Microfiltration (MF) of milk protein concentrate using ceramic membranes: Determination of limiting flux and serum protein (SP) removal at 8, 9 or 10% protein in the recirculation loop

Abstract Text:

In designing a MF process to separate SP from casein, both the amount of SP removed and flux are likely to be a function of the protein concentration in the recirculation loop (RL). Our objective was to determine the limiting flux and SP removal at 8, 9 or 10% protein in the RL using 0.1µm ceramic graded permeability membranes with 4mm channel diameters.  The MF feed was an 85% milk protein concentrate diluted to an average protein of 5.5±0.1%. The concentration factor was chosen to achieve 8, 9 or 10% protein in the RL. The MF was operated with a longitudinal pressure drop of 220kPa at 50^oC. The MF was started at an average flux of 55±2kg/m² per h and flushed until the protein concentration in the RL was near the target. Once the target protein was reached, the MF was run at 55kg/m² per h for 1h, the flux was then increased in steps and run for 1h at each new flux. The flux was increased until the new flux could not be maintained. The limiting flux was the last flux the MF could run at for 1h.Transmembrane pressure averaged 66kPa at 55kg/m² per h and 190kPa at the limiting flux and  transmembrane pressure did not vary with protein concentration in the RL (P>0.05). Retentates and permeates at each flux were analyzed for true protein (TP) using Kjeldahl methods. SP removal as a percentage of theoretical removal (SPR) was calculated as TP in the permeate divided by SP in the permeate portion of the feed. TP concentrations in the RL averaged: 8.2±0.1, 9.2±0.1 and 10.1±0.2%. Cross-flow velocities depended on the protein concentration in the RL (P<0.05) and were: 7.09, 7.01 and 6.90m/s at 8, 9 and 10% protein respectively. Limiting fluxes decreased with increasing protein in the RL (P<0.05) and were: 154±1, 133±1 and 117±6kg/m² per h. SPR was not a function of the protein concentration in the RL (P>0.05), but SPR decreased (P<0.05) from 80% at 55kg/m² per h to 75% at the limiting flux, indicating fouling may have impacted passage of SP through the membrane. The protein concentration in the RL that the MF operated at had an impact on the limiting flux, but not passage of SP. However, as flux increased to the limiting flux there was reduced SP passage through the membrane, likely caused by membrane fouling.

Keywords: limiting flux, microfiltration, milk protein concentrate