Mg-Cu bimetal system for selective C-F bond activation

Article abstract of DOI:10.1016/j.jfluchem.2013.02.001

Bimetal system of Mg-CuCl in DMI for selective C-F and C-Cl bonds activation has been examined. This article involves (1) a short overview of bimetal system of Mg-metals for selective C-F bond activation, (2) C-F bond activation of alkyl trifluoroacetates

Full text of DOI:10.1016/j.jfluchem.2013.02.001

Journal of Fluorine Chemistry 152 (2013) 84–89
Contents lists available at SciVerse ScienceDirect
Journal of Fluorine Chemistry
journal homepage: www.elsevier.com/locate/fluor
Mg–Cu bimetal system for selective C–F bond activation
*
Shinya Utsumi, Toshimasa Katagiri , Kenji Uneyama
Department Applied Chemistry, Faculty of Engineering, Okayama University, Okayama 700-8650, Japan
A R T I C L E I N F O
A B S T R A C T
Article history:
Bimetal system of Mg–CuCl in DMI for selective C–F and C–Cl bonds activation has been examined. This
article involves (1) a short overview of bimetal system of Mg-metals for selective C–F bond activation, (2)
C–F bond activation of alkyl trifluoroacetates leading to alkyl difluoro(trimethylsilyl)acetates via
difluoroketene silyl acetal, (3) selective C–Cl bond activation of 3- and 4-chloropentafluoroethylben-
zenes, (4) C–F bond activation of substituted benzotrifluorides and (5) microscopic observation of an
active site of magnesium surface formed in Mg–CuCl system and elucidation of reaction mechanism of
C–Cl and C–F bond activation at the active site.
Received 29 November 2012
Received in revised form 29 January 2013
Accepted 1 February 2013
Available online 28 February 2013
Keywords:
Magnesium
ß 2013 Elsevier B.V. All rights reserved.
Mg–Cu bimetal
Defluorination
Difluoro(trimethylsilyl)acetate
Grignard reagent
1. Introduction
the generated Mg(II). It is needed for us to devise reaction
conditions so as to overcome this unfavorable property.
C–F bond activation has been one of the current active subjects,
which has been utilized for an effective transformation of
trifluoromethyl or perfluoroalkyl groups to difluoro or monofluor-
oalkyl moieties. Even though the C–F bond activation is not easy in
general due to the C–F bond strength, mechanism-based rational
C–F bond activations have been proposed and actually utilized
successfully [1]. Among them, metal magnesium can highly
promote the C–F bond activation since it has a reduction potential
feasible for C–F bond cleavage as shown in Table 1. And also metal
magnesium is easily handled and its use is economically feasible
and environmentally friendly [2].
Another factor for controlling the Mg-promoted defluorination
is how to match the LUMO level of a substrate with practical
reducing power of metal Mg. The reactivity of trifluoro-
methyl(CF3)-bearing substrates for Mg(0)-promoted reductive
defluorination can be explained by LUMO energy levels of the
substrates as shown in Supplementary Figs.
2 and 3. The
reactivities of the substrates are deduced by reaction times,
conversion of the substrates, and product yields and they are lined
with the order of LUMO energy levels which are calculated by AM1
geometry optimization of substrates with MacSpartan plus.
Supplementary Figs. 2 and 3 indicate that lower LUMO level of
Some of the successful Mg-metal promoted transformations of
trifluoromethyl building blocks are shown in Fig. 1 [3].
