Fluid Phase Equilibria 335 (2012) 32–38

Contents lists available at SciVerse ScienceDirect

Fluid Phase Equilibria

journal homepage: www.elsevier.com/locate/fluid

Densities of a deep eutectic solvent based on choline chloride and glycerol and its
aqueous mixtures at elevated pressures
Rhoda B. Leron a , David Shan Hill Wong b , Meng-Hui Li a,∗
a
R&D Center for Membrane Technology and Department of Chemical Engineering, Chung Yuan Christian University, Chung Li 32023, Taiwan, ROC
b
Department of Chemical Engineering, National Tsing-Hua University, Hsinchu, Taiwan, ROC

a r t i c l e i n f o a b s t r a c t

Article history: Deep eutectic solvents are cheap and biodegradable alternatives to conventional ionic liquids and volatile
Received 4 May 2012 organic solvents. In this work, we present new measurements on the densities of the deep eutectic solvent
Received in revised form 9 August 2012 choline chloride:glycerol (GLY) and its aqueous mixtures in the temperature range (298.15–323.15) K
Accepted 20 August 2012
and pressures up to 50 MPa. The density was measured using a vibrating-tube density meter and the
Available online 28 August 2012
experimental uncertainty was estimated to be ±0.1%. Density data were used to derive other properties
such as isothermal compressibility, isobaric expansivity, and excess molar volume. The excess molar
Keywords:
volumes of the mixtures of GLY (1) + H2 O (2) were investigated and an empirical equation was used to
Deep eutectic solvent
Excess molar volume
represent their dependence on temperature, pressure, and composition. The density data at elevated
Glyceline pressures were correlated with the temperature, pressure, and composition by a Tait-type equation.
High-pressure density © 2012 Elsevier B.V. All rights reserved.
Tait-type equation

1. Introduction
Due to their unique properties, ionic liquids (ILs) are regarded
as green replacements to volatile solvents in many electrochemi-
cal, synthetic, analytical, and engineering processes [1]. However,
the disadvantage is they are costly; thus, they have not yet been
widely adapted for industrial applications [2]. Also, certain types of
ILs (i.e. imidazolium-based) are non-biodegradable and toxic [3–5].
Thus, researchers continue to search for new and even greener
alternatives to ILs that may overcome the said limitations.
An alternative class of ILs called deep eutectic solvents (DESs),
which are formed by the complexation of a substituted quater-
nary ammonium salt with hydrogen-bond-forming compounds,
has recently been investigated by Abbott et al. [6,7]. When mixed
at the correct molar ratio, these compounds form a mixture with a
substantially low melting point [7,8]. For example, the DES formed
from choline chloride (m.p. = 302 ◦ C) and glycerol (m.p. = 18 ◦ C) at
1:2 molar ratio (trade name glyceline) has a melting point of −40 ◦ C
[9]. DESs have also shown to have solvent properties similar to
conventional ILs’, including negligible volatility, high conductivity,
non-flammability, high thermal and chemical stability, and high
solvation capacity. But compared to ILs, DESs are desirable for large-
scale applications because they can be prepared with high purity at
lower costs [7]. DESs can easily be prepared by simply mixing their
∗ Corresponding author. Tel.: +886 3 265 4109; fax: +886 3 265 4199.
E-mail address: mhli@cycu.edu.tw (M.-H. Li).
individual components with gentle heating. Some DESs are already
being processed in bulk or in ton scale and available for commercial
use [10]. They are generally less moisture-sensitive compared to ILs
[8] and many of them are made from biodegradable components,
which exhibit low or non-toxicity [7,11].
Some DESs are currently being utilized in the electrodeposition,
electroplating, and electropolishing of metals while their potential
for other applications are also being explored [10]. Certain types
of DESs have been reported to be promising solvents in enzyme-
based biotransformations [12,13], solubilization of poorly soluble
compounds for early drug development [14], fabrication of nano-
structured Ni and composite films [15,16], and absorption of CO2
[17,18]. The use of glycerol-based DES as a promising solvent in
the efficient extraction of glycerol from biodiesel has also been
demonstrated by several authors [9,19–21].
To assess and design the possible industrial-scale applications
of DESs, it is imperative that their thermophysical properties are
known. Among them is the density, which is an important phy-
sical property of any material. Very few data on the density of
DESs have been reported in the literature and measurements are
limited at atmospheric condition [22–24]. The objective of this
work is to contribute to the database of thermodynamic proper-
ties of DESs which could be useful for future applications. Thus,
we report new measurements on the densities of the deep eutec-
tic solvent 1 mole choline chloride:2 moles glycerol (GLY) and its
aqueous mixtures for pressures from (0.1 to 50) MPa and tempera-
tures (298.15–323.15) K. Ten different compositions (x1 = 0.1–1.0)
were considered. A Tait-type equation was used to correlate the

