- Email: [email protected]
Evaporation from an ionic liquid emulsion
Journal of Colloid and Interface Science 307 (2007) 494–499 www.elsevier.com/locate/jcis
Evaporation from an ionic liquid emulsion Stig E. Friberg ∗ Chemistry Department, University of Virginia, 1695 Goldentree Place, Charlottesville, VA, USA Received 7 September 2006; accepted 7 December 2006 Available online 12 December 2006
Abstract The conditions during evaporation in a liquid crystal-in-ionic liquid microemulsion (LC/µEm) were estimated using the phase diagram of the system. The equations for selected tie lines were established and the coordinates calculated for the sites, at which the evaporation lines crossed the tie lines. These values combined with the coordinates for the phases connecting the tie lines were used to calculate the amounts and the composition of the fractions of the two phases present in the emulsion during the evaporation. One of the emulsion phases was a lamellar liquid crystal and high energy emulsification would lead to the liquid crystal being disrupted to form vesicles. Such a system tenders a unique opportunity to study the interaction between vesicles and normal micelles, which gradually change to inverse micelles over bi-continuous structures. The amount of vesicles in the liquid phase versus the fraction liquid crystal was calculated for two extreme cases of vesicle core size and shell thickness. The limit of evaporation while retaining the vesicle structure was calculated for emulsions of different original compositions assuming the minimum continuous liquid phase to be 50% of the emulsion. © 2007 Elsevier Inc. All rights reserved. Keywords: Emulsions; Microemulsions; Vesicles; Evaporation; Detergents; Amphiphiles
1. Introduction Ionic liquids (IL) [1,2] offer significant advantages as solvents because of their negligible vapor pressure and useful molecular interactions. These properties have enabled their use as reaction medium; even as catalysts [3], for a number of organic reactions such as polymerization, Diels–Alder reactions, hydrogenation and Friedel Crafts reactions [4–8]. Other applications involve electrochemical reactions [9] and as biopolymers [10]. The structure of the individual molecules forming the ionic liquid makes it possible to change its properties [11] from liquids of high polarity to entities with properties similar to those of hydrocarbons. They have, hence, been engaged in colloidal systems acting both as hydrocarbons [12,13] and as liquids of sufficiently high polarity to serve as the polar continuous part of a microemulsion [14–18]. The structures in the latter liquids have recently been systematically investigated by Zheng and his group [19] using a combination of phase diagram determi* Fax: +1 434 973 8826.
E-mail address: [email protected]. 0021-9797/$ – see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.jcis.2006.12.019
nation, dynamic light scattering and molecular spectroscopy to elucidate the structures involved in a non-aqueous microemulsion of this kind. Although the less polar ionic liquids have many properties in common with hydrocarbons, they are markedly different in having an extremely low vapor pressure. In fact, this feature opens the potential for exact information about the behavior of their colloidal systems during evaporation. This fact in combination with the results of recent investigations of complex emulsion systems [20,21] encouraged an investigation into the conditions during evaporation of a nontraditional emulsion containing an ionic liquid. The chosen emulsion contains two phases; one of which is a ionic liquid-inwater microemulsion stabilized by a nonionic surfactant and the second phase a lamellar liquid crystal formed by the same surfactant and water with solubilized ionic liquid. During the evaporation the ionic liquid-in-water microemulsion is gradually transformed to a water-in-ionic liquid one over bi-continuous structures. This unusual system was selected, because it offered a unique opportunity to establish the basic conditions for an arrangement in which the evaporation would give rise to an
S.E. Friberg / Journal of Colloid and Interface Science 307 (2007) 494–499
emulsion, in which microemulsion droplets would be combined with vesicles in an ionic liquid matrix of incessantly reduced volume containing normal micelles, whose structure gradually is reversed to become inverse micelles. In such a medium the phase diagram approach would provide unique information about the potential limits to retain the structure of co-existing vesicles and micelles; a matter of interest beyond the narrow area of ionic liquids. The ionic liquid, 1-butyl-3-methylimidazolium tetrafluoroborate, [C4 -mim][BF4 ] was selected, because its microemulsion with water and a non-ionic surfactant varies from O/W to W/O within one solubility region; a phenomenon investigated by Zheng’s group [19]. 2. The method The goal of the investigation is to establish simple expressions to illustrate the changes in the phase fractions and structures encountered during evaporation of emulsions of ionic liquids. The analysis is built on the phase diagram of the system [12] utilizing an algebraic system to extract information from phase diagram [20,21]. The evaporation line in a system of water, oil and a surfactant, in which only water evaporates, follows a general equation S = S0 (W − 1)/(W0 − 1).
(1)
In Eq. (1), and throughout this article, capitals mean weight fraction of the components and index 0 denotes the initial composition of an emulsion. It should be observed that the equation correctly denotes a line emanating from the water corner, because water is the sole component evaporating. The equation for the line between the compositions of two phases in equilibrium (WL , SL , ILL ) and (WLC , SLC , ILLC ) is S = SL + (SLC − SL )(W − WL )/(WLC − WL ).
