Curvature-dependent adsorption of water inside and outside armchair carbon nanotubes

TL;DR

This study investigates how water molecules adsorb inside and outside armchair carbon nanotubes with different curvatures using advanced quantum chemistry methods, revealing curvature-dependent stability and developing force field models.

Contribution

It provides detailed curvature-dependent adsorption energies of water on CNTs and introduces two Lennard-Jones parameter sets to accurately model these interactions.

Findings

Water inside CNTs is stabilized by electron correlation, with adsorption energy decreasing as curvature decreases.

Outside CNT water adsorption energies vary little with curvature, with hydrogen pointing towards the wall being most stable.

Two sets of Lennard-Jones parameters are needed to accurately reproduce water-CNT interactions across different curvatures.

Abstract

The curvature dependence of the physisorption properties of a water molecule inside and outside an armchair carbon nanotube (CNTs) is investigated by an incremental density-fitting local coupled cluster treatment with single and double excitations and perturbative triples (DF-LCCSD(T)) study. Our results show that a water molecule outside and inside (n, n) CNTs (n=4, 5, 6, 7, 8, 10) is stabilized by electron correlation. The adsorption energy of water inside CNTs decreases quickly with the decrease of curvature (increase of radius) and the configuration with the oxygen pointing towards the CNT wall is the most stable one. However, when the water molecule is adsorbed outside the CNT, the adsorption energy varies only slightly with the curvature and the configuration with hydrogens pointing towards the CNT wall is the most stable one. We also use the DF-LCCSD(T) results to parametrize Lennard-Jones (LJ) force fields for the interaction of water both with the inner and outer sides of CNTs and with graphene representing the zero curvature limit. It is not possible to reproduce all DF-LCCSD(T) results for water inside and outside CNTs of different curvature by a single set of LJ parameters, but two sets have to be used instead. Each of the two resulting sets can reproduce three out of four minima of the effective potential curves reasonably well. These LJ models are then used to calculate the water adsorption energies of larger CNTs, approaching the graphene limit, thus bridging the gap between CNTs of increasing radius and flat graphene sheets.