pubs.acs.org/Langmuir © 2010 American Chemical Society
Universal Surfactant for Water, Oils, and CO2 )
Azmi Mohamed,† Kieran Trickett,† Swee Yee Chin,‡ Stephen Cummings,† Masanobu Sagisaka,§ Laura Hudson,† Sandrine Nave,† Robert Dyer, Sarah E. Rogers,^ Richard K. Heenan,^ and Julian Eastoe*,†
)
† School of Chemistry, University of Bristol, Cantock’s Close, Bristol BS8 1TS, U.K., ‡School of Chemical Science and Food Technology, Faculty of Science and Technology, University Kebangsaan Malaysia (UKM), 43600 Bangi, Selangor Darul Ehsan, Malaysia, §Department of Frontier Materials Chemistry, Graduate School of Science and Technology, Hirosaki University, 3 Bunkyo-cho, Hirosaki, Aomori 036-8561, Japan, Krυ.ss Surface Science Centre, School of Chemistry, University of Bristol, Cantock’s Close, Bristol BS8 1TS, U.K., and ^ ISIS-STFC, Rutherford Appleton Laboratory, Chilton, Oxon OX11 0QX, U.K.
Received June 7, 2010. Revised Manuscript Received July 21, 2010 A trichain anionic surfactant sodium 1,4-bis(neopentyloxy)-3-(neopentyloxycarbonyl)-1,4-dioxobutane-2-sulfonate (TC14) is shown to aggregate in three different types of solvent: water, heptane, and liquid CO2. Small-angle neutron scattering (SANS) has been used to characterize the surfactant aggregates in water, heptane, and dense CO2. Surface tension measurements, and analyses, show that the addition of a third branched chain to the surfactant structural template is critical for sufficiently lowering the surface energy, tipping the balance between a CO2-incompatible surfactant (AOT) and CO2-philic compounds that will aggregate to form micelles in dense CO2 (TC14). These results highlight TC14 as one of the most adaptable and useful surfactants discovered to date, being compatible with a wide range of solvent types from high dielectric polar solvent water to alkanes with low dielectrics and even being active in the uncooperative and challenging solvent environment of liquid CO2.
Introduction Because of an environmental drive to cut emissions, developing approaches for the utilization of CO2 in commercial applications is a key challenge for researchers. There are already a number of industrial applications of liquid CO2, especially in extraction.1,2 However, CO2 is a weak, nonpolar solvent and has therefore not reached its full potential in terms of commercial uses.1 One way to overcome this problem would be to use CO2-philic surfactants to stabilize reverse micelles and/or microemulsion phases, whereby pools of water would enhance the solubility of solutes that are otherwise insoluble in CO2.3 Previous advances in this area centered on the use of fluorinated surfactants3-8 that are very CO2 soluble but are not ideal for applications, being both expensive and environmentally hazardous.3,9 More recently, some success has been achieved in developing CO2-philic hydrocarbon surfactants including highly branched and oxygenated surfactants.10 The anionic sulfosuccinate, a dichain surfactant AOT4 (sodium bis(3,5,5-trimethyl-1-hexyl) sulfosuccinate), has increased methylation at the chain tips compared to that of common CO2-insoluble AOT (sodium bis-(2-ethyl1-hexyl) sulfosuccinate); this structural switch gives rise to enhanced solubility for AOT4 in CO2 at 500 bar and 33 C.11 *Corresponding author. E-mail:
[email protected]. (1) Beckman, E. J. Ind. Eng. Chem. Res. 2003, 42, 1598. (2) DeSimone, J. M. Science 2002, 297, 799. (3) Eastoe, J.; Gold, S.; Steytler, D. C. Langmuir 2006, 22, 9832. (4) Eastoe, J.; Cazelles, B. M. H.; Steytler, D. C.; Holmes, J. D.; Pitt, A. R.; Wear, T. J.; Heenan, R. K. Langmuir 1997, 13, 6980. (5) Eastoe, J.; Paul, A.; Downer, A.; Steytler, D. C.; Rumsey, E. Langmuir 2002, 18, 3014. (6) Liu, Z. T.; Erkey, C. Langmuir 2001, 17, 274. (7) Lee, C. T.; Johnston, K. P.; Dai, H. J.; Cochran, H. D.; Melnichenko, Y. B.; Wignall, G. D. J. Phys. Chem. B 2001, 105, 3540. (8) Eastoe, J.; Gold, S. Phys. Chem. Chem. Phys. 2005, 7, 1352. (9) Renner, R. Science 2004, 306, 1887. (10) Eastoe, J.; Gold, S.; Rogers, S.; Wyatt, P.; Steytler, D. C.; Gurgel, A.; Heenan, R. K.; Fan, X.; Beckman, E. J.; Enick, R. M. Angew. Chem., Int. Ed. 2006, 45, 3675. (11) Eastoe, J.; Paul, A.; Nave, S.; Steytler, D. C.; Robinson, B. H.; Rumsey, E.; Thorpe, M.; Heenan, R. K. J. Am. Chem. Soc. 2001, 123, 988.
