H2O2_interfacial_water

H₂O₂ production at the interfacial water confined in waterdroplet-solid contacting regime, investigated by MLIP-AIMD across multiphase scales.

This repository contains the computational workflow and datasets for investigating the spontaneous generation of hydrogen peroxide (H₂O₂) at the hydrophobic polymer (PDMS) and water interface. The mechanism involves contact electrification (Contact Electro-Catalysis) and the resulting built-in interfacial electric field, studied through a multiscale combination of Machine Learning Interatomic Potentials (MLIP) and Ab Initio Molecular Dynamics (AIMD).

---

🔬 Scientific Background

Experimental evidence shows that contact between water microdroplets and hydrophobic polydimethylsiloxane (PDMS) surfaces can spontaneously generate H₂O₂. This project uses first-principles and MLIP simulations to model:

1. Contact Electrification: Spontaneous electron/ion transfer across the solid-liquid interface.
2. Built-in Electric Field (BIEF): Formation of a highly localized electric double layer.
3. Reaction Kinetics: Reduction in the activation energy for H₂O oxidation and •OH radical recombination under strong interfacial polarization.

---

📁 Repository Structure & Workflow

The simulation workflow is divided into five sequential phases, mimicking the experimental physical process from bulk materials to the electrified interface:

1.ModelBuilding/

Generation of the initial atomic topology and coordinates for both the PDMS polymer chains and the bulk water system.

2.Prerelax_Silicone_slab/

Classical and MLIP-driven pre-relaxation of the highly crosslinked PDMS slab.

• Engine: LAMMPS
• Potential: MACE-MH-1 (omol pre-trained model optimized for organic polymers)
• Protocol: NPT/NVT melting-quenching and densification protocol to achieve realistic polymer density.

3.Prerelax_MatchingWater_Box/

Pre-relaxation of the water box matching the lateral dimensions of the PDMS slab.

• Engine: LAMMPS
• Protocol: NPT equilibration at 298 K and 1 atm to obtain realistic liquid water density and hydrogen bond networks.

4.Assembly_Silicone_WaterBox/

Automated assembly scripts (assemble_interface.py) to pack the relaxed water box onto the densified PDMS slab, creating the final .xyz starting structure (assembled_interface.xyz) with appropriate periodic boundary conditions.

5.AIMD_CP2K_WHU-HPC/

First-principles production runs executed on the WHU High-Performance Computing cluster.

• Engine: CP2K (QS module)
• Functional: r2SCAN (Meta-GGA) + rVV10 (non-local dispersion correction)
• Configuration: UKS (Unrestricted Kohn-Sham) to allow for spontaneous radical formation.
• Sub-tasks:
  • no-field/: Healthy zero-field AIMD trajectory capturing spontaneous electron transfer and intrinsic built-in electric field (evaluated via Hartree potential slicing).
  • with-field_deprecated/: Archive of external displacement field testing ($3.6 \times 10^9$ V/m). Deprecated due to numerical field emission in the vacuum layer.

---

🚀 Quick Start & Usage

1. Environment Setup: Ensure you have LAMMPS (with MACE plugin) and CP2K installed. Python 3.x is required for assembly scripts.

2. Pre-trained Models: The MACE-MH-1 model (.pt) for pre-relaxation is located in the 3.Prerelax_MatchingWater_Box/water_box_prerelaxation/mace_pretained_models/ directory.

3. Running AIMD: Submit the CP2K job on your HPC using the provided SLURM scripts:

   cd 5.AIMD_CP2K_WHU-HPC/regular_production_without-O2/no-field/
   sbatch run_cp2k_aimd.slurm

---

📊 Next Steps (Ongoing)

• Bader Charge Analysis: Analyzing no-field trajectory snapshots to quantify cross-interface charge transfer.
• Hartree Potential Profiling: Extracting the exact intensity of the spontaneous built-in electric field $E_{\text{built-in}}$ at the PDMS-water interface.
• Metadynamics (Plumed): Reconstructing the free-energy surface ($\Delta G$) for H₂O₂ generation pathways.