First-principles study on the lithiation process of amorphous SiO anode for Li-ion batteries with Bayesian...

Area of Science:

• Materials Science
• Electrochemistry
• Computational Chemistry

Background:

• Amorphous silicon monoxide (a-SiO) is a promising anode material for lithium-ion batteries (LIBs) due to its high performance.
• Understanding atomic-scale structural changes during lithiation is crucial for optimizing a-SiO anodes.
• Existing experimental data provides limited insight into the detailed lithiation process.

Purpose of the Study:

• To investigate the lithiation process of a-SiO using first-principles simulations and machine learning.
• To develop an efficient computational method for identifying stable lithium insertion sites in amorphous structures.
• To compare a novel Bayesian optimization approach with conventional random schemes for simulating lithiation.

Main Methods:

• Developed a computational code using Bayesian optimization for efficient site identification in amorphous models.
• Employed first-principles simulations to study the lithiation process.
• Generated amorphous silicon monoxide models using neural network potentials.
• Compared Bayesian optimization with a conventional random scheme for simulating Li insertion.

Main Results:

• Bayesian optimization identified more stable lithiation structures with lower formation energies compared to the random scheme.
• Simulations showed lithium atoms preferentially inserting into the silicon phase after the silicon dioxide phase.
• Significant differences in structural evolution were observed between the two simulation schemes.
• The study confirmed experimental findings regarding the lithiation sequence.

Conclusions:

• The lithiation process significantly influences the structural transformation of a-SiO.
• The choice of simulation method impacts the predicted structural evolution and stability.
• This work provides a framework for enhancing the performance and lifetime of a-SiO based LIB anodes.