(2)
Following the recent analysis [20,21] the amount of the two phases is calculated using one coordinate only. In the present calculations the water fraction is chosen, because the variation with water content is an obvious parameter for the evaporation. This water fraction, WTL Em , of the total emulsion composition for each tie line (TL) during the evaporation is obtained by a combination of Eqs. (1) and (2): WTL Em = SL + S0 /(W0 − 1) − WL (SLC − SL ) (WLC − WL ) S0 (W0 − 1) − (SLC − SL )/(WLC − WL ) . (3)
495
of the latter is postulated not to change, when it forms the shell of the vesicle. For an emulsion, in which one of the phases is a lamellar liquid crystal, the emulsification method is of paramount importance for the structure of the emulsion. With gentle emulsification, such as manual shaking, the liquid crystal is dispersed as particles, prominently displayed as shiny particles when viewed between crossed polarizers in an optical microscope. With high energy emulsification the liquid crystal particles are dispersed as vesicles. The vesicles consist of a central core of the continuous liquid with a radius r, surrounded by a shell, with a thickness d, of bilayers or multilayers of the liquid crystalline structure. The vesicles are in fact the dispersed drops in a W/W emulsion, stabilized by an even number of multiple layers. The presence of these structures exerts a pronounced demand on the continuous phase [21] and it is instructive to analyze the effect the presence of such structures would have in an emulsion. In the following analysis the vesicles are assumed to have a solvent core radius of r and a shell thickness of d giving a total volume of 4π(r + d)3 /3. The volume fraction of the shell is [(r + d)3 − r 3 ]/(r + d)3 , which in this contribution is assumed to be identical to the volume of the liquid crystal used to form the shell. Since the ratio between volumes is independent of the dimensions per se of the core and the shell, the shell volume fraction may be written as Frshell = (1 + R)3 − 1 /(1 + R)3 , (5) in which R = d/r. 3. The emulsion The emulsion system is shown in Fig. 1. It consists of a lamellar liquid crystal, marked LC in Fig. 1, and a microemulsion, marked L, that reaches from the water corner to homogeneous surfactant in a continuous region. Hence, the liquid part of the emulsion changes from a water continuous liquid to a surfactant continuous one with no phase change [19]. The evaporation line that is given in the figure emanates from the water corner, because the ionic liquid does not evaporate to a degree influencing the direction of the line. The tie lines between the two phases are indicated on the diagram and the compositions in equilibrium are given in Table 1.
Index LC refers to fraction of the liquid crystalline phase [20,21] while index L indicates the fraction isotropic liquid microemulsion phase. LC = (WTLEL − WLC )/(WL − WLC ) = S0 (WL − 1) + SL (1 − W0 ) /S0 (WL − WLC ) + (SLC − SL )(W0 − 1).
(4)
In the following evaluations the density of the liquid phase is assumed equal to that of the liquid crystal phase and the density
Fig. 1. The emulsion system. The liquid part of the emulsion, L, is a microemulsion and the liquid crystal, LC, has a lamellar structure. The coordinates for the tie lines are given in Table 1.
496
S.E. Friberg / Journal of Colloid and Interface Science 307 (2007) 494–499
Table 1 Fractions in the phase compositions connected by tie lines Liquid phase
Liquid crystal phase
Water
Surfactant
Water
Surfactant
1.00 0.92 0.71 0.40 0.155 0.115
0 0.02 0.145 0.37 0.555 0.755
0.49 0.48 0.425 0.36 0.22 0.23
0.51 0.505 0.515 0.545 0.685 0.72
Table 2 Coefficients for the equations for tie lines Tie line #
A
B
0 1 2 3 4 5
1 1.0341 1.067 2.120 0.245 0.790
−1 −1.102 −1.298 −4.375 2.000 −0.304
The curves are a good illustration of the advantage of extracting information from a phase diagram using the algebraic method [20,21]. At first it provides quantitative evidence of the trend of the initial increase of the fraction liquid crystal during evaporation and its subsequent decline to zero. Secondly it offers an easy access to the change with the water content because the initial and final parts of the curves in Fig. 2 are linear in water fraction of the total emulsion. It would be attractive to use these features to estimate the times for the increase and reduction of the liquid crystal fractions, but such an analysis is prohibited for several reasons. At first the vapor pressure varies with the composition, although not to the degree of molecular solutions [22]. Secondly, the presence of the liquid crystal has a significant influence on the kinetics of the evaporation [23]. Thirdly, an additional advantage of the algebraic method is the fact that the presentation of results is easily changed from one variable to another. As an example the results in Fig. 2 are directly transposed to substitute dependent variables. So is the water fraction of the total emulsion changed to fraction evaporated water by the following relation: WEv = (WOr − W )/WOr (1 − W ),
Fig. 2. Weight fraction liquid crystal during evaporation for emulsions containing 5% surfactant. Original weight fraction water and ionic liquid: (Q) 0.9500 and 0, (E) 0.9412 and 0.0088, (") 0.9375 and 0.0125, (P) 0.9286 and 0.0214, (2) 0.8750 and 0.0750, (!) 0.7500 and 0.2000.
Applying Eq. (2) to the system the expressions are established for the tie line equations counted from left to right in Fig. 1. Replacing SL + (SLC − SL )/(WLC − WL ) in Eq. (2) with A and (SLC − SL )/(WLC − WL ) with B, the equation is written as S = A + BW.
(6)
The coefficients A and B for the different tie lines from left to right in Fig. 1 are given in Table 2. The values of the water fraction for the junction of the evaporation lines with the tie lines are immediately obtained from Eq. (3) and direct substitution into Eq. (4) provides the values of the weight fraction of the liquid crystal for the different evaporation lines as shown in Fig. 2. The results illustrate the original increase of the fraction liquid crystal and the sudden reduction of these values at the later part of the evaporation process.
(7)
in which W is the water fraction of the total emulsion at the evaporation step, WOr the water faction in the emulsion prior to evaporation and WEv the evaporated fraction of WOr . The curve is shown in Fig. 3 in addition to the liquid crystal fraction versus the surfactant fraction and the ionic liquid fraction. As mentioned earlier, the lamellar liquid crystal will be changed to vesicles during high energy emulsification and an analysis of the importance of such a change for the present system is appropriate. During evaporation the liquid fraction, forming the continuous phase, is reduced and the liquid reaches a packing limit for stability of the vesicle structure, as was early introduced by Kaler et al. [24]. The structural changes have been summarized by Kaler in a recent publication [25]. It is evident that a large number of structures may appear during the “compacting” process; the available thermodynamic analysis [27] may be summarized as [25,26] the large vesicles disappearing causing a more mono-disperse distribution during reduced fraction of the solvent. In the present evaluation the picture is further complicated by the presence of microemulsion droplets in the continuous phase. At first these will interact with the vesicles and a modification of both structures is a reasonable assumption and, in addition, the evaporation causes a change of the micellar structure per se. In this, the first article about these problems, the matter of the micellar structure and its potential influence on the vesicle structure will be neglected in order to limit the magnitude of the article and the following discussion is restricted to the vesicles. Estimating the limit of evaporation to retain the vesicle stability is arbitrary, but a maximum volume fraction of vesicles at 0.5 is realistic and will be used in the following discussion. With this stipulation the condition for maximum liquid crystal fraction to retain the vesicles is written LC = 0.5 1 − 1/(1 + R)3 . (8)
S.E. Friberg / Journal of Colloid and Interface Science 307 (2007) 494–499
(a)
497
Fig. 4. The maximum fraction of liquid crystal in emulsions, in which the liquid crystal has formed vesicles with retained stability. The initial surfactant fraction was 0.01 and the ionic liquid fraction adjusted to give the required value of R.
