Disjoining Pressure Study of Formamide Foam ... - ACS Publications

Mar 10, 2010 - The resulting Π-h curves were fitted with the DLVO theory from which we extracted surface charge densities q0 and surface potentials Î...
1 downloads 0 Views 1MB Size
Download PDF
pubs.acs.org/Langmuir © 2010 American Chemical Society

Disjoining Pressure Study of Formamide Foam Films Stabilized by Surfactants G. Andersson,*,† E. Carey,‡ and C. Stubenrauch‡,§ †

School of Chemistry, Physics and Earth Science, Flinders University, Adelaide, Australia, ‡School of Chemical and Bioprocess Engineering, University College Dublin, Belfield, Dublin 4, Ireland, and §Institut f€ ur Physikalische Chemie, Universit€ at Stuttgart, Pfaffenwaldring 55, 70569 Stuttgart, Germany Received December 1, 2009

A thin film pressure balance was used to investigate the disjoining pressure Π as a function of the film thickness h of surfactant-stabilized formamide foam films. Nonionic (alkylpoly(ethylene glycol)s) and cationic surfactant (alkyltrimethylammonium bromides (CnTAB) with n = 14 and 16) solutions were studied in the absence and presence of electrolyte. The resulting Π-h curves were fitted with the DLVO theory from which we extracted surface charge densities q0 and surface potentials Ψ0. Investigating formamide foam films is of interest for studying the electrostatic component of the stabilizing forces in foam films. We know that the aqueous foam films are stabilized via electrostatic forces. In this case the self-dissociation of water contributes to the charges in the foam film. As formamide has a dissociation constant which is about 2 orders of magnitude lower than that of water, the number of charges in the solution due to self-dissociation is much smaller, which, in turn, should lead to lower electrostatic forces. Indeed, we found that formamide solutions of nonionic surfactants did not form stable foam films at concentrations below the critical micelle concentration. Regarding the cationic surfactants, the main difference between the formamide and the aqueous foam films is the fact that the concentration of ionic surfactants to form stable foam films is about 2 orders of magnitude higher compared to water. Consequently, the screening length for the electrostatic interaction and thus the film thickness are much smaller compared to films formed by the respective aqueous solutions.

1. Introduction Surfactant-stabilized foam films are liquid systems under confinement in which the two surfaces of the film come into close proximity. If the films are thinner than about 100 nm, the film’s surfaces interact.1 The observed so-called equilibrium thickness of foam films h reflects the balance between an externally applied pressure and the disjoining pressure Π, which is a sum of various surface forces. Thus, foam films are an important tool to study surface and interfacial forces. The disjoining pressure in a foam film acts perpendicular to the film surfaces and is understood to mainly consist of electrostatic Πelec, van der Waals ΠvdW, and steric forces Πsteric.2 While the electrostatic and steric forces are repulsive, the van der Waals force is attractive in the case of a symmetric foam film (air-liquid-air). It holds for the disjoining pressure Π ¼ Πelec þ ΠvdW þ Πsteric

ð1Þ

For a 1:1 electrolyte, assuming small potentials in the middle of the film and very little overlap between the double layers, the electrostatic component can be calculated by3  Πelec ¼ 64RTc tanh

2

Fψ0 4RT

 expð -KhÞ

ð2Þ

where T is the temperature, c the electrolyte concentration, F the Faraday constant, ψ0 the interfacial potential, κ-1 the Debye *Corresponding author. (1) Exerowa, D.; Kruglyakov, P. M. Foam and Foam Films - Theory, Experiment, Application; Elsevier: Amsterdam, 1998. (2) Stubenrauch, C.; von Klitzing, R. J. Phys.: Condens. Matter 2003, 15, R1197. (3) Israelachvili, J. Intermolecular and Surface Forces, 2nd ed.; Academic Press: New York, 1991.

7752 DOI: 10.1021/la100586h

length, and h the film thickness. The van der Waals component is given by3 A ΠvdW ¼ 6πh3

ð3Þ

where A is the Hamaker constant which takes into account the dispersion force. The thin film pressure balance (TFPB) is a method to measure the relation between the force (or pressure) in foam films and their thickness,1,2 and this Π-h relation is usually fitted with the Derjaguin-Landau-Verwey-Overbeek (DLVO) theory. Note that in the case of nonionic surfactants an equation of state similar to the van der Waals equation of state4 has been developed to model the Π-h relation. Of special interest is the contribution of the electrostatic force to the disjoining pressure in foam films. The electrostatic force can be determined from the electrostatic potential and the charge distribution in the film. The charge distribution used in the DLVO theory is based on the double-layer model and the PoissonBoltzmann equation. It is important to note that the PoissonBoltzmann equation is based on first principles of the electrostatic forces, while the double layer model is still unproven and quite contentious. Charges in foam films can stem not only from the dissociation of the solute (e.g., electrolytes or ionic surfactants) but also from the self-dissociation of the solvent. In foam films which are stabilized by nonionic surfactants (in the absence of electrolyte) it is only the self-dissociation of the solvent which can be the origin of charges. The charge distribution at liquid surfaces has attracted large interest in the past decade, and it has been investigated experimentally, with computer simulations and theoretical calculations.5-12 (4) Stubenrauch, C.; Strey, R. J. Phys. Chem. B 2005, 109, 19798. (5) Jungwirth, P.; Tobias, D. J. J. Phys. Chem. B 2001, 105, 10468.

Published on Web 03/10/2010

Langmuir 2010, 26(11), 7752–7760

Andersson et al.

Article Table 1. Properties of Formamide and Watera

solvent

F/g mL-1 [20 °C]

η/Pa s [25 °C]

σ/mN m-1 [25 °C] pKa [25 °C]

n [20 °C]

ε [25 °C]

A/Jb

formamide (HCONH2) 1.1334 (ref 33) 3.343  10-3 (ref 33) 57.0 (ref 33) 16.8 (ref 34) 1.447 (ref 33) 109.6 (ref 28) 6.1  10-20 (ref 3) 0.998 (ref 33) 0.890  10-3 (ref 33) 72.0 (ref 33) 14 (ref 33) 1.332 (ref 33) 78.5 (ref 33) 3.7  10-20 (ref 3) water (H2O) a Structure, density F, viscosity η, surface tension σ, self-dissociation constant pKa, refractive index n, dielectric constant ε, and Hamaker constant A. b No temperature given.

