Eutectic ionic liquids as potential electrolytes in dye-sensitized solar cells: Physicochemical and conductivity studies

Article abstract of DOI:10.1016/j.molliq.2020.114381

Ionic liquids (ILs) have been introduced as the electrolyte in dye-sensitized solar cells (DSSC) to replace organic solvents due to their non-flammability, chemical, and thermal stability, as well as less prone to evaporation and leakage. A series of eutectic ILs mixtures composed of 1-ethyl-3-methylimidazolium iodide, [C₂C₁im]I, 1-ethylpyridinium iodide [C₂py]I, 1-propylpyridinium iodide, [C₃py]I and 1-butylpyridinium iodide, [C₄py]I was developed. The structures and purity of the compounds were validated by nuclear magnetic resonance and Karl Fischer Titration. Their physicochemical properties, such as viscosities and ionic conductivities were evaluated for their potential use as electrolytes in DSSCs. The melting temperature of pure ionic liquids and the mixtures were determined using differential scanning calorimetry in the temperature range of 123.15 K to 393.15 K. It is found that eutectic mixtures were formed at the molar ratios of {0.75[C₂py]I:0.25[C₂C₁im], {0.25[C₂py]I:0.75[C₂C₁im]I}, {0.50[C₂py]I:0.50[C₂C₁im]I}, {0.25[C₂C₁im]I:0.75 [C₄py]I}, {0.75[C₂C₁im]I:0.25[C₄py]I}, and {0.50[C₂C₁im]I:0.50[C₄py]I} with the latter forming a highly stable liquid with no apparent melting peak in the studied temperature range. To understand the formation of eutectic composition, a correlation between molecular structures and the interaction energies of the cations and anions of ILs was performed using the Conductor-like Screening Model for Real Solvents (COSMO-RS). Dynamic viscosities of the binary mixtures were evaluated in the temperature range of 283.15 K to 353.15 K with 10 K increment. The highest ionic conductivity was exhibited by the eutectic ILs electrolyte {0.50[C₂C₁im]I:0.50[C₄py]I:I₂(50 wt%)} at 9.67 mS·cm^?1 showing its potential for electrolyte application in DSSC.

Full text of DOI:10.1016/j.molliq.2020.114381

Journal of Molecular Liquids 320 (2020) 114381
Contents lists available at ScienceDirect
Journal of Molecular Liquids
journal homepage: www.elsevier.com/locate/molliq
Eutectic ionic liquids as potential electrolytes in dye-sensitized solar
cells: Physicochemical and conductivity studies
c
Nurfathiah Izzaty Mohd Faridz Hilmy a, Wan Zaireen Nisa Yahya a,b, , Kiki Adi Kurnia
⁎
a
Department of Chemical Engineering, Universiti Teknologi PETRONAS, 32610 Seri Iskandar, Perak, Malaysia
Centre of Research in Ionic Liquid, Universiti Teknologi PETRONAS, 32610 Seri Iskandar, Perak, Malaysia
b
c
Department of Marine, Faculty of Fisheries and Marines, Universitas Arlangga, Jalan Mulyorejo Kampus C, Surabaya 60115, Indonesia
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 31 July 2020
Received in revised form 13 September 2020
Accepted 19 September 2020
Available online 25 September 2020
Ionic liquids (ILs) have been introduced as the electrolyte in dye-sensitized solar cells (DSSC) to replace organic
solvents due to their non-flammability, chemical, and thermal stability, as well as less prone to evaporation
and leakage. A series of eutectic ILs mixtures composed of 1-ethyl-3-methylimidazolium iodide, [C2C1im]I,
1-ethylpyridinium iodide [C2py]I, 1-propylpyridinium iodide, [C3py]I and 1-butylpyridinium iodide, [C4py]I
was developed. The structures and purity of the compounds were validated by nuclear magnetic resonance
and Karl Fischer Titration. Their physicochemical properties, such as viscosities and ionic conductivities were
evaluated for their potential use as electrolytes in DSSCs. The melting temperature of pure ionic liquids and
the mixtures were determined using differential scanning calorimetry in the temperature range of 123.15 K to
393.15 K. It is found that eutectic mixtures were formed at the molar ratios of {0.75[C2py]I:0.25[C2C1im], {0.25
[C2py]I:0.75[C2C1im]I}, {0.50[C2py]I:0.50[C2C1im]I}, {0.25[C2C1im]I:0.75 [C4py]I}, {0.75[C2C1im]I:0.25[C4py]I},
and {0.50[C2C1im]I:0.50[C4py]I} with the latter forming a highly stable liquid with no apparent melting peak
in the studied temperature range. To understand the formation of eutectic composition, a correlation between
molecular structures and the interaction energies of the cations and anions of ILs was performed using the
Conductor-like Screening Model for Real Solvents (COSMO-RS). Dynamic viscosities of the binary mixtures
were evaluated in the temperature range of 283.15 K to 353.15 K with 10 K increment. The highest ionic conduc-
tivity was exhibited by the eutectic ILs electrolyte {0.50[C2C1im]I:0.50[C4py]I:I2(50 wt%)} at 9.67 mS·cm−1
showing its potential for electrolyte application in DSSC.
Keywords:
Ionic liquid
Eutectic
DSSC
Electrolyte
COSMO-RS
© 2020 Elsevier B.V. All rights reserved.
1. Introduction
the concerns is the volatility and relatively high toxicity limit of the or-
ganic solvents which will affect the long-term performance of the cell
Dye-sensitized solar cells (DSSCs) is a promising third-generation
photovoltaic technology as it is relatively cheap as compared to first-
and second-generation solar cells, easy to manufacture, lightweight,
and can be operated in indoor low-light conditions [1]. It is comprised
of a mesoporous layer of dye-sensitized nanocrystalline titanium diox-
ide (TiO2), a cathode usually made of platinum (Pt), and an electrolyte
system in between the electrodes. Typically, the electrolyte is composed
of inorganic salts such as lithium iodide (LiI) or potassium iodide (KI) as
well as iodine (I2) to form the redox iodide/triiodide (I−/I3−) couple and
dissolved in organic solvents namely acetonitrile [2]. Although this liq-
uid electrolyte composition gives out a high diffusion rate and maxi-
mizes the cell efficiency, the long-term stability of the cell has been
the subject of concerns [3]. From the industrial point of view, one of
[2]. Besides, common inorganic salts such as LiI and KI have been re-
ported to have limited solubility in the liquid electrolyte at ambient
temperature [4]. Ionic liquids (ILs) have then been introduced to over-
come this problem, however, most of these compounds are highly vis-
cous which can limit the diffusion of the redox couple, I−/I3− in the
system.
Ionic liquids are low melting point salts (lower than 100 °C) and
comprised entirely of ions; an organic cation and either an inorganic
or organic anion [5]. ILs have been widely used in DSSCs due to their
chemical and thermal stability, non-flammability, and extremely low
vapor pressure which can eliminate the risk of leakage and volatility
[6,7]. They are also known as designer solvents due to their tunable
chemical and physical properties. Studies have also demonstrated that
the mixing of two or more ILs can lead to a melting temperature depres-
sion until a minimum value when melting occurs without phase separa-
tion known as the eutectic point [8]. The great interest in this kind of
mixture is that the system can be formulated so that non-room-
temperature ILs can be mixed to give a eutectic mixture that is liquid
⁎
Corresponding author at: Department of Chemical Engineering, Universiti Teknologi
PETRONAS, 32610 Seri Iskandar, Perak, Malaysia.
E-mail addresses: nurfathiah_18002858@utp.edu.my (N.I. Mohd Faridz Hilmy),
zaireen.yahya@utp.edu.my (W.Z.N. Yahya), kiki.adi@fpk.unair.ac.id (K.A. Kurnia).
https://doi.org/10.1016/j.molliq.2020.114381
0167-7322/© 2020 Elsevier B.V. All rights reserved.

N.I. Mohd Faridz Hilmy, W.Z.N. Yahya and K.A. Kurnia
Journal of Molecular Liquids 320 (2020) 114381
at, or close to room temperature operating conditions. To reach such
target, the salt systems melting profiles' can be modified through the
right choice of cation-anion pairs, by manipulating the size and shape
of their chemical structure and by changing the repulsive/attractive
forces that build the lattice energy of the solid-state [9].
enhancement. To complement the experimental data, a quantum chem-
istry model namely Conductor-like Screening Model for Real Solvents
(COSMO-RS) was used to delve into the interaction between these bi-
nary eutectic ILs mixture through the charge density distribution of
the molecular surface and the polarity of the components [21].
COSMO-RS model is very promising and allows to predict the thermo-
dynamic properties of ILs in either pure or in mixture with other com-
pounds. The screening of charge density distributions, namely sigma
profile (σ-profile), is derived from the quantum chemical calculations.
Then, based on the obtained data from σ-profile, the component chem-
ical potentials were quantified from the statistical thermodynamics'
treatment of molecular interactions through the generated sigma po-
tential (σ-potential) [22,23].
