J. Chem. Thermodynamics xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

J. Chem. Thermodynamics

journal homepage: www.elsevier .com/locate / jc t

Physicochemical properties of low viscous lactam based ionic liquids

http://dx.doi.org/10.1016/j.jct.2014.02.0090021-9614/� 2014 Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Tel.: +91 44 2257 4248; fax: +91 44 2257 4202.E-mail address: gardas@iitm.ac.in (R.L. Gardas).URL: http://www.iitm.ac.in/info/fac/gardas (R.L. Gardas).

Please cite this article in press as: P.K. Chhotaray et al., J. Chem. Thermodyn. (2014), http://dx.doi.org/10.1016/j.jct.2014.02.009

Pratap K. Chhotaray, Shankar Jella, 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 a b s t r a c t

Article history:Received 21 January 2014Received in revised form 13 February 2014Accepted 15 February 2014Available online xxxx

Keywords:Ionic liquidDensityViscosityCaprolactamButyrolactam

Lactam based ionic liquids were synthesized by an atom economization process between a lactam such ascaprolactam or butyrolactam with a Brønsted acid such as formic acid, acetic acid or hexanoic acid. Thedensity, speed of sound and viscosity were measured at atmospheric pressure and as a function of tem-perature from T = (293.15 to 333.15) K. The experimental density and viscosity values were fitted withlinear and Vogel–Tamman–Fulcher (VTF) equations, respectively and found to be fitting well withinthe experimental error. Thermodynamically important derived properties such as the coefficient of ther-mal expansion (a) and isentropic compressibility (bs) were calculated from the experimental density andspeed of sound values. Lattice potential energy (UPOT) has been calculated to understand the strength ofionic interaction between the ions and the standard entropy (So) has been estimated to assess the disor-der within the fluids. The remarkably low values of viscosity of ionic liquids studied are discussed on thebasis of activation energy estimated from the Arrhenius equation. Furthermore, the effect of alkyl chainlength on the anion, geometry of the cation and temperature has been analysed for the propertiesstudied.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

The ionic liquid (IL), a molecular puzzle in the hand of scientificcommunity, has witnessed an exceptional growth of researchactivity over the last decade. Researchers from all disciplines re-lated to chemistry have now kicked off the studies to improvethe molecular design required for the specific application. Unlikemolecular solvents, their unmatched characteristic propertieswhich include low volatility, high range in the liquid state, goodthermal stability, and high solvating capacity put together, makeit a potential alternative solvent. The applications span optical[1], electrical [2], chemical [3] to thermal [4] and mechanical [5].

To date, most work has been focused on the imidazolium basedionic liquid [6,7]. Even so, some work also has been centred onpyridinium, pyrrolidinium and piperidinium [8–10], triazolium[11,12], ammonium and phosphonium [13–16] based ionic liquid.Always there has been an inquisitive search to develop a novelionic liquid that can meet the new challenges and extend the po-tential applications. In this context, ionic liquids based on e-capro-lactam and c-butyrolactam (i.e. 2-pyrrolidone) are relatively newas compared to the ionic liquids mentioned above. Compared toimidazolium based ionic liquids, caprolactam ionic liquid (CPIL)

and butyrolactam ionic liquid (BTIL) are relatively cheap and havelower intrinsic toxicity [17,18]. CPIL and BTIL can be potentiallyused for gas sweetening as they have high affinity towards NO,NO2 [19], SO2 [20,21] and H2S [22]. Also caprolactam hydrogen sul-fate ionic liquid is being used as catalyst and reaction media formore than a decade in large scale industrial processes for Beck-mann rearrangement of cyclohexanone oxime to caprolactam[23,24]. In addition, the presence of the carbonyl group in the lac-tam based ionic liquid might encourage specific interaction whilebeing used as reaction media or catalyst.

It is evident that carboxylate based ILs are suitable for the dis-solution of cellulose [25,26] and lignocellulose [27] during biofuelproduction. The carboxylate ion having high hydrogen bond basi-city (Kamlet–Taft parameter, b) can strongly coordinate with thehydroxyl group of carbohydrates and help in dissolution [27].Moreover the viscosity of ILs plays an important role for the disso-lution of biomass composites. The undesirable viscosity slowsdown the rate of enzyme catalysed hydrolysis [28] and also thedissolution of cellulose [29–31].This work may open a new win-dow as it satisfies both the condition of low viscosity and the pres-ence of the carboxylate ion for cellulose processing.

The knowledge of the physicochemical properties is essentialfor the design of reaction and processing units, which influence di-rectly equipment performance. Except for the work by Deng et al.[32] who discussed the thermophysical properties of lactam basedionic liquids only at room temperature, there is no open literature

SCHEME 1. Synthesis of lactam based ionic liquid. n = 1, Butyrolactam; n = 3,Caprolactam; R@H, CH3, C5H11.

2 P.K. Chhotaray et al. / J. Chem. Thermodynamics xxx (2014) xxx–xxx

available that describes the effect of temperature upon thermo-physical properties of lactam based ILs. In this work, we measuredthe density (q), viscosity (g) and speed of sound (u) over a range oftemperature from T = (293.15 to 333.15) K at atmospheric pressurefor six novel lactam based ILs. The effects of the ring size of the cat-ion and alkyl chain length of anion upon the properties are alsodiscussed.

2. Experimental

2.1. Synthesis of ionic liquids

The source and purity of all the organic acids, amides used inthis study are summarized in table 1 and used without furtherpurification. The typical synthetic procedure adopted for caprolac-tam formate is illustrated in scheme 1 and can also be used as anexample for the synthesis of the remaining ionic liquid. Benzene/water (30 mL) was added to the double necked round bottom flaskcontaining caprolactam (11.32 g, 0.1 mol) equipped with a refluxcondenser. During vigorous stirring, formic acid (11.4 g, 0.1 mol)was added slowly through a dripping funnel. The temperature ofthe bath was kept below 10 �C during the addition and then ele-vated subsequently to room temperature. The reaction lasted foranother 24 h under a nitrogen atmosphere. The Benzene/watermix was then removed using a rotary evaporator and further driedby high vacuum for 2 h. The name, abbreviations, the respectivecations and anions used in this work, are shown in table 2.

