The new evolution of protic ionic liquids: Antielectrostatic activity correlated with their surface properties

Journal of Industrial and Engineering Chemistry 41 (2016) 40–49

Contents lists available at ScienceDirect

Journal of Industrial and Engineering Chemistry journal homepage: www.elsevier.com/locate/jiec

The new evolution of protic ionic liquids: Antielectrostatic activity correlated with their surface properties Joanna Feder-Kubis a[8_TD$IF], Małgorzata Musiał b, Marzena Dzida b, Monika Geppert-Rybczyn´ska b,* a b

Faculty of Chemistry, Wrocław University of Science and Technology, Wybrzez˙e Wyspian´skiego 27, 50-370 Wrocław, Poland Institute of Chemistry, University of Silesia, Szkolna 9, 40-006 Katowice, Poland

A R T I C L E I N F O

Article history: Received 17 May 2016 Received in revised form 2 July 2016 Accepted 3 July 2016 Available online 13 July 2016 Keywords: Protic ionic liquid Antielectrostatic effect Surface tension Contact angle Parachor

A B S T R A C T

The primary aim of this study was to explain the antielectrostatic activity of a series of protic 3(alkoxymethyl)-1H-imidazol-3-ium salicylate ionic liquids as a function of their surface properties. The surface tension and contact angle on chosen surfaces were measured and discussed. The results were investigated as a function of the alkyl chain length of the alkoxymethyl substituent in the imidazolium cation. Despite the complex structure of the ions, the estimated surface tension was correlated with the corresponding experimental results. It was found that the alkyl chain length has a significant impact on the surface properties and wetting abilities of salicylates. ß 2016 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.

Introduction Since the 1990s, when interest in ionic liquids (ILs) was piqued [1,2], the general picture of this class of compound has been evaluated based from more-or-less accidental physicochemical characteristics to intentional investigations of the properties of selected compounds. Nowadays, the unusual properties of ILs are thought of as attractive if they can be turned into concrete applications. The term ‘‘task-specific ionic liquids’’ is not new, but it refers to the next generations of materials that are being synthesized for use in special projects (e.g., catalysis, synthesis, gas absorption, analysis, lubricants, antiwear properties, medicine, pharmacology, electrochemistry, and energetic materials) [2–4]. However, in order to determine new applications it is necessary to reveal the physicochemical properties of a given material. The surface tension of ILs has been frequently investigated [5]. Some examples of interfacial tension in mixtures of ILs with alkanols and alkanes (when the miscibility gap occurs) are also known [6–9]. These parameters are necessary to use ILs as solvents in synthesis [10,11] or for constructing the proper liquid-liquid

* Corresponding author. E-mail address: [email protected] (M. Geppert-Rybczyn´ska).

system for extraction [12–14]. On the other hand, contact angle measurements of ILs are very limited and are largely available only for commercial ILs (e.g., bis(trifluoromethylsulfonyl)imides [NTf2], tetrafluoroborates [BF4], hexafluorophosphates [PF6], ethyl sulfates [EtSO4] or dicyanamides [N(CN)2]) and surfaces such as glass, poly(tetrafluoroethylene) (PTFE), or silicone, or modified Si materials [15–17]. As a matter of fact, contact angles can be useful especially in heterogeneous systems or transport processes in capillary and porous surfaces, but there have been no attempts to correlate this quantity with such property, as is done in this work. Here, this manuscript discusses one of the series of protic ILs with a salicylate anion, namely 3-(alkoxymethyl)-1H-imidazol3-ium salicylates, [H-im-C1OCn][Sal]. The syntheses and thermal properties of considered salts have been previously described in the literature [18,19]. Another study reported the excellent results of 3-(alkoxymethyl)-1H-imidazol-3-ium salicylates as wood-preservative agents [20], where the salicylate ILs penetrate the wood surface and act effectively against fungi. In the same study, the authors noted the antielectrostatic effect of salicylates on some surfaces. This aspect seems to be extremely interesting, and our work accordingly focuses on it more specifically. Thus, some ILs have been shown to be effective antielectrostatic agents [21–24]. For instance, a series of imidazolium ILs containing

http://dx.doi.org/10.1016/j.jiec.2016.07.003 1226-086X/ß 2016 The Korean Society of Industrial and Engineering Chemistry. Published by Elsevier B.V. All rights reserved.

J. Feder-Kubis et al. / Journal of Industrial and Engineering Chemistry 41 (2016) 40–49

alkoxymethyl substituent were tested by Pernak et al. to determine their capacity to drain a surface electric charge [21]. These authors found that the type of anion strongly influenced the strength of the resulting antielectrostatic effect. This study presents not only the antielectrostatic activity of 3-(alkoxymethyl)-1H-imidazol-3-ium salicylates but also their surface and wetting behavior. The goal of this work was to determine a relation between these two kinds of properties. In particular, our investigation focuses on the influence of the alkyl chain length of the alkoxymethyl substituent in the imidazolium ring on these properties. From the point of view of applications, this question may be important since it allows for a reduction in the effort necessary to select the best candidate to improve the antielectrostatic properties of any material after simple surface tension and contact angle experiments. Despite the fact that the dependence between antielectrostatic activity and surface properties appears to be interesting, this subject has not been researched by other scientists yet. Experimental Chemicals, materials and synthesis The syntheses and thermal properties of 3-(alkoxymethyl)-1Himidazol-3-ium salicylate have been described in a previous publication [19]. Additional information about the material used for synthesis and preparation and the structure and purity of all of the investigated substances confirmed by spectral analysis (Table S.1) are collected in the Supporting Information. Prior to use, all of the chemicals were carefully dried under vacuum for approximately 24–48 h with heating (T = 333 K). The drying procedure was maintained until the water content, detected using a TitroLine 7500 Karl Fischer trace coulometer (SI Analytics GmbH, Germany), was wH2O ffi (3–6)104. The calibration and characteristics of the work of the coulometric Karl Fischer titrator have been already described by Skowronek et al. [25]. The samples were stored in an argon atmosphere in vessels closed with septa. Immediately prior to the experiments, the samples were degassed by heating them up to 323 K. The results of the water determinations of the samples after drying are listed in Table 1 together with the molar masses of the investigated ILs. Higher homologues of [H-im-C1OCn][Sal] with n = 9–10, as well as higher-melting-point polymorphic types of [H-im-C1OC11][Sal],

