(569a) Computational Study of the Diffusion of Chemical Warfare Agents in Metal-Organic Frameworks in the Presence of Water Molecules

Chemical warfare agents (CWAs) are highly toxic and lethal materials that can cause significant damage to humans and the environment and, in many cases, result in death. Unfortunately, these materials have been used in battle and terrorist attacks. Hence, developing protective materials to capture and degrade CWAs is crucial. Diverse materials and methods have been implemented for these purposes over the past few decades. Research studies demonstrated that metal-organic frameworks (MOFs) are promising candidates for adsorption and degradation of CWAs. Both the equilibrium capacity and kinetics of the adsorption of CWAs are essential in MOFs. One of the important aspects of kinetics is the diffusion coefficient, which has been studied rarely and is mostly unknown for these systems. Since these materials are toxic, implementing molecular simulation techniques such as molecular dynamics (MD) simulations can help study the diffusion of these materials. In addition, because of the high level of toxicity of CWAs, simulants with similar chemical and physical properties and with lower toxicity usually have been used in experimental studies. In MD simulations, we can study CWAs and their corresponding simulants and find if they have similar diffusion behavior. Moreover, the presence of adsorbed water molecules can affect the diffusivity and adsorption of CWAs or simulants.

In this work, we studied the self-diffusion of the CWAs sarin and soman and the simulants DMMP, DIMP, and DIFP in the MOFs NU-1000 and DMMP in a series of zirconium MOFs with different connectivities, topologies, and linkers such as MOF-808, UiO-67, and MOF-841. The MOF structures were minimized using DFT calculations, and partial charges were calculated by the Density Derived Electrostatic and Chemical (DDEC06) approach. Rigid frameworks were considered for all MOFs. the TIP4P/2005 model was applied for the water molecules. For each MOF, different loadings of agent molecules under various relative humidities, ranging from 0 to 70%, were considered for the MD simulations. The initial locations of agents and water molecules in each MOF were determined using NVT or grand canonical Monte Carlo simulations in RASPA. Then, five ns NVT MD simulations were conducted in LAMMPS for equilibration at 298 K, followed by 300 ns or longer NVT MD simulations for the production run. Each simulation was repeated three times with different initial velocities and different locations of the adsorbate molecules. In addition to calculating the self-diffusivities and the preferred locations of the guest molecules, we also analyzed the water-water, CWA-water, CWA-framework, and water-framework hydrogen bonds to provide physical insights about the behavior and dynamics of these systems.