Langmuir 2010, 26(17), 13861–13866
A further advance was the design of CO2 surfactants that incorporate a third highly methylated chain, having the effect of dramatically improving CO2 compatibility over that of the analogous dichain surfactant.12 Recently, the custom-made trichain surfactant TC14 (sodium 1,4-bis(neopentyloxy)-3-(neopentyloxycarbonyl)-1,4-dioxobutane-2-sulfonate) has been shown to aggregate and form reverse micelles in pure CO2.13 The compound represents a notable improvement over other CO2-soluble hydrocarbon compounds,14 having milder solubility conditions and stabilizing reverse micelles in CO2. The structure of TC14 is consistent with paradigms for designing CO2-soluble compounds, including those based on the fractional free volume15,16 and cohesive energy density arguments.17,18 Nonetheless, the fundamental chemical reasons for the improvement of CO2 solubility with this compound are still not fully understood. To help in that direction, it would be useful to determine the effect of these new CO2-philic surfactants on the water/CO2 (w/c) interfacial tension; however, such measurements at the w/c interface are extremely experimentally challenging. In fact, only a few reliable examples of tension data at the w/c interface exist in the literature.16,19-23 (12) Gold, S.; Eastoe, J.; Grilli, R.; Steytler, D. C. Colloid Polym. Sci. 2006, 284, 1333. (13) Hollamby, M. J.; Trickett, K.; Mohamed, A.; Cummings, S.; Tabor, R. F.; Myakonkaya, O.; Gold, S.; Rogers, S.; Heenan, R. K.; Eastoe, J. Angew. Chem., Int. Ed. 2009, 48, 4993. (14) Fan, X.; McLeod, M. C.; Enick, R. M.; Roberts, C. B. Ind. Eng. Chem. Res. 2006, 45, 3343. (15) Stone, M. T.; Smith, P. G.; Da Rocha, S. R. P.; Rossky, P. J.; Johnston, K. P. J. Phys. Chem. B 2004, 108, 1962. (16) Stone, M. T.; Da Rocha, S. R. P.; Rossky, P. J.; Johnston, K. P. J. Phys. Chem. B 2003, 107, 10185. (17) Sarbu, T.; Styranec, T.; Beckman, E. J. Nature 2000, 405, 165. (18) O’Neill, M. L.; Cao, Q.; Fang, R.; Johnston, K. P.; Wilkinson, S. P.; Smith, C. D.; Kerschner, J. L.; Jureller, S. H. Ind. Eng. Chem. Res. 1998, 37, 3067. (19) Dickson, J. L.; Smith, P. G.; Dhanuka, V. V.; Srinivasan, V.; Stone, M. T.; Rossky, P. J.; Behles, J. A.; Keiper, J. S.; Xu, B.; Johnson, C.; DeSimone, J. M.; Johnston, K. P. Ind. Eng. Chem. Res. 2005, 44, 1370. (20) Dhanuka, V. V.; Dickson, J. L.; Ryoo, W.; Johnston, K. P. J. Colloid Interface Sci. 2006, 298, 406. (21) Bharatwaj, B.; Wu, L. B.; Da Rocha, S. R. P. Langmuir 2007, 23, 12071. (22) Da Rocha, S. R. P.; Johnston, K. P. Langmuir 2000, 16, 3690. (23) Da Rocha, S. R. P.; Harrison, K. L.; Johnston, K. P. Langmuir 1999, 15, 419.
