Palladium(II)-catalyzed rearrangement of glycal trichloroacetimidates: Application to the stereoselective synthesis of glycosyl ureas

Article abstract of DOI:10.1021/ja803378k

The research on the area of glycosyl urea derivatives, in which the O- and N-glycosidic bonds are replaced with the urea-glycosidic linkages, has recently emerged with applications in the field of aminoglycoside antibiotics. We have developed a novel meth

Full text of DOI:10.1021/ja803378k

Published on Web 07/19/2008
Palladium(II)-Catalyzed Rearrangement of Glycal
Trichloroacetimidates: Application to the Stereoselective
Synthesis of Glycosyl Ureas
Gregory J. Mercer, Jaemoon Yang, Matthew J. McKay, and Hien M. Nguyen*
Department of Chemistry and Biochemistry, Montana State UniVersity,
Bozeman, Montana 59717
Received May 6, 2008; E-mail: hmnguyen@chemistry.montana.edu
Abstract: The research on the area of glycosyl urea derivatives, in which the O- and N-glycosidic bonds
are replaced with the urea-glycosidic linkages, has recently emerged with applications in the field of
aminoglycoside antibiotics. We have developed a novel method for the stereoselective synthesis of R- and
ꢀ-glycosyl ureas via Pd(II)-catalyzed rearrangement of glycal trichloroacetimidates. In our approach, the
R- and ꢀ-selectivity at the anomeric carbon of N-glycosyl trichloroacetamides depends on the nature of the
palladium-ligand catalyst. While the cationic Pd(II)-L-4 (2-trifluoroacetylphenol) complex promotes
R-selectivity, the neutral Pd(II)-TTMPP-L-5 (4-chloro-2-trifluoroacetylphenol) complex favors ꢀ-selectivity.
The resulting R- and ꢀ-N-glycosyl trichloroacetamides were further coupled with a diverse array of primary
and hindered secondary nitrogen nucleophiles to provide the corresponding glycosyl ureas in moderate to
good yields and with no loss of stereochemical integrity at the anomeric carbon. We have further
demonstrated the utility of N-glycosyl trichloroacetamides as robust and versatile intermediates in the
synthesis of unsymmetrical urea-linked disaccharides and trisaccharide.
ride chains are susceptible to chemical or enzymatic hydrolysis,7
thus causing the cleavage of O-glycosidic bonds as well as the
Introduction
The glycosyl ureas are found in nature as a structural unit of
the glycocinnamoylspermidine antibiotics.1 In recent years,
research on the area of N-linked-glycopeptide mimics,2 neogly-
coconjugates,3 and pseudooligosaccharides,4 in which the O-
and N-glycosidic bonds are replaced with the urea-glycosidic
linkages, has emerged with the application in the development
of antidiabetic agents5 and aminoglycoside antibiotics.6 The
modification of the structures of naturally occurring carbohy-
drates has received considerable attention because oligosaccha-
degradation of the glycoconjugates. Therefore, it is desirable
to have access to modified structures such as glycosyl ureas
that are more stable under biological conditions and still
maintain the properties of the natural compounds. There are
only a few methods reported, however, for the stereoselective
construction of glycosyl urea,8 and in particular, the stereose-
lective synthesis of R-glycosyl ureas is still limited.2a,3,9
We report herein a novel method for the stereoselective
synthesis of R- and ꢀ-glycosyl ureas via Pd(II)-catalyzed glycal
imidate rearrangement.10 In our approach, the nature of the
palladium-ligand complex controls the anomeric selectivity
(Scheme 1).11 The cationic Pd(II), which promotes ionization
of the glycal imidate 1 by coordinating to the imidate nitrogen,12
results in the formation of R-N-glycosyl trichloroacetamide 2.
In contrast, use of neutral Pd(II) promotes a concerted-type
(1) (a) Ellestad, G. A.; Cosulich, D. B.; Broschard, R. W.; Martin, J. H.;
Kunstmann, M. P.; Morton, G. O.; Lancaster, J. E.; Fulmor, W.; Lovell,
F. M. J. Am. Chem. Soc. 1978, 100, 2515–2524. (b) Dobashi, K.;
Nagaoka, K.; Wantanabe, Y.; Nishida, M.; Hamada, M.; Naganawa,
H.; Takita, T.; Takeuchi, T.; Umezawa, H. J. Antibiot. 1985, 38, 1166–
1170.
(2) (a) Ichikawa, Y.; Nishiyama, T.; Isobe, M. Synlett 2000, 1253–1256.
(b) Ichikawa, Y.; Ohara, F.; Kotsuki, H.; Nakano, K. Org. Lett. 2006,
8, 5009–5012.
(3) Ichikawa, Y.; Nishiyama, T.; Isobe, M. J. Org. Chem. 2001, 66, 4200–
4205.
(7) Pigman, W. In The Carbohydrates, Chemistry and Biochemistry, Part
II A; Horton, D., Ed.; Academic Press: New York, 1970; pp 37-110.
(8) (a) Benn, M. H.; Jones, A. S. J. Chem. Soc. 1960, 3837–3841. (b)
Garc´ıa-Ferna´ndez, J. M.; Ortiz-Mellet, C.; D´ıaz-Pe´rez, V. M.; Fuentes,
J.; Kova´cs, J.; Pinte´r, I. Tetrahedron Lett. 1997, 38, 4161–4164. (c)
Garc´ıa-Moreno, M. I.; Benito, J. M.; Ortiz-Mellet, C.; Garc´ıa-
Ferna´ndez, J. M. Tetrahedron: Asymmetry 2000, 11, 1331–1341. (d)
Nishiyama, T.; Isobe, M.; Ichikawa, Y. Angew. Chem., Int. Ed. 2005,
44, 4372–4375. (e) Ichikawa, Y.; Matsukawa, Y.; Isobe, M. J. Am.
Chem. Soc. 2006, 128, 3934–3938. (f) Ichikawa, Y.; Matsukawa, Y.;
Tamura, M.; Ohara, F.; Isobe, M.; Kotsuki, H. Chem. Asian. J. 2006,
1, 717–723.
(4) (a) Prosperi, D.; Ronchi, S.; Panza, L.; Rencurosi, A.; Russo, G. Synlett
2004, 1529–1532. (b) Prosperi, D.; Ronchi, S.; Lay, L.; Rencurosi,
A.; Russo, G. Eur. J. Org. Chem. 2004, 395–405.
(5) (a) Tewari, N.; Tiwari, V. K.; Mishra, R. C.; Tripathi, R. P.; Srivastava,
A. K.; Ahmad, R.; Srivastava, R.; Srivastava, B. S. Bioorg. Med. Chem.
2003, 11, 2911–2922. (b) Paulsen, H.; Todt, K. AdV. Carbohydr. Chem.
