Atomistic Materials Modeling Expert

What you'll do

Mercor wants people who have spent years setting up and debugging atomistic simulations to mark an AI model's homework in exactly that field. The model produces answers involving DFT calculations, molecular dynamics runs and materials property predictions; you decide whether the method, the parameters and the physics behind them hold up.

The ad's scope breaks into four kinds of judgment:

• Method choice: is the functional, the level of theory or the simulation approach appropriate for the question
• Setup: simulation parameters, convergence criteria and the physical assumptions baked into them
• Results: atomic structures, electronic properties, reaction energetics, diffusion, stability, defect formation, adsorption energies and surface or interface behaviour
• Reasoning: where the scientific argument goes wrong even when the numbers look plausible

Alongside the review, you write reference solutions, evaluation rubrics and rigorous explanations. That second half matters: a rubric for a DFT question has to separate "used a sensible k-point mesh and checked convergence" from "quoted a plausible number", and only someone who has been burned by an unconverged calculation writes that criterion well.

The ad does not say you will run calculations for the project. The output it describes is evaluation, reference answers and rubrics.

Who fits

• A PhD in computational materials science, materials engineering, computational chemistry, condensed matter physics, chemical engineering or a close field; an exceptional MSc with substantial experience may be considered
• Hands-on work with at least one of DFT, ab initio MD, classical MD, Monte Carlo, or first-principles materials modelling generally
• Working fluency in a code such as VASP, Quantum ESPRESSO, CP2K, GPAW or LAMMPS
• A solid grasp of electronic structure, interatomic interactions, thermodynamics and statistical mechanics
• Experience interpreting simulation output and validating predictions
• Strong Python for scientific computing or workflow automation

Preferred, per the ad: ASE or pymatgen, surface and interface modelling, heterogeneous catalysis, semiconductors, high-throughput screening, machine-learned interatomic potentials, publications, and experience tying characterisation data to simulation.

What it pays and the commitment

A flat $84 per hour, Mercor's published figure. The commitment is a minimum of 10 hours a week, with room for up to 40. Payment is weekly via Stripe or Wise, and Mercor cannot support H-1B or STEM OPT candidates.

Worth knowing

Good:

• No residence requirement is published, unlike Mercor's US-only Atomistic & Surface Modeling Experts listing at the same $84 rate; if you are outside the US, this is the one to look at
• 10 hours a week is a low floor for an expert contract, which makes it workable alongside a postdoc or industry job
• Machine-learned potentials experience counts as a plus, so people who sit between simulation and ML have an edge
• Weekly payment terms are stated on the listing

Less good:

• $84 flat, with no band or progression mentioned
• The PhD bar is firm in practice; the MSc route is described as exceptional
• H-1B and STEM OPT holders are excluded
• Independent contractor: projects can be extended, shortened or ended early, and no duration is given

About this listing

Posted by Mercor as a remote hourly contract and read from the live listing on 12 October 2026, when it was marked as a new opportunity. Pay, hours and qualifications above are the ad's own; the end client is not named. See Mercor.