(393g) Enhancing Reversibility in Adsorption Processes: The Hysteresis Challenge

Adsorption hysteresis can significantly impact the functionality of porous materials like metal-organic frameworks (MOFs), influencing the reversibility of adsorption in applications such as adsorption cooling and water harvesting, where reversibility is essential. Hysteresis is especially pronounced in MOFs with large pores, which hinders the use of these materials despite their potential for high adsorption capacity. Adsorption hysteresis arises from the presence of high energy barriers separating low- and high-density phases of the fluid confined in the pore and is coupled with the tendency of the system to remain in metastable phases.

In this study, we explore adsorption hysteresis through the van der Waals (vdW) loop, a metric that quantifies the stability limits of adsorption hysteresis. The vdW loop maps out the density of stable, metastable, and unstable phases of the confined fluid, utilizing spinodal and binodal points to define the stability limits of the phases and the equilibrium pressure for coexisting phases, respectively. To accurately predict the vdW loop, we employ transition matrix Monte Carlo (TMMC) simulations. These simulations offer a comprehensive description of the free energy landscape associated with adsorption, enabling the capture of all possible system macrostates, including unstable regions of the phase diagram (e.g., barriers between states). Our analysis leverages the gRASPA simulation software, developed in our group and optimized for GPU nodes, to significantly accelerate the adsorption simulations.

Our findings illustrate the effects of the simulation system's size on the width of the hysteresis loop. These effects, in turn, correlate with the barrier associated with the transition between low- and high-density phases. We further explore how specific interaction patterns on the surfaces of idealized pore geometries (such as hydrophobic and hydrophilic sites) can influence these properties. Patterns explored in model systems can be correlated with the distribution of functional groups in real MOFs. This relationship is important for the design of MOFs optimized for minimal adsorption hysteresis and for enhancing their efficiency in applications like adsorption cooling and water harvesting. Additionally, we establish connections between the hysteresis loop, the transition barrier, and particular pore shapes and system sizes in MOFs, offering new insights into the design of new materials showing reversible adsorption.