Published on Web 08/05/2010
DOI: 10.1021/la102303q
13861
Article Table 1. Hydrocarbon Anionic Surfactant Structures Used in This Work
Mohamed et al.
reproducibility with previously reported measured and derived parameters.26,27 The work shows that TC14 is a remarkable compound, being capable of aggregation in a wide range of solvents spanning water to alkanes to liquid CO2. Significantly, the results provide new insight into the molecular structure requirements for designing CO2-active hydrocarbon surfactants, which is important for the optimization of CO2-based processing and handling technologies such as enhanced oil recovery, sequestration, and dry cleaning.
Experimental Section
Instead, here surface tension data from the air-water (a/w) surface are examined and analyzed to gain information about a range of anionic surfactants that are based on a common molecular motif. These compounds are shown in Table 1, representing the structural evolution of closely related single- (SC4), double(AOT, AOT4), and triple-chain (TC14, TC4) surfactants. The compounds all bear known CO2-philic alkyl chains but exhibit widely different solubility and a tendency to micellize in CO2. Although it is far from a perfect model for the c/w interface, the a/w surface does provide a well-defined and easy to study reference environment for comparing the adsorption and surface-packing properties of known CO2-philes. In particular, the focus is on TC14, which has been shown to be an extremely efficient surfactant for CO2.13 For a range of eight different fluorinated surfactants, a clear correlation between the aqueous surface tension and compatibility with carbon dioxide has been previously established.5 Hence, studies of aqueous surface tensions are known to be helpful in rationalizing the phase stability of reversed micelles in CO2,5 and that approach is taken here but with new hydrocarbon CO2-philic surfactants. Newly obtained surface tension data for TC14 have been compared to that of other surfactants, for which data have been collected over a number of years by different members of this research group.24-27a To validate these data further and in ensuing analyses, the standard AOT surface tension isotherm has been repeated, showing excellent (24) Hudson, L. Ph.D. Thesis, University of Bristol, 2008 (25) Eastoe, J.; Downer, A.; Paul, A.; Steytler, D. C.; Rumsey, E.; Penfold, J.; Heenan, R. K. Phys. Chem. Chem. Phys. 2000, 2, 5235. (26) Nave, S.; Eastoe, J.; Penfold, J. Langmuir 2000, 16, 8733. (27) (a) Nave, S.; Eastoe, J.; Heenan, R. K.; Steytler, D.; Grillo, I. Langmuir 2000, 16, 8741. (b) Li, Z. X.; Lu, J. R.; Thomas, R. K. Langmuir 1997, 13, 3681.
