Anodic coupling reactions and the synthesis of C-glycosides

Article abstract of DOI:10.1021/ol100800u

A convenient, two-step procedure has been developed for converting sugar derivatives into C-glycosides containing a masked aldehyde functional group. The chemistry takes advantage of an anodic coupling reaction between an electron-rich olefin and an alcohol. The sequence works for the formation of both furanose and pyranose derivatives if less polarized vinyl sulfide derived radical cation intermediates are used. With more polarized enol ether derived radical cations, the cyclizations work best for the formation of furanose derivatives where the rate of five-membered ring formation precludes elimination reactions triggered by the radical cation.

Full text of DOI:10.1021/ol100800u

ORGANIC
LETTERS
2010
Vol. 12, No. 11
2590-2593
Anodic Coupling Reactions and the
Synthesis of C-Glycosides
Guoxi Xu and Kevin D. Moeller*
Department of Chemistry, Washington UniVersity, St. Louis, Missouri 63130
moeller@wustl.edu
Received April 7, 2010
ABSTRACT
A convenient, two-step procedure has been developed for converting sugar derivatives into C-glycosides containing a masked aldehyde
functional group. The chemistry takes advantage of an anodic coupling reaction between an electron-rich olefin and an alcohol. The sequence
works for the formation of both furanose and pyranose derivatives if less polarized vinyl sulfide derived radical cation intermediates are used.
With more polarized enol ether derived radical cations, the cyclizations work best for the formation of furanose derivatives where the rate of
five-membered ring formation precludes elimination reactions triggered by the radical cation.
In connection with efforts to assemble molecular libraries
that can be monitored in “real-time” for their binding to
biological receptors, we have been developing techniques
for locating molecules next to individually addressable
electrodes in a microelectrode array.1 This is accomplished
by coating the arrays with a functionalized porous reaction
layer2 and then using the electrodes in the arrays to make
chemical reagents and catalysts that trigger chemical reac-
tions on the surface of the array. The reactions are used to
couple substrates in the solution above the array to functional
groups on the porous reaction layer proximal to the elec-
trodes. To date, “site-selective” esterification reactions,1 Heck
reactions,3 Suzuki reactions,4 click-reactions,5 hetero-Michael
reactions,1 reductive amination reactions,6 and Diels-Alder
reactions7 have all been used for this purpose. The chemistry
has led to a variety of substrates being placed on the arrays
and the completion of initial experiments probing the utility
of the arrays as analytical tools.1,8 Because of the key role
carbohydrates play in many biological contexts, these efforts
are now being extended to the placement of sugars and
glycopeptides on a microelectrode array.
One of the first steps in achieving this aim is to
synthesize sugar derivatives that can be used to capitalize
on the synthetic methodology mentioned above. To this
end, C-glycosides like 1 and 2 (Figure 1) appear ideal.
(1) For an example, see: Stuart, M.; Maurer, K.; Moeller, K. D.
Bioconjugate Chem. 2008, 19, 1514.
Figure 1. C-Glycoside targets.
(2) For the development of a new reaction layer, see: Hu, L.; Bartles,
J. L.; Bartles, J. W.; Maurer, K.; Moeller, K. D. J. Am. Chem. Soc. 2009,
131, 16638.
(3) (a) Tian, J.; Maurer, K.; Tesfu, E.; Moeller, K. D. J. Am. Chem.
Soc. 2005, 127, 1392. (b) Tang, R.; Chen, C.; Moeller, K. D. Synthesis
2007, 3411.
C-Glycosides such as 1 and 2 have been used in the
synthesis of bioactive sugar derivatives,9 and the aldehyde
(4) Hu, L.; Maurer, K.; Moeller, K. D. Org. Lett. 2009, 11, 1273.
(5) Bartles, J. L.; Lu, P.; Walker, A.; Maurer, K.; Moeller, K. D. Chem.
Commun. 2009, 5573.
(6) Tesfu, E.; Maurer, K.; Moeller, K. D. J. Am. Chem. Soc. 2006, 128,
70.
10.1021/ol100800u  2010 American Chemical Society
Published on Web 05/12/2010

