Iridium-Catalyzed Dynamic Kinetic Isomerization: Expedient Synthesis of Carbohydrates from Achmatowicz Rearrangement Products

Article abstract of DOI:10.1002/anie.201503151

A highly stereoselective dynamic kinetic isomerization of Achmatowicz rearrangement products was discovered. This new internal redox isomerization provided ready access to key intermediates for the enantio- and diastereoselective synthesis of a series of naturally occurring sugars. The nature of the de novo synthesis also enables the preparation of both enantiomers.

Full text of DOI:10.1002/anie.201503151

.
Angewandte
Communications
International Edition: DOI: 10.1002/anie.201503151
German Edition: DOI: 10.1002/ange.201503151
Carbohydrate Synthesis
Iridium-Catalyzed Dynamic Kinetic Isomerization: Expedient
Synthesis of Carbohydrates from Achmatowicz Rearrangement
Products**
Hao-Yuan Wang, Ka Yang, Scott R. Bennett, Sheng-rong Guo,* and Weiping Tang*
Abstract: A highly stereoselective dynamic kinetic isomer-
ization of Achmatowicz rearrangement products was discov-
ered. This new internal redox isomerization provided ready
access to key intermediates for the enantio- and diastereose-
lective synthesis of a series of naturally occurring sugars. The
nature of the de novo synthesis also enables the preparation of
both enantiomers.
B
iogenic furans are considered to be among the most
promising sustainable raw materials for the production of fine
chemicals.[1] Besides the defunctionalization methods devel-
oped for the synthesis of simple chemicals from furans,[2] the
Achmatowicz rearrangement occupies a unique position as it
efficiently converts a feedstock furan 1 into a much more
structurally complex dihydropyranone 2 (Scheme 1a).[3] The
potential of the Achmatowicz rearrangement for the de novo
synthesis of carbohydrates was immediately recognized after
its initial discovery in the 1970s.[4] Later, the research groups
of Feringa[5] and OꢀDoherty[6] reported that esters of hemi-
acetal 2 could undergo palladium-catalyzed stereospecific
allylic alkylation, thus establishing a unique glycosidation
method.[7] The Achmatowicz rearrangement has also been
applied to the synthesis of diverse libraries[8] and complex
natural products,[9] including hexacyclic harringtonolide, the
synthesis of which we recently completed.[10]
Scheme 1. Achmatowicz rearrangement and dynamic kinetic isomer-
ization.
ulation of hemiacetal and ketone groups in 2, such as
protection or oxidation of the hemiacetal and reduction of
the ketone group.[14] However, most transformations involv-
ing the anomeric center are not highly stereoselective.
We herein report a highly stereoselective synthesis of
lactones 3 from the epimeric mixtures of Achmatowicz
rearrangement products 2 through an internal redox isomer-
ization process (Scheme 1b), and its application to the
synthesis of naturally occurring sugars.[15] The protection of
the hemiacetal and reduction of the ketone were accom-
plished in one step. More importantly, a dynamic kinetic
stereoselective transformation of both epimers of 2 into
product 3 was possible because two prerequisites were
fulfilled: the rate of equilibration between the trans and cis
hemiacetals 2 was much faster than the isomerization, and the
rate of redox isomerization was faster for one hemiacetal than
for the other.