A series of ionic liquids namely 1-ethyl-3-methylimidazolium iodide,
[C2C1im]I and alkylpyridinium iodide i.e. 1-ethylpyridinium iodide
[C2py]I, 1-propylpyridinium iodide, [C3py]I and 1-butylpyridinium iodide,
[C4py]I was synthesized and mixed at different molar ratios. The chemical
structures of the studied ILs are shown in Fig. 1. The system has been cho-
sen in such a way to investigate the interaction of commonly used
imidazolium-based IL, [C2C1im]I with pyridinium-based ILs at different
alkyl chain length. Larger size cations namely alkylpyridinium ILs are
selected as they may act as a shield towards TiO2/dye against triiodide re-
combination in combination with imidazolium IL [24]. It has been demon-
strated that the alkyl chain length variations will highly influence ILs
properties notably the melting point, the viscosity, and the ionic conduc-
tivity [25]. To further assess the viscosity and ionic conductivity of electro-
lytes in presence of triiodides ions, iodine (I2), or tetrabutylammonium
triiodide [TBA][I-I-I] is incorporated in the system to match with DSSC
electrolyte composition.
Many eutectic mixtures have been used in various applications espe-
cially in the electrochemical industry due to their ability to induce low
melting point and high decomposition temperature, which can meet
the power needs in extremely low and high-temperature weather con-
ditions or heavy electrical applications [10]. Other applications such as
CO2 solubility and soil heavy metal such as cadmium ion (Cd2+) reme-
diations are also using eutectic based ILs melts as they can be easily
mixed and converted without further purification [11,12]. Encouraging
solar cells performances on the solvent-free ILs electrolytes for DSSCs
were recently reported with improved ionic conductivity, power con-
version efficiency [3,13], and long-term stability [1,3,14]. Another strat-
egy to enhance the mass transportation of I−/I3− ions in the electrolyte is
to incorporate low viscosity ILs. Previous studies have shown that the
incorporation of imidazolium based ILs bearing thiocyanate anion in
the electrolytes can be one of the ways to reduce the mass transport lim-
itations [15]. Besides, the incorporation of ionic liquids in polymer elec-
trolytes has also been a subject of interest and almost a must-have
ingredient to increase the ionic conductivities and performance in
DSSCs [16–19].
Common organic salts such as imidazolium, pyridinium, piperidinium,
pyrrolidinium, and ammonium based ILs are non-volatile and stable
under atmospheric environment and for short alkyl chains, they mostly
exist in the solid-state at room temperature [15]. As consequence, their
usage in the application for electrolyte purpose is limited, although they
are more easily synthesized and purified. A major highlight has been re-
ported by Lau et al. in 2015, on a series of solvent-free eutectic ILs
which composed of 1,3-dimethylimidazolium iodide (DMII), guanidium
thiocyanate (GNCS), N-methylbenzimidazole (NMB), 2-triazolium iodide,
2-triazolium tricyanomethanide and 0.33 M I2 which recorded high
power conversion efficiencies (PCE) of 3.7–5.9%. Further incorporation
of a high boiling point solvent namely sulfolane into the mixture has
afforded considerably high PCE up to 7.1% [20]. The encouraging high per-
formance and high stability electrolytes for DSSCs necessitate further in-
vestigations on new eutectic ILs. Nonetheless, very few studies have
been conducted towards understanding the physicochemical properties
of the eutectic ILs in DSSCs applications, as more focus is highlighted on
the solar cells' performances rather than the factors that influenced
their performances.
2. Materials and methods
Four iodide-based ILs with one triiodide-based IL were synthe-
sized, namely, 1-ethyl-3-methylimidazolium iodide [C2C1mim]I; 1-
ethylpyridinium iodide [C2py]I; 1-propylpyridinium iodide [C3py]I;
1-butylpyridinium iodide [C4py]I; tetrabutylammonium triiodide
[TBA][I-I-I].
1-methylimidazole (99 wt% of purity; CAS-number: 616-47-7),
iodoethane (99 wt% of purity; CAS number: 75-03-6), pyridine (≥99 wt
% purity; CAS number: 110-86-1), 1-iodopropane (99 wt% purity; CAS
number: 107-08-4), 1-iodobutane (99 wt% purity; CAS number: 542-
69-8), tetrabutylammonium iodide (98 wt% purity; CAS number: 311-
28-4), iodine ((≥99.8 wt% purity; CAS number: 7553-56-2), methanol
(≥99.9 wt% purity; CAS number: 67-56-1), ethanol (95 wt% purity;
CAS number: 64-17-5), ethyl acetate (≥99.5 wt% purity; CAS number:
141-78-6), acetonitrile (anhydrous, 99.8%; CAS number: 75-05-8) and
To leverage the advantages of the eutectic melts, this study is fo-
cused on developing binary eutectic ILs mixtures and evaluating the
melting behavior, the density, the viscosity, and ionic conductivity
I–
N+
CH3
CH3
H3C
CH3
I–
N
1-ethylpyridinium iodide, [C2py]I
N+
I
I
I
N+
CH3
I–
CH3
N+
H3C
CH3
1-propylpyridiniumiodide, [C3py]I
tetrabutylammonium triiodide, [TBA][I-I-I]
1-ethyl-3-methylimidazolium iodide
[C2C1im]I
I–
N+
CH3
1-butylpyridinium iodide, [C4py]I
Fig. 1. Molecular structure and nomenclature of [C2C1im]I, [C2py]I, [C3py]I, [C4py]I and [TBA][I-I-I] as a triiodide provider.
2

N.I. Mohd Faridz Hilmy, W.Z.N. Yahya and K.A. Kurnia
Journal of Molecular Liquids 320 (2020) 114381
1-butyl-3-methylimidazolium thiocyanate [BMIM][SCN] (95 wt% purity;
CAS number: 344790-87-0) were purchased from Sigma Aldrich and
Merck, and were used without further purification.
The NMR analysis of synthesized compounds was recorded on a
Bruker AVANCE III 500 MHz spectrometer. The water content was ana-
lyzed by Karl-Fischer Titration (coulometric titrators). The ultraviolet-
visible spectroscopy (UV–Vis) was used to determine the absorbance
of triiodide (I−3 ) in [TBA][I-I-I], with iodide (I−) and iodine (I2)
were also tested to confirm the formation of [TBA][I-I-I]. All NMR and
UV–Vis spectra are presented in the Supplementary Information (SI).
weight. The binary mixtures were comprised of [C2C1im]I, [C2py]I,
[C3py]I, [C4py]I, and prepared based on the mole fraction of 0.25–0.75
of the corresponding ILs. The mixtures were stirred and heated at
about 10 οC above the highest observed melting points of the two ILs
until the mixtures reached complete melting and formed homogenous
mixtures. Then, the mixtures were subsequently cooled to room tem-
perature and stored under vacuum.
Only one eutectic ILs mixture which remains at the liquid phase at
room temperature was selected for further analysis as potential electro-
lytes. Electrolytes A–C contains different wt% of [TBA][I-I-I] with the
chosen eutectic ILs, while electrolytes D–I contains different wt% of I2.
The eutectic ILs mixture was also tested with a low viscosity [BMIM]
[SCN] in electrolyte H–I that contains 10 wt% with two triiodides ion
providers (I2 and [TBA][I-I-I]). The detailed compositions of the electro-
lytes are shown in Table 1.
2.1. Synthesis of ILs
2.1.1. Preparation of 1-ethyl-3-methylimidazolium iodide, [C2C1im] I
This procedure followed the method adapted from Cavallo et al., [26]
with slight modification. In brief, 1-ethyl-3-methylimidazolium iodide
was synthesized using 1-methylimidazole and iodoethane. First, about
5 g of 1-methylimidazole (0.06090 mol, 1 equivalent) and 12.35 g of
the iodoethane (0.07917 mol, 1.3 equivalent) were mixed and stirred
in 33 mL of acetonitrile. This mixture was then refluxed at temperature
(75–80 °C) for 12 h. After 12 h, the mixture was cooled to room temper-
ature and washed with ethyl acetate. The solvent was then removed
by using a rotary evaporator under reduced pressure. The IL was
further dried using the glass vacuum line for 24 h. The final product of
1-ethyl-3-methylimidazolium iodide was stored in a desiccator until
further use.
2.3. Characterizations of ILs
The melting temperature (Tfus) and enthalpy of fusion (ΔHfus) were
measured by differential scanning calorimetry technique (DSC) using
Mettler Toledo DSC 1 STAR System equipped with liquid nitrogen. The
calibration experiment was carried out with a 10 K/min heating rate
at the temperature range of 123.15 K–393.15 K with heating, cooling,
and heating cycle.
Density data were measured by using AccuPyc II TEC from
Micromeritics by putting the sample into the sample cell. The density
of the samples was measured through the pressure change from the
helium gas displacement. The data are collected at different mole ratios
of binary eutectic ILs (0.25–0.75) at 25 °C.
White-yellowish solid, yield ≥ 85%, water content (4120 ppm). 1
H
NMR δH(500 MHz; DMSO) ppm: 1.414(3H, t), 3.85(3H, s), 4.18(2H, q),
7.68(2H, d), 9.11(1H, s).
The dynamic viscosity measurements were carried out in DHR-1
Rheometer, using 1° upper cone angle geometry with a 20 mm Peltier
plate. The experiment was carried out within a temperature range of
2.1.2. Preparation of 1-alkylpyridinium iodide (C2-C4)
1-alkylpyridinium iodide was synthesized using pyridine and
iodoalkane (chain length C2 – C4) following the same procedure as
per the synthesis of 1-ethyl-3-methylimidazolium iodide with slight
modification.
