Chen-Chen Zhang, Zhi-Peng Huo, Mei-Lin Tang et al.
Tetrahedron Letters 68 (2021) 152946
ica gel column (Table 1, entries 1 and 2). When Cu(OTf)2 was used,
the desired product 12a was obtained in 26% yield[15e] (Table 1
entry 3). Other triflates, such as Sc(OTf)3, Zn(OTf)2, Yb(OTf)3, In
(OTf)3 and Ce(OTf)3, could also afford the desired product 12a with
36%-51% yields (Table 1, entries 4–8) because of incomplete reac-
tion. Similarly, strong Lewis acid TiCl4 could only afford trace
amount of desired product 12a (Table 1 entry 9). Delightfully, BF.3-
OEt2 (1.0 equiv.) and TMSOTf (1.0 equiv.) could lead to the desired
12a in moderate yields (Table 1, entries 10–11). When the TMSOTf
increased to 2 equiv., the desired product 12a was achieved in 80%
yield (Table 1, entry 12). Further warming the reaction mixture to
room temperature could not improve the yield of 12a (Table 1,
entry 13). In addition, other solvents like toluene, THF and Hexanes
were also screened, and none of them led to the better yields
(Table 1, entries 14–16). Especially in Hexanes, the reaction
became complex with great amount of side reaction.
With the above optimized reaction conditions in hand, different
N-Boc lactams 7a-e and ynamides 10a-f were examined for this
one-pot transformation, and the results were summarized in
Fig. 1. Natural products containing [1,2-c][1,3]oxazine skeleton.
Scheme 1. N-Boc c-lactam 7a could react with different ynamides
10a-e, affording the desired 12a-e in moderate yields (68%-80%).
N-Boc d-lactam 7b also worked well, albeit leading to the desired
products 12f and 12g in slightly lower yields (56% and 61%, respec-
tively). Notably, the reactions of N-Boc e-lactam 7c with N-(Ts)
alkyl or Bn ynamides could afford the desired products 12h, 12i
and 12j in excellent yields (70%–87%). Importantly, 7c could react
with oxazolidone-substituted ynamide 10f to afford 12k in 45%
yield. Substituted N-Boc
c-lactams were also investigated. The
reactions of 2-OTBS substituted 7d with ynamides 10a-b could
afford the desired 12l and 12m with excellent diastereoselectivi-
ties (dr > 99:1) and in moderate yields [9a]. It was worth
mentioning that N-Boc b-lactam 7f could not work for this
transformation, probably due to the small ring tension effect
during the reduction-[4 + 2] cycloaddition process.
Next, the benzoimides 8a-c were examined under optimal reac-
tion conditions, and the results are shown in Scheme 2. tert-Butyl
2-oxo-3,4-dihydroquinoline-1(2H)-carboxylate 8a (n = 1) could
react with ynamides 10a-d to give the desired product 13a-d in
moderate yields (43%-70%), whereas the reaction of tert-butyl 2-
oxoindoline-1-carboxylate 8b (n = 0) and ynamide 10a could not
lead to any desired product. Notably, tert-butyl 2-oxo-2,3,4,5-
tetrahydrobenzo[b]azepine-1-carboxylate 8c (n = 2) could react
with ynamides 10a and 10b to afford the desired 13e and 13f in
54% and 67% yields respectively. The structures of 13a-f were
unambiguously confirmed by the X-ray crystallographic analysis
of compound 13a (see Supporting Information).
Then we turned our attention to investigate the application of
linear imides for this one-pot transformation, and the results are
summarized in Scheme 3. The imide substrate methylamine
(tert-butyl acetyl(methyl)carbamate) 9a worked well with yna-
mides 10a-c, 10e, 10g under the optimized reaction conditions,
giving the desired 3,4-dihydro-1,3-oxazin-2-ones skeleton 14a-e
in moderate yields (57%–80%). Similar results were obtained for
the imide substrates 9b and 9c, and the desired products 14f-m
were afforded in moderate yields (57%-87%). Unfortunately, the
imide substrates 9d and 9e with phenyl group for either R1 or R2
led to no desired products.
Fig. 2. The reactions of N-acyliminium ions.
tinuous efforts in exploring novel heterocyclic compounds as
potential innovative drugs [16], we envisioned that functionalized
1,3-oxazin-2-one skeletons 1, 2 and 3 could be conveniently pre-
pared in a one-pot fashion from N-Boc protected lactams. Herein
we present our one-pot approach to access 1,3-oxazin-2-ones
through TMSOTf-mediated successive reduction-[4 + 2] cyclization
process (Fig. 2, d).
Finally, two chiral ynamides 11a and 11b were selected to
investigate the potential stereochemical control of this one-pot
transformation in Scheme 4. The achiral oxazolidone-substituted
ynamide 10f could react with linear imide 9a to give the desired
15a in 44% yield. When chiral oxazolidone-substituted ynamide
11a (R4 = Bn, R5 = H) was used, the desired 15b was produced in
59% yield and with moderate diastereoselectivity (dr = 78:22). Sim-
ilar diastereoselectivity (dr = 75:25) was observed for the desired
product 15c when another chiral oxazolidone-substituted ynamide
Our investigation started with the reaction of imide 7a with
ynamide 10a. The general one-pot transformation was conducted
as followed: After the imide 7a reacted with lithium triethylboro-
hydride [17] (1.1 equiv.) at À78 °C for 0.5 h, the resulting mixture,
without quenching, was treated with the ynamide 10a and Lewis
acid in turn. A variety of Lewis acids were screened for this process.
Initially, AgOTf and AgNTf2 both afforded a lot of by-products, and
trace amount of desired product which was hardly isolated by sil-
2