Bulletin of the American Physical Society
APS March Meeting 2020
Volume 65, Number 1
Monday–Friday, March 2–6, 2020; Denver, Colorado
Session L21: Machine Learning and Computational Modeling for Energy Materials |
Hide Abstracts |
Sponsoring Units: GERA Chair: Michelle Johannes, United States Naval Research Laboratory Room: 302 |
Wednesday, March 4, 2020 8:00AM - 8:12AM |
L21.00001: Devlopment of a Machine-Learned Density Functional Tight Binding for TiH2 Bulk and Surface Chemistry Nir Goldman Knowledge of chemical defect energies and kinetics is essential for assessing potential hydrogen storage materials like TiH2, where hydrogen point defects need to be assessed accurately and efficiently. The Density Functional Tight Binding (DFTB) method is a highly efficient semi-empirical quantum approach that can accurately probe these properties, but can be challenging to parameterize for each system of interest due to the different bonding types that can occur. Here, we have created a machine learning-based approach for determining the DFTB models which is both rapidly optimized, systematically improvable, and highly transferable. Our method leverages the Chebyshev Interaction Model for Efficient Simulation (ChIMES), a reactive many-body molecular dynamics force field where interactions are represented by linear combinations of Chebyshev polynomials. In this work, we discuss our ChIMES/DFTB models for TiH2, and show its accuracy for both bulk and surface properties. Our approach is easy to implement and can yield accurate DFTB models for a number of challenging materials and conditions where chemical events can be difficult to model with standard quantum approaches alone. |
Wednesday, March 4, 2020 8:12AM - 8:24AM |
L21.00002: Molecular Dynamics Modeling of Plasma Material Interactions using Machine Learned Interatomic Potentials Mary Alice Cusentino, Mitchell Wood, Aidan Thompson Multiple materials, namely tungsten and beryllium, will be present in future fusion reactors as plasma facing components. Experiments of beryllium implantation in tungsten indicate the formation of W-Be intermetallics on the surface, which can affect the performance of the tungsten divertor. Molecular dynamics (MD) can provide insight into physical processes related to experimental observations. However, MD is limited by the accuracy of the interatomic potential used. Recently, machine learning methods are being used to develop more accurate, quantum-informed potentials. |
Wednesday, March 4, 2020 8:24AM - 8:36AM |
L21.00003: Machine learning modeling of the Curie temperature for ferromagnetic intermetallics Hongbin Zhang, Teng Long, Nuno Fortunato, Yixuan Zhang, Oliver Gutfleisch Magnetic materials play an essential role in green energy and informatics applications, such as efficient energy harvesting and conversion and low energy cost spintronic devices.Currently the key challenge is how to optimize the performance of existing systems and to design novel materials for broader applications. In this talk, a random forest model is trained to classify ferromagnetic and antiferromagnetic orderings and to predict the transition temperature (TC) of the ferromagnets, using 2805 known intermetallic compounds. The resulting accuracy is 86% for classification and 92% for regression (with a mean absolute error of 58K), comparing favourably with first-principles methods. We apply these models to 5183 intermetallic compounds found in the Materials Project database, predicting their magnetic ordering and TC. This enables us to make reliable predictions, particularly by combing high throughput and machine learning methods, paving the way to accelerate the discovery of novel magnetic compounds for technological applications. |
Wednesday, March 4, 2020 8:36AM - 8:48AM |
L21.00004: Comparison of Classical Molecular Dynamics and Ab initio Molecular Dynamics with Different Equilibration Methods for Modeling Solvent - Lithium Salt Systems in Lithium Air Batteries Emily Crabb, Arthur France-Lanord, Graham Michael Leverick, Ryan Stephens, Yang Shao-Horn, Jeffrey C Grossman Lithium-air batteries are an active area of research because of their potential to have a much higher energy density than traditional lithium-ion batteries. However, they are not yet commercially viable due to poor efficiency, high charging voltages, and low cycle lifetimes. In Li-air batteries, O2 reduction starts when superoxide forms in solvent and reacts with Li+ to form lithium superoxide (LiO2). Solid Li2O2 then forms as the final discharge product on the cathode. Recent experimental work suggests that the choice of solvent and the presence of lithium salts in the system may have a large impact on how the discharge product forms at the cathode. We therefore modeled the clustering of lithium salt molecules in solvent without LiO2 present with explicit solvent calculations using both classical and ab initio molecular dynamics simulations. For each ab initio simulation, we also used one of two equilibration procedures: (1) performing a classical molecular dynamics simulation or (2) performing shorter ab initio simulations at higher temperatures. A comparison of these computational approaches for properties such as coordination numbers will be presented. |
