Journal of Molecular Liquids 215 (2016) 534–540

Contents lists available at ScienceDirect

Journal of Molecular Liquids

j ourna l homepage: www.e lsev ie r .com/ locate /mol l iq

Properties and thermal behavior of natural deep eutectic solvents

R. Craveiro a, I. Aroso b,c, V. Flammia a, T. Carvalho a, M.T. Viciosa d, M. Dionísio a, S. Barreiros a, R.L. Reis b,c,A.R.C. Duarte b,c, A. Paiva a,⁎a LAQV—REQUIMTE, Departamento de Química, Faculdade de Ciências e Tecnologia, Universidade Nova de Lisboa, 2829-516 Caparica, Portugalb 3B's Research Group— Biomaterials, Biodegradable and Biomimetic, University of Minho, Headquarters of the European Institute of Excellence on Tissue Engineering and Regenerative Medicine,Avepark, Barco, 4805-017 Guimarães, Portugalc ICVS/3B's PT Government Associated Laboratory, Braga/Guimarães, Portugald Centro de Química-Física Molecular and IN — Institute of Nanoscience and Nanotechnology, Instituto Superior Técnico, Univ. Lisboa, Av. Rovisco Pais, 1049-001 Lisboa, Portugal

⁎ Corresponding author.E-mail address: alexandre.paiva@fct.unl.pt (A. Paiva).

http://dx.doi.org/10.1016/j.molliq.2016.01.0380167-7322/© 2016 Elsevier B.V. All rights reserved.

Downloaded from http://www.elea

a b s t r a c t

a r t i c l e i n f o

Article history:Received 7 August 2015Accepted 11 January 2016Available online xxxx

Natural deep eutectic solvents (NADES) have shown to be promising sustainable media for a wide range ofapplications. Nonetheless, very limited data is available on the properties of these solvents. A more compre-hensive body of data on NADES is required for a deeper understanding of these solvents at molecular level,which will undoubtedly foster the development of new applications. NADES based on choline chloride, organicacids, amino acids and sugars were prepared, and their density, thermal behavior, conductivity and polaritywere assessed, for different NADES compositions. The NADES studied can be stable up to 170 °C, depending ontheir composition. The thermal characterization revealed that all the NADES are glass formers and some,after water removal, exhibit crystallinity. The morphological characterization of the crystallizable materialswas performed using polarized optical microscopy which also provided evidence of homogeneity/phaseseparation. The conductivity of the NADES was also assessed from 0 to 40 °C. The more polar, organic acid-based NADES presented the highest conductivities. The conductivity dependence on temperature was welldescribed by the Vogel–Fulcher–Tammann equation for some of the NADES studied.

© 2016 Elsevier B.V. All rights reserved.

Keywords:Deep eutectic solventsPhysical propertiesCholine chlorideThermal analysisConductivity

1. Introduction

Green technology requires new solvents to replace common organicmedia that present inherent toxicity and have high volatility. Over thepast two decades, ionic liquids (ILs) have attracted great attentionfrom the scientific community, and the number of articles focusing onILs has grown exponentially. ILs aremolten salts, liquid at room temper-ature, whose potential relies on the possibility to tune their propertiesthrough the combination of different cations and anions [1]. Never-theless the “green” character of ILs is often questioned, mainly due totheir poor biodegradability, biocompatibility and sustainability.

Deep Eutectic solvents (DES) are obtained upon mixing two com-pounds in such a ratio that the resulting substance has a significantlylower melting point than that of each individual component [2]. Themost common DES are based on choline chloride (ChCl), carboxylicacids and other hydrogen-bonddonors, such as urea, citric acid, succinicacid, and glycerol. DES may have similar characteristics to ILs, such aslow vapor pressure, but are cheaper to produce, both due to the lowercost of the required raw materials and the simplicity of the synthesis.Furthermore, they are less toxic and often biodegradable [3]. RecentlyDai and co-workers have reported on a large number of stable natural

rnica.ir

deep eutectic solvents (NADES), based on primary metabolites, suchas organic acids, amino acids, and sugars [3]. Paiva and co-workershave recently reported on the cytotoxicity of different NADES, showingthat it is much lower than that of commonly used imidazolium-basedILs [4]. Radošević et al. have also studied three ChCl-based DES and sug-gested that they can be classified as readily biodegradable presentinglow to moderate toxicity [5]. These characteristics have led to growinginterest in the research community in replacing ILswithDES, as solventsfor biocatalysis [6,7], extraction [8] and chemical conversion [9] oforganic compounds, and polymer synthesis. As regards polymer pro-cessing, DES have been shown to dissolve bioactive materials and bio-polymers [2]. Bioactive DES with active pharmaceutical ingredients(APIs), such as ibuprofen [10], can be incorporated in biopolymersthrough the doping of the biopolymer matrix.

