Assessment of micro-polarity anisotropy as a function of surfactant packing in sodium dodecyl sulphonate–hexane reverse micelles

Abstract

The micro-polarity anisotropy behaviour across the aqueous phase of a SDS (sodium dodecyl sulphonate)–hexane reverse micelle (RM) relies on the SDS packing in the oil–water interfacial self-assembled surfactant structure of the RM.

---

Reverse micelle (RM) systems are advantageous for a number of processes such as nanoparticle synthesis¹ and the liquid–liquid extraction of enzymes.² The effectiveness of these processes relies on the micro-polarity anisotropy (MPA) of the aqueous phase of a RM. The MPA of a RM varies according to its water : surfactant (ω₀) ratio.³ A reliable estimate of MPA is critical to predict and regulate the enzyme activity in RMs.^4,5 The common approaches assess MPA(r⃑) (r⃑ – the radial distance, Fig. 1) in terms of the microstructural characteristics of water in RMs⁶ and the electrostatic potential distribution (ψ(r⃑)) across the aqueous phase (Fig. 1b).^7,8 Spectroscopy studies and molecular dynamic simulations reveal that, in terms of the micro-polarity attributes, the RM aqueous phase comprises at least two spatially separable regions (Fig. 1a and b), the interfacial region (ordered water) and the core region (bulk water).^6,9 The ψ(r⃑) profile that exemplifies the MPA results from the oil–water interfacial charge density levels (per volume of water in a RM) of the surfactant head groups, and the distributional aspects of the counter ions in the aqueous phase. The ψ(r⃑) profiles exhibit an exponential form (Fig. 1b).^6–9 For the aqueous phases in the vicinity of the surfactant-stabilized oil–water interfaces, the ψ profile is attributed to the surfactant chemistries. The ψ profile for these interfacial systems is developed as per the electrical double layer (EDL) models.^10–13 Conversely, as the volume of the confined (nano-sized) aqueous phase of a RM is much lower, the EDL characteristics are governed by the interfacial charge density levels (per RM water content), which decrease as the ω₀ values become higher. The charge density level, even at a particular ω₀ value, could vary depending on the degree of surfactant packing (inter-surfactant distances) in the self-assembled surfactant structure (SASS) of a RM.¹⁴ Thus, the EDL characteristics that determine the MPA are attributed to the surfactant-types and -density/packing in a SASS.^7,8 We assessed the MPA(r⃑) as a function of surfactant density/packing in the SASSs for the SDS (sodium dodecyl sulphonate)–hexane RMs.¹⁵

Fig. 1 (a) An illustration of the spatially segregated regions of a RM: the SASS region (R1), the interfacial region (R2), and the core (R3). A ψ(r⃑) profile is shown (b). A Cryo SEM image (c) shows small RMs of sizes <50 nm (green circles); a few large RMs (red circles). AUC results (d) reveal the RMs as spherical (peak in red circle). R_RM and l_B are the RM radius and the Berrium length, respectively.

The RMs were devised by adding desired amounts of water to SDS–hexane solutions (ESI,† 1). Cryo-SEM (ESI,† 2) images indicated that the RM sizes were <50 nm (ω₀ ≈ 24; Fig. 1c). Analytical ultra-centrifuge (AUC) studies (ESI,† 2) reveal the RMs as spherical in shape (Fig. 1d). The RMs with ω₀ in the range of 5 to 25 were spherical and of sizes in the range of 10 to 50 nm. For ESR studies (ESI,† 1), a 16-DSA (16-DOXYL-stearic acid) probe containing the NO˙ radical was added to the RM samples in desired amounts. The hyperfine splitting constant (a_N) – MPA(r⃑), and the rotational correlation time (τ_c) – surfactant packing in the SASS, were estimated for these RM systems. The application of the ESR technique is a reliable option as the energy absorbance value of a probe with free radicals, which can locate in the aqueous and SASS regions, allows concurrent estimation of the MPA and SASS packing aspects. The microstructural aspects of RM water were studied using IR (attenuated total reflection-infrared) spectroscopy (ESI,† 1).