41

have a Tm of at least at 295 K [19]. Therefore, these compounds were measured in a super-cooled state. The materials that were used for the antielectrostatic investigation included polyethylene (LDPE) film (POLITEN II/003/GO produced by Chemical Institute Blachownia in Ke˛dzierzyn-Koz´le, Poland) and polypropylene (PP) nonwoven fabric (WIGOFIL/ MALEN P F-401 produced by the Wigolen joint-stock company in Cze˛stochowa, Poland). These materials were free of lubricants and antioxidants. Disks 12.5 mm in diameter were cut from the LDPE and PP and then washed in acetone and dried by placing them in an air-conditioned room. A disk was rubbed with a cotton swab soaked with a 0.5% chloroform solution of each of the salts that were studied. Next, each disk was hung so that the chloroform could evaporate spontaneously. The disks, covered with protic ILs, were stored for 24 h in an air-conditioned room at 294  2 K and a relative humidity of 55  5%. The materials used for contact angles measurements included PTFE film with a thickness of 0.1 mm and a density of 0.35 g cm3 (Bisan LLC, Poland) and clear glass PP with a thickness of 0.04 mm (Office Products, Poland). The substrates were only rinsed with distilled water and dried; they were then stored in an atmosphere with the same humidity as was used for the contact angle measurements (i.e., around 40%). Antielectrostatic properties The surface resistance Rs (V ), half charge decay time t1/2 (s) and maximum voltage Uind (V) induced on the disk surface of the polymers were determined. The surface resistance was measured using a Tralin III Statron electrometer, and the maximum voltage induced with a Type V 531 digital voltmeter (Meratronik, Poland). This apparatus and measurement procedure have been previously described in the literature [21]. The experiment was performed in duplicate. In each test, the surface resistance was measured at least three times on each side of the disk, and the other parameters (i.e., the half charge decay time and the maximum voltage) were measured twice. The relative error in the determination of these three quantities did not exceed 5%. Density, surface tension and contact angles measurements The density, r, necessary for carrying out the surface tension measurements according to the pendant drop method, was

Table 1 List of 3-(alkoxymethyl)-1H-imidazol-3-ium salicylate, [H-im-C1OCn][Sal], ILs with n = 3–11, molar masses, and initial and final mass fraction water content based on coulometric Karl Fisher titration.

[TD$INLE]

IL

R

M 103[5_TD$IF] (mol kg1)

wH2O104 (initial)

wH2O104 (final)

C3H7 C4H9 C5H11 C6H13 C7H15 C8H17 C9H19 C10H21 C11H23

278.30 292.33 306.36 320.38 334.41 348.44 362.46 376.49 390.52

98.6 6.1 191.5 81.1 87.6 29.9 154.9 41.5 77.0

6.0 5.7 6.3 5.4 4.7 6.3 5.5 4.9 3.1

[H-im-C1OCn][Sal] [H-im-C1OC3][Sal] [H-im-C1OC4][Sal] [H-im-C1OC5][Sal] [H-im-C1OC6][Sal] [H-im-C1OC7][Sal] [H-im-C1OC8][Sal] [H-im-C1OC9][Sal] [H-im-C1OC10][Sal] [H-im-C1OC11][Sal]

42

J. Feder-Kubis et al. / Journal of Industrial and Engineering Chemistry 41 (2016) 40–49

measured using an instrument based on the vibrating-tube method (i.e., an Anton Paar DMA 5000 densimeter, Austria). The temperature range for the r measurements was T = (288.15–363.15) K, and the temperature step was 5 K. The uncertainty of the measurements was 0.02 kg m3, but in these conditions the uncertainly was only better than 0.1 kgm3. It was found that the ILs were not very sensitive to humidity (the maximum increase in water content after the density measurements was wH2O  1104). The surface tension g was measured using a DSA 100S Kru¨ss Tensiometer (with Drop Shape Analysis Software, GmbH, Germany) based on the pendant drop method. The necessary method and instrumental details as well as the experimental procedures have been described previously in the literature [26,27]. The temperature range for the g measurements was (288.2–333.2) K, with a step of 5 K. For each temperature point, several drops were created and equilibrated in an argon atmosphere. The general uncertainty of the method for surface tension was given by the manufacturer (i.e., 0.1 mN m1). The standard deviation of the mean value (based on 7–10 points) was lower than 0.04 mN m1. In order to determine the influence of moisture on the measured quantities, the surface tension of a few samples of [Him-C1OC8][Sal] and [H-im-C1OC6][Sal] with small concentrations of water (i.e., below wH2O = 0.12) was also studied. The contact angles u of the water and the [H-im-C1OCn][Sal] ILs were determined according to the principle of the sessile drop method using a Drop Shape Analyzer–DSA 100S (Kru¨ss GmbH, Germany). Each drop was placed on the surface using a thermostated chamber at (298  0.5) K. The contact angle as a function of time was recorded until a constant value was obtained. Therefore, the angles that were measured can be assumed to be static contact angles. The reported u values for each IL represent the average of at least 10 to several dozen independent measurements. The atmosphere inside the chamber was filled with argon, and the humidity during the measurements was roughly 40%. The resolution of the contact angle measurements was 0.18, but the standard deviations were in the range of 0.4–2.38. The experiments were conducted using PTFE because it is a standard hydrophobic, inert, non-polar (dispersive), low-energy compound; the literature results of u on this surface are typically reproducible [16,28]. Polypropylene possesses similar properties as PTFE. In fact, some literature reports have shown that this material is not purely dispersive [29]. However, due to the very low values of non-dispersive part compared with the dispersive [30], the former contribution can be neglected. In this work glass clear PP was used for the measurements, since it is a substance common with PP nonwoven fabric for which antielectrostatic properties have also been determined. The contact angle could not be measured directly on the same PP nonwoven fabric due to its good sorption abilities (the time of sorption of the IL drops did not exceed a few seconds). To the best of our knowledge, there have been no studies reported in the literature concerning the contact angles of ILs on PP; u for ILs on LDPE have been studied only in one paper [15]. Results and discussion Antielectrostatic properties The antielectrostatic (antistatic) effects of the salts that were obtained in the experiments can be estimated on the basis of the surface resistance, the half charge decay time and the maximum voltage induced on the disk surface of the polymers. As previously noted, all of the details concerning the experimental details and measurement conditions, as well as equations used for the calculations, have been previously described in details in Pernak et al. [21]. Based on the parameters provided above, the

Table 2 Criteria for the estimation of the antielectrostatic effect based on the surface resistance Rs and the half charge decay time t1/2. log (Rs/V )

t1/2 (s)

