"Precursor Reaction Pathway Leading to BiFeO3 Formation: Insights from Text-Mining and Chemical Reaction Network Analyses "

"Precursor Reaction Pathway Leading to BiFeO3 Formation: Insights from Text-Mining and Chemical Reaction Network Analyses "

Abstract

BiFeO3 (BFO) is an emerging non-toxic multiferroic perovskite material with applications in sensors, memory devices, and spintronics. The crystallinity and structure of BFO critically influence its functional properties. Achieving phase-pure BFO through sol-gel synthesis is challenging due to gaps in our understanding of the interactions between metal complexes in precursor solutions. In this talk, I will discuss how combining text-mining with chemical reaction network (CRN) analysis provides new insights into the chemistry of BFO sol-gel precursors. Our text-mining analysis of 340 synthesis recipes reveals trends in precursor use, showing a preference for nitrates as metal salts, 2-methoxyethanol (2ME) as the dominant solvent, and the frequent use of citric acid as a chelating agent to achieve phase-pure BFO. CRN analysis indicates that the thermodynamically favored interaction between bismuth nitrate and 2ME involves partial solvation followed by dimerization, challenging previous assumptions. This dimerization likely seeds the final phase, and further oligomerization, facilitated by nitrite ion bridging, is crucial for forming the desired BFO crystal structure. A deeper examination of dimer-related phenomena reveals that surfactants hinder oligomerization and that solvents should stabilize de-nitrated complexes. This study highlights the importance of selecting appropriate precursor materials and treatment conditions to encourage oligomerization and prevent side reactions that compromise crystal purity.

 

Speaker

Viktoriia Baibakova

I am a PhD candidate in the Hacking Materials(link is external) group at the Energy Technologies Area, supervised by Dr. Anubhav Jain and Professor Gerbrand Ceder. My research focuses on applying artificial intelligence methods and ab initio calculations to materials science-related tasks. I have also worked on projects that apply computer vision to extract raw data from scientific plots and use large language models for generating crystal structures. This is my final semester at UC Berkeley, and I am currently writing my PhD dissertation. Looking ahead, I aim to continue doing research and applying AI within the materials science domain, specifically focusing on sustainability and environmental issues. I am always happy to connect on LinkedIn(link is external).

Date/Time
Monday, September 23, 2024 - 03:00pm to 03:30pm
Type
Seminar