Modellbasierte Optimierung der Struktureigenschaften von polymeren Adsorbentien für die chromatographische Trennung von Glucose und Xylose

  • Model-based optimization of structural properties of polymeric resins for the chromatographic separation of glycose and xylose

Leipnitz, Martin; Jupke, Andreas (Thesis advisor); Bathen, Dieter (Thesis advisor)

Aachen (2021)
Book, Dissertation / PhD Thesis

In: Aachener Verfahrenstechnik series AVT. FVT - fluid process engineering 5 (2021)
Page(s)/Article-Nr.: XVIII, 134 Seiten : Illustrationen, Diagramme

Dissertation, Rheinisch-Westfälische Technische Hochschule Aachen, 2021

Abstract

In preparative chromatography, the adsorbent is one of the most important influencing parameters which determines the overall performance of the designed apparatus. This performance can significantly be increased by applying model-based optimisation approaches in the designing stage of the separation unit. However, such approaches usually focus on the optimisation of the design and the operating point of an apparatus after selection of a suitable adsorbent. Even though, the selection of the adsorbent is crucial, the adsorbent is typically selected based on expert knowledge or experimental investigation of pre-selected promising candidates. Therefore, the selection of the adsorbent is not included into model-based optimisation approaches. In this contribution, a model-based approach for the optimised selection of an adsorbent was developed. As exemplary system to investigate, a mixture of glucose and xylose was chosen. Purification of monosaccharides using cation exchanger resins and water is established in preparative scale. The degree of cross-linking, ionic form and particle size were identified as the performance influencing structural properties of these resins. Using a semi-preparative chromatographic setup, model parameters of a transport dispersive model were determined for resins with degrees of cross-linking of 2% to 9 %, particle sizes of 55.5 μm to 350 μm and ionic forms of Na-, Ca-, Mg- and K-form. Based on the experimental data, correlations between model parameters and structural properties were determined. A discontinuous separation of glucose and xylose applying a fractionation strategy without waste fraction was investigated. Based on the derived mathematical correlations, structural properties as well as design and operating parameters were optimised using mechanistic modeling. The resulting performance of resins with optimised structural properties were compared to resins with industrially established structural property values. For all resins with optimised structural properties, significant increases in the volume specific productivity were achieved. Besides the optimisation of structural properties of a resin for a specific separation task, influences of constraints of the separation task on the values of the optimal structural properties were investigated. In different optimisation case studies, structural properties as well as design and operating parameters were optimised for varying demanded purities and yields considering xylose in the xylose-rich product fraction as well as for varying compositions of the feed mixture. For varying constraints of the separation task, different optimal structural properties of resins were identified. Concluding from these results, resins with highest selectivity factors do not necessarily correspond to the highest volumespecific productivity. It became rather apparent, that structural properties of the optimal resin has to be designed to the specific separation task including the consideration of its constraints.

Identifier