The references given here are a small selection of the large number of publications in this field. Disagreement prevails especially about the sign of the charge at liquid surfaces. Unfortunately, the sign of the charge at the surface cannot be directly determined with the TFPB. However, measurements with wetting films13 as well as with mixtures of nonionic and ionic surfactants14,15 clearly showed that the surfaces of foam films stabilized by nonionic surfactants are negatively charged. Moreover, it is known that foam films stabilized by cationic surfactants are negatively charged at very low and positively charged at higher surfactant concentrations,16 while film surfaces of anionic surfactants are always negatively charged. Using a solvent with a dissociation constant lower than that of water, like for example formamide, one expects that the number of charges in this solution is smaller compared to the respective aqueous solution. Thus, comparing foam films of aqueous and formamide solutions, one should learn more about the contribution of charges originating from the self-dissociation of the solvent to the electrostatic forces in foam films. The properties of formamide show similarities to water but have also distinct differences making formamide interesting for studies on the influence of the solvent on surface properties of solutions. The properties of formamide and water are compared in Table 1. Like water, formamide is a polar and protic solvent with a high surface tension. However, formamide forms a one-dimensional17 while water forms a three-dimensional hydrogen network. Moreover, formamide has a dissociation constant of about 2 orders of magnitude lower than that of water. Further, formamide has a low vapor pressure which allows using vacuum-based surface science methods like angle-resolved X-ray photoelectron spectroscopy (ARXPS) and neutral impact collision ion scattering spectroscopy (NICISS) to investigate the molecular structure of the surface. Indeed, ARXPS and NICISS have been used to measure the surface excess of the surfactant directly.18-20 In another study with formamide as solvent bubble coalescence was studied in the presence of various electrolytes, and it was found that the ion specifity of cations and anions in bubble coalescence is very (6) Jungwirth, P.; Tobias, D. J. Chem. Rev. 2006, 106, 1259. (7) Garrett, B. Science 2004, 303, 1146. (8) Petersen, P.; Saykally, R. Chem. Phys. Lett. 2004, 397, 51. (9) Marcelja, S. J. Phys. Chem. B 2006, 110, 13062. (10) Kunz, W.; Belloni, L.; Bernard, O.; Ninham, B. W. J. Phys. Chem. B 2004, 108, 2398. (11) Beattie, J. K. Phys. Chem. Chem. Phys. 2008, 10, 330. (12) Vacha, R.; Buch, V.; Milet, A.; Devlin, J. P.; Jungwirth, P. Phys. Chem. Chem. Phys. 2008, 10, 332. (13) Ciunel, K.; Armelin, M.; Findenegg, G. H.; v. Klitzing, R. Langmuir 2005, 21, 4790. (14) Carey, E.; Stubenrauch, C. J. Colloid Interface. Sci. 2010, 343, 314. (15) Buchavzov, N.; Stubenrauch, C. Langmuir 2007, 23, 5315. (16) Kolarov, T.; Yankov, R.; Esipova, N. E.; Exerowa, D.; Zorin, Z. M. Colloid Polym. Sci. 1993, 271, 519. (17) Ohtaki, H.; Katayama, N.; Ozutsumi, K.; Radnai, T. J. Mol. Liq. 2000, 88, 109. (18) Eschen, F.; Heyerhoff, M.; Morgner, H.; Vogt, J. J. Phys.: Condens. Matter 1995, 7, 1961. (19) Andersson, G.; Krebs, T.; Morgner, H. Phys. Chem. Chem. Phys. 2005, 7, 136. (20) Krebs, T.; Andersson, G.; Morgner, H. J. Phys. Chem. B 2006, 110, 24015. (21) Henry, C. L.; Craig, V. S. J. Langmuir 2008, 24, 7979.

Langmuir 2010, 26(11), 7752–7760

similar to that in water.21 However, differences were found for the transition concentration of bubble coalescence. Finally, it was found that both water and formamide show preferential adsorption of polarizable inorganic ions at liquid surfaces.22 As regards the properties of surfactants in water and formamide, the Krafft temperature is in a number of cases higher in formamide than in water (see e.g. ref 23), which will be discussed in more detail in the section about the surface tension measurements. The aim of the work here is to deepen our knowledge about the stabilizing forces in foam films by changing the properties of the solvent in the foam films. Using formamide instead of water as solvent reduces the number of charges in the solution as formamide has a much lower self-dissociation constant than water. It can be expected that the charge density is reduced at the surface of foam films stabilized by nonionic surfactants and thus the electrostatic component of the stabilizing forces as well. As a consequence of reducing the number of charges in the solution due to self-dissociation, we can study whether adding ions differing in their propensity to adsorb at surfaces influences the thickness and stability of the foam films. Sodium iodide is a suitable compound for such a study as sodium ions and iodide are found to be different in their propensity to adsorb at surfaces.5 Further, the higher Hamaker constant of formamide compared to water is expected to destabilize the foam films as the van der Waals forces are attractive forces. In ionic surfactant solutions the number of charges in the solution due to the solute is much larger than due to the selfdissociation of water. Thus, replacing water with formamide will influence foam films stabilized by ionic surfactants differently than foam films stabilized by nonionic surfactants. In the first case it is the separation of anion and cation at the surface of the foam film which is of interest. In this study we examined formamide foam films stabilized both by nonionic and ionic surfactants. We used a TFPB to measure Π-h curves of formamide solutions with various surfactants as solutes. The chosen nonionic surfactants are Brij 30, which is a mixture of C12Ej with an average headgroup size of j = 4, and the pure tetradecyltetraethylene glycol (C14E4). We studied the respective foam films in the absence and in the presence of added electrolyte. A lot of systematic studies with CiEj surfactants have been carried out (although not directly with C14E4 and Brij 30), and thus we will compare these results with those obtained for the formamide foam films. As ionic surfactants two cationic alkyltrimethylammonium bromides (CnTAB) with n = 14 and 16 were chosen. The reason for this choice was the fact that extensive studies have been performed recently on aqueous foam films stabilized by CnTAB,24,25 which allows a direct comparison of aqueous and formamide foam films. Our main goal was to study the influence of the solvent on the properties of foam films in general and on the electrostatic repulsion acting in these films in particular. (22) Andersson, G.; Morgner, H.; Cwiklik, L.; Jungwirth, P. J. Phys. Chem. C 2007, 111, 4379. (23) Rico, I.; Lattes, A. J. Phys. Chem. 1986, 90, 5870. (24) Bergeron, V. Langmuir 1997, 13, 3474. (25) Schulze-Schlarmann, J.; Buchavzov, N.; Stubenrauch, C. Soft Matter 2006, 2, 584.