Fig. 4 reveals the extremely small weight fraction allowed for small values R in accordance with Kaler [25]. For such values of R Eq. (8) arrives at Vmax ≈ 1.5R showing a linear behavior. For large values, lim VR⇒∞ = 0.5.
(b)
In real systems the lower limit of R is determined by the fact that a bimolecular layer is physically the thinnest membrane possible. Hence the dimensions of the shells are limited to greater than approximately 4 nm, with values of 8 and 12 nm as the subsequent dimensions. There is, of course, in principle, no upper limitation as to the thickness of the vesicle shells, but with such an intense emulsification as to form vesicles from the liquid crystal particles, an assumption of thicker shells than the mentioned largest value is not realistic [25]. The maximum fraction of liquid crystal to retain the stability of the vesicles may now be estimated with the following basic constraints. Maximum radius of the vesicle is 0.1 µm, a reasonable value considering the conditions of intense emulsification. In the same manner a rational minimum value is 25 nm. The combination of a core radius of 0.1 µm and a shell thickness of 4 nm results in an R value of 0.04. The opposite extreme is a core radius of 25 nm and a shell thickness of 12 nm with an R value of 0.48. Furthermore, with the conservative assumption of a 0.5 fraction of vesicles as the maximum with retained stability, the two limits 0.056 and 0.35 are obtained to the fraction of liquid crystal to be dispersed as vesicles. With these numbers as a basis it is possible to calculate the limits to evaporation for the lines in Fig. 2. For the two R values selected the water content of these limits is equal to WLimit = WL + k(WL − WLC ),
(c) Fig. 3. Weight fraction liquid crystal versus surfactant fraction (a), ionic liquid fraction (b) and evaporated water fraction (c) during evaporation of emulsions with 5% surfactant. Original weight fraction water and ionic liquid: (Q) 0.9500 and 0, (E) 0.9412 and 0.0088, (") 0.9375 and 0.0125, (P) 0.9286 and 0.0214, (2) 0.8750 and 0.0750, (!) 0.7500 and 0.2000.
(9)
in which k is equal to 0.056 or 0.35. Table 3 displays these numbers. The question of interest now becomes the limit of evaporation for different emulsions before the vesicle stability limit is reached. These limits are found by combining the equation for evaporation lines, Eq. (1), with equations describing the limits
498
S.E. Friberg / Journal of Colloid and Interface Science 307 (2007) 494–499
Table 3 Water fractions for maximum liquid crystal forming stable vesicles against coalescence during evaporation Liquid phase water
R = 0.04 Water
Surfactant
R = 0.48 Water
Surfactant
1.0 0.92 0.71 0.4 0.155 0.115
0.972 0.896 0.694 0.398 0.159 0.121
0.028 0.047 0.166 0.380 0.562 0.753
0.824 0.768 0.611 0.386 0.177 0.155
0.176 0.188 0.273 0.431 0.600 0.743
Fig. 5. The maximum evaporated water fraction with retained vesicle stability for emulsions with 0.01 fraction surfactant and varied fraction ionic liquid: R = 0.48 (!) and 0.040 (F).
according to the numbers in Table 3. These were amenable to fit to simple empirical equations. The values under R = 0.04 followed the equation S = 0.4149W 2 − 1.1905W + 0.7883 (R 2 = 0.999),
(10)
while the values for R = 0.48 obeyed the empirical relation S = 0.5770W 2 − 1.2884W + 0.8425 (R 2 = 0.999).