1968, 23, 115–232. (c) Truscheit, E.; Frommer, W.; Junge, B.; Mueller,
L.; Schmidt, D. D.; Wingender, W. Angew. Chem., Int. Ed. 1981, 20,
744–761. (d) Inouye, S.; Tsuruoka, T.; Ito, T.; Niida, T. Tetrahedron
1968, 24, 2125–2144. (e) Anzeveno, P. B.; Creemer, L. J.; Daniel,
J. K.; King, C. H. R.; Liu, P. S. J. Org. Chem. 1989, 54, 2539–2542.
(6) (a) Umezewa, S.; Tsuchiya, T. In Aminoglycosides Antibiotics;
Umezewa, H.; Hooper, R. I., Eds; Springer: Berlin, 1982; pp 37-
110. (b) Kirst, H. A. in Burger’s Medicinal Chemistry and Drug
DiscoVery; Wolff, M. E., Eds; Wiley: New York, 1996, pp 463-525.
(9) Bianchi, A.; Ferrario, D.; Bernardi, A. Carbohydr. Res. 2006, 341,
1438–1446.
(10) Yang, J.; Mercer, G. J.; Nguyen, H. M. Org. Lett. 2007, 9, 4231–
4234.
9
11210 J. AM. CHEM. SOC. 2008, 130, 11210–11218
10.1021/ja803378k CCC: $40.75
2008 American Chemical Society

Rearrangement of Glycal Trichloroacetimidates
A R T I C L E S
Scheme 1
mechanism to provide ꢀ-N-glycosyl trichloroacetamide 3.13 The
resulting R- and ꢀ-N-glycosyl trichloroacetamides 2 and 3 are
further transformed into the corresponding glycosyl ureas 4 and
5, respectively, in a two-step procedure. The advantage of using
the palladium catalysis is that both R- and ꢀ-N-glycosyl
trichloroacetamides can be synthesized from the same starting
material with good to excellent anomeric selectivity under mild
conditions. Additionally, the starting glycal trichloroacetimidate
can be readily prepared by the combination of glycal, trichlo-
roacetonitrile, and a catalytic amount of DBU. Although the
[3,3]-sigmatropic rearrangement of allylic trichloroacetimidates
is pioneered by Overman,14 this method has never been
employed in carbohydrate synthesis to control the anomeric
selectivity of the N-glycosyl trichloroacetamides.
Results and Discussion
A. Pd(II)-Catalyzed Glycal Imidate Rearrangement. A suc-
cessful Pd(II)-catalyzed rearrangement of glycal trichloroace-
timidates requires the development of a suitable palladium-ligand
complex that can differentiate between the cyclization-induced
rearrangement and ionization-recombination pathways (Figure
1). The cyclization-induced rearrangement of allylic trichloro-
acetimidates catalyzed by the chiral Pd(II) catalyst has been
thoroughly studied by Overman and Bergman.13 It was envi-
sioned that a complementary method for glycal trichloroace-
timidate rearrangement would arise from coordination of the
neutral Pd(II)-ligand catalyst to the double bond of 1 to form
π-complex 6, which is activated toward nucleophilic attack by
the imidate nitrogen to form σ-complex 7 (Figure 1). Subsequent
Grob-like fragmentation followed by dissociation of the amide
from palladium yields ꢀ-N-glycosyl trichloroacetamide 3.
Alternatively, because the cationic palladium(II) species acts
as Lewis acid catalyst,15 it would coordinate to the imidate
nitrogen of 1 to form 8 which subsequently undergoes ionization
to generate ion pair 9.12 The intermediate 9 then recombines in
Figure 1. Proposed mechanism for palladium(II)-catalyzed rearrangement
of glycal imidates.
a stereoelectronically favored mode to form R-N-glycosyl
trichloroacetamide 2.
The feasibility of the palladium-catalyzed allylic imidate
rearrangement outlined in Figure 1 was examined by treatment
of glucal imidate 10 with 2.5 mol % of Pd(PhCN)2Cl2 in CH2Cl2
at 25 °C for 2 h to provide a 1:1 mixture of R- and ꢀ-N-glycosyl
trichloroacetamide 11 in 60% yield (Table 1, entry 1). Addition
of 4 Å molecular sieves significantly increased the yield of 11
(entry 2). It was anticipated that the stereoselectivity at the
anomeric carbon would depend on the ligand on palladium.11
Accordingly, glucal imidate 10 was treated with a preformed
solution of Pd(PhCN)2Cl2 and Ph3P, and N-glycosyl trichloro-
acetamide 11 was isolated in 83% yield with R:ꢀ ) 1:2 (entry
3). A variety of Buchwald’s bulky biaryl phosphine ligands were
investigated because these ligands are resistant toward oxidation
by molecular oxygen.16 Since these ligands are electron-rich,
they could potentially reduce Pd(II) to Pd(0) in the reaction;
thus, only a 1:1 mixture of palladium and ligand was investi-
gated. With the use of RuPhos, t-BuX-Phos, and JohnPhos as
the phosphine ligands, the anomeric selectivity was improved,
favoring the ꢀ-anomer (entries 5-7). Employing tris(trimethox-
yphenyl)phosphine (TTMPP) as the ligand led to an improve-
ment of both the yield and the ꢀ-selectivity (entry 8). These
results clearly suggest that the bulky phosphine ligands promote
the cyclization-induced rearrangement13 over the ionization of
glycal imidate 1012 to favor the formation of ꢀ-N-glycosyl
trichloroacetamide 11.
(11) (a) Kim, H.; Men, H.; Lee, C. J. Am. Chem. Soc. 2004, 126, 1336–
1337. (b) Schuff, B. P.; Mercer, G. J.; Nguyen, H. M. Org. Lett. 2007,
9, 3173–3176.
(12) (a) Calter, M.; Hollis, T. K.; Overman, L. E.; Ziller, J.; Zipp, G. G. J.
Org. Chem. 1997, 62, 1449–1456. (b) Yougai, S.; Miwa, T. J. Chem.
Soc., Chem. Commun. 1983, 68–69.
(13) (a) Watson, M. P.; Overman, L. E.; Bergman, R. G. J. Am. Chem.
Soc. 2007, 129, 5031–5044. (b) Anderson, C. E.; Overman, L. E. J. Am.
Chem. Soc. 2003, 125, 12412–12413.
(14) (a) Overman, L. E. J. Am. Chem. Soc. 1974, 96, 597–599. (b) For a
comprehensive reivew, see: Overman, L. E.; Carpenter, N. E. In
Organic Reactions; Overman, L. E., Ed.; Wiley-VCH: Weinheim,
2005,Vol. 66, pp 1-107.
