Oil Adsorption of Some Surfactants - The Journal

Sep 4, 2008 - Laboratoire des IMRCP, UMR au CNRS N° 5623, Service Commun d'HPLC, FR 2599, and Service Commun de Spectroscopie IR et Raman, FR 2599, U...
1 downloads 13 Views 759KB Size
Download PDF
12318

J. Phys. Chem. B 2008, 112, 12318–12325

Partition and Water/Oil Adsorption of Some Surfactants Rawad Tadmouri,† Chantal Zedde,‡ Corinne Routaboul,§ Jean-Claude Micheau,† and Ve´ronique Pimienta*,† Laboratoire des IMRCP, UMR au CNRS N ° 5623, SerVice Commun d’HPLC, FR 2599, and SerVice Commun de Spectroscopie IR et Raman, FR 2599, UniVersite´ Paul Sabatier, 118, route de Narbonne, F-31062 Toulouse Cedex, France ReceiVed: May 27, 2008; ReVised Manuscript ReceiVed: July 16, 2008

Adsorption isotherms have been determined at the water/oil interface for five biphasic systems involving surfactants (non-ionic and ionic) present in both phases at partition equilibrium. The systems studied were polyoxyethylene(23)lauryl ether (Brij35) in water/hexane and four ionic surfactants, hexadecyltrimethylammonium bromide (CTAB), and a series of three tetraalkylammonium dodecylsulfate (TEADS, TPADS, and TBADS) in water/CH2Cl2. Interfacial tension measurements performed at the water/air and water/oil interfaces provided all the necessary information for the determination of the adsorption parameters by taking partition into account. These measurements also allowed the comparison of the adsorption properties at both interfaces which showed an increase of the adsorption equilibrium constant and a decrease of the maximum surface concentration at the water/oil interface compared to water/air. The values of the critical aggregation concentration showed, in all cases, that only the surfactant dissolved in the aqueous phase contribute to the decrease of the water/oil interfacial tension. In the case of the four ionic surfactants, the critical aggregation concentration obtained in biphasic conditions were lowered because of the formation of mixed surfactant-CH2Cl2 aggregates. Introduction Water, oil, and surfactants are associated in multitude of applications such as oil recovery, food industry, extraction or purification processes, or chemical synthesis. The interest of these ternary mixtures arises from their properties of adsorption, partition, and aggregation. This last aspect is the one that has raised, by far, the largest amount of investigations.1 Concerning adsorption, most of the available experimental data are dedicated to water/air adsorption, whereas a limited number of investigations have been devoted to water/oil interfaces. In these studies, a simplifying assumption was often applied: the surfactant was considered to be only soluble in one of the phases. The organic solvent chosen were nonpolar long-chain alcanes, and when ionic surfactants were used, they were considered to be entirely solubilized in the aqueous phase.2-6 For non-ionic surfactants, solubilization in the aqueous or in the organic phase was assumed depending on the properties of the surfactant and organic phase.7 This way, the authors get rid of the contribution of partition which is, however, a crucial process in important applications as extraction or phase transfer catalysis. Examples of the determination of equilibrium water/oil adsorption parameters (comparable to the ones obtained at the water/air interface) when taking partition into account are much scarce and generally devoted to non-ionic surfactants.8 We report here on the equilibrium properties of biphasic water/oil systems involving surfactants soluble in both phases. The first system involved a non-ionic surfactant: Brij35 in water/ hexane. In the following, four ionic surfactants were used: a cationic one, hexadecyltrimethylammonium bromide (CTAB), and three anionic surfactants, tetraalkylammonium dodecylsul* Corresponding author. E-mail: [email protected]. † UMR au CNRS N° 5623. ‡ Service Commun d’HPLC. § Service Commun de Spectroscopie IR et Raman.

