Syntheses of a wide family of new aryl based perfluorosulfonimide lithium salts. Electrochemical performances of the related polymer electrolytes

Article abstract of DOI:10.1016/j.jfluchem.2011.06.041

This paper reports both on a general multistep synthesis of a wide family of aryl substituted perfluorosulfonimides and on a preliminary electrochemical investigation of two lithium salts hosted by a poly(oxyethylene) homopolymer. Both salts have a cationic transference number more than twice that of LiTFSI. Additionally, one of these salts exhibits markedly higher cationic conductivities than POE/LiTFSI electrolytes. These preliminary data are very encouraging as, thanks to the aryl moiety, a wide variety of salts can be considered in order to still improve the performances of polymer electrolytes.

Full text of DOI:10.1016/j.jfluchem.2011.06.041

Journal of Fluorine Chemistry 132 (2011) 1213–1218
Contents lists available at ScienceDirect
Journal of Fluorine Chemistry
journal homepage: www.elsevier.com/locate/fluor
Syntheses of a wide family of new aryl based perfluorosulfonimide lithium salts.
Electrochemical performances of the related polymer electrolytes
E. Paillard a, F. Toulgoat b, R. Arvai c, C. Iojoiu d, L. Cointeaux d, M. Medebielle c, F. Alloin d,
B. Langlois c, J.-Y. Sanchez d,
*
a WWU. Institut fu¨r Physikalische Chemie, Corrensstra
b
e 28/30, 48149 Mu¨nster, Germany
b
´
UMR UDS/CNRS 7200, Faculte de Pharmacie Strasbourg, 74 route du Rhin, BP.24, 67401 Illkirch, France
c
ˆ
´
UMR 5246 ICBMS, Batiment Raulin, Universite Claude Bernard Lyon 1, 43 Bd du 11 Novembre 1918, 69622 Villeurbanne Cedex, France
d LEPMI, UMR 5279 CNRS – G-INP, Universite´ Joseph Fourier, Universite´ de Savoie, BP.75, 38402 Saint-Martin-d’He`res cedex, France
A R T I C L E I N F O
A B S T R A C T
Article history:
This paper reports both on a general multistep synthesis of a wide family of aryl substituted
perfluorosulfonimides and on a preliminary electrochemical investigation of two lithium salts hosted by
a poly(oxyethylene) homopolymer. Both salts have a cationic transference number more than twice that
of LiTFSI. Additionally, one of these salts exhibits markedly higher cationic conductivities than POE/LiTFSI
electrolytes. These preliminary data are very encouraging as, thanks to the aryl moiety, a wide variety of
salts can be considered in order to still improve the performances of polymer electrolytes.
ß 2011 Elsevier B.V. All rights reserved.
Received 31 March 2011
Received in revised form 28 June 2011
Accepted 29 June 2011
Available online 6 July 2011
Keywords:
Perfluorosulfonimide lithium salts
Multistep syntheses
Cationic transference numbers
Polymer electrolytes
POE
PEO
1. Introduction
dangerous degradation by-products [3]. This thermal stability is
crucial as it will be very difficult and probably costly to avoid the
Moving from lithium-ion batteries dedicated to the 4C mass
market (computers, cellular phones, camcorders, cordless tools) to
lithium batteries addressing the electric vehicle market requires (i)
improving the safety level (toxicity, flammability etc.) and (ii) a
clever battery recycling allowing most of the components to be
reused. In current lithium-ion batteries the salt providing the best
performances is the lithium hexafluorophosphate, LiPF6. The latter
belongs to the family of LiXFn and involves a coordinate bond or
dipolar bond. Following the formation of this bond, the whole of
the X–F bonds is indiscernible but the coordinate bond induces a
chemical weakness of the anion. This family of anions that include
BF4ꢀ, AsF6ꢀ, PF6ꢀ, SbF6ꢀ, are therefore subjected to disproportion-
ate into the Lewis acid XFnꢀ1 [1]. Among the related lithium salts,
the more appropriate one is LiPF6. Indeed LiBF4 leads to lower
electrolyte conductivities, LiAsF6 is suspected to be carcinogenic
and LiSbF6 has a limited stability in reduction. Additionally, it has
been recently reported that LiPF6, whose thermal stability (ꢁ50 8C
and around 80 8C in solution) is very limited [2], can lead to
heating of a battery pack. Last, its recycling would not allow LiPF6
to be easily reused. Among, the alternatives to the previous salts,
LiBOB, lithium bis(dioxalato) borate [4] is probably the most
famous. Its synthesis is non-costly and it provides a good SEI on
lithium graphite electrode but its dissolution in aprotic carbonate
solvents is limited, leading to fairly low salt concentrations in
current electrolytes. A review reports on various fluorinated forms
of LiBOB and on lithium tetrakis(polyfluoroalkoxy) aluminate [5].