p
-system of the substrate enables smooth electron transfer from
metal Mg and thus gives the defluorinated compounds in higher
yields; in the case of higher LUMO level of the -system, reductive
Meanwhile, metallic Mg possesses potentially unfavorable
property for reduction and Grignard reagent preparation, which
results in limited reductive power and unfavorable induction
period, respectively. To access this problem, some kinds of
activations such as elevated reaction temperature, use of an
activator such as iodine and 1,2-dibromoethane, ultrasonic
irradiation, and vigorous mechanical stirring have been employed
for smooth initiation of the desired reaction. The major unfavor-
able property arises from oxide layer of the Mg metal surface. The
activation described above would be dependent on either removal
of the oxide layer or cooperation of Lewis acidic self-catalysis by
p
defluorination does not occur or gives products in low yields. The
consistency of the product yields of Mg-promoted defluorination-
silylation reactions with the order of LUMO energy levels of the
substrates seems to be reasonable within the same type of
trifluoromethyl and pentafluoroethyl compounds. Meanwhile,
whether the defluorination occurs or not is not necessarily
consistent with the calculated LUMO energy levels on comparing
the reactions of different types of the substrates. For example,
defluorination of (pentafluoroethyl)benzene (LUMO = ꢀ0.48 eV)
does not occur, in contrast, that of trifluoromethyl alkyl ketone
(LUMO = ꢀ0.29 eV) dose under the examined experimental con-
ditions. Thus, the reactivity of the substrates in Mg(0)-promoted
defluorination reactions can not be explained only by the LUMO
energy levels of the substrates. Other factors such as interactions
between substrates and the surface of metal Mg and cooperative
* Corresponding author. Tel.: +81 86 251 8605; fax: +8 186 251 8021.
E-mail address: tkata@cc.okayama-u.ac.jp (T. Katagiri).
0022-1139/$ – see front matter ß 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jfluchem.2013.02.001

S. Utsumi et al. / Journal of Fluorine Chemistry 152 (2013) 84–89
85
Table 1
Reduction potential of metals. Mnþ þ neꢀ @ Mð0Þ Eo½Vꢁ.
M
Li
K
Na
Mg
2.4
Al
Mn
1.2
Zn
Ni
H
0
Cu
ꢀ0.3
ꢀEo
3.0
2.9
2.7
1.7
0.8
0.3
action of the Mg(II) ion with functional group of substrates may
affect feasibility of the substrate reaction.
silanes [12] are building blocks useful for nucleophilic polyfluor-
oalkylation like Ruppert–Prakash reagents [13]. Among them alkyl
difluoro(trimethylsilyl)acetates are useful for nucleophilic intro-
duction of difluoroacetate unit and are alternatives for the
corresponding Reformatsky and Ullmann reagents [14d,15]. They
can be prepared by electrochemical reductive [14] or SmI2-
promoted [16] silylation of alkyl trifluoroacetates 1. However,
more accessible Mg-promoted defluorination protocol is not
applicable for the purpose although it works for the corresponding
phenyl trifluoroacetate 1 (R = Ph) [17]. The LUMO of phenyl
trifluoroacetate is lower than those of alkyl trifluoroacetates,
which thus enables Mg-promoted reductive defluorination of
phenyl trifluoroacetate. The problem which must be solved is how
to activate the Mg-system so as to make defluorinative silylation of
alkyl trifluoroacetates 1 possible.
We tried Mg/CuCl/TMS-Cl bimetallic system for defluorinative
silylation since the conventional Mg/TMS-Cl system ever used
successfully for defluorination of trifluoromethyl ketones did not
work for the purpose. Metal Mg powder was pretreated with CuCl
in the presence of TMS-Cl in DMI at room temperature under Ar
atmosphere for 15 min and then alkyl trifluoroacetates 1 were
added to the suspension of Mg/CuCl/TMS-Cl in DMI. The reaction
mixture was stirred at room temperature for 4 h (Ref. [18] for
experimental details of removal of silyl acetal 3 and isolation of the
desired ethyl 2, 2-difluoro-(2-trimethylsilyl)acetates 2. Among Mg
metal, Mg powder was found to be more useful than Mg turning
and Mg ribbon.
As shown in Supplementary Figs. 2 and 3, Mg-promoted
defluorination of alkyl trifluoroacetates and benzotrifluoride does
not proceed under the conventional reaction conditions ever
examined in our laboratory. Question is how to activate metal
magnesium reduction system so as to realize defluorination of
these unreactive substrates.