fate (TAADS). A series of increasing chain length of the counterion, tetraethyl (TEA), tetrapropyl (TPA), and tetrabutyl (TBA) ammonium, was used. For the four ionic surfactants, the organic phase was CH2Cl2. In the case of ionic surfactants, a supplementary process, dissociation in water, needs to be taken into account together with partition and adsorption. This chemical process results in a nonlinear distribution of the surfactant as a function of initial concentration.9,10 The method proposed here for the determination of the adsorption properties at the water/oil interface is based on the modeling of interfacial tension measurements obtained at the water/air and water/oil interfaces. In the case of ionic surfactants, the adsorption data also provided information on the effect on the aggregation properties of the saturation of the aqueous phase by the organic solvent. Experimental Section All chemical reagents used were of analytical grade. Polyoxyethylene(23)lauryl ether, Brij35 (Aldrich), cetyltrimethylammonium bromide, CTAB (Aldrich, g 99%), CH2Cl2 (Aldrich, HPLC grade), tetraethylammonium bromide TEAB (Acros Organic, 99+%), tetrapropylammonium bromide TPAB (Acros Organic, 98+%), tetrabutylammonium bromide TBAB (Acros Organics, 99+%), and sodium dodecylsulfate SDS (Prolabo, 98%) were used as received. All the solutions were prepared with ultra-pure water (resistivity >16 MΩ · cm). TAADSs were extracted by CH2Cl2 from an equimolar solution of SDS and TAABs. The organic phase was then washed with water and dried over sodium sulfate. CH2Cl2 was finally evaporated. TEADS (yield 70%) was obtained as a white powder whereas TPADS (yield 90%) and TBADS (yield 90%) were colorless viscous liquids at room temperature. Purity was verified by mass spectroscopy, and sodium and bromides ions could not be detected.

10.1021/jp804674y CCC: $40.75  2008 American Chemical Society Published on Web 09/04/2008

Prtition and Water/Oil Adsorption of Surfactants Interfacial Tension. Equilibrium interfacial tension measurements at the water/CH2Cl2 interface were performed with a pendent drop tensiometer (KRUSS GmbH, model DSA 10MK2, Germany) at room temperature. The size and shape of the drop formed at the tip of a needle fixed on a glass syringe was analyzed by the Drop Shape Analysis software. The diameter of the needle was L ) 1.463 mm for water/air measurements and L ) 0.5 mm for water/oil measurements. The absolute error was estimate to be (0.02 mN.m-1. The value of γ0 obtained for pure water and at the water/oil interface depend on the apparatus adjustments and could vary by about 1 mN · m-1 from one experiment to an other; however, measurements for each compound were performed the same day, and care was taken that the same value was obtained. During an experiment, the blank was repeated several times with pure solvents in order to check any contamination of the syringe by the surfactants. Surface tension was first measured at the water/air interface on a set of samples of concentration C0 ranging, depending on the surfactant used, from 5.10-7 to 2.10-2 mol · L-1. A total of 8 mL of each sample was then put into contact with the same volume of organic phase (hexane or CH2Cl2). The two-phase system was then placed under moderate stirring for 15 h at room temperature until equilibrium was reached. We had verified that equilibrium was in fact completed after 7-8 h. The two phases were then separated. A fraction of the aqueous equilibrated solutions was used for new measurements at the water/air interface for the determination of partition. These fractions were previously submitted to water-saturated air bubbling in order to get rid of the organic solvent solubilized in the continuous phase. Fifteen minutes air bubbling was enough to get reproducible results. The remaining fractions of aqueous solutions were then used as the receiving phase for measurements at the water/ oil interface. ESI-LC-MS. The conditions for the quantitative analysis of TAADSs by ESI-LC-MS are given in refs 9 and 10. HPLC-ELSD. The aqueous and organic solutions of Brij35 and aqueous phases of CTAB obtained after partition were analyzed by using HPLC (Waters 2695) coupled to an evaporative light scattering detector (Waters W2420). The column (Sun Fire 100 × 3 mm, 5 µm) was brought to a fixed temperature of 30 °C. Eluents were 20% water and 80% acetonitrile for Brij35 and 55% of an aqueous solution containing 0.1% of formic acid and 45% acetonitrile for CTAB. The solvent flow rate was held constant at 0.6 mL · min-1. The temperature of the detector was 50 °C, pressure was 35 psi, and gain was fixed to 50. The volume of injected samples varied from 10 to 100 for Brij35 and from 2 to 10 µL for CTAB samples. Infra-Red Spectrophotometry. CTAB organic phases obtained after partition were analyzed by IR (Perkin-Elmer 1760x). CaF2 cuvettes of 197 µm optical path were used. The absorption band chosen corresponded to the asymmetrical CH stretching of CH2 moieties: νCH2,as ) 2927 cm-1. Two sharp bands detected at νOH ) 3686 and 3600 cm-1 were attributed to the OH stretching of water. The shape and position of the bands, which was not modified by the presence of surfactants correspond to free water in two different environments. No broadening of these bands, expect in the presence of aggregates (reverse micelles), in the organic phase was observed. This was also verified in the case of Brij35. Results Non-Ionic Surfactant: Brij35 in Water/Hexane. The method used to determine the water/hexane partition coefficient of Brij35

J. Phys. Chem. B, Vol. 112, No. 39, 2008 12319

Figure 1. Interfacial tension of Brij35 as a function of C0. Water/air: (O) before and (b) after partition; best fit (continuous line) obtained by using Langmuir model for the determination of partition (Table 1, Kp ) 5.3). Water/hexane (0).