0378-3812/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.fluid.2012.08.016
 R.B. Leron et al. / Fluid Phase Equilibria 335 (2012) 32–38
high-pressure density with temperature, pressure, and composi-
tion. The isothermal compressibilities and isobaric expansivities
of GLY were derived from experimental density data. The excess
molar volumes, VE , of the mixtures of GLY (1) + H2 O (2) were inves-
tigated, and an empirical equation proposed by Lugo et al. [25] was
used to represent VE as a function of temperature, pressure, and
composition.
2. Materials and methods
2.1. Materials
The deep eutectic solvent glyceline 200 (purity >0.98 mass frac-
tion), a mixture of choline chloride + glycerol (1:2 molar ratio),
was obtained from Scionix Ltd® . The water content of the sample
was measured using a Mettler Toledo Karl–Fischer (model DL31)
titrator, and was found to be <0.005 (mass fraction). The sodium
chloride used in the preparation of calibration solutions was Ame-
rican Chemical Society (ACS)-certified grade and was obtained from
J.T. Baker. The ionic liquid 1-butyl-3-methylimidazolium tetrafluo-
roborate, [Bmim][BF4 ], (purity >0.98 mass fraction) was supplied by
Tokyo Chemical Co., Ltd. Prior to use, the pure [Bmim][BF4 ] sample
was dried at 343 K for at least 48 h in an evacuated oven in order to
remove any volatile impurities and to reduce the water content. The
final water content of the sample was 392 ppm. Distilled deionized
water (Type I reagent-grade; resistivity = 18.3 M cm), which was
processed in a Barnstead Thermolyne (model Easy Pure 1052) water
system, was used in all the experiments. All aqueous solutions were
prepared from degassed samples and weight measurements were
done using a Mettler Toledo (model AL204) digital balance with
an accuracy of ±1 × 10−4 g. The description of the chemicals used
together with their sources is provided in Table 1.
2.2. Density measurements
Density measurements were performed using an Anton Paar
(model DMA HP) vibrating-tube density meter connected to a
model DMA 5000M master instrument. The DMA HP measured the
period of harmonic oscillation of a built-in U-tube containing the
sample. The temperature in the measuring cell was controlled by
an integrated Peltier thermostat. The accuracy of the temperature
measurement was ±0.05 K. The pressure was created using a pres-
sure generator from High Pressure Equipment Company (HIP model
#62-6-10), and was monitored with a GE Druck pressure gauge
(model DPI 104, 0–100 MPa) having an accuracy of 0.05% FS. The
period of oscillation, cell temperature, and pressure were recorded
using the DMA 5000 M master instrument which was connected
to the DMA HP via the DMA HP interface. All high pressure valves
and fittings were supplied by High Pressure Equipment Company
(USA) and Nova Swiss (Switzerland). A detailed schematic repre-
Table 1
Description of the chemicals used in this study.
Chemical Purity (% mass) Source
1-Butyl-3- >98 Tokyo Chemical
methylimidazolium Industry Co., Ltd.
tetrafluoroborate
[Bmim][BF4 ]
Glyceline 200 (GLY) >98 Scionix Ltd.
(1 mole choline
chloride + 2 moles
glycerol)
Sodium chloride (ACS 100.2 (NaCl) J.T. Baker
reagent Baker
analyzed)
33
sentation of the high pressure density installation can be found in
our previous work [26].
Before each experiment, the system was rinsed with deionized
water and ethanol, dried with nitrogen gas, and slowly evacuated
for at least 3 h. Then, the inlet valve was opened to fill the sys-
tem and the pressure generator cell with the sample, and the valve
was closed. To avoid any trapped air in the system, the outlet valve
was slightly opened and a few milliliters of the sample was allo-
wed to flow out by moving the pressure generator piston forward.
The outlet valve was then closed, and the system was brought to
the desired temperature and pressure. The temperature was allo-
wed to stabilize for at least 30 min for a 5-K interval. Upon thermal
and mechanical equilibrations, the period of oscillation was recor-
ded using the DMA 5000M master instrument, which was then
converted to the density.
The density meter was calibrated using distilled deionized water
and NaCl (aq) solutions of molalities (1 and 3) mol kg−1 . NaCl (aq)
solutions were used in this work because they have been recom-
mended by several authors as high-density calibration fluids for
high-pressure density meters [27–29]. The calibration was carried
out in the temperature range (298.15–323.15) K and over the pres-
sure range (0.1–50) MPa and a total of 198 data points were used.
The reference data reported by Wagner and Pruss [30] for water
and by Archer [31] for NaCl (aq) were used to correlate the density
with the temperature and pressure using the equation

= ajkl pj T k P l . (1)
j,k,l
where  is the reference density; p is the oscillation period; T is
the temperature; P is the pressure; and ai,j,k are fitting parameters,
with j = 0, 2; k = 0, 1, 2; and l = 0, 1, 2. The average absolute deviation
(AAD) between the reference and the calculated data was 0.003%.
The AAD is defined as
 
1  i,ref − i,cal 
N
AAD% = 100 × . (2)
N i,ref
i
where i,ref is the ith reference density data and i,cal is the den-
sity calculated by Eq. (1). The calibration was tested by measuring
the densities of NaCl (aq) solutions with concentrations (2 and
4.5) mol kg−1 within the considered temperature and pressure
ranges and by comparing them with the reference data taken from
Archer [31]. The AAD between the measured and the reference data
is less than 0.02%, which is well within the error of the experiment.
The calibration was further tested by measuring the densities of
[Bmim][BF4 ] at the considered temperature and pressure ranges.
This IL has been extensively studied, and the densities at varied
range of temperatures and pressures are well-reported in the lite-
rature. Our measured values agree very well with those reported
by Zafarani-Moattar and Shekaari [32], Sanmamed et al. [33] and
Soriano et al. [34] at 1 atm with AAD not exceeding 0.06%. Our
results are also close to the data reported by Gardas et al. [35]
(AAD = 0.04%) and the calculated values from the correlation pro-
posed by Matkowska and Hofman [36] (AAD = 0.02%) for different
range of temperatures and pressures. The measured densities are
in between the values calculated from the correlations proposed
by Jacquemin et al. [27] (AAD = 0.16%) and those calculated from
Rebelo et al. [37] (AAD = 0.28%). These deviations, which are typi-
cal of ILs, may be mainly due to the differences in the purities of
the samples used. The results of the validation experiments serve
to confirm the correctness of the apparatus calibration, accuracy
of the method as well as the operation of the equipment, and the
validity of the reported measurements.
Anton Paar DMA HP does not possess a built-in correction sys-
tem for viscosity effects. Such correction has been recommended
34 R.B. Leron et al. / Fluid Phase Equilibria 335 (2012) 32–38

Table 2 1.21
Experimental densities and calculated isothermal compressibilities and isobaric
expansivities of GLYa as functions of temperature and pressure.

