Extending Pummerer reaction chemistry. Asymmetric synthesis of spirocyclic oxindoles via chiral indole-2-sulfoxides

Article abstract of DOI:10.1021/ol0617244

The conversion of (S₈)-3-(ω-allylsilane and silyl enol ether)indole-2-sulfoxides to spirocyclic indolenines and then to oxindoles proceeds, in favorable cases, with moderate levels of chirality transfer from sulfur to C(3) of the indole core. A

Full text of DOI:10.1021/ol0617244

ORGANIC
LETTERS
2006
Vol. 8, No. 18
4137-4140
Extending Pummerer Reaction
Chemistry. Asymmetric Synthesis of
Spirocyclic Oxindoles via Chiral
Indole-2-sulfoxides
Ken S. Feldman* and Andrew G. Karatjas
104 Chemistry Building, Chemistry Department, The PennsylVania State UniVersity,
UniVersity Park, PennsylVania 16802
ksf@chem.psu.edu
Received July 13, 2006
ABSTRACT
The conversion of (Ss)-3-(
favorable cases, with moderate levels of chirality transfer from sulfur to C(3) of the indole core. A mechanistic model, which features either
an Sn2 -like additive Pummerer sequence or a tight ion pair generated by an Sn1-like vinylogous Pummerer transform, is proposed to rationalize
ω-allylsilane and silyl enol ether)indole-2-sulfoxides to spirocyclic indolenines and then to oxindoles proceeds, in
′
the sense of asymmetric induction.
Recent advances in the Pummerer chemistry of heteroarenes,
including imidazoles,1 furans,2 benzofurans,2b and thiophenes,2
in addition to indole-2-sulfoxides,3 have documented that
these processes are effective and efficient methods for
achieving formal oxidative nucleophilic additions to the het-
erocyclic core with complete control of regiochemistry and
with complete avoidance of product overoxidation. In con-
trast to the furan, benzofuran, and thiophene series, the
Pummerer reaction of 3-substituted indole-2-sulfoxides (cf.
1, Scheme 1) is not accompanied by rearomatization, and
the generation of a new stereogenic center at C(3) raises the
possibility of chirality transfer from sulfur-to-carbon in these
systems.
Scheme 1. Additive and Vinylogous Pummerer Mechanisms
for a Chiral Indole-2-sulfoxide
The asymmetric synthesis of C-Nu bonds via the Pum-
merer reaction of chiral sulfinate substrates has had an
(1) Feldman, K. S.; Skoumbourdis, A. P. Org. Lett. 2005, 7, 929-931.
(2) (a) Akai, S.; Kawashita, N.; Satoh, H.; Wada, Y.; Kakiguchi, K.;
Kuriwaki, I.; Kita, Y. Org. Lett. 2004, 6, 3793-3796. (b) Padwa, A.; Nara,
S.; Wang, Q. Tetrahedron Lett. 2006, 47, 595-597.
(3) Feldman, K. S.; Vidulova, D. B.; Karatjas, A. G. J. Org. Chem. 2005,
70, 6429-6440.
uneven history, with notable successes recorded by Marino,4
Kita,5 Garc´ıa Ruano,6 and Padwa6 amidst many failed
10.1021/ol0617244 CCC: $33.50
© 2006 American Chemical Society
Published on Web 08/08/2006

