Utilization of aldoses as a carbonyl source in cyclocarbonylation of enynes

Article abstract of DOI:10.1021/jo1012288

Figure presented. The reaction of enynes with acetyl-masked aldoses in the presence of a rhodium(I) catalyst resulted in cyclocarbonylation, thus avoiding the direct use of carbon monoxide, to afford bicyclic cyclopentenones. In rhodium catalysis, aldoses serve as a carbon monoxide equivalent by donating their carbonyl moieties on the acyclic aldehyde form to enynes. A variety of aldoses, including d-glucose, d-mannose, d-galactose, d-xylose, and d-ribose, can be used as a carbonyl source. Using the method, a wide variety of enynes were cyclocarbonylated in 22-67% yields. An asymmetric variant also proceeded with moderate to high enantioselectivity.

Full text of DOI:10.1021/jo1012288

pubs.acs.org/joc
the direct one-step synthesis of bicyclic cyclopentenones.1
Utilization of Aldoses as a Carbonyl Source in
Cyclocarbonylation of Enynes
In the original method, enynes were reacted with stoichio-
metric amounts of transition metals containing a carbonyl
ligand. Numerous subsequent studies reported on transition-
metal-catalyzed reactions in which carbon monoxide itself
was used as a carbonyl source. Recent progress has disclosed
that various formyl compounds, such as aldehydes2 and
formates,3 can be used as a substitute for carbon monoxide,
leading to much more approachable and easily handled
transformations, without the need to use toxic carbon mon-
oxide. The strategies involve the decarbonylation of the
formyl compounds by transition-metal catalysts to produce
an isolated carbonyl moiety or metal-carbonyl species,
which serves as a substitute for carbon monoxide. In this
study, we envisioned the use of aldoses, in which the alde-
hyde form is present in trace amounts at equilibrium, as a
carbonyl source in the cyclocarbonylation of enynes. Aldoses
are widespread in nature, in the forms of oligosaccharides,
polysaccharides, glycoproteins, lipopolysaccharides, nucleo-
tidyl sugars, and nucleic acids, and they are one of the most
reliable and sustainable carbon resources. Therefore, aldoses
can be regarded as the most favorable carbonyl source. Their
conventional synthetic applications can be broadly classified
into the following two types: synthetic tools such as organoca-
talysts, ligands, and auxiliaries for asymmetric synthesis,4 and
synthetic building blocks for the synthesis of biologically active
compounds and glycosides.5 Quite recently, Chung et al.
developed a cyclocarbonylation reaction of an enyne using just
glucose as a carbonyl source, which is based on the same
strategy as mentioned above.2l We describe herein a novel
utilization of aldoses in the cyclocarbonylation of enynes and
the development of the strategy into an asymmetric reaction,
based on the catalytic decarbonylation of the acyclic aldehyde
form.6 This represents the first asymmetric variant using
aldoses as a carbonyl source in the cyclocarbonylation of
enynes.
Keiichi Ikeda, Tsumoru Morimoto,* and Kiyomi Kakiuchi
Graduate School of Materials Science, Nara Institute of
Science and Technology (NAIST), Takayama, Ikoma,
Nara 630-0192, Japan
morimoto@ms.naist.jp
Received June 23, 2010
The reaction of enynes with acetyl-masked aldoses in the
presence of a rhodium(I) catalyst resulted in cyclocarbo-
nylation, thus avoiding the direct use of carbon monox-
ide, to afford bicyclic cyclopentenones. In rhodium
catalysis, aldoses serve as a carbon monoxide equivalent
by donating their carbonyl moieties on the acyclic alde-
hyde form to enynes. A variety of aldoses, including
D-glucose, D-mannose, D-galactose, D-xylose, and D-ribose,
can be used as a carbonyl source. Using the method, a wide
variety of enynes were cyclocarbonylated in 22-67%
yields. An asymmetric variant also proceeded with moder-
ate to high enantioselectivity.
We first attempted the cyclocarbonylation of enyne 1a
with D-glucose (2a) in the presence of 5 mol % of of [RhCl-
(cod)]2 and 10 mol % of BINAP in 1,4-dioxane/dimethyla-
cetamide (DMA) (1/1) at 130 °C for 40 h. The resulting
The cyclocarbonylation of enynes with carbon monoxide,
so-called the Pauson-Khand reaction, is a powerful tool for
(1) For recent reviews of the Pauson-Khand reaction, see: (a) Gibson,
S. E.; Stevenazzi, A. Angew. Chem., Int. Ed. 2003, 42, 1800–1810. (b) Shibata,
T. Adv. Synth. Catal. 2006, 348, 2328–2336. (c) Lee, H. W.; Kwong, F. Y.
Eur. J. Org. Chem. 2010, 789–811.
