Thermophysical properties of ammonium and hydroxylammonium protic ionic liquids

Thermophysical properties of ammonium and hydroxylammonium protic ionic liquids

J. Chem. Thermodynamics 72 (2014) 117–124 Contents lists available at ScienceDirect J. Chem. Thermodynamics journal homepage: www.elsevier.com/locat...

889KB Sizes 11 Downloads 200 Views

J. Chem. Thermodynamics 72 (2014) 117–124

Contents lists available at ScienceDirect

J. Chem. Thermodynamics journal homepage: www.elsevier.com/locate/jct

Thermophysical properties of ammonium and hydroxylammonium protic ionic liquids Pratap K. Chhotaray, Ramesh L. Gardas ⇑ Department of Chemistry, Indian Institute of Technology, Madras, Chennai 600 036, India

a r t i c l e

i n f o

Article history: Received 22 June 2013 Received in revised form 12 December 2013 Accepted 6 January 2014 Available online 11 January 2014 Keywords: Ionic liquid Density Viscosity Coefficient of thermal expansion Isentropic compressibility Glass transition temperature

a b s t r a c t In this work, five protic ionic liquids having propylammonium, 3-hydroxy propylammonium as cations and formate, acetate, trifluoroacetate as anions have been synthesized. Thermophysical properties such as density (q), viscosity (g) and sound velocity (u) have been measured at various temperatures ranging from (293.15 to 343.15) K at atmospheric pressure. The experimental density and viscosity were fitted with second order polynomial and Vogel–Tamman–Fulcher (VTF) equations, respectively. Also experimental densities were correlated with the estimated density proposed by Gardas and Coutinho model. The coefficient of thermal expansion (a) and isentropic compressibility (bs) values have been calculated from the experimental density and sound velocity data using empirical correlations. Lattice potential energy (UPOT) has been calculated to understand the strength of ionic interaction between the ions. Thermal decomposition temperature (Td) and glass transition temperature (Tg) along with crystallization and melting point were investigated using TGA and DSC analysis, respectively. The effect of alkyl chain length and electronegative fluorine atoms on anionic fragment as well as hydroxyl substituent on cationic side chain in the protic ionic liquids has been discussed for studied properties. The effect of DpKa over the studied properties has also been analyzed. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction Ionic liquids (ILs) are organic salts generally composed of ions and having melting points below 100 °C [1]. It draws extensive attention during last few decades due to their fascinating properties like low volatility, high solvation capacity, large electro-chemical window, high thermal and electrical conductivity both in academia and industry. Protic ionic liquids (PILs) are subgroup of ILs formed by proton transfer from a Brønsted acid to a Brønsted base [2]. The presence of an available proton which is responsible for hydrogen bonding makes PILs different from other ILs. Till date, PILs have not received much more attention as compared to their counterpart aprotic ionic liquids [3]. Nevertheless, these ILs have many beneficial properties and potential applications, basically due to their protic nature, such as self-assembly media [4–9], catalysts in chemical reactions [10–12], biological applications [13], and proton conducting electrolytes for polymer membrane fuel cells [14]. Hydroxylammonium ionic liquid can be used to dissolve zein, an industrially important natural polymer [15]. Many insoluble polymers such as polyaniline and polypyrrole are found to be

⇑ Corresponding author. Tel.: +91 44 2257 4248; fax: +91 44 2257 4202. E-mail address: [email protected] (R.L. Gardas). URL: http://www.iitm.ac.in/info/fac/gardas (R.L. Gardas). 0021-9614/$ - see front matter Ó 2014 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jct.2014.01.004

highly soluble in this class of ILs [16]. The effect of the hydroxyl (OH) group in this type of ionic liquid for the solvation with polar solvents has been revealed by the determination of solvato -chromic parameter [17]. Currently, aqueous monoethanolamine has been used for CO2 removal from natural gas in industrial processes [18]. Due to its serious environmental concerns related to volatility, recovery and corrosiveness [19–23], ILs seem to be better alternative solvents which can surpass the above difficulties along with high efficiency for CO2 absorption [24]. Ionic liquids having imidazolium and pyridinium cations are the most studied classes of ionic liquids for gas separation [25–31]. Recently, ammonium and hydroxylammonium ionic liquids have also drawn considerable attention towards CO2 absorption both from natural and flue gas [12,32,33]. Knowledge of thermophysical properties such as density, viscosity and thermal stability are important to determine the possible application of ionic liquids for gas sweetening. In spite of their importance and easy preparation technique there are very few literature available on the thermophysical properties of pure hydroxylammonium and ammonium ionic liquids [34–37]. There may exist a possible equilibrium due to incomplete proton transfer during neutralization reaction, resulting in the formation of neutral ion-pairs. MacFarlance et al. suggested that DpKa (the difference in pKa [38,39] value for the acid and base determined in dilute aqueous solutions) > 4 is sufficient for complete

118

P.K. Chhotaray, R.L. Gardas / J. Chem. Thermodynamics 72 (2014) 117–124

List of Symbols

q g u

a bs

Density Viscosity Sound velocity Coefficient of thermal expansion Isentropic compressibility

Td Tg Vm Lf K

proton transfer [40]. However the nature of ionic liquids which depends upon the difference in pKa values between the acid and the base is still a matter of debate in the scientific community. Specifically Angell et al. demonstrated that DpKa > 10 is required for complete proton transfer [41]. In this work, we have synthesized, purified and characterized five ammonium and hydroxylammonium based PILs. The density, viscosity and sound velocity have been measured at atmospheric pressure with temperature variation from (293.15 to 343.15) K. The sound velocity and hence the isentropic compressibility can be regarded as thermodynamic properties, as the ultrasonic absorption is negligible due to the use of low frequency (3 MHz) and low amplitude of the acoustic waves [42,43]. The experimental density and viscosity have been correlated with second order polynomial and Vogel–Tamman–Fulcher (VTF) equations, respectively. The coefficient of thermal expansion and isentropic compressibility have been calculated using the density and sound velocity data. Thermal decomposition temperature (Td) and glass transition temperature (Tg) of studied ionic liquids have been analyzed by TGA and DSC, respectively.