1-ethylpyridinium iodide, [C2py]I: Yellow solid, yield ≥ 85%, water
content (1229 ppm). 1H NMR δH(500 MHz; DMSO) ppm: 1.553(3H, t),
4.659(2H, s), 8.170(2H, t), 8.612(1H, t), 9.111(2H, d).
1-propylpyridinium iodide, [C3py]I: Brown-yellowish liquid, yield
≥ 85%, water content (1000 ppm). 1H NMR δH(500 MHz; D2O) ppm:
0.821(3H, t), 1.921(2H, t), 4.457(2H, s), 7.947(2H, t), 8.418(1H, t),
8.724(2H, d).
293.15 K–353.15 K at a constant shear rate of 1 s−1
.
The ionic conductivity measurements (σ) were conducted using
WonaTech electrochemical impedance spectroscopy (WEIS) potentiostat,
with the AC impedance method was applied at the range of 0.1 Hz
−103 Hz and 10 mV AC amplitude. The samples were thermally equili-
brated to room temperature for at least 1 h before the measurements.
ILs mixtures were then filled up between two mirror-finished stainless-
steel electrodes using a Teflon ring separator in a constant-volume cylin-
drical cell and sealed in a Teflon cylinder. The experimental data gener-
ated from the computer-interfaced WEIS were fitted using the NOVA
software for calculation purposes.
COSMO-RS is a computational method that calculates the thermody-
namic properties of fluids and solutions based on quantum mechanical
data. In this work, the solubility of binary eutectic mixtures of ILs was
interpreted through complex molecular interactions of binary ILs in σ-
profiles and σ-potential. The σ-profiles of all the involved cations and an-
ions of the ionic liquids were directly obtained from the COSMOthermX
Version 19.0.1 available database using triple-zeta valence with polariza-
tion (TZVP) parametrization. For the σ-profiles, all the compounds were
distributed within three polarity regions which are two polar regions
with a non-polar region in between them. The σ-potential is a measure
1-butylpyridinium iodide, [C4py]I: Brown-yellowish solid, yield
≥ 90%, water content (630 ppm). 1H NMR δH(500 MHz; DMSO) ppm:
0.98(3H, t), 1.231(2H, m), 1.861(2H, m), 4.479(2H, t), 7.926(2H, t),
8.399(1H, t), 8.710(2H, s).
2.1.3. Preparation of tetrabutylammonium triiodide
The procedure followed the method adapted from Chowdhury et al.
with slight modification [27]. First, about 6.0 g of tetrabutylammonium
iodide (0.01624 mol, 1 equivalent) was mixed with 4.1 g of iodine
(0.01624 mol, 1 equivalent) resublimed in 30 mL hot methanol then
stirred at 50 °C and 200 rpm for 30 min. A purplish-black crystal was
formed at the bottom of the round-bottom flask. The product was dis-
solved in 18 mL of boiling ethanol at 75 °C. Then, the product was cooled
in an ice bath for 25 min for crystallization purposes. The solvents were
removed by using a rotary evaporator at 80 °C under reduced pressure.
Black-purplish solid, yield ≥ 85%. 1H NMR δH(500 MHz; DMSO) ppm:
0.937(12H, t), 1.321(8H, m), 1.57(8H, m), 3.161(8H, t). UV–Vis absorp-
Table 1
The eutectic ionic liquids electrolyte formulations.
Samples
Electrolytes
A
B
C
D
E
F
[C4py]I: [C2C1im]I: [TBA][I-I-I] (1 wt%)
[C4py]I: [C2C1im]I: [TBA][I-I-I] (5 wt%)
[C4py]I: [C2C1im]I: [TBA] [I-I-I] (10 wt%)
[C4py]I: [C2C1im]I: I2 (5 wt%)
[C4py]I: [C2C1im] I: I2 (10 wt%)
[C4py]I: [C2C1im] I: I2 (25 wt%)
tion wavelength (ACN): [I−] 330 nm, [I2] 362 nm and 458.5 nm, [I3−
362 nm.
]
2.2. Binary eutectic ILs formulations
G
H
I
[C4py]I: [C2C1im] I: I2 (50 wt%)
[C4py]I: [C2C1im]I: I2 (50 wt%): [BMIM][SCN] (10 wt%)
[C4py]I: [C2C1im]I: [TBA][I-I-I] (5 wt%): [BMIM][SCN] (10 wt%)
The ILs binary mixtures were prepared at standard room tempera-
ture and pressure by mixing two pure components with predetermined
3

N.I. Mohd Faridz Hilmy, W.Z.N. Yahya and K.A. Kurnia
Journal of Molecular Liquids 320 (2020) 114381
Table 2
formation of a highly stable liquid at room temperature in the presence
of traces water. All the synthesized ILs were found to be hygroscopic as
shown by their water contents although they were vacuum dried for
24 h and stored in a desiccator before use. The result is also reported
by Stolarska et al., suggesting that an absence of the melting peaks for
ILs mixtures is caused by the high hygroscopicity of the ILs mixture
[29]. Zürner et al. on the other hand reported that the absence of crystal-
lization point in DSC measurement of ILs was explained due to their ther-
mal stability at room temperature [30]. Besides, the result also supported
by Gómez et al. that this type of ILs are good-glassy formers, which indi-
cates that they present a weak tendency for crystallization and melting
[31].
Almost in all of the ILs DSC curves, small endothermic peaks were
observed at temperature ca. 220–240 K with the enthalpy energy of
ca. 2–3 J/g. For instance, based on the Fig. 2, small endothermic peaks
were observed for {0.25[C2py]I:0.75[C2C1mim]I} and {0.75[C2py]I:0.25
[C2C1mim]I} at temperature of 231.15 K and 230.65 K respectively. In-
terpretation of these peaks is rather complex and may be attributed to
the presence of traces of water during the formation of the ILs mixtures
[32]. Although the peaks appeared at a temperature lower than the fu-
sion temperature of water at 273.15 K, the ILs-water system may expe-
rience itself a depression melting temperature like the water‑sodium
chloride salt system. Another plausible attribution for these peaks is
the molecular relaxation of the ILs that leads to a misinterpretation of
the glass transition temperature as a weak melting point. Therefore,
only the intense endothermic peak will be considered as the melting
point for the pure ILs and in the ILs mixtures.
Thermal properties of the investigated pure ILs: glass phase transition temperature (Tg)
and melting temperature (Tfus) measured by DSC.
ILs
Tg (K)
Tfus,1 (K)
ΔHfus,1 (J/g)
Tfus,2 (K)
ΔHfus,2 (J/g)
[C2C1im]I
[C2py]I
[C3py]I
[C4py]I
[TBA][I-I-I]
x1
236.82
–
230.98
–
–
350.32
380.32
–
246.32
343.82
53.15
37.91
–
2.64
163.94
–
–
–
–
–
–
339.48
–
43.09
–
{x1 [C2py] I: (1-x1) [C2C1im] I}
0.250
0.500
0.750
–
–
–
231.15
315.32
230.65
2.54
38.81
2.10
331.98
–
309.82
54.80
–
55.59
{x1[C3py] I: (1-x1) [C2C1im] I}
0.250
0.500
0.750
212.16
202.24
200.12
333.82
liquid
268.40
65.31
–
0.01348
–
–
–
–
–
–
{x1[C4py] I: (1-x1) [C2C1im] I}
0.25
0.5
213.37
211.06
–
332.82
liquid
225.65
48.78
–
2.88
–
–
–
–
–
–
0.75
of the affinity of a component in a mixture towards another [28]. Both σ-
profiles and σ-potential data were represented in a distribution
histogram.
3. Results and discussions
The melting points for the binary mixtures of {0.25[C2py]I:0.75
[C2C1mim]I}, {0.50[C2py]I:0.50[C2C1mim]I} and {0.75[C2py]I:0.25
[C2C1mim]I} are found at 331.98 K, 315.32 K and 309.82 K respec-
tively. The lowest melting point is found at the 0.75 molar ratio of
[C2py]I to [C2C1mim]I. It is rather surprising at a higher molar concentra-
tion of [C2py]I, which has a higher melting point than [C2C1mim]I, this
composition resulted in the lowest melting point. As both cations have
the same alkyl chain length, the larger pyridinium has indeed disrupted
the interaction between the ions and favorize more disorder, thus low-
ering the melting point [33]. It is also noteworthy that for the binary
mixture of {0.50[C2py]I:0.50[C2C1mim]I}, {0.25[C3py]I:0.75[C2C1mim]I}
and {0.25[C4py]I:0.75[C2C1mim]I} exothermic crystallization peaks
were observed during the second heating cycle at the temperature of
256.32 K, 273.15 K and 261.15 K as shown in Fig. 2. One general conclu-
sion that could be drawn is the quench cooling on the endothermic
curve has allowed a significant amorphous structure to rearrange upon
heating to form the crystalline structure before the melting point [34].
For [C3py]I, which is liquid at room temperature, when it is mixed
with [C2C1mim]I at the composition of {0.25[C3py]I:0.75[C2C1mim]I}, a
semi-solid was obtained with a melting point of 333.82 K. As the
3.1. Phase behavior of pure ILs and mixtures
The glass phase transition (Tg) and melting point (Tfus) temperatures
of the pure ILs and their respective binary mixtures at a mole ratio of
0.25–0.75 are presented in Table 2 and the DSC curves are shown in
Fig. 2. The pure ILs namely [C2C1im]I, [C2py]I, [C4py]I, and [TBA][I-I-I]
are all in the solid-state at ambient temperature, with the melting points
of 350.32 K, 380.32 K, 246.32 K, and 343.82 K respectively. Although
[TBA][I-I-I] has the second lowest melting point, it did not produce a sta-
ble formation of eutectic mixture with pyridinium or imidazolium ILs.