Amore detailed characterization of DES can lead to further scientificdevelopments. Dai et al. have characterized some NADES by nuclearmagnetic resonance (NMR) spectroscopy and concluded that waterplayed an important role in NADES formation [3]. In the case of theNADES composed by 1,2-propanediol, ChCl and water, the authors ob-served a strong interaction between the hydroxyl groups of all species.In addition, Dai and co-workers also determined the thermal and phys-ical characteristics of some of NADES with water in its composition [3].Florindo et al. also reported on the strong influence of water onthe properties of ChCl:carboxylic acid-based DES [11]. The thermal

535R. Craveiro et al. / Journal of Molecular Liquids 215 (2016) 534–540

properties of DES were also presented by those authors, with special at-tention to the glass transition temperature (Tg) and decomposition tem-perature (Td). Rengstl and co-workers have recently reported on thethermal behavior of DES based on different choline ILs [12].

Following thework of Dai et al. [3], we preparedNADES composed ofdifferent sugars, organic acids and ChCl, in the ratios reported and ac-cording to the procedures described by those authors. We measured anumber of properties of these NADES. We also used polarized opticalmicroscopy measurements coupled with differential scanning calorim-etry analysis to better understand the thermal behavior of NADES,e.g., the influence of water on the glass transition andmelting tempera-ture. Density, polarity and conductivity measurements were alsoperformed.

ChCl can form NADES with almost any kind of primary metabolitesand has been used to prepare most of the DES reported in the literature.Therefore ChCl is present in most of the NADES we selected for ourstudy. In addition, three sugar based NADES are also studied in thiswork.

2. Material and methods

2.1. Materials

Choline chloride (ChCl) (N98% purity, CAS number 67-48-1), D-(+)-xylose (99% purity, CAS number 58-86-6), citric acidmonohydrate (CASnumber 5949-29-1), D-(+)-glucose (99% purity, CAS number 50-99-7),Nile red (N98% purity, CAS number 7385-67-3), andHydranal CoulomatAG were obtained from Sigma-Aldrich. D-(+)-sucrose (98% purity, CASnumber 57-50-1) were obtained from Fluka. L-(+)-tartaric acid (N99%purity, CAS number 87-69-4) was obtained from Fisher scientific. Allchemicals were used without further purification.

2.2. Preparation of NADES

NADES were prepared according to Table 1. Weighed amounts ofeach component, as required to achieve the molar ratios indicated inthe table, were dissolved in water. The two solutions were mixed, andwater was removed in a rotary evaporator at 50 °C under vacuum,until a clear viscous liquid was obtained. NADES were then kept undervacuum for 24 h, after which they were stored in a desiccator.

2.3. Differential scanning calorimetry analysis (DSC)

To determine the degradation temperature of the NADES (Td), ex-periments were performed using a DSC Q100 equipment (TA Instru-ments — ELNOR). The experiments were conducted under a nitrogenatmosphere, with samples of 5–10 mg packed in aluminum pans. Thesamples were heated at a constant heating rate of 20 °C min−1, from−40 °C up to 250 °C. The results presented are the average of at leastthree measurements.

Table 1NADES prepared in this study, respective abbreviations and water content.

NADESMole ratio

Component 1 Component 2

Choline chloride D-(+)-glucose 1:1Choline chloride Citric acid 1:1Choline chloride D-(+)-sucrose 4:1

Choline chloride D-(+)-sucrose 1:1

Choline chloride L-(+)-tartaric acid 2:1

Choline chloride D-(+)-xylose 2:1

Choline chloride D-(+)-xylose 3:1

Citric acid D-(+)-sucrose 1:1

Citric acid D-(+)-glucose 1:1

D-(+)-glucose L-(+)-tartaric acid 1:1

To assess the thermal behavior of the NADES, calorimetric experi-ments were carried out with a DSC Q2000 from TA Instruments Inc.(Tzero™ DSC Technology) operating in the Heat Flow T4P option [13].Measurements were performed under dry high purity helium, at aflow rate of 50 mL ∙min−1. Less than 5 mg of each sample were encap-sulated in Tzero aluminum pans. The set was not hermetically sealedto allow free water evaporation. At least two scans at cooling andheating rates of 20 °C ∙min−1 were performed, covering the tempera-ture range from −90 °C to 120 °C. Each sample was kept for one addi-tional minute at 120 °C at the end of the scan, to ensure waterremoval. Also each samplewas kept for 10min at−90 °C in order to ob-tain a better signal of the glass transition temperature, when present.

2.4. Polarized optical microscopy measurements

Polarized optical microscopy was performed on an Olympus Bx51optical microscope equipped with a Linkam LTS360 liquid nitrogen-cooled cryostage. The microstructure of the samples was monitoredby taking microphotographs at appropriate temperatures, using anOlympus C5060 wide zoom camera. A drop of each sample was posi-tioned on a microscope slide and inserted in the hot stage. Before eachmeasurement, the samples were heated to 120 °C and kept at least10 min at this temperature to allow water removal; after this thermaltreatment a cover slip was placed on the top of the sample. Coolingand heating thermal treatments were carried out at a rate of20 °C ∙min−1.

2.5. Water content determination

The water content of the NADES was determined after drying undervacuum for 24 h upon preparation. A 831 KF Coulometerwith generatorelectrode and without diaphragm was used. The water content valuesgiven are an average of at least three measurements.