(11)
These equations are subsequently used to calculate the water content for evaporating emulsions, when the evaporation line reaches these limits. The initial fraction of surfactant was 0.01 for all the emulsions to be evaporated, while the amount of ionic liquid was varied. The results, Fig. 5, are presented as the maximum evaporated water fraction versus the initial ionic liquid fraction. In this case the fraction evaporated water was counted on the total emulsion weight; different from the representation in Fig. 3c. In that figure the evaporated water is given as the fraction of the fraction water in the original emulsion, while in Fig. 5 the evaporated water is given as the fraction of the total emulsion weight. The algebraic treatment of the phase diagram results offers a unique and useful feature for a problem like this. The combination of Eqs. (10) and (11) with Eq. (9) gives two equations that are solved with respect to the water faction. The numerical
results are
W = 1.435 − 1.205k ± (1.435 − 1.205k)2 1/2 − 1.900 + 2.410K ,
(12)
for R = 0.04, and
W = 1.117 − 0.867k ± (1.117 − 0.867k)2 1/2 − 1.460 + 1.733K ,
(13)
for R = 0.48. It is immediately obvious that the lower limit to the value of k for the evaporation line is equal to the solution of the expression under the root sign put equal to zero; k = 0.51 or 0.72. These two values agree with the maximum values of the fraction ionic liquid in Fig. 5. The consequences of the destabilization of vesicles during the evaporation of an emulsion of this kind are unambiguous; virtually any emulsion will cause the vesicles to coalesce ultimately forming liquid crystal particles in the system. Such a process, to reach the final state, must with necessity be initiated by flocculation and subsequent coalescence of the vesicles [28]. Different polyhedral structures as potential intermediates in such a transformation are of interest and have been analyzed [29], but such reports do not fully cover the case of evaporation of vesicle solutions with spherical vesicles, because coalescence would require a change in the total area of the vesicle; in principle not realistic for spherical micelles. However, the general validity of this objection is debatable. The fact is that vesicles formed during emulsification frequently have interfaces far in excess of the one needed for spherical shape and flocculated half-spheres are frequently observed in commercial emulsions [30]. This fact makes a discussion of potential structures of aggregated polyhedra of interest. One relevant example of such polyhedra is the octahedron, which can be close packed into a space filling arrangement. Since half spheres do exist in emulsions of this kind, a comparison between the interface demands for half spheres and octahedra with equal volume would be useful to estimate the potential for close packed polyhedra. The volume of a half sphere is 2πr 3 /3 and its area 3πr 2 . The volume of an octahedron with equal length d of the edges is d 3 21/2 /3 and its area 2d 2 31/2 . For identical volume one finds d = r(π21/2 )1/3 . The area of the octahedron area and that of the half sphere are virtually equal with a ratio between them of 0.99. Hence, the energy barrier to vesicles in an emulsion forming close packed gel structures without coalescence appears scarcely prohibitive. Of course, if the flocculation would be accompanied by coalescence to a degree that all the octahedra share surfaces, the change from half spheres to octahedra means a substantial reduction of the surface. It is certainly thought provoking to note that for such an arrangement of octahedra is within the realm of possibility and have actually been experimentally observed [28]. As for the algebraic treatment in this contribution, the solutions to Eqs. (12) and (13) also give a second point for the evaporation curve joining the limit for vesicles, now with greater water fraction of evaporation. The value of this point has not
S.E. Friberg / Journal of Colloid and Interface Science 307 (2007) 494–499
been included for two reasons. At first the validity of the empirical Eqs. (10) and (11) does not include lower values of water content and, secondly, the significance of the additional value satisfying Eqs. (12) and (13) is not explicitly established for the scenario of evaporation from actual emulsion. The potential for vesicles coalesced into a liquid crystalline phase during evaporation, spontaneously to form again without mechanical action is the focus of an ongoing discussion [25,29]. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12]
C.A. Angell, Science (1995) 1924. T. Welton, Chem. Rev. 99 (1999) 2071. C.M. Gordon, Appl. Catal. 222 (2001) 101. C.L. Hussey, Pure Appl. Chem. 60 (1988) 1763. Y. Chauvin, H. Oliver-Bourbigou, CHEMTECH 25 (1995) 26. K.R. Seddon, J. Chem. Tech. Biotech. 68 (1997) 351. N. Audic, H. Clavier, M. Mauduit, J.-C. Guillemin, J. Am. Chem. Soc. 125 (2003) 9248. T. Welton, Chem. Rev. 99 (1999) 2071. E.V. Dickinson, M.E. Williams, S.M. Hendrickson, H. Masui, R.W. Murray, J. Am. Chem. Soc. 121 (1999) 613. N. Kimizuka, T. Nakashima, Langmuir 17 (2001) 6759. B.P. Binks, A.K.F. Dyab, P.D.I. Fletcher, Chem. Commun. (2003) 2540. S.E. Friberg, Y. Qi, F. Patel, R.A. Mackay, J.D. Holbrey, K.R. Seddon, P.A. Aikens, J. Dispersion Sci. Technol. 21 (2000) 185.
499
[13] Y. Gao, B. Han, G. Li, D. Shen, Z. Li, W. Hou, G. Zhang, Langmuir 21 (2005) 5681. [14] J. Eastoe, S. Gold, S.E. Rogers, A. Paul, T. Welton, R.K. Heenan, I. Grillo, J. Am. Chem. Soc. 127 (2005) 7302. [15] H.X. Gao, J.C. Li, B.X. Han, W.N. Chen, J.L. Zhang, R. Zhang, D.D. Yan, Phys. Chem. Chem. Phys. 6 (2004) 2914. [16] H. Gao, J. Li, B. Han, W. Chen, J. Zhang, R. Zhang, D. Yang, Phys. Chem. Chem. Phys. 6 (2004) 1463. [17] J. Eastoe, S. Gold, A. Paul, T. Welton, R.K. Heenan, I. Grillo, J. Am. Chem. Soc. 127 (2005) 7302. [18] J. Li, J. Zhang, H. Gao, B. Han, L. Gao, Colloid Polym. Sci. 283 (2005) 1371. [19] S.Y. Gao, J. Zhang, H. Xu, X. Zhao, L. Zheng, Langmuir, in press. [20] S.E. Friberg, A. Al-Bawab, Langmuir 21 (2005) 9896. [21] S.E. Friberg, Colloids Surf. A 282–283 (2006) 369. [22] S.E. Friberg, Adv. Colloid Interface Sci. 75 (1998) 181. [23] B. Langlois, S.E. Friberg, J. Soc. Cosmet. Chem. 44 (1993) 1. [24] E.W. Kaler, K.L. Herrington, A.K. Murthy, J.A.N. Zasadzinski, J. Phys. Chem. B 96 (1992) 6698. [25] E.W. Kaler, K.A. Herrington, D.J. Iampietro, B.A. Coldren, H.-T. Jung, J.A. Zasadzinzki, in: M. Abe, J.F. Scamehorn (Eds.), Mixed Surfactant Systems, Marcel Dekker, New York, 2005, p. 289. [26] D. Simmons, M.E. Cates, J. Phys. II (France) 2 (1992) 1439. [27] S. Bailey, M. Longo, S. Chirovolu, J.A. Zasadzinski, Langmuir 6 (1990) 1326. [28] H.T. Jung, Y.S. Lee, E.W. Kaler, B. Goldren, J.A. Zasadzinski, Proc. Natl. Acad. Sci. 99 (2002) 15328. [29] M. Muller, H. Hoffman, in: M. Abe, J.F. Scamehorn (Eds.), Mixed Surfactant Systems, Marcel Dekker, New York, 2005, p. 403. [30] M. Almgren, personal information.
