Communication
In order to gain some insights into the mechanism of this re-
action, a series of control experiments were carried out
(Scheme 1). As shown in Scheme 1a, this transformation could
be terminated completely by introducing TEMPO. No desired
C
product 3a was observed, and CF3 radical trapping product 3
was obtained instead in 3.3% yield based on 19F NMR analysis,
which implied the radical nature of this reaction. When the re-
action was carried out in dark, the desired product 3a could
be generated in 71% yield, which showed that the light did
not affect this transformation. When the loading of NMM was
reduced to 0.1 equivalents, 3a could be obtained in 23% yield,
which suggested that NMM was served as an initiator or could
be regenerated. In the absence of NMM, the starting material
1a was fully recovered (Scheme 1b).
Figure 2. X-band EPR spectrum obtained in DMF at 298 K in the presence of
PBN. Line I: A solution of 1a (0.1 mmol), 2 (0.2 mmol), NMM (0.2 mmol), PBN
(0.2 mmol) in DMF (1 mL). Line II: A solution of 1a (0.1 mmol), 2 (0.2 mmol),
PBN (0.2 mmol) in DMF (1 mL). Line III: A solution of PBN in DMF (1 mL).
Scheme 1. Control experiments. The yield and conversions were determined
Based on our experimental observations, a possible mecha-
nism is proposed for this transformation (Figure 4). In contrast
to recent examples where the photo-activity of EDA complexes
is responsible,[4a–d] the reversible electron transfer (ET) in the
EDA complex 5 is activated thermally and lead collapse of 5
by 19F NMR with PhCF3 as the internal standard.
C
The CF3 radical could be observed using electron paramag-
netic resonance (EPR) in the presence of a spin trap tert-butyl-
a-phenylnitrone (PBN) (Figure 2). When PBN was added into
the reaction mixture under the standard conditions, a spectrum
at g=2.00646 (298 K) was recorded that could be clearly
traced to the CF3-PBN spin trap.[13]
into CF3 radical and NMM+ irreversibly.[4f–h,19] The CF3 radical
C
C
C
is then added onto arene 1a to give radical 7. Radical 7 is oxi-
C
dized by radical cation NMM+ to generate cation 8 and re-
generate NMM (path A). Ultimately deprotonation of cation 8
assisted by NMM yields the final product 3a. At this stage, oxi-
dation of 7 to 8 by Umemoto’s reagent 2 cannot be ruled out
completely (path B).
Given the weak reductive capacity of NMM (ENMM/NMM+
=
C
1.2 V vs SCE)[14] and oxidative capacity of Umemoto’s reagent 2
(E2/2 À =À0.35 V vs SCE),[15] intermolecular SET from NMM to 2
In summary, we have described originally a simple and effi-
cient method for direct CÀH trifluoromethylation of arenes
with Umemoto’s reagent. The CF3 radical is generated by an
EDA complex formed by Umomoto’s reagent and an amine,
a novel medium which is different from the reported methods.
Some experiments and theoretical calculations were furnished
to support the EDA complex. Transition-metal catalysts, direct-
ing groups and external oxidants can be avoided. The method-
ology presented here allows to access highly-functionalized
CF3-containing indoles, pyrroles, benzofurans and electron-rich
benzenes at room temperature in good chemical yields. Fur-
ther application using this EDA complex, as well as more de-
tailed mechanism investigation, are underway in our laborato-
ry.
C
is thermodynamically disfavored. It is known that tertiary
amines are able to form EDA complexes with electron-accept-
ing molecules, so this SET event can be facilitated by an EDA
complex of 2 with NMM (Figure 3a).[12] The existence of the
1
EDA complex could be supported by the H NMR spectroscopy
and UV/VIS spectrum.[1h,16] Using Job’s method of continuous
variations, a molar donor/acceptor ratio of 1:1 in solution for
EDA was readily established.[17] Concomitantly the equilibrium
constants KEDA (KEDA =18.7) for formation of the EDA complex
were determined spectrophotometrically (using the Benesi–Hil-
debrand method).[18] Theoretical calculations showed that the
formation of the EDA complex was thermodynamically favored
(for details, see Supporting information) (Figure 3d). This band
is associated with an electron-transfer transition from HOMO
(H) to LUMO (L) of this complex. For details, see Supporting In-
formation). The formation of the EDA complex was further sup-
ported by NMR titration (Figure 3b). The appearance of a new
Experimental Section
A 3 mL nap vial was equipped magnetic stir bar and was charged
with a solution of arene or heteroarene 1 (0.2 mmol, 1.0 equiv),
1
set of signals in the H NMR spectroscopy of the mixture of
NMM and 2 provided proof of the EDA complex.
Umemoto’s reagent 2 (0.4 mmol, 2.0 equiv), NMM (0.4 mmol,
Chem. Eur. J. 2015, 21, 1 – 6
3
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