group bound to the former anomeric carbon can be readily
converted into a wide variety of substrates for microelec-
trode-array reactions.
to anodic electrolysis condiditons.14 The first conditions tried
(entry 1) for the electrolysis were the optimized ones
developed for earlier enol ether-alcohol coupling reactions.15
The reaction led to only a low yield of product along with
recovered starting material. This was true even after 10 F
(5× the stoichiometric amount) of current had been passed.
The reaction could be optimized by first moving toward a
more polar solvent and then taking advantage of lithium
perchlorate as the electrolyte for the electrolysis. Under these
conditions, an 85% isolated yield of the desired cyclic
product 7 was obtained as a 1.5:1 ratio of diastereomers about
the newly formed bond. The major diastereomer had the
acetal group trans to the neighboring methoxy at C3 of the
ring. The formation of diastereomers was not a concern since
the center generated is epimerizable following hydrolysis of
the acetal. What was clear from the cyclization was that
electrolysis was compatible with the highly functionalized
substrate. The need for the more polar solvent and lithium
perchlorate electrolyte was rationalized by considering the
very polar nature of the reaction substrate. Electrolytes play
a key role in defining the nature of the reaction environment
surrounding an electrode.16 A “greasy” electrolyte will
exclude polar species relative to nonpolar ones. Hence with
tetraethylammonium tosylate and THF present, the very polar
substrate 6 might have difficulty approaching the electrode
surface leading to background oxidation of the methanol
solvent. As the polarity of the environment surrounding the
electrode is increased, more of the substrate can reach the
electrode surface and the efficiency of the reaction improves.
Support for this rationalization was gained when substrate
8 was studied (Table 2). In this case, a significant change in
Previous syntheses of C-glycosides such as 1 or 2 involved
either an initial conversion of a sugar to a cyano derivative
followed by reduction and hydrolysis10 or longer multistep
routes.11 As a complementary alternative, it is tempting to
suggest that C-glycosides can be rapidly assembled from
existing sugar derivatives by taking advantage of a Wittig
reactionsanodic oxidation sequence such as the one il-
lustrated in Scheme 1.12,13 But are anodic coupling reactions
Scheme 1
and the highly reactive radical cation intermediates generated
really compatible with such highly functionalized substrates?
To provide an initial answer for this question, a model
tetramethoxyfuranose substrate was selected (Table 1). To
Table 1. Initial Cyclizations
Table 2. Benzyl-Protected Substrate
entry electrolyte
solvent (v/v)
coulomb (F) yield (%)
1
2
3
4
Et4NOTs
Et4NOTs
Et4NOTs
LiClO4
MeOH/THF (3:7)
MeOH/THF (6:4)
MeOH
10.0
5.5
5.5
22
30
40
MeOH
2.6
85a
a A 1.5:1 ratio of diastereomers was formed.
entry electrolyte
solvent (v:v)
coulomb (F/mol) yield (%)
1
2
3
4
LiClO4
LiClO4
LiClO4
MeOH/THF (3:7)
MeOH/THF (6:4)
MeOH
5.5
3.6
3.0
3.4
54
60
this end, the trimethoxy sugar derivative was treated with
an ylide to form methoxy enol ether 6, and then 6 subjected
62a
50
Et4NOTs MeOH
a A 2:1 ratio of diastereomers was formed.
(7) Bi, B.; Maurer, K.; Moeller, K. D. Angew. Chem., Int. Ed. 2009,
48, 5872.
(8) Tesfu, E.; Roth, K.; Maurer, K.; Moeller, K. D. Org. Lett. 2006, 8,
709.
the oxygen-carbon balance of the substrate was made by
changing from methyl to benzyl protecting groups on the
oxygens. While the lithium perchlorate in methanol condi-
(9) Dondoni, A.; Mauer, A. Chem. ReV. 2000, 100, 4395.
(10) (a) Araki, S.; Morita, K.; Yamaguchi, M.; Zhao, Z.; Wilson, T. J.;
Lilley, D. M. J.; Harusawa, S. J. Org. Chem. 2009, 74, 2350. (b) Albrecht,
H. P.; Repke, D. B.; Moffatt, J. G. J. Org. Chem. 1973, 38, 1836.
(11) (a) Ueda, M.; Yamaura, M.; Ikeda, Y.; Suzuki, Y.; Yoshizato, K.;
Hayakawa, I.; Kigoshi, H. J. Org. Chem. 2009, 74, 3370. (b) Jung, J. H.;
Lee, E. Angew. Chem., Int. Ed. 2009, 48, 5698. (c) Popsavin, M.; Psopsavin,
V.; Vukojevic´, G.; Csana´di, J.; Miljkovic´, D. Carbohydr. Res. 1994, 260,
145. (d) Dondoni, A.; Scherrmann, M.-C. J. Org. Chem. 1994, 59, 6404.
(e) Dhavale, D. D.; Desai, V. N.; Sindkhedkar, M. D.; Mali, R. S.; Castellari,
C.; Trombini, C. Tetrahedron Asymmetry 1997, 8, 1475.
(12) For a review of anodic olefin coupling reactions, see: Moeller, K. D.
(13) For a general review of electrochemical approaches, see: Yoshida,
J.; Kataoka, K.; Horcajada, R.; Nagaki, A. Chem. ReV. 2009, 108, 2265.
(14) See the Supporting Information.
(15) (a) Liu, B.; Duan, S.; Sutterer, A. C.; Moeller, K. D. J. Am. Chem.
Soc. 2002, 124, 10101. (b) Xu, H.-C.; Brandt, J. D.; Moeller, K. D.
Tetrahedron Lett. 2008, 49, 3868.
(16) Synthetic Organic Electrochemistry, 2nd ed.; Fry, A. J., Ed.; John
Wiley and Sons, Inc: New York, 1989; pp 38-42 and 113-114.
Synlett. 2009, 8, 1208
.
Org. Lett., Vol. 12, No. 11, 2010
2591