TABLE 1: Adsorption Parameters Obtained by Using Langmuir Model for Brij35 at the Water/Air and Water/ Hexane Interface Brij35

Γm/mol · m-2 (Amin/A2)

KL/m3 · mol-1

water/air water/hexane

4.6 × 10-6 (36) 3.3 × 10-6 (55)

4.1 × 102 8.8 × 103

a Minimum area per molecule (Amin) is indicated between brackets.

was proposed by Ravera et al.11 These authors used water/air surface tension measurements as an assaying method to determine the aqueous bulk concentration of a non-ionic surfactant before and after partition. According to this procedure, surface tension measurements, represented by the open circles in Figure 1, were performed on a set of aqueous solutions of concentration C0. The value of the critical micellar concentration (CMC ) 5.7 × 10-5 mol · L-1) was in good agreement with the literature.12 To use the points obtained bellow CMC as a calibration curve, the experimental points had to be reproduced by a suitable function. Any function could have been used for this; however, because specially designed adsorption model are available, it is interesting to use them. They can reproduce very accurately the experimental data and have the advantage to provide physically meaningful parameters. For Brij 35, satisfactory fitting was obtained (Figure 1) by using the ideal Langmuir adsorption model given by eq 1.

γ ) γ0 - RTΓm ln(1 + KLC0)

(1)

where γ is the interfacial tension, γ0 is the value in the absence of surfactant, R is the gas law constant, T is the absolute temperature, Γm is the maximum surface concentration, and KL is the equilibrium adsorption constant. The values of Γm and KL obtained are gathered in Table 1. The filled circles in Figure 1 were measured on the same set of solutions after contact with an equal volume of hexane. This curve lies above the previous one because of the decrease of the concentration of surfactant in the aqueous phase after partition. The curve was fitted by using eq 2,

12320 J. Phys. Chem. B, Vol. 112, No. 39, 2008

γ ) γ0 - RTΓm ln(1 + KLCAQ)

Tadmouri et al.

(2)

for which the Γm and KL were fixed to the values determined before partition (Table 1). CAQ, the concentration in the aqueous phase at partition equilibrium, can be expressed as a function of C0 by using eq 3.

CAQ)C0 ⁄(1 + Kp)

(3)

Kp, the partition coefficient, is the parameter optimized to reproduce the curve obtained after partition. We found Kp ) 5.3, showing, as expected, a larger affinity of Brij35 for the organic phase. The value of C0 at which the CMC is observed after partition corresponds to a concentration in the aqueous phase of 3.5 × 10-4/6.3 ) 5.6 × 10-5 mol · L-1, in fair agreement with the CMC found before partition. In order to confirm these results and because the use of interfacial tension data restricts the domain of concentration to values lower than CMC, we have extended the domain of analysis by using another analytical method. The results obtained by HPLC-ELSD (see Experimental Section) are plotted in Figure 2. The continuous line, calculated by using the value of Kp obtained by surface tension measurements, confirms the accuracy of the method and, also, that aggregation does not modify partition. Partition is independent of the aggregation state in the aqueous phase, and the aggregation equilibrium is continuously shifted toward the formation of monomers which are themselves submitted to partition. In Figure 1, we have also plotted the interfacial tension measured at the water/oil interface. A correlation clearly appeared between the CMC obtained at the water/air interface after partition and the breakpoint observed at the water/oil interface. Because we have just seen that micellization, occurring in the aqueous phase, does not modify partition, the breakpoint at the water/oil interface had necessarily the same origin as that at the water/air interface: aggregation in the aqueous phase. The possibility that aggregation could occur in the organic phase for the same initial concentration is very improbable; its absence was verified by IR spectroscopy (see Experimental Section).13 The correlation between the water/air and water/oil curves indicates, then, that adsorption at the water/oil interface is representative of what occurs in the aqueous phase; surfactants

