Density, thermal expansion and viscosity of cholinium-derived ionic liquids

Article abstract of DOI:10.1002/cphc.201100852

Density and viscosity data of the N-alkyl-N,N-dimethyl-N-(2-hydroxyethyl) ammonium bis(trifluoromethylsulfonyl)imide ionic liquids homologous series [N_{1 1 n 2(OH)}][Ntf₂] with n=1, 2, 3, 4 and 5 have been measured at atmospheric pressure in the 2832 group on the alkyl chain of the cation of about 17.2 cm ³mol^-1, except for [N_{1 1 1 2(OH)}]⁺, which exhibits an outlier behaviour. Molecular dynamics simulation results are used to explain the volumetric behaviour along the homologous series from a molecular perspective. The predictive power of group contribution methods for density and viscosity is also tested.

Full text of DOI:10.1002/cphc.201100852

DOI: 10.1002/cphc.201100852
Density, Thermal Expansion and Viscosity of
Cholinium-Derived Ionic Liquids
Anabela J. L. Costa,[a] Mꢀrio R. C. Soromenho,[a] Karina Shimizu,[b] Isabel M. Marrucho,*[a]
Josꢁ M. S. S. EsperanÅa,*[a] Josꢁ N. Canongia Lopes,*[a, b] and Luꢂs Paulo N. Rebelo[a]
Density and viscosity data of the N-alkyl-N,N-dimethyl-N-(2-hy-
and anions is used to estimate the effective molar volume of
the different cations present in this study. The results agree
with data for other cation families that show a molar volume
increment per CH2 group on the alkyl chain of the cation of
about 17.2 cm3 molꢀ1, except for [N1 1 1 2(OH)]+, which exhibits an
outlier behaviour. Molecular dynamics simulation results are
used to explain the volumetric behaviour along the homolo-
gous series from a molecular perspective. The predictive
power of group contribution methods for density and viscosity
is also tested.
droxyethyl)ammonium bis(trifluoromethylsulfonyl)imide ionic
liquids homologous series [N1 1 2(OH)][Ntf2] with n=1, 2, 3, 4
n
and 5 have been measured at atmospheric pressure in the
283<T/K<373 temperature range and the corresponding iso-
baric thermal expansion coefficients have been calculated. This
work studies the effect of increasing the alkyl chain length of
the cholinium-based cation on the density, viscosity and relat-
ed properties of this family of ionic liquids. A volumetric pre-
dictive method based on the effective molar volume of cations
1. Introduction
Choline, N,N,N-trimethyl-N-(2-hydroxyethyl)ammonium chlo-
ride, is a water-soluble essential nutrient usually grouped
within the B-complex vitamins. It can be found in food ob-
tained from vegetables or animals, and supports several essen-
tial biological functions.[1] Studies show that this nutrient is es-
pecially important during pregnancy and lactation.[1–3] More-
over, Meck et al. and Jones et al. have observed that a choline-
supplemented diet in rats play a critical role in brain develop-
ment and memory function.[2,3]
combination with distinct cations, bistriflimide anions generally
yield ionic liquids with relatively high toxicities.[10] Moreover,
the increase of the length of the alkyl side chain also negative-
ly affects toxicity, due to the increasing hydrophobicity of the
ionic liquid.[11] There are already several studies dealing with
the toxicity and biodegradability of cholinium-based ionic liq-
uids with different anions.[12–16] However, only one of them[13]
presents toxicity data for the bistriflimide anion, showing their
higher toxicity when compared to that of their analogues hal-
ides. That study also shows the negative effect on toxicity of
increasing the alkyl side chain in the cholinium cation.
Choline is used nowadays as a starting material for the de-
velopment of environmentally friendly ionic liquids. The goal is
to obtain ionic liquids that can unite the most relevant ionic
liquid properties (such as a wide liquid range, non-flammabili-
ty,[4] and negligible vapour pressure at ambient conditions)[5–7]
with biocompatible properties. In this context, the presence of
the hydroxyl group of the choline cation increases the biode-
gradability of the compound.[8]
Nevertheless, the scientific interest in cholinium-based ionic
liquids is increasing as shown by several studies already pub-
lished on possible applications, for example, in catalytic reac-
tions,[17–21] as reaction media,[22] in processing of metals and
metal oxides,[23] as crosslinking agents[24] or as protein handling
media.[25] Only one study on the separation of aliphatic and ar-
omatic compounds with [N1 1 n 2(OH)][Ntf2] has been found in the
Unfortunately, it is known that, for a given anion, ionic liq-
uids based on the cholinium cation show higher melting
points than those based on 1-alkyl-3-methylimidazolium cat-
ions.[9] Herein we have chosen bis(trifluoromethanesulfonyl)-
imide (bistriflimide) as the anion so that one could have
a wider range of compounds in the liquid state at around
room temperature. Our objective is to show the effect of in-
creasing the alkyl chain length of one of the alkyl groups con-
nected to the nitrogen in the cholinium cation in terms of den-
sity and viscosity.
[a] A. J. L. Costa, M. R. C. Soromenho, Dr. I. M. Marrucho,
Dr. J. M. S. S. EsperanÅa, Dr. J. N. C. Lopes, Prof. L. P. N. Rebelo
Instituto de Tecnologia Quꢀmica e Biolꢁgica
Universidade Nova de Lisboa
Av. Repfflblica, 2780-157 Oeiras (Portugal)
E-mail: imarrucho@itqb.unl.pt
jmesp@itqb.unl.pt
jnlopes@ist.utl.pt
Homepage: http://www.itqb.unl.pt
[b] Dr. K. Shimizu, Dr. J. N. C. Lopes
Centro de Quꢀmica Estrutural
Instituto Superior TØcnico
Choline is thought to be safe for the environment and
human health. However, care must be taken regarding the
choice of the anion and the size/functionality of the alkyl side
chains of the cation in order to preserve the “green” status of
the final product. On one hand it is already known that, in
1049-001 Lisboa (Portugal)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cphc.201100852.
ChemPhysChem 0000, 00, 1 – 9
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
1
&
ÞÞ
These are not the final page numbers!

