With molecular-oxygen-activated Lewis acids: Dinuclear molybdenum complexes for aza-Diels-Alder reactions of acyl hydrazones

Article abstract of DOI:10.1002/anie.200600801

(Chemical Equation Presented) Easy activation: A Lewis acid catalyst based on molybdenum relies on activation by molecular oxygen for its high catalytic activity. The novel complex catalyzes the aza-Diels-Alder reaction of acyl hydrazones with Danishefsky

Full text of DOI:10.1002/anie.200600801

Communications
Transition-Metal Catalysis
DOI: 10.1002/anie.200600801
With Molecular-Oxygen-Activated Lewis Acids:
Dinuclear Molybdenum Complexes for Aza-
Diels–Alder Reactions of Acyl Hydrazones**
Yasuhiro Yamashita, Matthew M. Salter,
Kayoko Aoyama, and Shu¯ Kobayashi*
Lewis acids are amongst the most important activating groups
in chemistry, and their use is ubiquitous in synthesis. As a
result, the last two decades have seen a veritable explosion of
interest in the synthesis of Lewis acids based on almost every
suitable metal in the Periodic Table and their application as
catalysts in a broad range of reactions.[1] In addition to the
nature of the metal center, the activity of Lewis acids may be
influenced by external factors, such as coordinating additives
and variation of the associated counteranion; for example,
exchanging a highly coordinating halogen anion for the less
nucleophilic triflate anion is frequently used to increase the
activity of Lewis acids. Apart from these techniques, however,
there are few effective ways of increasing Lewis acidity. We
now report a new type of Lewis acid system that is activated
by molecular oxygen, which to the best of our knowledge
represents the first example of a Lewis acid that is activated in
this way. Whilst dinuclear molybdenum compounds [Mo2-
[2]
(OAc)4]have been investigated as inorganic compounds,
their use in synthesis has remained relatively unexplored.[3]
Furthermore, we demonstrate that this catalyst mediates the
hitherto unknown Diels–Alder reaction of aromatic acyl
hydrazones with Danishefsky dienes (Scheme 1).[4]
Scheme 1. Aza-Diels–Alder reaction catalyzed by a Lewis acid.
The aza-Diels–Alder reaction of imines with dienes is an
important method for the construction of substituted piper-
[*] Dr. Y. Yamashita, Dr. M. M. Salter, K. Aoyama, Prof. Dr. S. Kobayashi
Graduate School of Pharmaceutical Sciences
The University of Tokyo
Hongo, Bunkyo-ku, Tokyo 113–0033 (Japan)
Fax: (+81)3-5684-0634
E-mail: skobayas@mol.f.u-tokyo.ac.jp
[**] This work was partially supported by a Grant-in-Aid for Scientific
Research from the Japan Society for the Promotion of Science
(JSPS).
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
3816
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 3816 –3819

Angewandte
Chemie
idine derivatives. However, some imines are not suitable for
use in these reactions owing to their instability, which leads to
low yields of the isolated cyclic products. Hydrazones offer a
solution to this problem, as they are known to act as stable
exerted a profound effect on reactivity (entry 6) and that its
exclusion led to a dramatic reduction in yield (entry 7).[7] We
therefore deliberately sought to introduce molecular oxygen
into the reaction by controlled preparation of the catalytic
species, thus the catalyst solution was pretreated with O2 prior
to the addition of the diene.
