Preference of 4-exo ring closure in copper-catalyzed intramolecular coupling of vinyl bromides with alcohols

Article abstract of DOI:10.1021/ja072793w

The copper-catalyzed intramolecular O-vinylation of γ-bromohomoallylic alcohols was investigated. With 10 mol % of CuI as the catalyst and 20 mol % of 1,10-phenanthroline as the ligand, the reactions of 3-bromo-3-buten-1-ols in refluxing CH₃CN

Full text of DOI:10.1021/ja072793w

Published on Web 06/09/2007
Preference of 4-exo Ring Closure in Copper-Catalyzed Intramolecular
Coupling of Vinyl Bromides with Alcohols
Yewen Fang and Chaozhong Li*
Joint Laboratory of Green Synthetic Chemistry, Shanghai Institute of Organic Chemistry, Chinese Academy of
Sciences, 354 Fenglin Road, Shanghai 200032, China
Received April 21, 2007; E-mail: clig@mail.sioc.ac.cn
Table 1. Optimization of Reaction Conditions in the Synthesis of
The copper-catalyzed formation of C-X (X ) N, O, S, etc.)
2a
bonds has been successfully developed during the past few years.1
With the intramolecular Ullmann coupling (IUC) as the strategy,
the preparation of many medium- and even large-sized heterocycles
can be achieved.2 However, reports on the synthesis of strained
heterocycles such as four-membered rings are rare with this IUC
strategy.3 In fact, this problem seems universal in transition-metal-
catalyzed intramolecular coupling processes. We report here that
the uncommon 4-exo ring closure in the copper-catalyzed intramo-
lecular O-vinylation of γ-bromohomoallylic alcohols is not only
an efficient process leading to the convenient synthesis of 2-me-
thyleneoxetanes but also fundamentally preferred over other modes
(5-exo, 6-exo, and 6-endo) of cyclization. Moreover, this unique
selectivity is different from that of the corresponding palladium-
catalyzed processes.
entrya
ligandb
base
yield of 2a (%)c
yield of 3 (%)c
1
2
3
4
5
6
7
8
9
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
K2CO3
0
0
85
∼60
32
0
A-C
D
E
F
F
F
F
F
F
14
71
98
54
58
0
0
0
0
0
K3PO4
DABCO
NaOtBu
0
0
98d
0
10
2-Methyleneoxetanes are biologically interesting and promising
synthetic intermediates due to their unique combination of func-
tionalities.4 Unfortunately, there is a scarcity of methods for their
preparation.5 Owing to our interest in copper-catalyzed intramo-
lecular vinylation,3a,6 we envisioned that these compounds might
be synthesized via Cu(I)-catalyzed O-vinylation of γ-bromohomoal-
lylic alcohols. Thus, we explored this possibility with the use of
1a as the model substrate, and the reactions were carried out in
refluxing acetonitrile with the catalysis of CuI. The results are
summarized in Table 1. Our initial screening on the ligands showed
that 1,10-phenanthrolines (E and F) afforded the expected product
2a in high yields, while other typical ligands (A-D) gave mainly
alkyne 3 via direct elimination of HBr (entries 1-5, Table 1). 1,-
10-Phenanthroline F was particularly effective, and the product 2a
was achieved in almost quantitative yield under mild conditions
(entry 5, Table 1). We next examined the effect of base on the
coupling reaction. Cs2CO3 turned out to be better than K3PO4 or
K2CO3, while the organic base DABCO was ineffective (entries
5-8, Table 1). When NaOtBu was employed, 3 was obtained in
