Diastereoselective γ-vinyl butyrolactone synthesis via gold catalyzed cyclization of allylic acetate

Article abstract of DOI:10.1039/b913348h

An efficient method for the preparation of polysubstituted γ-vinyl butyrolactones was developed through gold catalyzed highly diastereoselective cyclization of the malonate substituted allylic acetates.

Full text of DOI:10.1039/b913348h

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Diastereoselective c-vinyl butyrolactone synthesis via gold catalyzed
cyclization of allylic acetatew
Ya-Hui Wang,a Li-Li Zhu,a Yu-Xin Zhangb and Zili Chen*a
Received (in Cambridge, UK) 6th July 2009, Accepted 17th November 2009
First published as an Advance Article on the web 4th December 2009
DOI: 10.1039/b913348h
An efficient method for the preparation of polysubstituted c-vinyl
butyrolactones was developed through gold catalyzed highly
diastereoselective cyclization of the malonate substituted allylic
acetates.
Table 1 Investigation of the reaction of compound 1a with different
catalystsa
The metal catalyzed Tsuji-Trost type reaction has been the
object of many studies,1 in which allylic esters are often used
as essential reaction partners. Many transition metals
(especially later transition metals, such as Ni, Pd, Ir, Rh etc.)
have been successfully utilized in these allyl coupling reactions
to prepare alkene-containing molecules with variable structural
characteristics and diversities. However, employing gold
complexes in the reaction of allylic esters has seldom been
reported.2 Herein we will describe an efficient method to
construct polysubstituted g-vinyl butyrolactone derivatives
through gold catalyzed highly diastereoselective cyclization
of the allylic acetates.
Catalyst
T/h
Sol./temp.
Yield (%)b/de
1
2
AuPPh3Cl/AgOTf
AuPPh3Cl/AgOTf
AuPPh3Cl/AgBF4
AuPPh3Cl/AgSbF6
AuPPh3Cl/AgPF6
AuPPh3Cl
AgSbF6
AuCl3
AuBr3
AuCl
1
2
2
3
1
1
2
2
1
1
2
2
DCM/rt.
0
24 (15/1)
15 (15/1)
80 (15/1)
DCE/70 1C
DCE/70 1C
DCE/70 1C
DCE/70 1C
DCE/70 1C
DCE/70 1C
DCE/70 1C
DCE/40 1C
DCE/70 1C
DCE/70 1C
DCE/70 1C
DCE/70 1C
DCE/70 1C
DCE/70 1C
DCE/70 1C
DCE/70 1C
DCE/70 1C
DCE/70 1C
3
4c
5
o5
6
7
8
0
Trace
o10
9d
10
11
12
13
14
15
16
17
18
19
81(15/1)
0
31(10/1)
Cu(OTf)2
BF3ꢀEt2O
43 (7/1)
o5
Homogeneous gold catalysis has attracted great attention in
recent years. Previous studies in this field mainly focused on
the transformations involving alkyne-, allene- or olefin-
containing derivatives.3,4 Only two results demonstrated the
allylic acetate’s isomerization,2a and its coupling with allylstannane
by using gold catalysts.2c Encouraged by these achievements,
we decided to investigate the gold catalyzed reaction of the
allylic acetates with the nucleophiles other than organometallic
reagents.5
TsOHe
2
HOAce
12
12
12
12
12
12
0
HCle
13
15
0
0
45
HOTfe
CF3COOHe
(CF3CO)2Oe
AuPPh3Cl/AgSbF6
f
1 eq. K2CO3
a
Unless noted, all reactions were carried out at 0.5 mmol scale in 5 mL
solvent with 2 mol% equivalent of catalyst, DCM = CH2Cl2, DCE =
b
C2H4Cl2. Isolated yields. Entry
c
Our initial effort was set to the gold catalyzed cyclization of
allylic acetates by using an ester as an intramolecular
nucleophile.z Syn-4-acetoxy cyclohexenyl malonate 1a was
chosen as the model system for our initial investigation. As
shown in Table 1, treating 1a with 2% mol AuPPh3Cl/AgOTf
in CH2Cl2 at rt. afforded no product (Table 1, entry 1). When
the reaction temperature was increased to 70 1C in DCE, to
our delight, a cyclized product 2a was obtained in 24% yield
with a high diastereoselectivity (de = 15/1). The structure of
2a was determined to be 3,4-anti-4,5-syn-3-methoxycarbonyl
tetrahydrobenzo butyrolactone by comparing with the
previous report.6
4 was used as condition A.
Entry 9 was used as condition B. 10% eq. of Brønsted acids were
d
e
f
used. 1.2 eq. K2CO3 was added.
