Modulating the stereochemical outcome of the Ireland-Claisen reaction of (E)- and (Z)-allylic glycolates

Article abstract of DOI:10.1016/j.tetlet.2011.12.011

The diastereoselectivity of Ireland-Claisen rearrangements of allylic glycolates is dependent on the E:Z ratio of the silyl ketene acetals, the alkene geometry in the allyl unit, and the transition state topography. High yields and excellent diastereoselectivities (>95:5) have been achieved for selected substrates, including those with R₂ = ethyl that results in a newly formed quaternary center. A discussion of the scope, selectivities, and transition state models will be presented.

Full text of DOI:10.1016/j.tetlet.2011.12.011

Tetrahedron Letters 53 (2012) 825–828
Contents lists available at SciVerse ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Modulating the stereochemical outcome of the Ireland–Claisen reaction
of (E)- and (Z)-allylic glycolates
⇑
Ken S. Feldman , Brandon R. Selfridge
Department of Chemistry, The Pennsylvania State University, University Park, PA 16802, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
The diastereoselectivity of Ireland–Claisen rearrangements of allylic glycolates is dependent on the E:Z
ratio of the silyl ketene acetals, the alkene geometry in the allyl unit, and the transition state topography.
High yields and excellent diastereoselectivities (>95:5) have been achieved for selected substrates,
including those with R2 = ethyl that results in a newly formed quaternary center. A discussion of the
scope, selectivities, and transition state models will be presented.
Received 17 November 2011
Revised 1 December 2011
Accepted 2 December 2011
Available online 8 December 2011
Ó 2011 Elsevier Ltd. All rights reserved.
Keywords:
Ireland–Claisen
Allylic glycolate
Silyl ketene acetal
Diastereoselectivity
Lewis acid
Introduction
(Fig. 1) was examined first. The stereochemical outcome of these
transformations with 1a and 1b can be rationalized by citing reac-
The Ireland–Claisen rearrangement is an important C–C bond
forming [3,3] sigmatropic process first described in 1972.1 Since its
introduction, many groups have utilized this transformation both
in the construction of natural products2,3 and in the further meth-
odology development, set forth nicely in numerous reviews.4–6
Most notably, contiguous stereogenic centers can be formed with
high diastereoselectivity. The stereochemical outcome is dictated
by the geometries of both the intermediate silyl ketene acetal
and allylic alkene, and the transition state topology. While Lewis
acids have been shown to promote the Ireland–Claisen reaction
with increased diastereoselectivity,7 no systematic studies have
tion through the canonical chair-like transition state 2 featuring a
(Z)-silyl ketene acetal. The (Z)-silyl ketene acetal geometry pre-
sumably results via chelation-controlled9 enolization of the glyco-
late. By utilizing TMSCl as the silylating agent and a base/acid
workup, easily isolable carboxylic acids were produced without
the need for further purification. Very little enhancement in the
diastereoselectivity was seen with the addition of catalytic
amounts of Lewis acids when using LHMDS as the base (Table 1).
For substrate 1a, SnCl4 appeared to be marginally more effective
by both increasing the yield from 94% to near quantitative and
the dr from 88% to 94% (Table 1, compare entries 1 and 6).10 Vari-
ation in the outcome of Ireland–Claisen rearrangement for sub-
strate 1b was a little more pronounced, with an increase in the
yield from 83% to 99% by the addition of catalytic ZnCl2 (Table 1,
compare entries 9 and 14).
been reported that examine the reactions of
a-substituted glyco-
lates under Lewis acid mediation. In addition, all documented
examples of Ireland–Claisen reactions of glycolates utilized TMSCl
as the silylating agent without investigating possibly superior
alternatives.8 This work focuses on the effects of base, silyl reagent,
solvent, and additional a-substitution (resulting in a newly formed
quaternary center) on overall yield and diastereoselectivity.
Results and discussion
Unsubstituted benzyloxy glycolates containing an (E)-allylic
alkene
The effect of Lewis acid mediation on the diastereoselectivity of
the Ireland–Claisen rearrangement of
a-unsubstituted glycolates
⇑
Corresponding author.
E-mail address: ksf@chem.psu.edu (K.S. Feldman).