I. M. Marrucho, J. M. S. S. EsperanÅa, J. N. Canongia Lopes, et al.
literature.[26] Studies of the volumetric and transport properties,
namely, density and viscosity, of cholinium based ionic liquids
are scarce due to the novelty of this family of ionic liquids, as
well as their high melting temperature.[27] However, there are
several papers that present the properties of mixtures of choli-
nium based ionic liquids with other solvents, like urea,[28] glyc-
erol,[29] carboxylic acids[30] and others,[31,32] called deep eutectic
solvents.
of two ILs present quasi-zero excess volumes.[40] This feature
led to the establishment[33–35] of a very simple, albeit very pre-
cise, rule-of-thumb for predicting the molar volume of any
aprotic IL. The volume is taken as the sum of the effective vol-
umes occupied by the constituent cations and anions (Vm =
V*c +V*a). The typical predictive power of this algorithm is
better than 0.5%. In order to establish a scale of effective vol-
umes of cations and anions, we had to anchor the value of
a reference ion. From calculations based on atomic radii, we
established a value of 73.7 cm3 molꢀ1 for the effective volume
occupied by [PF6]ꢀ at 298.15 K.[33] Since the molar volume of
the anion [Ntf2]ꢀ is well defined within the scheme of estimat-
ed effective ionic radii,[33–35] measuring 158.7 cm3 molꢀ1 at
298.15 K and atmospheric pressure, we used this methodology
to predict the volumes of the distinct cations studied based on
the Table SI1 (Supporting Information) experimental density
values. The volumes of the distinct cations are shown in bold
in Table 1.
Herein, N-alkyl-N,N-dimethyl-N-(2-hydroxyethyl) ammonium
bistriflimide ionic liquids, [N1 1 n 2(OH)][Ntf2] with n=1,2,3,4 and 5,
were synthesised and their density and viscosity were mea-
sured within an extended temperature range, 283<T/K<373.
Several predictive/interpretative methods were used and,
whenever necessary, new sets of parameters were calculat-
ed.[33–38]
2. Results and Discussion
2.1. Density
The experimental density data at atmospheric pressure for
Table 1. Effective size of several cations, V*c, and anions, V*a, at 298.15 K
and 1 bar.[33–35,41–46]
[N1
2(OH)][Ntf2], [N1
2(OH)][Ntf2], [N1
2(OH)][Ntf2],
3
1
1
1
2
1
[N1 1 4 2(OH)][Ntf2] and [N1 1 5 2(OH)][Ntf2] are presented in Figure 1
and Table SI1 (Supporting Information). The results follow the
IL Cation
V*c
[cm3 molꢀ1
IL Anion
V*a
[cm3 molꢀ1
]
]
[C0mim]+
[C2mim]+
[C4mim]+
[C6mim]+
[C8mim]+
[C10mim]+
[C12mim]+
64.8
99.2
[Cl]ꢀ
25.9
27.8
39.1
53.0
53.4
73.7
158.7
91.4
55.2
53.3
73.6
91.7
125.5
159.5
193.3
113.1
66.5
83.7
117.9
–
[Br]ꢀ
133.6
168.0
202.3
236.7
271.1
556.6
126.0
144.3
287.6
304.6
304.8
287.4
318.8
355.4
93.1
[NO3]ꢀ
[CH3CO2]ꢀ
[BF4]ꢀ
[PF6]ꢀ
[Ntf2]ꢀ
[(C6)3(C14)P]+
[CH3CH2SO4]ꢀ
[N(CN)2]ꢀ
[HSO4]ꢀ
[C1SO4]ꢀ
[C2SO4]ꢀ
[C4SO4]ꢀ
[C6SO4]ꢀ
[C8SO4]ꢀ
[FeCl4]ꢀ
[C1SO3]ꢀ
[C2SO3]ꢀ
[C4SO3]ꢀ
–
+
[N1114
]
[C4mpyrro]+
[C3-3-C1pyr]+
[C4-3-C1pyr]+
[C4-4-C1pyr]+
[C4pyr]+
[C6pyr]+
[C8pyr]+
[N1 1 1 2(OH)
+
]
Figure 1. Experimental density data, 1 [kgmꢀ3], of ionic liquids with
+
+
+
+
[N1 1 2 2(OH)
[N1 1 3 2(OH)
[N1 1 4 2(OH)
[N1 1 5 2(OH)
]
]
]
]
107.8
125.3
142.7
159.9
a common anion bis(trifluoromethanesulfonyl)imide, [Ntf2]ꢀ, and distinct N-
alkyl-N,N-dimethyl-N-(2-hydroxyethyl)ammonium cations, [N1 1 n 2(OH)]+, from
+
293.15 K to 363.15 K at atmospheric pressure: , [N1 1 1 2(OH)] ; , [N1 1 2 2(OH)]+;
^
^
–
–
+
, [N1 1 3 2(OH)]+; D, [N1 1 4 2(OH)] ; , [N1 1 5 2(OH)]+.
~
*
expected behaviour along the homologous series, with a sys-
tematic decrease in density as the alkyl side chain of the
cation increases.
The results show a CH2 group increment of 17.36 cm3 molꢀ1
for those ILs with an alkyl chain length longer than methyl,
which perfectly matches the range of values found in the liter-
ature for this group of 17.2ꢁ0.3 cm3 molꢀ1. This universal in-
crease occurs irrespective of the chemical nature and position
(cation or anion) of the substituted alkyl side chain.[35] The re-
sults also show that [N1 1 1 2(OH)][Ntf2] does not fit into this trend
To the best of our knowledge, there are no studies regarding
the effect of temperature on density for this cholinium-based
ionic liquid family. There is only one literature datum for the
density of [N1
2(OH)][Ntf2] at 298.15 K with a value of
2
1
1.4974 gcmꢀ3, which is 0.2% higher than our value.[39]
presenting instead
a
molar volume addition of only
Some methodologies based on the concept of effective
ionic volumes are extremely useful in predicting the density
and molar volume of ILs. Despite the highly complex liquid
structure of ILs, their volumetric properties have been shown
to behave very close to ideality. For instance, binary mixtures
14.7 cm3 molꢀ1 when changing from [N1
[N1 1 2 2(OH)][Ntf2].
2(OH)][Ntf2] to
1
1
The effective molar volume of distinct cations, V*c, and
anions, V*a, available in the literature were merged with the re-
sults obtained in this work and are listed in Table 1.[33–35,41–46]
&2
&
www.chemphyschem.org
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPhysChem 0000, 00, 1 – 9
ÝÝ
These are not the final page numbers!

