Filtration of milk fat globule membrane fragments from acid buttermilk cheese whey

The proteins and polar lipids present in milk fat globule membrane (MFGM) fragments are gaining attention for their technological and nutritional properties. These MFGM fragments are preferentially enriched in side streams of the dairy industry, like butter serum, buttermilk, and whey. The objective of this study was to recover MFGM fragments from whey by tangential filtration techniques. Acid buttermilk cheese whey was chosen as a source for purification by tangential membrane filtration because it is relatively rich in MFGM-fragments and because casein micelles are absent. Polyethersulfone and cellulose acetate membranes of different pore sizes were evaluated on polar lipid and MFGM-protein retention upon filtration at 40°C. All fractions were analyzed for dry matter, ash, lipids, proteins, reducing sugars, polar lipid content by HPLC, and for the presence of MFGM proteins by sodium dodecyl sulfate-PAGE. A fouling coefficient was calculated. It was found that a thermocalcic aggregation whey pretreatment was very effective in the clarification of the whey, but resulted in low permeate fluxes and high retention of ash and whey proteins. By means of an experimental design, the influence of pH and temperature on the fouling and the retention of polar lipids (and thus MFGM fragments), proteins, and total lipids upon microfiltration with 0.15 μM cellulose acetate membrane was investigated. All models were highly significant, and no outliers were observed. By increasing the pH from 4.6 to 7.5, polar lipid retention at 50°C increased from 64 to 98%, whereas fouling of the filtration membrane was minimized. A 3-step diafiltration of acid whey under these conditions resulted in a polar lipid concentration of 6.79 g/100 g of dry matter. As such, this study shows that tangential filtration techniques are suited for the purification of MFGM fragments.     

Gelation of casein micelles in ��-casein reduced milk prepared using membrane filtration

Pasteurized skim milk was subjected to membrane filtration using a molecular weight cut-off of 80 kDa and a plate and frame pilot scale system at temperatures below 10 °C. Via this process, transmission of whey proteins and ??-casein through the membrane was achieved. The milk was concentrated to two times (based on volume reduction), and whey protein-free permeate was added to return to the original volume fraction of casein micelles in milk. This diafiltration process was carried out four times, and the retentate obtained was nearly free of whey proteins and with approximately 20% of ??-casein removed. The same membrane filtration was also carried out at 25 °C to achieve transmission of whey protein but not of ??-casein, and to obtain whey protein-depleted milk without depletion of ??-casein.The gelling behaviour of these samples, reconstituted to the original casein volume fraction, was examined using rheology and diffusing wave spectroscopy. When compared to the original skim milk it was found that there were no statistically significant differences in gelation behaviour during acidification, but differences were noted in gelation time and final stiffness modulus for samples undergoing renneting. These differences were attributed mostly to the changes in ionic composition, as when the serum composition of the retentates was re-equilibrated against the original skim milk by dialysis; the gelation behaviour of the samples was comparable to that of skim milk. The results clearly indicate the importance of the milk's overall ionic balance in the early stages of aggregation of rennet-induced gelation of milk.     

Effect of microfiltration concentration factor on serum protein removal from skim milk using spiral-wound polymeric membranes