À
imine equivalents and the hydrazone N N bond can be
readily cleaved to give amine functionalities.[5] We have
previously demonstrated that acyl hydrazones derived from
aliphatic aldehydes can function as electrophiles in allylations
and Mannich-type reactions in the presence of Lewis acid
catalysts, such as [Sc(OTf)3](OTf = trifluoromethanesulfo-
nate).[6] However, when we attempted to extend this method-
ology to the aza-Diels–Alder reactions of hydrazones with
Danishefsky dienes, we found that the reaction proceeded
only sluggishly in the presence of conventional Lewis acids
The effect of varying the molybdenum source was also
investigated. Trials showed that use of dinuclear Mo sources
with a counterion that has a similar basicity to acetate, such as
[Mo2(OCOC5H11)4]and [Mo (OCOPh)4], gave very similar
2
results (entries 9 and 10); however, the use of [Mo2-
(OCOCF3)]or [Mo 2(OTf)4]resulted in complete shutting
down of the reaction (entries 11 and 12). Raising the reaction
temperature to 408C resulted in a marginal improvement in
yield, whereas decreasing the catalyst loading to 5 mol% at
408C raised the yield still further to 75% (entry 14). The
addition of dried 5-Š molecular sieves (MS) to the reaction
significantly improved the efficiency of the reaction and
afforded the desired adduct in 85% yield (entry 15); however,
attempts to decrease the catalyst loading down to 2 mol%
resulted in a steep drop in yield (entry 16). It is important to
note that use of the mononuclear oxidized molybdenum
species (entries 17 and 18) did not promote the reaction at
all.[9] This behavior suggests that the catalyst involves a
species with at least two metal centers in a well-defined
spatial array.[10]
such as [Sc(OTf)3], [Yb(OTf)3]
, BF·OEt2, and ZnCl2
3
(Table 1, entries 1–4). On further investigation, it was dis-
covered that when the catalyst was switched to [Mo2(OAc)4],
the reaction proceeded smoothly to afford the desired adduct
in good yield (entry 5). A screen of common laboratory
solvents revealed that acetonitrile and acetone gave the
optimum results. The use of other less polar (toluene, CH2Cl2,
EtOAc) or more polar solvents (MeNO2, dimethylformamide
(DMF)) did not lead to any improvement in yield (26–56%).
Whilst delighted by these results, we were concerned by
the poor reproducibility of the reaction. After careful
examination of the reaction conditions, it was found that
molecular oxygen present as a contaminant in the system
With these results in hand, we turned our attention to the
synthetic scope of the reaction (Table 2). Benzoylhydrazones
Table 1: Aza-Diels–Alder reaction of 1a with 2a catalyzed by a Lewis
acid.[a]
Table 2: Substrate scope in the reaction of hydrazone 1 with Danishefsky
diene 2 with nuclear Mo Lewis acid catalysis.[a]
Entry
Catalyst
Solvent
Yield of 3aa [%]
Entry
R1
Diene
Product
Yield [%][b]
1[b]
2[b]
[Sc(OTf)3]
[Yb(OTf)3]
BF3·OEt2
ZnCl2
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
acetone
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
–
[c]
8
22
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20c
21
22c
23
PhCH2CH2 (1a)
PhCH2CH2 (1a)
2a
2b
2a
2b
2a
2b
2a
2b
2a
2b
2a
2b
2a
2a
2b
2a
2b
2a
2b
2a
2b
2a
2b
3aa
3ab
3ba
3bb
3ca
3cb
3da
3db
3ea
3eb
3 fa
3 fb
3ga
3ha
3hb
3ia
85
86
88
77
85
76
85
82
81
70
70
50
16
62
73
86
68
84
61
65
72
64
42
3[b]
4[b]
trace amount
CH3 (1b)
À
5[b]
[Mo2(OAc)4]
[Mo2(OAc)4]
[Mo2(OAc)4]
[Mo2(OAc)4]
[Mo2(OCOC5H11)]
[Mo2(OBz)4]
[Mo2(OCOCF3)4]
[Mo2(OTf)4]·4AcOEt
[Mo2(OAc)4]
[Mo2(OAc)4]
[Mo2(OAc)4]
[Mo2(OAc)4]
[MoO2]
69
70
<10
70
CH3 (1b)
À
À
6
C3H7 (1c)
7[d]
C3H7 (1c)
À
8[e]
C5H11 (1d)
À
À
9
10
11
12
64
C5H11 (1d)
À
60
trace amount
no reaction
71
C7H15 (1e)
À
C7H15 (1e)
À
C9H19 (1 f)
13[f]
14[g]
15[g,h]
16[h,i]
17[k]
18[k]
C9H19 (1 f)
À
À
75
85
55
c-C6H10 (1g)
(CH3)2CHCH2 (1h)
À
À
(CH3)2CHCH2 (1h)
TBSOCH2CH2 (1i)
TBSOCH2CH2 (1i)
À
no reaction
no reaction
À
[MoO3]
3ib
3ja
3jb
3ka
3kb
3la
ꢀ À
TIPSC C (1j)
[a] In the case of the Mo2 catalysts, the Mo2 species was freshly prepared
from [Mo(CO)6] according to a literature procedure[8] and stored under Ar
in a glove box. For the preparation of the catalyst, the Mo2 salt (10 mol%)
and hydrazone were heated under O2 at 508C for 3 h unless otherwise
noted. After addition of the diene, the reaction was carried out under Ar
at 208C for 24 h. [b] Reaction carried out at room temperature using
20 mol% catalyst with no activation. No precautions to exclude oxygen
were taken. [c] Complex mixture. [d] Reaction performed under Ar in
degassed MeCN. [e] Treated with O2 at room temperature. [f] Reaction
carried out at 408C for 24 h. [g] 5 mol% of catalyst was used at 408C for
24 h. [h] Powdered 5-ꢀ MS were added (50 mg/0.4 mmol substrate).