2. Experimental section 2.1. Chemicals Acetic acid (P99%), formic acid (P95%), trifluoroacetic acid (99%), 3-amino-1-propanol (99%) and propyl amine (P99%) were obtained from Sigma Aldrich and were used without further purification.

2.2. Synthesis of ionic liquids All the five ILs were synthesized by exothermic neutralization of equimolar bases with different acids [16]. The bases were taken in a two necked round bottom flask equipped with reflux condenser and dropping funnel. Acids were then added drop wise under vigorous stirring to the round bottom flask kept in ice bath, maintaining the reaction mixture temperature below 10 °C. After complete addition, the temperature was raised to room temperature with further stirring up to 24 h. The resulting viscous liquid obtained was then connected to high vacuum for 48 h at room temperature with continuous stirring, to remove the water content and residual amine, taken in excess to ensure complete consump-

Thermal decomposition temperature Glass transition temperature Molecular volume Inter molecular free length Jacobson’s constant

tion of acid. The ionic liquids were then stored under N2 atmosphere. Their structure and abbreviations are shown in table 1. 2.3. Characterizations Proton, carbon NMR was recorded (see figure in the Supporting Information) on Brukar Avance 500 MHz spectrometer using deuterated DMSO as solvent. For 3-HPAF d = 8.44 ppm (s, 1H, HCOO), d = 5.96 ppm (broad, 4H, OH and NHþ 3 ), d = 3.46 ppm (t, 2H, CH2AN), d = 2.79 ppm (t, 2H, CH2AO), d = 1.66 ppm (qn, 2H, CH2AC). For 3-HPAAc d = 5.29 ppm (broad, 4H, OH and NHþ 3 ), d = 3.46 ppm (t, 2H, CH2AN), d = 2.73 ppm (t, 2H, CH2AO), d = 1.71 ppm (s, 3H, CH3AC), d = 1.61 ppm (qn, 2H, CH2AC). For 3-HPATFAc d = 5.28 ppm (broad, 4H, OH and NHþ 3 ), d = 3.46 ppm (t, 2H, CH2AN), d = 2.77 ppm (t, 2H, CH2AO), d = 1.64 ppm (qn, 2H, CH2AC). For PAF d = 8.44 ppm (s, 1H, HCOO), d = 3.56 ppm (broad, 3H, NHþ 3 ), d = 2.68 ppm (t, 2H, CH2AN), d = 1.53 (sx, 2H, CH2AC), d = 0.88 ppm (t, 3H, CH3AC). For PAAc d = 5.12 ppm (broad, 3H, NHþ 3 ), d = 2.60 ppm (t, 2H, CH2AN), d = 1.72 ppm (s, 3H, CH3ACO) d = 1.47 ppm (sx, 2H, CH2AC), d = 0.87 ppm (t, 3H, CH3AC). IR spectra were recorded by JASCO FT/IR-4100 spectrometer using NaCl disk. The device has a maximum resolution of 0.9 cm1 and have 22,000:1 signal to noise ratio. For all PILs the broad band appeared in (3600 to 2600) cm1 range exhibits the characteristic ammonium peak and OAH stretching vibration. A combined broad band of the N–H plane bending vibrations and carbonyl stretching is observed around 1600 cm1. 2.4. Apparatus and procedure Density and sound velocity of the studied ionic liquids were measured using Anton Paar (DSA 5000 M) vibrating tube digital density and sound velocity meter. It uses an inbuilt oscillating glass U tube for density measurement and stainless steel cell for sound velocity measurement. To ensure accurate and reproducible results as well as highly convenient sample handling, the instrument has a number of unique features like: (i) Density (in the range from (0 to 3) g  cm3) and sound velocity (in the range from (1000 to 2000) m  s1) can be measured simultaneously in the temperature range from (273.15 to 343.15) K with pressure variation from (0 to 3) bar. (ii) Sample filling errors are detected automatically. (iii) A PT-100 sensor is used to measure the temperature with an accuracy of ±0.01 K. (iv) It provides automatic viscosity correction across the sample’s entire viscosity range and (v) It has the facility

TABLE 1 Ionic liquids names and abbreviations used in this work. Cation

Anion

Name

Abbreviation

HCOO CH3COO

Propylammoniumformate Propylammonium acetate

PAF PAAc

HCOO CH3COO CF3COO

3-Hydroxypropylammonium formate 3-Hydroxypropylammonium acetate 3-Hydroxypropylammonium trifluoroacetate