AuPPh3Cl or AgSbF6 separately in the reaction yielded no
product. Among the non-phosphine-ligated gold catalysts,
AuBr3 had the highest catalytic activity (Table 1, entry 9,
condition B). Lewis acids such as BF3ꢀEt2O and Cu(OTf)2
gave 2a in moderate yields with diminished diastereo-
selectivities (Table 1, entry 11, 12). In the control experiments,
Brønsted acids tested afforded either trace amount of 2a or no
product (Table 1, entry 13–18). Whereas treating 1a in the
basic condition gave 2a in 45% yield after an increased
reaction time (Table 1, entry 19).
Combinations of gold catalyst with other silver salts were
then screened. AuPPh3Cl/AgSbF6 performed very well
(Table 1, entry 4, 80% yields, condition A). Employing
We then turned to examine the reaction’s scope and
limitations. As shown in Table 2, a range of cyclic and linear
substrates were tested, which provided the corresponding
polysubstituted mono- or bicyclic g-vinyl cyclobutyrolactones.
Based on the results in Table 2, two factors seemed to affect
the nucleophilicity of the malonate ester. First, the presence of
the hydrogen in the C3 position (see structure of 1a in Table 1)
and its high acidity favored the intramolecular nucleophilic
a Department of Chemistry, Renmin University of China,
Beijing 100872, China. E-mail: zilichen@ruc.edu.cn
b School of Chemistry and Environment, Beijing University of
Aeronautics and Astronautics, 100083, China
w Electronic supplementary information (ESI) available: Experimental
procedures and data for all new compounds. CCDC 750019. For ESI
and crystallographic data in CIF or other electronic format see DOI:
10.1039/b913348h
ꢁc
This journal is The Royal Society of Chemistry 2010
Chem. Commun., 2010, 46, 577–579 | 577

View Article Online
Table 2 Gold catalyzed cyclization of allylic acetates 1 to afford
g-vinyl butyrolactone derivativesa
entry 10). Substrate 1d’s acidity, which has only one methyl
ester on the C3 position, was lower than that of the other
substrates. As expected, 1d provided product 2d in relatively
low yield (45% yield Table 2, entry 4). Second, ester alkyl
group’s lability enhanced the carbonyl oxygen’s nucleophilicity,
and improved the substrate’s reactivity. For example:
compound 1b had a much higher reactivity than 1a did, and
afforded 2b in near quantitative yield (Table 2, entry 2);
Substrates 1k, 1m, and 1n, which had a methyl, allyl or
methoxycarbonyl methyl group at the C3 position, afforded
the desired products in moderate to good yields (2k, 74%
yield, 2m, 69% yield, 2n, 86% yield, Table 2, entry 11–13), as
compared to the poor reactivity of 1j.
Substrate
Product
Time/h Yield(%)b/de
1c
1
2
2a, 81 (15/1)e
2b, 99 (16/1)e
2c
3d
4d
5d
20
20
20
2c, 56 (1.5/1)e
Linear substrates 1e–1i, bearing an alkyl or phenyl group in
the alkene’s neighboring position, also worked well. Among
these compounds, substrates 1e and 1h, which contained a
phenyl group (Table 2, entry 5) or an electron donating
p-methoxyphenyl group (Table 2, entry 8), worked better than
1f and 1g (with alkyl or electro-withdrawing p-chlorophenyl
group, Table 2, entry 6, 7) did. However, compound 1i, which
had an electron donating 2,4,6-trimethoxy phenyl group,
produced 2i in 48% yield (1i, Table 2, entry 9). In cyclic
substrates, the additional ester group and allyl group at the C3
position in 1m and 1n did not affect the reaction (Table 2,
entry 12–13).
2d, 45
2e, 90 (19/1)e
6d
0.5
2f, 67 (10/1)e
2g, 65 (19/1)e
2h, 95 (21/1)e
7d
8d
17
4
High diastereoselectivities were observed in almost all
products, except 2c (Table 2, entry 3). This might be owing
to inefficient stereo-control by the non-neighboring steric
effect in open-chain molecules.7 Linear compounds 1e–1i
provided 3,4-trans-4,5-trans g-vinyl butyrolactones in high
diastereoselectivities. Products 1e, 1g, 1h, 1i, which contained
bulky phenyl groups at the neighboring position, had better
stereoselectivities than 1f did.8 Whereas cyclic substrates
afforded 3,4-trans-4,5-cis products, and gave de values ranging
from 15/1 to 99/1.
