🔬 materials science

Breaking Water at Graphene Defects

This study demonstrates that solvation qualitatively alters water dissociation at graphene single vacancies by enabling two competing acidic and basic pathways absent in the gas phase, thereby revealing unexpectedly rich interfacial chemistry with significant implications for carbon functionalization and nanofluidic transport.

Original authors: Samuel G. H. Brookes, IniOluwa C. Popoola, Fabian Berger, Kara D. Fong, Angelos Michaelides, Christoph Schran

Published 2026-09-11

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Samuel G. H. Brookes, IniOluwa C. Popoola, Fabian Berger, Kara D. Fong, Angelos Michaelides, Christoph Schran

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Water is a relentless chemist. Wherever it touches a solid surface, it can break apart, splitting into charged fragments that react with the material beneath. This process, known as dissociation, is the engine behind everything from the rusting of metal to the generation of clean energy in fuel cells. In a vacuum, breaking a water molecule apart is incredibly difficult, requiring a massive burst of energy. However, when water meets a surface, the surface can act as a catalyst, lowering the energy needed to split the bond. For decades, scientists have understood how this works on flat metal surfaces or metal oxides, where the geometry of the atoms guides the reaction. But a vast class of materials remains a mystery: two-dimensional sheets of carbon, known as graphene. These atom-thin materials are widely used in advanced membranes and electronic devices, yet we have lacked a clear picture of how water behaves when it encounters a flaw in the perfect carbon sheet. Specifically, we did not know if the surrounding liquid water changes the way the reaction happens compared to a single molecule floating in empty space.

To solve this, researchers at the University of Cambridge and their collaborators turned their attention to a specific type of flaw in graphene called a single vacancy. Imagine a perfect honeycomb pattern of carbon atoms; a single vacancy is simply the absence of one atom, leaving a hole where three neighbors are left dangling. This hole is not just a gap; it is a highly reactive site with unpaired electrons that make it eager to grab onto passing molecules. The team wanted to see what happens when a water molecule meets this hole, but they faced a significant hurdle. To understand the reaction, they needed to simulate not just one water molecule, but a full layer of liquid water surrounding the defect, all while tracking the breaking and forming of chemical bonds with extreme precision. Traditional computer models were too slow to run these simulations for the time required, and simpler models could not describe the complex electron behavior of the carbon atoms.

The researchers overcame this by training a new type of artificial intelligence model on high-level quantum mechanical data. This machine-learned potential allowed them to run simulations that were both fast enough to capture the movement of dozens of water molecules and accurate enough to describe the breaking of chemical bonds. They set up two scenarios: one with a single water molecule in a vacuum, and another with the same vacancy surrounded by a pool of about sixty water molecules. The results revealed a startling difference. In the vacuum, the reaction follows a single, straightforward path. The water molecule attacks the hole, and in one concerted motion, it splits, with one part attaching to the carbon and the other part staying attached to the oxygen. It is a direct, single-step process.

However, when the researchers introduced the liquid environment, the story changed completely. The surrounding water molecules acted as a stabilizing force for charged particles, opening up two entirely new pathways that do not exist in a vacuum. In the liquid, the water molecule can split in two different ways. In one route, the carbon hole grabs a hydrogen atom, leaving behind a negatively charged hydroxide ion that floats away in the water. In the other route, the carbon grabs the oxygen part, leaving a positively charged hydronium ion to drift off. These two competing pathways are significantly easier to start than the single path seen in a vacuum, requiring less energy to begin. The liquid environment essentially lowers the barrier for the reaction by stabilizing the charged pieces as they separate.

The study further showed that these different ways of splitting the water lead to different chemical structures on the graphene surface. Some of the resulting structures, particularly those where the carbon holds an oxygen group, act like magnets for water, holding onto the liquid for much longer than the bare hole or the original graphene sheet. This suggests that the defect transforms the surface from water-repelling to water-attracting. The researchers found that the acidic pathway, where a hydronium ion is released, is thermodynamically more stable, implying that these tiny holes in graphene could act as local sources of acidity, altering the chemical balance of the water right at the surface.

This work demonstrates that the presence of liquid water is not just a background detail but a fundamental driver of chemical reactivity on two-dimensional materials. By revealing that solvation opens up new, lower-energy routes for water to split, the study suggests that the behavior of graphene in real-world applications—such as in water filtration membranes or energy devices—cannot be predicted by studying the material in isolation. The liquid environment fundamentally rewrites the rules of the reaction. The findings indicate that by controlling these defects, scientists might be able to tune how water interacts with graphene, potentially allowing for the design of materials with specific wetting properties or enhanced abilities to transport ions. The research highlights that to truly understand chemistry at the interface of solids and liquids, one must look beyond the vacuum and into the complex, stabilizing embrace of the solvent itself.

