Separation of milk fat globules via microfiltration: Effect of diafiltration media and opportunities for stream valorization

Abstract

Milk fat globule membranes (MFGM) sourced in buttermilk have gained recent interest given their nutritional value and functional properties. However, production of isolated MFGM has been challenging given their size similarity with casein micelles, which limits attempts toward fractionation by size exclusion techniques. Therefore, the hypothesis underpinning this study is that the removal of proteins from cream before butter-making facilitates MFGM isolation. As such, milk fat globules were separated from raw whole milk via microfiltration (1.4-µm pore diameter and 0.005-m filtration surface area) by using 3 diafiltration media; namely, skim milk ultrafiltration permeate, saline, and water. Their effects on the stability of the milk fat globules and protein permeation was elucidated. Whereas a substantial reduction in protein concentration was achieved with all diafiltration media (~90% reduction), water and saline produced negligible membrane fouling with better filtration performance. Moreover, diafiltration with skim milk ultrafiltration permeate exhibited reduced permeate flux. Colloidal stability of the resultant milk decreased with all diafiltration solutions due to changing composition and reduced apparent viscosity. Overall, microfiltration was found to be an efficient method for separation of milk fat globules from whole milk, leading to increased MFGM fragment concentration in buttermilk dry matter, thus making it more suitable for industrial utilization.

Changes in milk fat globules and membrane lipids under the shear fields of microfiltration and centrifugation

“…The results for zeta potential coincided with those of particle size, indicating that the given processing modified the surface of the milk fat globules and possibly released MFGM components due to shear stress and change in the equilibria. A more negative zeta potential of milk fat globule upon microfiltration was reported previously by Jukkola et al, [23] and was explained by a decrease in mineral composition. Furthermore, Bourlieu et al [42] reported an accumulation of free fatty acids on the MFGM that made the zeta potential more negative (and increased the electrostatic repulsion).…”

“…The shear-induced damage was estimated based on particle size and zeta potential measurements. The particle size was expected to increase due to permeation of the smallest fat globules and casein micelles [23]. However, microfiltration decreased the mean fat globule size slightly after the first cycle (Table 1) after which the particle size remained constant.…”

“…A multichannel tubular ceramic membrane was adopted (1.4 μm pore size and 0.24 m 2 filtration surface area, Membralox®, Pall Corporation, France). The filtration system was cleaned before and after each experiment according to [23].…”

“…Particle size distribution was measured for creams before and after the processes with a Mastersizer 2000 (Malvern Instruments, Malvern, UK) based on light scattering according to [27] and [23]. A refractive index of 1.458 was adopted for the dispersed phase and the samples were analysed in triplicate.…”

“…The zeta potential of the milk fat globules was measured on the following day after processing using an electrophoretic mobility instrument (Zetasizer, Nano series ZS90, Malvern Instruments, Malvern, UK) according to Michalski et al [23]. The samples were diluted 500-fold in 10mM saline (pH 6.8) and measured in a Malvern Dip Cell at 20 °C.…”