2024 KIAS-NCTS Workshop on Ab Initio Approaches to Quantum Materials
April 17~19, 2024 / KIAS, Bldg.1, 5F, Rm.1503
|
| Program | Home > Program |
| 　 | April 17 (Wednesday) |
April 18 (Thursday) |
April 19 (Friday) |
| 09:30-10:00 | Chair: Young-Woo Son | Chair: Tze Tzen Ong | Chair: Sangkook Choi |
| Guang-Yu Guo | Feng-Chuan Chuang | Yong-Hoon Kim | |
| 10:00-10:30 | Joongoo Kang | Youngkuk Kim | Jer-Lai Kuo |
| 10:30-11:00 | Liang-Yan Hsu | Tay-Rong Chang | Coffee |
| 11:00-11:30 | Coffee | Coffee | Chair: Bo Gyu Jang |
| Chair: Guang-Yu Guo | Chair: Hongkee Yoon | Yu-Hui Tang | |
| 11:30-12:00 | Sooran Kim | Yea-Lee Lee | Bongjae Kim |
| 12:00-12:30 | Hongkee Yoon | Woo Seok Jeong | Han Hsu |
| 12:30-14:00 | Lunch | Closing | |
| 　 | |||
| 14:00-14:30 | Chair: Myung Joon Han | Chair: Tay-Rong Chang | 　 |
| Hyoung Joon Choi | Michitoshi Hayashi | ||
| 14:30-15:00 | Tsung-Han Lee | Se Young Park | |
| 15:00-15:30 | Bo Gyu Jang | Chi-Ruei Pan | |
| 15:30-16:00 | Coffee | Poster | |
| 16:00-16:30 | Chair: Feng-Chuan Chuang | ||
| Chao-Ping Hsu | |||
| 16:30-17:00 | Yeonghun Lee | ||
| 17:00-17:30 | Tze Tzen Ong | ||
| 17:30-18:00 | 　 | Banquet | |
| 18:00-20:00 | Dinner for Invited Speakers | ||
April 17 (Wednesday)
Session I Electron and phonon dynamics
Chair: Young-Woo Son (KIAS)
09:30 ~ 10:00
Guang-Yu Guo (Nat’l Taiwan Univ, Taiwan)
Bulk Photovoltaic Effect in Low-dimensional Semiconductors: Ab Initio Studies
Bulk photovoltaic effect (BPVE) refers to the generation of intrinsic dc photocurrent in a single-phase material without inversion (P) symmetry. In recent years, BPVE has received a resurge of interest due to its promising applications in, e.g., high efficient solar energy harvesting and high sensitive THz radiation detection. Interestingly, it has also been recently shown to be a powerful probe of geometric structure of quantum states in the materials [1-2]. In this talk, I will first present an introduction to this interesting topic. Then, I will report the results and findings of our ab initio investigations on the BPVE in PT-symmetric antiferromagnetic CrI3 bilayer [3], few-layer helical chainlike selenium and tellurium [4], few-layer pentagonal transition metal dichalcogenide semiconductors PdS2 and PdSe2 [5] as well as zigzag BN nanotubes. In particular, the calculated shift current conductivity in few-layer PdX2 (X= S and Se) is found to be large, being about 130 μA/V2 and comparable to that of well-known bulk materials [5]. In few-layer PdX2 (X= S and Se), the calculated injection current susceptibilities are 100 ×108 A/V2/s, again being large. Interestingly, our GW-BSE calculations reveal a large in-gap excitonic shift current in the BN nanotubes due to the excitation of the A exciton. Importantly, the direction of the shift current in the BN nanotubes is found to be independent of the tube chirality (n,0), contrary to the simple rule of sign (Jsh) = mod (n,0) predicted by the previous model Hamiltonian studies. These indicate that the low-dimensional semiconductors considered here may find promising applications in photovoltaic solar cells.
The speaker thanks many collaborators especially Y.-S. Huang, Y.-H. Chan, V. K. Gudelli, J. Ahn, N. Nagaosa and A. Vishwanath. He also acknowledges the support from the National Science and Technology Council of The ROC.
[1] J. Ahn, G.-Y. Guo and N. Nagaosa, Low-Frequency Divergence and Quantum Geometry of the Bulk Photovoltaic Effect in Topological Semimetals, Phys. Rev. X 10, 041041 (2020).
[2] J. Ahn, G.-Y. Guo, N. Nagaosa and A. Vishwanath, Riemannian geometry of resonant optical Responses, Nature Phys. 18, 290 (2022).
[3] V. K. Gudelli and G.-Y. Guo, Antiferromagnetism-induced second-order nonlinear optical responses of centrosymmetric bilayer CrI3, Chin. J. Phys. 68, 896 (2020).
[4] M. Cheng, Z.-Z. Zhu and G.-Y. Guo, Strong bulk photovoltaic effect and second-harmonic generation in two-dimensional selenium and tellurium, Phys. Rev. B 103, 245415 (2021)
[5] V. K. Gudelli and G.-Y. Guo, Large bulk photovoltaic effect and second-harmonic generationin few-layer pentagonal semiconductors PdS2 and PdSe2, New J. Phys. 23, 093028 (2021)
[6] Y.-S. Huang, Y.-H. Chan and G.-Y. Guo, Large shift currents via in-gap and charge-neutral excitons in a monolayer and nanotubes of BN, Phys. Rev. B 108, 075413 (2023).
