Anionic Surfactant

Sep 4, 2008 - TEM micrographic experiments illustrate that there is no direct relationship between the microstructures of the concentrated ... Cover I...
1 downloads 0 Views 2MB Size
Download PDF
12326

J. Phys. Chem. B 2008, 112, 12326–12337

Salt-Induced Phase Inversion in Aqueous Cationic/Anionic Surfactant Two-Phase Systems Yan-Qing Nan* and Li-Sheng Hao Department of Chemistry, Hunan Normal UniVersity, Changsha, 410081, People’s Republic of China ReceiVed: June 11, 2008; ReVised Manuscript ReceiVed: July 9, 2008

Phase inversion of aqueous two-phase systems with excess cationic surfactant (abbreviated as ATPS-C) formed by aqueous mixtures of 1,3-propanediyl bis(dodecyl dimethylammonium bromide) (abbreviated as 12-3-12) and sodium dodecyl sulfonate (abbreviated as AS) at 318.15 K was investigated. The experimental results indicate that addition of NaF, NaCl, NaHCO3, or NaNO3 can result in phase inversion of ATPS-C formed by 12-3-12/AS systems; however, addition of NaBr cannot lead to phase inversion. TEM micrographic experiments illustrate that there is no direct relationship between the microstructures of the concentrated phase in ATPS-C and phase inversion. To interpret the phase-inversion phenomena of ATPS-C, the phase composition, phase density, and phase volume ratio between the dilute phase and the concentrated phase in ATPS-C were investigated. Phase composition analysis results illustrate that for the ATPS-C formed by 0.10 mol · kg-1 12-3-12/AS mixed system, the concentration of Br- counterions in the dilute phase of ATPS-C increases with addition of NaF, NaCl, NaHCO3, or NaNO3. At the same time, the molar ratio between the F- (Cl-, HCO3-, or NO3-) counterions and Br- counterions in the concentrated phase of ATPS-C increases also. It illustrates that part of the bromide counterions which are the natural counterions of the surfactant 12-3-12 in excess are exchanged by other anionic counterions when an additional salt is added to the system. The investigation indicates that the common ground of the added F-, Cl-, HCO3-, or NO3- counterions is that they all make a smaller density contribution than that of Br- counterions, although they have a weaker or stronger counterion binding ability with the mixed positively charged aggregates in ATPS-C than that of Brcounterion. Density experiments illustrate that the density increase of the dilute phase is larger than that of the concentrated phase in the ATPS-C with addition of NaF, NaCl, NaHCO3, or NaNO3; thus, phase inversion occurs. The densities of the added inorganic sodium salt aqueous solution and the order of the Hofmeister series for the added inorganic anions with respect to the chaotropic headgroup of 12-3-12 play important roles in the phase inversion of ATPS-C. 1. Introduction Aqueous two-phase systems (ATPS) have attracted much attention since the pioneering work of Albertsson.1 It is well known that the mixtures of two incompatible polymers (for instance, poly(ethylene glycol) (PEG) and dextran) or of one polymer and one salt (for instance, PEG and Na2SO4) form a stable two-phase liquid system with water as solvent in both phases. Besides the above two kinds of ATPS, there are still many kinds of ATPS: two oppositely charged polymers ATPS, oppositely charged surfactant/polyelectrolyte ATPS, and two oppositely charged surfactants ATPS, etc. Aqueous two-phase systems present a powerful technique for the separation of biomolecules, cells, cell particles, anions, and cations, etc.2-7 Therefore, phase behaviors of these systems are necessary for design of extraction processes. For several polymer/salt ATPS it is interesting to note that an increase in the system temperature can lead to a phase inversion where the bottom phase becomes the polymer-rich phase and the top phase becomes the salt-rich phase at higher temperature, whereas at lower temperature the polymer-rich phase is the top phase and the salt-rich phase is the bottom phase.8-13 However, only very limited investigations have been performed. This kind of phase inversion has been deduced to the changing relative densities of the phases as temperature increases. For the aqueous two-phase systems containing poly(vinylpyrrolidone) (PVP) and sodium citrate, an * To whom correspondence should be addressed. E-mail: nanyq@ 21cn.com.

increase in temperature will drive water from the PVP-rich phase to the salt-rich phase, so the PVP-rich phase is concentrated while the salt-rich phase will be somewhat more diluted. Thus, by increasing the temperature the density of the PVP-rich phase increases while the density of the salt-rich phase decreases; at higher temperature, the density of the PVP-rich phase will be greater than that of the salt-rich phase; thus, phase inversion occurs.12 Aqueous mixtures of anionic and cationic surfactants exhibit rich microstructured phase behavior and many unique properties that arise from the strong electrostatic interactions between the oppositely charged head groups. For example, the mixtures have dramatically lower critical aggregation concentrations than do single pure surfactant and are often more surface active than either pure surfactant.14,15 The mixtures form a wide variety of microstructures, such as spherical and rodlike micelles, vesicles, lamellar phases, and flat discs, etc.16 Of particular interest is the spontaneous formation of an equilibrium phase of unilamellar vesicles.17 The mixing of two oppositely charged surfactants often results in associative phase separation, like precipitation16,17 or ATPS.18,19 Phase behavior research on aqueous mixtures of oppositely charged surfactants is rather complex because it constitutes a five-component system, i.e., two ionic surfactant salts, complex salt formed by oppositely charged surfactant ionic pair, inorganic salt, and water.17,20-23 Composition analysis23 indicates that the separated two phases in one ATPS-C cannot be treated as mixtures of cationic

10.1021/jp805126j CCC: $40.75  2008 American Chemical Society Published on Web 09/04/2008

Salt-Induced Phase Inversion surfactant, anionic surfactant, and water again; in either of the two separated phases of one ATPS-C, the stoichiometric relationships between surfactant ions and their counterions no longer exist; the four kinds of ions, i.e., surfactant cations, surfactant anions, inorganic cations, and inorganic anions, redistribute in the two separated phases. However, for the ATPS formed by different kinds of oppositely charged surfactants, the concentrated phase may be the bottom one (for example, in ATPS-C formed by 12-3-12/AS/water systems) or the top one (for example, in ATPS-C formed by cetyl trimethylammonium bromide (CTAB)/AS/water systems). What is the decisive factor that determines the concentrated phase in one ATPS as the top one or the bottom one? It is important to reveal the reason why some concentrated phases in ATPS are the bottom ones and some other concentrated phases in ATPS are the top ones. It may help us to understand the phase behavior and reveal the mystery behind phase separation of ATPS. Density measurement illustrates that the density difference between the top phase and the bottom phase in ATPS is very small.24 Some factors which influence the densities of the two-coexisted phases in one ATPS such as the changes of temperature or addition of inorganic salts may result in phase inversion. Electrostatic effects often dominate the phase behavior of oppositely charged surfactant systems. Addition of inorganic salt may cause a modification of both intermicellar and intramicellar interaction because of electrostatic screening. Consequently, the phase behavior as well as properties of the aqueous mixed oppositely charged surfactants undergo marked changes upon addition of salts.23-25 The changes are usually strongly dependent on the kinds of added salts. Specific ionic effects are ubiquitous in colloids field. It is well known that specific ionic (Hofmeister) effects play an important role in colloidal aggregation.26 The effect of anions on micelles has been extensively studied over the years, especially on positively charged micelles; for example, it was found that the presence of anions induces a spherical to rod-like micelle transition.27 It was shown that for aqueous two-phase systems of nonequimolar surfactant mixtures, the salt effect is mainly dependent on the oppositely charged counterion. The shift of the aqueous twophase region is strengthened following the Hofmeister series.28 Specific alkali cation effects were observed in the transition from micelle to vesicles for the catanionic surfactant solution composed of sodium dodecylsulfate (SDS) or sodium dodecylcarboxylate (SL) and dodecyltrimethylammonium bromide (DTAB) with an excess of anionic surfactant through salt addition.29,30 The effects were attributed to the reduced repulsion between the surfactant headgroups, induced by binding of the counterions on the micellar headgroups. However, the specific alkali cation effects follow the Hofmeister series for the SDS/ DTAB system; if the sulfate headgroup of the anionic surfactant is replaced by a carboxylic group, the order of the ions is reversed, i.e., it follows the reversed Hofmeister series. Collins’ “Law of Matching Water Affinities”31,32 has been employed to explain the specific ion effects in the two catanionic systems. The main idea of the Collins model is that small oppositely charged kosmotropic ions can come close together, forming inner-sphere ion pairs without intermediate water molecules. The same is supposed to be true for big chaotropic ions, whereas when a kosmotropic ion approaches a chaotropic counterion, the ions should remain separated by at least one water molecule. Kunz’s group29,30 interpreted their experimental results according to Collins’ concept of matching water affinities. For example, for the SL/DTAB system, the headgroups in excess are dedocylcarboxylate with a kosmotropic behavior. According to

