Optimal economic design and operation of single- and multi-column chromatographic processes

Single-column chromatography is widely used in the biopharmaceutical industries, although multi-column alternatives in the form of simulated moving bed (SMB) processes are now emerging. It may be difficult, however, to determine which column alternative will be best suited for a given application, and this work sets out to address this issue. A systematic approach is presented that is based on a full economic appraisal of each process alternative based on an optimization of the net annual profit. Single-column processes with and without recycling are considered, as are both the SMB and the Varicol process. The cyclic steady state for the SMB and Varicol processes is determined directly by complete discretization. The approach is applied to a case study based on a linear isotherm where it is found that for this particular system, a recycling policy is not necessary for the single column. When comparing the single-column process with the multi-column alternatives, the single column is the most economical provided the life time of the project is short; however, the economic benefits of the more capital-intensive multi-column processes are greater if the life time of the project is over 5 years. The SMB process is found to perform marginally better than the Varicol process over 15 years; however, this may be because not all extra degrees of freedom for the Varicol process were considered.     

Modified reactive SMB for production of high concentrated fructose syrup by isomerization of glucose to fructose

This work presents modifications to the Hashimoto's hybrid simulated moving bed reactor (SMBR) system which was used to produce 55% high fructose syrup (HFS55). The purpose of this study is to develop a new SMBR system to overcome the disadvantages of Hashimoto system (3-zone SMB with seven reactors), i.e., low utility of reactors when feed being a 50/50 blend of glucose and fructose. Two different configurations of modified system were presented in this paper: the first configuration is 4-zone SMB with one reactor, while the other one consists of one additional reactor. Both of these configurations aim at improving the concentration and purity of glucose at the inlet of the reactor, which will lead to both high productivity and high purity of fructose in the product. A state-of-the-art optimization technique, viz., non-dominated sorting genetic algorithm (NSGA) is used in finding the optimal design and operating parameters for the modified reactive SMB and Varicol processes. Compared with the Hashimoto's system, high productivity and purity of fructose can be achieved in these new systems using less number of reactors.     

Optimal design and experimental validation of a modified four-zone SMB process for the separation of a ternary amino acid mixture

For a standard four-zone simulated moving bed (SMB) chromatography, several process modifications have been made in previous studies such that its application scope could be extended to a ternary separation. One of the effective modifications reported was to (1) replace its closed-loop configuration with an open-loop configuration and (2) utilize the extract port for collecting both the intermediate-affinity and the highest-affinity components in regular sequence in every switching period. Most of previous researches on such a modified four-zone SMB (MF-SMB) have been limited to process simulation and optimization. The experimental validation of the MF-SMB process with linear isotherms was attempted in this article for the first time using a ternary amino acid mixture as a model system. First, the intrinsic parameters of three amino acids were estimated from a series of multiple-frontal experiments. The estimated parameter values were then used in the stage of the MF-SMB optimization, which was assisted by an up-to-date genetic algorithm. Based on the optimized conditions, the MF-SMB experiment was conducted and the assay results for product samples verified the attainment of a ternary separation. The experimental purities and concentrations were also found to agree closely with the model predictions.     

Development of a three-zone simulated moving bed process based on partial-discard strategy for continuous separation of valine from isoleucine with high purity,high yield,and high product concentration

The issue of separating valine from isoleucine has been a major concern in the biotechnological process for production of valine. To address this issue, an optimal three-zone simulated moving bed (SMB) process for continuous separation of valine was developed in this study. It was first found that an Amberchrom-CG161C resin was highly suitable for the adsorbent of such SMB process. The adsorption isotherm and mass-transfer parameters of valine and isoleucine on the Amberchrom-CG161C adsorbent were then determined through multiple frontal experiments. The determined parameters were used in the next stage of optimizing the SMB for valine separation, which was performed on the basis of genetic algorithm. For the optimized SMB process, a partial-discard strategy was applied to the raffinate port in order to make a further improvement in the valine product concentration. Finally, the optimized SMB based on the partial-discard strategy was tested experimentally using the self-assembled SMB equipment. The experimental results showed that the developed process in this study was highly effective in continuous separation of valine from isoleucine while ensuring the attainment of high product concentration. The experimental data for the SMB effluent histories and the SMB column profiles were also in close agreement with the model predictions.     

