Low-melting, low-viscous, hydrophobic ionic liquids: Aliphatic quaternary ammonium salts with perfluoroalkyltrifluoroborates

Article abstract of DOI:10.1002/chem.200400817

A novel class of low-melting, hydrophobic ionic liquids based on relatively small aliphatic quaternary ammonium cations ([R¹R²R ³NR]⁺, wherein R¹, R², R³ = CH₃ or C₂H₅, R = n-C₃H ₇, n-C₄H₉, CH₂CH₂OCH ₃) and perfluoroalkyltrifluoroborate anions ([R_FBF ₃]^-, R_F = CF₃, C₂F ₅, H-C₃F₇, n-C₄F₉) have been prepared and characterized. The important physicochemical and electrochemical properties of these salts, including melting point, glass transition, viscosity, density, ionic conductivity, thermal and electrochemical stability, have been determined and comparatively studied with those based on the corresponding [BF₄]^- and [(CF₃SO ₂)₂N]^- salts. The influence of the structure variation in the quaternary ammonium cation and perfluoroalkyltrifluoroborate ([R_FBF₃]^-) anion on the above physicochemical properties is discussed. Most of these salts are liquids at 25°C and exhibit low viscosities (58-210 cP at 25°C) and moderate conductivities (1.1-3.8 mS cm^-1). The electrochemical windows of these salts are much larger than those of the corresponding 1,3-dialkyimidazolium salts. Additionally, a number of [R_FBF₃]^- salts exhibit plastic crystal behavior.

Full text of DOI:10.1002/chem.200400817

Low-Melting, Low-Viscous, Hydrophobic Ionic Liquids: Aliphatic
Quaternary Ammonium Salts with Perfluoroalkyltrifluoroborates
Zhi-Bin Zhou, Hajime Matsumoto,* and Kuniaki Tatsumi[a]
Abstract: A novel class of low-melting,
hydrophobic ionic liquids based on rel-
atively small aliphatic quaternary am-
monium cations ([R1R2R3NR]+, where-
in R1, R2, R3 = CH3 or C2H5, R = n-
C3H7, n-C4H9, CH2CH2OCH3) and per-
ductivity, thermal and electrochemical
stability, have been determined and
comparatively studied with those based
on the corresponding [BF4]ꢁ and
[(CF3SO2)2N]ꢁ salts. The influence of
the structure variation in the quaterna-
ry ammonium cation and perfluoroal-
kyltrifluoroborate ([RFBF3]ꢁ) anion on
the above physicochemical properties
is discussed. Most of these salts are liq-
uids at 258C and exhibit low viscosities
(58–210 cP at 258C) and moderate con-
ductivities (1.1–3.8 mScmꢁ1). The elec-
trochemical windows of these salts are
much larger than those of the corre-
sponding 1,3-dialkyimidazolium salts.
Additionally, a number of [RFBF3]ꢁ
salts exhibit plastic crystal behavior.
fluoroalkyltrifluoroborate
anions
([RFBF3]ꢁ, RF = CF3, C2F5, n-C3F7, n-
C4F9) have been prepared and charac-
terized. The important physicochemical
and electrochemical properties of these
salts, including melting point, glass
transition, viscosity, density, ionic con-
Keywords: electrochemistry · ionic
liquids
borates
·
perfluoroalkyltrifluoro-
plastic crystals ·
·
quaternary ammonium
Introduction
lytes to replace conventional materials.[1–8] However, these
ILs cannot be used as electrolytes for high-energy electro-
chemical devices, such as 4 V class Li batteries (i.e. Li metal
or carbon used as anode material) because of the electro-
chemical instability of the 1,3-dialkylimidazolium cations
(i.e. the cathodic limit is ꢀ1 V versus Li+/Li, which ex-
cludes the use of Li metal or carbon as anode).[4,9] Com-
pared with the 1,3-dialkylimidazolium cations, saturated
quaternary ammoniums are more resistant against oxidation
and reduction, thereby their ILs with electrochemically
stable anions, such as bis(trifluoromethanesulfonyl)imide
[(CF3SO2)2N]ꢁ (TFSIꢁ), have much larger electrochemical
windows than the corresponding 1,3-dialkylimidazolium
compounds.[9] Furthermore, recent studies have shown that
the ILs based on the quarternary ammonium (QA) and
robust anions, such as TFSIꢁ and [BF4]ꢁ, have sufficient
electrochemical stability to support them as possible safety
electrolytes (nonvolatile and nonflammable) for high-energy
devices, such as Li batteries[10–12] and double-layer capaci-
tors,[13] and for electropositive metal deposition, such as Li
and Eu.[9b,14]
In recent years, ionic liquids (ILs) have received a consider-
able upsurge of interest as new media for organic reac-
tions,[1] biocatalytic transformations,[2] and separation tech-
nologies,[3] and as potential electrolytes for various electro-
chemical devices, such as Li-ion batteries,[4,5] dye-sensitized
solar cells,[6] electrochemical actuators, electrochromic win-
dows, and numerical displays.[7] This is mainly attributable
to their superior properties over conventional volatile or-
ganic solvents and electrolytes, such as nonvolatility, non-
flammability, high thermal stability, and wide liquid
ranges.[1–3,5,8]
Currently, the basic understanding and applications of ILs
are largely governed by imidazolium systems because the
ILs based on the imidazolium cations, especially 1,3-dialkyli-
midazolium cations, have been empirically demonstrated to
afford low viscosities and high conductivities, which are the
key requirements for ILs as new reaction media or electro-
[a] Dr. Z.-B. Zhou, Dr. H. Matsumoto, Dr. K. Tatsumi
Research Institute for Ubiquitous Energy Devices
National Institute of Advanced Industrial Science and Technology
1-8-31 Midorigaoka, Ikeda, Osaka 563–8577 (Japan)
Fax : (+81)72-751-9622
Besides a large electrochemical window, high ionic con-
ductivity and fast ion mobility (including the doped ions
such as Li+ for Li batteries) are the additional requirements
for ILs as supporting electrolytes in high-energy devices.
This means that the viscosity of ILs must be as low as possi-
E-mail: h-matsumoto@aist.go.jp
752
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/chem.200400817
Chem. Eur. J. 2005, 11, 752 – 766

FULL PAPER
ble and the anion involved in the ILs must have weakly co-
ordinating ability to depress ion pairing. Therefore, the QA
cation involved should be small because large ones have
been found to result in high viscosities.[15–17] To date, al-
though a large number of QA salts with a wide variety of
anions have been reported,[12–33] very few of them appear to
satisfy the above requirements. This is mainly due to a lack
of anions, which not only have chemical and electrochemical
stability but which can also form low-melting and low-vis-
cous ILs with relatively small QA cations (total C atoms
ꢂ10). For instance, several known electrochemically stable
anions tend to form either high-melting (e.g. >1008C for
high symmetry [BF4]ꢁ,[20,22,23] [PF6]ꢁ,[22,24] [AsF6]ꢁ[22]) or very
highly viscous salts (e.g. >300 cP for [BF4]ꢁ,[13,25] and
>1300 cP at 258C for large and rigid bis(oxalato)borate
[BOB]ꢁ[26]) with small QA cations, whereas those that can
form low-melting and low-viscous ILs with small QA cations
ions ([R1R2R3NR]+, wherein R1, R2, R3 = CH3 or C2H5, R
= n-C3H7, n-C4H9, CH2CH2OCH3) and perfluoroalkyltri-
fluoroborate anions ([RFBF3]ꢁ, RF = CF3, C2F5, n-C3F7, n-
C4F9). Most of them are liquids at 258C, and exhibit low vis-
cosities, moderate conductivities, and good electrochemical
stability. For comparison, the corresponding QA salts with
two representative anions, [BF4]ꢁ and TFSIꢁ, were also syn-
thesized and characterized.
Results and Discussion
Synthesis and characterization: In the present study, 36 qua-
ternary ammonium (QA) salts, comprising the relative small
QA cations [R1R2R3NR]+ (R1, R2, R3 = CH3 or C2H5, R =
n-C3H7, n-C4H9, CH2CH2OCH3) with the fluoroanions
[RFBF3]ꢁ (RF = CF3, C2F5, n-C3F7, n-C4F9), [BF4]ꢁ, and
TFSIꢁ, were synthesized according to Scheme 1. The synthe-
(e.g.
trifluoro-N-(trifluoromethylsulfonyl)acetamide
[(CF3SO2)(CF3CO)N]ꢁ,[9c,27] di-
cyanamide [N(CN)2]ꢁ[28]) suffer
from electrochemical instability.
Until recently, the development
of stable, low-melting and low-
viscous QA salts has been
almost limited to the anion
TFSIꢁ[9–13,16,29,30] because this
fluoroanion not only possesses
chemical and electrochemical
stability[9–13,21] but it also com-
bines low symmetry, high flexi-
bility, and a weakly coordinat-
ing nature,[34,35] all of which are
beneficial to the reduction of
the cohesive forces of the salts,
thus lowering the melting point
and viscosity. To further explore
and extend the potential appli-
cations of the QA-based ILs in
electrochemical devices and
other emerging research fields,
such as enzyme-catalyzed reac-
ꢁ
Scheme 1.
tions[30] and classic C C formation (e.g. Heck and Suzuki
coupling reactions),[36] new robust anions that can form low-
melting and low-viscous ILs with small QA cations are cer-
tainly needed.
sized QA salts are presented in Table 1. For convenience, a
special notation for the QA cations is used (see Scheme 1):
“N” denotes the QA cations and each number in the sub-
script denotes the number of carbons in each alkyl substitu-
ent. The subscript, “1O2”, denotes the methoxyethyl
(CH3OCH2CH2) substituent.
ILs are usually prepared from halide salts as the source of
the cation component. In this case, it is very difficult to
avoid halide impurities in the resulting ILs. It is well known
that the presence of halide impurities in the resultant ILs
significantly alters the physicochemical properties of the
ILs.[47] To depress halide contamination, the new salts
[R1R2R3NR][RFBF3] in the present study were prepared via
a modified route by neutralization of aqueous solutions of
[R1R2R3NR][OH] with aqueous solutions of H[RFBF3]
(Scheme 1). The aqueous [R1R2R3NR][OH] salts were pro-
Very recently, we found that a new class of chemically
and electrochemically stable and weakly coordinating fluo-
roanions, namely perfluoroalkyltrifluoroborates ([RFBF3]ꢁ)
(RF = CF3, C2F5, n-C3F7, n-C4F9),[37–42] can form low-melting
and low-viscous ILs, even with large 1,3-dialkylimidazolium
cations (<77 cP at 258C)[43–45] and relatively small aliphatic
QA cations (58–86 cP at 258C).[46] More importantly, they
are more resistant toward hydrolysis than their [BF4]ꢁ ana-
logues on account of their hydrophobicity.[43–45] Encouraged
by these results, we herein report the systematic synthesis
and characterization of a series of new hydrophobic and
low-melting salts based on relatively small aliphatic QA cat-
Chem. Eur. J. 2005, 11, 752 – 766
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
753

H. Matsumoto et al.
Table 1. Thermal properties of various prepared salts.
[a]
[c]
[d]
[e]
[f]
Entry
Salts
Tg
Tc[b]
[8C]
Ts-s
Tm
DSm
Td
[8C]
[8C]
[8C]
[Jmolꢁ1 Kꢁ1
]
[8C]
1
2
3
4
5
6
7
8
N
1223[CF3BF3]
ꢁ40, ꢁ30, ꢁ21
95
ꢁ3
77
7.4
9.5
252
212
186
163
174
210
315
320
326
307
322
345
315
307
284
291
275
351
273
314
307
283
287
305
393
392
376
377
372
372
404
400
382
388
390
386
N1224[CF3BF3]
N111.1O2[CF3BF3]
N112.1O2[CF3BF3]
ꢁ40
51
33.3
28.8
29.9
60.7
29.3
32.7
38.4
57.9
ꢁ7
8
N
122.1O2[CF3BF3]
ꢁ22
N222.1O2[CF3BF3]
N1223[C2F5BF3]
N
N111.1O2[C2F5BF3]
N112.1O2[C2F5BF3]
N122.1O2[C2F5BF3]
N222.1O2[C2F5BF3]
N1223[n-C3F7BF3]
N1224[n-C3F7BF3]
N111.1O2[n-C3F7BF3]
10
ꢁ67, ꢁ40
ꢁ49
54
15
30
ꢁ33
1224[C2F5BF3]
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
ꢁ117
ꢁ113
ꢁ98
ꢁ76, ꢁ65
-63, ꢁ51
3
57
50
23
88.4
15.7
41.1
45.9
ꢁ53, 25
ꢁ58
ꢁ86
N
112.1O2[n-C3F7BF3]
ꢁ113
ꢁ112
N122.1O2[n-C3F7BF3]
N222.1O2[n-C3F7BF3]
N1223[n-C4F9BF3]
N1224[n-C4F9BF3]
N111.1O2[n-C4F9BF3]
N112.1O2[n-C4F9BF3]
N122.1O2[n-C4F9BF3]
N222.1O2[n-C4F9BF3]
N1223[BF4]
N1224[BF4]
N111.1O2[BF4]
N112.1O2[BF4]
N
N222.1O2[BF4]
N1223[TFSI]
N1224[TFSI]
N111.1O2[TFSI]
N112.1O2[TFSI]
N122.1O2[TFSI]
N222.1O2[TFSI]
ꢁ33
6
54
60
50
ꢁ28
30.7
17.9
29.6
54.4
60.2
12, 24
ꢁ76, 3
ꢁ110
ꢁ108
ꢁ56
ꢁ11
11
186
165
26.1
13.6
20.6
45.8
55.0
61.0
57.0
15.8
87.6
85.6
54[g]
ꢁ97[h]
ꢁ95
ꢁ26
ꢁ51
4[i]
122.1O2[BF4]
8[j]
56
14
9
ꢁ21, ꢁ2
ꢁ93
ꢁ44
37
ꢁ96[k]
ꢁ95
ꢁ82
ꢁ43
20
81.3
[a] Glass transition temperature determined by DSC on heating. [b] Crystallization temperature determined by DSC on heating. [c] Solid–solid transition
temperature determined by DSC on heating. [d] Melting point determined by DSC on heating. [e] Entropy of melting (DSm = DHm/Tm, where DHm is
melting enthalpy at Tm [K]). [f] Decomposition temperature determined by TGA. [g] 64.[19b] [h] ꢁ97;[19b] [i] 13.[19b] [j] 9.[13] [k] ꢁ93.[25]
duced by anion exchange of the corresponding
[R1R2R3NR]Br salts, which were synthesized by reaction of
trialkylamines with an equivalent of the appropriate alkyl
bromide in anhydrous acetone in an autoclave. The excep-
tion was N111.1O2Br, which was directly prepared at room
temperature. Aqueous H[RFBF3] was prepared by cation ex-
change of the corresponding K[RFBF3].[38] The salts based
on [BF4]ꢁ and TFSIꢁ were similarly prepared, except aque-
ous H[BF4] and aqueous H[TFSI] were used instead of
aqueous H[RFBF3]. All the salts prepared by this modified
method (Table 1) were found to contain less than 30 ppm of
Brꢁ and K+ ions, as estimated by fluorescence X-ray spec-
trometry. In addition, the [RFBF3]ꢁ and TFSIꢁ salts are im-
miscible with water to some extent, and this allows these
salts to be separated directly from the aqueous phase. The
final yields of the prepared salts ranged from moderate to
high (66–92%). In the [RFBF3]ꢁ series, an increase in the
length of the perfluoroalkyl (RF) chain in the [RFBF3]ꢁ gen-
erally resulted in an increase in the yield, as expected by the
increased hydrophobicity.
The newly prepared salts in Table 1 were characterized by
1H, 19F, and 11B NMR, FAB-MS, and elemental analysis. All
the characterization data are consistent with the expected
compositions and structures. The water content in the
[RFBF3]ꢁ and TFSIꢁ salts was less than 50 ppm after drying
at 708C and 0.02 Torr for 24 h owing to their hydrophobicity,
whereas the [BF4]ꢁ salts are miscible with water and still
contained an appreciable amount of water (ꢀ400–600 ppm),
even after rigorously drying at 808C and 0.02 Torr for 48 h.
In our experience, it is rather difficult to reduce the water
content any further in these hydrophilic liquids because
drying at a higher temperature gives rise to a hydrolysis re-
action between the [BF4]ꢁ and the residual water.
The thermal properties of these prepared salts, including
the QA bromide, were determined by differential scanning
calorimetry (DSC), and thermal gravimetric analysis
(TGA). For the salts that are liquids at 258C, the fundamen-
tal properties including density (d), dynamic viscosity (h),
ionic conductivity (k), and electrochemical stability, were
also determined. Some of the characterization data are pre-
754
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 752 – 766

Hydrophobic Ionic Liquids
FULL PAPER
sented in Tables 1–3. The influence of the structure variation
in the cation and anion on the above properties will be de-
tailed in the following sections.
crystal phases in simple molecular compounds typically have
a DSm <20 JKꢁ1 molꢁ1. Recently, MacFarlane[29] and co-
workers reported that it is more appropriate to modify this
criterion to ꢀ40 JKꢁ1 molꢁ1 when a relative large and flexi-
ble anion (e.g. TFSIꢁ) exists in the ionic compound. This
new criterion appears to be applicable to the [RFBF3]ꢁ salts
(RF = CF3, C2F5, n-C3F7, n-C4F9), as these new anions con-
taining a perfluoroalkyl (RF) chain may have higher degrees
of freedom than those with simple symmetry (e.g. [BF4]ꢁ[23]
and [PF6]ꢁ[24]), and the salts with the former anions, there-
fore, may afford higher residual entropies of melting than
those with the latter anions. Thus, it seems reasonable that
many of the [RFBF3]ꢁ salts, which show solid–solid transi-
tions with a DSm <40 JKꢁ1 molꢁ1 (Table 1), are exhibiting
varying extents of plastic crystal behavior. It is of current in-
terest to explore conductive organic plastic crystals as an
all-solid-state electrolyte for high-energy devices, such as Li
batteries.[49,50] Therefore, our future work will investigate the
conductivity behavior of these salts and mixtures with their
corresponding lithium salts. Among the salts to be investi-
gated, N1223[CF3BF3] (entry 1) and N1223[C2F5BF3] (entry 7)
may be the particularly interesting ones that enter a plastic
crystal phase at a low temperature (ꢀꢁ408C for entry 7
and ꢀꢁ218C for entry 1) and exist in this phase up to a
high temperature (958C for entry 1 and 548C for entry 7)
before melting. The plastic crystal phase within these tem-
perature regions is very suitable for practical applications
because the electrolytes in many electrochemical devices are
required to operate in an ambient temperature range (from
ꢁ40 to 808C).
Phase transition: The solid–liquid phase transitions of these
QA salts were investigated by differential scanning calorim-
etry (DSC). The data for the glass transition temperature
(Tg), crystallization temperature (Tc), solid–solid transition
(Ts–s), and melting point (Tm), if appropriate, are collected
in Table 1. In general, four types of phase transition behav-
ior were observed for these QA salts on heating. The typical
DSC traces for these four types of behavior are exemplified
in Figure 1. The first type of behavior is characterized by
Figure 1. DSC traces on heating for a) N122.1O2[CF3BF3], b) N1224[n-
C3F7BF3], c) N1223[CF3BF3], d) N122.1O2[C2F5BF3], and e) N112.1O2[C2F5BF3].
Inset: an enlarged trace for e) N112.1O2[C2F5BF3] in the temperature range
close to Tg.
Melting point: It was reported that the melting point (Tm)
of a compound is determined by the strength of its crystal
lattice, which in turn is controlled by three main factors: in-
termolecular forces, molecular symmetry, and the conforma-
tional degrees of the freedom of a molecule.[51] For aliphatic
QA salts, the influence of structure variation in the cation
and anion on the Tm has been investigated previous-
ly.[15–17,22–24,27–29,31] Generally, low symmetry and a relatively
large size of the cation combined with low symmetry and ef-
fective charge distribution in the anion favor destabilization
of the packing efficiency in the crystal lattice of the QA
salts, thus decreasing the strength of the crystal lattice which
in turn lowers the Tm. Additionally, introducing an alkyl
ether to the QA cation also tends to depress the Tm in some
cases.[9,19,25,26]
Therefore, on the basis of this knowledge, we tried to pre-
pare low-melting QA salts from QA cations with relatively
low symmetry and/or having an alkyl ether substituent, as
shown in Scheme 1. Indeed, low melting points were ach-
ieved for most of the salts in our study (Table 1). With the
exceptions of N1223[BF4] and N1224[BF4], all the prepared
salts have a Tm of <1008C, 23 are liquids at room tempera-
ture, most of the liquids are those containing an alkyl ether
(CH3OCH2CH2) substituent in the cation.
the salts that have a single melting transition (Figure 1a, en-
tries 5, 6, 9, 21, 25–27, 30, and 33 in Table 1). The second
type of behavior is represented by a number of salts that ex-
hibited one or multiple solid–solid transitions (Ts–s) before
melting (Figure 1b and c, entries 1–4, 7, 8, 13–15, 18–20, 24,
and 31 in Table 1). As can be seen in Table 1, most of the
[RFBF3]ꢁ salts show this behavior; however, such behavior
was not observed among the [BF4]ꢁ salts. The unique prop-
erty associated with this behavior will be detailed below.
The third type of behavior is represented by the salts that
only exhibited glass transitions at a very low temperature in
the range ꢁ95 to ꢁ1138C without melting (Figure 1d, en-
tries 11, 16, 17, 23, 34, and 35 in Table 1). The fourth type of
behavior was observed in the remaining salts (entries 10, 12,
22, 28, 29, 32, and 36 in Table 1). A typical DSC trace for
this behavior is shown in Figure 1e. On heating, these salts
showed a glass transition (Tg) at a very low temperature
(from ꢁ95 to ꢁ1178C) to form a supercooled liquid, which
then crystallized (Tc, observed as an exothermic peak) at a
relatively high temperature, followed by a melting transition
(Tm) as the temperature increased further.
The melting entropies (DSm) of the prepared salts are also
Figure 2 shows the observed Tm of these salts. As a gener-
al trend, for each anion, the salts with cations of higher sym-
presented in Table 1. Timmermans[48] reported that plastic
Chem. Eur. J. 2005, 11, 752 – 766
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
755

H. Matsumoto et al.
ethoxyethyl
(CH3CH2OCH2),
and
(CH3CH2OCH2CH2).[9,19,25,26] Given that the alkyl ether
group is more flexible but more polar than its isoelectronic
alkyl group, and the flexibility tends to lower the Tm, where-
as the polarity has an opposite effect, a tentative explana-
tion for the lower Tm for the QA salts containing an alkyl
ether group is that the better flexibility of the alkyl ether
group might prevail over its polarity and thus plays a deci-
sive role in determining the Tm.
Glass transition: Figure 3 shows the glass transition temper-
ature (Tg) of the salts in Table 1. It appears that the Tg of
these QA salts is mainly governed by the strength of ion in-
Figure 2. Melting point of various salts.
metry show a higher Tm than those with cations of lower
+
symmetry. For example, the [N2222
]
salts with [RFBF3]ꢁ (RF
= C2F5, n-C3F7, n-C4F9) decompose directly at a high tem-
perature (278–2908C) instead of melting;[38] however, replac-
+
ing the [N2222
]
of these salts with [N1223]+, which has a
lower symmetry but the same formula weight, lowers the Tm
to the low-melting range (54–978C). This is also true for the
TFSI salts. For example, N2222[TFSI] (105–1098C)[16,17] versus
N
1223[TFSI] (148C). Likewise, the salts with lower symmetry
+
+
[N112.1O2
]
and [N122.1O2
]
(containing three different sub-
stituents) generally exhibit a lower Tm than the correspond-
+
+
ing salts with higher symmetry [N111.1O2
]
and [N222.1O2]
(containing two different substituents). Exchanging the
Figure 3. Glass transition temperature of various ionic liquids.
+
+
[N1223] of the salts for the larger [N1224] generally gave rise
to a further decrease of the Tm (e.g. entry 1 versus 2, 7
versus 8, 13 versus 14, 25 versus 26, 31 versus 32 in Table 1).
Likewise, for the same cation, all the salts with lower sym-
metry, [RFBF3]ꢁ and TFSIꢁ, exhibit a lower Tm than those
with higher symmetry [BF4]ꢁ.
Another notable effect on the Tm of the aliphatic QA salt
is to introduce a methoxyethyl (CH3OCH2CH2) group to the
QA cation, which was also observed in other publica-
tions.[19,25,26] However, the reasons for this effect are not well
understood. To gain an insight into the influence of the alkyl
teractions, the ion size, as well as the flexibility and polariza-
bility of the anion. As seen in Figure 3, the Tg of the salts
with large [RFBF3]ꢁ (ꢁ117 to ꢁ988C, RF = C2F5, n-C3F7, n-
C4F9) are all lower than those with the small [BF4]ꢁ (ꢁ978C
for entry 28 and ꢁ958C for entry 29) or the large, but a
little polarizable TFSIꢁ (ꢁ82 to ꢁ968C), and much lower
than those with the large nonfluorinated and more rigid and
polarizable [BOB]ꢁ (it comprises two rigid five-member
rings and each of them contains two carbonyl (C=O) groups,
Tg = ꢁ36 to ꢁ448C).[26]
ether chain on the Tm of the aliphatic QA salt, we therefore
+
carried out a comparative study of the Tm of the [N1224
]
For the same cation, the Tg regularly increases in the
order [C2F5BF3]ꢁ <[n-C3F7BF3]ꢁ <[n-C4F9BF3]ꢁ <[BF4]ꢁ ꢀ
TFSIꢁ ![BOB]ꢁ. A similar trend was also observed in the
related imidazolium salts.[44,45] This trend strongly suggests
that 1) replacing the [BF4]ꢁ with a larger [RFBF3]ꢁ weakens
the electrostatic interactions (including possible hydrogen
bonding) between the cation and anion in the salt, which in
turn decreases the Tg, 2) the decrease of the electrostatic in-
teractions dominates over the increase of van der Waals in-
teractions in the [RFBF3]ꢁ salts (RF = C2F5, n-C3F7, n-C4F9)
owing to the extremely low polarizability of fluorine atom,
and 3) the ion interactions in the salts with the small [BF4]ꢁ
are influenced by the electrostatic interactions to a greater
extent than by van der Waals interactions, while those in the
+
salts with those of the [N122.1O2
]
salts. As shown in Figure 2,
salts, replacing the n-butyl group in the
] with an isoelectronic methoxyethyl (CH3OCH2CH2)
+
for all the [N1224
]
+
[N1224
group resulted in a significant decrease in the Tm in all
cases. For example, N122.1O2[BF4] (88C) versus N1224[BF4]
(1658C),
N
122.1O2[CF3BF3] (ꢁ228C) versus N1224[CF3BF3]
+
(ꢁ38C), and only Tg (ꢁ95 to ꢁ1138C) for the [N122.1O2
]
salts with [RFBF3]ꢁ (RF = C2F5, n-C3F7, n-C4F9, entries 11,
17, and 23) and TFSIꢁ (entry 35). In addition, the salts with
the cations [N111.1O2]+, [N112.1O2]+, and [N222.1O2
]
also show a
+
low Tm. Moreover, low-melting events have also been ob-
served in the QA salts containing other alkyl ether groups,
such as methoxymethyl (CH3OCH2), ethoxymethyl
756
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 752 – 766

Hydrophobic Ionic Liquids
FULL PAPER
[RFBF3]ꢁ salts (RF = C2F5, n-C3F7, n-C4F9) are more under
the control of van der Waals interactions rather than electro-
static interactions because the Tg increased as the size of the
[RFBF3]ꢁ increased.
more stable than those containing [RFBF3]ꢁ, irrespective of
the cations, and the thermal stability of these salts generally
increases in the order [CF3BF3]ꢁ <[n-C3F7BF3]ꢁ ꢀ[n-
C4F9BF3]ꢁ ꢀ[C2F5BF3]ꢁ <[BF4]ꢁ ꢀTFSIꢁ. This trend indi-
cates that the anion has a significant impact on the thermal
stability of the QA salts, as observed in the imidazolium
salts with these anions.[45] Figure 5 displays thermogravimet-
On the other hand, one may note that the TFSIꢁ salts
show a slightly higher Tg than the corresponding [BF4]ꢁ
salts, and a much higher Tg than the corresponding [RFBF3]ꢁ
salts (Figure 3 and Table 1), although the TFSIꢁ has a much
better flexibility and charge distribution than the [BF4]ꢁ
salts and would be expected to give a lower Tg for the corre-
sponding salts. This result may be attributed to 1) the size of
the TFSIꢁ is much larger than that of the [BF4]ꢁ, and 2) the
partially fluorinated TFSIꢁ (it contains two sulfonyl (-SO2-)
groups) is more polarizable than the perfluorinated [BF4]ꢁ
and highly perfluorinated [RFBF3]ꢁ, both of which signifi-
cantly trade off the flexibility and charge distribution of the
TFSIꢁ and tend to increase the attractive forces (i.e. van -
der Waals interactions) in the TFSIꢁ salts, thus raising the
Tg. Moreover, comparing the Tg of the nonfluorinated
[BOB]ꢁ salts with that of the perfluorinated [n-C4F9BF3]ꢁ
salts (the size of the [n-C4F9BF3]ꢁ should be larger than that
of the [BOB]ꢁ viz. [B(C2O4)]ꢁ) more clearly manifests the
significant impact of rigidity and polarizability of the anion,
for example, N112.1O2[BOB] (ꢁ368C)[26] versus N112.1O2[n-
C4F9BF3] (ꢁ1108C).
+
Figure 5. TGA traces of the [N1224
]
salts with [RFBF3]ꢁ (RF = CF3, C2F5,
n-C3F7, n-C4F9), [BF4]ꢁ, and TFSIꢁ.
For the same anion, the Tg value of all the salts increased
as the cation size increased, as expected by the increased
van der Waals interactions. Because the glass transition tem-
perature (Tg) is a qualitative signature of ion mobility in ILs
and a low Tg is a sign of high ion mobility,[16b,25,26] the lower
Tg of the [RFBF3]ꢁ salts compared to that of the [BF4]ꢁ salts
may partly explain why the former exhibit lower viscosities
than the latter (see the viscosity section below).
+
ric traces of six [N1224
]
salts with different anions as exam-
ples to illustrate the different decomposition behavior of
these salts during pyrolysis. The thermal stability order and
decomposition behavior of these QA salts with [RFBF3]ꢁ is
concurrent with our recent observations with the corre-
sponding imidazolium salts.[45] The lower thermal stability of
the [RFBF3]ꢁ salts, relative to those containing [BF4]ꢁ and
TFSIꢁ, is essentially attributed to the pyrolysis of [RFBF3]ꢁ
at a low temperature.[45] We will now describe the decompo-
sition of the N1224[CF3BF3] in detail, as exemplified by the
family of [RFBF3]ꢁ salts. As seen in Figure 5, N1224[CF3BF3]
exhibited a distinct two-stage decomposition behavior, as
observed in the related imidazolium salts.[45] It started to de-
compose at ꢀ2008C, and ꢀ18% of the weight loss was ob-
served up to 3008C, which corresponds to the elimination of
a CF2 moiety from this salt, most probably [CF3BF3]ꢁ!
[BF4]ꢁ + other products (formed from CF2 carbene). Subse-
quently, the weight of the residue remained almost constant
until ꢀ3808C. It began to decompose again above ꢀ3908C,
and the subsequent decomposition behavior of the residue
was exactly in agreement with that of N1224[BF4]. Similar de-
composition behavior was also observed in the other QA
salts with [CF3BF3]ꢁ. These results strongly indicate that de-
composition of the [CF3BF3]ꢁ salts is initiated from the py-
rolysis of [CF3BF3]ꢁ.
Thermogravimetric analysis: The thermal stability of the
salts was determined by TGA. Figure 4 shows the decompo-
sition temperature (Td) of the QA salts in Table 1. As
shown in Figure 4, the salts containing TFSIꢁ and [BF4]ꢁ are
For a constant anion, it seems that the variation in the
QA cation has less effect on the thermal stability of the cor-
responding salts, with the exception of the [CF3BF3]ꢁ series,
where the salts containing the methoxyethyl group in the
Figure 4. Thermal stability of various ionic liquids.
Chem. Eur. J. 2005, 11, 752 – 766
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
757

H. Matsumoto et al.
QA cation show less thermal stability than their tetraalkyl-
ammonium analogues, as shown in Figure 4.
the electrostatic and van der Waals interactions), ion size,
and by the polarizability and flexibility of the anion.
Figure 7 shows the viscosities (h) of the liquid salts based on
the [RFBF3]ꢁ (RF = CF3, C2F5, n-C3F7, n-C4F9), [BF4]ꢁ, and
TFSIꢁ (Table 2) at 258C. The viscosities of the [RFBF3]ꢁ and
Density: Figure 6 shows the densities of the liquid QA salts
based on the [RFBF3]ꢁ (RF = CF3, C2F5, n-C3F7, n-C4F9),
[BF4]ꢁ, and TFSIꢁ at 258C. The density (d) lies in the range
Figure 7. Viscosity of various liquid salts at 258C.
Figure 6. Density of various liquid salts at 258C.
1.16–1.55 gmLꢁ1. Because a number of the liquid salts have
a methoxyethyl (CH3OCH2CH2) group in their cation, the
density trend is more noticeable for these salts. As seen in
Figure 6, for a given anion, the densities of the trialkyl(2-
methoxyethyl)ammonium salts decreased when the formula
weight (Mw) of the cation increased. The densities of the tri-
Table 2. Density, viscosity, and specific conductivity of liquid salts at
258C (water content: <50 ppm in [RFBF4]ꢁ and TFSIꢁ salts, and 400–
600 ppm in the [BF4]ꢁ salts).
Entry[a]
Salts
d[b]
h[c]
k[d]
[gmLꢁ1
]
[cP]
[mScmꢁ1
]
2
4
5
6
N
1224[CF3BF3]
1.18
1.27
1.25
1.22
1.25
1.33
1.31
1.28
1.41
1.39
1.37
1.34
1.45
1.42
1.40
1.21
1.20
1.42
1.38
1.45
1.42
1.40
210
97
108
151
104
58
68
87
76
70
2.1
2.5
3.0
2.0
2.3
3.8
3.2
2.4
2.5
2.6
1.9
1.8
1.5
1.3
1.1
1.7[f]
1.3
2.2
1.6
3.1
2.6
2.1
N112.1O2[CF3BF3]
N122.1O2[CF3BF3]
N222.1O2[CF3BF3]
N1224[C2F5BF3]
N112.1O2[C2F5BF3]
N122.1O2[C2F5BF3]
N222.1O2[C2F5BF3]
alkyl(2-methoxyethyl)ammonium salts are a little higher
+
than those of the [N1224
]
salts with the same anion. For ex-
ample, keeping the [C2F5BF3]ꢁ constant, the density falls in
the order 9 ([N111.1O2]+)>10 ([N112.1O2]+)>11 ([N122.1O2]+)>
12 ([N222.1O2]+)>8 ([N1224]+).
8
10
11
12
15
16
17
18
22
23
24
28
29
31
32
34
35
36
On the other hand, for a given cation, the density gradual-
ly increased as the bulkiness of the fluoroanion increased in
the order [BF4]ꢁ <[CF3BF3]ꢁ <[C2F5BF3]ꢁ <[n-C3F7BF3]ꢁ <
[n-C4F9BF3]ꢁ ꢀTFSIꢁ. This trend is consistent with our
recent results for the related imidazolium ILs.[45] Moreover,
as shown in Figure 5, in the family of the trialkyl(2-methox-
yethyl)ammonium ([Nxyz.1O2]+) with [CnF2n+1BF3]ꢁ (n = 0–
4), it seems that the density decreased linearly with increas-
ing number of CH2 units in the [Nxyz.1O2]+, and linearly in-
creased with increasing the number of CF2 units in the
[CnF2n+1BF3]ꢁ (n = 0–4). By a simple regression analysis,
the densities of these 15 liquids can be expressed linearly
[Eq. (1)], for m = 6–9, n = 0–4; r2 = 0.998, where m and n
represent the number of carbon atoms in the cation and
anion, respectively.
N
111.1O2[n-C3F7BF3]
N112.1O2[n-C3F7BF3]
N122.1O2[n-C3F7BF3]
N222.1O2[n-C3F7BF3]
N112.1O2[n-C4F9BF3]
N122.1O2[n-C4F9BF3]
N222.1O2[n-C4F9BF3]
N112.1O2[BF4]
N122.1O2[BF4]
N1223[TFSI]
N1224[TFSI]
N112.1O2[TFSI]
88
91
102
118
135
335[e]
426
94
120
60
69
N122.1O2[TFSI]
N222.1O2[TFSI]
85
[a] In accord with the entries in Table 1. [b] Density. [c] Viscosity. [d] Spe-
cific conductivity. [e] 1.7.[25] [f] ꢀ335.[25]
TFSIꢁ series of ILs are in the range 58–210 cP, all of which
are significantly lower by >100 cP than those of the [BF4]ꢁ
salts (335 cP for entry 28 and 426 cP for entry 29 at 258C),
and much lower than those of the QA salts with the non-
fluorinated [BOB]ꢁ (>1300 cP at 258C[26]).
d ðg mLꢁ1Þ ¼ 1:38 ꢁ 0:024 m þ 0:058 n
ð1Þ
Viscosity: The viscosity of the QA salts is essentially deter-
mined by the strength of ion interactions (mainly including
758
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 752 – 766

Hydrophobic Ionic Liquids
FULL PAPER
The lower viscosities for the [RFBF3]ꢁ and TFSIꢁ ILs than
those of the [BF4]ꢁ salts may be explained by the decreased
ion interactions and increased ion mobility owing to im-
proved charge distribution, lower symmetry, and lower po-
larizability for the [RFBF3]ꢁ salts, and to improved charge
distribution, larger flexibility, and lower symmetry for the
TFSIꢁ salts. Notwithstanding that the [BOB]ꢁ is among the
most weakly coordinating anions,[53] the extremely high vis-
cosities were observed for all [BOB]ꢁ salts,[26] which is essen-
tially attributed to high rigidity (comprising two five-mem-
bered rings) and some polarizability (containing four car-
bonyl (C=O) groups) and large size of the [BOB]ꢁ, as dis-
cussed for the Tg. The first factor reduces the flexibility and
the latter two ones increase the van der Waals interactions
in the salts, all of which in turn significantly increase the vis-
cosity. These impressive examples based on the [BOB]ꢁ and
its derivatives clearly indicate that an anion with high rigidi-
ty and polarizability and large size is not suitable for form-
ing low-viscous ILs, despite its weakly coordinating abili-
ty.[26]
tively low symmetry, high flexibility, and good charge distri-
bution, are expected to be discovered and investigated in
future.
The structure variation in the cation also influences the
viscosity. For the same anion, the viscosities of the QA salts
containing a methoxyethyl (CH3OCH2CH2) group are gen-
erally lower than those of the tetraalkylammonium salts
(e.g. entries 9–12 versus 8). This may be attributable to
having a flexible methoxyethyl group in the QA cation that
results in an increase in the ion mobility, as the Tg of the
former group of salts is lower than that of the latter one
(Figure 3). In the family of QA salts containing a methox-
yethyl group, the viscosity increased as the cation size in-
creased in the order [N112.1O2]+, [N111.1O2]+ <[N122.1O2]+ <
[N222.1O2]+, as expected by the increase of van der Waals in-
teractions. This trend suggests that, to reduce the viscosity
of the QA salts further, the cation involved should be small-
er, or at least not larger, than those used in this study.
It should be noted that the viscosities of all the QA salts
in the present study are all remarkably higher than those of
the related 1,3-dialkylimidazolium salts with a similar for-
mula weight.[43–45] For example, the viscosity of the
In the family of the fluoroborate ([RFBF3]ꢁ and [BF4]ꢁ)
ILs, for a given cation, the viscosities show the increasing
order
[C2F5BF3]ꢁ <[n-C3F7BF3]ꢁ <[CF3BF3]ꢁ <[n-
N
112.1O2[CF3BF3] (Mw 269) is 97 cP at 258C, which is ꢀ
C4F9BF3]ꢁ ![BF4]ꢁ, with the exception of N222.1O2[CF3BF3].
It is very interesting to note that this order of viscosity is ex-
actly consistent with that of the Tg (i.e. the lower the Tg, the
more fluid the ILs), with the exception of the salts contain-
ing the [CF3BF3]ꢁ (no Tg observed). These results further
suggest that the ion interactions in the salts with relative
large [RFBF3]ꢁ (RF = C2F5, n-C3F7, n-C4F9) are essentially
determined by van der Waals interactions, and a very large
anion in this class (e.g. [n-C4F9BF3]ꢁ) is therefore unsuitable
for forming low-viscous ILs, despite its good charge distribu-
tion and extremely low polarizability. The most unexpected
result is that the salts with the smaller [CF3BF3]ꢁ show a
1.3 times greater than that of the 1-methyl-2-methoxyethyli-
midazolium salt with [CF3BF3]ꢁ (43 cP at 258C, Mw
278).[44,45] It seems that the almost flat shape and good
charge distribution of the imidazolium ring plays a critical
role in lowering the viscosity. This may explain why the five-
membered cyclic QA cations, N-alkyl-N-methylpyrrolidini-
um, are preferred to the acyclic QA cations to form ILs in
recent studies.[11,12,29] The reason is that the former cations
show a quasiflat shape to some extent and can form more
fluid ILs than acyclic QA cations containing the same
number of carbon atoms, for example, P13[TFSI] ([P13]+ N-
methyl-N-propylpyrrolidinium, 63 cP at 258C) versus
higher viscosity than those with the larger [RFBF3]ꢁ (RF
=
N
1124[TFSI] (110 cP at 258C).[16c] Therefore, it seems reason-
C2F5, n-C3F7). It appears that 1) ion interactions in the
former are stronger than those in the latter and are still
under the influence of the electrostatic interactions owing to
the small size of the [CF3BF3]ꢁ, as in the case of the [BF4]ꢁ
able that [RFBF3]ꢁ could form low-viscous ILs with the pyr-
rolidinium cations.
Conductivity: It appears that the ionic conductivities of
these QA salts are mainly governed by their viscosities and
formula weight. Figure 8 shows the specific conductivities
(k) of the liquid salts based on the [RFBF3]ꢁ (RF = CF3,
C2F5, n-C3F7, n-C4F9), [BF4]ꢁ, and TFSIꢁ at 258C (Table 2).
salts; and 2) the [RFBF3]ꢁ with a longer RF group (RF
=
C2F5, n-C3F7) have more degrees of freedom than the
[CF3BF3]ꢁ, thus lowering the viscosity. The lowest viscosities
for the [C2F5BF3]ꢁ salts are essentially attributed to a good
charge distribution and medium size for the [C2F5BF3]ꢁ.
As shown in Figure 7, for the same cation, the viscosity
values of the TFSIꢁ salts are nearly equivalent to those of
the most fluid [C2F5BF3]ꢁ series (e.g. N112.1O2[TFSI] (60 cP)
versus N112.1O2[C2F5BF3] (58 cP) and N122.1O2[TFSI] (69 cP)
versus N122.1O2[C2F5BF3] (68 cP)), although the [TFSI]ꢁ has a
larger size than the [C2F5BF3]ꢁ. It seems that improved flex-
ibility and charge delocalization of the [TFSI]ꢁ plays a dom-
inant role over its size and polarizability in determining the
viscosity. All the above results indicate that the viscosities of
the QA salts are strongly dependent on the type of anion
present. Therefore, in the search of more low-viscous QA
salts, new stable anions, especially fluoroanions with rela-
For a given anion, the conductivity values decreased in the
+
order [N112.1O2]+ >[N122.1O2]+ >N1223
,
[N222.1O2]+ >[N1224
]
with the exception of the N122.1O2[CF3BF3], as expected by
the increased viscosity (Figure 7), and formula weight be-
cause the number of the ion carriers per unit volume gener-
ally decreased with increasing formula weight. For a given
cation, the conductivity decreased in the order [C2F5BF3]ꢁ >
TFSIꢁ, [n-C3F7BF3]ꢁ, [CF3BF3]ꢁ >[n-C4F9BF3]ꢁ ꢀ[BF4]ꢁ.
This trend strongly indicates that an anion, which has a
large formula weight (e.g. TFSIꢁ, [n-C4F9BF3]ꢁ) or results in
high viscosity (e.g. [BF4]ꢁ, [CF3BF3]ꢁ), is unfavorable for
producing highly conductive ILs. Among these QA salts, the
N
112.1O2[C2F5BF3] (entry 10) exhibits the highest conductivity
Chem. Eur. J. 2005, 11, 752 – 766
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
759

H. Matsumoto et al.
Table 3. Cathodic and anodic limits (Ecathodic and Eanodic versus Fc+/Fc at
1 mAcmꢁ2) and electrochemical windows (DE) for various ionic liquids
determined on a glassy carbon electrode.
Salts
Ecathodic
Eanodic
DE
[V]
[V]
[V]
N1114[TFSI]
N1224[CF3BF3]
ꢁ3.35
ꢁ3.43
ꢁ3.43
ꢁ3.44
ꢁ2.75
ꢁ2.71
ꢁ2.95
ꢁ3.13
ꢁ3.38
ꢁ3.32
ꢁ3.33
ꢁ3.44
ꢁ3.40
ꢁ3.33
ꢁ3.41
ꢁ3.41
ꢁ3.44
ꢁ2.49
2.50
2.32
2.21
2.25
2.24
2.36
2.21
2.35
2.32
2.29
2.28
2.21
2.24
2.34
2.29
2.22
2.29
2.14
5.85
5.75
5.64
5.69
4.99
5.07
5.16
5.48
5.70
5.61
5.61
5.65
5.63
5.67
5.70
5.63
5.73
4.63
N
1224[C2F5BF3]
N1224[TFSI]
N111.1O2[n-C3F7BF3]
N
111.1O2[TFSI][a]
N112.1O2[CF3BF3]
N112.1O2[TFSI]
N122.1O2[CF3BF3]
N122.1O2[C2F5BF3]
N122.1O2[n-C3F7BF3]
N122.1O2[n-C4F9BF3]
N122.1O2[BF4]
N
122.1O2[TFSI]
N222.1O2[CF3BF3]
N222.1O2[C2F5BF3]
N222.1O2[TFSI]
Figure 8. Specific conductivity of various liquid salts at 258C; the data in
parentheses is formula weight (Mw) of the cation.
[C2mmi][CF3BF3]
(3.8 mScmꢁ1 at 258C) because it has the lowest viscosity
(58 cP at 258C) and a relatively low formula weight.
On the other hand, all the QA salts in this study show
much lower conductivities than the imidazolium salts with a
similar formula weight, for example, the most conductive
QA salt N112.1O2[C2F5BF3] (3.8 mScmꢁ1 and 57 cP at 258C,
Mw 319) versus the 1-methyl-2-methoxyethylimidazolium
salt with [C2F5BF3]ꢁ[45] (6.1 mScmꢁ1 and 38 cP at 258C, Mw
328). This is mainly caused by the higher viscosities for the
former (57 cP at 258C) than for the latter[45] (38 cP at 258C).
Therefore, one of the significant challenges in developing
conductive QA salts is how to reduce their viscosity even
further.
[a] Melted at 358C and determined at room temperature (supercooled).
Electrochemical stability: The electrochemical stability of
most QA salts was investigated by linear sweep voltamme-
try (LSV) on a glassy carbon electrode under the same con-
ditions. The cathodic and anodic limits (Ecathodic and Eanodic
(V) versus ferrocene (Fc)/ferrocenium (Fc+) redox couple)
were arbitrarily defined as the potential at which the current
density reached 1 mAcmꢁ2 in the first scan and IR drop (i.e.
a voltage drop associated with the electrical resistance of
the electrolyte) was negligible on account of moderate con-
ductivities. The respective Ecathodic and Eanodic and electro-
chemical windows (DE (V) = EanodicꢁEcathodic) for these QA
salts are summarized in Table 3, including previously report-
Figure 9. Linear sweep voltammogram of various ionic liquids on a glassy
carbon electrode (surface area: 7.85ꢁ10ꢁ3 cmꢁ2) in the first scan at 258C;
scan rate: 50 mVsꢁ1; counterelectrode: Pt wire; potential (V) was refer-
red to ferrocene (Fc)/ferrocenium (Fc+) redox couple in each salt; the
data for [C2mim][CF3BF3] ([C2mim]+
were obtained from the literature.[44–45]
= 1-ethyl-3-methylimidazolium)
ed values for
N
1114[TFSI][9b,16c] and [C2mmi][CF3BF3]
the same cathodic stability, irrespective of the alkyl substitu-
ents in the cation.[20,21] These are also the cases for our QA
([C2mim]+ 1-ethyl-3-methylimidazolium)[44,45]. Figure 9
=
displays the polarization curves of seven salts, which mainly
manifests the impact of the cation structure on the Ecathodic
(i.e. the significant influence of having a methoxyethyl
group in the QA cations and the sizes of these cations on
the cathodic stability).
In general, the cathodic stability of the QA salts with
electrochemically stable anions, such as [BF4]ꢁ, TFSIꢁ, and
[PF6]ꢁ, is essentially limited by the decomposition of the
QA cation. The tetraalkylammonium cations show almost
salts. For example, the two tetraalkylammonium cations,
+
[N1114
]
and [N1224]+, show nearly the same Ecathodic value (ꢀ
ꢁ3.4 V). For the same cation, the Ecathodic values of the QA
salts were little affected by the anion species, for example,
+
all the [N122.1O2
]
salts with various anions ([RFBF3]ꢁ,
[BF4]ꢁ, and TFSIꢁ) show a nearly equivalent Ecathodic value
(ꢁ3.3 to ꢁ3.4 V). However, the cathodic stability of the
salts in this study is highly dependent on the nature of the
cations, as indicated by their different Ecathodic values
760
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 752 – 766

Hydrophobic Ionic Liquids
FULL PAPER
(Table 3 and Figure 9). All the QA cations exhibit better
ity of 1,3-dlialkyimidazolium salts with electrochemically
stable anions, such as [RFBF3]ꢁ, [BF4]ꢁ, and TFSIꢁ, are lim-
ited by the cations but not by the anions.[43,45] However, as
seen in Figure 9 and Table 3, the anodic stability of these
anions is hardly distinguishable, since the Eanodic values for
these QA salts are very close and show some dependence
on the cations. Additionally, it should be noted that the pre-
viously reported improved anodic stability for the inorganic
fluoroanion, [BF4]ꢁ, than for the organic fluoroanion,
TFSIꢁ, was not observed in this study.[13,21,22] This may be at-
tributed to the presence of relatively high amounts of water
in our [BF4]ꢁ salts (400–600 ppm), which reduces the anodic
stability of the salts, as reported in the literature.[54]
cathodic stability than the [C2mim]+ in the following order
+
(Table 3): [C2mim]+ (ꢀꢁ2.5 V)<[N111.1O2
]
(ꢀꢁ2.7 to
+
+
ꢁ2.8 V)<[N112.1O2
]
(ꢀꢁ3.0 to ꢁ3.1 V)<[N122.1O2
]
(ꢀꢁ3.3
+
+
+
to ꢁ3.4 V)<[N1114
]
(ꢀꢁ3.4 V) ꢀ[N222.1O2
]
ꢀ[N1224
]
(ꢀ
ꢁ3.4 V). This order clearly indicates that replacing the alkyl
in the tetraalkylammonium cations with an isoelectronic
alkyl ether tends to reduce the cathodic stability of the cati-
ons, which is more pronounced for the small cations, for ex-
+
+
ample, [N1114
]
(ꢀꢁ3.4 V) versus [N111.1O2
]
(ꢀꢁ2.7 to
ꢁ2.8 V), whereas increasing the sizes of trialkyl(methoxy-
ethyl)ammonium cations tends to increase their cathodic
stability, as indicated by the increased Ecathodic values. A pre-
liminary explanation for these results is that there are some
electrostatic interactions occurring between the lone-pair
electrons of oxygen atom of the methoxyethyl group and
positive charge nitrogen atom in the cation, which probably
increases the cathodic stability of the positively charged ni-
trogen atom through decreasing positive charge density on
the nitrogen atom; however, it reduces the cathodic stability
of the methoxyethyl group owing to a decrease of electron
density on its oxygen atom. These interactions are stronger
for the smaller trialkyl(methoxyethyl)ammonium on account
of its higher charge density. Thus, the lower cathodic stabili-
ty for the small trialkyl(methoxyethyl)ammonium cations
may be caused by electroreduction of the methoxyethyl
group in the cation at a relative high potential, while in-
creasing the trialkyl(methoxyethyl)ammonium size may ef-
fectively “bury” the positive charge on the nitrogen atom of
the ammonium, which in turn decreases the interactions
mentioned above, thus inhibiting the decomposition of the
methoxyethyl group (i.e. increasing the cathodic stability of
the cation), as indicated by the highest Ecathodic values for the
[N222.1O2]+. These explanations have some support from re-
sults reported by Aurbach and co-workers,[53] in which the
authors have identified that the electroreduction of the
Conclusion
A novel family of low-melting, hydrophobic ionic liquids
containing small aliphatic quaternary ammoniums
([R1R2R3NR]+, wherein R1, R2, R3 = CH3 or C2H5, R = n-
C3H7, n-C4H9, CH2CH2OCH3) and perfluoroalkyltrifluoro-
borates ([RFBF3]ꢁ, RF = CF3, C2F5, n-C3F7, n-C4F9) has
been prepared and characterized. Most of them are liquids
at room temperature and exhibit much lower viscosities
(58–210 cP at 258C) than those with [BF4]ꢁ (335–426 cP at
258C) owing to the relatively lower symmetry and better
charge distribution of the [RFBF3]ꢁ. Many of them exhibit
plastic crystal behavior, which has a fundamental impor-
tance in all-solid-state electrolytes. More importantly, all
these quaternary ammonium salts exhibit improved cathodic
and anodic stability, hence larger electrochemical windows,
than the corresponding 1,3-dialkyimidazolium salts. Because
of these desirable properties, these new salts not only enrich
the existing arsenal of quaternary ammonium ionic liquids,
but may also serve as new media for some special organic
and biocatalytic reactions and as electrolytes for high-
energy devices and electrochemical processes in which the
1,3-dialkylimidazolium salts cannot operate.
N
4444[ClO4] solutions of ether solvents, such as DME
(CH3OCH2CH2OCH3) and THF, was essentially associated
with the decomposition of the [N4444]+; however, addition of
Li+ ions to this electrolyte caused the electroreduction of
+
DME and THF to precede that of the [N4444
]
owing to the
pronounced interactions between the small Li+ and the
ether solvents. From these results, we may conclude that
having an alkyl ether chain in the small QA cations is unfav-
orable for producing electrochemically stable QA salts. The
Experimental Section
General: The following reagents and materials were used as received: di-
methylethylamine, diethylmethylamine, triethylamine, 1-bromopropane,
1-bromobutane, 2-bromoethyl methyl ether (Tokyo Kasei); 30% aqueous
trimethylamine; anhydrous diethyl ether, acetone, and tetrahydrofuran
(H2O<30 ppm) (Wako Pure Chem.); bis(trifluoromethanesulfonyl)imide
acid (Central Glass); 48% aqueous tetrafluoroboric acid (Aldrich);
acidic cation-exchange and basic anion-exchange resins (Mitsubishi
Chem., ion exchange capacity>2.0 meq.mLꢁ1).
distinctive cathodic stability of the tetraalkylammoniums
+
(e.g. [N1114
]
and [N1224]+) and large trialkyl(methoxyethyl)-
ammonium (e.g. [N222.1O2]+) may favor their salts with elec-
trochemically stable anions as promising candidates for elec-
trolytes in high-energy devices.
Potassium perfluoroalkyltrifluoroborates K[RFBF3] (RF = CF3, C2F5, n-
C3F7, and n-C4F9)[38,40] were prepared and purified according to the re-
ported procedures.
For anodic stability, as shown in Figure 9 and Table 3, all
the QA salts are obviously more resistant toward oxidation
than [C2mmi][CF3BF3]. This result suggests that the anodic
stability of the former is determined by the involved anions,
while that of the latter is essentially limited by the decompo-
sition of the cation [C2mmi]+. This result is in keeping our
previous conclusion that both the cathodic and anodic stabil-
1H, 19F, and 11B NMR spectra were recorded on a JEOL JNMAL400
spectrometer operating at 399.65, 376.05, and 128.15 MHz, respectively.
[D4]Methanol was used the as solvent for the halide salts, and [D6]ace-
tone for the salts based on [RFBF3]ꢁ, [BF4]ꢁ, and TFSIꢁ. Chemical shift
1
values are reported relative to the TMS internal reference for H, and ex-
ternal references CCl3F in [D6]acetone for 19F (dF = 0.00) and BF3·Et2O
Chem. Eur. J. 2005, 11, 752 – 766
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
761

H. Matsumoto et al.
in CDCl3 for 11B (dB = 0.00). FAB-MS were measured on a JEOL JMS-
HX110/110A spectrometer. Elemental analysis (C, H, and N) was per-
formed at the Center for Organic Elemental Microanalysis of Kyoto Uni-
versity.
The water content in the liquid salts was determined by Karl-Fischer ti-
tration (Mitsubishi Chem., model CA-07).
0.51 mol amine). The reaction mixture was stirred at room temperature
for three days. The excess amine and water was removed by evaporation
at 708C under reduced pressure. The residual solid was washed with di-
ethyl ether (2ꢁ25 mL). Recrystallization twice from acetone/2-butanol
(75:25) gave N111.1O2Br (50.5 g, yield 85%) as a white solid. Tm = 1498C;
Td = 2538C; 1H NMR: d = 3.21 (s, 3ꢁ3H), 3.40 (s, 3H), 3.62 (s, 2H),
3.82 ppm (s, 2H).
The amounts of residual halide and potassium ions in the liquid salts
were approximately estimated by a fluorescence X-ray spectrometer
(JEOL, model JSX-3201) and revealed to be less than 30 ppm.
N-Ethyl-N,N-dimethyl-N-(2-methoxyethyl)ammonium
bromide
(N112.1O2Br): The same procedure was followed as described for N1223Br,
except ethyldimethylamine (22.7 g, 0.31 mmol) and 2-bromoethyl methyl
ether (41.7 g, 0.30 mol) were used. Recrystallization twice from acetone/
Phase transition analysis: The calorimetric measurements were per-
formed with a differential scanning calorimeter (Perkin Elmer Pyris1
DSC equipped with a liquid nitrogen cryostatic cooling) in the tempera-
ture range ꢁ1508C to a predetermined temperature. An average sample
weight of 5–10 mg was sealed in an aluminum pan, and then heated and
cooled at a rate of 108Cminꢁ1 under a flow of helium. The glass transi-
tion temperature (Tg, onset of the heat capacity change), crystallization
temperature (Tc, onset of the exothermic peak), solid–solid transition
temperature (Ts-s, onset of the endothermic peak), and melting point (Tm,
onset of the endothermic peak) were taken during heating in the second
heating–cooling cycle.
2-butanol (85:15) gave N112.1O2Br (50 g, yield 78%) as a white solid. Tm
=
648C; Td = 2498C; 1H NMR: d = 1.37 (t, J = 7.2 Hz, 3H), 3.14 (s, 2ꢁ
3H), 3.39 (s, 3H), 3.50 (q, 2ꢁ2H), 3.58 (t, J = 4.8 Hz, 2H), 3.81 ppm (s,
2H).
N,N-Diethyl-N-methyl-N-(2-methoxyethyl)ammonium
bromide
(N122.1O2Br): The same procedure was followed as described for N1223Br,
except diethylmethylamine (27.0 g, 0.31 mol) and 2-bromoethyl methyl
ether (41.7 g, 0.30 mol) were used. Recrystallization twice from acetone
gave N122.1O2Br (52 g, yield 76%) as a white solid. Tm = 858C; Td
=
2368C; 1H NMR: d = 1.35 (t, J = 7.2 Hz, 2ꢁ3H), 3.08 (s, 3H), 3.39 (s,
3H), 3.48 (q, J = 7.1 Hz, 2ꢁ2H), 3.58 (t, J = 4.8 Hz, 2H), 3.82 ppm (s,
2H).
Thermal gravimetric analysis (TGA): TGA was performed on a thermal
analysis system (Seiko Instruments, TG/DTA6200). An average sample
weight of 5 mg was placed in a platinum pan and heated at 108Cminꢁ1
from ꢀ408C to 6008C under a flow of nitrogen. The onset of decomposi-
tion was defined as the decomposition temperature (Td).
N,N,N-Triethyl-N-(2-methoxyethyl)ammonium bromide (N222.1O2Br): The
same procedure was followed as described for N1223Br, except triethyl-
amine (31.4 g, 0.31 mol) and 2-bromoethyl methyl ether (41.7 g, 0.30 mol)
were used. Recrystallization twice from acetone/tetrahydrofuran (70:30)
gave N222.1O2Br (yield 80%) as a white solid. Tm = 1068C; Td = 2248C;
1H NMR: d = 1.30 (t, J = 7.4 Hz, 3ꢁ3H), 3.38 (s, 3H), 3.43 (q, J =
7.2 Hz, 3ꢁ2H), 3.51 (t, J = 4.8 Hz, 2H), 3.77 ppm (s, 2H).
Viscosity: The viscosity was measured with a spindle viscometer equip-
ped with a thermostated water bath (Brookfield, model DV-III+, cone
spindle: CP-52) at 258C. The measurements have a precision of ꢃ2%.
Density: The density was the mean of three measurement values, which
were approximately determined by measuring the weight of three sepa-
rate samples (1.0 mL) at 258C. The relative standard uncertainty of the
mean density was less than 5%.
Procedure for aqueous solutions of perfluoroalkyltrifluoroborate acid
(H[RFBF3]) and quaternary ammonium hydroxide ([QA][OH]):
Aqueous H[RFBF3]: Aqueous H[RFBF3] was prepared by passing the cor-
responding K[RFBF3] (RF = CF3, C2F5, n-C3F7, n-C4F9) through a column
filled in acidic ion-exchange resin, as described in the literature.[38] The
concentration of the aqueous H[RFBF3], determined by acid–base titra-
tion, was ꢀ0.2 moldmꢁ3. The resultant solution was collected in a PFA
bottle, and stored in a refrigerator until required.
Specific conductivity: The ionic conductivity (k) of the neat liquid salts
was measured by a conductivity meter (Radiometer Analytical, model
CDM230) in a sealed conductivity cell at 258C.
Electrochemical stability: Linear sweep voltammetry (LSV) was per-
formed with an automatic polarization system (ALS model600) in an
argon-filled glove box (O2 and water <5 ppm), with a 5 mL beaker-type
three-electrode cell equipped with a glassy carbon electrode (surface
area: 7.85ꢁ10ꢁ3 cmꢁ2), a Pt wire counterelectrode, and an I3ꢁ/Iꢁ reference
electrode consisting of Pt wire/0.015 moldmꢁ3 I2 + 0.060 moldmꢁ3 [(n-
C3H7)4N]I in [C2mim]TFSI ([C2mim]+ 1-ethyl-3-methylimidazolium). The
potential was referred to ferrocene (Fc)/ferrocenium (Fc+) redox couple
in each salt. The data for each salt were collected in the first cathodic
and anodic scan at 258C.
Aqueous [QA][OH]: The same procedure was followed for aqueous
H[RFBF3] as described above, except that the basic anion-exchange resin
and the corresponding quaternary ammonium bromide ([QA][Br]) were
used. The concentration of the aqueous [QA][OH] was ꢀ0.2 moldmꢁ3
.
The solution was collected in a plastic bottle, and stored in a refrigerator
until required.
Procedure for quaternary ammonium salts with perfluoroalkyltrifluoro-
borate ([QA][RFBF3]): The [QA][RFBF3] salts were prepared in
30 mmol scale as follows: aqueous [QA][OH] was neutralized with equi-
molar aqueous H[RFBF3] (RF = CF3, C2F5, n-C3F7, n-C4F9) in a PFA
flask until the pH was ꢀ6–7. The mixture was concentrated to ꢀ15 mL
at 408C by evaporation under reduced pressure. The liquid salt precipi-
tated in the bottom was separated directly, washed with deionized water
(5 mL) and toluene (10 mL) in turn, and dried at 708C for 24 h under
vacuum (0.03 Torr) to afford a colorless or pale yellow liquid (water con-
tent <50 ppm). The solid salt was collected by filtration, dried under
vacuum, and recrystallized from 2-butanol to afford a white product.
Procedure for quaternary ammonium bromides ([QA][Br]):
N,N-Diethyl-N-methyl-N-(n-propyl)ammonium bromide (N1223Br): This
compound was prepared in 0.30-mol scale in a 300 mL autoclave (Taiatsu
Techno) by reaction of 1-bromopropane (37.0 g, 0.30 mol) with a small
excess of ethyldimethylamine (27.0 g, 0.31 mol) in anhydrous acetone at
1008C and ꢀ6mPa for 12 h. The cooled red suspension was then poured
into diethyl ether (150 mL). After filtration, the obtained solid was
washed with diethyl ether (2ꢁ25 mL), and recrystallized twice from ace-
tone to afford N1223Br (50 g, yield 80%) as a white solid. Tm = 2708C
(decomp); 1H NMR: d = 1.02 (t, J = 7.2 Hz, 3H), 1.33 (t, J = 7.2 Hz,
2ꢁ3H), 1.78 (m, 2H), 3.00 (s, 3H), 3.23 (m, 2H), 3.38 ppm (q, J
7.3 Hz, 2ꢁ2H).
=
N,N-Diethyl-N-methyl-N-(n-propyl)ammonium trifluoromethyltrifluoro-
borate (N1223[CF3BF3]): Yield: 75%, white solid; 1H NMR: d = 1.01 (t, J
= 7.4 Hz, 3H), 1.42 (t, J = 7.4 Hz, 2ꢁ3H), 1.87 (m, 2H). 3.13 (s, 3H),
3.38 (m, 2H), 3.53 ppm (q, J = 7.3 Hz, 2ꢁ2H); 19F NMR: d = ꢁ74.6 (q,
2J(B,F) = 32.6 Hz, 3F; CF3), ꢁ155.5 ppm (q, 1J(B,F) = 39.6 Hz, 3F;
N-(n-Butyl)-N,N-diethyl-N-methylammonium bromide (N1224Br): The
same procedure was followed as described for N1223Br, except for the use
of 1-bromobutane (41.1 g, 0.30 mol) instead of 1-bromopropane. Recrys-
tallization twice from acetone/tetrahydrofuran (65:35) gave N1224Br (54 g,
yield 80%) as a white solid. Tm = 1798C; Td = 2368C; 1H NMR: d =
1.02 (t, J = 7.4 Hz, 3H), 1.33 (t, J = 7.4 Hz, 2ꢁ3H), 1.40 (m, 2H), 1.71
(m, 2H), 2.99 (s, 3H), 3.25 (m, 2H), 3.36 ppm (m, 2ꢁ2H).
BF3); 11B NMR: d
=
ꢁ0.49 ppm (qq, 1J(B,F)
= =
39.0 Hz, 2J(B,F)
32.3 Hz); FAB-MS: m/z (%): 130 (100) [N1223]+, 137 (100) [CF3BF3]ꢁ; el-
emental analysis calcd (%) for C9H20BF6N (267.06): C 40.5, H 7.6, N 5.2;
found: C 40.3, H 7.4, N 5.3.
N-(n-Butyl)-N,N-diethyl-N-methylammonium trifluoromethyltrifluorobo-
rate (N1224[CF3BF3]): Yield: 77%, colorless liquid; H NMR: d = 0.99 (t,
N,N,N-Trimethyl-N-(2-methoxyethyl)ammonium bromide (N111.1O2Br): 2-
Bromoethyl methyl ether (56.0 g, 0.40 mol) was added at room tempera-
1
ture to
a
solution of 30% aqueous trimethylamine (100 g solution,
J = 7.2 Hz, 3H), 1.40 (t, J = 7.2 Hz, 2ꢁ3H), 1.44 (t, J = 7.4 Hz, 2H),
762
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 752 – 766

Hydrophobic Ionic Liquids
FULL PAPER
1.81 (m, 2H), 3.12 (s, 3H), 3.38 (m, 2H), 3.50 ppm (q, J = 7.3 Hz, 2ꢁ
2H); 19F NMR: d = ꢁ74.6 (q, 2J(B,F) = 31.6 Hz, 3F; CF3), ꢁ155.5 ppm
N-Ethyl-N,N-dimethyl-N-(2-methoxyethyl)ammonium pentafluoroethyl-
trifluoroborate (N112.1O2 [C2F5BF3]): Yield: 88%, pale yellow liquid;
1H NMR: d = 1.45 (t, J = 7.2 Hz, 3H), 3.28 (s, 2ꢁ3H), 3.39 (s, 3H),
3.64 (q, J = 7.2 Hz, 2H), 3.71 (t, J = 4.8 Hz, 2H), 3.92 ppm (s, 2H);
19F NMR: d = ꢁ83.0 (s, 3F; CF3), ꢁ135.8 (q, 2J(B,F) = 19.3 Hz, 2F;
1
1
(q, J(B,F) = 38.6 Hz, 3F; BF3); 11B NMR: d = ꢁ0.48 ppm (qq, J(B,F)
= 39.9 Hz, 2J(B,F) = 32.3 Hz); FAB-MS: m/z (%): 144 (100) [N1224]+,
137 (100) [CF3BF3]ꢁ; elemental analysis calcd (%) for C10H22BF6N
(281.09): C 42.7, H 7.9, N 5.0; found: C 42.7, H 7.6, N 5.0.
CF2), ꢁ152.7 ppm (q, 1J(B,F)
= =
40.7 Hz, 3F; BF3); 11B NMR: d
1
2
0.21 ppm (qt, J(B,F) = 40.9 Hz, J(B,F) = 19.1 Hz); FAB-MS: m/z (%):
132 (100) [N112.1O2]+, 187 (100) [C2F5BF3]ꢁ; elemental analysis calcd (%)
for C9H18BF8NO (319.04): C 33.9, H 5.7, N 4.4; found: C 33.7, H 5.6, N
4.3.
N,N,N-Trimethyl-N-(2-methoxyethyl)ammonium trifluoromethyltrifluoro-
borate (N111.1O2[CF3BF3]): Yield: 80%, white solid; 1H NMR: d = 3.37
(s, 3ꢁ3H), 3.40 (s, 3H), 3.78 (s, 2H), 3.95 ppm (s, 2H); 19F NMR: d =
ꢁ74.6 (q, 2J(B,F)
=
31.6 Hz, 3F; CF3), ꢁ155.5 ppm (q, 1J(B,F)
=
38.6 Hz, 3F; BF3); 11B NMR: d = ꢁ0.52 ppm (qq, 1J(B,F) = 39.9 Hz,
2J(B,F) = 32.3 Hz); FAB-MS: m/z (%): 118 (100) [N111.1O2]+, 137 (100)
[CF3BF3]ꢁ; elemental analysis calcd (%) for C7H16BF6NO (255.01): C
33.0, H 6.3, N 5.5; found: C 32.7, H 6.3, N 5.6.
N,N-Diethyl-N-methyl-N-(2-methoxyethyl)ammonium pentafluoroethyl-
trifluoroborate (N122.1O2[n-C2F5BF3]): Yield: 90%, colorless liquid;
1H NMR: d = 1.41 (t, J = 7.2 Hz, 2ꢁ3H), 3.19 (s, 3H), 3.39 (s, 3H),
3.59 (q, J = 7.2 Hz, 2ꢁ2H), 3.67 (t, J = 4.8 Hz, 2H), 3.91 ppm (s, 2H);
19F NMR: d = ꢁ83.0 (s, 3F; CF3), ꢁ135.8 (q, 2J(B,F) = 20.3 Hz, 2F;
N-ethyl-N,N-Dimethyl-N-(2-methoxyethyl)ammonium trifluoromethyltri-
fluoroborate (N112.1O2 [CF3BF3]): Yield: 66%, pale yellow liquid;
1H NMR: d = 1.44 (t, J = 7.2 Hz, 3H), 3.26 (s, 2ꢁ3H), 3.39 (s, 3H),
3.62 (q, J = 7.2 Hz, 2H), 3.69 (t, J = 4.8 Hz, 2H), 3.91 ppm (s, 2H);
19F NMR: d = ꢁ74.7 (q, 2J(B,F) = 31.6 Hz, 3F; CF3), ꢁ155.0 ppm (q,
1J(B,F) = 39.6 Hz, 3F; BF3); 11B NMR: d = ꢁ0.53 ppm (qq, 1J(B,F) =
39.0 Hz, 2J(B,F) = 32.3 Hz); FAB-MS: m/z (%): 132 (100) [N112.1O2]+,
137 (100) [CF3BF3]ꢁ; elemental analysis calcd (%) for C8H18BF6NO
(269.03): C 35.7, H 6.7, N 5.2; found: C 35.3, H 6.7, N 5.3.
CF2), ꢁ152.8 ppm (q, 1J(B,F)
= =
40.7 Hz, 3F; BF3); 11B NMR: d
1
2
0.15 ppm (qt, J(B,F) = 40.8 Hz, J(B,F) = 19.1 Hz); FAB-MS: m/z (%):
146 (100) [N122.1O2]+, 187 (100) [C2F5BF3]ꢁ; elemental analysis calcd (%)
for C10H20BF8NO (333.07): C 36.1, H 6.1, N 4.2; found: C 35.8, H 5.9, N
4.1.
N,N,N-Triethyl-N-(2-methoxyethyl)ammonium pentafluoroethyltrifluoro-
borate (N222.1O2[C2F5BF3]): Yield: 90%, colorless liquid; 1H NMR: d =
1.37 (t, J = 7.2 Hz, 3ꢁ3H), 3.38 (s, 3H), 3.56 (q, J = 7.2 Hz, 3ꢁ2H),
3.63 (t, J = 4.8 Hz, 2H), 3.87 ppm (s, 2H); 19F NMR: d = ꢁ83.0 (s, 3F;
CF3), ꢁ135.8 (q, 2J(B,F) = 19.4 Hz, 2F; CF2), ꢁ153.0 ppm (q, 1J(B,F) =
N,N-Diethyl-N-methyl-N-(2-methoxyethyl)ammonium trifluoromethyltri-
fluoroborate (N122.1O2[CF3BF3]): Yield: 70%, colorless liquid; 1H NMR: d
= 1.41 (t, J = 7.2 Hz, 2ꢁ3H), 3.20 (s, 3H), 3.38 (s, 3H), 3.59 (q, J =
7.3 Hz, 2ꢁ2H), 3.68 (t, J = 4.8 Hz, 2H), 3.90 (s, 2H); 19F NMR: d =
39.7 Hz, 3F; BF3); 11B NMR: d
= = 40.9 Hz,
0.19 ppm (qt, 1J(B,F)
2J(B,F) = 19.1 Hz); FAB-MS: m/z (%): 160 (100) [N222.1O2]+, 187 (100)
[C2F5BF3]ꢁ; elemental analysis calcd (%) for C11H22BF8NO (347.1): C
38.1, H 6.4, N 4.0; found: C 38.1, H 6.4, N 4.0.
ꢁ74.7 (q, 2J(B,F)
=
32.6 Hz, 3F; CF3), ꢁ155.3 ppm (q, 1J(B,F)
=
38.7 Hz, 3F; BF3); 11B NMR: d = ꢁ0.49 ppm (qq, 1J(B,F) = 38.9 Hz,
2J(B,F) = 32.3 Hz); FAB-MS: m/z (%): 146 (100) [N122.1O2]+, 137 (100)
[CF3BF3]ꢁ; elemental analysis calcd (%) for C9H20BF6NO (283.06): C
38.2, H 7.1, N 5.0; found: C 38.0, H 7.1, N 5.1.
N,N-Diethyl-N-methyl-N-(n-propyl)ammonium (heptafluoro-n-propyl)tri-
fluoroborate (N1223[n-C3F7BF3]): Yield: 88%, white solid; 1H NMR: d =
1.01 (t, J = 7.4 Hz, 3H), 1.41 (t, J = 7.2 Hz, 2ꢁ3H), 1.88 (m, 2H), 3.15
(s, 3H), 3.38 (m, 2H), 3.53 (q, J = 7.2 Hz, 2ꢁ2H); 19F NMR: d = ꢁ80.5
(s, 3F; CF3), ꢁ127.6 (s, 2F; BCCF2-), ꢁ133.8 (s, 2F; CF2B), ꢁ152.4 ppm
(q, 1J(B,F) = 40.7 Hz, 3F; BF3); 11B NMR: d = 0.25 ppm (qt, 1J(B,F) =
40.3 Hz, 2J(B,F) = 19.9 Hz); FAB-MS: m/z (%): 130 (100) [N1223]+, 237
(100) [C3F7BF3]ꢁ; elemental analysis calcd (%) for C11H20BF10N (367.08):
C 36.0, H 5.5, N 3.8; found: C 35.8, H 5.2, N 3.9.
N,N,N-Triethyl-N-(2-methoxyethyl)ammonium trifluoromethyltrifluoro-
borate (N222.1O2[CF3BF3]): Yield: 85%, white solid; 1H NMR: d = 1.38
(t, J = 7.2 Hz, 3ꢁ3H), 3.38 (s, 3H), 3.54 (q, J = 7.2 Hz, 3ꢁ2H), 3.63 (t,
J = 4.8 Hz, 2H), 3.87 ppm (s, 2H); 19F NMR: d = ꢁ74.6 (q, 2J(B,F) =
31.5 Hz, 3F; CF3), ꢁ155.4 ppm (q, 1J(B,F)
= 39.6 Hz, 3F; BF3);
11B NMR: d = ꢁ0.48 ppm (qq, 1J(B,F) = 39.0 Hz, 2J(B,F) = 32.3 Hz);
FAB-MS: m/z (%): 160 (100) [N222.1O2]+, 137 (100) [CF3BF3]ꢁ; elemental
analysis calcd (%) for C10H22BF6NO (297.09): C 40.4, H 7.5, N 4.7;
found: C 40.5, H 7.5, N 4.7.
N-(n-Butyl)-N,N-diethyl-N-methylammonium (heptafluoro-N-propyl)tri-
fluoroborate (N1224[n-C3F7BF3]): Yield: 88%, white solid; 1H NMR: d =
0.99 (t, J = 7.2 Hz, 3H), 1.40 (t, J = 7.4 Hz, 2ꢁ3H), 1.46 (t, J = 7.4 Hz,
2H), 1.82 (m, 2H), 3.12 (s, 3H), 3.38 (m, 2H), 3.50 ppm (q, J = 7.2 Hz,
N,N-Diethyl-N-methyl-N-(n-propyl)ammonium pentafluoroethyltrifluoro-
borate (N1223[C2F5BF3]): Yield: 90%, white solid; 1H NMR: d = 1.02 (t,
J = 7.4 Hz, 3H), 1.42 (t, J = 7.2 Hz, 2ꢁ3H), 1.86 (m, 2H), 3.15 (s, 3H),
3.38 (t, 2H), 3.53 ppm (q, J = 7.2 Hz, 2ꢁ2H); 19F NMR: d = ꢁ83.1 (s,
2ꢁ2H); 19F NMR: d
=
ꢁ80.4 (s, 3F; CF3), ꢁ127.5 (s, 2F; BCCF2-),
ꢁ133.7 (s, 2F; CF2B), ꢁ152.3 ppm (q, 1J(B,F)
= 40.0 Hz, 3F; BF3);
11B NMR: d = 0.25 ppm (qt, 1J(B,F) = 40.6 Hz, 2J(B,F) = 19.1 Hz);
FAB-MS: m/z (%): 144 (100) [N1224]+, 237 (100) [C3F7BF3]ꢁ; elemental
analysis calcd (%) for C12H22BF10N (381.11): C 37.8, H 5.8, N 3.7; found:
C 37.7, H 5.6, N 3.9.
3F; CF3), ꢁ135.9 (q, 2J(B,F)
=
19.3 Hz, 2F; CF2), ꢁ153.1 ppm (q,
1J(B,F) = 41.0 Hz, 3F; BF3); 11B NMR: d = 0.21 ppm (qt, 1J(B,F) =
2
40.6 Hz, J(B,F) = 19.86 Hz); FAB-MS: m/z (%): 130 (100) [N1223]+, 187
(100) [C2F5BF3]ꢁ; elemental analysis calcd (%) for C10H20BF8N (317.07):
C 37.9, H 6.4, N 4.4; Found: C 37.8, H 6.1, N 4.5.
N,N,N-Trimethyl-N-(2-methoxyethyl)ammonium (heptafluoro-n-propyl)-
trifluoroborate (N111.1O2[n-C3F7BF3]): Yield: 88%, colorless liquid;
1H NMR: d = 3.36 (s, 3ꢁ3H), 3.40 (s, 3H), 3.77 (s, 2H), 3.95 ppm (s,
2H); 19F NMR: d = ꢁ80.5 (s, 3F; CF3), ꢁ127.6 (s, 2F; BCCF2-), ꢁ133.7
N-(n-Butyl)-N,N-diethyl-N-methylammonium pentafluoroethyltrifluoro-
1
borate (N1224[C2F5BF3]): Yield: 88%, colorless liquid; H NMR: d = 0.99
1
(s, 2F; CF2B), ꢁ152.3 ppm (q, J(B,F) = 39.6 Hz, 3F; BF3); 11B NMR: d
(t, J = 7.2 Hz, 3H), 1.40 (t, J = 7.4 Hz, 2ꢁ3H), 1.46 (t, J = 7.2 Hz,
2H), 1.81 (m, 2H), 3.12 (s, 3H), 3.38 (m, 2H), 3.50 ppm (q, J = 7.2 Hz,
= 0.25 ppm (qt, 1J(B,F) = 40.6 Hz, 2J(B,F) = 19.0 Hz); FAB-MS: m/z
(%): 118 (100) [N111.1O2]+, 237 (100) [C3F7BF3]ꢁ; elemental analysis calcd
(%) for C9H16BF10NO (355.03): C 30.5, H 4.5, N 4.0; found: C 30.4, H
4.5, N 4.2.
2
2ꢁ2H); 19F NMR: d = ꢁ83.0 (s, 3F; CF3), ꢁ135.8 (q, J(B,F) = 19.9 Hz,
2F; CF2), ꢁ152.8 ppm (q, 1J(B,F) = 39.6 Hz, 3F; BF3); 11B NMR: d =
1
2
0.21 ppm (qt, J(B,F) = 40.6 Hz, J(B,F) = 19.9 Hz); FAB-MS: m/z (%):
144 (100) [N1224]+, 187 (100) [C2F5BF3]ꢁ; elemental analysis calcd (%) for
C11H22BF8N (331.1): C 39.9, H, 6.7, N 4.2; found: C 39.7, H, 6.4, N 4.2.
N-ethyl-N,N-dimethyl-N-(2-methoxyethyl)ammonium (heptafluoro-n-pro-
pyl)trifluoroborate (N112.1O2[n-C3F7BF3]): Yield: 90%, pale yellow liquid;
1H NMR: d = 1.45 (t, J = 7.1 Hz, 3H), 3.28 (s, 2ꢁ3H), 3.39 (s, 3H),
3.65 (q, J = 7.1 Hz, 2H), 3.72 (t, J = 4.6 Hz, 2H), 3.93 ppm (s, 2H);
19F NMR: d = ꢁ80.5 (s, 3F; CF3), ꢁ127.5 (s, 2F; BCCF2-), ꢁ133.6 (s,
2F; CF2B), ꢁ152.3 ppm (q, 1J(B,F) = 40.6 Hz, 3F; BF3); 11B NMR: d =
N,N,N-Trimethyl-N-(2-methoxyethyl)ammonium pentafluoroethyltrifluor-
oborate (N111.1O2[C2F5BF3]): Yield: 86%, white solid; 1H NMR: d = 3.37
(s, 3ꢁ3H), 3.40 (s, 3H), 3.76 (s, 2H), 3.94 ppm (s, 2H); 19F NMR: d =
ꢁ83.0 (s, 3F; CF3), ꢁ135.8 (q, 2J(B,F) = 19.3 Hz, 2F; CF2), ꢁ153.0 ppm
(q, 1J(B,F) = 39.6 Hz, 3F; BF3); 11B NMR: d = 0.22 ppm (qt, 1J(B,F) =
40.9 Hz, 2J(B,F) = 19.1 Hz); FAB-MS: m/z (%): 118 (100) [N111.1O2]+,
187 (100) [C2F5BF3]ꢁ; elemental analysis calcd (%) for C8H16BF8NO
(305.02): C 31.5, H 5.3, N 4.6; found: C 31.2, H 5.2, N 4.6.
1
2
0.26 ppm (qt, J(B,F) = 40.6 Hz, J(B,F) = 19.0 Hz); FAB-MS: m/z (%):
132 (100) [N112.1O2]+, 237 (100) [C3F7BF3]ꢁ; elemental analysis calcd (%)
for C10H18BF10NO (369.05): C 32.5, H 4.9, N 3.8; found: C 32.3, H 4.7, N
4.1.
Chem. Eur. J. 2005, 11, 752 – 766
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
763

H. Matsumoto et al.
N,N-Diethyl-N-methyl-N-(2-methoxyethyl)ammonium
(heptafluoro-n-
N,N,N-Triethyl-N-(2-methoxyethyl)ammonium
(nonafluoro-n-butyl)tri-
propyl)trifluoroborate (N122.1O2[n-C3F7BF3]): Yield: 90%, colorless liquid;
1H NMR: d = 1.41 (t, J = 7.3 Hz, 2ꢁ3H), 3.20 (s, 3H), 3.38 (s, 3H),
3.59 (q, J = 7.2 Hz, 2ꢁ2H), 3.67 (t, J = 4.8 Hz, 2H), 3.91 ppm (s, 2H);
19F NMR: d = ꢁ80.3 (s, 3F; CF3), ꢁ127.5 (s, 2F; BCCF2-), ꢁ133.7 (s,
2F; CF2B), ꢁ152.3 ppm (q, 1J(B,F) = 38.7 Hz, 3F; BF3); 11B NMR: d =
fluoroborate (N222.1O2[n-C4F9BF3]): Yield: 92%, colorless liquid;
1H NMR: d = 1.36 (t, J = 7.2 Hz, 3ꢁ3H), 3.36 (s, 3H), 3.54 (q, J =
7.2 Hz, 3ꢁ2H), 3.63 (t, J = 4.6 Hz, 2H), 3.86 ppm (s, 2H); 19F NMR: d
= ꢁ80.9 (s, 3F; CF3), ꢁ123.8 (s, 2F; BCCCF2-), ꢁ125.8 (s, 2F; BCCF2-),
ꢁ133.0 (s, 2F; CF2B), ꢁ152.3 ppm (q, 1J(B,F)
= 40.6 Hz, 3F; BF3);
1
2
11B NMR: d = 0.23 ppm (qt, 1J(B,F) = 40.6 Hz, 2J(B,F) = 19.1 Hz);
FAB-MS: m/z (%): 160 (100) [N222.1O2]+, 287 (100) [C4F9BF3]ꢁ; elemental
analysis calcd (%) for C13H22BF12NO (447.12): C 34.9, H 5.0, N 3.1;
found: C 34.9, H 4.8, N 3.2.
0.25 ppm (qt, J(B,F) = 40.6 Hz, J(B,F) = 19.0 Hz); FAB-MS: m/z (%):
146 (100) [N122.1O2]+, 237 (100) [C3F7BF3]ꢁ; elemental analysis calcd (%)
for C11H20BF10NO (383.08): C 34.5, H 5.3, N 3.7; found: C 34.8, H 5.4, N
3.6.
N,N,N-Triethyl-N-(2-methoxyethyl)ammonium (heptafluoro-n-propyl)tri-
fluoroborate (N222.1O2[n-C3F7BF3]): Yield: 90%, colorless liquid;
1H NMR: d = 1.38 (t, J = 7.2 Hz, 3ꢁ3H), 3.38 (s, 3H), 3.56 (q, J =
7.3 Hz, 3ꢁ2H), 3.64 (t, J = 4.6 Hz, 2H), 3.88 ppm (s, 2H); 19F NMR: d
= ꢁ80.4 (s, 3F; CF3), ꢁ127.5 (s, 2F; BCCF2-), ꢁ133.7 (s, 2F; CF2B),
ꢁ152.5 ppm (q, 1J(B,F) = 40.7 Hz, 3F; BF3); 11B NMR: d = 0.24 ppm
Procedure for quaternary ammonium salts with tetrafluoroborate ([QA]
[BF4]): The [QA][BF4] salts were prepared in a 50 mmol scale as follows:
aqueous [QA][OH] was neutralized with equimolar aqueous H[BF4] in a
PFA flask until the pH was ꢀ6–7. The solution was evaporated at 30–
408C under reduced pressure to give a viscous liquid or solid. If a viscous
liquid was obtained, it was dried at 408C for 48 h at room temperature
under vacuum (0.03 Torr), then dissolved in anhydrous acetone (60 mL),
and filtered through a PTFE membrane filter (0.2 mm). The collected so-
lution was evaporated and dried at 808C for 48 h under vacuum
(0.03 Torr), yielding a pale yellow or colorless liquid (water content:
ꢀ400–600 ppm). If a solid product was obtained, it was recrystallization
from 2-butanol/methanol to give a white solid.
1
2
(qt, J(B,F) = 40.3 Hz, J(B,F) = 19.1 Hz); FAB-MS: m/z (%): 160 (100)
[N222.1O2]+, 237 (100) [C3F7BF3]ꢁ; elemental analysis calcd (%) for
C12H22BF10NO (397.11): C 36.3, H 5.6, N 3.5; found: C 36.3, H 5.5, N 3.6.
N,N-Diethyl-N-methyl-N-(n-propyl)ammonium (nonafluoro-n-butyl)tri-
fluoroborate (N1223[n-C4F9BF3]): Yield: 90%, white solid; 1H NMR: d =
1.01 (t, J = 7.4 Hz, 3H), 1.42 (t, J = 7.2 Hz, 2ꢁ3H), 1.86 (m, 2H). 3.15
(s, 3H), 3.38 (m, 2H), 3.53 ppm (q, J = 7.2 Hz, 2ꢁ2H); 19F NMR: d =
ꢁ80.9 (s, 3F; CF3), ꢁ123.9 (s, 2F; BCCCF2-), ꢁ125.9 (s, 2F; BCCF2-),
N,N-Diethyl-N-methyl-N-(n-propyl)ammonium
tetrafluoroborate
1
(N1223[BF4]): Yield: 88%, white solid; H NMR: d = 1.01 (t, J = 7.2 Hz,
3H), 1.39 (t, J = 7.2 Hz, 2ꢁ3H), 1.86 (m, 2H), 3.13 (s, 3H), 3.37 (m,
2H), 3.53 ppm (q, J = 7.3 Hz, 2ꢁ2H); 19F NMR: d = ꢁ150.0 ppm (s);
11B NMR: d = ꢁ0.23 ppm (s); FAB-MS: m/z (%): 130 (100) [N1223]+, 87
(100) [BF4]ꢁ; elemental analysis calcd (%) for C8H20BF4N (217.06): C
44.3, H 9.3, N 6.5; found: C 43.8, H 9.1, N 6.4.
ꢁ133.1 (s, 2F; CF2B), ꢁ152.2 ppm (q, 1J(B,F)
= 40.6 Hz, 3F; BF3);
11B NMR: d = 0.25 ppm (qt, 1J(B,F) = 40.3 Hz, 2J(B,F) = 19.0 Hz);
FAB-MS: m/z (%): 130 (100) [N1223]+, 287 (100) [C4F9BF3]ꢁ; elemental
analysis calcd (%) for C12H20BF12N (417.09): C 34.6, H 4.8, N 3.4; found:
C 34.3, H 4.7, N 3.4.
N-(n-Butyl)-N,N-diethyl-N-methylammonium
(nonafluoro-n-butyl)tri-
N-(n-Butyl)-N,N-diethyl-N-methylammonium
tetrafluoroborate
fluoroborate (N1224[n-C4F9BF3]): Yield: 92%, white solid; 1H NMR: d =
0.99 (t, J = 7.4 Hz, 3H), 1.40 (t, J = 7.4 Hz, 2ꢁ3H), 1.46 (t, J = 7.2 Hz,
2H), 1.82 (m, 2H), 3.14 (s, 3H), 3.41 (m, 2H), 3.52 ppm (q, J = 7.2 Hz,
2ꢁ2H); 19F NMR: d = ꢁ80.9 (s, 3F; CF3), ꢁ123.8 (s, 2F; BCCCF2-),
1
(N1224[BF4]): Yield: 91%, white solid; H NMR: d = 0.99 (t, J = 7.4 Hz,
3H), 1.41 (t, J = 7.2 Hz, 2ꢁ3H), 1.47 (t, J = 7.4 Hz, 2H), 1.82 (m, 2H),
3.15 (s, 3H), 3.41 (m, 2H), 3.53 ppm (q, J = 7.3 Hz, 2ꢁ2H); 19F NMR: d
= ꢁ150.2 ppm (s); 11B NMR: d = 0.21 ppm (s); FAB-MS: m/z (%): 144
(100) [N1223]+, 87 (100) [BF4]ꢁ; elemental analysis calcd (%) for
C9H22BF4N (231.09): C 46.8, H 9.6, N 6.1; found: C 46.5, H 9.5, N 6.0.
1
ꢁ125.8 (s, 2F; BCCF2-), ꢁ133.0 (s, 2F; CF2B), ꢁ152.2 ppm (q, J(B,F) =
38.6 Hz, 3F; BF3); 11B NMR: d
= = 40.3 Hz,
0.25 ppm (qt, 1J(B,F)
2J(B,F) = 20.0 Hz); FAB-MS: m/z (%): 144 (100) [N1224]+, 287 (100)
[C4F9BF3]ꢁ; elemental analysis calcd (%) for C13H22BF12N (431.12): C
36.2, H 5.1, N 3.3; found: C 36.0, H 5.0, N 3.5.
N,N,N-Trimethyl-N-(2-methoxyethyl)ammonium
tetrafluoroborate
(N111.1O2[BF4]): Yield: 89%, white solid; 1H NMR: d = 3.34 (s, 3ꢁ3H),
3.40 (s, 3H), 3.74 (s, 2H), 3.92 ppm (s, 2H); 19F NMR: d = ꢁ149.3 ppm
N,N,N-Trimethyl-N-(2-methoxyethyl)ammonium (nonafluoro-n-butyl)tri-
fluoroborate (N111.1O2[n-C4F9BF3]): Yield: 90%, white solid; H NMR: d
(s); 11B NMR:
d
=
ꢁ0.23 ppm (s); FAB-MS: m/z (%): 118 (100)
1
[N111.1O2]+, 87 (100) [BF4]ꢁ; elemental analysis calcd (%) for
C6H16BF4NO (205.01): C 35.2, H 7.9, N 6.8; found: C 35.2, H 7.9, N 6.8.
= 3.37 (s, 3ꢁ3H), 3.40 (s, 3H), 3.80 (s, 2H), 3.96 ppm (s, 2H); 19F NMR:
d = ꢁ80.9 (s, 3F; CF3), ꢁ123.8 (s, 2F; BCCCF2-), ꢁ125.8 (s, 2F; BCCF2-
), ꢁ133.0 (s, 2F; CF2B), ꢁ152.2 ppm (q, 1J(B,F) = 38.6 Hz, 3F; BF3);
11B NMR: d = 0.25 ppm (qt, 1J(B,F) = 40.3 Hz, 2J(B,F) = 19.0 Hz);
FAB-MS: m/z (%): 118 (100) [N111.1O2]+, 287 (100) [C4F9BF3]ꢁ; elemental
analysis calcd (%) for C10H16BF12NO (405.04): C 29.7, H 4.0, N 3.5;
found: C 29.4, H 3.9, N 3.7.
N-Ethyl-N,N-dimethyl-N-(2-methoxyethyl)ammonium tetrafluoroborate
(N112.1O2[BF4]): Yield: 90%, pale yellow liquid; 1H NMR: d = 1.42 (t, J
= 7.2 Hz, 3H), 3.24 (s, 2ꢁ3H), 3.39 (s, 3H), 3.62 (q, J = 7.2 Hz, 2H),
3.69 (t, J = 4.8 Hz, 2H), 3.90 ppm (s, 2H); 19F NMR: d = ꢁ149.5 ppm
(s); 11B NMR:
d
=
ꢁ0.25 ppm (s); FAB-MS: m/z (%): 132 (100)
[N112.1O2]+, 87 (100) [BF4]ꢁ; elemental analysis calcd (%) for
C7H18BF4NO (219.03): C 38.4, H 8.3, N 6.4; found: C 38.4, H 8.8, N 6.4.
N-Ethyl-N,N-dimethyl-N-(2-methoxyethyl)ammonium (nonafluoro-n-but-
yl)trifluoroborate (N112.1O2[n-C4F9BF3]): Yield: 90%, pale yellow liquid;
1H NMR: d = 1.46 (t, J = 7.4 Hz, 3H), 3.29 (s, 2ꢁ3H), 3.39 (s, 3H),
3.65 (q, J = 7.3 Hz, 2H), 3.72 (t, J = 4.8 Hz, 2H), 3.93 ppm (s, 2H);
19F NMR: d = ꢁ80.9 (s, 3F; CF3), ꢁ123.8 (s, 2F; BCCCF2-), ꢁ125.8 (s,
2F; BCCF2-), ꢁ133.0 (s, 2F; CF2B), ꢁ152.5 ppm (q, 1J(B,F) = 38.1 Hz,
3F; BF3); 11B NMR: d = 0.25 ppm (qt, 1J(B,F) = 40.3 Hz, 2J(B,F) =
19.9 Hz); FAB-MS: m/z (%): 132 (100) [N112.1O2]+, 287 (100) [C4F9BF3]ꢁ;
elemental analysis calcd (%) for C11H18BF12NO (419.06): C 31.5, H 4.3, N
3.3; found: C 31.2, H 4.3, N 3.7.
N,N-Diethyl-N-methyl-N-(2-methoxyethyl)ammonium tetrafluoroborate
(N122.1O2[BF4]): Yield: 90%, colorless liquid; 1H NMR: d = 1.39 (t, J =
7.2 Hz, 2ꢁ3H), 3.15 (s, 3H), 3.38 (s, 3H), 3.57 (q, J = 7.3 Hz, 2ꢁ2H),
3.63 (t, J = 4.8 Hz, 2H), 3.88 ppm (s, 2H); 19F NMR: d = ꢁ149.8 ppm
(s); 11B NMR:
d
=
ꢁ0.25 ppm (s); FAB-MS: m/z (%): 146 (100)
[N122.1O2]+, 87 (100) [BF4]ꢁ; elemental analysis calcd (%) for
C8H20BF4NO (233.06): C 41.2, H 8.7, N 6.0; found: C 40.9, H 9.2, N 6.0.
N,N,N-Triethyl-N-(2-methoxyethyl)ammonium
tetrafluoroborate
1.38 (t, J
(N222.1O2[BF4]): Yield: 92%, white solid; 1H NMR: d
=
=
N,N-Diethyl-N-methyl-N-(2-methoxyethyl)ammonium (nonafluoro-n-but-
yl)trifluoroborate (N122.1O2[n-C4F9BF3]): Yield: 91%, colorless liquid;
1H NMR: d = 1.41 (m, 2ꢁ3H), 3.21 (m, 3H), 3.38 (m, 3H), 3.60 (q, J =
7.2 Hz, 2ꢁ2H), 3.67 (t, J = 4.8 Hz, 2H), 3.91 ppm (s, 2H); 19F NMR: d
= ꢁ80.9 (s, 3F; CF3), ꢁ123.8 (s, 2F; BCCCF2-), ꢁ125.8 (s, 2F; BCCF2-),
7.2 Hz, 3ꢁ3H), 3.38 (s, 3H), 3.54 (q, J = 7.2 Hz, 3ꢁ2H), 3.64 (t, J =
4.8 Hz, 2H), 3.87 ppm (s, 2H); 19F NMR:
d
=
ꢁ150.3 ppm (s);
11B NMR: d = ꢁ0.23 ppm (s); FAB-MS: m/z (%): 160 (100) [N222.1O2]+,
87 (100) [BF4]ꢁ; elemental analysis calcd (%) for C9H22BF4NO (247.09):
C 43.8, H 9.0, N 5.7; found: C 43.9, H 9.4, N 5.7.
ꢁ133.1 (s, 2F; CF2B), ꢁ152.3 ppm (q, 1J(B,F)
= 38.7 Hz, 3F; BF3);
11B NMR: d = 0.23 ppm (qt, 1J(B,F) = 40.3 Hz, 2J(B,F) = 19.0 Hz);
FAB-MS: m/z (%): 146 (100) [N122.1O2]+, 287 (100) [C4F9BF3]ꢁ; elemental
analysis calcd (%) for C12H20BF12NO (433.09): C 33.3, H 4.7, N 3.2;
found: C 33.1; H 4.6, N 3.1.
Procedure for quaternary ammonium salts with bis(trifluoromethanesul-
fonyl)imide ([QA][TFSI]): The same procedure was followed as used for
the [QA][RFBF3] salts, as described above, except commercial H[TFSI]
was used instead of aqueous H[RFBF3].
764
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 752 – 766