Because of the importance of the Achmatowicz rear-
rangement, a variety of oxidation conditions have been
developed for the conversion of 1 into 2,[11] including recently
reported practical catalysts based on vanadium[12] and mono-
oxygenase.[13] Efforts have also been devoted to the manip-
[*] H.-Y. Wang, K. Yang, S. R. Bennett, Prof. Dr. S.-r. Guo,
Prof. Dr. W. Tang
Substrate 2a was prepared from acetylfuran by reduction
and Achmatowicz rearrangement. The hemiacetal exists as
a mixture of two epimers, and the diastereomeric ratio is
around 3:1 in favor of the cis isomer (Table 1). We first
screened a number of transition-metal catalysts, such as
[{Ru(cod)Cl2}n], [Cp*Ru(CH3CN)3], [{Rh(cod)Cl}2], [Rh-
(cod)]BF4, [Ir(cod)]BF4, [{Ir(cod)Cl}2], Pd(OAc)2, [Pd2-
(dba)3], [NiCl2{P(C6H11)3}2], AgOAc, and CuBr. Interestingly,
only iridium-based catalysts provided the desired isomer-
ization product 3a. Cationic [Ir(cod)]BF4 only yielded a trace
amount of product 3a. We obtained compound 3a in nearly
quantitative yield by using a catalytic amount of the complex
[{Ir(cod)Cl}2] (Table 1, entry 1). The diastereomeric ratio of
product 3a was also around 3:1 in favor of the cis isomer, thus
indicating that the rate of iridium-catalyzed isomerization is
much faster than the equilibration of the two epimers. The
School of Pharmacy, University of Wisconsin-Madison
Madison, WI 53705-2222 (USA)
E-mail: wtang@pharmacy.wisc.edu
Prof. Dr. S.-r. Guo
Department of Chemistry, Lishui University
Lishui, Zhejiang Province, 331200 (P.R. China)
E-mail: guosr9609@lsu.edu.cn
Prof. Dr. W. Tang
Department of Chemistry, University of Wisconsin-Madison
Madison, WI 53706 (USA)
[**] We thank the University of Wisconsin (UW) for funding and SAFC
Inc. for the generous donation of the Ir catalyst. S.-r.G. thanks Lishui
University for financial support of a visiting scholar position at UW.
We thank rotation students Hao Li, Jiaxuan Yan, and Minsoo Ju for
their help with some of the experiments.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201503151.
8756
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 8756 –8759

Angewandte
Chemie
Table 1: Screening of conditions for the isomerization of 2a to 3a.
Table 2: Scope of the iridium(I)-catalyzed dynamic kinetic isomeriza-
tion.[a]
Entry
Substrate
Product (d.r.)
Yield [%][b]
Entry Reaction conditions
Yield [%][a] d.r.[a]
1
2
3
2a, R=Me (d.r. 3:1)
2b, R=H
2c, R=nPr
2d, R=iPr
2e, R=tBu
2 f, R=cyclopropyl (d.r. 3:1)
2g, R=CH2OTBS
2h, R=phenyl (d.r. 5:1)
2i, R=p-FC6H4
2j, R=p-MeOC6H4
2k, R=p-CF3C6H4
2l, R=o-CH3C6H4
3a (>20:1)
3b
91
96
93
90
63
94
84
93
95
88
60
75
1
2
DCE, 508C
C6H4CO2H (50 mol%), DCE, 508C
96
97
3:1
12:1
9:1
3c (>20:1)
3d (>20:1)
3e (>20:1)
3 f (>20:1)
3g (>20:1)
3h (>20:1)
3i (>20:1)
3j (>20:1)
3k (>20:1)
3l (>20:1)
4
3
4
p-NO2C6H4CO2H (50 mol%), DCE, 508C 99
p-CH3C6H4CO2H (50 mol%), DCE, 508C 97
5[c]
6
10:1
14:1
14:1
14:1
16:1
14:1
18:1
>20:1
20:1
5
6
7
8
C6H4CO2H (50 mol%), DCE, RT
C6H4CO2H (100 mol%), DCE, RT
C6H4CO2H (5 mol%), DCE, RT
CH3CO2H (50 mol%), DCE, RT
CF3CO2H (50 mol%), DCE, RT
2,6-Cl2C6H3CO2H (50 mol%), DCE, RT
97
88
86
89
82
99
7[d]
8
9
10
11[e]
12
9
10
11
12
2,6-Cl2C6H3CO2H (50 mol%), CHCl3, RT 98
2,6-Cl2C6H3CO2H (50 mol%), CH2Cl2, RT 87
[a] Reactions conditions: [{Ir(cod)Cl}2] (2.5 mol%), CHCl3, room tem-
perature, 8–12 h, unless noted otherwise. [b] Yield of the isolated
product. [c] Substrate 2e was used without purification, and 63% was
the yield after three steps from furan and pivalaldehyde. [d] The reaction
was carried out at 508C. [e] Substrate 2k was used without purification,
and 60% was the yield after three steps from furan and 4-(trifluorome-
thyl)benzaldehyde. TBS=tert-butyldimethylsilyl.