Our objective was to determine the effect of concentration factor (CF) on the removal of serum protein (SP) from skim milk during microfiltration (MF) at 50°C using a 0.3-μm-pore-size spiral-wound (SW) polymeric polyvinylidene fluoride (PVDF) membrane. Pasteurized (72°C for 16 s) skim milk was MF (50°C) at 3 CF (1.50, 2.25, and 3.00×), each on a separate day of processing starting with skim milk. Two phases of MF were used at each CF, with an initial startup-stabilization phase (40 min in full recycle mode) to achieve the desired CF, followed by a steady-state phase (90-min feed-and-bleed with recycle) where data was collected. The experiment was replicated 3 times, and SP removal from skim milk was quantified at each CF. System pressures, flow rates, CF, and fluxes were monitored during the 90-min run. Permeate flux increased (12.8, 15.3, and 19.0 kg/m² per hour) with decreasing CF from 3.00 to 1.50×, whereas fouled water flux did not differ among CF, indicating that the effect of membrane fouling on hydraulic resistance of the membrane was similar at all CF. However, the CF used when microfiltering skim milk (50°C) with a 0.3-μm polymeric SW PVDF membrane did affect the percentage of SP removed. As CF increased from 1.50 to 3.00×, the percentage of SP removed from skim milk increased from 10.56 to 35.57%, in a single stage bleed-and-feed MF system. Percentage SP removal from skim milk was lower than the theoretical value. Rejection of SP during MF of skim milk with SW PVDF membranes was caused by fouling of the membrane, not by the membrane itself and differences in the foulant characteristic among CF influenced SP rejection more than it influenced hydraulic resistance. We hypothesize that differences in the conditions near the surface of the membrane and within the pores during the first few minutes of processing, when casein micelles pass through the membrane, influenced the rejection of SP because more pore size narrowing and plugging occurred at low CF than at high CF due to a slower rate of gel layer formation at low CF. It is possible that percentage removal of SP from skim milk at 50°C could be improved by optimization of the membrane pore size, feed solution composition and concentration, and controlling the rate of formation of the concentration polarization-derived gel layer at the surface of the membrane during the first few minutes of processing.     

A physicochemical investigation of membrane fouling in cold microfiltration of skim milk

The main challenge in microfiltration (MF) is membrane fouling, which leads to a significant decline in permeate flux and a change in membrane selectivity over time. This work aims to elucidate the mechanisms of membrane fouling in cold MF of skim milk by identifying and quantifying the proteins and minerals involved in external and internal membrane fouling. Microfiltration was conducted using a 1.4-μm ceramic membrane, at a temperature of 6 ± 1°C, cross-flow velocity of 6 m/s, and transmembrane pressure of 159 kPa, for 90 min. Internal and external foulants were extracted from a ceramic membrane both after a brief contact between the membrane and skim milk, to evaluate instantaneous adsorption of foulants, and after MF. Four foulant streams were collected: weakly attached external foulants, weakly attached internal foulants, strongly attached external foulants, and strongly attached internal foulants. Liquid chromatography coupled with tandem mass spectrometry analysis showed that all major milk proteins were present in all foulant streams. Proteins did appear to be the major cause of membrane fouling. Proteomics analysis of the foulants indicated elevated levels of serum proteins as compared with milk in the foulant fractions collected from the adsorption study. Caseins were preferentially introduced into the fouling layer during MF, when transmembrane pressure was applied, as confirmed both by proteomics and mineral analyses. The knowledge generated in this study advances the understanding of fouling mechanisms in cold MF of skim milk and can be used to identify solutions for minimizing membrane fouling and increasing the efficiency of milk MF.     

Compositional and functional properties of buttermilk: a comparison between sweet, sour, and whey buttermilk

Buttermilk is a dairy ingredient widely used in the food industry because of its emulsifying capacity and its positive impact on flavor. Commercial buttermilk is sweet buttermilk, a by-product from churning sweet cream into butter. However, other sources of buttermilk exist, including cultured and whey buttermilk obtained from churning of cultured cream and whey cream, respectively. The compositional and functional properties (protein solubility, viscosity, emulsifying and foaming properties) of sweet, sour, and whey buttermilk were determined at different pH levels and compared with those of skim milk and whey. Composition of sweet and cultured buttermilk was similar to skim milk, and composition of whey buttermilk was similar to whey, with the exception of fat content, which was higher in buttermilk than in skim milk or whey (6 to 20% vs. 0.3 to 0.4%). Functional properties of whey buttermilk were independent of pH, whereas sweet and cultured buttermilk exhibited lower protein solubility and emulsifying properties as well as a higher viscosity at low pH (pH ≤ 5). Sweet, sour, and whey buttermilks showed higher emulsifying properties and lower foaming capacity than milk and whey because of the presence of milk fat globule membrane components. Furthermore, among the various buttermilks, whey buttermilk was the one showing the highest emulsifying properties and the lowest foaming capacity. This could be due to a higher ratio of phospholipids to protein in whey buttermilk compared with cultured or sweet buttermilk. Whey buttermilk appears to be a promising and unique ingredient in the formulation of low pH foods.     