10:00 ~ 10:30
Joongoo Kang (DGIST, Korea)
First-principles theory of ionic thermoelectricity
Thermoelectricity (TE) is present in all electrically conductive materials such as “electron liquids” in solid metals (or doped semiconductors) and “ion liquids” in electronically gapped ionic systems. Despite the growing interest in harnessing the TE processes in ionic energy materials, a first-principles theory of fully ionic TE is still unknown. In this talk, we show that the symmetries of ionic TE—namely, the energy-gauge invariance, the Onsager reciprocal relation, and the topological quantization of adiabatic charge transport—uniquely determine how heat and charge fluxes should couple in the Green-Kubo formalism. Our theory provides a rigorous approach to extracting TE coefficients (e.g., Seebeck coefficient) from the thermal motions of constituent ions in equilibrium molecular dynamics simulations. As a proof-of-concept demonstration, we apply this approach to Cu2S (a solid-liquid hybrid phase) and Li3N (a solid-state electrolyte).
10:30 ~ 11:00
Liang-Yan Hsu (IAMS, AS, Taiwan)
Quantum-Electrodynamic Electron Transfer ReactionIn
this talk, I will briefly introduce my latest development in an emerging field “QED chemistry”. How to use quantum-electrodynamic (QED) effects to control chemical reactions is a grand challenge in chemistry. In this talk, first of all, to capture the effect of infinite photonic modes on electron-transfer reactions, we incorporate molecular quantum electrodynamics into the Marcus theory, develop a unified theory for describing radiative and non-radiative electron transfer processes, and establish the concept of “electron transfer overlap”. Furthermore, we demonstrate that electron transfer rates can be greatly enhanced by several orders of magnitude without cavities, which is implicitly supported by experimental reports. Second, to model the effect of dielectric medium in QED electron transfer reactions, we start from macroscopic quantum electrodynamics and derive an explicit Marcus-type expression, which helps us to capture the vacuum fluctuations of electromagnetic fields led by plasmonic modes and infinite cavity modes.
Session II Machine Learning 1
Chair: Guang-Yu Guo (NTU)
11:30 ~ 12:00
Sooran Kim (Kyungpook Nat’l Univ, Korea)
Machine-Learning-Guided Empirical Formula and Materials Design for Cuprates
Cuprates have been at the center of long debate regarding their superconducting mechanism; therefore, predicting the critical temperatures of cuprates remains elusive. On the other hand, the data-driven machine learning (ML) technique has been developed to predict the properties of materials without demanding a pre-known exact mechanism. In this talk, we introduce machine-learning-guided prediction models with functional forms. Using machine learning and first- principles calculations, we predict the maximum superconducting transition temperature (Tc,max) of hole-doped cuprates and suggest the functional form for Tc,max with the root-mean-square-error of 3.705 K and R2 of 0.969. We have found that the Bader charge of apical oxygen, the bond strength between apical atoms, and the number of superconducting layers are essential to estimate Tc,max. Furthermore, we predict the Tc,max of hypothetical cuprates generated by replacing apical cations with other elements. Among the hypothetical structures, the cuprates with Ga show the highest predicted Tc,max values, which are 71, 117, and 131 K for one, two, and three CuO2 layers, respectively. These findings suggest that machine learning could guide the design of new high-Tc superconductors in the future.
12:00 ~ 12:30
Hongkee Yoon (Kangwon Nat’l Univ, Korea)
Advancing Machine Learning Potentials Towards Physically Reliable Molecular Representation Learning
Estimating the energetic properties of molecular systems is a critical task in material design. Machine learning has shown remarkable promise in this task over classical force fields. However, a fully data-driven approach faces significant challenges due to the limited availability and skewed distribution of labeled data, which is predominantly concentrated around stable states. Moreover, generating an infinitely large dataset using Density Functional Theory (DFT) is not only impractical but also incurs substantial costs [1]. This highlights the importance of research focused on maximizing the efficiency of learning from the available data. In this presentation, we propose a molecular representation learning method that efficiently leverages the existing dataset and extrapolates well beyond the training distribution [2]. This is achieved through physics-driven parameter estimation from classical energy equations and innovative self-supervised learning techniques inspired by masked language modeling. To ensure the reliability of our model, we introduce novel evaluation schemes that assess its performance in a multifaceted manner, moving beyond the traditional metrics of energy or force accuracy. Our extensive experiments demonstrate that the proposed method effectively discovers molecular structures, significantly outperforming other baselines. Furthermore, we extend its application to chemical reaction pathways beyond stable states, marking a step towards physically reliable molecular representation learning.
[1] J. Lim, Y. Shim, J. Park, H. Yoon, M. Shim, Y.-G. Kim, and D. S. Kim, Molecular Dynamics Study of Silicon Carbide Using an Ab Initio-Based Neural Network Potential: Effect of Composition and Temperature on Crystallization Behavior, J. Phys. Chem. C 127, 22692 (2023).
[2] S. Yi, Y. Cho, J. Sul, S. W. Ko, S. K. Kim, J. Choo, H. Yoon, and J. Lee, Towards Physically Reliable Molecular Representation Learning, UAI 2023.