2.2. Characterizations

Proton NMR was recorded on a Brukar Avance 500 MHzspectrometer. For BTF (CDCl3) d = 2.14 ppm (qn, 2H), d = 2.39 ppm(t, 2H), d = 3.45 ppm (t, 2H), d = 7.34 ppm (s, 1H), d = 8.04 ppm(s, 1H), d = 9.08 ppm (s, 1H). For BTAc (DMSO-d6) d = 1.91 ppm (s,3H), d = 1.96 ppm (qn, 2H), d = 2.06 ppm (t, 2H), d = 3.2 ppm(t, 2H), d = 7.49 ppm (s, 1H). For BTH (DMSO-d6) d = 0.86 ppm (t,3H), d = 1.26 ppm (m, 4H), d = 1.49 ppm (qn, 2H), d = 1.96 ppm(qn, 2H), d = 2.06 ppm (t, 2H), d = 2.18 ppm (t, 2H), d = 3.20 ppm(t, 2H), d = 7.50 ppm (broad, 1H), d = 11.96 ppm (broad, 2H).For CPF (CDCl3) d = 1.67 ppm (m, 4H), d = 1.77 ppm (qn, 2H),d = 2.47 ppm (t, 2H), d = 3.22 ppm (qr, 2H), d = 7.53 ppm (broad,NH), d = 8.05 ppm (s, 1H). For CPAc (CDCl3) d = 1.65 ppm (m, 4H),d = 1.75 ppm (qn, 2H), d = 2.04 ppm (s, 3H), d = 2.45 ppm (t, 2H),d = 3.20 ppm (qr, 2H), d = 7.46 ppm (broad, NH). For CPH(CDCl3) d = 0.88 ppm (t, 3H), d = 1.31 ppm (m, 4H), d = 1.63 ppm(m, 4H), d = 1.68 ppm (qn, 2H), d = 1.75 ppm (qn, 2H),d = 2.30 ppm (t, 2H), d = 2.46 ppm (t, 2H), d = 3.21 ppm (qr, 2H),d = 7.37 ppm (broad, NH).

2.3. Measurement technique

The density and speed of sound were measured with an AntonPaar (DSA 5000 M) instrument which employs the well-knownoscillating U-tube principle (for density measurement). The instru-ment has a number of features for easy sample handling and toproduce reliable results. It can measure the density in the range

TABLE 1Provenance and mass fraction purity for each chemical sample used during this work.

Chemical name Source Mass fraction purity

e-Caprolactam Sigma Aldrich 0.99c-Butyrolactam Sigma Aldrich 0.99Formic acid Sigma Aldrich P0.95Acetic acid Sigma Aldrich 0.99Hexanoic acid Sigma Aldrich 0.99

Please cite this article in press as: P.K. Chhotaray et al., J. Chem. Thermodyn. (

of (0 to 3) g � cm�3 and speed of sound from (1000 to 2000) m � s�1,simultaneously at temperatures from T = (273.15 to 343.15) K,with a pressure variation from (0 to 0.3) MPa. A Lovis 2000 ME (An-ton Paar) was used to measure the dynamic viscosity of the ionicliquid. It can measure the viscosity from (0.3 to 10,000) mPa � susing the rolling ball technique. The temperature is controlled bya built-in precise Peltier thermostat with an accuracy ofT = 0.02 K. At regular intervals of time, the density meter was cali-brated with dry air and Millipore water as described in the manualand the Lovis instrument was calibrated with the reference liquid(S3, N26, N100 liquid for (1.59, 1.8 and 2.5) mm capillary, respec-tively) supplied by the Anton Paar Co., Austria. The samples wereloaded into the density, speed of sound cell as well as the Loviscapillary in one cycle and the measurements were carried out bythe slow equilibration mode, simultaneously. The values reportedhere are the average of three consecutive measurements carriedout between the temperatures T = (293.15 to 333.15) K with aninterval of T = 5 K at atmospheric pressure. To verify theinstrument, we measured all the intended properties of thereference IL, 1-hexyl-3-methylimidazolium bis(trifluoromethylsul-fonyl)imide, [C6mim][Tf2N] [33,34]. The standard uncertaintiesassociated with the measurements were estimated to be less thanT = 0.002 K for temperature, 0.007 kg �m�3 for density,0.005 mPa � s for viscosity and 0.5 m � s�1 for the speed of sound.

The Karl Fischer Titrator from Analab (Micro Aqua Cal 100) wasused to measure the water content. The instrument operates on theconductometric titration principle using dual platinum electrodesthat permits detection of water content from less than 10 � 10�6

to 100%. The instrument was calibrated with Millipore qualitywater.

Thermograms are recorded using the TGA instrument (TAinstruments Hi-Res. TGA Q500) with a weighing precision of±0.01% at a heating rate of 10 �C �min�1 under a nitrogen atmo-sphere with an open aluminium pan. Simultaneous measurementof sample temperature and heating rate control were carried outaccurately and precisely by the two thermocouples, positionedimmediately adjacent to the sample. The Curie point temperaturecalibration was carried out with Nickel metal.

3. Results and discussion

Before the measurement of thermophysical properties, sampleswere dried under high vacuum for 24 h, water content was thenmeasured and reported in table 3. The experimental density ofILs was measured from T = (293.15 to 333.15) K at 0.1 MPa and val-ues summarized in table 4. As shown in figure 1, the structural andtemperature dependence on density cover a wide range from967.34 kg �m�3 for CPH at T = 333.15 K to 1140.78 kg �m�3 forBTF at T = 293.15 K. As expected, increase in carbon chain length,ring size upon anion and cation, respectively shows decreasing ef-fect over the density. For instance, ILs having same butyrolactamcation, at T = 303.15 K the density decreases from 1131.51 kg �m�3

for formate (BTF) to 1010.57 kg �m�3 for hexanoate (BTH) type ILs.Similarly for the same acetate anion at T = 303.15 K, there is de-crease in density from 1090.91 kg �m�3 for butyrolactam (BTAc)to 1051.83 kg �m�3 for caprolactam (CPAc) type ILs. This decreasein density can be attributed to the difficulties in the formation of

2014), http://dx.doi.org/10.1016/j.jct.2014.02.009

TABLE 2Ionic liquids name and abbreviations used in this work.