[a] Yields and diastereomeric ratios (cis/trans) were calculated on the
basis of 1H NMR spectroscopy with CH2Br2 as the internal standard.
cod=1,5-cyclooctadiene, DCE=dichloroethane.
configuration of 3a and its trans isomer was assigned by
comparing their spectra with previously reported spectral
data.[16]
a number of carbohydrate lactones, such as gulonolactone and
allonolactone.[17] Neither electron-withdrawing nor electron-
donating groups on the aryl group impacted the yield
significantly (Table 2, entries 8–10). The trifluoromethyl-sub-
stituted substrate 2k was not stable and needed to be used
immediately after its preparation (Table 2, entry 11). An
ortho-substituted phenyl group was also compatible with the
transformation (Table 2, entry 12). In all cases, the cis isomer
was observed exclusively.
We next screened different Brønsted acids in an attempt
to accelerate the rate of equilibration between epimers cis-2a
and trans-2a (Table 1, entries 2–12). To our delight, the cis/
trans ratio immediately increased to 12:1 upon the addition of
benzoic acid (50 mol%; Table 1, entry 2). This result clearly
indicates that the rate of equilibration between the two
epimeric hemiacetals is accelerated by an acid additive, and
that a dynamic kinetic transformation is possible. The use of
benzoic acids with either an electron-withdrawing nitro group
or an electron-donating methyl group led to a slight decrease
in the diastereomeric ratio (Table 1, entries 3 and 4). The
diastereomeric ratio was increased slightly by carrying out the
reaction at room temperature (Table 1, entry 5). The yield
was slightly lower when the amount of benzoic acid was
increased to 100 mol% or decreased to 5 mol%, but the
diastereomeric ratio remained the same (Table 1, entries 6
and 7). By screening several other carboxylic acids and
different solvents (Table 1, entries 8–12), we found that the cis
isomer could be obtained exclusively in nearly quantitative
yield when 2,6-dichlorobenzoic acid was used as the cocata-
lyst (Table 1, entry 11).
A gem-dimethyl substituent was also tolerated in the
isomerization reaction (Scheme 2). Substrate 5 was prepared
in two steps from commercially available 2-acetylfuran (4).
The scope of the iridium-catalyzed redox isomerization is
shown in Table 2. Product 3a was isolated in 91% yield
(Table 2, entry 1) when the reaction was carried out under the
optimized conditions given in entry 11 of Table 1. The R
group in substrate 2 can also be a hydrogen atom (Table 2,
entry 2). Product 3e with a tert-butyl substituent was isolated
in 63% yield over three steps from furan and pivalaldehyde
(Table 2, entry 5). Substrate 2e was not stable and was
subjected to isomerization immediately after its preparation
by the addition of furanyl lithium to pivalaldehyde and
subsequent Achmatowicz rearrangement. Cyclopropyl and
silyl ether groups were also tolerated (Table 2, entries 6 and
7). Product 3g is a key intermediate in the synthesis of
Scheme 2. Formal synthesis of noviose: a) MeMgBr; b) NBS, 66%
over 2 steps; c) [{Ir(cod)Cl}2] (2.5 mol%), 508C, 89%. NBS=
N-bromosuccinimide.
The isomerization product 6 has been converted into noviose
in three steps through methylation, reduction, and dihydrox-
ylation.[18] The combination of the Achmatowicz rearrange-
ment and iridium-catalyzed isomerization provides a concise
de novo synthesis of noviose.[19]
The secondary alcohol 8 could be prepared in 98% yield
with 98% ee according to known protocols (Scheme 3).[20]
Angew. Chem. Int. Ed. 2015, 54, 8756 –8759
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
8757

.
Angewandte
Communications
from which products 3a and 20, respectively, are formed upon
hydrogenation. Overall, this transformation is a rare example
of stereoselective internal transfer hydrogenation. In the
absence of an acid, the ratio of 3a to 20 is around 3:1, which is
similar to the ratio of cis-2a to trans-2a. This observation
suggests that the rate of equilibration between the two
hemiketals is slower than the rate of internal transfer
hydrogenation in the absence of an acid. The rate of
interconversion between the two hemiacetals became signifi-
cantly faster than the internal transfer hydrogenation in the
presence of an acid additive, thus making the stereoselective
dynamic kinetic isomerization possible.