Wednesday, March 4, 2020 8:48AM - 9:00AM |
L21.00005: Benchmarking SCAN functional for two-dimensional crystal structures Gracie Chaney, Daniel Wines, Jaron Kropp, Fatih Ersan, Can Ataca It has recently been reported that the strongly constrained and appropriately normed (SCAN) meta-GGA functional has performed exceptionally well for density functional theory (DFT) calculations involving molecular and crystalline systems. In addition, the computational cost of SCAN is argued to be much less than that of the highly demanding hybrid functional methods with comparable accuracy. SCAN has been applied to several three-dimensional systems, but has not been widely used for two-dimensional materials such as transition metal (M) monochalcogenides (MX), M dichalcogenides (MX2), and M trichalcogenides (MX3). In this study, we provide a comprehensive set of data obtained by SCAN, hybrid functionals (HSE), and PBE. Specifically, we compare lattice constants, band gaps, electronic/thermal transport and magnetic properties, and computational cost. We also study optical properties with GW approximation, using DFT orbitals obtained from SCAN and PBE. Our goal is to benchmark results from SCAN, PBE, HSE, PBE+GW, and SCAN+GW and to create a detailed picture of how SCAN performs compared to other well established DFT functionals. This work is the terminal paper for benchmarking different DFT functionals and will guide further theoretical studies involving 2D materials |
Wednesday, March 4, 2020 9:00AM - 9:12AM |
L21.00006: Theoretical study on the electric-double layer formed at interlayer of transition-metal-carbide MXene by quantum-classical hybrid interface simulation Yasunobu Ando, Masashi Okubo, Minoru Otani, Atsuo Yamada Two-dimensional transition-metal carbide MXene has been known as a promising material of high-performance electrode for energy storage applications with both non-aqueous and aqueous electrolytes because it has electronic conductivity, ion capability in interlayer nanospace, and the redox activity of Ti. |
Wednesday, March 4, 2020 9:12AM - 9:24AM |
L21.00007: First-principles analysis of band offsets, phase stability, and alkali incorporation in (Ag,Cu)(In,Ga)Se2 solar cells Kostiantyn Sopiha, Jes Larsen, Jan Keller, Marika Edoff, Charlotte Platzer-Björkman, Clas Persson, Jonathan Scragg Over the years, thin-film photovoltaics based on Cu(In,Ga)Se2 (CIGS) have undergone continuous improvements, with recent efficiency records indicating that CIGS is yet to unleash its full potential. Ag-alloying (forming ACIGS) is emerging as an exciting focus area, since ACIGS demonstrates better crystallinity and lower VOC losses than CIGS alone. In this work, we investigate the fundamental origin of these improvements by performing detailed first-principles analysis of ACIGS alloys. Based on the computed electronic properties, we establish a correlation between the composition and conduction band minimum, which allows rational tuning of band offsets at ACIGS/buffer (CdS or Cd-free) interface with a desired band gap of the absorber. From a thermodynamic analysis, we reveal the existence of a miscibility gap that can induce phase separation and Ag grading during thin film growth. Furthermore, from a defect analysis, we show that Ag promotes solubility of alkalis during post-deposition treatment. The obtained results are verified by measuring device and material characteristics for different ACIGS films, which showed a good agreement with the theoretical predictions. |
Wednesday, March 4, 2020 9:24AM - 9:36AM |