(15) Mikiami, K.; Hatano, M.; Akiyama, K. Top. Organomet. Chem. 1985,
14, 270–321, and references therein.
(16) Barder, T. E.; Buchwald, S. L. J. Am. Chem. Soc. 2007, 129, 5096–
5101.
9
J. AM. CHEM. SOC. VOL. 130, NO. 33, 2008 11211

A R T I C L E S
Mercer et al.
Table 1. Neutral Pd(II)-Catalyzed Stereoselective Formation of
ꢀ-N-Glycosyl Trichloroacetamidea
to determine if the bulky phosphine ligand is necessary for the
ꢀ-selectivity (entry 5). A 1:2 mixture of R- and ꢀ-isomer 11
was detected in the reaction. Thus, the combination of the bulky
phosphine ligand and salicylaldehyde (L-1) increased the
ꢀ-selectivity as well as shortened the reaction time. The more
electron-rich ligands (L-2 and L-3) significantly increased the
reaction time and lowered the ꢀ-selectivity (entries 6 and 7).
On the other hand, switching to the more electron-withdrawing
ligands18 (L-4 and L-5) dramatically enhanced the ꢀ-selectivity
and shortened the reaction time (entries 8 and 9). The optimal
conditions with the use of 4-chloro-2-trifluoroacetylphenol (L-5)
as the ligand has been established in terms of anomeric
selectivity (R:ꢀ ) 1:16), yield (88%), and reaction time (2.5 h)
(entry 9). Thus, striking a balance between the electronic nature
of the σ-donor ligands is key to increasing the ꢀ-selectivity and
shortening the reaction time. To determine if the effect of
salicylaldehyde (L-1) and its derivatives (L-4 and L-5) on the
anomeric selectivity does not simply derive from incorporation
of a weak acid, a control experiment was performed with
4-hydroxybenzaldehyde (L-6), whose acidity is comparable to
that of salicylaldehyde. The desired N-glycosyl trichloroaceta-
mide 11 was isolated in 63% yield with R:ꢀ ) 1:3 (entry 10).
This result validates that coordination of salicylaldehyde (L-1)
and its derivatives (L-4 and L-5) to palladium is important for
the high stereoselectivity observed in the product.
When cationic palladium(II),15,19 Pd(CH3CN)4(BF4)2 (2.5 mol
%), was employed in the reaction, the desired R-N-glycosyl
trichloroacetamide 11 was obtained in 73% yield as the major
anomer (Table 3, entry 1). Thus, switching to the cationic
palladium reverses the anomeric selectivity, favoring the R-ano-
mer. Addition of 10 mol % of salicylaldehyde (L-1) significantly
increased the R-selectivity (entry 2). Further optimization
focused on the catalyst loading. Decreasing the catalyst loading
to 0.5 mol % from 2.5 mol % gave 82% of 11 with R:ꢀ ) 13:1
(entry 3). Lowering the catalyst further decreased the yield and
the R-selectivity (entry 4). To determine if HBF4, which may
generate from the cationic palladium catalyst, is the source of
catalysis, the rearrangement was performed in the presence of
2,6-di-tert-butylpyridine (DTBP) (5 mol %) as an acid scavenger
(entry 5). In the event, the formation of R-anomer 11 from glycal
imidate 10 proceeded in comparable yield and selectivity to
those of entries 1-4. Switching to the more electron-withdraw-
ing ligand, L-4 (2-trifluoroacetylphenol), showed improved yield
and R-selectivity (entry 6). A control experiment was performed
with 4-hydroxybenzaldehyde (L-6) to determine the effect of
the salicylaldehyde and its derivatives on the yield and anomeric
selectivity of the rearrangement (entry 8). The desired trichlo-
roacetamide 11 was isolated in 58% yield with R:ꢀ ) 10:1.
Because 4-hydroxybenzaldehyde (L-6) is a weak acid, it could
potentially decompose the acid-sensitive glycal trichloroace-
timidate starting material 10 during the course of the reaction.
In contrast, salicylaldehyde (L-1) and its derivatives (L-4 and
L-5) were responsible for the high yield and R-selectivity
observed in the product.
a All reactions were carried out in CH2Cl2 (0.2 M) with 2.5 mol %
Pd(PhCN)2Cl2 and 2.5 mol % phosphine ligand. b Isolated yield. c 1H
NMR ratio.
Although the bulky phosphine ligands provided the product
with moderate to good ꢀ-selectivity, it took 16-25 h for the
reaction to go to completion. This suggests that the rate of
nucleophilic attack by the imidate nitrogen to the palladium-
olefin complex such as 6 (Figure 1) may not be fast enough.
Additionally, the rate difference between the cyclization-induced
rearrangement pathway and the ionization pathway is not large
enough. An alternative approach is to increase the electrophi-
licity of double bonds of glycal imidates by altering the
electronic nature of the palladium-ligand complex. Further-
more, the catalyst must effectively control the relative population
of the R- and ꢀ-anomers. Toward this end, several σ-donor
ligands were explored (Scheme 2) in conjunction with the bulky
phosphine ligands.
A standard set of conditions that employed 2.5 mol %
Pd(PhCN)2Cl2, 2.5 mol % phosphine ligand, and 10 mol %
σ-donor ligand were adopted (Table 2). Gratifyingly, it was
found that addition of salicylaldehyde17 (L-1) significantly
shortened the reaction time and increased the ꢀ-selectivity
(entries 1 and 2), providing the desired N-glycosyl trichloro-
acetamide 11 in good yield. We also examined whether the
reaction temperature effected the selectivity; increasing or
decreasing the reaction temperature only decreased the ꢀ-se-
lectivity (entries 3 and 4). This trend with temperature suggests
that while heat can lead to faster rearrangement, it can also lead
to faster equilibration between the two pathways outlined in
Figure 1. On the other hand, cooling can lead to slower
cyclization-induced rearrangement pathway and allow equilibra-
tion between the two pathways; thus, the ꢀ-selectivity was
decreased. A control experiment without TTMPP was performed
Chart 1 summarizes the result of the application of the
optimized neutral palladium(II) conditions to a series of glycal
trichloroacetimidates. Both L-1 (salicylaldehyde) and L-5 (4-
chloro-2-trifluoroacetylphenol) ligands provided the desired
products 12-16 with good ꢀ-selectivity. In general, the rear-
rangement was faster with the L-5 ligand, and the glycosyl
(18) Brodsky, B. H.; Du Bois, J. J. Am. Chem. Soc. 2005, 127, 15391–
15393.
