Volcano Plot Generation

Overview

Volcano plots visualize the Sabatier principle: plotting catalytic activity (negative overpotential) against a binding energy descriptor to identify optimal catalysts at the peak of the volcano. This is the standard tool for computational catalyst screening.

Key applications:

• OER/HER/ORR catalyst screening: Compare overpotentials across catalyst compositions
• Scaling relation validation: Overlay theoretical volcano lines from Norskov scaling
• Descriptor identification: Find which adsorption energy best predicts activity
• High-throughput screening: Visualize hundreds of candidates in one plot

MCP Tool: catgo_catalysis action="volcano"

Generate Volcano Plot Data

Provide a list of catalyst results with descriptor values and overpotentials:

{"tool": "catgo_catalysis", "arguments": {
  "action": "volcano",
  "params": {
    "catalyst_results": [
      {"name": "RuO2(110)", "dG_OH": 1.45, "overpotential": 0.37},
      {"name": "IrO2(110)", "dG_OH": 1.52, "overpotential": 0.42},
      {"name": "MnO2(110)", "dG_OH": 0.95, "overpotential": 0.68},
      {"name": "TiO2(110)", "dG_OH": 2.10, "overpotential": 1.15},
      {"name": "Fe-NiOOH", "dG_OH": 1.30, "overpotential": 0.32}
    ],
    "reaction": "OER",
    "descriptor_x": "dG_OH"
  }
}}

Custom Descriptor Axes

Use any computed property as the x-axis descriptor:

{"tool": "catgo_catalysis", "arguments": {
  "action": "volcano",
  "params": {
    "catalyst_results": [
      {"name": "Pt(111)", "d_band_center": -2.25, "overpotential": 0.45},
      {"name": "Pd(111)", "d_band_center": -1.83, "overpotential": 0.52},
      {"name": "Ni(111)", "d_band_center": -1.29, "overpotential": 0.75}
    ],
    "reaction": "HER",
    "descriptor_x": "d_band_center"
  }
}}

Two-Descriptor Plot

Specify both x and y descriptors explicitly (instead of using overpotential for y):

{"tool": "catgo_catalysis", "arguments": {
  "action": "volcano",
  "params": {
    "catalyst_results": [
      {"name": "RuO2", "dG_OH": 1.45, "dG_O": 2.90},
      {"name": "IrO2", "dG_OH": 1.52, "dG_O": 3.10}
    ],
    "reaction": "OER",
    "descriptor_x": "dG_OH",
    "descriptor_y": "dG_O"
  }
}}

Parameters

Parameter	Type	Default	Description
catalyst_results	list[dict]	--	List of catalyst dicts with name, descriptor values, overpotential
reaction	string	"OER"	Reaction type: OER, HER, CO2RR, NRR
descriptor_x	string	"dG_OH"	Key for x-axis descriptor in result dicts
descriptor_y	string	null	Key for y-axis. If null, uses -overpotential

Catalyst Result Dict Fields

Each dict in catalyst_results should contain:

Field	Required	Description
name	yes	Catalyst identifier (plot label)
(descriptor_x key)	yes	X-axis value (e.g., dG_OH, d_band_center)
overpotential	yes*	Overpotential in V (*unless descriptor_y is set)

Return Format

{
  "points": [
    {"name": "RuO2(110)", "x": 1.45, "y": -0.37, "dG_OH": 1.45, "overpotential": 0.37}
  ],
  "ideal_line": {
    "x": [0.5, 0.505, ...],
    "y": [-0.23, -0.22, ...]
  },
  "descriptor_x": "dG_OH",
  "reaction": "OER"
}

The ideal_line is generated for OER using Norskov scaling relations:

• Left branch: limited by OH adsorption (step 1)
• Right branch: limited by OOH formation (step 4), using the scaling relation dG_OOH = 0.84 * dG_OH + 3.29

For other reactions, ideal_line is null (scaling relations not hard-coded).

Complete Workflow: OER Catalyst Screening

1. Compute overpotentials for each candidate

For each catalyst surface, run the full OER workflow (see OER skill) to obtain dG_OH, dG_O, dG_OOH, and the overpotential.

2. Collect results

Gather the results from all candidates into a list:

{"tool": "catgo_catalysis", "arguments": {
  "action": "oer",
  "params": {"dG_OH": 1.45, "dG_O": 2.90, "dG_OOH": 3.74}
}}

Repeat for each catalyst.

3. Generate volcano plot

{"tool": "catgo_catalysis", "arguments": {
  "action": "volcano",
  "params": {
    "catalyst_results": [
      {"name": "RuO2", "dG_OH": 1.45, "overpotential": 0.37},
      {"name": "IrO2", "dG_OH": 1.52, "overpotential": 0.42}
    ],
    "reaction": "OER",
    "descriptor_x": "dG_OH"
  }
}}

Common Pitfalls

1. All descriptor values must use consistent DFT settings (same functional, ENCUT, k-points). Mixing PBE and RPBE results on one volcano plot produces misleading comparisons.
2. The OER ideal volcano line assumes the universal OOH-OH scaling relation (dG_OOH = 0.84 * dG_OH + 3.29). This may not hold for non-oxide catalysts.
3. The y-axis convention is -overpotential (higher = better catalyst). A catalyst at the peak of the volcano has the lowest overpotential.
4. Catalyst results missing the descriptor_x key are silently skipped. Check that all result dicts have the expected keys.
5. For HER, the typical descriptor is dG_H (hydrogen binding energy). For CO2RR, dG_CO or dG_COOH is commonly used.