TMD Electronic Structure Analysis using GPAW DFT

Project Overview

This repository contains the computational framework and results for the systematic study of electronic properties of group-VI two-dimensional transition metal dichalcogenides (TMDs) using density functional theory (DFT). The project was developed as part of the MEMY-512 Computational Materials Science II course at the University of Crete.

📋 For Detailed Documentation: For a comprehensive description of our development process, theoretical background, computational setup, and detailed analysis methodology, please refer to the complete project report: tmd_report.pdf. The PDF contains extensive details on our thought process, validation procedures, and in-depth discussion of results that complement this markdown.

Research Objectives

The primary goal is to calculate and analyze the electronic structure properties of TMD monolayers, specifically:

• Band structure calculations along high-symmetry k-point paths
• Total density of states (DOS) analysis
• Orbital-projected DOS (PDOS) to understand electronic contributions
• Phase-dependent electronic properties across different structural polymorphs

Materials and Phases Studied

1H Phase Materials (Semiconducting):

• MoS₂ (Molybdenum disulfide)
• MoSe₂ (Molybdenum diselenide)
• WS₂ (Tungsten disulfide)
• WSe₂ (Tungsten diselenide)
• MoTe₂ (Molybdenum ditelluride)
• WTe₂ (Tungsten ditelluride)

Extended Phase Analysis:

• 1T phase WTe₂ (Metallic octahedral coordination)
• 1T' phase WTe₂ (Distorted 1T with potential topological properties)

Repository Structure

.
├── tmd_report.pdf           ← Project description
├── tmd_pres_slides.pdf.pdf  ← Presentation slides
├── requirements.yaml        ← Conda environment  
└── TMD_codes/               ← All scripts, data & workflows  
    ├── tmd_bs.py            ← Band structure calculator  
    ├── dos_tmd.py           ← DOS & PDOS plotting module  
    ├── mx2_utils.py         ← Custom ASE MX₂ builder + helpers  
    ├── tmd_prior_runs/      ← Reference `.gpw` runs & band‐path images  
    ├── fails/               ← Divergent or unconverged calculations  
    └── MX2_<metal><X>_<p>/  ← Material-phase subdirectories  
        ├── initial/         ← “Guess” structures & input scripts  
        ├── relax/           ← Relaxation outputs (`*.gpw`, `.traj`)  
        ├── gs/              ← Ground‐state densities (`*_gs.gpw`)  
        ├── bs/              ← Band‐structure JSON + plots  
        ├── dos/             ← DOS/PDOS data & figures  
        └── img/             ← All generated figures (band paths, DOS, SOC insets)

Directory Organization

The project employs a hierarchical directory structure that mirrors the computational workflow:

• Material-specific subdirectories follow the naming convention MX2_1p, where:

  • M = transition metal (Mo, W)
  • X = chalcogen (S, Se, Te)
  • p = structural phase (H, T, T')

• Systematic file naming encodes material identity and calculation type

• Complete computational provenance from initial structural guesses to converged geometries

• Visualization consolidation in img/ subdirectories containing DOS plots, band diagrams, and orbital projections

Computational Modules

1. Band Structure Calculator (tmd_bs.py)

Performs electronic band structure calculations for TMD monolayers with support for all three structural phases.

Key Features:

• Customized ASE MX2 builder distinguishing 1H from 2H phases
• Support for distorted 1T' phase construction
• Integrated workflow: structure building → relaxation → SCF → band calculation
• High-symmetry k-point path analysis (Γ-K-M-Γ)

Usage:

# Basic band structure calculation
mpirun -np 8 python tmd_bs.py --material MoS2 --phase "1H" --nprocs 4 --domain-parallel --kpt-parallel

# Skip relaxation for pre-optimized structures
mpirun -np 8 python tmd_bs.py --material WTe2 --phase "1T'" --skip-relax --band-parallel

Supported Arguments:

• --material: TMD compound (MoS2, MoSe2, WS2, WSe2, MoTe2, WTe2)
• --phase: Structural phase (1H, 1T, 1T')
• --nprocs: MPI process count
• --domain-parallel, --kpt-parallel, --band-parallel: Parallelization options
• --skip-relax: Skip geometry optimization

2. DOS Analysis Module (dos_tmd.py)

Computes total and projected density of states from completed GPAW calculations.