(17) Miller, K. J.; Bagg, J. H.; Abu-Omar, M. M. Inorg. Chem. 1999, 38,
4510–4514.
(19) Wayland, B. B.; Schramm, R. F. Inorg. Chem. 1969, 8, 971–976.
9
11212 J. AM. CHEM. SOC. VOL. 130, NO. 33, 2008

Rearrangement of Glycal Trichloroacetimidates
A R T I C L E S
Scheme 2
a-c
Table 2. Optimization for Stereoselective Synthesis of ꢀ-N-
Glycosyl Trichloroacetamide with σ-Donor Ligandsa
Chart 1. Synthesis of ꢀ-N-Glycosyl Trichloroacetamides
σ-donor
ligand
entry
phosphine ligand
time, h
temp, °C
yield,b
%
R:ꢀc
1
2
3
4
5
6
7
8
9
10
t-BuX-Phos
TTMPP
TTMPP
TTMPP
none
TTMPP
TTMPP
TTMPP
TTMPP
TTMPP
L-1
L-1
L-1
L-1
L-1
L-2
L-3
L-4
L-5
L-6d
6
4
8
2
1
24
7
2.5
2.5
3
25
25
0
40
25
25
25
25
25
25
70
86
30
75
71
82
70
95
88
63
1:7
1:9
1:6
1:4
1:2
1:8
1:7
1:12
1:16
1:3
a All reactions were carried out in CH2Cl2 (0.2 M) with 2.5 mol %
Pd(PhCN)2Cl2, 2.5 mol% phosphine ligand, and 10 mol % σ-donor
ligand except for entry 5. b Isolated yield. c 1H NMR ratio. d L6 )
4-hydroxybenzaldehyde.
Table 3. Cationic Pd(II)-Catalyzed Stereoselective Formation of
R-N-Glycosyl Trichloroacetamidea
a All reactions were performed with 2.5 mol % Pd(PhCN)2Cl2 and 2.5
c
mol % of TTMPP and 10 mol % of L-1 or L-5 ligands. b Isolated yield.
1
H NMR ratio.
provided the desired products 12-19 with excellent R-selectivity
(Chart 2). In general, the glycosyl trichloroacetamides were
obtained with higher R-selectivity when L-4 was employed as
ligand in the rearrangement. These results suggest that the
cationic palladium-ligand complex was responsible for the
observed R-selectivity at the anomeric center, and the nature
of the protecting groups on the glycal substrates had little effect
on the selectivity.
entry palladium, mol %
ligand
additive
time
yield, %b R:ꢀc
1
2
3
4
5
6
7
8
2.5
2.5
0.5
0.1
0.5
0.5
0.5
0.5
none
none
45 min
1 h
1 h
73
80
82
78
79
93
83
58
9:1
14:1
13:1
9:1
10 mol % L-1 none
2 mol % L-1 none
0.4 mol % L-1 none
2 h
We also investigated whether the glycal imidate rearrange-
ment could be catalyzed by a Lewis acid (Scheme 3). Accord-
ingly, treatment of glycal imidate 10 with 0.5 mol % TMSOTf
in CH2Cl2 only resulted in decomposition. This result suggests
that the Lewis-acid-catalyzed rearrangement is intolerant of the
acid-sensitive glycal imidates. Under thermal conditions, the
rearrangement provided 11 in a quantitative yield. However, it
took 6 h at 120 °C for the reaction to go to completion. Based
on the concerted nature of the thermal [3,3]-sigmatropic
rearrangement, the configuration at C(1) in 11 was assigned as
2 mol % L-1
2 mol % L-4
2 mol % L-5
DTBP 1 h
none
none
9:1
45 min
45 min
45 min
14:1
15:1
10:1
2 mol % L-6d none
a All reactions were carried out in CH2Cl2 (0.2 M) with
Pd(CH3CN)4(BF4)2 and σ-donor ligand (1:4) except for entry 1.
b Isolated yield. c 1H NMR ratio. d L6 ) 4-hydroxybenzaldehyde.
trichloroacetamides were obtained with higher ꢀ-selectivity.
Furthermore, these results suggest that the deactivating effect
of a 4,6-acetal protecting groups on these glycal trichloroace-
timidates restrict them in the tg conformations, thus limiting
ionization to favor ꢀ-anomers.20 On the other hand, glycal
imidates incorporating acyclic protecting groups gave a mixture
of R- and ꢀ-N-glycosyl trichloroacetamides such as 17 and 18.
Under the optimized cationic palladium(II) conditions, both
L-1 (salicylaldehyde) and L-4 (2-trifluoroacetylphenol) ligands
1
the ꢀ-anomer. The H NMR of 11 matches with that of the
neutral palladium(II)-catalyzed rearrangement of 10. The NH-
resonance of ꢀ-isomer 11 appeared at δ ) 6.92 ppm. In the
present series of ꢀ-N-glycosyl trichloroacetamides, we typically
found the NH-resonances at δ ) 6.87-6.97 ppm.
The chemical shifts of the NH proton are consistently unique
for the R-anomer products. In the present series of R-N-glycosyl
trichloroacetamides, we typically found the NH-resonances at
δ ) 7.15-7.26 ppm. To confirm the anomeric stereochemistry
in the R-N-glycosyl trichloroacetamide series, R-anomer 15 was
(20) (a) Jensen, H. H.; Nordstrøm, L. U.; Bols, M. J. Am. Chem. Soc. 2004,
126, 9205–9213. (b) Crich, D.; Chandrasekera, N. S. Angew. Chem.,
Int. Ed. 2004, 43, 5386–5389.
9
J. AM. CHEM. SOC. VOL. 130, NO. 33, 2008 11213

A R T I C L E S
Mercer et al.
a-c
increasingly recognized as a biologically relevant moiety.22
A
Chart 2. Synthesis of R-N-Glycosyl Trichloroacetamides
number of powerful tools have emerged recently for the
construction of this functionality.8,9 However, the stereoselective
synthesis of R- and ꢀ-glycosyl ureas is still lacking. Additionally,
glycosyl ureas are generally prepared in several steps starting
from glycosyl azides.8 In our methodology studies, glycosyl
ureas were formed by dihydroxylation of N-glycosyl trichloro-
acetamide and subsequent treatment of the resulting diol
intermediate with a base and a nitrogen nucleophile (Scheme
5). It was hypothesized that the trichloroacetamide proton could
be selectively deprotonated to generate an isocyanate in situ,
which would subsequently participate in the glycosyl urea
formation in the presence of a nucleophilic nitrogen. In this
strategy, there is no loss of stereochemical integrity at the C(1)-
anomeric position, a major drawback of other methodologies.