The surface of commercially available magnesium metal is
mostly corroded so that freshly generated magnesium like Rieke
Mg and Mg-metal alloys have been sometimes used for promoting
Grignard reaction of unreactive substrates. Rieke magnesium
generated by MgCl2–K–KI in THF enabled Grignard reaction of 4-
fluorotoluene [4]. The Mg–Cu alloy for Grignard reagent genera-
tions from n-butyl chloride [5], fluorobenzene [6], bromoalkenes
[7], and aryl bromides [8] was reported. Benzeneselenenyl cupper
generated by the reaction of diphenyl diselenide with Mg–Cu
bimetals was used for selenation of iodobenzene [9]. A bimetallic
system of Mg–MgCl2–FeCl2 in THF was also found to be effective
for Grignard reagent generation of aryl chlorides [10].
Keeping in mind the scope and limitation of the Mg-promoted
reductive reactions and the activation of metal magnesium ever
studied, we have examined Mg–CuCl system in detail so as to
enable (1) defluorinative silylation of alkyl trifluoroacetates, (2)
chemoselective either defluorination or dechlorination of 4-
chloro- and 3-chloro(pentafluoroethyl)benzenes, (3) defluorina-
tive silylation of substituted benzotrifluorides, and finally tried (4)
the rational explanation for the mechanism of the defluorination
based on the microscopic observation of the Cu-deposited metal
magnesium surface. This short account has been written on the
bases of doctor thesis of Dr. Shinya Utsumi, Okayama University,
Japan (March, 2012) and the summarized results of C–F bond
activation over fifteen years in our laboratory.
The sole use of metal Mg was totally ineffective. In contrast, the
combined use of metal Mg with CuCl (0.5 eq) induced the desired
reaction dramatically. Any other combinations of FeCl3, NiCl2, and
Zn dust with metal Mg were useless. Interestingly, metal Cu
instead of CuCl was also ineffective, suggesting freshly generated
Cu(0) on metal Mg may act effectively for reductive defluorination.
The microscopic observation of Mg–Cu bimetallic system is
discussed in Section 2.4.
2. Results and discussion
The solvent DMI was found to be the best for both conversion
and yield of 2 so far as examined; solvent (Conv. %, yield of 2, %)
DMI (100%, 62%), DMF (30%, 8%), NMP (64%, 14%), DMI/THF (22%,
5%), THF (24%, 2%), MeCN (14%, 4%). Optimization of the reaction
conditions resulted in the best conditions for defluorinative
silylation of alkyl trifluoroacetates [Mg (4 eq)/CuCl (0.5 eq)/LiCl
(8 eq)/TMS-Cl (8 eq) in DMI, room temperature for 4 h] [18,19]. The
desired products 2, 2-difluoro-2-(trimethylsilyl)acetates 2 were
obtained under the optimized conditions in the range of 60–68%
isolated yields as shown in Supplementary Table 2 [20]. The
proposed reaction mechanism is shown in Scheme 1 [18]. Stepwise
reaction sequence; first electron transfer ! trapping anion radical
with TMS-Cl ! second electron transfer leads to the formation of
intermediate 5 which undergoes either defluorination or trapping
with TMS-Cl [21] resulting in formation of 2 or 3, respectively
(Scheme 2).
2.1. Defluorinative silylation of alkyl trifluoroacetates
a,a-Difluoroacyl moieties (RCF2COX) are important key units
for inhibitory activity in pharmaceuticals [11]. And polyfluoroalkyl
R2
O
R1
R3
O-TMS
R
N(Ar)TMS
CO2Bn
F
F
F
F
F
O-TMS
N(Ar)TMS
TMS
X
F
F
OBu
F3C
R
F
F
X = NAr
X = O
NAr
2.2. Selective defluorination and dechlorination of 3- and 4-
chloro(pentafluoro-ethyl)benzenes
TMS-CF2CO2R4
R : CH=
X : CH-
F3C
TMS
F
F
Defluorinative silylation of 3-chloro(pentafluoroethyl)benzene
6 proceeds smoothly and quantitatively at 0 8C, affording 3-chloro-
[(10-trimethylsilyl)tetrafluoroethyl]benzene 9 under the conven-
tional Mg/TMS-Cl/DMPU system, in which no dechlorination was
observed [22]. Meanwhile, the corresponding 4-chloro-substrate
F
F
Fig. 1. Partially fluorinated building blocks prepared by Mg-promoted
defluorination of CF3-building blocks; TMS = SiMe3.

