Angewandte
Chemie
Table 1: Stability of 3[a] in different solvents and in the presence of
additives.
Entry Solvent
d(19F)
3d
7d
[ppm]
(RT)
(RT)
(12 h, 508C)
1
[D6]DMSO
À48.6
À48.6
À48.6
À49.7
À49.0
n.c.
n.c.
n.c.
n.c.
n.c.
n.c.
n.c.
n.c.
n.c.
n.c.
n.c.
–
–
n.c.
n.c.
2[b]
3[c]
4[d]
5[d]
6
[D6]DMSO
[D6]DMSO
[D7]DMF
CDCl3
CDCl3/DBU
À49.0 dec.: 40%[e] dec.[f]
–
n.c.
Scheme 2. Electrophilic trifluoromethylthiolating agents.
7
8
9
CDCl3/lutidine À49.0
n.c.
n.c.
n.c.
n.c.
n.c.
n.c.
n.c.
n.c.
–
–
–
n.c.
n.c.
n.c.
n.c.
–
CDCl3/Et3N
CDCl3/D2O
[D8]toluene
[D8]THF
À49.0
À49.0
À44.5
À50.6
À50.4
À50.1
À51.3
dec.: <5%
n.c.
10[d]
11[d]
12[d]
13[d]
14
n.c.
n.c.
n.c.
n.c.
explore its synthetic utility in direct trifluoromethylthiola-
tions of alkynes and boronic acids.
[D6]acetone
CD3CN
CD3OD
To develop an efficient synthesis of 4 we decided to
perform a series of studies. The cyclic voltammogram of N-
chlorophthalimide showed a reduction peak at À0.45 V
versus SCE in acetonitrile whereas the nucleophilic trifluoro-
methylthio reagents AgSCF3 and CuSCF3 were oxidized at
potentials of + 0.37 V and + 0.81 V, respectively. This data
showed the possibility of a redox transformation of formal
F3CSÀ to F3CS+ and reduction of formal Cl+ to ClÀ (see the
Supporting Information for spectra). Therefore, we decided
to firstly explore the reaction of N-chlorophthalimide with
AgSCF3 in acetonitrile (Scheme 2).
However, in this case, the expected redox reaction turned
out to be quite challenging because of the formation of
a substantial amount of F3CSSCF3 (5) as the main product,
thus suggesting a radical mechanism. Evaluation of various
reaction conditions such as solvent and temperature did not
show any improvement. Interestingly, when CuSCF3 in
acetonitrile was used, the desired product was formed in
high yield with almost no formation of 5; approximately 3%,
as determined by 19F NMR analysis of the crude reaction
mixture. This result might be explained by a nonradical
n.c.
[a] NMR experiment conditions: 4 (9.25 mg, 0.037 mmol) and ethyl
trifluoroacetate (internal standard, 5.3 mg, 0.037 mmol) in 0.75 mL of
appropriate solvent. Final concentration of reagent was 50 mm.
Decomposition as detected by signals with respect to the internal
standard. [b] 21.09 mg (5 equiv) of TFA was added. [c] 11.1 mg
(15 equiv) of D2O was added. [d] (Trifluoromethyl)benzene (5.5 mg,
0.037 mmol) was used as an internal standard. [e] 40% decomposition
after 15 min. [f] Complete decomposition after 24 h. DBU=1,8-
diazabicyclo[5.4.0] undec-7-ene, DMF=N,N-dimethylformamide,
DMSO=dimethylsulfoxide, n.c.=no change, dec.=decomposition,
THF=tetrahydrofuran.
3 h) showed no signs of decomposition. The presence of
transition-metal ions (e.g. Cu+, Cu2+, Fe2+, Fe3+, Pd2+) also
does not decompose 4.
To our delight, 4 was totally inactive toward acidic
conditions (e.g. AcOH, MsOH, TsOH, TfOH) and the
presence of water. In contrast, 4 can be sensitive to strong
bases. In the presence of weak bases such as lutidine there was
no decomposition, and Et3N as a moderate base gave less than
5% decomposition after 12 hours at 508C. In the presence of
a strong base such as DBU, 4 decomposed in short time.
Noteworthy, by 19F NMR spectroscopy we realized that
during this process new species were formed: F3CSSCF3
(d = À45.6 ppm), F3CSOH (d = À5.2 ppm), F3CSO2H (d =
À
pathway. Insertion of CuSCF3 into the N Cl bond could form
a copper(III) species, which undergoes reductive elimination
to provide the desired product (Scheme 2). There is no
evidence for the formation of the amino/CuIII/SCF3 species,
but Dobbie and Emeldus reported the preparation of
trifluoromethylthio(bistrifluoromethylamino)mercury,[19]
thus showing the possibility to form aminometallic SCF3
species and indicating that such a copper intermediate could
be feasible.
À
À81.8 ppm), and the corresponding anion F3CSO2 (d =
À86.1 ppm).[20]
Subsequently, the reaction was scaled up to over 5.0 grams
and the desired product 4 was isolated as colorless crystals in
90% yield after flash chromatography on silica gel and fully
characterized by spectroscopic techniques. The compound 4
melts without decomposition at 115–1168C (lit. 115–1178C)
and could be easily sublimed at 808C/1 mbar. To gain
evidence for the potential utility and possible limitations of
this reagent, we investigated its stability. Accordingly, 4 is not
light sensitive and it is stable in both the solid state and in
solution. NMR studies showed no decomposition of the
reagent under aerobic conditions in solvents such as
[D6]DMSO, [D7]DMF, CDCl3, [D8]toluene, [D8]THF,
[D6]acetone, CD3CN, and CD3OD (Table 1). Standing in
solution at room temperature for seven days, or heating of the
solutions at 508C for 12 hours (DMF and toluene at 1008C/
With this shelf-stable electrophilic trifluoromethylthiolat-
ing reagent 4 in hand, we started to test its usefulness in
reaction with representative C nucleophiles such as boronic
acids and terminal acetylenes. Initial attempts showed that
there was no product formation when heating in diglyme or
DCE (see the Supporting Information, and Tables 2 and 3)
Hence, we turned our attention to copper catalysis which is
widely applied in transformations involving boronic acids and
alkynes. We observed that the reactions proceeded only in the
presence of copper(I) salts whereas copper(II) salts showed
no activity. Further examination of various commercially
available copper sources revealed that CuCl and CuI were the
most appropriate for the reaction with boronic acids and only
CuI showed good catalytic activity for the functionalization of
alkynes. Interestingly, the ligand had a strong influence on the
Angew. Chem. Int. Ed. 2014, 53, 1650 –1653
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1651