Properties of Cholinium-Derived Ionic Liquids
Other methods are available for estimating the density of
ILs, namely the Ye and Shreeve method[36] and its extension
proposed by Gardas et al.,[37] who use a group contribution
method and Equation (1):
17.4 cm3 molꢀ1) with the one obtained recently for the series
1-ethyl-3-methylimidazolium
m3·moleculeꢀ1).[40]
alkylsulfate
(2.88·10ꢀ29
2.2. Isobaric Thermal Expansion Coefficient
1 ¼ M=½NVða þ bT þ cPÞꢂ
ð1Þ
For all ionic liquids investigated, a simple first-order polynomial
is insufficient to fit the data over the temperature range stud-
ied. Thus, a second-order polynomial equation was used to fit
the ln(1) vs T data [Eq. (3)]:
in which 1 is the density in kgmꢀ3, M is the molar mass in
kgmolꢀ1, N is the Avogadro constant, V is the molecular
volume in m3, T is the temperature in K and P is the pressure
in MPa. The coefficients a, b and c are 0.8005ꢁ0.0002, 6.652ꢄ
10ꢀ4 ꢁ0.007ꢄ10ꢀ4 Kꢀ1 and ꢀ5.919ꢄ10ꢀ4 ꢁ0.024ꢄ10ꢀ4 MPaꢀ1,
respectively, at 95% confidence level.[37]
2
lnð1=kg mꢀ3Þ ¼ A þ BðT=KÞ þ CðT=KÞ
ð3Þ
where A, B and C are fitting parameters and T is the tempera-
ture in K. The parameters obtained allow for the calculation of
the corresponding (ap) values [Eq. (4)]. The A, B and C parame-
ters for each ionic liquid, as well as the respective standard de-
viation, appear in the Table 3.
The group contributions for the molecular volumes of the
studied ILs were taken from the ref. [36] and are listed in
Table 2.
The isobaric thermal expansion coefficient, ap, can be de-
fined as the temperature derivative of ln(1) as expressed in
Equation (4):
Table 2. Comparison between the molecular volumes contributions from
the literature[36] and those calculated using the Ye and Shreeve method.
Species
Vmꢀ3
Literature
Calculated
apðKꢀ1Þ ¼ ꢀ½@ ln 1=@TðKÞꢂp¼ ꢀ½B þ 2CðT=KÞꢂ
ð4Þ
+
+
[N1 1 1 2(OH)
[N1 1 2 2(OH)
CH2 in [N1 1 n 2(OH)
]
]
1.72ꢄ10ꢀ28
2.00ꢄ10ꢀ28
2.80ꢄ10ꢀ29
1.70ꢄ10ꢀ28
1.95ꢄ10ꢀ28
2.89ꢄ10ꢀ29
The isobaric thermal expansion coefficients for this family of
ionic liquids are shown in Figure 2. The use of a second-order
polynomial function is warranted by the intrinsic precision
along the densimetry runs, determined using an apparatus
that allows for the measurement of density with five significant
digits, corresponding to 0.001% deviations. It must be
stressed, however, that the overall uncertainty of the density
measurements is estimated to be much higher (around 0.02%)
due to problems associated with the full characterization of
the samples regarding their purity. The corresponding estimat-
ed overall uncertainty in the values of ap is estimated at 2%,
but this fact does not preclude the following analysis of the ap
trends with temperature, evaluated from separate high-preci-
sion densimetry runs. Overall, the ap values increase with the
alkyl chain length of the cation.
+
] with n>2
To compare the predicted data with our experimental ones,
we used the mean percent deviation (MPD) for i data points,
given by Equation (2):
ꢀ
ꢀ
ꢀ
ꢀ
MPD½%ꢂ ¼ 100Si ð1cal ꢀ 1expÞ=1exp i=i
ð2Þ
The MPD of the predicted density data were between 1.2%
and 0.5%. Although a reasonable estimation of densities is
achieved by this method, it is valuable to re-estimate new
group contributions of each cation and consequently the CH2
contribution based on experimental data. The group contribu-
tion parameters of the anion are well defined in the litera-
ture,[36] so only those of the cations were estimated. The calcu-
lation of the new group contribution parameters was per-
formed by minimization of the MPD as defined in Equation (2).
The temperature dependence of the isobaric expansion co-
efficient is very small, with a negative slope in the studied tem-
perature range. This behaviour is not anomalous, with other
ionic liquids also presenting ap values with very modest T-de-
pendences and sometimes even minima in the ap versus T
plots not far from room-temperature.[40] For the
[N1 1 1 2(OH)][Ntf2] the slope of the ap with temperature is less ac-
The molecular volume of the cholinium cation, [N1
2(OH)]+,
1
1
and the CH2 group for the rest of the alkylcholinium-based
family of ionic liquids, [N1
1
n
2(OH)]+, with n>1, were calculat-
ed and appear in Table 2. The
MPD obtained with the new
group contributions is smaller
than 0.08% for all ionic liquids
used herein. It is interesting to
note the agreement in CH2
group contribution used in the
Table 3. Fitting parameters for density as a function of temperature using the quadratic polynomial equation
[Equation (3)] for the studied ILs with a common anion bis(trifluoromethanesulfonyl)imide, [Ntf2]ꢀ, and distinct
N-alkyl-N,N-dimethyl-N-(2-hydroxyethyl)ammonium cations, [N1
2(OH)]+. All fittings have a correlation coeffi-
1
n
cient higher than 0.9999.
IL Cation
[N1 1 1 2(OH)
[N1 1 2 2(OH)
[N1 1 3 2(OH)
[N1 1 4 2(OH)
[N1 1 5 2(OH)
Aꢁsa
Bꢁsb/104 Kꢀ1
Cꢁsc/107 Kꢀ2
+
+
+
+
+
]
7.5304
7.5062
7.4806
7.4565
7.4346
ꢁ0.0005
ꢁ0.0005
ꢁ0.0005
ꢁ0.0005
ꢁ0.0006
ꢀ7.220
ꢀ7.010
ꢀ7.163
ꢀ7.233
ꢀ7.275
ꢁ0.03
ꢁ0.03
ꢁ0.03
ꢁ0.03
ꢁ0.04
1.721
1.389
1.540
1.562
1.559
ꢁ0.05
ꢁ0.04
ꢁ0.05
ꢁ0.05
ꢁ0.06
]
]
]
]
density
prediction
(2.89ꢄ
10ꢀ29 m3·moleculeꢀ1
=
ChemPhysChem 0000, 00, 1 – 9
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemphyschem.org
&
3
&
ÞÞ
These are not the final page numbers!

I. M. Marrucho, J. M. S. S. EsperanÅa, J. N. Canongia Lopes, et al.
Figure 2. Isobaric thermal expansion coefficients, ap/Kꢀ1, of ionic liquids with
a common anion bis(trifluoromethanesulfonyl)imide, [Ntf2]ꢀ, and distinct N-
alkyl-N,N-dimethyl-N-(2-hydroxyethyl)ammonium cations, [N1 1 n 2(OH)]+, from
+
293.15 K to 363.15 K at atmospheric pressure: , [N1 1 1 2(OH)] ; , [N1 1 2 2(OH)]+;
^
^
+
, [N1 1 3 2(OH)]+; D, [N1 1 4 2(OH)] ; , [N1 1 5 2(OH)]+.