This approach can be flawed, however, by the competing
hydrolysis pathway to form the free amine at the C(1)-position,23
which then undergoes anomerization. Determined to showcase
N-glycosyl trichloroacetamide substrates as robust and versatile
intermediates, we screened a number of bases and solvents
(Table 4). Using NaH or K2CO3 as a base in THF resulted in
no reaction (entries 1 and 2). Performing the reaction in DMF
with K2CO3 as a base resulted in hydrolysis of the trichloroacetyl
group of 23 to provide the free amine, which underwent
anomerization to form ꢀ-anomer (entry 3). Ultimately, it was
found that the softer base Cs2CO3 with DMF as solvent to be
ideally suited to the task (entry 4). With use of n-butylamine
as a nitrogen nucleophile, the desired R-glycosyl urea 24 was
isolated in 74% yield. In contrast to other methods,8 our
approach requires only 2-3 steps for the synthesis of a
R-glycosyl urea starting from R-N-glycosyl trichloroacetamide.
Most importantly, the stereochemical integrity at the anomeric
carbon remains intact.24
a All reactions were performed with 0.5 mol % Pd(CH3CN)4(BF4)2 and
2.0 mol % L-1 or L-4 ligands. b Isolated yield. c 1H NMR ratio.
Scheme 3
With optimized reaction conditions in hand, we set out to
define the scope of this transformation. We discovered that a
diverse array of primary and secondary nitrogen nucleophiles
could be coupled to R-glycosyl trichloroacetamides 15 and 18
to form the corresponding R-glycosyl ureas 25-31 in moderate
to good yields and with no anomerization (Table 5). Notably,
the efficiency of this reaction is demonstrated by the ability to
couple methylpiperazine and pyridylpiperazine with R-glycosyl
trichloroacetamide 18 to form R-glycosyl ureas 30 and 31,
respectively (entries 6 and 7).
There are only a few examples of symmetrical urea-linked
disaccharides, in which two glucopyranoside units are bound
at the two anomeric positions.5,8a Furthermore, there are only
two reports on the synthesis of unsymmetrical urea-linked
disaccharides, in which the anomeric position of a glycopyra-
nosyl donor is connected to the nonanomeric position of a
glycopyranosyl acceptor.8f,25 Encouraged by the results obtained
with noncarbohydrate amines in Table 5, we decided to explore
the feasibility of our method with primary and hindered amines
Scheme 4
treated with OsO4 and NMO to provide the corresponding diol
1
20A (Scheme 4). The JCH coupling between the anomeric
carbon of 20A and its associated proton was measured, and 1JCH
) 170.2 Hz. It is known that, in a standard chair conformation,
O-glycopyranosides with a value of 173 Hz is diagnostic for
an R-linkage.21
(22) Ellestad, G. A.; Cosulich, D. B.; Broschard, R. W.; Martin, J. H.;
Kunstmann, M. P.; Morton, G. O.; Lancaster, J. E.; Fulmor, W.;
Loevell, G. M. J. Am. Chem. Soc. 1978, 100, 2515–2524.
(23) (a) Nishikawa, T.; Ohyabu, N.; Yamamoto, N.; Isobe, M. Tetrahedron
1999, 55, 4325–4340. (b) Urabe, D.; Sugino, K.; Nishikawa, T.; Isobe,
M. Tetrahedron. Lett. 2004, 45, 9405–9407.
B. Stereoselective Synthesis of Glycosyl Ureas. During the
course of our methodology studies, we became interested in
the synthesis of glycosyl ureas because their functionality is
(24) The five step non-selective method for the synthesis of glycosyl urea
from R-glycosyl azide was developed by Isobe and coworkers to
circumvent the unsuccessful attempt at forming the isocyanate in situ
from trichloroacetamide. Ichikawa, Y.; Nishiyama, T.; Isobe, M. Synlett
2000, 1253–1256.
(25) Propseri, D.; Ronchi, S.; Lay, L.; Rencurosi, A.; Russo, G. Eur. J.
Org. Chem. 2004, 395–405.
(21) Bock, K.; Pedersen, C. J. Chem. Soc., Perkin Trans. 2 1974, 293.
9
11214 J. AM. CHEM. SOC. VOL. 130, NO. 33, 2008

Rearrangement of Glycal Trichloroacetimidates
A R T I C L E S
Scheme 5
Table 4. Transformation of N-Glycosyl Trichloroacetamide into a Glycosyl Urea
entry
base
solvent
results
1
2
3
4
NaH
THF
THF
DMF
DMF
no reaction
no reaction
hydrolysis
74% yield
K2CO3
K2CO3
Cs2CO3
Table 5. Stereoselective Synthesis of R-Glycosyl Ureasa
Table 6. Synthesis of Unsymmetrical Urea-Linked Disaccharidesa
a All reactions were performed in DMF at 25 °C with Cs2CO3 (4
equiv) and R′-NH2 (2-3 equiv). b The polyol intermediate of 39 was
acylated to ease the purification process.
philes for the preparation of various unsymmetrical urea-linked
disaccharides. In order to achieve high conversion, an excess
amount of carbohydrate amines (2-3 equiv) were employed in
the reaction, and the unreacted amines were recovered after
chromatography. The most encouraging result was discovered
a All reactions were performed in DMF at 25 °C with Cs2CO3 (4
equiv) and R′-NH2 (2-3 equiv). b Yield included three steps.
of carbohydrate nucleophiles 32-35 (Table 6). Gratifyingly, it
was found that these carbohydrate amines are suitable nucleo-
9
J. AM. CHEM. SOC. VOL. 130, NO. 33, 2008 11215

A R T I C L E S
Mercer et al.
Table 7. Stereoselective Synthesis of ꢀ-Glycosyl Ureasa
resulting diol 48 with carbohydrate primary amine 32 provided
urea-linked trisaccharide 49 in 68% yield and with no epimer-
ization of the anomeric C-N bond.
