Surfactants for CO2 - Langmuir (ACS Publications)

---

9832

Langmuir 2006, 22, 9832-9842

InVited Feature Article Surfactants for CO2 Julian Eastoe,*,† Sarah Gold,† and David C. Steytler‡ School of Chemistry, UniVersity of Bristol, Bristol BS8 1TS, United Kingdom, and School of Chemical Sciences and Pharmacy, UniVersity of East Anglia, Norwich NR4 7TJ, United Kingdom ReceiVed March 21, 2006. In Final Form: June 16, 2006 For some 15 years the attainment of efficient, nonfluorinated CO2-active surfactants has been a Holy Grail for researchers spanning pure and applied chemical sciences. This article tells the story of small-molecule CO2-active surfactants, from the first tentative observations with fluorinated compounds in 1991 up to recently discovered fluorinefree oxygenated amphiphiles.

Introduction It is without doubt that waste solvents from chemical processing and related industries pose a huge environmental threat. One of the goals of green chemistry is to find viable replacements for volatile organic solvents and compounds (VOCs). In this respect, supercritical fluids (SCFs) are attractive because of the ease of solvent removal, recyclability, and tunability of properties such as density and solvent quality through temperature and pressure control. Unfortunately, the critical temperatures (Tc’s) of many potential SCF solvents can be inconveniently high (e.g., water Tc ≈ 374 °C); however, the critical point for carbon dioxide is more conveniently accessible (Pc ) 72.8 bar, Tc ) 31.1 °C). Furthermore, supercritical carbon dioxide (sc-CO2) is well suited to green applications, especially in the food and pharmaceutical industries, because it is an environmentally benign and biocompatible solvent. There are other benefits, making CO2 a key green solvent: it is nonflammable, nontoxic, cheap, abundant, and importantly one of the few solvents not regulated as a VOC by the U.S. Environmental Protection Agency (EPA). That is the spin. Now the truth: CO2 is generally a poor solvent, especially for polar and high-molecular-weight solutes, and this rules out many potential applications. Water and carbon dioxide exhibit a weak mutual solubility depending on pressure and temperature. (The solubility of water in CO2 at 15 °C and 450 bar is ∼0.1 wt %.) The solvent properties of sc-CO2 could be vastly improved by the incorporation of surfactant reverse micelles comprising polar nanodomains. Ideally, these amphiphiles would also lead to water-in-CO2 (w/c) microemulsions. Alas, very few common surfactants exhibit any significant solubility in CO2,1 a fact normally rationalized by the very different Hildebrandt solubility parameters for typical hydrocarbons and CO2. However, solubility is significantly enhanced for fluorinated alkanes, surfactants, and polymers.1-3 From the outset, in the early 1990s, the need for custom-designed surfactants was recognized; the search for CO2-active surfactants was given an important boost * Corresponding author. † University of Bristol. ‡ University of East Anglia. (1) Consani K. A.; Smith, R. D. J. Supercrit. Fluids 1990, 3, 51. (2) Hoefling, T. A.; Enick, R. M.; Beckman, E. J. J. Phys. Chem. 1991, 95, 7127. (3) Eastoe, J.; Gold, S. Phys. Chem. Chem. Phys. 2005, 7, 1353.

by Hoefling, Enick, and Beckman, who introduced heavily fluorinated dichain surfactants based on the aerosol-OT motif.2 This finding that F surfactants were CO2-soluble triggered significant efforts by groups in the U.S., Europe, Japan, and Asia with the aim of uncovering key molecular structures to promote efficient stabilization of w/c phases. This body of work up to 2004 has recently been reviewed by Gold and Eastoe,3 and the reader is directed there for further details. The current article sets in context research from Bristol on small-molecule surfactants with relevant output from other groups, setting the scene to introduce significant new findings with fluorine-free, oxygenated, CO2-active surfactants. It is appropriate to make the distinction between small-molecule CO2 surfactants (the subject of this article) and polymeric CO2philes, which have been the focus of other research groups over a similar time frame.4-7 An aim of the research described herein was to identify, through systematic studies, key CO2-compatible molecular structures that can be used to generate high-efficiency surfactants of any kind. For this, low-molecular-weight surfactants have some advantages over block copolymer amphiphiles; they can be more easily synthesized in monodisperse form, characterized,7 and also purified to the rarified levels necessary for reliable surface chemistry studies (especially surface/interfacial tension; Supporting Information). Important advances in the field of polymeric CO2-philes by DeSimone et al. were made4-7 that revealed guiding principles inspiring this work on small-molecule systems; in particular, the concept of CO2-philic and CO2-phobic groups5 and the need for high fluorine levels to achieve compatibility with this very “weak” solvent. The different surfactant classes and nomenclature featured in this article are shown in Table 1: each class is identified as 1a, 1b, and so forth; a short-hand name (di-CFn, di-HCFn, ...) is given, as are the specific structures that have been investigated (m and n). Details of the synthesis, chemical characterization, and purification protocols can be found in the accompanying references,2,8-15 as indicated in Table 1. This article tracks the (4) Guan, Z.; DeSimone, J. M, Elsbernd, C. S. Science 1992, 257, 945. (5) DeSimone, J. M.; Maury, E. E.; McClain, J. B.; Romack, T. J.; Combes, J. R.; Menceloglu, Y. Z. Science 1994, 265, 356. (6) Guan, Z.; DeSimone, J. M. Macromolecules 1994, 27, 5527. (7) Fulton, J. L.; Pfund, D. M.; McClain, J. B.; Romack, T. J.; Maury, E. E.; Combes, J. R.; Samulski, E. T.; DeSimone, J. M.; Capel, M. Langmuir 1995, 11, 4241.

