Water structuring at stacked graphene interfaces unveiled by machine-learning molecular dynamics

Dianwei Hou, Yevhen Horbatenko, Stefan Ringe, Minhaeng Cho · 2025

Read Water structuring at stacked graphene interfaces unveiled by machine-learning molecular dynamics by Dianwei Hou, Yevhen Horbatenko, Stefan Ringe, Minhaeng Cho (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2508-17685.txt.

Introduction

Does water bead on graphene, or spread? The answer has been fought over for years. It shifts with what sits under the sheet. Dianwei Hou and three co-authors settle part of the fight in 2025. Their simulations learn their forces from a machine model. That lets them run at near first-principles accuracy for far longer.

Their mock spectra clear up a puzzle. They match a lab method that reads surface water. A lone sheet on a water-loving base looks water-loving in tests. The team shows that look comes from water that has crept under the sheet. Its signal cancels part of the top signal. That under-layer is worth forming for a lone sheet, but not for a stack. The result gives a clear, physical read of how graphene meets water.

Catalog record

CREATORDianwei Hou, Yevhen Horbatenko, Stefan Ringe, Minhaeng Cho
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE61 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2508-17685

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Water structuring at stacked graphene interfaces unveiled by machine-learning molecular dynamics Dianwei Hou, Yevhen Horbatenko, Stefan Ringe, Minhaeng Cho Abstract The wettability of monolayer and multilayer graphene remains a topic of longstanding debate. Here, we combined first-principles molecular dynamics simulations accelerated with the atomic cluster expansion machine learning interatomic potential to investigate how substrate, graphene layer number, and intercalated water molecules influence graphene’s wettability. Simulated vibrational sum-frequency generation (vSFG) spectra revealed that the experimentally observed hydrophilic behavior of monolayer graphene on hydrophilic substrates arose not from wetting transparency, but from signal cancellation induced by intercalated water. Energetic analyses further showed that intercalated water molecules were thermodynamically favorable for monolayer graphene on hydrophilic substrates, but not for multilayer systems, leading to changes in the vSFG response in line with experimental observations. These results offer a mechanistic understanding of graphene-water interactions and have broad implications for the design of graphene-based interfaces and devices. 1 Introduction Since the successful mechanical exfoliation of graphene in 2004, RN592 two-dimensional (2D) materials have garnered widespread scientific interest owing to their unique physical and chemical properties, which are fundamentally distinct from those of their bulk counterparts. RN625 ; RN624 ; RN596 Graphene, a single layer of carbon atoms arranged in a hexagonal lattice with \mathrm{sp^{2}} hybridization, is one of the most representative 2D materials. RN648 It possesses a combination of remarkable properties, including high electrical and thermal conductivity, excellent optical transparency, superior chemical stability, outstanding mechanical strength, and complete impermeability to atoms and molecules. RN629 ; RN630 ; RN627 ; RN628 ; XU201873 ; RN631 These remarkable properties have enabled a wide range of applications, such as desalination of water, transparent conductive electrodes, electrical energy storage, and electrocatalysis. RN646 ; RN643 ; RN647 ; RN644 ; RN645 Notably, many of these applications involve the interaction of graphene with water or aqueous electrolytes. RN641 ; RN642 ; RN53 This interaction enforces a distinct molecular structure of water, expressed through orientation of water molecules and the hydrogen-bond network, both of which are crucial for the optimal performance of devices like proton exchange membrane fuel cells B808149M ; moilanen2008water ; xu2019transparent and nanofluidic devices sachar2021hydrogen ; joly2016strong . One of the most widely used experimental techniques for evaluating graphene-water interactions is the measurement of the water contact angle (WCA). RN609 ; RN607 ; RN608 However, WCA values are influenced by a variety of factors, including the presence of an underlying substrate, RN586 ; RN499 ; RN496 the defect density on both the graphene and the substrate, RN616 ; RN615 ; RN612 ; RN613 the number of graphene layers, RN91 ; RN496 and surface contaminants doi:10.1021/acs.jpcc.5b10492 ; RN599 ; RN614 . These factors can result in a wide range of reported WCA values, ranging from 30^{\circ} to 127^{\circ} . RN619 ; RN618 ; RN600 Given that materials with WCA below 90^{\circ} are considered hydrophilic and those above 90^{\circ} hydrophobic, there remains ongoing debate as to whether graphene is intrinsically hydrophilic or hydrophobic. BELYAEVA2020100482 Although