(3) (a) Park, K. H.; Son, S. U.; Chung, Y. K. Chem. Commun. 2003, 39,
1898–1899. (b) Lee, H. W.; Chan, A. S. C.; Kwong, F. Y. Chem. Commun.
2007, 43, 2633–2635.
(2) (a) Morimoto, T.; Fuji, K.; Tsutsumi, K.; Kakiuchi, K. J. Am. Chem.
Soc. 2002, 124, 3806–3807. (b) Shibata, T.; Toshida, N.; Takagi, K. Org. Lett.
2002, 4, 1619–1621. (c) Shibata, T.; Toshida, N.; Takagi, K. J. Org. Chem.
2002, 67, 7446–7450. (d) Fuji, K.; Morimoto, T.; Tsutsumi, K.; Kakiuchi, K.
Angew. Chem., Int. Ed. 2003, 42, 2409–2411. (e) Jeong, N.; Kim, D. H.; Choi,
J. H. Chem. Commun. 2004, 40, 1134–1135. (f) Fuji, K.; Morimoto, T.;
Tsutsumi, K.; Kakiuchi, K. Tetrahedron Lett. 2004, 45, 9163–9166. (g)
Kwong, F. Y.; Li, Y. M.; Lam, W. H.; Qiu, L.; Lee, H. W.; Yeung, C. H.;
Chan, K. S.; Chan, A. S. C. Chem.;Eur. J. 2005, 11, 3872–3880. (h) Shibata,
T.; Toshida, N.; Yamasaki, M.; Maekawa, S.; Takagi, K. Tetrahedron 2005,
61, 9974–9979. (i) Kwong, F. Y.; Lee, H. W.; Qiu, L.; Lam, W. H.; Li, Y.-M.;
Kwong, H. L.; Chan, A. S. C. Adv. Synth. Catal. 2005, 347, 1750–1754. (j)
Kwong, F. Y.; Lee, H. W.; Lam, W. H.; Qiu, L.; Chan, A. S. C. Tetrahedron:
Asymmetry 2006, 17, 1238–1252. (k) Morimoto, T.; Fujioka, M.; Fuji, K.;
Tsutsumi, K.; Kakiuchi, K. Pure Appl. Chem. 2008, 80, 1079–1087. (l) Quite
recently, Chung’s group has reported on the use of alcohols as a carbonyl
source in a rhodium-catalyzed cyclocarbonylation. In this case, aldehydes
generated in situ via dehydrogenation of alcohols act as a carbonyl source:
Park, J. H.; Cho, Y.; Chung, Y. K. Angew. Chem., Int. Ed. 2010, 49, 5138–
5141.
(4) For recent reviews, see: (a) Hollingsworth, R. I.; Wang, G. Chem. Rev.
ꢀ
ꢁ
2000, 100, 4267–4282. (b) Dieguez, M.; Pamies, O.; Claver, C. Chem. Rev.
2004, 104, 3189–3215. (c) Boysen, M. M. K. Chem.;Eur. J. 2007, 13, 8648–
8659.
(5) For reviews, see: (a) Inch, T. D. Adv. Carbohydr. Chem. Biochem.
1972, 27, 191–225. (b) Hanessian, S. Acc. Chem. Res. 1979, 12, 159–165. (c)
Frase-Reid, B.; Sun, K. M.; Tam, T. F. Bull. Soc. Chim. Fr. 1981, 238–246.
(d) Inch, T. D. Tetrahedron 1984, 40, 3161–3213. (e) Williams, N. R.;
Davidson, B. E.; Ferrier, R. J.; Furneaux, R. H. Carbohydr. Chem. 1985,
17, 244–255. (f) Nicolaou, K. C.; Mitchell, H. J. Angew. Chem., Int. Ed. 2001,
40, 1576–1624.
(6) For reports on decarbonylation of aldoses by a rhodium complex, see:
(a) Andrews, M. A.; Klaeren, S. A. J. Chem. Soc., Chem. Commun. 1988,
1266-1267 (stoichiometric). (b) Beck, R. H. F.; Elseviers, M. Lemmens,
H. O. J. EP 0716066A1, 1996 (stoichiometric). (c) Andrews, M. A.
Organometallics 1989, 8, 2703-2708 (stoichiometric). (d) Andrews, M. A.;
Gould, G. L.; Klaeren, S. A. J. Org. Chem. 1989, 54, 5257-5264
(stoichiometric). (e) Monrad, R. N.; Madsen, R. J. Org. Chem. 2007, 72,
9782-9785 (catalytic).