In order to overcome the LiPF6 issues, fluoro(perfluoroalkyl)
lithium phosphates were also proposed [6].
Lithium triflate LiCF3SO3, lithium methide used name of
tris(trifluoromethylsulfonyl)methane LiC(CF3SO2)3 and lithium
bis(trifluoromethylsulfonyl) amide abbreviated as LiTFSI
LiN(CF3SO2)2 are thermally, chemically and electrochemically
stable lithium salts. As in the previous salts i.e. LiXFn, the related
anions are the conjugated bases of superacids. Unfortunately,
the first one leads to poor conductivities in liquid and polymer
electrolytes while the second one, whose molar mass exceeds
400 g, provides high conductivities in polymer electrolyte [7]. As
for LiTFSI, it exhibits high conductivities in poly(oxyethylene)
based electrolytes [8]. TFSIꢀ, a fairly hard anion [9], benefits
from a large delocalization of its negative charge as evidenced
* Corresponding author. Tel.: +33 4 76 82 6560; fax: +33 4 76 82 6670.
E-mail address: Jean-Yves.Sanchez@lepmi.grenoble-inp.fr (J.-Y. Sanchez).
0022-1139/$ – see front matter ß 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.jfluchem.2011.06.041

1214
E. Paillard et al. / Journal of Fluorine Chemistry 132 (2011) 1213–1218
by ab initio calculations while the conformational analysis
showed that the total energy calculated for the rotation around
the nitrogen atom is very low [9,10]. The large delocalization of
the negative charge is an asset for the ion-pair dissociation in
aprotic medium that, in addition to the high flexibility of the
anion, leads to high conductivities. Another asset lies in its high
oxidative stability that is in agreement with the fairly high
hardness of TFSI anion. D. DesMarteau who invented the TFSI
anion did a fruitful review of perfluorosulfonimide anions [11].
The cationic transference numbers, T+, are however lower than
most of the lithium salts [12–14]. It can be emphasized that,
despite its anion bulkiness, lithium methide leads also to very
low cationic transference numbers in poly(oxyethylene), POE,
based host polymers [14].
Whatever its advantages TFSI anion cannot be modified in such
a way to incorporate extra functionalities as, for instance, polar
[15] or plasticizing moieties. In order to take advantage of the
outstanding properties of perfluorosulfonimides while modulating
their physicochemical properties a multistep synthesis strategy
has been conceived. In this contribution we will describe this
strategy and provide the preliminary data obtained with two
perfluorosulfonimide salts.
O O
S
F2
R = ArS(O)n-(CF2)m-1
O O
S
R
n = 0, 1, 2
m = 1, 2
C
N
CF3
= Li+, Et3NH+
M
M
1
Scheme 1. General structure of described sulfonimide salts.
nucleophilic substitution around the benzylic carbon, is due
to the huge leaving ability of the perfluoroalkylsulfonimide
moiety.
This method has been also extended to the synthesis of
[(CF3CF2O–(CF2)2SO2)(CF3SO2)N]ꢀ M+ (M = Li+, Et3NH+).
In a previous paper, we described complementary methods for
the deprotection of symmetrical N-benzyl or N-allyl (perfluor-
oalkyl)sulfonimides which allow preparing salts in which the
(perfluoroalkyl)sulfonimide anion is associated with a very large
panel of metallic or onium cations [18]. These methods have been
largely exemplified from N-benzyl or N-allyl triflimide but were
also applied to the preparation of symmetrical perfluoroalkylsul-
fonimides as TFSI [(CF3SO2)2N]ꢀ M+ (Scheme 4). Moreover, there is
no doubt that they can be used to debenzylate 3 under various
conditions.
2. Results and discussion
In order to synthesize dissymmetric perfluorosulfonimide
salts bearing an aryl moiety a multistep strategy starting from
thiophenol has been selected. Thereafter the performances
of polymer electrolytes based on two lithium salts were
investigated.
In order to obtain the broadest range cations and to use their
most available and convenient sources, three types of nucleophiles
have been used for the deprotection of N-benzyl (or allyl)
(perfluoroalkyl)sulfonimides: (1) as above, ethanol associated
with various bases such as metallic and ammonium hydroxides
(LiOH, KOH, Ca(OH)2, Zn(OH)2, Me4N+ OHꢀ), metallic oxides (Sc2O3,
Ag2O), metallic carbonates and hydrogenocarbonates (NaHCO3,
MgCO3), organometallics (ZnEt2, LiN(TMS)2) or amines (Et3N,
iPr2EtN); (2) onium halides (nBu4N+ Brꢀ, Me3S+ Iꢀ, MePh3P+ Brꢀ);
(3) amines, imidazoles or phosphines (Et3N, iPr2NH, N–Me
2.1. Synthesis of new sulfonamide lithium salts
Recently, we described
a new and general synthesis of
(trifluoromethyl)-(perfluoroalkyl)sulfonimide lithium and triethy-
+
lammonium salts
substituted, at the
1
in which the perfluoroalkyl chain is
imidazole, Ph3P) which led to benzylonium cations Bn–R3
(R = N, P).
v
position, by an aryl sulfanyl, an arylsulfinyl
or an arylsulfonyl moiety (Scheme 1) [16a,b].
