Controlling Amorphous Silica Surface Roughness.*

Amorphous silica (a-SiO₂) surfaces have found application in heterogeneous catalysts by grafting metals on the active sites of the functionalized surfaces. From a molecular modeling perspective, one challenge has been the creation of a-SiO₂ slab models with a wide range of surface functionalization and roughness. In this work, we present a method, based on Capillary Fluctuation Theory (CFT), to create and functionalize a-SiO₂ surfaces of controlled surface roughness using the ReaxFF potential, which is a classical reactive force field that allows bond forming, breaking and charge equilibration without the high cost of quantum mechanics methods. Currently, the most common method for generation of a-SiO₂ surfaces with ReaxFF involves a “melt-quench-cleave” protocol in which a-SiO2 is equilibrated a high temperature (above T_g) followed by quenching to room temperature and cleaving with a flat plane, a procedure that yields artificially flat surfaces. In this study, we modify this method to generate surfaces of controlled roughness, as measured by the mean-squared displacement (MSD) of the surface atom positions from a flat plane. In this work, we generate a sample set of 2-d grid meshes in Fourier space from the CFT distribution corresponding to a specific surface stiffness. Inverse Fourier transforms of each k-space mesh yields a set of real-space surface meshes of controlled mean-squared displacement. These real-space surface meshes are then used to cleave the quenched silica bulk samples to create silica surfaces of controlled roughness. We show that this procedure yields a linear relationship between the input surface stiffness and the MSD of the surface, consistent with CFT. To create rough functionalized surfaces, the slabs are then exposed to ReaxFF water to allow for autofunctionalization of the surface with silanol groups followed by structural characterization of the slabs.