Antielectrostatic effect

<9 9–9.99 10–10.99 11–11.99 12–12.99 >13

<0.5 0.51–2 2.1–10 10.1–100 >100 >600

Excellent Very good Good Sufficient Insufficient Lack of antielectrostatic properties

antielectrostatic effect can be estimated using the criteria listed in Table 2 [21]. The measurements results that were obtained using two different materials (i.e., LDPE film and PP nonwoven fabric) are presented in Table 3. For the investigated protic ILs, they strongly depend on the alkyl chain length of the alkoxymethyl group in the imidazolium cation. This dependence is much stronger than that presented in the literature for other homologues series of ILs [24,31]. The salts with long alkyl chains, [H-im-C1OC10][Sal] and [H-im-C1OC11][Sal], possess excellent antielectrostatic properties. As the alkyl chain length reduces, the capacity to drain surface electric charge of the discussed salts decreases significantly. In this way, ILs with octyl and nonyl alkyl chains, ([H-im-C1OC8][Sal] and [H-im-C1OC9][Sal]), demonstrate only very good antielectrostatic effects. Next, the compounds with six and seven carbon atoms in their alkyl chain exhibit a good or sufficient capacity to drain surface electric charge. Finally, salicylates with shorter alkyl chains (from propyl to pentyl) possess no antielectrostatic properties. In addition, it has been noted that hygroscopic ILs can be particularly active at facilitating charge outflow [32,33]. This property is supposedly due to the fact that hygroscopic ILs can adsorb moisture from the atmosphere, thereby forming a monomolecular conductive layer on their surfaces [33]. Despite the 3-(alkoxymethyl)-1H-imidazol-3-ium salicylates studied here are not very hygroscopic, some of them (particularly those with longer alkoxymethyl substituent in their cation) demonstrate a strong tendency to drain surface charge. Due to their unique solvent capabilities, salicylate ILs are likely able to penetrate the polymer surface and diffuse the electric charge; similar observations have already been reported in the literature [24,31]. It should be noted that the activities of 3-(decyloxymethyl)-1Himidazol-3-ium salicylate ([H-Im-C1OC10][Sal]) and 3-(undecyloxymethyl)-1H-imidazol-3-ium salicylate ([H-Im-C1OC11][Sal]) described above are similar to those of a known antistatic agent, Catanac 609 [American Cyanamin Co., N,N-bis(2-hydroxyethyl)-N(30 -dodecyloxy-20 -hydroxypropyl) methylammonium methosulfate; log (Rs/V) = 8.48, t1/2 = 0.25 s and Uind = 100 V] [21]. As shown in Table 3, the antielectrostatic effect depends only slightly on the type of the material that was covered with protic ILs. In the case of PP nonwoven fabric, the surface resistance as well as the half-charge decay time and the voltage-induced values for active ILs (from [H-im-C1OC6][Sal] to ([H-im-C1OC11][Sal]) were slightly lower than those for LDPE. This result is correlated supposedly with the better adsorption of the studied compounds on the PP nonwoven fabric than on LDPE. For this reason, and due to the popularity of PP as materials used in office products, this compound was selected for additional investigations of the contact angles of salicylate ILs. Density and surface tension measurements The results of the density measurements in the temperature range of (288.15–363.15) K have been collected in Table S.2 and are presented in Fig. S.1 in the Supporting Information.

J. Feder-Kubis et al. / Journal of Industrial and Engineering Chemistry 41 (2016) 40–49

43

Table 3 Antielectrostatic effect of salicylate-based protic ILs on low-density polyethylene (LDPE) film and polypropylene (PP) nonwoven fabric compounds. IL

LDPE (film)

PP (nonwoven fabric)

[H-im-C1OCn][Sal]

log (Rs/V )

t1/2 (s)

Uind (V)

Antielectrostatic effect

log (Rs/V )

t1/2 (s)

Uind (V)

Antielectrostatic effect

None [H-im-C1OC3][Sal] [H-im-C1OC4][Sal] [H-im-C1OC5][Sal] [H-im-C1OC6][Sal] [H-im-C1OC7][Sal] [H-im-C1OC8][Sal] [H-im-C1OC9][Sal] [H-im-C1OC10][Sal] [H-im-C1OC11][Sal]

>15 >14 >14 >14 11.8 11.2 9.9 9.2 8.9 8.8

>600 >600 >600 >600 60 42 1.9 1.29 0.3 0.25

970 404 845 646 796 431 405 338 275 134

Lack Lack Lack Lack Sufficient Sufficient Very good Very good Excellent Excellent

>15 >14 >14 >13 10.5 10.1 9.8 9.1 8.8 8.5

>600 >600 >600 >600 3.5 2.1 1.85 1.0 0.35 0.2

880 466 465 200 262 274 353 272 241 97

Lack Lack Lack Lack Good Good Very good Very good Excellent Excellent

The surface tension data of [H-im-C1OCn][Sal] ILs measured within the temperature range of (288.2–333.2) K have been collected in Table S.3 in the Supporting Information and are presented in Fig. 1. In Fig. 2, the surface tension of [H-imC1OCn][Sal] is shown along with, for comparison, the surface tension of a few other series of the most popular ILs, alkanes and 1-alkanols (1-alkyl-3-methylimidazolium tetrafluoroborates, [CnC1im][BF4] [34]; 1-alkyl-3-methylimidazolium hexafluorophosphates, [CnC1im][PF6] [35,36]; 1-alkyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imides [CnC1im][NTf2] [37,38]; 1-alkanols [39]; linear alkanes [38]) versus the number of methylene groups in a chain at 298.2 K. The surface tension of the ILs that has been investigated [H-im-C1OCn][Sal] ranged between 30.24103 N m1 and 41.7103 N m1. These data are in the range of the most typical values for ILs [5]. The surface tension decreases significantly with temperature and with elongation of the alkyl chain of the alkoxymethyl substituent in the imidazolium cation. Its variation with the length of the chain (n) is somewhat similar than that observed for 1-alkyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imides; [CnC1im][NTf2] and the highest homologues of

[(Fig._1)TD$IG]

both series exhibit very similar g values (Fig. 2). The g for tetrafluoroborates and hexafluorophosphates also decreases with n, but the slopes of the g(n) dependencies are larger than for other IL groups. The decrease in surface tension with lengthening chain in the cation is typical of ILs; for 1-alkanols and alkanes g generally increases with n. Based on the temperature dependence of the surface tension, the surface entropy, Sa, and the surface enthalpy, Ha, can be obtained from equations [40]: Sa ¼ 



@g @T

Ha ¼ g T



(1)

p



@g @T



(2)

p

where all of the symbols have the same meaning as listed above. The thermodynamic functions of the surface are collected in Table S.4 in the Supporting Information; the Sa values for [H-ImC1OCn][Sal], other series of ILs, 1-alkanols and alkanes are presented in Fig. 3.