Optimal design and experimental validation of a three-zone simulated moving bed process based on the Amberchrom-CG161C adsorbent for continuous removal of acetic acid from biomass hydrolyzate

In the production process of bio-ethanol from biomass, acetic acid is recognized as the key impurity to be removed from the sugar components that are generated by hydrolyzing biomass. In regard to this issue, it has recently been confirmed that the Amberchrom-CG161C resin was highly qualified as the adsorbent of a simulated moving bed (SMB) process for continuous separation of acetic acid from the biomass hydrolyzate, i.e., sugars. However, the previous study on the Amberchrom-CG161C SMB with the aforementioned separation goal has been limited to only a theoretical work, including some batch-chromatography tests. The experimental validation of such an Amberchrom-CG161C SMB process, including its optimal design, was attempted in this article. This task began by assembling the experimental unit of the SMB process with three zones. Its operating conditions were then optimized by using genetic algorithm. Under the optimized operating conditions, the relevant three-zone SMB experiment was conducted. The assay of all the resultant product samples verified that the SMB separation of interest was performed successfully as designed. The experimental data were also found to agree closely with the model predictions. Finally, a partial-discard strategy was applied to maintain the sugar product concentration as high as possible.     

Continuous separation of succinic acid and lactic acid by using a three-zone simulated moving bed process packed with Amberchrom-CG300C

The issue of separating succinic acid and lactic acid in a continuous mode has been a major concern in the biotechnological process for production of succinic acid. To address this issue, both the optimal design and the experimental validation of a three-zone simulated moving bed (SMB) process for such separation were attempted in this article using the Amberchrom-CG300C resin and a self-assembled SMB unit with three zones. First, the intrinsic parameters of the two organic acids on the Amberchrom-CG300 resin were estimated at 40 °C from a series of multiple frontal experiments. The resulting intrinsic parameters were then used in optimizing the experimental setting points for pump flow rates and switching time of the three-zone SMB equipment, which was assisted by an up-to-date genetic algorithm. Based on the optimized conditions, the relevant SMB experiment was conducted at 40 °C and all the resultant samples from the product ports and column outlets of the SMB unit were assayed. It was confirmed from the assay results that the continuous separation of succinic acid and lactic acid was performed successfully. The experimental data for the product concentration profiles and the internal concentration profiles were also in reasonable agreement with the model predictions.     

Experimental validation of the solvent-gradient simulated moving bed process for optimal separation of phenylalanine and tryptophan

A solvent-gradient simulated moving bed (SG-SMB) process for separation of two useful amino acids, phenylalanine and tryptophan, has been studied in previous researches, which confirmed that the SG-SMB outperformed the corresponding isocratic SMB for the same separation task. However, all of the previous studies on the SG-SMB for separation of the two amino acids have been limited to process simulation and optimization, including some batch-chromatography tests. The experimental validation of such an SG-SMB process was attempted in this article. This task began by assembling the experimental unit of the SG-SMB process. Its operating conditions were determined from the SG-SMB optimization tool based on genetic algorithm, in which the mass-transfer parameters as well as the adsorption isotherm parameters of the feed components (two amino acids) were entered as the mathematical models expressed as a function of liquid-phase composition (i.e., modifier concentration in the liquid phase). Based on the determined operating conditions, the relevant SG-SMB experiment was conducted until a cyclic steady-state was reached. The assay of all the resultant product samples verified that the SG-SMB separation of interest was performed successfully as designed. The experimental data were also found to agree closely with the model predictions.     

Genetic algorithm with local optimization

A hybrid of genetic algorithm and local optimization was tested on a massively multimodal spin-lattice problem involving a huge configuration space. The results are good, and global optima will probably be achieved in a sizeable proportion of cases, especially if a selection scheme is applied that maintains genetic diversity by introducing a spatial separation between the members of the population. If we use single-point cross-over, the performance of the algorithm depends strongly on the order of the units corresponding to individual spins in the bit strings that the genetic part of the algorithm processes. Due to some interplay between the genetic algorithm and local optimization, the best performance is achieved with a peculiar ordering, while the results with the most obvious ordering are much worse. I introduce an ordering-invariant crossover operation that gives excellent performance: it almost always yields states of the lowest energy. I expect this or some similar crossover operation to work well in the hybrid scheme for many other problems as well.     