DOI: 10.1021/la100586h

7753

Article

Andersson et al.

2. Experimental Part 2.1. Materials. The nonionic surfactants tetradecyltetraethylene glycol, C14E4 (purity g99%), and Brij 30, C12Ej whith j = 4 being dominant, were purchased from Fluka and Aldrich, respectively. The cationic surfactants tetradecyltrimethylammonium bromide, C14TAB (purity 99%), and hexadecyltrimethylammonium bromide, C16TAB (purity ∼99%), were obtained from Aldrich. Formamide was received from Aldrich (purity >99%). Further purification of formamide is not required as it was found that formamide of this grade does not contain a considerable amount of impurities that affect the surface composition of cationic solutions.19 All chemicals were used as received. Sodium chloride (NaCl) and sodium iodide (NaI) were obtained from Merck and Aldrich, respectively. Both salts were roasted at 500 °C overnight to drive off organic contaminants. All glassware (except the film holders) was cleaned with deconex UNIVERSAL 11 from Borer Chemie, rinsed thoroughly with Milli-Q water, and dried before use. Film holders were cleaned with hot distilled water, boiled in hot hydrochloric acid (HCl), again heated in hot distilled water, then heated at 450-500 °C for 2 h, and allowed to cool; eventually, at least 0.5 L of hot Milli-Q water was sucked through each disk. This process was seen as an ideal cleaning procedure for cationic surfactant contaminated film holders. The clean films holders were then sucked dry and placed in formamide solution overnight. 2.2. Surface Tension Measurements. The surface tensions were measured as a function of the surfactant concentration by the Du No€ uy ring method, using a STA1 tensiometer from Sinterface Technologies. For each surfactant a 100 mL flask of a stock solution at a concentration close to the maximum solubility was prepared. For each measurement 10-20 mL solution for washing the glassware was taken out, and 20 mL solution was used for the measurement. After each measurement the flask was filled up with formamide. The platinum ring was washed with water and annealed. The diluted solution was used for the next measurement. 2.3. Foamability Tests. In order to investigate the foaming properties, simple hand shaking tests were performed. Samples (4 mL) of either aqueous or formamide surfactant solutions were shaken by hand in a 20 mL stoppered test tube in a reproducible way 10 times for 10 s. All samples were measured at c = cmc bar C16TAB in formamide which was measured at its solubility limit, i.e., at c = 7.2  10-2 M < cmc. The change of the foam volume was then monitored over 20 min. 2.4. Thin Film Pressure Balance. Disjoining pressure versus thickness curves (Π-h curves) for thin free-standing foam films were measured with a thin film pressure balance (TFPB). The Π-h curve is obtained by subjecting a free-standing horizontal liquid foam film to a defined gas pressure and measuring its equilibrium thickness h interferometrically. This method is a unique technique to measure interaction forces, thickness (5100 nm), drainage, and stability of thin foam films. Experimental details have been published elsewhere.26 Briefly, foam films are formed in a special film holder, which is placed in a gastight measuring cell in such a way that the film is exposed to the gas pressure Pg and the free end of the glass tube to the reference pressure Pr. The pressure difference ΔP = Pg - Pr can be adjusted via two syringes and measured by a highly sensitive difference pressure transducer. The Π-h curves are generated by interferometrically measuring the equivalent film thickness heq after applying a fixed pressure in the cell. The “true film thickness” h can be obtained according to the three-layer model where the film is considered as a solvent core of refractive index ns surrounded by two surfactant layers of different refractive index (see ref 26 for details). We use the three-layer model as we want to compare the results of this study with the previously published data of aqueous (26) Stubenrauch, C.; Schlarmann, J.; Strey, R. Phys. Chem. Chem. Phys. 2002, 4, 4504; 2003, 5, 2736 (erratum).

7754 DOI: 10.1021/la100586h

Figure 1. Surface tension σ as a function of bulk surfactant concentration c for formamide solutions of C14TAB, C16TAB, Brij 30, and C14E4. The solid lines represent fits of the experimental data carried out with a polynomial of third order. foam films where the same model was used (e.g., those in ref 26). Before measuring the Π-h curves, the film holder was kept in the surfactant solution overnight in order to establish equilibrium. To make sure that the film holder did not pollute the solution, the surface tension of the solution was measured over a period of 2 h before each experiment. Eventually, the cell was filled with excess solution, the lid of the cell was treated with a commercial antifogging agent, and the experimental setup was assembled and left for at least 3 h to ensure vapor-liquid equilibrium. All Π-h curves were reproduced at least once. All measurements were carried out at room temperature. The measured Π-h curves were fitted with the interaction curves calculated within the framework of the DLVO theory in order to determine the surface charge densities q0. The electrostatic part of the disjoining pressure is given by the excess osmotic pressure of the ions in the midplane of the film.3 The excess osmotic pressure is calculated from the charge density in the midplane by solving the Poisson-Boltzmann equation, using the concept of the constant surface charge boundary condition and assuming the course of the electrostatic potential or counterion distribution in the film.3 For low charge concentrations the counterion distribution can be approximated by eq 2. The van der Waals component of the disjoining pressure was calculated with the Hamaker constant A = 3.7  10-20 J for the air/water/ air system and A = 6.1  10-20 J for the air/formamide/air system.3 To obtain the electrostatic component of the disjoining pressure, the nonlinear Poisson-Boltzmann equation was solved. The calculations were done with the algorithm of Chan et al.27 using constant charge boundary conditions and the theoretical Debye length κ-1.3 Debye lengths were calculated using a dielectric constant of ε = 78.53 and ε = 109.628 for the air/water/air system and air/formamide/air system, respectively. The parameter extracted from these calculations is the apparent surface potential ψ0, with which the corresponding surface charge density q0 can be calculated using the Grahame equation.3

3. Results and Discussion We investigated in total four different surfactants in formamide as solvent. The chosen nonionic surfactants were Brij 30 (C12Ej with an average j of 4) and the pure tetradecyltetraethylene glycol (C14E4). The used ionic surfactants are tetradecyltrimethylammonium (27) Chan, D. Y.; Pashley, R. M.; White, L. R. J. Colloid Interface Sci. 1980, 77, 283. (28) Hernandez-Luis, F.; Galleguillos, H. R.; Fernandez-Merida, L.; GonzalesDı´ az, O. G. Fluid Phase Equilib. 2009, 275, 116.