Only a small amount of it can interact with imidazolium or pyridinium-
based iodide forming a semi-solid state mixture and not a homogeneous
liquid eutectic mixture. Interestingly, for [C3py]I, which is liquid at ambi-
ent temperature, no apparent melting peak is observed in the DSC anal-
ysis operating window of 123.15 K to 393.15 K. Noteworthy that right
after the vacuum drying of [C3py]I, a semi-solid was formed but all of
the solid immediately melted when exposed to ambient air. It can be
suggested that as the [C3py]I is highly hygroscopic, this allows the
(c)
(a)
(b)
[C2C1im]I
[C2C1im]I
[C2C1im]I
Tg = 236.82 K
Tfus,1 = 350.32 K
Tg = 236.82 K
Tfus,1 = 350.32 K
Tfus,1 = 350.32 K
Tfus,1 = 332.82 K
{0.25 [C2py]I : 0.75[C2C1im]I}
{0.50 [C2py]I : 0.50 [C2C1im]I}
{0.25 [C4py]I : 0.75 [C2C1im]I}
{0.25 [C3py]I : 0.75[C2C1im]I}
Tfus,1
= 231.15 K
T= 261.15 K
Tfus,2 = 331.98 K
Tg = 213.37
Tg = 212.16 K
{0.50 [C3py]I : 0.50 [C2C1im]I}
Tg = 202.24 K
T = 273.15 K
Tfus,1 = 333.82 K
{0.50 [C4py]I : 0.50[C2C1im]I}
Tg = 211.06 K
{0.75 [C4py]I : 0.25 [C2C1im]I}
Tfus,1 = 225.65 K
T = 256.32 K
Tfus,1 = 315.32 K
{0.75 [C2py]I : 0.25[C2C1im]I}
[C2py]I
{0.75 [C3py]I : 0.25 [C2C1im]I}
Tg = 200.12 K
Tfus,1 = 230.65 K
Tfus,2 = 309.82 K
Tfus,1 = 268.4 K
[C4py]I
[C3py]I
Tfus,1 = 380.32
Tfus,1 = 246.32 K
Tfus,2 = 339.48 K
373.15
Tg = 230.98 K
123.15
173.15
223.15
273.15
323.15
373.15
123.15
173.15
223.15
273.15
323.15
123.15
173.15
223.15
273.15
323.15
373.15
Temperature (K)
Temperature (K)
Temperature (K)
Fig. 2. The DSC curves of ILs mixture of a) [C2py]I:[C2C1mim]I, b) [C3py]I:[C2C1mim]I and c) [C4py]I: [C2C1mim]I.
4

N.I. Mohd Faridz Hilmy, W.Z.N. Yahya and K.A. Kurnia
Journal of Molecular Liquids 320 (2020) 114381
concentration of [C3py]I increased, at 0.5 and 0.75 molar ratio, highly
stable liquids were obtained, and only glass phase transitions were ob-
served with no apparent melting peak were recorded. Similar findings
are observed for [C4py]I, where highly stable liquids were formed at
0.5 and 0.75 molar ratio. According to Valderrama et al., the odd alkyl
chain length is less stable than the even alkyl chain length, hence the
melting point will be relatively lowered [35]. Based on our results, for
short alkyl chain length, the increase of alkyl chain will favorize disorder
and hamper ring stacking thus lowering the melting point. Based on
these results, only the binary eutectic mixtures which form stable liquid
without any phase transition even at low temperature is chosen as po-
tential IL-based electrolyte for DSSC.
pyridinium cations have a distribution of charge densities in the non-
polar region (−1 < σ < 1) and close towards hydrogen bond donor
(HBD), which indicates the non-polar alkyl and aromatic groups prop-
erties. Besides, all cations show the low-intensity peaks at values
below the cutoff σ < −1. It can be observed that the polarized charge
of these cations is displaced towards more H-bond donor region in the
order of [C2C1mim] > [C2py] = [C3py] = [C4py] > [TBA]. These distribu-
tions are mainly associated with the nitrogen (N) element existed
within the molecules which make the molecules having the ability of
hydrogen bond donor (intensively blue colored in the sigma surface of
cations).
The iodide anion presents its high polarity region at 1.5 e·nm−2
(above cutoff σ > 1). Thus, it can be considered as a strong hydrogen
bond acceptor, which is corresponding to its red-colored sigma surface.
For [I-I-I], the highest peak can be seen at 0.7 e·nm−2 at a non-polar re-
gion which proves that [I-I-I] is less polar than iodide (yellow-green col-
ored in the sigma surface of anions). Furthermore, the low intensity
peak of [I-I-I] is displaced towards a positive region indicating this
anion ability to be a slightly hydrogen bond acceptor (HBA).
3.2. Correlations of physicochemical properties using COSMO-RS
3.2.1. σ-profile analysis
The basis of COSMO-RS as a thermodynamic model is it considers
the interaction energy of polarization charge densities on the surface
of the molecule. The sigma profile (σ-profile) which is related to the
charge density distribution on the molecular surface for [TBA][I-I-I]],
[C2C1mim]I and [C2py]I, [C3py]I and [C4py]I are generated. The polarity
of the component can be determined through the help of molecular
electrostatic potentials [23]. The interactions of molecular electrostatic
potential will incorporate the hydrogen bonding and hydrophobicity
of the structure through the data generated by COSMO-RS. Thus,
COSMO-RS can demonstrate the clear visual of polarity distribution
from σ-profile while creating multiple interactions of ILs according to
their likability nature towards each other through σ-potential.
In this work, [TBA][I-I-I] has been compared with [C2C1mim]I and
[C2py]I, [C3py]I and [C4py]I in terms of molecular similarity towards
each other by comparing their charge density and electrostatic poten-
tial. The σ-profile distribution can be divided into three regions, namely
hydrogen bond donor (σ < − 1 e. nm−2), the non-polar region extend-
ing between the two polar regions (−1 e. nm−2 < σ < 1 e. nm−2), and
hydrogen bond acceptor region (σ > − 1 e. nm−2). Here, the extent of
screening charge densities or sigma surface is varied from −3 e. nm−2
(blue) to 3 e. nm−2 (red) within the molecule, while the intermediate
region is represented by green and yellow colored [36].
3.2.2. σ-Potential analysis
The COSMO-RS also generates the σ-potential, which describes their
interaction energy as compounds. In the σ-potential plot, a higher neg-
ative value of μ(σ) suggests a stronger attraction between components,
while a higher positive value indicates an increase in repulsive behavior.
Similar to the σ-profile plot, there are three regions divided by the hy-
drogen bond threshold (σhb = + / − 1 e. nm−2) as shown in Fig. 4
[37]. The σ-potential of [TBA][I-I-I] shows a parabolic curve that indi-
cates the repulsive behavior towards both polar regions which is caused
by its long alkyl-chains [38]. However, it can be seen a similar non-polar
characteristic with [C2C1mim]I, [C2py]I, [C3py]I, and [C4py]I due to the
contribution of the alkyl chain. Thus, it makes a weak solubility between
[TBA][I-I-I] with the ILs as mentioned above will be possible because of
this interaction. [C2C1mim]I, [C2py]I, [C3py]I, and [C4py]I have the same
trend with an increasing attraction towards the HBDs due to the pres-
ence of [I] anion which possesses a strong HBA. In addition, the interac-
tions of [C2C1mim] I, [C2py]I, [C3py]I and [C4py]I towards HBA are less
repulsive compared to [TBA]. This is due to the presence of N element
in the [C2C1im] I, [C2py]I, [C3py]I, and [C4py]I that exhibits a polarized
charge on the HBD region.
The σ-profile of independent [I-I-I] and I anions with [C2C1mim],
[C2py], [C3py], [C4py], and [TBA] cations are presented in Fig. 3. The
H-bond donor
Region
50
Non-polar
Region
H-bond acceptor
Region
[C2C1mim]
45
40
[C2py]
[C3py]
[C4py]
I
[TBA]
[C3py]
35
30
25
20
15
10
5
[C2py]
[C4py]
[TBA]
[I-I-I]
I
[I-I-I]
[C2C1mim]
0
-3.0
-2.0
-1.0
0.0
1.0
2.0
3.0
σ /e.nm-2
Fig. 3. σ-profiles of cations [TBA], [C2C1mim], [C2py], [C3py], [C4py] and anions [I-I-I] and I-−
.
5

N.I. Mohd Faridz Hilmy, W.Z.N. Yahya and K.A. Kurnia
Journal of Molecular Liquids 320 (2020) 114381
Interacꢀon with H-
bond donors
Interacꢀon with
non-polar groups
Interacꢀon with H-
bond acceptors
200.0
150.0
100.0
50.0
0.0
-50.0
-100.0
-150.0
-200.0
[C2C1mim]I
[C2py]I
[C3py]I
[C4py]I
[TBA][I-I-I]
-250.0
-3.0
-2.0
-1.0
0.0
1.0
2.0
3.0
σ/e.nm-2
Fig. 4. σ-profiles of the different ILs: [TBA][I-I-I], [C2C1im]I, [C2py]I, [C3py]I, and [C4py]I.