3-HPAF 3-HPAAc 3-PATFAc

119

P.K. Chhotaray, R.L. Gardas / J. Chem. Thermodynamics 72 (2014) 117–124

for air adjustments (the local air pressure is correctly accounted for measurements). The measurements were taken with slow equilibration mode for better accuracy with successive increments of 5 K within the temperature range (293.15 to 343.15) K. The instrument was calibrated with Millipore quality water and dry air in frequent intervals of time. As we are dealing with high density and high viscosity samples, the instrument was also calibrated with reference ionic liquid namely, 1-hexyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide, [C6Mim][Tf2N] [44,45]. Dynamic viscosities of ILs were measured by Anton Paar microviscometer (Lovis 2000 ME) mounted on the master instrument DSA 5000 M. It can measure viscosity from 0.3 mPa  s to 10,000 mPa  s by rolling ball technique. The temperature range of this study was from (293.15 to 343.15) K, at 5 K intervals with an accuracy of 0.02 K which is controlled by a built-in precise Peltier thermostat. Before measurement, calibration was carried out using standard oils (S3, N26, N100 liquid for 1.59 mm, 1.8 mm and 2.5 mm capillary, respectively) provided by Anton Paar Co., Austria. The standard uncertainties associated with the measurements were estimated to be less than 0.002 K for temperature, 0.007 kg  m3 for density, 0.005 mPa  s for viscosity and 0.05 m  s1 for sound velocity. The water content was measured by Analab (Micro Aqua Cal 100) Karl Fischer Titrator and Karl Fischer reagent from Merck. Calibration was carried out with distilled water (by weight) with an average titer factor of 5.356. It can detect water content from less than 10 ppm to 100% by conductometric titration with dual platinum electrode. The thermal decomposition data were obtained using TGA instrument (TA instruments Hi-Res. TGA Q500) with weighing precision ±0.01% from room temperature to 300 °C at a heating rate of 10 °C  min1 under nitrogen atmosphere in a open platinum pan. Its established performance arises from a responsive low-mass furnace, ultra-sensitive thermobalance, and efficient horizontal purge gas system (with mass flow control). To ensure simultoneous measurement of sample temperature and heating rate control, accurately and precisely, two thermocouples are positioned immediately adjacent to the sample. Curie point temperature calibration was carried out with Nickel metal. The phase transition measurements were carried out by DSC instrument (TA instruments DSC Q200) having sensitivity 0.2 lW and temperature accuracy ±0.1 °C. The Tzero hermatic aluminum pan is used for better resolution. Multiple cooling devices can be added to the instrument, apart from that it has inbuilt superior cooling system where an array of highly conductive Nickel rods are connected to the Silver furnace. Heat flow (ASTM E968) and baseline calibration was carried out with Indium metal and Sapphire respectively. Scanning sequence involves, freezing the IL sample to 80 °C, maintaining this temperature for 20 min., and then heating the sample to 50 °C with a heating rate of 2 °C  min1 under nitrogen atmosphere with a flow rate of 50 cm3  min1.

3. Results and discussion Complete proton transfer has occurred in all of the synthesized ionic liquid as there is no residual peak of starting material in 1H NMR and also DpKa is greater than 4 as suggested by MacFarlance et al. [40] as a pre requisite condition for complete proton transfer. All the ionic liquids were dried before each measurement and the water content was less than 8000 ppm. The detailed water content, molecular weight and DpKa values are illustrated in table 2. The experimental density, viscosity and sound velocity presented in table 3 are the average of three measurements. As it can be seen from figure 1, the density of all ionic liquids decreases linearly over the studied range of temperatures from (293.15 to 343.15) K covering wide range of density from 1311 kg  m3 in case of 3-HPATFAc at 293.15 K to 959 kg  m3 for PAAc at 343.15 K. Ionic liquids having the same anion, hydroxylammonium ionic liquids have higher density as compared to their ammonium counterpart as can be seen from table 3, which may be mainly due to the additional hydrogen bonding. For the same cation, the ionic liquid having higher DpKa shows higher density [46], for instance with 3-hydroxypropylammonium, the ILs with HCOO as the conjugate anion (DpKa = 6.21) have higher density than those with CH3COO as the conjugate anion (DpKa = 5.2). Another cause for higher density in case of HCOO based ionic liquid can be attributed to the less steric hindrance as compared to the CH3COO based ionic liquids. At T = 303.15 K the densities of studied acetate based ILs (PAAc: 983.85 kg  m3; 3-HPAAc: 1109.46 kg  m3) cover a wide range as compared to acetate based ILs reported in literature ([C2im][Ac]: 1029.9 kg  m3; [C4mim][Ac]: 1049.2 kg  m3; [C4mpyr][Ac]: 1018.1 kg  m3; and [N0122][Ac]: 1012.2 kg  m3) [47] which indicates a significant effect of cationic structure upon density. Figure 1 illustrates that among all the studied ionic liquids, 3-HPATFAc is denser as compared to other ionic liquids. This can be attributed to its higher molar mass, and high electronegativity of fluorine atoms which may reduce the overall volume of molecules. For all ionic liquids, as expected, alkyl chain length on anion has a decreasing effect over the density due to steric hindrance. The experimental density (q) values were fitted by the method of least square fitting using the second order polynomial equation given by:

q ¼ A0 þ A1 T þ A2 T 2 ;

ð1Þ

where A0, A1 and A2 are correlation coefficients. The coefficient values are estimated and presented in table 4 along with average absolute relative deviation (ARD) defined by the following equation:

ARD ¼

! 1 X jqexp  qcal j 100; n qexp

ð2Þ

where n is the number of data points.

TABLE 2 Molecular weight (M), water content, decomposition temperature (Td), glass transition temperature (Tg), molecular volume (Vm), lattice potential energy (UPOT) and DpKa of ionic liquids.

a b

IL

M/g  mol1‘

Water content/ppm

Td/°C

Tg/°C

Vm/nm3a

UPOT/kJ  mol1a

DpKa

PAF PAAc 3-HPAF

105.14 119.16 121.14

103 113 140

 26 24

0.1757 0.2005 0.1755

523 505 523

6.85 5.84 6.21

3-HPAAc

135.16

136

26

0.2018

504

5.2

3-HPATFAc

189.13

6753 5436 7124 6600b 7538 7200b 7482

184

28

0.2401

481

9.73

T = 298 K. Values are from Pinkert et al. [35].

