Synthesis of an n-mesityl substituted chiral imidazolium salt for NHC-catalyzed reactions

Article abstract of DOI:10.1021/ol800006m

A new synthetic approach to chiral imidazolium salts makes possible the first synthesis of an N-mesityl substituted, aminoindanol-derived N-heterocyclic carbene precursor, 1-CIO₄. The successful synthesis allows the first direct comparison of o

Full text of DOI:10.1021/ol800006m

ORGANIC
LETTERS
2008
Vol. 10, No. 5
957-960
Synthesis of an N-Mesityl Substituted
Chiral Imidazolium Salt for
NHC-Catalyzed Reactions
Justin R. Struble, Juthanat Kaeobamrung, and Jeffrey W. Bode*
Roy and Diana Vagelos Laboratories, Department of Chemistry,
UniVersity of PennsylVania, Philadelphia, PennsylVania 19104-6232
bode@sas.upenn.edu
Received January 2, 2008
ABSTRACT
A new synthetic approach to chiral imidazolium salts makes possible the first synthesis of an N-mesityl substituted, aminoindanol-derived
N-heterocyclic carbene precursor, 1 ClO4. The successful synthesis allows the first direct comparison of otherwise identical imidazolium and
‚
triazolium precursors across a number of NHC-catalyzed processes. These studies confirm striking differences in reactivity and mechanism
between the two classes.
N-Heterocyclic carbenes derived from azolium salts are
remarkable and versatile catalysts for a wide range of organic
transformations.1 The classical benzoin and Stetter reactions
of simple aldehydes are catalyzed by thiazolium and tri-
azolium-derived carbenes.2 More recently, an extensive array
of new processes operating via reactive intermediates
catalytically generated by internal redox reactions of R-func-
tionalized aldehydes have been discovered.3 Successful
developments from our lab include enantioselective N-mesityl
substituted triazolium-catalyzed annulations affording cis-
disubstituted dihydropyranones and dihydropyridinones,4a,b
cyclopentenes,4c and bicyclic â-lactams,4d all with exceptional
levels of diastereo- and enantioselectivity. We and Glorius
also documented diastereoselective annulations of enals and
aldehydes to afford γ-butyrolactones using imidazolium
catalysts.5
The starting materials for many of these annulation
processes are similar or identical, but the reaction outcomes
differ dramatically by the choice of imidazolium vs triazo-
lium precatalysts. These disparate results have led to
considerable confusion over the difference between the
imidazolium and triazolilum precatalysts. Direct comparisons
between the reactivity and mechanistic pathways of these
two classes have been precluded by the use of structurally
different catalyst classes in each case.
(1) For a recent, comprehensive review, see: Enders, D.; Niemeier, O.;
Henseler, A. Chem. ReV. 2007, 107, 5606-5655.
(2) For recent reviews, see: (a) Johnson, J. S. Angew. Chem., Int. Ed.
2003, 42, 2534-2536. (b) Christmann, M. Angew. Chem., Int. Ed. 2005,
44, 2632-2634.
(4) (a) He, M.; Uc, G. J.; Bode, J. W. J. Am. Chem. Soc. 2006, 128,
15088-15089. (b) He, M.; Struble, J. R.; Bode, J. W. J. Am. Chem. Soc.
2006, 128, 8148-8150. (c) Chiang, P.-C.; Kaeobamrung, J.; Bode, J. W.
J. Am. Chem. Soc. 2007, 129, 3520-3521. (d) He, M.; Bode, J. W. J. Am.
Chem. Soc. 2008, 130, 418-419.
(3) For the first reports of this concept, see: (a) Chow, K. Y.-C.; Bode,
J. W. J. Am. Chem. Soc. 2004, 126, 8126-8127. (b) Reynolds, N. T.; Read
de Alaniz, J.; Rovis, T. J. Am. Chem. Soc. 2004, 126, 9518-9519. (c) For
a highlight, see: Zeitler, K. Angew. Chem., Int. Ed. 2005, 44, 7506-7510.
(5) (a) Sohn, S. S.; Rosen, E. L.; Bode, J. W. J. Am. Chem. Soc. 2004,
126, 14370-14371. (b) Burstein, C.; Glorius, F. Angew. Chem., Int. Ed.
2004, 43, 6205-6208. (c) Burstein, C.; Tschan, S.; Xie, X.; Glorius, F.
Synthesis 2006, 2418-2439.
10.1021/ol800006m CCC: $40.75
© 2008 American Chemical Society
Published on Web 02/12/2008