We compartmentalize a RM's environment into three spatially segregated and sequentially placed regions (Fig. 1a and b), and the MPA(r⃑) of these regions was assessed in terms of their contribution to the RM's ψ(r⃑) profile. The non-linear Poisson–Boltzmann description of the ψ(r⃑) profile that associates these three regions is:

A solution to eqn (1) would require the determination of at least two dimensionless variables: x = κr and ξ = κR_RM, where κ is the inverse of the Debye length, r is the radial distance, and R_RM is the RM size. A numerical solution to these dimensionless variables can be found elsewhere.¹⁰ A general form of ψ(r⃑) for a RM can be written as:where, .

The ξ term is dependent on the association coefficient of the counter ions, given as β.^10,19 The parameters that allow differentiation of the ψ(r⃑) profile for the regions R2 and R3 are:

where, l_B = Bjerrum length.

For R_RM ≤ l_B, ψ decreases in an exponential manner from the interface to the core, and a plateau region of the ψ(r⃑) profile is absent. For R_RM ≫ l_B, the plateau part of the ψ(r⃑) profile represents the r⃑ ≈ 0 − (R_RM − l_B) region (R3). And, this plateau region changes minimally with increase in RM sizes, as the micro-polarity contributions are due to the bulk-type water therein. In order to gain insights into how the ψ(r⃑) profiles differentiate for the RMs of different sizes, the responses of the NO˙ of 16-DSA in RMs of different sizes were determined. These responses were in terms of the energy absorbance values of the NO˙ in the aqueous phases of the RMs. We define the NO˙ radical (in a RM) system as a π-electron system; m π electrons and n core atoms; m + n = N. The charge and position vectors of the jth particle in this system are e_j and r_j, ρ^π_r is the π-electron density, and Q^H_CH is the σ–π parameter. The hyperfine splitting value, a^H_r, for a hydrogen-bonded carbon atom (rth carbon atom) of 16-DSA is:

a^H_r = Q^H_CHρ^π_r and ρ^π_r = c_kr² (4)

The term c_kr² relates to the energy level of the jth particle as:where, χ_r is the rth atomic orbital. The energy levels of the N particle system a^H_r are: , where R is the electric field of the RM aqueous phase, and as per the counter ion density levels in the R2 and R3 regions, and the charge density (surfactant head groups) levels in the R2 region. If NO˙ radicals reside in a spherical RM having an isotropic solvent (relative permittivity ε_r) with a point dipole at the core as p_r, the Block Walker reaction field R_R¹⁶ for NO˙ is:

R_R = (p_r/R_RM³)(ε_r⃑) (6)

The reaction field R_R is stronger as the RM size (R_RM) decreases, indicating that R_R becomes stronger with an increase in the head group charge density levels. As R_R depends on the head group and the counter ion densities in the RMs, the R_R values (as a_N value) exemplify the ψ(r⃑) acting on the NO˙ radicals. We determined the a_N values for RMs of different sizes. The changes in the micro-polarity of the r⃑ = 0 − (R_RM − l_B) region (bulk water) are negligible as a function of the R_RM values (ω₀ levels).^8,16 Thus, the a_N values vary as per the micro-polarity aspects of the R2 region. Therefore, the a_N values, as a function of R_RM, provide a measure of how a_N would be as a function of r⃑ for a RM. We developed the ‘a_Nvs. ω₀’ correlation, which is representative of the a_Nvs. r⃑ for the SDS RMs.

Due to the non-polarity attribute of the R1 region, the a_N values for this region are negligible in comparison to those for the R2 and R3 regions. The rotational coefficient τ_c is particularly pronounced for the non-polar R1 region. Thus, the changes in a_N for the R1 region, and τ_c for the R2 and R3 regions, as a function of ω₀, are negligible. The a_N and τ_c values can be accounted as:

〈Δτ_c〉 = p₁〈Δτ_c〉 + p₂₊₃〈Δτ_c〉 ≈ p₁〈Δτ_c〉 (7)

〈Δa_N〉 = p₁〈Δa_N〉 + p₂₊₃〈Δa_N〉 ≈ p₂₊₃〈Δa_N〉 (8)

where p₁, p₂ and p₃ are fractional contributions of the regions R1, R2 and R3 respectively.