almost quantitative yield within 10 min, which gradually isomerized
to allene 4 (entry 9, Table 1). This observation also supported the
oxidative addition mechanism of O-vinylation by excluding the
elimination-addition mechanism (via the intermediacy of 3 or 4).
With the optimized conditions in hand (10 mol % of CuI, 20
mol % of E, 200 mol % of Cs2CO3 in refluxing CH3CN), we then
examined the generality of this methodology. The results are listed
in Table 2. The primary, secondary, and even tertiary alcohols all
served as good substrates for the 4-exo ring closure (2b-e).
Substrates with various substitutions afforded the expected products
in good to excellent yields. A number of functional groups, such
as -OH and -OBz, were well-tolerated (2h-n). Moreover, the
configuration of the CdC double bond was nicely retained as
evidenced by the reactions of 1f and 1g. As a comparison, the chloro
analogue of 1a remained inert under the optimized conditions. The
results also indicated that the reactivity of alcohols decreases in
a Reaction conditions: 1a (0.3 mmol), CuI (0.03 mmol), ligand (0.06
mmol), base (0.6 mmol), CH3CN (3 mL), reflux, 8 h. b A: N,N′-
Dimethylethylenediamine; B: 2-aminopyridine; C: L-proline; D: 1,2-trans-
bis(pyridin-2-ylmethylene)cyclohexane-1,2-diamine; E: 3,4,7,8-tetramethyl-
1,10-phenanthroline; F: 1,10-phenanthroline. c Isolated yield based on 1a.
d The reaction time was 10 min.
the order of aliphatic > allylic > benzylic (2h-j), in accordance
with the literature observations,8a while 1, 2, and 3° alcohols are
of similar rates.8b The substitution on the CdC bond (1f and 1g)
discourages the cyclization, and the reactions had to be conducted
at higher temperatures (dioxane, reflux).
The above methodology was then successfully extended to other
modes of cyclization. The IUC in 5-exo (2o-q), 6-exo (2r), and
even 6-endo (2s) modes proceeded smoothly under the optimized
conditions (Table 2).
The ease of the 4-exo ring closure shown in Tables 1 and 2
prompted us to explore the competition among different modes of
cyclization. Thus, the reactions of substrates 5a-g, each having
two possible IUC pathways, were performed (Table 3). We were
surprised to find that, in all of the cases, oxetanes 6a-g via 4-exo
cyclization were the only detectable products whose structure was
unambiguously established by 2D NMR experiments. These results
clearly revealed that the 4-exo ring closure is fundamentally
preferred over the 5-exo (entries 1-5, Table 3), 6-exo (entry 6,
Table 3), and 6-endo (entry 7, Table 3) modes. To the best of our
knowledge, these are the first examples in transition-metal-catalyzed
chemistry describing the predominance of 4-exo cyclization.
To further understand the unique selectivity in the above copper-
catalyzed processes, we tried the possible Pd(0) catalysis in
O-vinylation (Scheme 1, see also Supporting Information for
details). Our efforts showed that, with Pd(OAc)2 (5 mol %)/BINAP
(7.5 mol %) as the catalyst and Cs2CO3 (2 equiv) as the base,
compound 1q underwent highly efficient cross-coupling (5-exo) in
refluxing THF to give the product 2q in 94% yield. However, under
the same conditions, the primary alcohol 1o gave only the â-hydride
9
8092
J. AM. CHEM. SOC. 2007, 129, 8092-8093
10.1021/ja072793w CCC: $37.00 © 2007 American Chemical Society