The effect of linear velocity and flux on performance of ceramic graded permeability membranes when processing skim milk at 50°C

Raw milk (about 500 kg) was cold (4°C) separated and then the skim milk was pasteurized at 72°C and a holding time of 16 s. The milk was cooled to 4°C and stored at ≤4°C until processing. The skim milk was microfiltered using a pilot-scale ceramic graded permeability (GP) microfilter system equipped with 0.1-µm nominal pore diameter ceramic Membralox membranes. First, about 155 kg of pasteurized skim milk was flushed through the system to push the water out of the system. Then, additional pasteurized skim milk (about 320 kg) was microfiltered (stage 1) in a continuous feed-and-bleed 3× process using the same membranes. The retentate from stage 1 was diluted with pasteurized reverse osmosis water in a 1:2 ratio and microfiltered (stage 2) with a GP system. This was repeated 3 times, with total of 3 stages in the process (stage 1 = microfiltration; stages 2 and 3 = diafiltration). The results from first 3 stages of the experiment were compared with previous data when processing skim milk at 50°C using ceramic uniform transmembrane pressure (UTP) membranes. Microfiltration of skim milk using ceramic UTP and GP membranes resulted in similar final retentate in terms of serum proteins (SP) removed. The SP removal rate (expressed by kilogram of SP removed per meter-squared of membrane area) was higher for GP membranes for each stage compared with UTP membranes. A higher passage of SP and SP removal rate for GP than UTP membranes was achieved by using a higher cross-flow velocity when processing skim milk. Increasing flux in subsequent stages did not affect membrane permeability and fouling. We operated under conditions that produced partial membrane fouling, due to using a flux that was less than limiting flux but higher than critical flux. Because the critical flux is a function of the cross-flow velocity, the difference in critical flux between UTP and GP membranes resulted only from operating under different cross-flow velocities (6.6 vs 7.12 for UTP and GP membranes, respectively). Conditions that allow microfiltration operation at higher flux will reduce the membrane surface area required to process the same amount of milk in the same length of time. Less membrane surface area reduces investment costs and uses less energy, water, and chemicals to clean the microfiltration system.     

Effect of whey protein concentration on the fouling and cleaning of a heat transfer surface

In the studies of fouling and cleaning of heat exchange surfaces in dairy plants, whey protein deposits and heat induced whey protein gels (HIWPG) are considered as suitable model material to simulate the proteinaceous based type “A” milk fouling. Protein concentration of the fouling solution may significantly influence the formation of milk deposits on heat exchange surfaces, hence affecting the cleaning efficiency. In this study, a laboratory produced heat induced whey protein gels (HIWPG) and a pilot plant heat exchanger fouling/cleaning were used to investigate the effect of protein concentration on formation and cleaning of dairy fouling. Here, HIWPGs made from different protein concentrations were formed in capsules and then dissolved in aqueous sodium hydroxide (0.5 wt%). The dissolution rate calculation based on the UV spectrophotometer analysis. In the pilot-scale plant study, whey protein fouling deposits were formed by recirculating whey protein solutions with different concentrations through the heat exchange section in different runs, respectively. The deposit layers were then removed by recirculating aqueous sodium hydroxide (0.5 wt%) and the cleaning efficiency was monitored in the form of the recovery of heat transfer coefficient while both fluid electric conductivity and turbidity were monitored as indications of cleaning completion. It was found that increasing the protein concentration of the HIWPG significantly increased the gel hardness and the dissolution time. In addition, increasing the protein concentration significantly increased both, the amount of the fouling on the pilot-scale plant and the time required to clean the fouling deposit.     