The 〈a_N〉 vs. ω₀(r⃑) profile was considered to assess the MPA(r⃑) of a RM of a particular size.¹⁷Fig. 2 (inset) shows that this profile exhibits a maxima. The lower 〈a_N〉 values are observed as r⃑ → R_RM (interface) and r⃑ → 0 (core). The lower 〈a_N〉 value at the ω₀ ≈ 5 level justifies highly ordered forms of water in the R2 region, and presumably due to the relative population of the head groups and the counter ions across the interfacial region being significantly higher. Across the 5 < ω₀ < 10 range, the 〈a_N〉 values are higher with ω₀, which is due to the increase in the volume of the R2 region having less ordered forms of water (in comparison to the highly ordered region at ω₀ ≈ 5). The drop in the 〈a_N〉 vs. r⃑ profile beyond the maxima, at the ω₀ > 10 levels, is due to the increase in the bulk-type water content of the R3 region. At ω₀ ≫ 10 levels, the bulk water content is higher, and the plateau region of the a_Nvs. r⃑ profile mostly accounts for the micro-polarity of the bulk-type water in the R3 region. Though these inferences that are made from the a_Nvs. r⃑(ω₀) profile provide an assessment of the MPA(r⃑), these inferences are based on the assumption that: the surfactant charge density levels in the SASSs, as per ω₀ levels, exhibit a linear relationship. The surfactant density level, according to the Debye–Stokes–Einstein equation is:

where η indicates micro-viscosity (inter-surfactant distances), NV_chain is the volume of the surfactant chains, and k and T represent the Boltzmann constant and temperature, respectively. As the surfactant density level (NV_chain) decreases proportionately with increase in the ω₀ values, the corresponding τ_c value would be expected to decrease in a linear fashion. Conversely, Fig. 2 (inset) shows that the 〈τ_c〉 value decreases in a non-linear fashion (exponentially) as the ω₀ value increases. This non-linearity confirms that the inter-surfactant distances (as η) in the SASSs vary with the ω₀ levels. This non-linearity can be justified by the fact that the surfactant content required to stabilize a certain volume of RM water need not be same, e.g., the RM shapes could be slightly different. Studies on RM systems, using small angle X-ray scattering, small angle neutron scattering and nuclear magnetic resonance techniques, have revealed that the packing-parameter of a surfactant relies on the effective oil–water interfacial area occupied by the surfactant head-group, which is specific according to the volume of RM water.¹⁸ Wherein, an increase in the polar volume for the RM system induces a slight increase in the interfacial area and the volume of the RM aqueous phase, which altogether justifies the non-linearity observed as a function of w₀ levels. Accordingly, as per the packing aspects of SASSs and, the charge density levels, the point dipole moment p_r and the counter ion distribution (ξ) would vary. As the 〈a_N〉 value for a ω₀ level could differ based on the SASS packing attributes, the a_Nvs. ω₀ profile fails to provide a reliable estimate of the MPA(r⃑). We considered a normalized value of 〈a_N〉 as 〈a_N〉/〈τ_c〉, which accounts for the variability associated with the degree of surfactant-packing in the SASSs for different ω₀ levels.

Fig. 2 Assessment of MPA(r⃑) as a function of surfactant-packing: 〈a_N〉/〈τ〉 vs. ω₀(r⃑). The inset shows a_Nvs. ω₀(r⃑) and 〈τ〉 vs. ω₀(r⃑) relationships. OW and BW represent ordered water and bulk water, respectively. The critical distance r⃑_c for RMs (without HRP) demarcates (blue line) the OW region from the BW region.