13862 DOI: 10.1021/la102303q
Surfactant Synthesis. The synthesis, purification, and chemical characterization of AOT4, TC14, and SC4 have been fully described elsewhere.11,13,24 Meanwhile, TC4 was synthesized following the same method as used in the preparation of TC14 but with a different alcohol precursor (Sigma-Aldrich). Further information concerning surfactant characterization can be found in the Supporting Information. AOT was purchased from Sigma-Aldrich and purified by Soxhlet extraction using dry ethyl acetate as the solvent. To remove any residual salts that remained after Soxhlet extraction, the surfactant was dissolved in the minimum amount of dry methanol and was subjected to repeated centrifugation.26 Deuterated water and heptane (98%) were purchased from Goss Scientific and used without further purification. Small-Angle Neutron Scattering (SANS). Experiments were carried out using the LOQ diffractometer3 and the newly commissioned SANS2d instrument at the Rutherford Appleton Laboratory, ISIS, U.K. Protocols for LOQ were as described before.3,26,27 For experiments on SANS2d, an incident wavelength range of 2.2-14 A˚, with the 1m2 detector offset sideways and vertically by 150 mm, was used resulting in an effective Q range of 0.005-0.7 A˚-1 was used. The measurements gave the absolute scattering cross section I(Q) (cm-1) as a function of momentum transfer Q (A˚-1). Absolute intensities ((5%) were determined by calibrating the received signal from a partially deuterated polymer standard, which was corrected for sample transmission and cell and solvent backgrounds, as reported previously.27 For aqueous micellar systems, deuterated water, D2O (FD2O = 6.330 1010 cm-2) was used to provide the necessary contrast. In SANS experiments for the study of reversed micellar (w/o microemulsion) systems, contrast was provided using a H surfactant and H2O against the deuterated heptane solvent (FC7D16 = 6.232 1010 cm-2). Ambient pressure experiments were run at 25 C in 2 mm Hellma quartz cells with a surfactant concentration of 0.10 M (mol dm-3). High-pressure experiments were conducted using a high-pressure cell (Thar) as described in the Supporting Information. Constant conditions were used for all samples in CO2: P = 380 bar and T = 25 C. As with the ambient pressure samples, the data were normalized for transmission, the empty cell, the solvent background, and pressure-induced changes in the cell volume.11,13 Neutrons are scattered by short-range interactions with sample nuclei, with the “scattering power” of different components being defined by a scattering length density (SLD), F (cm-2). For liquid CO2, FCO2 ≈ 2.50 mass density 1010 cm-2.28 At an experimental pressure of 380 bar, the CO2 density is ∼1.0 g cm-3 so that FCO2 is approximately 2.5 1010 cm-2. The contrast (FCO2 Fsurfactant/H2O) was provided by using a hydrogen-containing surfactant (H surfactant) with H2O and was calculated as a linear combination of contributions from the H surfactant (assuming that a density of 1.1 g cm-3 gives FTC14 = 0.812 1010 cm-2, FAOT4 = 0.486 1010 cm-2, and FTC4 = 0.503 1010 cm-2). Water (FH2O = -0.56 1010 cm-2) is assumed to distribute within the surfactant micelles. The data have been fitted using the FISH interactive fitting program, fixing the scattering length density differences as calculated above and fitting for the micellar volume (28) McClain, J. B.; Londono, D.; Combes, J. R.; Romack, T. J.; Canelas, D. A.; Betts, D. E.; Wignall, G. D.; Samulski, E. T.; DeSimone, J. M. J. Am. Chem. Soc. 1996, 118, 917.
Langmuir 2010, 26(17), 13861–13866
Mohamed et al.
Article
fraction and appropriate structural parameters as required by the different scattering laws. (Further details are in the Supporting Information and ref 29.) After extensive trial fitting with different approaches and on the basis of many reports in the literature,3-5,8,10-13,25-27,30 two main models have been employed: (1) for aqueous systems an ellipsoid form factor (P(Q)) multiplied by a Hayter-Penfold charged structure factor (S(Q))31 and (2) for reversed micelles in heptane and CO2, a spherical form factor taking into account a Schulz polydispersity distribution function, as described elsewhere.3,5,11,27 Surface Tension Measurements. Surface tension measurements were made using the Wilhelmy plate method on a Kr€ uss K11 or K100 instrument. All glassware and surface tension dishes were cleaned using 50% nitric acid solution and then rinsed thoroughly with distilled water and dried with compressed air. Before each measurement, the Pt plate was cleaned in distilled water and then heated for a few seconds in a blue Bunsen flame. The cleanliness of the plate and glassware were periodically checked by measuring the surface tension of pure water (Elga, 18 MΩ cm). All measurements were carried out at 25 C. For each concentration, 10 repeat measurements were obtained. In some cases, the chelating agent EDTA (99.5% tetrasodium salt hydrate, Sigma) was used at a constant EDTA to surfactant ratio in order to sequester trace impurities of divalent cationic species M2þ.26,27a New surface tension data were obtained for TC14, TC4, and AOT and then compared to data for other surfactants that have been collected over a number of years by different members of the Bristol research group (Nave (AOT, AOT4),26,27 Downer (DiCF4),25 and Hudson (SC4)24). To add validity to this collection of data, the newly obtained isotherm for AOT was compared to that previously determined by Nave. There was very good agreement in terms of the derived physical parameters, suggesting that valid comparisons with this compiled data set can be made. For experiments conducted by Nave and Downer, the measurements were made using a Du Nouy ring (Kr€ uss K10) and drop volume (Lauda TVT1) instruments, respectively. All experiments done by Hudson used the Lauda TVT1 instrument.24 In all previous cases, the protocols for cleaning equipment and preparing samples were as described here.
