Organic Letters
Letter
(0.5 mL) by adding 1a dropwise to the reaction mixture. To
our delight, the yield of 3a was increased to 51% (entry 10,
Table S1). To further explore the optimal conditions, different
1a:2a ratios from 1.2:1 to 4:1 in the presence of Rh2(esp)2 (2
mol %) were screened (entries 11−14, Table S1), and we
found that the desired product 3a was afforded in 85% yield
with a 1a:2a ratio of 2:1 (entry 14, Table S1). However,
changing the gas atmosphere to O2 or air furnished 3a in 64−
65% yield (Table S5). According to the screening of conditions
presented above, we determined that the reaction should be
conducted at 80 °C in DCE with the catalyst Rh2(esp)2 (2 mol
%). The structure of 3a was unambiguously determined by X-
ray crystallography.
In addition, D-galactals with Ac, Bz, and Bn protecting groups
were also good substrates for obtaining the desired PFGs (3i′,
88%; 3j′, 51%; 3k′, 55%). For L-rhamnal (3l′), the
cycloaddition could also successfully occur, and the structure
of 3l′ was unambiguously confirmed by X-ray crystallographic
analyses.
To demonstrate the potential of this chemistry, a gram scale
study was performed, and 3a was afforded in 51% yield
(Scheme 3a).
Scheme 3. Gram Scale Preparation and Synthetic
Applications of 3a
With the optimized reaction conditions in hand, we explored
the scope and generality of this methodology (Scheme 2).
Using 3,4,6-tri-O-acetyl-D-glucal (2a) as a model glycal, the
modified aryl groups of the triazoles (1) were first examined in
the Rh-catalyzed cycloaddition. Both electron-donating func-
tional groups (EDGs) and electron-withdrawing functional
groups (EWGs) on the benzene ring of 4-aryl-1-tosyl-1,2,3-
triazoles 1 with 2a could work well to form the desired
products 3a−3o in yields of 45−85%. The substituent position
and electronic effects obviously affected the reaction yields.
The products carrying EDGs (3b and 3l) and weak EWGs (3f,
3h, and 3j) were obtained in higher yields of 73−84%
compared to the moderate yields of those carrying strong
EWGs (3m and 3n). The products carrying the substituents at
the para position of the phenyl ring (3b, 3f, 3h, 3j, and 3l)
showed higher yields of 73−84% compared to the 61−67%
yields of those bearing the substituents at the meta position
(3c, 3e, 3g, 3i, and 3k). The substituent at the ortho position
of the phenyl ring (1d) afforded the product 3d in a much
lower yield of 45%. It was gratifying that the reaction of a
thienyl triazole (1o) also proceeded smoothly to produce the
desired 3o in a yield of 70%. However, when the substituent
was cyclohexyl, no desired product 3p was produced with the
consumption of triazole. A possible reason is that the stability
of cyclohexyl rhodium carbenoid was poor. Meanwhile, the
substituent was 1-naphthyl, the reaction also hardly occurred,
and a trace of product 3q was found.
At the same time, the N-sulfonyl groups of the triazole
substrates were also examined (3r−3z). When the substrates
with less reactive groups such as methanesulfonyl, isopropa-
nesulfonyl, and p-chlorobenzenesulfonyl groups were subjected
to the Rh-catalyzed cycloaddition, the desired products (3r−
3t) were produced in 61%, 57%, and 54% yields, respectively.
The reactions of N-phenylsulfonyl triazole (1u) and methyl-
and methoxyl-substituted phenylsulfonyl triazoles (1v−1x)
proceeded well in high yields of 70−80%. Unfortunately, we
did not obtain the desired product 3y when the phenyl ring
had a strong electron-withdrawing ester group. In addition,
when the aromatic ring is a naphthalene ring on the sulfonyl
group, the reaction hardly occurred.
Subsequently, diverse synthetic transformations from 3a
were carried out as illustrated in Scheme 3b−e. The
deprotection of 3a was attempted under alkaline conditions
with K2CO3 or NaOMe to afford 4 in 95% yield (Scheme 3b),
which constitutes the backbone of important biologically active
substances.7 Furthermore, oxidative addition of 3a with
mCPBA directly formed monoprotected diol 5 as a sole
stereoisomer in good yield,19 which was confirmed by HSQC,
HMBC NMR spectra, and a selective excitation experiment.
Then, the epoxidation of 3a with DMDO15 could yield 6 and
6′ in 90% yield with high diastereoselectivity (10/1 dr).
Next, we had a try to open the cyclic N,O-acetal moiety of 3
into polyhydroxy alkylpyrroles that possess potential biological
activities10,11 through an acid-catalyzed ring opening reaction.
When 10% TfOH,11 BF3·Et2O,20 or BCl3,21 was used in the
acid-catalyzed ring opening reactions, complex mixtures were
obtained, even at low temperatures. To our surprise, when 3a
was treated with TMSOTf,22 an unexpected product C-
nucleoside 7a was obtained in 67% yield instead of the ring
opening product polyhydroxy alkylpyrrole (Scheme 3e). The
structure of 7a was confirmed by mass spectra, HSQC, HMBC
NMR spectra, and a selective excitation experiment. Similarly,
7x can be obtained with a yield of 74% by treating 3x with
TMSOTf. C-Glycoside is an important structural unit with a
wide range of biological activities, such as cytotoxic and
antifungal activities.23 This unexpected transformation might
go by a three-step mechanism, including ring opening of the
epoxy, acetyl group migration,24 and nucleophilic substitution,
and provide a new alternative approach to such C-nucleosides
Finally, the scope of glycal substrates was examined. The Rh-
catalyzed cycloaddition of D-glycals containing various
protecting groups, such as methyl, benzyl (Bn), and benzoyl
(Bz), proceeded smoothly, forming the corresponding
products 3a′−3e′ in moderate to high yields (45−72%).
Notably, the allyl group could be tolerated, though the desired
N-glycoside 3f′ was afforded in 22% yield. The low yield of 3f′
might be attributed to the reactivity of the allyl group with
triazole. The glycal-containing Br or Ts groups were also able
to give the target products in high yields (3g′, 71%; 3h′, 80%).
C
Org. Lett. XXXX, XXX, XXX−XXX