Distribution of Sodium Cations in Faujasite-Type Zeolite:  A Canonical Parallel Tempering Simulation Study

Abstract

A Monte Carlo replica exchange (MCRE) algorithm was used to compute the sodium cation distribution in bare faujasite zeolite with a number of cations per unit cell ranging from zero (Si:Al → ∞) to 96 (Si:Al = 1). Eight independent realizations of the system were simulated simultaneously, in the temperature range 300-2325 K. The resulting distributions at room temperature were found to be in very good agreement with both available experiments and the analytical quasi-chemical model of Mortier and co-workers. A single canonical simulation at room temperature using site-to-site hopping yields identical results. The main advantage of the MCRE method is that no assumption is needed on the actual cation adsorption sites. One could thus, in principle, predict cation locations and distributions in a nanoporous solid for which the precise location of extraframework cations is not known.

Cation redistribution upon dehydration of Na58Y faujasite zeolite: a joint neutron diffraction and molecular simulation study

“…The parameters are taken from Jaramillo et al [12] for the zeolite framework (in particular, Si and Al atoms are described as "T atoms" carrying identical charges) and from Dang et al [13] for Na + cations, together with the TIP4P water model of water [14]. This force field was successfully used in previous studies of water adsorption in Y and X zeolites [4,5,6,15,16]. Periodic boundary conditions are used, electrostatic interactions are computed using Ewald summation and a cut-off equal to half the box length is used to compute Lennard-Jones interactions.…”

“…The sampling of configuration space is enhanced by using "jump moves", i.e. large amplitude trial displacements for the cations, and by simulating the system at 700 K. While this does not correspond to the target temperature, we observed that it provides in the cases under consideration the same results as parallel tempering [5,17] at a much lower computational cost (only one temperature is used, instead of the 14 replicas for our parallel tempering simulations). In particular, we have checked that this procedure allows us to obtain reproducible results starting from uncorrelated initial conditions, and to overcome possible free energy barriers separating metastable cation distributions, which may result in hysteresis in the hydration/dehydration processes.…”

“…One of the reasons for the discrepancy between cation location results is the strong sensitivity of cation distribution to experimental conditions (temperature, dehydration conditions, nature and quantity of sorbed molecules …) [1,3]. The presence of water has been shown to drastically modify the cation location in zeolites and in particular in faujasite zeolite [4,5,6,7,8]. The experimental determination of cation location in hydrated sample is complicated because of partial occupancies affecting low symmetry sites.…”

Development and Application of Molecular Simulation Methods for the Screening of Industrial Zeolite Adsorbents

“…Among the above moves, jump moves appear to be essential for the efficient finding of equilibrium cationic locations. Compared with our previous investigations [18], they allow for a quicker convergence without requiring exploration of higher temperatures.…”

“…The location of cations is also a delicate issue, because it is not always available from experimental measurements and it may be influenced by the presence of water, which is often co-adsorbed in laboratory measurements or in industrial processes. Application of Monte Carlo techniques has provided encouraging answers to this problem [18][19][20].…”