Cation Anion Name Abbreviation

γ-butyrolactam

formate

Butyrolactam formate BTF

acetate

Butyrolactam acetate BTAcButyrolactam hexanoate BTH

ε-caprolactam hexanoate

Caprolactam formate CPFCaprolactam acetate CPAcCaprolactam hexanoate CPH

TABLE 3Molar mass (M), water content, decomposition temperature (Td), molecular volume(Vm), standard entropy (So) and lattice potential energy (UPOT) of the ionic liquids.

IL M Water content TdaVm

aSo aUPOT

g �mol�1 10�6 �C nm3 J � K�1 �mol�1 kJ �mol�1

BTF 131.13 3832 110 0.192 267.8 510.72BTAc 145.15 3376 117 0.220 303.7 492.40BTH 201.26 3436 120 0.329 441.0 443.52CPF 159.19 4127 115 0.244 333.7 479.19CPAc 173.21 3872 119 0.272 369.0 465.72CPH 229.32 3129 126 0.383 506.6 426.90

a At T = 298 K.

FIGURE 1. Plot of the experimental values of density as a function of temperature.+, BTF; �, BTAc; ⁄, BTH; N, CPF; j, CPAc; �, CPH. The symbols representexperimental values and the solid lines correspond to the fit of the data byequation (1).

P.K. Chhotaray et al. / J. Chem. Thermodynamics xxx (2014) xxx–xxx 3

hydrogen bonds and packing and arises due to the asymmetry ofions. Having the same acetate anion, the density of the lactambased ILs studied at T = 303.15 K (BTAc: 1090.93 kg �m�3; CPAc:1051.83 kg �m�3) is greater than those of 1-ethylimidazoliumacetate [C2im][Ac], 1-butyl-3-methylimidazolium acetate[C4mim][Ac], 1-butyl-1-methylpyrrolidinium acetate [C4mpyr][Ac]and N,N-dimethyl-N-ethylammonium acetate [N0122][Ac] whichpossess densities of (1029.9, 1049.2, 1018.1 and 1012.2) kg �m�3,respectively [35]. The above fact may be related to the presenceof carbonyl group on caprolactam and butyrolactam cation, whichmay facilitate the formation of hydrogen bond and subsequentlyleads to increase in density as compared to other reported cationicmoiety. The experimental densities were fitted well with the linearequation given as follows:

q ¼ aþ b � T; ð1Þ

where a and b are empirical parameters obtained by the method ofleast squares fitting. The fitting parameters and average absoluterelative deviation (ARD) defined by the equation (2) are presentedin table 5.

ARD ¼ 1n

X jqexp � qcaljqexp

!100; ð2Þ

where n is the number of data points. From the experimental den-sity results, the isobaric coefficient of thermal expansion can be cal-culated by using the equation (3) as follows:

a ¼ � 1q

@q@T

� �P

: ð3Þ

Please cite this article in press as: P.K. Chhotaray et al., J. Chem. Thermodyn. (

The a values are presented in table 4. For the butyrolactam ionic li-quid (BTIL), with the change alkyl side chain length upon anions,the coefficients of thermal expansions are within experimentaluncertainty. However, caprolactam ionic liquid (CPIL) illustrates asmall increment of coefficient of thermal expansion upon increasingthe alkyl chain length of anion. This indicates the decrease in degreeof ordering of ILs with increasing carbon chain length. But inciden-tally, the coefficient of thermal expansion decreases with increasingsize of the ring. The reason may be due to the difference in molec-ular geometry of the cation. It is well accepted by quantum calcula-tion [36], IR, NMR [37], and XRD [38] that, e-caprolactam exists inits most stable chair conformation, whereas Warshel et al. [39] sug-gest that c-butyrolactam has the planar geometry. In spite of highernumber of carbon atoms in caprolactam ILs, the chair conformationmay lead to a well packed arrangement such that the coefficient ofthermal expansion becomes less compared to butyrolactam ILs. Ascan be seen, there are no considerable changes of thermal expan-sion coefficients (ca. 3%) with temperature over the studied intervalof T = 50 K. But having the same carboxylate anion, for the temper-ature interval of T = (303.15 to 333.15) K, the thermal expansioncoefficient is greater in the lactam ILs studied (2.46% for BTAc,

2014), http://dx.doi.org/10.1016/j.jct.2014.02.009

TABLE 4Experimental density (q), viscosity (g), speed of sound (u), coefficient of thermal expansion (a), isentropic compressibility (bs) of pure BTF, CPF, BTAc, CPAc, BTH and CPH fromT = (293.15 to 333.15) K at atmospheric pressure.