[i] 2 mol% of catalyst was used at 408C for 24 h. [k] Activated by heating
under O2 at 508C for 3 h prior to addition of the hydrazone.
ꢀ À
TIPSC C (1j)
À
C6H5 (1k)
À
C6H5 (1k)
À
p-ClC6H4 (1l)
À
p-ClC6H4 (1l)
3lb
[a] [Mo2(OAc)4] (5 mol%), the hydrazone substrate, and 5-ꢀ MS were
mixed in dry MeCN and heated to 508C under O2 atmosphere for 3 h.
After addition of the diene in MeCN (final reaction mixture concen-
tration=0.2 mm), the reaction was carried out at 408C under Ar for 24 h.
[b] Yield of the isolated product analysis. [c] Conditions: 10 mol%
catalyst, reaction time=72 h. TBS=tert-butyldimethylsilyl, TIPS=tri-
isopropylsilyl.
Angew. Chem. Int. Ed. 2006, 45, 3816 –3819ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3817

Communications
Table 3: Strecker reaction of (E)-N-benzylidenebenzenamine (8) with
TMSCN.
prepared from non-branched aldehydes performed well in the
reaction, thus affording the corresponding 4-piperidenone
products in good yields (entries 1–12). However, the intro-
duction of a- or b-branching had a detrimental effect on the
efficiency of the reaction (entries 13–15). The mild nature of
the procedure was attested to by the observation that a
benzylhydrazone bearing a silyloxy functionality could be
employed in the reaction without significant deprotection
(entries 16 and 17). Additionally, alkynyl hydrazones worked
well as they gave the desired adducts in good yields
(entries 18 and 19). It should be noted that aryl hydrazones
also functioned well in the reaction, although longer reaction
times were required to achieve good yields (entries 20–23).
The molecular-oxygen-activated molybdenum catalyst
was also found to be effective in other Lewis acid mediated
reactions (Scheme 2). Gratifyingly, the Mukaiyama aldol
Entry Mo source
Mo loading [mol%] O2 activation[a] Yield [%][b]
1[c]
2[c]
3
[Mo2(OAc)4]
[Mo2(OAc)4]
[MoO2]
5
5
yes
no
96
24
18
21
10
10
yes
yes
4
[MoO3]
[a] Mo source, imine, and MS heated under O2 for 3 h prior to the
addition of TMSCN. [b] Yields of the isolated products. [c] Freshly
prepared [Mo2(OAc)4] (bright yellow).
the presence of molecular oxygen and the mononuclear
[MoOx]species (entries 3 and 4). This observation further
reinforces the hypothesis that a binuclear array of molybde-
num centers is responsible for the efficient action of the novel
catalyst.
Next, we gave consideration to the structure of the active
catalyst species. Our working hypothesis involves an oxidized
dinuclear molybdenum species; therefore, as molybdenum
complexes are readily oxidized and Cotton et al. reported[10]
the preparation and X-ray crystallographic characterization
of [Mo2(h2-O2CCH3)2(m2-DXylF2,6)2(m2-O2)]( 10; DXylF =
N,N,’-di(2,6-xylylformamidine)), a dinuclear molybdenum
complex in which a diamidinate ligand is coordinated to the
metal centers and has two bridging oxygen atoms that span
À
the Mo Mo bond, we examined the species obtained by
heating [Mo2(OAc)4]with hydrazone 1a in MeCN by electro-
spray ionization mass spectroscopy (ESI MS; solvent: MeCN,
negative mode). A major peak at m/z 758 was observed,
which corresponds to a molecular formula of C32H30Mo2N4O6.