Session III Correlated Systems
Chair: Myung Joon Han (KAIST)
14:00 ~ 14:30
Hyoung Joon Choi (Yonsei Univ, Korea)
Development of an efficient and reliable DFT+DMFT method
We develop an efficient and reliable charge self-consistent density functional theory plus dynamical mean-field theory (DFT+DMFT) method based on pseudo-atomic orbitals [1]. In our DFT+DMFT method, electronic states are expressed with pseudo-atomic orbitals, the Anderson impurity model is constructed by projecting electronic states to Löwdin orthogonalized pseudo-atomic orbitals, and the exact double counting method is implemented. Our DFT+DMFT method improves the efficiency of the DFT+DMFT iterations by using the sparse-sampling approach and the causal optimization method [2]. By adopting the parameter-free analytic continuation method [3], our method produces the electronic spectral function in a reliable way. We demonstrate validity of our method by studying electronic structures of SrVO3 and various parent compounds of iron-based superconductors.
[1] Mancheon Han and Hyoung Joon Choi, to be published.
[2] Mancheon Han and Hyoung Joon Choi, Phys. Rev. B 104, 115112 (2021). [arXiv:2109.01593]
[3] Mancheon Han and Hyoung Joon Choi, Phys. Rev. B 106, 245150 (2022). [arXiv:2301.00129]
14:30 ~ 15:00
Tsung-Han Lee (Nat’l Chung Cheng Univ, Taiwan)
Ghost Rotationally-Invariant Slave-Boson: An Efficient and Accurate Approach to Strongly Correlated Materials
The rotationally-invariant slave-boson (RISB) approach is a highly efficient method for simulating strongly correlated systems [1]. When combined with density functional theory (DFT+RISB), it becomes a powerful tool for studying strong correlation effects in materials [2,3]. However, despite its efficiency, the RISB method sometimes suffers from insufficient accuracy, leading to inaccurate descriptions of material properties, such as an overestimated effective mass of Sr2RuO4 [4] and a larger critical Coulomb interaction for Mott transitions [5].
In this talk, I will introduce a systematic way to enhance the accuracy of RISB by introducing auxiliary ghost orbitals, which we refer to as the ghost-rotationally-invariant slave-boson (g-RISB) method or equivalently ghost-Gutzwiller approximation [6]. I will first present examples of transition metal oxides where DFT+RISB necessitates the use of unrealistic Coulomb parameters, significantly deviating from first-principle calculated values, to reproduce the experimental observations [5]. Subsequently, I will demonstrate how DFT+g-RISB offers a systematic approach to improve the accuracy of DFT+RISB, enabling accurate descriptions of correlated materials with realistic Coulomb interactions [7,8]. Moreover, I will compare the accuracy and efficiency of g-RISB with the well-established dynamical mean-field theory (DMFT) and discuss the advantages and disadvantages of g-RISB over DMFT.
[1] F. Lechermann, A. Georges, G. Kotliar, and O. Parcollet, Phys. Rev. B 76, 155102 (2007)
[2] C. Piefke and F. Lechermann, physica status solidi (b) 248, 2269 (2011)
[3] Nicola Lanatà, Yongxin Yao, Cai-Zhuang Wang, Kai-Ming Ho, and Gabriel Kotliar, Phys. Rev. X 5, 011008 (2015)
[4] J. I. Facio, J. Mravlje, L. Pourovskii, P. S. Cornaglia, and V. Vildosola, Phys. Rev. B 98, 085121 (2018)
[5] Nicola Lanatà, Tsung-Han Lee, Yong-Xin Yao, Vladan Stevanović, Vladimir Dobrosavljević, npj Computational Materials 5 (1), 1-6 (2019)
[6] Nicola Lanatà, Tsung-Han Lee, Yong-Xin Yao, and Vladimir Dobrosavljević, Phys. Rev. B 96, 195126 (2017)
[7] Tsung-Han Lee, Nicola Lanatà, and Gabriel Kotliar, Phys. Rev. B 107, L121104 (2023)
[8] Tsung-Han Lee, Corey Melnick, Ran Adler, Nicola Lanatà, Gabriel Kotliar, Phys. Rev. B 108, 245147 (2023)
15:00 ~ 15:30
Bo Gyu Jang (Kyunghee Univ, Korea)
Self-consistent DFT+U+V study of mixed-valence compounds
The DFT+U+V method is an extension of the DFT+U approach that includes inter-site Coulomb interactions V in addition to on-site Coulomb interaction U. This extension improves the accuracy of the DFT+U method in systems where the hybridization between orbitals at different atomic sites is important. Recently, this new method has received great attention due to its high accuracy in correcting self-interactions and its low computational costs [1]. It has successfully described various properties of solids such as semiconductor band gaps [1], Li-ion battery voltages [2], and perovskites defect states [3].
In this presentation, I will discuss the application of DFT+U+V approach to study mixed valence compounds. Firstly, I will examine the prototypical charge-ordered material BaBiO3, where the charge-ordered state can be suppressed by substituting Ba with K, leading to superconductivity. However, the conventional DFT method fails to describe its physical properties. Our study reveals that the inter-site Coulomb interaction V is essential to describe all their doping-dependent experimental features such as breathing instabilities and electron-phonon matrix element [4]. Secondly, I will discuss LiCu2O2 which contains Cu(II) (d9) and Cu(I) (d10) ions in its crystal structure. This material exhibits an unprecedented 250 % broadening of the bandwidth compared to the DFT prediction [5]. However, DFT+U+V successfully describes the experimental bandwidth, indicating the significance of inter-site Coulomb interaction in this system.
