Power = Heat is an equation embedded in every engineer’s subconscious supported by not only a more detailed theoretical knowledge but also experience -- whether one carries in a pocket a toasty smartphone sans cover, speedily “touch-types” on the uncomfortably warm keyboard of a laptop or designs applications that must contend with much higher power densities and temperatures than those encountered in consumer electronics.
Whether self-induced or environmentally placed, components often operate in harsh environments that challenge the performance of electronic devices that enable those applications. Turbine engine monitoring devices, power supplies in industrial environments, traction drives in electric trains and automobiles, fast chargers for electric buses or cars, and inverters in solar power plants are all such applications that come to mind.
However, the dominant semiconductor technology today involves silicon (Si), a material that does not perform well beyond the junction temperatures of 150°C. This junction temperature may be common for emerging high-power applications.1 Si, like other semiconductors, faces temperature dependent issues of intrinsic carrier density that may overwhelm lightly doped regions, exponential increase of reverse-bias junction leakage currents, thermionic leakage and degradation of carrier mobility.
And so, engineers must take steps to manage this heat.
Silicon: A drag on emerging applications?
Managing heat to keep silicon devices functioning is not an easy task. Thermal design not only consumes engineering time but weight and space allocations as well. A highly integrated power semiconductor solution may require a liquid-cooled heat sink that physically bloats the system to the point that it becomes cumbersome to transport or install. It may even affect the performance of the application itself.
For instance, an electric car’s Si-based drive system could easily weigh upwards of 33 lbs (15 kg). Most of this “dead weight” due to thermal management requirements the car must carry around. The result: shorter range between charges and therefore resistance to market entry.
A new dawn for silicon carbide
Wide bandgap (WBG) materials, like silicon carbide (SiC) and gallium nitride (GaN), offer temperature-dependent characteristics that take performance limits to levels not possible with Si. The nearly three times wider bandgaps of GaN and SiC mean devices using either of these materials can continue performing beyond Si’s range. Moreover, WBG materials exhibit higher breakdown fields, allowing thinner device structures and, therefore, lower on-resistance per unit area of the device.
SiC can be heavily doped to increase conductivity and yet maintain its high electric field breakdown. It is also a mechanically robust material: hard and inert with a low coefficient of thermal expansion (CTE) of 4.2.
GaN’s electron mobility is more than twice that of SiC, which makes it the sought-after technology today for very high-frequency applications. However, SiC’s thermal conductivity is nearly four times that of GaN, which makes thermal management easier.
SiC MOSFETs also operate at high frequencies compared to Si because they are majority-carrier devices exhibiting high edge rates during switching. The high switching frequencies mean lower switching losses and higher system efficiencies.
A future of highly efficient, high-temperature performance devices thus awaits the use of SiC.
Enabling electric transport
For the xEV automotive sector, comprising electric vehicles (EV) and plug-in hybrid vehicles (PHV), power systems are the most important with the battery subsystem being the most expensive and price inhibitive aspect of xEVs going from luxury to mainstream.
In EVs, the drive train is optimized for full-load capability but real-world driving conditions rarely allow the drive train to run constantly at full load. The higher efficiency enabled by SiC-based drives means that losses during non-ideal load conditions are lower. This translates directly to a longer EV range or lower costs of smaller batteries. Moreover, the high power density of SiC devices offers a solution to the limited space available in the engine compartment of EVs.
It is therefore no surprise that, according to reverse engineering efforts by System Plus Consulting in 2018, Tesla integrated a full SiC power module in its Model 3. In a more “demanding” EV application, Rohm designed a full-SiC power module-based inverter for the fourth season of the Formula E race series (2017). This enabled them to save approximately 13 lbs (6 kg) over a second-season (2015) conventional inverter and shave off 43 percent in volume.2
SiC benefits to EVs extend beyond traction to fast charging. The industry’s goal is to implement chargers that “fill-up” an EV in under 30 minutes. That would require the charger to deliver 80 kW to 100 kW DC, bypassing the onboard charger (OBC) of the car. SiC’s high power density and easier thermal management requirements enable such chargers to be smaller and more efficient. Similar advantages are available for OBCs as well; Renault-Nissan-Mitsubishi, for instance, plans to upgrade their 22 kW one-hour OBC for the Renault Zoe model using SiC power semiconductors.3
SiC-based traction drive applications extend to industrial uses, such as the N700S model of Japan’s high-speed Shinkansen bullet train. It is expected to roll out in 2020, 11 tons lighter than the previous version due partly to full SiC power drive and a natural air-cooling system.