L21.00008: Accelerated Discovery of Efficient Solar Cell Materials Using Quantum and Machine-Learning Methods Francesca Tavazza, Kamal Choudhary Solar energy plays an important role in solving serious environmental problems and meeting the high energy demand. However, the lack of suitable materials hinders further progress of this technology. Here, we present the largest inorganic solar cell material search till date using density functional theory (DFT) and machine-learning approaches. We calculated the spectroscopic limited maximum efficiency (SLME) using the Tran–Blaha-modified Becke–Johnson potential for 5097 nonmetallic materials and identified 1997 candidates with an SLME higher than 10%, including 934 candidates with a suitable convex-hull stability and an effective carrier mass. Screening for two-dimensional-layered cases, we found 58 potential materials and performed G0W0 calculations on a subset to estimate the prediction uncertainty. As the above DFT methods are still computationally expensive, we developed a high accuracy machine-learning model to prescreen efficient materials and applied it to over a million materials. The data and tools are publicly distributed at: https://www.ctcms.nist.gov/~knc6/JVASP.html, https://www.ctcms.nist.gov/jarvisml/, https://jarvis.nist.gov/, and https://github.com/usnistgov/jarvis. |
Wednesday, March 4, 2020 9:36AM - 9:48AM |
L21.00009: Electrical property dominated promising half-Heusler thermoelectrics through high-throughput material computations Shuping Guo, Tiantian Jia, Shashwat Anand, Yongsheng Zhang, Jeff Snyder Half-Heusler (HH) compounds are one of the state-of-the-art thermoelectric materials with high electrical properties. Here, we carry out high-throughput computations on 95 HH compounds. Using the thermoelectric properties of NbFeSb and ZrNiSn as the screening criterion, we filter out nine p-type and six n-type promising candidates with environmentally friendly elements. Scrutinizing their electronic structures, we find that the cooperative effects of high band degeneracy, small deformation potential, light band, and large phonon velocity contribute to the large power factor. It is interesting to notice that the electrical properties dominate the thermoelectric performance in HH compounds. Balancing the excellent electrical properties and relatively low thermal conductivities, three HH compounds are predicted to be promising thermoelectric candidates. Our work not only provides novel promising materials for future experimental investigation but also offers insights into understanding the underlying physical nature of high thermoelectric performance. |
Wednesday, March 4, 2020 9:48AM - 10:00AM |
L21.00010: Computational Discovery of an Enormous Class of Stable Quaternary Chalcogenides with Very Low Lattice Thermal Conductivity Koushik Pal, Christopher Mark Wolverton The development of efficient thermal energy management devices such as thermoelectrics, barrier coatings, and thermal data storage disks relies on compounds which possess low lattice thermal conductivity (κl ). Here, we present an enormous class of thermodynamically stable quaternary chalcogenides AMM'Q3 (A=Alkali, alkaline earth, post transition metals; M,M'=transition metals, lanthanides; Q= chalcogens) that possess intrinsically low κl using high-throughput DFT calculations. Leveraging the computed energetics of hundreds of thousands of multinary compounds in the Open Quantum Materials Database (OQMD), we discovered a large number (nearly 1000) of thermodynamically stable chalcogenides through successive screening based on the calculations on multiple crystallographic prototypes of the experimentally known AMM'Q3 compounds. We validate the low-κl in this family of compounds by calculating the lattice thermal conductivity taking several representative compounds using the highly accurate anharmonic lattice dynamics methods. Our predictions suggest new experimental research opportunities in the synthesis and characterization of these stable, low-κl compounds. |
Wednesday, March 4, 2020 10:00AM - 10:12AM |
L21.00011: Electron-phonon scattering effects on the transport properties of ZrS2 from first-principles Hitoshi Mori, Masayuki Ochi, Kazuhiko Kuroki Multi-valley structure is known as one of the favorable band structures for enhancing thermoelectric efficiency. N-type TiS2 has electronic structure with multi-valley character, and its power factor is relatively high: ~40 μW/cmK2 at room temperature. In previous experimental studies, the electrical resistivity of TiS2 has been found to exhibit strong temperature dependence of ~T2. One of the previous studies indicated that inter-valley scattering among conduction band valleys may be related to the peculiar temperature dependence of the electrical resistivity [1]. |
Wednesday, March 4, 2020 10:12AM - 10:24AM |
