If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Minor acidification of diafiltration water using various acidification agents affects the composition and rennet coagulation properties of the resulting microfiltration casein concentrate
Faculty of Chemistry, Biotechnology and Food Science (KBM), Norwegian University of Life Sciences (NMBU), Box 5003, N-1433 Ås, NorwayTINE SA R&D, Bedriftsveien 7 Kalbakken, 0901 Oslo, Norway
Cheese made from microfiltration (MF) retentate may suffer from textural defects due to a high Ca concentration. The reduction of colloidal minerals by the acidification of milk before MF at pH below 6.0 has been well documented in the literature. This process, however, creates less valuable side streams to the MF process and induces changes in the casein micelles that negatively affect their coagulation properties. The objective of this study was to determine whether a minor reduction in pH by using different acidifiers in the diafiltration (DF) water could induce changes in composition and renneting properties of the MF retentate. A 2-stage filtration process was used, with the first designed to increase the casein concentration to 8% and the second to slightly reduce the casein concentrate by 0.1 pH unit by DF, without influencing the total protein concentration. Four acidifying agents were tested during DF: lactic acid, hydrochloric acid, citric acid, and carbon dioxide. Diafiltration with water was used as a reference. At the start of DF, the retentates of acid DF had a slightly reduced pH, with an average of 0.09, whereas the pH of the reference retentate increased by an average of 0.07 unit. The reference retentate regained its starting pH by the end of DF. The carbonated retentate gradually increased in pH during processing, whereas the pH of the lactic, hydrochloric, and citric acid retentates remained constant. The permeate from the lactic acid and carbonated treatments had a reduced whey protein content compared with the reference. The total P and inorganic phosphate were lowered in the retentate by using carbonation. The total amount of Mg and Na were lowered in the retentate by using citric acid. The ionic Ca content in the retentate increased with use of lactic or hydrochloric acid. The type of acidifier used reduced the rennet clotting time. Combined acidified diafiltration with a slight reduction affects the permeate composition and improves the retentate clotting time despite the minimal mineral modification.
Microfiltration (MF) is of increasing interest for the cheese industry, mostly for its protein separation selectivity, where high casein retentates and native whey protein (WP) permeates can be obtained (
The MF process influences the chemical composition of the obtained retentate; as the casein concentration increases, the colloidal and casein-bound mineral contents and buffer capacity also increase (
). Therefore, adjustment of the mineral content of the concentrated cheese milk before cheesemaking is necessary. The reduction of mineral content in MF retentate could be achieved by acidification of the milk (
Effects of depleting ionic strength on 31P nuclear magnetic resonance spectra of micellar casein during membrane separation and diafiltration of skim milk.
). The effect of acidifying milk to pH 6.0 and below, where changes in mineral balance are apparent, is well documented in the literature. The solubilization of colloidal Ca phosphate (CCP) from the casein micelle upon acidification of milk causes an increase in the content of soluble minerals such as various forms of pH-dependent phosphates and Ca2+ (
). Colloidal phosphate is fully solubilized at pH 5.2, whereas Ca requires a further decrease in pH for complete solubilization. As casein concentration increases, an even further reduction in pH is required to induce solubilization of CCP (
). The type of acidifier used plays a role in the ionic composition of the serum phase, which further affects the functionality of the casein micelle (
). Moreover, filtration of highly acidified milk yields a relatively acidic permeate, the value of which for further applications is reduced. Thus, this process is not a practical solution for industries when the side streams of the MF process are intended for usage in other applications. The DF process, which aims to remove lactose and soluble minerals and increase protein content, alters the mineral balance of the casein micelle.
showed that the extent of DF (0.5 and 1 diavolume; DV) decreased the soluble Ca and P contents in the retentate. These authors also reported a disruption of the casein micelle and increased solubility of κ-, β-, and αS-CN, a reduced rennet gel stiffness, and an increased gelation time.
Effects of depleting ionic strength on 31P nuclear magnetic resonance spectra of micellar casein during membrane separation and diafiltration of skim milk.
reported that by altering the DF medium (i.e., by adding citrate to the DF water), the total P:Ca ratio increased, and this shifted the casein phosphate nanoclusters signals recorded by 31P nuclear magnetic resonance compared with using regular DF water. They also observed a shift toward an acidic pH with a reduction in Ca and P contents when using citric acid as the DF medium. Therefore, addition of modifying agents to the DF water and the DF factor could be used to modify the composition of casein concentrates.
