1180
I. Nakamura et al. / Tetrahedron Letters 55 (2014) 1178–1182
all; a trace amount (4%) of the imine 4a was obtained along with
65% of the recovered 1a (entry 7). The reaction of arenesulfonyl
isocyanate 2b, which possesses an electron-deficient aryl group,
resulted in the predominant formation of the undesired imine
4b, while those of N-benzoyl- and N-(p-nitrophenyl)isocyanate
were sluggish (entries 8–10). In terms of the ligands, bulky and
electron-donating phosphine ligands such as t-Bu3P and SPhos8 re-
sulted in higher yields of 3a (entries 11–13). Reaction conditions,
such as concentration and loading amounts of SPhos, were opti-
mized to ultimately afford 3a in an acceptable yield (entry 17).
During the optimization process, we noticed the formation of a
small amount of byproduct 6 due to the migration of the phenyl
group in all entries in Table 1, except for entries 7 and 9. For exam-
ple, 6a (R4 = Ts) was obtained in 9% yield under the optimized reac-
tion conditions (entry 17).
The ratio between the hydrogen-migration product 3 and the
aryl-migration product 6 is apparently dependent on the electronic
character of the oxime substituent R3, as summarized in Table 2.
Substrates 1b and 1c, which possess an electron-deficient aromatic
group, selectively favored hydrogen-migration, whereas substrates
with electron-rich aromatic groups shifted the reaction towards
aryl-migration (entries 1–4). It should be noted that, for the reac-
tions between 1e and 2a (entry 5), the migration of the cyclohexyl
group was not observed.
Next, the substitution effects on the propargyl group of oxime
(E)-1 were investigated, as summarized in Table 3. For substituents
on the alkyne terminus, the reaction was able to tolerate both aryl
and alkyl substituents (entries 1–4). Substitution at the propargylic
position by an electron-donating anisyl group resulted in a rapid
reaction, affording the desired product 3m in a good yield (entry
5). Arylaldoxime 1l, which possesses an electron-withdrawing
p-trifluoromethylphenyl group at R2, was converted in the absence
of SPhos into product 3o in an acceptable yield (entry 7); the
reaction of 1l and 2a in the presence of SPhos resulted in lowering
the chemical yield (34%). Substrate 1m, which possesses an alkyl
group at the propargylic position was significantly less reactive,
with the recovery of 70% of 1m (entry 8). It should be noted that
for every reaction (listed in Table 3), byproducts were also
detected; specifically, phenyl-migration product 6 (ca. 5–10%)
and N-tosylimine 4a (ca. 10–20%) (see Supporting information).
Using the optimized reaction conditions, the reaction of Z oxime
(Z)-1b and 2a afforded the same dihydropyrimidine 3e as that of E
oxime (E)-1b (Eq. 2 and Table 2, entry 1). In the absence of the
copper catalyst, however, the reaction favored the formation of
tosylimine 4e rather than 3e.
Ar
H
Ar
Ts
Ph
Ts
w/ or w/o Cu catalyst
DCE, 80°C
N
N
N
N
Ts
O
+
+
•
O
N
Ph
Ar
Ph
ð2Þ
H
4e
Ph
2a
(1.2 equiv)
3e
(Z)-1b
Ar: p-F3CC6H4
w/ CuBr, SPhos
w/o Cu catalyst
30%
<1 %
20%
43%
To gain insight into the mechanisms of the reaction, specifically
that of the hydrogen atom at the oxime moiety, the reaction was
carried out using deuterium-labeled substrates (Eqs. 3 and 4).
The reaction of deuterated (E)-1a-d (99%-deuterated at the oxime
moiety) with N-tosylisocyanate 2a afforded 3a-d (15% yield, 99%-
deuterated at the 5-position), along with the phenyl-migration
product 6a-d (15% yield) and N-tosylimine 4a-d (Eq. 3). The results
clearly indicate that the hydrogen atom at the 5-position of dihy-
dropyrimidine 3 is attributable to the aldoxime hydrogen atom
of substrate 1. Because the yield was significantly affected by deu-
terium-labeling (lower yield of 3a-d versus 3a, Table 1, entry 17),
the reaction was repeated using a 1:1 mixture of 1a and 1a-d (total
amount, 1 mmol) with isocyanate (1.2 mmol) under standard con-
ditions (Eq. 4). The results showed that the deuterium content of
the product 3a-d at the 5-position was 29%, suggesting that the
hydrogen-migration step was decelerated by deuterium-labeling.
Accordingly, the deuterium-to-hydrogen ratio of the phenyl-
migration product 6a-d, which formed without involving C-H(D)
bond cleavage, was higher than 50%. Furthermore, because the
deuterium content of recovered substrate 1a was 50%, there is no
difference in the consumption rate of 1a and 1a-d, which implies
that the hydrogen migration process occurs (at the least) after
the irreversible elemental process(es).
D
Ph
10 mol % CuBr
11 mol % SPhos
Ts
N
•
O
N
O
+
DCE (0.25 M), 80 ºC
18 h
Ph
2a
Ph
1a-d
(99% D)
Ph
D
ð3Þ
Ts
Ts
Ts
D
N
N
N
N
N
+
+
Ph
Ph
Ph
Ph
Ph
D
Ph
6a-d
15% yield
(99% D)
3a-d
15% yield
(99% D)
4a-d
23% yield
(99% D)
Table 3
Cu-catalyzed reactions of 1f–m and N-tosylisocyanate 2aa
H
Ph
Ph
H
O
Ph
D
O
Ph
10 mol % CuBr
Ts
11 mol % SPhos
N
Ts
N
N
N
N
Ts
O
+
+
•
+
O
N
as above
2 h
•
N
O
DCE, 80 °C
R2
R1
R2
Ph
Ph
H
3
R1
2a
Ph
Ph
2a
1.20 mmol
1
(E)-
1a
1a
-d
0.50 mmol
0.50 mmol
1
R1
R2
Time (h)
3
Yieldb (%)
ð4Þ
H/D
Ph
Ph
H/D
1
2
3
4
5
1f
p-MeOC6H4
p-ClC6H4
p-F3CC6H4
Cy
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
10
18
12
24
6
10
12
48
3i
3j
51
45
48
55
64
63
43
7d
Ts
Ts
Ts
N
O
1g
1h
1i
N
N
N
N
N
+
+
+
3k
3l
3m
3n
3o
3p
Ph
H/D
Ph
Ph
Ph
Ph
Ph
H/D
Ph
Ph
1j
p-MeOC6H4
p-MeC6H4
p-F3CC6H4
Cy
6
1k
1l
1m
1a
3a
6a
-d
0.10 mmol
(H/D=36/64)
4a
-d
0.14 mmol
(H/D=71/29)
-d
7c
8
0.26 mmol
0.14 mmol
(H/D=50/50)
(H/D=36/64)
a
The reactions of (E)-1 (0.20 mmol) and N-tosylisocyanate 2a (0.24 mmol) were
Based on our experimental results, a probable reaction mecha-
nism between 1 and 2a is illustrated in Scheme 2, which is based
on the -acidity and on the single-electron redox ability of the cop-
per catalyst. First, the -acidic copper catalyst coordinates to the
triple bond of 1 to form -complex 7. Next, the enhanced electro-
carried out in the presence of CuBr (10 mol %) and SPhos (11 mol %) in 1,2-dichlo-
roethane (0.8 mL) at 80 °C.
p
b
Isolated yields.
Without SPhos.
70% of 1m was recovered.
c
p
d
p