J. Phys. Chem. B, Vol. 112, No. 39, 2008 12327 Collins’ concept, alkylcarboxylates should come in close contact with kosmotropic ions like lithium, making it more possible to form inner-sphere ion pairs, whereas cesium ions remain further away. Therefore, it is expected that Li+ screens more efficiently the negative charge excess on the aggregates than Cs+, and this is precisely what is observed; strong cation specificity was found in assisting vesicle formation following the reversed Hofmeister series. Phase inversion in ATPS-C of 12-3-12/AS/water or 12-312/SDS/water was observed with addition of some alkali halides, whereas addition of NaBr cannot lead to phase inversion of these ATPS-C.22,33 For example,22 when the total surfactant concentration mT e 0.10 mol · kg-1, all top phases are dilute and all bottom phases are concentrated in surfactant for the ATPS-C of 12-3-12/AS mixtures with pure water or 0.10 mol · kg-1 NaBr aqueous solution as solvent at 318.15 K, whereas when 0.10 mol · kg-1 NaCl aqueous solution substitutes pure water as solvent, phase inversion occurs and all of the top phases are concentrated and the bottom phases are dilute, while when 0.10 mol · kg-1 NaF aqueous solution is used as solvent for parts of the ATPS-C phase inversion occurs. For the 12-3-12/AS mixtures, why does addition of NaF and NaCl cause phase inversion of the ATPS-C? Why does addition of NaBr not lead to phase inversion? Does addition of other kinds of sodium salts result in phase inversion? Do the microstructures of the concentrated phases change after phase inversion occurs? In other words, what is the decisive factor in the phase inversion of ATPS-C? Are there specific ionic effects in the phase inversion of ATPS-C? The resolution of these questions is very helpful for deeply understanding the phase separation mechanism of ATPS. To answer the questions, in this work, the mixed cationic gemini surfactant 12-3-12 and a conventional anionic surfactant AS system were used to form aqueous two-phase systems. NaF, NaCl, NaBr, NaHCO3, and NaNO3 aqueous solutions with different molality were used as solvents. Phase-inversion phenomena, phase composition, phase density, phase volume ratio, and microstructures of some aqueous two-phase systems were investigated. 2. Experimental Section 2.1. Materials. Sodium dodecyl sulfonate (AS) was supplied by Shanghai Chemical Reagent Corp. (purity g 97%) and recrystallized three times from ethanol and three times from water. The gemini surfactant 12-3-12 was synthesized by reaction of 1,3-dibromopropane and dodecyldimethylamine in dry ethanol under reflux at 353.15 K for 48 h;34 the product was then recrystallized in ethanol-ethyl acetate mixtures three times. NMR and elemental analysis were used to check the chemical composition. AS and 12-3-12 were dried in a vaccum desiccator at 333.15 K for 48 h before use. No minimum was found in the surface-tension determination before the cmc for the two kinds of surfactants. Sodium fluoride (NaF), AR, was supplied by Shanghai No. 3 Reagent Factory. Sodium chloride (NaCl), AR, was purchased from Qingdao Chemical Reagent Works. Sodium Bromide (NaBr), AR, was supplied by Shanghai No. 4 Reagent Factory. NaF, NaCl, and NaBr were baked at 773 K for 24 h before use. Sodium hydrogen carbonate (NaHCO3), AR, and sodium nitrate (NaNO3), AR, were used as obtained from Xian Chemical Reagent Works. Water was first deionized and then distilled twice from potassium permanganate solution. Methods. In this paper, for all the studied mixed 12-3-12/ AS systems, the total surfactant concentration mT is 0.10 mol · kg-1.

12328 J. Phys. Chem. B, Vol. 112, No. 39, 2008 Stock solutions (0.10 mol · kg-1) of 12-3-12 and AS were prepared using sodium salt aqueous solutions as solvent. About 15 samples with different molar ratios of 12-3-12 to AS were prepared by mixing stock solutions of 12-3-12 and AS in test tubes for each kind of solvent (for example, 0.07 mol · kg-1 NaCl aqueous solution was used as solvent). The tubes were then immersed in a water bath at 318.15 ( 0.01 K for about 1 week or longer until equilibrium was attained. Phase separation was judged visually. Phase equilibrium is attained when a clear interfacial boundary vertical to the wall is formed; the volumes and appearance of the top and the bottom phases no longer change with time. TEM Measurements. (1) Freeze-Fracture Technique: Fracturing and replication were carried out in a high-vacuum freezeetching system (Balzers BAF-400D) according to standard techniques. (2) Negative-Stained Method: One drop of the mixed surfactant solution was spread on a 200-mesh copper grid coated with a carbon film; another drop of the staining solution (2.0 wt. % of uranyl acetate in ethanol solvent) was then added. The excess solution was sucked away by filter paper. The sample was then air dried. (3) Transmission Electron Microscope: Replicas were examined in a transmission electron microscope (PHILIPS-FEI TECNA120). Negative-stained samples were examined in a transmission electron microscope (JEM-100CX II or JEOL-1230 electron microscope operating at 100 kV). Composition Measurements. First, a certain amount of solution of the top phase or bottom phase of an ATPS-C system was weighted and then kept at 323.15 ( 0.1 K in a water bath for about 24 h; water in the solution was evaporated thoroughly. The solid substance obtained was mixed thoroughly and dried in a vacuum desiccator at 323.15 K for 24 h and then weighted and analyzed by an elementary analyzer (elementar vario EL III) made in Germany. The concentration of Br- or Na+ ions in the dilute phase of the ATPS-C was measured by a DX280 Br- ion or DX223 Na+ ion-selective electrode and DX 200 reference electrode (METTLER TOLEDE SevenMulti), respectively. For standard solutions, the measurement error of the ion-selective electrode method is usually less than 1.5%. The concentration of Br- ions or the sum of the concentrations of Br- ions and Cl- ions were determined by the Mohr method also. The Mohr method uses potassium chromate as an indicator in the titration of bromide or chloride ions with a silver nitrate standard solution. After all the bromide or chloride has been precipitated, the first excess of titrant results in formation of a brick-red silver chromate precipitate, which signals the end point. For standard solutions, the measurement error of the Mohr method is usually less than 0.5%. Density Measurements. Densities of the solutions are measured using a density bottle with a volume of about 10 mL calibrated by redistilled water; the density at 318.15K is 0.990216 g · cm-3.35 To eliminate systematic error, densities of various solutions are measured using the same density bottle. The difference between duplicate measurements is less than (0.00005 g · cm-3. Temperature fluctuation in the thermostatic water bath during the measurement was controlled in less than (0.1 K. The electrical balance used has an accuracy of 0.1 mg. 3. Results 3.1. Phase Inversion in ATPS-C of 12-3-12/AS Mixtures with mT ) 0.10 mol · kg-1. Composition analysis23 and extraction experiments (Figure 1) verify that all of the bottom phases are concentrated in ATPS-C of 12-3-12/AS/water mixtures with a total surfactant concentration mT between 0.035 and 0.10 mol · kg-1 at 318.15 K.

Nan and Hao

Figure 1. Extraction photograph of ATPS-C formed by 0.10 mol · kg-1 12-3-12/AS/water mixture with a molar ratio of 12-3-12 to (12-3-12 + AS) MR12-3-12/(12-3-12+AS) ) 0.6471 at 318.15 K (dye Sudan III). (The purpose for giving the extraction photograph here is only for illustration since the oil-soluble dye Sudan III can be solubilized to the concentrated phase; the bottom phase with a pink color in the ATPS-C system is the concentrated phase. The concentrations of gemini surfactant cations in the top phase and bottom phase of the ATPS-C are 0.00009 and 0.07180 mol · kg-1, respectively, and those of dodecyl sulfonate are 0.00006 and 0.03875 mol · kg-1, respectively.)

Figure 2. Aqueous two-phase region and phase-inversion region in ATPS-C of 12-3-12/AS mixtures with mT ) 0.10 mol · kg-1 at NaCl aqueous solutions at 318.15 K. (The four points with MR12-3-12/(12-3-12+AS) ) 0.6670 correspond to the four ATPS-C chosen for phase density, phase composition measurements.)