of unique reactive intermediates by internal redox reactions
promoted by the association of an N-heterocyclic carbene
and the enal (Scheme 1). Depending on the structure of the
catalyst and the facility of proton-transfer events, the initially
formed Breslow intermediate i can ultimately serve as acyl
anion equivalent ii, homoenolate equivalent iii, or enolate
iv. In the latter two cases, the aldehyde is simultaneously
oxidized to the carboxylic acid oxidation state. As this
communication demonstrates, the fate of the Breslow inter-
mediate is intimately controlled not by the steric demands
of the catalyst but rather by the choice of imidazolium vs
triazolium-derived NHCs.
In the present communication, we address these important
issues through the synthesis of chiral imidazolium salt 1,
which corresponds to the N-mesityl aminoindanol-derived
triazolium salt 2 that has served as the most reactive and
selective precatalyst in numerous annulation processes. The
successful synthesis of 1 allows the first direct comparison
between these two catalyst classes, differing only by a single
atom at a site removed from the reactive center. The distinct
reactivity of these two classes provides valuable insight into
the mechanistic differences and the challenges associated
with developing enantioselective variants.
At the outset of our studies, we sought to address two
recurring issues in our development of novel, NHC-catalyzed
reactions. First, numerous annulation processes were pro-
moted exclusively by one catalyst type (imidazolium or tri-
azolium) but not the other. Even in the few instances where
a given transformation is promoted by either catalyst class,
such as cyclopentene-forming annulations, we and Nair have
demonstrated that different mechanisms may be operating.4c,6
Second, although highly enantioselective triazolium-catalyzed
processes have been achieved for diverse reaction types, no
examples of highly selective imidazolium-catalyzed processes
have been documented. For example, despite extensive
efforts from our group and others, no highly enantioselective
NHC-catalyzed γ-lactam and γ-lactone formations have been
achieved.
Scheme 1. Reactive Intermediates Catalytically Generated by
NHC-Promoted Redox Reactions of R,â-Unsaturated Aldehydes
A key feature of the chiral triazolium salts designed by
Leeper,7 Enders,8 and Rovis9 is their unsymmetrical, poly-
cyclic structure, which has proven exceptionally successful
in the levels of selectivity provided. They are also readily
prepared by a convenient, modular route from chiral amino
alcohols. Unfortunately, the synthesis of similar chiral imida-
zolium salts has proven challenging. With a few notable ex-
ceptions,10 the chiral imidazolium-derived carbenes reported
to date are constructed from either a chiral, C-2 symmetric
diamine backbone or with the chiral centers resident on the
N-substitutent.11 The requirement, in many of the annulation
reactions, of an ortho,ortho′-disubsituted aromatic on the
N-substituent further increases the synthetic challenge.
The basis of the novel annulation reactions of R,â-unsatu-
rated aldehydes and electrophiles is the catalytic generation
Scheme 2. Synthesis of Aminoindanol-Derived Imidazolium Salt 1‚ClO4
958
Org. Lett., Vol. 10, No. 5, 2008

