A Molecular Simulation Approach to Studying Mass Transfer Across Surface Barriers

Mass transfer in technologically important microporous materials such as zeolites and carbon molecular sieves (CMS) is currently a subject of great interest,1,2 since mass transfer properties are often determining factors in the efficiency of separation processes. The effort toward understanding these properties has largely concentrated on the mobility of molecules within the microporous regions of these materials, where the limiting resistance to mass transfer in a process is usually located. In many experimental and industrial situations, however, resistances which are external to the micropore structure are important and must be considered. A surface barrier effect has been observed for both zeolites and CMS.1 The nature of this barrier is not yet well understood, but it has been postulated that it is associated with a narrowing of the pore mouths at the boundaries between microporous and macroporous regions of adsorbent material. This narrowing is believed to cause a significant resistance to mass transfer due to the steric and energetic interactions between the diffusing molecules and those comprising the pore mouth. This is not the only possible cause of a surface barrier effect. For example, the structure of a pelletized adsorbent could be significantly different in a macroscopic region near the pellet surface. If this region is small compared to the pellet dimensions, an apparent barrier effect may arise although the resistance is not concentrated at the pore entrances but rather is distributed through an interfacial volume. Laminar boundary layer effects could also produce such a barrier under certain circumstances. It is the goal of the present work to evaluate the contribution of the pore mouth resistance to mass transfer and determine if it is sufficient to cause the surface barriers seen experimentally. In the following section, a brief review will be given of experimental observations of the surface barrier effect. Also, several previous attempts to model this behavior at a molecular level will be discussed. In subsequent sections we summarize two modeling approaches which we have used to study this phenomenon and we present ideas which we believe would profit from further study.