P (MPa) T (K)

298.15 303.15 308.15 313.15 318.15 323.15

1.20
b −3
 (g cm )
0.1 1.1921 1.1892 1.1862 1.1834 1.1805 1.1777
5.0 1.1936 1.1907 1.1877 1.1849 1.1820 1.1792
10.0 1.1950 1.1921 1.1892 1.1863 1.1835 1.1806
15.0 1.1963 1.1935 1.1906 1.1877 1.1849 1.1821

ρ / g.cm -3
20.0 1.1977 1.1948 1.1920 1.1891 1.1863 1.1835 1.19
25.0 1.1991 1.1962 1.1933 1.1905 1.1877 1.1849
30.0 1.2004 1.1976 1.1947 1.1919 1.1891 1.1863
35.0 1.2017 1.1989 1.1960 1.1933 1.1904 1.1877
40.0 1.2030 1.2002 1.1974 1.1946 1.1918 1.1891
45.0 1.2043 1.2015 1.1987 1.1959 1.1931 1.1904
50.0 1.2056 1.2028 1.2000 1.1972 1.1944 1.1917 1.18
T (GPa−1 )
0.1 0.245 0.247 0.250 0.252 0.255 0.258
5.0 0.240 0.243 0.245 0.248 0.251 0.253
10.0 0.237 0.239 0.241 0.244 0.246 0.249
15.0 0.233 0.235 0.237 0.240 0.242 0.245
20.0 0.229 0.231 0.233 0.236 0.238 0.240
1.17
25.0 0.225 0.228 0.230 0.232 0.234 0.237 0 10 20 30 40 50
30.0 0.222 0.224 0.226 0.228 0.230 0.233
35.0 0.219 0.221 0.223 0.225 0.227 0.229 P / MPa
40.0 0.215 0.217 0.219 0.221 0.223 0.225
45.0 0.212 0.214 0.216 0.218 0.220 0.222 Fig. 1. Density of GLY as a function of pressure at different temperatures: (this
50.0 0.209 0.211 0.213 0.215 0.217 0.219 work) 䊉, 298.15 K; , 303.15 K; , 308.15 K; , 313.15 K; , 318.15 K; , 323.15 K;
103 ˛P (K−1 ) , Shahbaz et al. [22]; lines, calculated by the Tait-type equation, Eq. (3).
0.1 0.497 0.493 0.489 0.485 0.480 0.476
5.0 0.495 0.491 0.486 0.482 0.478 0.473
that a good agreement exists between the two sets of data with an
10.0 0.492 0.488 0.484 0.479 0.475 0.471
15.0 0.490 0.486 0.481 0.477 0.473 0.468
AAD of 0.03%, which is well within the uncertainty of the measure-
20.0 0.488 0.483 0.479 0.475 0.470 0.466 ments. In our previous work [24], we also reported density data for
25.0 0.486 0.481 0.477 0.472 0.468 0.463 GLY at atmospheric pressure. We found that these data are consis-
30.0 0.484 0.479 0.475 0.470 0.466 0.461 tently lower than the present measurements (at 0.1 MPa) having
35.0 0.482 0.477 0.473 0.468 0.463 0.459
an AAD of 0.04%, and that the relative deviations tend to decrease
40.0 0.480 0.475 0.471 0.466 0.461 0.457
45.0 0.478 0.473 0.469 0.464 0.459 0.455 at higher temperatures. The previous measurements were perfor-
50.0 0.476 0.471 0.467 0.462 0.457 0.453 med using an Anton Paar DMA 5000M instrument which possessed
a
GLY = 1 mole choline chloride + 2 moles glycerol.
a built-in correction system for the effect of the liquid’s viscosity
b
Uncertainty = ±0.1%. and calibrated directly in Anton Paar. Therefore, we can say that the
observed deviations can be attributed to viscosity effects. This also
suggests that viscosity corrections would indeed be necessary if a
by many authors for density measurements obtained by vibrating- general correlation for such and viscosity of GLY would be available
tube density meters to account for the differences between the later on. To our knowledge, no literature data have been repor-
viscosity of the samples and those of the calibrating fluids [38–40]. ted on the densities of the studied DES at high pressures and the
Several authors used an empirical approach to account for such considered temperatures; thus, no further comparison was made.
effect using information provided by the manufacturer (i.e. for The experimental densities of GLY were represented as a func-
Anton Paar DMA 512P and DMA HPM). Unfortunately, there is no tion of temperature and pressure by a Tait-type equation of the
general correlation available for the DMA HP model. Anton Paar form
provides a graphical representation of the influence of viscosity on
0 (T )
the density measured by the instrument, which includes a rather (T, P) = . (3)
few data points [40,41]. However, the viscosity data required for 1 − C ln (B(T ) + P/B(T ) + Pref )
the density correction are not yet available for the studied systems. where 0 is the reference density; Pref is the reference pressure
Thus, no corrections on the measured densities have been perfor- (0.1 MPa); and C, B(T) are adjustable parameters determined by
med and raw density data were reported. The overall uncertainty fitting the data into Eq. (3) by applying the Marquardt method.
of the density measurements was estimated to be within ±0.1%. The parameter C was assumed to be independent of temperature
while B was assumed to be temperature-dependent. The measu-
3. Results and discussion red densities of the DES at the reference pressure were fitted to a
second-order equation of the form
The experimental densities of GLY as functions of pressure
0 = a0 + a1 T + a2 T 2 . (4)
and temperature are tabulated in Table 2. Results show that
the densities decrease as the temperature increases (at constant The fitting parameters were found to be a0 = 1.4260,
pressure) and increase as the pressure increases (at constant tem- a1 = −9.7598 × 10−4 and a2 = 6.4286 × 10−7 with an AAD of
perature). These behaviors are expected because the density of a 0.002%. The temperature-dependent parameter B(T) in Eq. (3) is
substance generally decreases at higher temperatures due to ther- expressed as
mal expansion while at increasing pressure density increases due
B(T ) = b0 + b1 T. (5)
to compression.
The densities of GLY at 0.1 MPa presented in this work were com- The values of the parameters b0 , b1 and C are presented in
pared to those reported by Shahbaz et al. [22,23]. It is shown in Fig. 1 Table 3, along with the corresponding AAD%. Like in many liquid
 R.B. Leron et al. / Fluid Phase Equilibria 335 (2012) 32–38 35