Figure 1. Ireland–Claisen reaction of benzyloxy glycolates 1a and 1b.
0040-4039/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2011.12.011

826
K. S. Feldman, B. R. Selfridge / Tetrahedron Letters 53 (2012) 825–828
Table 1
Table 2
Lewis acid survey and base comparison with 1a and 1b using TMSCl (1.5 equiv) as
silylating agent and THF as solvent (emboldened values reflect the best results
throughout)
Use of increased amounts of base in the Ireland–Claisen rearrangements of 1a and 1b
using TMSCl (3 equiv) and THF as solvent
Entry
Subst
Base (3 equiv)
Yield (%)
dra (%)
Entry
Subst
Base (1.5 equiv)
Additive (mol %)
Yield (%)
dra (%)
1
2
3
1a
1a
1a
NaHMDS
LHMDS
KHMDS
64
78
12
>95:5b
76
60
1
2
3
4
5
6
1a
1a
1a
1a
1a
1a
LHMDS
LHMDS
LHMDS
LHMDS
LHMDS
LHMDS
SnCl4 (2)
TiCl4 (2)
BF3ÁOEt2 (20)
Ti(i-OPr)4 (5)
ZnCl2 (5)
None
100
94
99
69
45
94
94
92
91
93
92
88
4
5
6
1b
1b
1b
NaHMDS
LHMDS
KHMDS
83
80
46
>95:5b
82
90
a
>95:5b
—
dr = 3a/4a (from 1a) or 3b/4b (from 1b) ratio.
7
8
1a
1a
KHMDS
NaHMDS
None
None
7
b
1H NMR detection limit.
—
9
10
11
12
13
14
15
16
1b
1b
1b
1b
1b
1b
1b
1b
LHMDS
LHMDS
LHMDS
LHMDS
LHMDS
LHMDS
KHMDS
NaHMDS
ZnCl2 (5)
SnCl4 (2)
BF3ÁOEt2 (20)
Ti(i-OPr)4 (5)
TiCl4 (2)
None
99
93
92
92
84
83
27
—
>95:5b
>95:5b
>95:5b
>95:5b
>95:5b
>95:5b
84
None
None
—
a
dr = 3a/4a (from 1a) or 3b/4b (from 1b) ratio.
b
1H NMR detection limit.
The most notable result that emerged from these experiments
was the superiority of LHMDS (1.5 equiv vs substrate) as a base
for 1a and 1b when compared to NaHMDS or KHMDS. The latter
two bases at 1.5 equiv vs substrate delivered rearrangement prod-
uct in only poor yield when they worked at all. It is possible that
the kinetics of glycolate deprotonation could play a role in rational-
izing these observations.
Figure 2. Ireland–Claisen reaction of ethyl-substituted benzyloxy glycolates 5a and
5b.
Increasing the amount of base used in the deprotonation im-
pacted on both the yield and diastereoselectivity of this transform
(Table 2). Using 3 equivalents of NaHMDS and KHMDS led to much
higher yields of rearrangement products compared to the 1.5 equiv
of base as reported in Table 1. In addition, the best diastereoselec-
tivity now was obtained with a sodium counterion for both 1a and
1b. Overall, the best conditions for the Ireland–Claisen rearrange-
ment of unsubstituted benzyloxy glycolates 1a and 1b utilize
1.5 equiv LHMDS, 1.5 equiv of TMSCl, and catalytic amounts of
SnCl4 and ZnCl2, respectively.
of TMS (Table 3, entry 15). Interestingly, using NaHMDS reversed
the diastereoselectivity of the reaction (Table 3, entry 16). It is not
clear where the effects of the switch to Na+ as a counterion are man-
ifest, that is, silyl ketene acetal geometry or the transition stage
topography. Overall, the incorporation of an a-ethyl appendage on
the substrates 1a and 1b neither derailed the Ireland–Claisen rear-
rangement of the derived species 5a and 5b nor diminished the
diastereoselectivity attainable, but the chemical yields did suffer a
little.