陶瓷膜对脱脂乳中酪蛋白与其他组分的高效分离

量化200 nm陶瓷膜脱除乳清蛋白、乳糖、灰分和钙的能力。在50℃条件下,对脱脂乳进行3倍浓缩,之后连续2次补水至原体积进行稀释过滤,浓缩倍数均为3,最终得到3次滤液,计算各组分总脱除率。结果表明乳糖脱除率为85.81%,α-乳白蛋白脱除率为79.27%,β-乳球蛋白脱除率为71.64%,灰分脱除率为62.16%,钙脱除率为35.64%。稀释过滤完毕后膜的纯水膜通量衰减系数为89.27%,使用质量分数为2%氢氧化钠和0.5%的硝酸溶液进行清洗,膜通量的恢复系数为99.07%。     

基于陶瓷膜技术的羊乳酪蛋白和其他组分的高效分级分离

为量化0.05 μm陶瓷膜脱除羊乳中乳清蛋白、乳糖、灰分、钙和磷的能力,在50 ℃条件下,脱脂乳进行3 倍浓缩,之后2 次间歇补水至原体积进行清洗过滤,最终得到1 份截留液、3 份透过液,并计算各组分总脱除率。结果表明：乳清蛋白脱除率为96.17%,乳糖脱除率为86.42%,灰分脱除率为73.39%,钙脱除率为34.90%,磷脱除率为55%。稀释过滤完毕后膜的纯水膜通量衰减系数为55.57%,使用质量分数为2%氢氧化钠和1%的硝酸溶液进行清洗,膜通量的恢复系数为99.21%。0.05 μm陶瓷膜可以实现羊乳酪蛋白和其他组分的有效分离,该技术适合在没有干酪乳清的条件下,以生鲜乳为原料加工酪蛋白胶束粉、乳清蛋白粉、乳糖等乳基配料产品。     

Effect of skim milk treated with high hydrostatic pressure on permeate flux and fouling during ultrafiltration

Ultrafiltration (UF) is largely used in the dairy industry to generate milk and whey protein concentrate for standardization of milk or production of dairy ingredients. Recently, it was demonstrated that high hydrostatic pressure (HHP) extended the shelf life of milk and improved rennet coagulation and cheese yield. Pressurization also modified casein micelle size distribution and promoted aggregation of whey proteins. These changes are likely to affect UF performance. Consequently, this study determined the effect of skim milk pressurization (300 and 600 MPa, 5 min) on UF performance in terms of permeate flux decline and fouling. The effect of HHP on milk proteins was first studied and UF was performed in total recycle mode at different transmembrane pressures to determine optimal UF operational parameters and to evaluate the effect of pressurization on critical and limiting fluxes. Ultrafiltration was also performed in concentration mode at a transmembrane pressure of 345 kPa for 130 or 140 min to evaluate the decline of permeate flux and to determine fouling resistances. It was observed that average casein micelle size decreased by 32 and 38%, whereas β-lactoglobulin denaturation reached 30 and 70% at 300 and 600 MPa, respectively. These results were directly related to UF performance because initial permeate fluxes in total recycle mode decreased by 25% at 300 and 600 MPa compared with nonpressurized milk, critical flux, and limiting flux, which were lower during UF of milk treated with HHP. During UF in concentration mode, initial permeate fluxes were 30% lower at 300 and 600 MPa compared with the control, but the total flux decline was higher for nonpressurized milk (62%) compared with pressure-treated milk (30%). Fouling resistances were similar, whatever the treatment, except at 600 MPa where irreversible fouling was higher. Characterization of the fouling layer showed that caseins and β-lactoglobulin were mainly involved in membrane fouling after UF of pressure-treated milk. Our results demonstrate that HHP treatment of skim milk drastically decreased UF performance.     

