Competitive [2,3]-and [1,2]-oxonium ylide rearrangements. Concerted or stepwise?

Article abstract of DOI:10.1021/ol300213u

The axial-equatorial conformational isomer distribution of the reactant diazoacetoacetate or its metal carbene intermediate is reflected in Rh(II) catalyzed oxonium ylide forming reactions of 3-(trans-2-arylvinyl) tetrahydropyranone-5-diazoacetoacetates that afford diastereoisomeric products for both the symmetry-allowed [2,3]-and the formally symmetry-forbidden [1,2]-oxonium ylide rearrangements.

Full text of DOI:10.1021/ol300213u

ORGANIC
LETTERS
2012
Vol. 14, No. 7
1676–1679
Competitive [2,3]- and [1,2]-Oxonium Ylide
Rearrangements. Concerted or Stepwise?
Deana M. Jaber,‡ Ryan N. Burgin,† Matthew Helper,† Peter Y. Zavalij,† and Michael P. Doyle*,†
Department of Chemistry and Biochemistry, University of Maryland, College Park,
Maryland 20742, United States, and Department of Biology and Physical Sciences,
Marymount University, Arlington, Virginia 22207, United States
mdoyle3@umd.edu
Received January 26, 2012
ABSTRACT
The axialꢀequatorial conformational isomer distribution of the reactant diazoacetoacetate or its metal carbene intermediate is reflected in Rh(II)
catalyzed oxonium ylide forming reactions of 3-(trans-2-arylvinyl)tetrahydropyranone-5-diazoacetoacetates that afford diastereoisomeric
products for both the symmetry-allowed [2,3]- and the formally symmetry-forbidden [1,2]-oxonium ylide rearrangements.
We have reported that dirhodium(II) catalyzed reac-
tions of racemic trans-aryl-substituted tetrahydropyra-
none diazoacetoacetates 1 produce two diastereoisomeric
products (syn-4 and anti-4) that are proposed to arise from
two noninterconvertable oxonium ylide intermediates (3A
and 3B) via what appeared to be a concerted [1,2]-rearrange-
ment (Scheme1).1 Product ratios were independent of ligand
size or electron-withdrawing influence, and although
para-substituents on the aromatic ring affected the yields of
the ylide rearrangement products, they had limited effect on
their ratios (stereoselectivity). However, increasing the steric
bulk of the aryl group (ortho substituents) led exclusively to
the formation of one diastereoisomer and suggested that
conformational isomers in the reactant and the intermediate
metal carbene are responsible for the formation of diaste-
reoisomeric products.
The mechanism of the [1,2]-rearrangement of oxonium
ylides has been interpreted as occurring via either a cleav-
age/recombination (either homolytic or heterolytic) or a
concerted pathway.2 Because this reaction is formally
considered to be symmetry forbidden,3 the rearrangement
is regarded to occur by cleavage/recombination,4 although
other mechanistic possibilities exist.5ꢀ7 However, con-
vincing evidence regarding the cleavage/recombination
(4) (a) Stewart, C.; McDonald, R.; West, F. G. Org. Lett. 2011, 13,
720. (b) Murphy, G. K.; Marmsaeter, F. P.; West, F. G. Can. J. Chem.
2006, 84, 1470. (c) Murphy, G. K.; West, F. G. Org. Lett. 2005, 7, 1801.
(d) Marmsaeter, F. P.; West, F. G. J. Am. Chem. Soc. 2001, 123, 5144. (e)
Feldman, K. S.; Wrobleski, M. L. J. Org. Chem. 2000, 65, 8659. (f)
Doyle, M. P.; Ene, D. G.; Forbes, D. C.; Tedrow, J. S. Tetrahedron
Lett. 1997, 38, 4367. (g) Doyle, M. P.; Griffin, J. H.; Chinn, M. S.;
Van Leusen, D. J. Org. Chem. 1984, 49, 1917.
(5) A concerted mechanism, including those with metal association:
(a) Lu, C.-D.; Liu, H.; Chen, Z.-Y.; Hu, W.-H.; Mi, A.-Q. Org. Lett.