However, the automotive market is expected to be the biggest beneficiary of SiC devices. That industry is expected to spend over $300 billion in xEV development, according to market research company Yole Développement.4 Yole predicts the SiC power semiconductor market’s value will approach $2 billion by 2024 at a compound annual growth rate (CAGR) of 29 percent for the 2018-2024 period. And they believe the automotive market will come to hold about half of the SiC device market share.
Charging the solar market
Solar power generation often uses 50 kW to 60 kW boost converters that offer efficiencies in the range of 98 percent. Using SiC, the switching frequencies can be increased by two to three times over Si and efficiency driven up to nearly 99.5 percent. This also reduces the size of the magnetics and other components, like link capacitors, for estimated three times space and ten times weight savings.5
The one-percent efficiency increase from SiC means that, if it were employed across the United States’ 60 GW installed solar generation, it would save 600 MW annually.6
Next-gen telecommunications
Next-generation 5G requirements, including the trend toward high frequencies, offer new opportunities for WBG devices. 5G allows for over 1 GHz bands to increase data transfer speeds and uses two new frequency bands at 3.5 GHz and 4.8 GHz.
Silicon technologies, like LDMOS, perform well at below 3 GHz but at higher frequencies, GaN offers higher efficiencies. Like SiC devices, GaN-on-SiC has much better thermal characteristics than Si, leading to higher power densities and therefore savings in size, weight and heat sink related costs.
GaN-on-SiC dominates the GaN RF market, according to Yole Développement, and is already being used by the 4G LTE infrastructure segment. The firm sees RF GaN, including GaN-on-SiC, particularly well-suited for power amplifiers (PA) in remote radio heads (RRH) of 5G networks at sub-6 GHz frequencies.7
The future is hot for SiC
The power electronics market is driven not only by the need for higher efficiency, as required by solar energy and data centers, but the unique requirements of less space and weight along with high temperature in electric transportation and industrial traction. Combined with higher frequency requirements in 5G and defense markets, the time is ripe for WBG devices.
Previously held back by high costs and low volumes, SiC is coming into its own with new investments in wafer supply, and continuing improvements in materials and manufacturing processes, such as II-VI’s prototyping of 200 mm semi-insulating SiC substrates for GaN-on-SiC PAs in 5G antennas.8
Therefore, the time is also ripe for engineers to discover how they can migrate their technology of choice to SiC.
References
- A. Elasser and T.P. Chow. “Silicon carbide benefits and advantages for power electronics circuits and systems”. In: Proceedings of the IEEE (June 2002).
- Rohm EV Solutions ver.1.2 catalog.
- STMicroelectronics Press Release, September 9, 2019: https://investors.st.com/news-releases/news-release-details/stmicroelectronics-supply-advanced-silicon-carbide-power.
- Power SiC 2019: Materials, Devices, And Applications, Market & Technology Report - July 2019, Yole Développement.
- Guy Moxey, Wolfspeed. “Silicon Carbide: Transforming the Future of Power,” February 21, 2019.
- “Silicon Carbide in Solar Energy,” Office of Energy Efficiency & Renewable Energy, U.S. Department of Energy: https://www.energy.gov/eere/solar/silicon-carbide-solar-energy.
- RF GaN Market: Applications, Players, Devices, and Technologies 2019-2024, Yole Développement.
- II-VI Pre Press Release, October 10, 2019:https://www.ii-vi.com/news/ii-vi-incorporated-unveils-the-worlds-first-200-mm-semi-insulating-sic-substrates-for-rf-power-amplifiers-in-5g-antennas/.
