Predicting Adsorption Affinities of Small Molecules on Carbon Nanotubes Using Molecular Dynamics Simulation.

Computational techniques have the potential to accelerate the design and optimization of nanomaterials for applications such as drug delivery and contaminant removal; however, the success of such techniques requires reliable models of nanomaterial surfaces as well as accurate descriptions of their interactions with relevant solutes. In the present work, we evaluate the ability of selected models of naked and hydroxylated carbon nanotubes to predict adsorption equilibrium constants for about 30 small aromatic compounds with a variety of functional groups. The equilibrium constants determined using molecular dynamics coupled with free-energy calculation techniques are directly compared to those derived from experimental measurements. The calculations are highly predictive of the relative adsorption affinities of the compounds, with excellent correlation (r ≥ 0.9) between calculated and measured values of the logarithm of the adsorption equilibrium constant. Moreover, the agreement in absolute terms is also reasonable, with average errors of less than one decade. We also explore possible effects of surface loading, although we demonstrate that they are negligible for the experimental conditions considered. Given the degree of reliability demonstrated, we move on to employing the in silico techniques in the design of nanomaterials, using the optimization of adsorption affinity for the herbacide atrazine as an example. Our simulations suggest that, compared to other modifications of graphenic carbon, polyvinylpyrrolidone conjugation gives the highest affinity for atrazine-substantially greater than that of graphenic carbon alone-and may be useful as a nanomaterial for delivery or sequestration of atrazine.