Keio University

TCAD Research and Development Center

Publish: June 30, 2025
KGRI

Center Director: Kohei M. Itoh (Professor, Faculty of Science and Technology)

Primary Campus: Shin-Kawasaki Town Campus (K2)

Center Overview

TCAD (Technology Computer Aided Design) is a research field that contributes to the elucidation of physical phenomena in semiconductors and the development of the semiconductor industry by predicting semiconductor device performance through process and device simulations based on physical and chemical models. Researchers at Keio University have contributed significantly to its development. To lead its further advancement as a fundamental academic discipline and its application in industry, the TCAD Research and Development Center will be established as a contract research hub jointly funded by multiple private companies. This center will serve as a base for TCAD researchers within Keio and will fulfill its mission as the core of a TCAD collaboration network that promotes cooperation with TCAD researchers from participating companies. It will bring together researchers from academia and industry to promote TCAD research and development in the fields of semiconductor physics and semiconductor engineering. In particular, to advance TCAD technology, it will create an environment where TCAD researchers from participating companies and researchers from within Keio can work closely together.

Additionally, the center will engage in activities to connect researchers from public research institutions like AIST and domestic universities with participating companies, building a framework for R&D using TCAD through industry-academia collaboration. Furthermore, it will expand the TCAD user base by extending the use of TCAD simulation to fields other than semiconductors (e.g., nuclear power, space engineering, biochemistry, pharmacology, and medical sciences), and will also provide educational opportunities in semiconductor physics for students by having universities use TCAD as a practical training tool. Ultimately, the research outcomes will be provided to participating companies as TCAD systems, contributing to the development of domestic industries.

Keywords and Main Research Themes

TCAD, physics, chemistry, applied physics, electrical engineering, mechanical engineering

Fiscal Year 2018 Business Plan

■Continuing Activities from the Previous Fiscal Year: Background, Rationale, and Goals

As an applied development of fundamental technologies, we plan to apply domain decomposition calculations to structures that are close to actual commercial devices.

Specifically, with the goal of further scaling up, we will increase the number of devices and, in structures that also consider the peripheral circuit structure, aim to improve convergence and reduce calculation time using domain decomposition techniques.

To this end, we will investigate a determinant solver compatible with GPU (Graphics Processing Unit) computing technology. We will also aim for a practical level of acceleration in GPU-compatible Monte Carlo ion implantation simulation technology.

Meanwhile, regarding the physical model for SiC, we will investigate the physical model within the SiC substrate. Specifically, we are considering the introduction of an SiC defect model and the investigation of an impurity diffusion model in SiC.

■New Activity Goals and Content for Fiscal Year 2018, and Background for Implementation

To improve the simulation accuracy of SiC power devices, we will introduce a physical model that takes into account external influences on power devices, such as stress, heat, and electromagnetic fields. For this purpose, we will develop a collaborative analysis technology using the API functions of a multiphysics-compatible simulator and a TCAD simulator.

In addition, to accelerate large-scale analysis, we will consider introducing parallel distributed processing technology in addition to refining the domain decomposition technique. Furthermore, we aim to expand the functionality of the TCAD simulator, considering its application to other power devices.

Fiscal Year 2018 Business Report

■Implementation Details for the Fiscal Year's Business Plan, Research Outcomes, and Degree of Achievement

We conducted research and development on large-scale, large-domain, and ultra-high-speed simulations. We also developed functions not available in commercial TCAD simulators. The specific results are shown below.

  1. Large-scale, large-domain, ultra-high-speed domain decomposition technology: It became possible to perform batch calculations on a test case of a vertical MOS with 36 (6x6) cells and peripheral structures using the domain decomposition method. Hybrid mesh generation technology: By introducing a composite mesh of cubes and cuboids instead of the conventional orthogonal mesh, the number of mesh points was reduced to 1/8, calculation time was shortened to 1/13, and memory usage was reduced to 1/6, enabling large-scale, large-domain calculations. Monte Carlo ion implantation simulation using GPU (Graphics Processing Unit): Using GPU technology, the calculation time for ion implantation simulation on a Si substrate was accelerated 100-fold compared to a single CPU. The prediction accuracy of the impurity concentration distribution was almost identical to that of conventional calculations, achieving a significant speed-up.

  2. API (Application Programming Interface) function: The introduction of the API function has made it easier to incorporate physical models and computer science technologies that are difficult to handle with commercial simulators. We enhanced the semiconductor physical models, including models for SiC, small-signal analysis, piezoresistance, thermal conductivity, and specific heat. We also improved calculation control functions, including convergence determination, control of Newton's method, and calculation time control.

  3. Increased precision of diffusion models: Added implanted impurity species (F, C, N, Al). Added physical information to the diffusion model. Increased precision by adding variables for impurity diffusion calculation (from 8 previously to 30).

  4. Documentation improvement: Improved usability by creating a maintenance manual and program manuals.

  5. TCAD Academic Council activities: For 50 domestic universities and public research institutions, we provided licenses for a new TCAD simulator with API functionality and held training sessions for users. This aimed to expand the user base by facilitating the introduction of new extended physical models and to create opportunities for student education.

■Number of Published Papers (with number and names of major journals), Number of Conference Presentations (Domestic/International), and Social Contributions such as Events (Date, Location)

  • Number of conference presentations: 9 (domestic)

  • June 6, 2018: Held a council and research meeting (TCAD Academic Council) on the use of the TCAD simulator by domestic universities and public research institutions (50 institutions in total) at the Shin-Kawasaki Town Campus (K2).

■Noteworthy Achievements through Center Activities

  1. As an outcome of contract research, released the TCAD system (6 times) to parties within Keio and to the commissioning research companies: HyENEXSS V8.0K-HyENEXSS V8.5K.

  2. Announced the release of the TCAD simulator with new features to members of the TCAD Academic Council, which led to the further expanded use of the TCAD software at domestic research institutions (increase in users) and expanded use in different fields other than semiconductor physics (particle physics, space engineering, catalytic chemistry, computer science).

  3. As a result of joint research, we were able to add functions that could not be handled by commercial TCAD.

Project Members

Principal Investigator

Kohei M. Itoh

ProfessorDepartment of Applied Physics and Physico-Informatics, Faculty of Science and Technology

Hideharu Amano

ProfessorDepartment of Information and Computer Science, Faculty of Science and Technology

Yuji Awano

ProfessorDepartment of Electronics and Electrical Engineering, Faculty of Science and Technology