Erstellt am 20. Juli 2026
PostDoc-Computational Materials Science & Scientific Software Engineering all genders /Part-Fulltime
Merck Group
Darmstadt, Germany
Vollzeit
Reference: 2115212696
Your role: Post Doc - Computational Materials Science & Scientific Software Engineering
You will join an MLIP-powered computational platform for semiconductor materials discovery, working at the intersection of computational chemistry, scientific software engineering, and modern AI. You will design and implement production-grade Python workflows that connect first-principles calculations with machine-learned interatomic potentials to accelerate materials screening and deepen process understanding. You will develop end-to-end simulation pipelines, including slab generation, adsorption energy screening, and molecular dynamics, ensuring code is modular, tested, and well-documented. You will run and analyze DFT calculations with Quantum ESPRESSO and VASP via ASE, generating high-quality training data for MLIPs and validating results against experimental benchmarks. You will evaluate and deploy MLIP frameworks (such as MACE or UMA), building robust training pipelines, validation protocols, and model-selection workflows. You will implement cheminformatics steps for molecular input preparation, SMILES handling, 3D conformer generation, binding site identification, and NEB-based transition-state searches, linking gas-phase properties to surface workflows. You will operate on HPC infrastructure with SLURM, containerization, and workflow orchestration to ensure reproducibility and scalability. You will translate domain expert requirements into maintainable, production-grade software and contribute to coding standards, reviews, and CI/CD practices. You will stay curious about AI tooling and be ready to integrate new approaches into scientific workflows. You will collaborate across disciplines, communicating clearly with experimentalists, data scientists, and external partners while delivering tangible software that ships.
Who you are:
EL-CT-S Materials AI
Expert 2
You will join an MLIP-powered computational platform for semiconductor materials discovery, working at the intersection of computational chemistry, scientific software engineering, and modern AI. You will design and implement production-grade Python workflows that connect first-principles calculations with machine-learned interatomic potentials to accelerate materials screening and deepen process understanding. You will develop end-to-end simulation pipelines, including slab generation, adsorption energy screening, and molecular dynamics, ensuring code is modular, tested, and well-documented. You will run and analyze DFT calculations with Quantum ESPRESSO and VASP via ASE, generating high-quality training data for MLIPs and validating results against experimental benchmarks. You will evaluate and deploy MLIP frameworks (such as MACE or UMA), building robust training pipelines, validation protocols, and model-selection workflows. You will implement cheminformatics steps for molecular input preparation, SMILES handling, 3D conformer generation, binding site identification, and NEB-based transition-state searches, linking gas-phase properties to surface workflows. You will operate on HPC infrastructure with SLURM, containerization, and workflow orchestration to ensure reproducibility and scalability. You will translate domain expert requirements into maintainable, production-grade software and contribute to coding standards, reviews, and CI/CD practices. You will stay curious about AI tooling and be ready to integrate new approaches into scientific workflows. You will collaborate across disciplines, communicating clearly with experimentalists, data scientists, and external partners while delivering tangible software that ships.
Who you are:
- PhD in Computational Chemistry, Quantum Chemistry, Materials Science, Physics, or a closely related field with a strong computational component.
- Solid grounding in quantum chemistry and surface science, including DFT, thermodynamics/kinetics, slab models, and periodic boundary conditions.
- Hands-on experience with DFT codes (Quantum ESPRESSO, VASP) and ASE as a simulation interface, plus familiarity with MLIPs (MACE, UMA, or equivalent) and training-data pipelines.
- Basic cheminformatics skills (SMILES handling, 3D conformer generation with RDKit/CREST, binding-site identification, NEB transition-state searches).
- Production-grade Python software engineering: type hints, docstrings, pytest, linting, modular design, and strong version control with Git and CI/CD.
- HPC proficiency: SLURM/PBS, array jobs, Apptainer/Docker, and workflow orchestration (Snakemake, Prefect, or equivalent).
- Ability to translate research ideas into robust, well-documented code and to work at the research-engineering interface.
- Strong communication skills and a collaborative mindset, comfortable working with cross-disciplinary teams and external partners.
EL-CT-S Materials AI
Expert 2
