Reversible light-induced critical separation

Abstract

A reversible, critical phase separation induced by UV light at 350 nm has been achieved by doping a small amount of photosensitive surfactant into an AOT-stabilised water-in-decane microemulsion. The initial and separated phases were analysed by small-angle neutron scattering (SANS), which suggests irradiation produces two coexisting water-in-oil microemulsions with relative droplet concentrations of 1 : 3. To better understand the effect of the photosurfactant, drop shape analysis tensiometry was used, showing a change in the effectiveness of the interfacial film under different irradiation conditions.

---

The quest for clean, low-energy separation and recovery techniques for colloidal and nanoparticle dispersions, as well as catalytic systems and speciality organic syntheses, has recently attracted attention.^1,2 For many applications, traditional separation methods such as temperature-induced instabilities, centrifugation, addition of electrolytes or alcohols are often undesirable due to the high energy consumption and significant changes in system composition, resulting in complex product mixtures.

The so-called ‘critical’ phase separation,³ whereby a monophasic system may be separated into two coexisting phases, differing in density and particle concentration, represents an attractive approach for concentrating colloids from continuous phases. For water-in-decane (w/d) microemulsions stabilised by the common anionic surfactant sodium bis(2-ethyl-hexyl)sulfosuccinate (AOT), temperature can be used as an accessible control parameter to induce such a ‘liquid–liquid’ critical-type phase separation.^3–5

Here it is shown that critical behaviour can also be achieved using UV light irradiation of an azobenzene-containing photosurfactant, sodium 4-butylphenyldiazenylphenoxybutane-1-sulfonate (AZOT-S4, Fig. 1) which is doped into a standard AOT-stabilised w/d microemulsion. The azobenzene functional group is a well known photoisomerisable chromophore which can be reversibly switched between trans and cis isomers with appropriate UV-visible radiation.

Fig. 1 Structure of the photosurfactant AZOT-S4.

The isomerisation is reversible, free of side-reactions and considered to be one of the cleanest photo-reactions,⁶ making azobenzenes attractive for photoswitchable applications. The trans isomer is around 50 kJ mol⁻¹ more stable than the cis form and the activation barrier to conversion is approximately 200 kJ mol⁻¹,⁷ although these values are influenced by the nature of substituents attached to the azobenzene moiety and solvent environment. It is believed that the photoisomerisation leads to a decrease in the distance between the 4 and 4′ ring positions from 0.9 nm to 0.55 nm on converting the trans to the cis isomer, affecting dipole moment, hydrophobicity and light absorption, allowing changes in system properties such as viscosity,⁸ surface tension⁹ and aggregation behaviour.^10,11 In addition, the extended π-conjugation and variety of azobenzene-containing molecules make them useful in the production of well-ordered, nanoscale materials.¹² To this end, azobenzenes have been used in mixtures and assemblies with surfactants, polymers and peptides in order to create functional supramolecular structures suggesting applications such as molecular motors,¹³ polarising filters,¹² optical storage devices¹⁴ and light-driven actuators,¹⁵ amongst others.

This work is a significant progression from previous studies into photo-separable colloidal systems^16–19 with the added advantage of reversibility. In the photosensitive microemulsions described here, UV light causes a critical phase separation (photographic inset to Fig. 2). Using λ = 350 nm incident light, the predominantly trans AZOT-S4 incorporated into the w/d microemulsion can be converted to the cis form. This photoisomerisation drives a reduction in the water–decane interfacial tension, γ_w/d, resulting in a separation of the single phase microemulsion into two coexisting microemulsion phases (i.e. a liquid–liquid critical separation). Recovery of the single phase microemulsion can be achieved by irradiation at λ = 450 nm (or ambient/broad-spectrum illumination or a temperature increase) to regenerate trans AZOT-S4.

Fig. 2 Photograph and small-angle neutron scattering (SANS) profiles for the single phase pre-separation and photo-irradiated phase-separated microemulsions. [AOT] = 88 mmol dm⁻³, [AZOT-S4] = 1.5 mmol dm⁻³ (molecular ratio AOT : AZOT-S4 = 60 : 1), w = 67 (w.r.t. total surfactant concentration), T = 25 °C. Solid lines are model fits, generated as described in the ESI,† allowing determination of average droplet radius, R^av, polydispersity index, σ/R^av, and droplet volume fraction, ϕ_d.

