Silicon Carbide Set To Reduce Size Of Hybrid Electric Engines
Since the introduction of the Toyota Prius in Japan in 1997, worldwide HEV sales have grown rapidly year on year. In 2003 they represented 0.15% of the total automobile market, and in 2005 this figure is predicted to rise to 0.5%. Car buyers are now being offered an increased range of models. In addition to Toyota, Ford and Honda have joined the hybrid revolution and DaimlerChrysler, General Motors, Hyundai, Nissan and others are expected to follow with the introduction of their own hybrids in coming years.
The advantages of HEV
With fuel efficiency ratings as high as 60 mpg, HEVs consume up to 50% less fuel per mile than many other cars, thereby reducing fossil-fuel emissions. As a result, government subsidies are available in many countries for HEV drivers living in high-congestion areas. Sales suggest that an increasing number of consumers are attracted to the efficiency, the environmentally friendly operation, and the subsidies associated with HEVs. But despite the clear attractions, HEV growth is projected to stagnate at around 3% of the total automobile market unless HEV prices fall to compete with traditional cars containing internal combustion engines (ICEs).
Current HEV platforms, which use silicon-based power electronics, are faced with two major challenges: size and weight. In addition to an ICE, HEVs must also accommodate power electronics, energy storage, and an electric motor in the predefined volume of the automobile platform. Engineering solutions to combat these challenges, such as using alternative frame or body materials and reducing either passenger or cargo space, result in less capable, more expensive HEVs than their traditional ICE counterparts.
The HEV s motor drive - a power-electronics component that converts stored energy into an alternating-current (AC) source needed to operate the electric motor - is one of the main contributors to the system s size and weight. Typically, HEV motor drives use silicon insulated-gate bipolar transistors (IGBTs) for the primary switching element, with silicon pin diodes as the fly-back diode, configured in a module designed to control three-phase motors. The module is positioned inside the engine compartment as close to the electric motor as possible to minimize parasitic inductance and reduce cabling weight.
The combination of silicon IGBTs and pin diodes is ideal for high-power applications, because the devices can be scaled to handle hundreds of amps per die. But, like all silicon devices, they are limited to junction temperatures of 150-175 °C. Controlling the junction temperature of the silicon electronics in the engine compartment s harsh environment requires large heat sinks and liquid cooling, but both of these solutions are costly and difficult to integrate into the volume available within the engine compartment.
The temperature limitations inherent to silicon technology mean that state-of-the-art silicon electronic components cannot meet the demands of HEV platforms and represent an area for significant improvement if the HEV market is to continue to grow.
This is where the opportunity for SiC lies. The ability of SiC-based power electronics to address these issues more efficiently than their silicon counterparts is a fundamental strength of the technology.
SiC is a wide-bandgap semiconductor that has been considered suitable for next-generation power electronics for many years. Fundamental material advantages, such as higher breakdown voltage and reduced thermal generation of intrinsic free carriers, separate SiC from traditional semiconductor materials such as silicon and GaAs (see table). As a result, SiC electronics can operate at substantially higher temperatures, power densities and frequencies than conventional silicon. The combination of these three strengths translates into smaller, lighter, and simpler electrical systems for HEVs.
Of course, SiC semiconductor devices face some challenges that must be overcome before they can reach the power levels demanded by HEVs and become economically feasible. SiC substrate costs have been high for many years and availability has been low, although this situation is starting to change as additional substrate suppliers create competition for customers.
Substrate quality and diameter have also limited the advancement of SiC devices. Crystal defects called micropipes have long been the nemesis of the material. They cause catastrophic device failure and inhibit chip scaling to higher current levels. However, over the last several years, micropipe densities have been reduced to one-hundredth of previous levels, and this trend is set to continue as the technology matures. Further improvements to SiC material quality and availability are also expected as the industry migrates to 4 inch substrates and beyond.
Another area of difficulty is that the device topologies exploited in silicon technology - such as CMOS and IGBT devices - are not directly transferable to their SiC counterparts for practical and physical reasons. Conventional silicon CMOS relies heavily on dopant diffusion. This process is unsuitable for the SiC material system, because dopant diffusion is negligible at practical processing temperatures of below 1500 °C. Current research is investigating dopant diffusion in SiC at temperatures above 1800 °C, but semiconductor processing equipment operating in this regime is practically non-existent, and heating substrates to these temperatures is likely to cause them to bow, warp and crack.
There is also an absence of commercially available SiC p-type substrates that are required for IGBT structures. Although researchers in Germany have produced p-doped SiC substrates by using a growth chamber with additional gas-flow, this work is still in its infancy (see "Additional pipework opens up transistor applications for SiC" Compound Semiconductor March 2005 p23).
The difficulties associated with dopant diffusion, and the lack of availability of p-type substrates, are just two revealing examples of why alternative control circuitry and power-switching device topologies must be developed for the unique SiC material system. One such alternative is the SiC Smart Power Chip, an all-SiC three-phase motor drive developed by Mississippi State University spin-out SemiSouth. This component will be able to operate in high-temperature environments, such as an engine compartment, without a large, complex cooling system. As a result, it should greatly reduce the size, weight and complexity of HEV system design.
This future product is the result of the combined efforts of SemiSouth and the US National Institute of Standards and Technology (NIST). Under NIST s Advanced Technology Program, SemiSouth is working to develop and integrate advanced SiC control circuitry and power transistors in a compact power module for use in HEVs. By using SiC, the HEV motor-drive-inverter volume can be cut by more than 50% and the liquid cooling system can be eliminated.
Although SiC smart-power technology is a well discussed topic, it has received limited scientific investigation. Proposed solutions have been met with skepticism, mainly related to the operation of the power switch. SiC bipolar junction transistors (BJTs) suffer from low current gain and forward-voltage degradation. Similarly, SiC MOSFETs are plagued with high-temperature reliability concerns and threshold voltage shifts.
SemiSouth has also developed a SiC power FET technology that overcomes the MOSFET and BJT problems. The technology is ideal for efficient power conversion at high temperatures because the absence of the metal-oxide-semiconductor region improves device reliability, and its unipolar nature eliminates any effects from forward and reverse recovery. SemiSouth believes that this technology will be capable of operating at much higher temperatures (300-500 °C) and frequencies (≥1 MHz) than conventional silicon, as well as reducing the overall size and weight of HEVs. Some progress has already been made in this area, with SemiSouth s engineers developing a variety of high-power devices that have operated at 500 °C.
Through continued research and advancements in technology, such as those undertaken by SemiSouth, future control electronics will be realized through the monolithic integration of SiC FET devices to form logic gates and operational amplifiers. These basic building blocks will combine to create ICs for controlling HEV motor drives. In addition, discrete, high-power SiC FETs will be integrated with the monolithic control elements to form a complete multichip power module.
All of these technological advances are expected to be integrated into HEVs during the next three to five years, and together they should lead to lighter, smaller, and more fuel-efficient cars that will in turn promote further growth in this industry.