Thermomorphic phase separation in ionic liquid-organic liquid systems - Conductivity and spectroscopic characterization

Article abstract of DOI:10.1039/b505588a

Electrical conductivity, FT-Raman and NMR measurements are demonstrated as useful tools to probe and determine phase behavior of thermomorphic ionic liquid-organic liquid systems. To illustrate the methods, consecutive conductivity measurements of a thermomorphic methoxyethoxyethyl-imidazolium ionic liquid/ 1-hexanol system are performed in the temperature interval 25-80°C using a specially constructed double-electrode cell. In addition. FT-Raman and ¹H-NMR spectroscopic studies performed on the phase-separable system in the same temperature interval confirm the mutual solubility of the components in the system, the liquid liquid equilibrium phase diagram of the binary mixture, and signify the importance of hydrogen bonding between the ionic liquid and the hydroxyl group of the alcohol. The Owner Societies 2005.

Full text of DOI:10.1039/b505588a

View Article Online / Journal Homepage / Table of Contents for this issue
R E S E A R C H P A P E R
Thermomorphic phase separation in ionic liquid–organic liquid
systems—conductivity and spectroscopic characterization
Anders Riisager,*ab Rasmus Fehrmann, Rolf W. Berg, Roy van Halw and
a
a
bc
Peter Wasserscheid*c
a
Department of Chemistry and Interdisciplinary Research Center for Catalysis (ICAT),
Technical University of Denmark, DK-2800 Kgs. Lyngby, Denmark. E-mail: ar@kemi.dtu.dk;
Fax: (þ45) 45252235; Tel: (þ45) 45252233
b
Institut fu¨r Technische Chemie und Makromolekulare Chemie, RWTH-Aachen, Worringer
Weg 1, D-52074 Aachen, Germany
c
Lehrstuhl fu¨r Chemische Reaktionstechnik, Universita¨t Erlangen-Nu¨rnberg, Egerlandstr. 3,
D-91058 Erlangen, Germany. E-mail: peter.wasserscheid@crt.cbi.uni-erlangen.de;
Fax: (þ49) 9131 8527421; Tel: (þ49) 9131 8527420
Received 21st April 2005, Accepted 29th June 2005
First published as an Advance Article on the web 21st July 2005
Electrical conductivity, FT-Raman and NMR measurements are demonstrated as useful tools to probe and
determine phase behavior of thermomorphic ionic liquid–organic liquid systems. To illustrate the methods,
consecutive conductivity measurements of a thermomorphic methoxyethoxyethyl-imidazolium ionic liquid/
1
-hexanol system are performed in the temperature interval 25–80 1C using a specially constructed double-
1
electrode cell. In addition, FT-Raman and H-NMR spectroscopic studies performed on the phase-separable
system in the same temperature interval confirm the mutual solubility of the components in the system, the
liquid–liquid equilibrium phase diagram of the binary mixture, and signify the importance of hydrogen bonding
between the ionic liquid and the hydroxyl group of the alcohol.
1
8–21
phine ligands.
Also, aqueous-organic systems exhibiting
temperature-dependent reversible phase separation due to
Introduction
In recent years, several concepts applying ionic liquids have
1
been developed for phase-separable catalytic applications and
22
preferential water-solubility of products are known.
Ionic liquids are entirely composed of ions which facilitate
their study by conductivity measurements. This has previously
been used extensively for phase equilibria studies of solid/
liquid vanadium alkali pyrosulfate molten melts (i.e. high-
temperature ionic liquids) to determine phase diagrams of
2
separation technology, by using the unique properties of ionic
3
liquids in comparison to traditional liquid reaction media,
,4
5
e.g. non-volatility, polar but yet non-coordinating nature,
6
high thermal stability and large liquidus range. Some of the
more recent catalytic developments include the use of transi-
7
23
24
the mixtures. In addition, ionic liquids are polar highly
25–27
–9
tion metal catalyzed two-phase
and analogous ‘‘hetero-
1
ordered hydrogen-bonded fluids.
As a consequence, a
0,11
genized’’ supported ionic liquid phase (SILP)
1
systems as
2
–ionic liquid systems. For selective separations
solvent used to dissolve an ionic liquid in a system possessing
thermomorphic phase behavior must be quite polar and have a
comparatively high permittivity (dielectric constant), since the
ability of a solute to form hydrogen bonds or other possible
relatively strong interactions with a solvent is an important
feature of this behavior. Thus, it must be a good insulator to
lower the attraction between the ions in the ionic liquid. In
addition, the ionic liquid may interact with the species of the
other phase in competition with the intermolecular forces of
the molecules in that phase and thereby facilitate phase mis-
cibility. A common class of solvents which possesses such
characteristics is alcohols. Alcohols are composed of struc-
tured molecular entities associated via hydrogen bonding,
which is clearly reflected by their relatively high boiling points
compared to hydrocarbons having comparable molecular
weight and structure.
2
well as scCO
recent developments include the use of task specific ionic
liquids (TSILs) for, e.g. aqueous coordinative extraction of
1
3,14
heavy metals with polyether-modified ionic liquids,
moval of hetero-aromatics from fuel and diesel fractions
re-
,15
7
1
6
and adsorptive desulfurization of flue gas.
In search for alternative methodologies to perform organic
liquid phase-separable catalytic reactions using ionic liquid
technology, we have been attracted by an approach involving
thermomorphic ionic liquid–organic systems, i.e. composite
systems which change their phase behavior with temperature.
Such systems can eliminate mass-transfer limitations prevailing
in interfacial catalytic reactions (like in biphasic and possibly
also in SILP approaches), if the systems consist of a homo-
geneous phase at reaction conditions (e.g. elevated tempera-
tures) but of separated ionic liquid catalyst phase and organic
phase at lower temperatures. Thermo-regulated transition
metal catalysts are well-known from, e.g. fluorous phase-
Liquid–liquid equilibria (LLE) in binary mixtures of com-
mon [1-alkyl-3-methylimidazolium][X] ionic liquids (X ¼
28
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
3
SO )
[PF
6
] , [BF
4
] , [(CF
3
SO
2
)
2
N] , [(CN)
2
N] and CF
3
1
7
organic liquid systems and for aqueous-organic systems using
catalysts based on nonionic polyether-functionalized phos-
and 1-butanol or other linear and branched alcohols with
25,29
various chain lengths
have previously been examined using
visual cloud point determinations. In all cases, typically LLE
behavior of partially miscible two-component systems with an
upper critical solution temperature (UCST) was observed, with
w Present address: Solvent Innovation GmbH, Nattermannallee 1,
D-50829 Cologne, Germany.
3
052
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 3 0 5 2 – 3 0 5 8
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5

View Article Online
In the present study, we apply primarily conductivity as a
quantitative method for measuring the phase separation tem-
peratures rather than the more qualitative visual method of
cloud point determination.
Experimental
General comments
All solvents and reagents used for syntheses were purchased
from commercial sources in high quality and used without
further purification. For the conductivity measurements the
double electrode cell with the gold electrodes (described below
Scheme 1 Preparation of 1,3-bis-[2-(2-methoxyalkoxy)ethyl]imidazo-
lium bis-trifluoromethylsulfonylamide ionic liquids.
in detail) was cleaned before use by 4 M HNO
3
, water and
ethanol and finally oven dried (110 1C). Moreover, 1-hexanol
˚
(
Fluka, Z 99%) was dried with 4 A molecular sieves before
relatively low solubility of the ionic liquid in the alcohol but
high solubility of the alcohol in the ionic liquid (sometimes
more than 50 mol%). Moreover, the alcohol solubility in the
ionic liquids followed the order of alcohol branching, i.e.
tertiary 4 secondary 4 primary alcohols, which correlates
with the relative basicity of the alcohols (i.e. hydrogen bond
acceptor capability). In addition, a longer alkyl chain of the
alcohol resulted in an increase in the UCST due to less
interaction with the ionic liquid through, e.g. hydrogen-bond-
ing, while an increase in the length of the alkyl chain on the
imidazolium cation resulted in higher mutual solubility, i.e.
decrease in UCST.
In order to exemplify the methodology of using electrical
conductivity as a probe of thermomorphic behavior, we have
measured the conductivity in a binary liquid system composed
of the ionic liquid 1,3-bis[2-(2-methoxyethoxy)ethyl]imidazo-
lium bis-trifluoromethylsulfonylamide, 3 (Scheme 1 with n ¼ 1)
and 1-hexanol, as a representative system. More generally, this
ionic liquid-alcohol system belongs to a series of thermo-
morphic composite systems containing various linear and
use. All NMR spectra were measured on a Varian Mercury 300
MHz spectrometer at 25 1C unless otherwise specified. The
density of the ionic liquid 3 was determined using a pycnometer
(
V ¼ 5.0 ml), the water content determined by Karl–Fisher
titration (Hydranal-Coulomat AG anolyte, Riedel-de Haen),
and the melting point (T ) and decomposition temperature
T ) by thermal analysis (Perkin Elmer PYRIS diamond DSC,
¨
m
2
9
(
d
ꢀ
1
temperature ramp: 5 1C min ).
Methanesulfonic acid 2-(2-methoxyethoxy)ethyl ester, 1 (n = 1)
To an ice-cooled solution of 93.7 ml 2-(2-methoxyethoxy)etha-
nol (95.52 g, 0.795 mol) and 110.2 ml triethylamine (80.45 g,
0.795 mol) in 1.5 l anhydrous diethyl ether was 61.7 ml
methanesulfonyl chloride (91.07 g, 0.795 mol) slowly added
over approximately 2 h under vigorous stirring. After contin-
uous stirring overnight, the precipitated triethylammonium
chloride was removed by filtration and the diethyl ether
removed under reduced pressure with a rotation evaporator.
The crude liquid ester product was then re-dissolved in di-
chloromethane, washed with water and finally isolated
by removing the solvent under reduced pressure again.
4
branched C –C12 alcohols and the di-substituted imidazolium
ionic liquid, which may be prepared by stepwise alkylation of
imidazole with methanesulfonic acid 2-(2-alkoxyethoxy)ethyl
esters 1 via mono-substituted 1-[2-(2-methoxyalkoxy)ethyl]imi-
dazole compounds 2 (Scheme 1). Using a similar synthetic
strategy we have also prepared [1-(2-(2-methoxyalkoxy)ethyl)-
Yield ¼ 152.43 g (97%).
1
H-NMR (300 MHz), CDCl /TMS: d/ppm ¼ 2.876 (3H, s, –
3
OCH ), 3.155 (3H, s, –SCH ), 3.337 (2H, t, J ¼ 4.5 Hz, CH O–
3
3
3
CH –CH –), 3.453 (2H, t, J ¼ 4.5 Hz, CH O–CH –CH –),
2
2
3
2
2
3
dialkyl phosphate or [(CF SO ) N] after additional anion
-alkylimidazolium][X] ionic liquids (where X ¼ alkyl sulfate,
3.556 (2H, t, J ¼ 4.5 Hz, –SO –CH –CH –), 4.162 (2H, t, J ¼
3 2 2 3
3
2
2
ꢀ
13
4.5 Hz, –SO –CH –CH –). C-NMR (75.34 MHz), CDCl /
3
2 2
exchange) by alkylations using the corresponding methyl and
ethyl compounds. More details on these ionic liquids will be
reported elsewhere.
TMS: d/ppm ¼ 37.54 (–OCH ), 58.93 (–SCH ), 69.11 (CH O–
2 2 3 2 2 3 2
3
3
3
CH –CH –), 69.83 (CH O–CH –CH –), 70.60 (–SO –CH –
CH –), 71.94 (–SO –CH –CH –).
2
3
2
2
3
0
Analogous bis-alkoxymethyl substituted and 1,3-bis-al-
1
3
koxymethyl-2-alkyl substituted imidazolium ionic liquids
ꢀ
ꢀ
ꢀ
ꢀ
1-[2-(2-Methoxyethoxy)ethyl]imidazole, 2 (n = 1)
containing Cl , [BF ] , [PF ] or [(CF SO ) N] anions have
4
6
3
2 2
previously been prepared by imidazole alkylation using halo-
genmethyl alkyl ethers followed by anion metathesis. Also,
closely related 1-methyl-3-alkoxyethyl substituted imidazolium
Under stirring and cooling in an ice-water bath 238.5 g sodium
hydroxide (5.96 mol) was added to a solution of 4.90 g tetra-
ethylammonium bromide (0.023 mol) in 300 ml water. After
most of the base was dissolved, 52.35 g imidazole (0.769 mol)
was added to the suspension followed by 152.43 g methane-
sulfonic acid 2-(2-methoxyethoxy)ethyl ester (1, n ¼ 1) (0.769
mol). The mixture was then reacted by stirring overnight at
room temperature, where the mixture was allowed to gradually
warm up by melting of the ice in the cooling bath. Finally, the
product was separated from the reaction mixture by extraction
with dichloromethane four times, where after the liquid, mono-
alkylated product was isolated by removal of the dichloro-
3
2
ionic liquids based on the same anions or on more hydro-
ꢀ
3
phobic perfluoroalkyltrifluoroborate anions (i.e. [R
3
F
BF
3
] ),
respectively, have previously been reported by similar proce-
dures and examined with respect to their physicochemical
properties including, e.g. visual solubility characteristics and
ionic conductivity. Notably, it was here realized that the
presence of ether functional alkyl groups in the imidazolium
ionic liquids considerably modified their solubility compared
to the similar alkyl-substituted compounds, while altering of
3
2
the anion did not have any significant influence. In addition,
also more hydrophilic 1-alkyl-3-poly(ethyleneglycol)imidazo-
methane under reduced pressure. Yield ¼ 64.4 g (49%).
1
H-NMR (300 MHz), CD
3
CN/TMS: d/ppm ¼ 3.328 (3H, s,
ꢀ
ꢀ
ꢀ
lium ionic liquids with [BF
4
] , [PF
6
] or [(CF
3
2
SO )
2
N] anions
–OCH
(2H, t, J ¼ 4.5 Hz, CH
Hz, –N–CH –CH
–), 4.115 (2H, t, J ¼ 5.2 Hz, –N–CH
6.949 (1H, broad s, –N–CHQCH–N), 7.093 (1H, broad s, –N–
CHQCH–N), 7.530 (1H, s, –NQCH–N–). C-NMR (75.34
CN/TMS: d/ppm ¼ 46.14 (–OCH
MHz), CD ), 57.64 (CH O–
CH –CH –), 69.64 (CH O–CH –CH –), 71.18 (–N–CH –CH –),
3
), 3.472 (2H, t, J ¼ 4.5 Hz, CH
O–CH –CH
–), 3.714 (2H, t, J ¼ 5.2
–CH –),
3 2 2
O–CH –CH –), 3.554
have been prepared by microwave mediated alkylation of
alkyl-imidazolium with chloropolyethoxy-ethanol compounds
3
2
2
2
2
2
2
3
4
followed by anion exchange reactions. Importantly, however,
so far none of the previous studies involving the hydroxyl- or
ether-modified imidazolium ionic liquids have included exam-
ination of thermomorphic phase-separable properties.
1
3
3
3
3
2
2
3
2
2
2
2
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 3 0 5 2 – 3 0 5 8
3053

View Article Online
1
2 2
17.10 (–N–CH –CH –), 119.21 (–N–CHQCH–N–), 128.11
(
–N–CHQCH–N–), 137.22 (–NQCH–N–).
1,3-bis-[2-(2-Methoxyethoxy)ethyl]imidazolium
bis-trifluoromethylsulfonylamide, 3 (n = 1)
A mixture of 7.56 g 1-[2-(2-methoxyethoxy)ethyl]imidazole (2,
n ¼ 1) (44.4 mmol) and 8.80 g methanesulfonic acid 2-(2-
methoxyethoxy)ethyl ester (1, n ¼ 1) (44.4. mmol) was heated
at 60 1C for approximately 100 h under argon atmosphere until
1
all imidazole was quaternized (determined by H-NMR of
samples taken during the reaction). Then, a solution of 12.91 g
lithium bis-trifluoromethylsulfonylamide (45.0 mmol) in 50 ml
water was added to the reaction mixture along with 50 ml
dichloromethane. After stirring the two-phase mixture vigor-
ously for 2 h the aqueous phase was separated from the organic
product phase, washed twice with dichloromethane and dis-
charged. The dichloromethane washing phases were then
combined with the organic reaction phase where after the
volatile solvent was removed under reduced pressure. Finally,
the non-volatile ionic liquid product left behind was further
dried in vacuo (1 mbar, 50 1C, 6 h). Yield ¼ 21.3 g (87%). r
ꢀ
3
(
ꢀ
22 1C) ¼ 1.330 g cm , [H
60 1C, T
¼ 344 1C.
H-NMR (300 MHz), CDCl
OCH
), 3.468 (4H, t, J ¼ 4.2 Hz, CH
O–CH –CH
–), 3.781 (4H, t, J ¼ 4.5 Hz,
2
O] ¼ 7.7 mM (104 ppm), T o
m
d
1
Fig. 1 Double-electrode pair conductivity cell. (a) Stainless steel
cannula injection tube, (b) PEEK Rotulex top with tightening screws,
3
/TMS: d/ppm ¼ 3.305 (6H, s,
–
3
3
O–CH –CH –), 3.587
2
2
(
c) PEEK Rotulex cap, (d) Pyrex glass tube (l ¼ 100 mm), (e) vertical
(
–
4H, t, J ¼ 4.2 Hz, CH
3
2
2
adjustable Accofal-coated gold electrode (d ¼ 1 mm), (f) PEEK
positioning plate, (g) gold electrode plate (8 ꢁ 12.5 ꢁ 2 mm), (h)
magnetic stirring bar.
N–CH –CH –), 4.301 (4H, t, J ¼ 4.6 Hz, –N–CH –CH –),
2
2
2
2
7
.419 (2H, s, –N–CHQCH–N–), 8.709 (1H, s, –NQCH–N–).
C-NMR (75.45 MHz), CDCl /TMS: d/ppm ¼ 49.96 (–OCH ),
1
3
3
3
5
9.02 (–N–CH
2
–CH –O–), 68.66 (–N–CH
2
–CH –O–), 70.42
2 2
electrode pair in the cell) used to obtain the specific conductiv-
(
CH O–CH –CH ), 71.70 (CH O–CH –CH ), 122.92 (–N–CHQ
3 2 2 3 2 2
19
ꢀ1
ities k (in mS cm with k ¼ LK) was calculated as the mean
value of cell constants determined from measuring the con-
ductivity of a 0.1 D KCl solution before and after each series of
measurements on the ionic liquid-alcohol system, as described
CH–N–), 123.08 (–NQCH–N–), 136.34 (–SO
NMR (282.33 MHz), CDCl /TMS: d/ppm ¼ ꢀ79.52 (-SO -CF ).
2
–CF
3
).
F-
3
2
3
Conductivity measurements
35
in the literature.
Each series of conductivity measurements was started by
measuring on heated (about 80 1C), monophasic systems after
allowing the system to equilibrate to constant conductivity
without stirring. Subsequently, the following measurements
were done at gradually decreased temperatures after similar
equilibration periods. Consecutive measurements were per-
formed on mixed systems with increasingly molar ratios of
The conductivity (L) was measured on the binary 3/1-hexanol
system using a Radiometer CDM230 conductivity meter and a
compact Pyrex glass cell (dinner ¼ 16 mm, l ¼ 100 mm, V E 11
3
cm ) equipped with two sets of gold electrodes penetrating a
gas tight (sealed by Viton O-ring) and electrically insulating
polyetheretherketone (PEEK) polymer top and a Rotulex cap
connected by tightening screws (Fig. 1). The vertically adjus-
table gold electrode pairs (8 ꢁ 12.5 ꢁ 2 mm Au plate, 1 mm Au
wire) were fixed and made gas tight (Viton O-rings around the
electrodes) by tightening the screws between the top and the
cap during measurements. To maintain the relative position of
the electrode plates during measurements, the electrode wires
were penetrating a PEEK positioning plate placed between the
upper electrode pair and the cell cap. Furthermore, in order to
perform conductivity measurements with the lower electrode
plates only, the gold wires were partly coated (from 3 mm
above the electrode plate to 20 mm from the top of the wire)
with a non-conducting PTFE-like layer (ca. 50 mm Accofal
1
-hexanol, i.e. X(1-hexanol) from 0 to 1, by injecting known
amounts of the alcohol directly through the cannula injection
tube into the ionic liquid contained in the cell. The specific
conductivities could be measured within ꢂ0.1%.
Spectroscopic measurements
The ionic liquid 3, 1-hexanol and a mixture thereof were
studied by FT-Raman spectroscopy at 25 1C (Bruker IF S66
FRA-106 FT-spectrometer, 1064 nm NIR Nd-YAG laser, 100
1
mW, liquid nitrogen cooled Ge-diode detector) and by H-
2
G54 layer, Accoat Coating, Denmark) possessing a high
NMR spectroscopy in the temperature interval 20–70 1C. The
variable temperature NMR measurements were performed on
the liquids (without spinning) in dried NMR capillary tubes
penetration resistance for liquids and high physico-chemical
and thermal (ꢀ240 to 250 1C) stability. Moreover, a stainless
steel cannula injection tube fitted with a septum positioned in
the PEEK top allowed successive alcohol addition without
manipulation of the cell during measurement series.
6
containing a sealed capillary tube with d -DMSO for signal
locking and as internal chemical shift reference, after rigorous
mixing, temperature equilibration and phase separation (if
present). The temperature of the NMR probe was precali-
brated to ꢂ0.1 1C by measuring on an ethyleneglycol standard.
Chemical shifts are determined to ꢂ0.01 ppm and the inte-
grated peaks to ꢂ2%.
To perform the conductivity measurements the cell contain-
ing a magnetic stirring bar was placed in a stirred (700–1000
rpm) and thermostatic silicon oil bath (Heidolph MR3003
magnetic stirrer with Pt100 thermo element regulation). The
actual cell temperature during the measurements was measured
with an external Pt100 thermo element connected to an Omega
DP460 control unit (ꢂ0.1 1C) positioned at the outer wall of
Results and discussion
ꢀ1
the glass cell. Cell conductivity constants K (about 0.5 cm
Consecutive conductivity was measured on the binary 3/
1-hexanol system at 13 different compositions in the full
depending on the actual position of the electrodes for each
3
054
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 3 0 5 2 – 3 0 5 8
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5

View Article Online
a
Table 1 Composition and CSTs of 3/1-hexanol mixtures
b
CSTs of 40 to 52 (ꢂ0.5) 1C depending on the composition
(
see Table 1).
In the examined temperature interval the conductivity of the
Composition no. Molar ratio 1-hexanol/3 X(1-hexanol) CST (1C)
3
alcohol-rich phase was increased by a factor of ca. 10 com-
pared to the non-ionic pure alcohol, as a result of a significant
solubility of strong ionic liquid electrolyte in the very weak
alcohol electrolyte. The presence of ionic liquid in the alcohol-
rich phase at room temperature was confirmed by the existence
1
2
3
4
5
6
7
8
9
1
1
1
1
a
0.0000
0.055(4)
0.095(1)
0.188(1)
0.443(7)
0.922(0)
1.805(8)
2.354(4)
3.189(7)
3.578(0)
7.207(2)
14.062(7)
—
0.0000
—
0.052(5)
0.086(9)
0.158(3)
0.307(3)
0.479(7)
0.643(6)
0.701(9)
0.761(3)
0.781(6)
0.878(2)
0.933(4)
1.0000
o20
o20
o20
o20
o20
o20
40
ꢀ
1
of the characteristic FT-Raman band at 742 cm assigned to
normal modes of coupled S–N and C–S stretching bands (and
ꢀ
perhaps CF
3
deformation) of the [(CF
3
SO
2
)
2
N] anion (in
3
6
48
51
accordance with the literature ), by comparison to the corre-
sponding solid lithium-salt (Fig. 3, see Table 2 for further
assignments of the observed FT-Raman bands to the species
present). Similarly, an increase in the relative intensity of FT-
0
1
2
3
52
51
ꢀ
Raman bands at, e.g. 2900 cm (assigned to CH and CH
2 3
1
—
stretching), indicated a considerable content of alcohol in the
ionic liquid-rich phase at room temperature. For this phase,
however, a less pronounced change in conductivity was ob-
served compared to the alcohol-rich phase despite the dissolu-
tion, indicating only minor interference of the alcohol with the
ion pairing in the ionic liquid at ambient temperature.
Assuming that only ionic liquid species are charge carrying
in the binary system and that the activity coefficients of the
species remain unchanged with temperature, the relative con-
Inaccuracies in compositions are indicated by the numbers in par-
b
entheses. CSTs (critical solution temperatures) determined by differ-
entiation of the conductivity equations (see text for more details).
composition range (see Table 1 for details) and in the tem-
perature range of 25–80 1C with a double electrode glass cell. In
Fig. 2 representative results are shown for the measured
specific conductivities (on a logarithmic scale) versus tempera-
ture in both phases of a system having a composition corre-
as pl oo nn ed. ing to X(1-hexanol) ¼ 0.878(2) and of the ionic liquid
The specific conductivities measured with the two sets of
electrode pairs in the 3/1-hexanol mixture were lower and
different than the values for the pure ionic liquid, as a result
of a large difference in concentration of the charge carrying
species. When heated from room temperature to a CST of
centration of the ionic liquid (C ) in the system can be
rel
estimated, as function of the temperature, from the ratio
between the measured specific conductivities of the ionic liquid
alone and the two phases, as shown in Fig. 4.
Here, it can be seen that Crel in the alcohol-rich and ionic
liquid-rich phases changed linearly with temperature from
initially 0.33 and 0.66 at room temperature, respectively, i.e.
concentration of the ionic liquid in the ionic liquid-rich phase
was twice the concentration in the alcohol-rich phase, to being
essentially equal at 0.50 above the CST (as expected for one
single phase).
5
2 1C specific conductivities progressively approached the same
values for the lower ionic liquid-rich and upper alcohol-rich
phase, respectively (Fig. 2). In contrast, significantly lower—
but identical—specific conductivity was measured with the two
electrode pairs in the cell above the CST when the mixture
became homogeneous. This observation strongly indicated
that phase miscibility was accompanied by a significant change
in mobility of the conducting species as a result of intermole-
cular interactions between the ionic liquid and alcohol species.
1
This was further supported by complementary H-NMR spec-
troscopic examinations (vide infra). For the examined alcohol-
rich systems with compositions X(1-hexanol) 4 0.643(6)
a similar phase separation was observed as above, with
Fig. 3 Normalized FT-Raman spectra recorded at 25 1C of solid
Li[(CF SO ) N] salt, 1-hexanol, ionic liquid 3, and the upper and
3 2 2
Fig. 2 Specific conductivity as function of the temperature of the ionic
liquid
J), respectively, of a 3/1-hexanol mixture with mole fraction
X(1-hexanol) ¼ 0.878(2).
3
(ꢁ) and of the upper phase (K) and lower phase
lower phase of a 3/1-hexanol system having X(1-hexanol) ¼ 0.878(2)
(arrows indicate characteristic FT-Raman bands of 1-hexanol and 3,
respectively).
(
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 3 0 5 2 – 3 0 5 8
3055

View Article Online
a
Table 2 Observed FT-Raman bands and their assignment
b
1-Hexanol
Ionic liquid 3
173 w, 3112 vw
Li[(CF
3
SO
2
)
2
N] (solid)
Assignment
Cation
3
2
2
2
955 w sh, 2928 w sh
900 vs
2963 vs, 2928 vw
2886 vs, 2831 s
2774 vw, 2728 vw
CH
CH
CH
x
x
x
str
str
str
874 vs, 2732 w
1
1
569 w, 1473 vw sh
451 m, 1425 m
Cation
Cation
1450 m
1-Hexanol
Anion
Anion
1359 vw
1345 w, 1319 m
1
340 m
1303 m, 1223 vw
1-Hexanol
Cation
Anion
1
1
1
294 w
243 s
218 vw br
1248 s
1212 w, 1229 w
Anion
Anion
1
1
153 w
131 w
1
137 m
Cation
Anion
1
1
188 vw, 1119 w
077 w, 1061 w
1-Hexanol
1-Hexanol
Cation
1
108 w, 1031 m
1027 w, 996 vw
1-Hexanol
Cation
9
62 w, 918 vw br
953 vw , 919 vw
888 m, 866 vw, 854 vw
1-Hexanol
1-Hexanol
Cation
8
8
7
48 m
17 w br
42 vs
811 vw br
748 vs
Anion
Anion
820 w, 754 vw
1-Hexanol
Anion
Anion
6
6
5
5
5
58 vw
32 vw
98 m
72 m
54 m
669 w
625 w
600 m
573 sh
560 w, 555 w
Anion
Anion
Anion
4
4
55 vw
1-Hexanol
Anion
Cation
4
28 w br
4
03 m
09 vw, 363 w
1-Hexanol
Anion
Anion
389 m
346 w sh
333 s, 317 w
3
3
39 w sh
28 m
Anion
317 vw
1-Hexanol
Cation
Anion
3
2
12 m
80 s
280 s
a
Intensity codes: w ¼ weak, m ¼ medium, s ¼ strong, sh ¼ shoulder, v ¼ very, br ¼ broad. Calibration with cyclohexane lines to a precision of
ꢀ1 b
.
1
cm
Assignments of Raman bands to group vibrations are difficult because couplings are intense. However, the bis-trifluoromethylsulfonyl-
37
36
amide anion vibrations have previously been studied and those in 1,3-dialkylimidazolium cation systems also recently. Cation ¼ 1,3-bis-[2-
(2-methoxyethoxy)ethyl]imidazolium, anion ¼ bis-trifluoromethylsulfonylamide.
The specific conductivities measured at temperatures be-
tween 20 and 80 1C (i.e. including data below and above the
UCSTs) in the ionic liquid-rich phase of 3/1-hexanol mixtures
having different compositions can be well fitted to empirical
5
4
3
2
equations of the type k(T) ¼ aT þ bT þ cT þ dT þ eT þ f
2
(
R factors Z 0.996). The equations obtained can further be
used to calculate conductivities at specific temperatures as
function of the alcohol molar ratio. In Fig. 5 calculated
isothermal curves using this method are shown.
In order to get further information on the composition and
the effect of the various interactions present in the mixture of
1
-hexanol and 3 undergoing phase change, H-NMR spectra
1
obtained in the temperature interval 20–70 1C of the pure
alcohol and of the lower, ionic liquid-rich phase (or homo-
geneous mixture at higher temperatures) in a system having an
initial composition X(1-hexanol) ¼ 0.807(0) were analyzed.
1
The measurements of H chemical shifts (d
H
) and comparison
Fig. 4 Relative concentration (Crel) of the ionic liquid 3 as function of
the temperature in the upper phase (K) and lower phase (J), res-
pectively, of a 3/1-hexanol mixture with mole fraction X(1-hexanol) ¼
0.878(2) (dotted line indicate the CST determined for the mixture).
of the integrals of clearly separated proton signals of the
alcohol and the ionic liquid, respectively, allowed determina-
tion of the temperature and composition dependant change of
3
056
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 3 0 5 2 – 3 0 5 8
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5

View Article Online
Fig. 5 Specific conductivities as function of 1-hexanol mole fractions
in lower ionic liquid-rich phases of 3/1-hexanol mixtures calculated at
isothermal temperatures of (from bottom to top): 20 1C (&), 30 1C (B ),
Fig. 7 Liquid–liquid equilibrium phase diagram showing upper cri-
tical solution temperatures (UCSTs) above 20 1C determined for the 3/
1
40 1C (D), 50 1C (ꢁ), 55 1C (L), 60 1C (–), 70 1C (J) and
1-hexanol binary system by conductivity (ꢂ0.5 1C) (K) and H-NMR
80 1C (þ).
(ꢂ1.0 1C) (J) measurements, respectively, as function of the 1-hexanol
mole fraction (dotted line represents extrapolation).
dH. Hence, in Fig. 6, representative compositions determined
as the molar 1-hexanol/3 ratios and the alcohol mole fractions
as function of the temperature calculated from the integral
construction of a LLE phase diagram of the 3/1-hexanol
system from temperatures above 20 1C, as illustrated in Fig. 7.
The determined phase diagram disclosed good consistency
between the complementary measurements, and revealed a
marked phase miscibility gap in the binary system. Notably,
all ionic liquid-rich compositions of the mixture had CSTs
below room temperature due to the very high alcohol solubility
in the ionic liquid already at room temperature (almost 50
mol%), as also previously observed for some related alcohol/
ratios of the alcohol-CH
3-NCH CH O or 3-OCH protons, respectively, are shown.
3
protons and the 3-NCHQCHN,
2
2
3
From the proton integrals an excellent agreement was found
between the three selected signals, revealing an increased
amount of alcohol in the saturated ionic liquid-rich phase,
corresponding to mole fractions of 1-hexanol between 0.53 and
0
.79, when the mixture was heated from room temperature to a
2
9
1
,3-dialkyl-imidazolium ionic liquid systems.
CST estimated to be 51.5 1C (Fig. 6). This CST value was
nearly the same as that found by conductivity measurements
for mixtures having related compositions (vide supra), and
confirmed a very high alcohol content in the ionic liquid-rich
phase, as also suggested by the relative ionic liquid concentra-
tions determined from the conductivity (see Fig. 4). Above the
CST, the mixture became homogeneous and, accordingly, the
composition remained unchanged in good agreement with the
initial value of X(1-hexanol) ¼ 0.807(0) of the mixture.
The temperature dependence of d on the alcohol-OH
H
proton (DdOH), where DdOH ¼ dOH(20 1C) ꢀ dOH(T), was also
determined for the system having a composition corresponding
to X(1-hexanol) ¼ 0.807(0) (Fig. 8).
For the pure alcohol DdOH was positive and changed almost
linearly upon heating, as a consequence of a gradual up-field
shift of dOH related to a steady structural change caused by a
progressive breaking of hydrogen bonds in the liquid at higher
temperatures (Fig. 8). Hence, at a temperature of 51.5 1C,
corresponding to the CST for the mixture, the DdOH value was
Combined, the CSTs determined from the conductivity and
the NMR examinations together with the phase saturation
temperatures found by the NMR experiment, allowed the
0
.47 ppm for the pure alcohol. In comparison, all other alcohol
protons were only shifted slightly down-field in the same
temperature interval, due to increased thermally molecular
Fig. 6 Composition of the lower phase of 3/1-hexanol system with
X(1-hexanol) ¼ 0.807(0) as function of the temperature determined by
Fig. 8 Temperature dependence of the chemical shift of the alcohol
proton (DdOH) in pure 1-hexanol (J) and in the lower phase of a 3/1-
hexanol system with X(1-hexanol) ¼ 0.807(0) (K) (dotted line indicates
the CST determined for the mixture from the conductivity measure-
ments).
1
H-NMR spectroscopy from comparison of alcohol-CH
integrals of 3-NCH ¼ CHN (&), 3-NCH
3
integral with
2 2 3
CH O (D) and 3-OCH (J),
respectively (dotted line indicate the CST determined by conductivity).
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 3 0 5 2 – 3 0 5 8
3057

View Article Online
3
4
J. S. Wilkes, J. Mol. Catal. A: Chem., 2004, 214, 11–17.
K. N. Marsh, J. A. Boxall and R. Lichtenthaler, Fluid Phase
Equilib., 2004, 219, 93–98.
M. J. Muldoon, C. M. Gordon and I. R. Dunkin, J. Chem. Soc.,
Perkin Trans. 2, 2001, 433–435.
J. G. Huddleston, A. E. Visser, W. M. Reichert, H. D. Willauer,
G. A. Broker and R. D. Rogers, Green Chem., 2001, 3, 156–164.
J. Esser, P. Wasserscheid and A. Jess, Green Chem., 2004, 6,
vibrations resulting in DdOH values of about ꢀ0.02 to ꢀ0.03
(
not shown).
In contrast to the situation in the pure alcohol, the variation
5
6
7
8
9
in the measured DdOH for the mixtures levelled off in the
temperature interval from 30 1C to the CST, reaching a much
lower value of DdOH ¼ 0.30 ppm (36% lower) at CST. At
temperatures above the CST DdOH increased practically line-
arly again with essentially the same slope as for the pure
alcohol, however, with a DdOH value corresponding to a shift
of approximately 10 1C when compared to the pure alcohol.
Combined, these observations clearly demonstrated that the
alcohol intra-molecular hydrogen bonding structure was not
preserved when the alcohol dissolved in the ionic liquid-rich
phase at intermediate temperatures below the CST. Most
likely, this can be attributed to competitive hydrogen bonding
of the alcohol to the ionic liquid anion or to the polyether alkyl
groups in the ionic liquid imidazolium cation. This also
suggests that mutual phase miscibilities are accompanied by
a notable change in mobility of the conducting ionic liquid
species, as previously indicated by the complementary con-
ductivity measurements.
3
16–322.
H. Olivier-Bourbigou and L. Magna, J. Mol. Catal. A: Chem.,
002, 182–183, 419–437.
2
J. Dupont, R. F. de Souza and P. A. Z. Suarez, Chem. Rev., 2002,
102, 3667–3692.
10 A. Riisager, R. Fehrmann, S. Flicker, R. van Hal, M. Haumann
and P. Wasserscheid, Angew. Chem., Int. Ed., 2005, 44, 815–819.
C. P. Mehnert, Chem.–Eur. J., 2005, 11, 50–56.
1
1
1
2
S. V. Dzyuba and R. A. Bartsch, Angew. Chem., Int. Ed., 2003, 42,
48–150, and cited references.
1
13
14
15
A. E. Visser, R. P. Swatloski, W. M. Reichert, R. Mayton, S.
Sheff, A. Wierzbicki, J. H. Davis Jr. and R. D. Rogers, Chem.
Commun., 2001, 135–136.
J. D. Holbrey, A. E. Visser, S. K. Spear, W. M. Reichert, R. P.
Swatloski, G. A. Broker and R. D. Rogers, Green Chem., 2003, 5,
1
29–135.