T q g U a � 104 bs q g u a � 104 bs

K kg �m�3 mPa � s m � s�1 K�1 TPa�1 kg �m�3 mPa � s m � s�1 K�1 TPa�1

BTF CPF293.15 1140.78 7.17 1531.18 8.12 373.89 1087.05 30.96 1551.98 7.70 381.92298.15 1136.14 6.14 1515.70 8.15 383.13 1082.87 23.92 1535.98 7.73 391.43303.15 1131.51 5.32 1500.02 8.19 392.78 1078.69 18.86 1520.15 7.76 401.17308.15 1126.89 4.64 1484.38 8.22 402.74 1074.51 15.14 1504.37 7.79 411.23313.15 1122.25 4.09 1468.45 8.25 413.23 1070.32 12.36 1488.45 7.82 421.71318.15 1117.61 3.63 1452.90 8.29 423.88 1066.13 10.24 1472.80 7.85 432.41323.15 1112.99 3.25 1437.45 8.32 434.84 1061.95 8.59 1457.93 7.88 443.02328.15 1108.35 2.92 1422.07 8.36 446.15 1057.77 7.30 1441.09 7.91 455.23333.15 1103.72 2.65 1406.80 8.39 457.80 1053.59 6.27 1425.08 7.94 467.36

BTAc CPAc293.15 1099.80 9.36 1514.48 8.08 396.42 1060.03 39.81 1523.08 7.72 406.67298.15 1095.35 7.84 1496.73 8.12 407.53 1055.92 29.03 1507.01 7.75 417.00303.15 1090.91 6.66 1480.86 8.15 418.01 1051.83 21.92 1489.70 7.78 428.41308.15 1086.47 5.72 1463.96 8.18 429.46 1047.74 17.03 1472.85 7.81 439.98313.15 1082.03 4.96 1447.15 8.22 441.30 1043.65 13.68 1456.08 7.84 451.93318.15 1077.57 4.35 1429.33 8.25 454.24 1039.56 11.10 1439.26 7.87 464.38323.15 1073.12 3.83 1412.54 8.28 467.04 1035.46 9.19 1422.59 7.90 477.21328.15 1068.69 3.40 1395.74 8.32 480.33 1031.37 7.67 1406.04 7.94 490.44333.15 1064.24 3.06 1378.97 8.35 494.14 1027.28 6.48 1389.49 7.97 504.20

BTH CPH293.15 1018.78 14.67 1429.33 8.05 480.46 998.59 42.00 1472.69 7.83 461.73298.15 1014.67 12.13 1412.80 8.08 493.76 994.69 30.85 1455.95 7.86 474.26303.15 1010.57 10.16 1396.20 8.11 507.62 990.78 23.40 1439.03 7.89 487.40308.15 1006.47 8.63 1379.64 8.15 522.00 986.88 18.11 1421.93 7.92 501.17313.15 1002.37 7.40 1363.22 8.18 536.83 982.97 14.25 1404.80 7.95 515.50318.15 998.27 6.42 1346.92 8.21 552.16 979.04 11.59 1387.83 7.98 530.30323.15 994.17 5.61 1330.72 8.25 568.02 975.15 9.44 1370.84 8.01 545.70328.15 990.08 4.94 1314.67 8.28 584.38 971.24 7.91 1353.43 8.05 562.09333.15 985.98 4.39 1298.59 8.32 601.43 967.34 6.64 1336.79 8.08 578.49

Standard uncertainties u are u(T) = 0.002 K, u(q) = 7 � 10�3 kg �m�3, u(u) = 0.5 m � s�1, u(g) = 0.005 mPa � s and the combined standard uncertainties uc are uc(bs) = 0.05 TPa�1,and uc(a) = 0.002 kK�1.

TABLE 5Fitting parameters of equation (1) for density and ARD.

PILs a � 101 b � 10�3 ARD � 103

BTF �9.264 1.412 0.328BTAc �8.890 1.360 0.477BTH �8.200 1.259 0.300CPF �8.367 1.332 0.158CPAc �8.185 1.300 0.246CPH �7.815 1.228 0.356

4 P.K. Chhotaray et al. / J. Chem. Thermodynamics xxx (2014) xxx–xxx

2.56% for BTH) as compared to [C4mim][Ac] (0.06%) [40], 2-hydrox-yethylammonium acetate (HEAc) (1.8%), bis-(2-hydroxyethyl)ammonium acetate (BHEAc) (1.9%) [41] and choline hexanoate(1.76%) [42] as reported in literature. This difference in expansioncoefficient may be arises due to structural differences in cation.

Figure 2 represents the temperature dependent speed of soundfrom T = (293.15 to 333.15) K at atmospheric pressure. As a normaltrend, the speed of sound decreases with increasing temperature.However, the addition of a carbon chain on either cation or anionproduces the opposite trend over the speed of sound. While theincrement of carbon chain length on the cationic side works in fa-vour of speed of sound, the increment on the anionic side reducesthe speed of sound in the ionic fluid. Table 4 also illustrates animportant thermodynamic property of the ionic liquids, viz. theisentropic compressibility (bs), as calculated by well-known New-ton–Laplace equation.

bs ¼1

qu2 ; ð4Þ

where q is the density and u is the speed of sound. The termisentropic means that as the sound wave pass through a fluid, thepressure and temperature fluctuate within each microscopic vol-ume but the entropy of the system as a whole remains constant.This condition holds true when we measure the speed of soundusing a low frequency generator (DSA 5000 M transducer producedapproximately 3 MHz frequency) [43], at high frequency(>100 MHz) there is velocity dispersion as well as absorption ofsound waves due to the coupling with molecular processes withinthe fluid [44,45]. As expected, isentropic compressibility increaseswith temperature in all of the ionic liquids studied. Moreover, itcan also be observed that increasing carbon chain length on the

Please cite this article in press as: P.K. Chhotaray et al., J. Chem. Thermodyn. (

anion leads to higher isentropic compressibility, which reflectsthe fact that ionic liquids are more structurally disordered andhence less densely packed. Unexpectedly, CPAc and CPF show high-er compressibility as compared to their butyrolactam counterpart,probably due to the combined effect of geometry of caprolactamand smaller size of the carboxylate ion.

To understand the intermolecular interaction, Jacobson sug-gested an empirical relation for calculating inter molecular freelength (Lf) [46]. According to which Lf is given by

Lf ¼ Kffiffiffiffiffibs

p; ð5Þ

where K is called as temperature dependent Jacobson’s constant.The values of intermolecular free length (Lf) for the ionic liquidsstudied at different temperatures are presented in table 6.