3.3. Density/molar volume for binary eutectic ILs
Since the pure ILs are solid and some of the mixtures were semi-solid
at room temperature, it was not possible to obtain the data in the tem-
perature range due to the limitation of the apparatus which cannot be
run at higher temperatures [41].
When two miscible liquids are mixed, the volume of the mixture
may not be equal to the sum of the volume of the individual component,
except for an ideal mixture. Hence, a binary ILs mixture is more likely to
have a deviation from ideality due to the molecular interactions be-
tween two compounds in the mixture. Furthermore, the binary ILs mix-
ture volume can either increase or decrease as a function of components
in the mixture [39]. This work provides interesting information regard-
ing the study of volumetric properties, where the criterion to measure
the molecular interactions at the molecular level is through the excess
molar volume [40]. The density values have been used to calculate ex-
cess molar volumes (VEm) to evaluate the deviation from the ideal binary
ILs mixture using the following Eq. (1):
Based on Fig. 5, the VEm shows a positive deviation for both binary eu-
tectic ILs of {[C3py]I: [C2C1mim]I} and {[C4py]I: [C2C1mim]I} over the en-
tire composition range. It shows that the molar volume of the mixture is
larger than the molar volume of ideal mixing. This suggests that there
are more free spaces in these mixtures and the sum of the favorable in-
teractions is fewer as expected [41]. As the mole fraction of [C4py] I de-
crease, the VEm increases. In other words, due to the disruption of
hydrogen bonding interactions between the hydrogen atoms of N-CH3
and the electron rich center of the anion from [C2C1mim]I, the interac-
tion between [C4py]I and [C2C1mim]I is the weakest when the amount
of [C2C1mim]I increases. A study by Dong et al. stated that the H-
bonds in imidazolium-based ILs perturb the Coulomb forces to deviate
from charge symmetry, resulting in the enhanced fluidization of ILs
[42]. The {[C3py]I:[C2C1im]I} mixture also followed the same trend of
having the increasing VEm due to the increasing amount of [C2C1im]I
while the mole fraction of [C3py]I decreases. Thus, a higher amount of
[C2C1im]I provides a higher positive deviation from the ideal mixture.
Apart from that, a large positive deviation of VEm for {[C4py]I:
[C2C1mim]I} as compared to {[C3py]I: [C2C1mim]I} can be attributed to
the dispersion forces between molecules due to the increasing alkyl
chain length, which is compensated by the weak interactions between
dissimilar molecules. Also, the higher alkyl chain length may hinder
the ions to pack closely and cause a decrease in density. Moreover
Motlagh et al., supported that the increase in alkyl chain length can in-
crease the flexibility of the alkyl chain. As a result, this may increase
the degree of freedom of the IL components in the volume and increas-
ing the volume [38]. Thus, when the density decreases, the molecular
volume will be increased and VEm will be increased.
VmE
cm3
:mol−1 ¼ Vmreal−Vmideal
ꢀ
ꢁ
ꢀ
ꢁ
ðx1M1 þ x2M2Þ
x1M1
ρ1
x2M2
ρ2
¼
−
−
ð1Þ
ρ12
where Vmreal is the molar volume of real fluid, Vmideal is the molar vol-
ume of ideal, ρ12 is the density of the mixture x1, M1, ρ1 and x2, M2, ρ2 are
the mole fractions, the molecular weights, and the densities of pure
components 1 and 2, respectively.
The experimental values of densities (ρ) of binary eutectic ILs at
298.15 K are listed as a function of mole fraction in Table 3. The investi-
gated binary ILs mixture are only focused on {[C3py] I:[C2C1mim]I} and
{[C4py]I:[C2C1mim]I} which form stable liquids at room temperature.
Table 3
Experimental values of densities (ρ) of binary eutectic ILs at
298.15 K as a function of mole fraction.
ILs
ρ (g·cm−3
)
3.4. Dynamic viscosity of binary eutectic ILs
[C2C1mim]I
[C3py]I
[C4py]I
1.7059
1.5712
1.5682
The viscosity measurement was conducted to study the degree of bi-
nary eutectic ILs' resistance to flow. The graphical representation of
temperature dependence of experimental dynamic viscosities for bi-
nary eutectic ILs from 293.15 K -353.15 K can be shown in Fig. 6. The
pure [C3py]I exhibits the highest viscosity as compared to other binary
eutectic ILs mixtures although it is liquid at room temperature. This
proves that the formation of eutectic ILs mixtures did decrease the over-
all flow resistance as compared to the pure ILs. Both [C2C1mim]I and
[C4py]I viscosity values cannot be determined due to its solid nature at
{x1[C3py] I: (1-x1) [C2C1im] I}
0.25
0.5
1.5974
1.5912
1.5783
0.75
{x1[C4py] I: (1-x1) [C2C1im] I}
0.25
0.5
1.5777
1.5531
1.5301
0.75
6

N.I. Mohd Faridz Hilmy, W.Z.N. Yahya and K.A. Kurnia
Journal of Molecular Liquids 320 (2020) 114381
10
9
8
7
6
5
4
3
2
1
0
0
{x1[C3py]I : (1-x1)[C2C1mim]I}
{x1[C4py]I : (1-x1)[C2C1mim]I}
0.25
0.5
x1
0.75
1
Fig. 5. Plot of excess molar volume (VEm) against mole fraction of [C3py]I and [C4py]I with [C2C1mim]I at T = 298.15 K.
2500
2000
1500
1000
500
studied range of 293.15 K in which {0.5[C4py]I:0.5[C2C1mim]I} has
lower viscosity value compared to {0.5[C3py]I:0.5[C2C1mim]I}. How-
ever, both eutectic mixtures exhibit comparable values at all studied
temperature range. The temperature dependence of the dynamic vis-
cosity can be further correlated as per Eq. (2).
[C3py]I
{0.5[C4py]I : 0.5[C2C1mim]I}
{0.75[C4py]I : 0.25[C2C1mim]I}
{0.5[C3pyr]I : 0.5[C2C1mim]I}
{0.75[C3py]I : 0.25[C2C1mim]I}
ln η ¼ ln A þ Ea=RT
ð2Þ
where η is the dynamic viscosity, A is the collision frequency, Ea is the
activation energy, T is the temperature of the ILs, and R is the universal
gas constant (8.314 J/mol·K). The viscosity of the binary ILs shows a
temperature dependence indeed following the Arrhenius model.
To further estimate the application of {0.5[C4py]I:0.5[C2C1mim]I} as
a potential electrolyte in DSSCs, nine different electrolytes are formu-
lated with the addition of [TBA][I-I-I] or iodine as triiodide ions provider
with the different composition of electrolytes can be referred in Table 1.
The viscosities of the electrolytes A-I as compared to the eutectic IL {0.5
[C4py]I:0.5[C2C1mim]I} are plotted in Fig. 7. For the electrolytes A-C
where [TBA][I-I-I] is tested as a triiodide provider, a significant decrease
of viscosities values is observed as compared to the {0.5[C4py]I:0.5
[C2C1mim]I} eutectic IL. Moreover, increasing the weight percentage of
[TBA][I-I-I] only results in increasing the viscosity of the mixture. The ef-
fect of higher viscosity at a high concentration of bulky [TBA][I-I-I] salt is
more likely due to the reduced free volume and the increase of inter-
ionic interactions which further decrease the fluidity of the mixture.
0
290
310
330
350
Temperature (K)
Fig. 6. Experimental viscosity, ŋ for [C3py]I, {0.5[C4py]I: 0.5[C2C1mim]I}, {0.75[C4py]I: 0.25
[C2C1mim]I}, {0.5[C3py]I: 0.5[C2C1mim]I} and {0.75[C3py]I: 0.25[C2C1mim]I} as function of
temperature.
room temperature. Both {0.75[C4py]I:0.25[C2C1mim]I} and {0.75[C3py]
I:0.25[C2C1mim]I} are the second most viscous ILs mixture due to the
steric hindrance of the pyridinium molecular structure which increases
the flow resistance. There is a slight difference of viscosity at the lowest
750
700
650
600
550
500
450
400
350
300
250
200
150
100
50
{0.5[C4pyr]I : 0.5[C2C1mim]I}
A
B
C
D
E
F
G
H
I
0
290
310
330
350
Temperature (K)
Fig. 7. Experimental viscosity, ŋ for eutectic salt based ionic liquids electrolyte as function of temperature.
7

N.I. Mohd Faridz Hilmy, W.Z.N. Yahya and K.A. Kurnia
Journal of Molecular Liquids 320 (2020) 114381
Table 4
Arrhenius Model. The Ea values represent the energy barriers that need to
be overcome by ions or mass transport. Based on Fig. 8, the steeper the
slope of the Arrhenius plot, the higher the Ea. Based on the viscosity-
temperature dependent trend, dynamic viscosity decreases with temper-
ature and increases with the increase in mole fraction. This is important to
be highlighted as a significant decrease in viscosity at a higher tempera-
ture will further enhance DSSCs performance under solar irradiation.
Activation energy value (Ea), collision frequency (A) and linear fitting parameters, R2 of
eutectic ionic liquids electrolytes.
