(228c) Improving the Accuracy of Surface Area Determination of Porous Materials: A Weighted Average BET Approach

Porous materials, including metal-organic frameworks (MOFs) and zeolites, exhibit considerable potential for a range of applications, particularly in gas storage, gas separation, and catalysis, due to their distinctive properties, such as structural diversity and substantial porosity. The surface area of porous materials is an important indicator for evaluating their performance in adsorption-related application, and the Brunauer-Emmett-Teller (BET) method is widely recognized as the most well-known and extensively used technique for determining this key property. However, when applying the BET method, it is essential to select an appropriate region for surface area calculation. The subjective choice of the linear region can introduce uncertainties in the results, leading to variations in the computed area even for the same structure. Besides, the BET method has been reported to greatly overestimate the surface area, especially for highly porous materials. To address the challenges associated with the conventional BET method, a new “weighted average (WA)” BET approach is proposed herein to mitigate the uncertainty and overestimation. Instead of trying to identify a single, hopefully “correct” linear region, this WA BET approach considers all possible linear regions to determining the surface area by summing their weighted contribution. To assess the effectiveness of the WA BET method, more than 260 structural diverse MOFs are investigated with their Ar adsorption isotherms at 87 K computed by grand canonical Monte Carlo (GCMC) simulations. An improvement in accuracy of approximately 4% in determining the surface areas is achieved using the WA BET method. More importantly, for those MOFs with a surface area greater than 3,500 m²/g, the improvement can reach nearly up to 10%. The accuracy of the WA BET approach is also found to be superior to the excess sorption work (ESW) approach and the inversion point (I-point) method, both of which do not require the selection of linear regions. Overall, the findings of this work are anticipated to aid the future structural characterization of newly synthesized porous materials.