10.1021/la060764d CCC: $33.50 © 2006 American Chemical Society Published on Web 10/27/2006

Surfactants for CO2 Table 1. CO2-Compatible Surfactants Described in This Article

Langmuir, Vol. 22, No. 24, 2006 9833

progress in reducing levels of harmful and costly fluorine from very effective molecules 1 to 2 and 3 to purely hydrocarbon surfactants 4 and 5 that show only modest activity in CO2. Finally, the research leads to oxygenated compounds 6 and 7, which are shown to be very efficient fluorine-free CO2 surfactants. Supporting information13 is an important aspect of this article. The SI section provides synthetic details as well as protocols and criteria for obtaining surfactants of sufficiently high surface chemical purity, with links to relevant literature. Also included in SI are experimental details on two important techniques for

interrogating w/c phasessnear infrared spectroscopy (NIR) and high-pressure small-angle neutron scattering (HP-SANS or SANS). There is no generally accepted definition of the term “microemulsion”, and this has been a disputed area for more than 20 years. For example and relevant here, with water-insolvent systems there is no clear delineation between low-watercontent hydrated reverse micelles and more swollen surfactantcoated water droplets. In the latter systems, with higher water loadings, the state of water in the pools approaches that of the bulk liquid. This has obviously caused confusion; some authors identify the low-water-content hydrated micelles as “microemulsions”, and others feel that “free” water is an essential feature. The situation is further complicated by the fact that different surfactant types have separate headgroup hydration requirements: for example, ethylene oxide groups are believed to demand three water molecules per EO,16 whereas an anionic sulfonate group with a sodium ion takes around four waters of hydration.17 In terms of the water-to-surfactant molar ratio w ) [water]/ [surfactant] and as a rule of thumb, microemulsion droplets containing free water pools can be assumed to have w values of greater than 4 for the sodium sulfonates described in this article and a value of 3 × nEO for typical CiEj nonionics. In this article, the term microemulsion refers to swollen reverse micelles stabilized by sodium sulfosuccinate surfactants with water loadings above w ) 4.17 Results from NIR and HP-SANS are described in the main article because these methods provide key and irrefutable evidence for the presence of such microemulsified water pools: NIR through overtone bands in the region of 1300-1600 nm characteristic of free water, and SANS owing to the efficient contrasting of surfactant-stabilized D2O nanodomains (used in place of H2O) against the CO2 continuum. In the case of SANS, unmistakable fingerprint I(Q) intensity profiles also arise for D2O nanodroplet cores, stabilized by an outer hydrocarbon surfactant monolayer shell (the so-called coreshell form factor). The origin and evolution of these I(Q) curves to provide a plausible model for the microemulsion droplet structure is also explained in SI. The central issue is whether any given surfactant can be considered sufficiently active at the water-CO2 interface. Highpressure measurements of interfacial tensions γw/c between water and CO2 in the presence of surfactant and polymers have been reported.18,19 However, because this is an experimentally demanding technique and the interpretation of γw/c versus concentration data is fraught with difficulties,8,10,12,18,19 it has not found widespread application. Over the years, numerous reports have emerged suggesting microemulsion formation in CO2 with commercially available nonionic hydrocarbon surfactants20,21 and medium-chain alcohol-AOT mixtures.22 These