Forward osmosis concentration of milk: Product quality and processing considerations

Concentration of milk in the dairy industry is typically achieved by thermal evaporation or reverse osmosis (RO). Heat concentration is energy intensive and leads to cooked flavor and color changes in the final product, and RO is affected by fouling, which limits the final achievable concentration of the product. The main objective of this work was to evaluate forward osmosis (FO) as an alternative method for concentrating milk. The effects of fat content and temperature on the process were evaluated, and the physicochemical properties and sensory qualities of the final product were assessed. Commercially pasteurized skim and whole milk samples were concentrated at 4, 15, and 25°C using a benchtop FO unit. The FO process was assessed by monitoring water flux and product concentration. The color of the milk concentrates was also evaluated. A sensory panel compared the FO concentrated and thermally concentrated milks, diluted to single strength, with high temperature, short time pasteurized milk. The FO experimental runs were conducted in triplicate, and data were analyzed by single-factor ANOVA. Water flux during FO decreased with time under all processing conditions. Higher temperatures led to faster concentration and higher concentration factors for both skim and whole milk. After 5.75 h of FO processing, the concentration factors achieved for skim milk were 2.68 ± 0.08 at 25°C, 2.68 ± 0.09 at 15°C, and 2.36 ± 0.08 at 4°C. For whole milk, after 5.75 h of FO processing, concentration factors of 2.32 ± 0.12 at 25°C, 2.12 ± 0.36 at 15°C, and 1.91 ± 0.15 at 4°C were obtained. Overall, maximum concentration levels of 40.15% total solids for skim milk and 40.94% total solids for whole milk were achieved. Additionally, a triangle sensory test showed no significant differences between regular milk and FO concentrated milk diluted to single strength. This work shows that FO is a viable nonthermal processing method for concentrating milk, but some technical challenges need to be overcome to facilitate commercial utilization.     

Effects of caseinate, whey and milk powders on the texture and microstructure of emulsified chicken meat batters

The effects of adding dry caseinate, whole milk, skim milk, regular, and modified whey protein powders, at a level of 2 g/100 g, to meat batters were studied. All dairy additives, except for the regular whey, significantly reduced cook loss (30-50% reduction). Caseinate and modified whey formed distinct dairy protein gel regions within the meat batters, as revealed by light microscopy. Both also contributed more to enhancing the textural properties of the meat batters compared to the other dairy proteins; i.e., increasing texture profile analysis fracturability, and hardness, respectively. Overall, the most cost-effective ingredient appeared to be the modified whey, which also provided the best moisture retention.     

The effect of ultrasound on casein micelle integrity

Samples of fresh skim milk, reconstituted micellar casein, and casein powder were sonicated at 20 kHz to investigate the effect of ultrasonication. For fresh skim milk, the average size of the remaining fat globules was reduced by approximately 10 nm after 60 min of sonication; however, the size of the casein micelles was determined to be unchanged. A small increase in soluble whey protein and a corresponding decrease in viscosity also occurred within the first few minutes of sonication, which could be attributed to the breakup of casein-whey protein aggregates. No measurable changes in free casein content could be detected in ultracentrifuged skim milk samples sonicated for up to 60 min. A small, temporary decrease in pH resulted from sonication; however, no measurable change in soluble calcium concentration was observed. Therefore, casein micelles in fresh skim milk were stable during the exposure to ultrasonication. Similar results were obtained for reconstituted micellar casein, whereas larger viscosity changes were observed as whey protein content was increased. Controlled application of ultrasound can be usefully applied to reverse process-induced protein aggregation without affecting the native state of casein micelles.     