2.6. Density measurements

The density of the NADES was measured following a simple gravi-metric procedure, using a calibrated volume at 23 °C.

2.7. Conductivity measurements

The conductivity of the different NADES was assessed by dielectricrelaxation spectroscopy (DRS). For the DRS measurements, sampleswere placed between two stainless steel electrodes (10 mm diameter)in a BDS 1200 parallel plate capacitor, using two 50 μm silicon spacersto maintain sample thickness. The sample cell was mounted on a BDS1100 cryostat, and exposed to a gas stream resulting from the evapora-tion of liquid nitrogen in a Dewar. Temperature control was ensured bya Quatro Cryosystem controller and performed to within ±0.5 °C (allmodules supplied by Novocontrol). Measurements were carried out

Sample nameWater content/wt.%

Density/g ∙mL−1

ChCl:gluc (1:1) 5.5 1.27ChCl:ca (1:1) 0.2 1.30ChCl:suc (4:1) 0.2 1.22

ChCl:suc (1:1) 0.2 1.35

ChCl:ta (2:1) 1.9 1.26

ChCl:xyl (2:1) 3.8 1.23

ChCl:xyl (3:1) 0.2 1.22

ca:suc (1:1) 1.2 1.43

ca:gluc (1:1) 0.5 1.45

gluc:ta (1:1) 0.4 1.46

536 R. Craveiro et al. / Journal of Molecular Liquids 215 (2016) 534–540

using an Alpha-N analyzer also from Novocontrol, covering a frequencyrange from 10−1 Hz to 1 MHz. After a first cooling ramp from roomtemperature to 0 °C, isothermal spectra were collected from 0 °C to40 °C, in steps of 5 °C.

2.8. Polarity measurements

Polarity measurements were carried out using Nile red assolvatochromic probe. A stock solution was prepared by dissolving1 g ∙L−1 of Nile red in ethanol, and stored at 4 °C. The NADES samplewas placed in a 1 cm2 quartz cuvette, and a blank was recorded. Afterthat 50–100 μL of Nile red stock solutionwere added to theNADES sam-ple. The cuvette was placed under a gentle high-purity nitrogen gasstream to evaporate the solvent, and the UV spectra were immediatelyacquired, at 23 °C. Spectra of Nile red solutions with different amountsof water were also recorded in order to study the influence of thewater amount on the wavelength of maximum absorbance (λmax) ofNile red.

3. Results and discussion

3.1. Thermal characterization and polarized light microscopy (POM)

The thermal stability of NADES was evaluated by DSC up to 250 °C.All NADES present a single degradation peak at temperatures aboveca. 120 °C (decomposition temperatures presented in Table 2). DSCwas further used to probe thermally induced transformations over sev-eral scans; the individual components of the NADESwere also analyzed.

For all the NADES studied in the present work, a discontinuity in theheat flux is observed at lower temperatures in the first scan, due to theglass transition. This is a second order transition, with no latent heat as-sociated, and in DSC it appears as a step transition. It corresponds to achange in the structure of the material from a glass-like state to arubber-like state, or vice-versa, reflected in a jump in heat capacity.The detection of a glass transition fromwhich a glass transition temper-ature (Tg) value can be determined allows us to classify all the NADEStested as glass formers; this is also true for all the pure components,with the exception of ChCl. In most cases, Tg was lowest in the firstscan, and shifted to higher temperatures in following scans, until it be-came constant. This effect can be attributed to water evaporation. Theglass transition temperature decreases upon hydration due to a plasti-cizing effect of water, as already observed [14]. In order to determine

Table 2Thermal properties obtained from DSC data for the NADES under study, namely degrada-tion temperature (Td), melt and cold crystallization temperatures (Tc,melt and Tc,cold), melt-ing temperature (Tm), and the temperature of the onset of the glass transition (Tg),obtained in the last DSC cycle.

NADESTd/°C

Tc,melt

/°CTc,cold/°C

Tm/°C

Tg/°C

ChCl:gluc (1:1) 129.8 – – −28.4ChCl:ca (1:1) 171.3 −9.7 76.0 −21.4ChCl:suc (4:1) 141.7 −33.9 51.8 79.2 −42.0ChCl:suc (1:1) 126.8 – – −15.8ChCl:ta (2:1) 130.8 – – −41.6ChCl:xyl (3:1) 165.2 20.1 59.9 78.5 −46.4ChCl:xyl (2:1) 172.7 – 56.9 78.3 −51.2ca:suc (1:1) 121.2 – – −14.0ca:gluc (1:1) 130.1 – – 9.8/48.7a

gluc:ta (1:1) 117.5 – – −18.3Glucose −39.2b/−37.0c

Sucrose −39.1b

Xylose −43.8b

Citric acid 64.2 −38.2Tartaric acid −40.1b

Choline chloride 46.5–33.2 90.8 –

a Two Tg values are observed in the thermogram for this sample.b The flux discontinuity across the glass transition is ill-defined.c Sample with 12.8 wt.% of water.

the effect of water in the glass transition, water was added to theChCl:xyl (2:1) NADES. As can be observed in Table 3, up to 5 wt.% ofwater added Tg decreases about 4 °C, thus confirming the plasticizing ef-fect of water in NADES.

