znin2s4-computational

Computational workflows and analysis supporting the study of precursor-derived ZnIn2S4 thin-film architectures for photocatalysis.

Overview

This repository is the computational component of an Imperial College London final-year chemistry project. It investigates how crystal structure, surface chemistry, charge management, and interfaces influence photocatalytic reduction chemistry in ZnIn2S4 thin films.

The intended device architecture is NiOx / ZnIn2S4 / Co-Pi. The computational programme compares pristine ZnIn2S4, NiOx/ZnIn2S4, ZnIn2S4/Co-Pi, and NiOx/ZnIn2S4/Co-Pi to determine how buried-interface and reaction-facing modifications affect charge separation and surface reactivity.

Research Questions

• How do ZnIn2S4 polymorphs and polytypes change structural stability and electronic properties?
• How strongly do surface and facet choices affect adsorption and reaction chemistry?
• Can H2O2 formation through two-electron oxygen reduction serve as a preliminary probe of reduction-side photogenerated-electron utilisation?
• Which early steps control CO2 activation and reduction on ZnIn2S4 surfaces?
• Does a buried NiOx interface improve hole management without introducing detrimental trap states?
• Does reaction-facing Co-Pi improve hole management while preserving access to reduction sites?
• Are bottom-interface and top-surface engineering complementary, redundant, or competitive?

Computational Scope

The planned scope includes bulk ZnIn2S4 structure benchmarking, surface models, adsorption and reaction intermediates, NiOx/ZnIn2S4 interfaces, Co-Pi/ZnIn2S4 models, and charge-density and electronic-structure analysis. Supporting molecular calculations may be added where they provide a useful complement to periodic models.

The current periodic electronic-structure work uses VASP. Structure generation, crystallographic validation, and post-processing may additionally use tools such as Python, spglib, pymatgen, ASE, Bader analysis, LOBSTER, and VASPsol; molecular calculations may use ORCA when introduced. Mention of these tools describes the intended project scope and does not imply that every workflow has been completed.

Current Work

The implemented work is currently concentrated in WP1, the ZnIn2S4 polymorph/polytype benchmark. The starting structure set contains:

• experimental cubic spinel ZnIn2S4;
• ordered alpha1 derived from the experimentally disordered IIIa parent;
• reconstructed beta derived from experimental IIb;
• legacy experimental IIa as the starting structure for an unconstrained IIa to IIa-prime relaxation;
• experimental IIb.

The repository includes the source CIFs, reconstructed structures, the alpha1 ordering script, structure-provenance documentation, and PBE-D3(BJ) bulk geometry-optimisation inputs and retained outputs for the five WP1 structures. Later surface, reaction, and interface work described in the roadmap remains planned unless corresponding files are added.

Repository Structure

.
├── README.md
├── LICENSE
└── vasp/
    ├── ROADMAP.md
    └── wp1_polymorph_polytype_benchmark/
        ├── calculation/   # WP1 bulk inputs and retained outputs
        ├── script/        # structure-generation code
        └── structure/     # raw structures and provenance records

Computational Methodology

The current WP1 bulk inputs use VASP with PBE, D3(BJ) dispersion correction, PAW datasets, a 500 eV plane-wave cutoff, and full optimisation of atomic positions and lattice vectors. The electronic and ionic convergence targets are 1 x 10^-6 eV and 0.01 eV/angstrom, respectively, with structure-dependent Gamma-centred k-point meshes.

These settings document the present WP1 calculations rather than a universal prescription for future surface, interface, magnetic, charged, or molecular models. Method-specific convergence and validation are required as the project expands.

Reproducibility

• Experimental source structures retain their database and literature provenance.
• Reconstructed structures are distinguished from directly experimental structures and are generated by scripts where available.
• Curated structures, computational inputs, scripts, and scientifically useful outputs of manageable size are version controlled.
• Licensed VASP POTCAR files and large regenerable runtime artefacts are excluded from Git.

Status

This is an active research workspace for the 2026–2027 Imperial College London final-year project. Repository contents reflect work completed to date and should not be interpreted as completion of the full roadmap.

License

Original code and documentation in this repository are released under the MIT License. Source structures from the Crystallography Open Database (COD) retain their CC0/public-domain status and associated provenance. VASP outputs are provided as generated research data; neither VASP itself nor licensed POTCAR datasets are distributed here.

Author

Angze Li
MSci Chemistry, Imperial College London

GitHub: @angzeli