Fig. 1. Surface tension of [H-im-C1OCn][Sal] ILs (n = 3–11) at T = (288.2–333.2) K; open circles: n = 3; open squares: n = 4; open diamonds: n = 5; open triangles: n = 6; filled circles: n = 7; filled squares: n = 8; filled diamonds: n = 9; filled triangles: n = 10; stars: n = 11; lines–linear fit.

[(Fig._2)TD$IG]

44

J. Feder-Kubis et al. / Journal of Industrial and Engineering Chemistry 41 (2016) 40–49

Fig. 2. Surface tension at 298.2 K vs. the number of methylene groups in the alkyl chain of the alkoxymethyl substituent of investigated [H-Im-C1OCn][Sal] ILs and other rows of substances; filled circles: [H-im-C1OCn][Sal]; open squares: 1-alkyl-3-methylimidazolium tetrafluoroborates, [CnC1im][BF4] [34]; filled triangles: 1-alkyl-3methylimidazolium hexafluorophosphates, [CnC1im][PF6] [35,36]; open diamonds and filled diamonds: lower and higher homologues from 1-alkyl-3methylimidazolium bis(trifluoromethylsulfonyl)imides [CnC1im][NTf2] [37,38]; plus signs: 1-alkanols [39]; stars: linear alkanes [38].

The surface entropy and enthalpy of the investigated [H-ImC1OCn][Sal] ILs decrease regularly with the lengthening of the alkyl chain of the alkoxymethyl group. Based on Fig. 3, one can see that Sa decreases with n, which denotes an increase in the mutual orientation of longer alkyl chains of the alkoxymethyl substituent. This situation is supposedly similar to that which occurs in

[(Fig._3)TD$IG]

1-alkanols and alkanes. For other groups of ILs, the change in the surface entropy in the homologues series is sometimes not regular, as it is for [CnC1im][PF6] [36]. For some ILs, such as [CnC1im][BF4] and [CnC1im][NTf2], Sa increases with n. The critical temperature, Tc, of the ILs studied was estimated based on the temperature dependence of the surface tension and

Fig. 3. Surface entropy vs. the number of methylene groups in the alkyl chain of the alkoxymethyl substituent of investigated [H-Im-C1OCn][Sal] ILs (filled circles in the figure and in the insert) and other rows of substances; open squares: [CnC1im][BF4] [34]; filled triangles: [CnC1im][PF6] [35,36]; open and filled diamonds: [CnC1im][NTf2] [37,38]; plus signs: 1-alkanols [39]; stars: alkanes [38].

J. Feder-Kubis et al. / Journal of Industrial and Engineering Chemistry 41 (2016) 40–49 Table 4 Estimated critical temperatures, Tc (K), for [H-Im-C1OCn][Sal] using the Eo¨tvo¨s equation (Eq. 3). Ionic liquid

(Tc  d) (K)

[H-im-C1OC3][Sal] [H-im-C1OC4][Sal] [H-im-C1OC5][Sal] [H-im-C1OC6][Sal] [H-im-C1OC7][Sal] [H-im-C1OC8][Sal] [H-im-C1OC9][Sal] [H-im-C1OC10][Sal] [H-im-C1OC11][Sal]

900  4 1054  4 1107 3 1158  16 1141  6 1156  17 1147  7 1150  6 1148  4

n for salicylates resembles the Tc(n) dependence for alkanes and 1-alkanols. The most reasonable basis for the prediction of surface tension is a parachor concept [43–45]. The experimental parachor, Pexp, was calculated using density and surface tension data from Tables S.2 and S.3 according to the relation: P¼

the molar volume (molar surface energy, g  V2/3[10_TD$IF]9) using the Eo¨tvo¨s equation [41]:

g V 2=3 ¼ kðT c TÞ

45

(3)

where V is the molar volume of the liquid, Tc is the critical temperature and k is an empirical constant. Estimated values of Tc are listed in Table 4; k, which can be treated as a measure of the polarity of a substance [15], will be provided and discussed in the next part of this work (Table 6). The critical temperature is difficult to determine, even for protic ILs, due to their low volatility. Therefore, the estimation carried out using the Eo¨tvo¨s relation is a very convenient way of determining any knowledge about Tc. For the investigated [H-Im-C1OCn][Sal] ILs, the critical temperature generally was higher for higher homologues, which is inconsistent with observations of 1-alkyl-3methylimidazolium bis(trifluoromethylsulfonyl)[2_TD$IF]imides [37]. Similarly, the estimated Tc for 1-butyl-1-methylpyrrolidinium bis(trifluoromethylsulfonyl)[2_TD$IF]imide, [C4C1pyr][NTf2] (1095  22) K is lower than that of the first liquid homologue 1-propyl-1-methylpyrrolidinium bis(trifluoromethylsulfonyl)imide, [C3C1pyr][NTf2] (1328  5) K [42]. The increase in the critical temperature with [(Fig._4)TD$IG]

M

r

g 1=4

(4)

where M is the molar mass of the IL (Table 1). The theoretical values of the parachor, Pest, were calculated using its contributions for atoms and groups of atoms [45]. However, for the methylene group, for homologues [H-ImC1OCn][Sal] n = 4–11 the adjusted value PCH2 = 36.25 was used instead of the suggested value of 39.90 [[1_TD$IF]45]. The criterion for this correction was the minimum of the average value of the relative deviation between the experimental and estimated parachors for all [H-Im-C1OCn][Sal] homologues n = 4–11 at 298.2 K being equal to 0.005. For the first homologue [H-Im-C1OC3][Sal], such a modification was not necessary, and the original value of PCH2 = 39.90 was adopted [45] because of the acceptably low relative deviation between the experimental and estimated parachor obtained for this substance. The necessity of fitting the contribution from methylene group in order to predict the surface tension for ILs has already been postulated [46–50]. The value PCH2 = 36.25 used here is lower than that of 38.7 [50] applied for 1alkyl-3-methylimidazolium serine [Ser] ILs, but higher than that proposed for some series of 1-alkyl-3-methylimidazolium tetrafluoroborates [BF4], bis(trifluoromethylsulfonyl)imides [NTf2[12_TD$IF]], hexafluorophosphates [PF6], dicyanamides [DCA] and thiocyanates [SCN]: 32.8 [47]. The larger value (PCH2 = 44.0) was suggested for different cation cores of ILs with tricyanomethanide anion [C(CN)3] [46]. Pest versus Pexp for [H-Im-C1OCn][Sal] ILs is presented in Fig. S.2 in the Supporting Information. In this figure, one can also note the same dependences for [CnC1im][NTf2] ILs [34], for 1-alkanols [51] and alkanes (experimental values at

Fig. 4. Surface tension of mixtures of chosen [H-Im-C1OCn][Sal] with water at 298.2 K: open circles: [H-Im-C1OC8][Sal] + water; open squares: [H-Im-C1OC6][Sal] with water; lines–linear fits; the error bars correspond to the uncertainty of surface tension  0.1103 N m1.