Table 3 also shows the effect ofwater on Tmof theNADES.NADES areobtained by the complexation of a hydrogen-bond acceptor and ahydrogen-bonddonor. Therefore, waterwill have a significant influenceon the complex, altering the properties of the NADES and interfering inthe liquidus and solidus lines of the eutectic phase diagram. As forChCl:xyl (2:1), with the addition of water only one thermal event corre-sponding to melting is observed. At 4 wt.% water added to the NADES,Tm suffers a depression of 1 °C, while at higherwater contents the oppo-site effect occurs and an increase in the melting temperature is ob-served. In fact, Dai and co-workers observed a strong interactionbetween the hydroxyl groups of water and the hydroxyl groups of xy-lose and choline, and with the chloride anion [3]. The destabilizationof the supramolecular structure of NADES bywater increases themobil-ity of both components of the NADES, which translates into a decreasein Tg. This agrees with Dai et al. who reported a Tg of −81.8 °C forChCl:xyl (2:1) with 7.74% water content. The impact of the water con-tent on Tg observed for NADES is not observed with the pure compo-nents, for which almost no variation occurs in the respective Tg(Table 2). E.g. after the addition of 12.8 wt.% water to glucose, only asmall change in Tg occurs.

For gluc:ta (1:1), illustrated in Fig. 1 (a), and most of the NADESstudied, the glass transition is the only thermal event detected(Table 2). Polarized optical microscopy (POM)was also used to observethe phase transformations detected by DSC. Fig. 1 (a) includes a micro-photograph taken by POM at 0 °C. It illustrates the micro-cracks thatemergedon heating at−57 °C (well below theTg), covering all the sam-ple. Observation of the cracks confirms the vitreous state of the sample,as reported for amorphous polymers [15]. The cracks start to disappearwhen the sample is further heated slightly above the glass transition,giving rise to a homogeneous dark image, characteristic of the super-cooled and isotropic liquid.

The ca:gluc (1:1) NADES has a different behavior, exhibiting twoglass transitions (Fig. 1 (b)). This suggests that two glasses were formedon cooling. The phase separation is clearly shown in the POM micro-graphs of Fig. 1 (b), where two phases are observed in both the vitreousstate and supercooled state. Nevertheless, cracks only emerge in one ofthe sample phases in the sub-glass region. The microphotograph takenby POM near −55 °C, a temperature below the two calorimetric glasstransitions, reveals the coexistence of two different glasses by the ob-servation of micro-cracks that only emerge in a definite region of thesample (inset of Fig. 1 (b)). Crack formation was also detected by DSCthrough a sharp discontinuity in the heat flux near −50 °C, deep inthe glassy state (blue line in Fig. 1 (b)). A similar behavior was foundfor other lowmolecular weight glass formers [16,17]. The two domainsare still observed at a temperature above the two glass transitions de-tected by DSC, evidencing that phase separation persists when thewhole sample is in the supercooled liquid state.

In the case of ChCl:ca (1:1), ChCl:suc (4:1), ChCl:xyl (2:1) andChCl:xyl (3:1), a more complex thermal behavior is observed. In addi-tion to the glass transition, these NADES exhibit exo and endothermicevent peaks due to crystallization andmelting. Upon thermal treatmentto 120 °C, ChCl:ca (1:1) undergoes partial crystallization on cooling

Table 3Influence of the amount of water on the Tg and Tm values, measured by DSC. Data forChCl:xyl (2:1) NADES.

Water content (wt.%) Tg first scan (°C) Tm (°C)

2.2 −58.5 77.13.8 −59.1 76.25.0 −61.3 77.07.74 [3] −81.8 [3] –

Fig. 1. (a)Thermogramobtained at 20 °Cmin−1 for gluc:ta (1:1). Themicrophotograph includedwas takenbyPOMat 0 °Cduring a heating cycle. (b) Thermogramobtained at 20 °Cmin−1

for ca:gluc (1:1). The microphotographs included were taken by POM at−70 °C during a cooling cycle (top left), and at 95 °C during a heating cycle (bottom right). Cooling and heatingrates of 20 °C min−1. Degree of magnification of POMmicrophotographs: 40×. (For interpretation of the references to color in this figure, the reader is referred to the web version of thisarticle.)

537R. Craveiro et al. / Journal of Molecular Liquids 215 (2016) 534–540

from the liquid state, which is called melt crystallization [13], and therespective melting is observed in the following heating scan. On theother hand, ChCl:xyl (2:1) fails to crystallize on cooling, but crystallizesupon heating from the glass, i.e., it exhibits cold-crystallization. This is atype of crystallization undergone by an amorphous material uponreheating above its glass transition [13]. Fig. 2 illustrates all the phasetransformations for ChCl:xyl (3:1). Interestingly, ChCl:suc (4:1) andChCl:xyl (3:1) undergo both melt and cold crystallizations, in additionto the glass transition.