Evaporation from an ionic liquid emulsion Stig E. Friberg ∗ Chemistry Department, University of Virginia, 1695 Goldentree Place, Charlottesville, VA, USA Received 7 September 2006; accepted 7 December 2006 Available online 12 December 2006
Abstract The conditions during evaporation in a liquid crystal-in-ionic liquid microemulsion (LC/µEm) were estimated using the phase diagram of the system. The equations for selected tie lines were established and the coordinates calculated for the sites, at which the evaporation lines crossed the tie lines. These values combined with the coordinates for the phases connecting the tie lines were used to calculate the amounts and the composition of the fractions of the two phases present in the emulsion during the evaporation. One of the emulsion phases was a lamellar liquid crystal and high energy emulsification would lead to the liquid crystal being disrupted to form vesicles. Such a system tenders a unique opportunity to study the interaction between vesicles and normal micelles, which gradually change to inverse micelles over bi-continuous structures. The amount of vesicles in the liquid phase versus the fraction liquid crystal was calculated for two extreme cases of vesicle core size and shell thickness. The limit of evaporation while retaining the vesicle structure was calculated for emulsions of different original compositions assuming the minimum continuous liquid phase to be 50% of the emulsion. © 2007 Elsevier Inc. All rights reserved. Keywords: Emulsions; Microemulsions; Vesicles; Evaporation; Detergents; Amphiphiles
1. Introduction Ionic liquids (IL) [1,2] offer significant advantages as solvents because of their negligible vapor pressure and useful molecular interactions. These properties have enabled their use as reaction medium; even as catalysts [3], for a number of organic reactions such as polymerization, Diels–Alder reactions, hydrogenation and Friedel Crafts reactions [4–8]. Other applications involve electrochemical reactions [9] and as biopolymers [10]. The structure of the individual molecules forming the ionic liquid makes it possible to change its properties [11] from liquids of high polarity to entities with properties similar to those of hydrocarbons. They have, hence, been engaged in colloidal systems acting both as hydrocarbons [12,13] and as liquids of sufficiently high polarity to serve as the polar continuous part of a microemulsion [14–18]. The structures in the latter liquids have recently been systematically investigated by Zheng and his group [19] using a combination of phase diagram determi* Fax: +1 434 973 8826.
E-mail address: [email protected]. 0021-9797/$ – see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.jcis.2006.12.019
nation, dynamic light scattering and molecular spectroscopy to elucidate the structures involved in a non-aqueous microemulsion of this kind. Although the less polar ionic liquids have many properties in common with hydrocarbons, they are markedly different in having an extremely low vapor pressure. In fact, this feature opens the potential for exact information about the behavior of their colloidal systems during evaporation. This fact in combination with the results of recent investigations of complex emulsion systems [20,21] encouraged an investigation into the conditions during evaporation of a nontraditional emulsion containing an ionic liquid. The chosen emulsion contains two phases; one of which is a ionic liquid-inwater microemulsion stabilized by a nonionic surfactant and the second phase a lamellar liquid crystal formed by the same surfactant and water with solubilized ionic liquid. During the evaporation the ionic liquid-in-water microemulsion is gradually transformed to a water-in-ionic liquid one over bi-continuous structures. This unusual system was selected, because it offered a unique opportunity to establish the basic conditions for an arrangement in which the evaporation would give rise to an
S.E. Friberg / Journal of Colloid and Interface Science 307 (2007) 494–499
emulsion, in which microemulsion droplets would be combined with vesicles in an ionic liquid matrix of incessantly reduced volume containing normal micelles, whose structure gradually is reversed to become inverse micelles. In such a medium the phase diagram approach would provide unique information about the potential limits to retain the structure of co-existing vesicles and micelles; a matter of interest beyond the narrow area of ionic liquids. The ionic liquid, 1-butyl-3-methylimidazolium tetrafluoroborate, [C4 -mim][BF4 ] was selected, because its microemulsion with water and a non-ionic surfactant varies from O/W to W/O within one solubility region; a phenomenon investigated by Zheng’s group [19]. 2. The method The goal of the investigation is to establish simple expressions to illustrate the changes in the phase fractions and structures encountered during evaporation of emulsions of ionic liquids. The analysis is built on the phase diagram of the system [12] utilizing an algebraic system to extract information from phase diagram [20,21]. The evaporation line in a system of water, oil and a surfactant, in which only water evaporates, follows a general equation S = S0 (W − 1)/(W0 − 1).
(1)
In Eq. (1), and throughout this article, capitals mean weight fraction of the components and index 0 denotes the initial composition of an emulsion. It should be observed that the equation correctly denotes a line emanating from the water corner, because water is the sole component evaporating. The equation for the line between the compositions of two phases in equilibrium (WL , SL , ILL ) and (WLC , SLC , ILLC ) is S = SL + (SLC − SL )(W − WL )/(WLC − WL ).