86
S. Utsumi et al. / Journal of Fluorine Chemistry 152 (2013) 84–89
O
O
OTMS
+
Mg - CuCl - LiCl
TMS
TMS-F2C
OR
F3C
F3C
OR
TMS-Cl - DMI
RT - 4 h
OR
1
2
3
ET from Mg
O-TMS
OR
-
TMS-Cl
-F
F
2
O-TMS
F
4
F3C
OR
5
3
TMS-Cl
Scheme 1. Mg-promoted defluorination silylation.
11 provides a mixture of defluorination (14, 34%) and dechlorina-
tion (12, 44%) products. Keeping in mind the remarkable rate
acceleration by CuCl and LiCl in Mg/TMS-Cl defluorination system,
it is interesting and worthwhile to find reaction conditions which
make selective dechlorination of 6 possible.
DMPU induces the selective defluorination of 6, while no
reaction of 6 occurs in THF. However, combination of metal Mg/
TMS-Cl with CuCl in THF enables complete conversion of 6 at 32 8C
within 1.5 h to provide defluorination product 9 as a major product
in 88% yield along with small amount (5%) of dechlorination
product 7. The most interesting result is preferential formation of
dechlorination product 7 in less polar solvent such as a mixture of
ether-THF (9:1) (Scheme 3) [23]. The same phenomenon was
observed in the reaction of 11.
As a summary in this section, polar aprotic solvent such as
DMPU accelerates reaction rate and enhances defluorination of
benzylic fluorine of 6 and 11. Meanwhile, dechlorination occurs
preferentially in the presence of CuCl in less polar solvent like a
mixture of ether and THF [23]. Effect of CuCl for rate acceleration
and solvent dependent selectivity of either defluorination or
dechlorination are discussed in Section 2.4.
Scheme 3. Chemo – selective dehalogenation.
system may modify the reaction site circumstances in solid–liquid
interface to bring out reaction conditions favorable for the desired
reductive defluorination (Table 2).
As shown in Table 3, substrates with an electron-withdrawing
group were completely converted to difluoro(trimethylsilyl)-
methylarenes or fluoro(bistrimethylsilyl)methylarenes at 25 8C
although they never reacted in the absence of CuCl (Scheme 4).
Some substrates with an electron-donating group such as 3-NMe2
and 4-Me undergo the reaction slowly at elevated temperature
(50 8C), but substrates with a more strongly electron-donating
group such as 4-NMe2 and 4-OMe were never defluorinated even
at 50 8C. These data demonstrate the effect of CuCl is remarkable.
Chemical behavior of both 3-chloro and 4-chlorobenzotrifluorides
is somewhat complicated due to their two reaction sites toward
metal Mg. Mostly both substrates undergo dechlorinative silyla-
tion at first (70% for 3-chloro and 90% for 4-chloro) and then
defluorinative silylation at the second step [23].
2.3. Defluorination of substituted benzotrifluorides
2.4. Mechanism of defluorination on the Cu-deposited metal Mg
surface
Defluorination of benzotrifluoride by Mg/TMS-Cl/DMF or DMI
system is not successful even though electrochemical reductive
defluorination-alkylation [24], silylation [25], and hydrogenation
[26] proceed effectively. The electrochemical reduction potential is
tunable so as to adjust potential required for benzotrifluoride,
while metal Mg has its inherent reduction potential. However, the
facts described in Sections 2.1 and 2.2 suggest Mg–Cu bimetallic
Microscopic observation of the Cu-deposited metal Mg surface
was tried so as to elucidate a possible active site and the reaction
mechanisms of reductive defluorinative and dechlorinative silyla-
tions of 3- or 4-chloro(pentafluoroethyl)benzenes (6 and 11) and
also facile defluorinative silylation of alkyl trifluoroacetates. The
Mg ribbon surface washed with TMS-Cl is shown in Fig. 4A, which
TMS
TMS
CF2CF3
CF2CF3
CFCF3
CF2CF3
CFCF3
Mg / CuCl / TMS-Cl
Solvent-temp
+
+
+
TMS
Cl
Cl
H
TMS
9
6
7
8
10
TMS
TMS
CF2CF3
CF2CF3
CF2CF3
CFCF3
CFCF3
Mg / CuCl / TMS-Cl
Solvent-temp
+
+
+
TMS
Cl
H
TMS
Cl
12
14
11
15
13
Scheme 2. Defluorination and dechlorination of 6 and 11.