These salts were prepared from original
fluoroalkylsulfonyl fluorides (or their sulfinyl or sulfonyl con-
geners) 2, arising from -(arylsulfanyl)-perfluoroalkyl bromides,
It must be noticed that the two last methods are the most
appropriated to reach anhydrous (perfluoroalkyl)sulfonimide salts.
v
-(arylsulfanyl)per-
v
2.2. Polymer electrolytes based on aryl lithium arylsulfonimide salts
as we previously described (Scheme 2) [17a,b].
The key intermediate from 2 to 1 (which does not need to be
isolated) is the corresponding N-benzyl-sulfonimide 3 which,
surprisingly, was easily debenzylated, under very mild condi-
tions and without any hydrogen or metallic catalyst, by a simple
treatment, at room temperature, with a nucleophile such as
ethanol, followed by neutralization of the resulting (trifluor-
omethyl)(perfluoroalkyl)sulfonimide oxonium salt (Scheme 3)
[16]. Such an easy debenzylation, occurring through a simple
Two salts, Salts 1 and 2, whose structures are provided in
Scheme 5, were investigated.
Due to its high crystallinity and to its poor mechanical
performances above the melting point, poly(oxyethylene) host
polymer POE is not the best polymeric matrix, but it remains the
reference host polymer for polymer electrolytes free of molecular
solvent. Therefore the preliminary assessments of these new salts
were performed on the POE complexes.
Br(CF2)mBr
m = 1 (3 eq)
CsF (1.1 eq)
SO2 (3 eq)
Mg (2 eq)
m = 2 (1.25 eq)
TMSCl (4 eq)
PhSH
PhS-(CF2)mBr
isolated
m = 1; 62 %
m = 2; 90 %
PhS-(CF2)mSiMe3
isolated
NaH (1.5 eq)
DMF
THF
CH3CN
-40°C to r.t.
-78°C to r.t.
-40°C to r.t.
m = 1; 92 %
m = 2; 90 %
F-TEDA (1 eq)
PhS-(CF2)mSO2
not isolated
Cs
PhS-(CF2)mSO2F
isolated m = 1; 57 %
m = 2; 68 %
CH3CN
-40°C to r.t.
2
Scheme 2. Synthesis of
v-(arylsulfanyl)perfluoroalkylsulfonyl fluorides.

E. Paillard et al. / Journal of Fluorine Chemistry 132 (2011) 1213–1218
1215
1) Bn-NH2
O O
O O
S
F2
O O
S
O O
S
F2
DCE / 50 °C
Tf2O / DIEA
R
R
S
R
C
N
CF3
C
NH
Ph
C
F
2) HCl
DCM
-10°C to r.t.
F2
Ph
2
3
not isolated
O O
S
F2
O O
S
F2
O O
S
O O
S
M+ OH-
r.t.
R
R
EtOH
r.t.
C
N
CF3
Et
C
N
CF3
M
O
H
Ph
1
Scheme 3. (Trifluoromethyl)-(perfluoroalkyl)sulfonimide salts.
O
O
O O
S
O
O
R-NH2
S
S
Rf
N
R
Rf
Rf
Rf = CF3, X = TfO
Rf = ArS(O)n(CF2)m, X = F
X
R = Bn, allyl
O
O
O O
S
M+ = Li+, Na+, K+, 1/2 Ca2+, 1/2 Mg2+
1/2 Zn2+, Ag+, 1/3 Sc3+
1) EtOH
2) M+ B-
S
Rf
N
Rf
Me4N+, Et3NH+, iPr2EtNH+
-
2-
B- = OH-, HCO3 , CO3
M
Alk-, R2N-
MB = Et3N, iPr2EtN
RnA+ X-
O
O
O
O
O O
S
O O
S
S
S
A = N, S, P
A = N, P
Rf
N
Rf
Rf
N
R
Rf
A = N, S, P
X- = Br-, I-
RnA
R = Bn, allyl
O
O
O O
R1nA
S
S
Rf
N
Rf
R-AR1
A = N, P
n
Scheme 4. Access to various (perfluoroalkyl)sulfonimides salts.