~
*
centuated than those of the other members of this IL family,
suggesting that [N1 1 1 2(OH)][Ntf2] will reach its lowest ap value at
a lower temperature than the other ionic liquids within the
series.
2.3. Structural Data from Molecular Dynamics Results
Figure 3. Radial distribution functions between selected pairs of interaction
centres in pure [N1 1 n 2(OH)][Ntf2] with n=1, 2, 3, 4 and 5, ionic liquids: (a) CT–
CT rdfs; (b) N–NBT and NBT–NBT rdfs. The out-of-phase behaviour of the rdfs
of (b) signals the charge ordering effect imposed by the iconicity of the sys-
tems. The outlier behaviour of [N1 1 1 2(OH)][Ntf2] is highlighted using a thicker
black curve in both figures.
The outlying behaviour of [N1 1 1 2(OH)][Ntf2] within the series (al-
ready found in terms of predicted densities using the effective
ionic volume scheme and exacerbated in the case of the ther-
mal expansion trends with temperature) can be interpreted as
a consequence of a different structural organization within the
liquid. This issue can be investigated using atomistic simula-
tions.
Molecular dynamics simulations have been extensively used
to predict the complex structural organization of ionic liquids,
where local charge ordering effects caused by the ubiquitous
presence of the ions that form the polar network of the ionic
liquid have to co-exist with non-polar domains composed of
the alkyl moieties that are also present in some ions.[47] In the
case of the cholinium cation and its alkylated derivatives,
[N1 1 n 2(OH)][Ntf2] with n=1, 2, 3, 4 and 5, new simulations of the
corresponding liquids have been performed at 298 K (Experi-
mental Section). The equilibrated simulation trajectories have
been used to calculate radial distribution functions (rdfs) for
selected pairs of interaction centres within each system
(Figure 3).
chains it is possible that such association might lead to the for-
mation of a second continuous non-polar network, permeating
the ever-present polar network. On the other hand for
[N1 1 1 2(OH)][Ntf2] no first peak is clearly visible. In fact the methyl
(or methylene) groups directly attached to the central nitrogen
atom of tetra-alkylammonium cations can be regarded as still
part of the polar network of the ionic liquid. Moreover, the
symmetry of the three methyl chains in [N1 1 1 2(OH)][Ntf2] (as op-
posed to the non-symmetrical substitution for all other studied
cases) implies that the polar network can interconnect in
a more isotropic, compact way. This can be inferred from Fig-
ure 3b, where the rdfs between selected charged interaction
centres in the anion (NBT) and in the cation (N) are analysed.
Whereas the cation–anion (NBT–N) show typical distances
common throughout the entire homologous series [the two
peaks of the rdf correspond to direct interactions between N
and NBT (0.45 nm) and indirect interactions between those
two atoms mediated by an oxygen atom of bistriflamide
(0.60 nm)], the anion-anion distances (corresponding to the
charge-ordered second shell of the polar network) are smaller
Figure 3a shows the rdfs between the terminal carbon
atoms (CT) of the longest unsubstituted alkyl chain of each
cation. The first peaks (slightly above 0.4 nm) of this type of
rdf are a fingerprint of the degree of segregation between the
polar network and the non-polar domains. Even for
[N1 1 2 2(OH)][Ntf2] there is already a non-negligible positive corre-
lation between CT atoms. For [N1 1 3 2(OH)][Ntf2], [N1 1 4 2(OH)][Ntf2]
and [N1 1 2(OH)][Ntf2], the association between CT groups be-
for [N1
2(OH)][Ntf2] (0.80 nm) than for the remainder of the
1
5
1
comes progressively more pronounced and for longer alkyl
series (around 0.84 nm).
&4
&
www.chemphyschem.org
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPhysChem 0000, 00, 1 – 9
ÝÝ
These are not the final page numbers!

Properties of Cholinium-Derived Ionic Liquids
Figure 4. Selected snapshots of simulation runs in the systems (from left to right) [N1 1 1 2(OH)][Ntf2], [N1 1 2 2(OH)][Ntf2], [N1 1 3 2(OH)][Ntf2], [N1 1 4 2(OH)][Ntf2] and
[N1 1 5 2(OH)][Ntf2]. The colour coding highlights the presence of the polar network (anions, cations=dark grey) and the growth of non-polar domains in its
midst (long non-substituted alkyl side chains of the cations=light grey).
Finally, the structural changes along the homologous series
and their concomitant impact on the volumetric properties of
the neat, bulk fluids can also be appreciated from a qualitative
perspective using the sequence of color-coded snapshots pre-
sented in Figure 4. The appearance of small non-polar clusters
containing the terminal atoms of the alkyl side chains only
Bh parameters are obtained using
a
group contribution
method based on Equation (6).
Ah ¼ Siniai;n
Bh ¼ Sinibi;n
ð6Þ
takes place from [N1 1 2(OH)][Ntf2] onwards. [N1 1 1 2(OH)][Ntf2] (a
2
pure polar network) remains a true outsider in relation to this
progression of increasingly larger non-polar pools dispersed in
the continuous ionic liquid polar network.
There are no parameters defined for the cholinium-based cat-
ions and their values were calculated using the fit to the exper-
imental viscosity data and the available literature ai,h and bi,h
values for the [Ntf2] anion.[38] The group contribution parame-
ters for the cations as a whole are given in Table 4. The results
2.4. Viscosity
Viscosity data, h(T), for all studied ionic liquids are presented
in Figure 5 and Table SI2 (Supporting Information). As expect-
ed, viscosity decreases with increasing temperature. The tem-
perature-dependence of the data can be analysed using the
Vogel-Fulcher-Tammann method represented by Equation (5),
Table 4. Group contribution parameters ai;h and bi;h calculated using the
Vogel–Tammann–Fulcher method.
Species
ai,h
bi,h/K
+
+
+
+
+
[N1 1 1 2(OH)
]
ꢀ7.54
ꢀ7.64
ꢀ8.05
ꢀ8.21
ꢀ8.32
773.68
776.33
840.94
892.85
931.55
[N1 1 2 2(OH)
[N1 1 3 2(OH)
[N1 1 4 2(OH)
[N1 1 5 2(OH)
]
]
]
]
À
Á
ln h ¼ Ah þ Bh= T ꢀ T0h
ð5Þ
where h is the viscosity in Pa·s, T is the temperature in K and
Ah, Bh and T0h are adjustable parameters.