For an ATPS-C with one concentrated bottom phase, how can we inverse the concentrated phase from the bottom to the top? For the ATPS-C of the mixed 12-3-12/AS systems, what is the relationship between phase inversion and the amount of added NaCl or NaF? Does addition of NaHCO3 or NaNO3, in which HCO3- or NO3- ions have a weaker or stronger counterion binding ability than Br- ions, lead to phase inversion in ATPS-C? Does the increase of the amount of added NaBr bring about phase inversion of ATPS-C? To answer the questions, the following investigation was performed. 3.1.1. Phase InWersion in ATPS-C of 12-3-12/AS/NaCl Aqueous Solutions. Figure 2 presents the aqueous two-phase region of 12-3-12/AS mixtures with mT ) 0.10 mol · kg-1 at different NaCl aqueous solutions. The two solid lines are the two boundaries of the aqueous two-phase region with excess 12-3-12. Regions 1 and 2 are the phase-inversion region and the nonphaseinversion region in ATPS-C region, respectively. The dashed line is the boundary that divides the phase-inversion region 1 and

Salt-Induced Phase Inversion

Figure 3. Extraction photographs of some ATPS-C formed by 12-312/AS/NaCl aqueous solution with mT ) 0.10 mol · kg-1 and MR12-3-12/(12-3-12+AS) ) 0.6670 at 318.15 K (dye Sudan III): (A) bNaCl ) 0.03, (B) 0.06, (C) 0.07, (D) 0.10 mol · kg-1.

nonphase-inversion region 2 in the ATPS-C region. The dashed line was obtained by the following method. First, about 15 samples with different MR12-3-12/(12-3-12+AS) were prepared for each kind of solvent (such as 0.07 mol · kg-1 NaCl aqueous solution). Second, the samples were kept at a constant temperature of 318.15 ( 0.01 K for about 1 week or longer until equilibrium was attained. Third, Sudan III was added to the aqueous two-phase systems in equilibrium, and then it was observed which phase in ATPS-C is the concentrated phase through extraction results for each sample. Fourth, the value of the turning MR12-3-12/(112-3-12+AS) of the ATPS-C was determined; when MR12-3-12/(12-3-12+AS) is smaller than the value, the concentrated phase in ATPS-C is the top one; when MR12-312/(12-3-12+AS) is larger than the value, the concentrated phase in ATPS-C is the bottom one. Last, the dashed line was obtained by linking these turning points at NaCl aqueous solutions with different concentration. The phase regions 3 and 4 adjacent the ATPS-C region are the liquid-crystalline phase region and isotropic homogeneous phase region, respectively. The experimental results indicate that phase inversion in ATPS-C formed by 0.10 mol · kg-1 12-3-12/AS/(NaCl aqueous solution) at 318.15 K begins to occur at about bNaCl g 0.065 mol · kg-1 (bNaCl is the molality of NaCl aqueous solution used as solvent), and at about bNaCl g 0.095 mol · kg-1 for all of the ATPS-C, phase inversion was observed. Extraction photographs of four ATPS-C (see the four points with MR12-3-12/(12-3-12+AS) ) 0.6670 in the ATPS-C region in Figure 2) formed by 12-3-12/AS/NaCl aqueous solution with different bNaCl are given in Figure 3; the results illustrate clearly that for the ATPS-C with a certain mT and MR12-3-12/(12-3-12+AS) as the amount of NaCl added to the system increases, phase inversion occurs. 3.1.2. Phase InWersion in ATPS-C of 12-3-12/AS/Sodium Salt Aqueous Solutions. The aqueous two-phase region of 123-12/AS mixtures with mT ) 0.10 mol · kg-1 at NaF aqueous solutions, NaHCO3 aqueous solutions, and NaNO3 aqueous solutions are given in Figures 4, 5, and 6. The notations are the same as in Figure 2. Experimental results indicate that phase inversion in ATPS-C formed by 0.10 mol · kg-1 12-3-12/AS/ (NaF aqueous solution) at 318.15 K begins to occur at about bNaF g 0.095 mol · kg-1, and at about bNaF g 0.135 mol · kg-1, for all of the ATPS-C, phase inversion is observed. For those ATPS-C with NaHCO3 aqueous solution as solvent, phase inversion originally occurs at about bNaHCO3 g 0.0875 mol · kg-1, and at about bNaHCO3 g 0.155 mol · kg-1, for all of the ATPSC, phase inversion takes place. When NaNO3 aqueous solution is used as solvent, phase inversion is originally observed at about bNaNO3 g 0.085 mol · kg-1, and at about bNaNO3 g 0.165

J. Phys. Chem. B, Vol. 112, No. 39, 2008 12329

Figure 4. Aqueous two-phase region and phase-inversion region in ATPS-C of 12-3-12/AS mixtures with mT ) 0.10 mol · kg-1 at NaF aqueous solutions at 318.15 K. (The four points with MR12-3-12/(12-3-12+AS) ) 0.6634 correspond to the four ATPS-C chosen for phase density, phase composition measurements.)

Figure 5. Aqueous two-phase region and phase-inversion region in ATPS-C of 12-3-12/AS mixtures with mT ) 0.10 mol · kg-1 at NaHCO3 aqueous solutions at 318.15 K. (The four points with MR12-3-12/(12-3-12+AS) ) 0.6928 correspond to the four ATPS-C chosen for phase density, phase composition measurements.)

Figure 6. Aqueous two-phase region and phase-inversion region in ATPS-C of 12-3-12/AS mixtures with mT ) 0.10 mol · kg-1 at NaNO3 aqueous solutions at 318.15 K. (The four points with MR12-3-12/(12-3-12+AS) ) 0.7945 correspond to the four ATPS-C chosen for phase density, phase composition measurements.)

mol · kg-1, for all of the ATPS-C, phase inversion occurs. For example, when one of the three kinds of 0.20 mol · kg-1 sodium salt aqueous solutions was used as solvent, phase inversion was observed for all of the ATPS-C at 318.15 K (Figure 7).

12330 J. Phys. Chem. B, Vol. 112, No. 39, 2008

Nan and Hao

Figure 7. Extraction photographs of ATPS-C formed by 0.10 mol · kg-1 12-3-12/AS/(0.20 mol · kg-1 aqueous solution) at 318.15 K (dye Sudan III): (A) MR12-3-12/(12-3-12+AS) ) 0.6980 (NaF); (B) MR12-3-12/(12-3-12+AS) ) 0.7126 (NaHCO3); (C) MR12-3-12/(12-3-12+AS) ) 0.8966 (NaNO3). Figure 9. Aqueous two-phase systems formed by 0.10 mol · kg-1 12-3-12/AS/(NaBr aqueous solution) at 318.15 K (dye Sudan III): (A) MR12-3-12/(12-3-12+AS) ) 0.6823 and bNaBr ) 0.10 mol · kg-1, (B) MR12-3-12/(12-3-12+AS) ) 0.9019 and bNaBr ) 0.20 mol · kg-1, (C) MR12-3-12/(12-3-12+AS)) 0.9939 and bNaBr ) 0.30 mol · kg-1.

Figure 8. Aqueous two-phase region of 0.10 mol · kg-1 12-3-12/AS/ NaBr aqueous solution at 318.15 K.