For a given surfactant concentration, the phase behaviour is sensitive to temperature and the composition parameter, w = [water]/[surfactant].²⁰ The single phase microemulsions were formulated to be located close to the higher temperature phase boundary (often called the haze point or cloud point, T_cloud). For the AOT-only system (no added AZOT-S4), a T_cloud separation occurs on heating above the phase boundary; note these are not the conditions needed for the critical instability, which occurs for a pure AOT–decane–water microemulsion at w = 40.8 and T = 40 °C.⁴ The samples studied here are w = 67, [AOT] = 88 mmol dm⁻³, [AZOT-S4] = 1.5 mmol dm⁻³ at T = 25 °C, and, in the absence of incident UV irradiation, the trace level of added AZOT-S4 had no noticeable effect on the location and nature of the T_cloud boundary. Furthermore, heating mixed AOT/AZOT-S4 single phase microemulsions resulted in ‘classic’ cloud point phase instabilities, rather than critical-type liquid–liquid separations.

However, after UV irradiation critical-type phase separations were observed, and characterised by visual transparency of the equilibrium upper and lower phases and their relative volumes (upper : lower ≈ 2 : 1). Small-angle neutron scattering (SANS) shows that the pre-separation microemulsion experiences weak attractive inter-droplet interactions [κ (coefficient related to isothermal compressibility) = 1.32, ζ (correlation length) = 11 nm—see ESI†], whereas attractive interactions are essentially absent in the two separated coexisting phases. The droplet size is very similar in pre- and post-separation phases (radius ≈ 6 nm), consistent with previous work on similar, photosurfactant-free AOT w/d systems.⁵ The fitted radius is smaller than that found for a pure AOT-stabilised microemulsion (expected to be around 8.5 nm²¹), consistent with a minor effect of AZOT-S4 on the preferred interfacial curvature. The ratio of droplet concentrations in the equilibrium phases, determined from the absolute scattering intensity is around 1 : 3 (top phase : bottom phase).

In order to better understand the function of AZOT-S4 photosurfactant in this light-induced critical transition, water–decane interfacial tensions, γ_w/d were determined by drop shape analysis (DSA) tensiometry (Fig. 3). The systems studied contained (1) AZOT-S4 only and (2) AOT and AZOT-S4 at a ratio of 60 : 1. The interfacial tension of a pure water–decane interface is around 50 mN m⁻¹,²² so it can be seen that AZOT-S4 alone is a relatively weak surfactant giving rise to γ_w/d ≈ 16 mN m⁻¹. Starting in primarily the trans configuration, illumination at λ = 350 nm (dotted line) generating the cis form induces an increase in γ_w/d; azobenzenes are known to generally be less surface-active in the cis form.⁹ Irradiation at λ = 450 nm (dashed line) reverts the surfactant to the trans form and the equilibrium interfacial tension is eventually recovered.

Fig. 3 Drop shape analysis tensiometry of pure AZOT-S4 and AZOT-S4 + AOT at the water–decane interface. Vertical dotted/dashed lines indicate irradiation.

Interestingly, mixtures of AOT and AZOT-S4 exhibit the opposite behaviour: UV irradiation at λ = 350 nm gives a marked decrease in the γ_w/d, which then relaxes back under illumination at λ = 450 nm. A possible explanation is that the cis isomer has a higher dipole moment across the azobenzene group, being therefore more water-soluble than the trans,^9,23 and hence is displaced from the interface by the more surface-active AOT, leading to an overall decrease in interfacial tension. When the cis-AZOT-S4 relaxes back to the trans isomer, it may re-partition into a mixed interface with AOT once more.

These light-induced changes in γ_w/d can account for the optically triggered liquid–liquid phase behaviour of the w/d systems (Fig. 2). Addition of trans-AZOT-S4, which is a poorer oil–water emulsifier than AOT, reduces the overall surfacant efficiency, thereby nudging the system away from the critical transition point. Irradiation to generate the cis photosurfactant increases the aqueous phase solubility; cis-AZOT-S4 may then partition more strongly into the aqueous droplets (compared to the stabilising film) thereby shifting the interfacial surfactant composition in favour of AOT. If the value of γ_w/d decreases, stronger droplet–droplet interactions are expected that would drive the liquid–liquid phase separation.⁴ Irradiation at λ = 450 nm regenerates the trans-AZOT-S4, restabilising the homogeneous single phase w/d microemulsion. For these photosurfactant systems it would be very difficult to experimentally determine changes in interfacial film compositions; however, it is unlikely that UV irradiation generates a pure AOT-only interface, since the microemulsion phase behaviour of the AZOT-S4-containing systems after irradiation is distinctly different from that seen with AOT alone (as mentioned above).