S. Utsumi et al. / Journal of Fluorine Chemistry 152 (2013) 84–89
87
Table 2
Effect of solvent and additive to defluorination and dechlorination of 6 and 11.a
Substrate
6
Solvent
Additive
Temp. ( 8C)
Time (h)
Conv. (%)
Products and yield (%)
DMPU
THF
–
0
5
100
0
7 (0)
(0)
8 (0)
(0)
9 (99)
(0)
10 (0)
(0)
–
55
5
THF
CuCl
CuCl
CuBr
–
32
1.5
10
5
100
100
25
(5)
(2)
(88)
(4)
(2)
Ether-THFb
Ether
reflux
reflux
0
(79)
(8)
(13)
(9)
(0)
(0)
(0)
11
DMPU
DMF
5
100
81
12 (0)
(24)
(0)
13 (0)
(4)
14 (32)
(32)
(0)
15 (41)
(1)
–
0
5
THF
–
55
5
0
(0)
(0)
THF
CuCl
CuCl
CuBr
32
5
100
85
(33)
(56)
(14)
(12)
(8)
(10)
(0)
(38)
(0)
Ether-THFb
Ether
30
5
Reflux
24
43
(0)
(0)
(0)
a
Substrate/Mg/CuCl/TMS-Cl =1:4:0.5:8 (eq).
Ether-THF (9:1).
b
exhibits a metallic luster. Fig. 4B shows Mg metal surface treated
with CuCl and TMS-Cl in THF, on which several micrometer-sized
Cu metal-deposits were observed. Then, the Cu-deposited metal
Mg was subjected to the reaction of 3-chloro(pentafluoroethyl)-
benzene. The surface observed after the reaction is shown in Fig.
4C. The dark brown-colored Cu metal deposits were found at the
bottom of the black-colored and roughly sculpted pits of Mg metal
surface. Fig. 4D is a cross-sectional view observed along the
straight line on Fig. 4C. Meanwhile, the Mg metal surface distal to
the Cu-deposits retained its smooth surface with metallic luster.
The black pits were observed to grow vertically rather than
horizontally, while the consumption of Mg metal in the absence of
metal Cu proceeded horizontally as reported by Whitesides [27].
These observations suggest that the consumption of the Mg metal
proceeds in closely proximate to the Cu-deposits where both
freshly sculpted Mg metal and Mg2+ ion may be accumulated to
activate the substrate as shown in Fig. 2. Moreover, Cu metal is
more noble than Mg metal so that electron from the freshly
sculpted Mg metal would be injected to the substrate through
deposited metal Cu.
3-chloro(pentafluoroethyl)benzene 6 are elucidated as follows; in
polar aprotic solvents like DMPU and DMI, defluorinative silylation
occurs exclusively. Two electron transfer from metal Mg to
substrate through Cu(0) metal deposit generates the ionic
intermediate, which diffuses to the bulk solution, releases fluoride
ion, and finally traps TMS-Cl to produce 3-chloro-(10-trimethylsi-
lyltetrafluoroethyl)benzene (Fig. 3). In contrast, less polar solvent
like a mixture of ether-THF (9:1) induces dechlorinative Grignard
reagent formation preferentially in which injecting electrons from
metal Mg to substrate and pulling out chloride ion from the
substrate by the aid of Lewis acidic Mg2+ ion would occur
concertedly (Fig. 4) rather than stepwise via the dianion or anion
radicals due to instability of the ionic intermediate in less polar
solvent.
In Section 2.1, scope and limitation of Mg promoted defluor-
inative silylation of alkyl trifluoroacetates in Mg/TMS-Cl/DMF
system where trifluoromethyl ketones and aryl trifluoroacetates
are successfully defluorinated, while alkyl trifluoroacetates are
not. In contrast, addition of CuCl (0.5 eq) in Mg/TMS-Cl/DMI
system enables smooth defluorination of alkyl trifluoroacetates.