According to the literature,[38] the optimal value of T0h for
the application of this correlation to ILs is 165.06 K. The Ah and
show that a more refined group contribution method based
on the addition of CH2 groups is not recommended due to the
non-monotonous behaviour of the viscosity data (Figure 6)—
[N1 1 1 2(OH)][Ntf2] and [N1 1 2 2(OH)][Ntf2] are outliers and it would be
spurious to try a fitting with just the remaining three cations.
The use of this new group contribution parameterization
allows us to correlate the viscosity data using the Vogel–Tam-
mann–Fulcher method with an MPD between 0.1% and 0.7%.
Figure 6 shows the viscosity trends as a function of the un-
substituted alkyl side chain length, n, of the cation at two se-
lected temperatures (298 K and 333 K). The viscosity data show
non-monotonous behaviour at both selected temperatures,
with viscosity minima at [N1
2(OH)][Ntf2] (298 K) and
2
1
[N1 1 3 2(OH)][Ntf2] (333 K). The observed behaviour confirms the
inadequacy of a general VTF method using a single T0h temper-
ature to account for the trends along the homologous series
(in other words, larger contributions from more CH2 groups
along the series cannot account for the non-monotonous be-
haviour of the trends).
Figure 5. Experimental viscosity data, ln (h/mPas), of ionic liquids with
a common anion bis(trifluoromethanesulfonyl)imide, [Ntf2]ꢀ, and distinct N-
alkyl-N,N-dimethyl-N-(2-hydroxyethyl)ammonium cations, [N1 1 n 2(OH)]+, from
+
283.15 K to 373.15 K at atmospheric pressure: , [N1 1 1 2(OH)] ; , [N1 1 2 2(OH)]+;
^
^
+
, [N1 1 3 2(OH)]+; D, [N1 1 4 2(OH)] ; , [N1 1 5 2(OH)]+.
~
*
ChemPhysChem 0000, 00, 1 – 9
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemphyschem.org
&
5
&
ÞÞ
These are not the final page numbers!

I. M. Marrucho, J. M. S. S. EsperanÅa, J. N. Canongia Lopes, et al.
found in the literature for the estimation of viscosity by the
Vogel–Tammann–Fulcher method. New contribution group pa-
rameters were derived for both methods, despite the reasona-
ble agreement of density data using the Ye–Schreeve method.
Molecular dynamics studies show that the existence of small
non-polar clusters containing the terminal atoms of the alkyl
side chains only takes place from [N1 1 2 2(OH)][Ntf2] onwards. This
structural change along the series is at the origin of the non-
conformal behaviour of [N1
series of ionic liquids.
2(OH)][Ntf2] within the studied
1
1
Experimental Section
Chemicals
The N-alkyl-N,N-dimethyl-N-(2-hydroxyethyl)ammonium bistrifli-
mide ionic liquids, [N1 1 n 2(OH)][Ntf2] with n=1, 2, 3, 4 and 5, used
herein were synthesised using the procedures described below.
Figure 6. Experimental viscosity values, (h/mPas), of ILs with a common
anion bis(trifluoromethanesulfonyl)imide, [Ntf2]-, and distinct N-alkyl-N,N-di-
+
methyl-N-(2-hydroxyethyl)ammonium cations, [N1 1 n 2(OH)
]
at atmospheric
*
*
pressure: , at 298.15 K; , at 333.15 K.
Commercially available reagents with analytical grade quality were
used as received. Lithium bistriflimide (LiNtf2) (mole fraction purity
99%) was purchased from IoLiTec. 1-bromoethane, 1-bromopro-
pane, 1-bromobutane and 1-bromopentane, all with mole fraction
purity ꢃ99% and dimethylaminoethanol with a mole fraction
purity ꢃ98% were purchased from Sigma–Aldrich.
As in the case of the volumetric properties, the dynamics
along this particular homologous series are also a consequence
of the complex nature of the systems. Apart from the structur-
al heterogeneities at a molecular level discussed in the previ-
ous section, one also has to consider the delicate balance be-
tween the electrostatic interactions characteristic of ionic
media, the progressively stronger van der Waals interactions
between the alkyl side chains of the cations, the possibility of
hydrogen-bonding via the hydroxyl groups of the choline
cation (with the oxygen atoms of the anion but also with
other hydroxyl groups of other cations), and even the presence
of weak van der Waals interactions between the terminal tri-
fluoromethyl groups of the bistriflamide anion. All these inter-
actions will exhibit differentiated temperature dependences[27]
and their relative strength will be a function of the degree of
polar-to-non-polar ratio along the homologous series.[47,48]
The salts [N1
2(OH)][Br] were prepared first. The alkyl halide
n
1
(CnH2n+1Br, n=2, 3, 4, 5) (1.1 mol equiv.) and 2-dimethylaminoetha-
nol (1.0 mol equiv.) were mixed in round-bottomed flask, dissolved
in n-hexane, under vigorous stirring, warmed-up to 808C under
1
reflux and N2 atmosphere for several hours (monitored by H NMR).
The formation of the product can be observed by its precipitation
from n-hexane. After completion of the reaction, the mixture was
then filtered and the reaction product was washed extensively
with n-hexane in order to remove unreacted compounds and then
dried under vacuum. The resulting [N1 1 2(OH)][Br] (yield 96–99%)
n
was used without further purification.
The synthesis of the [N1 1 n 2(OH)][Ntf2] series was then performed by
a metathesis reaction between the corresponding [N1
2(OH)][Br]
n
1
(1.0 equiv) and a slight excess of LiNtf2 (1.05 equiv). In a typical re-
action, 25 mL of LiNtf2 (5.25m in water) was added, at room tem-
perature, to an equal amount of a stirred solution of [N1 1 n 2(OH)][Br]
(5.0m in water). The mixture was stirred for an additional 30 min,
then extracted with dichloromethane (3ꢄ15 mL), and washed with
doubly-distilled deionized water (MilliQ) (3ꢄ5 mL). After evapora-
tion of the solvent, the product obtained was dried under vacuum.
