4
L. Luo et al. / Tetrahedron Letters xxx (2013) xxx–xxx
Table 3 (continued)
Entry
Reactant 12
Reactant 7
Product 11
Yieldsb (%)
NC
NH2
S
S
8
12g
7c
85
O
11m
O
O
O
NH2
O
O
O
N
N
H
S
S
O
O
9
7a
76
O
11n
12i
a
Reaction conditions: 12 (1 mmol), NBS (1 mmol), 3 mL EtOH, rt, 10 min (30 min for 12b); then 7 (1 mmol), rt, 20 min.
Isolated yields.
b
O
S
CN
O
O
O
O
R3
R4
Thorpe-Ziegler
Cyclization
A
Bromination
R3
R4
R3
R4
7
NC
S
R1
Br
A
SN2 substitution
1 equiv NBS
R1
O
14
13
12
O
R3
R1
S
O
Retro Ene Reaction
A
R4
O
Path A
R4CO2Et
N
H
H
O
O
NH
16
R3
R4
EtOH
S
R3
R1
A
O
R3
S
A
O
R1
S
NH2
R1
O
15
11
A
PhNCO
Path B
R4=NHPh
N
N
Ph
Retro Ene Reaction
H
H
17
Scheme 2. Proposed mechanisms for tandem reaction.
To test our hypothesis, we selected reagent 7a and pentane-2,4-
dione 12a for the model reaction. To our delight, -halogenation of
ally, the stronger electronic withdrawing group in 1,3-dicarbonyl
compounds 12 was preferentially removed. For example, CH3CO–
was easier to be eliminated than C2H5OCO– and PhCO– (Table 2,
entries 2 and 4). However, C6H5NHCO– was unexpectedly easier
to leave than CH3CO– and BnCO–(entries 6–7, Table 2). And the
presence of active hydrogen in the amide was essential for the high
leaving reactivity of C6H5NHCO– (Table 2, entry 6 vs entry 8).
According to the experimental results, the order of the leaving
reactivity of –COR groups was determined as: C6H5NHCO– > CH3-
CO– > C6H5CO– > C2H5OCO–, C6H5N(CH3)CO–.
Subsequently, the scope of the transformation was further ex-
panded with different mercaptonitriles 7 and typical 1,3-dicar-
bonyl compounds, such as b-diketones 12a, b-ketoester 12b, and
b-ketoamides 12g and 12i, respectively (Table 3).14 7a-b reacted
with 12 affording the thiazole derivatives, while using 7c–e as
reactants led to the thiophene derivatives. All the reactions were
performed smoothly affording the desired products in good yields
(72–86%) except 11l (53%) within 1 h at room temperature. Among
them, 11j (81%, Table 3, entry 5) and 11n (76%, Table 3, entry 9)
were regarded as key intermediates of PI3K inhibitors 5 and
tubulin polymerization inhibitor 1, respectively. According to the
literature, 11j was prepared in 67% yield from odorous ethyl
a
12a with NCS was rapidly completed in ethanol at room tempera-
ture within 10 min. After addition of 7a and Et3N, the mixture was
stirred for 20 min, affording the desired product 11a in 74% yield
(Table 1, entry 1). Encouraged by this success, the reaction was fur-
ther optimized by changing halogenating agent, solvent, tempera-
ture, and base (Table 1, entries 2–10). When NBS was used instead
of NCS, the yield was improved (81%, Table 1, entry 2). And using I2
as the halogenating agent led to a poor yield (25%, entry 3, Table 1).
Thus, NBS is a more suitable halogenating agent for this transfor-
mation. However, changing the amount of NBS did not improve
the yields (Table 1, entries 4–5). Then the solvents were screened
and ethanol was found to be superior to others (Table 1, entries
6–9). It was worth noting that the product was obtained in similar
good yield under base-free condition (Table 1, entry 10).
With the optimized conditions in hand (Table 1, entry 10), we
next investigated the substrate scope of 1,3-dicarbonyl com-
pounds. As shown in Table 2, 7a reacted smoothly with various
1,3-dicarbonyl compounds 12, including b-diketones 12a and
12c-d, b-ketoesters 12b and 12e, and b-ketoamides 12f–h, afford-
ing 2-methylthio-4-aminothiazoles 11a–e in 63–82% yields. Usu-