Full Paper
volved in the reaction, the reaction was then carried out under
open air. To our delight, compound 4a was obtained in 57%
yield with good diastereoselectivity (19:1) (Table 1, entry 17).
With optimal reaction conditions in hand, we first surveyed
the substrate scope of the thermally induced rearrangement of
1 (Table 2). Other N-protecting groups such as 4-bromobenze-
nesulfonyl (Bs) and methylsulfonyl (Ms) were well compatible
and the corresponding products 2b and 2c were obtained in
71 and 54% yields, respectively. For substrate 1d with ethoxy
instead of methoxy, the reaction gave the desired product 2d
in moderate yield, perhaps due to increased steric hindrance.
Substrates with electron-donating groups (R3 or R4) led to rela-
tively higher yields (2 f–2g, 2k–2m) than those with electron-
withdrawing ones (2h–2j, 2n), suggesting that a cationic inter-
mediate may be formed during the reaction process. In addi-
tion, R4 could be extended to vinyl or naphthyl groups and the
reactions delivered the corresponding products 2o and 2p in
45 and 75% yields. The reaction was sluggish upon changing
R4 to a furyl ring, and the corresponding product 3q was ob-
tained only in 14% yield.
and 2). Substrates with a larger steric hindrance on the oxygen
atom suffered lower efficiency and diastereoselectivity (Table 3,
entries 3 and 4). It should be noted that BnOH was detected in
the system when 1e was used as the substrate.[19] The elec-
tronic effect (R3 and R4) was also investigated. Generally, elec-
tron-rich substrates resulted in better reaction outcomes than
those electron-deficient substrates (Table 3, entries 5–14). In
particular, substrate 1j bearing a fluorine atom on R3 was less
reactive under the standard conditions, and only a trace of the
1
corresponding product was detected by H NMR spectroscopy
when the reaction was carried out at even higher temperature
(Table 3, entry 10). Substrates with a naphthalene ring, furyl
ring, or vinyl group were also tolerated in the reaction (Table 3,
entries 15 and 16). For substrate 1t, which bears methyl group
as R4, the corresponding products 3t could be furnished as the
only product under more meticulous conditions, whereas the
standard conditions resulted in complex mixtures (Table 3, en-
tries 18). Interestingly, using thiophene-based triazole 1r as
substrate, the corresponding product 4r could be obtained ex-
clusively in 39% yield along with good diastereoselectivity
(Table 3, entry 19). All these cyclized products 3 and 4 could
easily be separated by silica gel flash-column chromatography
or preparative thin layer chromatography (PTLC).
Next, we turned our effort to investigate the substrate scope
of the rhodium(II)-catalyzed rearrangement of N-sulfonyl tria-
zoles 1 (Table 3). Substrates with Bs and Ms N-protecting
groups were suitable in this reaction, giving the desired prod-
ucts in 46 and 63% total yields, respectively (Table 3, entries 1
To investigate the mechanisms of these reactions, several
deuterium labeling experiments were conducted.[19] By adding
D2O into the thermally induced reaction, the corre-
sponding deuterated product [D]-2a was obtained
Table 2. Reaction scope: Thermally induced rearrangement.[a]
(Scheme 2a). For the rhodium-catalyzed reaction, no
deuterium incorporation was observed in all the iso-
lated products due to the proton exchange in silica
gel chromatography during the purification, where-
as the deuterium-labeled product [D]-4a could be
1
detected by H NMR spectroscopy of the crude reac-
tion residue (Scheme 2b). For substrate [D]-1a, the
thermal-induced rearrangement reaction gave the
corresponding deuterated product [D]-2a, indicating
that a 1,2-H shift takes place in the reaction process
(Scheme 2c). In the rhodium-catalyzed reaction of
[D]-1a, the deuterated product [D]-3a was obtained
together with 4a (Scheme 2d).
Based on the above results and the previously re-
ported literature,[11–17] plausible mechanisms for
these reactions have been proposed using 1a as
a model substrate in Schemes 3 and 4, respectively.
As illustrated in Scheme 3, the thermally induced
transformation begins with rapid denitrogenation of
the N-sulfonyl triazole 1a, affording carbene inter-
mediate A, which then induces a 1,2-H shift to give
ketenimine intermediate A’. As shown in path a, the
methoxy group may attack the imino carbon to
form intermediate B. The following ring opening
and rearrangement give the corresponding zwitter-
ionic intermediate C and its resonanced intermedi-
ates C’ and C’’, in which the carbocation could be
stabilized by the adjacent aromatic rings. The subse-
[a] Reaction conditions: Under argon, 1 (0.2 mmol) was heated in ClCH2CN (2 mL) in
a sealed tube at 80oC for 12–24 h. Yields are isolated yields of purified products.
quent cyclization gives product 2a. Alternatively, ke-
tenimine A can also undergo a 1,5-OMe shift to give
Chem. Eur. J. 2016, 22, 5727 – 5733
5729
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim