Simulation of Forces between Humid Amorphous Silica Surfaces: A Comparison of Empirical Atomistic Force Fields

Abstract

Atmospheric humidity strongly influences the interactions between dry granular particles in process containers. To reduce the energy loss in industrial production processes caused by particle agglomeration, a basic understanding of the dependence of particle interactions on humidity is necessary. Hence, in this study, molecular dynamic simulations were carried out to calculate the adhesion between silica surfaces in the presence of adsorbed water. For a realistic description, the choice of force field is crucial. Because of their frequent use and transferability to biochemical systems, the Clay and CWCA force fields were investigated with respect to their ability to describe the water–silica interface in comparison to the more advanced Reax force field, ab initio calculations, and experiments.

Molecular Simulation of CO2- and CO3-Brine-Mineral Systems

“…The apparent discrepancy between the monolayer coverage of silica at P/ P sat ~0.7 predicted by GCMC simulation (Leroch and Wendland 2012) and the bilayer coverage observed experimentally at the same relative vapor pressure (Asay et al 2009) may arise from the pH-dependence of silica surface charge: at near-neutral pH values, silica has a negative surface charge density because of the deprotonation of silanol groups (>Si-OH = >Si-O − + H + , with pK a = 7.0 ± 0.6; Sonnefeld et al 2001;Carroll et al 2002;Dove and Craven 2005). The uncharged silica surface (with fully protonated silanol groups) simulated by Leroch and Wendland (2012) is representative of acidic conditions, the point of zero net charge of silica being located near pH 3 (Wang et al 2012b). Figure 11b includes data obtained in systems where P/P sat < 1 was imposed by placing a bubble of water vapor (Wensink et al 2000;Bagherzadeh et al 2012) or CO 2 (Bagherzadeh et al 2012) in contact with the adsorbed water film.…”

“…Grand Canonical Monte Carlo simulation predictions of water film thickness vs. P/P sat on silica (Leroch and Wendland 2012), mica (Malani and Ayappa 2009), and calcite surfaces (Rahaman et al 2008) in mineral-water-vapor systems predict that these surfaces have similar hydration properties at P/P sat values up to 0.7, where they carry a single statistical water monolayer (Fig. 11b).…”

“…The figure shows that adsorbed water molecules are attracted to the forsterite surface because of their affinity for structural Mg 2+ ions (light gray spheres). (b) Compilation of MD and GCMC simulation predictions of the thickness of adsorbed water films on flat mineral surfaces at T = 298 to 300 K on silica (Bagherzadeh et al 2012;Leroch and Wendland 2012;Wensink et al 2000), calcite (Rahaman et al 2008), and mica (Malani and Ayappa 2009). Values of film thickness and P/P sat in the simulations of Wensink et al (2000) and Bagherzadeh et al (2012) were estimated by visual inspection of MD simulation snapshots and by applying Equation (1) with γ gw = 54.7 mN m −1 and r m = 1.1 ± 0.2 nm (Wensink et al 2000) or with γ gw = 80.1 or 70.1 mN m −1 (in the absence or presence of CO 2 , respectively) and r m = 1.93 ± 0.4 nm (Bagherzadeh et al 2012) [γ gw values were roughly estimated from the results of Vega and deMiguel (2007) water monolayer (Leroch and Wendland 2012), and uncharged silica surfaces exposed to a bulk CO 2 fluid carry zero water monolayers (Bagherzadeh et al 2012).…”

“…Bagherzadeh et al (2012) used a water model (TIP4P/Ice) that significantly overestimates the surface tension of liquid water (Vega and de Miguel 2007) and a silica-water interaction model (Lopes et al 2006) that was not designed for use with the TIP4P/Ice water model and that performs less than optimally when used with at least one other water model (Skelton et al 2011). In recent years, the CLAYFF model (Cygan et al 2004b) has emerged as a remarkably accurate model of the interaction between silicate minerals and liquid water (Bourg and Sposito 2010;Ferrage et al 2011;Marry et al 2011;Skelton et al 2011;Bourg and Steefel 2012;Leroch and Wendland 2012), but this model has been exclusively used with two water models (the SPC and SPC/E water models; Berendsen et al 1981Berendsen et al , 1987) that underestimate the surface tension of liquid water (Vega and de Miguel 2007;Nielsen et al 2012). When used with the SPC water model, the CLAYFF model underestimates the adhesive energy caused by a water meniscus between silica surfaces in humid air (P/P sat = 0.3 to 0.8) by 5 to 22% (Leroch and Wendland 2012).…”

Influence of Capillary Bridge Formation onto the Silica Nanoparticle Interaction Studied by Grand Canonical Monte Carlo Simulations

“…Despite its simplicity, it reproduces the amorphous silica structure very accurately and moreover predicts the correct crystal-glass transition temperature. For the water–silica interface we decided to use the conventional Clay force field 28 which has shown 18 a reasonably well described hydrogen-bond network on the silica surface as well as adhesion energies in agreement with experiments.…”