Table 3 Table 4
Fitting parameters bi and ci of the Tait-type equation, Eq. (3). Experimental density data for GLY (1) + H2 O (2) mixtures as functions of pressure,
temperature, and concentration.
System i Parameters AADa (%)
P (MPa) T (K)
bi ci
298.15 303.15 308.15 313.15 318.15 323.15
GLY 0 438.5 0.0666 0.003
1 −0.5572 x1 = 0.1002 a

GLY (1) + H2 O (2) 0 532.6 0.1279 0.02 0.1 1.0742b 1.0720 1.0697 1.0674 1.0648 1.0623
1 −0.5395 −0.1317 5.0 1.0759 1.0737 1.0714 1.0691 1.0666 1.0640
2 0.0999 10.0 1.0776 1.0754 1.0731 1.0707 1.0682 1.0657
 

N
i,exp −i,cal  15.0
20.0
1.0793
1.0810
1.0771
1.0788
1.0748
1.0766
1.0724
1.0741
1.0699
1.0716
1.0674
1.0691
1
a
mart AAD = 100 × N i,exp
.
25.0 1.0827 1.0805 1.0782 1.0759 1.0734 1.0709
i 30.0 1.0844 1.0822 1.0799 1.0775 1.0751 1.0726
35.0 1.0859 1.0837 1.0814 1.0791 1.0766 1.0742
40.0 1.0877 1.0854 1.0832 1.0808 1.0784 1.0759
systems, the parameter B(T) was found to be a decreasing function 45.0 1.0893 1.0870 1.0848 1.0824 1.0800 1.0775
of temperature [42]. The densities of GLY were accurately correla- 50.0 1.0909 1.0886 1.0864 1.0840 1.0816 1.0791
ted with the temperature and pressure by the Tait-type equation at x1 = 0.2001
an AAD of 0.003%. The calculated values are shown as lines in Fig. 1. 0.1 1.1150 1.1125 1.1099 1.1073 1.1046 1.1019
5.0 1.1165 1.1140 1.1114 1.1088 1.1061 1.1034
The experimental densities of GLY were also used to derive other 10.0 1.1180 1.1156 1.1130 1.1104 1.1077 1.1050
important properties such as isothermal compressibility and iso- 15.0 1.1195 1.1171 1.1145 1.1119 1.1092 1.1065
baric expansivity. These can provide important information on the 20.0 1.1209 1.1185 1.1160 1.1134 1.1107 1.1080
dependence of the volumetric properties on temperature and pres- 25.0 1.1224 1.1199 1.1174 1.1149 1.1122 1.1095
30.0 1.1239 1.1215 1.1190 1.1164 1.1138 1.1110
sure. The isothermal compressibility, T , can describe the effect of
35.0 1.1253 1.1228 1.1203 1.1178 1.1151 1.1125
pressure on the density based on the equation 40.0 1.1269 1.1244 1.1219 1.1194 1.1168 1.1141
  45.0 1.1283 1.1259 1.1234 1.1208 1.1182 1.1155
1 ∂
T = . (6) 50.0 1.1297 1.1273 1.1248 1.1222 1.1196 1.1169
 ∂P x1 = 0.3000
T
0.1 1.1395 1.1369 1.1342 1.1315 1.1287 1.1259
The isobaric expansivity, ˛P , is related to the variation of the den- 5.0 1.1409 1.1383 1.1356 1.1329 1.1301 1.1273
sity with the temperature (at constant pressure) which is described 10.0 1.1424 1.1398 1.1371 1.1343 1.1316 1.1288
as 15.0 1.1438 1.1412 1.1385 1.1358 1.1331 1.1303
  20.0 1.1452 1.1426 1.1400 1.1372 1.1345 1.1317
1 ∂ 25.0 1.1466 1.1440 1.1413 1.1387 1.1360 1.1333
˛P = − . (7) 30.0 1.1480 1.1454 1.1428 1.1401 1.1374 1.1347
 ∂T
P 35.0 1.1494 1.1468 1.1441 1.1415 1.1388 1.1361
40.0 1.1508 1.1482 1.1456 1.1429 1.1402 1.1376
In this work, the isothermal compressibilities of GLY at different
45.0 1.1521 1.1495 1.1469 1.1442 1.1415 1.1388
temperatures and pressures were calculated by taking the deriva- 50.0 1.1534 1.1508 1.1482 1.1456 1.1429 1.1402
tive of Eq. (3). Since the Tait equation is an integrated form of an x1 = 0.3998
empirical equation that represents the isothermal compressibility 0.1 1.1555 1.1528 1.1500 1.1473 1.1445 1.1417
as a function of pressure [43], T can be conveniently derived from 5.0 1.1569 1.1542 1.1514 1.1487 1.1459 1.1431
10.0 1.1583 1.1556 1.1528 1.1501 1.1473 1.1445
Eq. (3) using Eq. (4), together with the parameters given in Table 3. 15.0 1.1597 1.1570 1.1543 1.1515 1.1488 1.1460
The resulting expression is 20.0 1.1611 1.1584 1.1557 1.1529 1.1502 1.1474
C
 25.0 1.1624 1.1597 1.1570 1.1543 1.1516 1.1488
T = . (8) 30.0 1.1638 1.1612 1.1584 1.1557 1.1529 1.1501
B(T ) + P 0 35.0 1.1651 1.1625 1.1597 1.1571 1.1543 1.1516
40.0 1.1665 1.1639 1.1612 1.1585 1.1557 1.1530
It can be observed that the values of T in Table 2 increase
45.0 1.1677 1.1651 1.1624 1.1597 1.1570 1.1543
with increasing temperature and decrease with increasing pres- 50.0 1.1691 1.1664 1.1637 1.1610 1.1583 1.1556
sure. These trends are consistent with those observed in some ILs x1 = 0.5004
reported in the literature [27,43–45]. 0.1 1.1667 1.1639 1.1610 1.1583 1.1555 1.1527
Similarly, the isobaric thermal expansivities of GLY different 5.0 1.1681 1.1653 1.1625 1.1597 1.1569 1.1540
10.0 1.1695 1.1667 1.1639 1.1611 1.1583 1.1554
temperatures and pressures were obtained analytically by diffe- 15.0 1.1708 1.1681 1.1653 1.1625 1.1597 1.1569
rentiating the Tait-type equation together with Eqs. (4) and (5), 20.0 1.1722 1.1694 1.1667 1.1639 1.1611 1.1583
which leads to the relation 25.0 1.1736 1.1708 1.1680 1.1653 1.1625 1.1597
    30.0 1.1748 1.1721 1.1693 1.1666 1.1638 1.1610
d0 /dT dB/dT (P − 0.1)
˛P = − +C . (9) 35.0 1.1761 1.1734 1.1706 1.1679 1.1651 1.1624
0 [1 − C ln (B + P/B + 0.1)](B + P)(B + 0.1)
40.0 1.1774 1.1747 1.1722 1.1693 1.1665 1.1638
where d0 /dT = a1 + 2a2 T and dB/dT = b1 . The values of ˛P given in 45.0 1.1787 1.1760 1.1733 1.1706 1.1678 1.1651
Table 2 decrease with increasing temperature and pressure. Such 50.0 1.1800 1.1773 1.1746 1.1719 1.1691 1.1664
x1 = 0.5998
variation may be related to changes in the molecular distance or
0.1 1.1746 1.1718 1.1689 1.1662 1.1633 1.1605
free-volume within the liquid [46]. 5.0 1.1760 1.1732 1.1703 1.1675 1.1646 1.1618
The densities of the mixtures of GLY (1) + H2 O (2) as a function of 10.0 1.1774 1.1746 1.1717 1.1689 1.1661 1.1632
temperature, pressure, and composition are presented in Table 4. 15.0 1.1787 1.1760 1.1732 1.1703 1.1675 1.1647
The densities of the aqueous mixtures tend to decrease as the tem- 20.0 1.1801 1.1773 1.1745 1.1717 1.1689 1.1661
25.0 1.1815 1.1787 1.1758 1.1730 1.1702 1.1674
perature rises and increase as the pressure increases. The values 30.0 1.1827 1.1800 1.1772 1.1744 1.1716 1.1688
also increase as the mole fraction of GLY in the solution increases, 35.0 1.1841 1.1813 1.1785 1.1758 1.1729 1.1702
which is expected since the density of the DES is higher than that 40.0 1.1854 1.1826 1.1799 1.1771 1.1743 1.1716
of water. The present measurements obtained at 0.1 MPa were also 45.0 1.1866 1.1839 1.1812 1.1784 1.1756 1.1728
50.0 1.1879 1.1852 1.1825 1.1797 1.1769 1.1742
compared to the atmospheric density data reported in our previous
work where viscosity corrections were included [24]. The present
36 R.B. Leron et al. / Fluid Phase Equilibria 335 (2012) 32–38