DOI: 10.1021/jo1012288
r
Published on Web 08/25/2010
J. Org. Chem. 2010, 75, 6279–6282 6279
2010 American Chemical Society

Ikeda et al.
JOCNote
SCHEME 1. 13C Labeling Experiment
SCHEME 2. Effect of Anomeric Configuration on Reaction
Efficiency
mixture included the targeted carbonylated product 3a in
only 7% yield, along with the dimer of 1a as the main
product, which is formed via a [2 þ 2 þ 2] cycloaddition.7
During the course of the reaction, the reaction mixture was
heterogeneous because of the extremely low solubility of 2a
in organic solvents.8 In order to increase the participation of
the CO substitute in the reaction, we used tetracetyl-D-
glucose 2b as a carbonyl source, which can be readily pre-
pared in two steps from D-glucose (2a).9 The use of 2b resulted
in a dramatic increase in the yield, leading to the formation of
the carbonylated product 3a in 55% yield. Under the condi-
tions, the cationic precursor, [Rh(cod)2]OTf, which is reported
to catalyze efficiently even under a lower concentration of
carbon monoxide,10 did not catalyze the sequence of the
decarbonylation-cyclocarbonylation to the production of 3a.
In order to verify the origin of the carbonyl groups
introduced into the product 3a, we performed the reaction
using acetyl-masked D-glucose 2b0, enriched at C-1 (the
anomeric position) with 13C (>99%). Under the aforemen-
tioned conditions, the catalytic reaction of 1a with 2b0
resulted in cyclocarbonylation to give the carbonylated
product 3a0 in 46% yield (Scheme 1). In this reaction, the
introduced carbonyl group in 3a0 was labeled with >99% 13C.
Thus, this clearly indicates that the origin of the carbonyl
moiety is not from the acetyl group of DMA (solvent) or the
masked glucose, but from the anomeric carbon of the masked
glucose 2b0.
We further investigated the effect of the stereochemistry
of the anomeric carbon on the efficiency of the reaction
(Scheme 2). When either a mixture of R- and β-anomers
(R/β = 30/70) or the β-anomer was used, the desired product
was obtained in 52% yield in both cases, and therefore, no
difference between the reactivities of the different anomers
was observed. These results can be rationalized by the
observation that, in 1,4-dioxane/DMA (1/1) at 130 °C, both
the R/β-mixture and the β-anomer are converted quantita-
tively within 1 h into a mixture of R- and β-anomers in an R/β
ratio of 77/23. Extending the reaction time (40 h) led to no
change in the ratio between R- and β-anomers. Thus, R/β-
interconversion via an acyclic aldehyde form would occur
rapidly and reach equilibrium under the conditions of the
present catalysis, despite the initial ratio of the mixture, and
CHART 1
it can be safely concluded that both anomers function
equally well as a carbonyl source.
The present reaction tolerated the use of other acetyl-
masked aldoses 2c-f, which are easily produced from readily
available hexoses (D-galactose and D-mannose) as well as
pentoses (D-xylose and D-ribose).11 Under otherwise identi-
cal conditions, 2c-f all functioned as a carbonyl source in the
cyclocarbonylation reactions of enyne 1a to produce the
carbonylated product 3a, although the yields were slightly
lower (2c 39%; 2d 29%; 2e 43%; 2f 39%) than that for the
masked D-glucose 2b (Chart 1).
The cyclocarbonylation reactions of various enynes were
explored using aldose derivatives. Some selected results are
shown in Table 1. Some enynes preferred the use of the
masked xylose 2e to the glucose analogue 2b as a carbonyl
source and/or xylene to the mixed solvent (1,4-dioxane/
DMA). Replacement of the phenyl group in 1a with a butyl
substituent (1b) gave the carbonylated product 3b in 67%
yield (entry 4). When various substituents with different elec-
tronic properties were introduced into the aromatic ring of
1a, the corresponding bicyclic cyclopentenones 3c-h were
obtained in moderate to high yields (entries 5, 7, 9, 11, 13,
and 15). For enynes 1e-h having an electron-withdrawing
group, a less polar solvent, xylene, was more suitable for this
reaction than a polar mixed solvent, 1,4-dioxane/DMA,
because the latter solvent accelerated the dimerization of
enynes.7 Enyne 1i, containing a 1,1-disubstituted alkene unit,
reacted slowly to give 3i in 23% yield, along with 53% of the
starting material (entry 17). The reaction of enyne 1j, which is
easily prepared from malonic acid ester, also proceeded
slowly to yield 23% of 3j with 56% of unreacted 1j after
40 h (entry 19). The N-tosylamide-tethered enyne 1k reacted
relatively smoothly to afford 3k in 54% yield (entry 21).