tions still led to the highest yield in this case, the effect of
the change was not nearly as pronounced, suggesting that
the less polar substrate was better able to approach the
electrode surface when less polar reaction conditions were
used. The yield and the current efficiency for the reactions
using the benzyl protected ribose substrate were lower than
those obtained with 6. Proton NMR data for the crude
reaction mixture suggested that this observation resulted from
background oxidation of the benzyl protecting groups. Once
again, the presence of the oxygens on the ribose backbone
did not interfere with the cyclization.
an oxidative cleavage of the hemithioacetal.17 Overall, the
mixture of acetals was not an issue since the products will
be hydrolyzed to the aldehyde for future development. Hence,
the use of the vinyl sulfide proved desirable because of the
higher yield associated with the substrate synthesis.
With the synthesis of furanose-derived C-glycosides in
place, attention was turned to the synthesis of pyranose
derivatives. Once again, the study began by examining the
anodic oxidation of a polymethoxy ether substrate (Scheme
3). In this case, the anodic oxidation led to none of the
Attention was next turned toward exploring the compat-
ibility of a substrate having the oxygens differentially
protected. Two such substrates were examined, one with an
enol ether as the site for the initial oxidation and the second
with a vinyl sulfide serving in this capacity (Scheme 2). Both
Scheme 3
Scheme 2
desired cyclic product. Instead, a complex mixture of
products was obtained out of which 10% of a product (15)
generated by fragmentation of the initial radical cation was
isolated. Presumably, product 15 arose due to the slower six-
membered ring formation relative to the earlier five-
membered ring formation in the furanose cases.
The effects of the slower six-membered ring cyclization
were also seen when a substrate lacking the C3 and C4
methoxy groups was oxidized (Scheme 4). In this example,
substrates were synthesized from the protected sugar using
a Wittig reaction. Formation of the vinyl sulfide consistently
led to higher yields of electrolysis substrate. The electrolysis
of both substrates proceeded in a similar manner with both
substrates benefiting from the use of LiClO4 as electrolyte
and 30% MeOH/THF as the solvent. With the very hydro-
phobic tert-butyldiphenylsilyl protecting group on the mol-
ecule, the reactions did benefit from the use of THF as a
cosolvent. In the case of methoxy enol ether substrate 11a,
the reaction gave the product in a yield close to 100% when
measured by NMR (coumarin added as an internal standard
to the crude reaction material). The isolated yield was 70%
with the loss of material being attributed to the instability
of the acetal product. The reaction was compatible with the
synthesis of multigram quantities of the product. In this case,
a 2.4:1 ratio of diastereomers about the newly formed
carbon-oxygen bond was formed.
Scheme 4
In the case of the vinyl sulfide substrate 11b, a nearly
identical result was obtained with a 73% isolated yield of
cyclized product being produced in a 2.5:1 ratio of diaster-
eomers about the newly formed carbon-oxygen bond. The
cyclized material was obtained as a mixture of the monothiol
acetal and the dimethoxy acetal product. The thioacetal could
be converted into the dimethoxy acetal using HgCl2, HgO,
and methanol. In our hands, this method worked better than
there was no ether substituent to trigger the fragmentation.
However, the yield of the cyclization was still low, this time
(17) For alternative methods, see: Saito, I.; Takayama, M.; Sakurai, T.
J. Am. Chem. Soc. 1994, 116, 2653. (b) Yoshida, J.; Isoe, S. Chem. Lett.
1987, 16, 631. (c) Yoshida, J.; Sugawara, M.; Tatsumi, M.; Kise, N. J.
Org. Chem. 1998, 63, 5950.
2592
Org. Lett., Vol. 12, No. 11, 2010

