210
G. Gonfa et al. / Journal of Molecular Liquids 238 (2017) 208–214
136.23. Elemental analysis (%): C (62.31%), H (5.65%), N (18.18%), S
(13.86%).
The relative volatilities (α21) were calculated from the VLE data ob-
tained from HSGC analyses. Since the peak area obtained from the HSGC
(Ai) for component (i) is proportional to the partial pressure of the com-
ponent in the vapour phase (Pi), then:
1-(2-Hydroxyethyl)-3-methylimidazolium thiocyanate: 1H NMR
(CDCl3): δ (ppm) = 3.86 (3H, s), 3.89 (2H, t), 4.27 (2H, t), 7.41 (1H,
d), 7.46 (1H, d), 8.97 (1H, s). 13CNMR (CDCl3): δ (ppm) = 35.7, 51.8,
60.5, 121.5, 122.7, 123.9, 136.4. Elemental analysis: C (45.42%), H
(6.01%), N (22.71%), S (17.41%).
Ai ¼ ciPi
ð1Þ
where, ci is a calibration factor for component i. According to modified
Raoult's law, the partial pressure of a component (i) in the vapour
phase is given by:
3-(3-Butyl-1H-imidazol-3-ium-1-yl)propanenitrile thiocyanate: 1H
NMR (DMSO-d6): δ = 0.84 (3H, t), 1.25 (2H, m), 1.65 (2H, m), 3.15
(2H, t), 4.35 (2H, t), 4.32 (2H, t), 4.45 (1H, s), 7.65 (1H, s), and 9.05
(1H, s). 13C NMR: δ (ppm) = 14.1, 18.3, 21.7, 29.4, 48.5, 50.7, 120.5,
121.9, 124.3, 134.5. Elemental analyses: C (55.87%), H (6.83%), N
(23.72%), S (13.58%).
3-[3-(2-Hydroxyethyl)-1H–imidazol-3-ium-1-yl]-propanenitrile
thiocyanate: 1H NMR (DMSO-d6): δ = 3.25 (2H, t), 3.75 (2H, t), 4.22
(2H, t), 5.37 (2H, d), 5.45 (1H, t), 7.91 (1H, s), 7.85 (1H, s), 9.45 (1H, s).
13C NMR (DMSO-d6): δ (ppm) =17.5, 47.6, 54.1, 59.3, 118.7, 119.3,
121.5, 122.9, 124.9. Elemental analysis (%): C (48.2%), H (5.3%), N
(24.9%), S (14.3%).
Pi ¼ xiγiPsiat
ð2Þ
where, Psiat is the vapour pressure of pure component (i), xi and γi are
the composition and activity coefficient of the component in the liquid
phase, respectively. Therefore, Eq. (1) becomes,
Ai ¼ cixiγiPsiat
ð3Þ
If the vial contains only pure component (i) (xi and γi=1), the peak
area (Aoi ) is proportional to the vapour pressure of the component at the
prevailing temperature. Therefore, Eq. (3) becomes:
3-(3-Allyl-1H-imidazol-3-ium-1-yl)propanenitrile thiocyanate: 1H
NMR (DMSO-d6): δ = 3.20 (2H, t), 4.55 (2H, t), 4.91 (2H, d), 5.35 (2H,
d), 6.15 (1H, m), 7.75 (1H, s), 7.89 (1H, s), 9.15 (1H, s). 13C NMR
(DMSO-d6): δ (ppm) = 18.2, 46.9, 54.1, 114.9, 118.5, 119.4, 121.7,
122.6, 125.8, 133.3. Elemental analysis: C (54.5%), H (5.4%), N (25.4%),
S (14.5%).
Aoi ¼ ciPsat
ð4Þ
i
Combining Eqs. (3) and (4) the relative volatility can be calculated
3-(3-Benzyl-1H-imidazol-3-ium-1-yl)propanenitrile thiocyanate:
1H NMR (DMSO-d6): δ = 3.25 (2H, t), 4.59 (2H, t),5.45 (2H, s), 7.47
(5H, m), 7.91 (1H, s), 7.95 (1H, s), 9.65 (1H, s). 13C NMR (DMSO-d6): δ
(ppm) = 17.9, 47.2, 56.7, 118.3119.4, 121.9, 122.3, 124.5, 125.6, 129.1,
129.2, 130.1, 130.3, 138.6. Elemental analysis: C (62.2%), H (5.2%), N
(20.7%), S (11.8%).
