Viktoriia Baibakova1,Kevin Cruse1,Carolin Sutter-Fella1,Anubhav Jain1,Samuel Blau1
Lawrence Berkeley National Laboratory1
Viktoriia Baibakova1,Kevin Cruse1,Carolin Sutter-Fella1,Anubhav Jain1,Samuel Blau1
Lawrence Berkeley National Laboratory1
BiFeO3 thin films have potential applications in memory devices and photovoltaics due to their multiferroic properties [1]. They can be synthesized through spin-coating of a sol-gel [2], but this process is complex and can result in impurity phases that negatively impact physical and device properties. In this study, we aim to improve our understanding of the mechanisms involved in sol-gel synthesis of BiFeO3 by using chemical reaction networks [3]. We also seek to propose a different method for managing precursor materials in synthesis recipe databases that can produce defining features effectively processed by machine learning models. We present preliminary results obtained from constructing and analyzing the reaction network for the reaction between bismuth(III) nitrate pentahydrate and 2-methoxyethanol. For instance, our study suggests that replacing the nitrites and water in the outer layer of bismuth with 2-methoxyethanol may be hindered by kinetics, but becomes easier once hydrogen bonds are weakened. This finding may help explain why the reaction does not occur at room temperature in experiments [4]. We also report our efforts to explore preferred chemical routes based on energy gain using high-throughput computations with xTB [5] and QChem [6]. Additionally, we discuss alternative representations of materials that can be used to improve the success of chemical reaction networks and contribute to a data-driven understanding of the effects of different chemical components on the outcome. Our findings contribute to a more predictive and mechanistic understanding of the sol-gel synthesis process, which could lead to the production of high-quality BiFeO3 thin films.<br/><br/>REFERENCES<br/>[1] Zhang, Qi, Daniel Sando, and Valanoor Nagarajan. "Chemical route derived bismuth ferrite thin films and nanomaterials." <i>Journal of Materials Chemistry C</i> 4.19 (2016): 4092-4124.<br/>[2] Danks, Ashleigh E., Simon R. Hall, and Z. J. M. H. Schnepp. "The evolution of ‘sol–gel’chemistry as a technique for materials synthesis." <i>Materials Horizons</i> 3.2 (2016): 91-112.<br/>[3] Barter, Daniel, et al. "Predictive stochastic analysis of massive filter-based electrochemical reaction networks." <i>Digital Discovery</i> (2022).<br/>[4] Zhang, Qi, Nagarajan Valanoor, and Owen Standard. "Epitaxial (001) BiFeO 3 thin-films with excellent ferroelectric properties by chemical solution deposition-the role of gelation." <i>Journal of Materials Chemistry C</i> 3.3 (2015): 582-595.<br/>[5] Bannwarth, Christoph, et al. "Extended tight-binding quantum chemistry methods." <i>Wiley Interdisciplinary Reviews: Computational Molecular Science</i> 11.2 (2021): e1493.<br/>[6] Shao, Yihan, et al. "Advances in molecular quantum chemistry contained in the Q-Chem 4 program package." <i>Molecular Physics</i> 113.2 (2015): 184-215.