Figure 3. Interfacial tension of Brij35 at the water/hexane interface as a function of the concentration in the aqueous phase at partition equilibrium, CAQ. 0, experimental point; continuous line, best fit obtained by using the Langmuir model (see Table 1).

solubized in the organic phase do not contribute to the decrease of the interfacial tension. Therefore, the concentration in the aqueous phase (CAQ) is the one that has to be taken into account for the determination of the adsorption parameters (eq 2). The values obtained by best fit (Figure 3) are reported in Table 1. The maximum concentration of adsorbed surfactant is lower at the water/hexane interface than at the water/air interface. These values correspond to a minimum area per adsorbed molecule, Amin, increasing from 36 A2 at the water/air interface to 55 A2 at the water/oil interface. The adsorption constant is increased by a factor 20 when replacing air by hexane. Ionic Surfactants: CTAB, TEADS, TPADS, and TBADS in Water/CH2Cl2. Partition. The protocol followed to determine partition in the case of ionic surfactants is similar to that for the non-ionic surfactant. Water/air adsorption was first measured on a set of aqueous solutions of concentration C0 (Figure 4, open circles). These calibration curves were fitted by using, this time, the Frumkin isotherm. This model, which takes into account lateral interactions between adsorbed surfactant, improved noticeably the quality of fitting. For a pure ionic surfactant in the absence of added electrolyte, the usual Frumkin model is given by

γ ) γ0 + (2RTΓm)[ln(1 - θ) + aθ2] KFC0 )

Figure 2. Concentration of Brij 35 at partition equilibrium in the aqueous phase (O) and in the organic phase (0) as a function of the initial concentration in the aqueous phase C0. Simulation (continuous line) obtained by using Kp ) 5.3 determined by surface tension measurements. Best fit (dotted line) on the present experimental points. CMC is indicated by the arrow.

θ exp(-2aθ) (1 - θ)

(4) (5)

where θ is the degree of coverage of the surface layer, defined by θ ) Γ/Γm, and KF is the equilibrium adsorption constant. The parameter a accounts for intermolecular interactions. The factor 2, that appears in the equation of state for ionic surfactants, takes into account the adsorption of both the ionic surfactant and its counterion.14 The parameters obtained, together with the values of CMC, are gathered in Table 2. The value of CMC obtained for CTAB is in agreement with the literature. The values obtained for the three TAADS are close to the values reported by Benrraou et al.15 As expected, CMC decreases with increasing chain length and therefore with the hydrophobic character of the surfactant. The maximum concentration at interface, Γm, is slightly higher for CTAB than for the three TAADSs, and similar values are obtained for the three compounds. The corresponding Amin are given between brackets. The adsorption equilibrium constant KF is comparable for all species, showing however an increase with chain length

Prtition and Water/Oil Adsorption of Surfactants

J. Phys. Chem. B, Vol. 112, No. 39, 2008 12321

Figure 4. Surface tension at the water/air interface before (O) and after (b) partition; interfacial tension at the water/CH2Cl2 interface (0) for (A) CTAB, (B) TEADS, (C) TPADS, and (D) TBADS. Best fit obtained by using Frumkin model (continuous lines). The corresponding parameters are given in Tables 2 and 3. Zooms have been added for TPADS and TBADS (inserts).

TABLE 2: Water/Air Adsorption Parameters (Frumkin Model) and CMC of the Four Ionic Surfactants Γm/mol · m

-2

CTAB TEADS TPADS TBADS

2

3

-1

(Amin/Å ) KF/m .mol

3.1 × (53) 2.1 × 10-6 (78) 2.1 × 10-6 (78) 2.0 × 10-6 (84) 10-6

3.7 1.6 2.1 7.1

a

CMC/mol.L-1

1.8 1.1 1.6 1.3

8.9 × 10-4 5.0 × 10-3 2.0 × 10-3 1.6 × 10-3

TABLE 3: Dissociation Constant in Water (Kd), Water/ CH2Cl2 Partition (Kp), and Apparent Partition Constant (Kapp) of CTAB and the Three TAADSsa CTAB TEADS TPADS TBADS

Kdb/mol · m-3

Kp

Kapp /m3 · mol-1

Kapp /m3 · mol-1

50 1,34 0,06

350 1160 5200

73 7 8.7 × 102 8.3 × 104

78c 6.8d 8.2 × 102d 7.8 × 104d

a The values reported in the last column were obtained by using different analytical methods. b From ref 16. c HPLC-ELSD. d LC-MS.

