COHP Analysis

SkillDev tools

Use when the user asks about COHP (Crystal Orbital Hamilton Population), chemical bonding analysis, LOBSTER output, orbital-resolved bonding, or bonding/antibonding character between atoms.

Available today. Use it from your connected AI after setup.

Connect ahel once, and every AI you use reads what you have installed.

Then ask your AI: use the COHP Analysis skill

What this skill tells your AI

The instructions your AI receives, as published by hello-qm/catgo-lrg in .claude/skills/cohp/SKILL.md and read by ahel’s review.

Overview

Crystal Orbital Hamilton Population (COHP) analysis quantifies bonding and antibonding interactions between atom pairs in a crystal. COHP data is generated by LOBSTER from DFT output (typically VASP). CatGo reads COHPCAR.lobster files and provides interactive visualization and data extraction.

Key applications:

  • Bonding character: Identify bonding (negative COHP) vs antibonding (positive COHP) regions
  • Orbital-resolved analysis: Decompose into s-s, p-d, d-d contributions
  • Catalyst design: Understand adsorbate-surface bonding (e.g., CO-Pt d-band interactions)
  • Integrated COHP (ICOHP): Quantitative bond strength from ICOHPLIST.lobster

MCP Tool: catgo_analyze (workflow node: cohp_analysis)

COHP analysis in the Claude Code MCP uses the workflow node cohp_analysis or the dedicated catgo_cohp_data tool (in the full MCP server). The workflow engine automatically reads LOBSTER output from a parent calculation step.

Workflow-Based COHP Analysis

Add a cohp_analysis node downstream of a VASP calculation that produces LOBSTER output:

{"tool": "catgo_workflow", "arguments": {
  "action": "add_node", "workflow_id": "wf_bonding",
  "node_type": "cohp_analysis",
  "params": {"source": "parent_step"}
}}

Upload COHPCAR File Directly

If you have a COHPCAR.lobster file, upload it to create a session:

{"tool": "catgo_analyze", "arguments": {
  "action": "dos",
  "params": {"file_type": "COHPCAR", "remote_path": "/path/to/COHPCAR.lobster"}
}}

The upload returns a session_id along with metadata: nspin, npoints, efermi, and a list of available bonds with atom labels and distances.

Retrieve COHP Data for Specific Bonds

Once you have a session_id, query COHP data for specific bonds:

{"tool": "catgo_analyze", "arguments": {
  "action": "dos",
  "params": {
    "session_id": "abc-123",
    "bond_indices": [1, 3],
    "include_orbitals": false
  }
}}

Orbital-Resolved COHP

To decompose bonding into orbital contributions (e.g., p-d interactions):

{"tool": "catgo_analyze", "arguments": {
  "action": "dos",
  "params": {
    "session_id": "abc-123",
    "bond_indices": [1],
    "include_orbitals": true,
    "orbital_filter": ["p-d", "s-d"]
  }
}}

Aggregate Orbital COHP

Sum orbital contributions for a cleaner view:

{"tool": "catgo_analyze", "arguments": {
  "action": "dos",
  "params": {
    "session_id": "abc-123",
    "bond_indices": [1],
    "aggregate_orbitals": true,
    "orbital_filter": ["d-d"]
  }
}}

Parameters

ParameterTypeDescription
session_idstringSession ID from COHPCAR upload
bond_indicesint[]1-based bond numbers to retrieve (from upload metadata)
include_orbitalsboolInclude orbital-decomposed COHP (default false)
orbital_filterstring[]Filter orbital pairs, e.g. ["p-d", "s-d"]
aggregate_orbitalsboolSum filtered orbital contributions (default false)

Complete Workflow: VASP + LOBSTER COHP

1. Run VASP static calculation with LOBSTER-compatible settings

{"tool": "catgo_workflow", "arguments": {
  "action": "create", "name": "COHP bonding analysis"
}}
{"tool": "catgo_workflow", "arguments": {
  "action": "add_node", "workflow_id": "wf_cohp",
  "node_type": "single_point",
  "params": {
    "software": "vasp",
    "ENCUT": 520,
    "ISYM": -1,
    "NBANDS": 200,
    "system_name": "static for LOBSTER"
  }
}}

2. Add COHP analysis node

{"tool": "catgo_workflow", "arguments": {
  "action": "add_node", "workflow_id": "wf_cohp",
  "node_type": "cohp_analysis",
  "params": {"source": "parent_step"}
}}

3. Connect and run

{"tool": "catgo_workflow", "arguments": {
  "action": "connect",
  "workflow_id": "wf_cohp",
  "from_id": "static_node_id",
  "to_id": "cohp_node_id",
  "from_handle": "data",
  "to_handle": "data"
}}

Return Format

COHP data is returned as:

{
  "energies": [-10.0, -9.95, ...],
  "series": [
    {
      "label": "Fe1-O2 (2.01 A)",
      "spin_up": [0.12, 0.15, ...],
      "spin_down": [-0.08, -0.10, ...],
      "bond_index": 1,
      "is_total": true
    }
  ],
  "efermi": -2.35
}

Negative COHP values indicate bonding interactions; positive values indicate antibonding.

Common Pitfalls

  1. LOBSTER requires specific VASP settings: ISYM = -1 (no symmetry), sufficient NBANDS, and LWAVE = .TRUE. for wavefunction output.
  2. Bond indices are 1-based (matching LOBSTER convention). Index 0 is the "Average" across all bonds.
  3. Large COHPCAR files (>100 MB) may take several seconds to parse. The session is cached for 30 minutes after last access.
  4. For spin-polarized calculations, both spin_up and spin_down channels are returned. Non-magnetic calculations have spin_down = null.
  5. ICOHP (integrated COHP) from ICOHPLIST.lobster provides a single number per bond -- use it for quick bond strength comparisons without needing the full energy-resolved COHP.

Signals

GitHub stars
196
Forks
23
Last commit
Sep 2026
Advanced
Catalog kind
skill
Gateway key
cohp-analysis
Source
github.com/hello-qm/catgo-lrg