Table 4 (Continued) 0.0

P (MPa) T (K)

298.15 303.15 308.15 313.15 318.15 323.15

x1 = 0.7001
0.1 1.1807 1.1778 1.1748 1.1720 1.1691 1.1663 -0.1
5.0 1.1821 1.1792 1.1762 1.1734 1.1705 1.1677
10.0 1.1835 1.1806 1.1777 1.1748 1.1720 1.1691
15.0 1.1848 1.1820 1.1791 1.1762 1.1734 1.1706
20.0 1.1861 1.1832 1.1804 1.1775 1.1747 1.1719

-1
V / cm. mol
25.0 1.1874 1.1846 1.1817 1.1789 1.1761 1.1733
30.0 1.1887 1.1860 1.1831 1.1803 1.1775 1.1747 -0.2

3
35.0 1.1900 1.1873 1.1844 1.1817 1.1788 1.1761
40.0 1.1913 1.1886 1.1858 1.1830 1.1802 1.1775

E
45.0 1.1926 1.1899 1.1871 1.1843 1.1815 1.1788
50.0 1.1939 1.1912 1.1884 1.1856 1.1828 1.1801
x1 = 0.8001
0.1 1.1853 1.1824 1.1794 1.1766 1.1737 1.1709 -0.3
5.0 1.1867 1.1838 1.1808 1.1780 1.1751 1.1723
10.0 1.1881 1.1852 1.1823 1.1794 1.1766 1.1737
15.0 1.1894 1.1866 1.1837 1.1808 1.1780 1.1752
20.0 1.1907 1.1878 1.1850 1.1821 1.1793 1.1765
25.0 1.1921 1.1892 1.1863 1.1835 1.1807 1.1779
30.0 1.1934 1.1906 1.1877 1.1849 1.1821 1.1793 -0.4
35.0 1.1947 1.1919 1.1890 1.1863 1.1834 1.1807 0.0 0.2 0.4 0.6 0.8 1.0
40.0 1.1960 1.1932 1.1904 1.1876 1.1848 1.1821 x1
45.0 1.1973 1.1945 1.1917 1.1889 1.1861 1.1834
50.0 1.1986 1.1958 1.1930 1.1902 1.1874 1.1847
x1 = 0.9000 Fig. 2. Excess molar volumes of GLY (1) + H2 O (2) mixtures as a function of x1
0.1 1.1890 1.1861 1.1831 1.1803 1.1774 1.1746 at 0.1 MPa and different temperatures: 䊉, 298.15 K; , 303.15 K; , 308.15 K; ,
5.0 1.1905 1.1876 1.1846 1.1818 1.1789 1.1761 313.15 K; , 318.15 K; , 323.15 K; lines, calculated by Eq. (11).
10.0 1.1918 1.1890 1.1861 1.1832 1.1804 1.1775
15.0 1.1931 1.1903 1.1874 1.1845 1.1817 1.1789
20.0 1.1945 1.1916 1.1888 1.1859 1.1831 1.1803
the excess molar volumes decrease (more positive) as the pressure
25.0 1.1959 1.1930 1.1901 1.1873 1.1845 1.1817 increases at constant temperature.
30.0 1.1972 1.1944 1.1915 1.1887 1.1859 1.1831 The values of VE were then correlated with the temperature,
35.0 1.1985 1.1957 1.1928 1.1901 1.1872 1.1845 pressure, and composition using an empirical equation proposed
40.0 1.1998 1.1970 1.1942 1.1914 1.1886 1.1859
by Lugo et al. [25]
45.0 1.2011 1.1983 1.1955 1.1927 1.1899 1.1872
50.0 1.2024 1.1996 1.1968 1.1940 1.1912 1.1885
 1
 