for the TAADS series. The values of a show a deviation from ideal behavior for all compounds because of attractive interactions between the alkyl chains. In Figure 4, water/air surface tension obtained after partition are represented by the filled circles. All the curves obtained at partition equilibrium present a remarkable difference with the previous case. For Brij35, the value of γ at CMC was the same

before and after partition, whereas in this case, γ at CMC is higher after partition. For TPADS, and most of all, for TBADS, the amplitude of the curve gets very small at the scale of the plot. We will come back to this point in the third paragraph of this chapter and focus first on the points measured before aggregation. A zoom on the data obtained for TPADS and TBADS (inserts in Figure 4) shows that the experimental accuracy is sufficient to exploit these data for the determination of partition. As already mentioned, for ionic surfactant, dissociation influences the distribution ratio and needs to be taken into account to model partition. Dissociation could occur in both phases; however, we have shown in ref 9 that for CTAB, this process was negligible in CH2Cl2. This assumption can be reasonably applied to the three TAADS which show lower dissociation than CTAB in the aqueous phase. The processes at equilibrium can be represented by processes [1] and [2] + ABAQ a AAQ + BAQ

ABAQ a ABORG A+ AQ

BAQ

where and are the dissociated ions in water and ABAQ and ABORG are the associated forms which are the species undergoing partition. At equilibrium, the concentrations x1 ) + [AAQ ]eq, y1 ) [ABAQ]eq, and y2 ) [ABORG]eq are linked by the relations

12322 J. Phys. Chem. B, Vol. 112, No. 39, 2008

Tadmouri et al.

x21 Kd ) y1

(6)

y2 y1

(7)

Kp )

Kd being the dissociation constant and Kp being the partition coefficient. The distribution ratio at equilibrium is no longer constant and depends on the initial concentration. Combining eqs 6 and 7 to the equation of conservation of mass, the concentration of each specie can be expressed as a function of the initial concentration in the aqueous phase C0.

1+ x1 ) Kd



y1 )

1 + Kp Kd 2(1 + Kp) 1 + 4C0

C0 - x1 1 + Kp

y2 ) Kpy1 ) Kp

C0 - x1 1 + Kp

(8) (9) (10)

The determination of partition deserves now the knowledge of the dissociation constant in water. For the three TAADS, Kd

values were provided in previous work.16 For CTAB, which is considered to be fully dissociated in water, we could not determine Kd and could only reach the value of the apparent constant Kapp ) Kp/Kd ) y2/x21. The parameters extracted from best fit on the experimental curves (Figure 4) are gathered in Table 3. As expected, for the three TAADS, Kp increases with chain length. If we compare the values of Kapp, CTAB lays between TEADS and TPADS. The values reported in the last column of Table 3 were obtained by other analytical techniques allowing the titration on a larger domain of concentration (except for TEADS for which the domains explored overlap). For CTAB, HPLC coupled to an ELSD was used to assay the aqueous phase, and IR was used for the organic samples (Figure 5). For the three TAADS, the aqueous phase was analyzed by LC-MS (Figure 6). We have chosen to represent these plots by using a linear scale to allow the visualization of the nonlinear evolution of the concentration in the aqueous phase. The continuous lines represent the results calculated by using the parameters obtained by surface tension measurements, confirming the accuracy of the method also in this case. The arrows in Figures 5 and 6 indicate the value of the critical aggregation concentration (that we will call CAC to distinguish it from the CMC observed in

Figure 5. CTAB concentration at partition equilibrium as a function of C0. (A) Aqueous phase; 0, HPLC-ELSD; CAC is indicated by the arrow. Dotted line: best fit on the present experimental points (last column Table 3). (B) Organic phase; O, IR. Continuous line: simulation obtained by using the parameters obtained by surface tension measurements (Table 3).

Figure 6. TAADS concentration at partition equilibrium in the aqueous phase measured by LC-MS. (A) 4, TEADS. (B) 0, TPADS; [, TBADS, as a function of C0. Continuous line: simulation obtained by using the parameters obtained by surface tension measurements (Table 3). Dotted line: best fit on the present experimental points (last column Table 3). CAC are indicated by the arrows.

Prtition and Water/Oil Adsorption of Surfactants

J. Phys. Chem. B, Vol. 112, No. 39, 2008 12323

Figure 7. Water/CH2Cl2 interfacial tension versus CAQ at partition equilibrium. (A) 0, CTAB and O, TPADS; (B) O, TEADS and 0, TBADS. Continuous line: best fit obtained by using Langmuir model (eq 2).

