https://scholar.dgist.ac.kr/handle/20.500.11750/81 2026-04-04T18:06:45Z https://scholar.dgist.ac.kr/handle/20.500.11750/60021 Title: Theory of slidetronics in ferroelectric van der Waals layers Author(s): Lee, Byeoksong; Lee, Minki; Kang, Joongoo Abstract: Ferroelectricity can emerge in vertically stacked two-dimensional materials even when their constituent monolayers are nonferroelectric. In these sliding ferroelectrics, polarization switching is driven by small lateral displacements between layers. Here, we develop a comprehensive materials design framework for slidetronics founded on a symmetry principle: any sliding-induced polarization change from a state P to P' can be equivalently described by applying an appropriate point-group operator, or "generator" G, to the entire system, such that P' = GP. This generator-based framework classifies all possible sliding-induced transformations, establishes the necessary symmetry conditions for switchable polarization components, and provides design strategies for realizing targeted switching behaviors. A central result is that complete polarization inversion is symmetry forbidden in bilayers but becomes possible in multilayers. First-principles calculations confirm these predictions, revealing novel phenomena including dipole-locked ferroelectricity in cellulose bilayers, in-plane switching in As2S3-based systems, and full polarization reversal in a PdSe2 trilayer. 2025-11-30T15:00:00Z https://scholar.dgist.ac.kr/handle/20.500.11750/59351 Title: Topological Machine Learning Unveils Hidden Reaction Pathways in Nanocrystal Synthesis Author(s): Lee, Byeoksong; Choi, Mahnmin; Shin, Jibin; Ha, Hyunwook; Shim, Doeun; Jeong, Sohee; Kang, Joongoo Abstract: Uncovering reaction pathways from analytical data such as UV-vis spectra remains a central challenge in nanocrystal synthesis, where transient and ill-defined intermediates complicate mechanistic analysis. Conventional approaches, reliant on manual spectral feature extraction and expert interpretation, are prone to bias and often overlook critical events. Here we present a machine learning framework that integrates transformer-based data augmentation with topological manifold learning to objectively elucidate reaction pathways directly from raw, high-dimensional spectroscopic data. The key insight is that the topology of the data manifold reflects the structure of the underlying reaction pathway. Applied to ex-situ UV-vis data sets of indium arsenide nanocrystal synthesis, this approach reconstructs the complete reaction landscape, identifying previously unreported metastable intermediates and revealing how chemical additives modulate intermediate formation to steer pathway selection. Broadly adaptable to diverse analytical data, this topological learning framework provides a generalizable strategy for mechanistic discovery and predictive control in complex chemical systems. 2025-11-30T15:00:00Z https://scholar.dgist.ac.kr/handle/20.500.11750/58965 Title: Targeted Selenite Adsorption Using Defective Fe-BTC: Effective in Acidic and Alkaline Conditions Author(s): Byun, Asong; Lee, Byeoksong; Jeong, Yujin; Kang, Joongoo; Park, Jinkyu; Park, Jinhee Abstract: Amorphous Fe-BTC, characterized by entirely defective metal nodes, has been employed for the effective adsorption of toxic selenite anions from aqueous solutions. Remarkably, Fe-BTC maintains high adsorption efficiency across a broad pH range (2-12), achieving a maximum adsorption capacity of 491 mg g-1, ranking among the highest recorded for adsorbents, including MOFs. The adsorption process involves distinct chemical interactions depending on pH: weak and variable interactions under acidic conditions (pH 2) and strong, diverse coordination modes under alkaline conditions (pH 11). Notably, the strong coordination ability of selenite ensures high selectivity over selenate and competing anions such as Cl-, NO2-, NO3-, CO32-, SO42-, and PO43-. The abundance of metal defects endows Fe-BTC with superior adsorption capacity compared to crystalline Fe-MOF, MIL-100(Fe). This study provides a comparative analysis of selenite adsorption on Fe-BTC under acidic and alkaline conditions, emphasizing pH-dependent adsorption mechanisms and their implications for designing effective adsorbents for toxic species removal. 2025-08-31T15:00:00Z https://scholar.dgist.ac.kr/handle/20.500.11750/57241 Title: Molecular Surface Doping of Cellulose Nanocrystals: A High-Throughput Computational Study Author(s): Lee, Juhyung; Lee, Byeoksong; Park, Nae-Man; Park, Ji-Sang; Kang, Joongoo Abstract: Cellulose, a linear polymer of glucose residues, is the most abundant biopolymer on Earth. However, its inability to conduct electricity limits its applications in flexible electronics and energy storage devices. Here, we performed high-throughput first-principles computational screening to identify promising molecules for surface doping of cellulose nanocrystals (CNCs). We examined over 1600 molecules, including those from the TABS database, to find candidates for p-type and n-type doping. Our results identified several p-type dopants, such as hexacyano-trimethylene-cyclopropane (CN6-CP) and octacyanoquinodimethane (OCNQ). However, no suitable n-type dopants were found due to the low electron affinity of cellulose. We constructed atomic models of CNCs of cellulose Iα and Iβ crystals, showing how their electronic band structures depend on surface hydrogen bond reconstructions. We propose a novel mechanism for photocurrent generation in CNC Iα surfaces by manipulating the hydrogen bond network at the surfaces. The selection of potential p-type dopants was further refined through the first-principles calculations of the CNC models with molecular dopants adsorbed on the surface. Finally, we demonstrate that suitable surface functionalization can enhance the electron affinity of CNCs, partially overcoming the challenges of n-type doping. © 2024 American Chemical Society. 2024-10-31T15:00:00Z