(8) Downer, A.; Eastoe, J.; Pitt, A. R.; Simister, E. A.; Penfold, J. Langmuir 1999, 15, 7591. (9) Guo, W.; Li, Z.; Fung, B. M. J. Phys. Chem. 1992, 96, 6738. Yoshino, N.; Hamano, K.; Omiya, Y. Langmuir 1995, 11, 466. Guo, W.; Li, Z.; Fung, B. M.; O’Rear, E. A.; Harwell, J. H. J. Phys. Chem. 1992, 96, 6738. Guo, W.; Li, Z.; Fung B. M.; O’Rear, E. A. J. Phys. Chem. 1992, 96, 10068. (10) Eastoe, J.; Dupont, A.; Murray, M.; Martin, L.; Guittard, F.; Taffin de Givenchy, E.; Heenan, R. K. Langmuir 2004, 20, 9953.Eastoe, J.; Dupont, A.; Murray, M.; Martin, L.; Guittard, F.; Taffin de Givenchy, E.; Heenan, R. K. Langmuir 2004, 20, 9960. (11) Steytler, D. C.; Rumsey, E.; Thorpe, M.; Eastoe, J.; Paul, A.; Heenan, R. K. Langmuir 2001, 17, 7948. (12) Nave, S.; Eastoe, J.; Penfold, J. Langmuir 2000, 16, 8733. Nave, S.; Eastoe, J.; Heenan, R. K.; Steytler, D. C.; Grillo, I. Langmuir 2000, 16, 8741. (13) See Supporting Information for this article. (14) 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. 2006, 45, 3675. (15) Fan, X.; Potluri, V. K.; McLeod, M. C.; Wang, Y.; Lui, J.; Enick, R. M.; Hamilton, A. D.; Roberts, C. B.; John ston. J. K.; Beckman, E. J. J. Am. Chem. Soc. 2005, 127, 11754.

(16) Sedev, R. Langmuir 2001, 17, 562. (17) MacDonald, H.; Bedwell, B.; Gulari, E. Langmuir 1986, 2, 704. (18) Harrison, K. L.; Johnston, K. P.; Sanchez, I. C. Langmuir 1996, 12, 2637. da Rocha, S. R. P.; Harrison, K. L.; Johnston, K. P. Langmuir 1999, 15, 419. da Rocha, S. R. P.; Johnston, K. P. Langmuir 2000, 16, 3690. Psathas, P. A.; Sander, E. A.; Lee, M. Y.; Lim, K. T.; Johnston, K. P. J. Disper. Sci. Technol. 2002, 23, 65. Psathas, P. A.; Sander, E. A.; Ryoo, W.; Mitchell, D.; Lagow, R. J.; Lim, K. T.; Johnston, K. P. J. Disper. Sci. Technol. 2002, 23, 81. Hebach, A.; Oberhof, A.; Dahmen, N.; Ko¨gel, A.; Ederer, H.; Dinjus, E. J. Chem. Eng. Data 2002, 47, 1540. (19) Sagisaka, M.; Fujii, T.; Ozaki, Y.; Yoda, S.; Takebayashi, Y.; Kondo, Y.; Yoshino, N.; Sakai, H.; Abe, M.; Otake, K. Langmuir 2004, 20, 2560. (20) Ryoo, W.; Webber, S. E.; Johnston, K. P. Ind. Eng. Chem. Res. 2003, 42, 6348. (21) Liu, J.; Han, B.; Zhang, J.; Li, G.; Zhang, X.; Wang, J.; Dong, B. Chem.s Eur. J. 2002, 8, 1356. Liu, J.; Han, B.; Li, G.; Zhang, X.; He, J.; Liu, Z. Langmuir 2001, 17, 8040. Liu, J.; Zhang, J.; Tiancheng, M.; Han, B.; Li, G.; Wang, J.; Dong, B. J. Supercrit. Fluid. 2003, 26, 275. Liu, J.; Han, B.; Wang, Z.; Zhang, J.; Li, G.; Yang, G. Langmuir 2002 18, 3086. (22) Hutton, B. H.; Perera, J. M.; Grieser, F.; Stevens, G. W. Colloids Surf., A 1999, 146, 227. Hutton, B. H.; Perera, J. M.; Grieser, F.; Stevens, G. W. Colloids Surf., A 2001, 189, 177.

9834 Langmuir, Vol. 22, No. 24, 2006

Eastoe et al.

potentially important findings could have had significant impacts on industrial applications of CO2, hence they warranted detailed investigations by NIR and HP-SANS. Unfortunately, no clear evidence for the presence of microemulsion water nanodroplets was found with any of these surfactants, although low-hydrationnumber reverse micelles are undoubtedly formed in some circumstances. The upshot is that no easy fixes have been found to obtain CO2-active surfactants. The only certain route has been to design and custom synthesize likely candidates comprising known CO2philic groups. After nearly 15 years of attempts by various groups, some advances have been made recently; this article ends by demonstrating that fluorine-free oxygenated compounds 6 and 7 are CO2-active surfactants capable of stabilizing mixed water and CO2 phases. New surfactants of this kind could herald a new era in green solvent chemistry and applications of CO2-based technologies. Experimental Section Materials. Details on the reagents, synthetic and purification procedures, and chemical characterization for all of the surfactants can be found elsewhere.2,8-15 H2O (18.2 MΩ cm) was obtained from a Purite, Elga, or Millipore system, and deuterium oxide (99.9% D atom Fluorochem Ltd) was distilled before use. Liquid CO2 (B.O.C. instrument grade) was used as received. Water/CO2 Microemulsion Phase Stability. Studies of the w/c microemulsion phase behavior were conducted using a stainless steel variable-volume high-pressure cell fitted with sapphire windows, as used in previous work.10,11,13 The cell operation and protocols for sample preparation are detailed elsewhere.10,13 The microemulsion composition is defined in terms of the surfactant concentration (typically in the range of 0.05-0.15 mol dm-3), and the solubilization capacity, in terms of the water-to-surfactant molar ratio w ) [water]/ [surfactant]. Models have been developed to account for the T-Pdependent solubility of water in CO2.23 The weak water-CO2 solubility typically reduces the effective w values by between 1.5 and 2.5 w units for the systems under study, depending on surfactant concentration, pressure, and temperature. For reporting phase behavior, uncorrected w values are quoted (Figure 3), whereas with near-infrared spectra and SANS data corrected w values (wcorr) are given, taking into account the solubility of water in the background CO2 solvent. Pressure- and temperature-induced changes in the turbidity of the microemulsion samples were noted by visual inspection. Above the phase-transition pressure, Ptrans, clear onephase microemulsion regions were observed. On lowering the pressure isothermally, samples become cloudy on the approach to Ptrans, eventually turning opaque below the stability boundary. At each temperature, Ptrans was determined ((5 bar) by reducing the pressure from a single-phase region until phase separation; no signs of hysteresis were noted when comparing heating/cooling curves. Near-Infrared Spectroscopy. Near-infrared spectra of binary surfactant/CO2 mixtures, water/surfactant/CO2, and water/surfactant/ alcohol/CO2 phases were obtained using a Thermo-Nicolet Nexus NIR spectrometer as described elsewhere.13 Interfacial Tension Measurements. Interfacial tension measurements at the CO2-water interface were measured by the pendant drop method using a CAM 200 tensiometer (KSV instruments) under saturated surfactant absorption conditions for AOT and AOT 4 (4). A sufficient quantity of surfactant was initially placed in the highpressure optical cell to maintain saturation of the CO2 phase over the P-T conditions examined. Solutions of selected surfactant at a concentration above the cmc were introduced into the pressure cell using a high-pressure injection system to form the pendant droplets. Sufficient ethylenediamine tetraacetic acid (EDTA, tetrasodium salt, Aldrich) was added to the aqueous solutions to eliminate adsorption (23) Wiebe, R.; Gaddy, V. L. Chem. ReV. 1941, 63, 475. Coan, C. R.; King, A. D., Jr. J. Am. Chem. Soc. 1971, 93, 1857. King, M. B.; Mubarak, A.; Kim, J. D.; Bott, T. R. J. Supercrit. Fluids 1992, 5, 296.

Figure 1. (a) HP-SANS profiles for an F7-H7 (2) stabilized waterin-CO2 microemulsion at 25 °C. Composition 3 wt % (50 mmol dm-3) and 3.5 wt % D2O (∼1.75 mol dm-3) and wcorr ≈ 33.0. Pressures: 500 (0), 260 (b), and 120 bar (O). Lines are fits described in ref 43. (Plot a originally appeared in ref 43.)ss(b) HP-SANS data and error bars for a w ) 20 w/c microemulsion stabilized by di-HCF4 (1b). [di-HCF4] ) 0.10 mol dm-3; 15 °C and 500 bar. Lines are fits described in ref 44: - I(Q); -- P(Q); -‚ S(Q) ss(Plot b originally appeared in ref 44.) effects of trace amounts of divalent impurities as described elsewhere.10,12 For each drop formed, the average interfacial tension (γ j ) was obtained by curve fitting the Young-Laplace equation to at least eight images of the droplet.10 Measurements were repeated over a time frame of up to 1 h to ensure that equilibrium had been achieved. The reproducibility of the method was excellent, and all values agreed with negligible deviation. Small-Angle Neutron Scattering and Data Analysis. Smallangle neutron scattering (SANS) was used to characterize the microemulsion structure. Experiments were conducted on the timeof-flight small-angle diffractometer LOQ at ISIS, Rutherford Appleton Laboratory, Didcot, U.K., in conjunction with the pressure cell mentioned above as outlined in Supporting Information. Standard procedures13 for the normalization of raw neutron counts yielded absolute scattering intensities I(Q) in cm-1, where the momentum transfer is Q ) (4π/λ)sin(θ/2) and θ is the scattering angle (