By analogy with complex 10, we tentatively assigned this
species as having one of the two possible structures shown in
Scheme 3. Moreover, the aza-Diels–Alder reaction pro-
Scheme 2. Mukaiyama aldol and Mannich-type reactions catalyzed by
the [Mo2(OAc)4]/O2 system.
reaction between benzaldehyde and silyl enol ether 4
proceeded smoothly in the presence of 5 mol% [Mo2(OAc)4]
to afford the desired b-hydroxy ester 5 in excellent yield.
Similarly, the corresponding Mannich-type addition of 4 to N-
o-methoxyphenyl imine 6 was also catalyzed by the oxidized
molybdenum system to give b-imino ester 7 cleanly and
efficiently. In both cases, it was found that activation of the
catalyst by pretreatment with molecular oxygen was essential
for an efficient reaction: the corresponding reactions carried
out without O2 activation proceeded to give the products in
much lower yields than those observed under the oxygen-
activated conditions.
Scheme 3. Plausible structures for the catalytic species formed from
[Mo2(OAc)4] and hydrazone 1a under O2 in MeCN.
In addition, it was found that the [Mo2(OAc)4]/O2 system
could also be applied to the catalysis of the Strecker reaction
of imine 8 with trimethylsilyl cyanide (TMSCN) to give the
corresponding a-cyano imine 9 (Table 3). In these cases as
well, the controlled introduction of molecular oxygen was
found to be essential for an efficient reaction. Oxidation of
freshly prepared [Mo2(OAc)4][8] prior to use afforded the
desired addition product in very high yield (entry 1), whereas
omission of the oxidative conditioning step resulted in a
drastic decrease in the efficiency of the reaction (entry 2).
Interestingly, the reaction proceeded only very sluggishly in
ceeded in quite low yield (< 10%) in the presence of
50 mol% [Mo2(OAc)4]. This result provided further circum-
stantial evidence for our hypothesis and indicated that the two
hydrazone molecules that coordinate to the Mo2 species were
less reactive and that the dimolybdenum–hydrazone complex
activates free the hydrazone towards cyclization.
In summary, we have discovered a new type of Lewis acid
catalyst based on molybdenum, which relies on activation by
molecular oxygen for its function. The novel complex
3818
www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 3816 –3819

Angewandte
Chemie
Tetrahedron Lett. 1995, 36, 6351; b) A. V. Malkov, I. Baxendale,
D. J. Mansfield, P. Kocovsky. Tetrahedron Lett. 1997, 4895;
catalyzes the aza-Diels–Alder reaction of a range of acyl
hydrazones with Danishefsky dienes to afford 4-piperidinones
in good yields. Additionally, the same catalyst efficiently
mediates Mukaiyama aldol and Mannich-type reactions as
well as the addition of TMSCN to imines. In all cases,
activation by molecular oxygen was found to be crucial for
high catalytic activity; the absence of such activation resulted
in a dramatic decrease in efficiency and yield. Further
investigations into the synthetic scope of the reaction and
catalyst structure are in progress.
ˇ
c) A. V. Malkov, I. R. Baxendale, D. Dvorꢀk, D. J. Mansfield, P.
Kocovsky, J. Org. Chem. 1999, 64, 2737; d) A. V. Malkov, S. L.
Davis, W. L. Mitchell, P. Kocovsky, J. Org. Chem. 1999, 64, 2751;
e) A. Sakakura, R. Kondon, K. Ishihara, Org. Lett. 2005, 7, 1974,
and references therein.
[10]F. A. Cotton, L. M. Daniels, C. A. Murillo, J. G. Slaton J. Am.
Chem. Soc. 2002, 124, 2878.
Received: March 1, 2006
Published online: May 3, 2006
Keywords: aza-Diels–Alder reaction · homogeneous catalysis ·
.