C O M M U N I C A T I O N S
Table 2. Intramolecular Coupling of Vinyl Bromides with Alcohols
Scheme 1. Pd(OAc)2/BINAP-Catalyzed Intramolecular Coupling
the 4-exo cyclization product 6b, along with the formation of diene
9 via intramolecular Heck reaction.9
The above difference between Pd(0) and Cu(I) in O-vinylation
is truly remarkable. While the Pd-catalyzed method is sensitive to
substrate structures and suffers from the competing â-hydride
elimination,10 the Cu-catalyzed processes enjoy their generality to
include a wide range of substrates. More importantly, the selectivity
between 4-exo and 5-exo cyclization is completely reversed by
switching the catalyst from Cu to Pd.
Although the reason for the preference of 4-exo cyclization in
Cu(I)-catalyzed O-vinylation remains unclear, it might be possible
that the Cu(I) coordinates to the alkoxide prior to its oxidative
insertion into the C-Br bond. Furthermore, the transition state for
4-exo cyclization as a Cu-containing five-membered ring structure
could be sterically more easily accessible and thermodynamically
more stable than that of the corresponding 5-exo, 6-exo, or 6-endo
cyclization. Presumably, the analogous Pd-containing five-mem-
bered ring might not be sterically favorable as Pd(0) is much larger
in size than Cu(I). Further studies to elucidate the mechanism of
this unusual preference and to understand the difference between
Cu(I) and Pd(0) catalysis in O-vinylation are currently underway
in our laboratory.
a Isolated yield based on 1. b The reaction was conducted in refluxing
dioxane. c The reaction was conducted in refluxing toluene.
Acknowledgment. This project was supported by the National
Table 3. Competition Experiments in the O-Vinylation of Alcohols
NSF of China (Grant Nos. 20325207 and 20472109).
Supporting Information Available: Typical procedures for the
IUC of γ-bromohomoallylic alcohols and for the Pd(0)-catalyzed
reactions, the characterizations of 1-9. This material is available free
of charge via the Internet at http://pubs.acs.org.
References
(1) For reviews, see: (a) Kunz, K.; Scholz, U.; Ganzer, D. Synlett 2003, 2428.
(b) Ley, S. V.; Thomas, A. W. Angew. Chem., Int. Ed. 2003, 42, 5400.
(c) Beletskaya, I. P.; Cheprakov, A. V. Coord. Chem. ReV. 2004, 248,
2337. (d) Dehli, J. R.; Legros, J.; Bolm, C. Chem. Commun. 2005, 973.
(2) For the latest selected examples, see: (a) Toumi, M.; Couty, F.; Evano,
G. Angew. Chem., Int. Ed. 2007, 46, 572. (b) Okano, K.; Tokuyama, H.;
Fukuyama, T. J. Am. Chem. Soc. 2006, 128, 7136. (c) Cai, Q.; Zou, B.;
Ma, D. Angew. Chem., Int. Ed. 2006, 45, 1276.
(3) (a) Lu, H.; Li, C. Org. Lett. 2006, 8, 5365. (b) Joyeau, R.; Kobaiter, R.;
Sadet, J.; Wakselman, M. Tetrahedron Lett. 1989, 30, 337.
(4) (a) Dollinger, L. M.; Howell, A. R. Bioorg. Med. Chem. Lett. 1998, 8,
977. (b) Dollinger, L. M.; Howell, A. R. J. Org. Chem. 1998, 63, 6782.
(5) (a) Howell, A. R.; Truong, A. Tetrahedron Lett. 1996, 37, 8651. (b)
Dollinger, L. M.; Howell, A. R. J. Org. Chem. 1996, 61, 7248.
(6) (a) Fang, Y.; Li, C. Chem. Commun. 2005, 3574. (b) Hu, T.; Li, C. Org.
Lett. 2005, 7, 2035. (c) Pan, Y.; Lu, H.; Fang, Y.; Fang, X.; Chen, L.;
Qian, J.; Wang, J.; Li, C. Synthesis 2007, 1242.
(7) Taillefer, M.; Ouali, A.; Renard, B.; Spindler, J. Chem.sEur. J. 2006,
12, 5301.
(8) (a) Nordmann, G.; Buchwald, S. L. J. Am. Chem. Soc. 2003, 125, 4978.
(b) Miyamoto, H.; Okawa, Y.; Nakazaki, A.; Kobayashi, S. Angew. Chem.,
Int. Ed. 2006, 45, 2274.
(9) Grigg, R.; Stevenson, P.; Worakun, T. Tetrahedron 1988, 44, 2049.
(10) Kuwabe, S.; Torraca, K. E.; Buchwald, S. L. J. Am. Chem. Soc. 2001,
123, 12202.
a Isolated yield based on 5.
elimination product 7 and no reaction occurred for 1a. Moreover,
the Pd(OAc)2/BINAP-catalyzed reaction of 5b in refluxing 1,2-
dimethoxyethane gave the 5-exo cyclization product 8 rather than
JA072793W
9
J. AM. CHEM. SOC. VOL. 129, NO. 26, 2007 8093