The temperatures of melting and crystallization, in the cases wheresuch processes were observed, are given in Table 2. All the NADES thatundergo crystallization present a sharp endothermic melting peak attemperatures close to 80 °C. Only the NADES with ChCl in their compo-sition exhibit crystallization. This can be an indication that the compo-nent that determines crystallization is ChCl. Pure ChCl melts between85.2 °C and 90.8 °C, meaning that mixing with the second componentinfluencesmelting by slightly decreasing Tm. The effect ismore dramaticin the case of crystallization. For pure ChCl, only melt crystallization isobserved between 46.5 and 33.2 °C. On the other hand, NADES

Fig. 2. Thermogram obtained at 20 °C ∙min−1 for ChCl:xyl (3:1) after water removal at100 °C. The microphotographs included were taken by POM at −10 °C during a coolingcycle (top left), and at 30 °C during a heating cycle (bottom right). Cooling and heatingrates of 10 °C min−1. Degree of magnification of POM microphotographs: 40×.

containing ChCl exhibit both melt and cold crystallizations, and theseprocesses occur at temperatures quite different than for pure ChCl.The differences found for the crystallization phenomenon vs. meltingare not completely unexpected sincewhile melting is purely thermody-namic in nature, crystallization is also controlled by kinetic factors thatinfluence nuclei formation and crystal growth [18]. Using transmittedlight microscopy with cross polarizers, it is possible to obtain informa-tion on the isotropy of the NADES at different experimental conditions.

3.2. Density

All the NADES have densities over 20% higher than that of water,NADES containing ChCl having comparatively lower densities, asshown in Table 2.

3.3. Conductivity

The NADES were studied over a range of temperatures and frequen-cies by dielectric relaxation spectroscopy (DRS) which, for a sampleunder the influence of an oscillating electrical field, mainly probesreorientationalmovements of dipoles and propagation ofmobile chargecarriers. Themigration of charge carriers is due to translational diffusionthrough hopping movements of electrons, holes and ions, giving rise toconductivity. The conductivity of each sample was measured from10−1 Hz to 106 Hz, in a range of temperatures from 0 to 40 °C, in stepsof 5 °C. The property under measurement is the complex dielectricfunction, ε*:

ε� ωð Þ ¼ ε0ωð Þ � iε″ ωð Þ ð1Þ

where Ω is the angular frequency. The real part of ε* is related withenergy stored by the system, while the imaginary part accounts for theenergy dissipated inside the material. Ω is given by:

ω ¼ 2πf ð2Þ

where f is the frequency of the outer electrical field. One of thealternative representations of the dielectric response is the complexconductivity, σ⁎ [19]:

σ� ωð Þ ¼ iωε0ε� ωð Þ ð3Þ

where ε0 is the vacuum permittivity. Since the propagation of mobilecharge carriers also contributes to the complex function, it can be

Table 4Fitting parameters obtained by applying the VFT equation (Eq. 5).

NADESσ∞

/S ∙cm−1B/K

T0/K

ChCl:xyl (3:1) 1063 ± 812 1107 ± 301 136 ± 2ca:suc (1:1) 556 ± 333 1556 ± 223 127 ± 1ca:gluc (1:1) 1.3 ± 0.7 397 ± 84 227 ± 6

NOTE: The uncertainties are the statistical errors given by the fitting program. For eachmaterial, the similarity between B and T0 estimated throughσdc (T) indicates a parallelismbetween these two quantities.

Fig. 3. Conductivity as a function of frequency for the sample ca:suc (1:1).

538 R. Craveiro et al. / Journal of Molecular Liquids 215 (2016) 534–540

advantageous to analyze the dielectric response through the complexconductivity function:

σ � ωð Þ ¼ σ0ωð Þ þ iσ ″ ωð Þ ð4Þ

Conductivity data can elucidate the charge transport mechanism ofthe material [19]. In the present study, the analysis was based mainlyon the real part of the conductivity, which has a universal behavior fora variety of disordered conductive materials, exhibiting a plateau atthe lowest frequencies and highest temperatures. At this plateau, theconductivity is frequency independent and by extrapolating to Ω → 0,the dc conductivity, σdc, can be determined. At higher frequencies, theconductivity bends off into a frequency dependent regime, with a pro-nounced increase of the conductivity with increasing frequency [20].Conductivity data were taken from this plateau for each isotherm,i.e., a set of very similar data points were extracted from the plateau,for which an average value along with respective standard deviationwere calculated.

Fig. 4. Temperature dependence of the pure conductivity, σdc, of NADES. • ChCl:xyl (3:1),✮ ca:suc (1:1), ✰ ca:gluc (1:1). The solid lines were obtained by fitting with the VFTequation.

An important feature in the σ′(ν) plot, where ν is the frequency ofthe applied oscillating electrical field, is the occurrence of a plateau, asshown in Fig. 3 for ca:suc (1:1).

