Fructo-oligosaccharides separation and purification by simulated moving bed chromatography

The interest on oligosaccharides such as fructo-oligosaccharides (FOS) has strongly increased in recent years for food and pharmaceutical applications, mainly due to their improved technological and functional properties. FOS can be produced by fermentative processes from sucrose, and can be found in mixture with other mono- and di-saccharides and salts, at the end of the process [1]. Unlike FOS, the small saccharides (SGF), namely fructose, glucose and sucrose in the mixture, are known to be cariogenic, caloric and do not present prebiotic activity. The purification of FOS from the other sugars can represent and important increment on the economic value of the final product, which can be further used in diabetic and dietetic food [2]. Different strategies have been developed for this purpose, including microbial treatment [3], ultra and nano-filtration, activated charcoal systems [4], or ion-exchange chromatography [5]. Ion exchange resins may be then used in batch or continuous chromatographic processes, as Simulated Moving Bed (SMB) chromatography, to purify sugars. A screening of different commercial resins was previously done in order to select the most suitable to separate the oligosaccharides [5]. The resin Diaion 535Ca showed an increased recovery yield and purity of FOS (92 and 90%, respectively). In the present work, the separation process was implemented in the SMB, using the selected resin, namely. Equilibrium adsorption isotherms were determined by the Retention Time Method (RTM), for each single sugar. The resin was afterwards packed in eight SMB columns, and tested in the pilot plant. Different operation parameters, including switching time, extra time, internal flow-rates and operating pump flow-rates for feed, raffinate, desorbent, eluent and recycling streams, were tested in the plant. The separation of fructose from glucose and FOS from the SGF was evaluated. Firstly, the separation of a binary sugar mixture of fructose/sucrose (~ 50/50%) was performed followed by the separation of FOS from a fermentative broth. Fructose was purified from 53 to 76% and sucrose from 47 to 77%. FOS and SGF were purified from 50 to 67%. The implementation of UV detectors between the SMB columns allowed following the sugar concentration profile online during the separation process. The accurate selection of the operating parameters was made using the concentration signal obtained and showed to be a crucial step for an improved separation