Our initial attempts to prepare chiral, bicyclic imidazolium
salts bearing the essential N-mesityl substituent were unsuc-
cessful using established methods such as those reported by
Arduengo,12 Kuhn,13 and Waymouth and Hendrick.14 We
were intrigued, however, by a recent, elegant approach to
unsymmetrical imidazolium salts disclosed by Fu¨rstner.15
This approach allowed for a redesign of our synthetic
strategy, by reducing the synthetic challenge to the prepara-
tion of N-formyl bicyclic ketone 8. This precursor, in turn,
was prepared from cis-aminoindanol 3 by alkylation, formy-
lation, epoxidation, intramolecular epoxide opening,16 and
oxidation as shown in Scheme 2. With 8 in hand, a slight
modification of Fu¨rstner’s protocol allowed the introduction
of the N-mesityl substituent and formation of the imidazolium
ring. This approach proved to be robust and reliable and
allowed for the preparation of a range of N-substituents.17
With chiral imidazolium precatalyst 1 in hand, we
undertook a systematic comparison of imidazolium vs
triazolium precatalysts in reactions known to be catalyzed
by N-heterocyclic carbenes. Our previous studies employed
triazolium salt 2‚Cl as the precatalyst, but our corresponding
imidazolilum salt was most easily prepared and handled as
the perchlorate salt (1‚ClO4). To avoid any concern over
counterion effects, we prepared and employed triazolium
2‚ClO4 for comparison. A brief survey showed that this salt
had identical reactivity as the chloride variant.
Scheme 3. Reactions Catalyzed Preferentialy by Triazolium
a
NHC-Precatalyst 2‚ClO4
The results of the catalyst comparison confirmed, for the
first time, that profound reactivity differences exist between
the two classes of NHC-precatalysts. Of nine discrete reaction
types screened, six processes were promoted almost exclu-
sively by the triazolium precatalyst (2‚ClO4) (Scheme 3).
Another three processes, all of which invoke catalytically
generated homoenolates as the putative reactive intermedi-
ates, were preferentially promoted by the imidazolium salt
1‚ClO4 (Scheme 4). In all cases, the reaction conditions
screened were identical, save for the catalyst structure, and
no attempts were made to optimize the reacton conditions
for improved yields or stereoselectivities. As we anticiapted
a All reactions run under identical conditions with the two
different precatalysts. No attempts were made to optimize the
reaction conditions. bUnreacted enal was detected by 1H NMR
analysis of the unpurified reaction mixture. c% ee not determined.
(6) Nair, V.; Vellalath, S.; Poonoth, M.; Suresh, E. J. Am. Chem. Soc.
2006, 128, 8736-8737.
(7) Knight, R. L.; Leeper, F. J. J. Chem. Soc., Perkins 1 1998, 1891-
1893.
(8) (a) Enders, D.; Breuer, K.; Runsink, J.; Teles, J. H. HelV. Chim. Acta.
1996, 79, 1899-1902. (b) Enders, D.; Kallfass, U. Angew. Chem., Int. Ed.
2002, 41, 1743-1745.
(9) (a) Kerr, M. S.; Read de Alaniz, J.; Rovis, T. J. Am. Chem. Soc.
2002, 124, 10298-10299. (b) Kerr, M. S.; Read de Alaniz, J.; Rovis, T. J.
Org. Chem. 2005, 70, 5725-5728.
from our results using achiral imidazolium salts,18 precatalyst
1‚ClO4 was ineffective for NHC-catalyzed inverse electron-
demand Diels-Alder processes. This was true regardless of
how the key catalyst-bound enolate was generated, either
via redox reactions of electron-deficient enals (Scheme 3a)
or R-chloro aldehydes (Scheme 3b).19 They were also
unreactive in our recently disclosed annulation of enals and
electron-deficient enones (Scheme 3c) or N-sulfonyl imines
(10) Glorius, F.; Altenhoff, G.; Goddard, R.; Lehmann, C. Chem.
Commun. 2002, 2704-2705.
(11) For an excellent review of chiral N-heterocyclic carbenes, see: Perry,
M. C.; Burgess, K. Tetrahedron: Asymmetry 2003, 14, 951-961.
(12) Arduengo, A. J.; Krafczyk, R.; Schmutzler, R.; Craig, H. A.;
Goerlich, J. R.; Marshall, W. J.; Unverzagt, M. Tetrahedron 1999, 55,
14523-14534.
(13) Kuhn, N.; Kratz, T. Synthesis 1993, 561-562.
(14) Nyce, G. W.; Csihony, S.; Waymouth, R. M.; Hedrick, J. L. Chem.-
Eur. J. 2004, 10, 4073-4079.
(15) Fu¨rstner, A.; Alcarazo, M.; Ce´sar, V.; Lehmann, C. W. Chem.
Commun. 2006, 2176-2178.
(18) The most common achiral imidazolium salt employed for homoeno-
late based annulations is IMesCl:
(16) Potter, G. W. H.; Monro, A. M. J. Heterocycl. Chem. 1972, 299-
301.
(17) See the Supporting Information for complete experimental proce-
dures for the synthesis of 1‚ClO4. Further details on the development of
this route and its use to prepare a range of N-substituted derivatives is the
subject of forthcoming full account.
Org. Lett., Vol. 10, No. 5, 2008
959