Langmuir 2010, 26(11), 7752–7760

Andersson et al.

Article

Table 2. Critical Micelle Concentration cmc, Surface Tension at the cmc σcmc, and Surface Excess Γ of Nonionic Alkylpoly(ethylene glycol) Surfactants CiEj in Both Formamide and Aqueous Solutions formamide cmc/M C10E4

1.2  10-1 (ref 35)

water -1

σcmc/mN m

-6

Γ/10

mol m

33 (ref 35)

C12E4 Brij 30 (1-2)  10-2 a C14E4 1.1  10-2 a a Current study.

27.5 29.2

4.6a 3.4a

bromide (C14TAB) and hexadecyltrimethylammonium bromide (C16TAB). Surface tensions of all systems were measured in order to determine the critical micelle concentration (cmc) and to estimate the surface excess Γ. We measured Π-h curves for all systems at concentrations close to the cmc. For a better understanding of the results we carried out preliminary foaming tests. We found for the two cationic surfactants that the foaming properties (i.e., foamability and foam stability) of the formamide solutions were significantly reduced compared to those of the aqueous solutions. In the case of the two nonionic surfactants we observed just the opposite. We will come back to these results when discussing the surface tension data and the Π-h curves. 3.1. Surface Tension. In Figure 1 the surface tensions σ as a function of the bulk concentration c for all investigated systems are shown. The σ-c curves were measured up to the solubility limit of the surfactant in formamide or beyond the cmc, respectively. For all surfactants the cmc is much higher in the formamide solution than in the respective aqueous solution. A difference of up to 3 orders of magnitude is found (see Tables 2 and 4). Similar to the aqueous solutions, the cmc increases with decreasing length of the alkyl chain. In the case of Brij 30 a significant dip of 3.5 mN m-1 is observed at the cmc as expected for a surfactant mixture. The cmc of C16TAB could not be measured as the solubility limit has been reached. The data in Tables 2 and 4 show that the concentration of the formamide solutions are about 2 orders of magnitude larger than those of the aqueous solutions when comparing data points of about the same surface tension. This means that the solvophobicity of all surfactants used is less in formamide than in water, which, in turn, is directly reflected in the cmc values. In the following we will evaluate and discuss the results for the nonionic and the ionic surfactants. 3.1.1. Nonionic Surfactants CiEj. We have fitted the surface tension (σ-c) curves with polynomials of third order and calculated the respective Gibbs adsorption isotherm, which, in turn, allows for the calculation of the surface coverage at each concentration. The coefficients of the polynomials are given in the Supporting Information. Assuming that the solutions are ideal over the entire concentration range, we carried out the calculations with an activity coefficient of f = 1. A constant activity coefficient seems to be a reasonable choice as the cmc for Brij and C14E4 are low enough (around 10 mM). One has to keep in mind that we cannot prove that f = 1 but that we simply use an assumption which is in line with literature data. For example, measurements of the nonionic surfactant 1-palmitoyl-2-oleoyl-snglycero-3-phosphocholine in the polar solvent 3-hydroxypropionitrile revealed a cmc of 0.28 mM and a constant activity coefficient over the entire concentration range.20 The surface excess Γ of Brij 30 and C14E4 formamide solutions calculated via the Gibbs equation is shown in Figure 2, where Γ is plotted versus the bulk concentration c. The dashed lines indicate the cmc values. For C14E4 the maximum surface excess of the Langmuir 2010, 26(11), 7752–7760

-2

cmc/M

σcmc/mN m-1

Γ/10-6 mol m-2

8.6  10-4 (ref 36) 7.3  10-4 (ref 31) 6.9  10-5 (ref 37) 6.4  10-5 (ref 38) (2-3)  10-5 a 5.8  10-6 (ref 30)

30 (ref 36) 28 (ref 31) 29 (ref 38)

3.3 (ref 36) 3.6 (ref 31) 3.8 (ref 39)

30a 30 (ref 30)

3.7a 4.4 (ref 30)

Figure 2. Surface excess Γ as a function of bulk surfactant concentration c for formamide solutions of Brij 30 and C14E4. The solid lines represent fits of the experimental σ-c curves carried out with a polynomial of third order (see Figure 1). Dashed lines represent cmc values. Note that all fits were carried out assuming an activity coefficient of f = 1.

aqueous solution Γmax,water = 4.4  10-6 mol m-2 is greater than the one in the formamide solution where Γmax,FA = 3.4  10-6 mol m-2 is found (see Table 2). For Brij 30, however, the maximum surface excess of the aqueous solution Γmax,water = 3.7  10-6 mol m-2 is less than the one in the formamide solution where Γmax,FA = 4.6  10-6 mol m-2 is found (see Table 2). A comparison between the two nonionic surfactants reveals that the maximum surface coverage Γmax of C14E4 (3.4  10-6 mol m-2) is lower than that of Brij 30 (4.6  10-6 mol m-2), while the cmc values are similar although the alkyl chain length of C14E4 is larger than that of Brij. Note that in water one observes the opposite ; C14E4 has a higher surface coverage and a lower cmc as expected due to the longer alkyl chain (see for comparison the values for C10E4 and C12E4 in Table 2). We have no explanation yet for the unexpected behavior in formamide. We also investigated the foaming properties of the respective aqueous and formamide solutions. For Brij 30 we found a direct relation between the foaming properties on the one side and the surface tension and surface coverage results on the other side. However, as such a relation could not be found for C14E4, we interpret the foaming results in terms of surfactant depletion in the bulk when generating new surfaces while generating the foam. 3.1.2. Ionic Surfactants CnTAB. In Figure 3A the surface excess Γ of C14TAB and in Figure 3B the surface excess Γ of C16TAB are shown. For each surfactant the data for both formamide and aqueous solutions are shown. The solubility limit of C16TAB in formamide was reached at concentrations below the cmc under the given experimental DOI: 10.1021/la100586h

