Shape and topology optimization of enzymatic microreactors

Structural optimization methods have been used by mechanical and civil engineers over the yearsto find the optimal structures. Structural optimization is a series of computational techniqueswhich include shape and topology optimization. Shape optimization is directly applied to theboundaries of a structure and results in the deformation of the configuration. Topologyoptimization contributes to the improvement of the layout of the material in a domain. Themechanical performance of a structure is evaluated by an objective function which can be for example maximizing its stiffness.The need for effective and cost efficient reactors for pharmaceutical processes forces the industry to search for better technologies. In biochemical engineering, the used reactor design in a given process is usually limited to a range of well-established configurations and layouts. Usually the implemented reactors in a chemical process do not always yield in the best reaction conditions.This thesis develops an innovative application of topology and shape optimization methods to achemical engineering problem. The main goal is to design a reactor according to the limitations of the reaction system by modifying the reactor configuration. In this thesis structural optimization methods were exclusively applied to enzymatic microreactors. The case studies were chosen such that they can be experimentally tested afterwards. In this way, the design of the reactor is customized to the reaction system and itcontributes to the reduction of extensive experimental work to find the best reactor configuration.Shape optimization has been applied to an YY-microreactor with a rectangular cross-section withthe intention to investigate the shape influence on the active mixing of substances and consequently in the reaction yield. The inlet and the outlet are located at the respective ends of the reactor. Both inlet and outlet have a Y shape where two streams meet at the entrance of the reaction chamber and two streams are split again at the exit. The optimization routine focuses onthe modification of the microreactor shape parameters such as height and width. This is achieved by a computational fluid dynamic (CFD) simulation study, which investigates a biocatalyticreaction for the production of optically pure chiral amines in the reactor system. The routine implements kinetic models into a CFD framework (ANSYS CFX®), which is coupled with a selfprogrammedMATLAB® code. ANSYS CFX® performs the discretization of the microreactorinto finite volume elements and calculates the main reactor outputs. The MATLAB® routine performs the optimization by changing the geometry. Furthermore, it includes the evaluation of the objective function, the new definition and execution of the next simulation for each new microreactor shape. Afterwards, the performance of the system is evaluated by comparing theobjective function (reaction yield) with the previous best configuration. If the geometry changesiiresult in a better reaction yield, this geometry is selected as the best and the old configuration is discarded. The optimization routine continues until a constraint is fulfilled or the optimization converges. The changes of the geometry are performed by a gradient-free method named random search. The random search modifies the design variables by sampling in an arbitrary manner froma vector which sets the variation limits. Subsequently, the same coupled routine between ANSYS CFX® and MATLAB® is applied to topology optimization. The method was used as a novel technique to computationally discover the best spatial distribution of an enzyme inside microreactors. Usually, the enzyme is uniformly distributed inside a reactor, which can mean either at a wall surface or in a packed bed reactor or free in solution. Therefore, these three applications are studied.The aim is to improve the product formation per same amount of enzyme in the reactor. The Evolutionary Structural Optimization (ESO) method is adapted to perform the optimization. The ESO method removes inefficient elements from a structure by a gradual and iterative procedureaccording to a rejection criterion which determines the elements that should be removed everyiteration.The MATLAB® routine is featuring the adaptation of the ESO method to the biocatalytic reactor. The two-dimensional topology optimization is applied to a microreactor with immobilized enzyme on the wall surface. The selected reactor geometry is an adaptation of a previously scientific documented shape used in topology optimization of microreactors. The threedimensional topology is computationally applied to the distribution of enzyme in a miniaturized packed bed reactor as well as to a microreactor with free enzyme in the volume.In the last part of the thesis, the topology of microreactors is the experimentally studied. This is achieved by using the peroxidase-catalyzed oxidation of 2,2’-azino-bis(3-ethylbenzthiazoline-6-sulfonic acid) (ABTS) to its radical form by reduction of hydrogen peroxide. The determination of the kinetic mechanism is required in order to validate the optimized microreactors. Two microreactor shapes are topology optimized for posterior experimental validation. The first shape corresponds to the shape with immobilized peroxidase on the wall surface. The experimental validation was attempted by using a photochemical reaction. The reaction attaches linkage molecules to a masked surface, which has an immobilization pattern. The linkage molecules will thereafter react with the enzyme molecules binding them covalently to the surface.The second microreactor configuration corresponds to a square shaped cross section microchannel with free enzyme in solution. For this case study, a well-mixed solution of enzymeand substrate is considered to enter the microreactor. The experimental comparison is performed by comparing an improved inlet configuration with a reference system. The configurations were selected and fabricated as a compromise considering
the outcome of the topology optimization and the limitations of the fabrication process.