Optimal design of a tandem simulated moving bed process for separation of paclitaxel, 13-dehydroxybaccatin III,and 10-deacetylpaclitaxel

Paclitaxel, 13-dehydroxybaccatin III (13-DHBIII), and 10-deacetylpaclitaxel (10-DAP), which came from the plant cell culture, have been regarded as highly valuable because of their efficacy in the anticancer treatments. Due to these values, the necessity for separating the three components in an economical way has been a matter of grave concern in industry. In this study, a tandem simulated moving bed (SMB) process that consisted of two four-zone SMB units in series was applied to such a ternary separation. First, a series of pulse injection experiments were performed for estimation of the adsorption isotherm and mass-transfer parameters. The estimated parameters were utilized in the tandem SMB optimization tool that was prepared based on the standing wave design principle. During the optimization of interest, the throughput of the tandem SMB was maximized while meeting the requirements on product purities and pressure drop. The results proved that the most economical strategy of utilizing the tandem SMB was to recover paclitaxel in the first SMB unit and then separate the remaining two components (13-DHBIII and 10-DAP) in the second SMB unit. Furthermore, such a strategy was also found to result in better tandem SMB performances over the entire region of a pressure drop limit.     

Comparison of Amberchrom-CG161C and Dowex99 as the adsorbent of a four-zone simulated moving bed process for removal of acetic acid from biomass hydrolyzate

One of the important steps in the application of biomass to producing sugars, which can be converted into bio-ethanol and other valuable chemicals by fermentation, is to hydrolyze the biomass components by sulfuric acid. It was reported that such a hydrolysis entailed the generation of acetic acid, which has been recognized as a key impurity to be surely removed from the biomass hydrolyzate for ensuring high fermentability of the hydrolyzed sugars. Regarding such a removal task, there has been a previous application of a simulated moving bed (SMB) process based on the Dowex99 adsorbent, whose performance, however, was limited by low selectivity between acetic acid and sugars. To overcome such a limitation, another adsorbent alternative to Dowex99 was searched in this study. It was found that Amberchrom-CG161C allowed higher selectivity between acetic acid and sugars than Dowex99. To investigate the relative superiority of Amberchrom-CG161C over Dowex99 as the adsorbent of an SMB process for removing acetic acid from the biomass hydrolyzate, the two SMB processes based on Amberchrom-CG161C and Dowex99 were optimized using the SMB optimization tool based on standing wave design (SWD) method. The optimization results revealed that the Amberchrom-CG161C SMB outperformed the Dowex99 SMB by a wide margin.     

The application of preparative batch HPLC, supercritical fluid chromatography, steady-state recycling, and simulated moving bed for the resolution of a racemic pharmaceutical intermediate

This article discusses the chromatographic resolution of a racemic pharmaceutical intermediate. Preparative batch high performance liquid chromatography (HPLC), supercritical fluid chromatography (SFC), steady-state recycling (SSR), and simulated moving bed (SMB) were used to resolve a total of 12.2 kg of a racemic pharmaceutical intermediate. In this study, a first batch of 0.8 kg of racemate was separated on the preparative batch HPLC and SFC, and subsequently another 5.9 kg of racemate was separated on the SSR. Lastly, a third batch of 5.5 kg was separated on the SMB. The separation conditions and results of these techniques are discussed. The productivities and solvent costs of SFC versus HPLC are compared. The productivities and solvent costs of SMB, SSR, and HPLC are also compared. The analytical method development and process optimization of these processes are also discussed in this article.     

An integrated approach to optimization of Escherichia coli fermentations using historical data

Using a fermentation database for Escherichia coli producing green fluorescent protein (GFP), we have implemented a novel three-step optimization method to identify the process input variables most important in modeling the fermentation, as well as the values of those critical input variables that result in an increase in the desired output. In the first step of this algorithm, we use either decision-tree analysis (DTA) or information theoretic subset selection (ITSS) as a database mining technique to identify which process input variables best classify each of the process outputs (maximum cell concentration, maximum product concentration, and productivity) monitored in the experimental fermentations. The second step of the optimization method is to train an artificial neural network (ANN) model of the process input-output data, using the critical inputs identified in the first step. Finally, a hybrid genetic algorithm (hybrid GA), which includes both gradient and stochastic search methods, is used to identify the maximum output modeled by the ANN and the values of the input conditions that result in that maximum. The results of the database mining techniques are compared, both in terms of the inputs selected and the subsequent ANN performance. For the E. coli process used in this study, we identified 6 inputs from the original 13 that resulted in an ANN that best modeled the GFP fluorescence outputs of an independent test set. Values of the six inputs that resulted in a modeled maximum fluorescence were identified by applying a hybrid GA to the ANN model developed. When these conditions were tested in laboratory fermentors, an actual maximum fluorescence of 2.16E6 AU was obtained. The previous high value of fluorescence that was observed was 1.51E6 AU. Thus, this input condition set that was suggested by implementing the proposed optimization scheme on the available historical database increased the maximum fluorescence by 55%.     