It can be seen clearly from table 6 that inter molecular freelength increases with increasing temperature, carbon chain lengthof anion as well as with ring size of cation. The speed of sound andinter molecular free length behave in a reverse manner as can beseen from tables 4 and 6. As the temperature increases, an increaseof inter molecular distance results, thereby increasing the distancebetween surfaces of the two molecules which in turn leads to

2014), http://dx.doi.org/10.1016/j.jct.2014.02.009

FIGURE 2. Plot of the experimental values of speed of sound as a function oftemperature. +, BTF; �, BTAc; ⁄, BTH; N, CPF; j, CPAc; �, CPH. The symbolsrepresent experimental values and dashed line in the figure is a guide for eyes.

So

= 34.5 nc

+ 334.1

b

P.K. Chhotaray et al. / J. Chem. Thermodynamics xxx (2014) xxx–xxx 5

decreasing the speed of sound. Similarly the addition of carbon onthe chain length on the anion side also adds more asymmetry tothe structure and reflects in their higher values of free lengthand lower value of speed of sound. As obvious, non-planar capro-lactam ILs shows the higher inter molecular separation as com-pared to planar butyrolactam ILs.

Molecular volumes (Vm) were calculated (at T = 298.15 K), fromthe experimental density as given by the following equation (6).

Vm ¼MNq

; ð6Þ

where N is the Avogadro’s number, M is the molar mass and q is thedensity (at T = 298.15 K) of the ionic liquid. As can be seen fromtable 3, the BTAc molecule occupies 0.029 nm3 more volume ascompared to BTF due to the presence of an additional methylenegroup. A similar increment of molecular volume (0.028 nm3) is alsoobserved from CPF to CPAc. Furthermore, there is a molecular vol-ume increment of approximately 0.026 nm3 per methylene groupbetween caprolactam and butyrolacatm ILs having the same anionand also 0.028 nm3 augmentation (per methylene group) in molec-ular volume observed between acetate and hexanoate based lactamILs. These experimental results are supported by the literature forionic liquids [40,47,48] and for other organic materials, where themean contribution due to methylene (–CH2) groups are as follows,for alcohols (0.028 nm3), n-amines (0.027 nm3), n-paraffins(0.027 nm3) [49]. These observations suggest that –CH2 contribu-tion to the molar volume is independent of the nature of both an-ions and cations of ILs.

Yang and co-workers [50] suggest that standard entropy can becalculated by using equation (7), established by Glasser [49].

S0 � 1246:5Vm þ 29:5: ð7Þ

Table 3 presents the calculated standard entropy at T = 298 K for io-nic liquids studied whereas figure 3 (left side) shows the linear

TABLE 6Inter molecular free length (10 � Lf/nm) at different temperatures.

T/K BTF BTAc BTH CPF CPAc CPH

293.15 0.378 0.389 0.428 0.382 0.394 0.420298.15 0.387 0.399 0.439 0.391 0.404 0.430303.15 0.395 0.408 0.450 0.400 0.413 0.441313.15 0.413 0.426 0.470 0.417 0.432 0.461323.15 0.430 0.446 0.491 0.434 0.450 0.482

Please cite this article in press as: P.K. Chhotaray et al., J. Chem. Thermodyn. (

dependencies of standard entropy with number of carbon atomon the carboxylate anion. Figure 3 (left side) depicts the increasingdisorder in IL molecules and hence the standard entropy as thenumber of carbon atom increases on carboxylate anion. In a similarfashion, the molecular volume also increases linearly with increas-ing number of carbon atom on carboxylate anion as shown in figure3 (right side). It is obvious that butyrolactam ILs contribute less tothe standard entropy due to its smaller size and planar structureas compared to chair conformation of caprolactam ILs. From the leftside of figure 3, it can be seen that the slope of the line is34.6 J � K�1 �mol�1 for butyrolactam ILs and 34.5 J � K�1 �mol�1 forcaprolactam ILs, which are close to the literature values obtainedfor [Cnmim][Ala] (34.6 J � K�1 �mol�1) [50], [Cnmim][BF4](33.9 J � K�1 �mol�1), [Cnmim][Tf2N] (35.1 J � K�1 �mol�1) [49].These slope values signify that each –CH2 group in the alkyl chainof the anions contributes to an increase of approximately34.5 J � K�1 �mol�1 to the standard entropy of the carboxylate ILs.

Glasser developed a method for the evaluation of lattice poten-tial energies (UPOT) for simple ionic solids of type MX with chargeratio (1:1) [49].

UPOT ¼ cqM

� �1=3þ d; ð8Þ

where q is the density, M is the molar mass, c and d are the con-stants. Equation (8) can also be applied to the condensed phase io-nic liquid with the constant values c = 1981.2 kJ �mol�1 � cm andd = 103.8 kJ �mol�1. The calculated values of the lattice energy atT = 298.15 K of the ILs studied are given in table 3 which indicatea monotonic decrease of lattice potential energy with increasingcarbon chain length over carboxylate ions due to the weaker pack-ing efficiency. The calculated values of lattice potential energy forionic liquids are smaller than those of inorganic salts. For instance,fused CsI has the lattice energy 613 kJ �mol�1, nearly 20% more thanthe highest observed lattice energy for lactam based ILs studied.This lower lattice energy in the case of ILs may be one of the pri-mary reasons for its existence as liquid at room temperature.

The temperature dependent dynamic viscosities measured fromT = (293.15 to 333.15) K with an increment of T = 5 K, at 0.1 MPaare presented in table 4. From figure 4, it can be seen that the long-er is the alkyl chain length of the anion, and larger is the cationicring size, a greater values of the viscosity results. That is the reasonwhy at T = 293.15 K CPH shows highest viscosity (42 mPa � s) andBTF (7.2 mPa � s) has the lowest. Increasing number of carbon oneither side of the ion not only makes the entity heavier and bulkier

a

So

= 34.6 nc

+ 268.4

IGURE 3. Plot of molar volume (Vm) and standard entropy (So) as a function of theumber of carbon atoms on the carboxylate anion of ionic liquids at T = 298.15 K. d,tandard entropy; s, molar volume. (a) Butyrolactam ILs; (b) Caprolactam ILs.