Results and Discussion Small-Angle Neutron Scattering (SANS). SANS is a powerful technique for obtaining structural information, especially for micellar aggregates.32 To enable a close comparison, the measurements were conducted at the same surfactant concentration and temperature. For example, data for TC4 in water have not been shown, owing to the formation of a biphasic system, because of high hydrophobicity and poor solubility (three long hydrophobic chains and only one hydrophilic group). The singlechain SC4 is essentially insoluble in heptane and CO2, giving no discernible scattering, so it has not been shown (see below for surface tension data). Figure 1 shows the SANS profiles for TC14, TC4, AOT4, and AOT in three very different solvents: water, heptane, and CO2. In water, the scattering profiles are indicative of distributions of charged micelles of TC14, AOT, and AOT4 (Figure 1a).31 It is generally accepted that an ellipsoidal form factor adequately accounts for intramicellar scattering31,32 and the peak arises owing to an electrostatic interparticle structure factor S(Q) arising from the charge stabilization of micelles.31 Parameters derived from model fits are given in Table 2, with principal radii R1 for
Figure 1. SANS profiles for TC14, TC4, AOT4, and AOT showing changes in aggregate structure in three different solvents: (a) Normal micelles in water at 0.10 M surfactant. The inset shows I(Q) scaled by a reduced concentration Cr = (C - cmc)/cmc to account for differing levels of micellized surfactants owing to the widely different cmc’s, and for presentation purposes. (b) Reversed micelles in heptane for w = [water]/[surf ] = 5 and a surfactant concentration of 0.10 M. (c) Dry reversed micelles in CO2 for w = 0 and [surf ] = 0.04 M obtained at 380 bar and 25 C. Data for AOT4 and TC4 are new to this study; that for TC14 and AOT has been replotted from ref 13. Lines through data points are model fits, and parameters are listed in Table 2. Characteristic error bars are shown for certain samples.
(29) Heenan, R. K. Rutherford Appleton Laboratory Report, RAL-89-129, 1989. (30) Eastoe, J.; Bayazit, Z.; Martel, S.; Steytler, D. C.; Heenan, R. K. Langmuir 1996, 12, 1423. (31) Hayter, J. B.; Penfold, J. Colloid Polym. Sci. 1983, 261, 1022. (32) Eastoe, J. Surfactant Chemistry; Wuhan University Press: Wuhan, China, 2003.
TC14, AOT4, and AOT being 11, 13, and 13 A˚, respectively, (2 A˚, consistent with the molecular length. Also listed in Table 2 are the fitted ellipsoid aspect ratios (X = R3/R1, with R3 being the secondary axis). With heptane as the solvent, hydrated reversed micelles could be stabilized by all of the double- and triple-chained surfactants at
Langmuir 2010, 26(17), 13861–13866
DOI: 10.1021/la102303q
13863
Article
Mohamed et al.
Table 2. Parameters Fitted to SANS Data from Micelles in Water, Heptane, and CO2a solvent/radius (2 A˚ D2O surfactant
R1
X
d-heptane w=5
CO2 w=0
TC14 11 1.7 15 11 AOT4 13 2.0 21 13 TC4 16 13 AOT 13 1.9 19 a For the aqueous systems, values of the principal ellipsoid radius R1 and the aspect ratio X (= R3/R1, with R3 being a secondary axis) are listed. The absence of fitted values from the TC4 data in water and AOT in CO2 are due to low solubilities.
