Microscopic and macroscopic investigations of the transient hydrodynamic memory behavior of chromatographic columns

在制备色谱柱的工业应用过程中，经常遇到不希望的现象，如介质壁脱离、部分床沉降、流动通道的形成和颗粒床的不可逆压实。它表明，在自己以前的工作中，这些现象是由填充过程所造成的颗粒床的不均匀性的结果。在第一个3年的资助期内，开发了一种新的微色谱柱，使流体动力学行为的详细实验检查。将计算流体力学（CFD）和离散元法（DEM）相结合，建立了一种新的确定性三维混合仿真模型。该模型能够模拟单个色谱颗粒之间以及颗粒与流体之间的相互作用。模拟结果与实验数据吻合较好。研究了色谱填料在流动压缩和机械压缩过程中复杂的压缩/松弛行为。由此产生的轴向流动轮廓对于流动压缩和机械压缩是不同的。在流动压缩过程中，观察到沿柱高沿着线性增加的压力分布，最高压缩发生在柱出口附近。然而，在机械压缩过程中，压力曲线沿着柱高呈指数分布，最高压缩发生在适配器下方的柱入口附近。因此，在第二个12个月的供资期内，将对流动压缩和机械压缩包装战略的适当建模组合进行研究，以获得更均匀的包装结构。在这种情况下，经验上建立的和工业上应用的包装程序将在可实现的包装均匀性方面得到改进。扩展模拟模型的额外引入的模型参数将通过单个参数研究来确定。新开发的模型将使调查的包装压实行为由于颗粒重排和颗粒迁移的第一次。通过应用粗晶粒概念，仿真模型所需的计算时间将显著减少。这将使得能够以高精度预测制备色谱柱的流体动力学行为，以避免填料的不可逆劣化，并基于模拟模型改进填料程序。

During the industrial application of preparative chromatography columns, unwanted phenomena are frequently encountered like medium wall detachment, partial bed subsidence, formation of flow channels, and irreversible compaction of the particle bed. It was shown in own previous work that these phenomena are a result of inhomogeneities of the particle bed which are caused by the packing process. During the first 3 year long funding period, a novel micro-chromatography column was developed which enabled a detailed experimental examination of the hydrodynamic behavior. A novel deterministic, three-dimensional, hybrid simulation model was developed by coupling computational fluid dynamics (CFD) and the discrete element method (DEM). This model enabled to simulate the interactions between single chromatography particles as well as between particles and fluid. A good agreement of simulation results and experimental data was obtained. The complex compression/relaxation behavior of the chromatography packing during flow compression and mechanical compression was studied. The resulting axial flow profiles turned out to be different for flow compression and mechanical compression. During flow compression, a linearly increasing pressure profile along the column height was observed with the highest compression taking place near the column outlet. During mechanical compression, however, the pressure profile along the column height was exponential with the highest compression occurring near the column inlet below the adaptor. Therefore, during the second 12-month long funding period, proper modelling-based combinations of flow compression and mechanical compression packing strategies are to be investigated in order to obtain a more homogeneous packing structure. In this context, empirically established and industrially applied packing procedures will be improved in terms of the achievable packing homogeneity. The additionally introduced model parameters of the extended simulation model will be identified by individual parameter studies. The newly developed model will enable to investigate the packing compaction behavior due to particle rearrangement and particle migration for the first time. The required computational time of the simulation model will be reduced significantly by applying the coarse grain concept. This will enable the prediction of the hydrodynamic behavior of preparative chromatography columns with high accuracy in order to avoid irreversible deterioration of the packing and to improve the packing procedures on the basis of the simulation model.