Stability of water-in-oil-in-water emulsions formed by membrane emulsification : a thesis presented in partial fulfilment of the requirements for the degree of Master of Technology in Food Technology at Massey University, Palmerston North, New Zealand

Abstract

The main objectives of this study were to determine i. The effectiveness of encapsulating whey protein concentrates (WPC) within water-in-oil-in-water multiple emulsions produced by membrane emulsification. ii. The effect of the primary and secondary emulsification conditions and membrane operating parameters on the multiple emulsion properties; of particular concern were the yield and physical stability of the emulsions. The multiple emulsions were prepared by a two-stage emulsification process. The emulsification conditions were varied widely to determine the optimum conditions for the production of multiple emulsions. Ultra-Turrax, ultrasound and valve homogenisation were tried for the preparation of the primary emulsion; an Ultra-Turrax and Shirazu porous glass (SPG) membrane emulsification were used for secondary emulsification. The standard primary emulsions (water-in-oil) were prepared with 10% of the WPC and 6.4% glucose (as a marker) in the water phase with a water-in-oil volume fraction of 0.25. The oil phase (soybean oil) consisted of 10% hydrophobic emulsifier (Polyglycerol polyricinoleate [PGPR] or Span 80). Typically pre-emulsification was carried out in an Ultra-Turrax at 20500 rpm and the primary emulsification was done in a homogeniser at a pressure of 500/100 bars in the two stages; secondary emulsions were prepared with an SPG membrane of pore size 2μm or 3.8µm with the dispersed phase (water-in-oil) being pushed through the pores of the membrane at a specific transmembrane pressure (125-150 kPa) and dispersed phase flux into a continuous phase (water with a hydrophilic emulsifier concentration, usually Tween 80, of 1%) flowing through the inside of the membrane at a particular velocity (standard of 1 m/s). The yield of the multiple emulsions formed was estimated by measuring glucose release using an Advantage Glucose meter. Unlike Span 80, PGPR was able to form stable o/w emulsions and hence the initial yield of the multiple emulsions varied from 80% at 2.5% to 100% at 10% PGPR. The higher the concentration of water in the inner phase the lower the yield of the multiple emulsions and the higher the droplet size of the primary emulsions. A valve homogeniser gave the best results for primary emulsification. Of the 3 homogenisation pressures (250 bar, 500 bar, 1000 bar) tried, the w/o emulsion produced with 500 bar and 10% PGPR was taken as the standard as this was found to be stable for 6 months without physical damage. A 30% maximum loading of the WPC in the inner water phase was also determined. A further increase may destabilise the process by causing blockage to the membrane pores. The yield as well as the droplet size of the multiple emulsions was found to increase as the membrane pore-size was increased from 1.4 µm to 3.8 µm. Transmembrane pressure and continuous phase velocity did not have much influence on the yield of the multiple emulsions. However an increase in continuous phase velocity increased the opacity of the serum layer formed indicating that an increased amount of smaller droplets were formed. The dispersed phase flux was increased by increases in any of the transmembrane pressure, PGPR concentration and membrane pore-size. Hydrophilic emulsifiers (whey protein isolate, soy protein isolate and sodium caseinate) did not influence the yield; however the Tween 80 stabilised multiple emulsions showed a smaller droplet size. An increase in temperature from 20 - 50°C resulted in a lower yield as well as a higher droplet size. The osmotic gradient set up by glucose and WPC in the inner phase of the emulsion resulting in an influx of water from the outer phase causing bulging of the droplets. Sorbitol added at 1.7% in the outer phase gave a high initial yield (100%) as well as a low droplet size (2-3 µm). The cream layer formed as a result of storage was found to decrease with increase in sorbitol concentration (to 5.9%) due to the lower size ot the droplets formed. The key issues identified were to find an alternative to PGPR with lower Accepted Daily Intake (ADI) value without compromising the emulsification properties and to standardise ways to analyse droplet size of w/o emulsions. Overall the study proved that functional ingredients can be encapsulated using stable w/o/w multiple emulsions prepared using SPG membranes under standardised conditions and hence appears to offer promise for manufacture of commercial products.