Technical Summary: Breaking Water at Graphene Defects

Problem Statement
Water dissociation at solid surfaces is a fundamental process underpinning corrosion, catalysis, electrochemistry, and photovoltaics. While defect-mediated reactivity and solvent-assisted proton transfer are well-established for metals and metal oxides, the coupling of these phenomena in two-dimensional (2D) materials like graphene remains poorly understood. Specifically, it is unclear how solvation alters the mechanism and energetics of water dissociation at graphene defects, and how the resulting chemisorbed species modify the local water environment. Traditional Density Functional Theory (DFT) is too computationally expensive for modeling fully solvated systems over the nanosecond timescales required for enhanced sampling, while classical force fields fail to describe bond-breaking and bond-making events accurately.

Methodology
To address these limitations, the authors developed a machine-learned interatomic potential (MLIP) based on the equivariant MACE architecture. This potential was trained on spin-polarized hybrid DFT data using the revPBE0-D3 functional, which is essential for correctly describing the localized $\sigma$ and delocalized $\pi$ defect states of the graphene single vacancy (SV) without suffering from the self-interaction errors common in semi-local functionals.

The study employed multiple-walker well-tempered metadynamics to explore the free energy landscape of water dissociation at a graphene SV. Two distinct environments were simulated:

Gas-phase: A single water molecule interacting with the SV.
Solvated: Approximately 60 water molecules surrounding the SV.

The reaction progress was tracked using two collective variables: the C–H coordination number and the C–O coordination number at the vacancy site. Simulations were conducted at 300 K, with cumulative simulation times of 6 ns for the gas-phase and 40 ns for the solvated system. Water residence lifetimes were further analyzed using unbiased NVT simulations over 2 ns.

Key Contributions and Results
The study reveals that solvation qualitatively transforms the mechanism of water dissociation at a graphene single vacancy:

Mechanistic Shift: In the gas phase, water dissociation proceeds via a single concerted channel where the water molecule attacks the vacancy, breaking an O–H bond and forming C–OH and C–H bonds simultaneously. This process has a forward free energy barrier of 28.4 kcal/mol.
Solvated Pathways: In the presence of bulk water, the reaction splits into two competing heterolytic pathways, both of which have lower barriers than the gas-phase concerted process:
1. Basic Route: The SV abstracts a proton to form a chemisorbed SV–H intermediate and a solvated hydroxide ion ( $OH^-$ ). The barrier is 21.9 kcal/mol.
2. Acidic Route: Water attacks the vacancy oxygen-first to form a chemisorbed SV–OH intermediate, transferring a proton to the solvent via the Grotthuss mechanism to form a solvated hydronium ion ( $H_3O^+$ ). The barrier is 23.7 kcal/mol.
Thermodynamic Preference: The acidic pathway (SV–OH + $H_3O^+$ ) is thermodynamically favored, with the intermediate being approximately 20 kcal/mol more stable in free energy than the basic intermediate. This suggests that solvated SVs may act as local sources of hydronium ions, contributing to interfacial acidity.
Impact on Wettability: The chemisorbed intermediates (Hydroxy-SV and SV-Ketone) significantly enhance water adsorption compared to the bare vacancy or pristine graphene. The Hydroxy-SV state exhibits a water residence time of 55 ps, and the SV-Ketone state 27 ps, compared to much shorter lifetimes for non-reactive sites. This indicates that defect chemistry converts the local environment from hydrophobic to hydrophilic.

Significance and Claims
The paper claims that a single atomic vacancy in graphene is sufficient to transform the local environment from physisorption to reactive chemisorption, a transition driven by the stabilization of charged intermediates by the solvent. The authors highlight that the dual reactivity of the solvated SV—acting as both a proton acceptor and donor—resembles the ambiphilic character of carbenes in molecular chemistry.

The study emphasizes the necessity of explicit solvation in simulating reactive interfacial systems, extending this understanding from metals and oxides to 2D carbon surfaces. The authors suggest that these findings imply that reaction mechanisms characterized only under vacuum conditions may require reassessment for other 2D materials. Furthermore, the distinct water adsorption properties and residence times associated with different chemisorbed states suggest that controlled defect engineering could offer a route to tuning wettability, friction, and ion transport in 2D membrane materials. The work demonstrates that machine-learned potentials trained on hybrid DFT, combined with enhanced-sampling techniques, are now feasible for resolving the full free energy landscape of rare reactive events at explicitly solvated interfaces.

More like this