Fns

2014), http://dx.doi.org/10.1016/j.jct.2014.02.009

TABLE 7Adjustable parameters for the experimental viscosity of the ILs to VTF equation (9)along with ARD.

PILs g0/mPa � s B/K T0/K ARD

BTF 0.1243 500.6 169.8 0.216BTAc 0.0921 578.1 168.1 0.590BTH 0.1041 627.8 165.8 0.695CPF 0.0392 853.6 164.8 0.280CPAc 0.0236 928.6 167.4 0.834CPH 0.0199 957.0 167.9 0.802

FIGURE 5. Plot of the experimental values of viscosity (symbols) as a function oftemperature, linear approximations using the VTF equation (dashed lines) andArrhenius equation (solid lines) for the ILs.

TABLE 8Activation energies (Eg), infinite temperature viscosities (g1) and ARD, obtained byfitting experimental viscosity data using equation (10).

PILs Eg ± ra 104 � g1 ± ra ARDJ �mol�1 mPa � s

BTF 20.2 ± 0.3 17.71 ± 2.00 1.017BTAc 22.7 ± 0.4 8.31 ± 1.18 0.992BTH 24.1 ± 0.3 7.14 ± 0.79 1.032

6 P.K. Chhotaray et al. / J. Chem. Thermodynamics xxx (2014) xxx–xxx

but also leads to higher van der Waals attractions [51,52]. Also thenon-planar geometry of caprolactam moiety may further add diffi-culties to the transport of CPILs. The viscosities reported in thiswork are remarkably low compared to some well studied ILs. Forexample at T = 298.15 K, the viscosities are, 546 mPa � s for [N4][Ac](n-butylammonium acetate) [53], 441 mPa � s for [C4mim][Ac] [54]and 2827 mPa � s for [3-HPA][Ac] (3-hydroxypropylammoniumacetate) [55] as given in literature. The experimental viscositiesare fitted with the most widely used non Arrhenius Vogel–Tammann–Fulcher (VTF) equation:

g ¼ g0 expB

T � T0

� �; ð9Þ

where g0, B and T0 are adjustable parameters. These parameters arecalculated by fitting experimental data points and the values arepresented in table 7 along with average absolute relative deviation(ARD). Although the quality of fitting of ILs viscosities are good inthe case of VTF equation as compared to Arrhenius equation[35,56,57] yet we have used this equation to calculate activation en-ergy: a measure of energy required for the free movement of ions ofionic liquids. The logarithmic form of Arrhenius equation can be ex-plained as follows:

ln g ¼ ln g1 þEg

RT; ð10Þ

where R is the universal gas constant (8.314 J � K�1 �mol�1), g1 isthe infinite temperature viscosity, and Eg (J �mol�1) is the activationenergy. The Eg and g1 values were calculated from the slope andintercept, respectively of Arrhenius plot (figure 5, Solid lines) andpresented in table 8.

Figure 5 illustrates the fitting of experimental viscosity datawith VTF as well as the Arrhenius equation. In general, the Arrhe-nius equation prduces a good fitting, while a careful observationdepicts a non-linear concave dependence on viscosity as observedin other ILs [56,58,59]. Even though, the relative deviation fromexperimental viscosity is higher in the case of Arrhenius fittingas compared to VTF fitting (from tables 7 and 8), the activation en-ergy and the infinite temperature viscosity were calculated as afirst approximation to characterise the ionic liquids and comparedwith literature data [56]. The activation energy (Eg) is the mini-mum energy required for the ions to move across each other. Thelower is the activation energy value, the more easily ions are able

FIGURE 4. Plot of the experimental values of viscosity as a function of temperature.+, BTF; �, BTAc; ⁄, BTH; N, CPF; j, CPAc; �, CPH. The symbols representexperimental values and the solid lines correspond to the fit of the data byequation (9).

CPF 32.1 ± 0.5 0.56 ± 0.13 1.998CPAc 36.0 ± 0.8 0.14 ± 0.05 2.920CPH 37.2 ± 0.8 0.09 ± 0.04 3.056

a Standard deviation.

Please cite this article in press as: P.K. Chhotaray et al., J. Chem. Thermodyn. (

to move past each other. As can be seen from table 8, the activationenergy of the ILs studied are significantly lower (and hence theviscosity) by a third order magnitude while comparing with theliterature [C2mim][Ac] (38.9 kJ �mol�1) [60], [C4mim][Ac](44.4 kJ �mol�1), [N0112][Ac] (24.0 kJ �mol�1) [35], [C4mim][PF6](23.9 kJ �mol�1) [56]. This finding can be regarded as a direct con-sequence of the structural arrangements, lower symmetry andlower interactions among ions of the ionic liquid. An increasingcarbon chain length on either side of ion leads to an increase inactivation energy. The reason may be due to the inhibition of lam-inar flow as a result of increased molar mass and change in molec-ular geometry of ions. As can be seen from table 8, the increase incarbon chain length (five –CH2) on anion side from BTF to BTHcauses an increment of 3.9 J �mol�1 in activation energy, whereasaddition of only two –CH2 upon cationic side of BTF (to formCPF) leads to an increment of 11.9 J �mol�1 in the activation en-ergy. The above observation indicates that geometry of the ionic

2014), http://dx.doi.org/10.1016/j.jct.2014.02.009

FIGURE 6. Plot of weight per cent against temperature to illustrate the thermaldecomposition of lactam based PILs. (a) , BTF; (b) , CPF; (c) , BTAc; (d)

, CPAc; (e) , BTH; (f) ——, CPH.

P.K. Chhotaray et al. / J. Chem. Thermodynamics xxx (2014) xxx–xxx 7

fragments plays an important role for the viscosity behaviourrather addition only of the alkyl chain length. At infinite tempera-ture, as the interactions which contribute to the viscosity are nolonger effective, the resultant viscosity is due to the geometricstructure of the ions. Therefore, the g1 value describes the struc-tural contribution of the ions to the dynamic viscosity and gener-ally decreases with increase in activation energy as illustrated intable 8.