46

J. Feder-Kubis et al. / Journal of Industrial and Engineering Chemistry 41 (2016) 40–49

293.15 K) [44]. In Table S.5 in the Supporting Information, the Pest and Pexp values are listed together with their relative deviations. Fig. S.2 (Supporting Information) reveals that the estimated and experimental parachors for [H-Im-C1OCn][Sal] are very close to the Pest(Pexp) line found for alkanes, despite P of the investigated salicylates being much higher than that of the 1 alkanols and alkanes. This finding confirms the applicability of the group contribution model for estimating parachors, even for such complex substances as the ILs being tested. Table S.5 (Supporting Information) also contains experimental and estimated surface tension data at 298.2 K for the series of [H-Im-C1OCn][Sal] ILs with their relative deviations. The largest relative deviations between the experimental and predicted surface tension were observed for homologues with n = 5 and 7, which supports, to some extent, the idea that odd-numbered homologues may sometimes fulfill different correlation rules than even-numbered homologues [46,52].

Based on Fig. 4, one may presume that the influence of a small amount of water influences g values of [H-Im-C1OCn][Sal] only slightly; however, none of the investigated ILs attained the maximum water content. Nevertheless, the results of the analysis of the temperature dependence of g, based on comparisons of the surface entropy calculated for four binary mixtures of [H-ImC1OC8][Sal], revealed that in this case the variation of Sa[3_TD$IF] with xH2O concentration was not meaningless (the lowest Sa value was (54.04  0.43)106 J m2 K1 for water concentration in IL: xH2O = 0.1169 and the highest one was (56.35  0.33)106 J m2 K1 K1 for a water concentration of IL: xH2O = 0.0517). Since the Sa values of some homologues of [H-Im-C1OCn][Sal] ILs are very similar (Table S.4 in the Supporting Information), the reported difference in Sa for two water concentrations is too large to be neglected. In summary, special care should be taken during surface tension measurements to avoid any contact between the IL sample and water. Contact angles

Influence of water on surface tension The influence of a small amount of water on surface tension has been already investigated by, for instance, Freire et al. [53] and Tong et al. [49]. These investigations did not provide an unambiguous answer to the question of how water in small amounts influences the surface tension of ILs. It seems that this effect depends also on the constitution of ions and the hydration processes. Freire et al. [53] postulated, however, that the water content largely does not influence the temperature dependence of the surface tension. According to this hypothesis, the surface entropy and surface enthalpy should not change significantly, when the concentration of water slightly increases. In this work, a similar investigation was conducted using four different solutions of [H-im-C1OC8][Sal] + water and two solutions of [H-imC1OC6][Sal] + water. The surface tension data of these mixtures at T = (288.2–333.2) K are provided in Table S.6 in the Supporting Information and the results at 298.2 K are presented in Fig. 4.

The contact angles of water and [H-Im-C1OCn][Sal] ILs, n = 3–10 on PTFE and PP were measured at 298 K using the sessile drop method; the data are collected in Table S.7 in the Supporting Information. The homologue [H-Im-C1OC11][Sal] had a tendency to turn solid on the investigated surfaces. Therefore, this work does not include contact angle measurements for this compound. The sinking time of [H-Im-C1OCn][Sal] in the PP nonwoven fabric did not exceed a few seconds, and it was accordingly not possible to measure the contact angle of salicylate ILs on this material. For this reason, our investigation makes uses of readily available clear glass PP. The u values that were measured for both materials are also shown in Fig. 5 together with the contact angles of the series of [CnC1im][NTf2] on PTFE at 298.2 K [17]. For the last homologous series, the dependence of u on the length of the alkyl chain in the imidazolium cation can be neglected. The scatter in the values of the contact angles of water and ILs on some surfaces provided in the literature is often beyond the

[(Fig._5)TD$IG]

Fig. 5. Contact angles of ILs and water on PTFE and PP at 298 K: open circles: [H-Im-C1OCn][Sal] on PTFE; [5_TD$IF]open squares: [H-Im-C1OCn][Sal] on PP; [6_TD$IF]filled circles: [CnC1im][NTf2] on PTFE at 298.2 K [17]; lines: water on PTFE or on PP.

J. Feder-Kubis et al. / Journal of Industrial and Engineering Chemistry 41 (2016) 40–49

declared uncertainty of the measurements [16,28]. The reason for this situation is that u depends very strongly on a variety of factors including solid samples, ILs and the experimental details; therefore, the decision was made to limit the preparation of surfaces for measurements and use the same conditions for the water as well as for all of the samples. The u measurements for water on PTFE obtained in this work ((118.2  1.2)8 at 298 K) fell between the literature values that amounted to 109.58 at room temperature [54], 1118 at 293 K [55], 1188 at 298 K [15], and 1228 at room temperature [56] and at 298 K [57]. Furthermore, u for water on PP (97.0  0.6)8 fell between some literature results: 87.58 at room temperature [30], 898 at room temperature [58], and 988 [59] at an unknown temperature. In summary, these values are rather near the upper limit of the range of the literature values. Typically, even a small amount of roughness of the solid increases the contact angle [60]. Based on an inspection of Fig. 5, it is evident that the wetting behavior of the investigated salicylates on two materials (PTFE and PP) was better than that of water and increased with the length of the alkyl chain of the alkoxymethyl group in the cation. However, the quoted [CnC1im][NTf2] ILs [17] have lower contact angles on PTFE (and better wetting behavior) than [H-Im-C1OCn][Sal], which may result from the different IL anion and the alkoxymethyl group in the cation instead of the alkyl chain. According to Pereira et al. [17], larger contact angles on PTFE are characteristic for ILs with non-fluorinated anions. Generally, the wetting behavior of ILs depends on the basicity of the anion hydrogen bonds; the higher the relative basicity of the anion, the larger the contact angle of such an IL on a non-polar surface [17]. It is interesting to note, however, the much smaller contact angles of salicylate on PP compared with PTFE. In order to determine an explanation for this observation, some additional calculations were carried out. Independent contributions to the surface tension (or, more generally, the interfacial tension) arising from intermolecular interactions [61] are dispersive, g di and non-dispersive (polar), g nd i :

g i ¼ g di þ g nd i

(5)

where i refers to liquid, L, or solid, S. Since both solid materials are defined as low-energy solids (g nd i ffi 0), the knowledge of contact angles of [H-Im-C1OCn][Sal] ILs on them makes it possible to divide their surface tension into two parts, i.e., a non-dispersive (polar) and dispersive (non-polar). Finally, the polarity fraction of liquid, gnd[13_TD$IF]/g, can also be calculated. When the geometric mean rule to describe the dispersive forces between pairs of different molecules is applied, the following equation is obtained [29,62]: 1=2

g L ð1 þ cos uÞ ¼ 2ðg dS g dL Þ

(6)

