Periodic Reporting for period 2 - AddMorePower (Advanced modelling and characterization for power semiconductor materials and technologies)
Okres sprawozdawczy: 2024-01-01 do 2025-06-30
With the rapid and massive spread of power electronics based on wide-bandgap semiconductors to enable the digitalization and the electrification of our society as well as its supply with sustainable energy, new requirements arise to the conception and integration of wide-bandgap semiconductors and related interconnect materials in next-generation power electronic systems. AddMorePower will provide the necessary characterization and modelling techniques that meet the particular needs of these upcoming power semiconductor technology generations.
The AddMorePower project aims to tackle physical peculiarities of new power semiconductor materials like GaN and SiC. To that AddMorePower’s mission is to overcome the limitations currently present in the power GaN and related Cu metallization technologies by vastly improving related characterization and modelling techniques. By overcoming theses current limitations, the project aims to increase EU shares in power semiconductor production and thereby to enable a more resilient European power electronics industry that will provide solutions to important industrial and social areas.
Thereby, AddMorePower will focus on the following objectives:
• Developing novel X-ray and electron-probe based characterization workflows and protocols for power semiconductor materials.
• Setting up model concepts for better characterization and lifetime prediction of power semiconductor interconnect materials.
• Establishing FAIR and open data practices to enable efficient data workflows between characterization and modelling techniques.
The AddMorePower project aims to tackle physical peculiarities of new power semiconductor materials like GaN and SiC. To that AddMorePower’s mission is to overcome the limitations currently present in the power GaN and related Cu metallization technologies by vastly improving related characterization and modelling techniques. By overcoming theses current limitations, the project aims to increase EU shares in power semiconductor production and thereby to enable a more resilient European power electronics industry that will provide solutions to important industrial and social areas.
Thereby, AddMorePower will focus on the following objectives:
• Developing novel X-ray and electron-probe based characterization workflows and protocols for power semiconductor materials.
• Setting up model concepts for better characterization and lifetime prediction of power semiconductor interconnect materials.
• Establishing FAIR and open data practices to enable efficient data workflows between characterization and modelling techniques.
A robust project management was established, including mailing lists for communication, document sharing platform, regular work package meetings and a monthly technical meeting. Moreover, three overall on-site project meetings have been organized so far to review progress and align on open and emerging tasks.
To enable open knowledge transfer between research facilities and to support FAIR data management, a data policy and data management plan were established. The consortium uses the NOMAD repository for internal data sharing and external data publication. Data parsers have been developed to ease the upload of the manifold project data to project-internal NOMAD Oasis’ and the public NOMAD repository.
Data schemes underpinning domain-level ontologies for materials modelling and characterization were reviewed or created, resulting in AddMorePower CHADAs and MODAs. To that, the AddMorePower consortium so far has produced three CHADAs based on new CWA 17815 (e.g. on electron backscatter diffraction measurements and nano X-ray tomography) and three MODAs (e.g. on impurity diffusion and void growth modelling).
Concerning the experimental and modelling parts of the project, all intended specific workflows and protocols for X-ray and electron-probe characterization techniques as well as modelling approaches were initiated and are already yielding results (see also results beyond the state-of-art).
For communication and dissemination, a website (AddMorePower.eu) social media channels and newsletter are actively used for external communication. For visual communication, a project video as well as video interviews with project partners were created.
To enable open knowledge transfer between research facilities and to support FAIR data management, a data policy and data management plan were established. The consortium uses the NOMAD repository for internal data sharing and external data publication. Data parsers have been developed to ease the upload of the manifold project data to project-internal NOMAD Oasis’ and the public NOMAD repository.
Data schemes underpinning domain-level ontologies for materials modelling and characterization were reviewed or created, resulting in AddMorePower CHADAs and MODAs. To that, the AddMorePower consortium so far has produced three CHADAs based on new CWA 17815 (e.g. on electron backscatter diffraction measurements and nano X-ray tomography) and three MODAs (e.g. on impurity diffusion and void growth modelling).
Concerning the experimental and modelling parts of the project, all intended specific workflows and protocols for X-ray and electron-probe characterization techniques as well as modelling approaches were initiated and are already yielding results (see also results beyond the state-of-art).
For communication and dissemination, a website (AddMorePower.eu) social media channels and newsletter are actively used for external communication. For visual communication, a project video as well as video interviews with project partners were created.