120

P.K. Chhotaray, R.L. Gardas / J. Chem. Thermodynamics 72 (2014) 117–124

TABLE 3 Experimental density (q), viscosity (g), sound velocity (u), coefficient of thermal expansion (a), isentropic compressibility (bs) of pure PAF, PAAc, 3-HPAF, 3-HPAAc, and 3-HPATFAc from T = (293.15 to 343.15) K at pressure p = 0.1 MPa. T/K

q/kg  m3

g/mPa  s

u/m  s1

a/104  K1

bs/TPa1

293.15 298.15 303.15 308.15 313.15 318.15 323.15 328.15 333.15 338.15 343.15

996.18 993.47 990.77 988.09 985.42 982.75 980.10 977.45 974.80 972.15 969.49

96.77 78.60 63.58 52.67 43.97 37.16 31.43 26.92 23.32 20.18 17.57

PAF 1553.06 1542.72 1532.43 1522.21 1511.96 1501.79 1491.69 1481.63 1471.65 1461.78 1451.97

5.35 5.37 5.38 5.40 5.41 5.42 5.44 5.45 5.47 5.48 5.50

416.19 422.93 429.80 436.77 443.91 451.17 458.54 466.04 473.67 481.40 489.26

293.15 298.15 303.15 308.15 313.15 318.15 323.15 328.15 333.15 338.15 343.15

989.97 986.91 983.85 980.77 977.69 974.60 971.50 968.40 965.30 962.17 959.03

932.22 627.41 435.42 309.95 226.35 167.69 128.15 99.03 78.03 61.96 50.02

PAAc 1517.12 1507.96 1497.08 1487.62 1477.02 1467.75 1457.45 1447.92 1437.29 1427.42 1417.47

6.25 6.27 6.29 6.31 6.33 6.35 6.37 6.39 6.41 6.43 6.45

438.87 445.60 453.50 460.73 468.84 476.29 484.58 492.55 501.48 510.09 518.97

4.19 4.20 4.20 4.21 4.22 4.23 4.24 4.25 4.26 4.27 4.28

261.68 264.53 267.47 270.38 273.42 276.17 279.52 282.68 285.81 289.00 292.21

293.15 298.15 303.15 308.15 313.15 318.15 323.15 328.15 333.15 338.15 343.15

1148.29 1145.84 1143.39 1140.95 1138.53 1136.12 1133.73 1131.34 1128.97 1126.61 1124.25

3-HPAF 339.04 1824.27 257.15 1816.35 198.39 1808.29 154.49 1800.45 122.56 1792.32 98.17 1785.25 79.84 1776.40 65.37 1768.31 53.98 1760.43 45.36 1752.54 38.38 1744.69

293.15 298.15 303.15 308.15 313.15 318.15 323.15 328.15 333.15 338.15 343.15

1114.58 1112.02 1109.46 1106.91 1104.36 1101.80 1099.24 1096.68 1094.13 1091.59 1089.03

3-HPAAc 4261.70 1779.78 2827.01 1770.84 1937.03 1761.87 1348.55 1753.29 970.90 1745.22 714.44 1736.37 526.00 1727.81 400.11 1719.35 305.97 1710.34 238.21 1702.84 190.40 1693.60

4.58 4.60 4.61 4.62 4.63 4.64 4.65 4.66 4.67 4.68 4.69

283.24 286.77 290.36 293.89 297.30 301.03 304.73 308.45 312.44 315.93 320.14

293.15 298.15 303.15 308.15 313.15 318.15 323.15 328.15 333.15 338.15 343.15

1311.44 1307.81 1304.20 1300.58 1296.96 1293.34 1289.72 1286.10 1282.49 1278.88 1275.25

3-HPATFAc 1430.20 1487.90 970.33 1476.51 677.32 1465.53 483.03 1454.92 352.31 1444.53 262.07 1434.31 198.79 1424.15 153.05 1413.92 119.86 1403.56 94.92 1393.20 76.72 1382.79

5.52 5.53 5.55 5.56 5.58 5.60 5.61 5.63 5.64 5.66 5.68

344.43 350.74 357.00 363.23 369.50 375.84 382.29 388.93 395.81 402.85 410.10

FIGURE 1. Density of ionic liquids as a function of temperature from (293.15 to 343.15) K. PAF; d PAAc;  3-HPAF; j 3-HPAAc; N 3-HPATFAc. The symbols represent experimental values and the solid lines represent the values calculated from equation (1).

TABLE 4 Prediction coefficients for density and ARD of equation (1) and MAD of equation (4).

a b

PILs

A0  103

A1  101

A2  104

ARDa  104

MADb

ARDb

PAF PAAc 3-HPAF 3-HPAAc 3-HPATFAc

1.1656 1.1557 1.3132 1.2653 1.5248

6.1613 5.1969 6.3258 5.1682 7.3132

1.3049 1.5538 2.3865 0.0909 0.1189

0.808 0.433 0.319 0.423 0.280

0.37 0.16 0.91 0.72 0.29

0.17 0.09 0.40 0.33 0.14

Using equation (1). Using equation (4).

where N is Avogadro’s number, M is molecular weight and q is density (at 298.15 K) of ionic liquid. From table 2, it can be seen that 3-HPAAc molecule occupy 0.0263 nm3 more volume as compare to 3-HPAF molecule due to the presence of an additional methylene

Standard uncertainties u are u(T) = 0.002 K, u(q) = 5  103  kg  m3, u(u) = 0.05 m  s1, u(g) = 0.005 mPa  s and the combined standard uncertainties uc are uc(bs) = 0.05 TPa1, and uc(a) = 0.002 K1.

From the experimental density, molecular volumes (Vm) are calculated using the equation (3)

M Vm ¼ ; Nq

ð3Þ

FIGURE 2. Correlation between experimental and calculated density of studied ionic liquids at temperatures (293.15 to 343.15) K, using Gardas and Coutinho equation. PAF; s PAAc; e 3-HPAF; h 3-HPAAc; 4 3-HPATFAc. The solid line represent the x = y trend line.