Ethyl-substituted benzyloxy glycolates containing an (E)-allylic
alkene
Lewis acid mediation along with base variation and silylating
agent variation was examined for the Ireland–Claisen reaction of
the ethyl-substituted benzyloxy glycolates 5a and 5b (Fig. 2). These
reactions result in the creation of a quaternary stereogenic center
with excellent diastereoselectivity, as depicted in 7 and 8. By going
through a transition state featuring a (Z)-silyl ketene acetal in a chair
conformation, 6, the phenyl-bearing species 5a is transformed with-
out Lewis acid present primarily into 7a in modest yield and with
mediocre diastereoselectivity (Table 3, entry 1). Including Lewis
acids in the reaction solution did not improve the outcome
(Table 3, entries 2–5). Overall, the most notable improvements in
yield and diastereoselectivity resulted when the bulkier TIPS group
replaced the TMS group (Table 3, entry 7).
However, the addition of select Lewis acids did marginally
improve the observed diastereoselectivity with the i-Pr-bearing
substrate 5b. When using LHMDS, the diastereoselectivity achieved
in going from 5b to 7b was increased by the addition of catalytic
SnCl4 (Table 3, entry 9). This same level of diastereoselectivity was
achieved using KHMDS without the need of a Lewis acid (Table 3,
entry 14), with the added benefit of a slight increase in yield. Once
again, as with the phenyl-bearing substrate 5a, the isopropyl-
substituted substrate 5b proceeded to product 7b with the highest
yield and greatest diastereoselectivity when TIPS was used instead
Table 3
Lewis acid survey and base comparison of 5a and 5b using TMSCl (1.5 equiv) and THF
as solvent
Entry
Subst
Base (1.5 equiv)
Additive (mol %)
Yield (%)
dra (%)
1
2
3
4
5
5a
5a
5a
5a
5a
LHMDS
LHMDS
LHMDS
LHMDS
LHMDS
None
48
36
44
20
13
84
82
84
60
75
TiCl4 (2)
SnCl4 (2)
Ti(i-OPr)4 (5)
ZnCl2 (5)
6
7c
8
5a
5a
5a
KHMDS
KHMDS
NaHMDS
None
None
None
—
67
16
—
93
65
9
10
11
12
13
5b
5b
5b
5b
5b
LHMDS
LHMDS
LHMDS
LHMDS
LHMDS
SnCl4 (2)
TiCl4 (2)
Ti(i-OPr)4 (5)
ZnCl2 (5)
None
43
34
42
47
40
>95:5b
>95:5b
93
87
88
14
15c
16
5b
5b
5b
KHMDS
KHMDS
NaHMDS
None
None
None
53
57
22
>95:5a
>95:5
59
a
b
c
dr = 7a/8a (from 5a) or 7b/8b (from 5b) ratio.
1H NMR detection limit.
Entries 7 and 15 utilized TIPSOTf as the silylating agent, 1:1 THF/toluene as the
solvent, and a final nBu4NF treatment to form the acid product.

K. S. Feldman, B. R. Selfridge / Tetrahedron Letters 53 (2012) 825–828
827
Figure 4. Determination of silyl ketene acetal geometry via nOe analysis.
Figure 3. Ireland–Claisen reaction of ethyl substituted PMB glycolate 9 containing a
(Z)-allylic alkene.
Ethyl substituted (p-methoxy)benzyloxy glycolates containing a
(Z)-allylic alkene
Figure 5. Stereochemical assignment of 3a.
The diastereoselectivity of the Ireland–Claisen rearrangement
for the phenyl-bearing (Z)-alkene substrate 9 predictably was re-
versed compared to the (E)-alkene series 5a/5b (Fig. 3). Upon the
addition of substrate 9 to a mixture of base and silylating reagent,
Ireland–Claisen rearrangement proceeded to furnish the corre-
sponding silyl ester, presumably through a Z-silyl ketene acetal
in a chair conformation, 10. Desilylation of this isolable silyl ester
afforded carboxylic acid 11 in generally good yield and with high
diastereoselectivity over the two-step sequence (Table 4). The
use of the PMB ether should impart more flexibility into the ter-
tiary alcohol deprotection chemistry, which may prove valuable
in the downstream synthesis projects.