2005, 7, 83. (b) Qu, Z.; Shi, W.; Wang, J. J. Org. Chem. 2004, 69, 217. (c)
Kitagaki, S.; Yanamoto, Y.; Tsutsui, H.; Anada, M.; Nakajima, M.;
Hashimoto, S. Tetrahedron Lett. 2001, 42, 6361. (d) Roskamp, E. J.;
Johnson, C. R. J. Am. Chem. Soc. 1986, 108, 6062.
† University of Maryland.
‡ Marymount University.
(1) Jaber, D. M.; Burgin, R. N.; Hepler, M.; Zavalij, P.; Doyle, M. P.
Chem. Commun. 2011, 47, 7623.
(2) (a) Wang, J. In Comprehensive Organometallic Chemistry;
Crabtree, R. H., Mingos, D. M., Eds.; Elsevier: Oxford, 2007; Vol. 11,
pp 151ꢀ178. (b) West, F. G. In Modern Rhodium-Catalyzed Organic
Reactions; Evans, P. A., Ed.; Wiley-VCH: New York, 2005; Chapter 18.
(c) Hodgson, D. M.; Pierard, F. Y. T. M.; Stupple, P. A. Chem. Soc. Rev. 2001,
30, 50. (d) Doyle, M. P.; Mckervey, M. A.; Ye, T. Modern Catalytic
Methods for Organic Synthesis with Diazo Compounds; Wiley: New York,
1998. (e) Doyle, M. P.; Forbes, D. C. Chem. Rev. 1998, 98, 911.
(3) Woodward, R. B.; Hoffmann, R. Angew. Chem., Int. Ed. Engl.
1969, 8, 781.
(6) Through an oxonium metal enolate: (a) Ji, J.; Zhang, X.; Zhu, Y.;
Qian, Y.; Zhou, J.; Yang, L.; Hu, W. J. Org. Chem. 2011, 76, 5821. (b)
Sawada, Y.; Mori, T.; Oku, A. J. Org. Chem. 2003, 68, 10040.
r
10.1021/ol300213u
Published on Web 03/12/2012
2012 American Chemical Society

that treatment of 5 with dinitrogen extrusion catalysts
would form oxonium ylide(s), with the resultant reaction-
(s) leading to either or both products from the [2,3]-
sigmatropic or the [1,2]-Stevens rearrangement, expecting
the former to dominate.12 Compounds 5 were prepared
Scheme 1. Conformation-Dependent Pathway for the Forma-
tion of Two Diastereoisomers via [1,2]-Oxonium Ylide Rear-
rangements
solely as the trans-diastereoisomers by BF3 Et2O-mediated
3
hetero-DielsꢀAlder reactions between the cinnamalde-
hyde and Danishefsky’s diene13 followed by zinc tri-
flate catalyzed MukaiyamaꢀMichael reaction with methyl
3-(tert-butyldimethylsilanoxy)-2-diazo-3-butenoate (eq 1).14
Rhodium(II) catalyzed decomposition of 5a, using
1.0 mol % of dirhodium octanoate [Rh2(oct)4] in refluxing
dichloromethane, afforded a mixture of [2,3]-sigmatropic
6a and [1,2]-Stevens 7a rearrangement products (eq 2) in
47% and 41% yield, respectively (54:46 ratio). After care-
ful chromatographic separations and spectroscopic ana-
lyses of the reaction mixture, two diastereoisomers for
both rearrangement products were isolated and identified.
