Micellization of economically viable surfactants in CO2

Journal of Colloid and Interface Science 258 (2003) 367–373www.elsevier.com/locate/jcis

Micellization of economically viable surfactants in CO2

Julian Eastoe,a,∗ Audrey Dupont,a David C. Steytler,b,∗ Matthew Thorpe,b Alexandre Gurgel,b

and Richard K. Heenanc

a School of Chemistry, University of Bristol, Bristol BS8 1TS, UKb School of Chemical Sciences, University of East Anglia, Norwich NR4 7TJ, UK

c ISIS-CLRC, Rutherford Appleton Laboratory, Chilton, Oxon OX11 0QX, UK

Received 22 March 2002; accepted 28 October 2002

Abstract

Stability and aggregation structures of various economically viable surfactants for CO2 are reported. The compounds are eithercommercially available octylphenol nonionics (Triton X-100, X-100 reduced, and X-45) or custom-made analogues of aerosol-OT (J. Am.Chem. Soc. 123 (2001) 988). These were selected to reveal the influence of chain terminal group structure, namely highly methylatedt-butylunits, on solubility and aggregation in CO2. In addition the mean ethylene oxide block length is varied for the Triton surfactants (X-100∼ EO10, X-45 ∼ EO8). High-pressure small-angle neutron scattering (SANS) experiments revealed the presence of aggregates, consistentwith spheroidal reverse micelles. The nonionics show a temperature and pressure dependence on solubility. These results confirm the specialaffinity of highly methyl-branched tails for CO2. However, none of these systems were able to disperse significant amounts of water orbrine; therefore hydrated reversed micelles or microemulsion droplets were not stabilized. Hence the utility of these cheap methyl-branchedsurfactants in CO2 is limited, and so groups of greater CO2-philicity are needed to achieve the goal of water–hydrocarbon surfactant–CO2dispersions. 2003 Elsevier Science (USA). All rights reserved.

Keywords:CO2; Triton; AOT; Aggregates; SANS

1. Introduction

In recent years, supercritical CO2 (sc-CO2) has grown inpopularity since it is a potential alternative to organic sol-vents that are employed in multifarious chemical processes.For these applications, CO2 has a number of attractiveproperties: low cost, environmental friendliness, nonflam-mability, and nontoxicity for food and pharmaceutical gradeuses. Furthermore, as a supercritical fluid, its critical pres-sure and temperature (respectively 73.8 bar and 31.1◦C) canbe easily reached so that solvent quality can be fine-tunedby T –P variation. Therefore sc-CO2 offers great oppor-tunities for application in a variety of domains includingextractions [1] and polymer processing [2,3]. More recentapplications in nanotechnology include the production of sil-ver nanoparticles [4,5], and the preparation of nanocrystals[6] and of carbon nanotubes [7].

* Corresponding authors.E-mail address:[email protected] (J. Eastoe).

However, CO2 differs significantly in nature from con-ventional organic solvents. It is essentially nonpolar, witha very low dielectric constant, and exhibits low polariz-ability and weak Van der Waals forces. It is therefore apoor solvent for polar molecules, for which CO2-activesurfactants have been developed to boost solubility levelsthrough solubilization in micelles and water-in-CO2 (w/c)microemulsions. Preliminary studies [8] concluded that mostconventional surfactants are insoluble in sc-CO2. On theother hand (for sound reasons [9]), certain fluorocarbon[10,11] and to a lesser extent silicone surfactants [12,13]are known to dissolve in CO2. However, such compoundsoften require specialized synthesis and therefore remain ex-pensive.