Independent evidence for the (Z) configuration of the intermedi-
ate silyl ketene acetal can be found through the use of a Claisen-
inactive model system 13 (Fig. 4). The derived silyl ketene acetal
14 was isolated and analyzed by 1H NMR nOe experiments. An
observed correlation between the PMB CH2 and the TIPS C–H estab-
lished the silyl ketene acetal geometry in this species, and by infer-
ence the Claisen-active silyl ketene acetals formed from 9, as (Z).
The influence of different silylating reagents on diastereoselec-
tivity was explored with substrate 9 (Table 4). After observing
lower yields utilizing TMSCl as the silylating agent for ethyl substi-
tuted glycolates (Table 3), the more reactive triflates were exam-
ined. Both TIPSOTf and TESOTf greatly increased the yield and
diastereoselectivity of the 9 ? 11/12 conversion (Table 4, entries
1 and 8). These larger silyl ethers may impact on the yield and dia-
stereoselectivity of the rearrangement at several key junctures; (a)
their greater electrophilicity (compared to the corresponding chlo-
rides) might promote more rapid enolate trapping, thus minimiz-
ing competitive side reactions, (b) the greater steric bulk
Figure 6. Stereochemical assignment of 11.
compared to TMS may make alternative transition states to 10
more energetically penalizing, and (c) the resultant silyl ester
products may be more resistant to adventitious cleavage/yield loss
compared to the trimethylsilyl alternative. It also should be noted
that comparable yields were achieved using TMSCl in the presence
of Et3N, but the diastereoselectivity was a little lower.
A survey of several different bases again was conducted.
KHMDS was determined to be the superior base for the transfor-
mation of 9 to 11 when using TIPSOTf as the silylating agent. NaH-
MDS failed to give any product. Furthermore, it was observed that
mixing THF with toluene (ꢀ1:1 ratio) as the solvent system in-
creased the diastereoselectivity of the reaction.
Stereochemical assignments by nOe
Stereochemical assignment of the a-unsubstituted acid 3a was
accomplished by first performing a selenolactonization to give a
selenium ether, followed by oxidative elimination within that
ether to give 15 (Fig. 5) The stereochemical assignment of 15,
and hence 3a, follows from nOe analysis. The stereochemical
assignment of the major isomer of the isopropyl series, 3b, was
Table 4
Silylating reagent and solvent effects for transformation of 9 into 11 and 12, respectively
Entry
Base (3 equiv)
Solvent
Silyl reagent (3 equiv)
Additive (mol %)
Yield (%)
dra (%)
1
2
3
4
5
6
7
KHMDS
KHMDS
KHMDS
LHMDS
NaHMDS
KHMDS
KHMDS
Toluene/THF
Toluene/THF
THF
THF
THF
TIPSOTf
TIPSOTf
TIPSOTf
TIPSOTf
TIPSOTf
TIPSOTf
TIPSCl
TiCl4(10)
73
67
34
50
—
95
—
—
—
—
—
—
>95:5b
>95:5b
93
—
87
75
Toluene
Toluene
71
61
8
KHMDS
Toluene
TESOTf
—
75
92
9
10
11
KHMDS
KHMDS
KHMDS
Toluene
Toluene
Toluene
TMSOTf
TMSCl
TMSCl
—
Et3N
—
—
71
13
—
87
78
a
dr = 11/12 ratio.
b
1H NMR detection limit.

828
K. S. Feldman, B. R. Selfridge / Tetrahedron Letters 53 (2012) 825–828
inferred by the comparison of its 1H NMR spectral data with those
of 3a.
7b, 9, 11, 13, 14, 15, and 16.) associated with this article can be
found, in the online version, at doi:10.1016/j.tetlet.2011.12.011.
Stereochemical assignments within the ethyl-substituted series
were accomplished by first iodolactonizing the corresponding car-
boxylic acid 11, followed by reductive removal of the iodide to give
a single compound 16 that was analyzed by 1H NMR nOe (Fig. 6).
Similarly, the stereochemical assignments of the major isomer 7a
from the benzyl ether/(E)-alkene series were inferred from 1H
NMR spectral data comparisons with 11.