The aim of the present study was to find an intermediate solution with capabilities to both (1) allow a slight reduction in milk pH as a prepreparation step for further cheesemaking to improve the texture of cheese made from MF milk without disruption of the casein micelle and (2) maintain a valorized stream of the MF process (i.e., resulting in a nonacidic permeate). The objective of this study was thus to investigate an MF-DF process, where different acidifiers added to the DF water were tested for their potential effect on the resulting casein concentrate retentate and permeate. To the authors' knowledge, this is the first comprehensive study comparing the effect of different acidifying agents on the retentate and permeate composition following an acid DF process.
MATERIALS AND METHODS
Sample Preparation
Figure 1A illustrates the 2-stage production process of the casein concentrate used in this study. Fresh milk, obtained from the Animal Production Experimental Centre at the Norwegian University of Life Sciences (Ås), was skimmed (Westfalia Separator AG, MSD50-01-076, Oelde, Germany), pasteurized (A3-HRB, Alfa Laval, Lund, Sweden), and microfiltered (UF/MF pilot MCC RV 01118340, APV, Slikeborg, Denmark) at 50°C using a 0.14-µm ceramic membrane (Inside Céram, Tami Industries, GEA, Nyons, France) at uniform transmembrane pressure (51 ± 4 kPa) to 8 ± 0.1% (wt/wt) casein concentrates, as previously described by
Effect of freezing temperatures and time on mineral balance, particle size, rennet and acid coagulation of casein concentrates produced by microfiltration.
. The macrocomposition of the retentate during MF was determined by a MilkoScan FT1 (CombiFoss 6500, Foss, Hillerød, Denmark) using Fourier transform infrared analysis as a fast method to monitor the protein concentration. The casein concentrates were distributed in 1-L sterile containers and kept at 4°C before further filtration.
Figure 1Schematic representation of (A) 2-stage filtration process of microfiltration (MF) and diafiltration (DF) and (B) details of DF treatments and nomination. DV = diavolume.
The DF process was initiated after MF concentration by adding 0.3 DV of DF medium to the MF retentate. The DV was chosen based on a target lactose content of the MF-DF retentate of 3.9 ± 0.1%. Tap water was used as the reference DF medium to mimic industrial practice. Acidified DF medium was tap water with the addition of agents as described below. Filtration was then continued at 50°C using a Labscale TFF system (Millipore, Oslo, Norway), with a Pellicon XL Cassette and a Biomax membrane (500 kDa, cat. no. PBX500C50; Millipore) at an inlet and outlet pump pressure of ∼3 bar. The recirculation of the casein concentrates in the feed tank and collection of permeate were initiated after discarding the first 30 mL of retentate and permeate from the system. The concentration continued until the volume of the collected permeate corresponded to the volume of added acidified water (measured using a graduated cylinder). The membranes were cleaned according to the manufacturer's instructions.
The individual acidifying agents in tap water used for DF were (1) 10 mM lactic acid (dl-lactic acid, 85% wt/wt, Sigma-Aldrich, St. Louis, MO), (2) 10 mM HCl (hydrochloric acid, ACS reagent, 37%, Sigma-Aldrich), (3) 10 mM citric acid (10% wt/wt citric acid monohydrate solution, citric acid monohydrate ACS reagent, ≥99.0%, Sigma-Aldrich), and (4) 1.69 g/L CO2. The carbonation of water with CO2 was performed using a SodaStream (Aqvia, AGA, Stockholm, Sweden). Adjustment for the concentration of CO2 was measured using an PBA-B instrument connected to a CarboQC ME module with a PFD filling device and operated through Generation M instrument software version 2.42 (Anton Paar, Graz, Austria). The concentration of CO2 used was based on a decrease in the pH of the DF water to 4.6. The selected concentration of each acidifying agent was previously adjusted to correspond to the amount required to obtain a decrease in pH of the casein concentrate by 0.1 unit after the addition of acidified water. Continuous pH measurements were performed during acid DF using a 742020 Hach SensIon+ PH31 meter with 5011T probe (Lange GmbH, Dusseldorf, Germany) connected to LabCom V2.1 software (Lange GmbH).
