Platinum Nanoparticle Database: 11,400 Structures from Two Generation Protocols

11,400 platinum nanoparticle structures generated via two distinct computational protocols, 'cooking' and 'crystallization'. Jan Kloppenburg from Aalto University created this database using a Gaussian Approximation Potential (GAP) interatomic potential for platinum. The data is provided in ASE's extended XYZ format for use in materials science and machine learning research.

Use Cases

• Training machine learning interatomic potentials based on the provided nanoparticle geometries and energies.
• Benchmarking molecular dynamics simulation methods using the two distinct generation protocols described.
• Studying the structural properties of platinum nanoparticles with sizes between 10 and 349 atoms.
• Validating computational 'cooking' and 'crystallization' protocols for nanostructure generation.

Strengths

• Contains 11,400 total structures (8,000 from 'cooking' and 3,400 from 'crystallization' protocols).
• Provides atomic structures generated with a specific, documented Gaussian Approximation Potential (GAP).
• Includes two distinct generation methods with detailed temperature and time parameters (e.g., annealing at 1500 K for 20 ps).

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count for each individual file is known, but the total data size and specific features per row are not detailed.
• Last update date is unknown; freshness unverified.

Provenance

Source

Aalto University

Collection Method

Computationally generated using a Gaussian Approximation Potential (GAP) interatomic potential via 'cooking' and 'crystallization' protocols.