TABLE 4: Adsorption Parameters Obtained by Using Langmuir Model at the Water/Oil Interfacea

CTAB TEADS TPADS TBADS

Γm/mol · m-2 (Amin/Å2)

KL/m3 · mol-1

CAC/mol · L-1 (CMC/CAC)

2.1 × 10-6 (80) 1.8 × 10-6 (92) 1.0 × 10-6 (160) 0.5 × 10-6 (320)

110 16 70 210

1.1 × 10-4 (8) 1.7 × 10-3 (3) 6.2 × 10-5 (30) 5.6 × 10-6 (280)

a Last column: CAC in the biphasic system; ratio CMC in pure water over CAC in the presence of CH2Cl2 in brackets (biphasic system).

pure water). They show, as before for the non-ionic surfactant, that aggregation does not modify partition. Water/Oil Adsorption. As in the previous case, for the nonionic surfactant, the critical concentration at the water/air interface after partition coincides with the break point observed at the water/oil interface. The plateau observed at the water/oil interface shows again the only contribution of the aqueous phase to adsorption. Water/oil measurements have been plotted in Figures 7 as a function of the concentration in the aqueous phase at partition equilibrium (CAQ). At low concentration in the aqueous phase, which is the case after partition, the four surfactants can be considered to be fully dissociated in the aqueous phase, and CAQ ) x1 + y1 ≈ x1. The values of CAC, obtained directly from these graphs, are given in Table 4. The points before aggregation were fitted by using the Langmuir isotherm (eq 2). The values obtained are gathered in Table 4. This time, the use of the Frumkin isotherm did not result in a significant improvement of the quality of fitting, in agreement with previous observations17,18 showing, for ionic surfactant, a non-ideal behavior at the water/air interface and an ideal one at the water/oil interface. This is explained by a better solvation of the hydrophobic tails in the organic phase. Concerning adsorption, the higher value of Γm is, as in the case at the water/air interface, obtained for CTAB. But in this case, this parameter varies significantly for the TAADS series. The corresponding minimum area per adsorbed molecule increases from 92 A2 for TEADS to 320 A2 for TBADS. In the series, the adsorption equilibrium constant KL also increases more significantly. All values are higher than at the water/air interface, however following the same order: TEADS < TPADS < CTAB < TBADS.

Figure 8. CTAB surface tension at the water/air interface before (O), after partition/15 min air bubbling (b), after partition/15 h air bubbling (≤).

In each case, the critical aggregation concentration was decreased. The higher variation was observed for TBADS, the CMC of which was divided by a factor 280 in the biphasic system. The decrease of the CMC is explained by a well-known phenomenon which is the formation of mixed aggregates observed when hydrophobic compound are added to a surfactant solution.19 This effect is, in our case, due to the solubilization of CH2Cl2 in the water phase. The fact that the same critical aggregation concentration was obtained at the water/air interface after partition indicated that although our sample had been submitted to air bubbling for 15 min, which was sufficient to remove the CH2Cl2 solubilized in the continuous phase (see Experimental Section), it was not sufficient to expel to organic molecules solubized in the mixed aggregates. For CTAB, a much longer bubbling (15 h) was performed on the aqueous samples obtained after partition. Surface tension measured in these conditions have been plotted together with previous ones in Figure 8. The new points come to their expected values, and γCMC is restored. Discussion For all the systems analyzed here, Brij35 in water/hexane or the four ionic surfactants in water/ CH2Cl2, adsorption at the water/oil interface is qualitatively comparable. Only surfactants dissolved in the aqueous phase contribute to the decrease of

12324 J. Phys. Chem. B, Vol. 112, No. 39, 2008

Tadmouri et al.

the interfacial tension, which is evidenced by the coincidence of the breakpoints in the γ-versus-log C0 curve at the water/air interface after partition and at the water/oil interface. However, in biphasic systems, all concentrations being coupled by equilibrium relations, the interfacial tension could be expressed as a function of the concentration in the organic phase, CORG. In the case of the non-ionic surfactant, the linear relation between CAQ and CORG gives rise to an equation of state similar to eq 2 in which CAQ is replaced by CORG/Kp. If represented as a function of CORG, fitting on the interfacial tension data provides the value of KL/Kp.20 For ionic surfactants, dissociation in water is coupled to adsorption and partition, and the dependence of CORG on CAQ is no more linear. In the present case, because the concentration of the molecular species can be neglected in water (CAQ ≈ x1), the interfacial tension as a function of CORG is given by eq 11.

