Continuous Separation of Multicomponent Aromatic/Aliphatic Mixtures By Simulated Moving Bed Adsorption: Combined Modeling and Experimental Approach

Simulated moving bed (SMB) chromatography has potential for high-efficiency separation of many valuable chemical mixtures. Much work in SMB-based separations has focused on binary mixtures, but considerably less attention has been devoted to the separation of multicomponent feeds. In this talk, we take a combined experimental and modeling approach to study the multicomponent separation of aromatics and aliphatics using a nano-porous silica adsorbent, which is of relevance in the separation of complex streams in petrochemical refineries. Our approach is based upon a more rigorous determination of adsorption, mass transfer, and SMB process model parameters based upon detailed experimental input (including column concentration profiles) from SMB experiments carried out in a 16-column “mini-plant” designed and constructed in-house at the Georgia Institute of Technology. Rigorous SMB process modeling with a simpler model mixture of components, along with SMB operation experiments with column-level composition sampling, allows us to demonstrate reliable prediction and validation of the effects of operational parameter changes on the compositional profiles and separation performance. We then generalize this approach by adding more components to the mixture, and obtain a very good agreement between model and experimental concentration profiles for a range of operating conditions and representative aliphatic and aromatic components in the feed. Hence, we demonstrate a robust multicomponent SMB process model with the capability to quantitatively predict the influence of key operating parameters such as extract, feed, desorbent and recycle flow rates, desorbent/feed ratio, and switch time on SMB separation results as well as column concentration profiles. Furthermore, conditions for clear separations of complex mixtures are also predicted and demonstrated.