N-heterocyclic carbene-catalyzed cyclotetramerization of acrylates

Article abstract of DOI:10.1021/ol4021942

N-Heterocyclic carbenes (NHCs) were found to catalyze the unprecedented cyclotetramerization of acrylates, producing the trisubstituted cyclopentenones in moderate yields. The proton or deuterium adducts of the deoxy-Breslow intermediate derived from NHC and two molecules of methyl acrylate were obtained. A reaction mechanism involving the new umpolung/cyclization sequence is proposed.

Full text of DOI:10.1021/ol4021942

ORGANIC
LETTERS
2013
Vol. 15, No. 23
5916–5919
N‑Heterocyclic Carbene-Catalyzed
Cyclotetramerization of Acrylates
Shin-ichi Matsuoka,* Shoko Namera, Atsushi Washio, Koji Takagi, and Masato Suzuki
Department of Materials Science and Engineering, Graduate School of Engineering,
Nagoya Institute of Technology, Gokiso-cho, Showa-ku, Nagoya 466-8555, Japan
matsuoka.shinichi@nitech.ac.jp
Received August 2, 2013; Revised Manuscript Received November 1, 2013
ABSTRACT
N-Heterocyclic carbenes (NHCs) were found to catalyze the unprecedented cyclotetramerization of acrylates, producing the trisubstituted
cyclopentenones in moderate yields. The proton or deuterium adducts of the deoxy-Breslow intermediate derived from NHC and two molecules of
methyl acrylate were obtained. A reaction mechanism involving the new umpolung/cyclization sequence is proposed.
Catalytic oligomerizations of unsaturated compounds
are important bond-forming reactions because they pro-
vide access to valuable chemicals such as higher olefins
and cyclic products from simple chemical feedstocks. Among
the activated olefins, methyl acrylate (MA) has been one
of the most studied substrates due to its high reactivity
toward nucleophiles or transition-metal complexes.1 Indeed,
the RauhutꢀCurrier reaction,2 i.e., the head-to-tail dimer-
ization, of MA is catalyzed by the trialkyl phosphines3 or the
metal complexes4 to give the vinylidene dimer or trimer,
which can be used as monomers for radical polymerizations.5
Alternatively, the vinylene diester was obtained from MA by
the tail-to-tail dimerization catalyzed by the transition
metal complexes,6 which can be an alternative synthetic
pathway to adipinic acid. In addition, the catalytic trisannu-
lation and cyclodimerization of acrylates to produce
1,3,5-benzenetricarboxylates7 and a coumalate8 have
been reported. To the best of our knowledge, however,
there are no other types of oligomerizations of MA or
the catalytic tetramerization of activated olefins.9
N-Heterocyclic carbenes (NHCs) have been very useful
catalysts for the umpolung reactions of aldehydes.10
Recently, the reactions of NHCs with activated olefins
have also gained attention.11 The zwitterions generated
from NHC withsucholefins show an alternative reactivity:
(1) proton transfer to generate enediamines, i.e., the deoxy-
Breslow intermediates,12ꢀ15 or (2) the further addition to
(1) For a review, see: Tembe, G. L.; Bandyopadhyay, A. R.; Ganeshpure,
P. A.; Satish, S. Catal. Rev.-Sci. Eng. 1996, 38, 299.
(2) For a review, see: Aroyan, C. E.; Dermenci, A.; Miller, S. J.
(7) (a) Tamaso, K.; Hatamoto, Y.; Sakaguchi, S.; Obora, Y.; Ishii, Y.
J. Org. Chem. 2007, 72, 3603. (b) Jiang, H.-F.; Shen, Y.-X.; Wang, Z.-Y.
Tetrahedron Lett. 2007, 48, 7542.
Tetrahedron 2009, 65, 4069.
(3) For a selected example, see: (a) Trumbo, D. L.; Zander, R. A.
J. Polym. Sci., Polym. Chem. 1991, 29, 1053. (b) Su, W.; McLeod, D.;
Verkade, J. G. J. Org. Chem. 2003, 68, 9499.
(4) For selected examples, see: (a) Yi., C. S.; Liu, N. J. Organomet.
Chem. 1998, 553, 157. (b) Aresta, M.; Dibenedetto, A.; Quaranta, E.