γ ) γ0 - RT Γm ln(1 + (KL⁄Kapp1 ⁄ 2)CORG1 ⁄ 2)

(11)

According to eq 11, to obtain the adsorption parameters, the experimental data should be plotted as a function of the square root of CORG. This was verified by our experimental results. If the experimental points were plotted as a function of CORG, the Langmuir isotherm did not provide a satisfactory fitting, and the Frumkin model could reproduce the experimental data but delivering unrealistic values of Γm and of the interaction parameter a (high negative values). When γ was plotted as a function of the square root of CORG, perfect fitting was obtained by using the ideal model, and Γm and KL were identical as when plotted as a function of CAQ, which confirmed the validity of our approach. An other common property of the non-ionic surfactant and the ionic ones concerns the comparison of the adsorption parameters at the water/air and water/oil interfaces. For all compounds, the value of Γm are lower and KL values are higher at the water/oil than at the water/air interface. In the case of non-ionic surfactant, we could not find in the literature data allowing the comparison between the two interfaces. Our results are however comparable to previous studies in which ionic surfactants, although not submitted to partition, are solubilized in an aqueous phase in contact with an organic phase. Investigations devoted to SDS showed an increase of the water/oil adsorption constant compared to water/air. This increase was of a factor 2 when heptane18 was used as the organic solvent and reached a factor 10 with heptadecane.6 The maximum concentration at interface, Γm, was found, as in our case, to be lower at the water/oil interface. To compare this parameter, authors usually prefer to use the corresponding minimum area per molecule (Amin) which is more convenient for a description at the molecular level. According to Borkwankar et al.,18 Amin was increased by a factor 2 at the water/heptane interface. At the water/hexane interface, Ivanov et al.6 found an increase of a factor 1.3 in agreement with the value obtained by Oh et al.3 Medrzycka et al.4 observed a similar behavior for a series of long chain alkyltrimethylammonium bromide and found an increase of Amin by a factor 1.4 for CTAB at the water/ hexadecane interface, which is comparable to 1.5 found in the present work for the same surfactant. Several interpretation are proposed to explain the difference between Amin at the water/ air and water/oil interfaces. According to Ivanov et al.,6 the balance between the adsorption energy (related to the adsorption constant) and the electrostatic repulsive interactions between the adsorbed surfactants would be at the origin of this difference. At the water/air interface, the electrostatic interactions could overcome the energy of adsorption and some surfactants would

be pushed out of the interfacial plane. The polar heads of these surfactants would penetrate deeper in the electrical double layer, hence decreasing the interaction energy and leading to higher packing and smaller area per molecule. At the water/oil interface, higher adsorption energy would maintain the surfactant in the plane on the same level, increasing direct repulsion between the polar heads and resulting in larger values of Amin. However, this interpretation, based on electrostatic interactions, can only apply to ionic surfactants; it cannot explain the variation of Amin observed for the non-ionic surfactant Brij35. The interpretation, proposed by Medrzycka et al.4 based, this time, on hydrophobic interactions could be suitable for both non-ionic and ionic amphiphiles. For these authors, solvation of the alkyl chains of the surfactant by the organic phase would lead to the intercalation of solvent molecules between the chains, increasing the distance between the adsorbed molecules. This interpretation is, in fact, the same as the one proposed to explain the nonideal behavior of surfactant at the water/air interface (due to attractive interactions of the alkyl chains in air) versus the ideal behavior observed at the water/oil interface (in which these interaction are suppressed by solvation of the chains). This interpretation could apply, in our case, to the differences observed for Brij35, CTAB, and also for TEADS, but it cannot explain the evolution of this parameter when the series of TAADS is considered. For these three surfactants, which only differ by the size of their counterion, the alkyl chain being the same, the evolution of Amin cannot be justified by geometrical considerations. Although for TEADS, the ratio between Amin at the water/oil and water/air interfaces is of 1.2, it gets to 2 for TPADS and reaches almost 4 for TBADS. Interpretations, based on molecular packing at saturated interfaces, do not hold anymore. In this case, Γm is in fact limited by the occurrence of aggregation. The formation of mixed aggregates, determined by bulk properties, is the process limiting the extent of adsorption well before saturation is reached. The variation of Amin between the two interfaces increases in the order TEADS < CTAB < TPADS < TBADS. The same ranking is obtained for the decrease of the critical aggregation concentration (CMC/ CAC in Table 4) in biphasic conditions and also for the variation of Kapp (Table 3). Partition being related to the hydrophobic character of the amphiphile, this property appears to be also determining for the formation of mixed aggregates and, therefore, for the limit of adsorption at the water/oil interface. Conclusion We have shown that interfacial tension measurements performed at the water/air interface before and after partition is a reliable technique for the determination of partition of both nonionic and ionic surfactants. This method is even in some cases, when the concentration in the aqueous phase becomes very low because of high affinity for the organic phase, more precise than heavy analytical techniques. For TBADS, for instance, standard deviation is much more important when using LC-MS than interfacial tension measurements. Our measurement also showed a privileged contribution of the aqueous phase to the water/oil adsorption. This property, which is valid for the five surfactants involved here, cannot immediately be generalized to all surfactants in biphasic systems; it can depend on both the nature of the solvent and surfactant. However, water non-miscible solvents are in most cases not, or slightly, polar, and one can think that when classical surfactant are used, the processes described above could be applicable in numerous cases.