2.2.1. Thermal analyses of POE electrolytes
2.2.1.2. TGA analyses. As the thermal stability of POE in air is
questionable above 130 8C, the thermogravimetric analyses
(TGA) were restricted to both salts. The TGA traces show
that the onset of weight loss starts from 160 and 225 8C
respectively for Salts 1 and 2. The presence of an additional CF2
induces a neat thermal stability increase in air. The latter
can be related to an increase in the electron withdrawing
effect induced, in Salt 2, by the additional CF2 that decreases
the electronic density on the sulfur atom of the thioether. Both
salts are however sufficiently stable (i) to meet the thermosta-
bility requirements of a lithium-polymer battery and (ii) to
allow an easy recovery to be performed following a battery
recycling.
2.2.1.1. DSC analyses of POE electrolytes. Both new salts induce a
decrease in the melting points of the polymer electrolytes, as
compared to POE. This decrease depends on the salt concentration
and seems to be similar for both electrolytes (Table 1).
It is well-known that the salt interaction with the POE matrix
induces an increase in the glass transition temperature Tg. This effect
is often called transient cross-linking. This Tg increase which is
clearly evidenced in amorphous electrolytes is however partially
screened in semi-crystalline electrolytes as the crystalline phase
constrains the amorphous one. The comparison of POE electrolyte’s
Tg based on both salts is provided in Table 1.
Table 1
CF3
S
S
CF2 SO2 N- SO2
Salt 1
Thermal characteristics of Salts 1 and 2.
Li+
O/Li (mol/mol)
Salt 1
Salt 2
Tg (8C)
Tf (8C)
D
H (J/g)
Tg (8C)
Tf (8C)
DH (J/g)
CF CF SO
2
N- SO2
CF3
Salt 2
2
2
12
20
30
ꢀ30
ꢀ30
ꢀ32
33
47
51
30
65
80
ꢀ31
ꢀ31
–
36
48
–
40
72
–
Li+
Scheme 5. Formula of Salts 1 and 2.

1216
E. Paillard et al. / Journal of Fluorine Chemistry 132 (2011) 1213–1218
1.E-03
1.E-04
1.E-05
1.E-06
1.E-07
2.8
2.9
3
3.1
3.2
3.3
3.4
3.5
1000/T (K-1)
Salt 1. O/Li = 12; O/Li = 20; O/Li = 30
Salt 2 . O/Li = 14; O/Li = 23; O/Li = 35
Fig. 1. Conductivity dependence on temperature.
2.2.2. Conductivity measurements
those determined, in the same conditions, for the lithium sulfonate
analogues of Salts 1 and 2 [19]. On the other hand they are
markedly higher than those of POE/LiTFSI that laboriously reach
0.1 [12,14]. The high cationic conductivity maxima determined for
POE/Salt 1 must be emphasized as it is roughly the double of the
value obtained with POE/LiTFSI.
The conductivities were measured in the temperature range
20–80 8C. The Arrhenius plots of POE electrolytes based on both
salts i.e. Salts 1 and 2 are gathered in Fig. 1. As usual in POE
electrolytes the conductivity dependence on temperature follows
an Arrhenius law and a VTF one respectively below and above the
melting point Tm. The conductivity decrease around the melting
temperature is more pronounced at the lowest salt concentrations
i.e. 30 and 35, in relation with a plasticizing effect of the salts. The
conductivities, at 80 8C, of the two types of electrolytes exceed
0.2 mS/cm making them candidates for the application. The
conductivity maxima, measured at 80 8C and at a concentration
O/Li = 20 for POE/Salt 1, reaches a very encouraging value i.e.
0.6 mS/cm. This one is however lower than that obtained using
LiTFSI/POE electrolytes. While in the analogues based on lithium
sulfonates i.e. C6H5–S–CF2–SO3Li and C6H5–S–CF2–CF2–SO3Li [19]
the POE electrolytes based on the former were found slightly less
conductive than the latter, in the case of sulfonimides, POE/Salt 1
that bears only one CF2, is roughly twice more conductive than
POE/Salt 2. This result was unexpected as the tetrafluoroethylene
moiety should (i) increase the electron withdrawing effect and
therefore the ion pair dissociation and (ii) increase the anion
flexibility. It seems therefore that the molecular weight of the
anion has a higher impact on the electrolyte conductivity than the
dissociation, which is already probably sufficient in Salt 1.
2.2.4. Cyclic voltammetry
The cyclic voltammetry was performed using
2 support
electrolytes. The first one, acetonitrile + NBu4PF6, is well-adapted
to check the oxidative stability of the salts but unsuitable to
explore the stability of the salts at low potentials. The second one
was a solution of N(Et)4ClO4 in propylene carbonate. The cathodic
and anodic peaks are presented in Table 3.