This formula, derived i.e. in [62], is a start point for calculations of the dispersive and non-dispersive part of the surface tension of a liquid from its contact angle on a purely dispersive solid, when g dS Table 5 Dispersive, gd, and the non-dispersive, gnd, components of the surface tension of the liquids, at 298 K, calculated using Eqs. 5 and 6 and the contact angles measured on two non-polar surfaces: PTFE and PP. Substance [H-im-C1OC3][Sal] [H-im-C1OC4][Sal] [H-im-C1OC5][Sal] [H-im-C1OC6][Sal] [H-im-C1OC7][Sal] [H-im-C1OC8][Sal] [H-im-C1OC9][Sal] [H-im-C1OC10][Sal]

gd[7_TD$IF] PTFE103

gnd PTFE103

gd PP103

gnd PP103

(N m1)

(N m1)

(N m1)

(N m1)

18.0 17.1 16.7 17.9 18.4 19.2 20.7 22.2

22.9 20.4 19.1 17.4 16.2 14.3 12.1 10.5

17.6 17.2 16.7 17.9 17.7 17.0 16.3 15.7

23.3 20.3 19.2 17.3 16.8 16.5 16.5 17.0

47

Table 6 The parameter k from the Eo¨tvo¨s equation (Eq. 3) and the polarity fraction, gnd/g at 298 K. Substance

k107 (mol2/3 J K1)

gnd/g

[1_TD$IF][H-im-C1OC3][Sal] [H-im-C1OC4][Sal] [H-im-C1OC5][Sal] [H-im-C1OC6][Sal] [H-im-C1OC7][Sal] [H-im-C1OC8][Sal] [H-im-C1OC9][Sal] [H-im-C1OC10][Sal]

2.60 1.99 1.85 1.78 1.84 1.83 1.87 1.92

0.56 0.54 0.53 0.49 0.48 0.46 0.44 0.42

is known (Eq. (5)). This can be extracted indirectly from Eq. 6 using the dispersive part of water (as a standard) on the solid surfaces from the literature 21.8103 N m1 [54], and the measured surface tension of pure water 71.74103 N m1 at 298.2 K (this work). The g dS that was obtained for PTFE was 16.42103 N m1 at 298 K, which was slightly lower than the value reported in the literature: 17.5103 N m1 at 298 K [54]. The gd and gnd values of the investigated [H-Im-C1OCn][Sal] ILs on two solids calculated from Eqs. 5 and 6 are collected in Table 5; in Table 6 the k parameter from the Eo¨tvo¨s equation (Eq. (3)) and the average values for the polarity fraction, gnd/g on solid materials, are listed. An analysis of the values listed in Table 5 reveals that the nondispersive parts of the surface tension of the lower homologues of [H-Im-C1OCn][Sal] are larger than the dispersive parts of the surface tension. This trend is the opposite division to those found in the literature for typical aprotic ILs [15], where both contributions to g are comparable or the dispersive one prevails. For higher [HIm-C1OCn][Sal], the situation changes since gd of the investigated liquids on PTFE increases significantly and gd on PP changes only slightly. In this way, for higher homologues [H-im-C1OCn][Sal] on PTFE gd  2gnd; on PP gd  gnd. The mutual balance between gd and gnd of salicylates, which differs depending on the material, may be the reason for the difference in their wettability. It seems that this behavior is a special feature of the investigated protic salicylates and their interactions with the solid surface. Excluding [H-im-C1OC3][Sal], all of k values were larger than that of water (1.4124107 mol2/3 J K1), but lower than that of non-polar substance such as hexane (2.2038107 mol2/3 J K1) [41]. The k data collected in Table 6 allow one to presume that the investigated ILs reveal a moderate polarity (the most polar would be the [H-im-C1OC6][Sal], [H-im-C1OC7][Sal] and [H-imC1OC8][Sal] ILs). On the other hand, the polarity fractions, which are defined as the ratio between the non-dispersive component of the surface tension and the surface tension of a liquid gnd/g, fall between 0.56 and 0.42, For water and glycerol, these ratios are 0.7 and 0.47, respectively [15]. Therefore, despite k changing nonmonotonically with the length of the alkyl chain of alkoxymethyl group in the imidazolium cation of [H-im-C1OCn][Sal] as the gnd/g values do, the average values of these two factors reveal that the investigated series of ILs can be classified as substances of medium polarity that are closer to water than the ILs investigated by Restolho et al. [15]. This fact from above can be an additional argument for an explanation of the decrease in contact angles of [H-im-C1OCn][Sal] ILs with n on PTFE and PP. Good wettability of PP by [H-im-C1OCn][Sal] ILs is rather desirable when they are used to modify the antistatic properties of this solid compound. Conclusions In this work, the density, surface tension and contact angle on two non-polar surfaces (PTFE and PP) of the homologous series of [H-Im-C1OCn][Sal] ILs were measured. There is also a discussion of