25 C. Shown in Figure 1b is a family of curves for w=[water]/ [surfactant] = 5, demonstrating that reversed micelles of all four surfactants can be swollen with water in an organic solvent. The underlying structure of the surfactants, with bulky hydrocarbon chains and a small ionic headgroup (Table 1), facilitates the formation of reverse micelles or w/o microemulsions. Moving now to CO2, trichain TC14 exhibited good compatibility with liquid CO2 at 0.04 M, resulting in homogeneous optically transparent solutions with a reversible cloud-point transition at Ptrans = 160 bar, 25 C. However, Ptrans was higher for the longerchain TC4 with a translucent appearance evident at 400 bar and 25 C. These high-pressure SANS results show that TC14, AOT4, and TC4 clearly form reverse micelles in CO2 whereas no scattering is observed for AOT, consistent with the absence of any aggregation for the commercially available analogue. Fitted curves gave the micellar radii summarized in Table 2. It is clear to see that for trichain compounds TC14 and TC4, in both in deuterated heptane and CO2, the SANS profiles have similar forms. It is recognized that the three systems shown in Figure 1 are not directly comparable in terms of concentration and composition. Although attempted, it was found to be impossible to do that across these solvents of such widely differing polarity while also generating healthy SANS signals that could be reliably analyzed. For example, there are phase stability and solubility limits specific to each of the three solvents; furthermore, in CO2, TC14 and AOT4 could stabilize w = 5 hydrated micelles,13 but that was not possible with the other trichain analogue TC4, which phase separated after the addition of water. Hence, the systems under study represent a fair compromise between concentration (to optimize SANS intensities) and composition (to optimize phase stability). It is obviously wrong to attempt direct comparisons between these systems presented; despite these limitations, the overarching observation is that from this subset of common (AOT) and custom-made surfactants the trichain TC14 represents a special case that is capable of stabilizing micellar aggregates at extremes of solvent polarity. Significantly, TC14 clearly stabilizes aggregates across this wide range of solvent types, spanning water, heptane, and liquid CO2. As can be seen in Figure 1 and considering the extensive literature in this field,1-23 TC14 appears to represent a unique surfactant being compatible with a spectrum of solvents, importantly, liquid CO2 as well as commonly used water and alkanes. Surface Tension Measurements. It is of interest to understand these observations made above and why these trichain compounds, TC14 in particular, are so effective in CO2. Previously, for fluorocarbon CO2-philic surfactants, there have been notable successes in correlating the aqueous surface tension behavior with performance and reversed micelle formation in liquid CO2.5 By extending that approach now to these new hydrocarbon CO2-philes, an extensive set of air-water surface tension data were acquired and 13864 DOI: 10.1021/la102303q
Figure 2. Air-water surface tension γcmc vs ln(concentration) for
various surfactants at 25 C. Quadratic lines are fitted to pre-cmc data, with linear fits to post-cmc data.
analyzed. The results and analyses provide new and useful insights into the molecular structure requirements for generating CO2compatible hydrocarbon surfactants. (Justifications for taking the a/w surface as a first approximation model for the c/w interface have been given in the Introduction and ref 5.) cmc’s. Figure 2 shows the surface tension measurements for SC4, AOT, AOT4, TC14, and TC4. In all cases, there is a sharp break at the cmc and no evidence of surface tension minima, which is consistent with surfactants of high purity. The new cmc values are in good agreement with those previously reported.11-13,26,27 In addition a comparison of the individual cmc’s shows that they follow a general trend of decreasing cmc with increasing effective chain length (i.e., considering the total number of carbons in the longest linear portion only). However, because the surfactants differ in terms of degrees of chain branching, the exact relationship between the cmc and the alkyl chain length is more subtle (and not the purpose of this study). More information