Salts 1 and 2 exhibit a wide electrochemical stability window
and, in particular, a good stability towards the reduction. They also
show a better stability towards the oxidation than their sulfonate
analogues, whose anodic peaks are respectively 5.1 and 5.3 V vs Li/
Li+ [19]. The sulfonimide moiety shifts therefore the stability
towards the oxidation by at least 200 mV. On the other hand, the
Salt 2 is more stable towards the oxidation than the Salt 1. It may
be inferred that the additional CF2 in Salt 2 decreases the electronic
density on the sulfur atom, thus increasing its stability towards the
oxidation.
Table 2
Cationic transference numbers and cationic conductivities of polymer electrolytes.
2.2.3. Cationic transference numbers
T+
s
(S cmꢀ1
)
s+
=
s
ꢂT+ (S cmꢀ1
)
Although obtaining high conductivities is essential to minimize
the polymer electrolyte contribution to the internal resistance of
the battery, the knowledge of its cationic transference number, T+,
is required to assess their performances in batteries. Indeed, on the
one hand the cationic transference number has an impact on the
formation of concentration gradients in the electrolyte during the
battery operation and, on the other hand, it allows the cationic
conductivity s+ to be calculated, which is the relevant parameter
for an electrolyte. Cationic transference numbers, T+, and cationic
conductivities, s+, are presented in Table 2.
Salt 1
Salt 2
O/Li = 30
O/Li = 20
O/Li = 12
0.21
0.26
–
3.6 ꢃ 10ꢀ4
5.8 ꢃ 10ꢀ4
–
7.6 ꢃ 10ꢀ5
1.5 ꢃ 10ꢀ4
–
O/Li = 34.8
O/Li = 23.2
O/Li = 13.4
0.25
0.20
0.19
2.1 ꢃ 10ꢀ4
2.9 ꢃ 10ꢀ4
2.2 ꢃ 10ꢀ4
5.2 ꢃ 10ꢀ5
5.8 ꢃ 10ꢀ5
4.2 ꢃ 10ꢀ5
Table 3
The cationic transference numbers of the polymer electrolytes
(measured at 70 8C) roughly range between 0.2 and 0.26. They
depend both on the type of POE/salt electrolytes and on the salt
concentration. The measured T+ are close to those previously
measured, in poly(oxyethylene) host polymers, with usual lithium
salts as lithium perchlorate or lithium iodide [14] but lower than
Electrochemical stability windows of Salts 1 and 2.
Lithium sulfonimide
salts
Cathodic peak
(V vs Li/Li+)
Anodic peak (V vs Li/Li+)
Salt 1
Salt 2
<0.3
<0.3
5.4
5.6

E. Paillard et al. / Journal of Fluorine Chemistry 132 (2011) 1213–1218
1217
3. Conclusion
0 8C, the mixture was stirred for 30 min at constant temperature
then 1 h at room temperature (followed by TLC and 19F NMR). Then
the products were evaporated, and the residue was dissolved in hot
pentane and the supernatant was isolated. The solvent was
evaporated and the benzyl sulfonimide was obtained.
These new multistep organic syntheses provide routes to a
variety new fluorinated organic anions. The starting thiophenol
synthon allows, through a clever selection of suitable substituted
thiophenols, a variety of salts based on multifunctional anions to
be designed as, for instance, ionic liquids, which could be useful in
several electrochemical energy storage and conversion applica-
tions. At this stage, due to the organic constraints, only a few
amounts of materials were prepared, restricting therefore the
scope of our investigations. Nevertheless both lithium salts
investigated demonstrated high thermal stability in air and an
outstanding electrochemical stability window. Additionally, one of
the salts i.e. Salt 1 exhibited a very high cationic conductivity.
N-Benzyl-difluoro-(phenylsulfanyl)-N-trifluoromethanesulfonyl-
methanesulfonimide. Precursor of Salt 1. Yield = 89%. White solid,
mp = 80–84 8C. C15H12F5NO4S3. M = 461.45 g molꢀ1 19F NMR
. d
ꢀ72.59(s, 2F), ꢀ72.96 (s, 3F). 1H NMR
d5.03 (s, 2H, H7) 7.34–7.56 (m,
8H, HAr), 7.69 (d, 2H, 3JH
¼ 7:3 Hz, H3). 13C NMR
d 56.3 (s, C7),
ꢀH
2
3
119.0 (q, 1JC–F = 324.9 Hz, CF3), 122.4 (t, 3JC–F = 3.0 Hz, CAr4 ), 128.8 (s,
C
Ar), 128.9 (t, 1JC–F = 327.9 Hz, CF2), 129.5 (s, CAr), 129.7 (s, CAr), 130.0
(s, CAr), 131.7 (s, CAr), 132.7 (s, CAr8 ), 137.4 (t, 4JC–F = 1.1 Hz, CAr3 ).
