Finding Reaction Pathways and Transition States: r-ARTn and d-ARTn as an Efficient and Versatile Alternative to...

Area of Science:

• Computational materials science
• Chemical physics
• Theoretical chemistry

Background:

• Characterizing transition states and diffusion pathways is crucial for understanding materials evolution and chemical reactions.
• High computational costs of *ab initio* methods hinder energy landscape exploration.

Purpose of the Study:

• To reduce the computational cost of finding transition states and diffusion pathways.
• To propose three adapted versions of the Activation-Relaxation technique (ARTn): ARTn, refining ART (r-ART), and directed ART (d-ART).

Main Methods:

• Revisiting and adapting the Activation-Relaxation technique (ARTn) for *ab initio* calculations.
• Developing ARTn for exploring energy landscapes from a single minimum.
• Developing r-ART for identifying transition states between two minima.
• Developing d-ART for reconstructing pathways between two states.

Main Results:

• The adapted ARTn methods efficiently explore complex material energy landscapes.
• Demonstrated applications on benchmark examples and complex silicon defects.
• Achieved 2-6 times speedup compared to climbing image nudged elastic band (CI-NEB) method with higher precision.

Conclusions:

• The improved ARTn algorithms offer a computationally efficient and accurate approach for characterizing transition states and diffusion pathways.
• These methods are particularly beneficial for complex systems like defects in silicon.
• ARTn variants provide a powerful alternative to existing string methods for materials and chemical reaction pathway analysis.