In summary, we have discovered a novel iridium-catalyzed
stereoselective dynamic kinetic internal transfer hydrogena-
tion reaction. This new method provides a practical and
unified approach to the synthesis of deoxy- and amino sugars
from simple furan derivatives.
Scheme 3. Catalytic asymmetric synthesis of deoxysugars:
a) [{Cp*RhCl2}2] (0.05 mol%), (R,R)-TsDPEN (0.12 mol%), HCO2Na,
408C, 98%; b) NBS, 85%; c) [{Ir(cod)Cl}2] (2.5 mol%), room temper-
ature, 76%; d) H2, Pd/C; e) DIBAL-H, 40% over 2 steps; f) PPh3,
p-NO2C6H4CO2H, DEAD, then MeOH, Et3N, 75% over 2 steps.
Cp*=1,2,3,4,5-pentamethylcyclopentadienyl, DEAD=diethyl azodicar-
boxylate, DIBAL-H=diisobutylaluminum hydride, TsDPEN=N-p-tosyl-
1,2-diphenylethylenediamine.
Keywords: carbohydrates · deoxysugars · furans · iridium ·
isomerization
How to cite: Angew. Chem. Int. Ed. 2015, 54, 8756–8759
Angew. Chem. 2015, 127, 8880–8883
The stereochemical integrity was retained under the isomer-
ization conditions in a gram-scale synthesis. Reduction of the
key intermediate 10 completed the enantioselective synthesis
of l-rhodinose (11).[21] The same intermediate has been
converted into l-oliose (12) in three steps.[16] Mitsunobu
inversion of allylic alcohol 10, followed by hydrolysis,[14a]
afforded osmundalactone (13).[16,22] The synthesis of l-ami-
cetose (14), l-digitoxose (15), and l-canarose (16) from
intermediate 13 has been reported previously.[14b,23] Inter-
mediates 10 and 13 have also been converted into amino sugar
derivatives, such as daunosamine and ristoamine.[24]
[1] A. J. Ragauskas, C. K. Williams, B. H. Davison, G. Britovsek, J.
Cairney, C. A. Eckert, W. J. Frederick, J. P. Hallett, D. J. Leak,
C. L. Liotta, J. R. Mielenz, R. Murphy, R. Templer, T. Tscha-
plinski, Science 2006, 311, 484.
[2] For selected reviews, see: a) A. Corma, S. Iborra, A. Velty,
Chem. Rev. 2007, 107, 2411; b) M. Besson, P. Gallezot, C. Pinel,
Chem. Rev. 2014, 114, 1827; c) A. J. J. Straathof, Chem. Rev.
2014, 114, 1871; d) R. A. Sheldon, Green Chem. 2014, 16, 950.
[3] O. Achmatowicz, P. Bukowski, B. Szechner, Z. Zwierzch, A.
Zamojski, Tetrahedron 1971, 27, 1973.
[4] For selected examples, see: a) O. Achmatowicz, Jr., P. Bukowski,
Can. J. Chem. 1975, 53, 2524; b) O. Achmatowicz, B. Szechner,
Carbohydr. Res. 1976, 50, 23; c) R. Bognµr, P. Herczegh,
Carbohydr. Res. 1976, 52, 11; d) N. L. Holder, Chem. Rev. 1982,
82, 287; e) M. Takeuchi, T. Taniguchi, K. Ogasawara, Synthesis
1999, 341.
A trace amount of ketolactone 17[25] was isolated when the
isomerization was conducted on a larger scale (Scheme 4).
This intermediate decomposes quickly at room temperature.
A mechanism for the dynamic kinetic isomerization is
proposed in Scheme 4. Thus, the dehydrogenation of cis-2a
and trans-2a is thought to afford metal complexes 18 and 19,
[5] A. C. Comely, R. Eelkema, A. J. Minnaard, B. L. Feringa, J. Am.
Chem. Soc. 2003, 125, 8714.
[6] a) R. S. Babu, G. A. OꢀDoherty, J. Am. Chem. Soc. 2003, 125,
12406; b) R. S. Babu, M. Zhou, G. A. OꢀDoherty, J. Am. Chem.
Soc. 2004, 126, 3428.
[7] For recent reviews on transition-metal-catalyzed glycosylation,
see: a) X. Li, J. Zhu, J. Carbohydr. Chem. 2012, 31, 284; b) M. J.