Eutectic ionic liquids electrolytes
Ea (kJ/mol)
A
R2
{0.5[C4py]I: 0.[C2C1mim]I}
44.657
36.662
31.825
34.652
41.439
38.643
34.818
26.744
24.389
36.132
121,419.00
11,012.700
1524.920
4660.140
55,882.100
26,635.500
9026.320
600.763
0.9971
0.9972
0.9847
0.9951
0.9913
0.9926
0.9929
0.9945
0.9583
0.9987
A
B
C
D
E
F
G
H
I
3.5. Ionic conductivity
The ionic conductivity of the ILs originates from the inherent motion
of cations and anions in ILs under electric potential differences [44]. The
electrochemical impedance spectroscopy was carried out to investigate
the charge transfer ability. The ionic conductivity of {0.5[C4py]I:0.5
[C2C1im]I} and the electrolyte A – I were calculated using Eq. (3):
324.375
9656.630
The addition of I2 at different mass ratios is also evaluated to be com-
pared with [TBA][I-I-I] for the selection of the best triiodide provider.
When the mass ratio of I2 is increased, the viscosities values decreased
significantly. Compared to the electrolyte A-C, the electrolyte G with a
concentration of 50 wt% of I2 possessed the lowest viscosity and is
found to be lower compared to the literature viscosity data [3]. Pengfei
et al. reported a series of eutectic ILs electrolytes which comprised of
1,3-dimethylimidazolium iodide, 1-ethyl-3-methylimidazolium io-
dide, methylpyridinium iodide, ethylpyridinium iodide (DEMPI) in
presence of iodine at 50 wt%. The viscosity value was at 617 mPa·s
which is six-fold higher than our reported viscosity for electrolyte
G at 102.58 mPa·s. The effect of increasing concentration of I2 towards
eutectic ILs {0.5[C4py]I:0.5[C2C1mim]I} is therefore completely the op-
posite as compared to [TBA][I-I-I], which can be explained by the differ-
ent nature of these two salts. This can simply be explained by the larger
cation size of [TBA][I-I-I] which reduces the mobility as compared to I2
which in presence of iodide will form triiodide ions.
Furthermore, according to Fang et al., having a low viscous ILs can
help to enhance the diffusion of redox couple [43]. In the attempt to im-
prove the electrolyte performance, 1-butyl-3-methylimidazolium thio-
cyanate [BMIM][SCN] was added as a low-viscous IL to form electrolyte
H and I. As predicted, the addition of a [BMIM][SCN] decreases the vis-
cosity of the resulting electrolytes H and I. Some previous studies re-
ported that thiocyanate based ILs have enhanced the DSSC device
performance by reducing the diffusion coefficient of triiodides of the re-
sultant ILs electrolytes [3,15]. The eutectic ILs electrolytes are further
correlated with the Arrhenius model as per the Eq. (2). The value of
the activation energy, Ea, the pre-exponential factor, A, and the R-
square, R2, for the eutectic ILs electrolytes are shown in Table 4 with
the curves as shown in Fig. 8. The high R2 value which is closed to
0.99 indicated that the ILs mixtures were well-fitted with the proposed
l
σ ¼
ð3Þ
RsS
where σ is the ionic conductivity mS·cm−1, Rs is the ohmic resistance of
the electrolyte, S is the area of the electrodes, and l is the distance be-
tween two electrodes.
The electrochemical impedance values were measured and com-
pared, as shown in Fig. 9, with the data of solution resistance, Rs and
ionic conductivity, σ, are presented in Table 5. The equivalent circuit
model illustrates as inset, in which the high-frequency intercept on
the real axis represents the Rs; the impedance curve from the first semi-
circle located at the high-frequency region indicates the charge transfer
resistance for the I−3 reduction at the counter electrode (CE)-electrolyte
interface. Meanwhile, the second semicircle located relatively at the
lower-frequency region indicates the diffusion resistance of the I−/I3−
redox species. It can be found that the CE with the smaller charge trans-
fer resistance which corresponds to smaller impedance curves pos-
sesses the less overpotential for an electron transferring from the CE
to the electrolyte and great electrochemical ability [45,46]. The imped-
ance plot, such as the Nyquist plot was fitted using NOVA software,
and the values of Rs were determined from the generated equivalent cir-
cuits. All the chi-squared (X2) values, which is the fitting parameter of
the impedance curves are well-fitted by having the X2 that was closed
to (≤0.01).
Based on Fig. 9, the intercept of the curve with the real (Z′) axis gives
an estimation of the solution resistance, Rs. The electrolyte G displays
the smallest impedance curves for both semicircles and has the lowest
resistance value compared to others. It means that the higher addition
of iodine can increase the I−/I3− diffusion which can regenerate the
dye faster and improve the ionic conductivity [47,48].
7
6.5
6
{0.5[C4py]I : 0.5[C2C1mim]I}
A
B
C
D
E
F
5.5
5
4.5
4
3.5
3
G
H
I
2.5
2.8
3
3.2
3.4
3.6
1000/K
Fig. 8. Arrhenius plot of eutectic ionic liquids electrolyte.
8

N.I. Mohd Faridz Hilmy, W.Z.N. Yahya and K.A. Kurnia
Journal of Molecular Liquids 320 (2020) 114381
10000
9000
8000
7000
6000
5000
4000
3000
2000
1000
{0.5[C4py]I : 0.5[C2C1mim]I}
A
B
C
D
E
F
G
H
I
0
0
500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Z' (Ω.cm2)
Fig. 9. Nyquist plot for electrolyte A - I.
The ionic conductivity is frequently correlated with the viscosity as
both are influenced by the ion mobility of the mixture. A low viscosity
electrolyte will result in high ionic conductivity due to the increase of
the ion mobility of the mixtures [49,50]. For the electrolyte A-C, the in-
crease in [TBA][I-I-I], has caused the ionic conductivity of the mixture to
decrease due to an increase in viscosity of the electrolytes. Apart from
that, a higher addition of I2 into the mixture will increase the ionic con-
ductivity of the mixture correlated with the results of viscosity. It is due
to the increase of I−3 concentration towards the mixture. Previous re-
search reported that because of the ion aggregation/pairing, a large
ion size could cause the reduction of ion mobility and the reduction of
available charge carriers [44]. Hence, although the increase in the
[TBA][I-I-I] should make the ionic conductivity to be increased as the
I−3 concentration increases, due to the large size of tetrabutylammonium
cation, this has caused the ionic conductivity to be decreased.
All the tested electrolytes have followed the same trend of having
the less viscosity mixture with high ionic conductivity as displayed in
Table 5, except for the electrolyte H. The electrolyte H showed lower
ionic conductivity when [BMIM][SCN] was added to the eutectic mix-
ture. The previous study reported that the addition of a low viscous
non-active ionic liquid such as thiocyanate based ILs will help to in-
crease the ionic conductivity [3]. However, the electrolyte H exhibited
lower ionic conductivity as compared to corresponding electrolyte G
even though electrolyte H has the lowest viscosity as compared to
other electrolytes. Yuan et al. reported that besides the obvious influ-
ence of viscosity, the effect of ion size and formula weight must also
be stressed out [44]. Hence, these factors could be the cause of the de-
creasing ionic conductivity of Electrolyte H.
function of electrolyte is as a charge-transfer carrier, and its perfor-
mance is closely relevant to the viscosity and conductivity which deter-
mine the rate and quality of charge transfer respectively [51]. Several
studies also reported the successful incorporation of eutectic ILs in elec-
trolytes for DSSCs [3,52]. As mentioned earlier, the reported eutectic ILs
(DEMPI) by Pengfei et al. with the viscosity value of 617 mPa·s and ionic
conductivity of 1.54 mS·cm−1 have recorded PCE of 4.48% [3]. The com-
parable electrolyte G which is far less viscous (102 mPa·s) has also
shown higher ionic conductivity of 9.67 mS·cm−1 compared to the
previous work. Additionally, Bai et al. recorded eutectic ILs electrolytes
composed of 1,3-dimethylimidazolium iodide (DMII), 1-ethyl-3-
methylimidazolium iodide (EMII), and 1-allyl-3-methylimidazolium io-
dide (AMII) in presence of iodine, resulting in ionic conductivity of
3.16 mS·cm−1 and high power conversion efficiency of 7.1% with the ad-
ditives guanidinium thiocyanate (GNCS) and N-butylbenzimidazole
(NBB) [52]. All in all, the eutectic ILs reported here are highly potential
candidates to be introduced as electrolytes in DSSCs. Further studies on
the application of the eutectic ILs as electrolytes in DSSCs will be de-
scribed in future reports.
4. Conclusions
A series of alkylpyridinium iodide (C2 – C4) were synthesized and
mixed with [C2C1mim]I to form stable eutectic ionic liquid mixtures.
[C2C1mim]I, [C2py]I and [C4py]I are all in solid-state at room tempera-
ture while [C3py]I is in liquid state. From the thermal analysis of pure
ILs and the binary ILs mixtures, the depreciation of melting temperature
are observed for the binary eutectic ILs of {[C2py]I:[C2C1mim]I} with the
lowest melting point at equimolar ratio. For {[C3py]I:[C2C1mim]I} and
{[C4py]I:[C2C1mim]I at equimolar ratios, the eutectic mixtures have re-
sulted to the formation of highly stable liquid, with the lowest viscosity
and highest ionic conductivity. Therefore, these two compositions are
highly suitable for electrolyte application in DSSC. COSMO-RS and den-
sity measurement have confirmed that the reasons for the interaction
between ILs are the nitrogen (N element) that existed within the com-
ponents which can become a hydrogen bond donor and also its alkyl
chain length effect. The longer the alkyl chain length, the higher the pos-
itive deviation of VEm, thus, the density will be decreased. The studies of
the eutectic ILs as electrolytes with the incorporation of triiodide ions
have demonstrated a significant decrease in viscosity and an increase
in ionic conductivity with the highest value of 9.6751 mS·cm−1. These
results highlight that it is possible to prepare highly stable eutectic ILs
mixture with improved performance, from solid ILs at ambient temper-
ature thus allowing the design of new ILs as potential electrolytes com-
ponent for DSSCs application.