Likewise, carbonyl group of the trifluoroacetates would be
activated with Mg2+ ion, which may make electron injection from
Based on the microscopic observation of the Mg metal surface,
the mechanisms of defluorinative and dechlorinative silylations of
metal Mg through Cu deposits to the substrates easy. The
a-
carbanion intermediates 5 undergo either defluorination leading to
ketene silyl acetals 4 (Fig. 5) or trapping TMS-Cl resulting in the
formation of silyl acetals 3 (Scheme 1) [21].
Table 3
Effect of substituent in defluorinativesilylation of substituted benzotrifluorides16.a
Substituent
Product
yieldb
and
Substrate
X
sm or sp
Time (h)
conv. (%)
17 (%)
18 (%)
16a
16b
16c
16d
16e
16f
3-F
0.34
0.12
5
15
15
15
5
100
100
100
100
100
100
42d
78d
0d
2
91
85
90
82
96
2
3-OMe
4-F
0.06
0
H
0
6
3-TMS
4-TMS
3-NMe2
4-Me
4-OMe
4-NMe2
ꢀ0.04
ꢀ0.07
ꢀ0.16
ꢀ0.17
ꢀ0.27
ꢀ0.83
16
94
0
c
15
15
15
15
15
–
16g
16h
16i
32
68
0
0
0
16j
0d
0
0
a
Substrate/Mg/CuCl/TMS-Cl =1:4:0.5:8 (eq).
Isolated yield in 5 mmol scale reaction.
Could not be isolated.
b
c
d
Reaction temperature is 50 8C.
CF3
CF2TMS
CF(TMS)2
Mg / CuCl / TMS-Cl
X
+
X
X
DMI, 25 o
C
16
17
18
Fig. 2. An image of freshly hollowed Mg pit which becomes wider and deeper in the
Scheme 4. Defluorinative silylation of substituted benzotrifluorides 16a.
course of reaction time.

88
S. Utsumi et al. / Journal of Fluorine Chemistry 152 (2013) 84–89
in polar solution (DMPU)
benzotrifluorides where trifluoromethyl group can be transformed
to difluoro(trimethylsilyl)methyl moiety (CF3 ! Me3SiCF2) which
is a building unit widely usable for introducing functionalized CF2
unit. The Mg–CuCl bimetallic system is applicable even in less
polar solvent such as ether-THF (9:1) so that Grignard reagent
formation by chemoselective dechlorination of 3-chloro(penta-
fluoroethyl)benzene predominates where electron transfer and
dechlorination would occur concertedly. The present protocol (Mg/
CuCl/TMS-Cl/DMI) would be widely usable for defluorination of
trifluoromethyl group and Grignard reagent formation.
TMS
CF2CF3
2-
CFCF3
TMS-Cl
Cl
Cl
-
F
diffusion
to
CF2CF3
bulk solution
Appendix A. Supplementary data
-
e
+2
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.jfluchem.2013.
02.001.
Cl
Cu
Mg2+
-
Mg
Mg
Mg Mg Mg
Cu
e
+
Cu
References
[1] (a) H. Amii, K. Uneyama, Chem. Rev. 109 (2009) 2119–2183;
(b) K. Uneyama, Organofluorine Chemistry, Blackwell Publishing, Oxford, 2006.
[2] (a) G.S. Silverman, P.E. Rakita, Handbook of Grignard Reagents, Dekker, New
York, 1996;
+
Mg2
Mg
Fig. 3. ET followed by defluorination in bulk solution.
(b) B.J. Wakefield, Organomagnesium Methods in Organic Synthesis, Academic
Press, London, 1995.