L21.00012: Li and Na migration in Guest-free, type I clathrates evaluated via ab initio methods Xihong Peng, Andrew Dopilka, Candace Chan Open, cage-like clathrates Tt46 (Tt = Si, Ge, Sn) has a potential as anodes in Li/Na-ion batteries. Density-functional theory calculations were performed to explore ionic mobility of Li/Na through clathrate crystals. Local energy minima of Li/Na locations inside the clathrates were determined, the Li/Na migration paths/barriers were calculated using nudged elastic band method. It was found that it is favorable for Li to occupy the Tt20 cage center while preferring the off-center positions in larger Tt24 cages. The lowest Li migration barriers were found to be 0.35, 0.13 and 0.37 eV for Si46, Ge46, and Sn46, respectively. In contrast, Na shows preference for the cage centers and higher migration barriers than Li. In general, the Tt24 channels in the guest-free, type I clathrates are ideal for fast Li diffusion, while Na is too large to migrate effectively between cages. The energy landscape for Li inside the clathrates is uniquely different than that in diamond allotropes, leading to significantly lower energy barriers for Li migration. These results suggest that open frameworks of intermetallic clathrates may enable facile Li migration and have potential as anodes in Li-ion batteries. |
Wednesday, March 4, 2020 10:24AM - 10:36AM |
L21.00013: Effect of Bulk Diffusivities of Co-solvents on the Microscopic Dynamics of an Ionic Liquid Naresh C Osti, Ray Matsumoto, Matthew Thompson, Peter Thomas Cummings, Madhusudan Tyagi, Eugene Mamontov Room temperature ionic liquids (RTILs) are considered superior over organic electrolytes in terms of their thermal stability, volatility, voltage window, and device life expectancy, making them promising electrolyte materials for electrical double layer capacitors. However, pure RTILs often exhibit high viscosity, low conductivity, and poor diffusivity, which can affect the charging and discharging rates, impacting device performances. The addition of co-solvents is found to change the viscosity of RTILs, leading to improved physico-chemical characteristics that affect their bulk and interfacial properties. Even though a direct correlation of co-solvent polarity with cation diffusivity has been established, an impact of co-solvent bulk diffusivity on the cation dynamics of RTILs has never been explored. Here, using different neutron scattering spectrometers and molecular dynamics simulation, we showed a presence of a phase separation into an ionic liquid-rich and a solvent-rich phase of 1-butyl-3-methyl-imidazolium bis(trifluromethylsulfonyl)imide, [Bmim+ ][TFSI−], mixed with four different co-solvents of nearly the same dipole moment, where we observed a scaling of long-range translational mobility of the [Bmim+] cation with the bulk diffusivities of the organic solvents. |
Wednesday, March 4, 2020 10:36AM - 10:48AM |
L21.00014: Properties of V2O5 polymorphs from first principles Sakthi Kasthurirengan, Hartwin Peelaers V2O5 is a very promising battery electrode material that can intercalate not only Li, but also more abundant alkaline metals such as Na and K, and even multivalent ions such as Mg, Ca, Zn, and Al. V2O5 can occur in several polymorphs, with at least 5 different polymorphs observed. During intercalation phase transitions can take place, and such phase transitions can be detrimental to battery performance. Understanding these transitions requires knowledge of the energetics and structural properties of the various V2O5 polymorphs. |
Wednesday, March 4, 2020 10:48AM - 11:00AM |
L21.00015: Computational Design and Study of Small D-A Type Organic Molecule with Ambipolar Characteristics and Rich Linear and Non-Linear Optical Properties Dwaipayan Chakraborty, Priya Johari Push-pull type small organic molecule has recently gained a huge scientific research owing to their remarkable charge transfer properties, high non-linear optical response, reduced HOMO-LUMO gap and hence broad range of absorption spectrum, air stability etc. which collectively promotes this class of molecules as potential candidate for non-linear optical devices, OFETs and organic solar cells2. In this effort, we therefore rationally designed a promising Donor(D)-π-Acceptor(A) (i.e, push-pull) type molecule NNDM-NH2, a trans-stilbene derivative. We predicted its crystal structure starting from the experimental crystal structure of another stilbene derivative and calculated the charge transport properties, electronic band structure, gas-phase linear and non-linear optical properties. We also did the Hirshfeld surface analysis and plotted the molecular electrostatic potential to get insight into the structure-property correlation. We found that this new organic semiconductor owns a high charge carrier mobility for hole and electron, together with desirable electronic and linear and non-linear optical properties revealing NNDM-NH2 as a potential candidate for the opto-electronic devices. |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2024 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700