3 
3
a
Uncertainty in x1 = ±10−4 . VE = x1 x2 D ij (2x1 − 1) i−1 (T − T0 )j−1 . (11)
b
Uncertainty in  = ±0.1%. d0 + d1 P
i=1 i=j

data are systematically higher than those we reported earlier, with where T0 = 273.15 K and d0 , d1 , Dij are adjustable parameters pre-
a maximum deviation of 0.06%. Relatively higher deviations were sented in Table 5. The applied correlation successfully represented
found at low temperatures and high DES concentrations, where the
mixtures are expected to exhibit higher viscosities. 0.0
The excess molar volumes of the GLY (1) + H2 O (2) mixtures were
also investigated. For each mixture at a temperature and pressure,
VE was calculated using the equation


2
-0.1
VE = xi Mi (−1 − i−1 ). (10)
i=1

where  is the density of the mixture and i , xi , and Mi are the

-1

density, mole fraction, and molecular weight of component i (GLY

V / cm mol

or water), respectively. The molecular weight of GLY was calculated -0.2

3

from its individual components based on the method suggested by

Abbott et al. [9].
E

For all the mixtures, the excess molar volumes are negative over
the whole composition range at all temperatures and pressures
with minima virtually occurring at x1 ≈ 0.40. For representation, the -0.3
excess molar volumes are plotted as a function of mole fraction in
Figs. 2 and 3. These results imply increased densities of the mixture
compared with the ideal solution which has been observed in many
IL-solvent systems [47,48] and in our earlier work [24]. The varia-
-0.4
tion of the VE with composition for various isotherms at P = 0.1 MPa 0.0 0.2 0.4 0.6 0.8 1.0
is shown in Fig. 2. The values of VE decrease (more positive) with
x1
increasing temperature at constant pressure. This may probably be
due to the weakening of the hydrogen bonds in the solutions, resul-
Fig. 3. Excess molar volume of GLY (1) + H2 O (2) mixtures as a function of x1 at
ting from increased temperatures. Fig. 3 illustrates the variation of T = 323.15 K and different pressures: 䊉, 0.1 MPa; , 10 MPa; , 20 MPa; , 30 MPa;
VE with composition for various isobars at T = 323.15 K. As shown, , 40 MPa; , 50 MPa; lines, calculated by Eq. (11).
 R.B. Leron et al. / Fluid Phase Equilibria 335 (2012) 32–38 37

Table 5
Fitting parameters of Eq. (11) for the excess molar volumes of GLY (1) + H2 O (2)
mixtures.
1.20
i j Dij d0 102 d1 AAD (%)

1 1 −1.1836 0.8207 0.6207 1.1

2 4.5993 × 10−3
3 −2.0627 × 10−5
1.16
2 1 0.5648
2 −3.2352 × 10−3
3 1.6066 × 10−5
−0.2461

-3
3 1

g.cm
2 1.0439 × 10−2

ρ/g
3 −8.7552 × 10−5 1.12

1.20
1.08

1.16
1.04
0 10 20 30 40 50
P / MPa
-3
g cm

1.12 Fig. 5. Density of GLY (1) + H2 O (2) mixtures as a function of pressure at T = 298.15 K
and different mole fractions: 䊉, 0.1002; , 0.2001; , 0.3000; , 0.3998; , 0.5004;
, 0.5998; , 0.7001; , 0.8001; , 0.9000; lines, calculated by the modified form of
Eq. (3).

1.08
densities of the aqueous mixtures at 0.1 MPa. The figure also illus-
trates the excellent agreement between the calculated densities
and the experimental data with an AAD of 0.003%. The fitting para-
meters bi and ci for Eqs. (5) and (12) are presented in Table 3. For the
1.04 aqueous mixtures, the parameter B(T) decreases with increasing
temperature while C(x1 ) generally decreases with increasing DES
300 310 320
mole fraction. Excellent agreement between the experimental and
T/K the calculated densities of the aqueous mixtures can also observed,
as shown in Fig. 5 for a selected isotherm (T = 298.15 K). The cal-
Fig. 4. Density of GLY (1) + H2 O (2) mixtures as a function of temperature at 0.1 MPa
and different mole fractions: 䊉, 0.1002; , 0.2001; , 0.3000; , 0.3998; , 0.5004; culated AAD between the experimental and the predicted values
, 0.5998; , 0.7001; , 0.8001; , 0.9000; lines, calculated using Eq. (13). is 0.02% which suggests that the proposed correlation accurately
represents the experimental density data as functions of tempera-
ture, pressure, and composition.
the experimental data (total of 594 data points) at an AAD of
1.2% with only 11 fitting parameters. The good agreement bet-
ween the experimental and the calculated values is also depicted 4. Conclusions
in Figs. 2 and 3.
Finally, to obtain a general correlation for the densities of the The densities of the deep eutectic solvent GLY and its aqueous
GLY (1) + H2 O (2) mixtures as a function of all three variables – mixtures in the temperature range (298.15–323.15) K and up to
temperature, pressure, and composition – the Tait-type equation 50 MPa were reported in this work. The densities of GLY as well
earlier presented as Eq. (3) was modified by applying an approach as its aqueous mixtures decreased with increasing temperature
similar to that by Abdulagatov et al. [28]. Here, we assumed that and increased with increasing pressure and DES mole fraction.
B remained the temperature-dependent parameter represented by Experimental data were used to derive other properties such as iso-
Eq. (5), and that C was concentration-dependent. The latter is be thermal compressibilities, isobaric expansivities, and excess molar
represented by the function volumes. The isothermal compressibilities of GLY increased as the
temperature increased and decreased as the pressure increased.