Organometallics 2000, 19, 4199. (c) Nakagawa, H.; Sakaguchi, S.; Ishii,
Y. Chem. Commun. 2003, 502.
(5) (a) Kobatake, S.; Yamada, B. J. Polym. Sci., Polym. Chem. 1996,
34, 95. (b) Hirano, T.; Yamada, B. Polymer 2003, 44, 347.
(6) For selected examples, see: (a) Pertici, P.; Ballantini, V.; Salvadori,
P.; Bennett, M. A. Organometallics 1995, 14, 2565. (b) DiRenzo,
G. M.; White, P. S.; Brookhart, M. J. Am. Chem. Soc. 1996, 118, 6225.
(c) Kaneko, Y.; Kanke, T.; Kiyooka, S.; Isobe, K. Chem. Lett. 1997, 23.
(d) Braunstein, P.; Chetcuti, M. J.; Welter, R. Chem. Commun. 2001, 2508.
(e) Wang, C.-C.; Lin, P.-S.; Cheng, C.-H. Tetrahedron Lett. 2004, 45,
6203. (f) Hirano, M.; Sakate, Y.; Komine, N.; Komiya, S.; Bennett, M. A.
Organometallics 2009, 28, 4902.
(8) Maeda, S.; Obora, Y.; Ishii, Y. Eur. J. Org. Chem. 2009, 4067.
(9) Catalytic cyclotetramerization of 1,3-butadiene, an unactivated
olefin: (a) Bosch, M.; Brookhart, M. S.; Ilg, K.; Werner, H. Angew.
Chem., Int. Ed. 2000, 39, 2304. (b) Tobisch, S.; Werner, H. Dalton Trans.
2004, 2963. (c) Bosch, M.; Werner, H. Organometallics 2010, 29, 5646.
(10) For selected recent reviews, see: (a) Vora, H. U.; Rovis, T.
Aldrichimica Acta 2011, 44, 3. (b) Biju, A. T.; Kuhl, N.; Glorius, F.
Acc. Chem. Res. 2011, 44, 1182. (c) Hirano, K.; Piel, I.; Glorius, F. Chem.
Lett. 2011, 40, 786. (d) Grossmann, A.; Enders, D. Angew. Chem., Int.
Ed. 2012, 51, 314. (e) Bugaut, X.; Glorius, F. Chem. Soc. Rev. 2012, 41,
3511.
(11) For a review, see: Ryan, S. J.; Candish, L.; Lupton, D. W. Chem.
Soc. Rev. 2013, 42, 4906.
(12) Fischer, C.; Smith, S. W.; Powell, D. A.; Fu, G. C. J. Am. Chem.
Soc. 2006, 128, 1472.
r
10.1021/ol4021942
Published on Web 11/15/2013
2013 American Chemical Society

electrophiles. In the former case, the further nucleophilic
addition of the intermediates enables the umpolung reac-
tions such as the cyclization,12 tail-to-tail dimerization,13
rearrangement,14 and three-component reactions.15 The
latter also leads to the useful reactions such as the aza-
MoritaꢀBaylisꢀHillman reaction,16 cyclizations,17
dimerization,18 and polymerizations.19 Although the
NHC catalysis of activated olefins appears promising, the
examined substrates are still limited to the disubstituted
olefins, i.e., vinylenes and vinylidenes. We then turned our
attention to highly reactive vinyl compounds as substrates
and developed the selective cyclotetramerization of MA to
form the trisubstituted cyclopentenone. We now report this
unprecedented NHC catalysis and propose an umpolung/
cyclization mechanism involving the deoxy-Breslow inter-
mediate derived from the MA dimer.
Table 1. Catalyst Screening for the Cyclotetramerization of MAa
The reaction of MA was initially evaluated using 10 mol %
of IDipp in THF at 60 °C for 24 h (entry 4 in Table 1).