[1] S.-H. Lee and Y.-W. Son, Phys. Rev. Res. 2, 043410 (2020).
[2] M. Cococcioni and N. Marzari, Phys. Rev. Mater. 3, 033801 (2019).
[3] C. Ricca, I. Timrov, M. Cococcioni, N. Marzari, and U. Aschauer, Phys. Rev. Res. 2, 023313 (2020).
[4] B. G. Jang, M. Kim, S.-H. Lee, W. Yang, S.-H. Ji, and Y.-W. Son, Phys. Rev. Lett. 130, 136401 (2023).
[5] M. Moser et al., Phys. Rev. Lett. 118, 176404 (2017).
Session IV Excited states and topological systems
Chair: Feng-Chuan Chuang (NSYSU)
16:00 ~ 16:30
Chao-Ping Hsu (IOC, AS, Taiwan)
Excited-State Charge transfer Coupling from Quasiparticle Energy Density Functional Theory
Electric transfer (ET) coupling is an important parameter in understanding and predicting the rate of charge transfer processes. Under Fermi's golden rule, the rate of ET is proportional to the amplitude square of the coupling values. ET coupling involving excited states is often calculated via single-excitation theories such as configuration interaction singles or time-dependent density functional theory (TDDFT), both being feasible for large molecules but rather limited in their quality of prediction. This challenge was addressed with the recently developed Quasiparticle Energy Density Functional Theory (QE) for excited state energy calculation. In present work, the QE method is further extended to calculating ET couplings involving excited states. For the intermolecular ET between a furan and 1,1-dicyanoethylene, the two possible ET couplings, between a pair of symmetric locally excited (LE) and charge-transfer (CT) states and another pair of asymmetric states are studied with both generalized Mülliken-Hush (GMH) and fragment charge difference (FCD) schemes. The coupling values exhibits smooth exponential distance dependence, and they are close to those obtained through TDDFT, but with only a small fraction of computational costs. Our work shows that QE approach is promising for calculation of ET couplings involving excited states for large and complex systems.
16:30 ~ 17:00
Yeonghun Lee (Incheon Nat’l Univ, Korea)
Two-Dimensional Topological Transistors
In this talk, we present our modeling work of two-dimensional (2D) topological transistors based on the gate-controlled phase transition of the quantum spin Hall insulators. For the device operation, the gate bias induces an out-of-plane electric field and breaks the inversion symmetry of the 2D topological insulators, resulting in the phase transition to the normal insulating phase. The concept of the topological insulator field-effect transistor was proposed approximately a decade ago, promising high performance with unique edge transport due to the symmetrically forbidden backscattering. While several experimental works have validated the topological phase transition, the demonstration of the actual device operation remains elusive. Our work addresses the challenges through comprehensive material and device modeling, providing insights into the intricate interplay between the gate-induced topological phase transition and device operation. By bridging the gap between condensed matter physics and device engineering, our work aims to unlock the full potential of the 2D topological transistors.
17:00 ~ 17:30
Tze Tzen Ong (IOP, AS, Taiwan)
Feature Spectrum Topology
Topology is a fundamental aspect of quantum physics, and it has led to key breakthroughs and results in various fields of quantum materials. In condensed matters, this has culminated in the recent discovery of symmetry-protected topological phases. However, symmetry-based topological characterizations rely heavily on symmetry analysis and are incapable of detecting the topological phases in systems where the symmetry is broken, thus missing a large portion of interesting topological physics. Here, we propose a new approach to understanding the topological nature of quantum materials, which we call feature spectrum topology. In this framework, the ground-state is separated into different partitions by the eigenspectrum of a feature, a particular chosen internal quantum degree of freedom, such as spin or pseudo-spin, and the topological properties are determined by analysis of these ground-state partitions. We show that bulk-boundary correspondence guarantees gapless spectral flows in either one of the energy or feature spectrum. Most importantly, such ‘feature-energy duality’ of gapless spectral flows serves as a fundamental manifestation of a topological phase, thereby paving a new way towards topological characterizations beyond symmetry considerations. Our development reveals the topological nature of a quantum ground state hidden outside symmetry-based characterizations, hence, providing a platform for a more refined search of unconventional topological materials.
April 18 (Thursday)
Session V Topological Physics
Chair: Tze Tzen Ong (IOP, AS)
09:30 ~ 10:00
Feng-Chuan Chuang (Nat’l Sun Yat-sen Univ, Taiwan)
Engineering Topological Phases of 2D Layered Materials
In this presentation, we explore the fields of topological insulators (TIs) and two-dimensional materials (2D) materials, which offer promising applications in technology. We highlight the collaborative nature of experimental and theoretical approaches in accurately predicting the properties of quantum materials. Our focus centers on 2D materials with honeycomb structures which showcase the emergence of unique properties through heterojunctions formed by layer stacking. Central to our research is the investigation of TIs, unveiling surface conducting states within insulating bulk materials. We elucidate topology principles, showcasing their invariance under continuous deformations. Through theoretical calculations and experimental validation, we uncover distinctive properties and potential applications in spintronics and quantum computing. Furthermore, we explore the manipulation of topological properties through external factors like strain, substrate engineering, and material functionalization. By adjusting lattice structures and chemical compositions, we demonstrate transitions between trivial and non-trivial topological phases in various 2D materials. Our research extends beyond theory, achieving successful experimental realization of proposed materials. Collaborating with experimental groups, we validate our findings using scanning tunneling microscopy (STM) and angle-resolved photoemission spectroscopy (ARPES) techniques. In summary, our work advances understanding of topological phases and band engineering in 2D materials, promising innovations across various technological fields. By uncovering principles behind topological insulators and exploring novel material designs, we contribute to the expanding realm of quantum materials, paving the way for future breakthroughs.