WCA measurement is a useful tool for characterizing surface wettability, it is a macroscopic observable and does not directly reveal molecular-level details of the interfacial water structure. Therefore, a comprehensive understanding of the wettability of graphene requires molecular-scale insight into the graphene-water interactions that govern its observed WCA behavior. Vibrational sum-frequency generation (vSFG) spectroscopy RN621 ; RN620 ; RN26 ; RN15 ; RN622 is a powerful technique for probing the molecular structure of water at interfaces. RN606 ; RN103 ; RN605 ; RN604 ; RN603 ; RN602 ; RN107 As a second-order non-linear optical process, vSFG requires breaking the inversion symmetry to generate a finite signal. This requirement makes the technique inherently surface-specific, as bulk water is centrosymmetric and thus does not contribute to the optical response. Consequently, the vibrational signals detected by vSFG arise exclusively from interfacial water molecules. Experimental and simulation results from Wang et al. RN493 , along with independent experimental measurements by Xu et al. RN669 , both demonstrated that pristine graphene floating on water exhibited hydrophobic behavior. In contrast, when supported on a calcium fluoride ( \text{CaF}{\vphantom{\text{X}}}_{\smash[t]{\text{2}}} ) substrate, the measured vSFG spectra indicated a hydrophilic character of the graphene layer. RN97 ; RN551 ; RN91 ; RN525 Given the hydrophilicity of the substrate and the atomic thickness of graphene, some researchers proposed that the wettability of the underlying substrate significantly influences the overall interfacial properties of graphene and water. This phenomenon was referred to as wetting transparency and was proposed to explain the observed wettability of the graphene surface. RN499 ; RN496 However, the extent to which graphene is transparent to wetting remains a subject of ongoing debate. Graphene has been reported to be fully wetting-transparent, RN91 ; RN496 ; RN503 ; RN504 partially wetting-transparent, RN598 ; RN500 ; RN599 ; RN611 or fully wetting-opaque RN589 ; RN501 ; RN601 ; RN600 ; D0NR08843A . In addition to substrate effects, intercalated water molecules may also influence the shape of vSFG spectra. Ohto et al. RN52 simulated the vSFG spectra of pristine graphene and found that when all water molecules were located on one side of the graphene sheet, the spectrum closely resembled that of the water-air interface, consistent with the hydrophobic nature of graphene. However, the presence of even a small number of water molecules on the opposite side of the graphene layer significantly reduced the peak amplitude of the dangling O-H bond, indicating a diminished hydrophobic response. In experimental work, Temmen et al. RN515 ; RN74 used non-contact atomic force microscopy and Kelvin probe force microscopy to characterize the thickness of graphene sheets and the intercalated water layer. They found that even after annealing to 750 K, it was nearly impossible to completely remove the water molecules trapped between graphene and the substrate \text{CaF}{\vphantom{\text{X}}}_{\smash[t]{\text{2}}} . Li et al. RN528 investigated the behavior of intercalated water by calculating the migration energy barrier of a single water molecule confined between monolayer graphene and substrate \text{CaF}{\vphantom{\text{X}}}_{\smash[t]{\text{2}}} , using the climbing image nudged elastic band method. They reported a migration barrier of approximately 0.19 eV. They also calculated the diffusion coefficient of water at 300 K, finding it to be 8.35 × 10^{-6} cm 2 /s. These results, a low migration barrier, a high diffusion rate, and the persistence of water even after high-temperature annealing, collectively suggest the presence of intercalated water between graphene and the \text{CaF}{\vphantom{\text{X}}}_{\smash[t]{\text{2}}} substrate. Beyond substrate effects and intercalated water, the number of graphene layers can also influence the experimentally measured vSFG spectra. Kim et al. RN91 systematically investigated the WCA and vSFG spectra of multilayer graphene. Their vSFG measurements revealed that when the number of graphene layers exceeded four, the vSFG spectra exhibited pronounced changes. These changes were attributed to the transition of the graphene film from wetting transparent to translucent and finally to opaque as the number of layers increases from one to six. However, the fundamental mechanisms underlying these dramatic changes in the vSFG response with increasing graphene thickness were not fully understood. Although experimental vSFG provides valuable information about interfacial water structure, it remains challenging to control atomically the interface to disentangle the various factors, such as graphene-water interactions, graphene layer effects, substrate influence, and water penetration. To achieve a comprehensive microscopic understanding, atomic-scale simulations are

Read the complete text