Search

In this work, the Outlet Stream Swing (OSS) non-conventional operation technique is compared with a classical Simulated Moving Bed (SMB) unit, for the separation of a racemic mixture of guaifenesin. A methodic approach, based on the complete separation and regeneration regions, is used to evaluate the performance of the OSS extract-raffinate, as well as the OSS raffinate-extract, operating strategies with the classical SMB in terms of purity, productivity and eluent consumption requirements. The variants of the OSS technique: OSS extract-0 and OSS raffinate-0 are presented, simulated, and used not only to introduce more simplistic OSS modes of operation, but also to better understand the influence of specific OSS operating parameters. The experimental validation of OSS extract-0, operated by means of the FlexSMB-LSRE® unit, is also presented as part of this work, demonstrating the potential, as limitations, of such operating techniques. [ABSTRACT FROM AUTHOR]

The design of a simulated moving bed (SMB) process relies on valid thermodynamic predictions of multicomponent adsorption built up from accurate binary adsorption equilibrium data. Experimental adsorption equilibria of binary mixtures constituted by propane, propylene, isobutane and 1-butene on 13X zeolite were determined using breakthrough experiments at 373 K and 150 kPa. In addition, these binary adsorption experiments allow to confirm the choice of isobutane as an interesting desorbent for the separation of propane-propylene by SMB, since it has an intermediate selectivity between the two species to separate. Various prediction models are available in the literature but only a few of them have both physical and thermodynamical consistency. The ideal adsorbed solution theory (IAST), the thermodynamically consistent extended Toth model (TCET), and the physically-consistent extended Toth isotherm (PCET) were used to predict binary adsorption equilibria from pure component adsorption isotherms parameters. The PCET model was found suitable for representing the adsorption equilibrium of the different hydrocarbon mixtures with a reasonably good accuracy. [ABSTRACT FROM AUTHOR]

Monoclonal antibodies (mAbs) are among the most important therapeutic proteins for the treatment of cancer. They also find wide application in the field of diagnostics. The market for mAbs was US$ 5.4 billion in 2002 and is expected to triple by 2010 (1). MAbs are mainly produced in cell culture and purified by chromatography. Apart from the mAb, the downstream processing is confronted with a multitude of impurities from the upstream process. Some of these impurities bind strongly to the chromatographic stationary phase and can only be removed by separate cleaning solutions. Depending on the type of media used, e.g. affinity or ion exchange, differing cleaning regimes may have to be applied. In addition, sanitization is required to prevent microbial contamination. Both steps are closely tied to the mAb production step and need to be repeated in regular intervals. In this work, the impact of irreversible impurity adsorption on the separation performance is investigated experimentally; first for a batch column and then for a continuous chromatographic process. At first, the retention time of a model substance is established as a suitable measure to describe the degree of irreversible adsorption. The impact of the presence and the absence of cleaning-in-place (CIP) on the retention time of the model substance is demonstrated. Since the issue of column cleaning is particularly important if processes are operated in continuous mode, in the second part of this work the introduction of a CIP step in the continuous multicolumn countercurrent solvent gradient purification (MCSGP) process is investigated. The process purifies a mAb from a cell culture supernatant that also contains irreversibly adsorbing impurities—or impurities that desorb in a later cycle of the process as a contamination. Following the same approach as for the batch experiments, the MCSGP process is operated in the presence and the absence of CIP in order to track prospective changes of the process performance caused by changes in the product retention time. Instead of the model substance, a cell culture supernatant, providing a real case impurity profile in an industrial application is used for the process. Although an impact of the product retention time change on the process yield is not observed it is shown that CIP is required for a stable long-term operation in terms of purity and system pressure drop. With the introduction of a CIP step the process was successfully operated for 9000 min (6 days) without interruption. [ABSTRACT FROM AUTHOR]

Chromatography and SMB processes were studied to isolate only the intermediate retained component from ternary mixtures. The separation of ternary phenol and amino acid mixtures was simulated with Aspen Chromatography®. Simulation results for the phenol system with 10 µm packing showed that recycled center-cut chromatography usually produced the highest purity compared to cascades with two 4-zone SMBs and a 2-zone SMB/chromatography. For the amino acid mixture at ΔPmax=1.0 bar and purity-recovery index=95%, the recycled cascade of two 4-zone SMBs had approximately a 25% larger productivity compared to the simpler recycled center-cut chromatography. [ABSTRACT FROM AUTHOR]

A gradient SMB model is used to analyze the separation performance of proteins in a step-wise salt gradient ion-exchange simulated moving bed (IE-SMB) for the nonlinear adsorption equilibrium isotherms. The strategy of the selection of salt gradient in IE-SMB chromatography is discussed based on the adsorption equilibrium isotherms of both BSA and myoglobin on Q-Sepharose FF anion exchanger in the pH 8 Tris buffer. The separation region is used to select the flow rates in the salt gradient IE-SMB chromatography for the separation of proteins. The separation region is calculated by the gradient SMB model where both the mass transfer resistance in the ion exchangers and axial liquid dispersion in columns are taken into account. Some strategies are suggested for the selections of salt gradient and flow rates in the IE-SMB chromatography with open loop configuration for two cases: (a) binary separation of proteins with the requirement of the complete recovery from extract stream and raffinate stream; and (b) protein purification from a stream with some impurities. [ABSTRACT FROM AUTHOR]

A new simulated moving bed (SMB) operation technique is introduced, the outlet streams swing (OSS), based on the dynamic collection of fronts from the equivalent true moving bed (TMB). This new assumption improves the product outlet purities, when compared with the classic operating mode, as a consequence of an increased separation between collection fronts and the contaminating ones. The strategy is dynamic and discontinuous, a collection front expansion is performed in zone 1 and/or 4, obtained by the violation of the classic respective design constraints during a first step and contraction during the second step using at least the classic design operating parameter for zones 1 and 4. Another non-conventional technique, the MultiFeed (MF) SMB operation, is simulated and analyzed here as a solo technique as well as a powerful modelling technique for the particular case of inner sections Varicol® (only the zones 2 and 3 have variable number of columns). [ABSTRACT FROM AUTHOR]

Analogs, recycled chromatography systems that mimic SMBs, for both a cascade of two four‐zone SMBs and five‐zone SMBs performing ternary separations are designed and analyzed. The basic Analog consists of one column and several mixed tanks equal to the number of columns in the mimicked SMB. Analogs with mixed tanks give lower purities than the SMBs because of mixing in the tanks. For example, the five‐zone SMB with one column per zone achieved a purity index (PI) of 88.3, while the Analog with five mixed tanks has a PI of 83.4. If the tanks are divided into smaller tanks the purity improves, and for the Analog to be the five‐zone SMB using 10 tanks the PI was 85.4. Analogs with a single unmixed reservoir were also designed for a cascade of two four‐zone SMBs with one column per zone. This reservoir reduces the mixing and improves the separation. For example, a cascade of two four‐SMBs with one column per zone achieved a PI of 86.7; the Analog to this SMB with 8 mixed tanks achieved a PI of 82.9, while the Analog with an unmixed reservoir achieved a PI of 84.6. [ABSTRACT FROM AUTHOR]