.; Linstrumelle, G. Synlett 1997, 922.
(12) (a) Marshall, J. A.; Yanik, M. M. Org. Lett. 2000, 2, 2173. (b)
(10) It is conceivable that a competitive intermolecular hydrosilylation
competes to form the undesired isomer.
Tamao, K.; Maeda K.; Tanaka, T.; Ito, Y. Tetrahedron Lett. 1988, 29, 6955.
64
Org. Lett., Vol. 3, No. 1, 2001