The same procedure was adopted for the synthesis of
[N1 1 1 2(OH)][Ntf2], but using as a reagent a commercial sample of [N1
2(OH)][Cl], which was supplied by Sigma Aldrich with a stated
3. Conclusions
Density and viscosity data of N-alkyl-N,N-dimethyl-N-(2-
hydroxyethyl)ammonium
bis(trifluoromethanesulfonyl)imide
ionic liquids, with alkyl side chain n=1, 2, 3, 4 and 5 from
283.15 up to 373.15 K were measured. The objective was to
study the effect of increasing the alkyl chain length of the
cholinium-based cation on the density, viscosity and related
properties of this family of ionic liquids. As expected, density
decreases as the alkyl chain length of the cation increases
while a more complex behaviour was found for viscosity. Ther-
mal expansion coefficients were calculated from the measured
density data and for the [N1 1 1 2(OH)][Ntf2] the slope of the ap
with temperature is less negative than those of the others
members of this IL family.
The increase obtained in molar volume of 17.28 cm3 molꢀ1
per CH2 group added to the alkyl chain of the cation compares
well with the literature value except for [N1 1 1 2(OH)][Ntf2]. The
Ye–Schreeve group contribution method was used to estimate
density. No group contribution parameters for the cation were
1
1
purity better than 98%.
Table 5 lists the ILs, their respective acronyms and chemical struc-
ture. Proton NMR spectra were recorded with
a BRUKER
Avance 400 Ultrashield Plus spectrometer, operating at 400 MHz
and 298 K. The spectra, given in Table SI3, confirmed purity levels
better than 99% for all synthesised samples.
In order to reduce water and volatile compounds to negligible
values, all samples were dried under vacuum (0.1 Pa) and vigorous-
ly stirred at moderate temperature (330 K) for at least a day imme-
diately before their use. Coulometric Karl-Fischer titrations revealed
levels of water, always below d=300 ppm. It must be stressed that
cholinium-based ionic liquids (even when combined with hydro-
phobic anions such as [Ntf2]ꢀ) tend to be more hydrophilic than
&6
&
www.chemphyschem.org
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPhysChem 0000, 00, 1 – 9
ÝÝ
These are not the final page numbers!

Properties of Cholinium-Derived Ionic Liquids
Measurements were carried out at
temperatures between 283.15 and
373.15 K at atmospheric pressure.
At least 3 measurements were per-
formed for each IL and the report-
ed result is the average value with
Table 5. Acronyms and chemical structures of the ILs studied herein. Other acronyms used throughout the
text (not depicted) include the cations: [Cnmim]+ =1-alkyl-3-methylimidazolium; [P6 6 6 14 + =trihexyl-tetradecyl-
]
phosphonium; [N1 1 1 4
=1-alkyl-3-methylpyridinium; [C4-4-C1pyr]+ =1-butyl-4-methylpyridinium; and [Cnpyr]+ =N-alkylpyridinium.
]
+ =trimethyl-butylammonium; [C4mpyrro]=N-buty-N-methylpyrrolidinium; [Cn-3-C1pyr]+
Ionic Liquid
Acronym
Cation
Anion
a
mean deviation smaller than
N,N,N-trimethyl-N-(2-hydroxyethyl)ammonium
bis(trifluoromethanesulfonyl)
1.0%.
[N1 1 1 2(OH)] [Ntf2]
imide
N-ethyl-N,N-dimethyl- N-(2-hydroxyethyl)ammonium
bis(trifluoromethanesulfonyl)
imide
N-propyl-N,N-dimethyl- N-(2-hydroxyethyl)ammonium
bis(trifluoromethanesulfonyl)
[N1 1 2 2(OH)] [Ntf2]
[N1 1 3 2(OH)] [Ntf2]
[N1 1 4 2(OH)] [Ntf2]
[N1 1 5 2(OH)] [Ntf2]
Molecular Dynamics Simulations
Molecular dynamics simulations
were carried out for all ionic liq-
uids in this work. The simulations
were performed using the DL_
POLY code[50] and an all-atom force
field based on the CLaP force-
field,[51,52] specially tailored to en-
compass this and other ionic liquid
homologous series. For each ionic
liquid herein, we started from low-
density initial configurations com-
posed of 150 ion pairs. The boxes
imide
N-butyl-N,N-dimethyl- N-(2-hydroxyethyl)ammonium
bis(trifluoromethanesulfonyl)
imide
N-pentyl-N,N-dimethyl- N-(2-hydroxyethyl)ammonium
bis(trifluoromethanesulfonyl)
imide
their imidazolium-based counterparts. Titrations for [N1 1 1 2(OH)][Ntf2]
(the most hydrophilic of the studied ionic liquids) have yielded d=
164, 142 and 117 ppm after different experiences/recovery cycles.
The stated value of d=300 ppm is a conservative estimate that
takes into account the uncertainty associated with the handling of
the samples.
were equilibrated under isothermal-isobaric ensemble conditions
for 700 ps at 300 K using the Nosꢁ–Hoover thermostat and isotrop-
ic barostat with time constants of 0.5 and 2 ps, respectively. Fur-
ther simulation runs of 300 ps each and a total simulation time
above 1.6 ns were used to produce equilibrated systems at the
studied temperature. Electrostatic interactions were treated using
the Ewald summation method considering six reciprocal-space vec-
tors, and repulsive-dispersive interactions were explicitly calculated
below a cutoff distance of 1.6 nm (long-range corrections were ap-
plied assuming the system has an uniform density beyond that
cutoff radius). Details concerning this type of simulation can be
found elsewhere.[51,52]
Densimetry
The density measurements were performed at atmospheric pres-
sure with an Anton Paar vibrating tube densimeter, model
DMA 5000, at temperatures ranging from 293 K to 363 K. The
densimeter was internally calibrated by measuring the densities of
atmospheric air and bi-distilled water, according to the manufac-
turer’s recommendations. The DMA 5000 cell is embedded in
a cavity inside a metal block, controlled for temperature by several
Peltier units resulting in a temperature stability better than ꢁ2 mK
for periods over 10 min. Under these operating conditions, the
density measurement was repeatability better than 0.04 kgmꢀ3
Acknowledgements
The work was partially supported by a grant from Iceland, Liech-
tenstein and Norway through the EEA financial mechanism (Proj-
ect PT015). Financial support was also provided by Fundażo
para a CiÞncia e Tecnologia (FCT) through projects PEst-OE/EQB/
LA0004/2011, PTDC/CTM/73850/2006, PTDC/QUI/71331/2006,
PTDC/QUI/72903/2006 and PTDC/QUI-QUI/101794/2008. J.M.S.S.E.
and I.M.M. acknowledge FCT for a contract under Programa
CiÞncia 2007. K.S. acknowledges a post-doctoral grant SFRH/BPD/
38339/2007. The National NMR Network (REDE/1517/RMN/2005),
was supported by POCI 2010 and Fundażo para a CiÞncia
e a Tecnologia.
and the expanded uncertainty was estimated to be ꢁ0.3 kgmꢀ3
.