The 〈a_N〉/〈τ_c〉 vs. r⃑(ω₀) profile was examined to study the MPA(r⃑). The slopes of this profile, (Δ(〈a_N〉/〈τ_c〉)/Δr⃑), provide a direct estimate of how the MPA(r⃑) varies as per the differences in the packing attributes of the SASSs. The slope value was higher at the ω₀ > 15 levels, than that at the ω₀ < 10 levels (Fig. 2, dotted lines). At the ω₀ < 10 levels, the lower slope value is due to the ordered form of water in the R2 region, which could be interpreted from the a_Nvs. r⃑(ω₀) profile as well Fig. 2 (inset). At ω₀ > 15 levels, R_RM ≫ l_B, the higher slope value is due to the plateau region of the ψ(r⃑) profile that represents the bulk water in the region |r⃑| ≤ 0 − (R − l_B). Instead of the plateau region, as seen in the 〈a_N〉 vs. r⃑(ω₀) profile, a higher slope value is seen for the 〈a_N〉/〈τ_c〉 vs. r⃑(ω₀) profile as τ_c decreases with increase in ω₀ – decrease in the SASS volume. A non-linearity in the 〈a_N〉/〈τ_c〉 vs. r⃑(ω₀) profile – the deviation shown in a circle in Fig. 2 – demonstrates that the aqueous phase constitutes two spatially segregated regions, having markedly different micro-polarity behaviour. This non-linearity in the 〈a_N〉/〈τ_c〉 vs. r⃑(ω₀) profile is pronounced at a critical distance, r⃑_c, and demarcates the two regions of a RM. Such a demarcation is the maxima (Fig. 2 (inset)) in the 〈a_N〉 vs. r⃑(ω₀) profile; the degree of surfactant packing in SASSs is not considered in making this demarcation. Thus, the 〈a_N〉/〈τ_c〉 vs. r⃑(ω₀) profile allows to reliably demarcate the two spatially separable R2 and R3 regions for SDS–hexane RMs, ω₀ ≈ 12. Though here, the slope values were similar across the ω₀ < 10 and ω₀ > 15 ranges, it could well be possible that the slope values could vary significantly due to marked differences in the SASS packing aspects.¹⁰

The 〈a_N〉/〈τ_c〉 vs. r⃑ profiles were developed for the RMs (ω₀ levels) having HRP enzymes at different concentration levels. In comparison to the 〈a_N〉/〈τ_c〉 values for the RMs alone, the 〈a_N〉/〈τ_c〉 values were significantly lower for RMs having HRP enzymes (Fig. 2). The lower values of 〈a_N〉/〈τ_c〉 are due to the HRP-concentration dependent micro-polarity attributes. The differences in the slope values were almost an order of magnitude less with HRP in the RMs (Fig. 2), which strongly suggested that the RM water exhibits an ordered form not only in the R2 region but in the R3 region as well. Wherein, the bulk-type water is in negligible amounts in RMs having HRP enzymes. In order to examine the microstructural aspects of the ordered form of water, we studied the O–H/O–D stretching vibration spectroscopic (ATR-IR) absorbance line shapes. In terms of the HRP structural dynamics, there were significant conformational changes of the HRP enzymes in the RMs (ESI,† 4). The amide I peak of HRP in RMs shifted to lower frequencies, and suggested the transformation of the α-helix structures of HRP to β-sheet types. Based on the negligible changes in the amide III peak positions and shapes, the possibilities of HRP denaturation in the RMs were eliminated. In terms of the microstructural features of water, for the RM of the ω₀ ≈ 20 level, the blue shifts of the OD- and OH-bands positions (Fig. 3a and b) were prominent for RMs with HRP. These shifts confirmed the ordered forms of water in the R2 and R3 regions. This ordered forms of water could be ascribed to the 〈a_N〉/〈τ_c〉 values. In the cases where other shapes of RM are possible and, depending whether the RMs coalesce, the slope values are likely to be different. It would be interesting to see how the 〈a_N〉/〈τ_c〉 vs. r⃑(ω₀) profile could be applicable for RM–enzyme systems to study the dependency of the MPA(r⃑) on the surfactant-packing in a SASS with surfactants of different chemistries.

Fig. 3 ATR-FTIR spectra of stretching vibrations of OD (a) and OH (b) stretching vibrations of D₂O and H₂O in RMs, and in the presence of HRP. The blue shift of the spectrum (green color spectral lines) indicates the ordered microstructure of the aqueous phase in the RM.

Acknowledgements

Support from the NSF I/UCR Center for Particulate and Surfactant Systems is acknowledged. The authors are thankful to Dr William Rice at the New York Structural Biology Center for the Cryo-SEM studies and Dr Jun Wu for Analytical Ultra centrifuge studies.