121

P.K. Chhotaray, R.L. Gardas / J. Chem. Thermodynamics 72 (2014) 117–124

group. Similarly, PAAc molecule possesses 0.0248 nm3 more volume than PAF molecule. The values indicates that the mean contribution due to methylene (ACH2) groups are in a close agreement with the values for alcohols (0.0280 nm3), n-amines (0.0272 nm3), n-paraffins (0.0267 nm3) [48] and for ionic liquids as reported in literature [49,50]. Density of studied ionic liquids were predicted using the Gardas and Coutinho model [51] proposed for the estimation of density of ionic liquid over a wide range of temperature from (273.15 to 393.15) K and (0.10 to 100) MPa pressure, is given by the equation (4)

qcal ¼

M ; NV m ða þ bT þ cPÞ

ð4Þ

where qcal is the density in kg  m3, M is the molar mass in kg  mol1, N is the Avogadro’s number, Vm is the molecular volume in nm3, T is the temperature in K, and P is the pressure in MPa. The model parameters a, b, and c were fitted to experimental data obtained from literature [52,53] and equation (4) used obtain the values as 0.8005 ± 0.00023, (6.6520 ± 0.0069)  104  K1, and (5.919 ± 0.024)  106  MPa1, respectively, at 95% confidence level [51]. The molecular volumes are obtained by using equation (3). The maximum average relative deviation (MAD) and ARD are provided in table 4. The calculated density (qcal) shows good agreement with the experimental density (qexp) as can be seen from figure 2, and were related via qcal = (0.9986 ± 0.0008) qexp with a confidence level of 95% and R2 = 0.9993. From the above obtained results it may be concluded that the equation proposed by Gardas and Coutinho [51] can also be applied for new families of ionic liquids [54], other than those used in the development of correlation, with a good level of confidence. Ultrasonic velocity and related thermodynamic parameter helps us in characterizing thermodynamic and physico-chemical aspects of inter molecular interaction. Figure 3 shows that the ultrasonic velocity decreases linearly with respect to temperature over the studied temperature range. As shown in figure 3 and illustrated in table 3, the increase in carbon chain length at anionic side in both ammonium and hydroxylammonium ionic liquids has a decreasing effect upon sound velocity. The cause can be attributed to the increase in spatial distance between molecules due to steric hindrance.

FIGURE 3. Ultrasonic sound velocity of ionic liquids as a function of temperature from (293.15 to 343.15) K. PAF; d PAAc;  3-HPAF; j 3-HPAAc; N 3-HPATFAc. The symbols represent experimental values and dashed line in the figure is a guide for eyes.

TABLE 5 Inter molecular free length (Lf/Å) at different temperature. T/K

PAF

PAAc

3-HPAF

3-HPAAc

3-HPATFAc

293.15 298.15 303.15 313.15 323.15

0.399 0.406 0.414 0.428 0.442

0.409 0.417 0.425 0.440 0.454

0.316 0.321 0.326 0.336 0.345

0.329 0.335 0.340 0.350 0.360

0.363 0.370 0.377 0.390 0.403

Some frequently used derived values for industrial operations are the temperature and pressure dependence of volume which can be expressed as thermal expansion coefficient (a) and isentropic compressibility (bs), respectively. Equations (5) and (6) used the experimental density and sound velocity values to determine these thermodynamic properties.

a¼

bs ¼

  1 @q ; q @T P 1

qu2

;

ð5Þ

ð6Þ

where a, bs are coefficient of thermal expansion and isentropic compressibility, respectively, u is the sound velocity, q is the density, T is the temperature and p is the pressure. As can be seen from table 3, both coefficient of thermal expansion (a) and isentropic compressibility (bs) show similar behavior with little increment against temperature for all the ionic liquids. PAAc shows highest value of isentropic compressibility among all the ionic liquids and is firmly supported by the free length values illustrated in table 5. Among the hydroxylammonium ionic liquids, 3-HPATFAc shows the highest availability of free space between the molecules which reflects in the higher value of isentropic compressibility. This feature arises probably due to less intermolecular interaction with the neighboring ionic liquid molecules. Moreover, this fact is supported by the free length data presented in table 5. As obvious, alkyl chain length has an increasing effect over both the derived properties, coefficient of thermal expansion (a) and isentropic compressibility (bs). Although the difference in density between 3-HPATFAc and PAF is reasonably high, rather they show

FIGURE 4. Viscosity of ionic liquids as a function of temperature from (293.15 to 343.15) K. PAF; d PAAc;  3-HPAF; j 3-HPAAc; N 3-HPATFAc. The symbols represent experimental values and the solid curves represent the values calculated from equation (7).

122

P.K. Chhotaray, R.L. Gardas / J. Chem. Thermodynamics 72 (2014) 117–124

similar behavior with temperature as can be seen from their close proximity of thermal coefficient values. To understand the intermolecular interaction in liquid, Jacobson suggested an empirical relation for calculating inter molecular free length (Lf) [55]. According to which Lf is given by

Lf ¼ K

pffiffiffiffiffi bs ;

