Novel GaN Power Devices and Packaging Technologies for 300C Ambient Operation

Lead Research Organisation: Heriot-Watt University
Department Name: Sch of Engineering and Physical Science

Abstract

Abstracts are not currently available in GtR for all funded research. This is normally because the abstract was not required at the time of proposal submission, but may be because it included sensitive information such as personal details.
 
Description The key objective of the project was to develop a laser based method for assembly (die-attach or mounting) of a chip on a substrate for high power and/or high temperature applications. The laser based approach is potentially a faster and energy efficient method. Through the development of the process and comparison with the conventional method using a hotplate for the heating process, it has been shown that the new approach is capable of producing strong bonding between the chip and the substrate using a silver nanoparticle paste material. The nanoparticle material can be sintered at a relatively low temperature (< 300C) to form a joint between a microchip and a substrate in advanced packaging of the wide bandgap (WBG) semiconductor (SiC and GaN) based power transistor devices and modules. The sintered silver layer has a melting temperature close to that of the bulk silver material (962C) and therefore is an enabling material for high power and high temperature application of the advanced power devices in electrical vehicles and aerospace systems. Silver sintering is already replacing the conventional assembly method using solder based materials in advanced applications.

In the investigation and development of the laser based sintering/bonding process, a systematic approach was applied to the study of the effects of process parameters (laser power, sintering/bonding pressure and time, surface metallisations (copper or silver) on the quality of the laser sintered bonding layer between chips and the DBC substrates as well as silicon substrates with coated metal films with the top layer being a copper film or silver film for comparison. The main quality indicator is the shear strength of the interfacial bonds as well as the silver layer between a chip and the substrate. The findings show that a strong bond can be produced using an efficient, low cost, high power diode laser system to generate a temperature greater than 300C. A high sintering pressure produces a strong bond. We have investigated sintering pressures from a minimal value (< 0.2 MPa, pressureless sintering) and 1.5 MPa. The sintered silver layer is more dense at higher sintering pressures resulting a high shear strength. With copper to copper (copper surface both chip and substrate), a shear strength of up to 20 MPa was obtained at the sintering pressure of 3 MPa and a sintering time of 5 minutes. This is a good result for the moderate level of sintering pressure. In comparison, at the same sintering pressure the conventional hotplate sintering method produced a shear strength of ~15 MPa at a longer sintering time of 30 minutes. It was also found that it is possible to produce a good shear strength of 10 MPa at the sintering time of 1 minute. This is a reduction of the sintering time by a factor of 30 as compared the hotplate based sintering method. If the method is used in a production process, it would be an efficient manufacturing process. The results were published in the leading journal of the field of research.

In the study of the effect of surface metallisation on shear strength, the findings show that with silver film on the surface of the chip shear strength values higher than 30 MPa can be achieved. In this case the fracture interface is between the nanoparticle silver layer and the copper surface of the DBC indicating it is the weaker bonding interface. While in copper-copper bonding, the fracture interface is between the chip surface and the sintered silver layer. This is because the silver paste was printed on the DBC substrate and there is better interaction with the copper surface of the DBC since the paste was wet and has solvent and other organic additives which assist initial contact of the nanoparticles with the copper surface. But for the interaction of the copper surface on chip with the silver layer, the silver layer was dried in order to remove the solvent before pressure sintering, therefore the contact between the silver particles and the copper surface of the chip is not as good as that of the paste state with the DBC surface. With both surfaces being silver, the strength is even higher, a value of ~ 40 MPa was achieved. Therefore the findings show that silver surfaces are better for producing high quality interfacial bonds for die-attach in power electronic packaging.

One key finding is the successful demonstration of the laser assembly method for chips with metal contact pads on the top surfaces of chips. In this work, silicon chips with contact pads with gold surface were used in the laser bonding work. The contact pads face the laser the beam in the bonding process so the laser radiation incident on the pad areas is reflected and not absorbed by the silicon chip to contribute to the laser heating process for sintering of the silver paste layer, and as a result the areas of the silver particle layer shadowed by the pads could experience a lower temperature than the other parts of the layer and hence possibility of insufficient sintering. However the findings from shear strength test and fracture analysis show successful sintering the silver paste layer producing a reliable joint. It is necessary to increase the laser power in the sintering/bonding process.