(2)
Following the recent analysis [20,21] the amount of the two phases is calculated using one coordinate only. In the present calculations the water fraction is chosen, because the variation with water content is an obvious parameter for the evaporation. This water fraction, WTL Em , of the total emulsion composition for each tie line (TL) during the evaporation is obtained by a combination of Eqs. (1) and (2): WTL Em = SL + S0 /(W0 − 1) − WL (SLC − SL ) (WLC − WL ) S0 (W0 − 1) − (SLC − SL )/(WLC − WL ) . (3)
495
of the latter is postulated not to change, when it forms the shell of the vesicle. For an emulsion, in which one of the phases is a lamellar liquid crystal, the emulsification method is of paramount importance for the structure of the emulsion. With gentle emulsification, such as manual shaking, the liquid crystal is dispersed as particles, prominently displayed as shiny particles when viewed between crossed polarizers in an optical microscope. With high energy emulsification the liquid crystal particles are dispersed as vesicles. The vesicles consist of a central core of the continuous liquid with a radius r, surrounded by a shell, with a thickness d, of bilayers or multilayers of the liquid crystalline structure. The vesicles are in fact the dispersed drops in a W/W emulsion, stabilized by an even number of multiple layers. The presence of these structures exerts a pronounced demand on the continuous phase [21] and it is instructive to analyze the effect the presence of such structures would have in an emulsion. In the following analysis the vesicles are assumed to have a solvent core radius of r and a shell thickness of d giving a total volume of 4π(r + d)3 /3. The volume fraction of the shell is [(r + d)3 − r 3 ]/(r + d)3 , which in this contribution is assumed to be identical to the volume of the liquid crystal used to form the shell. Since the ratio between volumes is independent of the dimensions per se of the core and the shell, the shell volume fraction may be written as Frshell = (1 + R)3 − 1 /(1 + R)3 , (5) in which R = d/r. 3. The emulsion The emulsion system is shown in Fig. 1. It consists of a lamellar liquid crystal, marked LC in Fig. 1, and a microemulsion, marked L, that reaches from the water corner to homogeneous surfactant in a continuous region. Hence, the liquid part of the emulsion changes from a water continuous liquid to a surfactant continuous one with no phase change [19]. The evaporation line that is given in the figure emanates from the water corner, because the ionic liquid does not evaporate to a degree influencing the direction of the line. The tie lines between the two phases are indicated on the diagram and the compositions in equilibrium are given in Table 1.
Index LC refers to fraction of the liquid crystalline phase [20,21] while index L indicates the fraction isotropic liquid microemulsion phase. LC = (WTLEL − WLC )/(WL − WLC ) = S0 (WL − 1) + SL (1 − W0 ) /S0 (WL − WLC ) + (SLC − SL )(W0 − 1).
(4)
In the following evaluations the density of the liquid phase is assumed equal to that of the liquid crystal phase and the density
Fig. 1. The emulsion system. The liquid part of the emulsion, L, is a microemulsion and the liquid crystal, LC, has a lamellar structure. The coordinates for the tie lines are given in Table 1.
496
S.E. Friberg / Journal of Colloid and Interface Science 307 (2007) 494–499
Table 1 Fractions in the phase compositions connected by tie lines Liquid phase
Liquid crystal phase
Water
Surfactant
Water
Surfactant
1.00 0.92 0.71 0.40 0.155 0.115
0 0.02 0.145 0.37 0.555 0.755
0.49 0.48 0.425 0.36 0.22 0.23
0.51 0.505 0.515 0.545 0.685 0.72
Table 2 Coefficients for the equations for tie lines Tie line #
A
B
0 1 2 3 4 5
1 1.0341 1.067 2.120 0.245 0.790
−1 −1.102 −1.298 −4.375 2.000 −0.304
The curves are a good illustration of the advantage of extracting information from a phase diagram using the algebraic method [20,21]. At first it provides quantitative evidence of the trend of the initial increase of the fraction liquid crystal during evaporation and its subsequent decline to zero. Secondly it offers an easy access to the change with the water content because the initial and final parts of the curves in Fig. 2 are linear in water fraction of the total emulsion. It would be attractive to use these features to estimate the times for the increase and reduction of the liquid crystal fractions, but such an analysis is prohibited for several reasons. At first the vapor pressure varies with the composition, although not to the degree of molecular solutions [22]. Secondly, the presence of the liquid crystal has a significant influence on the kinetics of the evaporation [23]. Thirdly, an additional advantage of the algebraic method is the fact that the presentation of results is easily changed from one variable to another. As an example the results in Fig. 2 are directly transposed to substitute dependent variables. So is the water fraction of the total emulsion changed to fraction evaporated water by the following relation: WEv = (WOr − W )/WOr (1 − W ),
Fig. 2. Weight fraction liquid crystal during evaporation for emulsions containing 5% surfactant. Original weight fraction water and ionic liquid: (Q) 0.9500 and 0, (E) 0.9412 and 0.0088, (") 0.9375 and 0.0125, (P) 0.9286 and 0.0214, (2) 0.8750 and 0.0750, (!) 0.7500 and 0.2000.
Applying Eq. (2) to the system the expressions are established for the tie line equations counted from left to right in Fig. 1. Replacing SL + (SLC − SL )/(WLC − WL ) in Eq. (2) with A and (SLC − SL )/(WLC − WL ) with B, the equation is written as S = A + BW.
(6)
The coefficients A and B for the different tie lines from left to right in Fig. 1 are given in Table 2. The values of the water fraction for the junction of the evaporation lines with the tie lines are immediately obtained from Eq. (3) and direct substitution into Eq. (4) provides the values of the weight fraction of the liquid crystal for the different evaporation lines as shown in Fig. 2. The results illustrate the original increase of the fraction liquid crystal and the sudden reduction of these values at the later part of the evaporation process.
(7)
in which W is the water fraction of the total emulsion at the evaporation step, WOr the water faction in the emulsion prior to evaporation and WEv the evaporated fraction of WOr . The curve is shown in Fig. 3 in addition to the liquid crystal fraction versus the surfactant fraction and the ionic liquid fraction. As mentioned earlier, the lamellar liquid crystal will be changed to vesicles during high energy emulsification and an analysis of the importance of such a change for the present system is appropriate. During evaporation the liquid fraction, forming the continuous phase, is reduced and the liquid reaches a packing limit for stability of the vesicle structure, as was early introduced by Kaler et al. [24]. The structural changes have been summarized by Kaler in a recent publication [25]. It is evident that a large number of structures may appear during the “compacting” process; the available thermodynamic analysis [27] may be summarized as [25,26] the large vesicles disappearing causing a more mono-disperse distribution during reduced fraction of the solvent. In the present evaluation the picture is further complicated by the presence of microemulsion droplets in the continuous phase. At first these will interact with the vesicles and a modification of both structures is a reasonable assumption and, in addition, the evaporation causes a change of the micellar structure per se. In this, the first article about these problems, the matter of the micellar structure and its potential influence on the vesicle structure will be neglected in order to limit the magnitude of the article and the following discussion is restricted to the vesicles. Estimating the limit of evaporation to retain the vesicle stability is arbitrary, but a maximum volume fraction of vesicles at 0.5 is realistic and will be used in the following discussion. With this stipulation the condition for maximum liquid crystal fraction to retain the vesicles is written LC = 0.5 1 − 1/(1 + R)3 . (8)
S.E. Friberg / Journal of Colloid and Interface Science 307 (2007) 494–499
(a)
497
Fig. 4. The maximum fraction of liquid crystal in emulsions, in which the liquid crystal has formed vesicles with retained stability. The initial surfactant fraction was 0.01 and the ionic liquid fraction adjusted to give the required value of R.