Thermal decomposition temperatures reported in table 3 repre-sent the onset decomposition temperature. The thermal stability ofILs primarily depends upon the strength of their heteroatom–car-bon and heteroatom–hydrogen bonds. As can be seen from figure6, the protic ionic liquids studied are less thermally stable com-pared to imidazolium type of aprotic ionic liquids [61]. This maybe due to proton back transfer, which proceeds through equilib-rium shifting towards neutral components. It is also clear from fig-ure 6 that all the ILs studied show almost similar behaviour whileundergoing thermal treatment with a heating rate of 10 �C �min�1.For all ionic liquids, the plot shows a steep shouldering near about100 �C followed by an initial weight loss. Among the ILs studied,CPH shows the highest thermal stability (126 �C) and BTF(110 �C) the lowest. Having the same carboxylate anion, caprolac-tam ionic liquids show a higher thermal decomposition tempera-ture than their butyrolactam counterpart. Also as the number ofcarbon atoms increase over those in the carboxylate anion, thermalstability increases from formate to hexanoate ILs. The above obser-vation indicates that the longer is the carbon chain length on car-boxylate anion, the greater is the thermal stability. Also the chairconformation of caprolactam ILs may favuor the formation ofhydrogen bonding, which affects their higher thermal stability ascompared to planar geometry of butyrolactam ILs.

4. Conclusions

This work presents a systematic thermophysical property studyof six lactam based ionic liquids at temperatures from T = (293.15to 333.15) K at atmospheric pressure. The experimental densityand viscosity values were fitted with linear and VTF equations,respectively and found to be fitting well. The molecular volume,lattice potential energy and standard entropy were estimated fromexperimental density values. Caprolactam ILs are thermally morestable than butyrolactam ILs. Also higher alkyl chain length on an-ion favours thermal stability. As expected, it appears that density

Please cite this article in press as: P.K. Chhotaray et al., J. Chem. Thermodyn. (

and speed of sound decreases while thermal expansion coefficientand isentropic compressibility increases linearly with temperature.Furthermore the viscosity shows an asymptotic decrement overstudied temperature range. The observed significantly lowviscosity of studied lactam based ILs is discussed on the basis ofactivation energy estimated from the Arrhenius equation. As thecarbon chain length increases over the anion, the density decreaseswhereas viscosity increases. This behaviour may be correlated tothe balance between van der Waals and electrostatic forces in-volved in ILs. The geometry of caprolactam (chair conformation)and butyrolactam (planar) plays a crucial role in guiding theproperties of studied ILs. The non-planar geometry of caprolactamcation may be responsible for the higher viscosity and lowerdensity values of caprolactam ILs as compared to its counterpartbutyrolactam ILs.

Acknowledgments

The author would like to acknowledge Council of Scientific andIndustrial Research (CSIR), and IIT Madras for their financialsupport.

References

[1] R.E.D. Sesto, C. Corley, A. Robertson, J.S. Wilkes, J. Organomet. Chem. 690(2005) 2536–2542.

[2] E. Dickinson, M.E. Williams, S.M. Hendrickson, H. Masui, R.W. Murray, J. Am.Chem. Soc. 121 (1999) 613–616.

[3] L.A. Blanchard, D. Hancu, E.J. Beckman, J.F. Brennecke, Nature 399 (1999) 28–29.

[4] M. Villanueva, A. Coronas, J. García, J. Salgado, Ind. Eng. Chem. Res. 52 (2013)15718–15727.

[5] M. Choa, H. Seo, J. Nam, H. Choi, J. Koo, Y. Lee, Sens. Actuators B 128 (2007) 70–74.

[6] J.J. Chen, W.W. Li, X.L. Li, H.Q. Yu, Phys. Chem. Chem. Phys. 14 (2012) 4589–4596.

[7] J. Kiefer, M.M. Molina, K. Noack, ChemPhysChem 13 (2012) 1213–1220.[8] R.C. Vieira, D.E. Falvey, J. Am. Chem. Soc. 130 (2008) 1552–1553.[9] 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.[10] P.K. Mandal, A. Samanta, J. Phys. Chem. B 109 (2005) 15172–15177.[11] S. Sanghi, E. Willett, C. Versek, M. Tuominen, E.B. Coughlin, R. Soc, Chem. Adv. 2

(2012) 848–853.[12] M.W. Schmidt, M.S. Gordon, J.A. Boatz, J. Phys. Chem. A 109 (2005) 7285–7295.[13] Z. Man, Y.A. Elsheikh, M.A. Bustam, S. Yusup, M.I.A. Mutalib, N. Muhammad,

Ind. Crops Prod. 41 (2013) 144–149.[14] U. Domanska, A. Pobudkowska, M. Królikowski, Fluid Phase Equilib. 259

(2007) 173–179.[15] R.L. Gardas, R. Ge, P. Goodrich, C. Hardacre, A. Hussain, D.W. Rooney, J. Chem.

Eng. Data 55 (2010) 1505–1515.[16] U. Domanska, L.M. Casás, J. Phys. Chem. B 111 (2007) 4109–4115.[17] E. Duan, Z. Li, J. Liu, Y. Liu, Y. Song, J. Guan, K. Yang, R. Soc, Chem. Adv. 3 (2013)

18817–18820.[18] J. Yang, Q. Zhang, L. Zhu, S. Zhang, J. Li, X. Zhang, Y. Deng, Chem. Mater. 19

(2007) 2544–2550.[19] E. Duan, B. Guo, D. Zhang, L. Shi, H. Sun, Y. Wang, J. Air Waste Manage. Assoc.