We further studied an enantioselective reaction using
aldose derivatives as a carbonyl source. For the former
(7) (a) Oh, C. H.; Sung, H. R.; Jung, S. H.; Lim, Y. M. Tetrahedron Lett.
2001, 42, 5493–5495. (b) Yamamoto, Y.; Kuwabara, S.; Ando, Y.; Nagata,
H.; Nishiyama, H.; Itoh, K. J. Org. Chem. 2004, 69, 6697–6705. (c) Chopade,
P. R.; Louie, J. Adv. Synth. Catal. 2006, 348, 2307–2327. (d) Shibata, T.;
Otomo, M.; Tahara, Y.; Endo, K. Org. Biomol. Chem. 2008, 6, 4296–4298.
(8) The use of H2O or 1-butanol as a solvent, which can dissolve at the
reaction temperature, afforded no carbonylated product.
(9) For the first step, see: (a) Robert, S. W.; Rainier, J. D. Org. Lett. 2007,
9, 2227–2230. For the second step, see: (b) Sim, M. M.; Kondo, H.; Wong,
C. H. J. Am. Chem. Soc. 1993, 115, 2260–2267.
(11) For the synthesis of 2c and 2d, see ref 9. For the synthesis of 2e and 2f,
see: Itoh, T.; Takamura, H.; Watanabe, K.; Araki, Y.; Ishido, Y. Carbohydr.
Res. 1986, 156, 241–246.
^
(10) Kim, D. E.; Kim, I. S.; Ratovelomanana-Vidal, V.; Genet, J.-P.;
Jeong, N. J. Org. Chem. 2008, 73, 7985-7989.
6280 J. Org. Chem. Vol. 75, No. 18, 2010

Ikeda et al.
JOCNote
TABLE 1. Rh(I)-Catalyzed Cyclocarbonylation of Various Enynes Using Aldose Derivativesa
aConditions: enyne (0.50 mmol), acetyl-masked aldose (1.0 mmol), [RhCl(cod)]2 (0.025 mmol), rac-BINAP (0.050 mmol), and solvent (2 mL) at
c
130 °C for 40 h in a sealed tube. b(S)-BINAP was used as a ligand. Isolated yield. The value in parentheses is the yield of the recovered enyne.
dEnantiomeric excess and absolute configuration were determined by HPLC using chiral stationary columns and specific optical rotation using a
polarimeter, respectively.
reaction using racemic BINAP as a ligand, no enantioselec-
tivity was observed in the formation of 3a. Thus, the chi-
ralities of 2b and the decarbonylated residue do not affect the
enantioselectivity of the reaction. In the reaction of 1a with
2b, the use of either (S)- or (R)-BINAP led to enantioselective
cyclocarbonylation to afford 3a in the same chemical yields
and moderate enantiomeric excesses: for (S)-BINAP, 44 and
59% ee (S);for(R)-BINAP, 44 and 56% ee (R). The reaction of
1a with 0.2 or 2.0 equiv of D-glucose in the presence of 8 mol %
of RhCl(CO)((S)-BINAP)12 in diglyme at 160 °C for 12 h
resulted in a complete consumption of 1a to give a messy
mixture, including less than 10% yield of the cyclocarbonylated
product 3a without enantioselectivity. In all of the reactions
examined using (S)-BINAP, the cyclocarbonylation proceeded
slightly less efficiently than when racemic BINAP was used,
although with moderate to high enantioselectivity, to give the
corresponding bicyclic cyclopentenones (entries 2, 4, 6, 8, 10,
12, 14, 16, 18, 20, and 22). For the reactions of less reactive
enynes, such as entries 18 and 20, higher catalyst loading
(10 mol % of [RhCl(cod)]2 and 20 mol % of (S)-BINAP)
had almost no remarkable effect on the reactivities and
selectivities.13 It would be caused by the deactivation of the
catalysts by the high reducing ability of the aldose.14
A reaction pathway for the present cyclocarbonylation is
proposed, which consists of two rhodium-catalyzed pro-
cesses, as follows: the decarbonylation of the masked aldose
in the acyclic aldehyde form leading to the formation of a
rhodium-carbonyl species,15 and the subsequent carbony-
lation of the enyne utilizing the resulting carbonyl moiety
(Scheme 3). Although, in general, the acyclic aldehyde form
is present at extremely low concentrations in an equilibrium
mixture of the dissolved aldoses,16 the rhodium catalyst is
able to capture it efficiently and to consequently donate the
carbonyl moiety in the formyl group to the enyne. On the
basis of our previous studies,2a,k we postulate that the
(13) For entry 18, 29% yield (32% recovery) and 86% ee; for entry 20,
31% yield (29% recovery) and 71% ee.
