Computational evaluation of aluminophosphate zeotypes for CO₂/N₂ separation

Adsorbent materials are essential for capturing carbon dioxide from industrial emissions. Zeolites and similar zeotype structures have been studied for their potential in selective CO2 adsorption. AlPOs, a subclass of zeotypes, have received less attention compared to aluminosilicate zeolites. Experimental studies on AlPOs remain limited, especially regarding their CO2 adsorption behavior. This gap motivated researchers to explore computational methods for evaluating AlPOs. Prior research has shown that zeolites can selectively capture CO2, but the role of AlPOs in this context is unclear. Computational models offer a way to screen many structures efficiently. This study aimed to fill that knowledge gap by using simulations to assess CO2/N2 separation potential in AlPOs. The findings could help prioritize which AlPOs to investigate experimentally.

This study aimed to computationally evaluate the CO2/N2 separation performance of aluminophosphate (AlPO) frameworks. The goal was to identify promising candidates for selective CO2 adsorption from flue gases. CO2/N2 mixtures are a relevant model system for this application. The researchers used simulations to predict adsorption behavior under different conditions. They focused on comparing selectivity and working capacity across various AlPO structures. The motivation was to guide future experimental work on AlPOs. By simulating a large number of AlPO frameworks, the study sought to identify top-performing structures. The results could help reduce the need for costly and time-consuming experimental trials.

The researchers used grand-canonical Monte Carlo (GCMC) simulations to model CO2/N2 adsorption in AlPO frameworks. They first optimized the structures using dispersion-corrected density-functional theory calculations. A total of 51 AlPO frameworks were considered, including pure AlPOs and derivatives with heteroatoms. The simulations focused on Henry constants and CO2 uptake at various pressures. The team selected 21 AlPOs based on preliminary results for detailed isotherm calculations. Adsorption isotherms were computed for a 15:85 CO2/N2 mixture up to 10 bar. Interaction energy maps were generated to analyze host-guest interactions. The approach allowed the team to compare selectivity and working capacity across different structures.

The simulations revealed significant variation in CO2/N2 separation performance among the 51 AlPOs. Four topologies—GIS, ATN, ATT, and SIV—showed the highest selectivity (75–140) at 1 bar pressure. These structures also had reasonable CO2 working capacities (1–1.7 mmol g⁻¹). Under vacuum-swing adsorption conditions, GIS performed particularly well. However, under pressure-swing adsorption, selectivity and working capacity were inversely related. Frameworks like AFY, KFI, and SAV had high working capacities (above 2 mmol g⁻¹) but lower selectivities (25–35). The results suggest a tradeoff between these two performance metrics. Interaction energy maps provided insights into the molecular-level interactions driving adsorption. These findings help identify which AlPOs are most promising for experimental validation.

The computational study identified several AlPO frameworks with high potential for CO2/N2 separation. GIS, ATN, ATT, and SIV topologies showed the best combination of selectivity and working capacity under vacuum-swing conditions. However, under pressure-swing conditions, higher working capacities came at the expense of lower selectivity. The authors suggest that these findings can guide future experimental efforts in developing AlPO-based adsorbents. The results highlight the importance of matching adsorption conditions to the intended application. The study also emphasizes the value of computational screening in reducing the number of experimental trials needed. Interaction energy maps provided atomic-level insights into CO2 adsorption mechanisms. These conclusions are based on the simulation data and do not propose new experimental methods or future research directions.