Notes and references

1. J. Eastoe, M. J. Hollamby and L. Hudson, Recent advances in nanoparticle synthesis with reversed micelles, Adv. Colloid Interface Sci., 2006, 128, 5–15,  DOI:10.1016/j.cis.2006.11.009.
2. M. Dekker, et al., Enzyme Recovery by Liquid-Liquid-Extraction Using Reversed Micelles, Chem. Eng. J. Biochem. Eng. J., 1986, 33, B27–B33,  DOI:10.1016/0300-9467(86)80050-8.
3. P. L. Luisi, Enzymes Hosted in Reverse Micelles in Hydrocarbon Solution, Angew. Chem., Int. Ed. Engl., 1985, 24, 439–450,  DOI:10.1002/anie.198504393.
4. A. Dasgupta, D. Das and P. K. Das, Reactivity of trypsin in reverse micelles: neglected role of aggregate size compared to water-pool components, Biochimie, 2005, 87, 1111–1119,  DOI:10.1016/j.biochi.2005.05.006.
5. E. P. Melo, M. R. Aires-Barros and J. M. Cabral, Reverse micelles and protein biotechnology, Biotechnol. Annu. Rev., 2001, 7, 87–129.
6. D. E. Moilanen, E. E. Fenn, D. Wong and M. D. Fayer, Water dynamics in large and small reverse micelles: From two ensembles to collective behavior, J. Chem. Phys., 2009, 131, 014704,  DOI:10.1063/1.3159779.
7. D. Bratko, A. Luzar and S. H. Chen, Electrostatic Model for Protein Reverse Micelle Complexation, J. Chem. Phys., 1988, 89, 545–550,  DOI:10.1063/1.455443.
8. F. Akoum and O. Parodi, Electrostatic Interactions inside the Aqueous Core of Spherical Reversed Micelles (L2-Phase), J. Phys. (Paris), 1985, 46, 1675–1681,  DOI:10.1051/jphys:0198500460100167500.
9. R. Biswas, J. Furtado and B. Bagchi, Layerwise decomposition of water dynamics in reverse micelles: A simulation study of two-dimensional infrared spectrum, J. Chem. Phys., 2013, 139, 144906,  DOI:10.1063/1.4824446.
10. R. Vacha and S. Roke, Sodium dodecyl sulfate at water-hydrophobic interfaces: a simulation study, J. Phys. Chem. B, 2012, 116, 11936–11942,  DOI:10.1021/jp304900z.
11. R. Vacha, et al., The Orientation and Charge of Water at the Hydrophobic Oil Droplet-Water Interface, J. Am. Chem. Soc., 2011, 133, 10204–10210,  DOI:10.1021/Ja202081x.
12. D. Das, S. Roy, R. N. Mitra, A. Dasgupta and P. K. Das, Head-group size or hydrophilicity of surfactants: the major regulator of lipase activity in cationic water-in-oil microemulsions, Chemistry, 2005, 11, 4881–4889,  DOI:10.1002/chem.200500244.
13. A. Dasgupta, D. Das, R. N. Mitra and P. K. Das, Surfactant tail length-dependent lipase activity profile in cationic water-in-oil microemulsions, J. Colloid Interface Sci., 2005, 289, 566–573,  DOI:10.1016/j.jcis.2005.03.083.
14. D. Kurad, G. Jeschke and D. Marsh, Lipid membrane polarity profiles by high-field EPR, Biophys. J., 2003, 85, 1025–1033,  DOI:10.1016/S0006-3495(03)74541-X.
15. A. L. Creagh, J. M. Prausnitz and H. W. Blanch, Structural and Catalytic Properties of Enzymes in Reverse Micelles, Enzyme Microb. Technol., 1993, 15, 383–392,  DOI:10.1016/0141-0229(93)90124-K.
16. H. Block and S. M. Walker, Modification of Onsager Theory for a Dielectric, Chem. Phys. Lett., 1973, 19, 363–364,  DOI:10.1016/0009-2614(73)80380-X.
17. H. Caldararu, A. Caragheorgheopol, M. Vasilescu, I. Dragutan and H. Lemmetyinen, Structure of the Polar Core in Reverse Micelles of Nonionic Poly(Oxyethylene) Surfactants, as Studied by Spin-Probe and Fluorescence Probe Techniques, J. Phys. Chem., 1994, 98, 5320–5331,  DOI:10.1021/J100071a024.
18. M. P. Pileni, Reverse Micelles as Microreactors, J. Phys. Chem., 1993, 97, 6961–6973,  DOI:10.1021/J100129a008.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5sm01950h

---