attempts.7 As exemplified with the chiral indole-2-sulfoxide
1, significant sulfur-to-carbon transfer of stereochemical
information is possible only if the intermediacy of an achiral
thionium ion 6 can be avoided. In earlier studies on other
systems, both concerted bond-breaking/bond-making schemes4
and, in independent work, tight ion pairs that preserve
(planar) chirality5,6 have been invoked as the means to steer
reaction away from an achiral thionium ion.
With unsaturated sulfoxide substrates, the course of the
Pummerer reaction embodies a mechanistic dichotomy: re-
arrangement via either an additiVe or a Vinylogous pathway,
as exemplified with the indole-2-sulfoxide substrate in
Scheme 1. The additive pathway offers an intrinsic conduit
for passing stereochemical information from sulfur to the
newly forming C-Nu1 bond (2f3), at least in principle (cf.
Marino’s work4). On the other hand, the vinylogous alterna-
tive can proceed either with loss of stereochemistry through
an achiral thionium ion 6 or with preservation of stereo-
chemical information via a planar chiral tight ion pair 5. In
the case of the indole-2-sulfoxides, the nucleophilic addition
products derived from either pathway are identical, prevent-
ing a simple assignment of mechanism based on the product
formed.
excesses of the ketone precursors to 9a/b were assayed at >
98% by the chiral shift reagent (S)-2,2,2-trifluoroanthryl
ethanol, as expected by the use of enantiopure 10.
The Pummerer chemistry commenced with the simple
allylsilane-bearing sulfoxide 7a, Table 1. Unfortunately, 1H
Table 1. Pummerer-Initiated Cyclization of the
Allylsilane-Bearing Chiral (S)-Indole-2-sulfoxide 10a
polaritya
(viscos)b
yield 13
(%)
eec
(%)
entry
solvent
T (°C)
1
2
3
4
5
6
7
8
CF3CH2OH
n-C5H11OH
C2H2Cl4
CH2Cl2
Et2O
toluene
toluene
Et2O
59.8 (2.0)
49.1 (25.4)
39.4 (3.7)
40.7 (0.73)
34.5 (0.28)
33.9 (1.2)
-40
-78
-40
-78
-78
-78
-90
-110
18
17
17
57
39
41
NR
NR
0
0
26
38
49
55
Six chiral indole-2-sulfoxides 7a/b, 8a/b, and 9a/b were
chosen to study the possibility of S f C chirality transfer,
Figure 1 (see the Supporting Information for syntheses). The
a ET(30) values (kcal/mol) from: Reichardt, C. Chem. ReV. 1994, 94,
2319-2358. b Centipoise, from: CRC Handbook of Chemistry and Physics,
85th ed.; Lide, D. R., Ed.; CRC Press: Boca Raton, 2004; 2004-2005.
c Average of two independent trials.
NMR chiral shift reagent studies with thioimidate 11 did not
lead to identification of a characteristic signal that permitted
differentiation of the enantiomers. Only upon conversion of
11 into 13 were the enantiomers distinguishable by use of
Eu(tfc)3 (see the Supporting Information), but the absolute
configuration of the newly formed quaternary stereogenic
center could not be assigned at this time.
Both yield and enantiomeric excess with 7a responded to
solvent variation, as the less polar solvents led to product in
both higher yield and greater ee. The alcohol solvents
performed poorly by any criteria, although the role of
competitive solvent sulfonylation in diverting the reaction
was not assessed. Switching to the non-hydroxylic and less
polar solvents C2H2Cl4 and CH2Cl2 provided the first glimpse
of success by delivering product in modest ee. By switching
to the slightly less polar but more viscous toluene as solvent,
the maximum ee (55%) was observed. Temperatures lower
than -78 °C did not lead to productive reaction for these
substrates.
Examination of 7b (R ) OCH3) was undertaken with the
expectation that the electronic character of R would have
little impact on a Pummerer reaction proceeding through an
additive Sn2′-like path but might exert some influence on
the reaction through the vinylogous Sn1-like path as a
consequence of a Hammond postulate-type argument. Greater
incursion of the vinylogous path from 7b might promote
reaction through achiral 6, leading to a drop in ee’s compared
Figure 1. Chiral indole-2-sulfoxide substrates for Pummerer-
mediated oxidative cyclization.
key transformation in each of these substrate syntheses
involved the sulfinylation of an indole C(2) anion with the
chiral sulfoxide transfer reagent 10 introduced by Evans,8
following the protocols described in the pioneering studies
of Marino.4 The enantiomeric excesses (>98%, detection
limit) of the chiral sulfoxides 7a/b were verified by chiral
shift reagent studies with Eu(tfc)3, whereas the enantiomeric
(4) (a) Marino, J. P.; Perez, A. D. J. Am. Chem. Soc. 1984, 106, 7643-
7644. (b) Marino, J. P.; Bogdan, S.; Kimura, K. J. Am. Chem. Soc. 1992,
114, 5566-5572.
(5) Kita, Y.; Shibata, N.; Fukui, S.; Masahiko, B.; Fujita, S. J. Chem.
Soc., Perkin Trans. 1 1997, 1763-1767.
(6) (a) Garc´ıa Ruano, J. L.; Alema´n, J.; Aranda, M. T.; Are´valo, M. J.;
Padwa, A. Org. Lett. 2005, 7, 19-22. (b) Garc´ıa Ruano, J. L.; Alema´n, J.;
Aranda, M. T.; Are´valo, M. J.; Padwa, A. Phosphorus, Sulfur, Silicon 2005,
180, 1497-1498. (c) Garc´ıa Ruano, J. L.; Alema´n, J.; Padwa, A. Org. Lett.
2004, 6, 1757-1760.
(7) Feldman, K. S. Tetrahedron 2006, 62, 5003-5034.
(8) Evans, D. A.; Faul, M. M.; Colombo, L.; Bisaha, J. J.; Clardy, J.;
Cherry, D. J. Am. Chem. Soc. 1992, 114, 5977-5985.
4138
Org. Lett., Vol. 8, No. 18, 2006