Fig. 4 reveals the extremely small weight fraction allowed for small values R in accordance with Kaler [25]. For such values of R Eq. (8) arrives at Vmax ≈ 1.5R showing a linear behavior. For large values, lim VR⇒∞ = 0.5.
(b)
In real systems the lower limit of R is determined by the fact that a bimolecular layer is physically the thinnest membrane possible. Hence the dimensions of the shells are limited to greater than approximately 4 nm, with values of 8 and 12 nm as the subsequent dimensions. There is, of course, in principle, no upper limitation as to the thickness of the vesicle shells, but with such an intense emulsification as to form vesicles from the liquid crystal particles, an assumption of thicker shells than the mentioned largest value is not realistic [25]. The maximum fraction of liquid crystal to retain the stability of the vesicles may now be estimated with the following basic constraints. Maximum radius of the vesicle is 0.1 µm, a reasonable value considering the conditions of intense emulsification. In the same manner a rational minimum value is 25 nm. The combination of a core radius of 0.1 µm and a shell thickness of 4 nm results in an R value of 0.04. The opposite extreme is a core radius of 25 nm and a shell thickness of 12 nm with an R value of 0.48. Furthermore, with the conservative assumption of a 0.5 fraction of vesicles as the maximum with retained stability, the two limits 0.056 and 0.35 are obtained to the fraction of liquid crystal to be dispersed as vesicles. With these numbers as a basis it is possible to calculate the limits to evaporation for the lines in Fig. 2. For the two R values selected the water content of these limits is equal to WLimit = WL + k(WL − WLC ),
(c) Fig. 3. Weight fraction liquid crystal versus surfactant fraction (a), ionic liquid fraction (b) and evaporated water fraction (c) during evaporation of emulsions with 5% surfactant. Original weight fraction water and ionic liquid: (Q) 0.9500 and 0, (E) 0.9412 and 0.0088, (") 0.9375 and 0.0125, (P) 0.9286 and 0.0214, (2) 0.8750 and 0.0750, (!) 0.7500 and 0.2000.
(9)
in which k is equal to 0.056 or 0.35. Table 3 displays these numbers. The question of interest now becomes the limit of evaporation for different emulsions before the vesicle stability limit is reached. These limits are found by combining the equation for evaporation lines, Eq. (1), with equations describing the limits
498
S.E. Friberg / Journal of Colloid and Interface Science 307 (2007) 494–499
Table 3 Water fractions for maximum liquid crystal forming stable vesicles against coalescence during evaporation Liquid phase water
R = 0.04 Water
Surfactant
R = 0.48 Water
Surfactant
1.0 0.92 0.71 0.4 0.155 0.115
0.972 0.896 0.694 0.398 0.159 0.121
0.028 0.047 0.166 0.380 0.562 0.753
0.824 0.768 0.611 0.386 0.177 0.155
0.176 0.188 0.273 0.431 0.600 0.743
Fig. 5. The maximum evaporated water fraction with retained vesicle stability for emulsions with 0.01 fraction surfactant and varied fraction ionic liquid: R = 0.48 (!) and 0.040 (F).
according to the numbers in Table 3. These were amenable to fit to simple empirical equations. The values under R = 0.04 followed the equation S = 0.4149W 2 − 1.1905W + 0.7883 (R 2 = 0.999),
(10)
while the values for R = 0.48 obeyed the empirical relation S = 0.5770W 2 − 1.2884W + 0.8425 (R 2 = 0.999).