This plateau gives σdc, the direct conductivity. At lower tem-peratures, or higher frequencies, the conductivity becomes frequencydependent and the σ′(ν) plot presents a pronounced increase. The con-ductivities of the NADES studied, are an average of the values at theplateau. ChCl:suc (1:1) and ChCl:suc (4:1) were found to exhibit thehighest conductivities with values of 1.35 S ∙cm−1 and 1.28 S ∙cm−1 at25 °C respectively. The conductivities of these two NADES are ofthe same order of magnitude as those of some ionic liquids (e.g.[BMIM][BF4] has a conductivity of 3.55 S ∙cm−1 at room temperature)[21]. All other NADES have conductivities of at least one order ofmagni-tude lower (see supplementary material) and the amplitude of the var-iation of σdc is relatively small within the studied temperature range.

However for the NADES ca:suc (1:1), ca:gluc (1:1) and ChCl:xyl(3:1) it is possible to simulate the temperature dependence of the directconductivity by the Vogel–Fulcher–Tammann (VFT) equation. The VFTequation usually describes the temperature dependence of the relaxa-tion time associated with the dynamic glass transition, but it wasfound to describe quite well σdc (1/T) for a variety of materials, in-cluding ionic liquids [18,22,23]. The VFT equation was fitted to σdc

(1/T), according to:

σdc ¼ σ∞∙ exp� B

T�T0

� �ð5Þ

where B is an empirical parameter accounting for the plot curvature,σ∞ is the high temperature limit of the conductivity, and T0 is the Vogeltemperature at which the conductivity goes to zero. When the VFTequation is used to simulate the glassy dynamics relaxation time, T0 isinterpreted as the glass transition temperature of an ideal glass, i.e., aglass obtainedwith an infinitely slow cooling rate [24]. Thefitting is rep-resented in Fig. 4 as solid lines, evidencing that the temperature depen-dence of the conductivity is adequately described by the VFT equation.The VFT fitting parameters are given in Table 4.

Fig. 5. ENR values obtained for NADES in this study. The [BMIM][BF4] ILwas included in thestudy for comparison.

Fig. 6. Effect of the amount of water on λmax of Nile red.

539R. Craveiro et al. / Journal of Molecular Liquids 215 (2016) 534–540

The fact that the VFT equation gives an adequate representation ofthe experimental data can be taken as an indication that the ionconducting motions are governed by the glassy dynamics, as found forrelated systems [18,22,23]. The comparison between the Vogel temper-ature and the glass transition temperature gives the same variation, i.e.,the highest T0 values are found for ca:suc (1:1) and ca:glu (1:1), whilethe lowest value is found for ChCl:xyl (3:1); the Tg values observe thesame order for NADES. This seems to corroborate the coupling betweenthe charge transport mechanism and the structural relaxation associat-ed with the dynamic glass transition.

3.4. Polarity

The polarity of the NADES under study was measured with thesolvatochromic dye Nile red, which has been used for weak acids andprotic molecular solvents [25] and ILs [26].

For each sample λmax was determined, and the related parameterENR was calculated, using the following equation [27]:

ENR ¼ 28;591=λmax ð6Þ

with ENR in kcal mol−1 and λmax in nm.Solvents with higher polarity shift λmax of the dye to higher

wavelength values, yielding lower ENR values, according to Eq. (6). Bymeasuring changes in λmax of Nile red relative to the value it exhibitsin a reference solvent, it is possible to calculate the relative polarity ofthe solvent of interest [28]. Fig. 5 gives the polarity of the NADES stud-ied. The polarity of the [BMIM][BF4] IL was included for the sake ofcomparison.

According to Eq. (6), our experimental value of ENR for [BMIM][BF4]yields a λmax of 556.0 nm, which differs by less than 1% from the valuereported in the literature [26]. Our data indicates that NADES composedof ChCl and organic acids are more polar than those combining ChClwith a sugar, in agreement with results previously reported [3]. The

Table 5Influence of the water content of ChCl:xyl (2:1) and ChCl:ca (1:1) on λmax and ENR of Nilered value.

NADES Water content(wt.%)

λmax

(nm)ENR(kcal mol−1)

ChCl:xyl (2:1)

2.2 563 50,783.8 564 50,696.0 570 50,16

7.74 [3] – 49.72 [3]

ChCl:ca (1:1)4.8 592 48.305.2 596 47.945.9 597 47.89

slightly higher polarity of gluc:ta (1:1) relative to gluc:ca (1:1) can bedue to its shorter alkyl chain length. Also, tartaric acid is a strongeracid (pKa = 2.98) than citric acid (pKa = 3.14). As expected, a higherproportion of organic acid brings about an increase in polarity.

Another factor that needs to be taken into account is the water con-tent of the sample. Dai and co-workers have shown the influence of theamount of water of the NADES on their polarity [3]. In this work, to ac-count for that effect, measurements with Nile red solutions with differ-ent amounts of water from 0.1% to 15% were performed. A bathocromicshift is visible, since increasing water content shifts λmax to higherwavelengths. The deviation of Nile red's λmax was under 1.2% and thechanges in ENR values are small (Fig. 6).

Lower ENR values imply higher polarity. The same is observed inTable 5 for the case of ChCl:xyl (2:1) and ChCl:ca (1:1), where the var-iation of water content results in small variations in the ENR value. Nev-ertheless, these small changes in the ENR values can alter the relativepolarity scale, meaning that the control of the water content in theseNADES is of extreme importance.