Prtition and Water/Oil Adsorption of Surfactants References and Notes (1) Schwuger, M. J.; Stickdorn, K.; Schomaecker, R. Chem. ReV. 1995, 95, 849. (2) Gillap, W. R.; Weiner, N. D.; Gibaldi, M. J. Phys. Chem. 1968, 72, 2222. (3) Oh, S. G.; Shah, D. O. J. Phys. Chem. 1993, 97, 284. (4) Medrzycka, K.; Zwierzykowski, W. J. Colloid Interface Sci. 2000, 230, 67. (5) Volkova-Gugeshashvili, M. I.; Volkov, A. G.; Markin, V. S. Russ. Electrochem. 2006, 42, 1073. (6) Ivanov, I. B.; Ananthapadmanabhan, K. P.; Lips, A. AdV. Colloid Interface Sci. 2006, 123-126, 189. (7) (a) Peltonen, L.; Hirvonen, J.; Yliruusi, J. J. Colloid Interface Sci. 2001, 240, 272. (b) Abe´cassis, B.; Testard, F.; Zemb, Th.; Berthon, L.; Madic, C. Langmuir 2003, 19, 6638. (c) Di Lorenzo, M.; Vinagre, H. T. M.; Joseph, D. D. Colloid Surf. A 2001, 180, 121. (8) Ferrari, M.; Liggieri, L.; Ravera, F. J. Phys. Chem. B 1998, 102, 10521. (9) Pradines, V.; Despoux, S.; Claparols, C.; Martins, N.; Micheau, J. C.; Lavabre, D.; Pimienta, V. J. Phys. Org. Chem. 2006, 19, 350.

J. Phys. Chem. B, Vol. 112, No. 39, 2008 12325 (10) Pradines, V.; Tadmouri, R.; Lavabre, D.; Micheau, J. C.; Pimienta, V. Langmuir 2007, 23, 11664. (11) Ravera, F.; Ferrari, M.; Liggieri, L.; Miller, R.; Passerone, A. Langmuir 1997, 13, 4817. (12) Patist, A.; Bhagwat, S. S.; Penfield, K. W.; Aikens, P.; Shah, D. O. Surf. Deterg. 2000, 3, 53. (13) Quan, L.; Weng, S.; Wu, J.; Zhou, N. J. Phys. Chem. B 1998, 102, 3168. (14) Ravera, F.; Ferrari, M.; Liggieri, L. AdV. Colloid Interface Sci. 2000, 88, 129. (15) Benrraou, M.; Bales, L.; Zana, R. J. Phys. Chem. B 2003, 107, 13432. (16) Pradines, V.; Lavabre, D.; Micheau, J. C.; Pimienta, V. Langmuir 2005, 21, 11167. (17) Lucassen-Reynders, E. H. J. Phys. Chem. 1966, 70, 1777. (18) Borwankar, R. P.; Wasan, D. T. Chem. Eng. Sci. 1988, 43, 1323. (19) Solubilization in surfactant aggregates; Christian, S. D., Scamehorn, J. E., Eds.; Marcel Dekker, INC.: New York and Basel, 1995. (20) Miller, R.; Kretzschmar, G. AdV. Colloid Interface Sci. 1991, 37, 97.

JP804674Y