N-Benzyl-2-(phenylsulfanyl)-1,1,2,2-tetrafluoro-N-(trifluoro-
methylsulfonyl)-ethanesulfonimide. Precursor of Salt 2. Yield = 92%.
4. Experimental
White solid; mp = 58 8C. C16H12F7NO4S3. M = 511.45 g molꢀ1
NMR
CF2SO2N). 1H NMR
d5.07 (s, 2H, H8), 7.37–7.55 (m, 8H, HAr), 7.65 (d,
.
19F
d
ꢀ72.82 (s, 3F, CF3), ꢀ85.89 (s br, 2F, SCF2), ꢀ102.85 (s br, 2F,
4.1. Commercial products
1
2H, 3JH
¼ 7:4 Hz, H3). 13C NMR
d 56.6 (s, C8), 116.7 (tt, JF–
ꢀH
2
3
High molecular weight POE, Mw = 5.106 g molꢀ1 was purchased
from Janssen. Acetonitrile from Acros Organics was further dried
C = 305.2 Hz, JF–C = 39.0 Hz, CF2), 119.1 (q, JF–C = 324.9 Hz, C7),
2
1
1
2
3
122.5 (tt, JF–C = 292.4 Hz, JF–C = 31.3 Hz, CF2), 122.9 (t, JF–
˚
over an activated molecular sieves (3 A) before use (<15 ppm H2O)
C = 3.3 Hz, CAr ), 128.9 (s, CAr), 129.6 (s, CAr), 129.7 (s, CAr), 130.1
4
and stored in a glove box.
(s, CAr), 131.4 (s, CAr), 132.3 (s, CAr ), 137.5 (s, CAr3 ).
9
(Trifluoromethyl)-(perfluoroalkyl)sulfonimide lithium. A solution
of ethanol (C = 0.2 M) and benzyl sulfonamide (1 equiv.) was
stirred at room temperature for 8 h then LiOHꢂH2O (1 equiv.) was
added. The mixture was stirred overnight and dried evaporated.
The residue was dissolved in diethyl ether and filtrated. The filtrate
was evaporated and the residue was washed with pentane to
obtain the lithium salt. This compound was hygroscopic and the
melting point cannot be determinate.
4.2. Salt syntheses and characterizations
4.2.1. General procedures
All non-aqueous reactions were performed using oven-dried
glassware under an atmosphere of argon. Standard inert atmo-
sphere techniques were used in handling all air and moisture
sensitive reagents. THF was freshly distilled from sodium
benzophenone ketyl. Unless otherwise stated, reagents were
purchased from chemical companies and used without prior
purification. For chromatographic purification, reagent grade
solvents were used as received. Product purification by flash
chromatography was performed using Merck Geduran SI60 Silica
Gel 60 M (230–400 mesh). Mass spectrometry (LRMS) was
recorded on a ThermoFinnigan Mat 95xL apparatus (mode of
ionization: electrospray, chemical ionization or electron impact).
Melting points were determined with a Bu¨chi apparatus and are
given uncorrected. Reactions were monitored by thin layer
chromatography (TLC) using aluminiumbacked silica gel plates
(Merck, Kieselgel 60 F254). TLC spots were viewed under
ultraviolet light. 1H NMR (300 MHz), 19F NMR (282 MHz) and
13C NMR (75 MHz) spectra were run on a Bruker Advance300
spectrometer, and obtained using CDCl3 or acetone-D6. Chemical
Difluoro-phenylsulfanyl-N-(trifluoromethanesulfonyl)-methane-
sulfonimide lithium salt. Salt 1. Yield = 85%. White solid. C8H5F5Li-
NO4S3. M = 377.26 g molꢀ1
.
19F NMR (Acetone-d6)
d
ꢀ79.00 (s, 2F),
ꢀ80.08 (s, 3F). 1H NMR (Acetone-d6)
d 4.55 (s br), 7.41–7.53 (m,
3H, HAr), 7.65 (m, 2H, H3). 13C NMR (Acetone-d6)
d 121.0 (q, JC–
1
1
F = 321.8 Hz, CF3), 126.0 (m, CAr4 ), 128.7 (t, JC–F = 319.4 Hz, CF2),
130.0 (s, CAr), 131.1 (s, CAr), 137.4 (s, CAr). SM (ESI-MeOH) m/
z = 370.0 (Mꢀ, 100%), 371.0 [(M+1)ꢀ, 12%], 372.0 [(M+2)ꢀ, 12%],
746.8 [(2M+Li)ꢀ, 55%], 746.8 [(2M+1 + Li)ꢀ, 10%], 746.8
[(2M+2 + Li)ꢀ, 10%], 762.7 [(2M+Na)ꢀ, 12%]. SMHR calcd for
C8H5F5NO4S3: 369.9301; found: 369.9306.