The retentates and permeates obtained from the MF and DF were abbreviated as follows (Figure 1B): (1) MF retentate (MFR) and permeate (MFP), (2) DF reference retentate (RDR) and permeate (RDP), (3) DF lactic acidified retentate (LDR) and permeate (LDP), (4) DF HCl retentate (HDR) and permeate (HDP), (5) DF citric acid retentate (CDR) and permeate (CDP), and (6) DF carbonated retentate (ODR) and permeate (ODP). The skim milk (
) was calculated by subtracting NPN from total N. Casein was calculated by subtracting noncasein N from total N. Undenatured WP was calculated by subtracting NPN from noncasein N. The results were multiplied by the factor 6.38 to calculate the percentage of protein. The identification and quantification of the protein composition were performed using capillary electrophoresis in combination with the Kjeldahl results as described by
. Inorganic phosphate (Pi) was analyzed using Agilent (G1600AX) capillary electrophoresis with ChemStation software (Agilent Technologies, Waldbronn, Germany) according to
Effect of freezing temperatures and time on mineral balance, particle size, rennet and acid coagulation of casein concentrates produced by microfiltration.
. Calcium ion activity was determined using an Orion 97-20 calcium ion selective electrode (Calcium Ionplus Sure-Flow Plastic Membrane Combination ISE, Thermo Scientific, Chelmsford, MA) with an mV meter (PHM290, pH-STAT Controller, MeterLab, Radiometer Analytica, Copenhagen, Denmark).
Rennet Coagulation Properties
The rennet coagulation properties of the skim milk and retentates from MF and DF were measured using a Formagraph (Lattodinamografo; Foss Italia SpA, Padova, Italy) as described by
. A 10-mL aliquot of a casein concentrate sample was incubated at 32°C for 30 min in the sample cuvette, followed by addition of 200 μL of rennet (CHY-MAX; Chr. Hansen A/S, Hørsholm, Denmark) previously diluted 1:50 in acetate buffer (pH 5.6), and was immediately analyzed for 45 min at 32°C. Coagulation analysis included the rennet clotting time (RCT; min), curd firming time shown as time from RCT until the width of the bifurcate was 20 mm, and curd strength (in mm of distance) of the bifurcate after 30 min of coagulation.
Statistical Analysis
The entire experiment was carried out in 3 filtration replicate blocks with 3 separate milk batches. Samples obtained from MF and DF of each batch were analyzed in triplicate. The results are the averages of those obtained from the 3 replicate blocks. Statistical significances for the DF treatments were evaluated using ANOVA at P < 0.05 with the DF treatment as a fixed factor and the replicate block as a random factor. The mean values were compared using a Tukey pairwise comparison test, and all statistical data were processed using packages and functions in R Studio (version 1.1.456, https://rstudio.com/products/rstudio/download/).
RESULTS
Change in Retentate pH During DF
Throughout the DF process, the pH value differed significantly (P < 0.05) between the reference and the different acidification treatments (Figure 2). By adding acidified DF water to the MF retentate, the pH of the retentate instantly decreased by ∼0.1 unit, whereas the reference sample increased by ∼0.1 pH unit after water addition (Table 1). Throughout the DF process and removal of permeate, the pH of the reference retentate decreased to regain its starting value of pH 6.6 ± 0.03. In contrast, the pH of HDR and CDR remained relatively stable at 6.47 ± 0.02 and 6.43 ± 0.06, respectively. The LDR showed a slight reduction in pH during DF to 6.43 ± 0.1, and the ODR pH increased throughout the DF process to 6.5 ± 0.01. The LDR and CDR samples had a significantly (P < 0.05) lower end pH compared with the reference sample.
Figure 2Change in pH during the diafiltration process of microfiltration retentate. − reference; −·− lactic acid; − − HCl; −▴− citric acid; ···· carbonated water. Means ± SD.
Table 1Effect of acidifying agents used during diafiltration (DF) of casein concentrates on 0.1-pH changes at instant (Change) and at the end point (End) of the DF process and on rennet coagulation properties (RCP) as measured by Formagraph
The protein concentration returned to 8% (wt/wt) after DF (Table 2). The contents of true protein, casein, and protein fractions (Supplemental Table S1, https://doi.org/10.3168/jds.2020-18237) in the retentates were not influenced by the DF medium. However, the total WP content in the retentates was significantly (P < 0.05) reduced in the HDR and ODR compared with the RDR.
Table 2Effect of different acidifying agents used for diafiltration (DF) on the composition of the retentate and the permeate compared with skim milk (SM) and microfiltered retentate (MFR) and permeate (MFP)
The protein compositional analysis of the permeate (Table 2) showed no casein leakage after MF and DF, but the different acid DF influenced the amount of true protein (total WP) in the permeate. The specific contents of α-LA and β-LG in the permeate (Table 3) were significantly (P < 0.05) reduced in the LDP and ODP samples compared with the RDP.