kcal/mol), but the ee’s do, with the greater ee attending
reaction in the more polar solvent. For the R ) OCH3 case
8b, both yields and ee’s slightly increase in response to
increasing solvent polarity, and again, the ee is maximized
in CH3CN.
Table 2. Pummerer-Initiated Cyclization of the
Allylsilane-Bearing Chiral (S)- Indole-2-sulfoxide 7b
The final substrates chosen for study, the silyl enol ethers
9a and 9b (Table 4), represent a marked change in reactivity
Table 4. Pummerer-Initiated Cyclization of the
Allylsilane-Bearing Chiral (S)-Indole-2-sulfoxides 9a and 9b
yield (%)
entry
solvent
T (°C)
14
15
16
ee (%)
1
2
3
CH2Cl2
Et2O
toluene
-78
-78
-78
47
47
42
50
50
82
44
50
60
10
28
56
entry
solvent
CH2Cl2
T (°C)
yield 17a/b (%)
eea (%)
1
2
3
4
5
6
7
8
9a
9a
9a
9a
9b
9b
9b
9b
-78
-78
-78
-110
-78
-78
33
33
60
58
80
76
80
33
43
67
51
86
13
38
43
74
with the less cation-stabilizing case 7a. The data (Table 2)
do sustain this line of thinking, with lower ee’s attending
reaction of 7b compared with 7a in both CH2Cl2 and Et2O,
although no effect was seen in toluene. In practice, better
overall yields of the N-methyl oxindole 16 (and 13, Table
1) result from the use of crude reaction products in each
step.
toluene
Et2O
Et2O
CH2Cl2
toluene
Et2O
-78
-110
Et2O
The N-methylallylsilane substrates 8a and 8b represent
a stringent test for thionium ion formation, as the absence
of a labile N-H proton guarantees that any material
passing through a vinylogous Pummerer pathway must access
a presumably high-energy dicationic thionium ion intermedi-
ate dN+(CH3)-CdS+Ph.3 These substrates delivered the
oxindole products 13 and 16, respectively, directly through
in situ hydrolysis of a putative C(3)-cyclized thionium ion
intermediate, Table 3. For the R ) H case 8a, the yields
do not vary as per solvent polarity (CH3CN ET(30) ) 45.6
a Average of two or three independent trials.
of the pendant nucleophile (Mayr N values: allylsilane ∼
1.8; silyl enol ether ∼ 5.4).9 For these compounds, enantio-
meric excesses were measured on the oxindoles 17a/b that
were prepared by hydrolysis of the first-formed thioimidates.
For the R ) H substrate 9a, reasonable ee’s were obtained
in all solvents tested. However, it was gratifying to observe
that by the simple expedient of dropping the temperature to
-110 °C, the transformation proceeded smoothly to furnish
thioimidate and then oxindole product both in good yield
and in relatively high ee (1H NMR, (S)-2,2,2-trifluorom-
ethylanthryl ethanol; see the Supporting Information). The
R ) OCH3 substrate 9b proceeded with lower ee under
comparable conditions (-78 °C). At the low-temperature
limit of -110 °C, however, even this compound delivered
Pummerer product with significantly improved ee.
Table 3. Pummerer-Initiated Cyclization of the
Allylsilane-Bearing Chiral (S)-N-methylindole-2-sulfoxides 8a
and 8b
A clear benefit can be ascribed to the silyl enol ether
nucleophile terminator compared to the allylsilane analogue
in terms of both chemical reactivity (compare with Table 1,
entries 7 and 8) and enantiomeric excess obtained. The lack
of reactivity of allylsilane indole-2-sulfoxide 7a at -110 °C
is curious, as it is not clear why the remote nucleophile
should have any influence on the ability of the sulfoxide’s
oxygen to react with Tf2O. Perhaps this unexpected observa-
tion can be rationalized if either (1) the initial sulfonylation
step (1 f 2) is reversible, or (2) the triflate is hydrolyzed
entry
solvent
T (°C)
yield 13/16 (%)
eea (%)
1
2
3
4
5
8a
8a
8b
8b
8b
CH2Cl2
CH3CN
toluene
Et2O
-78
-40
-78
-78
-40
63
63
41
45
59
7
46
20
28
32
CH3CN
a Average of two or three independent trials.
(9) Mayr, H.; Kempf, B.; Ofial, A. R. Acc. Chem. Res. 2003, 36, 66-
77.
Org. Lett., Vol. 8, No. 18, 2006
4139