these processes work well with this catalyst, only a single
report of an enantioselective variant has appeared in the work
of Glorius, who obtained a maximum of 25% ee for a related
annulation.5c Despite the high reactivity of triazolium-derived
carbenes for the processes shown in Scheme 3, 2‚ClO4 and
related triazolium salts are almost ineffective for the reactions
shown in Scheme 4. The difference in reactivity between
the annulations shown in Scheme 3d and Scheme 4b is
particularly striking.
Scheme 4. Reactions Catalyzed Preferentialy by Imidazolium
NHC-Precatalyst 1a
The cyclopentanone-forming annulation shown in Scheme
4c was recently reported by Nair and co-workers,21 who
believe that it proceeds via the intermediacy of a catalytically
generated homoenolate. In support of this hypothesis, tri-
azolioum 2, which is typically a poor catalyst for homoeno-
late-based annulations, is almost unreactive. In contrast,
imidazolium 1 effects the expected transformation in 85%
ee, the highest level of enantioselectivity yet reported for
any transformation catalyzed by a chiral imidazolium-derived
NHC.
In summary, we have reported the synthesis of chiral
N-mesityl substituted imidazolium salt 1. In addition to
providing a viable, potentially modular route to synthetically
challenging chiral imidazolium salts, its successful synthesis
allowed for the first time direct comparisons of otherwise
structurally identical chiral imidazolium and triazolium NHC-
precatalysts. The distinct reactivity profiles of these two
classes reveal pronounced and unexpected differences in
reactivity from a molecular replacement remote from the
catalyst active site. The origin of this discrepancy, and the
development of new catalytic reactions and enantioselective
variants are the subject of ongoing investigations.
a All reactions run under identical conditions with the two
different precatalysts. No attempts were made to optimize the
b
reaction conditions. Absolute configuration not determined. Un-
1
reacted enal was detected by H NMR analysis of the unpurified
c
reaction mixture. % ee not determined.
(Scheme 3d), processes that we believe proceed via a
benzoin-oxy-Cope reaction. Interestingly, triazolium pre-
catalyst 2, but not its imidazolium counterpart, was highly
effective for intramolecular Stetter (Scheme 3e) and inter-
molecular benzoin reactions (Scheme 3f). Although these
reactions are well-known to be catalysed by other triazolium-
derived NHCs, this was the first time the N-mesityl substi-
tuted variant had been employed. These results strongly
suggest that the lack of benzoin products observed when
N-mesityl substituted imidazolium-derived NHCs are em-
ployed stem from the properties of the imidazolium ring
rather than steric restrictions of the N-mesityl moeity.
In contrast, NHC-catalyzed annulation process that we and
others have postulated to occur via catalytically generated
homoenolate equivalents operate preferentially with imida-
zolium-derived precatalysts. The three reactions shown in
Scheme 2 have all been reported to proceed in good yields
in the presence of catalytic amounts of IMesCl. Although
Acknowledgment. Partial support for this work was
provided by the National Science Foundation (CHE-
0449587). Unrestricted support from Amgen, AstraZeneca,
Eli Lilly, Boehringer-Ingelheim, and Bristol Myers Squibb
is gratefully acknowledged. We thank Donald Gauthier of
Merck & Co. for a generous gift of aminoindanol. J.K. thanks
the Royal Thai Government for a graduate fellowship. J.W.B
is a fellow of the Packard Foundation, the Beckman
Foundation, the Sloan Foundation, and a Cottrell Scholar.
Supporting Information Available: Full experimental
procedures and characterization for the preparation of catalyst
1. This material is available free of charge via the Internet
at http://pubs.acs.org.
OL800006M
(20) (a) He, M.; Bode, J. W. Org. Lett. 2005, 7, 3103-3106. (b) For
another racemic variant, see: Chan, A.; Scheidt, K. J. Am. Chem. Soc. 2007,
129, 5334-5335.
(21) Nair, V.; Pattoorpadi Babu, B.; Vellalath, S.; Suresh, E. Chem.
Commun. 2008, DOI: 10.1039/b715733a.
(19) Under these reaction conditions the chloroaldehyde was not com-
sumed when imidazolium salt 1‚ClO4 was employed. When DBU was used
as base, the aldehyde was consumed but the dihydropyranone product was
not observed.
960
Org. Lett., Vol. 10, No. 5, 2008