In comparison with NaCl aqueous solution, when NaF aqueous solution, NaHCO3 aqueous solution, or NaNO3 aqueous solution is used as solvent, both the lowest concentration of the salt solution that phase inversion is originally observed and the concentration of the salt solution that for all of the ATPS-C phase inversion occurs are larger. The increase of the amount of added NaF, NaCl, NaHCO3, or NaNO3 is favorable for phase inversion of the ATPS-C. 3.1.3. ATPS-C Region of 12-3-12/AS/NaBr Aqueous Solution. Figure 8 presents the aqueous two-phase region of 0.10 mol · kg-1 12-3-12/AS/(NaBr aqueous solution) at 318.15 K. When 0.30 mol · kg-1 NaBr aqueous solution is used as solvent, the solubility of 12-3-12 at this solvent is less than 0.10 mol · kg-1 at 318.15 K. When 0.10 mol · kg-112-3-12 singlephase solution with 0.30 mol · kg-1 NaBr aqueous solution as solvent at higher temperature was put into a water bath at 318.15 ( 0.01 K and kept constant temperature for a certain time, phase separation was observed and aqueous two-phase system forms; the case is similar to the biphasic phenomenon of the KBr/122-12/D2O system.36 Therefore, NaBr aqueous solution with higher concentration (i.e., bNaBr > 0.30 mol · kg-1) was not used. For all the ATPS-C studied, the concentrated phase is the bottom one (Figure 9); no phase inversion is observed. The increase of the amount of added NaBr cannot cause phase inversion of ATPS-C. 3.2. TEM Micrographs of ATPS-C. TEM micrographs of some concentrated phases in ATPS-C with NaCl aqueous solution, NaF aqueous solution, or water as solvent are given in Figures 10 and 11. The two ATPS-C presented in Figure 10A and 10B have the same MR12-3-12/(12-3-12+AS) and different bNaCl. For the one with the lower bNaCl, the concentrated phase is the bottom one. For the other with a higher bNaCl, its concentrated phase is the top one. The results indicate that

vesicles exist in these concentrated phases, similar to the case in the literature;37 addition of inorganic salt results in fusion of the vesicles; in Figure 10B the shapes of the vesicles ranges from spheroidal to almost tubular. For the concentrated top phase in ATPS-C formed by 0.10 mol · kg-1 12-3-12/AS/(0.20 mol · kg-1 NaF aqueous solution) with MR12-3-12/(12-3-12+AS) ) 0.6670, the presence of vesicles was determined and confirmed by both FF-TEM and negative-stained TEM. Vesicles and vesicle fusion were also observed in the two concentrated phases of ATPS-C in Figure 11C and 11D. The above TEM results illustrate that vesicles and fusion of vesicles are observed for the concentrated phase in ATPS-C of 12-3-12/AS/(sodium halide aqueous solution) whether the concentrated phase is the bottom one or the top one. There is no direct relationship between the microstructures of the concentrated phase in ATPS-C and phase inversion. 3.3. Phase Densities in ATPS-C of 12-3-12/AS Mixtures. Addition of NaF, NaCl, NaHCO3, or NaNO3 leads new inorganic anions into the mixed systems. However, addition of NaBr does not change the kinds of ions in the mixed systems but just increases the concentration of Na+ and Br-. For the mixed 12-3-12/AS systems, in the ATPS-C region the mixed aggregates are positively charged and the counterions are inorganic anions such as Br-, Cl-, F-, HCO3-, or NO3- ions. Literature density data35,38-43 illustrate that at a certain range of concentration the densities of several sodium salt aqueous solutions increase linearly as the molality b of the salts increases at 298.15, 318.15, and 323.15 K (Figure 12). Our previous experimental results24 also indicated that when mT is larger than the critical micelle concentration (cmc), the densities of four kinds of mixed 12-3-12/AS monophasic solutions, i.e., AS solution, 12-3-12 solution, liquid crystalline solution, and isotropic solution near the two boundaries of ATPS-C region, all increase linearly as mT increases. These linear relationships may indicate that the contributions to the density of the ions and water can be considered nearly additive. For the five kinds of sodium salt aqueous solutions in Figure 12, the larger the slope of the line is, the higher the density contribution of the monovalent anion is. The density contribution of the anions follows the sequence of Cl- < F- < NO3-< HCO3- < Br-. Since there are two Br- ions in one 12-3-12 molecule, the density (F) of 12-3-12 aqueous solution is large; for example, F ) 0.990216, 0.99254, and 0.99438 g · cm-3 for 0, 0.05, and 0.10 mol · kg-1 12-3-12 aqueous solutions at 318.15 K, respectively. A higher concentration of 12-3-12 results in a higher

Salt-Induced Phase Inversion

J. Phys. Chem. B, Vol. 112, No. 39, 2008 12331

Figure 10. FF-TEM micrographs of the concentrated phase in ATPS-C formed by 0.10 mol · kg-1 12-3-12/AS/(NaCl aqueous solution). (A) Bottom phase in ATPS-C with MR12-3-12/(12-3-12+AS) ) 0.7181 and bNaCl ) 0.07 mol · kg-1. (B) Top phase in ATPS-C with MR12-3-12/(12-3-12+AS) ) 0.7181 and bNaCl ) 0.20 mol · kg-1.

Figure 11. TEM micrographs of the concentrated phase in ATPS-C formed by 0.10 mol · kg-1 12-3-12/AS at different solvent at 318.15 K. (A and B) Top phase in ATPS-C with MR12-3-12/(12-3-12+AS) ) 0.6670 and bNaF ) 0.20 mol · kg-1. (C) The bottom phase in ATPS-C with MR12-3-12/(12-3-12+AS) ) 0.6928 and bNaF ) 0.10 mol · kg-1. (D) The bottom phase in ATPS-C with MR12-3-12/(12-3-12+AS) ) 0.6453 and pure water as solvent. (B and C) Determined with a JEOL-1230 electron microscope. (D) Determined with a JEM-100CX II electron microscope.

density; therefore, in comparison with the dilute phase, the concentrated phase with higher density is the bottom phase of ATPS-C. When NaBr aqueous solution substituted water as solvent, more Br- ions bind with the mixed aggregates; thus, the density of the mixed aggregates increases, no phase inversion occurs, and the density difference between the two coexisting phases in one ATPS-C increases24 (Figure 13). In order to determine the reason why addition of NaF, NaCl, NaHCO3, or NaNO3 can result in phase inversion of

ATPS-C formed by 12-3-12/AS systems, the densities of the two coexisting phases in some ATPS-C with one definite MR12-3-12/(12-3-12+AS) were investigated. The effect of bsalt (the molality of the sodium salt aqueous solution used as solvent) increase on the density changes was studied. The experimental results presented by Figures 14-17 indicate that although both the densities of the two coexisting phases in ATPS-C increase with bsalt, the density increase of the dilute phase is larger than that of the concentrated phase; thus, phase inversion in ATPS-C occurs.

12332 J. Phys. Chem. B, Vol. 112, No. 39, 2008

Figure 12. Densities of sodium aqueous solutions versus molality of the salts.

Figure 13. Densities of the concentrated phase and dilute phase in ATPS-C with mT ) 0.10 mol · kg-1 at different solvents at 318.15 K: (closed symbols) concentrated phase; (open symbols) dilute phase.

Figure 14. Densities of the concentrated phase and dilute phase in ATPS-C with MR12-3-12/(12-3-12+AS) ) 0.6634 at NaF aqueous solutions with different bNaF at 318.15 K.

3.4. Phase Compositions in ATPS-C of 12-3-12/AS Mixtures. In order to interpret the above phase density changes of ATPS-C in section 3.3, phase compositions of the above ATPS-C were determined. The experimental results are presented in Table 1. For the dilute phase in ATPS-C, the amounts of elemental carbon, nitrogen, and sulfur are too low to be detected. The elementary analysis results indicate that the contents of organic components are very low in the dilute phase of ATPS-C. These

Nan and Hao

Figure 15. Densities of the concentrated phase and dilute phase in ATPS-C with MR12-3-12/(12-3-12+AS) ) 0.6670 at NaCl aqueous solutions with different bNaCl at 318.15 K.

Figure 16. Densities of the concentrated phase and dilute phase in ATPS-C with MR12-3-12/(12-3-12+AS) ) 0.6928 at NaHCO3 aqueous solutions with different bNaHCO3 at 318.15 K.