7755

Article

Andersson et al. Table 3. Volume of Foam (Vfoam) as a Function of Time for Four Surfactants (C14TAB, C16TAB, Brij 30, C14E4) in Formamide (FA) and Watera Vfoam(FA)/mL

Vfoam(H2O)/mL

t/min C14TAB C16TAB Brij 30 C14E4 C14TAB C16TAB Brij 30 C14E4 0 5 10 15 20

31 4 3 2 1

8 3 2 2 1

22 19 18 18 18

15 10 7 6 3

51 49 46 45 45

61 56 50 49 49

5 4 2 2 2

2 1 0 0 0

a All samples were measured at c = cmc apart from the C16TAB/FA solution which was measured at c = 7.2  10-2 M < cmc. The time t = 0 is the time at which the shaking of the test tube was stopped.

Figure 3. Surface excess Γ as a function of bulk surfactant concentration c for C14TAB (A) and C16TAB (B) in formamide solutions. For comparison, the calculation for the aqueous solutions with data taken from ref 24 are also shown. For both surfactants three fits (solid lines) of the experimental σ-c curves were carried out with a polynomial of third order to obtain the shown Γ-c curves, namely fits of (a) the aqueous systems with an activity coefficient of f = 1, (b) the formamide system with f = 1, and (c) the formamide systems with f = variable, as indicated by the right y-axis and in the figure legend. The activity coefficients are those of Bu4PBr solutions in formamide and taken from ref 19.

conditions As is the case for water, one must work above the Krafft temperature Tk to form micelles in formamide. In the case of C16TAB in formamide Tk = 43 °C,23 which is why the cmc cannot be measured at T = 20 °C. By decreasing the alkyl chain length, Tk decreases and thus for C14TAB the formation of micelles at T = 20 °C is possible. As can be seen in Table 4, the cmc in the formamide solutions is about 2 orders of magnitude higher than in the respective aqueous solution. The surface excess as shown in Figure.3 is calculated via the Gibbs equation. When applying the Gibbs equation, the activity coefficients of the substances in the solution have to be known. We used two different assumptions for choosing the activity coefficient. In the first case we assume that the ionic surfactant solutions are ideal with a constant activity coefficient of f = 1 (as was the case for the nonionic surfactant solutions in the section above). In the second case we assume that the solutions are not ideal. In this case we used the activity coefficients of tetrabutylphosphonium bromide (Bu4PBr) in formamide solutions, i.e., of a system with the same solvent and an ionic solute.19 The reason for 7756 DOI: 10.1021/la100586h

carrying out the calculations with different activity coefficient is the observation made in ref 19 where a significant deviation from constant activity coefficients was found. As the molecular structures of Bu4PBr and C14TAB (same number of CH3 and almost the same number of CH2 groups) as well as of Bu4PBr and C16TAB (same number of CH3 and CH2 groups) are similar, the assumption of nonideal C14TAB and C16TAB formamide solutions is reasonable. We are aware that even though the molecular structures are very similar, it does not necessarily mean that the activity coefficients must be the same. However, the reason for using the activity coefficient of Bu4PBr is not to perform an accurate calculation of the surface excess but to give an indication of the possible range of the surface excess of C14TAB and C16TAB formamide solutions. The so-estimated possible range of the surface excess is larger for C14TAB than for C16TAB as the activity coefficient used decreases monotonously with increasing concentration and the concentration of the C14TAB is higher than that of the C16TAB solutions. For comparison, we show the surface excess of the respective aqueous solutions calculated via the Gibbs equation with the data for σ-c curves taken from refs 24 and 29. Looking at Table 4, one sees that 3.6  10-6 mol m-2 > Γmax > 1.8  10-6 mol m-2 for C14TAB in formamide, while Γmax = 3.6  10-6 mol m-2 for the aqueous solution. This result is in line with the foam properties where we found that the formamide foams have a much smaller foamability and foam stability. As regards C16TAB, we are not able to calculate Γmax or to compare the properties with those of the aqueous system due to the solubility limit. However, the same general trend as for C14TAB is expected. Note that the surface coverage given in Table 4 for C16TAB is the one at the solubility limit, which is much smaller than the one expected for a densely packed monolayer. 3.2. Foamability Tests. 3.2.1. Nonionic Surfactants CiEj. For the nonionic surfactants the foamability of the aqueous solutions was lower than that of the formamide solutions and the stability of the generated aqueous foams was lower than that of the formamide foams. Note that this behavior is opposite to that of the CnTAB solutions (see below for details). One would expect that this observation is due to a higher σcmc value (more energy is required to increase the surface = lower foamability) and/or a lower surface coverage Γ (less densely packed monolayer = lower foam stability) of the aqueous solution. Indeed, a look at Table 3, Figure 1, and Figure 4 reveals that in the case of Brij 30 the surface coverage in the aqueous solution is lower compared to the formamide solution, while the σcmc values are about the same. (29) Stubenrauch, C.; Khristov, Kh. J. Colloid Interface Sci. 2005, 286, 710. (30) Islam, M. N.; Kato, T. Langmuir 2003, 19, 7201. (31) Karraker, K.; Radke, C. J. Adv. Colloid Interface Sci. 2002, 96, 231.

Langmuir 2010, 26(11), 7752–7760

Andersson et al.