Conclusions
In summary, a novel method for palladium(II)-catalyzed
stereoselective formation of R- and ꢀ-N-glycosyl trichloroac-
etamides has been developed. The R- and ꢀ-selectivity at the
anomeric carbon of N-glycosyl trichloroacetamides depends on
the nature of the palladium-ligand catalyst. While the cationic
palladium-L-4 (2-trifluoroacetylphenol) complex promotes the
R-selectivity, the neutral palladium-TTMPP-L-5 (4-fluoro-2-
trifluoroacetylphenol) complex favors the ꢀ-selectivity. Because
of its substrate tolerance and mild conditions, this palladium
method is applicable to a wide range of glycal imidates. The
resulting R- and ꢀ-N-glycosyl trichloroacetamides were further
coupled with a diverse array of primary and hindered secondary
nitrogen nucleophiles to provide the corresponding glycosyl
ureas in moderate to good yields and with no loss of stereo-
chemical integrity at the anomeric carbon. We have further
demonstrated the utility of N-glycosyl trichloroacetamides as
robust and versatile intermediates in the synthesis of unsym-
metrical urea-linked disaccharides and trisaccharide. These
glycosyl ureas will be evaluated for their antibacterial activity.
a All reactions were performed in DMF at 25 °C with Cs2CO3 (4
equiv) and R′-NH2 (2-3 equiv). The diol intermediates of glycosyl
ureas were acylated to ease the purification process.
when the hindered amine of tetra-O-acetyl glucosamine 35 was
efficiently coupled with R-glycosyl trichloroacetamide 15 to
form unsymmetrical urea-linked disaccharide 39 in 68% yield.
In this reaction, the bis(tert-butylsilyl) group was replaced with
the acetyl group during the course of the reaction.
Experimental Section
Representative experimental procedures are listed here. Full
experimental details and spectral data for all new compounds can
be found within the Supporting Information.
These results were encouraging because they clearly showed
that our methodology is amenable to most amines. We also
examined this chemistry with ꢀ-N-glycosyl trichloroacetamides
(Table 7). Under standard conditions, coupling of 15ꢀ with
primary and secondary amines provided the corresponding
ꢀ-glycosyl ureas 40-42 in moderate yields.
Preparation of ꢀ-N-Glycosyl Trichloroacetamide 11ꢀ with
L-5 (4-chloro-2-trifluoroacetylphenol) as Ligand. A 10 mL oven-
dried Schlenk flask was charged with Pd(PhCN)2Cl2 (2.9 mg, 0.0075
mmol, 2.5 mol%), TTMPP (4.0 mg, 0.0075 mmol, 2.5% mol), and
CH2Cl2 (1.5 mL). The solution was stirred at 25 °C for 2 h, and
L-5 (6.7 mg, 0.03 mmol, 10 mol%) was then added. The resulting
mixture was stirred for 1 h, and glucal imidate 10 (99 mg, 0.3 mmol,
1 equiv) was added. The reaction mixture was stirred for 90 min,
diluted with benzene (2 mL), concentrated to about 1 mL, and
purified by silica gel flash chromatography (8/1 f 4/1, hexane/
ethyl acetate) to give 11 (88 mg, 88%, R:ꢀ ) 1:16) as a white
solid: mp ) 121-122 °C; Rf ) 0.33 (4/1, hexane/ethyl acetate);
1H NMR (CDCl3, 500 MHz): δ ) 6.92 (d, J ) 8.5 Hz, 1H, NH),
6.10 (d, J ) 10.0 Hz, 1H, H2), 5.97 (d, J ) 7.0 Hz, 1H, H1), 5.64
(d, J ) 10.5 Hz, 1H, H3), 4.33 (d, J ) 7.5 Hz, 1H, H4), 3.92 (dd,
J ) 10.5, 5.5 Hz, 1H, H6), 3.79 (t, J ) 10.5 Hz, 1H, H6), 3.61
(ddd, J ) 10.5, 7.5, 5.5, 1H, H5), 1.51 (s, 3H), 1.41 (s, 3H); 13C
NMR (CDCl3, 125 MHz): δ ) 161.3, 133.2, 125.8, 100.0, 92.0,
77.7, 72.3, 66.8, 62.4, 29.1, 19.0; IR (film, cm-1): V ) 3329, 2923,
1713, 1512, 1095; HRMS (ESI): calcd for C11H14Cl3N1O4Na [M
+ Na] 351.9881; found: 351.9882.
To this point, the efficacy of transforming R- and ꢀ-glycosyl
trichloroacetamides into R- and ꢀ-glycosyl ureas has been
demonstrated only in the preparation of urea-linked disaccha-
rides. The true test of the versatility of this concept is its ability
to function in the synthesis of complex pseudooligosaccharides.
At the same time, the synthesis of urea-linked oligosaccharides
would showcase our methodology of Pd(II)-catalyzed rear-
rangement of glycal trichloroacetimidates. Therefore, synthesis
of urea-linked trisaccharide 49 was undertaken (Scheme 6).
Preparation of urea-linked pseudotrisaccharide 49 was tar-
geted so that we can determine its bactericidal activity as well
as stability compared to other pseudodisaccharides. Under our
conditions of cationic Pd(II)-catalyzed stereoselective glycosy-
lation,26 coupling of glycosyl trichloroacetimidate donor 4327
with nucleophilic acceptor 44 provided disaccharide 45 in 70%
yield exclusively as R-anomer (Scheme 6). Hydrolysis of the
acetyl group followed by treatment of the resulting allylic
alcohol intermediate with trichloroacetonitrile and DBU afforded
the corresponding disaccharide trichloroacetimidate 46 in 91%
yield over two steps. With 46 in hand, the next step is to employ
our method of Pd(II)-catalyzed glycal imidate rearrangement.
The desired disaccharide trichloroacetamide 47 was isolated in
91% yield with an anomeric ratio of 5:1, favoring the R-anomer.
Subsequent dihydroxylation of 47 followed by coupling of the
Preparation of r-N-Glycosyl Trichloroacetamide 11 with
L-4 (2-trifluoroacetylphenol) as Ligand. A 10 mL oven-dried
Schlenk flask was charged with glucal imidate 10 (99 mg, 0.3 mmol,
1 equiv) and CH2Cl2 (1.0 mL). A preformed solution of
Pd(CH3CN)4(BF4)2 (0.67 mg, 0.0015 mmol, 0.5 mol%) and L-4
(1.14 mg, 0.006 mmol, 2.0 mol%) in CH2Cl2 (0.5 mL), which had
been stirring at 25 °C for 2 h, was added. The resulting mixture
was stirred at 25 °C for 45 min, diluted with benzene (2 mL),
concentrated to about 1 mL, and purified by silica gel flash
chromatography (8/1 f 4/1, hexane/ethyl acetate) to give 11 (92.2
mg, 93%, R:ꢀ ) 14:1) as a white solid: mp ) 125-128 °C; Rf)
1
0.38 (4/1, hexane/ethyl acetate); H NMR (CDCl3, 500 MHz): δ
) 7.22 (bs, 1H, NH), 6.18 (d, J ) 10.0 Hz, 1H, H2), 5.81 (dd, J
) 9.0, 2.0 Hz, 1H, H1), 5.72 (dt, J ) 10.0, 2.5 Hz, 1H, H3), 4.25
(d, J ) 9.0 Hz, 1H, H4), 3.92 (dd, J ) 11.0, 5.5 Hz, 1H, H6), 3.80
(t, J ) 11.0 Hz, 1H, H6), 3.47 (ddd, J ) 11.0, 9.0, 5.5, 1H, H5),
(26) Yang, J.; Cooper-Vanosdell, C.; Mensah, E. A.; Nguyen, H. M. J.