In study of failure modes, the findings show that in pressureless sintering the fracture in shear test occurs in the sintered silver layer. Since there is minimal pressure in sintering, the sintered silver layer has a large amount of voids with a more coarse microstructure and hence it has a weaker mechanical strength and thus the fracture occurs in the material in shear test. For copper-copper bonding as stated before the fracture is at the interface between the copper surface of the chip and the sintered silver layer. But for silver-silver bonding, the failure interface is between the silver surface of the chip and the silver surface of the remetallised of DBC substrate. This is for the same reason as in the case of copper-copper bonding since the silver paste was printed on to the DBC substrate with a remetallised silver layer.

In the investigation of reliability in thermal storage experiments, selected samples were stored in a furnace with a chamber temperature of 300C. Shear tests were conducted to determine the strength of samples after 100, 300 and 500 hours of thermal storage at 300C. The findings show significant degradation of shear strength in samples produced on DBC substrate with the copper surface. The reason is the effect of thermal oxidation of the copper surface of the DBC. But for the samples produced with both surfaces of the chip and the substrate being silver, the samples showed no degradation of shear strength after 500 hundred hours of thermal storage. The findings show that for high temperature applications, silver metallisation on both chips and substrates are necessary for long term reliability.

We also studied two different silver particle paste materials to compare their performances for power electronic packaging applications. The findings show the material with a mixture of nanoscale and microscale particles and high viscosity has much better performance than the other material with micron size particles and low viscosity. The high viscosity paste is better for stencil printing and the paste formulation produced the results that are as good as those reported in the literature and it is an excellent material for the laser based sintering process for die attach in power electronic packaging.

The findings of the research pave the way for adoption of the laser based sintering method for manufacturing of power electronic packages and modules based on the WBG semiconductor (SiC and GaN) devices for applications in the PEMD (Power Electronics, Machines and Drives) sectors that are of strategic and economic importance to the UK and internationally.

The project also led to the development of a 3D printing method for fabrication of MEMS based accelerometers and force sensors. The findings shows that it is possible to use 3D printing for fabrication/manufacture of sensor devices. The findings were published in IEEE Sensors and other journals.
Exploitation Route We are exploring technology transfer opportunities to bring the outcomes of the research to the UK industry for application/implementation in manufacturing of power electronic devices and modules in the UK. We had previous contact with two main UK based companies (Alter Technologies and Microchip, not to be made public) and resulted in the submission of an EPSRC proposal in 2021 to take the research outcomes forward to develop novel packaging processes. The research proposal received positive reviews but unsuccessful due to the limited amount of EPSRC funding available in the funding round. We are now conducting joint research (self-funded) with MCS Ltd and Karl Zeiss UK/US (confidential information) to investigate the use of advanced X-ray microscopy technology for nondestructive inspection and the advanced methods from MCS in material analysis for electronics. If the collaboration is successful it will create an advanced method for commercial application to take forward the outcomes of the research.

The project results were presented at a one-day national conference (Microtech'21) on 25 March 2021 organised by IMAPS UK. The annual Microtech conference is a major industrial event at national level in electronics manufacturing. The new research outcomes will be presented at Microtech'23 in March 2023. The results will be disseminated to the UK industry in the electronics manufacturing sector and for possible industrial takeup of the research outcomes.
Sectors Aerospace, Defence and Marine,Electronics,Energy,Manufacturing, including Industrial Biotechology
 
Description Our findings are being explored by two companies to develop a joint method in high resolution nondestructive and destructive inspection of microstructures of sintered nanoparticle paste layer in assembly (die-attach) of wide bandgap semiconductor (GaN and SiC) based power devices. Confidentially the two companies are MCS Ltd and Karl Zeiss UK/US. MCS is a leading service provider in material analysis in microelectronics manufacturing, Karl Zeiss is world leading provider of advanced optical and X-ray microscopes. We have provided unique samples to Karl Zeiss to use their facilities in US to study nondestructive analysis of the defects (voids) and microstructures in the laser sintered silver nanoparticle layer. It is not possible to use the conventional X-ray Micro-CT methods developed for solder based electronic assemblies. After the regions of interest have been found using the X-ray microscopy, MCS will carry out destructive studies of the cross-sections of the regions using their proprietary method and high resolution scanning electron microscopy (SEM) facility. If the collaboration is successful, an advanced material analysis method will be developed and used in commercial applications in advanced electronics manufacturing industries.
First Year Of Impact 2023
Sector Aerospace, Defence and Marine,Electronics,Manufacturing, including Industrial Biotechology
Impact Types Economic