2
The isobaric expansivities decreased with increasing temperature
C(x1 ) = ci x1i . (12) and pressure, which may be associated to the changes in the mole-
i=0 cular distance or free-volume within the liquid. A six-parameter
Tait-type equation accurately correlated the high-pressure densi-
The reference density data, 0 (T, x1 , 0.1 MPa), were calculated
ties of GLY to temperature and pressure with an AAD of 0.003%.
using the equation
The excess molar volumes of the aqueous mixtures were nega-
M tive over the whole range of compositions at all temperatures and
0 (T, x1 , 0.1 MPa) = 2 . (13)
pressures. The values decreased with increasing temperature (at
VE + x (Mi /0,i )
i=1 i
constant pressure) and pressure (at constant temperature) which
where VE is the excess molar volume obtained by Eq. (11); M is the may be related to the dependence of the strength of hydrogen
molecular weight of the mixture; 0,i is the reference density of bonding in the liquid on temperature, and to the decrease in the
component i calculated from Eq. (4); and Mi is the molecular weight molecular distance and the free-volume of the mixtures. A modi-
of the component i. Fig. 4 shows the temperature dependence of the fied form of the Tait equation and an empirical equation by Lugo
38 R.B. Leron et al. / Fluid Phase Equilibria 335 (2012) 32–38

