- 인공지능 & 에너지 소재 연구실 (Artificial Intelligence & Energy Materials Group)
- 02-705-8479(R607)
- sback@sogang.ac.kr
- https://seoinback.owlstown.net/
- 한양대학교 공학사(2011)
KAIST 공학박사(2017)
백서인Seoin Back
Research Areas
Artificial Intelligence, Computational Chemistry, Energy Materials, Catalysis, Batteries
Research Interests
- Fundamentals of Energy Materials
To develop materials design strategies for energy applications, it is essential to understand chemical and physical processes in atomic-scale. Using density functional theory (DFT) and quantum chemistry (QC) calculations, we simulate elementary reaction steps (e.g., adsorption/desorption, coupling/dissociation, (de)protonation, (de)intercalation) and evaluate their energetics/kinetics. Based on the fundamental understanding of the chemical processes, we establish descriptors for predicting properties to enable an efficient materials discovery. We also collaborate with multiple experimental groups to provide atomic-scale interpretations for the observed materials properties.
- Artificial Intelligence for Materials Discovery
In the current approach toward materials discovery, candidate materials are chosen from databases using human intuition, and their properties are computationally predicted. Thus, it is practically impossible to discover innovative materials via the current approach since databases do not contain such materials and computational predictions of all the deposited materials require too many resources. To overcome these limitations, our group is working on developing an inverse design strategy using machine learning. We are also interested in applying the developed strategy to discover new materials for energy applications including, but not limited to, fuel cells, water electrolyzer and batteries.
To develop materials design strategies for energy applications, it is essential to understand chemical and physical processes in atomic-scale. Using density functional theory (DFT) and quantum chemistry (QC) calculations, we simulate elementary reaction steps (e.g., adsorption/desorption, coupling/dissociation, (de)protonation, (de)intercalation) and evaluate their energetics/kinetics. Based on the fundamental understanding of the chemical processes, we establish descriptors for predicting properties to enable an efficient materials discovery. We also collaborate with multiple experimental groups to provide atomic-scale interpretations for the observed materials properties.
- Artificial Intelligence for Materials Discovery
In the current approach toward materials discovery, candidate materials are chosen from databases using human intuition, and their properties are computationally predicted. Thus, it is practically impossible to discover innovative materials via the current approach since databases do not contain such materials and computational predictions of all the deposited materials require too many resources. To overcome these limitations, our group is working on developing an inverse design strategy using machine learning. We are also interested in applying the developed strategy to discover new materials for energy applications including, but not limited to, fuel cells, water electrolyzer and batteries.
Selected Publications
"Electrochemical Hydrogen Peroxide Synthesis from Selective Oxygen Reduction over Metal Selenide Catalysts", Nano Letters 22 (2022) 1257-1264.
"Facet-Defined Strain-Free Spinel Oxide for Oxygen Reduction", Nano Letters 22 (2022) 3636-3644.
“Structural Insights into Multi-Metal Spinel Oxide Nanoparticles for Boosting Oxygen Reduction Electrocatalysis”, Advanced Materials 34 (2021) 2107868.
“Confined local oxygen gas promotes electrochemical water oxidation to hydrogen peroxide”, Nature Catalysis 3 (2020) 125-134.
“Efficient Discovery of Active, Selective, and Stable Catalysts for Electrochemical H2O2 Synthesis through Active Motif Screening”, ACS Catalysis 11 (2021) 2483-2491.
"Convolutional Neural Network of Atomic Surface Structures To Predict Binding Energies for High-Throughput Screening of Catalysts", The Journal of Physical Chemistry Letters, 10 (2019) 4401-4408.
“Toward a design of active oxygen evolution catalysts: insights from automated density functional theory calculations and machine learning”, ACS Catalysis 9 (2019) 7651-7659.
"Facet-Defined Strain-Free Spinel Oxide for Oxygen Reduction", Nano Letters 22 (2022) 3636-3644.
“Structural Insights into Multi-Metal Spinel Oxide Nanoparticles for Boosting Oxygen Reduction Electrocatalysis”, Advanced Materials 34 (2021) 2107868.
“Confined local oxygen gas promotes electrochemical water oxidation to hydrogen peroxide”, Nature Catalysis 3 (2020) 125-134.
“Efficient Discovery of Active, Selective, and Stable Catalysts for Electrochemical H2O2 Synthesis through Active Motif Screening”, ACS Catalysis 11 (2021) 2483-2491.
"Convolutional Neural Network of Atomic Surface Structures To Predict Binding Energies for High-Throughput Screening of Catalysts", The Journal of Physical Chemistry Letters, 10 (2019) 4401-4408.
“Toward a design of active oxygen evolution catalysts: insights from automated density functional theory calculations and machine learning”, ACS Catalysis 9 (2019) 7651-7659.
Professional Experience
2018-2020 Carnegie Mellon University 박사후연구원
2017-2018 Stanford University 박사후연구원
2017-2018 Stanford University 박사후연구원