As semiconductor devices continue to scale and become increasingly integrated, and as AI, IoT, and power electronics rapidly advance, thin-film electronic materials and heterogeneous interfaces are becoming increasingly critical to device performance and reliability. Our group conducts integrated research spanning thin-film synthesis, structural and physical-property characterization, and interface electronic-state analysis, with a focus on dielectric, semiconductor, and functional oxide materials. By combining combinatorial synthesis, automated characterization, data science and AI, and operando analysis, we aim to accelerate materials discovery and elucidate relationships among materials, processes, interfaces, and device properties. Through these efforts, we seek to develop novel thin-film materials and interface technologies for next-generation nanoelectronics, while establishing an advanced research platform for their efficient discovery, understanding, and optimization.
APPROACH
Data-Driven Development of High-Performance Dielectric Thin-Film Materials
High-performance dielectrics require the simultaneous achievement of multiple properties, including high permittivity, thermal stability, and low dielectric loss. We integrate combinatorial thin-film synthesis, automated mapping characterization, and MI/AI to efficiently explore materials and process spaces. By feeding correlations among composition, structure, and properties back into materials design, we promote data-driven materials development for multi-property optimization.
Analysis and Control of Advanced Semiconductor Interfaces Using Operando Photoelectron Spectroscopy
In advanced electronic devices, electronic states and defects at heterogeneous interfaces strongly influence device performance and reliability. We employ operando-HAXPES combining multi-energy measurements, electrical bias, and temperature control to investigate the electronic and chemical states of buried interfaces under operating conditions. The resulting insights are fed back into materials and process design, advancing Interface Engineering from interface understanding toward active control and design.
Data-Driven Materials Development and Operando Interface Science — Discovery, Understanding, and Control of Next-Generation Electronic Materials —
Overview
Next-generation semiconductor devices require not only high-performance materials but also a fundamental understanding and precise control of electronic states, defects, and potential distributions at nanoscale heterogeneous interfaces. Our research focuses on two complementary areas: data-driven development of high-performance dielectric thin-film materials and analysis and control of advanced semiconductor interfaces using operando photoelectron spectroscopy. By integrating materials exploration based on combinatorial thin-film synthesis, automated characterization, and materials informatics (MI) with electronic-state analysis of interfaces under device operating conditions using hard x-ray photoelectron spectroscopy (HAXPES), we aim to establish an integrated research platform connecting materials discovery, interface formation, and device evaluation.
Characteristics
Systematic and rapid exploration of multielement and multicompositional materials spaces on a single substrate using combinatorial thin-film synthesis
High-precision automated mapping techniques for acquiring composition, structure, electrical properties, and other characteristics across the same substrate
Data-driven thin-film materials exploration integrating automated characterization, data infrastructure, MI, and AI
Multi-objective exploration of dielectric materials considering not only high permittivity but also thermal stability, dielectric loss, leakage current, and insulating properties
Multidimensional operando-HAXPES combining multi-energy measurements, electrical bias, and temperature control to analyze potential distributions and chemical states at buried interfaces
Extension from the observation and characterization of materials and interfaces to interface engineering, in which acquired knowledge is fed back into materials and process design
Major reserch 1
Data-Driven Development of High-Performance Dielectric Thin-Film Materials
High-performance dielectric materials must simultaneously achieve multiple properties, including high permittivity, reduced film thickness, high-temperature stability, low dielectric loss, and low leakage current. We efficiently explore multidimensional materials and process spaces by combining combinatorial synthesis of compositionally graded thin films, automated mapping characterization, and MI/AI. For A2B2O7-based oxides, we have controlled composition and crystal structure and demonstrated Sr2Ta2O7–La2Ti2O7 thin films exhibiting a relative permittivity above 100 while maintaining the change in permittivity below 10% from room temperature to 300°C. We are further expanding the materials space toward multielement systems with enhanced dielectric properties and non-oxide high-k materials, advancing data-driven dielectric materials development for the simultaneous optimization of multiple functional properties.
Major reserch 2
Analysis and Control of Advanced Semiconductor Interfaces Using Operando Photoelectron Spectroscopy
In advanced electronic devices, electronic states, defects, and potential distributions at buried dielectric/semiconductor interfaces strongly influence device performance and reliability. We combine hard X-ray photoelectron spectroscopy (HAXPES) with variable photon energy, electrical bias, and temperature control to non-destructively investigate interfaces under device operating conditions. In In2O3/Al2O3 thin-film transistors, operando-HAXPES enables direct observation of local band bending and electronic-state changes during device operation. For GaN, we have clarified the formation of Ga–F bonds and associated changes in electronic states induced by fluorine-based surface termination. Through these studies, we are extending interface characterization toward Interface Engineering, in which electronic states and defects are actively understood and controlled to provide feedback for materials and process design.
Summary
By integrating Materials Discovery and Operando Interface Science, we aim to establish a unified approach to electronic materials research, spanning materials exploration, interface understanding, and interface control. We will further advance closed-loop materials exploration by integrating combinatorial synthesis, automated characterization, and MI/AI, while extending multi-energy, bias-dependent, and temperature-dependent operando-HAXPES to GaN, Ga2O3, oxide semiconductors, and two-dimensional semiconductors. By establishing a research cycle of Materials Discovery → Interface Formation → Operando Analysis → Device Evaluation → Feedback to Materials and Process Design, we seek to create high-performance, highly reliable materials and interface technologies for next-generation electronics.