from unreactive 2 upon workup to reform 1. Under these
speculative scenarios, further reaction by any pathway from
2 then depends on the ability of the pendant nucleophile (silyl
enol ether . allylsilane) to satisfy the burgeoning electron
demand by participation.
The absolute stereochemistries at C(3) of the oxindole
products 17a/b, 13, and 15 were determined by a combina-
tion of derivatization and X-ray crystallography, Scheme 2.
With the N-H allylsilane series 7a/7b, the ee vs solvent
polarity trends are just the opposite of those seen with the
N-CH3 analogues 8a/8b, an observation suggestive of
different mechanisms operating with the two types of
substrates. By process of elimination, the spotlight then
focuses on a vinylogous pathway/tight ion pair (cf. 2 f 5)
for 7a/7b as a means to generate thioimidate product with
high ee. This speculation is supported by the observation
that with solvents of nearly identical polarity (Et2O and
toluene), the more viscous of the two (toluene), which
presumably differentially retards dissociation of chiral 5 into
achiral 6, provides product with higher ee. In addition, the
observed solvent polarity trends fall within the scope of this
explanation to the extent that the more polar solvents should
accelerate the conversion of 5 into 6 and thus provide product
with lower ee. None of these data, however, rule out the
additive Sn2′ process for 7a/7b.
Scheme 2. Determination of Absolute Stereochemistry at C(3)
in the Oxindole Products
The yield and ee trends with the enol ether substrates 9a/
9b do not lend themselves to ready interpretation. The fact
that the “nucleophilicity” of the nucleophile seems to matter
(yields and ee’s of 17a/17b are higher than those of the
allylsilane analogues 11/14 under identical conditions) argues
for a role in the rate determining step for this unit, a picture
consistent with the Sn2′-like additive pathway. On the other
hand, the greater yields but lower ee’s of the R ) OCH3
species 9b compared to the R ) H analogue 9a lend weight
to a mechanistic argument relying on the vinylogous pathway
through a thionium ion (5 or 6). Thus, the mechanistic path-
(s) taken by the enol ether substrates cannot be disentangled
at this point.
In summary, the formation of 3,3-spirocyclic oxindole
products in modest-to-moderate enantiomeric excess by
Pummerer-mediated oxidative cyclization of chiral indole-
2-sulfoxides has been observed. Evidence in support of two
distinct mechanistic paths, depending upon the substrate’s
structural details, was garnered. Efforts to incorporate these
Pummerer methodology advances in natural product syn-
thesis are ongoing, and results will be reported in due course.
The absolute configuation at C(3) for both major enantiomers
formed from the Pummerer sequence is (S) as shown. The
absolute configurations of the oxindoles derived from the
allylsilane terminator systems 7a, 7b, 8a, and 8b were
established by chemical correlation, 17a f 12 f 13, and
17b f 15. In all cases, the (S) absolute configuration at C(3)
was found.
Revisiting the mechanistic picture depicted in Scheme 1
in light of the results described above permits some refine-
ment, but ultimately does not allow any definitive conclu-
sions to be drawn. At the very least, these results provide
indisputable evidence that these Pummerer reactions cannot
be proceeding exclusively or even predominantly through
the commonly cited but achiral free thionium ion intermedi-
ate (cf. 6). The most compelling evidence in support of the
additive Sn2′-like pathway can be found in the results of
the N-CH3 allylsilane series 8a/8b, where (1) the yield is
insensitive to R (R ) H or R ) OCH3), (2) the ee’s are
marginally higher when R ) H compared with R ) OCH3,
and (3) the ee’s increase (again, marginally) as the solvent
polarity increases.
Acknowledgment. Financial support from the National
Science Foundation (CHE 041877) is gratefully acknowl-
edged. The X-ray crystallographic analyses were conducted
by Dr. H. Yennawar under the auspices of NSF CHE
0131112.
Supporting Information Available: Experimental pro-
cedures for 7a/b-9a/b, 13-17a/b, and 19a/b; copies of 1H
and 13C NMR spectra for 7b, 8b, 9b, 14-17b, and 19a/b;
X-ray data for 19a and 19b; ee determinations for 13, 16,
17a, and 17b. This material is available free of charge via
the Internet at http://pubs.acs.org.
OL0617244
4140
Org. Lett., Vol. 8, No. 18, 2006