After the purification of the crude product by silica gel
column chromatography, compound 1 was isolated as
a pale yellow oil in 22% yield. The high-resolution ESI-
MS indicated the sodium adduct ion at m/z 335.1094,
suggesting the molecular formula of C15H20O7. The IR
spectrum showed absorption bands due to unconjugated
ester carbonyls at 1732 cmꢀ1 and the carbonyl of 2-cyclo-
pentene-1-one at 1702 cmꢀ1. The 1H NMR spectrum
(Figure S1, Supporting Information) showed a singlet
olefinic proton (δ 7.40, H-3), three methoxy protons
(δ 3.66, 3.67, 3.69), and diastereotopic methylene doublets
(δ 2.49, 3.11, H-4). The protons of two methylene groups
(H-11, 12) were observed as a singlet at the same chemical
shift (δ 2.54), which was confirmed by an HMQC experi-
ment (Figure S6). In the COSY spectrum (Figure S4,
Supporting Information), the olefinic proton (H-3) and
the methylene protons (H-7) were coupled to the diastereo-
topic protons (H-4) and the methylene protons (H-8),
respectively. The 13C NMR spectrum (Figure S4, Supporting
Information) indicated the presence of a ketone carbonyl
carbon (δ 204.37, C-1), tertiary and quaternary olefinic
carbons (δ 157.18, C-3; δ 142.67, C-2), and a quaternary
aliphatic carbon (δ 57.30, C-5). The 2D INADEQUATE
(Figure S7, Supporting Information) displayed 12
entry
cat.
IDipp
base
solvent temp (°C) yieldb (%)
1
toluene
toluene
THF
rt
60
rt
0
2
IDipp
IDipp
IDipp
TPT
0
3
trace
22
0
4
THF
60
60
80
60
60
60
60
60
60
60
60
5
THF
6
TPT
bulk
0
7
IDippHCl tBuOK
IMesHCl tBuOK
THF
28
35
0
8
THF
9
2
3
4
5
6
7
tBuOK or DBU THF
tBuOK or DBU THF
tBuOK or DBU THF
tBuOK or DBU THF
tBuOK or DBU THF
tBuOK or DBU THF
10
11
12
13
14
0
0
0
0
0
a 10 mol % of catalyst, 10 mol % of base, for 24 h. b Isolated yield.
(Figures S8ꢀS10, Supporting Information) showed 12
2J and 18 J reasonable correlations. Collectively, these
experiments proved the chemical structure of 1.
3
The catalyst screening and optimization were then per-
formed (Tables 1 and 2). The reaction with 10 mol % of
IDipp in THF or toluene at rt or 60 °C suggested that high
temperatures and polar solvents are required (entries 1ꢀ4
in Table 1). This reaction also proceeded with the NHC
precursors IDippHCl or IMesHCl and tBuOK; however,
IMesHCl showed the better activity (entries 7 and 8 in
Table 1). TPT and the NHC precursors 2-7 with tBuOK or
DBU did not produce 1 (entries 5, 6 and 9ꢀ14 in Table 1).
The optimization was performed using 5 mol % of IM-
esHCl for 24 h (Table 2). When the reactions were carried
13
13Cꢀ C correlations; thus, all signals in the 13C NMR
spectrum were assigned. In addition, the HMBC spectrum
(13) (a) Matsuoka, S.; Ota, Y.; Washio, A.; Katada, A.; Ichioka, K.;
Takagi, K.; Suzuki, M. Org. Lett. 2011, 13, 3722. (b) Biju, A. T.;
Padmanaban, M.; Wurz, N. E.; Glorius, F. Angew. Chem., Int. Ed.
2011, 50, 8412. (c) Kato, T.; Ota, Y.; Matsuoka, S.; Takagi, K.; Suzuki,
M. J. Org. Chem. 2013, 78, 8739.
(14) (a) Atienza, R. L.; Roth, H. S.; Scheidt, K. A. Chem. Sci. 2011, 2,
1772. (b) Atienza, R. L.; Scheidt, K. A. Aust. J. Chem. 2011, 64, 1158.
(15) Matsuoka, S.; Tochigi, Y.; Takagi, K.; Suzuki, M. Tetrahedron
2012, 68, 9836.
(16) (a) He, L.; Jian, T.-Y.; Ye, S. J. Org. Chem. 2007, 72, 7466.
(b) He, L.; Zhang, Y.-R.; Huang, X.-L.; Ye, S. Synthesis 2008, 2825.
(17) (a) Raveendran, A. E.; Paul, R. R.; Suresh, E.; Nair, V. Org.