Key Features:

• Total DOS with spin-channel resolution
• Orbital-projected DOS (metal d-orbitals, chalcogen p-orbitals)
• Spin-orbit coupling (SOC) analysis for heavy elements
• Band inversion analysis for 1T' phases

Usage:

# Basic DOS calculation
python dos_tmd.py mos2_1H_gs.gpw

# Full analysis with PDOS and SOC
python dos_tmd.py wte2_1T_gs.gpw --pdos --soc --width 0.05

# Custom energy range
python dos_tmd.py "mote2_1T'_gs.gpw" --emin -8 --emax 8 --resolution 0.01

Analysis Options:

• --pdos: Enable orbital-projected DOS
• --soc: Spin-orbit coupling analysis
• --width: DOS broadening (default: 0.1 eV)
• --emin/emax: Energy window relative to Fermi level

Phase-Specific Analysis

1H Phase (Semiconducting)

• Electronic character: Direct-gap semiconductors
• Critical points: Γ and K valley analysis
• Orbital contributions: Metal d-orbital and chalcogen p-orbital hybridization

1T Phase (Metallic)

• Electronic character: Metallic or semi-metallic
• Band features: d-orbital dominated conduction
• Magnetic properties: Potential spin-polarization

1T' Phase (Topological)

• Electronic character: Distorted 1T with metal-metal chains
• SOC effects: Spin-orbit driven band inversions
• Topological analysis: Band inversion around ±0.5 eV

Environment Setup

Prerequisites

• GPAW ≥ 25.1.0 with MPI support
• ASE (Atomic Simulation Environment)
• Python environment with scientific computing stack

Installation

# Create conda environment
conda env create -f requirements.yaml
conda activate tmd-dft

# Verify GPAW installation
python -c "import gpaw; print(gpaw.__version__)"

Computational Requirements

• MPI-enabled GPAW installation for parallel calculations
• Memory: Varies by system size and k-point sampling
• Time: Band structure calculations typically require 2-8 hours on modern clusters

Reproducibility and Validation

The repository is designed for complete reproducibility of results presented in tmd_report.pdf. Users can:

1. Verify existing calculations using provided reference data in tmd_prior_runs/
2. Reproduce specific results by following the systematic workflow
3. Extend to new materials using the modular calculation framework
4. Test alternative parameters guided by the analysis in fails/ directory

Data Integrity

• Complete provenance tracking from initial guesses to final results
• Systematic naming conventions for easy data management
• Benchmark comparisons through established reference calculations
• Parameter sensitivity analysis documented in failed calculation logs

Output and Visualization

The computational workflow generates:

1. Band structure plots along high-symmetry paths
2. Total DOS with spin-channel resolution
3. Orbital-projected DOS showing elemental and orbital contributions
4. SOC analysis plots for heavy-element compounds
5. Phase comparison visualizations highlighting electronic differences

Specialized Analyses

• 1H phases: Band gap characterization and valley physics
• 1T phases: Metallic band analysis and d-orbital splitting
• 1T' phases: Topological band inversion and SOC-driven transitions

Contributing and Extension

The modular design enables straightforward extension to:

• Additional TMD materials (group-V TMDs, heterostructures)
• Enhanced DFT methods (hybrid functionals (e.g., B3LYP), GW-BSE corrections)
• Advanced analysis techniques (topological invariants, transport properties)

References and Acknowledgments

This work was completed as part of the Computational Materials Science II course curriculum, utilizing the open-source GPAW DFT package and following established best practices for reproducible computational materials research.

📜 License & Citation

If you use these scripts or data, please cite:

Vourvachakis S. Georgios, “Band structure, total DOS, and orbital-projected DOS of group VI 2D TMDs…”, Department of Materials Science and Engineering, University of Crete, 2025.