References and notes
1. Ireland, R. E.; Mueller, R. H. J. Am. Chem. Soc. 1972, 94, 5897–5898.
2. Koch, G.; Janser, P.; Kottirsch, G.; Romero-Giron, E. Tetrahedron Lett. 2002, 43,
4837–4840.
3. Kallmerten, J.; Gould, T. J. Org. Chem. 1985, 50, 1128–1131.
4. McFarland, C. M.; McIntosh, M. C. In The Claisen Rearrangement: Methods and
Applications; Hiersemann, M., Nubbemeyer, U., Eds.; Wiley-VCH Verlag GmbH
& Co. KGaA: Weinheim, Germany, 2007; pp 117–120.
5. Castro, A. M. Chem. Rev. 2004, 104, 2939–3002.
6. Chia, Y.; Hong, S.; Lindsay, H. A.; McFarland, C.; McIntosh, M. C. Tetrahedron
2003, 58, 2905–2928.
7. Bedel, O.; Haudrechy, A.; Langlois, Y. Eur. J. Org. Chem. 2004, 18, 3813–3819.
8. For example, see: Jung, D. Y.; Kang, S.; Chang, S.; Kim, Y. H. Synlett 2006, 86–90.
9. Burke, S.; Fobare, W.; Pacofsky, G. J. Org. Chem. 1983, 48, 5221–5228.
10. Representative experimental: 2-Benzyloxy-3-methyl-5-phenyl-pent-4-enoic acid
Conclusions
The Ireland–Claisen rearrangement of both (E)- and (Z)-alkenyl
a-ethylglycolates can proceed with high levels of diastereoselec-
tivity, and in excellent yields, under certain experimental condi-
tions. The best results were obtained by using bulkier silylating
reagents (e.g., TIPS), potassium hexamethyldisilazide as a base,
and in a 1:1 toluene-to-THF solvent system. Catalytic Lewis acid
additives did not have significant impact on either the yield or dia-
stereoselectivity of the rearrangement for the systems examined.
(3a). To a stirring solution of LHMDS (1.0 M in THF, 405
2.8 mL of THF at À78 °C was added freshly distilled TMSCl (53
dropwise. A solution of benzyloxy-acetic acid 1-phenyl-but-2-enyl ester (1a)
(0.080 g, 0.269 mmol) in 200 L of THF was added dropwise, followed by SnCl4
(1.0 M in CH2Cl2, 11
L, 0.011 mmol). The solution was stirred at À78 °C for
l
L, 0.40 mmol) in
lL, 0.40 mmol)
l
l
30 min, 0 °C for 30 min, and then warmed to room temperature. After 14 h, 1 M
NaOH was added, stirred vigorously for 1 h, and Et2O was added. The resulting
solution was partitioned between Et2O and 1 M NaOH and the organic layer
was extracted with 1 M NaOH. The combined aqueous fractions were acidified
with 3 M HCl to pH 3, extracted with EtOAc, dried over Na2SO4, filtered, and
concentrated in vacuo to give 2-benzyloxy-3-methyl-5-phenyl-pent-4-enoic
Acknowledgment
Financial support from the National Science Foundation (CHE
0956458) is gratefully acknowledged.
acid (3a) (0.080 g, 100%, 94% dr) as a colorless oil. IR (thin film) 1713 cmÀ1 1H
;
NMR (400 MHz, CDCl3) d 7.37–7.19 (m, 10H), 6.47 (d, J = 15.9 Hz, 1H), 6.19 (dd,
J = 15.9, 8.0 Hz, 1H), 4.77 (d, J = 11.7 Hz, 1H), 4.48 (d, J = 11.7 Hz, 1H), 4.00 (d,
J = 4.7 Hz, 1H), 2.90 (m, 1H), 1.21 (d, J = 6.9 Hz, 3H); 13C NMR (100 MHz, CDCl3)
d 177.9, 137.7, 137.5, 131.35, 131.26, 129.0, 128.9, 128.6, 128.5, 127.8, 126.7,
82.0, 73.5, 40.9, 15.7; LRMS (ESI) m/z (relative intensity) 314.2 (100%, M+NH4+).
Supplementary data
Supplementary data (Experimental procedures, spectral data,
and copies of 1H and 13C NMR spectra for 1b, 3a, 3b, 5a, 5b, 7a,