The crystal structure of the chromatographically separated
major isomer of the [1,2]-Stevens rearrangement product
was obtained (Figure 1b), and like compound 4 (Scheme 1)
the major isomer was the syn-isomer. The minor isomer of
the [1,2]-Stevens rearrangement product and the major
isomer of the [2,3]-sigmatropic rearrangement product
were isolated together and characterized spectroscopically,
while the minor isomer of the [2,3]-rearrangement sigma-
tropic product (syn-isomer) was isolated in a separate
chromatographic fraction, and its crystal structure was
also obtained (Figure 1a). Table 1 reports the negligible
variance in the ratio of products from [2,3]- and [1,2]-
oxonium ylide rearrangements and
pathway is scarce. The most relevant evidence for a
homolytic pathway in oxonium ylide reactions comes from
detection of a “homo-dimer” in the rhodium acetate
catalyzed reactions of 4-benzyloxy-1-diazo-2-butanones,8
but similar ylide-producing reactions of the next higher
homologue, 5-benzyloxy-1-diazo-2-pentanone, were not
reported.9 More elaborate evidence for radical pair and/
or zwitterionic intermediates comes from ammonium
ylide [1,2]-rearrangements2,10 where, recently, substituent
effects on stereoselectivity (from 4:1 to 20:1 d.r.) were
obtained from aryl substituents of aryldiazoacetates in
the rhodium acetate catalyzed reactions of configuration-
€
11
ally stable Troger bases. Our proposed mechanism for
oxonium ylide formation from 1 and subsequent [1,2]-
Stevens rearrangement (Scheme 1) that suggests a con-
certed reaction is in conflict with a homolytic/heterolytic
cleavage/recombination mechanism (i.e., absence of sub-
stituent effects, and steric influence on stereoselectivity),
although a very fast cleavage/recombination process that
is barely distinguishable from a concerted reaction is
also possible. We now report analogous investigations
with compounds that are structurally similar to 1 whose
oxonium ylides could undergo symmetry-allowed [2,3]-
sigmatropic rearrangements that are not subject to orbital
symmetry restrictions.
3-(trans-Styryl)tetrahydropyranone-5-diazoacetoacetate
and its para-substituted derivatives 5aꢀc were selected to
investigate the competition between the [2,3]-sigmatropic
and [1,2]-Stevens rearrangement pathways. We anticipated
(7) Via alkyl/aryl transfer from the metal associated oxonium ylide to
the metal, then reductive elimination: (a) West, F. G.; Naidu, B. N.;
Tester, R. W. J. Org. Chem. 1994, 59, 6892. (b) Karche, N. P.; Jachak,
S. M.; Dhavale, D. D. J. Org. Chem. 2001, 66, 6323.
(8) Eberlein, T. H.; West, F. G.; Tester, R. W. J. Org. Chem. 1992, 57,
3479.
in the ratios of diastereoisomers formed by each process
from rhodium(II) octanoate catalyzed reactions with
(9) West, F. G.; Naidu, B. N.; Tester, R. W. J. Org. Chem. 1994, 59,
6892.
(10) (a) Sweeney, J. B. Chem. Soc. Rev. 2009, 38, 1027. (b) Tuzina, P.;
Somfai, P. Org. Lett. 2009, 11, 919.
(11) Sharma, A.; Guenee, L.; Naubron, J.-V.; Lacour, J. Angew.
(12) Kitagaki, S.; Yanamoto, Y.; Tsutsui, H.; Anada, M.; Nakajima,
M.; Hashimoto, S. Tetrahedron Lett. 2001, 42, 6361.
(13) Danishefsky, S.; Larson, E.; Askin, D.; Kato, N. J. Am. Chem.
Soc. 1985, 107, 1246.
(14) Liu, Y.; Zhang, Y.; Jee, N.; Doyle, M. P. Org. Lett. 2008, 10,
1605.
Chem., Int. Ed. 2011, 50, 3677.
Org. Lett., Vol. 14, No. 7, 2012
1677

para-substituted aryl derivatives 5aꢀb. The lower yield
and the difference in the ratio of products from
the para-methoxy derivative 5c is due to the forma-
tion of elimination products that, although evident
in the reaction mixture, were not isolated and fully
characterized.
Table 2. Effect of Dirhodium(II) Ligands on Yields and Product
Ratios from Oxonium Ylide Reactions of 3-(trans-Styryl)tetra-
hydropyranone-5-diazoacetoacetate 5aa
anti-6a:
syn-6a
syn-7a:
anti-7a
yield, %
(6aþ7a)c
catalyst
(6a:7a)b
Rh2(oct)4
Rh2(OAc)4
Rh2(pfb)4
Rh2(piv)4
Rh2(tpa)4
Rh2(tfa)4
54:46
51:49
55:45
55:45
53:47
51:49
78:22
80:20
80:20
78:22
73:27
79:21
78:22
77:23
77:23
76:24
75:25
75:25
88
94
46
94
93
41
a Reactions were performed in refluxing dichloromethane for 2 h
using 1.0 mol % of catalyst. Results reported are averages of two or more
reactions. b Product ratio determined by 1H NMR spectral analysis with
a variance of (2%. c NMR yield using benzaldehyde as an internal
standard.