on these chain branching effects on the cmc can be found elsewhere.12,24,26,27a Limiting Surface Tensions. The effectiveness of a surfactant at reducing surface tension can be expressed by the limiting surface tension at the cmc (γcmc), with reductions in γ encountered by ionic surfactants after the cmc is considered to be less significant.33 Of the surfactants tested, TC14 has the lowest γcmc = 27.0 mN m-1, being very low for a hydrocarbon surfactant, and is even comparable to some dichain fluorocarbon CO2-active surfactants (DiHCF4 = 26.8 mN m-1).5 A general trend of decreasing γcmc with carbon number as a result of the increase in hydrophobicity has previously been reported.5 Given that the effective chain length of TC14 (γcmc = 27.0 mN m-1) is only three carbons, this makes the limiting surface tension even more remarkable. The low γcmc is believed to be due to the incorporation of chain terminal methyls because CH3 groups have lower surface energies than CH2 groups.34 Comparing γcmc for the series of surfactants SC4 (monochain), AOT4 (dichain), and TC14 (trichain), a clear trend of decreasing γcmc with an increasing number of tails is evident, as reported previously with an analogous set of surfactants.34 It should be noted that the previous study was conducted in aqueous gelatin systems (not pure water) so absolute values cannot be directly compared to the ones presented here.34 The lowering of γcmc was attributed to the hydrocarbon tail packing efficiency balanced against electrostatic headgroup repulsions and an increase in the number of low-energy (33) Rosen, M. J. Surfactants and Interfacial Phenomena, 3rd ed.; Wiley-Interscience: Hoboken, NJ, 2004. (34) Pitt, A. R.; Morley, S. D.; Burbidge, N. J.; Quickenden, E. L. Colloids Surf., A 1996, 114, 321.
Langmuir 2010, 26(17), 13861–13866
Mohamed et al.
Article
with widely different cmc’s, a reduced concentration axis has been used (concentration/cmc). Values for Acmc were calculated from eq 2 and are detailed in Table 3, along with other physical parameters derived from the surface tension curves. The molecular area Acmc does not increase linearly with the number of surfactant chains, suggesting that the packing efficiency increases as additional chains are added onto the surfactant headgroup “shoulders”. This is most evident from the effective area occupied per individual chain, which decreases from 45.0 A˚2 for SC4 (monochain) to 35.0 A˚2 for AOT4 (dichain) and then to 26.7 A˚2 for TC14 (trichain). The FFV index is defined as FFV ¼ 1 Figure 3. Adsorption isotherms for various surfactants derived from the γcmc vs ln(concentration) plot (Figure 2). For ease of comparison, the data have been plotted on a reduced-concentration x axis. Symbols the same as for Figure 2.
methyl groups per headgroup.34 It was also shown previously that variations in γcmc followed the same trends as observed for surface energies of corresponding surfactant-coated solids.34 For example, when varying the chain-tip chemistry, both the surface energy of a surfactant-coated solid and γcmc decrease following the series CF3CF2 < H(CF2)2 < CH3CH2 < CH2CH2 < phenyl.34 The reported good correlation between γcmc and the surface energy of such model solids suggests that γcmc is strongly dependent on dispersive interactions between hydrophobic tails.34 Given that weak interactions exist between CO2 and hydrocarbon tails, this must be an important factor for minimizing tail-tail interactions in order to improve the solubility in CO2.15,16 It is conceivable, therefore, that a lower γcmc may suggest reduced dispersive interactions between hydrophobic tails and consequently an improvement in CO2 solubility. This is indeed the case for TC14, which has the lowest γcmc and the highest compatibility (and hence tendency to micellize) of the hydrocarbon surfactants tested. Similar observations linking the aqueous surface tension to performance in CO2 have been previously reported.5,25 Molecular Packing at the a/w Interface. Johnston proposed that a crucial factor in the success of fluorocarbon surfactants for stabilizing dispersions in CO2 is the larger volume occupied by fluorocarbon tails in comparison to that occupied by hydrocarbon tails.15 Simulations suggested that a better separation between CO2 and water by the bulkier fluorocarbon chains would result in lower interfacial tensions and consequently stable w/c microemulsions.16 The fractional free volume (FFV, see eq 3 below) is a measure of this, being calculated directly from the tail geometry and surfactant headgroup areas derived from surface tension data.15 One way to characterize the molecular packing efficiency of an adsorbed layer is to consider the surface excess and the area per headgroup at the cmc (Acmc). To obtain headgroup areas from the surface tension curve, the data in the pre-cmc region were fitted to quadratics, which were then used to generate adsorption isotherms using a prefactor of m = 2 (dissociating 1:1 ionic surfactants) in the Gibbs equation (eq 1).25,26,32 Γ ¼ -
