materials-screening

name	materials-screening
description	材料科学筛选技能。用于固态电解质、催化剂等材料的候选筛选、 属性预测和多目标优化排序。触发词包括"材料筛选"、"materials screening"等。
triggers	["材料筛选","materials screening","固态电解质","solid electrolyte","材料发现","materials discovery"]
domain	materials_science
version	1.0

Materials Screening Skill

Overview

AI-powered materials discovery workflow for screening candidate materials based on target properties and constraints.

Trigger Conditions

• User mentions materials screening, discovery
• Solid-state electrolyte research
• Catalyst design tasks

Workflow

Step 1: Define Objectives

Specify target properties and constraints:

• Ionic conductivity (max)
• Formation energy (min)
• Band gap (range)
• Stability threshold

Step 2: Search Candidates

Query materials databases:

• Materials Project
• AFLOW
• OQMD

Script: scripts/search_candidates.py

See: STRUCTURE_FAMILIES.md for common structure families

Step 3: Property Prediction

Use ML models for property prediction:

• GNN-based property prediction
• Formation energy estimation
• Stability assessment

Script: scripts/predict_properties.py

Step 4: Ranking & Selection

Multi-objective optimization:

• Pareto frontier analysis
• Weighted scoring
• Constraint filtering

Script: scripts/rank_materials.py

Output Format

{
  "query": "solid electrolyte Li-ion",
  "candidates": [
    {
      "formula": "Li7La3Zr2O12",
      "structure_family": "garnet",
      "predicted_conductivity": 1e-3,
      "formation_energy": -3.2,
      "stability": "stable",
      "rank": 1
    }
  ],
  "total_screened": 1000,
  "passed_constraints": 50
}