Table 3Calculations of protein compositions in skim milk, microfiltration (MF) retentate, and diafiltered (DF) permeate based on capillary electrophoresis results
RC = relative concentration (%); calculated as peak area divided by total peak area. Adj TPC = protein concentration (%) as adjusted for true protein content in the sample. Adj MEC = protein concentration (%) as adjusted for true protein content in the sample and molar extinction coefficient (absorbance) of the protein.
Means of protein concentration within the same row with different superscripts differ (P < 0.05).
Sum
100
100
100
100
100
a–c Means of protein concentration within the same row with different superscripts differ (P < 0.05).
1 RC = relative concentration (%); calculated as peak area divided by total peak area. Adj TPC = protein concentration (%) as adjusted for true protein content in the sample. Adj MEC = protein concentration (%) as adjusted for true protein content in the sample and molar extinction coefficient (absorbance) of the protein.
The mineral composition of the retentates and the permeates from the 2-stage filtration process is shown in Table 4 and Figure 3, respectively. The relatively large milk batch variation in mineral composition caused large standard deviations within treatments. The total contents of Ca and K in the retentate were not affected by the method of DF compared with the reference. The total contents of Mg, Na, and P in the retentate were influenced by the type of acid DF, and DF with carbonated water had the greatest effect compared with other treatments. Acidifying the DF water with CO2 reduced the total Mg and Na contents of the ODR, which were significantly (P < 0.05) lower than in CDR, which had the highest Mg and Na contents in the DF retentates. The CDR also had a significantly (P < 0.05) higher total Na content compared with LDR. The P and Pi contents were influenced by the DF medium, and ODR had significantly (P < 0.05) reduced total P and Pi contents compared with RDR. Diafiltration influenced the Ca2+ content of the DF retentate; LDR and HDR had a significantly (P < 0.05) higher Ca2+ concentration than RDR, which had the lowest Ca2+ content. The ODR did not significantly (P < 0.05) differ in Ca2+ concentration from any of the other retentates.
Table 4Effect of acidifying agent used for diafiltration (DF) on the mineral content (mM) of the retentate
Figure 3Mineral content (mM) of permeate from the diafiltration process of microfiltered retentate. CDP = citric acid permeate; HDP = HCl permeate; LDP = lactic acid permeate; ODP = carbonated permeate; RDP = reference. Means and SD of 3 batches. Letters (a, b) within the same plot represent significance differences (P < 0.05) between the diafiltration treatments. Boxes describe the variations in the data, mid lines are the median that splits the data between the lower 50% of observation from the upper 50% of observation, the lower whisker is the minimum and represents the lowest observation, and the upper whisker is the maximum and represents the highest observation in the dataset.
In the permeate, the total Ca and K were significantly (P < 0.05) influenced by some of the acid DF treatments, with CDP having a significantly (P < 0.05) higher total Ca content than RDP. In addition, HDP and LDP had a significantly (P < 0.05) higher total K content than RDP. The concentration of Ca2+ in the permeate was consequently influenced by the acid DF treatment, with HDP, LDP, and ODP showing a significant increase in Ca ions compared with RDP due to the change in pH. The P measured in the permeate most likely represented Pi, as no casein was detected in the DF permeate. The levels of P showed a slight variation between the DF treatments, but no significant differences between the treatments were found. Furthermore, total Mg or Na content in the permeate was significantly (P < 0.05) influenced by the DF treatments.
Organic Acids and Lactose in the Retentate and Permeate
As expected, DF with lactic acid and citric acid significantly (P < 0.05) increased their contents in their respective retentates and permeates (Table 1), whereas in RDR there was a tendency toward a reduced citric acid content during DF. The lactose content in the retentate seemed to be more reduced within RDR compared with acid DF, whereas its content in the permeate was not significantly (P < 0.05) influenced by the various treatments (Table 1).
Rennet Coagulation Properties
The use of acidified DF water significantly (P < 0.05) reduced the RCT of the resulting retentate compared with the reference retentate, as shown in Table 4, which was to some extent related to the pH. The LDR obtained the shortest RCT. Neither the firmness of the formed gels nor the curd firming time were significantly (P < 0.05) influenced by the treatments.
DISCUSSION
The acid DF step changed the pH of the retentate slightly, as anticipated, whereas the pH of the retentate was only temporarily changed by using only water during DF (RDR). An increase in the retentate pH by addition of DF water has been previously reported (
Effects of depleting ionic strength on 31P nuclear magnetic resonance spectra of micellar casein during membrane separation and diafiltration of skim milk.