Up to now little work has been reported in the quest forcheaper CO2-compatible surfactants. Dilute w/c emulsionshave been formed with block copolymers generated fromether carbonate [14] or from polyethylene glycol and cy-clohexene oxide (Pluronic 17R4) [15]. The aggregation ofstraight-chain nonionic polyethylene glycoln-alkyl ethers(C12E3 and C12E8 [16], C8E5 [17], and C12E4 [18]) in

0021-9797/03/$ – see front matter 2003 Elsevier Science (USA). All rights reserved.doi:10.1016/S0021-9797(02)00104-2

http://www.elsevier.com/locate/jcis

368 J. Eastoe et al. / Journal of Colloid and Interface Science 258 (2003) 367–373

Fig. 1. Molecular structures of the surfactants used in this study.

sc-CO2 was investigated by Fourier transform infrared spec-troscopy (FTIR) [16,17] along with solvatochromic tech-niques [17] and solubility studies [18]. However, in thesestudies little water was incorporated into the polar core ofthe micelles and addition of a cosurfactant or even a cosol-vent was often required. Ihara et al. [19] and Hutton et al.[20] reported that AOT (sodium bis (2-ethylhexyl) sulfos-uccinate) is sc-CO2 soluble and forms w/c microemulsions,but only in the presence of ethanol or pentanol cosurfac-tants. Along similar lines Eastoe et al. [21] demonstrated thathighly methyl-branched analogues of AOT formed reversemicelles in sc-CO2 (e.g., AOT4 in Fig. 1). These compoundsexhibit low surface tension in water at the critical micelliza-tion concentration (cmc), and they associate weakly in CO2

with favorable tail-sc-CO2 interactions.Here micellization of various low-cost hydrocarbon sur-

factants in sc-CO2 is reported. These surfactants, shownin Fig. 1, were selected because of their low cost andCO2-compatible chain tips, which are highly methylated.Triton surfactants are commercially available, being exten-sively used in industrial [22,23] and pharmaceutical formu-lations [24,25], and in biochemical research [26]. AOT4 wascustom-made, and its synthesis is now well established. Theseries of Triton surfactants was also chosen so that ethyleneoxide chain length effects could be studied. The aggrega-tion behavior of these surfactants in sc-CO2 was investigatedby the direct method of high-pressure small-angle neutronscattering (SANS). It can be emphasized that this studywidely contrasts with other works where an indirect spectro-scopic method is used to provide evidence for aggregation[16,17,27].

2. Experimental

2.1. Chemicals

TX-100 (Sigma), TX-45 (BDH), and TX-100R (Aldrich)were used as received. AOT4 was prepared and purifiedusing procedures described elsewhere [28,29].

2.2. Instruments and methods

2.2.1. Phase behaviorPhase transition measurements were made in a variable-

volume high-pressure cell having sapphire windows. A mov-able piston allowed the pressure to be varied in the range1–600 bar and a displacement transducer was used to detectthe piston position and thereby obtain the sample volume.Depending on pressure, this is usually in the range 12–15 cm3. The pressure was measured to within±2 bar,using a Dynisco pressure transducer inserted into the cell.The cell contents were mixed using a magnetic stirringbar driven by a powerful magnetic stirrer. A more techni-cal description may be found elsewhere [30]. In a typicalexperiment, the vessel was first cooled to 15◦C; then aknown mass of surfactant was introduced and liquid CO2

(B.O.C, Instrument Grade) was slowly added. Pressure- andtemperature-induced changes in the turbidity of the sampleswere noted by visual inspection.

To determine the effect of surfactant fractionation a two-phase system of CO2 (12 ml) and TX-100 (2 ml) wasexamined. After stirring and settling at fixed temperature andpressure, 100-µl samples were removed from the CO2 phaseand analyzed by HPLC.

2.2.2. HPLCThe samples taken from the pressure cell were diluted

to 2 ml with a mixture of isopropanol :n-heptane : methanolin a 70 : 10 : 20 volume ratio and injected (100 µl) into aperfusion chromatography system. Separation was achievedon an Adsorbosphere amino column (10-µm packing) usingthe above solvent mixture as mobile phase at a flow rateof 1.0 ml min−1 in isocratic mode. The components werequantified by UV detection at 276 nm using standard peakintegration software (BioCad Sprint). All solvents wereultrasonically degassed for 2 h before use.

2.2.3. Surface tensionThe surface chemical purity of AOT4 was assessed by

tensiometry as described elsewhere [29,31]. Measurementswere performed at 25◦C on a drop volume tensiome-ter (DVT) Lauda TVT1 operating in a dynamic mode.The adsorption isotherm and the estimated cmc (1.08 ×10−3 mol dm−3) were in agreement with previous work [29]and indicative of a surface chemically pure compound.

J. Eastoe et al. / Journal of Colloid and Interface Science 258 (2003) 367–373 369

2.2.4. High-pressure small-angle neutronscattering (SANS)

Scattering length density calculations at 25◦C indicatedsufficient (but weak) neutron contrast between hydrocarbonsurfactants and CO2. Experiments were conducted with theLOQ time-of-flight instrument [32] at the CLRC RutherfordAppleton Laboratory at ISIS UK, in conjunction with theabove stirred high-pressure cell [33]. SANS determinesthe scattering cross-sectionI(Q) (cm−1) as a function ofthe momentum transferQ (Å−1) = (4π/λ)sin(θ/2) whereλ is the incident neutron wavelength (2.2 → 10 Å) andθ the scattering angle (<7◦). The procedure for datatreatment [34], including correction factors [35], is describedelsewhere. This generally results in an uncertainty of∼15%on the absolute intensity. During data acquisition the contentof the cell was not stirred.

2.2.5. SANS data fittingThe scattering curves were analyzed using the FISH

program [36], which is based on an iterative least-squaresalgorithm. This was used to determine the best structuralparameters, and also to provide a measure of the residuals(sum of weighted squared errors SWSE). An approximateQ

resolution function for LOQ is also included [37]. A single-sphere model proved to be appropriate.

3. Results and discussion

3.1. Triton–CO2 binary systems

3.1.1. SolubilityDetermination of phase transition pressures for the liq-

uid nonionic surfactants examined was found to be morecomplex than for the w/c microemulsions studied previously.Solubility pressures as a function of temperature are shownin Fig. 2, for the different Triton surfactants, at a concentra-tion of 0.05 mol dm−3. The observations with TX-45 and

Fig. 2. Solubility pressures as a function of temperature for Tritonsurfactants in CO2 at 0.05 mol dm−3: (•) TX-45 (P1), (�) TX-100R (P1),(�) TX-100 (P2). See text for definition ofP1 andP2.

TX-100R were similar. At sufficiently high pressures, theliquid surfactants rapidly vanished to form an optically clearsingle phase. With decreasing pressure at constant temper-ature, an onset of turbidity at a pressureP1 appeared: thispoint is consistent with a phase boundary. As is the casewith most solutes in near critical fluids, the solubility of sur-factants studied increases with pressure at fixed temperature.The phase boundary pressureP1 is consistent with the initialseparation of surfactant corresponding to the solubility limitof the surfactant concentration in the cell (0.05 mol dm−3).Below this pressure, phase separation gives rise to an emul-sion of undissolved liquid surfactant in CO2. Such behavioris not uncommon with nonionic surfactants precipitatingfrom solution. The macroscopic droplets scatter light andimpart turbidity that increases with droplet growth and/orconcentration with decreasing pressure. Eventually, at thelower pressureP2, the emulsion becomes unstable, givingrise to a dramatic increase in turbidity. It should be noted thatexperimentally, the lower pressureP2 is difficult to assess vi-sually with great accuracy; therefore the results presented inFig. 2 for TX-100 exhibit a greater error of measurement.The emulsion formed on pressure reduction betweenP1 andP2 may persist for protracted periods of time before a finalseparation into two clear phases results. With TX-100, a sol-ubility of 0.05 mol dm−3 was never thoroughly achieved andthe system remained biphasic. For this surfactant decreasingpressure below 500 bar always resulted in progressive sur-factant precipitation and increased turbidity.

