Interfacial Behavior from the Atomic Blueprint: Machine Learning-Guided Design of Spatially Functionalized a-SiO2 Surfaces

TL;DR

This work demonstrates that the spatial arrangement of OH and CH3 groups on α-SiO2(0001) surfaces—not just overall composition—drives interfacial energetics, hydrogen-bond networks, and vibrational properties. By integrating DFT, ab initio MD, and on-the-fly learned force fields, it maps how different functionalization patterns alter mixing enthalpy, revealing a thermodynamically favored unpaired configuration near 67% CH3. The findings reproduce and rationalize experimental trends in surface energy and OH-stretch shifts, highlighting patterning as a design principle for silica-based coatings and interfaces. The approach provides a quantitative framework for tailoring wettability and interfacial behavior through atomic-scale spatial control of surface functionalities.

Abstract

a-Quartz surfaces functionalized with hydroxyl and methyl groups provide a versatile platform for controlling interfacial properties critical to applications such as catalysis, protective coatings, and energy conversion. The arrangement of these functional groups strongly influences interfacial interactions at solid-liquid interfaces, highlighting their relevance to colloid and interface science. However, conventional models often treat surface functionalization as spatially homogeneous, overlooking the atomic-scale organization of surface groups. We hypothesize that this spatial distribution, beyond overall composition, plays a decisive role in governing surface stability and interfacial behavior. To test this hypothesis, we employ a multi-scale simulation workflow combining density functional theory, ab initio molecular dynamics (AIMD), and machine-learned force fields (MLFFs). This approach allows us to explore a range of spatial patterns of OH/CH3 functionalization on the a-quartz (0001) surface. We evaluate the impact of spatial arrangements on mixing energy, hydrogen bonding networks, and vibrational properties with high accuracy and robustness. Our results reveal that spatial patterning strongly influences surface stability and interfacial structure. A thermodynamically favored unpaired configuration emerges near 67 % CH3 substitution, where isolated OH groups form secondary hydrogen bonds through reorientation toward subsurface oxygen atoms. This rearrangement induces a characteristic blue shift in OH stretching frequencies, indicating weaker H-bonding. These effects are absent in clustered arrangements. By establishing a clear link between functional group patterning and interfacial behavior, our work uncovers the underlying mechanisms to guide and accelerate the rational design of silica-based materials and coatings, directly relevant to colloid and interface science.

Figures (7)

• Figure 1: Snapshots of the mixed functionalization of the $\alpha$-SiO2 (0001) surface with varying OH/CH3 ratios. The DFT-optimized structures illustrate the spatial distribution of functional groups across different compositions. The silicon atoms are shown in yellow, framework oxygen atoms (SiO2) in orange, hydroxyl oxygen atoms in red, hydrogen atoms in white, and carbon atoms in grey.
• Figure 2: The external validation of the ML potential by assessing the forces and energy errors calculated by MLFF and verified by means of single point DFT calculation.
• Figure 3: Mixing enthalpy of OH/CH3 functional group distributions on the $\alpha$-SiO2 (0001) surface, calculated using the average total free energy from MD simulations at $300$ K. Each data point represents the average over $2$ independent surface models, with statistics collected from $80000$ MD steps. Green represents unpaired OH-Si-CH3 configurations, while blue and yellow indicate paired Si-(OH)2/Si-(CH3)2 spread and clustered configurations, respectively. Red denotes DFT validation points. The mixing enthalpy is normalized by the number of surface Si atoms. Black rhombs correspond to experimental surface mixing enthalpies extracted from the work of Yang et al. Yang2009
• Figure 4: Snapshots of the hydrogen bonding arrangement for the functionalized $\alpha$-SiO2 (0001) surface with varying OH/(OH+CH3) ratios: A - $100$ % OH, B - $67$ % OH (paired, cluster), C - $33$ % OH (unpaired) and D - $33$ % OH (paired). The silicon atoms are shown in yellow, framework oxygen atoms (SiO2) in orange, hydroxyl oxygen atoms in red, hydrogen atoms in white, and carbon atoms in grey. The arrows indicate the direction of OH groups forming the hydrogen bond network. The solid black line denotes the characteristic zig-zag signature of the hydrogen‐bonding network on the fully hydroxylated surface.
• Figure 5: Average tilt of OH groups depending on the OH/CH3 functional group distributions on the $\alpha$-SiO2 (0001) surface, calculated from MD simulations at $300$ K. Each data point represents the average over $2$ independent surface models, with statistics collected from $80000$ MD steps. Green colour represents unpaired OH-Si-CH3 configurations, while blue indicates paired Si-(OH)2/Si-(CH3)2 configurations.