Communication
Using 1a’ with a methyl group at the propargylic position as
substrate, a small amount of hydrotrifluoromethylation prod-
uct was formed, as shown by 19F NMR and GC-MS spectra of
the crude reaction mixture (see the Supporting Information).
Considering the Thorpe–Ingold effect,[12] which demon-
strates that increasing the size of two substituents on a tetrahe-
dral center leads to enhanced reactions between parts of the
other two substituents in the cyclization reaction, we prepared
alkyne 1b with two methyl substituents at the propargyl posi-
tion and continued to study its carbotrifluoromethylation. Grat-
ifyingly, the corresponding five-membered cyclic products 3b
were obtained as a mixture of two isomers in 61% overall
yield with a ratio of 11:1 under the same conditions (Table 1,
total yield of products E- and Z-3b was slightly lowered. Ume-
moto’s reagent (2c) failed to give any cyclic product with full
recovery of the starting materials (Table 1, entries 2 and 3).
Next, we focused on the catalyst and solvent effects. However,
various copper salt catalysts, such as CuCl, CuI, [(CH3CN)4Cu]-
[PF6], CuBr, Cu(OTf)2, and CuBr2, could not improve the yield of
3b in DCE (Table 1, entries 4–9). Using dichloromethane (DCM)
as solvent, the CuI-catalyzed reaction afforded 3b in 63% over-
all yield with high selectivity, whereas the total yield was only
41% with CuBr·Me2S (Table 1, entries 10 and 11). Therefore, CuI
was chosen as the catalyst to further study the solvent effect.
Much lower yields were obtained with N-methyl-2-pyrrolidone
(NMP), acetonitrile, dioxane or dimethyl sulfoxide (DMSO) as
solvent (Table 1, entries 12–15). Ethyl acetate and tert-butyl al-
cohol were also less effective for the reaction (Table 1, en-
tries 16 and 17). Therefore, DCM was identified as the best sol-
vent for this reaction among the solvents tested. Further inves-
tigation showed that the catalyst loading could be lowered to
10 mol% without any loss in yield or stereoselectivity (Table 1,
entry 18). Carrying out the reaction in DCM under reflux and
increasing the amount of 1b to 2.0 equivalents gave isomers
E- and Z-3b in 67% combined yield (Table 1, entry 19). At-
tempts to use PPh3 or PCy3 as an additive did not further im-
prove the yield (Table 1, entries 20 and 21). Surprisingly, the
total yield of E- and Z-3b was improved to 74% in the pres-
ence of B2Pin2 (Table 1, entry 22). Control experiments indicat-
ed that the copper salt was important for this reaction
(Table 1, entries 23 and 24). Thus, the optimal conditions were
set to 10 mol% of CuI, 2.0 equivalents of 1b, and 10 mol% of
B2Pin2 in DCM under reflux.
Table 1. Optimization of the reaction conditions.[a]
Entry
[Cu]
Additive
Solvent
E-3b/Z-3b
Yield [%][b]
1
CuBr·Me2S
CuBr·Me2S
CuBr·Me2S
CuCl
—
—
—
DCE
DCE
DCE
56:5
53:5
NR
2[c]
3[d]
4
—
DCE
41:4
47:5
47:5
54:5
38:4
44:4
37:4
57:6
13:<5
14:<5
19:<5
37:4
49:5
50:5
58:6
61:6
24:<5
52:5
67:7
NR
5
6
7
8
9
10
11
12
13
14
15
16
CuI
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
DCE
DCE
DCE
DCE
[(CH3CN)4Cu][PF6]
CuBr
Cu(OTf)2
CuBr2
CuBr·Me2S
CuI
With the optimized reaction conditions in hand, we studied
the scope of alkyne substrates for this copper-catalyzed carbo-
trifluoromethylation reaction. As shown in Table 2, a range of
terminal alkynes 1 with different substituents in the benzene
ring could react with 2a and give the expected carbotrifluoro-
methylation products in moderate to good yields. Both elec-
tron-rich and electron-deficient alkynes are compatible with
DCE
DCM
DCM
NMP
CH3CN
dioxane
DMSO
EA
tBuOH
DCM
DCM
DCM
DCM
DCM
DCM
DCM
CuI
CuI
CuI
CuI
CuI
CuI
CuI
CuI
CuI
CuI
CuI
—
17
18[e]
19[e,f,g]
20[e,f,g]
21[e,f,g]
22[e,f,g]
23[g]
24[g]
PPh3
PCy3
B2Pin2
B2Pin2
—
—
NR
[a] Reaction conditions: 1b (0.15 mmol), 2a (0.1 mmol), copper catalyst
(0.02 mmol), solvent (1 mL), 608C, 4 h, under N2; [b] determined by
19F NMR spectroscopy using hexafluorobenzene as internal standard;
[c] Togni reagent I was used as trifluoromethylating reagent; [d] Umemo-
to reagent was used as trifluoromethylating reagent; [e] 0.1 equivalents
of CuI were used; [f] 0.2 mmol of 1b was used; [g] the reaction was car-
ried out in DCM at reflux.
entry 1). The stereochemistry of the newly formed double
bond in the major product was established as E by NOE experi-
ments. Inspired by the result, we screened different conditions
to find a suitable protocol for the selective formation of cycli-
zation products. Using Togni reagent I (2b) instead of 2a, the
Figure 1. ORTEP representation of major isomer 3k.
Chem. Eur. J. 2014, 20, 15315 – 15319
15316
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim