Mesh Phases in a Ternary Nonionic Surfactant, Oil, and Water System

Nov 15, 2006 - Lancashire, Preston PR1 2HE, Lancashire, England, and Physical Chemistry 1, Lund UniVersity,. S-221 00 Lund, Sweden. ReceiVed March 29 ...
3 downloads 0 Views 204KB Size
Langmuir 2006, 22, 10951-10957

10951

Mesh Phases in a Ternary Nonionic Surfactant, Oil, and Water System Ying Wang,† Michael C. Holmes,*,† Marc S. Leaver,† and Andrew Fogden‡,§ Centre for Materials Science, Department of Physics, Astronomy and Mathematics, UniVersity of Central Lancashire, Preston PR1 2HE, Lancashire, England, and Physical Chemistry 1, Lund UniVersity, S-221 00 Lund, Sweden ReceiVed March 29, 2006. In Final Form: August 16, 2006 The binary system of hexaethylene glycol n-hexadecyl ether (C16EO6) and water (2H2O) has a complex, temperaturedependent lyotropic phase sequence, in the concentration region of 48-62 wt %. On cooling it shows the sequence lamellar phase, LR, random mesh phase Mh1(0), rhombohedral mesh phase, Mh1(R3hm), bicontinuous cubic phase, V1(Ia3hd), and a two-phase hexagonal region, H1 + Lβ. On heating from the latter two-phase region the phase sequence is V1(Ia3hd), Mh1(0), and LR. Polarizing optical microscopy, 2H nuclear magnetic resonance, and small-angle X-ray scattering have been used to study the stability of these phases, their sequence, and their physical parameters with the addition of the oils, 1-hexene, decane, and octadecane. The oils are located within the alkyl chain regions of the mesophase structures. Depending on whether the added oil is “penetrating” or “swelling”, it may reside in the region between the C16 alkyl chains of the surfactant or at the center of the bilayer and affect phase stability. Oils affect both the volume of the alkyl chain region (at fixed surfactant water mole ratio) and the rigidity of the interfacial region. Both effects can influence the phase structures and their ranges of stability. Adding different types of oil to the mesh phases gives an opportunity to understand the factors that are important in their formation. The transition from the Mh1(R3hm) phase to Mh1(0) phase is triggered by the hydrocarbon region swelling to a critical volume fraction of 0.32, a surfactant rod radius of ∼1.75 nm, and a critical water layer thickness of ∼2.5 nm. The latter is most likely responsible for a weakening of the interlayer headgroup overlap interaction and the loss of correlation between the layers. The lamellar phase becomes the only stable phase at high oil content.

1. Introduction Intermediate phase formation in surfactant-water systems, as well as in block copolymer melts, is an area which is attracting attention. They were once thought to be anomalies in the wellestablished phase sequence of hexagonal to lamellar phases. However the number of systems in which intermediate phases either replace or exist with bicontinuous cubic phases between the hexagonal and lamellar phases is increasing.1-16 There are * Corresponding author. E-mail: [email protected]. † University of Central Lancashire. ‡ Lund University. § Current address: Department of Applied Mathematics, Australian National University, Canberra ACT 0200, Australia. (1) Auvray, X.; Petipas, C.; Rico, I.; Lattes, A. Liq. Cryst. 1994, 17, 109-126. (2) Burgoyne, J.; Holmes, M. C.; Tiddy, G. J. T. J. Phys. Chem. 1995, 99, 6054-6063. (3) Chidichimo, G.; Vaz, N. A. P.; Yaniv, Z.; Doane, J. W. Phys. ReV. Lett. 1982, 49, 1950-1954. (4) Fairhurst, C. E.; Fuller, S.; Gray, J.; Holmes, M. C.; Tiddy, G. J. T. In Handbook of Liquid Crystals; Demus, D., Gray, G. W., Goodby, J. W., Spiess, H. W., Vill, V., Eds.; Wiley-VCH Verlagsgesellschaft mbH: Weinheim, Germany, 1998; pp 341-392. (5) Fogden, A. S.; Stenluka, M.; Fairhurst, C. E.; Holmes, M. C.; Leaver, M. S. Prog. Colloid Polym. Sci. 1998, 108, 129-138. (6) Hagsla¨tt, H.; So¨derman, O.; Jo¨nsson, B. Liq. Cryst. 1992, 12, 667-688. (7) Hall, C.; Tiddy, G. J. T. In Surfactants in Solution, 8th ed.; Mittal, K. L., Ed.; Plenum Press: New York, 1989; pp 9-23. (8) Hamley, I. W.; Gehlsen, M. D.; Khandpur, A. K.; Koppi, K. A.; Rosedale, J. H.; Schulz, M. F.; Bates, F. S.; Almdal, K.; Mortensen, K. J. Phys. II 1994, 4, 2161-2186. (9) Hashimoto, T.; Koizumi, S.; Hasegawa, H.; Izumitani, T.; Hyde, S. T. Macromolecules 1992, 25, 1433-1439. (10) Henriksson, U.; Blackmore, E. S.; Tiddy, G. J. T.; So¨derman, O. J. Phys. Chem. 1992, 96, 3894-3902. (11) Holmes, M. C. Curr. Opin. Colloid Interface Sci. 1998, 3, 485-492. (12) Hyde, S. T. Unpublished conference proceedings, 1990. (13) Ke´kicheff, P.; Cabane, B. J. Phys. (Paris) 1987, 48, 1571-1583. (14) Ke´kicheff, P.; Tiddy, G. J. T. J. Phys. Chem. 1989, 93, 2520-2526. (15) Laradji, M.; Shi, A. C.; Noolandi, J.; Desai, R. C. Macromolecules 1997, 30, 3242-3255. (16) Seddon, J. M. Biochim. Biophys. Acta 1990, 1031, 1-69.

three structural classes: ribbons; meshes; bicontinuous.3,5,6,11,17-20 The number of these structures is limited, and indeed, the structure of a bicontinuous intermediate is yet to be reliably confirmed, which reflects the difficulty in identifying these phases rather than their rarity. Studies of binary nonionic surfactant systems containing polyoxyethylene headgroups show that when the surfactant molecule is “balanced”, i.e., the polar and nonpolar parts of the molecule have comparable lengths, intermediate phases can replace the bicontinuous cubic phases found in shorter alkyl chain homologues.2,21-23 The hexaethylene glycol n-hexadecyl ether (C16EO6) and water (2H2O) system exhibits the following phase sequence at 55 wt % of surfactant.21 cooling:

LR f 45 °C f Mh1(0) f33 °C f Mh1 (R3hm) f 29 °C f V1(Ia3hd) f 27 °C f H1 + Lβ heating:

H1 + Lβ f 27 °C f V1(Ia3hd) f 33 °C f Mh1 (0) f 45 °C f LR Here LR is a classical lamellar phase, Mh1(0) is a disrupted lamellar (17) Anderson, D. M.; Davis, H. T.; Scriven, L. E.; Nitsche, J. C. C. In AdVances in Chemical Physics; Prigogine, I., Rice, S. A., Eds.; John Wiley: New York, 1990; Vol. LXXVII, pp 337-396. (18) Fogden, A.; Hyde, S. T. Eur. Phys. J., B 1999, 7, 91-104. (19) Hyde, S. T. Curr. Opin. Solid. State Mater. Sci. 1996, 1, 653-662. (20) Holmes, M. C.; Leaver, M. S. In Bicontinuous Liquid Crystals; Lynch, M. L., Spicer, P. T., Eds.; CRC Press: Taylor & Francis Group, Boca Raton, FL, 2005; pp 15-39. (21) Fairhurst, C. E.; Holmes, M. C.; Leaver, M. S. Langmuir 1997, 13, 49644975. (22) Funari, S. S.; Holmes, M. C.; Tiddy, G. J. T. J. Phys. Chem. 1994, 98, 3015-3023. (23) Leaver, M. S.; Fogden, A.; Holmes, M. C.; Fairhurst, C. E. Langmuir 2001, 17, 35-46.

10.1021/la060840c CCC: $33.50 © 2006 American Chemical Society Published on Web 11/15/2006

10952 Langmuir, Vol. 22, No. 26, 2006

or random mesh phase, Mh1(R3hm) is a rhombohedral mesh phase, V1(Ia3hd) is a normal bicontinuous cubic phase, and H1 + Lβ is a two-phase hexagonal plus Lβ phase region, with the nomenclature of the mesh phases adopted from ref 11. The curvature of the surfactant-water interface, at fixed concentration and pressure, increases with decreasing temperature as the headgroups become more hydrated.24-26 The metastable Mh1(R3hm) phase has a mesh structure in which the three connected nodes of the layers are correlated between layers to give an ABC packing and a rhombohedral symmetry.5,23 The cubic phase also consists of three connected nodes, although these are not planar but rather form the familiar Ia3d space group.27 Thermodynamically the Mh1(R3hm) and cubic phases are finely balanced in favor of the latter.5,21 This balance can be perturbed by the addition of a third component. The addition of oil to surfactant-water mixtures has been widely studied and leads to the formation of microemulsions28-32 and the more concentrated oil swollen lyotropic liquid crystalline phases.33,34 Most lyotropic phases observed in binary systems also form with added oil, but the more complex the phase the lower the swelling limit. Ninham et al. observed that, for ionic surfactants, microemulsions are formed at lower water content depending on the nature of the oil.35 This is related to the uptake of oil in the system. The observed behavior can be simply explained with reference to the surfactant parameter a concept developed to explain the choice of aggregate formed in solution by a surfactant on the basis of its molecular architecture.36,37 They proposed that saturated hydrocarbon oils, with chain lengths much shorter than that of the surfactant, will be able to reside in the region between the surfactant chains. These they termed “penetrating” oils, which will have a direct effect on the surfactant parameter and the interfacial curvature of the phase.33 Conversely saturated oils with tail lengths longer than the surfactant cannot penetrate into the surfactant tail region and therefore have no real effect on the surfactant parameter; rather they simply thicken the layer or swell the micelle.33 The penetration of the oil can be enhanced further by using unsaturated oil since it is more hydrophilic.35 The swelling and penetrating nature of oil can be compared with the work done on the spreading of alkanes on aqueous solutions of surfactants. It has been shown38,39 that the ability of an oil to spread on a surface is determined both by the oil (24) Mitchell, D. J.; Tiddy, G. J. T.; Waring, L.; Bostock, T.; Mcdonald, M. P. J. Chem. Soc., Faraday Trans. 1 1983, 79, 975-1000. (25) Andersson, M.; Karlstro¨m, G. J. Phys. Chem. 1985, 89, 4957-4962. (26) Karlstro¨m, G. J. Phys. Chem. 1985, 89, 4962-4964. (27) Seddon, J. M.; Templer, R. H. Philos. Trans. R. Soc. London, Ser. A 1993, 344, 377-401. (28) Kahlweit, M.; Strey, R.; Haase, D.; Kunieda, H.; Schmeling, T.; Faulhaber, B.; Borkovec, M.; Eicke, H.-F.; Busse, G.; Eggers, F.; Funck, T.; Richmann, H.; Magid, L.; So¨derman, O.; Stilbs, P.; Winkler, J.; Dittrich, A.; Jahn, W. J. Colloid Interface Sci. 1987, 118, 436-453. (29) Shinoda, K.; Lindman, B. Langmuir 1987, 3, 135-149. (30) Olsson, U.; Wennerstrom, H. AdV. Colloid Interface Sci. 1994, 49, 113146. (31) Eicke, H. F.; Meier, W.; Hammerich, H. Colloids Surf., A 1996, 118, 141-148. (32) Fletcher, P. D. I. Curr. Opin. Colloid Interface Sci. 1996, 1, 101-106. (33) Kunieda, H.; Ozawa, K.; Huang, K. L. J. Phys. Chem. B 1998, 102, 831-838. (34) Leaver, M. S.; Olsson, U.; Wennerstrom, H.; Strey, R.; Wurz, U. J. Chem. Soc., Faraday. Trans. 1995, 91, 4269-4274. (35) Ninham, B. W.; Chen, S. J.; Evans, D. F. J. Phys. Chem. 1984, 88, 58555857. (36) Israelachvili, J.; Mitchell, D. J.; Ninham, B. W. J. Chem. Soc., Faraday Trans. 2 1976, 72, 1525-1568. (37) Mitchell, D. J.; Ninham, B. W. J. Chem. Soc., Faraday Trans. 2 1981, 77, 601-629. (38) Aveyard, R.; Binks, B. P.; Fletcher, P. D. I.; MacNab, J. R. Langmuir 1995, 11, 2515-2524. (39) Matsubara, H.; Aratono, A.; Wilkinson, K. M.; Bain, C. D. Langmuir 2006, 22, 982-988.

Wang et al.

Figure 1. Schematic phase diagram on cooling for the three systems of the C16EO6/2H2O/1-hexene (solid line + vertical fill), C16EO6/ 2H O/decane (dashed line + horizontal fill), and C EO /2H O/ 2 16 6 2 octadecane (dotted line + no fill). Note that the octadecane concentration limit for the V1(Ia3hd) phase was not determined but was less than 0.0094 mole fraction.

chain length and by the structure of the surfactant. It has been concluded that the penetration of oils into the surface monolayer of an aqueous solution of a surfactant depends on the surfactant alkyl chain structure40 and upon the oil molecules filling the voids between the chains.41 It is likely that the ability of an oil to penetrate between the surfactant alkyl chains will be controlled by similar factors. If oil is penetrating, there is a concomitant increase in the surfactant parameter, assuming the surface area/molecule constraints at the interface remain more or less unaltered, favoring phases with lower interfacial curvature or even a change in the sign of the curvature. If the oil is swelling, the phase structures must swell, and phases such as cubic or mesh will be lost. The swollen layers can no longer form complex three-dimensional structures either because of a simple steric inability to “pack” or because of an increase in the rigidity of the structural units. In this paper results are presented from the addition of octadecane, decane, and 1-hexene (in increasing order of penetration) to the binary C16EO6 system. Polarizing optical microscopy and 2H nuclear magnetic resonance (NMR) measurements were used to establish the extent of single-phase regions. The structure of the phases was established using a combination of NMR and small-angle X-ray scattering (SAXS). 2. Experimental Section 2.1. Samples. The hexaethylene glycol n-hexadecyl ether (C16EO6) was synthesized by Nikko Chemicals (Tokyo, Japan) and supplied by Chesham Chemicals (Harrow, U.K.). The surfactant had a quoted purity >98% and was used as received. In the context of this paper “water” refers to deuterium oxide (2H2O) (purity >99.8%) which was supplied by Fluorochem Ltd. (Old Glossop, U.K.) and was also used as received. Octadecane (purity >99%), decane (purity >99%), and 1-hexene (purity >97%) were supplied and used as received by Aldrich (Dorset, U.K.) and Lancaster Synthesis (Lancaster, U.K.). In the binary system, the interesting phase transitions are thermally induced for samples that contain between 51 and 59 wt % surfactant. In all of the experiments the surfactant to water mole ratio was (40) Eastoe, J.; Hetherington, K. J.; Sharpe, D.; Dong, J. F.; Heenan, R. K.; Steytler, D. Langmuir 1996, 12, 3876-3880. (41) Binks, B. P.; Crichton, D.; Fletcher, P. D. I.; MacNab, J. R.; Li, Z. X.; Thomas, R. K.; Penfold, J. Colloids Surf., A 1999, 146, 299-313.

Nonionic Surfactant, Oil, and Water Mesh Phase

Langmuir, Vol. 22, No. 26, 2006 10953

maintained equivalent to that of a 55 wt % C16EO6 binary sample (mole ratio of surfactant to water of 0.052), where the intermediate phase possesses its maximum thermal extension upon cooling from the Mh1(0) phase.21 All the oil addition samples were made ensuring that this surfactant-to-water ratio remained constant and the third component was introduced as required. The amount of added oil is specified by the mole fraction of oil defined by noil/(noil + nsurf), where noil is the number of moles of oil and nsurf is the number of moles of surfactant. Samples were prepared by weighing the oil into glass tubes containing constrictions. Masses of C16EO6 and 2H2O were then calculated for this mass of oil so that the surfactant-to-water ratio was maintained. These masses were added to the tube which was flame sealed. Samples were mixed by heating to 50 °C and repeatedly centrifuging the sample through the constriction. Once the samples were homogeneous (judged by visual inspection), they were left to equilibrate in the lamellar phase for several days to ensure complete mixing. Once opened, 0.5 or 0.7 mm Lindman capillaries, 5 mm NMR tubes, and 0.2 mm path length flat capillaries or microslides were filled from the sample. Tubes and capillaries were flame sealed and the sample level marked. Prior to experiments being carried out the samples were left at in an oven at 50 °C overnight to ensure no solvent was lost due to incomplete sealing. For the SAXS experiments, the volume fraction of hydrophobic materials (φa) now has to contain the oil added to the system (which implicitly assumes that the oil is totally insoluble in water). We define this quantity as φa )

volume of hydrophobic moieties ) total volume Mchain Moil + Fchain Foil Mchain Moil MEO Mw + + + Fchain Foil FEO Fw

where Ma, Mo, MEO, and Mw are the relative molecular masses of the surfactant alkyl chain, oil, surfactant headgroup, and water, respectively. The corresponding densities for the surfactant moieties and water are 832 (alkyl chain), 1150 (headgroup), and 1100 kg m-3. The densities of the oils are 777 (octadecane), 730 (decane), and 673 (1-hexene) k gm-3 as provided in the chemical data sheets of the supplier. 2.2. Experimental Techniques. Details of the optical polarizing microscopy, 2H NMR, and small-angle X-ray scattering (SAXS) have been described elsewhere.21 In systems containing nonionic surfactants, Rendall et al.42 showed that the 2H NMR quadrupolar splitting (∆ν) obtained from 2H2O bound to the first one or two ethylene oxide (EO) groups of the surfactant molecule can be expressed as 1 ∆ν ) ∆ν0pbSb 2

(1)

where pb is the fraction of bound 2H2O molecules, Sb is the orientational order parameter of the bound water, and ∆νo is the splitting associated with a bound 2H2O molecule. In the binary system the variation of the quadrupolar splitting as a function of temperature and surfactant concentration helps to delineate the phase transitions and distinguishes between single and coexisting phases.21 Samples show Pake powder patterns indicating no particular orientations of the phases, although there was some reproducible sharpening of the powder pattern indicating ordering of the samples at the transition into the Mh1(R3hm) phase. This was used as an additional indicator for the formation of the phase. The indexing and analysis of SAXS patterns has been described elsewhere.5,21,23 (42) Rendall, K.; Tiddy, G. J. T. J. Chem. Soc., Faraday Trans. 1 1984, 80, 3339-3357.

Figure 2. Comparison of the LR phase bilayer spacing do with the two ideal swelling and penetration states for (a) octadecane and (b) 1-hexene systems at 50 °C. The upper solid line corresponds to a completely swelling state calculated from eq 3, and the lower line corresponds to a completely penetrating state calculated from eq 2, using νchain ) 0.45 nm3, νoctadecane ) 0.544 nm3, ν1-hexene ) 0.206 nm3, doo ) 6.2 nm, and φchaino ) 0.304.

Figure 3. Variation of the first-order peak position 1/do° in the LR phase as a function of the volume fraction of 2H2O, φw, at 50 °C.

3. Results and Discussion 3.1. Phase Behavior. Slices of the phase prism have been mapped by combining the results from optical microscopy and 2H NMR. Since there are now three components in the system, there are multiphase regions, the extent and nature of which are difficult to establish. Therefore, the phase behavior presented is schematic, showing the major single phase regions and progressing from most to least penetrating oil in the sequence 1-hexene, decane, and octadecane, Figure 1. Notice that Figure 1 shows a representation of the phase diagram on cooling; on heating the Mh1(R3hm) phase is replaced by the extended V1(Ia3hd) phase. The detailed method of determining the phase boundaries by 2H NMR can be found in ref 21. 3.1.1. 1-Hexene System. The characteristic optical textures and 2H NMR behavior reported in the binary system21 were

10954 Langmuir, Vol. 22, No. 26, 2006

Wang et al.

Figure 4. Variation of ∆ν as a function of the mole fraction of oil at 50 °C in the LR phase. Error bars are indicated on the 1-hexane results but are the same for other samples. Solid lines are a guide to the eye.

Figure 5. Surface area/surfactant molecule, sa, at 50 °C for oils 1-hexene, decane, and octadecane. Solid lines are a guide to the eye. Errors are indicated on the 1-hexene results. Table 1. Bilayer Spacing, do, Surface Area/Molecule, sa, and 2H NMR Splitting, ∆ν, in the Binary System Compared to Those Found in the 0.5 Mole Fraction Oil Samples at 50 °Ca param

binary

1-hexene

decane

octadecane

(do ( 0.1) nm-1 (sa ( 0.01) nm-2 (∆ν ( 5) Hz

6.2 0.48 708

6.4 0.53 788

(6.8) 0.53 773

8.3 0.49 714

a The parentheses refer to a value extrapolated from adjacent data points.

observed in this system irrespective of the 1-hexene concentrations. As the concentration of 1-hexene was increased, all the phase transition temperatures of the binary system were first lowered and then the cubic and mesh phases lost in the order cubic, Mh1(R3hm) phase and Mh1(0) phase. The cubic phase is unstable above 0.077 mole fraction of added 1-hexene, the Mh1(R3hm) phase is unstable above 0.2 mole fraction, and the Mh1(0) phase is stable below 0.5 mole fraction. Above 0.5 mole fraction of added 1-hexene only the lamellar phase is stable in the temperature interval studied (50-20 °C), either heating or cooling. 3.1.2. Decane System. Again both the Mh1(R3hm) and cubic phases are lost as the concentration of decane is increased. The cubic phase is least stable to the addition of decane and is lost in samples with oil mole fractions greater than 0.032. The Mh1(R3hm) and Mh1(0) phases are stable to 0.08 and 0.167 mole fraction, respectively. The classical lamellar phase is the only phase observed upon heating and cooling for mole fractions of added oil above 0.5. 3.1.3. Octadecane System. For this oil the octadecane concentration limit for the V1(Ia3hd) phase was not determined but was less than 0.0094 mole fraction. The Mh1(R3hm) and Mh1(0) phases are lost with 0.013 and 0.15 mole fraction of oil,

Figure 6. ∆ν vs mole fraction of (a) 1-hexene at temperatures of 50, 40, and 30 °C, (b) decane at 50, 40, and 34 °C, and (c) octadecane at 50, 40, and 34 °C. The vertical arrows show the approximate positions of the transitions from Mh1(0) phase (steeper gradient) to the LR phase (shallower gradient). Error bars are included for the 50 °C results. Solid lines are a guide to the eye.

respectively. As with all the oil addition experiments the Mh1(R3hm) phase is only observed on cooling from the Mh1(0) phase and never upon heating, but interestingly in this series of experiments the cubic phase does not replace the intermediate phase upon heating; rather a two-phase region in which Mh1(0) phase is observed. 3.2. Effect of Oil on the Phases Observed. The phase behavior, 2H NMR, and SAXS results provide information about the evolution of key structural parameters, such as the surface area/ molecule. For the classical lamellar phase the results will be compared with predictions made for ideal oil swelling or penetration. 3.2.1. Lamellar Phase. A classical LR phase is present in all the ternary systems irrespective of oil type or concentration at 50 °C. The SAXS pattern for all the samples in the LR phase exhibit well-defined first- and second-order peaks with repeat distances (d spacings) in the characteristic ratio of 1:2, reflecting the one-dimensional layer structure of the phase. The relationship

Nonionic Surfactant, Oil, and Water Mesh Phase

Langmuir, Vol. 22, No. 26, 2006 10955

Figure 7. Variation in σ as a function of added 1-hexene at 40 °C (calculated from eq 4 within the text). Lines are a guide to the eye. The break point is where phase behavior studies show a transition from the Mh1(0) to LR phase.

Figure 9. 3-connected mesh: (a) triangular box and rod units which form the mesh; (b) single layer of mesh, where a is the unit cell dimension, R ) 60° and β ) 120°; (c) unit cell showing the position of mesh nodes of each mesh layer (the triangular boxes located at the nodes); (d) schematic representation of the molecular arrangement within the mesh structure.

Figure 8. SAXS from the 0.019 mole fraction of 1-hexene sample at (a) 50 °C (La phase), (b) 40 °C (Mh1(0) phase), (c) 31 °C (Mh1(R3hm) phase), and (d) 28 °C (V1(Ia3hd) phase). All patterns recorded on cooling the sample.

between the main lamellar repeat distance, do, and the hydrophobic volume fraction of surfactant alkyl chains, φa, is given by

do )

dhc φa

where dhc is the thickness of the surfactant layer. The corresponding surface area/molecule, sa, can be calculated from

sa )

2Vchain 2Vchain ) dhc φado

where νchain is the volume of the surfactant alkyl chain/molecule, which is 0.45 nm3. In the lamellar phase of the C16EO6/2H2O binary system at 50 °C for a 55 wt % sample, dhc is 2.05 ( 0.01 nm and sa is 0.44 ( 0.02 nm2.5,21 These values are more or less constant for the lamellar phase formed in C16EO6 at 50 °C irrespective of surfactant concentration as would be expected in a classical lamellar phase formed with nonionic surfactants. Addition of an oil leads to

dhc do ) φchain + φoil and

sa )

2(Vchain + fVoil) dhc

Table 2. Approximate Limits of the Stability of the Mh1(R3hM) Phase Shown in Terms of Mole Fraction of Oil and Alkyl Chain Volume Fractiona param

binary

1-hexene

decane

octadecane

mol fractn of oil alkyl chain vol fractn, φa approx rhc (nm)

0 0.304 1.64

0.17 0.32 1.76

0.08 0.32 1.75

0.01 0.31 1.71

a The rod radius is calculated by solving the equation (3)(31/2)r 2(4r hc hc + π(a - 2rhc)) ) (31/2a2cφa)/2.

where νoil is the volume of an oil molecule and where f is defined as f ) (φoil/Voil)/(φchain/Vchain) and corresponds to the number of added oil molecules/surfactant chain in the system. In systems with added oil, there are two ideal states for the oil, namely completely penetrating or swelling.33 For a completely penetrating oil the repeat distance of a lamellar phase is given by eq 2, namely,

(

do ) doo

)

Vchain + fVoilφchaino Vchain + fVoil

(2)

where doo and φchaino are the values recorded in a 55 wt % binary C16EO6 sample under identical external conditions. The corresponding repeat distance predicted for a completely swelling oil is given by eq 3, namely,

(

do ) doo 1 +

)

fνoilφchaino νchain

(3)

Equation 2 predicts that for a completely penetrating oil there should be an observed decrease in do as the amount of oil is incorporated into the system assuming no phase change occurs. Under the same conditions the addition of a completely swelling oil is predicted to increase do, eq 3. These two cases give the ideal types of behavior that can be compared with the measured

10956 Langmuir, Vol. 22, No. 26, 2006

Wang et al.

Table 3. Lattice Parameters for the Mh1(R3hM) Phase Shown as a Function of Oil Type and Concentration 1-hexene

decane

octadecane

mol fractn

a ((0.1 nm)

c ((0.1 nm)

a ((0.1 nm)

c ((0.1 nm)

a ((0.1 nm)

c ((0.1 nm)