Chicory juice clarification by membrane filtration using rotating disk module

Clarification is the first step of inulin production from chicory juice, and membrane filtration as an alternative can greatly simplify this process, increase juice yield, improve product quality, and reduce the cost and waste volume. In this study, a rotating disk module (RDM) was used to investigate the clarification of chicory juice by four micro- and ultrafiltration membranes. Compared with dead end filtration, the RDM had a much higher permeate flux and product quality. High rotating speeds produced high permeate fluxes and reduced flux decline, because of the strong back transport of foulant from fouling layer to feed solution. At high rotating speeds of 1500–2000 rpm, the permeate flux increased with membrane pore size and transmembrane pressure (TMP), while at low rotating speeds (<1000 rpm), permeate flux was independent of membrane type and TMP due to a thick deposited fouling layer as a dominant filtration resistance, while carbohydrate transmission decreased at higher TMP because of denser cake layer as an additional selective membrane. The highest carbohydrate transmission (∼98%) and desirable permeate turbidity (2.4 NTU) was obtained at a TMP of 75 kPa and a rotating speed of 2000 rpm for FSM0.45PP membrane. With the RDM, the Volume Reduction Ratio (VRR) could reach 10 with a high permeate flux (106 L m⁻² h⁻¹) in the concentration test, and permeate was still rich in carbohydrate and well clarified. Chemical cleaning with 0.5% P3-ultrasil 10 detergent solution was able to recover 90% water flux of fouled membrane.     

Effect of heating on the distribution of transforming growth factor-��2 in bovine milk

The objective of this study was to characterize the impact of heat treatments on the distribution of transforming growth factor-beta (TGF-??2) between cream and skim milk and between the casein and whey fractions of skim milk. Skimming removed 45% and 62% of the TGF-??2 from raw and pasteurized milks and only 8% of the total TGF-??2 in skimmed pasteurized milk was found in whey, compared to 37% in whey from raw skimmed milk. The TGF-??2 content of whey decreased as the heat treatment of the milk increased in intensity (thermization > pasteurization > UHT sterilization). Using milk held for 1 or 2 min at temperatures ranging from 57 to 84 °C, it was shown that TGF-??2 in the whey portion decreases at temperatures above 66 °C and becomes undetectable at temperatures higher than 76 °C. Altogether, these data on the heat-induced changes in TGF-??2 content of cream, skim milk, casein and whey reveal a potentially negative impact of certain heat treatments in developing TGF-??2-enriched fractions from milk.     

USE OF ULTRAFILTRATION/REVERSE OSMOSIS SYSTEMS FOR THE CONCENTRATION AND FRACTIONATION OF WHEY

SUMMARY– Existing ultrafiltration/reverse osmosis technology provides a means of fractionating and concentrating cheese whey into liquid fractions containing a variety of protein: lactose ratios. These ratios may range from about 1:8 (raw whey) through 3:5 (a "skim milk equivalent") to 2:1 or higher. If a two- or three-stage ultrafiltration system were used with water injection between stages, a product with a protein:lactose ratio of 20:1 could be obtained. The exact protein:lactose ratio in the concentrate stream is a function of the permeability and selectivity characteristics of the membrane, and the system design and operating conditions. Some of the sanitation problems associated with the introduction of these new unit operations in the dairy and food processing industries are also treated at length.     

REVERSE OSMOSIS — CONCENTRATION APPLICATIONS

An examination is made of the influence of the main processing variables in the reverse osmosis (RO) concentration of both whey and skim milk. Up to a solids concentration ceiling of ca. 25 per cent, RO offers an economic advantage over thermal evaporation as a concentration technique, in terms of lower operating costs. Furthermore, RO possesses the additional advantage that it concentrates without heat, thereby retaining the native properties of the original raw material before concentration. In particular, there is no heat denaturation of the thermolabile whey proteins and, hence, no reduction of valuable functional properties and no cooked flavour development.
Based on these inherent advantages of RO over thermal evaporation, various process and product applications of RO to whey and skim milk utilization have been suggested. Additional features of membrane processing by RO have also been discussed.     