48

J. Feder-Kubis et al. / Journal of Industrial and Engineering Chemistry 41 (2016) 40–49

these properties as the important implementation of the antistatic properties of tested ILs toward PP and LDPE. The salicylates that were investigated revealed a variable antistatic effect on the solid compounds noted above. It changed from the least sufficient for [H-Im-C1OC6][Sal] and [H-ImC1OC7][Sal] to excellent for [H-Im-C1OC10][Sal] and [H-ImC1OC11][Sal]. For lower [H-Im-C1OCn][Sal] homologues, the values of surface resistance and half charge decay time that were observed did not indicate antistatic behavior. This finding implies a significant increase in the ability of ILs to drain surface charge with lengthening of the alkyl chain of the alkoxymethyl substituent in the imidazolium cation of the investigated salicylates. The surface tension of [H-Im-C1OCn][Sal] ILs decreased regularly with the length of the alkyl chain of the alkoxymethyl substituent in the imidazolium cation, and the surface entropy calculated from the temperature dependence of the surface tension changed in the same way. The decrease in the surface entropy (and increase in the surface order) of salicylates, such as for 1-alkanols and normal alkanes, is not common for typical ILs in which the alkyl or alkoxymethyl chains are attached to a large imidazolium cation [35,36]. The prediction of surface tension using the parachor concept by Sugden [43] and other groups’ contributions based on data from organic compounds have also been successful; however, some corrections concerning the contribution of the methylene group must be taken into account [47]. The water content (up to 7000 ppm) does not affect the surface tension significantly; however, special care must be taken if the variation in the surface entropy in a homologous series of ILs is not large. The contact angles [H-Im-C1OCn][Sal] of ILs determined on two non-polar, low-energy surfaces yielded different results. It was found that the wettability in PP according to the investigated ILs was much higher than in PTFE and increased in both cases for higher homologous. The smaller contact angles on PP may have been caused by the nature of investigated protic salicylate ILs that were more polar than the aprotic ILs presented in the literature [15]. It is also possible that some unspecified additions could have been present in the commercially available PP materials. The decrease in the surface tension and contact angle on some solid dispersive materials with the lengthening of the alkyl chain of alkoxymethyl group in the imidazolium cation of the IL is not a revealing result in and of itself. g behaves similarly for many homologues series of ILs [5]. Since the increase in the size of the cation results in dispersion of the ion charge, a reduction in the hydrogen bond energy between ions occurs and the surface tension decreases [37]. In terms of the contact angle, since the affinity of less-polar salicylates (such as higher homologues) for non-polar surface such as PTFE and PP is higher, the wettability of these surfaces by IL with a longer alkoxymethyl substituent is also better. Much more interesting is, however, the comparison of the order of these two effects between salicylate ILs and bis(trifluoromethylsulfonyl)imide ILs. In the case of the surface tension of [CnC1im][NTf2] ILs, from n = 2 to n = 10, Dg = 4.82103 N m1 (293.15 K); for [H-Im-C1OCn][Sal], from n = 3 to n = 11, Dg = 8.92103[14_TD$IF] N m1 (293.15 K). The contact angles of the [CnC1im][NTf2] ILs on PTFE changed somewhat [15_TD$IF]slightly, but generally [16_TD$IF]decrease with n; the [17_TD$IF]highest value for the n = 2 homologue ([18_TD$IF]66.938) was noted and the [19_TD$IF]lowest value for the n = [20_TD$IF]9 homologue ([21_TD$IF]64.148) (298.2 K) was noted. Therefore, Du = [2_TD$IF]2.798 [17]. In the case of the salicylates that were investigated (i.e., those from n = 3 to n = 10), Du = 18.78 (298.2 K); thus, the contact angles[23_TD$IF] also decrease with n. In addition, the antistatic properties of salicylates that were investigated also varied significantly. This behavior is recalled here again since the antistatic activity in the homologues series reported in the literature is often constant; this property seems to be rather anion-dependent [24,31]. In this way,

one may assume that a significant decrease in the surface tension and contact angle on PP in the homologues series of [H-ImC1OCn][Sal] ILs corresponds to a significant improvement in the antistatic properties of these substances toward PP. In summary, this work was demonstrated a connection between improvements in the antistatic properties of materials in the presence of salicylate and surface properties of these ILs. All of the results confirm that the investigated ILs can be used for PP coatings in order to improve the antistatic properties of PP. Acknowledgements The work was financed by the Narodowe Centrum Nauki grant no. 2013/09/D/ST5/03904. One of us (M. Geppert-Rybczyn´ska) would like to express also her gratitude to Dorota Gawliczek for her assistance during the surface tension measurements.

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.jiec.2016.07.003. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19]

[20] [21] [22] [23] [24] [25]

[26] [27] [28] [29] [30] [31] [32] [33] [34]

J.S. Wilkes, Green Chem. 4 (2002) 73. R. Giernoth, Angew. Chem. Int. Ed. 49 (2010) 2834. S. Tang, A. Gary, G.S. Baker, H. Zhao, Chem. Soc. Rev. 41 (2012) 4030. S. Zhang, Q. Zhang, Y. Zhang, Z. Chen, M. Watanabe, Y. Deng, Prog. Mater. Sci. 77 (2016) 80. M. Tariq, M.G. Freire, B. Saramago, J.A.P. Coutinho, J.N. Canongia Lopes, L.P.N. Rebelo, Chem. Soc. Rev. 41 (2012) 829. Ch. Wertz, A. Tschersich, J.K. Lehmann, A. Heintz, J. Mol. Liq. 131–132 (2007) 2. V.R. Vale, B. Rathke, S. Will, W. Schro¨er, J. Chem. Eng. Data 55 (2010) 4195. R.L. Gardas, R. Ge, N. Ab Manan, D.W. Rooney, Ch. Hardacre, Fluid Phase Equilibr. 49 (2010) 139. M. Geppert-Rybczyn´ska, A. Knorr, J.K. Lehmann, A. Heintz, J. Chem. Eng. Data 57 (2012) 1923. A. Yang, W. Pan, Enzyme Microb. Technol. 37 (2005) 19. G. Quijano, A. Couvert, A. Amrane, Bioresour. Technol. 101 (2010) 8923. S. Corderı´, E.J. Gonza´lez, N. Calvar, A´. Domı´nguez, J. Chem. Thermodyn. 53 (2012) 60. M. Larriba, P. Navarro, J. Garcı´a, F. Rodrı´guez, J. Chem. Thermodyn. 79 (2014) 266. P.F. Requejo, E. Go´mez, N. Calvar, [24_TD$IF]A´. Domı´nguez, Ind. Eng. Chem. Res. 54 (2015) 1342. J. Restolho, J.L. Mata, B. Saramago, J. Colloid Interface Sci. 340 (2009) 82. I. Delcheva, J. Ralston, D.A. Beattie, M. Krasowska, Adv. Colloid Interface 222 (2015) 162. M.M. Pereira, K.A. Kurnia, F.L. Sousa, N.J.O. Silva, J.A. Lopes-da-Silva, J.A.P. Coutinho, M.G. Freire, Phys. Chem. Chem. Phys. 17 (2015) 31653. J. Pernak, I. Goc, Polish J. Chem. 77 (2003) 975. J. Jacquemin, J. Feder-Kubis, M. Zore˛bski, K. Grzybowska, M. Chora˛z˙ewski, S. Hensel-Bielo´wka, E. Zore˛bski, M. Paluch, M. Dzida, Phys. Chem. Chem. Phys. 16 (2014) 3549. J. Pernak, I. Goc, A. Fojutowski, Holzforschung 59 (2005) 473. J. Pernak, A. Czepukowicz, R. Poz´niak, Ind. Eng. Chem. Res. 40 (2001) 2379. J. Pernak, K. Sobaszkiewicz, J. Foksowicz-Flaczyk, Chem. Eur. J. 10 (2004) 3479. A. Cieniecka-Rosłonkiewicz, J. Pernak, J. Feder-Kubis, A. Ramani, A. Robertson, K. Seddon, Green Chem. 7 (2005) 855. J. Feder-Kubis, M. Kubicki, J. Pernak, Tetrahedron-Asymmetry 21 (2010) 2709. J. Skowronek, M. Geppert-Rybczyn´ska, J. Jacquemin, P. Goodrich, J. Alvarez Vicente, M. Chora˛z˙ewski, S. Je˛z˙ak, E. Zore˛bski, M. Zore˛bski, M. Z˙arska, W. Kaca, P. Berdyczko, M. Dzida, J. Solut. Chem. 44 (2015) 824. A. Wandschneider, J.K. Lehmann, A. Heintz, J. Chem. Eng. Data 53 (2008) 596. M. Geppert-Rybczyn´ska, J.K. Lehmann, A. Heintz, J. Chem. Eng. Data 56 (2011) 1443. R. Sedev, Curr. Opin. Colloid Interface Sci. 16 (2011) 310. R.N. Shimizu, N.R. Demarquette, J. Appl. Polym. Sci. 76 (2000) 1831. E. Occhiello, M. Morra, G. Morini, F. Garbassi, D. Johnson, J. Appl. Polym. Sci. 42 (1991) 2045. M. Pie˛tka-Ottlik, R. Fra˛ckowiak, I. Maliszewska, B. Kołwzan, K.A. Wilk, Chemosphere 89 (2012) 1103. J. Chlebicki, J. We˛grzyn´ska, K.A. Wilk, J. Coll. Int. Sci. 323 (2008) 372. J. Pernak, J. Feder-Kubis, A. Cieniecka-Rosłonkiewicz, C. Fischmeister, S.T. Griffin, R.D. Rogers, N. J. Chem. 31 (2007) 879. W.-G. Xu, L. Li, X.-X. Ma, J. Wei, W.-B. Duan, W. Guan, J.-Z. Yang, J. Chem. Eng. Data 57 (2012) 2177.