Figure 17. Densities of the concentrated phase and dilute phase in ATPS-C with MR12-3-12/(12-3-12+AS) ) 0.7945 at NaNO3 aqueous solutions with different bNaNO3 at 318.15 K.

dilute phases in ATPS-C are almost composed of inorganic compounds. However, the bubble properties of the dilute phases illustrate that there are some surfactant molecules in the dilute phases of ATPS-C. Therefore, these results should mean that the surfactant concentration in the dilute phases is very low, maybe close to the critical micelle concentration of the mixed 12-3-12/AS system. Considering the strong synergistic effect

Salt-Induced Phase Inversion

J. Phys. Chem. B, Vol. 112, No. 39, 2008 12333

TABLE 1: Phase Compositions of the Two Coexisting Phases in some ATPS-C with mT ) 0.10 mol · kg-1 at Different Sodium Salt Aqueous Solutions at 318.15 K concentration of the concentrated phase (mol · kg-1)

concentration of the dilute phase (mol · kg-1) salt MR12-3-12/(12-3-12+AS) in ATPS-C NaF 0.6634

NaCl 0.6670

NaHCO3 0.6928

NaNO3 0.7945

bsalt mol · kg-1 0.09 0.11 0.12 0.14 0.03 0.06 0.07 0.10 0.08 0.12 0.14 0.17 0.09 0.12 0.13 0.16

serial number ATPS-C1 ATPS-C2 ATPS-C3 ATPS-C4 ATPS-C5 ATPS-C6 ATPS-C7 ATPS-C8 ATPS-C9 ATPS-C10 ATPS-C11 ATPS-C12 ATPS-C13 ATPS-C14 ATPS-C15 ATPS-C16

Br-

X-

Na+

B-

X-

Na+

12-3-122+

AS-

Vd/Vc

0.0564 0.0591 0.0619 0.0629 0.0498 0.0585 0.0591 0.0652 0.0610 0.0667 0.0686 0.0731 0.0859 0.0901 0.0949 0.1012

0.0789 0.0978 0.1067 0.1256 0.0217 0.0464 0.0576 0.0807 0.0660 0.1008 0.1194 0.1459 0.0397 0.0690 0.0764 0.0986

0.1353 0.1570 0.1686 0.1885 0.0741 0.1065 0.1182 0.1475 0.1270 0.1675 0.1880 0.2190 0.1256 0.1591 0.1713 0.1998

0.1907 0.1911 0.2017 0.2080 0.1868 0.2141 0.2316 0.2500 0.2874 0.2509 0.2083 0.2133 0.1986 0.2569 0.2832 0.3251

0.0983 0.1195 0.1327 0.1554 0.0353 0.0745 0.0862 0.1324 0.1064 0.1491 0.1596 0.1960 0.1161 0.1898 0.2299 0.3283

0.1151 0.1335 0.1397 0.1584 0.0565 0.0795 0.0842 0.1100 0.0806 0.1261 0.1550 0.1819 0.1030 0.1157 0.1125 0.1289

0.1165 0.1185 0.1308 0.1374 0.1093 0.1382 0.1544 0.1797 0.2011 0.1760 0.1368 0.1460 0.1216 0.1901 0.2300 0.3012

0.0591 0.0601 0.0662 0.0697 0.0545 0.0690 0.0771 0.0897 0.0892 0.0780 0.0607 0.0648 0.0315 0.0492 0.0595 0.0779

0.694 0.721 0.880 0.969 0.591 0.973 1.181 1.498 1.661 1.356 0.873 0.983 0.481 1.223 1.629 2.307

of the mixed cationic/anionic surfactants, the critical micelle concentration (cmc) of this kind of mixture is very low. For example, the cmc of the mixed tetradecyltrimethylammonium system is about 10-5 mol · L-1.44 Thus, in these dilute phases of ATPS-C the concentration of 12-3-122+ (C12H25(CH3)2N+(CH2)3N+(CH3)2C12H25) and AS- (C12H25SO3-) may be on the order of 10-5 mol · kg-1, and these concentration values were not given in Table 1. The amounts of 12-3-122+ and AS- in the concentrated phases of ATPS-C are obtained directly by the amounts of elemental nitrogen, sulfur, and carbon. The concentrations of Br- and Na+ in the dilute phases of ATPS-C were determined by the ion-selective electrode method. When NaCl aqueous solution is used as solvent, the sum of the concentrations of Br- and Cl- was obtained by the Mohr method. Thus, the concentration of Br- subtracted from the sum of the concentrations of Br- and Cl- gives the concentration of Cl-. The results in Table 1 indicate that the concentration of Na+ is very close to the sum of the concentrations of Br- and Cl-. This result is consistent with the fact that these dilute phases are almost composed of inorganic compounds. Therefore, the concentration of Cl- can also be obtained by subtracting the concentration of Br- from the concentration of Na+. When NaX (such as NaF, NaHCO3, or NaNO3) aqueous solutions were used as solvent, the concentration of Br- in the dilute phases was also determined by the Mohr method. The concentration of Br- obtained by the ion-selective electrode method and the Mohr method are very close to each other (see Supporting Information). The concentration of Br- presented by Table 1 was determined by the Mohr method. The concentration of Br- subtracted from the concentration of Na+ gives the concentration of X-. The X- ion represents the above univalent anion other than Br- ion. The concentration of inorganic ions in concentrated phases of ATPS-C was calculated by the material balance. For a certain amount of the total ATPS-C with a certain MR12-3-12/AS and bsalt, the total amount of Na+, Br-, and X- in ATPS-C is known; the phase densities of the two coexisting phases in ATPS-C and the phase volume ratio of the dilute phase to the concentrated phase Vd/Vc have been determined. Combined with the experimentally determined concentrations of Na+, Br-, and Xin the dilute phase, the amount of Na+, Br-, and X- in the dilute phase can be calculated; therefore, the amount of Na+,

Br-, and X- and their concentrations in the concentrated phase can be obtained. The experimental results indicate that for the three kinds of inorganic ions in ATPS-C the concentration of Na+ coions in dilute phase is larger than that in the coexisting concentrated phase of ATPS-C; however, both concentrations of the two kinds of counterions Br- and X- in the concentrated phase are larger than those in the coexisting dilute phase. The partition of these inorganic ions in the two coexisting phases of ATPS-C may be related to the electrostatic interactions between inorganic ions and the positively charged mixed aggregates. Electrostatic attraction between the counterions Br- or X- and the aggregates promotes the enrichment of these counterions Br- and X- in the concentrated phase of ATPS-C. Electrostatic repulsion between the Na+ coions and the aggregates causes the higher concentration of Na+ in the dilute phase of ATPS-C. The experimental results in Table 1 illustrate that for the different kinds of inorganic sodium salt aqueous solution used as solvents the concentrations of the three kinds of inorganic ions in the dilute phase of ATPS-C all increases with bsalt. For the same kind of inorganic sodium salt aqueous solution used as solvent, the only difference among the four ATPS-C is that the concentration of the inorganic sodium salt aqueous solution used as solvent bsalt is different. For example, for ATPS-C5 and ATPS-C6 the difference is that the solvent of ATPS-C5 is 0.03 mol · kg-1 NaCl aqueous solution and the solvent of ATPS-C6 is 0.06 mol · kg-1 NaCl aqueous solution. Considering the nearly additive contributions to the density of the ions and water, the density difference between ATPS-C5 and ATPS-C6 is close to the density difference between 0.03 mol · kg-1 NaCl aqueous solution and 0.06 mol · kg-1 NaCl aqueous solution. In other words, the density difference between ATPS-C5 and ATPSC6 is close to the sum of the density contribution of 0.03 mol · kg-1 Na+ and that of 0.03 mol · kg-1 Cl-. For the dilute phase in ATPS-C5 and that in ATPS-C6, when bNaCl changes from 0.03 to 0.06 mol · kg-1, the concentration of Na+ increases from 0.0741 to 0.1065 mol · kg-1 and the sum of the concentration of Br- and Cl- increases from 0.0715 to 0.1049 mol · kg-1, in which the increase of the concentration of Br- is 0.0087 mol · kg-1. These experimental results illustrate that for the two dilute phases in ATPS-C5 and ATPS-C6 their density difference is close to the sum of the density contribution of 0.0324 mol · kg-1 Na+, 0.0247 mol · kg-1 Cl-, and 0.0087 mol · kg-1

12334 J. Phys. Chem. B, Vol. 112, No. 39, 2008

Nan and Hao

Figure 18. Polarization photographs of two samples located at the liquid-crystalline single-phase region formed by 0.10 mol · kg-1 12-3-12/AS/ (0.10 mol · kg-1 sodium salt aqueous solution) at 318.15 K: (A) NaF, MR12-3-12/(12-3-12+AS) ) 0.6324; (B) NaBr, MR12-3-12/(12-3-12+AS) ) 0.7027.