2-(Phenylsulfanyl)-1,1,2,2-tetrafluoro-N-(trifluoromethylsulfo-
nyl)-ethanesulfonimide lithium salt. Salt 2. Yield = 90%. White solid.
C9H5F7LiNO4S3. M = 427.27 g molꢀ1
.
19F NMR (Acetone-d6)
d
3
ꢀ80.29 (s, 3F, CF3), ꢀ84.94 (t, 2F, SCF2, JF–F = 5.7 Hz), ꢀ112.04
shifts (
d
) are given in ppm vs. TMS (1H, 13C) or CFCl3 (19F), used as
(t, 2F, CF2SO2N, 3JF–F = 5.7 Hz). 1H NMR (Acetone-d6)
7.46–7.58 (m, 3H, HAr), 7.69 (m, 2H, H3). 13C NMR (Acetone-d6)
d 3.43 (s br),
internal references. Coupling constants J are reported in hertz (Hz).
The substitution pattern of the different carbons were determined
by a ‘‘DEPT135’’ sequence. Multiplicities are reported as follows:
s = singlet, d = doublet, t = triplet, q = quadruplet, m = multiplet,
tt = triplet of triplet, br = broad.
N-benzyl-sulfonimide 3. First step: To a solution of anhydrous
1,2-dichloroethane (C = 0.2 M) and sulfonyl fluoride (1 equiv.),
under inert atmosphere, was added freshly distilled benzylamine
(5 equiv.). The mixture was stirred at 50 8C during 20 h and was
followed by TLC and 19F NMR until sulfonyl fluoride completely
disappeared. At room temperature, an aqueous HCl solution (10%)
was added, and the mixture was extracted with dichloromethane.
The organic phases were dried over MgSO4, filtered and
evaporated. The residue was purified by column chromatography
on silica gel with gradient eluent (Pentane/AcOEt 1/0 to 4/1) to
obtain a benzyl sulfonamide.
d
1
2
1
115.1 (tt, JF–C = 293.1 Hz, JF–C = 34.4 Hz, CF2), 120.6 (q, JF–
1
2
C = 321.5 Hz, CF3), 124.0 (tt, JF–C = 290.4 Hz, JF–C = 32.5 Hz, CF3),
124.4 (m, CAr ), 130.2 (s, CAr), 131.6 (s, CAr), 137.9 (s, CAr3 ). SM (ESI-
4
MeOH) m/z = 420.0 (Mꢀ, 100%), 421.0 [(M+1)ꢀ, 10%], 422.0
[(M+2)ꢀ, 10%], 846.7 [(2M+Li)ꢀ, 62%], 862.7 [(2M+Na)ꢀ, 20%].
SMHR calcd for C9H5F7NO4S3: 419.9269; found: 419.9266.
4.3. Film processing
The electrolytes were prepared in a glove box by dissolving both
POE and the salt in acetonitrile. The resulting solutions were then
stirred overnight and casted on a Teflon plate. Solvent evaporation
was carried out in a glove box for 18 hrs. The films were then dried
under vacuum at 80 8C for 48 hrs and stored in a glove box under
argon.
Second step: To a solution of dichloromethane (C = 0.2 M) and
benzyl sulfonamide (1 equiv.), under inert atmosphere, was added
DIEA (1 equiv.). Then triflic anhydride (1.5 equiv.) was added at
The lithium salt concentration in the film is indicated by the
number n = O/Li, which corresponds to the oxyethylene/lithium
molar ratio.

1218
E. Paillard et al. / Journal of Fluorine Chemistry 132 (2011) 1213–1218
4.4. Thermal analysis
saturated calomel electrode as reference. The salts were dissolved
in ACN at a concentration ranging between 2 and 4 mmol. The use
of ACN allows exploring the stability towards the oxidation up to
5.8 V vs Li/Li+. To explore the stability of the salts towards the
reduction a 0.1 M solution of N(Et)4ClO4 in propylene carbonate
was used as a support electrolyte.
Glass transition temperatures, Tg, and melting temperatures,
Tm, were measured in nitrogen flow using a TA Instruments DSC
2920 modulated DSC. In a typical procedure, samples were heated
to 100 8C and cooled rapidly to ꢀ100 8C and, then the samples were
heated at 5 8C/min up to 100 8C. The oscillation period was 60 s and
amplitude was ꢄ1 8C. Tg and Tm were taken as the inflection point of
the specific heat increment at the glass-rubber transition and at the
onset of the melting peak, respectively. Around 10 mg of sample were
placed in a DSC aluminium crucible in a glove box. Thermogravi-
metric measurements were carried out with a Netzsch STA409
thermal analyzer. A few milligrams of the sample were heated from
room temperature up to 400 8C at 5 8C/min under air flow.