All reported density data were corrected for the viscosity effect
using the internal calibration of the densimeter.
To prevent any contamination by water or air bubbles, we filled
a syringe with (ca. 2 cm3) of the liquid under nitrogen flow and in-
jected it into the densimeter. Three samples (from different sec-
tions of the syringe) were measured at 293 K and if the density
values agreed within 0.04 kgmꢀ3, the last sample was used to mea-
sure the density for the entire temperature range.
Keywords: choline
· density · ionic liquids · molecular
dynamics simulations · viscosity
Viscosimetry
Measurements of viscosity were performed using an automated
SVM 3000 Anton Paar rotational Stabinger viscometer-densimeter.
The device uses Peltier elements for fast and efficient thermostati-
zation. Temperature uncertainty is ꢁ0.02 K. The precision of the
dynamic viscosity measurements is ꢁ0.5%. The overall uncertainty
of the measurements (taking into account the purity and handling
of the samples) is estimated to be 2%.[49]
[1] J. K. Blusztajn, Science 1998, 281, 794–795.
[2] J. P. Jones III, W. H. Meck, C. L. Williams, W. A. Wilson, H. S. Swartzwelder,
Dev. Brain Res. 1999, 118, 159–167.
[3] W. H. Meck, C. L. Williams, Dev. Brain Res. 1999, 118, 51–59.
[4] M. Smiglak, W. M. Reichert, J. D. Holbrey, J. S. Wilkes, L. Y. Sun, J. S.
Thrasher, K. Kirichenko, S. Singh, A. R. Katritzky, R. D. Rogers, Chem.
Commun. 2006, 2554–2556.
ChemPhysChem 0000, 00, 1 – 9
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemphyschem.org
&
7
&
ÞÞ
These are not the final page numbers!

I. M. Marrucho, J. M. S. S. EsperanÅa, J. N. Canongia Lopes, et al.
[5] L. P. N. Rebelo, J. N. C. Lopes, J. M. S. S. EsperanÅa, E. Filipe, J. Phys.
Chem. B 2005, 109, 6040–6043.
[30] A. P. Abbott, D. Boothby, G. Capper, D. L. Davies, R. K. Rasheed, J. Am.
Chem. Soc. 2004, 126, 9142–9147.
[6] M. J. Earle, J. M. S. S. EsperanÅa, M. A. Gilea, J. N. C. Lopes, L. P. N.
Rebelo, J. W. Magee, K. R. Seddon, J. A. Widegren, Nature 2006, 439,
831–834.
[7] J. M. S. S. EsperanÅa, J. N. C. Lopes, M. Tariq, L. M. N. B. F. Santos, J. W.
Magee, L. P. N. Rebelo, J. Chem. Eng. Data 2010, 55, 3–12.
[8] N. Gathergood, M. T. Garcia, P. J. Scammells, Green Chem. 2004, 6, 166–
175.
[31] A. P. Abbott, G. Capper, D. L. Davies, R. K. Rasheed, Chem. Eur. J. 2004,
10, 3769–3774.
[32] H. Zhao, G. A. Baker, S. Holmes, Org. Biomol. Chem. 2011, 9, 1908–1916.
[33] L. P. N. Rebelo, V. Najdanovic-Visak, R. Gomes de Azevedo, J. M. S. S. Es-
peranÅa, M. Nunes da Ponte, H. J. R. Guedes, Z. P. Visak, H. C. de Sousa,
J. Szydlowski, J. N. C. Lopes, T. C. Cordeiro in Ionic Liquids, IIIA: Funda-
mentals, Progress, Challenges, and Opportunities-Properties and Structure,
ACS Symposium Series 901 (Eds.: R. D. Rogers, K. R. Seddon), American
Chemical Society, Washington DC, 2005, pp. 270–291.
[34] J. M. S. S. EsperanÅa, H. J. R. Guedes, M. Blesic, L. P. N. Rebelo, J. Chem.
Eng. Data 2006, 51, 237–242.
[35] L. P. N. Rebelo, J. N. C. Lopes, J. M. S. S. EsperanÅa, H. J. R. Guedes, J.
Lachwa, V. Najdanovic-Visak, Z. P. Visak, Acc. Chem. Res. 2007, 40, 1114 –
1121.
´
[9] U. Domanska, Thermochim. Acta 2006, 448, 19–30.
[10] S. Stolte, J. Arning, U. Bottin-Weber, M. Matzke, F. Stock, K. Thiele, M.
Uerdingen, B. Jastorff, J. Ranke, Green Chem. 2006, 8, 621–629.
[11] S. Stolte, J. Arning, U. Bottin-Weber, A. Mꢅller, W. R. Pitner, U. Welz-Bier-
mann, B. Jastorff, J. Ranke, Green Chem. 2007, 9, 760–767.
[12] P. Nockemann, B. Thijs, K. Driesen, C. R. Janssen, K. V. Hecke, L. V. Meer-
velt, S. Kossmann, B. Kirchner, K. Binnemans, J. Phys. Chem. B 2007, 111,
5254–5263.
[36] C. Ye, J. M. Shreeve, J. Phys. Chem. A 2007, 111, 1456–1461.
[37] R. L. Gardas, J. A. P. Coutinho, Fluid Phase Equilib. 2008, 263, 26–32.
[38] R. L. Gardas, J. A. P. Coutinho, AIChE J. 2009, 55, 1274–1287.
[13] X. Wang, C. A. Ohlin, Q. Lu, Z. Fei, J. Hub, P. J. Dyson, Green Chem. 2007,
9, 1191–1197.
´
[14] Y. Yu, X. Lu, Q. Zhou, K. Dong, H. Yao, S. Zhang, Chem. Eur. J. 2008, 14,
11174–11182.
[39] U. Domanska, A. Marciniak, M. Krolikowski, J. Phys. Chem. B 2008, 112,
1218–1225.
[15] M. Petkovic, J. L. Ferguson, H. K. N. Gunaratne, R. Ferreira, M. C. Leit¼o,
K. R. Seddon, L. P. N. Rebelo, C. P. Pereira, Green Chem. 2010, 12, 643–
649.
[16] K. D. Weaver, H. J. Kim, J. Sun, D. R. MacFarlane, G. D. Elliott, Green
Chem. 2010, 12, 507–513.
[17] S. Abellꢆ, F. Medina, X. Rodrꢂguez, Y. Cesteros, P. Salagre, J. E. Sueiras, D.
Tichit, B. Coq, Chem. Commun. 2004, 1096–1097.
