Efficient Emulsification of Viscous Oils at High Drop Volume Fraction

Abstract

It is shown experimentally in this study that the increase of drop volume fraction can be used as an efficient tool for emulsification of viscous oils in turbulent flow. In a systematic series of experiments, the effects of drop volume fraction and viscosity of the dispersed phase on the mean, d(32), and maximum, d(V95), diameters of the drops, formed during emulsification, are quantified. The volume fraction, Φ, of the dispersed oily phase is varied between 1% and 90%, and oils with viscosity varying between 3 and 10,000 mPa.s are studied. All experiments are performed at sufficiently high surfactant concentration, as to avoid possible drop-drop coalescence during emulsification. The analysis of the experimental data shows that there is a threshold drop volume fraction, Φ(TR), at which a transition from inertial turbulent regime into viscous turbulent regime of emulsification occurs, due to the increased overall viscosity of the emulsion. At Φ < Φ(TR), d(32) and d(V95) depend weakly on Φ and are well described by known theoretical expression for emulsification in inertial turbulent regime (Davies, Chem. Eng. Sci. 1985, 40, 839), which accounts for the effects of oil viscosity and interfacial tension. At Φ > Φ(TR), both d(32) and polydispersity of the formed emulsions decrease very significantly with the increase of Φ (for the oils with η(D) > 10 mPa.s). Thus, very efficient emulsification of the viscous oils is realized. Very surprisingly, a third regime of emulsification is observed in the range of concentrated emulsions with Φ > 75%, where the mean drop size and emulsion polydispersity are found experimentally to be very similar for all oils and surfactants studied-an experimental fact that does not comply with any of the existing models of drop breakup during emulsification. Possible mechanistic explanations of this result are discussed. The experimental data for semiconcentrated and concentrated emulsions with Φ > Φ(TR) are described by a simple scaling expression, which accounts for the effects of all main factors studied.