1 dγ mRT ln c
ð1Þ
1 ΓNa
ð2Þ
Acmc ¼
Figure 3 summarizes the different adsorption isotherms derived using eq 1, and to compare the different surfactants easily Langmuir 2010, 26(17), 13861–13866
V tAcmc
ð3Þ
where V is the volume and t is the length of the surfactant chain, with Acmc as defined above. The FFV values were calculated according to eq 3,15 with hydrocarbon chain volumes and lengths determined using the SPARTAN molecular modeling program, as before.15 The molecules were created with all bonds in a trans configuration, followed by an energy-minimization calculation using the MMFF force field.15 The Acmc values were derived from the surface tension data as described above. In fact, FFV values have been previously reported for some CO2-philic surfactants;15 however, in that previous study Acmc values were assumed in all cases to be 100 A˚2. Here, Acmc values were determined from the surface tension data, so care should be taken in comparing these values with those previously reported.15 Normal AOT has a relatively high FFV (0.58) for a hydrocarbon surfactant, which is mainly due to the side-chain ethyl substituents. Although not a direct trimethylated analogue of AOT, the FFV of AOT4 is lower (FFV = 0.46), presumably because of the incorporation of the trimethylated chain tips (i.e., AOT4 fills space more efficiently than its common relative, AOT). There is also a large decrease in the FFV from the single-chain SC4 (FFV = 0.58) to its direct dichain analogue AOT4 (FFV = 0.46) and further still then to the shorter trichain surfactant TC14 (FFV = 0.26). Table 3 details some phase-transition pressures in CO2, with a lower pressure giving an indication of an improvement in the CO2-philicity of the surfactant. It is clear for hydrocarbon surfactants that a decrease in FFV broadly correlates to an improvement in the CO2 compatibility of the surfactant. However, it is also conceivable that this might be due to the increase in the number of chain terminal low-energy CH3 groups, which obviously goes hand in hand with an increase in the number of chains. The fact that TC14 has a considerably lower FFV than AOT4 may be one plausible reason for why TC14 forms stable hydrated micelles in CO213 by improving the separation of the small penetrating CO2 molecules from water, as required on the basis of computer simulations.16 It is interesting that the FFV value of TC14 is even lower than that of the fluorinated surfactant DiCF4 (FFV = 0.42). Given that DiCF4 is capable of stabilizing microemulsions up to w ≈ 30 and outperforms TC14 in CO2, this points to some limitations of considering FFV as the critical parameter for the design of CO2-philic surfactants. The interfacial tension at the pure w/c interface is typically 20-30 mN m-1. This is much lower than for typical hydrocarbon oil/water interfaces, which usually have γo/w ≈ 50 mN m-1. The smaller driving force for adsorption at the w/c interface due to the lower interfacial tension suggests that Acmc will be much larger. This has been repeatedly observed experimentally.4,22,23,25,30 For the fluorosurfactant DiCF4, for example, Acmc ≈ 63 A˚2 at the a/w interface, but it has been measured to be 61% larger than that at DOI: 10.1021/la102303q
13865
Article
Mohamed et al. Table 3. Parameters Derived from Surface Tension Measurementsa
surfactant
cmc/mM ( 0.03
γcmc/(mN m-1) ( 1
Acmc/A˚2( 2
area per chain/A˚2 ( 2
FFV a/w
FFV w/cg
Ptrans/bar
91.10 29.8 45 45.0 0.58 0.74 insoluble SC4b AOT 2.49 31.8 75 37.5 0.51 0.69 insoluble c 2.56 30.8 75 37.5 0.51 0.69 insoluble AOT 1.10 28.0 70 35.0 0.46 0.67 500 (1 wt % 40 C) f AOT4d TC14 21.63 27.0 80 26.7 0.26 0.54 160 (2 wt % 25 C) TC4 0.0125 29.7 53 17.6 0.0 0.34 translucent e 1.63 17.9 62 31.0 0.42 0.64