chem-dft-orca-optimization

Perform DFT geometry optimization on molecular structures using ORCA.

144|21|Updated Jan 8, 2026
One-click install
npx skills add https://github.com/learningmatter-mit/AtomisticSkills --skill chem-dft-orca-optimization
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: chem-dft-orca-optimization
Source: https://github.com/learningmatter-mit/AtomisticSkills/tree/main/.agents/skills/chem-dft-orca-optimization
Command: npx skills add https://github.com/learningmatter-mit/AtomisticSkills --skill chem-dft-orca-optimization

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

Chemists and computational researchers need a reliable way to optimize molecular geometries at the DFT level, both for energy minimization and for locating transition-state saddle points, without manually managing low-level optimizer workflows.

Core Features & Use Cases

  • Geometry minimization and single-ended TS search: Runs ORCA optimizations in either minimization mode or single-ended transition state mode (TS optimization).
  • SCINE/ReaDuct wrapper orchestration: Uses the SCINE/ReaDuct stack to manage optimization steps, convergence, and (optionally) Hessian evaluation at the end.
  • Reproducible structured outputs: Produces optimized geometries and a results JSON including energies, force metrics, and optional Hessian-derived diagnostics (including imaginary mode counting for TS verification).
  • Use Case: Optimize an incoming molecular structure to find a local minimum, then verify a candidate TS by requesting the final Hessian and checking that exactly one imaginary mode is present.

Quick Start

Use the chem-dft-orca-optimization skill to optimize your structure file with DFT by providing a structure path, choosing opt_type min or ts, and setting ORCA options like functional, basis set, and nprocs.

Frequently Asked Questions about chem-dft-orca-optimization

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run a DFT geometry optimization on a molecular structure using ORCA?

To run a DFT geometry optimization, you provide an ASE-readable input structure file along with parameters like charge, spin multiplicity, functional, and basis set. The task uses the SCINE/ReaDuct wrapper to execute ORCA and returns optimized geometries with energy results.

Can I find a transition state with ORCA using a single-ended search?

Yes, you can perform a single-ended transition state search by setting the optimization type to 'ts'. The process runs ORCA through the SCINE/ReaDuct wrapper and optionally evaluates the final Hessian to verify the saddle point by counting imaginary modes.

How do I verify a candidate transition state geometry using Hessian analysis?

Hessian analysis verifies a candidate transition state by checking that exactly one imaginary mode is present in the final output. You request this Hessian evaluation at the end of the ORCA transition state optimization via the SCINE/ReaDuct wrapper.

What do I need to set up an orca-agent environment for DFT calculations?

Setting up an orca-agent environment requires an installed ORCA executable and the ORCA_BINARY_PATH variable configured to point to it. You also need an ASE-readable input structure file and parameters such as charge, spin multiplicity, functional, and basis set.

Does the SCINE/ReaDuct wrapper support force and Hessian-based convergence diagnostics?

Yes, the SCINE/ReaDuct wrapper supports force and Hessian-based convergence diagnostics for ORCA optimizations. It manages optimization steps, tracks energy minimization or transition state convergence, and produces structured JSON outputs with force metrics and imaginary mode counts.

Why is my ORCA transition state optimization not converging properly?

ORCA transition state optimization convergence issues often relate to insufficient force metrics or incorrect parameter settings like functional and basis set choices. The SCINE/ReaDuct wrapper manages these convergence limits and provides structured JSON outputs to help diagnose the problem.