Previous research has proved that the trend of power conversion ef-
ficiency (PCE) is in good accordance with the results obtained from the
viscosity and ionic conductivity studies [19]. Gu et al. supported that the
Table 5
Ionic conductivity (σ) and RS values of Electrolyte A – I.
Electrolytes
Rs
σ (mS. cm−1
)
{0.5[C4py]I: 0.5[C2C1mim]I}
681
387
447
541
441
217
181
65.8
122
379
0.9348
1.6450
1.4242
1.1767
1.4436
2.9337
3.5172
9.6751
5.2182
1.6796
A
B
C
D
E
F
G
H
I
9

N.I. Mohd Faridz Hilmy, W.Z.N. Yahya and K.A. Kurnia
Journal of Molecular Liquids 320 (2020) 114381
devices: a brief review on their limits and applications, Polym. (Basel) 12 (2020)
918, https://doi.org/10.3390/polym12040918.
CRediT authorship contribution statement
[14] X. Huang, D. Qin, X. Zhang, Y. Luo, S. Huang, D. Li, Q. Meng, The potential of eutectic
mixtures as environmentally friendly, solvent-free electrolytes for dye-sensitized
solar cells, RSC Adv. 3 (2013) 6922, https://doi.org/10.1039/C3RA00137G.
[15] S.M.M. Yusof, W.Z.N. Yahya, Binary ionic liquid electrolyte for dye-sensitized solar
cells, Procedia Eng. 148 (2016) 100, https://doi.org/10.1016/j.proeng.2016.06.453.
[16] P.K. Singh, R.K. Nagarale, S.P. Pandey, H.W. Rhee, B. Bhattacharya, Adv. Nat. Sci.
Nanosci. Nanotechnol. 2 (2011), 023002, https://doi.org/10.1088/2043-6262/2/2/
023002.
Nurfathiah Izzaty Mohd Faridz Hilmy: Investigation, Data curation,
Writing - original draft. Wan Zaireen Nisa Yahya: Supervision, Concep-
tualization, Methodology, Data curation, Writing - review & editing.
Kiki Adi Kurnia: Conceptualization, Validation, Writing - review &
editing.
[17] M.S. Suait, M.Y.A. Rahman, A. Ahmad, Review on polymer electrolyte in dye-
sensitized solar cells (DSSCs), Sol. Energy 115 (2015) 452–470, https://doi.org/10.
1016/j.solener.2015.02.043.
Declaration of competing interest
[18] W. Yahya, W. Meng, M. Khatani, A. Samsudin, N. Mohamed, Bio-based chitosan/
PVdF-HFP polymer-blend for quasi-solid state electrolyte dye-sensitized solar
cells, e-Polymers 17 (5) (2017) 355–361, https://doi.org/10.1515/epoly-2016-0305.
[19] C.Y. Tan, N.M. Saidi, N. Farhana, F.S. Omar, M.M. Algaradah, S. Bashir, S. Ramesh, K.
Ramesh, Improved ionic conductivity and efficiency of dye-sensitized solar cells
with the incorporation of 1-methyl-3-propylimidazolium iodide, Ionics 1 (2020)
https://doi.org/10.1007/s11581-020-03447-2.
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
Acknowledgments
[20] G.P. Lau, J.-D. Décoppet, T. Moehl, S.M. Zakeeruddin, M. Grätzel, P.J. Dyson, Robust
high-performance dye-sensitized solar cells based on ionic liquid-sulfolane compos-
ite electrolytes, Sci. Rep. 5 (1) (2015)https://doi.org/10.1038/srep18158.
[21] S. Rezaei Motlagh, R. Harun, D.R. Awang Biak, S.A. Hussain, W.A. Wan Ab Karim
Ghani, R. Khezri, C.D. Wilfred, A.A.M. Elgharbawy, Screening of suitable ionic liquids
as green solvents for extraction of eicosapentaenoic acid (EPA) from microalagae
biomass using COSMO-RS model, Molecules 24 (2019) 713, https://doi.org/10.
3390/molecules24040713.
[22] H.W. Khan, A.V.B. Reddy, M.M.E. Nasef, M.A. Bustam, M. Goto, M. Moniruzzaman,
Screening of ionic liquids for the extraction of biologically active compound using
emulsion liquid membrane:COSMO-RS presiction and experiments, J. Mol. Liq. 309
(2020), 113122, https://doi.org/10.1016/j.molliq.2020.113122.
[23] T. Lemaoui, A.S. Darwish, N.E.H. Hammoudi, F. Abu Hatab, A. Attoui, I.M. Alnashef, Y.
Benguerba, Prediction of electrical conductivity of deep eutectic solvents using
COSMO-RS sigma profiles as molecular descriptors: a quantitative structure-
property relationship study, Ind. Eng. Chem. Res. (2020)https://doi.org/10.1021/
acs.iecr.0c02542.
This research was funded by Ministry of Higher Education (MOHE)
Malaysia through Fundamental Research Grant Scheme (Ref. FRGS/1/
2017/TK10/UTP/03/3) and partially funded by the Indonesian Ministry
of Research and Technology/National Agency for Research and Innova-
tion, and Indonesian Ministry of Education and Culture under World
Class University (WCU) Program managed by Institut Teknologi Ban-
dung. The authors would like to thank the various people who were in-
volved directly or indirectly to make this project done notably from
Chemical Engineering Department and Centre of Ionic Liquid (CORIL)
for the technical support.
Appendix A. Supplementary data
[24] S. Kuwahara, S. Taya, N. Osada, Q. Shen, T. Toyoda, K. Katayama, Effect of electrolyte
constituents on the motion of ionic species and recombination kinetics in dye-
sensitized solar cells, Phys. Chem. Chem. Phys. 16 (2014) 5242, https://doi.org/10.
1039/C3CP53964D.
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.molliq.2020.114381.
[25] Z.K. Koi, W.Z.N. Yahya, R.A. Abu Talip, K.A. Kurnia, Prediction of the viscosity of
imidazolium-based ionic liquids at different temperature using the quantitative
structure property relationship approach, New J. Chem. 43 (2019), 16207,
https://doi.org/10.1039/C9NJ03436F.
[26] G. Cavallo, D.W. Bruce, G. Terraneo, G. Resnati, P. Metrangolo, From molecules to
materials: engineering new ionic liquid crystals through halogen bonding, JoVE
(2018), e55636, https://doi.org/10.3791/55636.
[27] F.I. Chowdhury, M.H. Buraidah, A.K. Arof, B.E. Mellander, I.M. Noor, Impact of
tetrabutylammonium, iodide and triiodide ions conductivity in polyacrylonitrile
based electrolyte on DSSC perfromance, Sol. Energy 196 (2020) 379, https://doi.
org/10.1016/j.solener.2019.12.033.
[28] E. Boli, E. Voutsas, Ionic liquids as entrainers for the seperation of azeotropic mix-
tures: experimental measurements and COSMO-RS predictions, Chem. Eng. Sci.
219 (2020), 115579, https://doi.org/10.1016/j.ces.2020.115579.
[29] O. Stolarska, A. Soto, H. Rodríguez, M. Smiglak, Properties modification by eutectic
formation in mixtures of ionic liquids, RSC Adv. 5 (2015), 22178, https://doi.org/
10.1039/C4RA17268J.
[30] P. Zürner, H. Schmidt, S. Bette, J. Wagler, G. Frisch, Ionic liquid, glass or crystalline
solid? Structures and thermal behavior of (C4mim)2CuCl3, Dalton Trans. 45 (2016)
3327.
[31] E. Gómez, N. Calvar, Á. Domínguez, Thermal behaviour of pure ionic liquids, Ionic
Liquids-Current State of the Art 20 (2015) 199, https://doi.org/10.5772/59271.
[32] L. Timperman, A. Vigeant, M. Anouti, Eutectic mixture of protic ionic liquids as an
electrolyte for activated carbon-based supercapacitors, Electrochim. Acta 155
(2015) 164, https://doi.org/10.1016/j.electacta.2014.12.130.
[33] E. Siedlecka, M. Czerwicka, S. Stolte, P. Stepnowski, Stability of ionic liquids in appli-
cation conditions, Curr. Org. Chem. 15 (2011) 1974, https://doi.org/10.2174/
138527211795703630.
[34] J.E.K. Schawe, J.F. Löffler, Existance of multiple critical cooling rates which generate
different types of monolithic metallic glass, Nat. Commun. 10 (2019) 1337, https://
doi.org/10.1038/s41467-018-07930-3.
[35] J.O. Valderrama, R.A. Campusano, Melting properties of molten salts and ionic liq-
uids. Chemical homology, correlation, and prediction, C. R. Chimie 19 (2016) 654,
https://doi.org/10.1016/j.crci.2016.02.002.