Development of an Automated Material Evaluation System
Overview
With the advancement of high-performance materials, a wide range of evaluations is required. At the same time, a comprehensive perspective is necessary to properly understand trade-offs in material properties. To address this issue, in addition to specialized evaluation measurements related to high functionality, combining simple and diverse measuring devices is a potential solution. This study developed a material sequencer system that enables the integration of simple and diverse measuring devices by miniaturizing and modularizing each evaluation device while ensuring unified data acquisition across the entire system. Furthermore, since large amounts of spectral data are generated through measurements, an automated analysis technology (including algorithms and implementation) has been developed. In particular, an algorithm was designed to automatically estimate the threshold of spectra that follow a power law.
Characteristics
Development of an automated measurement system for fundamental properties (optical, electrical, and magnetic) in atmospheric conditions
Miniaturization of a Kerr effect device for evaluating magnetic properties (soft and ferromagnetic)
Development and implementation of an automated spectral analysis algorithm for power-law distributions
Major reserch 1
Simple Automated Material Evaluation System
A "Material Sequencer" was developed to automatically and continuously evaluate various inorganic materials. The image shows the external appearance of the device. This system is designed as a "health checkup" for materials, allowing diverse evaluations using simple devices to provide an overview of high-performance materials. The evaluation data is collected and integrated with the assumption of machine learning analysis. Each measurement device is modularized, fitting within a 30 cm cube, with standardized input/output controls and sample access methods. The samples targeted are inorganic materials with a 0.5-inch size. Currently, the system is capable of image measurement, reflection measurement, resistance measurement, and magnetic Kerr effect measurement. The magnetic Kerr effect device has been miniaturized using a pulse magnet to fit within this compact space.
Major reserch 2
Automated Analysis of Spectra Interpretable by Power Law
Automated spectral analysis technology has been developed for physical phenomena. In spectral data interpreted by the power law, the threshold value and power exponent serve as parameters. Examples include the critical current density obtained from current-voltage measurements of superconducting materials, the bandgap from visible-ultraviolet absorbance, and the ionization potential from photoelectron yield spectroscopy. The figure illustrates an algorithm for estimating the power exponent and threshold, along with its simulation. Power-law spectra are typically analyzed using double-logarithmic plots, but when the threshold shifts, a simple logarithmic transformation does not yield an accurate power exponent. Therefore, an adjusted plot incorporating the threshold shift is used for accurate estimation. Additionally, if the power exponent is fixed for a given phenomenon, the threshold can be estimated using an extended ReLU function and an absolute error-based fitting method.
Summary
Development of a system and a set of simple measurement devices for comprehensive material evaluation
Development of a compact Kerr effect device using a pulse magnet for determining magnetic properties
Development of an automated spectral analysis algorithm for power-law distributions
physical properties of material discriminating points
Overview
Discriminating points in materials often determine the physical properties of materials. For examples, insulator/semiconductor interfaces, metal/semiconductor interfaces, defects, catalysts, and dopants in semiconductors are material discriminating points. Since the density of discriminating points is very low generally, it is very difficult to elucidate their physical properties. Therefore, we have been studying discriminating points of the atomic structure and the chemical state using synchrotron-based X-ray absorption fine structural method and photoelectron holography. we also have been investigating the physical properties of the discriminating points of the material using operando methods to accurately understand the physical properties of the discriminating points of materials.
Characteristics
Physical properties of solid-solid interfaces by X-ray absorption and emission method
Elucidation of atomic structures and chemical states of material discriminating points by photoelectron holography (eg. defects, catalytically active sites, active and inactive atoms of dopants, etc.)
Elucidation of physical properties of material discriminating points by operando methods
Major reserch
Though there are three inequivalent Ga sites in the Si-doped k-Ga2O3 unit cell, it is not clear which Ga site is replaced by a Si atom as a dopant. Photoelectron holography revealed that the substitution sites are mixture of three inequivalent Ga sites and the active Si dopant site is the tetrahedral substitution site.
Using operando bulk sensitive photoelectron spectroscopy, we revealed that electron occupation and occupation behavior in Pt/SiO2/4H-SiC based power devices. Combined with theoretical calculations, we clarified the atomic structures and the energy levels of interfacial defects at the SiO2/4H-SiC interface.
Using X-ray absorption and emission spectroscopy, we clarified the site specific and the element specific valence band at the SiO2/Si interface. Based on the results, we reveled the site-specific dielectric constant at the vicinity of the interface.
1
2
Summary
We clarified the atomic structures and the chemical states of material discriminating points using photoelectron holography and X-ray absorption emission spectroscopy. We also revealed the dynamic properties of interface defects at solid-solid using operando photoelectron spectroscopy. In the near future, we would like to efficiently reveal the physical properties of material discriminating points by introducing combinatorial methods.
NAGATA.Takahiro@nims.go.jp
Fields of Electronic and Photofunctional Materials Research Center
Optical Single Crystals Group