Br-. These results indicate that the increase of the concentration of Na+ in the dilute phase is very close to the increase of bNaCl. Meanwhile, the increase of the total concentration of inorganic anions in the dilute phase is also very close to the increase of bNaCl. However, the concentration increase of the inorganic anions in the dilute phase comes from the sum of the concentration increase of Br- and Cl-. It is obviously different from the case in the whole ATPS-C, in which the concentration increase of the inorganic anions is entirely dependent on the concentration increase of Cl-. Because the density contribution of Br- is larger than that of Cl-, the density increase of the dilute phase in ATPS-C with bNaCl is larger than that of the whole ATPS-C, and therefore, it is also larger than the density increase of the concentrated phase in ATPS-C. That is why phase inversion can be observed as bNaCl increases further. When the other three kinds of inorganic sodium salts aqueous solutions are used as solvents, the situation is similar. 4. Discussion 4.1. Position of the Aqueous Two-Phase Region in Phase Diagrams. The results in Figures 2, 4-6, and 8 illustrate that the position of the aqueous two-phase regions in phase diagrams are different for different solvents. The phenomenon is analogous to the case in CTAB/AS mixed systems.45 When sodium salt aqueous solution substitutes water as solvent, the aqueous two-phase region shifts to higher MR12-3-12/(12-3-12+AS). Both the shift and the area enlargement of the aqueous two-phase region are strengthened following the Hofmeister series F- < HCO3< Cl- < Br- < NO3-. For dilute surfactant systems, the phase behavior is generally assumed to be mainly characterized by the local aggregate curvature determined by geometrical constraints embodied in the critical packing parameter P ) V/a0lc, where V and lc are the volume and length of the hydrophobic chain, respectively, and a0 is the area per headgroup.46-48 The results in these phase diagrams indicate that the aqueous two-phase region is located between a liquid-crystalline single-phase region (the polarization photographs presented by Figure 18 verify the liquid crystal is lamellar, P ≈ 1) and an isotropic micellar solutions phase region (for example, spherical micelles and rodlike micelles, P < 1/2), meaning that formation of ATPS-C is favorable at a certain P range (1/2 < P < 1). a0 describes the effective headgroup size, which in the case of ionic surfactants or mixed cationic/anionic surfactant systems is largely determined by repulsive electrostatic forces. Addition of salt causes a decrease of the headgroup

repulsions because of electrostatic screening, thus resulting in a decrease in the value a0 and consequently an increase of the critical packing parameter P.29,30 To maintain the P range forming the ATPS-C, both boundaries of the ATPS-C region move toward higher MR12-3-12/(12-3-12+AS). According to Collins’ law of matching water affinity,31,32 the oppositely charged ions with matching water affinity are those which will most readily form inner-sphere ion pairs. NH4+ ions exhibit chaotropic character, and substitution of hydrogen atoms by alkyl chains should not change the chaotropic behavior of the headgroup. For example, N(CH3)4+ ions exhibit strong chaotropic character.49 For the ATPS-C region, cationic surfactant 12-3-12 is in excess, the mixed aggregates are positively charged, and 12-3122+ (i.e., C12H25(CH3)2N+(CH2)3N+(CH3)2C12H25) ions should exhibit chaotropic character also. According to Collins’ concept, 12-3-122+ ions should come in close contact with chaotropic counterions like NO3-, making it more possible to form innersphere ion pairs, whereas kosmotropic F- ions remain further away. Therefore, it is expected that NO3- screens more efficiently the positive charge excess on the aggregates than F-; addition of NaNO3 decreases more strongly the critical packing parameter P of mixed aggregates with certain MR12-3-12/(12-3-12+AS) than that of NaF; the Hofmeister series order F- < HCO3- < Cl- < Br- < NO3- is observed in the shift of the aqueous two-phase regions for 12-3-12/AS/H2O systems with addition of uniunivalent sodium salt. 4.2. Effect of Counterion Binding Competition and Density Contribution of the Counterions on Phase Inversion in ATPS-C. The selective adsorption of counterions at charged surfactant interfaces has been observed in a number of systems, including micelles, vesicles, and air/solution interfaces.50 Competitive binding of two ions A- and B- to oppositely charged aggregates is described by a selectivity coefficient, KAB defined for univalent ions as

KBA )

cAcB cBcA

(1)

where cA and cB are the concentrations of counterions A and B. Quantities with a bar denote the interface phase, and those without denote the bulk solution phase. The literature50,51 indicates that addition of electrolyte and changing of surfactant headgroup type will have little effect on measured selectivities under most conditions. For example,50 the selectivity coefficient of Br- over Cl- is observed to be independent of solution composition, surface excess, and surfactant headgroup type. The

Salt-Induced Phase Inversion selectivity coefficients for the air/solution interface are in fair agreement with those observed in micellar systems. Although the micellar counterion competition experiments differ from the experiments at the air/solution interface in several ways, the origin of selectivity in micellar solutions should be the same as that at the air/solution interface. Warr51 developed a microscopic site-binding ion-exchange model for the binding of counterions to surfactant-coated interfaces. The predictions by the model have been tested against experimental ion selectivities, and the model has been successfully used to explain the dependence of the measured selectivity coefficient on electrolyte concentration and the discrepancies in published data. The plot KAB against ionic strength indicates that KAB is dependent on ionic strength and at high ionic strength (the ionic strength is about 0.10 mol · L-1) KAB has a limiting value. It indicates that at high ionic strength, the value of KAB is a constant. In the work by Bartet et al.52 the relative association degrees of different anions to micelles of hexadecyltrimethylammonium bromide (CTA) were measured by spectrophotometrically determining the amount of p-toluenesulfonate (TOS) or benzenesulfonate (BS) anions desorbed from the micelles by addition of increasing amounts of NaNO3, NaBr, NaC1, and ClNO3FNaF, etc. KBr, KBr, and KBrare 0.044, 0.20, and 1.10, respectively. For cationic micelle, anion affinities seem to follow an order akin to the Hofmeister series.53 Although we did not HCO3know the value of KBr, we can deduce that its value may be between 0.044 and 0.20 according to the order of the Hofmeister series. Therefore, for the mixed 12-3-12/AS aggregates with cationic surfactant 12-3-12 in excess, the degree of counterion binding follows the sequence F- < HCO3- < Cl- < Br- < NO3-. Competitive binding of the ions X- against Br- to positively charged 12-3-12/AS mixed aggregates is in accordance with the following order: F- < HCO3- < Cl- < NO3-. Considering that all the studied 12-3-12/AS mixed systems are at ionic strengths higher than 0.10 mol · kg-1 and addition of electrolyte has little effect on measured selectivities under most conditions,50,51 the values of KXBr-can be considered as constants. For each kind of the above-mentioned uniunivalent sodium salt, in order to investigate the salt effect on phase inversion of ATPS-C, the four ATPS-C chosen for density measurement and composition analysis have the same MR12-3-12/(12-3-12+AS) and different bsalt. For example, when NaF aqueous solution is used as solvent, the MR12-3-12/(12-3-12+AS) of the four ATPS-C is 0.6634 and the values of bNaF are 0.09, 0.11, 0.12, and 0.14 mol · kg-1. For the whole 12-3-12/AS mixed ATPS-C systems, the total surfactant concentration is 0.10 mol · kg-1. For a certain amount of mixed system, the amounts of 12-3-122+, AS-, and Br- are certain; as the concentration of NaF aqueous solution increases, the amounts of F- and Na+ increase. Considering the counterion binding competition of F- against Br- with the positively charged 12-3-12/AS aggregates, the amount of Br-, which is counterion binding with the mixed aggregates, decreases as bNaF increases and part of the Br-, which is originally counterion binding with the mixed aggregates, is replaced by F- and released to the bulk solution phase. Composition analysis indicates that the amounts of surfactant ions 12-3-122+ and ASin the dilute phase of ATPS-C are very low; almost all of the surfactant ions are enriched in the concentrated phase of the ATPS-C. Therefore, the concentrated phase of ATPS-C comes from the coacervation of the mixed aggregates, and the counterion binding competition between F- and Br- with the positively charged 12-3-12/AS aggregates will result in more and more Br- replaced by F- and released to the dilute phase

