Résumé
Even in the absence of surfactant, polymers or particles, spontaneous emulsions produced by dilution only, without energy input, can be very stable. Industrial applications of this “Ouzo effect” [1] range from food to pharmaceutics. Spontaneous emulsification is observed when diluting from a domain where dynamic nanometer-size aggregates are present. Why Ostwald ripening is not effective as destabilizing mechanism producing creaming remained unknown.Using in-situ auto-dilution, we followed by small-angle X-ray scattering (SAXS) over 3+ decades in reciprocal space the morphological changes occurring when going through the ternary phase diagram water-ethanol-octanol, from the alcohol binary to the water corner. This allows for the first time a full characterization of the multiple-scale coexisting microstructures. Synchrotron-SAXS data were complemented by Small Angle Neutron Scattering (SANS) profiles at particular compositions to benefit from isotopic contrast variation.[2]The aqueous phase is an ultra-flexible microemulsions[3,4] with nanometre-size polydisperse clusters confirmed by MD[5]. In the metastable regime, when the Ouzo effect leads to large droplets, coexisting phases are both ternary solutions structured at nanometre scale. The oil-rich micrometer-droplets contain up to 30 % water by mass and are always structured at nanoscale when the emulsion is stable.The phase transition is asymmetric around the plait point. When the initial concentration of hydrotrope is below the minimum hydrotrope concentration (MHC), the emulsification fails: emulsions cream immediately because the oil-rich domain does not exhibit direct clusters. Beyond MHC, the low surface tension between coexisting ternary fluids that both scatter with an Ornstein-Zernicke contribution results in Laplace pressure driving towards macroscopic phase separation below 100 Pa, explaining the puzzling resilience of spontaneous emulsion versus the universal mechanism of Ostwald ripening.References[1] S. Vitale, J. Katz, Langmuir, 19 (10) 2003, 4105-4110. DOI: 10.1021/la026842o[2] O. Diat et al., J Applied Cryst, 46 2013, 1665-1669. DOI: 10.1107/S002188981302606X[3] Th. Zemb et al., PNAS, 113 (16) 2016, 4260-4265. DOI: 10.1073/pnas.1515708113[4] S. Prévost et al., J Applied Cryst, 49 2016, 2063-2072. DOI: 10.1107/S1600576716016150[5] S. Schöttl et al., JCIS, 540 2019, 623-633. DOI: 10.1016/j.jcis.2019.01.037