Dai and co-workers have reported an ENR value of 49.72 kcal mol−1

for theNADESwith 7.74wt.%water already referred. In Table 5, it can beobserved that increasing the water content increases the polarity of theNADES. A decrease in water content of about 4 wt.% results in an in-crease of ca. 0.9 kcal ∙mol−1 in the ENR value.

This trend was also observed and reported for 1,2-propanediol:ChCl:water (1:1:1) NADES [3]. The authors of thatstudy suggested that the dilution of the NADES (up to water contentof 50%) caused a dramatic change in the structure of the NADES, dueto the rupture of the hydrogen bonding network formed between thetwo components. A careful control of the water content is necessarywhen polarity measurements are being performed, and also its effectin the overall relative polarity scale can be dependent on the NADEScomposition.

4. Conclusions

In this work we present a study of the thermophysical properties ofdifferent NADES. The determination of properties such as density, vis-cosity, thermal behavior, polarity, and conductivity are essential for abetter understanding of these solvents and for unveiling potential appli-cations of what is now considered to be the new generation of ILs. Un-derstanding the mechanisms leading to NADES formation is essentialfor the production of tailor-made NADES. The knowledge of thermo-physical data is essential for further developments inmodeling andmo-lecular simulation, which will undoubtedly provide new cues on thepossible combinations of compounds to prepare NADES with specificproperties.

Acknowledgments

Rita Craveiro, Tânia Carvalho and Alexandre Paiva are grateful forthe financial support from Fundação para a Ciência e a Tecnologia(FCT/MEC) through the grants PTDC/EQUEPR/12191/2010/ENIGMA,SFRH/BD/47088/2008 and SFRH / BPD / 44946 / 2008. We further ac-knowledge the financial support of FCT/MEC through the projectENIGMA — PTDC/EQU-EPR/121491/2010, and the project PEst-C/EQB/LA0006/2013 and FCOMP-01-0124-FEDER-020646. We also ac-knowledge the funding from the European Union Seventh Frame-work Programme (FP7/2007–2013) under grant agreement no.REGPOT-CT2012-316331-POLARIS, and from the Project “Novelsmart and biomimetic materials for innovative regenerative medi-cine approaches (Ref.: RL1 — ABMR — NORTE-01-0124-FEDER-000016)” co-financed by the North Portugal Regional OperationalProgramme (ON.2 — O Novo Norte), under the National StrategicReference Framework (NSRF), through the European Regional De-velopment Fund (ERDF).

540 R. Craveiro et al. / Journal of Molecular Liquids 215 (2016) 534–540

Appendix A. Supplementary data

Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.molliq.2016.01.038.

References

[1] C. Chiappe, D. Pieraccini, Ionic liquids: solvent properties and organic reactivity, J.Phys. Org. Chem. 18 (2005) 275–297, http://dx.doi.org/10.1002/poc.863.

[2] M. Zdanowicz, T. Spychaj, Ionic liquids as starch plasticizers or solvents, Polimery 56(2011) 861–864.

[3] Y. Dai, J. van Spronsen, G.-J. Witkamp, R. Verpoorte, Y.H. Choi, Natural deep eutecticsolvents as new potential media for green technology, Anal. Chim. Acta 766 (2013)61–68, http://dx.doi.org/10.1016/j.aca.2012.12.019.

[4] A. Paiva, R. Craveiro, I. Aroso, M. Martins, R.L. Reis, A.R.C. Duarte, et al., Natural deepeutectic solvents — solvents for the 21st century, Sustain. Chem. Eng. (2014)http://dx.doi.org/10.1021/sc500096j (ASAP).

[5] K. Rado, M. Cvjetko, D. Grgas, T. Landeka, I. Radoj, Ecotoxicology and environmentalsafety evaluation of toxicity and biodegradability of choline chloride based deep eu-tectic solvents, 112 (2015) 46–53, http://dx.doi.org/10.1016/j.ecoenv.2014.09.034.

[6] H. Zhao, G.a. Baker, S. Holmes, New eutectic ionic liquids for lipase activation andenzymatic preparation of biodiesel, Org. Biomol. Chem. 9 (2011) 1908–1916,http://dx.doi.org/10.1039/c0ob01011a.

[7] D. Carriazo, M.C. Serrano, M.C. Gutiérrez, M.L. Ferrer, F. del Monte, Deep-eutectic sol-vents playing multiple roles in the synthesis of polymers and related materials,Chem. Soc. Rev. 41 (2012) 4996–5014, http://dx.doi.org/10.1039/c2cs15353j.

[8] P.S. Kulkarni, L.C. Branco, J.G. Crespo, M.C. Nunes, A. Raymundo, C.a.M. Afonso, Com-parison of physicochemical properties of new ionic liquids based on imidazolium,quaternary ammonium, and guanidinium cations, Chemistry 13 (2007)8478–8488, http://dx.doi.org/10.1002/chem.200700965.