J. Phys. Chem. B, Vol. 112, No. 39, 2008 12335 of ATPS-C with the increase of bNaF. It results that the molar ratio between the F- counterions and Br- counterions in the concentrated phase of ATPS-C increases. This may be the reason why the concentration of Br- in the dilute phase increases with the increase of bNaF. Meanwhile, the increase of F- in the dilute phase is a little less than the increase of bNaF, and the increase of Na+ in the dilute phase is close to the increase of bNaF. Considering the larger density contribution of Br- than that of F-, the density increase of the dilute phase in ATPS-C is larger than that of the whole ATPS-C, and thus, it is larger than that of the concentrated phase in ATPS-C with the increase of bNaF. Therefore, as bNaF increases further, phase inversion occurs. For the other three kinds of sodium salts NaCl, NaHCO3, and NaNO3 the situation is analogous to the case with NaF aqueous solution as solvent. However, in comparison with F-, these anions present stronger counterion binding competition ability against Br-, and the shift of the aqueous two-phase region is in the order F- < HCO3- < Cl- < NO3-, as discussed in section 4.1. The results in the phase diagraphs (Figures 2 and 4-6) indicate that for a certain kind of sodium salt, such as NaF aqueous solution used as solvent, phase inversion of ATPS-C with higher MR12-3-12/(12-3-12+AS) was observed at higher bsalt. For different kinds of sodium salts, three factors play important roles in the phase inversion of ATPS-C. The first is that part of the bromide counterions, which are the natural counterions of the surfactant 12-3-12 in excess, are exchanged by the X- counterions when NaX is added to the 12-3-12/AS/ water system. The second is that the density of NaX aqueous solution is less than that of NaBr aqueous solution with the same concentration. All of the above four kinds of sodium salts satisfy this condition. The third is the counterion binding competition ability of the anion against Br-. The smaller density of NaX aqueous solution and stronger counterion binding competition ability of the anion are in favor of the phase inversion of the ATPS-C. For example, the density contribution of Cl- is a little smaller than that of F-, and meanwhile, the counterion binding competition ability of Cl- against Br- is stronger than that of F-; therefore, the phase inversion of ATPS-C in NaCl aqueous solution was observed at lower bsalt. The values of bsalt that phase inversion in ATPS-C was originally observed are 0.065 and 0.095 mol · kg-1 in NaCl and NaF aqueous solutions, respectively. The corresponding bsalt values for all of the ATPS-C where phase inversion occurs are g0.095 and g0.135 mol · kg-1 in NaCl and NaF aqueous solutions, respectively. For Cl- and HCO3-, it is analogous to the case for Cl- and F-. For Cl- and NO3-, the density contribution of Cl- is smaller than that of NO3-; however, the counterion binding competition ability of NO3- against Br- is stronger than that of Cl-. From the viewpoint of density contribution, addition of NaCl is more in favor of the phase inversion of ATPS-C, whereas from the viewpoint of counterion binding ability, addition of NaNO3 is more in favor of the phase inversion of ATPS-C. Because of the strong counterion binding ability of NO3-, the shift extent of the aqueous two-phase region for 123-12/AS mixed systems is very large in NaNO3 aqueous solution. For example, when 0.10 mol · kg-1 NaNO3 aqueous solution is used as solvent, the value of MR12-3-12/(12-3-12+AS) for the two boundaries of aqueous two-phase region is 0.715 and 0.824, whereas when water or 0.10 mol · kg-1 NaCl aqueous solution is used as solvent, the corresponding values are 0.630 and 0.654 or 0.661 and 0.737, respectively. Since for ATPS-C with higher MR12-3-12/(12-3-12+AS) phase inversion will be observed at higher bsalt, combination of the above two factors indicates

12336 J. Phys. Chem. B, Vol. 112, No. 39, 2008 that addition of NaCl to the mixed 12-3-12/AS systems is more in favor of the phase inversion of ATPS-C. When NaNO3 aqueous solutions are used as solvents, the phase inversion of ATPS-C was originally observed as bNaNO3 is 0.085 mol · kg-1, and for all the ATPS-C, phase inversion occurs as bNaNO3 g 0.165 mol · kg-1; the two values of bNaNO3 are larger than those with NaCl aqueous solution as solvent. Counterion binding competition ability of the four X- against Br with the positively charged 12-3-12/AS mixed aggregates can be qualitatively verified by the composition analysis results. For example, for systems ATPS-C2 and ATPS-C8 in Table 1, when 0.11 mol · kg-1 NaF aqueous solution is used as solvent, the concentration of Br- in the dilute phase is 0.0591 mol · kg-1. However, when 0.10 mol · kg-1 NaCl aqueous solution is used as solvent, the concentration of Br- in the dilute phase is 0.0652 mol · kg-1. This illustrates that in comparison with 0.11 mol · kg-1 F-, 0.10 mol · kg-1 Cl- has a stronger counterion binding competition ability against Br-; thus, both the amount of Br- (originally counterion binding with the mixed aggregates) substituted by Cland the amount of Br- in the dilute phase are larger. When 0.12 mol · kg-1 sodium salt aqueous solutions are used as solvents, the concentrations of Br- in the corresponding dilute phases of ATPS-C are 0.0619, 0.0667, and 0.0901 mol · kg-1, respectively, for the three kinds anions F-, HCO3-, and NO3-. This indicates that the competitive binding of the ions X- against Br- to positive charged 12-3-12/AS mixed aggregates follows the order F- < HCO3- < NO3-. This sequence is in accord with that by the selectivity coefficients. From above discussion we may conclude that the phase inversion in ATPS-C depends on the density contributions of the various ions in the concentrated phase and in the dilute phase. If the sum of the density contributions of the various ions in the concentrated phase is larger than that in the dilute phase, the concentrated phase in ATPS-C is the bottom one and, on the contrary, the concentrated phase is the top one. Therefore, the factors which can influence the relative magnitude of the densities in the concentrated phase and in the dilute phase of ATPS-C may cause phase inversion of ATPS-C. In this paper, addition of some inorganic sodium salts such as NaF, NaCl, NaHCO3, and NaNO3 into the mixed 12-3-12/AS/water systems can cause phase inversion of ATPS-C. Now, the phenomenon that the concentrated phase is the top one in ATPS formed by CTAB/AS/water systems and the concentrated phase is the bottom one in ATPS-C formed by 123-12/AS/water systems at 318.15 K could be qualitatively explained as follows. The difference between the two systems is that the cationic surfactant used is different: one system is CTAB and the other is 12-3-12. Different from 12-3-12, there is one Br- ion in one CTAB molecule; the density of CTAB aqueous solution is only a little larger than that of water, for example, the density of water at 318.15 K is 0.990216 g · cm-3;35 our experimental results for the densities for 0.05 and 0.10 mol · kg-1 CTAB aqueous solutions at 318.15 K are 0.99034 and 0.99043 g · cm-3, respectively. These results indicate that the density contribution of CTAB is near to zero; considering the larger density contribution of Br-, this means that the density contribution of CTA+ is negative. Composition analysis21,23 indicates that the concentration of Brin the concentrated phase of ATPS-C is smaller than that of CTA+, and the concentration of Br- in the dilute phase of ATPS-C is lager than that of CTA+. Therefore, the density of the concentrated phase is lower than that of the dilute phase, and the concentrated phase in the ATPS-C formed by CTAB/AS/water systems is the top one. Combined with the discussion in section 3.3 about the ATPS-C formed by 12-3-12/AS/water systems, we can conclude

Nan and Hao that the decisive factor of the concentrated phase in ATPS is the top one or the bottom one depends on the density contributions of the various ions. 5. Conclusions In this paper, phase inversion in ATPS-C formed by 0.10 mol · kg-1 12-3-12/AS mixtures was investigated. Experimental results illustrate that for 12-3-12/AS/water mixtures at 318.15 K addition of NaF, NaCl, NaHCO3, or NaNO3 can lead to phase inversion in ATPS-C; however, addition of NaBr cannot lead to phase inversion in ATPS-C. TEM experimental results indicates that vesicles and fusion of vesicles were observed in the concentrated phase of ATPS-C formed by 12-3-12/AS at different solvents whether the concentrated phase is the top one or the bottom one. No direct relationship between the microstructures of the concentrated phase in ATPS-C and the phase inversion was found. Composition analysis results indicate that the concentrations of the surfactants in the dilute phases of ATPS-C are very low; the dilute phases are almost composed of inorganic compounds when inorganic sodium salt aqueous solutions are used as solvents, and almost all of the surfactant ions are enriched in the concentrated phases of ATPS-C. The concentrated phase of ATPS-C comes from the coacervation of the mixed aggregates. For the four kinds of additional sodium salts NaX (i.e., NaF, NaCl, NaHCO3, and NaNO3) three factors play an important role in the phase inversion of ATPS-C. The first is that part of the bromide counterions, which are the natural counterions of the surfactant 12-3-12 in excess, is exchanged by the Xcounterions when NaX is added to the 12-3-12/AS/water system. The second is that the density of NaX aqueous solution is less than that of NaBr aqueous solution with the same concentration. The third is the counterion binding competition ability of the X- anion against Br-. For the mixed 12-3-12/AS aggregates with cationic surfactant 12-3-12 in excess, the degree of counterion binding with the chaotropic headgroups follows the order of Hofmeister series: F- < HCO3- < Cl- < Br- < NO3-. The exchange of Br- by NO3- is more pronounced because NO3the selectivity coefficient KBris larger than 1. The exchange of Br by F or HCO3 or Cl- is also detected, although the exchange is unfavorable since KXBr-is smaller than 1. The smaller density of NaX aqueous solution and the stronger counterion binding competition of X- against Br- with the positively charged 12-3-12/AS mixed aggregates are in favor of the phase inversion in ATPS-C. Acknowledgment. We gratefully acknowledge the National Natural Science Foundation of China (20676029) for financial support of this project. Supporting Information Available: Table showing concentrations of Br- in the dilute phase of some ATPS-C with mT ) 0.10 mol · kg-1 at different sodium salt aqueous solutions measured by the Mohr method and ion-selective electrode method at 318.15 K. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Albertsson, P-Å. Nature 1956, 177, 771. (2) Albertsson, P.-Å. Partitioning of cell particles and macromolecules, 3rd ed.; John Wiley and Sons: New York, 1986. (3) Haghtalab, A.; Mokhtarani, B.; Maurer, G. J. Chem. Eng. Data 2003, 48, 1170. (4) Griffin, S. T.; Dilip, M.; Spear, S. K.; Huddleston, J. G.; Rogers, R. D. J. Chromatogr. B 2006, 844, 23. (5) Long, M. S.; Keating, C. D. Anal. Chem. 2006, 78, 379.