ð7Þ

where K is called as Jacobson’s constant, which depends on temperature. The values of intermolecular free length (Lf) for all ionic liquids at different temperatures are presented in table 5. Table 5 clearly shows that Lf values increase with increasing temperature while the pressure is nearly constant. According to Eyring’s liquid state theory, the acoustic wave which was excited in the fluid is transmitted momentarily to the intermolecular length. In general, when the value of free length Lf in liquid is large, the ultrasonic velocity will have a lower value. The above fact can be validated from tables 3 and 5 where ultrasonic sound velocity decreases with the increase in intermolecular free length for all of the ionic liquids with only exception of 3-HPATFAc. Rise in temperature results in an increase of inter molecular distance, thereby increasing the distance between surfaces of the two molecules and ultimately decreasing the sound velocity as depicted in figure 3. Viscosity provides an important macroscopic property of ionic liquids, which is the result of numerous microscopic interactions such as columbic, van der Waals, hydrogen bonding and is quite dependent upon the shape and size of the constituent ions. The experimental viscosities of the ionic liquids from (293.15 to 343.15) K are shown in figure 4 and summarized in table 3. Customarily, for the studied ionic liquids, the viscosity increases rather rapidly at low temperature and decreases asymptotically to a smaller value as the temperature increases. The viscosity varies from as high 4262 mPa  s for 3-HPAAc to as low 18 mPa  s for PAF over the studied temperature range. As can be seen from table 3 at 303.15 K the viscosities of 3-hydroxypropylammonium acetate, 3-hydroxypropylammonium formate are (1937.03 and 198.39) mPa  s, respectively and as reported in literature, the viscosity of 2-hydroxyethylammonium acetate is found to be 224.68 mPa  s [37] at the same temperature. The above results indicate that the viscosity increases with increasing alkyl chain length on either of the ion. Longer alkyl chain lengths which not only leads to heavier and bulkier entity but also give rise to increasing the van der Waals attractions among the aliphatic alkyl chains, which causes the higher viscosity, in case of longer carbon chain length ionic liquids [56]. According to Seddon et al. ionic liquids generally show a non-Arrhenius behavior, hence the variation of the viscosity as a function of temperature is well described Vogel–Tamman–Fulcher (VTF) equation [57]:

g ¼ g0 exp

 B ; T  T0

FIGURE 5. Comparison of TGA behaviors for ammonium and hydroxylammonium PILs. (a) PAF; (b) PAAc; (c) 3-HPAAc; (d) 3-HPAF; (e) 3-HPATFAc.

ionic liquids along with other inter-ionic interactions between the protonated cation and the anion. To predict the relative stabilities of ionic liquids, Glasser proposed a method for estimating lattice potential energies (UPOT) [48]. The procedure rely only on the chemical formula, ionic charges and density (or molecular volume) of the materials involved. It neither requires nor depends upon any other structural information; therefore the equations apply equally well to amorphous solids and to ionic liquids as to their well tested application for ionic solids, since these procedure rely on columbic interactions being the principal contributors to the lattice energy. Glasser’s empirical equation is given by:

 q 1=3 U POT ¼ c þ d; M

ð9Þ

where q is the density, M is the molar mass, c and d are constants with values 1981.2 kJ  mol1  cm and 103.8 kJ  mol1, respectively. The lattice energy of all the ionic liquids were calculated



ð8Þ

where g0 (mPa  s), B (K) and T0 (K) are adjustable parameters, and the values are given in table 6 along with average absolute relative deviation (ARD). Hydroxylammonium ionic liquids show higher viscosity than their counter part ammonium ionic liquids. This is probably because of the additional hydrogen bonding possible for these

TABLE 6 Adjustable parameters of equation (7) for viscosity and ARD. PILs

g0/mPa  s

B/K

T0/K

ARD

PAF PAAc 3-HPAF 3-HPAAc 3-HPATFAc

0.0552 0.0365 0.0293 0.0365 0.0334

1264 1253 1534 1606 1408

123.9 169.7 129.2 155.5 161.2

0.230 0.109 0.190 0.335 0.121

FIGURE 6. Comparison of DSC behaviors for ammonium and hydroxylammonium PILs. (a) PAF; (b) PAAc; (c) 3-HPAAc; (d) 3-HPAF; (e) 3-HPATFAc.

P.K. Chhotaray, R.L. Gardas / J. Chem. Thermodynamics 72 (2014) 117–124

using equation (9) at 298 K and shown in table 2. The calculated values between (480 and 520) kJ  mol1 show much lower lattice energy than the CsI which has the lowest lattice energy (613 kJ  mol1) among all the alkyl halides [50]. This low lattice energy in case of ionic liquids may be attributed to their liquid nature at room temperature. Also from table 2 it can be seen that lattice energy decreases near about 19 kJ  mol1 in both type of ionic liquids which demonstrates a decrease in interaction energy with increasing alkyl chain length on anionic side chain. The thermo gravimetric analysis curves of the studied protic ILs with a heating rate of 10 °C  min1 are shown in figure 5. The thermal stability is seen to vary with the (cation + anion) combination under study, being the highest in case of 3-HPATFAc ionic liquid. The onset decomposition temperatures (Td) are illustrated in table 2. The thermal decomposition (Td) of PILs proceeds through equilibrium shifting towards neutral components. As shown in figure 5, in case of hydroxylammonium ILs, higher the DpKa higher is the thermal decomposition which is attributed to the strength of NAH bonding arises due to efficient proton transfer [40] but inconsistent results observed in case of ammonium ILs, even though PAAc has lower DpKa it shows higher thermal decomposition as compared to its counterpart PAF. For PAF and PAAc there is an initial weight loss at around 70 °C followed by a steep shouldering and the weight lost almost completely near about 140 °C. Hydroxyl substituent on the alkyl chains increases the onset decomposition temperature for 3-HPAF and 3-HPAAc as can be seen from figure 5, which shows a steep shouldering from (100 to 180) °C following an initial weight loss. Among all the studied ionic liquids, 3-HPATFAc shows the highest thermal stability up to 140 °C followed by constant weight loss up to 235 °C, where it decomposes completely. The DSC plots of the hydroxylammonium and ammonium ionic liquids are illustrated in figure 6 and the glass transition temperatures (Tg) are presented in table 2. Samples are cooled down to 80 °C and successively heated at the rate of 2 °C  min1. As the temperature increases, all of them shows a second order endothermic transition near about 25 °C, which could be assigned to the corresponding glass-transition (Tg) points, with only exception of PAF. After glass transition, molecules may obtain enough freedom of motion and have undergone a transition from a lower energy, hard and relatively brittle state into a rubber like higher energy state. In each of the three repeated cycle, all ionic liquids associated with a small exothermic peak (more prominent in PAAc) before undergoing the endothermic glass transition. We are working onto conform whether this exotherm is a characteristics peak of ionic liquid or it is due to the presence of contaminants. A higher glass transition point is observed from 3-HPAF to 3-HPAAc as the anion has a longer alkyl side chain. As shown in figure 6 the ionic liquid PAF shows an exothermic cold crystallization peak centered around 60 °C with heat changes of 112.8 J  g1. The endothermic transition centered at around 34 °C may correspond to melting point with a latent heat of 86.16 J  g1. According to our best knowledge, no literature data available for velocity of sound for studied five ILs, whereas Pinkert et al. [35] measured density and viscosity of two ILs (3-HPAAc and 3-HPAF) in the temperature range from (278 to 348) K at the interval of 10 K. For both ILs, 3-HPAAc and 3-HPAF, our density values are in good agreement (maximum absolute relative deviation of 0.9%) with Pinkert et al. [35]. However, large absolute relative deviation in viscosity data are observed (for 3-HPAF are up to 7% and for 3-HPAAc are up to 23%) because the experimental viscosities are substantially lower than Pinkert et al. [35] which is mainly due to difference in water content (table 2) and measurement methods. Presence of the water molecules decreases the viscosity remarkably by reducing the electrostatic attractions among the ions and hence overall cohesive energy of the system.

