Development of a multi-scale model for the adsorption of molecular target substances to magnetic carrier particles using the Lattice-Boltzmann methods

生物技术被认为是生产药品、特种化学品和食品的关键技术。许多药物产品是通过在基因修饰的细菌或细胞中表达来生产的。因此，从发酵液中分离所需产物起着重要的作用。分离的过程是困难的。除了靶蛋白外，还有不同的污染蛋白，它们的行为相似，只能通过利用等电点或特异性结合位点差异的高选择性方法进行分离。具有功能化表面的磁性颗粒是用于分离的有效选择。目前还没有详细的研究吸附过程及其对整个分离过程的影响，特别是对最大体积流量。本项目的目的是研究磁性功能化粒子对分子目标物质的吸附和分离过程。不同数量级的不同物理现象（吸附、种间效应和从流体中分离）对建模和模拟都具有挑战性。该过程通常由微观和宏观模型分别描述，并通过边界条件耦合。这些关系是复杂的，并没有在一个单一的方法来描述。由于计算成本高，模拟仅限于局部小区域和部分子过程。在所提出的项目中，物理现象是通过使用格子玻尔兹曼方法（LBM）进行模拟的介观建模联系起来的。因此，将要开发的模型将分子现象与宏观流体效应以及它们对毫米尺度粒子运动的影响联系起来。一种常见的方法是微观描述的颗粒动力学，离散的离散元方法，和一个宏观模型，离散的有限体积法。新方法允许更有效的模拟，并因此允许更大几何尺度的模拟，包括整个分离装置。通过模拟静态混合器中的吸附过程来进行校准和验证的建模方法。采用灵敏度法对模型进行了简化，并在现代磁选机上进行了模拟和实验验证。最后，欧拉-欧拉-欧拉模型是可转移的，并且可以用于其他多尺度问题，例如将来的光生物反应器的模拟。

Biotechnology is considered a key technology for the production of pharmaceuticals, specialty chemicals and food. Many pharmaceutical products are produced by expression in genetically modified bacteria or cells. Thereby, the separation of the desired products from fermentation broth plays an important role. The separation process is difficult. Besides the target protein, there are different contaminating proteins which behave similarly and which are separated only by means of highly selective methods taking advantage of differences in isoelectric point or specific binding sites. Magnetic particles with functionalized surface are an efficient option for the separation. Currently there are no detailed investigations of the adsorption process and its influence on the whole separation process, specifically on the maximum volume flow. However, they are necessary to develop efficient procedures for industrial use.The aim of the project is to investigate the process of adsorption and separation of molecular target substances on magnetic functionalized particles. The different physical phenomena in different orders of magnitude (adsorption, inter-particular effects and separation from fluid) are challenging concerning modeling as well as simulation. The process is usually described separately by micro- and macroscopic models and coupled via boundary conditions. The relationships are complex and have not been described in a single approach. Due to high computational costs simulations were limited both locally to small areas and partially to sub-processes. In the proposed project, the physical phenomena are linked by mesoscopic modeling using Lattice Boltzmann methods (LBM) for simulation. Thus, the model to be developed links molecular phenomenon in Ångström scale with macroscopic fluid effects as well as their influence on the particle motion in the mm scale. A common approach is the microscopic description of particle dynamics, discretized by a Discrete Element Method, and a macroscopic model, discretized by a Finite Volume Method. The new approach allows a more efficient simulation and thus the simulation of a larger geometric scale, including the entire separation apparatus. The modeling approach of the calibration and validation takes place by simulation of adsorption processes in a static mixer. Then the model is reduced with sensitivity-based approaches, the process is simulated on a modern magnetic separator and experimentally validated. To final Euler-Euler-Euler model is transferable and may be used for other multi-scale issues like e.g. the simulation of photo- bioreactors in the future.