Article

Table 4. Surface Potential ψ0 and Surface Charge Density q0 from DLVO Fitting of the Π-h Curves of C14TAB and C16TABa solvent

surfactant

c/M

cmc/M

σcmc/mN m-1

ψ0/mV

q0/mC m-2

Γ/10-6 mol m-2

FA* C14TAB 2.1  10-1 2.5  10-1 39.8 27 ( 2 (34 ( 2) 26 ( 3 (45 ( 3) 3.6 (1.8) C14TAB 3.5  10-3 3.5  10-3 38.2 130 46 3.6 H2O24 25 -3 -3 C14TAB 3.5  10 3.5  10 37.7 90 ( 10 20 ( 5 3.5 H2O 5.0  10-2 28 ( 2 (30 ( 2) 17 ( 2 (19 ( 2) 2.0 (1.7) FA** C16TAB 7.2  10-2 29 ( 2 (33 ( 2) 21 ( 2 (26 ( 2) 2.3 (2.1) FA*** C16TAB C16TAB 9.0  10-4 9.0  10-4 36.7 145 32 3.9 H2O24 a Surface excess Γ is obtained from fitting the surface tension isotherm with a third-order polynomial and applying the Gibbs adsorption isotherm. The data listed were obtained by using a constant and variable activity coefficients, respectively. In general, we consider the interpretation of results based on assuming nonideal solutions as the more realistic results due to the high ionic surfactant concentrations that had to be used.Note that for the surface potential ψ0, surface charge density q0, and surface excess Γ two values are given, namely values with brackets obtained for a constant activity coefficient of f = 1 and values without brackets obtained for f 6¼ 1. In the latter case, activity coefficients used are * f = 0.55, ** f = 0.93, and *** f = 0.86.

As regards C14E4, one sees that the surface excess in the formamide solution is lower compared to that of aqueous solution, and the surface tensions at the cmc are again about the same for both solvents. Thus, in this case a more stable aqueous foam is expected. A reasonable explanation for the observation that the formamide foam is more stable could be the very low cmc value of the nonionic surfactant C14E4 in water. If one generates large amounts of surface at these low surfactant concentrations, the bulk solution is depleted, which in turn increases the surface tension. In other words, once a foam is generated in the aqueous solution, we may assume that the σ value is indeed higher and the Γ value lower than for the respective values of the formamide solution, which results in a lower foamability and foam stability of the aqueous solutions. Finally, the higher foamability and higher foam stability of Brij 30 compared to C14E4 formamide solutions mirror the slightly lower σcmc and the higher Γ value of the Brij 30 solution (see Table 3). 3.2.2. Ionic Surfactants CnTAB. As can be seen in Table 3, the foamability and foam stability slightly increase with increasing alkyl chain length in the case of aqueous CnTAB solutions. A nearly negligible influence of the chain length has been reported earlier for C14TAB and C16TAB foams29 and is in line with our very inaccurate “shaking test”. As the solubility of C16TAB in formamide limited our study to c = 7.2  10-2 M < cmc, it was not possible to examine the influence of the alkyl chain length in formamide. Note that both the foamability and foam stability are quite low for the C16TAB formamide solution, which indicates that the chosen concentration (= solubility limit) is well below the cmc, which is also reflected in the σ-c curve (see Figure 1). As regards the influence of the solvent on the foam properties, we learn from the C14TAB data that the foamability of aqueous solutions is larger than that of the formamide solutions and that the stability of the generated aqueous foams is much higher than that of the formamide foams. Both observations can be related to the slightly lower σcmc values and the higher Γmax value of the aqueous solutions (see Table 4). However, a much more detailed foam study is required to understand the different foaming behavior of formamide and aqueous solutions on the one hand and of cationic and nonionic surfactants on the other. Differences in viscosity between water and formamide could influence the foamability as this is a dynamic phenomenon. However, as we found cases were the formamide solutions show higher and cases were the formamide solutions show lower foamability, we assume that the influence on the foamability study shown here is of secondary importance. 3.3. TFPB Data. 3.3.1. Nonionic Surfactants CiEj. In Figure 5 A the Π-h curves for Brij 30 stabilized foam films without and with 5  10-4 NaI are shown. Both foam films have a thickness of ∼6 nm, and their thickness is constant over the entire measured pressure range of up to 8000 Pa. Because of the small and constant thickness, these films have to be considered as Langmuir 2010, 26(11), 7752–7760

Figure 4. Surface tension σ as a function of bulk surfactant concentration c for aqueous Brij 30. The solid line drawn at concentrations below the cmc represents a fit of the experimental data with a polynomial of third order, while the horizontal line guides the eyes for concentrations above the cmc. Note that the cmc is only a rough estimate as Brij 30 is no well-defined surfactant.

Newton Black Films (NBFs). Consequently, the surface charge density and the surface potential are zero. In Figure 5 B the Π-h curves for C14E4-stabilized foam films are shown. The concentration of the surfactant solutions was 30 mM, which corresponds to 2.3cmc. As was the case for Brij 30, the thickness of both films is again ∼6 nm and constant over the entire measured pressure range of up to 8000 Pa. Thus, these films can also be considered as NBFs with zero surface charge density and zero surface potential. At c = 15 mM (c = 0.14cmc) no foam film could be stabilized in the absence of electrolyte, while films with 1, 10, and 100 mM NaI led to stable foam films for P e 5000 Pa. However, the formation of black spots was observed. The black spots remained in the thicker film for several hours, which, in turn, means that these films did not reach an equilibrium thickness. As these foam films are not in equilibrium, they are not depicted in Figure 5. Summarizing the observations at c e cmc, one can say that in contrast to the aqueous solutions the formamide solutions do not form stable common black films (CBFs) below the cmc neither in the absence nor in the presence of electrolyte. The fact that the C14E4 formamide solution forms a stable NBF at c > cmc is in line with the observations made for aqueous foam films stabilized by nonionic surfactants (see refs 1 and 2 and references therein). At c > cmc the surface coverage is obviously sufficiently high to stabilize a formamide foam film although the surface excess of C14E4 in formamide (3.4  10-6 mol m-2) is less than in aqueous solutions (4.4  10-6 mol m-2 30). DOI: 10.1021/la100586h

7757

Article

Figure 5. Disjoining pressure Π as a function of film thickness h for formamide solutions of (A) 3.0  10-2 M C14E4 solutions at different electrolyte concentrations (2, 0 M NaI; 4, 1  10-3 M NaI). Π-h curves of 1.5  10-2 M C14E4 solutions are stable and are very similar to the 3.0  10-2 M C14E4 and hence not shown. (B) Π-h curves of 2.0  10-2 M Brij 30 in formamide without (b) and with (O) electrolyte (5  10-4 M NaI). Note that all measurements are carried out at concentrations above the cmc as indicated in the figure legends. The vertical dashed line represents an NBF of 5 ( 0.5 nm thickness.