0.0 0.0094 0.0103 0.0131 0.0192 0.0769 0.1667

8.30 8.33

17.0 17.4

8.30

17.0

8.30

17.0

8.40

17.4 8.48

17.4

8.47 8.68

17.5 17.7

8.54 8.46 8.77

17.6 17.6 17.7

lamellar repeat distance obtained from SAXS experiments for octadecane (a swelling oil33), Figure 2a, and 1-hexene (a penetrating oil35), Figure 2b, at 50 °C. Note that decane shows intermediate behavior between those of 1-hexene and octadecane. Figure 2a shows that octadecane behaves like an ideal swelling oil. However, Figure 2b shows that 1-hexene is far from being an ideal penetrating oil but has characteristics of both swelling and penetrating. Further evidence for the ideal swelling behavior of octadecane is seen in Figure 3. For ideal swelling the thickness of the water layer, dw, will remain constant for a fixed mole ratio of water to surfactant (sa remains unchanged) even though the oil content is increasing and thereby do. Since dw ) φwdo, where φw is the volume fraction of water, a plot of 1/do against φw should be linear and pass through the origin. As φw approaches zero the lamellar phase (if it were stable) would be all oil and the separation of the bilayers would become very large; i.e. 1/do goes to zero as well. Figure 3 shows that while the octadecane results extrapolate to zero indicating it to be a swelling oil, the other two oils, although linear with φw, do not pass through the origin which indicates some penetrating nature. From eq 1, changes in ∆ν at fixed temperature with variations in oil mole fraction can be explained either by changes in the fraction of 2H2O molecules bound to the EO6 headgroups, pb, or by changes in the order parameter of the bound water, Sb, or a combination of both effects. In the LR phase at 50 °C, ∆ν is either constant (octadecane) or only gradually increasing (1hexene) with an increasing mole fraction of oil, Figure 4. A penetrating oil located between the surfactant chains should cause the surface area/molecule, sa, to increase while a swelling oil should cause no change.33 Figure 5 shows sa as a function of oil mole fraction at 50 °C calculated from do. ∆ν in the 1-hexene system increases with increasing oil mole fraction reflecting the increase in sa calculated from SAXS, whereas both ∆ν and sa are rather insensitive to the addition of octadecane. Table 1 summarizes the bilayer spacing, do, the surface area/molecule, sa, and the 2H NMR splitting ∆ν for the LR phase in the binary sample and ternary samples with 0.5 mole fraction oil at 50 °C. There are some clear trends. While all three oils cause the bilayer spacing, do, to expand, it is the ideal swelling oil, octadecane, that has the biggest effect and 1-hexene that has the least effect; an ideally penetrating oil would have decreased the layer spacing; see Figure 2. From the 2H NMR results it can be seen that the addition of 1-hexene increases ∆ν by approximately 10% since it is expanding the surface area/molecule and increasing either pb or Sb or both. This increase in ∆ν is significant and is equivalent to a decrease in temperature of 15 °C. In contrast octadecane has little or no effect on either sa or ∆ν; therefore, it may be concluded that it does not affect the molecular arrangement at the interfacial region. 3.2.2. Random Mesh Phase, Mh1(0). The effect of oil on the Mh1(0) phase is more difficult to quantify since its structure is that of lamellar planes pierced by water-filled defects. The phase does not possess a simple swelling relationship with either added water21 or swelling oil, and it is also difficult to calculate sa

without making substantial assumptions about the mesh structure.23 However the results from the phase diagram, 2H NMR, and SAXS show that the addition of oil to the Mh1(0) phase causes it to be replaced by the classical lamellar phase. 2H NMR has previously been shown to be sensitive to this subtle structural transition21 and can be identified by SAXS.21,23 A decrease in ∆ν as the temperature is decreased in the binary system marks the onset of the formation of pores in the lamellae and, hence, the transition into the Mh1(0) phase. Figure 6a-c shows ∆ν for 1-hexene, decane, and octadecane, respectively. At a fixed temperature irrespective of the type of oil added, increasing the concentration of the oil in the Mh1(0) phase causes the water-filled defects to be lost and the LR phase to be recovered. The results at 50 °C are shown for comparison. Here optical microscopy and SAXS reveal that all the samples are in the classical LR phase and that ∆ν is nearly constant or only slightly increasing with oil addition. At lower temperatures, 40 and 34 °C, the variation of ∆ν as a function of increasing oil concentration for all three oils is similar. At low oil concentrations, ∆ν increases rapidly prior to leveling out and recovering a ∆ν behavior that is similar to that in the LR phase at higher temperature. Since the behavior of the oils is qualitatively similar, only the results for the 1-hexene will be discussed in detail. At 50 °C ∆ν increases as 1-hexene is added to the LR phase, Figure 6a. This may be due to increases in either pb or Sb in eq 1. At 40 °C in the LR phase a similar behavior is observed as 1-hexene is added. However in the Mh1(0) phase, at oil mole fractions