Biomol. Chem. 2010, 8, 901. (b) Chen, X.-Y.; Sun, L.-H.; Ye, S. Chem.;
Eur. J. 2013, 19, 4441.
t
out in THF at 60 °C using BuOK, K2CO3, Cs2CO3,
and DBU as bases, the best yield was obtained with
tBuOK (entries 1ꢀ4 in Table 2). Among THF, CH2Cl2,
1,2-dimethoxyethane, and 1,4-dioxane at 40ꢀ80 °C using
tBuOK, the best isolated yield (47%) was obtained in
1,4-dioxane at 80 °C (entries 1 and 5ꢀ7 in Table 2). The
conversion and 1H NMR yield were 92 and 60%, respec-
tively(entry7). Underthese optimized conditions, theyield
was moderate (49%) even at a 10 mol % catalyst loading
(18) Matsuoka, S.; Shimakawa, S.; Takagi, K.; Suzuki, M. Tetra-
hedron Lett. 2011, 52, 6835.
(19) (a) Zhang, Y.; Miyake, G. M.; Chen, E. Y.-X. Angew. Chem.,
Int. Ed. 2010, 49, 10158. (b) Zhang, Y.; Chen, E. Y.-X. Angew. Chem.,
Int. Ed. 2012, 51, 2465.
Org. Lett., Vol. 15, No. 23, 2013
5917

involves the intermediate II (see Figure 1) derived from the
NHC and two molecules of MA. This also means the
nucleophilic reaction of II with the third MMA being slow.
We thus propose two possible reaction mechanisms in
Figure 1. The Michael addition of NHC with two mole-
cules of MA generates the zwitterionic ester enolate I.
The proton transfer then results in the formation of the key
deoxy-Breslow intermediate, II, in which the NHCs turn
the β-carbon of MA nucleophilic. The head-to-tail dimer
10 was not detected by GC analysis during the catalysis,
indicating that the R proton of I was not transferred.
The umpolung and the subsequent Michael addition
allow the bond formation between the β-carbons of MA.
This Michael addition (IIfVI) is relatively slow in com-
parison with the consecutive conjugate addition (NHCfI)
and the proton transfer (IfII) processes, as suggested by
the results shown in Scheme 1. In the proposed cycle A, II
reacts with two molecules of MA to generate the enolate
III. The proton transfer of III generates IV, which then
undergoes cyclization to form V. The deprotonation by the
methoxide ion produces 1 and regenerates the NHC. In the
proposed cycle B, the reaction ofIIwith MAtogenerateVI
is followed by the cyclization and the subsequent elimina-
tion of NHC to give VIII. The NHC-catalyzed Michael
addition of VIII to MA produced 1.21 When the reac-
tion mixture, produced under the conditions of entry 8 in
Table 1 for 1 h, was subjected to ESI-MS spectrometry, we
observed five intense signals corresponding to [1 þ Na]þ,
[IMes þ 2MA þ H]þ, [IMes þ 3MA þ H]þ, [IMes þ 4MA ꢀ
MeO]þ, and [IMes þ 4MA þ H]þ (Figure S20, Support-
ing Information). The latter two ions correspond to
V and IV in cycle A. In contrast, for the intermediates
in cycle B, the sodium or proton adduct ions of VIII were
not observed, and [IMes þ3MA ꢀ MeO]þ correspond-
ing to VII was detected but with a low intensity. Thus, we
postulated that cycle A is preferred.
Table 2. Cyclotetramerization of Acrylates Catalyzed by IMesa
[IMesHCl]
(mol %)
temp yieldd
base solventc (°C)
(%)
entry substrateb
1
MA
MA
MA
MA
MA
MA
MA
MA
MA
EA
5
5
tBuOK THF
K2CO3 THF
Cs2CO3 THF
60 30
2
60
60
60
9
9
5
3
5
4
5
DBU
THF
5
5
tBuOK CH2Cl2 40 trace
6
5
tBuOK DME
tBuOK DOX
tBuOK DOX
tBuOK DOX
tBuOK DOX
tBuOK DOX
tBuOK DOX
80 34
80 47 (60)e
80 49
80 24
80 37
80 23
80 31
7
5
8
10
2
9
10
11
12
5
i-BA
MEA
5
5
a For 24 h. b EA = ethyl acrylate, i-BA = isobutyl acrylate, MEA =
2-methoxyethyl acrylate. c DME = 1,2-dimethoxyethane, DOX = 1,4-
dioxane. d Isolated. e The value in parentheses indicates 1H NMR yield.
and decreased to 24% at 2 mol % (entries 8 and 9 in
Table 2). No dimers, trimers, and other tetramers were
observed in the GC analysisof the crude product. A portion
of MA was recovered as 1:2, 1:3 and 1:4 adducts of NHC
and MA, which were detected by ESI-MS (Figure S20,
Supporting Information). Under the optimal conditions
(entry 7 in Table 2), the cyclotetramerizations of ethyl,
isobutyl, and 2-methoxyethyl acrylates produced the cor-
responding cyclic tetramers in 37, 23, and 31% isolated
yields, respectively (entries 10ꢀ12 in Table 2).