A. Bosmann, L. Datsevich, A. Jess, A. Lauter, C. Schmitz and
P. Wasserscheid, Chem. Commun., 2001, 2494–2495.
¨
Conclusion
16 W. Wu, B. Han, H. Gao, Z. Liu, T. Jiang and J. Huang, Angew.
Chem., Int. Ed., 2004, 43, 2415–2417.
In this work we have demonstrated the use of conductivity
measurements as a simple technique for investigating thermo-
morphic behavior of binary ionic liquid-organic liquid systems,
exemplified by measurements on a binary methoxyethoxyethyl-
imidazolium ionic liquid/1-hexanol system. Moreover, we have
1
7
D. J. Brauer, K. W. Kottsieper, C. Liek, O. Stelzer, H. Waffensch-
midt and P. Wasserscheid, J. Organomet. Chem., 2001, 630,
177–184.
1
8
X. Zheng, J. Jiang, X. Liu and Z. Jin, Catal. Today, 1998, 44,
1
75–182.
19 J. Jiang, Y. Wang, C. Liu, F. Han and Z. Jin, J. Mol. Catal. A:
Chem., 1999, 147, 131–136.
1
shown how temperature dependent H-NMR and FT-Raman
measurements combined with the conductivity examinations
can provide (i) information about miscibility, (ii) be used as
complimentary methods for determining critical solution tem-
peratures (CSTs) as function of composition and the phase
diagram, and (iii) provide information about structural
changes occurring for the compounds upon mixing.
In perspective, we believe that the introduced conductivity
methodology can be implemented to monitor reactions in
biphasic ionic liquid catalysis, and as an analytical tool for
determining cross contamination in extraction and separation
technology based on ionic liquids. Furthermore, we envisage
that thermomorphic ionic liquid–organic liquid binary systems
may have application for important catalytic processes where
catalyst-product/reactant separation is problematic. Such
possible reactions will be explored in future work.
2
0
J. Jiang, Y. Wang, C. Liu, Q. Xiao and Z. Jin, J. Mol. Catal. A:
Chem., 2001, 171, 85–89.
21 Y. Wang, J. Jiang, Q. Miao, X. Wu and Z. Jin, Catal. Today, 2002,
4, 85–90.
P. J. Dyson, D. J. Ellis and T. Welton, Can. J. Chem., 2001, 79,
05–708.
7
2
2
3
7
2
S. B. Rasmussen, H. Hamma, O. B. Lapina, D. F. Khabibulin, K.
M. Eriksen, R. W. Berg, G. Hatem and R. Fehrmann, J. Phys.
Chem. B, 2003, 107, 13823–13830, and cited references.
24 S. N. V. K. Aki, J. F. Brennecke and A. Samanta, Chem.
Commun., 2001, 413–414.
2
5
U. Domanska and A. Marciniak, J. Phys. Chem. B, 2004, 108,
376–2382.
S. A. Katsyuba, P. J. Dyson, E. E. Vandyukova, A. V. Chernova
2
2
6
and A. Vidis, Helv. Chim. Acta, 2004, 87, 2556–2565.
27 J.-F. Huang, P.-Y. Chen, I.-W. Sun and S. P. Wang, Inorg. Chim.
Acta, 2001, 320, 7–11.
2
2
3
3
8
9
0
1
C.-T. Wu, K. N. Marsh, A. V. Deev and J. A. Boxall, J. Chem.
Eng. Data, 2003, 48, 486–491.
J. M. Crosthwaite, S. N. V. K. Aki, E. J. Maginn and J. F.
Brennecke, J. Phys. Chem. B, 2004, 108, 5113–5119.
J. Pernak, K. Sobaszkiewicz and J. Foksowicz-Flaczyk, Chem.–
Eur. J., 2004, 10, 3479–3485.
Acknowledgements
This work was supported by the ConNeCat project ‘‘Smart
ligands-smart solvents’’ financed by the German Federal Min-
istry for Education and Research (BMBF), and the Danish
Technical Research Council regarding part of the work carried
out at the Technical University of Denmark. The authors
thank Bodil Holten and Charlotte H. Gotfredsen (Department
of Chemistry, Technical University of Denmark), and Lykke
Ryelund (Department of Chemistry, University of Copenha-
gen) for technical assistance.
J. Pernak, A. Olszowka and R. Olszewski, Pol. J. Chem., 2003, 77,
179–187.
32 L. C. Branco, J. N. Rosa, J. J. Moura Ramos and C. A. M.
Afonso, Chem.–Eur. J., 2002, 8, 3671–3677.
Z.-B. Zhou, H. Matsumoto and K. Tatsumi, Chem.–Eur. J., 2004,
0, 6581–6591.
33
34
35
36
1
J. Fraga-Dubreuil, M.-H. Famelart and J. P. Bazureau, Org. Proc.
Res. Dev., 2002, 6, 374–378.
G. Jones and B. C. Bradshaw, J. Am. Chem. Soc., 1933, 55,
1
780–1800.
References
D. Brouillette, D. E. Irish, N. J. Taylor, G. Perron, M. Odziem-
kowski and J. E. Desnoyers, Phys. Chem. Chem. Phys., 2002, 4,
6063–6071.
1
C. C. Tzschucke, C. Markert, W. Bannwarth, S. Roller, A. Hebel
and R. Haag, Angew. Chem., Int. Ed., 2002, 41, 3964–4000.
Ionic Liquids in Synthesis, ed. P. Wasserscheid and T. Welton,
VCH-Wiley, Weinheim, 1st edn., 2003.
2
37 R. W. Berg, M. Deetlefs, K. R. Seddon, I. Shim and J. M.
Thompson, J. Phys. Chem. B, submitted.
3
058
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 3 0 5 2 – 3 0 5 8
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5