Although we cannot compare the formamide C14E4 films with the respective aqueous film as TFPB data for the latter solution have not been published so far, comparison with various other CiEj surfactants such as C12E6, C10E4, C10E6, and C10E8 can be done, which all form CBFs at c < cmc (unsaturated surface).4,31 In all of these systems it was found that the CBFs are stabilized via electrostatic repulsion, and it is argued that in aqueous solutions the self-dissociation of the solvent provides the charges causing the electrostatic stabilization. At c < cmc the presence of electrolyte has an effect on the foam films, and it has be found that increasing the electrolyte concentration reduces the film thickness and may result in the formation of an NBF. However, at c < cmc the formed NBFs are not stable, which simply means that a densely packed monolayer is required to stabilize an NBF which is not the case at c < cmc. Thus, at c < cmc the only stabilizing mechanism of a foam film is an electrostatic repulsion, which is screened at high electrolyte concentrations and thus destabilizes the foam film. Coming back to the formamide solutions, we need to underline that C14E4 formamide foam films are not stable at c e cmc. An explanation for the different behavior in formamide com7758 DOI: 10.1021/la100586h

Andersson et al.

Figure 6. Disjoining pressure Π as a function of film thickness h for one concentration of C14TAB (O, 2.1  10-1 M) and two concentrations of C16TAB (3, 5.0  10-2 M; 2, 7.2  10-2 M) in formamide. The solid lines are calculated according to the DLVO theory from which the surface charge densities q0 of the respective CBFs are obtained. (A) A constant activity coefficient (f = 1) is used for the calculations; i.e., the activity is set equal to the concentration. (B) Activity coefficients as outlined for evaluating the surface tension curves in Figure 3 are used.

pared to water can be the number of charges arising from the self-dissociation of the solvent, which is much lower in formamide than in water. Thus, the lower self-dissociation of formamide may be considered as the reason that no stable films are formed at c e cmc. Adding low amounts of inorganic electrolytes increases the number of charges present in the foam film and could lead to a surface charge if one type of the electrolyte ions preferentially adsorbs at the surface of the foam film. NaI seems to be a suitable salt for this experiment as computer simulations showed charge separation at aqueous5 and at formamide surfaces,22 respectively. However, the presence of NaI in the 7.5 mM and in the 15 mM C14E4 formamide solutions did not lead to stable equilibrium foam films but either to foam films with black spots (i.e., nonequilibrium films) or to foam films with a slightly longer lifetime compared to the salt-free systems. Thus, NaI seems to have indeed a weak stabilizing effect. We assume that the origin of the weak stabilizing effect is the difference in propensity of iodide and sodium ions to adsorb at the film surface as reported in ref 5. Another destabilizing effect is the higher Hamaker constant of formamide compared to water, possibly also the lower surface coverage with surfactant. Langmuir 2010, 26(11), 7752–7760

Andersson et al.

From Table 1 it can be seen that the viscosity of formamide is about 3 times larger than that of water, and the question could arises whether the increase in viscosity influences TFPB measurements. We assume that the only influence of the higher viscosity on the TFPB measurements is the time to reach equilibrium during the TFPB measurements which was found to be longer in case of the formamide films compared to the aqueous films. 3.3.2. Cationic Surfactants CnTAB. In Figure 6 the disjoining pressure Π as a function of film thickness h for two concentrations of C16TAB and one concentration of C14TAB is shown. The C14TAB-stabilized foam film has a concentration of 0.21 M, which corresponds to 0.84cmc. At lower concentrations no stable films were formed. The film has a maximum thickness of ∼6 nm at the lowest applied pressure of 250 Pa. The thickness reduces to ∼5 nm at a pressure of 8000 Pa. The change of the film thickness is small but significant within the error bars. The thickness of the foam film is only 2 times more than twice the length of a C14TAB molecule in the all-trans conformation when assuming 0.15 nm per double methylene group and 0.2 nm for the headgroup. This indicates that the surfactant monolayers of the foam film are in very close contact. Both investigated concentrations of the C16TAB solutions are close to the maximum solubility. From the surface tension measurements we know that at both concentrations no micelles are formed. The C16TAB-stabilized foam film at the higher concentration of 72 mM has a maximum thickness of ∼11 nm at the lowest applied pressure of 250 Pa. The thickness reduces down to ∼8 nm at a pressure of 3000 Pa. The thickness of the film at the lower concentration of 50 mM is slightly larger (ca. ∼1 nm over the entire pressure range). As both the C14TAB and C16TAB foam films are very thin the question arises whether these films are CBFs or NBFs. However, as decreasing film thickness with increasing pressure is characteristic for CBFs, all of studied films have to be considered as CBF. In two other cases thin CBFs have been observed before. For aqueous foam films stabilized by 0.03 M C12TAB (2cmc) film thicknesses ranging from 11 to 14 nm were observed.14 In the case of 0.1 M SDS (11cmc) a CBF was measured for P < 60 kPa with thicknesses less than 15 nm, while an NBF is formed at higher pressures.32 As the C14TAB and C16TAB foam film have to be considered as CBFs, the Π-h curves in Figure 6 are fitted with the DLVO theory as outlined in section 2.3. In the calculation the bulk electrolyte concentration is one of the parameters. In section 3.1.2 we explained why the activity coefficient of the C14TAB and the C16TAB solutions must not necessarily be considered as constant. Therefore, two different fits were used in order to examine the influence of the activity coefficient f on the DLVO fits. In the first case, the bulk surfactant concentration c has been used (i.e., an activity coefficient of f = 1.0 is assumed; see Figure 6A), while in the second case a nonconstant activity coefficient has been used (see Figure 6B). In the latter case the same activity coefficients as (32) Bergeron, V.; Radke, C. J. Langmuir 1992, 8, 3020. (33) Handbook of Chemistry and Physics, 82nd ed.; Lide, D. R., Ed.; CRC Press: Boca Raton, FL, 2001/2002. (34) Manfredi, A.; Ranucci, E.; Suardi, M. J. Bioact. Compat. Polym. 2007, 22, 219. (35) Schubert, K.-V.; Busse, G.; Strey, R.; Kahlweit, M. J. Phys. Chem. 1993, 97, 248. (36) Schlarmann, J.; Stubenrauch, C.; Strey, R. Phys. Chem. Chem. Phys. 2003, 5, 184. (37) Lu, J. R.; Li, Z. X.; Su, T. J.; Thomas, R. K. Langmuir 1993, 9, 2408. (38) Rosen, M. J.; Cohen, A. W.; Dahanyake, M.; Hua, X. J. Phys. Chem. 1982, 86, 541. (39) Schubert, K.-V.; Strey, R.; Kahlweit, M. J. Colloid Interface Sci. 1991, 141, 21.

