AIRSS

Official Resources

• Homepage: https://www.mtg.msm.cam.ac.uk/Codes/AIRSS
• Documentation: https://airss-docs.github.io/
• Source Repository: Distributed with license
• License: Academic License (Open source components available)

Overview

AIRSS is a simple yet powerful method and software package for predicting crystal structures. The core idea is to generate random structures (sensibly constrained by density, symmetry, and distances) and relax them using ab-initio forces. Developed by Chris Pickard and collaborators, AIRSS is robust, easy to parallelize, and effective for a wide range of materials, especially under high pressure.

Scientific domain: Random structure searching, crystal prediction, materials discovery
Target user community: Computational materials scientists, high-pressure physicists

Theoretical Methods

• Random Structure Searching (RSS)
• Density functional theory relaxation
• Symmetry constraints (random space groups)
• Species swapping
• Shaking (for finding nearby minima)

Capabilities (CRITICAL)

• Prediction of crystal structures from composition
• High-pressure phase prediction
• Point defect structure searching
• Surface and interface reconstruction
• Cluster structure search
• Constraint handling (distances, coordination)
• Integration with CASTEP (primary), VASP, Quantum ESPRESSO

Sources: AIRSS website, J. Phys.: Condens. Matter 23, 053201 (2011)

Inputs & Outputs

• Input formats: airss.pl command line arguments, seed files
• Output data types: Relaxed structures (.res, .cell), energy rankings

Interfaces & Ecosystem

• CASTEP: Deep integration and optimization
• Command line: Script-based workflow (airss.pl, castep_relax, etc.)
• Analysis: Tools for clustering and analyzing found structures

Workflow and Usage

1. Generate random structures: airss.pl -seed ...
2. Relax structures: Submit to DFT code (e.g., CASTEP)
3. Collect results: Rank by enthalpy
4. Analyze: Identify unique structures and ground state

Performance Characteristics

• Embarrassingly parallel (each random search is independent)
• High throughput
• Robust exploration of energy landscape
• Scales well to HPC clusters

Application Areas

• High-pressure hydrides and oxides
• Defect structures in semiconductors
• Grain boundaries
• Battery materials
• Molecular crystals

Community and Support

• Developed at University of Cambridge / UCL
• Active research use
• Workshops and tutorials

Verification & Sources

Primary sources:

1. Homepage: https://www.mtg.msm.cam.ac.uk/Codes/AIRSS
2. Documentation: https://airss-docs.github.io/
3. Publication: C. J. Pickard and R. J. Needs, J. Phys.: Condens. Matter 23, 053201 (2011)

Confidence: VERIFIED

Verification status: ✅ VERIFIED

• Website: ACTIVE
• Documentation: COMPREHENSIVE
• Method: STANDARD (RSS)
• Development: ACTIVE (Pickard Group)
• Applications: Random structure searching, high pressure, defects, CASTEP integration