The ultrafiltration/diafiltration (UF/DF) process is a crucial step in the canola protein isolation process from rapeseed meal. The process involves using a multi-stage membrane system to separate components of the mixture. As diafiltration dilutes the feed stream in the ultrafiltration system, a large amount of diafiltration water is required. Reducing the diafiltrate for sustainability reasons can be done by carefully selecting process variables or using recycle streams. However, finding the optimum process variables can be a meticulous process if performed experimentally or via trial and error. In this study, we performed an optimization using a mechanistic model integrated with a genetic algorithm to aid in finding an optimum combination of process variables. The mechanistic ultrafiltration model was derived by taking into account transport phenomena within the filtration system. Parameters were characterized experimentally in term of viscosity coefficient, membrane resistance, cake porosity, aggregate diameter equivalent, and material compressibility factors. Using the mechanistic model-based optimization in combination with actual experimental values, the performance of a four-stage UF/DF system could already be improved despite fixing the configuration, albeit at the cost of a reduction in purity. Further improvement could be achieved by using recycle streams. The optimized system achieved a diafiltrate reduction of up to 79%, an increase of purity of up to 31%, and an increase of dry matter content of up to 18%, while maintaining the product purity of the reference set-up.

Finding the best purification process is a challenging task. Recently, mechanistic models that can accelerate the development of chromatographic unit operations, the most important purification units, became widely available. In previous work, several chromatographic models have been linked together to simulate and optimize integrated processes. However, considering only chromatographic steps may lead to a suboptimal process. Consequently, the aim of this study was to include models for ultra- and diafiltration units into the optimization approach to account for buffer exchange steps before or between chromatography units. This approach was applied to an industrial case, the purification of a monoclonal antibody, where cation exchange, hydrophobic interaction and mixed mode were the possible chromatographic separation modes. It turned out that only the duration of the total filtration step and the duration of the ultrafiltration step were crucial variables for the optimization of the ultra- and diafiltration steps. The ‘best’ in silico purification process was found based on the performance criteria yield and solvent usage. The purity was required to be at least 99.9%.

The application of mechanistic models for chromatography requires accurate model parameters. Especially for complex feedstocks such as a clarified cell harvest, this can still be an obstacle limiting the use of mechanistic models. Another commonly encountered obstacle is a limited amount of sample material and time to determine all needed parameters. Therefore, this study aimed at implementing an approach on a robotic liquid handling system that starts directly with a complex feedstock containing a monoclonal antibody. The approach was tested by comparing independent experimental data sets with predictions generated by the mechanistic model using all parameters determined in this study. An excellent agreement between prediction and experimental data was found verifying the approach. Thus, it can be concluded that RoboColumns with a bed volume of 200 μL can well be used to determine isotherm parameters for predictions of larger scale columns. Overall, this approach offers a new way to determine crucial model input parameters for mechanistic modelling of chromatography for complex biological feedstocks.

Downstream process development is a major area of importance within the field of bioengineering. During the design of such a downstream process, important decisions have to be made regarding the type of unit operations as well as their sequence and their operating conditions. Current computational approaches addressing these issues either show a high level of simplification or struggle with computational speed. Therefore, this article presents a new approach that combines detailed mechanistic models and speed-enhancing artificial neural networks. This approach was able to simultaneously optimize a process with three different chromatographic columns toward yield with a minimum purity of 99.9%. The addition of artificial neural networks greatly accelerated this optimization. Due to high computational speed, the approach is easily extendable to include more unit operations. Therefore, it can be of great help in the acceleration of downstream process development.