(11)
These equations are subsequently used to calculate the water content for evaporating emulsions, when the evaporation line reaches these limits. The initial fraction of surfactant was 0.01 for all the emulsions to be evaporated, while the amount of ionic liquid was varied. The results, Fig. 5, are presented as the maximum evaporated water fraction versus the initial ionic liquid fraction. In this case the fraction evaporated water was counted on the total emulsion weight; different from the representation in Fig. 3c. In that figure the evaporated water is given as the fraction of the fraction water in the original emulsion, while in Fig. 5 the evaporated water is given as the fraction of the total emulsion weight. The algebraic treatment of the phase diagram results offers a unique and useful feature for a problem like this. The combination of Eqs. (10) and (11) with Eq. (9) gives two equations that are solved with respect to the water faction. The numerical
results are
W = 1.435 − 1.205k ± (1.435 − 1.205k)2 1/2 − 1.900 + 2.410K ,
(12)
for R = 0.04, and
W = 1.117 − 0.867k ± (1.117 − 0.867k)2 1/2 − 1.460 + 1.733K ,
(13)
for R = 0.48. It is immediately obvious that the lower limit to the value of k for the evaporation line is equal to the solution of the expression under the root sign put equal to zero; k = 0.51 or 0.72. These two values agree with the maximum values of the fraction ionic liquid in Fig. 5. The consequences of the destabilization of vesicles during the evaporation of an emulsion of this kind are unambiguous; virtually any emulsion will cause the vesicles to coalesce ultimately forming liquid crystal particles in the system. Such a process, to reach the final state, must with necessity be initiated by flocculation and subsequent coalescence of the vesicles [28]. Different polyhedral structures as potential intermediates in such a transformation are of interest and have been analyzed [29], but such reports do not fully cover the case of evaporation of vesicle solutions with spherical vesicles, because coalescence would require a change in the total area of the vesicle; in principle not realistic for spherical micelles. However, the general validity of this objection is debatable. The fact is that vesicles formed during emulsification frequently have interfaces far in excess of the one needed for spherical shape and flocculated half-spheres are frequently observed in commercial emulsions [30]. This fact makes a discussion of potential structures of aggregated polyhedra of interest. One relevant example of such polyhedra is the octahedron, which can be close packed into a space filling arrangement. Since half spheres do exist in emulsions of this kind, a comparison between the interface demands for half spheres and octahedra with equal volume would be useful to estimate the potential for close packed polyhedra. The volume of a half sphere is 2πr 3 /3 and its area 3πr 2 . The volume of an octahedron with equal length d of the edges is d 3 21/2 /3 and its area 2d 2 31/2 . For identical volume one finds d = r(π21/2 )1/3 . The area of the octahedron area and that of the half sphere are virtually equal with a ratio between them of 0.99. Hence, the energy barrier to vesicles in an emulsion forming close packed gel structures without coalescence appears scarcely prohibitive. Of course, if the flocculation would be accompanied by coalescence to a degree that all the octahedra share surfaces, the change from half spheres to octahedra means a substantial reduction of the surface. It is certainly thought provoking to note that for such an arrangement of octahedra is within the realm of possibility and have actually been experimentally observed [28]. As for the algebraic treatment in this contribution, the solutions to Eqs. (12) and (13) also give a second point for the evaporation curve joining the limit for vesicles, now with greater water fraction of evaporation. The value of this point has not
S.E. Friberg / Journal of Colloid and Interface Science 307 (2007) 494–499
been included for two reasons. At first the validity of the empirical Eqs. (10) and (11) does not include lower values of water content and, secondly, the significance of the additional value satisfying Eqs. (12) and (13) is not explicitly established for the scenario of evaporation from actual emulsion. The potential for vesicles coalesced into a liquid crystalline phase during evaporation, spontaneously to form again without mechanical action is the focus of an ongoing discussion [25,29]. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12]
C.A. Angell, Science (1995) 1924. T. Welton, Chem. Rev. 99 (1999) 2071. C.M. Gordon, Appl. Catal. 222 (2001) 101. C.L. Hussey, Pure Appl. Chem. 60 (1988) 1763. Y. Chauvin, H. Oliver-Bourbigou, CHEMTECH 25 (1995) 26. K.R. Seddon, J. Chem. Tech. Biotech. 68 (1997) 351. N. Audic, H. Clavier, M. Mauduit, J.-C. Guillemin, J. Am. Chem. Soc. 125 (2003) 9248. T. Welton, Chem. Rev. 99 (1999) 2071. E.V. Dickinson, M.E. Williams, S.M. Hendrickson, H. Masui, R.W. Murray, J. Am. Chem. Soc. 121 (1999) 613. N. Kimizuka, T. Nakashima, Langmuir 17 (2001) 6759. B.P. Binks, A.K.F. Dyab, P.D.I. Fletcher, Chem. Commun. (2003) 2540. S.E. Friberg, Y. Qi, F. Patel, R.A. Mackay, J.D. Holbrey, K.R. Seddon, P.A. Aikens, J. Dispersion Sci. Technol. 21 (2000) 185.
499
[13] Y. Gao, B. Han, G. Li, D. Shen, Z. Li, W. Hou, G. Zhang, Langmuir 21 (2005) 5681. [14] J. Eastoe, S. Gold, S.E. Rogers, A. Paul, T. Welton, R.K. Heenan, I. Grillo, J. Am. Chem. Soc. 127 (2005) 7302. [15] H.X. Gao, J.C. Li, B.X. Han, W.N. Chen, J.L. Zhang, R. Zhang, D.D. Yan, Phys. Chem. Chem. Phys. 6 (2004) 2914. [16] H. Gao, J. Li, B. Han, W. Chen, J. Zhang, R. Zhang, D. Yang, Phys. Chem. Chem. Phys. 6 (2004) 1463. [17] J. Eastoe, S. Gold, A. Paul, T. Welton, R.K. Heenan, I. Grillo, J. Am. Chem. Soc. 127 (2005) 7302. [18] J. Li, J. Zhang, H. Gao, B. Han, L. Gao, Colloid Polym. Sci. 283 (2005) 1371. [19] S.Y. Gao, J. Zhang, H. Xu, X. Zhao, L. Zheng, Langmuir, in press. [20] S.E. Friberg, A. Al-Bawab, Langmuir 21 (2005) 9896. [21] S.E. Friberg, Colloids Surf. A 282–283 (2006) 369. [22] S.E. Friberg, Adv. Colloid Interface Sci. 75 (1998) 181. [23] B. Langlois, S.E. Friberg, J. Soc. Cosmet. Chem. 44 (1993) 1. [24] E.W. Kaler, K.L. Herrington, A.K. Murthy, J.A.N. Zasadzinski, J. Phys. Chem. B 96 (1992) 6698. [25] E.W. Kaler, K.A. Herrington, D.J. Iampietro, B.A. Coldren, H.-T. Jung, J.A. Zasadzinzki, in: M. Abe, J.F. Scamehorn (Eds.), Mixed Surfactant Systems, Marcel Dekker, New York, 2005, p. 289. [26] D. Simmons, M.E. Cates, J. Phys. II (France) 2 (1992) 1439. [27] S. Bailey, M. Longo, S. Chirovolu, J.A. Zasadzinski, Langmuir 6 (1990) 1326. [28] H.T. Jung, Y.S. Lee, E.W. Kaler, B. Goldren, J.A. Zasadzinski, Proc. Natl. Acad. Sci. 99 (2002) 15328. [29] M. Muller, H. Hoffman, in: M. Abe, J.F. Scamehorn (Eds.), Mixed Surfactant Systems, Marcel Dekker, New York, 2005, p. 403. [30] M. Almgren, personal information.