by
y2=x2 γ2Psat A2Psat=x2A0
y1=x1
2
2
2
α21
¼
¼
¼
ð5Þ
γ1Ps1at
A1Ps1at=x1A01
where, α21 is the relative volatility of Cy to Bz, 1 and 2 refer to benzene
The ILs were rigorously purified using ethyl acetate and diethyl ether
and dried in a vacuum oven at 80 °C for 48 h before use. The water con-
tents were measured using coulometric Karl Fischer titrator (model
DL39). The water contents of the studied ILs varies from 250 to
350 ppm.
and cyclohexane, respectively.
2.4. Computational details
2.4.1. Interaction energy
To investigate the interaction between the cations and thiocyanate
anion and the ILs and the solutes, geometric optimizations and single-
point energy calculations were performed through DFT computations
using DMol3 module of Materials Studio software (version 2016) [23].
Full structural optimization for the cations, thiocyanate anion, the cat-
ion-anion dimer and the IL-solute combinations were carried out. All
the calculations were performed with density functional theory (DFT)
using the generalized gradient approximation (GGA) and Perdew and
Wang functional 91 (PW91) exchange-correlation functional [24]. The
calculations were performed using a double numerical plus polarization
function (DNP) as basis set [25]. A real-space orbital global cutoff of 3.7
Å was used for the geometry optimization and the SCF convergence
was set to 10−6. The convergence criteria were set as follows: energy
= 1 × 10−5 Ha; force = 2 × 10−3 Ha/Å; displacement = 0.005 Å. The
calculation process includes both geometry optimizations and single-
point energy calculation. First, full geometry optimizations were per-
formed to optimize the geometries of the ILs. Then, ILs were combined
with the solutes (Bz and Cy) to find the minimum energy configura-
tions. Then, the energies for each cation-anion dimer and IL-solute com-
binations were calculated using the methods consistent with the
geometry optimizations.
2.3. Vapour-liquid equilibrium
The isothermal vapour-liquid equilibrium (VLE) data for binary sys-
tem Bz (1) + Cy (2) and ternary systems Bz (1) + Cy (2) + solvent (3)
were measured using head space GC (Trace GC-2000). Headspace gas
chromatography (HSGC) is an efficient method for measuring VLE of
volatile compounds in non-volatile solvents [19–21]. The apparatus
consists of a head space sampler with 20-sample vial tray and gas chro-
matography. Samples of 12 ml with predefined molar compositions
were prepared gravimetrically. The uncertainty of the gravimetric mea-
surement was 1 × 10−4 g. The samples were filled to 20 ml vials and
closed tightly with the cap and septum. The vials were tempered in the
head space-sampler oven at 323.15, 338.15 and 353.15 K until the va-
pour-liquid equilibrium is attained. The temperature accuracy is better
than 0.1 K. When the equilibrium between the liquid and the vapour
phase is reached, the vapour phase of the vials were analyzed by gas
chromatography. Since the solvents (ILs, DMSO and DMF) have negligi-
ble vapour pressure at these temperatures, only two peaks are observed
in the ternary systems. The vapour phase compositions were deter-
mined using flame ionization detector (FID). Capillary column, BP20
(polyethylene glycol of 30 m × 0.25 mm × 0.25 μm) was used to sepa-
rate the components. The HSGC was calibrated by total vaporization
technique where volume of samples that can be totally evaporated
added to the vials [22]. Calibration curve was established between the
peak area ratio versus vapour phase mole composition. The correlation
coefficient (R2) of the calibration curve is about 0.9994. The reproduc-
ibility of the experimental results was confirmed by at least three inde-
pendent measurements and the average values were reported. The
compound uncertainty of the head space GC analysis was about 0.0035.
2.4.2. COSMO-RS computation
The structural optimization was performed using TmoleX 4.2 quan-
tum mechanics package [26]. The geometry optimization were per-
formed using density functional theory (DFT) with BPVP86 functional
[27] using triple zeta valence potential (TZVP) basis set and resolution
of identity standard (RI) approximation [28]. All calculations were per-
formed with COSMOthermX (Version C3.0) using the parameter file
BP_TZVP_C30_1601 (COSMOlogic GmbH
& Co KG, Leverkusen,