[9] S. Gore, S. Baskaran, B. Koenig, Efficient synthesis of 3,4-dihydropyrimidin-2-ones inlowmelting tartaric acid–ureamixtures, Green Chem. 13 (2011) 1009–1013, http://dx.doi.org/10.1039/c1gc00009h.

[10] S. Tuntarawongsa, T. Phaechamud, Polymeric eutectic drug delivery system, J. Met.Mater. Miner. 22 (2012) 27–32.

[11] C. Florindo, F.S. Oliveira, L.P.N. Rebelo, A.M. Fernandes, I.M. Marrucho, Insights intothe synthesis and properties of deep eutectic solvents based on cholinium chlorideand carboxylic acids, ACS Sustain. Chem. Eng. 2 (2014) 2416–2425, http://dx.doi.org/10.1021/sc500439w.

[12] D. Rengstl, V. Fischer,W. Kunz, Low-meltingmixtures based on choline ionic liquids,Phys. Chem. Chem. Phys. 16 (2014) 22815–22822, http://dx.doi.org/10.1039/C4CP02860K.

[13] M., R.B. Prime, Thermal Analysis of Polymers, Fundamentals and Applications, Hobo-ken, New Jersey, 2009.

[14] B. Carvalho, T.,. Augusto, V.,. Brás, A.R.,. Lourenço, N.M.T.,. Afonso, C.A.M.,. Barreiros,S.,. Correia, N.T.,. Vidinha, P.,. Cabrita, E.J.,. Dias, C.J.,. Dionísio, M.,. Roling, T.Carvalho, V. Augusto, a.R. Brás, N.M.T. Lourenço, C.a.M. Afonso, et al., Understandingthe ion jelly conductivity mechanism, J. Phys. Chem. B 116 (2012) 2664–2676,http://dx.doi.org/10.1021/jp2108768.

[15] R. Ramani, G. Shariff, Influence of gamma irradiation on the formation of methanolinduced micro-cracks in polycarbonate, J. Mater.Sci. 8 (2003) 1431–1438.

[16] V. Legrand, M. Descamps, C. Alba-Simionesco, Glass-forming meta-toluidine: a ther-mal and structural analysis of its crystalline polymorphism and devitrification,Thermochim. Acta 307 (1997) 77–83, http://dx.doi.org/10.1016/S0040-6031(97)00271-2.

[17] E. Dudognon, F. Danède, M. Descamps, N.T. Correia, Evidence for a new crystallinephase of racemic ibuprofen, Pharm. Res. 25 (2008) 2853–2858, http://dx.doi.org/10.1007/s11095-008-9655-7.

[18] J.J. Diogo, H.P.;.M. Ramos, Are the crystallization and the melting the reverse of eachother? J. Chem. Educ. 83 (2006) 1389–1392.

[19] A. Kremer, F. Schönals, Broadband dielectric spectroscopy, in: A. Kremer, F. Schönals(Eds.), Broadband Dielectr, Spectrosc., Springer Verlag, Berlin, 2003.

[20] C.J. Bowlas, D.W. Bruceb, K.R. Seddon, Liquid-crystalline ionic liquids, Chem.Commun. (Camb.) (1996) 1625–1626.

[21] X. Fu, Polymer Electrolytes, Elsevier, 2010http://dx.doi.org/10.1533/9781845699772.2.471.

[22] J.R. Sangoro, C. Iacob, A. Serghei, C. Friedrich, F. Kremer, Universal scaling of chargetransport in glass-forming ionic liquids, Phys. Chem. Chem. Phys. 11 (2009)913–916, http://dx.doi.org/10.1039/b816106b.

[23] J. Leys, M. Wübbenhorst, C. Preethy Menon, R. Rajesh, J. Thoen, C. Glorieux, et al.,Temperature dependence of the electrical conductivity of imidazolium ionic liquids,J. Chem. Phys. 128 (2008) 064509, http://dx.doi.org/10.1063/1.2827462.

[24] J.F. Mano, M. Dionísio, Handbook of Thermal Analysis and Calorimetry, Elsevier,2008.

[25] W. Ogihara, T. Aoyama, H. Ohno, Polarity measurement for ionic liquids containingdissociable protons, Chem. Lett. 33 (2004) 1414–1415.

[26] A.J. Carmichael, K.R. Seddon, Polarity study of some 1-alkyl-3-methylimidazoliumambient-temperature ionic liquids with the solvatochromic dye, Nile Red, J. Phys.Org. Chem. 13 (2000) 591–595, http://dx.doi.org/10.1002/1099-1395(200010)13:10b591::AID-POC305N3.0.CO;2-2.

[27] P.G. Jessop, D.a. Jessop, D. Fu, L. Phan, Solvatochromic parameters for solvents of in-terest in green chemistry, Green Chem. 14 (2012) 1245, http://dx.doi.org/10.1039/c2gc16670d.

[28] Y. Chen, Y. Cao, X. Lu, C. Zhao, C. Yan, T. Mu, Water sorption in protic ionic liquids:correlation between hygroscopicity and polarity, New J. Chem. 37 (2013) 1959,http://dx.doi.org/10.1039/c3nj00213f.