Org. Chem. 2008, 73, 794–800.
(27) Drew, K. N.; Gross, P. H. J. Org. Chem. 1991, 56, 509–513.
9
11216 J. AM. CHEM. SOC. VOL. 130, NO. 33, 2008

Rearrangement of Glycal Trichloroacetimidates
A R T I C L E S
Scheme 6
1.52 (s, 3H), 1.43 (s, 3H); 13C NMR (CDCl3, 125 MHz): δ ) 161.4,
134.0, 124.6, 100.1, 92.2, 74.5, 72.3, 67.0, 66.7, 62.9, 29.1, 18.9;
IR (film, cm-1): V ) 3329, 2923, 1710, 1498, 1099, 1038; HRMS
(ESI): calcd for C11H14Cl3N1O4Na [M + Na] 351.9881; found:
351.9882.
(98 µL, 0.9 mmol, 6 equiv) and N,N-dimethylformamide (3 mL).
The cloudy solution was stirred at 25 °C for 20 h. This heteroge-
neous mixture was poured into a saturated aqueous solution of
NaHCO3 (20 mL) with an ethyl acetate wash of the reaction flask.
The crude product was extracted with ethyl acetate (5 × 30 mL).
The combined organic extracts were washed with brine, dried over
MgSO4, and concentrated. To a scintillation vial the crude urea
was combined with pyridine (1.5 mL), acetic anhydride (200 µL,
1.8 mmol, 6 equiv), and one crystal of N,N-dimethylaminopyridine.
The solution was stirred at 25 °C for 21 h and then concentrated in
Vacuo. The resulting residue was purified by silica gel chromatog-
raphy (1/1, hexane/ethyl acetate) to give 25 (44 mg, 55%). Rf)
0.65 (1/1, hexane/ethyl acetate); 1H NMR (CDCl3, 500 MHz, ppm):
δ ) 7.35-7.20 (m, 5H), 5.49 (d, J ) 9.5 Hz, 1H), 5.32 (d, J ) 2.5
Hz, 1H), 5.21 (d, J ) 9.5 Hz, 1H), 4.97 (dd, J ) 9.5, 4.0 Hz, 1H),
4.91-4.86 (m, 1H), 4.32 (ddd, J ) 16.0, 10.0, 4.5 Hz, 2H), 4.17
(dd, J ) 10.0, 5.0 Hz, 1H), 3.99 (t, J ) 9.5 Hz, 1H), 3.85 (t, J )
10.0 Hz, 1H), 3.56 (td, J ) 10.0, 5.0 Hz, 1H), 2.11 (s, 3H), 2.01
(s, 3H), 1.02 (s, 9H), 0.95 (s, 9H); 13C NMR (CDCl3, 125 MHz,
ppm): δ ) 170.2, 170.0, 155.8, 138.2, 128.7, 127.6, 127.5, 78.1,
73.4, 72.7, 71.5, 71.0, 66.1, 44.6, 27.3, 26.9, 22.6, 20.8, 20.6, 19.9;
IR (film, cm-1): V ) 3347 (br, w), 2834 (m), 2893 (w), 2860 (m),
1753 (s), 1675 (m), 1645 (s), 1560 (s), 1236 (s), 1236 (s), 1077
(s); HRMS (ESI): calcd for C26H40N2O8SiNa (M + Na) 559.2446;
found: 559.2411.
Preparation of r-N-Glycosyl Urea 24. Compound 23 (75 mg,
0.10 mmol, 1 equiv) was added to a Schlenk flask as a solution in
THF, and the THF was removed in Vacuo. The alcohol was
subsequently azeotroped with toluene (3 × 1.5 mL). Cesium
carbonate (132 mg, 0.40 mmol, 4 equiv) was added to the flask
followed by n-butylamine (79 µL, 0.6 mmol, 6 equiv) and N,N-
dimethylformamide (2 mL). The cloudy solution was stirred at 25
°C for 20 h. This heterogeneous mixture was poured into a saturated
aqueous solution of NaHCO3 (20 mL) with an ethyl acetate wash
of the reaction flask. The crude product was extracted with ethyl
acetate (5 × 30 mL). The combined organic extracts were washed
with brine, dried over MgSO4, and concentrated. The resulting
residue was purified by silica gel chromatography (2/1, hexane/
ethyl acetate) to give 24 (51 mg, 74%). Rf) 0.31 (3/1, hexane/
ethyl acetate); 1H NMR (CDCl3, 500 MHz, ppm): δ ) 7.50-7.15
(m, 20H), 5.07 (t, J ) 10 Hz, 1H), 4.56 (dd, J ) 26, 12 Hz, 2H),
4.17 (s, 1H), 4.04 (d, J ) 2 Hz, 1H), 3.53-3.47 (m, 2H), 3.22 (dd,
J ) 10, 7 Hz, 1H), 3.18-3.12 (m, 1H), 2.93 (d, J ) 10 Hz, 1H),
2.81-2.75 (m, 1H), 2.42 (bs, 1H), 1.62-1.49 (m, 2H), 1.46-1.36
(m, 2H), 0.93 (t, J ) 7 Hz, 3H), 0.89 (s, 9H); 13C NMR (CDCl3,
75 MHz, ppm): δ ) 144.0, 128.8, 128.5, 128.0, 127.8, 126.9, 87.6,
80.2, 75.8, 71.8, 69.1, 67.9, 63.2, 45.4, 32.5, 2639, 20.6, 14.1; IR
(film, cm-1): V ) 2957 (s), 2929 (s), 2871 (s), 1734 (vs), 1491
(m), 1478 (m), 1449 (s), 1279 (m), 1153 (vs), 1106 (s), 1062 (s),
1037 (s), 699 (vs); HRMS (ESI): calcd for C42H50N2NaO7 (M +
Na) 717.3516; found: 717.3526.