Salt-Induced Phase Inversion (6) Xiao, J.-X.; Sivars, U.; Tjerneld, F. J. Chromatogr. B 2000, 743, 327. (7) da Silva, L. H. M.; da Silva, M. C. H.; de Aquino, R. A. N.; Francisco, K. R.; Cardoso, M. V. C.; Minim, L. A.; Coimbra, J. S. R. J. Phys. Chem. B 2006, 110, 23540. (8) Eiteman, M. A. J. Chromatogr. A 1994, 668, 13. (9) Zafarani-Moattar, M. T.; Sadeghi, R. J. Chem. Eng. Data 2004, 49, 297. (10) Pereira, M.; Wu, Y.-T.; Vena`ncio, A.; Teixeira, J. Biochem. Eng. J. 2003, 15, 1318. (11) Zafarani-Moattar, M. T.; Sadeghi, R. Fluid Phase Equilib. 2005, 238, 129. (12) Sadeghi, R.; Rafiei, H. R.; Motamedi, M. Theromochim. Acta 2006, 451, 163. (13) Huddleston, J.; Veide, A.; Ko¨hler, K.; Flanagan, J.; Enfors, S. O.; Lyddiatt, A. Trends Biotechnol. 1991, 9 (1), 381. (14) Lucassen-Reynders, E. H.; Lucassen, J.; Gile, D. J. Colloid Interface Sci. 1981, 81, 150. (15) Li, G.-Z.; Li, F.; Zheng, L.-Q.; Wang, H.-L. Colloids Surf. A: Physicochem. Eng. Aspects 1993, 76, 257. (16) Hao, J.; Hoffmann, H. Curr. Opin. Colloid Interface Sci. 2004, 9, 279. (17) Sˇegota, S.; Tezˇak. AdV. Colloid Interface Sci 2006, 121, 51. (18) Mao, M.; Huang, J.; Zhu, B.; Ye, J. J. Phys. Chem. B 2002, 106, 219. (19) Shang, Y. Z.; Liu, H. L.; Hu, Y.; Prausnitz, J. M. Colloids Surf. A: Physicochem. Eng. Aspects 2007, 294, 203. (20) Yatcilla, M. T.; Herrington, K. L.; Brasher, L. L.; Kaler, E. W.; Chiruvolu, S.; Zasadzinski, J. A. J. Phys. Chem. B 1996, 100, 5874. (21) Nan, Y.; Liu, H.; Hu, Y. Colloids Surf. A: Physicochem. Eng. Aspects 2005, 269, 101. (22) Nan, Y.-Q.; Liu, H.-L.; Hu, Y. Colloids Surf. A: Physicochem. Eng. Aspects 2006, 277, 230. (23) Nan, Y.-Q.; Liu, H.-L.; Hu, Y. J. Colloid Interface Sci. 2006, 293, 464. (24) Nan, Y.-Q.; Liu, H.-L.; Hu, Y. J. Dispersion Sci. Technol. 2007, 28, 255. (25) Brasher, L. L.; Herrington, K. L.; Kaler, E. W. Langmuir 1995, 11, 4267. (26) Leontidis, E. Curr. Opin. Colloid Interface Sci. 2002, 7, 81. (27) Aroti, A.; Leontidis, E.; Dubois, M.; Zemb, T.; Brezesinski, G. Colloids Surf. A: Physicochem. Eng. Aspects 2007, 303, 144. (28) Hao, L. S.; Nan, Y. Q.; Liu, H. L.; Hu, Y. J. Dispersion Sci. Technol. 2006, 27 (2), 271.

J. Phys. Chem. B, Vol. 112, No. 39, 2008 12337 (29) Renoncourt, A.; Vlachy, N.; Bauduin, P.; Drechsler, M.; Touraud, D.; Verbavatz, J.-M.; Dubois, M.; Kunz, W.; Ninham, B. W. Langmuir 2007, 23, 2376. (30) Vlachy, N.; Drechsler, M.; Verbavatz, J.-M.; Touraud, D.; Kunz, W. J. Colloid Interface Sci. 2008, 319, 542. (31) Collins, K. D. Methods 2004, 34, 300. (32) Collins, K. D. BioPhys. Chem. 2006, 119 (3), 271. (33) Shang, Y.; Liu, H.; Hu, Y.; Prausnitz, J. M. Colloids Surf. A: Physicochem. Eng. Aspects 2007, 302, 58. (34) Zana, R.; Benrraou, M.; Rueff, R. Langmuir 1991, 7, 1072. (35) Isono, T. J. Chem. Eng. Data 1984, 29, 45. (36) Buhler, E.; Mendes, E.; Boltenhagen, P.; Munch, J. P.; Zana, R.; Candau, S. J. Langmuir 1997, 13, 3096. (37) Lu, T.; Han, F.; Mao, G.; Lin, G.; Huang, J. B.; Huang, X.; Wang, Y.; Fu, H. Langmuir 2007, 23, 2932. (38) Lalibuerte´, M.; Cooper, W. E. A model for calculating the density of aqueous electrolyte solutions. J. Chem. Eng. Data 2004, 49, 1141 (Supporting Information). (39) Millero, F. J.; Ward, G. K.; Chetirkin, P. V. J. Acoust. Soc. Am. 1977, 61, 1492. (40) Millero, F. J.; Knox, J. H. J. Chem. Eng. Data 1973, 18, 407. (41) Isono, T. Rikagaku Kenkyusho Hokoku 1980, 56, 103. (42) Goncalves, F.; Kestin, J. Ber. Bunsen-Ges. Phys. Chem. 1977, 81, 1156. (43) Hershey, J. P.; Sotolongo, S.; Millero, F. J. J. Sol. Chem. 1983, 12, 233. (44) Graciaa, A.; Ben Ghoulam, M.; Marion, G.; Lachaise, J. J. Phys. Chem. 1989, 93, 4167. (45) Hao, L.; Nan, Y.; Liu, H.; Hu, Y. J. Dispersion Sci. Technol. 2006, 27, 271. (46) Tanford, C. The Hydrophobic Effect: Wiley: New York, 1980. (47) Israelachvili, J. N.; Mitchell, D. J.; Ninham, B. W. J. Chem. Soc., Faraday Trans. II 1976, 72, 1525. (48) Israelachvili, J. N. Intermolecular and Surface Forces; Academic Press: New York, 1985. (49) Lo´pez-Leo´n, T.; Jo´dar-Reyes, A. B.; Ortega-Vinuesa, J. L.; BastosGonza´lez, D. J. Colloid Interface Sci. 2005, 284, 139. (50) Morgan, J. D.; Napper, D. H.; Warr, G. G.; Nicol, S. K. Langmuir 1994, 10, 797. (51) Warr, G. G. Langmuir 1997, 13, 1451. (52) Bartet, D.; Gamboa, C.; Sepu´lveda, L. J. Phys. Chem. 1980, 84, 272. (53) Blasko´, A.; Bunton, C. A.; Cericbelli, G.; McKenzie, D. C. J. Phys. Chem. 1993, 97, 11324.

JP805126J