), as well as a decrease in pH when using acidified DF water. For most of the acid DF retentates, the pH remained relatively stable throughout the DF process, with a tendency for the LDR to further decrease and a tendency for ODR to increase over time during DF. The temperature of the carbonated water was low (∼4°C) when added to the retentate for DF. The temperature of the retentate, which at that moment was 50°C, automatically dropped together with the pH due to the addition of carbonated water. Throughout the DF process the temperature of the retentate increased to the set DF temperature (50°C). Therefore, the solubility of CO2 was gradually reduced due to the increase in temperature and equilibration with the atmospheric pCO2 (pressure change;
In concentrated milk, with higher casein and colloidal mineral contents compared with nonconcentrated milk, a shift toward a lower pH value would be necessary for increased solubilization of Ca, Pi, and Mg compared with unconcentrated milk (
). In this study, the protein content of the retentate was 8% and the added acidifiers targeted only a 0.1-unit decrease in pH as the DF was a preparation for cheesemaking. This provided retentates with a final pH ranging from 6.4 to 6.5, and major changes in the mineral balance were not expected, resulting in a nonsignificant change in total Ca, and changes in total P, Pi, and Mg for only some acid DF (CDR and ODR) compared with RDR. The total and soluble Ca contents in the permeate, however, reflected the influence of acid DF and indicated a trend toward their increase in the acidified permeates compared with the reference (RDP). Citric acid treatment expressed the most significant increase in soluble Ca content of the permeate. These results are consistent with the findings of
Variation of the calcium content in skim milk by diafiltration and ion exchange—Effects on permeation rate and structure of deposited layers in the RO.
). However, the Ca2+ content of the retentates in this experiment is attributed more to the type of acidifying agent used during DF and their affinity to bind Ca2+.
reported a decrease in Ca ion activity upon addition of citrate to a concentrated micellar casein solution when studying the effect of citrate on the physical changes in casein micelles in a concentrated micellar casein solution; this is due to the high Ca-chelating ability of citrate. The use of citric acid with DF, therefore, is expected to shift the Ca equilibrium toward soluble Ca citrate chelates (
explained that in skim milk depressurized after CO2 addition, the mineral composition is restored to its colloidal state. The influence of the solubilization by CO2 on the CCP is reversible at pH values above 5.8, with no insoluble forms of salt formed, after which Ca binds back to Pi and shifts back into the micelle. The observed significant increase in Ca2+ in the ODP compared with the reference (RDP) may seem contradictory to the abovementioned results. This phenomenon is explained by the fast and continuous removal of Ca2+ toward the permeate during the DF process while the depressurization of CO2 from the retentate is not yet fully complete or established, leading to a significantly (P < 0.05) higher Ca2+ content in the permeate compared with the reference.
The acid DF permeate and retentate might have formed other compounds that were not measured and could contribute to the explanation for the compositional changes occurring between the retentate and permeate phases. The hydrocarbonate (HCO3−), in the case of DF with CO2 for example, can easily bind to solubilized Ca and Mg and form carbonates.
The distribution of minerals between the permeate and retentate, as influenced by the combined DF and addition of acidifiers, is summarized in Figure 4. The MF process increases the micellar Ca and P contents in the retentate as a result of the concentration of casein micelles and the removal of milk serum. The type of DF medium influences the equilibrium of cations, anions, and soluble salts within the soluble or micellar phase of the retentate, which are then able to permeate the membrane or remain in the retentate.
Figure 4Schematic representation of the distribution of minerals between the permeate and retentate as influenced by the decrease in pH due to addition of acidifiers. DF = diafiltration.
): the lower the pH and the higher the Ca2+ concentration, the shorter the RCT. Lactic acid diafiltered retentate having the highest Ca2+ content and the lowest pH value explains the shortest RCT. A similar phenomenon was observed by
when comparing the RCT of lactic acid versus CO2-treated milk during cheesemaking. Lactic acid has the ability to form more soluble Ca2+ compared with CO2 and citric acid. The latter 2 form Ca carbonates and Ca citrate, respectively, which have lower solubility.
suggested that CO2 follows a mechanism that involves a change in the form of CCP and reorganization of the casein micelle and its surface activity in addition to the Ca2+ levels. They also reported that only carbonation to a pH below 5.8 improved the rheological properties of renneted gels, which is in line with the current finding showing that the different DF treatments did not influence the gel firmness compared with the reference. Despite the lack of effect by acid DF on curd firmness, it is important to consider the potential contribution of the slightly lower pH and change in Ca2+ to the further development of texture during cheesemaking due to the level of protein solvation (
). Another consideration would be the significantly (P < 0.05) higher levels of lactic and citric acid in the LDR and CDR, respectively, which may influence the cheese composition and texture characteristics. Citric acid can contribute to soft curd during cheesemaking (
), and milk acidified with CO2 has been reported to negatively influence the proteolytic and lipolytic activities of the starter culture and thereby cheese ripening (
). An important note is that despite the higher citric acid content in the CDR samples compared with the other DF treatments, the level of citric acid was similar to the original content in the skim milk, which might be of importance in cheese varieties in which the degradation of citrate to CO2 and diacetyl is important for cheese ripening.