McKay, H. M. Nguyen, ACS Catal. 2012, 2, 1563.
[8] a) M. D. Burke, E. M. Berger, S. L. Schreiber, Science 2003, 302,
613; b) M. D. Burke, E. M. Berger, S. L. Schreiber, J. Am. Chem.
Soc. 2004, 126, 14095.
[9] For recent reviews, see: a) K. E. O. Ylijoki, J. M. Stryker, Chem.
Rev. 2013, 113, 2244; b) T. V. Nguyen, J. M. Hartmann, D.
Enders, Synthesis 2013, 45, 845.
[10] M. Zhang, N. Liu, W. Tang, J. Am. Chem. Soc. 2013, 135, 12434.
[11] P. Merino, T. Tejero, J. I. Delso, R. Matute, Curr. Org. Chem.
2007, 11, 1076.
[12] Y. Ji, T. Benkovics, G. L. Beutner, C. Sfouggatakis, M. D.
Eastgate, D. G. Blackmond, J. Org. Chem. 2015, 80, 1696.
[13] D. Thiel, D. Doknic, J. Deska, Nat. Commun. 2014, 5, 5278.
[14] a) J. M. Harris, G. A. OꢀDoherty, Tetrahedron Lett. 2000, 41, 183;
b) L. Z. Zhu, A. Talukdar, G. S. Zhang, J. P. Kedenburg, P. G.
Wang, Synlett 2005, 1547.
Scheme 4. Proposed mechanism for the iridium-catalyzed isomeriza-
tion.
8758
www.angewandte.org
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 8756 –8759

Angewandte
Chemie
[15] For selected applications of lactone 3 or its derivatives in natural
product synthesis, see: a) J. M. Harris, G. A. OꢀDoherty, Org.
Lett. 2000, 2, 2983; b) J. M. Harris, G. A. OꢀDoherty, Tetrahedron
2001, 57, 5161; c) J. M. Harris, G. A. OꢀDoherty, Tetrahedron
Lett. 2002, 43, 8195; d) M. S. Li, J. Scott, G. A. OꢀDoherty,
Tetrahedron Lett. 2004, 45, 1005; e) M. S. Li, G. A. OꢀDoherty,
Tetrahedron Lett. 2004, 45, 6407; f) C. P. Burke, N. Haq, D. L.
Boger, J. Am. Chem. Soc. 2010, 132, 2157; g) C. P. Burke, M. R.
Swingle, R. E. Honkanen, D. L. Boger, J. Org. Chem. 2010, 75,
7505.
[20] a) X. Wu, X. Li, A. Zanotti-Gerosa, A. Pettman, J. Liu, A. J.
Mills, J. Xiao, Chem. Eur. J. 2008, 14, 2209; b) M. P. Croatt, E. M.
Carreira, Org. Lett. 2011, 13, 1390.
[21] R. H. Schlessinger, D. D. Graves, Tetrahedron Lett. 1987, 28,
4381.
[22] K. H. Hollenbeak, M. E. Kuehne, Tetrahedron 1974, 30, 2307.
[23] L. H. Zhu, J. P. Kedenburg, M. Xian, P. G. Wang, Tetrahedron
Lett. 2005, 46, 811.
[24] Y. Matsushima, J. Kino, Eur. J. Org. Chem. 2010, 2206.
[25] M. P. Georgiadis, S. A. Haroutounian, C. D. Apostolopoulos,
Synthesis 1991, 379.
[16] G. Zhang, L. Shi, Q. Liu, J. Wang, L. Li, X. Liu, Tetrahedron
2007, 63, 9705.
[17] J. M. Harris, M. D. Keranen, H. Nguyen, V. G. Young, G. A.
OꢀDoherty, Carbohydr. Res. 2000, 328, 17.
[18] B. Schmidt, S. Hauke, Eur. J. Org. Chem. 2014, 1951.
[19] For recent syntheses of noviose, see: a) S. Hanessian, L. Auzzas,
Org. Lett. 2008, 10, 261; b) Y. Matsushima, J. Kino, Synthesis
2011, 1290.
Received: April 6, 2015
Published online: June 1, 2015
Angew. Chem. Int. Ed. 2015, 54, 8756 –8759
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
8759