[3] (a) H. Amii, T. Kobayashi, Y. Hatamoto, K. Uneyama, Chem. Commun. (1999)
1323–1324;
in less polar solvent
Ether -THF
(b) K. Uneyama, G. Mizutani, Chem. Commun. (1999) 613–614;
(c) H. Amii, T. Kobayashi, H. Terasawa, K. Uneyama, Org. Lett. 3 (2001) 3103–
3105;
(d) T. Kobayashi, T. Nakagawa, H. Amii, K. Uneyama, Org. Lett. 5 (2003) 4297–
4300;
CF2CF3
CF2CF3
CF2CF3
(e) H. Amii, Y. Hatamoto, M. Seo, K. Uneyama, J. Org. Chem. 66 (2001) 7216–7218;
(f) G. Takikawa, T. Katagiri, K. Uneyama, J. Org. Chem. 70 (2005) 8811–8816;
(g) G. Takikawa, K. Toma, K. Uneyama, Tetrahedron Lett. 47 (2006) 6509–6511;
(h) Y. Nakamura, Y. Ozeki, K. Uneyama, J. Fluorine Chem. 129 (2008) 274–279;
(i) For a review, see K. Uneyama, H. Amii, J. Fluorine Chem. 114 (2002) 127–131.
[4] R.D. Rieke, S.E. Bales, J. Chem. Soc. Chem. Commun. (1973) 879–880.
[5] M.S. Kharasch, O. Reinmuth, Grignard Reactions of Nonmetallic Substances,
Prentice-Hall, New York, 1954, pp 12.
[6] H. Gilman, L.L. Heck, J. Am. Chem. Soc. 53 (1931) 377–378.
[7] C.D. Hurd, C.N. Webb, J. Am. Chem. Soc. 49 (1927) 546–559.
[8] H. Gilman, E.A. Zoellner, J. Am. Chem. Soc. 53 (1931) 1581–1583.
[9] N. Taniguchi, T. Onami, J. Org. Chem. 69 (2004) 915–920.
[10] B. Bogdanovic, M. Schwickardi, Angew. Chem. Int. Ed. 39 (2000) 4610–4612.
[11] (a) M.H. Gelb, J.P. Svaren, R.H. Abeles, Biochemistry 24 (1985) 1813–1817;
(b) M.H. Gelb, J. Am. Chem. Soc. 108 (1986) 3146–3147;
(c) H.L. Sham, Renin inhibitors with fluorine-containing amino acids, in: V.P.
Kukahr, V.A. Soloshonok (Eds.), Fluorine-containing Amino Acids, John Wiley &
Sons, Chichester, 1995, p. 333;
-
e
+
-
Cl
+
e
Cu
.
MgCl
+
MgCl
-
Mg2
Mg
e
Mg Mg Mg Mg
Cu
Mg
+
Cu
Mg
+
Mg2
Fig. 4. Dechlorination in less polar solution (ether–THF = 9:1).
(d) D. Schirlin, S. Baltzer, V. VanDorsselaer, F. Weber, C. Weill, J.M. Altenburger, B.
Neises, G. Flynn, J.M. Rimy, C. Tarnus, Bioorg. Med. Chem. Lett. 3 (1993) 253–258;
(e) J.T. Welch, Tetrahedron 43 (1987) 3123–3197;
(f) C. Doucet, I. Vergely, M. Reboud-Ravaux, J. Guilhem, R. Kobaiter, R. Joyeau, M.
Wakselman, Tetrahedron: Asymmetr. 8 (1997) 739–751;
(g) R. Joyeau, H. Molines, R. Labia, M. Wakselman, J. Med. Chem. 31 (1988) 370–
374.
F
F3C
F3C
-
F
OR
Mg
-F
TMS-Cl
OR
OR
-
e
Cu
O
O
TMSO
ClMg
5
[12] K. Uneyama, J. Fluorine Chem. 129 (2008) 550–576.
[13] (a) G.K.S. Prakash, A.K. Yudin, Chem. Rev. 97 (1997) 757–786;
(b) R.P. Singh, J.M. Shreeve, Tetrahedron 56 (2000) 7613–7632.