Observation of two diastereoisomeric products from
the [2,3]-sigmatropic rearrangement with nearly identi-
cal diastereoisomer ratios (anti-6:syn-6) as those for the
[1,2]-Stevens rearrangement (syn-7:anti-7) is inconsistent
with a cleavage/rearrangement/recombination pathway in
products from the [2,3]-sigmatropic rearrangementbut not
from the Stevens rearrangement. The fact that the dia-
stereomer ratios are nearly identical suggests that the two
processes originate from the same intermediate(s). The
absence of a substituent effect on the ratio of products
from [2,3]- and [1,2]-rearrangements (6:7) is unexpected if
both sets of products arise from a cleavage/rearrangement/
recombination pathway. Instead, the selectivity match
between the two pathways is more likely due to the
conformational equilibrium in the reactants (Scheme 1)
that results in two noninterconvertable oxonium ylide
intermediates and a subsequent competition between the
symmetry-allowed [2,3]-sigmatropic rearrangement and
the formally symmetry-forbidden [1,2]-Stevens rearrange-
ment (Scheme 2).
Figure 1. (a) Crystal structure (left) of syn-6a (minor isomer). (b)
Crystal structure (right) of syn-7a (major isomer).
Table 1. Effect of Ring Substitution on Product Formation in
the Rhodium(II) Octanoate Catalyzed Reactions of 3-(trans-
Styryl) Tetrahydropyranone-5-diazoacetoacetates 5aꢀca
yield,
anti-6:
syn-6
syn-7:
anti-7
%
Ar
C6H5
6:7b
(6þ7)c
5a
5b
5c
54:46
52:48
69:31
78:22
76:24
79:21
78:22
76:24
70:30
88
85
60
p-NO2C6H4
p-MeOC6H4
a Reactions were performed in refluxing CH2Cl2 for 2 h using 1.0 mol %
of Rh2(oct)4. Results reported are averages of two or more reactions
(2%. b Product ratio determined by 1H NMR spectral analysis with
a variance of (2%. c NMR yield using benzaldehyde as an internal
standard.
We also investigated ligand effects from dirhodium cata-
lysts on both the [2,3]-/[1,2]-reaction pathways (6a:7a) and
product diastereoisomer ratios from the [2,3]- and [1,2]-
rearrangements to determine if the oxonium ylide inter-
mediate was metal-associated or a free ylide. Results were
obtained from reactions of 5a with a wide spectrum of
catalysts (Table 2), and they show minimal dependence on
catalyst ligands. No reaction occurred with rhodium ca-
prolactamate under the same conditions. The diastereoi-
somer product ratio from the [2,3]-sigmatropic rearrange-
ment (anti-6a:syn-6a), as well as that from the [1,2]-Stevens
rearrangement (syn-7a:anti-7a), was invariant with com-
mon ligands on dirhodium (piv = pivalate, pfb = per-
fluorobutyrate, tfa = trifluoroacetate, tpa = triphenyl-
acetate) that cover a broad range of electronic influences,
although results with Rh2(tpa)4 for (anti-6a:syn-6a) sug-
gest a minor steric effect on selectivity. These composite
dataare consistent with anoxonium ylide intermediatefree
from association with the ligated metal catalyst during
product formation.
Scheme 2. Conformation-Dependent Pathway for the Forma-
tion of Two Diastereoisomers via [2,3]- and [1,2]-Oxonium Ylide
Rearrangements
Competition between [2,3]- and [1,2]-oxonium ylide
rearrangements is relatively common, especially with
1678
Org. Lett., Vol. 14, No. 7, 2012

diazoketones.15,16 This competition is generally thought to
be due to steric influences in structurally rigid systems.
Diastereoisomeric products from the [1,2]-Stevens rear-
rangement have been reported, and their formation has
been attributed to a cleavage/rearrangement/recombina-
tion pathway. However, the formation of two diastereoi-
somers for the [2,3]-sigmatropic rearrangement pathway is
virtually unprecedented. The exception is a report from
Clark and co-workers in a single example of structurally
well-defined diastereoisomeric products for both the [1,2]-
Stevens and the [2,3]-sigmatropic rearrangement path-
ways,16 but the origin of these diastereoisomeric products
was not explained; the explanation that we have advanced
adequately accounts for their “fascinating” results.
