Hydroxylation of silica nanoclusters (SiO2)M(H2O)N, M = 4, 8, 16, 24: stability and structural trends

Authors Organisations
Type Article
Original languageEnglish
Pages (from-to)20438-20443
JournalPhysical Chemistry Chemical Physics
Volume15
Issue number47
Early online date05 Nov 2013
DOI
Publication statusPublished - 2013
Links
Permanent link
View graph of relations
Citation formats

Abstract

Employing global optimisation and ab initio calculations, we follow the step-wise molecular hydroxylation of (SiO2)M(H2O)N, M = 4, 8, 16, 24, cluster species from their anhydrous state up to a N:M ratio (RN/M) of ≥0.5. Increasing N from zero for low RN/M values, significantly and progressively, energetically stabilises all cluster sizes. In all cases, this initial steep decrease in energy levels off at a well-defined threshold RN/M value to a linear regime where the decrease in energy per hydroxylation by a water molecule reaches a stable minimum value. Analysis of the structures of the globally optimised clusters for each size, M, and hydroxylation, N, reveals that the initially anhydrous cluster structures have increasingly tetrahedral SiO4 centres until the transition to the linear hydroxylation regime, whereupon the average deviation from tetrahedrality starts to increase. With increasing RN/M the smaller clusters (M = 4, 8) tend to open up and incorporate increasing numbers of Si–OH groups, seemingly approaching the limit of separate Si(OH)4 monomers. The larger clusters considered (M = 16, 24), however, are more resistant to structural disruption and with increasing RN/M energetically prefer not to form more Si–OH groups but, instead to form hydrogen bonds with subsequent water molecules on their surfaces. Such behaviour is also found to be more energetically favourable than the formation of fully-hydroxylated Si16O24(OH)16 and Si24O36(OH)24 cage isomers for RN/M = 0.5. We further found that the threshold RN/M value at which the transition to the linear hydroxylation regime is encountered follows an inverse power law with respect to increasing cluster size M which may indicate the existence of a more general fundamental basis underlying our results.