J. Feder-Kubis et al. / Journal of Industrial and Engineering Chemistry 41 (2016) 40–49 [35] J. Klomfar, M. Soucˇkova´, J. Pa´tek, J. Chem. Eng. Data 54 (2009) 1389. [36] J. Klomfar, M. Soucˇkova´, J. Pa´tek, Fluid Phase Equilibr. 385 (2015) 62. [37] P.J. Carvalho, M.G. Freire, M.I. Marrucho, A.J. Queimada, J.A.P. Coutinho, J. Chem. Eng. Data 53 (2008) 1346. [38] H.F.D. Almeida, M.G. Freire, A.M. Fernandes, J.A. Lopes-da-Silva, P. Morgado, K. Shimizu, E.J.M. Filipe, J.N. Canongia Lopes, L.M.N.B.F. Santos, J.A.P. Coutinho, Langmuir 30 (2014) 6408. [39] Ch. Wohlfarth, B. Wohlfarth, Surface Tension of Pure Liquids and Binary Liquid Mixtures, Landolt-Bo¨rnstein, Numerical Data and Functional Relationships in Science and Technology, Group IV: Physical Chemistry, vol. 16, Springer, 1997. [40] A.W. Adamson, A.P. Gast, Physical Chemistry of Surfaces, 6th ed., A WileyInterscience Publication, John Wiley and Sons. Inc., New York, Chichester, Weinheim, Brisbane, Singapore, Toronto, 1997. [41] I. Pa´szli, K. La´szlo´, Colloid Polym. Sci. 285 (2007) 1505. [42] P.J. Carvalho, C.M.S.S. Neves, J.A.P. Coutinho, J. Chem. Eng. Data 55 (2010) 3807. [43] S. Sugden, J. Chem. Soc. Trans. 125 (1924) 32. [44] O.R. Quayle, Chem. Rev. 53 (1953) 439. [45] T.A. Knotts, W.V. Wilding, J.L. Oscarson, R.L. Rowley, J. Chem. Eng. Data [4_TD$IF]46 (2001) 1007. [46] M. Soucˇkova´, J. Klomfar, J. Pa´tek, Fluid Phase Equilibr. 406 (2015) 181. [47] M. Soucˇkova´, J. Klomfar, J. Pa´tek, J. Chem. Thermodyn. 83 (2015) 52.

49

[48] J. Tong, Q.-S. Liu, W.-G. Xu, D.-W. Fang, J.-Z. Yang, J. Phys. Chem. B 112 (2008) 4381. [49] J. Tong, M. Hong, Ch. Liu, A. Sun, W. Guan, J.-Z. Yang, Ind. Eng. Chem. Res. 52 (2013) 4967. [50] J. Wei, Ch. Chang, Y. Zhang, S. Hou, D. Fang, W. Guan, J. Chem. Thermodyn. 90 (2015) 310. [51] S.A. Mumford, J.W.C. Phillips, J. Chem. Soc. (1950) 75. [52] G. Adamova´, J.N. Canongia Lopes, L.P.N. Rebelo, L.M.N.B. Santos, K.R. Seddon, K. Shimizu, Phys. Chem. Chem. Phys. 16 (2014) 4033. [53] M.G. Freire, P.J. Carvalho, A.M. Fernandes, I.M. Marrucho, A.J. Queimada, J.A.P. Coutinho, J. Colloid Interface Sci. 314 (2007) 621. [54] F. Liu, W. Shen, Colloid Surf. A 316 (2008) 62. [55] B. Jan´czuk, T. Białopiotrowicz, W. Wo´jcik, J. Colloid Interface Sci. 127 (1989) 59. [56] L.S. Penn, B.A. Miller, J. Colloid Interface Sci. 78 (1980) 238. [57] J.R. Dann, J. Colloid Interface Sci. 32 (1970) 302. [58] C. Bonnerup, P. Gatenholm, J. Adhes. Sci. Technol. 7 (1993) 247. [59] N. Inagaki, S. Tasaka, Y. Horikawa, Polym. Bull. 26 (1991) 283. [60] W.A. Zisman, [25_TD$IF]Adv. [26_TD$IF]Chem. [27_TD$IF]Ser. 43 (1964) 1. [61] F.M. Fowkes, Ind. Eng. Chem. 56 (1964) 40. [62] N. Correia, J.J. Moura Ramos, B.J.V. Saramago, J.C.G. Calado, J. Colloid Interface Sci. 189 (1997) 361.