10:00 ~ 10:30
Youngkuk Kim (SKKU, Korea)
Replicated Higher-Order Topology in the Hofstadter Butterfly of Twisted Bilayer Graphene
In this presentation, we study the complex energy spectrum of twisted bilayer graphene (TBG), particularly focusing on their Hofstadter energy spectrum and distinctive recursive higher-order topological phases [1]. Our calculations reveal the emergence of higher-order topological insulator (HOTI) phases within TBG, featuring localized corner states. These phases are intriguingly replicated in the Hofstadter spectrum, demonstrating the inherent self-similarity of the energy spectrum. We prove the presence of an exact flux translational symmetry in TBG across all commensurate angles. Leveraging this discovery, we observe that the original HOTI phase, initially identified at zero flux, it re-emerges with a half-flux periodicity, maintaining an effective twofold rotational symmetry. Furthermore, our study uncovers several replicas of the original HOTIs at flux values lacking protective symmetries. These replicated HOTIs, akin to their original counterparts, exhibit localized corner states and edge-localized real-space topological markers. The emergence of these replicas can be attributed to the varying interaction scales in TBG, particularly the intralayer and interlayer couplings. Our research not only sheds light on the topological dimensions of Hofstadter butterflies but also underscores the role of the symmetry-protected topology in the realm of quantum fractals.
[1] Kim, Sun-Woo, et al. "Replica higher-order topology of Hofstadter butterflies in twisted bilayer graphene." npj Computational Materials 9.1 (2023): 152.
10:30 ~ 11:00
Tay-Rong Chang (Nat’l Cheng Kung Univ, Taiwan)
Feature-energy duality of topological boundary states in multilayer quantum spin Hall insulator
Gapless topological boundary states characterize nontrivial topological phases arising from the bulk-boundary correspondence in symmetry-protected topological materials, such as the emergence of helical edge states in a Z_2 topological insulator. However, the incorporation of symmetry-breaking perturbation terms in the Hamiltonian leads to the gapping of these edge bands, resulting in missing these crucial topological boundary states. In this talk, I will introduce our recent results on the critical issue of bulk-boundary correspondence in the quantum spin Hall insulator (QSHI) via novel-approaching feature spectrum topology [1]. Our findings present a comprehensive understanding of feature-energy duality, illustrating that the aggregate number of gapless edge states in the energy-momentum map and the non-trivial edge states in the feature spectrum equals the spin Chern number of multilayer QSHI. We identify a van der Waals material Bi4Br4 as a promising candidate through first-principles calculations. Our work not only unravels the intricacies of bulk-boundary correspondence but also charts a course for exploring quantum spin Hall insulators with high spin-Chern number.
[1] Yueh-Ting Yao et al., arXiv:2312.11794 (2023).
Session VI Machine Learning 2
Chair: Hongkee Yoon (KNU)
11:30 ~ 12:00
Yea-Lee Lee (KRICT, Korea)
Empowering materials research through machine-learning and data platforms
The advent of machine learning technology is revolutionizing material development research. Through the application of machine learning, researchers can achieve faster and more accurate results than conventional methods in diverse areas, including design of new materials, improvement of measurement techniques, optimization of process conditions, and predictions for future research. In this presentation, we discuss our investigation into machine learning studies aimed at accelerating materials research. Our focus includes materials design, particularly for SnSe doping systems in thermoelectric materials, and enhancing measurement efficiency for 2D materials using photoluminescence (PL) data. Collaborating with experimental groups, we generated the experimental datasets and systematically collected these data through web-based data platforms. We propose novel doping compositions for SnSe that significantly enhance thermoelectric performance. We also demonstrate that the enhancement of PL resolution by machine learning dramatically reduces the measurement time. Additionally, we introduce our emerging automated laboratories integrated with machine learning, anticipating they will pave the way for systematic data collection, efficient exploration of space, and accelerated development of new materials.
12:00 ~ 12:30
Woo Seok Jeong (KIER, Korea)
Active Learning Configuration Interaction to Accelerate Electronic Structure Calculations for Aromatic Molecules
An accurate description of the electronic ground and excited states of aromatic compounds is of great importance for the development of optoelectronic, photocatalytic and photovoltaic applications. One of the most popular methods for calculating excited states of molecules is time-dependent density functional theory (TD-DFT), but it has been shown that TD-DFT can significantly fail to compute the energy of strongly correlated systems. For such molecular systems, multiconfigurational methods have been developed and are becoming increasingly popular but have only been used to study relatively small systems due to the demanding computational cost. In this talk, I will present the active learning based configuration interaction (ALCI) method, which combines machine learning approaches with the selected configuration interaction (SCI) theory. SCI theory is a conceptually simple and robust way to account for electron correlations, but it can only give reliable results if all or most of the energetically important electronic configurations are used to construct the wave function. In other words, the key step for the SCI calculations is how to find important configurations. In our ALCI scheme, we utilized machine learning (ML) model to identify important configurations, and the performance of the ML model is designed to be improved by iterations based on an active learning scheme. The ALCI method has been shown to successfully predict the lowest singlet-singlet vertical excitation energies for small- and medium-sized polycyclic aromatic hydrocarbons (i.e., from naphthalene, anthracene and pyrene) compared to complete active space CI (CASCI) results. Furthermore, the ALCI approach allows us to calculate the excitation states of larger systems such as tetracene, pentacene and hexacene with significantly reduced computational cost. We will also show that the ALCI method can be adopted to investigate different excited states (not limited to singlet-singlet transition) of various cyclic organic molecules.
References:
W. Jeong, C. A. Gaggioli, L. Gagliardi. J. Chem. Theory Comput., 2021, 17, 7518–7530.
Session VII Low dimensional systems
Chair: Tay-Rong Chang (NCKU)
14:00 ~ 14:30
Michitoshi Hayashi (CCMS, NTU, Taiwan)
Enhanced Photocatalytic Activity of V-Doped MoS2 with S Vacancies: Insights from Adsorption Studies and Reaction Pathways
The initial step in photocatalytic reactions involves the adsorption of reactants onto catalyst surfaces. The limited activity of monolayer MoS2 for CO2 reduction reactions (CO2RR) stems from its inert basal plane covered by S atoms. However, previous studies have shown that S vacancies on MoS2 surfaces can enhance photocatalytic activity for CO2 hydrogenation [1].
In this study, we systematically examined the adsorption activity of V-doped MoS2 systems with S vacancies for both H2O and CO2, utilizing 4 × 4 monolayer MoS2 models with varying Vanadium doping ratios. We found that MoS2 with an S vacancy is unable to chemisorb H2O molecules, whereas when the vacancy is located at a V-substituted site, the H2O molecule chemically bonds with the metals, resulting in significantly stronger adsorption energy. However, the adsorption of CO2 molecules in these systems still demonstrates physisorption with weak energies. To investigate further, we employed the climbing image nudged elastic band (CI-NEB) method to identify potential chemisorption intermediates arising from physisorbed CO2, comparing reaction barriers for its conversion into CO and O across three MoS2-based models. We observed reduced barriers for CO2 conversion in V-doped MoS2, with the most significant reduction observed in 4% V-doped MoS2 with 2 S vacancies.
Compared with pure MoS2, the effects of V-doping can be seen in (i) improved H2O adsorption and (ii) reduced barriers for converting physisorbed CO2 molecules.
References
1. J. Hu, L. Yu, et al. Nature Catalysis 4, 242 (2021).
14:30 ~ 15:00
Se Young Park (Soongsil Univ, Korea)
Spin-orbit-splitting driven nonlinear Hall effect in NbIrTe4
The Berry curvature dipole (BCD) is considered one of the mechanisms that induce nonlinear Hall effects (NLHE) in the materials having the time-reversal symmetry, where the low point group symmetry allows the net BCD, giving rise to the Hall effect quadratically depending on the applied electric field. We investigate the NLHE in NbIrTe4 thin films that persists above room temperature, exhibiting sign change in the Hall conductivity at 150 K. First-principles calculations combined with angle-resolved photoemission spectroscopy (ARPES) measurements show that partially occupied spin-orbit-split bands mainly contribute BCD. Moreover, the sign change in NLHE is driven by chemical potential shifts associated with the change in temperature, which reverses the overall sign of BCD. Our findings highlight the correlation between BCD and the electronic band structure, providing a viable route to create and engineer the Hall effect by tuning the geometric properties of quasiparticles in transition-metal chalcogen compounds.
15:00 ~ 15:30
Chi-Ruei Pan (IAMS, Taiwan)
Moiré Patterns in Hexagonal Boron Nitride on Cu(111) and Ru(0001) and the Effect of Selenium Intercalation
Van der Waal heterostructures have become a promising playground for two-dimensional (2D) materials in creating novel and functional devices. The intricate moiré patterns resulting from the lattice mismatch between 2D materials and metal substrates offer a unique platform for tailoring electronic and optical properties. Although moiré patterns are widely observed during 2D heterostructure fabrication, their microscopic topographies and effects on the electronic properties are not fully explored. Here, we study a single layer hexagonal boron nitride (hBN), grown on Cu(111) [1] and Ru(0001) [2], with a focus on the intercalation of Se atoms in the latter case [3]. The presentation will be mainly composed of two parts. For the first part, we will present the modulated electronic properties resulting from the moiré pattern. In particular, we will further discuss the amplitude of the modulated properties as a function of the rotation angle for hBN-Cu since its weak interaction and various domains with different lattice orientations can be observed. For the second part, we will then focus on the intercalation of Se atoms into the interface between corrugated hBN and Ru(0001). Three systems are introduced for investigating the effect of intercalation on the structural and electronic properties of the system. Adsorption, intercalation, and formation energies are calculated, respectively for each system, at different Se coverages. We find that at a coverage of approximately 28%-33%, the intercalated Se atoms have the lowest energy per atom. In addition, a convex hull is constructed based on the formation energy, indicating that the thermodynamically stable intercalated region has a coverage of about 28%-33% with the simultaneous presence of the region without intercalation. This coexisting phase has a large built-in work function change of about 1 eV between two regions, providing a platform for semiconducting engineering. Our theoretical results explain well the experimental observation of an abrupt change in the scanning tunneling microscopy image and the measured work function change when Se is intercalated into the hBN/Ru interface.