• X-ray measurements were carried out on GaN/Si heterostructures and Cu polyheaters. At ESRF, several in-situ experiments on GaN-structures, partially also time-resolved at extremely high data rates, have been carried out using three different X-ray-based methods i.e. SXDM, XBIC and FFXDM.
• Within the X-ray based characterization task, the consortium has designed an in-situ setup for laboratory nano-X-ray tomography. The data allows 3D reconstruction and thus, the quantitative 3D visualizing of micro-cracks. Moreover, amongst others, the strain and mosaicity of individual Cu grains were mapped by DFXM. Lastly, an in-situ setup for TXM and DFXM has been developed which is currently employed in first lab-measurements. Last but not least, XBIC was verified as a method to detect electrical degradation.
• In terms of Electron-probe based characterization using SEM-based techniques like ECCI and HR-EBSD, simulation- and machine learning-based methods opened a path to characterize dislocations in active GaN top layers in a non-destructive way. This was confirmed by conventional indexation in TEM (Transmission Electron Microscopy). Beyond, time-resolved cathodoluminescence (TRCL) and electron-beam induced current (EBIC) methods are underway to discriminate and characterize harmless from potentially electrically active defects. The GaN-derived methods are being transferred to SiC samples, too.
• Copper-based metallic layers and contacts were tested under different conditions and boundary conditions (free- standing foils, sections attached on wafers, XRM and TEM polyheaters) to deeper understand their ageing mechanisms. In-situ and post-test analysis showed that the microstructure is evolving slowly under thermo-mechanical stresses and that pore formation depends most probably on grain boundary diffusion.
• Combined TEM in-situ experiments and FEM crystal plasticity calculations have demonstrated that dislocation movement-based mechanisms are not mainly responsible for the ageing of the metal under thermo-mechanical cycling.
• In modelling, a FEM solver for crystal plasticity modeling was adapted and existing multi-physics codes for simulations were re-structured to allow for additional solutions needed in AddMorePower. In addition, the simulation of wedge grain boundary arrangement for curved grain boundaries were modified. It showed that curved grain boundaries might be sufficient to produce tensile strains in the copper. Thus, hypotheses for degradation mechanisms in copper during thermal straining were developed. Moreover, first simulation results of elastic fields of single threading dislocation in GaN were presented.
For communication and dissemination, a website (AddMorePower.eu) social media channels and newsletter are actively used for external communication. For visual communication, a project video as well as video interviews with project partners were created.
• Within the X-ray based characterization task, the consortium has designed an in-situ setup for laboratory nano-X-ray tomography. The data allows 3D reconstruction and thus, the quantitative 3D visualizing of micro-cracks. Moreover, amongst others, the strain and mosaicity of individual Cu grains were mapped by DFXM. Lastly, an in-situ setup for TXM and DFXM has been developed which is currently employed in first lab-measurements. Last but not least, XBIC was verified as a method to detect electrical degradation.
• In terms of Electron-probe based characterization using SEM-based techniques like ECCI and HR-EBSD, simulation- and machine learning-based methods opened a path to characterize dislocations in active GaN top layers in a non-destructive way. This was confirmed by conventional indexation in TEM (Transmission Electron Microscopy). Beyond, time-resolved cathodoluminescence (TRCL) and electron-beam induced current (EBIC) methods are underway to discriminate and characterize harmless from potentially electrically active defects. The GaN-derived methods are being transferred to SiC samples, too.
• Copper-based metallic layers and contacts were tested under different conditions and boundary conditions (free- standing foils, sections attached on wafers, XRM and TEM polyheaters) to deeper understand their ageing mechanisms. In-situ and post-test analysis showed that the microstructure is evolving slowly under thermo-mechanical stresses and that pore formation depends most probably on grain boundary diffusion.
• Combined TEM in-situ experiments and FEM crystal plasticity calculations have demonstrated that dislocation movement-based mechanisms are not mainly responsible for the ageing of the metal under thermo-mechanical cycling.
• In modelling, a FEM solver for crystal plasticity modeling was adapted and existing multi-physics codes for simulations were re-structured to allow for additional solutions needed in AddMorePower. In addition, the simulation of wedge grain boundary arrangement for curved grain boundaries were modified. It showed that curved grain boundaries might be sufficient to produce tensile strains in the copper. Thus, hypotheses for degradation mechanisms in copper during thermal straining were developed. Moreover, first simulation results of elastic fields of single threading dislocation in GaN were presented.
For communication and dissemination, a website (AddMorePower.eu) social media channels and newsletter are actively used for external communication. For visual communication, a project video as well as video interviews with project partners were created.