This explanation also accounts for the critical behaviour observed: the AZOT-S4 actually has a inhibiting effect on the surfactancy in the system. The critical phase separation occurs because of increased interactions between microemulsion droplets, causing transient aggregates and large local differences in droplet concentration.⁴ These interactions are different depending on how the interface is populated by surfactant, and hence when trans AZOT-S4 is also present at the interface, it mitigates inter-droplet forces. Upon irradiation, the more water-soluble cis form would partition from the interface into the watery interior of the droplet, and hence AOT interactions are stimulated, resulting in the phase separation. Once relaxed to the trans form, the mixed interface can re-form, and the system is again stable as a single microemulsion phase.

To summarise, a reversible, critical phase separation induced by UV light at 350 nm has been achieved by doping a small amount of photosensitive surfactant into an AOT-stabilised water-in-decane microemulsion. This ability to concentrate colloids from their solvent has direct applications in recovery of high value synthetic products, catalytic nanoparticles and other speciality colloids. The significance of using light as the trigger is that the recovery can be low-energy, low-cost, clean and most importantly, reversible. The next steps are to achieve higher phase concentration ratios, by design of more efficient photosurfactants and formulation of optimised microemulsions.

Acknowledgements

RT would like to thank Infineum UK Ltd, the University of Bristol for funding and the STFC for allocation of beamtime at the ILL. RJO thanks the EPSRC and the University of Bristol for financial support.

References

1. H. Cheng and D. A. Sabatini, Sep. Sci. Technol., 2007, 42, 453.
2. S. Vesaratchanon, A. Nikolov and D. T. Wasan, Adv. Colloid Interface Sci., 2007, 134–135, 268.
3. J. S. Huang and M. W. Kim, Phys. Rev. Lett., 1981, 47, 1462.
4. M. Kotlarchyk, S.-H. Chen and J. S. Huang, Phys. Rev. A, 1983, 28, 508.
5. M. Kotlarchyk, S.-H. Chen, J. S. Huang and M. W. Kim, Phys. Rev. A, 1984, 29, 2054.
6. H. Rau, in Photochemistry and Photophysics, ed. J. F. Rabek, CRC Press, Boca Raton, 1990, vol. 2, pp. 119.
7. R. H. E. Halabieh, O. Mermut and C. J. Barrett, Pure Appl. Chem., 2004, 76, 1445.
8. C. T. Lee Jr, K. A. Smith and T. A. Hatton, Macromolecules, 2004, 37, 5397.
9. T. Shang, K. A. Smith and T. A. Hatton, Langmuir, 2003, 19, 10764.
10. H. Sakai, A. Matsumura, S. Yokoyama, T. Saji and M. Abe, J. Phys. Chem. B, 1999, 103, 10737.
11. T. Shang, K. A. Smith and T. A. Hatton, Langmuir, 2006, 22, 1436.
12. Y. Zakrevskyy, J. Stumpe and C. F. J. Faul, Adv. Mater., 2006, 18, 2133.
13. W. R. Browne and B. L. Feringa, Nat. Nanotechnol., 2006, 1, 25.
14. P. H. Rasmussen, P. S. Ramanujam, S. Hvilsted and R. H. Berg, J. Am. Chem. Soc., 1999, 121, 4738.
15. T. Ikeda, J. Mamiya and Y. Yu, Angew. Chem., Int. Ed., 2007, 46, 506.
16. A. Salabat, J. Eastoe, A. Vesperinas, R. F. Tabor and K. J. Mutch, Langmuir, 2008, 24, 1829.
17. A. Vesperinas, J. Eastoe, P. Wyatt, I. Grillo, R. Heenan, J. Richards and G. Bell, J. Am. Chem. Soc., 2006, 128, 1468.
18. J. Eastoe, P. Wyatt, M. Sanchez-Dominguez, A. Vesperinas, A. Paul, R. Heenan and I. Grillo, Chem. Commun., 2005, 2785.
19. A. Vesperinas, J. Eastoe, S. Jackson and P. Wyatt, Chem. Commun., 2007, 3912.
20. P. D. I. Fletcher, A. M. Howe and B. H. Robinson, J. Chem. Soc., Faraday Trans. 1, 1987, 83, 985.
21. S. Nave, J. Eastoe, R. Heenan, D. Steytler and I. Grillo, Langmuir, 2000, 16, 8741.
22. S. Susnar, H. Hamza and A. Neumann, Colloids Surf., A, 1994, 89, 169.
23. L. Yang, N. Takisawa, T. Hayashita and K. Shirahama, J. Phys. Chem., 1995, 99, 8799.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental details; the synthetic route to AZOT-S4 (including NMR and mass spectral data for AZOT-S4); experimental details for SANS; and neutron modelling parameters and uncertainties. See DOI: 10.1039/b813234h

---