Simulated moving bed chromatography with supercritical fluids for the resolution of bi-naphthol enantiomers and phytol isomers

The combination of the simulated moving bed (SMB) technique with supercritical fluid chromatography (SFC) leads to a process with unique features. Besides the known advantages of the SMB process, the use of supercritical carbon dioxide as the mobile phase offers the advantages of reduction in organic solvents and an easy eluent/solute separation. Because of the low viscosity and high diffusion coefficients of supercritical fluids, a high efficiency is possible. The steps of process development for SMB SFC are presented using the separations of the bi-naphthol enantiomers and phytol isomers as examples. The development of a packed column SFC method at an analytical scale is shown for the separation of the bi-naphthol enantiomers on a chiral stationary phase and CO(2) with a modifier as the mobile phase. The influence of the modifier, modifier content, and column configuration on productivity of the SMB SFC process was investigated by simulation. The first set of experiments was performed in the SMB separation of phytol isomers at low concentration to test the feasibility of the SMB SFC high purity separation of the binary mixtures. In the second set of experiments, the productivity of the process was increased by increasing the feed concentration up to 54 grams feed per liter stationary phase (SP) and hour (g(feed)/l(SP) h).     

Improving the performance of nadolol stereoisomers' preparative separation using Chiralpak IA by SMB chromatography

The pseudobinary preparative separation of nadolol stereoisomers is performed by simulated moving bed chromatography (SMB). Using the Chiralpak IA adsorbent, a new 25:75:0.1 (v/v/v) methanol‐acetonitrile‐diethylamine solvent composition was selected to perform the experimental SMB separation and compare it with the previous results obtained using pure methanol. Using a 2 g L^?1 total feed concentration of an equimolar mixture of the four stereoisomers of nadolol, the more retained component was fully recovered (100% purity and 100% recovery), with a system productivity of 0.77 g L^?1 hour^?1 and a solvent consumption of 9.62 L g^?1. Comparing these results with the ones previously reported using 100:0.1 methanol‐diethylamine solvent composition, this work shows that the 25:75:0.1 methanol‐acetonitrile‐diethylamine is a better alternative for the preparative separation of nadolol stereoisomers by SMB chromatography. These results are confirmed by simulation of the SMB operation for higher feed concentrations, by comparing the performances of the two solvent compositions using the data obtained experimentally through the measurement of the adsorption equilibrium isotherms and the kinetic data obtained for both solvents. The new experimental and simulation results stress out that the performance of the preparative separation can be improved by a careful selection of the solvent composition.     

Comparing the performance of one-column process and four-zone simulated moving bed by computer simulation

A new one-column chromatography process, analogous to a four-zone simulated moving bed (SMB), was presented. The basic principle of the process was identical to that of a four-zone SMB. The process consisted of one chromatographic column and four tanks, instead of the four columns in the four-zone SMB (1-1-1-1), and has been used for the separation of two amino acids, phenylalanine and tryptophan, using an ion exchange resin. The operating parameters for the one-column process and four-zone SMB were obtained from equilibrium theory. Computer simulations were used to compare the performances of the new one column process to that of the general four-zone SMB, using Aspen Chromatography^™ v 11.1. The differences between the one-column and SMB processes in terms of the purities and yields of phenylalanine and tryptophan were less than 4 and about 6%, respectively. The lower purities of the one-column process were due to the loss of the developed concentration profiles in the column when the liquid was stored in tanks. The one-column process gave great flexibility, and would be useful for reconstructing an existing conventional chromatography process to one of a SMB.     

Adaptive optimization of fed-batch culture of yeast by using genetic algorithms

A simulation and experimental study has been carried out on the adaptive optimization of fed-batch culture of yeast. In the simulation study, three genetic algorithms based on different optimization strategies were developed. The performance of those three algorithms were compared with one another and with that of a variational calculus approach. The one that showed the best performance was selected to be used in the subsequent experimental study. To confer an adaptability, an online adaptation (or model update) algorithm was developed and incorporated into the selected optimization algorithm. The resulting adaptive algorithm was experimentally applied to fed-batch cultures of a recombinant yeast producing salmon calcitonin, to maximize the cell mass production. It followed the actual process quite well and gave a much higher value of performance index than the simple genetic algorithm with no adaptability.     