[18] A. P. Abbott, T. J. Bell, S. Handa, B. Stoddart, Green Chem. 2005, 7, 705–
707.
[19] S. Hu, T. Jiang, Z. Zhang, A. Zhu, B. Han, J. Song, Y. Xie, W. Li, Tetrahe-
dron Lett. 2007, 48, 5613–5617.
[20] P. Moriel, E. J. Garcꢂa-Suꢀrez, M. Martꢂnez, A. B. Garcꢂa, M. A. Montes-
Morꢀn, V. Calvino-Casilda, M. A. BaÇares, Tetrahedron Lett. 2010, 51,
4877–4881.
[21] J. Gorke, F. Srienc, R. Kazlauskas, Biotechnol. Bioprocess Eng. 2010, 15,
40–53.
[40] M. Tariq, A. P. Serro, J. L. Mata, B. Saramago, J. M. S. S. EsperanÅa, J. N. C.
Lopes, L. P. N. Rebelo, Fluid Phase Equilib. 2010, 294, 157–179.
[41] H. J. R. Guedes, T. C. Cordeiro, J. N. C. Lopes, J. M. S. S. EsperanÅa, L. P. N.
Rebelo, S. Huq, K. R. Seddon, J. Phys. Chem. B 2005, 109, 3519–3525.
[42] A. B. Pereiro, H. I. M. Veiga, J. M. S. S. EsperanÅa, A. Rodrꢂguez, J. Chem.
Thermodyn. 2009, 41, 1419–1423.
[43] A. J. L. Costa, J. M. S. S. EsperanÅa, I. M. Marrucho, L. P. N. Rebelo, J.
Chem. Eng. Data 2011, 56, 3433–3441.
[44] M. M. Cruz, R. P. Borges, M. Godinho, C. S. Marques, M. J. V. LourenÅo, M.
Macatr¼o, C. A. Nieto de Castro, M. Tariq, J. M. S. S. EsperanÅa, J. N. C.
Lopes, C. A. M. Afonso, L. P. N. Rebelo, Magnetic and thermophysical
studies of two paramagnetic liquid salts: [bmim][FeCl4] and
[P6,6,6,14][FeCl4]. in preparation.
[45] M. Blesic, M. Swadzba-Kwasny, T. Belhocine, H. Q. N. Gunaratne, J. N. C.
Lopes, M. F. C. Gomes, A. A. H. Pꢀdua, K. R. Seddon, L. P. N. Rebelo, Phys.
Chem. Chem. Phys. 2009, 11, 8939–8948.
[22] M. Avalos, R. Babiano, P. Cintas, J. L. Jimenez, J. C. Palacios, Angew.
Chem. 2006, 118, 4008–4012; Angew. Chem. Int. Ed. 2006, 45, 3904–
3908.
[23] A. P. Abbott, G. Frisch, J. Hartley, K. S. Ryder, Green Chem. 2011, 13,
471–481.
[24] R. Vijayaraghavan, B. C. Thompson, D. R. MacFarlane, R. Kumar, M. Suria-
narayanan, S. Aishwary, P. K. Sehgal, Chem. Commun. 2010, 46, 294–
296.
[46] F. S. Oliveira, M. C. Freire, P. I. Carvalho, J. A. P. Coutinho, J. N. C. Lopes,
L. P. N. Rebelo, I. M. Marrucho, J. Chem. Eng. Data 2010, 55, 4514–4520.
[47] J. N. A. Canongia Lopes, A. A. H. Pꢀdua, J. Phys. Chem. B 2006, 110,
3330–3335.
[48] K. Shimizu, A. A. H. Pꢀdua, J. N. C. Lopes, J. Phys. Chem. B 2010, 114,
15635–15641.
[49] M. Tariq, P. J. Carvalho, J. A. P. Coutinho, I. M. Marrucho, J. N. C. Lopes,
L. P. N. Rebelo, Fluid Phase Equilib. 2011, 301, 22–32.
[25] K. Fujita, D. R. MacFarlane, M. Forsyth, M. Yoshizawa-Fujita, K. Murata, N.
Nakamura, H. Ohno, Biomacromolecules 2007, 8, 2080–2086.
[50] W. Smith, T. R. Forester, I. T. Todorov, The DL-POLY 2 User Manual, Ver-
sion 2.20; STFC Daresbury Laboratory: Warrington, U. K., 2009.
[51] J. N. Canongia Lopes, A. A. H. Pꢀdua, J. Phys. Chem. B 2004, 108, 16893–
16898.
´
[26] U. Domanska, A. Pobudkowska, M. Krꢆlikowski, Fluid Phase Equilib.
2007, 259, 173–179.
[27] P. Nockemann, K. Binnemans, B. Thijs, T. N. Parac-Vogt, K. Merz, A. V.
Mudring, P. C. Menon, R. N. Rajesh, G. Cordoyiannis, J. Thoen, J. Leys, C.
Glorieux, J. Phys. Chem. B 2009, 113, 1429–1437.
[52] K. Shimizu, D. Almantariotis, M. F. Costa Gomes, A. A. H. Pꢀdua, J. N. C.
Lopes, J. Phys. Chem. B 2010, 114, 3592–3600.
[28] A. P. Abbott, G. Capper, D. L. Davies, R. K. Rasheed, V. Tambyrajah, Chem.
Commun. 2003, 70–71.
[29] A. P. Abbott, R. C. Harris, K. S. Ryder, C. D’Agostino, L. F. Gladden, M. D.
Mantle, Green Chem. 2011, 13, 82–90.
Received: October 26, 2011
Revised: January 20, 2012
Published online on && &&, 2012
&8
&
www.chemphyschem.org
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPhysChem 0000, 00, 1 – 9
ÝÝ
These are not the final page numbers!

ARTICLES
A. J. L. Costa, M. R. C. Soromenho,
K. Shimizu, I. M. Marrucho,*
J. M. S. S. EsperanÅa,* J. N. C. Lopes,*
L. P. N. Rebelo
&& – &&
Density, thermal expansion coefficient
and viscosity data are reported for
cholinium-based ionic liquids of the N-
alkyl-N,N-dimethyl-N-(2-hydroxyethyl)-
ammonium bis(trifluoromethylsulfon-
yl)imide series (see picture). The outlier
character of the first member of the ho-
mologous series is interpreted at a mo-
lecular level by means of MD simula-
tions.
Density, Thermal Expansion and
Viscosity ofCholinium-Derived Ionic
Liquids
ChemPhysChem 0000, 00, 1 – 9
ꢃ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemphyschem.org
&
9
&
ÞÞ
These are not the final page numbers!