due to the elimination and methanol trapping pathways that
are more traditional side reactions for an anodically generated
radical cation. Attempts to optimize the yield of the cycliza-
tion using the same set of conditions shown in Tables 1 and
2 led to a decrease in the yield of the cyclized product and
the overall mass balance of the reaction. In some ways, the
low yield of the cyclization was not a surprise. The yields
obtained for previous anodic cyclizations coupling enol ether
radical cations and alcohols leading to tetrahydropyran rings
benefited from the use of Et4NOTs as the electrolyte and
were sensitive to substituents on the tether connecting the
reactive groups.15a Presumably for these cyclizations, the use
of the Et4NOTs excludes methanol from the region sur-
rounding the electrode providing more time for the cycliza-
tion. With the more polar substrates used in the current study,
this approach proved problematic due to the lower efficiency
of the desired oxidation. What was needed was either a
method for making a less polar substrate so that Et4NOTs
could be used as the electrolyte or a method for accelerating
the cyclization.
site for the initial oxidation was synthesized. The anodic
oxidation of this substrate led cleanly to the cyclized material
(Scheme 5, eq 2). A 68% yield of the desired product was
obtained as a mixture of acetal products. Once again, this
mixture was not important since carrying the product forward
involves hydrolysis of the acetal.
The success of this reaction turned our attention back to
the fully substituted sugar derivative. Did the use of the vinyl
sulfide derived radical cation accelerate the cyclizaton to a
point where the fragmentation could be avoided? To our
delight, the answer to this question is yes (Scheme 6).
Scheme 6
An attempt to improve the cyclization using larger
hydrocarbon protecting groups was not successful (Scheme
5, eq 1). The reaction behaved in a fashion identical to the
When the fully methoxylated pyranose was converted into
a vinyl sulfide and then oxidized at an RVC anode, a 71%
isolated yield of the cyclized product was obtained. Only a
trace amount of fragmentation product 15 was observed.
In conclusion, a two-step Wittig reaction-anodic coupling
reaction strategy provides a very rapid, convenient method
for converting protected furanose and pyranose sugars into
C-glycosides. Currently, work to expand this chemistry to
include transformations involving sugar oligomers and efforts
to place and monitor the behavior of sugars site-selectively
on microelectrode arrays are underway.
Scheme 5
Acknowledgment. We thank the National Science Foun-
dation (CHE-0809142) for their generous support of this
work.
Supporting Information Available: Sample experimental
procedure for the oxidation reaction along with characteriza-
tion data for all new compounds. This material is available
free of charge via the Internet at http://pubs.acs.org.
oxidation of 16. While still larger groups might help in this
regard, the overall approach was abandoned because of the
limitations it placed on substrate generality.
Instead, a strategy to increase the rate of the cyclization
was pursued. Since the trapping of radical cation intermedi-
ates by heteroatoms is accelerated by the use of less polarized
radical cations,18,19 a substrate having a vinyl sulfide as a
OL100800U
(18) (a) Tang, F.; Moeller, K. D. J. Am. Chem. Soc. 2007, 129, 12414.
(b) Tang, F.; Moeller, K. D. Tetrahedron 2009, 65, 10863.
(19) Xu, H.-C.; Moeller, K. D. J. Am. Chem. Soc. 2010, 132, 2839.
Org. Lett., Vol. 12, No. 11, 2010
2593