Production of low-phytate soy protein isolate by membrane technologies: Impact of salt addition to the extract on the purification process

Previously, the advantages of combining electrodialysis with bipolar membranes (EDBM) to acidify a soy protein extract to pH 6 with ultrafiltration/diafiltration (UF/DF) using a 100 kDa membrane to produce a soy protein isolate (SPI) with low phytic acid to protein ratio were demonstrated. However, some limitations related to the fouling of ED spacers by precipitated proteins and to the flux decline during UF/DF at pH 6 were observed. Therefore the purpose of this work was to study the impact of adding KCl (0.12 M or 0.24 M) to the starting extract on the efficiency of both processes and on the solubility of the resulting isolates. Results indicated that adding KCl to the initial extract increased the productivity of EDBM by approximately two times and reduced proportionally its energy requirement. This was in large part due to the prevention of ED spacer fouling. Permeate flux during UF/DF was improved by as much as 20% when compared to the flux measured for the purification of pH 6 extracts without added salt. The phytic acid to protein ratio for pH 6 isolates obtained from the addition of KCl to the initial extract were also found to be lower than for isolates produced at pH 9 or by isoelectric precipitation at pH 4.5, but higher than for the pH 6 isolates obtained without addition of KCl to the initial extract. For the pH range 2-3.5, solubility of these isolates was improved when compared to the one of isolate pH 9, but it was similar to the one of the isolate pH 4.5. Industrial relevance: Production of added-value soy protein isolate is limited by the presence of high amount of phytic acid in the isolate produced by traditional isoelectric precipitation process. We presented previously an approach combining electrodialysis with bipolar membranes (EDBM) with ultrafiltration/diafiltration (UF/DF) for the production of soy protein isolate with low phytic acid to protein ratio and with improved solubility. However, fouling of the electrodialysis spacers by precipitated proteins was observed in the electrodialysis cell which limited practical industrial application of the approach. This problem was solved by the addition of KCl (0.12 M KCl or 0.24 M KCl) to the initial soy protein extract at pH 9 prior to its pH adjustment to 6 by EDBM. It resulted in the improvement of the EDBM productivity by more than 2 times and in a proportional decrease in its energy consumption. Flux during UF/DF at pH 6 was also improved by as much as 20% when KCl was added to the starting extract. In addition, the added KCl was found to be easily removed by UF/DF with the resulting isolates having a final protein content of 94 % which was found similar to the isolate obtained without addition of salt to the starting extract. Although the isolates produced from extracts with added KCl were found to have slightly lower solubility for the pH range of 2 to 3.5 than isolates produced from extract without addition of salt, their solubility was similar or better that for the isolates produced by traditional isoelectric precipitation process and with a lower phytic acid to protein ratio. These isolates could be considered as valuable ingredients for the formulation of fruit juice beverages or power juices, considering that the pH of these liquid food products is around 3.5. Copyright © 2011 Published by Elsevier Ltd. All rights reserved.