Evaluation of commercially available, wide-pore ultrafiltration membranes for production of α-lactalbumin-enriched whey protein concentrate

Commercially available, wide-pore ultrafiltration membranes were evaluated for production of α-lactalbumin (α-LA)-enriched whey protein concentrate (WPC). In this study microfiltration was used to produce a prepurified feed that was devoid of casein fines, lipid materials, and aggregated proteins. This prepurified feed was subsequently subjected to a wide-pore ultrafiltration process that produced an α-LA-enriched fraction in the permeate. We evaluated the performance of 3 membrane types and a range of transmembrane pressures. We determined that the optimal process used a polyvinylidene fluoride membrane (molecular weight cut-off of 50 kDa) operated at transmembrane pressure (TMP) of 207 kPa. This membrane type and operating pressure resulted in α-LA purity of 0.63, α-LA:β-LG ratio of 1.41, α-LA yield of 21.27%, and overall flux of 49.46 L/m²·h. The manufacturing cost of the process for a hypothetical plant indicated that α-LA-enriched WPC 80 (i.e., with 80% protein) could be produced at $17.92/kg when the price of whey was considered as an input cost. This price came down to $16.46/kg when the price of whey was not considered as an input cost. The results of this study indicate that production of a commercially viable α-LA-enriched WPC is possible and the process developed can be used to meet worldwide demand for α-LA-enriched whey protein.     

Flux pattern and fouling of membranes during ultrafiltration of some dairy products

The flux pattern of milk (whole milk, skimmed milk and buttermilk) showed a distinct contrast to whey (sweet whey and acid casein whey) systems during ultrafiltration (UF) at constant composition. For milk systems, the initial flux was lower than for whey systems, but the flux stabilised within a few minutes of operation. However, for both acid and sweet whey, the flux continued to fall with time for nearly 1 h. The fouling coefficient (FC) for buttermilk was 0.6 after 5 min, rising to 0.68 after 60 min. During this time flux did not decline, suggesting that concentration polarisation (CP), rather than fouling, was controlling the flux rate. In contrast, the FC for whey was 0.5 after 5 min with a progressive rise over the next 55 min, but the flux also fell throughout this period suggesting that flux was controlled by fouling rather than CP. The higher concentration of protein in milk systems appeared to be responsible for CP, rather than fouling, being the controlling mechanism.     

Development and optimization of a carbon dioxide-aided cold microfiltration process for the physical removal of microorganisms and somatic cells from skim milk

Physical removal of microorganisms from skim milk by microfiltration (MF) is becoming increasingly attractive to the dairy industry. Typically, this process is performed at temperatures of approximately 50°C. Additional shelf-life and quality benefits might be gained by conducting the MF process at low temperatures. Cold MF could also minimize microbial fouling of the membrane and prevent the germination of thermophilic spores. The objective of this study was to optimize a cold MF process for the effective removal of microbial and somatic cells from skim milk. An experimental MF setup containing a tubular Tami ceramic membrane with a nominal pore size of 1.4 μm was used for MF of raw skim milk at a temperature of 6 ± 1°C. The processing conditions used were cross-flow velocities of 5 to 7 m/s, and transmembrane pressures of 52 to 131 kPa. All MF experiments were performed in triplicate. The permeate flux was determined gravimetrically. Microbiological, chemical, and somatic cell analyses were performed to evaluate the effect of MF on the composition of skim milk. The permeate flux increased drastically when velocity was increased from 5 to 7 m/s. The critical transmembrane pressure range conducive to maximum fluxes was 60 to 85 kPa. When MF was conducted under optimal conditions, very efficient removal of vegetative bacteria, spores, and somatic cells, as well as near complete transmission of proteins into the MF milk, was achieved. To further enhance the flux, a CO₂ backpulsing system was developed. This technique is able both to increase the flux and to maintain it steadily for an extended period of time. The CO₂-aided cold MF process has the potential to become economically attractive to the dairy industry, with direct benefits for the quality and shelf life of dairy products.