From Fig. 2, it may also be seen that an increase inthe number of ethylene oxide groups reduces solubility inCO2. Both TX-100 and TX-100R have the same EO numberand differences in their phase behavior follow the generaltrend of decreasing solubility of organic compounds in CO2with increasing aromaticity [38]. Commercial grade non-ionics such as Triton surfactants represent a considerabledistribution of EO numbers in the polyethylene oxide chain.The decreasing solubility with increasing EO number notedabove may therefore result in a pressure-dependent frac-tionation of the surfactant [39] belowP1. This effect wasexamined by sampling the clear CO2 phase in equilibriumwith excess surfactant and is illustrated in Fig. 3 for TX-100.The data show the change in distribution of the EO numberprofile with pressure for a surfactant-saturated CO2 solution(i.e., excess TX-100). At low pressure the distribution ofdissolved surfactant is heavily skewed toward the more solu-ble, lower EO number components. With increasing pressurethe EO number distribution approaches that of the TX-100batch used but remains richer in the lower molecular weightcomponents. This behavior is reminiscent of fractionationtechniques that have previously been exploited with super-critical fluids [40]. It will be general and confirms that, belowP1, the distribution of surfactant oligomers in micelles mayvary with pressure, temperature, and overall surfactant con-centration.

In general, these observations contrast widely with phasebehavior seen for water-in-oil (w/o) [30] and water-in-CO2


Fig. 3. Distribution of EO oligomers in CO2 saturated with TX-100 asresolved by HPLC.T = 37◦C. (A) 220 bar, (B) 340 bar, (C) 400 bar,(D) TX-100 as used.

Table 1Fitted form factor and Guinier radii for the Triton surfactants

Triton surfactant FittedR ± 1 (Å) GuinierR ± 1 (Å)

TX-45 16 13TX-100R 18 16TX-100 14 11

(w/c) [10,36] microemulsions stabilized by chemically well-defined surfactants. In these three-component systems, re-versible catastrophic phase separation is normally observedat a single pressure, below which all the surfactant and waterseparate.

3.1.2. SANSIn Fig. 4, high-pressure SANS data confirm the presence

of spheroidal aggregates in sc-CO2. Compared to fluo-rocarbon surfactant-stabilized w/c systems [10,36,41], thescattering intensity is very low. Radii were obtained fromthe fitted curves and also deduced using the Guinier ap-proximation (Table 1); the results are consistent with reversemicelles. A general observation with all systems was thatbelow the phase boundaryP2 the scattering disappeared,confirming that this condition represents a complete separa-tion of micelles. The quality of the form-factor fits and theiroverall agreement with the Guinier method demonstrate theabsence of significant attractive interactions between the mi-celles. This provides further evidence for CO2 compatibilityof these branched-chained nonionic surfactants.

Fig. 4. SANS data obtained after subtraction of the cell and sc-CO2background for Triton surfactants:() TX-45, (✷) TX-100R,(◦) TX-100at 0.05 mol dm−3 and at 40◦C and 500 bar. Example error bars are shownfor TX-100 only.

Fig. 5. Solubility pressures (P1 vsT ) of binary systems of TX-100R in CO2as a function of concentration:(�) 0.05 and(�) 0.10 mol dm−3.

3.1.3. Concentration effectFigures 5 and 6 show respectively the phase behavior

and the neutron-scattering curves for TX-100R at differ-ent concentrations. In Fig. 5, the isopleths show solubilityincreasing with pressure (at constant temperature) and de-creasing with temperature (at constant pressure). This isnormal behavior in supercritical fluids and similar resultswere obtained for Dynol 604 in sc-CO2, which is also amethyl-branched acetylenic glycol-based nonionic surfac-tant [27]. According to the phase behavior in Fig. 5, theSANS experiments in Fig. 6 were carried out well withinthe single region (aboveP1) for the 0.06 mol dm−3 sam-ple, but were between the pressuresP1 andP2 for TX-100Rat 0.1 and 0.17 mol dm−3. The P1 phase transition couldbe higher still for TX-100R at 0.17 mol dm−3. However, inFig. 6, the scattering intensity increases with concentration.This behavior can be attributed to a combination of effects.First, because of the weak intensity, it is difficult to assessthe background accurately. Second, with TX-100R at 0.1


Fig. 6. SANS data for TX-100R as a function of concentration at 0.17(◦)(below P1), 0.11(✷) (below P1), and 0.06() (single phase) mol dm−3

at 40◦C and 500 bar. Example error bars are shown for the highestconcentration only.

and 0.17 mol dm−3, the systems are biphasic (betweenP1andP2) and fractionation effects may occur as observed forTX-100 (Fig. 3). Increasing the amount of excess surfactantin this system may result in an increased concentration of themore soluble, low-EO-number components.

The behavior of these micelle–CO2 systems again con-trasts with w/c microemulsions, for which water dropletconcentration relates simply to surfactant concentration [10,36,41].

3.2. Anionic–nonionic surfactant mixtures

Mixtures of TX-45 and AOT4 were also examined tosee whether synergistic effects may enhance overall solubil-ity [42]. For a mixture, surfactant composition is defined bya mole fractionX,

X = [TX-45][TX-45] + [AOT4] .

Figures 7 and 8 show, respectively, phase behavior andSANS data for TX-45 and AOT4 on their own in sc-CO2 andin a mixture atX = 0.4. Owing to the limitations of the high-pressure cell, the phase boundary of AOT4 at 0.05 mol dm−3

could not be established. Furthermore, in Fig. 7, the onset ofturbidityP1 for the mixture of AOT4 and TX-45 was close tothe cell operating limit. Hence, it was only possible to carryout SANS measurements in the vicinity of the phase bound-ary (500 bar), just belowP1 where cloudiness develops. Asshown in Fig. 8, on addition of AOT4 the scattering of themixed micelles increases above that for pure TX-45 aggre-gates. This is consistent with a difference in the backgroundmonomer solubility level between the two surfactants: thenonionic would then have an effectively higher “cmc” thanthe ionic. (This could also be due to changes in neutron con-trast, but calculations of the sum of scattering lengths

∑b

do not explain the factor of three changes in intensity.) Inspite of the proximity of cloud points, there are clearly ag-

Fig. 7. Solubility pressures for TX-45 and a mixture of AOT4 and TX-45 inCO2. The total surfactant concentration is 0.05 mol dm−3. (�) X = 0.4,[TX] = 0.02 mol dm−3, [AOT] = 0.03 mol dm−3; (�) X = 1 (pureTX-45).

Fig. 8. SANS data obtained at 40◦C and 500 bar for a mixture of AOT4 andTX-45 and the individual components. The total surfactant concentration is0.05 mol dm−3. Example error bars are shown for AOT4 only. (�) X =0.4, [TX] = 0.02 mol dm−3, [AOT] = 0.03 mol dm−3; (�) X = 1 (pureTX-45); (•) X = 0 (pure AOT4).

Table 2Fitted radii form factor for a mixture of AOT4 and TX-45 and the individualcomponents in sc-CO2

[AOT4] (mol dm−3) [TX-45] (mol dm−3) X R ± 1 (Å)

0.05 0 0 140 0.05 1 160.03 0.02 0.4 17

gregates present in CO2 for both the pure surfactants and themixture. The fitted radii are summarized in Table 2. Thesevalues are consistent with spheroidal reverse micelles of ap-proximately the same size, irrespective of composition.


3.3. Water uptake

It would be of great interest if water could be dispersedin the polar cores of the systems described above. Waterwas added to the binary systems (W = [H2O]/[TX] = 5),with an additional amount to account for the small watersolubility in CO2. It was noted that in some cases addedwater caused the precipitation of the surfactant. This haspreviously been observed with mixtures of the surfactantpentaethylene glycoln-octyl ether in sc-CO2 [17]. Near-infrared (NIR) studies [10] showed that water in CO2 alonegives rise to a sharp-centred peak at∼1390 nm, whereasthe signal from water in w/c microemulsion droplets is amuch broader band at∼1440 nm. However, NIR spectraobtained for the Triton surfactants in the presence of waterindicated no clear evidence of water solubilization in themicelles and the signals were consistent with water mainlylocated in the bulk CO2. Further attempts with brine (0.05and 0.10 mol dm−3) proved to be no more successful.

4. Summary and conclusions

In this work, some light has been shed on propertiesof cheap hydrocarbon surfactants for CO2. High-pressureSANS experiments demonstrated that commercial nonionicTriton surfactants and AOT4 form spheroidal reverse mi-celles in sc-CO2. These surfactants are similar in structure,as they possess highly methyl-branched tails. It is nowwell established that this chemical structure shows a favor-able affinity for CO2. Experiments with Triton surfactantsrevealed that solubility was temperature and pressure depen-dent. As may be expected, our measurements demonstratethat shorter ethylene oxide chain surfactants dissolve toform micelles at lower pressures. Because of their polydis-persity in EO groups, biphasic mixtures of these nonionicsurfactants with CO2 are also affected by fractionation, withselective solubilization of the more soluble lower EO num-ber components at lower pressures. Formation of mixedmicelles of AOT4 and TX-45 again demonstrates the ef-fectiveness of the highly methyl-branched tail for micelleformation in CO2.

Unfortunately no water or brine could be incorporatedinto the polar cores of the micelles, but these systems havefound application in metal ion extraction into CO2 fromwater [43].

To conclude, the hydrocarbon surfactants presented inthis study aggregate in CO2. However, despite their low cost,the solubility conditions remain extreme, and a major limi-tation is an inability to disperse water. Although aerosol–OTdoes not dissolve in CO2, highly methyl-branched ana-logues form spheroidal aggregates in sc-CO2 (e.g., AOT4).In the case of ethylene oxide nonionic surfactants, it alsoappears that the degree of methylation of the chain tips pro-motes solubility in sc-CO2 commensurate with the ethyleneoxide head. Therefore,t-butyl tips can be considered CO2-

compatible groups. It is to be hoped that these principles,combined with previous knowledge of CO2-active groups,will be helpful in the future in designing “CO2-philic” hy-drocarbon surfactants.

Acknowledgments

This work was funded under EPSRC Grants GR/LO5532and GR/L25653. The EPSRC and the Universities of Bris-tol and East Anglia are thanked for studentship support. Wethank CLRC for allocating beam time at ISIS and contribu-tions towards consumables and travel. Studentship supportfor Alexandre Gurgel from CNPq-Brazil is also gratefullyacknowledged.

References

[1] M.L. Campbell, D.L. Apocada, M.Z. Yates, T.M. McCleskey, E.R.Birnbaum, Langmuir 17 (2001) 5458.

[2] H. Shiho, J.M. DeSimone, Macromolecules 34 (2001) 1198.[3] A.I. Cooper, C.D. Wood, A.B. Holmes, Ind. Eng. Chem. Res. 39

(2000) 4741.[4] H. Ohde, J.M. Rodriguez, X.R. Ye, C.M. Wai, Chem. Commun. (2000)

2353.[5] Y.-P. Sun, P. Atorngitjawat, M.J. Meziani, Langmuir 17 (2001) 5707.[6] P.S. Shah, S. Hussain, K.P. Johnston, B.A. Korgel, J. Phys. Chem.

B 105 (2001) 9433.[7] M. Motiei, Y. Rosenfeld Hacohen, J. Calderon-Moreno, A. Gedanken,

J. Am. Chem. Soc. 123 (2001) 8624.[8] K.A. Consani, R.D. Smith, J. Supercrit. Fluids 3 (1990) 51.[9] T.A. Hoefling, R.M. Enick, E.J. Beckman, J. Phys. Chem. 95 (1991)

7127.[10] J. Eastoe, A. Downer, A. Paul, D.C. Steytler, E. Rumsey, J. Penfold,

R.K. Heenan, Phys. Chem. Chem. Phys. 2 (2000) 5235.[11] Z.T. Liu, C. Erkey, Langmuir 17 (2001) 274.[12] R. Fink, E.J. Beckman, J. Supercrit. Fluids 18 (2000) 101.[13] P.A. Psatthas, S.R.P. da Rocha, C.T. Lee, K.P. Johnston, K.T. Lim,

S. Weber, Ind. Eng. Chem. Res. 39 (2000) 2655.[14] T. Sarbu, T. Styranec, E.J. Beckman, Nature 405 (2000) 165.[15] S.R.P. da Rocha, K.L. Harrison, K.P. Johnston, Langmuir 15 (1999)

419.[16] G.G. Yee, J.L. Fulton, R.D. Smith, Langmuir 8 (1992) 377.[17] G.J. McFann, K.P. Johnston, S.M. Howdle, AIChE J. 40 (1994) 543.[18] J. Liu, B. Han, G. Li, Z. Liu, J. He, G. Yang, Fluid Phase Equilib-

ria 187–188 (2001) 247.[19] T. Ihara, N. Suzuki, K. Maeda, K. Sagara, T. Hobo, Chem. Pharm.

Bull. 43 (1995) 626.[20] B.H. Hutton, J.M. Perera, F. Grieser, G.W. Stevens, Colloids Surf.

A Physicochem. Eng. Aspects 146 (1999) 227.[21] J. Eastoe, A. Paul, S. Nave, D.C. Steytler, E. Rumsey, R.K. Heenan,

J. Am. Chem. Soc. 123 (2001) 988.[22] E.J. Kim, S.H. Hahn, Mater. Sci. Eng. A Struct. Mater. Prop. Micro-

struct. Process. 303 (2001) 24.[23] X. Fu, S. Qutubuddin, Colloids Surf. A Physicochem. Eng. As-

pects 179 (2001) 65.[24] Z. Liu, R. Bendayan, X.Y. Wu, J. Pharm. Pharmacol. 53 (2001) 779.[25] T. Nyholm, J.P. Slotte, Langmuir 17 (2001) 4724.[26] H. Hagerstrand, J. Bobacka, M. Bobrowska-Hagerstrand, V. Kralj-

Iglic, M. Fosnaric, A. Iglic, Cell. Mol. Biol. Lett. 6 (2001) 161.


[27] J. Liu, B. Han, G. Li, X. Zhang, J. He, Z. Liu, Langmuir 17 (2001)8040.

[28] N. Yoshino, N. Komine, J. Suzuki, Y. Arima, Bull. Chem. Soc. Jpn. 64(1991) 3262.

[29] S. Nave, J. Eastoe, J. Penfold, Langmuir 16 (2000) 8733.[30] J. Eastoe, B.H. Robinson, D.C. Steytler, J. Chem. Soc. Faraday

Trans. 86 (1990) 511.[31] J. Eastoe, S. Nave, A. Downer, A. Paul, A. Rankin, K. Tribe,

J. Penfold, Langmuir 16 (2000) 4511.[32] S.M. King, R.K. Heenan, The LOQ Instrument Handbook, Rutherford

Appleton Laboratory Report RAL-TR-96-036, CCLRC, Didcot, UK,1996.

[33] J. Eastoe, B.H. Robinson, W.K. Young, D.C. Steytler, J. Chem. Soc.Faraday Trans. 86 (1990) 2883.

[34] R.K. Heenan, J. Penfold, S.M. King, J. Appl. Crystallogr. 30 (1997)1140.

[35] G.D. Wignall, F.S. Bates, J. Appl. Crystallogr. 20 (1987) 28.[36] J. Eastoe, B.M.H. Cazelles, D.C. Steytler, J.D. Holmes, A.R. Pitt,

T.J. Wear, R.K. Heenan, Langmuir 13 (1997) 6980.[37] P.G. Cummins, E. Staples, J. Penfold, J. Phys. Chem. 94 (1990) 3740.[38] J. Chrastil, J. Phys. Chem. 86 (1982) 3016.[39] C.A. Eckert, M.P. Ekart, B.L. Knutson, K.P. Payne, C.L. Liotta,

N.R. Foster, Ind. Eng. Chem. Res. 31 (1992) 1105.[40] J.A. Pratt, M.A. McHugh, J. Supercrit. Fluids 9 (1996) 61.[41] J. Eastoe, A.M. Downer, A. Paul, D.C. Steytler, E. Rumsey, Prog.

Colloid Polym. Sci. 115 (2000) 214.[42] B.P. Binks, P.D.I. Fletcher, D.J.F. Taylor, Langmuir 13 (1997) 7030.[43] J.D. Holmes, submitted for publication.

Documents

Micellization of economically viable surfactants in CO2