61 (2011) 1393–1397.[20] E. Duana, B. Guo, M. Zhang, Y. Guan, H. Sun, J. Han, J. Hazard. Mater. 194 (2011)

48–52.[21] B. Guo, E. Duan, A. Ren, Y. Wang, H. Liu, J. Chem. Eng. Data 55 (2010) 1398–

1401.[22] B. Guo, E. Duan, Y. Zhong, L. Gao, X. Zhang, D. Zhao, Energy Fuels 25 (2011)

159–161.[23] V. Fábos, D. Lantos, A. Bodor, A. Bálint, L.T. Mika, O.E. Sielcken, A. Cuiper, I.T.

Horváth, ChemSusChem 1 (2008) 189–192.[24] S. Guo, Z. Du, S. Zhang, D. Li, Z. Li, Y. Deng, Green Chem. 8 (2006) 296–300.[25] S. Fendt, S. Padmanabhan, H.W. Blanch, J.M. Prausnitz, J. Chem. Eng. Data 56

(2011) 31–34.[26] B. Zhao, L. Greiner, W. Leitner, R. Soc, Chem. Adv. 2 (2012) 2476–2479.[27] A. Brandt, J.P. Hallett, D.J. Leak, R.J. Murphy, T. Welton, Green Chem. 12 (2010)

672–679.[28] S. Bose, D.W. Armstrong, J.W. Petrich, J. Phys. Chem. B 114 (2010) 8221–8227.[29] O.A.E. Seoud, A. Koschella, L.C. Fidale, S. Dorn, T. Heinze, Biomacromolecules 8

(2007) 2629–2647.[30] A. Pinkert, K.N. Marsh, S.S. Pang, M.P. Staiger, Chem. Rev. 109 (2009) 6712–

6728.[31] M.E. Zakrzewska, E. Bogel-Lukasik, R. Bogel-Lukasik, Energy Fuels 24 (2010)

737–745.

2014), http://dx.doi.org/10.1016/j.jct.2014.02.009

8 P.K. Chhotaray et al. / J. Chem. Thermodynamics xxx (2014) xxx–xxx

[32] Z. Du, Z. Li, S. Guo, J. Zhang, L. Zhu, Y. Deng, J. Phys. Chem. B 109 (2005) 19542–19546.

[33] 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.

[34] R.D. Chirico, V. Diky, J.W. Magee, M. Frenkel, K.N. Marsh, Pure Appl. Chem. 81(2009) 791–828.

[35] 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.

[36] F. Groenewald, J. Dillen, Struct. Chem. 23 (2012) 723–732.[37] G. Borgen, F. Rise, Magn. Reson. Chem. 31 (1993) 51–53.[38] F.K. Winkler, J.D. Dunitz, Acta Crystallogr. Sect. B: Struct. Sci. 31 (1975) 268–

269.[39] A. Warshel, M. Levitt, S. Lifson, J. Mol. Spectrosc. 33 (1970) 84–99.[40] A. Xu, J. Wang, Y. Zhang, Q. Chen, Ind. Eng. Chem. Res. 51 (2012) 3458–3465.[41] K.A. Kurnia, C.D. Wilfred, T. Murugesan, J. Chem. Thermodyn. 41 (2009) 517–

521.[42] N. Muhammad, M.I. Hossain, Z. Man, M. El-Harbawi, M.A. Bustam, Y.A.

Noaman, N.B.M. Alitheen, M.K. Ng, G. Hefter, C.Y. Yin, J. Chem. Eng. Data 57(2012) 2191–2196.

[43] T.J. Fortin, A. Laesecke, M. Freund, S. Outcalt, J. Chem. Thermodyn. 57 (2013)276–285.

[44] A.T.J. Hayward, Nature 221 (1969) 1047.[45] M.J. Blandamer, Introduction to Chemical Ultrasonics, Academic Press Inc.,

1973.[46] B. Jacobson, J. Chem. Phys. 20 (1952) 927–928.[47] 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.

Please cite this article in press as: P.K. Chhotaray et al., J. Chem. Thermodyn. (

[48] 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.

[49] L. Glasser, Thermochim. Acta 421 (2004) 87–93.[50] D.W. Fang, W. Guan, J. Tong, Z.W. Wang, J.Z. Yang, J. Phys. Chem. B 112 (2008)

7499–7505.[51] D. Rooney, J. Jacquemin, R. Gardas, Top. Curr. Chem. 290 (2009) 185–212.[52] J. Jacquemin, P. Husson, A.A.H. Padua, V. Majer, Green Chem. 8 (2006) 172–

180.[53] Y. Xu, J. Yao, C. Wang, H. Li, J. Chem. Eng. Data 57 (2012) 298–308.[54] C.P. Fredlake, J.M. Crosthwaite, D.G. Hert, S.N.V.K. Aki, J.F. Brennecke, J. Chem.

Eng. Data 49 (2004) 954–964.[55] P.K. Chhotaray, R.L. Gardas, J. Chem. Thermodyn. 72 (2014) 117–124.[56] O.O. Okoturo, T.J. VanderNoot, J. Electroanal. Chem. 568 (2004) 167–181.[57] M.H. Ghatee, M. Bahrami, N. Khanjari, H. Firouzabadi, Y. Ahmadi, J. Chem. Eng.

Data 57 (2012) 2095–2101.[58] R. Sescousse, K.A. Le, M.E. Ries, T. Budtova, J. Phys. Chem. B 114 (2010) 7222–

7228.[59] M. Gericke, K. Schlufter, T. Liebert, T. Heinze, T. Budtova, Biomacromolecules

10 (2009) 1188–1194.[60] M.G. Freire, A.R.R. Teles, M.A.A. Rocha, B. Schröder, C.M.S.S. Neves, P.J.

Carvalho, D.V. Evtuguin, L.M.N.B.F. Santos, J.A.P. Coutinho, J. Chem. Eng. Data56 (2011) 4813–4822.

[61] J.S. Wilkes, J. Mol. Catal. A: Chem. 214 (2004) 11–17.

JCT 14-44

2014), http://dx.doi.org/10.1016/j.jct.2014.02.009