Based on our previousdemonstration thatincreasing the
steric bulk of the aryl group of trans-aryl-substituted
tetrahydropyranone diazoacetoacetates 1 results in the
production of only one diastereoisomer in the [1,2]-Stevens
rearrangement, we constructed the di- and triphenyl ana-
loguesof 5ainordertoevaluatediastereoisomer formation
from the [2,3]-sigmatropic and [1,2]-Stevens rearrange-
ment pathways, as well the influence of size on the
[2,3]-/[1,2]-rearrangement ratio of products. The syntheses
of diazoacetoacetates 9 and 10 were easily accessible by
employing the previously described methods.
In addition to the [1,2]-Stevens rearrangement products,
2-methyl-6-vinylpyran-4-ones 13 and 14 were observed as
byproducts. Although these byproducts were not formed
in high yields, they are ubiquitous in catalytic reactions
of tetrahydropyranone- and tetrahydrofuranone-diazo-
acetoacetates.17
In conclusion, we have discovered that both the
[2,3]- and [1,2]-oxonium ylide rearrangement reactions of
3-(trans-styryl)tetrahydropyranone-5-diazoacetoacetate and
its para-substituted derivatives 5aꢀc form diastereoisomeric
products in the same ratios and that their diastereoisomer
ratios are independent of electronic influences from para
substituents or of catalyst ligands. These results are consistent
with the involvement of catalyst-free oxonium ylides whose
composition reflects the axialꢀequatorial conformational
isomer distribution of the reactant diazoacetoacetate or its
metal carbene intermediate. Furthermore, increasing the size
of the styryl group of the 3-(trans-styryl)tetrahydropyranone-
5-diazo-acetoacetate with phenyl substituents 9 and 10 re-
moves competition with the [2,3]-oxonium ylide rearrange-
ment, allowing the [1,2]-oxonium ylide rearrangement to
occur with the production of two diastereoisomers in ratios
that are identical with that from 5aꢀc.
Acknowledgment. Support for this research from the
National Institutes of Health (GM 46503) and the
National Science Foundation (CHE-0748121) is gratefully
acknowledged.
Rhodium(II) catalyzed reactions of 9 and 10 afforded
[1,2]-Stevens rearrangement products 11 and 12, respec-
tively, in 73% and 80% yield (eq 3). The [2,3]-sigmatropic
1
rearrangement products were not detected by either H
NMR spectra of the reaction mixtures or by HPLC
analyses. However, unlike the previously established re-
sults with diortho-substituted aryl derivatives of 1, two
diastereoisomers for each of the [1,2]-Stevens rearrange-
ment products were formed in amounts that corresponded
to those found with their monophenyl counterpart 5a. The
(syn/anti) ratios of 11 and 12 are consistent with those
ratios of 7aꢀc.
Supporting Information Available. General experimen-
tal procedures, X-ray structures of syn-6a and syn-7a, and
spectroscopic data for all new compounds. This material
is available free of charge via the Internet at http://pubs.
acs.org.
(15) (a) Marmsater Fredrik, P.; Vanecko John, A.; West, F. G. Org.
Lett. 2004, 6, 1657. (b) Clark, J. S.; Bate, A. L.; Grinter, T. Chem.
Commun. 2001, 459. (c) Brogan, J. B.; Zercher, C. K. Tetrahedron Lett.
1998, 39, 1691. (d) Brogan, J. B.; Zercher, C. K.; Bauer, C. B.; Rogers,
R. D. J. Org. Chem. 1997, 62, 3902.
(16) Clark, J. S.; Walls, S. B.; Wilson, C.; East, S. P.; Drysdale, M. J.
Eur. J. Org. Chem. 2006, 323.
(17) Doyle, M. P.; Dyatkin, A. B.; Autry, C. L. J. Chem. Soc., Perkin
Trans. 1 1995, 619.
The authors declare no competing financial interest.
Org. Lett., Vol. 14, No. 7, 2012
1679