Preparation of Urea-Linked Pseudodisaccharide 36. Diol 20A
(72 mg, 0.15 mmol, 1.0 equiv) was added to a Schlenk flask as a
solution in THF, and the THF was removed in Vacuo. The diol
was subsequently azeotroped with toluene (3 × 1.5 mL). Cesium
carbonate (146 mg, 0.45 mmol, 3.0 equiv) was added to the flask
followed by (-)-6-amino-6-deoxy-1,2:3,4-di-O-isopropylidene-R-
D-galactopyranose28 (106 mg, 0.41 mmol, 2.7 equiv) and N,N-
dimethylformamide (2.1 mL). The cloudy solution was stirred at
25 °C for 96 h. This heterogeneous mixture was poured into a
Preparation of r-N-Glycosyl Urea 25. Diol 20A (72 mg, 0.15
mmol, 1 equiv) was added to a Schlenk flask as a solution in THF,
and the THF was removed in Vacuo. The diol was subsequently
azeotroped with toluene (3 × 1.5 mL). Cesium carbonate (195 mg,
0.6 mmol, 4 equiv) was added to the flask followed by benzyl amine
(28) Streicher, B.; Wunsch, B. Carbohydr. Res. 2003, 338, 2375–2385.
9
J. AM. CHEM. SOC. VOL. 130, NO. 33, 2008 11217

A R T I C L E S
Mercer et al.
saturated aqueous solution of NaHCO3 (30 mL) with an ethyl
acetate wash of the reaction flask. The crude product was extracted
with ethyl acetate (5 × 30 mL). This organic phase was washed
with brine, dried over MgSO4, and concentrated. The resulting
residue was purified by silica gel chromatography (1/9, hexane/
ethyl acetate) to give 36 (68 mg, 75%) as a white solid: mp ) 126
was extracted with ethyl acetate (5 × 30 mL). This organic phase
was washed with brine, dried over MgSO4, and concentrated. The
resulting residue was purified by silica gel chromatography (1/1,
hexane/ethyl acetate) to give 49 (61 mg, 68%). Rf ) 0.12 (1/1
hexane/ethyl acetate); 1H NMR (CDCl3, 500 MHz, ppm): δ )
7.49-7.19 (m, 15H), 5.97 (bs, 1H), 5.76 (bs, 1H), 5.34-5.28 (m,
1H), 5.32 (s, 1H), 5.13 (bs, 1H), 4.61-4.54 (m, 2H), 4.47 (d, J )
5.5 Hz, 1H), 4.03-3.95 (m, 2H), 3.95-3.89 (m, 1H), 3.86 (t, J )
4 Hz, 1H), 3.84-3.78 (m, 3H), 3.69 (dd, J ) 7, 3.5 Hz, 1H), 3.51
(dd, J ) 10.5, 3 Hz, 1H), 3.47-3.27 (m, 5H), 1.45 (s, 3H), 1.42
(s, 3H), 1.40 (s, 3H), 1.35 (s, 3H), 1.34 (s, 3H), 1.32 (s, 3H), 1.29
(s, 6H), 1.26 (s, 3H); 13C NMR (CDCl3, 125 MHz, ppm): δ )
158.75, 143.85, 128.78, 127.80, 126.93, 112.61, 109.44, 109.02,
108.77, 105.92, 95.17, 86.82, 85.30, 80.91, 79.43, 73.03, 72.49,
71.29, 70.69, 70.53, 70.33, 66.74, 62.66, 29.68, 26.71, 25.99, 25.89,
25.17, 24.94, 24.60, 24.49; IR (film, cm-1): V ) 3378 (br, s), 2986
(s), 2931 (s), 1666 (s), 1549 (m), 1449 (m), 1381 (s), 1256 (m),
1211 (s), 1070 (vs); HRMS (ESI): calcd for C50H64N2NaO16 (M +
Na) 971.4148; found: 971.4186.
1
°C; Rf) 0.27 (1/1, hexane/ethyl acetate); H NMR (CDCl3, 500
MHz, ppm): δ ) 6.11 (d, J ) 9.5 Hz, 1H), 5.48 (d, J ) 5.0 Hz,
1H), 5.32 (dd, J ) 7.5, 4.0 Hz, 1H), 5.16 (d, J ) 9.5 Hz, 1H), 4.56
(dd, J ) 8.0, 2.0 Hz, 1H), 4.27 (dd, J ) 5.0, 2.5 Hz, 1H), 4.17 (dd,
J ) 7.5, 1.0 Hz, 1H), 4.13 (dd, J ) 10.0, 5.0, Hz, 1H), 3.96
(apparent s, 1H), 3.90 (t, J ) 9.0 Hz, 1H), 3.88-3.82 (m, 2H),
3.63 (apparent d, J ) 9.0 Hz, 1H), 3.54-3.40 (m, 3H), 3.25-3.17
(m, 1H), 3.06 (apparent s, 1H), 1.46 (s, 3H), 1.41 (s, 3H), 1.30 (s,
3H), 1.28 (s, 3H), 1.01 (s, 9H), 0.95 (s, 9H); 13C NMR (CDCl3, 75
MHz, ppm): δ ) 156.9, 109.5, 108.9, 96.3, 79.4, 74.6, 74.3, 71.7,
71.6, 70.8, 70.7, 67.1, 66.1, 40.8, 27.5, 27.0, 26.1, 26.0, 25.1, 24.4,
22.6, 20.0; IR (film, cm-1): V ) 3387 (br, m), 2964 (m), 2934 (s),
2897 (m), 2860 (s), 1661 (m), 1553 (m), 1473 (w), 1384 (m), 1212
(m), 1071 (m); HRMS (ESI): calcd for C27H48N2O11Si Na (M +
Na) 627.2920; found: 627.2889.
Preparation of Urea-Linked Pseudotrisaccharide 49. Diol 48
(77 mg, 0.095 mmol, 1.0 equiv) was added to a Schlenk flask as a
solution in THF, and the THF was removed in Vacuo. The diol
was subsequently azeotroped with toluene (3 × 1.5 mL). Cesium
carbonate (vaccuum-dried at 100 °C for 12 h) (124 mg, 0.380 mmol,
4.0 equiv) was added to the flask followed by (-)-6-amino-6-deoxy-
1,2:3,4-di-O-isopropylidene-R-D-galactopyranose28 (99 mg, 0.380
mmol, 4.0 equiv) and N,N-dimethylformamide (1.4 mL). The cloudy
solution was stirred at 25 °C for 96 h. This heterogeneous mixture
was poured into a saturated aqueous solution of NaHCO3 (30 mL)
with an ethyl acetate wash of the reaction flask. The crude product
Acknowledgment. This research is supported by the National
Science Foundation (NSF) under CHE-0809200. We thank Montana
State University and NSF EPSCoR for financial support. We also
thank Professor Justin Du Bois and Nichole Litvinas (Stanford
University) for the generous gift of L-5 ligand.
Supporting Information Available: Experimental procedures
1
and H NMR and 13C NMR spectra of all new compounds.
This material is available free of charge via the Internet at http://
pubs.acs.org.
JA803378K
9
11218 J. AM. CHEM. SOC. VOL. 130, NO. 33, 2008