9d
1
2i, 48 (20/1)e
10c
11c
24
NR
2k, 74 (27/1)f
We then investigated the gold catalyzed intermolecular
nucleophilic addition onto acetate 3. As shown in Scheme 1,
the reaction of allyl alcohol 4c with acetate 3 provided diallyl
ether 5c in 95% yield.9 Benzyl alcohol 4d or allyl amine 4e
afforded 5d or 5e in relatively low yields (5d in 26%, and 5e
in 30%), together with a certain amount of isomerized
product 6. The reaction of methanol or tert-butanol, however,
gave only trace amounts of products.
0.5
12c
0.5
0.5
2m, 69 (499/1)g
2n, 86 (25/1)f
13c
It was assumed that the enhanced reactivity of 4d, 4e
and especially 4c might be a result from the formation of
alkene- or phenyl-auric complexes in the reaction, which
actually led to the nucleophilic addition of these conju-
gated substrates onto acetate 3 to proceed in a pseudo-
intramolecular fashion.
a
Unless noted, all reactions were carried out at 0.5 mmol scale in 5 mL
b
CH2ClCH2Cl with 2% mol equivalent of catalyst. Isolated yields.
c
Condition B, 2% mol eqivalent of AuBr3 at 40 1C; Condition B
doesn’t work well for the linear substrates. Condition A, 2% mol
d
e
equivalent of AuPPh3Cl/AgSbF6 at 70 1C. The de value were
determined by 1H NMR spectra. The de value were determined by
f
g
separated yields. No isomer product detected.
addition of the ester onto the alkene. Compound 1j was an
example of opposite reactivity, in which C3 hydrogen was
replaced by a methyl group. As a result, when 1j was treated
with AuPPh3Cl/AgSbF6 or AuBr3 even at an elevated
temperature, no desired product was obtained (1j, Table 2,
Scheme 1 The reaction of acetate 3 with several alcohols and amines.
ꢁc
This journal is The Royal Society of Chemistry 2010
578 | Chem. Commun., 2010, 46, 577–579

View Article Online
Notes and references
z Typical procedure for the gold catalyzed cyclization of allylic
acetates: the described compound. (Procedure A) The gold catalyst
was generated in an oven-dried schlenk tube containing a magnetic stir
bar under N2 by addition of AgSbF6 (0.02 eq.), AuPPh3Cl (0.02 eq.),
and 1 mL 1,2-dichloroethane. After 2 min, a solution of substrate 1a
(0.5 mmol) in 4 mL 1,2-dichloroethane was added. The resulting
mixture was allowed to stand at 70 1C until complete consumption
of the starting material with TLC monitoring. After removal of the
solvent, the residue was purified by column chromatography on silica
gel to afford 2a in 80% yield.
Scheme 2 The effect of the substrate 1a’s stereo configuration on the
reaction yield.
1 (a) B. M. Trost and T. R. Verhoeven, in Comprehensive
Organometallic Chemistry, ed. G. Wilkinson, F. G. A. E. W. Stone
and Abel, Pergamon, Oxford, 1982, vol. 8, pp. 799–938;
(b) S. A. Godelski, in Comprehensive Organic synthesis,
ed. B. M. Trost and I. Fleming, Pergamon, Oxford, 1991, vol. 4,
pp. 585–661; (c) L. S. Hegedus, in Organometallics in synthesis,
ed. M. Schlosser, Wiley, Chichester, 1994, pp. 427–444;
(d) L. S. Hegedus, in Organische Synthesemit ubergangsmetallen,
VCH, Weinhein, 1995.
¨
2 (a) N. Marion, R. Gealageas and S. P. Nolan, Org. Lett., 2007, 9,
2653; Erratum: N. Marion, R. Gealageas and S. P. Nolan, Org.
Lett., 2008, 10, 1037; (b) C. Gourlaouen, N. Marion, S. P. Nolan
and F. Maseras, Org. Lett., 2009, 11, 81; (c) S. Porcel, V. Lopez-
Carrillo, C. Garca-Yebra and A. M. Echavarren, Angew. Chem.,
Int. Ed., 2008, 47, 1883.
3 For recent general reviews, see: (a) Z. Li, C. Brouwer and C. He,
Chem. Rev., 2008, 108, 3239; (b) A. Arcadi, Chem. Rev., 2008, 108,
3266; (c) D. J. Gorin and F. D. Toste, Nature, 2007, 446, 395;
(d) A. S. K. Hashmi, Chem. Rev., 2007, 107, 3180; (e) A. Furstner
and P. W. Davies, Angew. Chem., Int. Ed., 2007, 46, 3410;
(f) R. Skouta and C.-J. Li, Tetrahedron, 2008, 64, 4917.