Integrated framework of system dynamics and meta-heuristic for multi-objective land use planning problem

Land use optimization as a resource allocation problem can be defined as the process of assigning different land uses to a region. Sustainable development also involves the exploitation of environmental resources, investment orientation, technology development, and industrial changes in a coordinated form. This paper studies the multi-objective sustainable land use planning problem and proposes an integrated framework, including simulation, forecasting, and optimization approaches for this problem. Land use optimization, a multifaceted process, requires complex decisions, including selection of land uses, forecasting land use allocation percentage, and assigning locations to land uses. The land use allocation percentage in the selected horizons is simulated and predicted by designing a System Dynamics (SD) model based on socio-economic variables. Furthermore, land use assignment is accomplished with a multi-objective integer programming model that is solved using augmented ε-constraint and non-dominated sorting genetic algorithm II (NSGA-II) methods. According to the results of the SD model, land use changes depend on population growth rate and labor productivity variables. Among the possible scenarios, a scenario focusing more on sustainable planning is chosen and the forecasting results of this scenario are used for optimal land use allocation. The computational results show that the augmented ε-constraint method cannot solve this problem even for medium sizes. The NSGA-II method not only solves the problem at large sizes over a reasonable time, but also generates good-quality solutions. NSGA-II showed better performance in metrics, including number of non-dominated Pareto solutions (NNPS), mean ideal distance (MID), and dispersion metric (DM). Integrated framework is implemented to allocate four types of land uses consisting of residential, commercial, industrial, and agricultural to a given region with 900 cells.

     

A novel design of simulated moving bed (SMB) chromatography for separation of ketoprofen enantiomer

A simulated moving bed (SMB) chromatography system is a powerful tool for preparative scale separation, which can be applied to the separation of chiral compound. We have designed our own lab-scale SMB chromatography using 5 HPLC pumps, 6 stainless steel columns and 4 multi-position valves, to separate a racemic mixture of ketoprofen in to its enantiomers. Our design has the characteristics of the low cost for assembly for the SMB chromatography and easy repair of the unit, which differs from the designs suggested by other investigators. It is possible for the flow path through each column to be independently changed by computer control, using 4 multi-position rotary valves and 5 HPLC solvent delivery pumps. In order to prove the operability of our SMB system, attempts were made to separate the (S)-ketoprofen enantiomer from a ketoprofen racemic mixture. The operating parameters of the SMB chromatography were calculated for ketoprofen separation from a batch chromatography experiment as well as by the triangle theory. With a feed concentration of 1 mg/mL, (S)-ketoprofen was obtained with a purity of 96% under the calculated operating conditions.     

The role of carbon flux and biometric observations in constraining a terrestrial ecosystem model: a case study in disturbed forests in East Asia

The process of confining unnecessary freedom is a step toward advanced ecosystem modeling. This study demonstrates the importance of carbon flux and biometric observation in constraining a terrestrial ecosystem model with a simple optimization scheme. At the selected sites from AsiaFlux network, a simultaneous optimization scheme for both carbon flux and biomass was compared with carbon flux-oriented and biomass-oriented optimization schemes using the Biome-BGC model. The optimization scheme oriented to either carbon flux or biomass provided simulation results that were consistent with observations, but with reduced performance in unconstrained variables. The simultaneous optimization scheme yielded results that were consistent with observations for both carbon flux and biomass. By comparing long-term projections simulated by three schemes, it was found that the optimization oriented only to carbon flux has limited performance because misrepresented biomass significantly affected a projection of carbon exchange through heterotrophic respiration. From these experiments, we found that (1) a process model like Biome-BGC is capable of reproducing both carbon flux and biomass within acceptable proximity, (2) constraining biomass is importance not just because it is one of carbon cycle components, but also it significantly affects simulations of carbon flux. Thus, it is important to invest more effort to improve simulation of biomass as well as carbon flux.     

Preparative scale enantioseparation of a chiral epoxide: Comparison of liquid chromatography and simulated moving bed adsorption technology

The feasibility of using simulated moving bed technology (SMB) for chiral separation on cellulose triacetate is demonstrated on the preparative scale: 1 kg of a chiral epoxide has been separated. On comparing SMB technology with conventional liquid chromatography it turns out that the main advantage of SMB lies in the significant reduction of mobile phase consumption. The process design for SMB is made theoretically and the predictions are confirmed by our pilot study. © 1993 Wiley-Liss, Inc.