Langmuir 2010, 26(11), 7752–7760

Article

those used for evaluating the σ-c curves were used (see section 3.1.2). Using the activity instead of the bulk concentration, one calculates with the available amount of solute which should give a more realistic number of available molecules, i.e. ions, in the bulk. Taking f = 1 for the calculations, one assumes an ideal solution and overestimates the number of dissociated molecules. The surface charge densities q0 along with the corresponding surface potentials ψ0, which are a result of fitting the Π-h curves with the DLVO theory, are given in Table 4. It has to be noted that the values for the surface charge and surface potential depend on the model used to fit the Π-h curves; especially the thickness of the solvent layer in the model influences the values. Also shown are the surface excess Γ calculated from the Gibbs equation. In all cases the activity coefficients used for the specific calculation are given. In general, we consider the interpretation of the results based on assuming nonideal solutions as the more realistic results due to the high concentrations of ionic surfactant. For comparison, also the values for the respective aqueous foam films are given. Comparing the formamide foam films stabilized by C14TAB and C16TAB, we find that the surface charge density q0 of the C14TAB films is higher than that of the C16TAB films, while the surface potentials Ψ0 are very similar. This trend is the same as that found for the corresponding aqueous systems24 and simply due to the higher surfactant concentration of the C14TAB systems. Comparing foam films stabilized by the same surfactant, one sees that the formamide films are thinner than the corresponding aqueous films, which is reflected in the lower surface charge densities and the lower surface potentials. When comparing aqueous and formamide films, the dielectric constant most likely does not play an important role, as water and formamide have dielectric constants that differ for only about 20% (see Table 1).28 However, this has to be considered with some caution as the information available about the dielectric constant at the surface of both liquids is still very limited. The finding that the formamide foam films are thinner compared to the respective aqueous foam films can be explained by the higher bulk concentration of the ionic surfactant in the former case, which reduces the screening length for the electrostatic interaction. The same trend was found when increasing the NaCl concentration in aqueous C14TAB solutions.25 In order to compare the surface potential and the surface charge density of the aqueous and the formamide films, we need to consider the underlying model used to fit the Π-h curves and the fitting procedure. The surface potential Ψ0 is determined by fitting the Π-h curves with eqs 1-3. Subsequently, the surface charge density q0 is calculated. Note that q0 is an apparent surface charge density, which models the charge separation at the surface. Thus, the apparent surface charge density is influenced by the surfactant density at the film surface, the separation of the charges, the orientation of the surfactant molecules, and the dielectric constant in the film. In the framework of the applied DLVO theory the lower surface charge density and the lower surface potential of the formamide films could be due to either less charge separation of the anions and the cations or a lower density of surfactant molecules at the surface of the foam films. A clarification about the surface charge of a given system will be brought only by measuring the concentration depth profiles of the charges directly.

4. Conclusion We investigated formamide foam films stabilized by nonionic and ionic surfactants. Formamide foam films stabilized by the DOI: 10.1021/la100586h

7759

Article

nonionic alkylpoly(ethylene glycol)s at concentrations above the cmc were all found to be Newton black films. This is in agreement with the observations made for aqueous films stabilized by nonionic surfactants. However, at concentrations below the cmc no stable foam films are formed, which is different than aqueous solutions where CBFs are formed below the cmc. A possible reason for this observation may be the lower self-dissociation constant of formamide compared to water, which means that less ions are present in the solution. Assuming that these ions are responsible for the surface charge density q0 in foam films stabilized by nonionic surfactants, one expects a lower q0 and thus a lower tendency to stabilize a CBF in formamide compared to aqueous solutions. We assume that another destabilizing effect is the higher Hamaker constant of formamide compared to water, possibly also the lower surface coverage with surfactant. Foam films stabilized by the ionic surfactants CnTAB showed the same general trend as the respective aqueous foam films. First, in all cases CBFs are formed at surfactant concentrations c e cmc. Hence it is concluded that the films are stabilized by electrostatic forces. Second, in the case of C16TAB, the Π-h curve shifts toward lower film thickness, the surface charge density q0 increases, and the slope of the Π-h curves increases with increasing surfactant concentration. This is simply due to the fact that the surfactant itself is an electrolyte, which, in turn, reduces the Debye length and as a consequence reduces the strength of the

7760 DOI: 10.1021/la100586h

Andersson et al.

electrostatic component of the disjoining pressure. A third observation is the increasing stability of the foam films with increasing C16TAB concentration. The pressure at which the films rupture is 1000 Pa for 5.0  10-2 M and increases up to 2500 Pa for 7.2  10-2 M. Unfortunately, a rigorous comparison between C16TAB and C14TAB cannot be made as the cmc of the former in formamide is not known. However, as speculative it may be, it seems as if formamide foam films of C16TAB are thicker at similar c/cmc values as is the case in aqueous solutions.24 Again, this difference is simply due to the different electrolyte concentrations and thus Debye screening lengths. Finally, formamide foam films have a lower surface potential and lower surface charge density compared to the respective aqueous counterparts. This could be due to either less charge separation of the anions and the cations or a lower density of surfactant molecules at the surface of the foam films. Acknowledgment. Gunther Andersson thanks the Australian Academy of Science for funding the visit at the University College Dublin. Enda Carey acknowledges UCD Ad Astra Research Scholarship funding. Supporting Information Available: Surface tension data evaluation. This material is available free of charge via the Internet at http://pubs.acs.org.

Langmuir 2010, 26(11), 7752–7760