This article was published in the journal, Food Research International [© Elsevier]. The definitive version is available at: http://dx.doi.org/10.1016/j.foodres.2013.04.012Food grade sunflower oil was microencapsulated using cold water fish gelatine (FG)–gum arabic (GA) complex coacervation in combination with a batch stirred cell or continuous pulsed flow membrane emulsification system. Oil droplets with a controllable median size of 40–240 μm and a particle span as low as 0.46 were generated using a microengineered membrane with a pore size of 10 μm and a pore spacing of 200 μm at the shear stress of 1.3–24 Pa. A biopolymer shell around the oil droplets was formed under room temperature conditions at pH 2.7–4.5 and a total biopolymer concentration lower than 4% w/w using weight ratios of FG to GA from 40:60 to 80:20. The maximum coacervate yield was achieved at pH 3.5 and a weight ratio of FG to GA of 50:50. The liquid biopolymer coating around the droplets was crosslinked with glutaraldehyde (GTA) to form a solid shell. A minimum concentration of GTA of 1.4 M was necessary to promote the crosslinking reaction between FG and GTA and the optimal GTA concentration was 24 M. The developed method allows a continuous production of complex coacervate microcapsules of controlled size, under mild shear stress conditions, using considerably less energy when compared to alternative gelatine types and production methods

This article was published in the serial, Journal of Membrane Science [© Elsevier]. The definitive article is available at: http://www.sciencedirect.com/science/article/pii/S0376738808004742Membrane emulsification of unrefined pumpkin seed oil was performed using microengineered flat disc membranes on top of which a paddle blade stirrer was operated to induce surface shear. The membranes used were fabricated by galvanic deposition of nickel onto a photolithographic template and contained hexagonal arrays of uniform cylindrical pores with a diameter of 19 or 40 mu m and a pore spacing of 140 mu m. The uniformly sized pumpkin seed oil drops with span values less than 0.4 were obtained at oil fluxes up to 640 L m(-2) h(-1) using 2 wt.% Tween 20 (polyoxyethylene sorbitan monolaurate) or 2-10 wt.% Pluronic F-68 (polyoxyethylene-polyoxypropylen copolymer) as an aqueous surfactant solution. Pumpkin seed oil is rich in surface active ingredients that can be adsorbed on the membrane surface, such as free fatty acids, phospholipids, and chlorophyll. The adsorption of these components on the membrane surface gradually led to membrane wetting by the oil phase and the formation of uniform drops was achieved only for dispersed phase contents less than 10 vol.% At high oil fluxes, Pluronic F-68 molecules present at a concentration of 2 wt.% could not adsorb fast enough, on the newly formed oil drops, to stabilise the expanding inter-face. (c) 2008 Elsevier B.V. All rights reserved

This article was published in the serial, Journal of Membrane Science [© Elsevier B.V.]. The definitive version is available at: http://www.sciencedirect.com/science/article/pii/S0376738811009008Stirred cell membrane emulsification was used to create double water-in-oil-in-water emulsions at the dispersed phase flux of up to 3200 L m−2 h−1. The oil phase was unrefined pumpkin seed oil or sunflower oil containing 30% by volume of internal water phase and the external water phase was 2% Tween 20 (polyoxyethylene sorbitan monolaurate) or 2% Pluronic F-68 (polyoxyethylene-polyoxypropylen copolymer). Using microengineered nickel membranes with pore sizes between 15 and 40 μm, median drop sizes of double emulsion droplets were in the range between 100 and 430 μm, depending mainly on the shear at the membrane surface and dispersed phase injection rate. In most cases the drops were very uniform, with span (i.e. 90% drop size minus 10% drop size divided by median size) values of around 0.5. This data is similar to what was obtained previously for simple O/W emulsions of the same materials. Hence, the internal water phase, and internal surfactant, 5% PGPR (polyglycerol polyricinoleate), did not adversely influence the emulsification process. A marker material, copper sulfate, was added to the internal water phase and the release of copper was monitored with respect to time. For both lipid systems, at the larger droplet size, there was a significant period of no copper release, followed by almost linear release with time. This initial period was absent when the drop size was close to 100 μm. The initial entrapment efficiency of the copper, in all experiments, was higher than 94%

Tubular micrometer pore sized sieve type membranes with internal diameter of 14 mm and length of 60 mm containing uniform pores of diameter 10 and 20 μm were used to generate emulsions of sunflower oil dispersed in water and stabilized by Tween 20 using oscillatory flow of the continuous phase. Drop diameters between 30 and 300 μm could be produced, in a controllable way and with span values of down to 0.4. By using pulsed flow it was possible to provide dispersed phase concentrations of up to 45% v/v in a single pass over the membrane, that is, without the need to recirculate the continuous phase through the membrane tube. It was possible to correlate the drop size produced with the shear conditions at the membrane surface using the wave shear stress equation. The oscillatory Reynolds number indicated flow varying from laminar to substantially turbulent, but the change in flow conditions did not show a notable influence on the drop diameters produced, over what is predicted by the varying wall shear stress applied to the wave equation. However, the 20 μm pore sized sieve type membrane appeared to allow the passage of the pressure pulse through the membrane pores, under certain operating conditions, which did lead to finer drop sizes produced than would be predicted. These through-membrane pulsations could be suppressed by changes in operating conditions: a higher dispersed phase injection rate or more viscous continuous phase, and they did not occur under similar operating conditions used with the 10 μm pore sized sieve type of membrane. Generating emulsions of this size using pulsed continuous phase flow provides opportunities for combining drop generation at high dispersed phase concentration, by membrane emulsification, with downstream processing such as reaction in plug flow reactors.

This paper was accepted for publication in the journal Chemical Engineering Journal and the definitive published version is available at http://dx.doi.org/10.1016/j.cej.2015.12.024© 2015 Elsevier B.V. The Dispersion Cell membrane emulsification technique was used for the production of w/o emulsions with controlled droplet size and narrow size distribution. The influence of the operating parameters of the process was investigated. Varying the dispersed phase flux (10-1250 L h-1 m-2) and the shear stress (2-59 Pa), droplets between 30 and 280 μm were produced with CV's as low as 18%. Nickel and stainless steel membranes were used for the membrane emulsification. Pore geometry influenced the droplet size as well as uniformity and a normally hydrophilic stainless steel membrane with sharp pore openings produced more uniform and smaller drops compared to a PTFE coated hydrophobic nickel membrane with a conical pore surface. For the dispersed phase 15 wt.% PVA or 1-3 wt.% chitosan as well as their blends in water were used. Surfactants PGPR and ABIL EM90 were tested to determine their capability to form stable emulsions in Miglyol 840. PGPR could not be used to stabilize the emulsion with chitosan as the dispersed phase, probably due to the chemical interference between the carboxyl group present in the PGPR and chitosan. Solid microparticles were obtained by chemical crosslinking with glutaraldehyde (GA) at different concentrations (1-50 vol.%). Particles crosslinked using less than 10 vol.% GA were able to swell and release encapsulated compounds. Acid sensitive particles were produced by blending the PVA and chitosan. Up to 80% of Cu2+ and 20% of sodium salicylate was released from the particles under acidic conditions. No significant release was determined under neutral conditions