[36] K.A. Kurnia, S.P. Pinho, J.A.P. Coutinho, Evaluation of the conductor-like screening
model for real solvents for the prediction of the water activity coefficient at infinite
dilution in ionic liquids, Ind. Eng. Chem. Res. 53 (2014), 12466, https://doi.org/10.
1021/ie5021415.
[37] H. Cheng, C. Liu, J. Zhang, L. Chen, B. Zhang, Z. Qi, Screening deep eutectic solvents
for extractive desulfurization of fuel based on COSMO-RS model, Chem. Eng. Pro-
cess. 125 (2018) 246, https://doi.org/10.1016/j.cep.2018.02.006.
References
[1] H. Iftikhar, G.G. Sonai, S.G. Hashmi, A.F. Nogueira, P.D. Lund, Progress on electrolytes
development in dye-sensitized solar cells, Materials 12 (2019) 1998, https://doi.org/
10.3390/ma12121998.
[2] J. Wu, Z. Lan, J. Lin, M. Huang, Y. Huang, L. Fan, G. Luo, Electrolytes in dye-sensitized
solar cells, Chem. Rev. 115 (2015) 2136, https://doi.org/10.1021/cr400675m.
[3] S. Pengfei, P. Bingkun, Z. Jie, W. Yun, Z. Fayun, R. Senlin, Z. Guifu, Multi-component
eutectic salts to enhance the conductivity of solvent-free ionic liquid electrolytes for
dye-sensitized solar cells, Electrochim. Acta 314 (2019) 219, https://doi.org/10.
1016/j.electacta.2019.05.055.
[4] M. Buraidah, S. Shah, L. Teo, F.I. Chowdhury, M. Careem, I. Albinsson, B.-E. Mellander,
A. Arof, High efficient dye sensitized solar cells using phthaloylchitisan based gel
polymer electrolytes, Electrochim. Acta 245 (2017) 846, https://doi.org/10.1016/j.
electacta.2017.06.011.
[5] M. Watanabe, M.L. Thomas, S. Zhang, K. Ueno, T. Yasuda, K. Dokko, Application of
ionic liquids to energy storage and conversion materials and devices, Chem. Rev.
117 (2017) 7190, https://doi.org/10.1021/acs.chemrev.6b00504.
[6] J.-D. Peng, C.-P. Lee, D. Velayutham, V. Suryanarayanan, K.-C. Ho, Dye-sensitized
solar cells containing mesoporous TiO2 spheres as photoanodes and methyl sulfate
anion based biionic liquid electrolyts, J. Mater. Chem. A 3 (2015) 6383, https://doi.
org/10.1039/C5TA00316D.
[7] S. Pan, M. Yao, J. Zhang, B. Li, C. Xing, X. Song, P. Su, H. Zhang, Recongnition of ionic
liquids as high-voltage electrolytes for supercapacitors, Front. Chem. 8 (2020)
https://doi.org/10.3389/fchem.2020.00261.
[8] A. Alhadid, L. Mokrushina, M. Minceva, Design of deep eutectic systems: a simple
approach for preselecting eutectic mixture constituents, Molecules 25 (2020)
1077, https://doi.org/10.3390/molecules25051077.
[9] G.J. Maximo, R.J.B.N. Santos, P. Brandão, J.M.S.S. Esperança, M.C. Costa, A.J.A.
Meirelles, M.G. Freire, J.A.P. Coutinho, Generating ionic liquids from ionic solids:
an investigation of the melting behavior of binary mixtures of ionic liquids, Cryst.
Growth Des. 14 (2014) 4270, https://doi.org/10.1021/cg500655s.
[10] F. Elhi, M. Gantman, G. Nurk, et al., Influence of carboxylate anions on phase behav-
ior of choline ionic liquid mixtures, Molecules 25 (7) (2020) 1691, https://doi.org/
10.3390/molecules25071691.
[11] R.J. Isaifan, A. Amhamed, Review on carbon dioxide absorption by choline chloride/
urea deep eutectic solvents, Adv. Chem. 2018 (2018) 2675659, https://doi.org/10.
1155/2018/2675659.
[12] B.Y. Liu, F. Li, J. Tian, H.L. Yang, Properties of acetamide-alanine eutectic ionic liquid
and its application in soil heavy metal Cd2+ remediation, Key Eng. Mater. 842
(2020) 115, https://doi.org/10.4028/www.scientific.net/KEM.842.115.
[13] K. Karuppasamy, J. Theerthagiri, D. Vikraman, C.-J. Yim, S. Hussain, R. Sharma, T.
Maiyalagan, J. Qin, H.-S. Kim, Ionic liquid-based electrolytes for energy storage
[38] Z. Wu, C. Liu, H. Cheng, L. Chen, Z. Qi, Tuned extraction and regeneration process for
separation of hydrophobic compounds by aqueous ionic liquid, J. Mol. Liq. 308
(2020), 113032, https://doi.org/10.1016/j.molliq.2020.113032.
10

N.I. Mohd Faridz Hilmy, W.Z.N. Yahya and K.A. Kurnia
Journal of Molecular Liquids 320 (2020) 114381
[39] A.R.S. Christabel, D.J.P.M. Reji, A. Ramalingam, Evaluation of solution thermody-
namic properties of mixed ionic liquids at different temperatures (293.15 K -
343.15 K), Laboratory Unit Operations and Experimental Methods in Chemical Engi-
neering (2018) 123, https://doi.org/10.5772/intechopen.77016.
[40] R.S. Pinheiro, F.M.R. Mesquita, F.X. Feitosa, H.B.d. Sant’Ana, R.S.d. Santiago-Aguiar,
Density, viscosity and excess properties of binary mixtures of protic ionic liquid
(2-HDEAF, 2-HDEAA) + water at different temperatures, Braz. J. Chem. Eng. 35
(2018) 383, https://doi.org/10.1590/0104-6632.20180352s20160280.
[41] M.T. Clough, C.R. Crick, J. Gräsvik, P.A. Hunt, H. Niedermeyer, T. Welton, O.P.
Whitaker, A physicochemical investigation of ionic liquid mixtures, Chem. Sci. 6
(2015) 1101, https://doi.org/10.1039/C4SC02931C.
[46] A. Sacco, Electrochemical impedance spectroscopy: fundamentals and application in
dye-sensitized solar cells, Renew. Sust. Energ. Rev. 79 (2017) 814, https://doi.org/
10.1016/j.rser.2017.05.159.
[47] F. Javed, F. Ullah, M.R. Zakaria, H.M. Akil, An approach to classification and hi-tech
applications of room-temperature ionic liquids (RTILs): a review, J. Mol. Liq. 271
(2018) 403, https://doi.org/10.1016/j.molliq.2018.09.005.
[48] R. Atasiei, M. Raicopol, C. Andronescu, A. Hanganu, A.L. Alexe-Ionescu, G. Barbero, In-
vestigation of the conduction properties of ionic liquid crystal electrolyte used in
dye sensitized solar cells, J. Mol. Liq. 267 (2018) 81, https://doi.org/10.1016/j.
molliq.2018.01.010.
[49] S.N.F. Yusuf, R. Yahya, A.K. Arof, Ionic liquid enhancement of polymer electrolyte
conductivity and their effects on the performance of electrochemical devices, Prog-
ress and developments in ionic liquids, Rijeka: InTech 157 (2017)https://doi.org/10.
5772/65752.
[42] K. Dong, S. Zhang, J. Wang, Understanding the hydrogen bonds in ionic liquids and
their roles in properties and reactions, ChemComm 52 (2016) 6744, https://doi.org/
10.1039/C5CC10120D.
[43] Y. Fang, P. Ma, H. Cheng, G. Tan, J. Wu, J. Zheng, X. Zhou, S. Fang, Y. Dai, Y. Lin, Syn-
thesis of low-viscosity ionic liquids for application in dye-sensitized solar cells,
Chem. Asian J. 14 (2019) 4201, https://doi.org/10.1002/asia.201901130.
[44] W.-L. Yuan, X. Yang, L. He, Y. Xue, S. Qin, G.-H. Tao, Viscosity, conductivity, and elec-
trochemical property of dicyanamide ionic liquids, Front. Chem. 6 (2018)https://doi.
org/10.3389/fchem.2018.00059.
[50] K.H. Seol, S.J. Lee, K.G. Cho, K. Hong, K.H. Lee, Highly conductive, binary ionic liquid-
solvent mixture ion gels for effective switching of electrolyte-gated transistors, J.
Mater C 6 (2018), 10987, https://doi.org/10.1039/C8TC03076F.
[51] P. Gu, D. Yang, X. Zhu, H. Sun, P. Wangyang, J. Li, H. Tian, Influence of electrolyte pro-
portion on the performance of dye-sensitized solar cells, AIP Adv. 7 (2017), 105219,
https://doi.org/10.1063/1.5000564.
[45] X. Ma, G. Yue, J. Wu, Z. Lan, Efficient dye-sensitized solar cells made from high cat-
alytic ability of polypyrrole@platinum counter electrode, Nanoscale Res. Lett. 10
(2015) 1015, https://doi.org/10.1186/s11671-015-1015-z.
[52] Y. Bai, Y. Cao, J. Zhang, M. Wang, R. Li, P. Wang, S.M. Zakeeruddin, M. Grätzel, High
performance dye-sensitized solar cells on solvent-free electrolytes produced from
eutectic melts, Nat. 7 (2008) 626, https://doi.org/10.1038/nmat2224.
11