123

4. Conclusions In the present work, experimental density, viscosity and sound velocity of ammonium and hydroxylammonium ionic liquids were measured. Experimental densities were fitted well with the second order polynomial as well as Gardas and Coutinho model. The coefficient of thermal expansion, isentropic compressibility, molecular volume and lattice energy were determined from experimental density and sound velocity data. It was found that the variation of volume expansion of these ionic liquids could be considered as independent of temperature within the studied temperature range. Among same moieties, either ammonium or hydroxylammonium ionic liquids those having higher DpKa, show higher density, lower viscosity, with an exception of 3-HPATFAc. This exception may be attributed to the presence of three highly electronegative fluorine atoms. The effect of alkyl chain length on the properties also discussed. As expected the density and sound velocity decreases whereas viscosity increases upon increasing the alkyl chain length on anionic side. There is no appreciable effect of alkyl chain length found on thermal decomposition. Hydroxyl substituent on cationic side chain has a pronounced effect on all the experimental properties studied due to additional hydrogen bonding apart from van der Waals and columbic interactions. The calculated lattice potential energy was found to be lower than that in ionic solids. DSC shows second order phase changes for all studied ionic liquids except PAF which involve first order phase change of crystallization and melting. Acknowledgements We are thankful to Council of Scientific and Industrial Research (CSIR), India and New Faculty Scheme, IIT Madras for their generous research support. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jct.2014.01.004. References [1] T. Welton, Chem. Rev. 99 (1999) 2071–2083. [2] T.L. Greaves, A. Weerawardena, I. Krodkiewska, C.J. Drummond, J. Phys. Chem. B 112 (2008) 896–905. [3] T.L. Greaves, C.J. Drummond, Chem. Rev. 108 (2008) 206–237. [4] W. Tamura-Lis, L.J. Lis, P.J. Quinn, Biophys. J. 53 (1988) 489–492. [5] M.U. Araos, G.G. Warr, J. Phys. Chem. B 109 (2005) 14275–14277. [6] R. Atkin, G.G. Warr, J. Am. Chem. Soc. 127 (2005) (1941) 11940–11941. [7] W. Tamura-Lis, L.J. Lis, P.J. Quinn, J. Colloid Interface Sci. 150 (1992) 200–207. [8] J.C. Gálvez-Ruiz, G. Holl, K. Karaghiosoff, T.M. Klapötke, K. Löhnwitz, P. Mayer, H. Nöth, K. Polborn, C.J. Rohbogner, M. Suter, J.J. Weigand, J. Inorg. Chem. 44 (2005) 4237–4253. [9] W. Tamura-Lis, L.J. Lis, J. Phys. Chem. 91 (1987) 4625–4627. [10] I. Cota, R. Gonzalez-Olmos, M. Iglesias, F. Medina, J. Phys. Chem. B 111 (2007) 12468–12477. [11] C. Yue, A. Mao, Y. Wei, M. Lu, Catal. Commun. 9 (2008) 1571–1574. [12] Y.O. Sharma, M.S. Degani, Green Chem. 11 (2009) 526–530. [13] J. Pernak, I. Goc, I. Mirska, Green Chem. 6 (2004) 323–329. [14] Md.A.B.H. Susan, A. Noda, S. Mitsushima, M. Watanabe, Chem. Commun. 8 (2003) 938–939. [15] H.M. Choi, I. Kwon, Ind. Eng. Chem. Res. 50 (2011) 2452–2454. [16] N. Bicak, J. Mol. Liq. 116 (2005) 15–18. [17] H. Salari, A.R. Harifi-Mood, M.R. Elahifard, M.R. Gholami, J. Solution Chem. 39 (2010) 1509–1519. [18] J.E. Bara, D.E. Camper, D.L. Gin, R.D. Noble, Acc. Chem. Res. 43 (2010) 152–159. [19] N. McCann, M. Maeder, M. Attalla, Ind. Eng. Chem. Res. 47 (2008) 2002–2009. [20] J. Alejandre, J.L. Rivera, M.A. Mora, V. de la Garza, J. Phys. Chem. B 104 (2000) 1332–1337. [21] K.E. Gutowski, E.J. Maginn, J. Am. Chem. Soc. 130 (2008) 14690–14704. [22] E.F. da Silva, H.F. Svendsen, Ind. Eng. Chem. Res. 45 (2005) 2497–2504.