(14) Madsen has described the similar discussion on the rhodium-
catalyzed decarbonylation of aldoses in the previous report. See ref 6e.
(15) At present, we have failed unfortunately to recover the decarbony-
lated residue.
(16) For example, the acyclic aldehyde form of D-glucose exists only
0.019% in aqueous solution. See: Maple, R. R.; Allerhand, A. J. Am. Chem.
Soc. 1987, 109, 3168–3169.
€
(12) Bunten, K. A.; Farrer, D. H.; Poe, A. J.; Lough, A. Organometallics
2002, 21, 3344–3350. synthesis of RhCl(CO((S)-BINAP.
J. Org. Chem. Vol. 75, No. 18, 2010 6281

Ikeda et al.
JOCNote
SCHEME 3. Plausible Reaction Pathway
carbonyl moiety is transferred directly from the decarbony-
lation process to the cyclocarbonylation process, resulting in
a CO gas-free cyclocarbonylation of the enyne.
freeze-pump-thaw method and sealed under N2, and the
mixture was stirred at room temperature for 20 min and then
at 130 °C for 40 h. The reaction was analyzed by GC. The
reaction mixture was concentrated in vacuo, and the residue was
purified by silica-gel column chromatography (eluent; hexane/
AcOEt = 4/1) to give bicyclic cyclopentenone 2a (Rf 0.31, 55.0
mg, 0.28 mmol) in 55% yield as colorless oil:2a 1H NMR
(CDCl3) δ 2.34 (dd, J = 3.7 Hz, J = 18 Hz, 1H), 2.85 (dd,
J = 6.1 Hz, J = 18 Hz, 1H), 3.24 (dd, J = 7.9 Hz, J = 12 Hz,
1H), 3.29-3.37 (m, 1H), 4.37 (t, J = 7.9 Hz, 1H), 4.59 (d, J = 17
Hz, 1H), 4.94 (d, J = 17 Hz, 1H), 7.33-7.55 (m, 5H); 13C NMR
(CDCl3) δ 40.3, 43.3, 66.3, 71.3, 128.0, 128.0, 128.5, 128.6, 130.6,
134.7, 177.4, 206.8.
In conclusion, we report on the development of a new
method for the cyclocarbonylation of enynes using readily
available aldose derivatives. This represents a novel utiliza-
tion of aldoses as a carbon monoxide equivalent.2l Various
aldoses, such as D-glucose, D-galactose, D-mannose, D-xy-
lose, and D-ribose, can be used in the present method. The
13C labeling experiment verifies that the anomeric carbon of
aldoses is exclusively introduced into products under the
conditions of the rhodium catalysis. Thus, the reaction
involves the catalytic decarbonylation of the acyclic alde-
hyde forms of aldoses as a key step. The use of chiral BINAP
as a ligand to the rhodium catalyst led to the asymmetric
formation of various bicyclic cyclopentenones.
Acknowledgment. This work was financially supported, in
part, by a Grant-in-Aid for Scientific Research from MEXT
and JSPS, and a Sasagawa Scientific Research Grant from
The Japan Science Society. We thank Ms. Yoshiko Nishi-
kawa for assistance in obtaining HRMS.
Experimental Section
Typical Procedure. In a 7 mL screw-capped tube were placed
[RhCl(cod)]2 (12.3 mg, 0.025 mmol), BINAP (31.8 mg, 0.050
mmol), 2,3,4,6-tetraacetyl-D-glucose (2b) (348.3 mg, 1.0 mmol),
enyne 1a (86.1 mg, 0.50 mmol), and 1,4-dioxane/N,N-dimethy-
lacetamide (2.0 mL, v/v = 1/1). The tube was degassed using
Supporting Information Available: General experimental
methods, additional experimental procedures, and compound
characterization data. This material is available free of charge
via the Internet at http://pubs.acs.org.
6282 J. Org. Chem. Vol. 75, No. 18, 2010