References
[1] M. Armand, J.-Y. Sanchez, M. Gauthier, Y. Choquette, in: J. Lipkowski, P.N. Ross
(Eds.), The Electrochemistry of Novel Materials, vol. III, VCH Publishers, 1993, pp.
65–110.
[2] K. Xu, Chem. Rev. 104 (2004) 4303.
[3] A. Hammami, N. Raymond, M. Armand, Nature 424 (2003) 635.
[4] K. Xu, S. Zhang, T.R. Jow, W. Xu, C.A. Angell, Electrochem. Solid. Lett. 5 (2002)
A26.
[5] B.G. Nolan, S. Tsujioka, S.H. Strauss, in: T. Nakajima, H. Groult (Eds.), Fluorinated
Materials for Energy Conversion, Elsevier, 2005, pp. 195–221 (Chapter 9).
[6] P. Sartori, N. Ignat’ev, DE 19641138, Merck GMBH, 1998, Chem. Abstr.128:
294896.
[7] D. Benrabah, D. Baril, J.-Y. Sanchez, M. Armand, G. Gard, Faraday Trans. 89 (2)
(1993) 355.
[8] S. Sylla, J.-Y. Sanchez, M. Armand, Electrochim. Acta 37 (9) (1992) 1699–1701.
[9] D. Benrabah, R. Arnaud, J.-Y. Sanchez, Electrochim. Acta 40 (13–14) (1995)
2437–2443.
4.5. Conductivity measurements
Conductivities were determined by electrochemical impedance
spectroscopy using an HP 4192A Impedance Analyser in the
frequency range 5 Hz - 13 MHz. The samples were placed between
two stainless steel electrodes, under argon, in a Swagelok cell
equipped with Teflon joints and measurements were performed
from 100 8C to 20 8C. The temperature was equilibrated for 2 h
before each measurement.
[10] R. Arnaud, D. Benrabah, J.-Y. Sanchez, J. Phys. Chem.
10882–10891.
B 100 (26) (1996)
[11] O.E. Geiculescu, S.E. Creager, D.D. DesMarteau, in: T. Nakajima, H. Groult (Eds.),
Fluorinated Materials for Energy Conversion, Elsevier, 2005, pp. 223–251 (Chap-
ter 10).
[12] M. Kono, E. Hayashi, M. Watanabe, J. Electrochem. Soc. 5 (1998) 1521.
[13] Y. Aihara, K. Hayamizu, K. Sugimoto, T. Bando, T. Iguchi, J. Kuratomi, T. Ono, K.
Kuwana, J. Power Sources 97–98 (2001) 628.
[14] F. Alloin, J.-Y. Sanchez, J. Power Sources 81–82 (1999) 795–804.
[15] X. Ollivrin, F. Alloin, J.-F. Le Nest, D. Benrabah, J.-Y. Sanchez, Electrochim. Acta 48
(14–16) (2003) 1961–1969.
[16] (a) F. Toulgoat, B.R. Langlois, M. Me´debielle, J.Y. Sanchez, J. Org. Chem. 73 (2008)
5613–5616;
4.6. Transference numbers
The transference numbers were determined by the method
proposed by Vincent et al. [20], namely through a combination of
complex impedance and potentiostatic polarization measure-
ments. Symmetrical Li/electrolyte/Li cells were assembled in a
glove box in a Swagelok cell. The experiments were carried out at
70 8C. The polarization voltage was kept at 10 mV using a
computer-monitored Solartron 1470 battery test unit.
(b) E. Paillard, C. Iojoiu, F. Alloin, F. Toulgoat, M. Medebielle, B. Langlois, J.-Y.
Sanchez, Patent, FR2903692, 2008-01-18, EP2084166, 2009-09-10.
[17] (a) F. Toulgoat, B.R. Langlois, M. Me´debielle, J.Y. Sanchez, J. Org. Chem. 72 (2007)
9046–9052;
(b) J.-Y. Sanchez, B. Langlois, M. Medebielle, F. Toulgoat, Patent FR 2006-6466,
WO 2007-FR1226.
4.7. Cyclic voltammetry
´
[18] R. Arvai, F. Toulgoat, B.R. Langlois, J.Y. Sanchez, M. Medebielle, Tetrahedron 65
(2009) 5361–5368.
Cyclic voltammetry was performed, using as support electro-
lyte a 0.1 M solution of N(Bu)4PF6 in acetonitrile (ACN), a glassy
carbon as working electrode, a platinum auxiliary electrode and a
[19] E. Paillard, F. Toulgoat, C. Iojoiu, F. Alloin, J. Guindet, M. Medebielle, B. Langlois, J.-
Y. Sanchez, J. Fluorine Chem, doi:10.1016/j.jfluchem.2011.06.020, in press.
[20] J. Evans, C.A. Vincent, P.G. Bruce, Polymer 28 (1987) 2324–2328.