The apparent stability in lactose content of the retentate following acid DF is similar to findings reported by
, who recorded the same lactose concentration between retentate powders produced by regular DF-MF or lactic acid-acidified MF-DF retentate. However, a tendency toward an increased retention of lactose in the acid DF retentate was observed in this study compared with the nonacid DF retentate. The use of a 0.3-DV DF avoided disruption to the casein micelles, as findings by
Effects of depleting ionic strength on 31P nuclear magnetic resonance spectra of micellar casein during membrane separation and diafiltration of skim milk.
showed casein micelle disruption at a DV of 0.5 and above. The absence of caseins in the permeate and their nonsignificant change in the retentate attain that. The WP content, however, was influenced in both the retentate and the permeate following acid DF. The decrease in total WP content of the acid DF retentate has been previously reported by
. In the present study, HDR and ODR had the lowest WP content compared with the reference.
Protein analysis of the permeate showed that DF water had an enhanced ability to transfer more WP to the permeate than acid DF water with lactic acid and CO2. The reduced total WP content in LDP and ODP was reflected by the reduced concentration of α-LA and β-LG compared with RDP. This result could be explained by the ability of dissolved CO2 to react with the amine groups of protein, forming carbamate (
). This study showed that the transfer of WP to the permeate was altered by the addition of acidifiers to the DF water.
CONCLUSIONS
This study showed that the added acidifiers (lactic acid, HCl, citric acid, and CO2) changed the retentate and permeate composition after acid DF of MF casein concentrates. This was observed even at a small pH reduction of 0.1 unit and a relatively small DF factor. The most noticeable change in the retentate was an increase in Ca2+ content and reduced clotting time when using lactic acid and HCl to acidify the DF water. The permeate issued from acid DF had an overall reduced WP content compared with the DF with only water as well as an increased mineral content. Citric acid presented the highest affinity to bind Ca, with a significant increase in total Ca content in the permeate and lower Ca2+ content in the retentate and permeate. Each acidifying agent had slightly different effects on the retentate and permeate composition. The choice of acidifier in the DF process should be considered based on its positive contribution to the textural attribute of the desired dairy product or the quality of the side-stream permeate.
ACKNOWLEDGMENTS
This work was financed by the Norwegian University of Life Science (NMBU, Ås, Norway) within the framework of a PhD fellowship grant. The research was associated with the project “Cheese making from casein concentrates” (grant 269064) financed by Norwegian Research Council (NRC, Oslo, Norway) and TINE SA R&D (Oslo, Norway). The authors thank TINE SA for lending the Labscale Filtration unit. The authors acknowledge Geirfinn Lund and Ola Tjåland (NMBU) for assistance during milk processing, Ahmed Abdelghani Faculty of Chemistry, Biotechnology and Food Science (KBM, NMBU) for help in developing the capillary electrophoresis analysis and for valuable discussions, Kristian Hovde Liland at the Faculty of Science and Technology (RealTek, NMBU) for valuable input into the statistical analysis, and Solfrid Lohne at the Faculty of Environmental Sciences and Nature Management (MINA, NMBU) for mineral analysis. This research used the pilot plant facilities that received a grant from the Norwegian research council (NFR; grant 208674/F50). The authors have not stated any conflicts of interest.
Effects of depleting ionic strength on 31P nuclear magnetic resonance spectra of micellar casein during membrane separation and diafiltration of skim milk.
Effect of freezing temperatures and time on mineral balance, particle size, rennet and acid coagulation of casein concentrates produced by microfiltration.
Variation of the calcium content in skim milk by diafiltration and ion exchange—Effects on permeation rate and structure of deposited layers in the RO.
30536-1&id=gr1.jpg){kind=link}
30536-1&id=gr2.jpg){kind=link}
30536-1&id=gr3.jpg){kind=link}
30536-1&id=gr4.jpg){kind=link}