Hydrophobic Ionic Liquids
FULL PAPER
N,N-Diethyl-N-methyl-N-(n-propyl)ammonium bis(trifluoromethanesul-
fonyl)imide (N1223[TFSI]): Yield: 85%, colorless liquid; 1H NMR: d =
1.02 (t, J = 7.4 Hz, 3H), 1.43 (t, J = 7.2 Hz, 2ꢁ3H), 1.90 (m, 2H), 3.18
(s, 3H), 3.38 (m, 2H), 3.55 ppm (q, J = 7.3 Hz, 2ꢁ2H); 19F NMR: d =
ꢁ78.9 ppm (s); elemental analysis calcd (%) for C10H20F6N2O4S2 (410.4):
C 29.3, H 4.9, N 6.8; found: C 29.1, H 4.9, N 6.9.
[6] a) P. Wang, S. M. Zakeeruddin, P. Comte, I. Exnar, M. Gratzel, J.
Am. Chem. Soc. 2003, 125, 1166–1167; b) H. Matsumoto, T. Matsu-
da, T. Tsuda, R. Hagiwara, Y. Ito, Y. Miyazaki, Chem. Lett. 2001,
26–27.
[7] W. Lu, A. G. Fadeev, B. Qi, E. Smela, B. R. Mattes, J. Ding, G. M.
Spinks, J. Mazurkiewicz, D. Zhou, G. G. Wallace, D. R. MacFarlane,
S. A. Forsyth, M. Forsyth, Science 2002, 297, 983–987.
[8] R. Hagiwara, Y. Ito, J. Fluorine Chem. 2000, 105, 221–227.
[9] H. Matsumoto, M. Yanagida, K. Tanimoto, M. Nomura, Y. Kitaga-
wa, Y. Miyazaki, Chem. Lett. 2000, 922–923; b) H. Matsumoto, H.
Yanagida, K. Tanimoto, T. Kojima, Y. Tamiya, Y. Miyazaki in Proc.
Electrochem. Soc. (Molten Salts XII), Vol. 99–41 (Eds.: P. C. Trulove,
H. C. De Long, G. R. Stafford, S. Deki), 2000, pp. 186–192; c) H.
Matsumoto, H. Kageyama, Y. Miyazaki in Proc. Electrochem. Soc.
(Molten Salts XIII), Vol. 2002–19 (Eds.: P. C. Trulove, H. C.
De Long, R. A. Mantz, G. R. Stafford, M. Matsunaga), 2002,
pp. 1057–1065.
[10] H. Sakaebe, H. Matsumoto, Electrochem. Commun. 2003, 5, 594–
598.
[11] J. H. Shin, W. A. Henderson, S. Passerini, Electrochem. Commun.
2003, 5, 1016–1020.
[12] P. C. Howlett, D. R. MacFarlane, A. F. Hollenkamp, Electrochem.
Solid-State Lett. 2004, 7, A97-A101.
[13] T. Sato, G. Masuda, K. Takagi, Electrochim. Acta 2004, 49, 3603–
3611.
[14] A. I. Bhatt, I. May, V. A. Volkovich, M. E. Hetherington, B. Lewin,
R. C. Thied, N. Ertokc, J. Chem. Soc. Dalton Trans. 2002, 4532–
4534.
[15] a) J. E. Gordon, J. Am. Chem. Soc. 1965, 87, 4347–4358; b) J. E.
Gordon, G. N. Subba Rao, J. Am. Chem. Soc. 1978, 100, 7445–7454;
c) W. T. Ford, R. J. Hauri, D. J. Hart, J. Org. Chem. 1973, 38, 3916–
3918.
[16] a) J. Sun, M. Forsyth, D. R. MacFarlane, Ionics 1997, 3, 356–362;
b) J. Sun, M. Forsyth, D. R. MacFarlane, J. Phys. Chem. B 1998, 102,
8858–8864; c) D. R. McFarlane, J. Sun, J. Golding, P. Meakin, M.
Forsyth, Electrochim. Acta 2000, 45, 1271–1278.
[17] H. Matsumoto, H. Kageyama, Y. Miyazaki, Chem. Lett. 2001, 182–
183.
N-(n-Butyl)-N,N-diethyl-N-methylammonium
bis(trifluoromethanesul-
fonyl)imide (N1224[TFSI]): Yield: 87%, colorless liquid; 1H NMR: d =
0.99 (t, J = 7.4 Hz, 3H), 1.42 (t, J = 7.2 Hz, 2ꢁ3H), 1.46 (t, J = 7.4 Hz,
2H), 1.84 (m, 2H), 3.18 (s, 3H), 3.43 (m, 2H), 3.56 ppm (q, J = 7.3 Hz,
2ꢁ2H); 19F NMR: d = ꢁ78.9 ppm (s); elemental analysis calcd (%) for
C11H22F6N2O4S2 (424.42): C 31.1, H 5.2, N 6.6; found: C 31.1, H 5.3, N
6.6.
N,N,N-Trimethyl-N-(2-methoxyethyl)ammonium bis(trifluoromethanesul-
1
fonyl)imide (N111.1O2[TFSI]): Yield: 87%, white solid; H NMR: d = 3.39
(m, 4ꢁ3H), 3.80 (s, 2H), 3.96 ppm (s, 2H); 19F NMR: d = ꢁ78.9 ppm
(s); elemental analysis calcd (%) for C8H16F6N2O5S2 (398.34): C 24.1, H
4.1, N 7.0; found: C 24.1, H 4.1, N 7.0.
N-Ethyl-N,N-dimethyl-N-(2-methoxyethyl)ammonium
bis(trifluorome-
thanesulfonyl)imide (N112.1O2[TFSI]): Yield: 85%, colorless liquid;
1H NMR: d = 1.47 (t, J = 7.2 Hz, 3H), 3.31 (s, 2ꢁ3H), 3.39 (s, 3H),
3.67 (q, J = 7.2 Hz, 2H), 3.74 (t, J = 4.8 Hz, 2H), 3.94 ppm (s, 2H);
19F NMR:
d
=
ꢁ78.9 ppm (s); elemental analysis calcd (%) for
C9H18F6N2O5S2 (412.37): C 26.2, H 4.4, N 6.8; found: C 26.1, H 4.4, N 6.8.
N,N-Diethyl-N-methyl-N-(2-methoxyethyl)ammonium bis(trifluorome-
thanesulfonyl)imide (N122.1O2[TFSI]): Yield: 86%, colorless liquid;
1H NMR: d = 1.44 (t, J = 7.2 Hz, 2ꢁ3H), 3.24 (s, 3H), 3.39 (s, 3H),
3.62 (q, J = 7.3 Hz, 2ꢁ2H), 3.71 (t, J = 4.8 Hz, 2H), 3.92 ppm (s, 2H);
19F NMR:
d
=
ꢁ78.9 ppm (s); elemental analysis calcd (%) for
C10H20F6N2O5S2 (426.4): C 28.2, H 4.7, N 6.6; found: C 27.9, H 4.8, N 6.6.
N,N,N-Triethyl-N-(2-methoxyethyl)ammonium bis(trifluoromethanesulfo-
nyl)imide (N222.1O2[TFSI]): Yield: 88%, colorless liquid; 1H NMR: d =
1.40 (t, J = 7.2 Hz, 3ꢁ3H), 3.38 (s, 3H), 3.57 (q, J = 7.2 Hz, 3ꢁ2H),
3.67 (t, J = 4.8 Hz, 2H), 3.89 ppm (s, 2H); 19F NMR: d = ꢁ78.9 ppm
(s); elemental analysis calcd (%) for C11H22F6N2O5S2 (440.42): C 30.0, H
5.0, N 6.4; found: C 29.8, H 4.7, N 6.1.
[18] J. E. Gordon in Techniques and Methods of Organic and Organome-
tallic Chemistry, Vol. 1 (Ed.: R. E. Denny), Marcel Dekker, New
York, 1969, pp. 51–188.
[19] a) E. I. Cooper, C. A. Angell, Solid State Ionics 1983, 9/10, 617–622;
b) E. I. Cooper, C. A. Angell, Solid State Ionics 1986, 18/19, 570–
576.
[20] M. Ue, K. Ida, S. Mori, J. Electrochem. Soc. 1994, 141, 2989–2996.
[21] M. Ue, M. Takeda, M. Takehara, S. Mori, J. Electrochem. Soc. 1997,
144, 2684–2688.
Acknowledgement
This work was supported by R&D project for Li batteries by METI and
NEDO of Japan. We are grateful to Ms Kazuko Igarashi and Ms Etsuko
Hazama for their assistance in this work.
[22] K. Xu, M. S. Ding, T. R. Jow, J. Electrochem. Soc. 2001, 148, A267-
A274.
[23] S. Forsyth, J. Golding, D. R. MacFarlane, M. Forsyth, Electrochim.
Acta 2001, 46, 1753–1757.
[24] J. Golding, N. Hamid, D. R. MacFarlane, M. Forsyth, C. Forsyth, C.
Collins, J. Huang, Chem. Mater. 2001, 13, 558–564.
[25] W. Xu, E. I. Cooper, C. A. Angell, J. Phys. Chem. B 2003, 107,
6170–6178.
[26] W. Xu, L. M. Wang, R. A. Nieman, C. A. Angell, J. Phys. Chem. B
2003, 107, 11749–11756.
[1] a) D. Holbrey, K. R. Seddon, Clean Prod. Process 1999, 1, 223–236;
b) T. Welton, Chem. Rev. 1999, 99, 2071–2084; c) P. Wasserscheid,
W. Keim, Angew. Chem. 2000, 112, 3926–3945; Angew. Chem. Int.
Ed. 2000, 39, 3772–3789; d) J. Dupont, R. F. de Souza, P. A. Z.
Suarez, Chem. Rev. 2002, 102, 3667–3692.
[2] a) R. A. Sheldon, R. M. Lau, M. J. Sorgedrager, F. van-Rantwijk,
K. R. Seddon, Green Chem. 2002, 4, 147–151; b) U. Kragel, M. Eck-
stein, N. Kaftzik, Curr. Opin. Biotechnol. 2002, 13, 565–571; c) S.
Park, R. J. Kazlauskas, Curr. Opin. Biotechnol. 2003, 14, 432–437.
[3] a) C. C. Tzschucke, C. Markert, W. Bannwarth, S. Roller, A. Hebel,
R. Haag, Angew. Chem. 2002, 114, 4136–4173; Angew. Chem. Int.
Ed. 2002, 41, 3964–4000; b) C. F. Poole, J. Chromatogr. A 2004,
1037, 49–82.
[27] H. Matsumoto, H. Kageyama, Y. Miyazaki, Chem. Commun. 2002,
1726–1727.
[28] D. R. MacFarlane, J. Golding, S. Forsyth, M. Forsyth, G. B. Deacon,
Chem. Commun. 2001, 1430–1431.
[29] D. R. MacFarlane, P. Meakin, J. Sun, N. Amini, M. Forsyth, J. Phys.
Chem. B. 1999, 103, 4164–4170.
[30] P. Lozano, T. de Diego, S. Gmouh, M. Vaultier, J. L. Iborra, Biotech-
nol. Prog. 2004, 20, 661–669.
[4] a) N. Nakagawa, S. Izuchi, K. Kuwana, T. Nukuda, Y. Aihara, J.
Electrochem. Soc. 2003, 150, A695-A700; b) B. Garcia, S. Lavallee,
G. Perron, C. Michot, M. Armand, Electrochim. Acta 2004, 49,
4583–4588.
[31] a) O. D. Gupta, P. D. Armstrong, J. M. Shreeve, Tetrahedron Lett.
2003, 44, 9367–9370; b) J. Kim, R. P. Singh, J. M. Shreeve, Inorg.
Chem. 2004, 43, 2960–2966.
[5] A. Weber, G. E. Blomgrem in Advances in Lithium-Ion Batteries
(Eds.: W. A. Van Schalkwijk, B. Scrosati), Kluwer Academic/Plenum
Publishers, New York, 2002, pp. 185–232.
Chem. Eur. J. 2005, 11, 752 – 766
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
765

H. Matsumoto et al.
[32] K. Kim, S. Choi, D. Demberelnyamba, H. Lee, J. Oh, B. Lee, S.
Mun, Chem. Commun. 2004, 828–829.
[33] a) A. P. Abbott, G. Capper, D. L. Davies, R. Rasheed, Inorg. Chem.
2004, 43, 3447–3452; b) P. Abbott, G. Capper, D. L. Davies, H. L.
Munro, R. K. Rasheed, V. Tambyrajah, Chem. Commun. 2001,
2010–2011.
[34] M. Armand, W. Gorecki, R. Andreani, in Second International Sym-
posium on Polymer Electrolytes (Ed.: B. Scrosati), Elsevier, London,
1990, p. 91–97.
[35] a) P. Johansson, S. P. Gejji, J. Tegenfeldt, J. Lindgren, Electrochim.
Acta 1998, 43, 1375–1379; b) P. Johansson, P. Jacobsson, J. Phys.
Chem. A 2001, 105, 8504–8509.
[36] a) V. P. W. Bohm, W. A. Herrmann, Chem. Eur. J. 2000, 6, 1017–
1025; b) G. Zou, Z. Wang, J. Zhu, J. Tang, M. Y. He, J. Mol. Catal.
A: Chem. 2003, 206, 193–198.
[37] H. J. Frohn, V. V. Bardin, Z. Anorg. Allg. Chem. 2001, 627, 15–16.
[38] Z. B. Zhou, M. Takeda, M. Ue, J. Fluorine Chem. 2003, 123, 127–
131.
[39] A. A. Kolomeitsev, A. A. Kadyrov, J. Szczepkowska-Sztolcman, M.
Milewska, H. Koroniak, G. Bissky, J. A. Barten, G. V. Roschenthal-
er, Tetrahedron Lett. 2003, 44, 8273–8277.
[40] G. A. Molander, B. P. Hoag, Organometallics 2003, 22, 3313–3315.
[41] a) V. V. Bardin, S. G. Idemskaya, H. J. Frohn, Z. Anorg. Allg. Chem.
2002, 628, 883–890; b) M. Takeda, M. Takehara, M. Ue (Mitsubishi
Chem.), JP63934A, 2002; c) M. Schmidt, A. Kuhner, K.-D. Franz,
G.-V. Roschenthaler, G. Bissky, A. Kolomeitsev, A. Kadyrov (Merck
GmbH), US0160261A1, 2002.
[42] a) V. V. Bardin, H. J. Frohn, Main Group Met. Chem. 2002, 25, 589–
613; b) G. Pawelke, H. Burger, Coord. Chem. Rev. 2001, 215, 243–
266.
[43] Z. B. Zhou, M. Takeda, M. Ue, J. Fluorine Chem. 2004, 125, 471–
476.
[44] Z. B. Zhou, H. Matsumoto, K. Tatsumi, Chem. Lett. 2004, 680–681.
[45] Z. B. Zhou, H. Matsumoto, K. Tatsumi, Chem. Eur. J. 2004, 10,
6581–6591.
[46] Z. B. Zhou, H. Matsumoto, K. Tatsumi, Chem. Lett. 2004, 886–887.
[47] K. R. Seddon, A. Stark, M. J. Torres, Pure Appl. Chem. 2000, 72,
2275–2287.
[48] J. Timmermans, J. Phys. Chem. Solids 1961, 18, 1–8.
[49] a) D. MacFarlane, J. Huang, M. Forsyth, Nature 1999, 402, 792–794;
b) D. MacFarlane, M. Forsyth, Adv. Mater. 2001, 13, 957–966; c) D.
MacFarlane, P. Meakin, N. Amini, M. Forsyth, J. Phys. Condens.
Matter 2001, 13, 8257–8267.
[50] a) A. Abouimrane, Y. Abu-Lebdeh, P. J. Alarco, M. Armand, J.
Electrochem. Soc. 2004, 151, A1028–A1028; b) Y. Abu-Lebdeh,
P. J. Alarco, M. Armand, Angew. Chem. 2003, 115, 4637–4639;
Angew. Chem. Int. Ed. 2003, 42, 4499–4501.
[51] a) J. C. Dearden, Sci. Total Environ. 1991, 109/110, 59–68; b) R.
Abramowitz, S. H. Yalkowsky, Pharm. Res. 1990, 7, 942–947.
[52] W. Xu, C. A. Angell, Electrochem. Solid-State Lett. 2001, 4, E1-E4.
[53] D. Aurbach, H. Gottlieb, Electrochim. Acta 1989, 34, 141–156.
[54] U. Schroder, J. D. Wadhawan, R. G. Compton, F. Marken, P. A. Z.
Suarez, C. S. Consorti, R. F. de Souza, J. Dupont, New J. Chem.
2000, 24, 1009–1015.
Received: August 10, 2004
Published online: December 6, 2004
766
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 752 – 766