References
[1] Zhang, Qiang, et al. "Tuning band gap and work function modulations in monolayer hBN/Cu (111)
heterostructures with Moiré patterns." ACS nano 12(9), 9355-9362 (2018).
[2] Zhang, Qiang, et al. "Bandgap renormalization and work function tuning in MoSe2/hBN/Ru (0001)
heterostructures." Nature communications 7(1), 13843 (2016).
[3] Holbrook, Madisen, et al. "Creating a Nanoscale Lateral Junction in a Semiconductor Monolayer
with a Large Built-in Potential." ACS nano 17(7), 6966-6972 (2023).
April 19 (Friday)
Session VIII Surface and interface
Chair: Sangkook Choi (KIAS)
09:30 ~ 10:00
Yong-Hoon Kim (KAIST, Korea)
Theory of electric enthalpy in electrified interface
For the first-principles simulation of non-equilibrium open junction systems under finite bias, there exist several motivations to seek an alternative to the standard approach of combining density functional theory (DFT) and nonequilibrium Green’s function (NEGF) formalisms. For example, for graphene electrode-based van der Waals 2D tunneling transistors in operando conditions, DFT-N EGF simulations are not possible due to the inherent limitations of the formalism [1-2]. To overcome this challenge, in the past decade plus, we have developed multi-space constrained-search DFT (MS-DFT) [1-4], which as an alternative to the standard grand-canonical Landauer picture takes a microcanonical viewpoint that maps quantum transport processes to space-resolved optical excitation counterparts.
In this presentation, I discuss an ab initio theory to calculate the electric enthalpy of electrified interfaces based on MS-DFT [5]. Being a microcanonical theory, MS-DFT allows one to calculate the nonequilibrium total energy of a nanoscale electrode-channel-electrode junction. An additional MS-DFT calculation for the electrode-only counterpart that maintains the same bias voltage allows one to identify the internal energy of the channel as well as the electric field and the channel polarization, which together determine the electric enthalpy and the nonequilibrium adsorption energy. Application of the developed scheme to the water-Au and water-graphene interface models shows that the Au and graphene electrodes induce very different behaviors in terms of the electrode potential-dependent stabilization of water configurations. Applications of the developed formalism to ferroelectric tunnel junctions will be also discussed.
References
[1] H. S. Kim & Y.-H. Kim, https://doi.org/10.48550/arXiv.1808.03608 (2018).
[2] T. H. Kim, J. Lee, R. Lee, & Y.-H. Kim, Npj Comput. Mater. 8, 50 (2022).
[3] J. Lee, H. S. Kim, and Y.-H. Kim, Adv. Sci. 7, 2001038 (2020).
[4] J. Lee, H. Yeo, and Y.-H. Kim, Proc. Natl. Acad. Sci. U. S. A. 117, 10142 (2020)
[5] J. Lee, H. Yeo, R.-G. Lee, & Y.-H. Kim, https://doi.org/10.48550/arXiv.2402.15103 (2024).
10:00 ~ 10:30
Jer-Lai Kuo (IAMS, AS, Taiwan)
Development of Integrative Computational Tools to Link Potential Surface with Experimental Vibrational Spectra
Vibrational spectroscopy is a commonly used experimental method to probe the structures and dynamics of molecular systems. To interpret vibrational spectra, the standard procedure is to use an accurate and efficient electronic method to locate local minima on the potential surface (PES), compute their second derivatives and linear terms of the dipole to simulate IR spectra for comparison. However, anharmonic coupling, such as Fermi resonance, are known to lead to non-trivial and complex features in the high-frequency part of experimentally observed spectra in mid-IR [1]. There are several parallel developments on the theories and algorithms to address different aspects of anharmonic coupling and dynamical effects in linear and non-linear spectroscopy [2-4]. In this talk, I will present a team effort to develop an integrative theoretical scheme to (i) extract simple effective Hamiltonian directly from PES without any fitting parameters covering a frequency range from THz, mid-IR to NIR [5] and (ii) to enable direct simulations of the non-linear spectra and identification of specific spectral signals to elucidate the key roles of anharmonic coupling [6].
References
[1] EL Sibert, Annu. Rev. Phys. Chem. 74, 219 (2023)
[2] Q-R Huang, Y-C Li, T. Nishigori, M. Katada, A. Fujii, and J-L Kuo, J Phys. Chem. Lett. 11, 10067 (2020)
[3] M-T Wong and Y-C Cheng, J. Chem. Phys., 154, 154107 (2021).