[14] (a) K. Uneyama, K. Maeda, T. Kato, T. Katagiri, Tetrahedron Lett. 39 (1998) 3741–
3744;
4
-
+
Mg2
e
Mg Mg Mg
Cu
Mg
+
Cu
(b) A.A. Stepanov, T.V. Minyaeva, B.I. Martynov, Tetrahedron Lett. 40 (1999)
2203–2204;
+
Mg2
(c) K. Uneyama, G. Mizutani, K. Maeda, T. Kato, J. Org. Chem. 64 (1999) 6717–
6723;
Mg
(d) M. Bordeau, P. Clavel, A. Barba, M. Bertande, C. Biran, N. Roques, Tetrahedron
Lett. 44 (2003) 3741–3744;
Fig. 5. Defluorination of alkyl trifluoroacetates in Mg/CuCl/LiCl/TMS-Cl/DMI system.
(e) From ClCF2CO2Et P. Clavel, C. Biran, M. Bordeau, N. Roques, S. Trevin, Tetra-
hedron Lett. 41 (2000) 8763–8767.
[15] (a) For reaction with imines leading to difluorolactam M. Bordeau, F. Frebault, M.
Gobet, J.-P. Picard, Eur. J. Org. Chem. (2006) 4147–4154;
(b) For cross-coupling with aryl iodides K. Fujikawa, Y. Fujioka, A. Kobayashi, H.
Amii, Org. Lett. 13 (2011) 5560–5563.
[16] (a) J. Wettergren, T. Ankner, G. Hilmersson, Chem. Commun. (2010) 7596–7597;
(b) A. Otaka, J. Watanabe, A. Yukimasa, Y. Sasaki, H. Watanabe, T. Kinoshita, S.
Oishi, H. Tamamura, N. Fujii, J. Org. Chem. 69 (2004) 1634–1645.
[17] H. Amii, T. Kobayashi, K. Uneyama, Synthesis (2000) 2001–2003.
[18] S. Utsumi, T. Katagiri, K. Uneyama, Tetrahedron 68 (2012) 580–583.
3. Conclusion
The Mg–Cu bimetallic system created in Mg/CuCl/TMS-Cl/DMI
system greatly accelerates electron transfer from metal Mg to the
substrates which are not reactive in the absence of CuCl. Reaction
rate enhancement by Mg–CuCl bimetallic system enables defluor-
inative silylation of both alkyl trifluoroacetates and substituted

S. Utsumi et al. / Journal of Fluorine Chemistry 152 (2013) 84–89
89
[19] Reaction enhancement with LiCl in zincation of aryl iodides is known and the role
of the salt is proposed; The salt LiCl rapidly removes the formed organozinc
reagent from the metal surface by generating highly soluble RZnX LiCl complexes,
thus allowing a rapid reaction of remained molecule of R–X A. Krasovskiy, V.
Malakhov, A. Gavryushin, P. Knochel, Angew. Chem. Int. Ed. 45 (2006) 6040–6044.
[20] Transformation of ketene silyl acetals 4–2 is known [14b].
[21] Y. Ishino, H. Maekawa, H. Takeuchi, K. Sukata, I. Nishiguchi, Chem. Lett. (1995)
829–830.
[22] Y. Nakamura, K. Uneyama, J. Org. Chem. 72 (2007) 5894–5897.
[23] S. Utsumi, T. Katagiri, K. Uneyama, Tetrahedron 68 (2012) 1085–1091.
[24] (a) C. Saboureau, C. Troupel, S. Sibille, S. Perichon, Chem. Commun. (1989) 1138–
1139;
(b) Y. Yamauchi, T. Fukuhara, S. Hara, H. Senboku, Synthesis (2008) 438–442.
[25] P. Clavel, M.-P. Leger-Lambert, C. Biran, F. Serei-Spirau, M. Bordeau, N. Roques, H.
Marzouk, Synthesis (1999) 829–834.
[26] C.-P. Andrieux, C. Combellas, F. Kanoufi, J.-M. Saveant, A. Thiebault, J. Am. Chem.
Soc. 119 (1997) 9527–9540.
[27] (a) C.L. Hill, J.B. Vander Sande, G.M. Whitesides, J. Org. Chem. 45 (1980) 1020–1028;
(b) S.E. Koon, C.E. Oyler, J.H. Hill, W.J. Bowyer, J. Org. Chem. 58 (1993) 3225–3226;
(c) C.E. Teerlinck, W.J. Bowyer, J. Org. Chem. 61 (1996) 1059–1064.