4 The reaction of allenyl carbinol esters or allyl alcohol:
(a) A. K. Buzas, F. M. Istrate and F. Gagosz, Org. Lett., 2007,
9, 985; (b) A. Aponick, C.-Y. Li and B. Biannic, Org. Lett., 2008,
10, 669.
Scheme 3 Investigation of the reaction of chiral substrates 7a and 8b.
The effect of the substrates’ stereochemistry on the reaction
yield were then examined. Treating trans-1a with AuBr3 in
DCE at 40 1C for 2 h, gave 2a in 29% yield (Scheme 2), which
is much lower than that of cis-1a (Table 2, entry 1). It might
indicate that the reaction was favored to take a stereochemical
coordinative route.10
To elucidate the reaction mechanism, two chiral substrates
7a and 8b were prepared and investigated (Scheme 3).11,12 It
was found that the reaction of 7a afforded product 9 in 31%
yield with a low ee value. While the reaction of the isomerized
substrates 8b yielded the racemic butyrolactone 10.
Therefore, this reaction most probably started from the
generation of the allyl cation intermediates from the allylic
acetates in the condition of the gold catalyst.13 The following
nucleophilic addition was facilitated by the formation of the
malonate’s enol tautomer in the dimethyl substrates and the
lability of the benzyl group in the C3-tetrasubstituted dibenzyl
substrates.
5 The reaction of allylic acetates with allylstannanes or allylsilanes
have been reported, see ref. 3c.
6 The structure of 2a was determined by comparing its 1H and 13C
NMR spectra with previous reports. (a) L. Lamarque,
M. Campredon, A. Meou, P. Brun and R. Faure, Magn. Reson.
Chem., 1998, 36, 463.
7 (a) Stereo control by neighboring group: P. A. Bartlett, Tetrahedron,
1980, 46, 3; (b) M. D. Dowle and D. I. Davies, Chem. Soc. Rev.,
1979, 8, 171; (c) G. Cardillo and M. Orena, Tetrahedron, 1990, 46,
3321.
Subsequent demethylation of the in situ formed alkylidene
ketal A (Scheme 4) and debenzylation of the ionic intermediate
B afforded the desired g-vinyl butyrolactones.14 According to
the control experiment results (Table 1, entry 14 and 19),
in situ generated acetic acid did not affect the reaction.
In conclusion, we have developed an efficient method to
construct polysubstituted g-vinyl butyrolactones from the gold
catalyzed cyclization of the allylic acetates. This is the first
report, to our knowledge, of a gold catalyzed intramolecular
nucleophilic addition of esters onto allylic acetates. The effort
to extend this reaction and broaden its application is still
under way in this lab.
8 The structure of compound 2e was determined by comparing its 1H
and 13C NMR and spectra with previous reports: E. Fillion,
S. Carret, L. G. Mercier and V. E. Trepanier, Org. Lett., 2008,
10, 437.
9 The reaction of allyl silane with allyl acetate has ever been reported
by Echavarren’s group. See ref. 3c.
10 SN20 process has ever been assumed based on the results in
Scheme 3. For an comprehensive review on the SN20 reaction,
seeR. M. Magid, Tetrahedron, 1980, 36, 1901.
11 The absolute configuration of substrate 7a was induced by the
optical rotation data of 7a and 7b. Method to predict a small
molecule’s absolute configuration on the basis of its optical
rotation sign: (1)D. Z.-G. Wang, Tetrahedron, 2005, 61, 7125;
D. Z.-G. Wang, Tetrahedron, 2005, 61, 7134. The absolute
configuration of substrate 8b was determined by the crystal
structure of the diastereomer 8a.
Support of this work by the grant from the National
Sciences Foundation of China (No. 20872176) is gratefully
acknowledged.
12 Crystallographic data for compound 8a: C25H29NO6S, M =
471.55, Monoclinic, a = 11.814(2) A, alpha = 90 deg, b =
5.9898(12) A, beta
gamma = 90 deg, Volume 1185.3(4) A3, T = 113(2) K, space
group P21, Z = 2, 8646 reflections measured, 3921 unique [Rint
= 106.02(3) deg, c = 17.425(4) A,
=
0.0266]. The final wR(F2) was 0.0904 (all data). CCDC no 750019.
See supporting information for details.
13 C.-C. Lin, T.-M. Teng, A. Odedra and R.-S. Liu, J. Am. Chem.
Soc., 2007, 129, 3798.
14 A plausible explanation for the diastereoselectivities was provided
in the supporting information section.
Scheme 4 Two possible intermediates A and B.
ꢁc
This journal is The Royal Society of Chemistry 2010
Chem. Commun., 2010, 46, 577–579 | 579