[4] C-K Lin, Q-R Huang, M. Hayashi and J-L Kuo, Phys. Chem. Chem. Phys. 23, 257364 (2021)
[5] Q-R Huang, K. Yano, Y. Yang, A. Fujii, J-L Kuo, Phys. Chem. Chem. Phys., (submitted)
[6] J-H Yu, M-T Wong, Q-R Huang, J-L Kuo, and Y-C Cheng (in preparation)
Session IX Magnetic systems
Chair: Bo Gyu Jang (KHU)
11:00 ~ 11:30
Yu-Hui Tang (Nat’l Central Univ, Taiwan)
Determination of perpendicular magnetic anisotropy in magnetic heterostructures: DFT-based spin-orbit torque method
In order to increase the storage capacity of magnetoresistive random-access memories (MRAMs), the reduction of size of memory cell requires the enhancement of the magnetic anisotropy to meet the criterion of thermal stability. The recently discovered perpendicular magnetic anisotropy (PMA) at the ferromagnetic metal/oxide and ferromagnetic metal/heavy metal interfaces plays a crucial role in MRAM applications. However, both first-principles calculated total energy and fitted experimental data remain difficult to quantitatively evaluate the interfacial magnetic anisotropy energy, especially in magnetic heterostructures.
An alternative solution is the so-called spin-orbit torque (SOT) method. In this study, the validity of our DFT-based self-developed JunPy package [1] with SOT calculation has been rigorously confirmed in three kinds of PMA systems, including iron thin films [2], Fe/MgO/Fe magnetic tunnel junction [3] and Co/Pd magnetic bilayer. Our results not only agree with the experimental measurements and conventional MAE calculation but also provide deeper insights into atomistic spin dynamics of local magnetic moments.
References
[1] Y. -H. Tang et al., Phys. Rev. Research 3, 033264 (2021).
[2] B. -H. Huang et al., AIP Advances, 13, 015034 (2023).
[3] B. -H. Huang et al., Journal of Magnetism and Magnetic Materials, 585, 171098 (2023).
11:30 ~ 12:00
Bongjae Kim (Kyungpook Nat’l Univ, Korea)
Competing magnetisms and onset of the static magnetic order in Sr2RuO4
Sr2RuO4 has long been a system of interest due to its unconventional superconducting properties. While paramagnetic metal in its normal state, the system harbors interesting landscape of competing magnetic instabilities that are intricately linked to its superconducting pairing symmetries. In this talk, starting with a brief overview of the system, I will discuss the various magnetic features we have investigated in the Sr2RuO4 and with the intimate connection to the experimental findings. Specifically, we attempt to explain the onset of a static magnetic order beyond a critical strain by combining the first-principles density functional theory and Moriya’s self-consistent renormalization theory.
12:00 ~ 12:30
Han Hsu (Nat’l Central Univ, Taiwan)
Magnetic properties of SrCoO3 under pressure and epitaxial strain
In recent years, bulk and thin-film SrCoO3 have attracted significant attention due to their unique magnetic properties. At ambient conditions, bulk SrCoO3 is a ferromagnetic (FM) metal in cubic perovskite structure. Theory has shown that bulk SrCoO3 contains intermediate-spin (IS, t2g5eg1-like) Co exhibiting d6L character (nearly Co3+ accompanied by O 2p ligand holes). By contrast, magnetic properties of tensile-strained stoichiometric SrCoO3 thin films remain controversial. Previous calculations had predicted that thin-film SrCoO3 undergoes a strain-induced FM–antiferromagnetic (AFM) transition at tensile strain ε≈ 2% and remains AFM for ε≥ 2% [1]. Despite great efforts, such predicted AFM state has never been confirmed in experiments. Moreover, recent experiments with high-quality stoichiometric SrCoO3 thin films showed that SrCoO3 remains FM up to ε= 3% and undergoes a metal–insulator transition at ε= 2–3%, in direct contradiction with theoretical predictions [2]. So far, the underlying mechanism of this metal–insulator transition and the subsequent FM insulating state remains unclear. In this talk, I will discuss two major predictive results obtained by our calculations. (1) Upon compression to ~7 GPa, bulk SrCoO3 undergoes a simultaneous spin and metal–half-metal transition to a FM half-metallic low-spin (LS) state. Compared to the metallic IS state at ambient conditions, this half-metallic LS state exhibits even more prominent d6L character, including nearly nonmagnetic Co3+ and exceptionally large oxygen magnetic moments, which contribute most of the magnetization [3]. (2) At tensile strain ε≈ 2.5% a simultaneous structural, spin, orbital, and metal–insulator transition occurs, and a FM insulating state with high-spin (HS) Co in a complicated orbital ordering emerges, consistent with experiments. The energy gap of this state is opened via complicated orbital ordering, cooperative Jahn–Teller distortion, and octahedral tilting about all three crystal axes.
References
[1] Lee, J.H. and Rabe, K.M. Coupled magnetic-ferroelectric metal-insulator transition in epitaxially strained SrCoO3 from first principles. Phys. Rev. Lett. 107, 067601 (2011).
[2] Wang, Y. et al. Robust ferromagnetism in highly strained SrCoO3 thin films. Phys. Rev. X 10, 021030 (2020).
[3] Hsu, H. and Huang, S.C. Simultaneous metal–half-metal and spin transition in SrCoO3 under compression. Phys. Rev. Materials 2, 111401(R).
[4] Huang, S.C., Sarkar, K., Wentzcovitch, R.M., and Hsu, H. Orbital-ordered ferromagnetic insulating state in tensile-strained SrCoO3 thin films. Submitted (available at arXiv.2211.10404).