Scheme 1. Reaction of MA with IDipp Quenched by CF3CO2X
(X = H or D)
NHC has a nucleophilic reactivity similar to trialkyl
phosphines but promotes the distinct catalysis for MA.
Trialkyl phosphine catalyzes the Rauhut-Currier reaction
of MA to give the dimer 10 through the transfer of the
R- proton of the zwitterionic ester enolate intermediate
(IX) (eq 1).2,3 The transfer of the β-proton of IX to form
the phosphorus ylide is unfavorable. In contrast, NHC is
ineffective for the Rauhut-Currier reaction of MA. Since
the deoxy-Breslow intermediates are relatively stable as
previously pointed out,22 the transfer of the β-proton of I is
predominant over the R-proton.
The reaction mechanism was next considered. The reac-
tion for 10 min under the conditions of entry 4 in Table 1,
followed by the addition of CF3CO2H, produced com-
pound 820 in 52% yield (Scheme 1) which was estimated
by 1H NMR spectroscopy. The use of CF3CO2D instead
resulted in the formation of 9 with the selective deuterium
incorporation at the β carbon not the R carbon of the
terminal ester. These results indicated that this catalysis
It is also noteworthy that the structure of NHC has
significant effects on the catalysis of activated olefins.
(20) Compound 8 was alternatively synthesized from the reaction of
IDipp with the MA dimer 10 in 99% yield.
(21) The NHC-catalyzed Michael additions of 1,3-dicarbonyl com-
pounds were reported; see: (a) Boddaert, T.; Coquerel, Y.; Rodriguez, J.
Adv. Synth. Catal. 2009, 351, 1744. (b) Boddaert, T.; Coquerel, Y.;
Rodriguez, J. Chem.;Eur. J. 2011, 17, 2266.
(22) Zhao, L.; Chen, X. Y.; Ye, S.; Wang, Z. X. J. Org. Chem. 2011,
76, 2733.
5918
Org. Lett., Vol. 15, No. 23, 2013

Figure 1. Possible reaction mechanisms. Cycle A is preferred.
Indeed, IMes catalyzes the cyclotetramerization of MA
and produces the stable single adduct with methyl
methacrylate.19b In contrast, TPT catalyzes the tail-to-tail
dimerizations of methacrylates through the enolate inter-
mediate X (eqs 2 and 3).13 Although this enolate is an
analogue of IV and VI, it does not undergo cyclization
to generate XI (eq 2) but instead the proton transfer with
the elimination of TPT to produce 11 (eq 3). One reason
for this difference may be that the R-proton of X shows a
higher acidity than those of IV and VI due to the relatively
stronger electron-withdrawing ability of the triazolium
ring. This interesting difference between these two major
NHC catalysts suggests that the design of the NHC struc-
ture is important for future reaction discovery.
trisubstituted cyclopentenones. The key deoxy-Breslow
intermediate was obtained as the proton or deuterium
adducts, and the other intermediates were detected by
ESI-MS, leading us to propose the reaction mechanism.
This NHC catalysis is remarkable since (1) this is the first
cyclotetramerization of an activated olefin and (2) the first
reaction of NHC with a vinyl compound, which involves
(3) the first example of the Breslow type intermediate
derived from two electrophiles and (4) is the first catalytic
cascade of the umpolung of Michael acceptors followed by
cyclization. Further work is now in progress to expand the
scope of this cascade.
Acknowledgment. We thank Professor Akihiro Yoshino
and Masato Taki at Nagoya Institute of Technology for the
measurement of 2D INADEQUATE. This work was par-
tially supported by a Grant-in-Aid for Scientific Research
for Young Scientists (B) No. 24750102.
Supporting Information Available. Experimental section;
1H, 13C, and 2D NMR and ESI-MS spectra. This
material is available free of charge via the Internet at
http://pubs.acs.org.
In conclusion, we have determined that the cyclotetra-
merization of acrylates catalyzed by IMes produces the
The authors declare no competing financial interest.
Org. Lett., Vol. 15, No. 23, 2013
5919