Lewis acids · molybdenum · oxygen activation
[1] Lewis Acids in Organic Synthesis (Ed.: H. Yamamoto), Wiley-
VCH, Weinheim, 2000.
[2]a) C. D. Garner in Comprehensive Coordination Chemistry,
Vol. 3 (Eds.: G. Wilkinson, R. D. Gillard, J. A. McCleverty),
Pergamon, Oxford, 1987, chap. 36.3, p. 1301; b) F. H. Jardine,
P. S. Sheridan in Comprehensive Coordination Chemistry, Vol. 4
(Eds.: G. Wilkinson, R. D. Gillard, J. A. McCleverty), Pergamon,
Oxford, 1987, chap. 48, p. 901; c) J. L. Eglin in Comprehensive
Coordination Chemistry II, Vol. 4 (Eds.: J.A. McCleverty, T. J.
Meyer), Pergamon, Oxford, 2004, chap. 4.9, p. 575.
[3]For organic reactions using molybdenum dimer complexes, see:
a) M. McCann in Catalysis by Di- and Polynuclear Metal Cluster
Complexes (Eds.: A. D. Adams, F. A. Cotton), Wiley-VCH, New
York, 1998, chap. 5, p. 145; for asymmetric hetero Diels – Alder
reactions using chiral Rh dimer complexes, see b) M. P. Doyle,
I. M. Phillips, W. Hu, J. Am. Chem. Soc. 2001, 123, 5366; c) M.
Valenzuela, M. P. Doyle, C. Hedberg, W. Hu, A. Holmstrom,
Synlett 2004, 2425; d) M. P. Doyle, M. Valenzuela, P. Huang,
Proc. Natl. Acad. Sci. USA 2004, 101, 5391; e) R. E. Forslund, J.
Cain, J. Colyer, M. P. Doyle, Adv. Synth. Catal. 2005, 347, 87;
f) M. Anada, T. Washio, N. Shimada, S. Kitagaki, M. Nakajima,
M. Shiro, S. Hashimoto, Angew. Chem. 2004, 116, 2719; Angew.
Chem. Int. Ed. 2004, 43, 2665; for a reaction using [Mo2(OAc)4]
as a stoichiometric Lewis acid, see: g) H. Hirao, Y. Fujihara, S.
Tsuno, Y. Ohshiro, T. Agawa, Chem. Lett. 1984, 367.
[4]Y. Yamashita, Y. Mizuki, S. Kobayashi, Tetrahedron Lett. 2005,
46, 1803.
[5]M. J. Burk, J. E. Feaster, J. Am. Chem. Soc. 1992, 114, 6266, and
also see Ref. [6].
[6]a) H. Oyamada, S. Kobayashi, Synlett 1998, 249; b) S. Kobayashi,
T. Furuta, K. Sugita, H. Oyamada, Synlett 1998, 1019; c) S.
Kobayashi, Y. Hasegawa, H. Ishitani, H. Chem. Lett. 1998, 1131;
d) S. Kobayashi, K. Sugita, H. Oyamada, Synlett 1999, 138; e) K.
Manabe, H. Oyamada, K. Sugita, S. Kobayashi, J. Org. Chem.
1999, 64, 8054; f) O. Okitsu, H. Oyamada, T. Furuta, S.
Kobayashi, Heterocycles 2000, 52, 1143; g) M. Sugiura, S.
Kobayashi, Angew. Chem. Int. Ed. 2005, 44, 5176.
À
[7]An alternative radical pathway that accounts for C C bond-
forming reactions in the presence of metal catalysts and
molecular oxygen can be proposed. We investigated the effect
of adding
a radical scavenger (2,4,6-tri-tert-butylphenol,
20 mol%) to the reaction; almost no effect was observed and
the product heterocycle was obtained in 71% yield under the
conditions shown in Table 1, entry 14.
[8]A. B. Brignole, F. A. Cotton, Inorg. Synth. 1972, 13, 81.
[9]The use of monomeric Mo species as Lewis acids has been
ˇ
ˇ
ˇ
documented: a) H. Dvorꢀkova, D. Dvorꢀk, J. Srogi, P. Kocovsky,
Angew. Chem. Int. Ed. 2006, 45, 3816 –3819ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3819