et al. were successfully used to represent the high-pressure den- [3] R.J. Bernot, E.A. Kennedy, G.A. Lamberti, Environ. Toxicol. Chem. 24 (2005)
sities (AAD = 0.02%) and excess molar volumes (AAD = 1.2%) of the 1759–1765.
[4] C. Pretti, C. Chiappe, D. Pieraccini, M. Gregori, F. Abramo, G. Monni, L. Intorre,
aqueous GLY mixtures as a function of all three variables, namely Green Chem. 8 (2006) 238–240.
temperature, pressure and composition. [5] M. Hasib-ur-Rahman, M. Siaj, F. Larachi, Chem. Eng. Process. 49 (2010) 313–322.
[6] A.P. Abbott, G. Capper, D.L. Davies, R.K. Rasheed, V. Tambyrajah, Chem. Com-
mun. (2003) 70–71.
List of symbols [7] A.P. Abbott, D. Boothby, G. Capper, D.L. Davies, R.K. Rasheed, J. Am. Chem. Soc.
126 (2004) 9142–9147.
[8] A.P. Abbott, G. Capper, D.L. Davies, H.L. Munro, R.K. Rasheed, V. Tambyrajah,
Chem. Commun. (2001) 2010–2011.
aj,k,l calibration constants of the density meter [9] A.P. Abbott, R.C. Harris, K.S. Ryder, C. D’Agostino, L.F. Gladden, M.D. Mantle,
a0 , a1 , a2 fitting parameters for Eq. (4) (g cm−3 , g cm−3 K−1 , Green Chem. 13 (2011) 82–90.
g cm−3 K−2 ) [10] A.P. Abbott, IeMRC & SUMEEPnet Scientific Conference, 2008.
[11] M. Avalos, R. Babiano, P. Cintas, J.L. Jimenez, J.C. Palacios, Angew. Chem. Int. Ed.
AAD average absolute deviation (%) 45 (2006) 3904–3908.
b0 , b1 fitting parameters for Eq. (5) (MPa, MPa K−1 ) [12] S.Z.E. Abedin, F. Endres, Acc. Chem. Res. 40 (2007) 1106–1113.
ci fitting parameters for Eq. (12) [13] M.C. Gutierrez, M.L. Ferrer, L. Yuste, F. Rojo, F. del Monte, Angew. Chem. Int. Ed.
49 (2010) 2158–2162.
B(T) temperature-dependent parameter of the Tait-type equa- [14] H.G. Morrison, C.C. Sun, S. Neervannan, Int. J. Pharm. 378 (2009) 136–139.
tion, Eq. (3) [15] C. Gu, J. Tu, Langmuir 27 (2011) 10132–10140.
C(x) concentration-dependent parameter of the Tait-type [16] C.-D. Gu, J.-P. Tu, RSC Adv. 1 (2011) 1220–1227.
[17] X. Li, M. Hou, B. Han, X. Wang, L. Zou, J. Chem. Eng. Data 53 (2008) 548–550.
equation, Eq. (3) [18] W.C. Su, D.S.H. Wong, M.H. Li, J. Chem. Eng. Data 54 (2009) 1951–1955.
d0 , d1 pressure-dependent fitting parameters for Eq. (11) (unit- [19] A.P. Abbott, P.M. Cullis, M.J. Gibson, R.C. Harris, E. Raven, Green Chem. 9 (2007)
less, MPa−1 ) 868–872.
[20] M. Hayyan, F.S. Mjalli, M.A. Hashim, I.M. AlNashef, Fuel Process. Technol. 91
Dij temperature- and composition-dependent parameter for
(2010) 116–120.
equation, Eq. (11) (cm3 mol−1 K1−j ) [21] K. Shahbaz, F.S. Mjalli, M.A. Hashim, I.M. ALNashef, J. Appl. Sci. 10 (2010)
DES deep eutectic solvent 3349–3354.
GLY glyceline [22] K. Shahbaz, F.S. Mjalli, M.A. Hashim, I.M. AlNashef, Thermochim. Acta 515
(2011) 67–72.
IL ionic liquid [23] K. Shahbaz, S. Baroutian, F.S. Mjalli, M.A. Hashim, I.M. AlNashef, Thermochim.
M molecular weight of the DES mixture (g mol−1 ) Acta 527 (2012) 59–66.
Mi molecular weight of component i (g mol−1 ) [24] R.B. Leron, A.N. Soriano, M.-H. Li, J. Taiwan Inst. Chem. Eng. 43 (2012) 551–557.
[25] L. Lugo, M.J.P. Comunas, E.R. Lopez, J. Fernandez, Fluid Phase Equilib. 199 (2002)
NaCl (aq) aqueous sodium chloride solution 135–145.
N number of data points [26] R.B. Leron, M.-H. Li, J. Chem. Thermodyn. 54 (2012) 293–301.
T temperature (K) [27] J. Jacquemin, P. Husson, V. Mayer, I. Cibulka, J. Chem. Eng. Data 52 (2007)
2204–2211.
T0 reference temperature for Eq. (11) [28] I.M. Abdulagatov, A. Tekin, J. Safarov, A. Shahverdiyev, E. Hassel, J. Chem. Ther-
p period of oscillation (␮s) modyn. 40 (2008) 1386–1401.
P pressure (MPa) [29] J. Safarov, F. Millero, R. Feistel, A. Heintz, E. Hassel, Ocean Sci. 5 (2009) 235–246.
[30] R. Wagner, A. Pruss, J. Phys. Chem. Ref. Data 31 (2002) 387–535.
Pref reference pressure (0.1 MPa) [31] D.G. Archer, J. Phys. Chem. Ref. Data 21 (1992) 793–829.
VE excess molar volume (cm3 mol−1 ) [32] M.T. Zafarani-Moattar, H. Shekaari, J. Chem. Thermodyn. 38 (2006) 1377–1384.
x mole fraction [33] Y.A. Sanmamed, D. Gonzalez-Salgado, J. Troncoso, C.A. Cerdeirina, L. Romani,
Fluid Phase Equilib. 252 (2007) 96–102.
[34] A.N. Soriano, B.T. Doma Jr., M.-H. Li, J. Chem. Thermodyn. 41 (2009) 301–307.
Greek letters [35] R.L. Gardas, M.G. Freire, P.J. Carvalho, I.M. Marrucho, I.M.A. Fonseca, A.G.M.
˛P isobaric expansivity (K−1 ) Ferreira, J.A.P. Coutinho, J. Chem. Eng. Data 52 (2007) 80–88.
[36] D. Matkowska, T. Hofman, J. Mol. Liq. 165 (2012) 161–167.
T isothermal compressibility (GPa−1 ) [37] L.P.N. Rebelo, V. Najdanovic-Visak, Z.P. Visak, M. Nunes da Ponte, J. Szydlowski,
 density (g cm−3 ) C.A. Cerdeirina, J. Troncoso, L. Romani, J.M.S.S. Esperanca, H.J.R. Guedes, H.C. de
0 density at the reference pressure (0.1 MPa) (g cm−3 ) Sousa, Green Chem. 6 (2004).
[38] T. Retsina, S.M. Richardson, W.A. Wakeham, Appl. Sci. Res. 43 (1986) 127–158.
i,ref ith reference density data (g cm−3 ) [39] R. Lundstrum, A.R.H. Goodwin, K. Hsu, M. Frels, D.R. Caudwell, J.P.M. Trusler,
i,cal calculated density from Eq. (1) (g cm−3 ) K.N. Marsh, J. Chem. Eng. Data 50 (2005) 1377–1388.
[40] S.B.e. Abreu, H.M.N.T. Avelino, F.J.P. Caetano, J.M.N.A. Fareleira, J. Chem. Eng.
Data 55 (2010) 3525–3531.
Acknowledgements [41] Anton Paar GmbH, Austria, personal communication.
[42] J.H. Dymond, R. Malhotra, Int. J. Thermophys. 9 (1988) 941–951.
This research was supported by a Grant (NSC 101-3113-E-007- [43] R.L. Gardas, H.F. Costa, M.G. Freire, P.J. Carvalho, I.M. Marrucho, I.M.A. Fonseca,
A.G.M. Ferreira, J.A.P. Coutinho, J. Chem. Eng. Data 53 (2008) 805–811.
005) from the National Science Council of the Republic of China. [44] R. Gomes de Azevedo, J.M.S.S. Esperanca, V. Najdanovic-Visak, Z.P. Visak,
H.J.R. Guedes, M. Nunes da Ponte, L.P.N. Rebelo, J. Chem. Eng. Data 50 (2005)
References 997–1008.
[45] Z. Gu, J.F. Brennecke, J. Chem. Eng. Data 47 (2002) 339–345.
[46] G. Watson, C.K. Zéberg-Mikkelsen, A. Baylaucq, C. Boned, J. Chem. Eng. Data 51
[1] P. Wasserscheid, Nature 439 (2006) 797.
(2005) 112–118.
[2] A.H. Whitehead, M. Polzler, B. Gollas, J. Electrochem. Soc. 157 (2010)
[47] E. Gomez, B. Gonzalez, N. Calvar, E. Tojo, A. Dominguez, J. Chem. Eng. Data 51
D328–D334.
(2006) 2096–2102.
[48] J. Lehmann, M.H. Rausch, A. Leipertz, A.P. Froba, J. Chem. Eng. Data 55 (2010)
4068–4074.