124

P.K. Chhotaray, R.L. Gardas / J. Chem. Thermodynamics 72 (2014) 117–124

[23] R. López-Rendón, M.A. Mora, J. Alejandre, M.E. Tuckerman, J. Phys. Chem. B 110 (2006) 14652–14658. [24] F. Karadas, M. Atilhan, S. Aparicio, Energy Fuels 24 (2010) 5817–5828. [25] Y.S. Kim, W.Y. Choi, J.H. Jang, K.P. Yoo, C.S. Lee, Fluid Phase Equilib. 228–229 (2005) 439–445. [26] S.N.V.K. Aki, B.R. Mellein, E.M. Saurer, J.F. Brennecke, J. Phys. Chem. B 108 (2004) 20355–20365. [27] C. Cadena, J.L. Anthony, J.K. Shah, T.I. Morrow, J.F. Brennecke, E.J. Maginn, J. Am. Chem. Soc. 126 (2004) 5300–5308. [28] M. Ramdin, T.W.D. Loos, T.J.H. Vlugt, Ind. Eng. Chem. Res. 51 (2012) 8149– 8177. [29] A.P.S. Kamps, D. Tuma, J.Z. Xia, G. Maurer, J. Chem. Eng. Data 48 (2003) 746– 749. [30] L.A. Blanchard, Z.Y. Gu, J.F. Brennecke, J. Phys. Chem. B 105 (2001) 2437–2444. [31] S.J. Zhang, X.L. Yuan, Y.H. Chen, X.P. Zhang, J. Chem. Eng. Data 50 (2005) 1582– 1585. [32] X.L. Yuan, S.J. Zhang, J. Liu, X. Lu, Fluid Phase Equilib. 257 (2007) 195–200. [33] S. Aparicio, M. Atilhan, M. Khraisheh, R. Alcalde, J. Fernandez, J. Phys. Chem. B 115 (2011) 12487–12498. [34] M. Anouti, M.C. Caravanier, C.L. Floch, D. Lemordant, J. Phys. Chem. B 112 (2008) 9406–9411. [35] A. Pinkert, K.L. Ang, K.N. Marsh, S. Pang, Phys. Chem. Chem. Phys. 13 (2011) 5136–5143. [36] Y. Xu, J. Yao, C. Wang, H. Li, J. Chem. Eng. Data 57 (2012) 298–308. [37] K.A. Kurnia, C.D. Wilfred, T. Murugesan, J. Chem. Thermodyn. 41 (2009) 517– 521. [38] D.R. Lide, CRC Handbook of Chemistry and Physics, 90th ed., CRC Press/Taylor and Francis, Florida, 2009 (CD-ROM Version 2010). [39] J.A. Dean, Lange’s Handbook of Chemistry, 15th ed., McGraw-Hill Inc., New York, 1999.

[40] J. Stoimenovski, E.I. Izgorodina, D.R. MacFarlane, Phys. Chem. Chem. Phys. 12 (2010) 10341–10347. [41] M. Yoshizawa, W. Xu, C.A. Angell, J. Am. Chem. Soc. 125 (2003) 15411–15419. [42] L.D. Landau, E.M. Lifschitz, Fluid Mechanics, second ed., Pergamon Press, 1987. [43] J.S. Rowlinson, Liquids and Liquid Mixtures, Butterworths, London, 1959. [44] T.J. Fortin, A. Laesecke, M. Freund, S. Outcalt, J. Chem. Thermodyn. 57 (2013) 276–285. [45] K.N. Marsh, J.F. Brennecke, R.D. Chirico, M. Frenkel, A. Heintz, J.W. Magee, C.J. Peters, L.P.N. Rebelo, K.R. Seddon, Pure Appl. Chem. 81 (2009) 781–790. [46] M.S. Miran, H. Kinoshita, T. Yasuda, A.B.H. Susan, M. Watanabe, Phys. Chem. Chem. Phys. 14 (2012) 5178–5186. [47] H.F.D. Almeida, H. Passos, J.A. Lopes-da-Silva, A.M. Fernandes, M.G. Freire, J.A.P. Coutinho, J. Chem. Eng. Data 57 (2012) 3005–3013. [48] L. Glasser, Thermochim. Acta. 421 (2004) 87–93. [49] J. Tong, Q.-S. Liu, W.-G. Xu, D.-W. Fang, J.-Z. Yang, J. Phys. Chem. B 112 (2008) 4381–4386. [50] J.-Y. Wang, F.-Y. Zhao, R.-J. Liu, Y.-Q. Hu, J. Chem. Thermodyn. 43 (2011) 47–50. [51] R.L. Gardas, J.A.P. Coutinho, Fluid Phase Equilib. 263 (2008) 26–32. [52] R.L. Gardas, M.G. Freire, P.J. Carvalho, I.M. Marrucho, I.M.A. Fonseca, A.G.M. Ferreira, J.A.P. Coutinho, J. Chem. Eng. Data 52 (2007) 1881–1888. [53] R.L. Gardas, M.G. Freire, P.J. Carvalho, I.M. Marrucho, I.M.A. Fonseca, A.G.M. Ferreira, J.A.P. Coutinho, J. Chem. Eng. Data 52 (2007) 80–88. [54] R.L. Gardas, H.F. Costa, M.G. Freire, P.J. Carvalho, I.M. Marrucho, I.M.A. Fonseca, A.G.M. Ferreira, J.A.P. Coutinho, J. Chem. Eng. Data 53 (2008) 805–811. [55] B. Jacobson, J. Chem. Phys. 20 (1952) 927–928. [56] J. Jacquemin, P. Husson, A.A.H. Pádua, V. Majer, Green Chem. 8 (2006) 172– 180. [57] K.R. Seddon, A. Stark, M.J. Torres, Pure Appl. Chem. 72 (2000) 2275–2287.

JCT 13-366