| Popis: |
Separation and purification of functional ingredients from raw or waste streams are often done via processes that include a chromatographic step using a packed bed of resin particles that have affinity for the ingredients to be separated. A column packed with these particles presents numerous bottlenecks when dealing with untreated or large streams: a trade-off between mass transfer and hydraulic permeability, a high pressure drop and susceptibility to plugging and fouling. The large equipment (column diameters) and volume of resin needed for a moderate pressure drop and a high capacity, poses problems of elevated costs and complex operation. Other technologies such as radial flow chromatography and polymeric resins membranes may represent an improvement in other applications (e.g. pharma or fine chemicals), but at this point their capacity and costs do not seem to be feasible for the separation of small molecules from larger food streams. The aim of the research discussed in this thesis was to find the principles that determine the suitability of different structured adsorbents, such as monoliths, for the selective adsorption and recovery of high-added value food ingredients of relatively low molecular weight, such as oligosaccharides and bioactive peptides. To ensure a cost-effective process and high capacity for small molecules, we demonstrated the feasibility of using activated carbon, and compared its adsorptive and hydraulic performance in two different structures: porous particles and channeled monoliths (”honeycomb” structures). Furthermore, we assessed the feasibility and window of operation of monoliths in terms of adsorbent and column volume required, compared to packed beds. To demonstrate the isolation of bioactive peptides from crude mixtures with activated carbon, we used activated carbon to recover a lacto-tripeptide IPP from a commercial hydrolyzate (1.5% w/w) in chapter 2. The purity of the initial crude mixture was doubled in the isolate, to up to 35% with a recovery of IPP of about 80% in the first cycles of adsorption. This was repeated over many consecutive adsorption-desorption cycles until the activated carbon packed bed column was exhausted. This exhaustion was found to be caused not only by the occupation of irreversible sites but also by pore blockage. Finally, guidelines were given for the competitive exhaustion of the adsorbent for process optimization in order to obtain higher purity and yield. In chapter 3 we showed the benefits of using channeled monoliths for processing untreated streams. We compared the use of channeled monoliths with a packed bed, both made of the same type of activated carbon, for the adsorption of the lactotripeptide IPP from a crude hydrolyzate. The results showed similar productivity and dynamic adsorptive capacities at comparable linear velocities and residence times, but the packed bed showed a strong pressure drop increase during continuous loading of the column and the same consecutive adsorption-desorption cycles as studied in chapter 2. This indicates the occurrence of pore blockage and plugging of the column. These fouling mechanisms were confirmed with two semi-empirical model analogies: one analogous to membrane fouling and another using an analogy with a set of parallel channels. The strong pressure drop increase was even more noticeable at high velocities (and short residence times). These trends were not observed in the channeled monoliths: no significant pressure drop increase was found here, and high velocities were eminently feasible. In chapter 4 the adsorption of lactose onto a bed of activated carbon particles and activated carbon channeled monoliths was described with a detailed chromatographic model, taking into account the different mass transfer resistances. First, the single component adsorption isotherm parameters were obtained using frontal analysis on both adsorbents. Second, the kinetics of adsorption of lactose on both activated carbon adsorbents were estimated using the shallow bed method, assuming an infinite bath. The uptake curves were fitted to the homogeneous surface diffusion model and the linear driving force approximation. The estimation of the intraparticle diffusion coefficient and the film mass transfer coefficient showed a similar intraparticle mass transfer performance during the uptake adsorptive process. Fitting of the breakthrough data to the general rate model describing the full column operation showed differences in performance during the overall column operation. These differences could be related to higher axial dispersion in the squared channeled monoliths. The difference between the experimentally-derived axial dispersion and he expected assuming tubular coated tubes, suggested that the squared shape was responsible for the inhomogeneity of the flow. In chapter 5, we presented guidelines for the configuration of industrial scale chromatographic separation of small molecules. A window was identified that defines the feasible configurations to use for the highest productivity for a given set of process requirements. The performance of different axial packed beds, channeled monoliths and a continuous monolith assuming silica as base material were compared by means of HETP (height equivalent of theoretical plates) and pressure drop relations. The relations as a function of velocity were used to calculate the resultant velocity and packing length for different conditions (efficiency, pressure drop, affinity constant and throughput). The specific productivity of channeled monoliths was shown to be up to 2.5 orders of magnitude higher than that of a packed bed. Therefore, at large scales (in which the pressure drops need to be limited, and the flow rate is high), channeled monoliths are preferred since they may reduce the equipment size up to 100 times and the required adsorbent volume up to 1000 times. Finally, in chapter 6 we discussed the suitability of activated carbon regarding its re-usability and purification potential in the separation of small food ingredients. The suitability of channeled monoliths for certain applications was also highlighted. Finally, other suitable adsorbents were suggested, and some future prospects in the selection of adsorbents were given. |
