Nothing to lose – Efficiency to gain
Firma zum Thema
The blue LED was a based on SiC, meaning that the light emission was coming from a SiC pn junction. However, SiC blue LEDs were relatively inefficient because SiC is an “indirect” bandgap semiconductor material, meaning that some heat has to be given up in order to release a photon. People knew that GaN was a “direct” bandgap semiconductor, meaning that it could be much more efficient, but there were very difficult issues in making p-type GaN at the time, so SiC was the only viable option.
Palmour: “SiC was a very stable and manufacturable product, but it was not very efficient, and would not have ever been practical for lighting applications. It was predominantly used for indicators lights, signage, and making R-G-B displays. One of the 2014 Nobel prize winners was Prof. Akasaki, who was the first to demonstrate how to make p-type GaN. Cree was pursuing GaN in the background, but once Nakamura (another Nobelist) made p-type GaN a manufacturable process, it accelerated our efforts. Nakamura and Nichia were the first to release a GaN blue LED, which had much higher efficiency than Cree’s SiC blue LED. Cree released its first GaN LED, grown on a SiC substrate, a little less than one year later.”
Largely ignored as an economically viable technology in high power applications, SiC struggled to gain the respect it deserved outside. After years of continuous quality improvements, Cree finally achieved the cost breakthrough it needed to drive larger-scale adoption of SiC into more advanced power applications.
In 2002, Cree introduced its first 600V SiC Schottky Diode. The new diodes made power factor correction in switch mode power supplies (e.g. power supplies for blade servers) much more efficient, which allowed server manufacturers to meet energy efficiency standards set by the ENERGY STAR program. The diodes also improved payback time on solar power installations by making the transfer from low-voltage DC current to high-voltage AC current much more efficient.
In 2011, the company introduced its first SiC MOSFET, bringing the product full circle for Cree. Ironically, before Cree marketed its first blue LED, Palmour and team built a SiC MOSFET that operated at 650°C to demonstrate the semiconductor’s capability.
“There is always a need for a more efficient switch and I believe our SiC MOSFET is poised for growth,” he said. “Our second generation of our SiC MOSFET improved performance at nearly half the price, and I expect our third and fourth generation MOSFETS will continue that trend.”
Palmour is confident that SiC-based switches and diodes will continue to outperform incumbent technologies such as silicon, which should lead to broader adoption and thus further cost reductions. But what really gets Palmour excited are the applications no one has considered yet.
HV applications presents whole new realm of possibilities
“Silicon carbide has inherent advantages in high-voltage applications where incumbent technologies aren’t considered a viable option. We are still challenged with gaining acceptance in today’s marketplace, but I think high-voltage applications present a whole new realm of possibilities for us.”
These new possibilities are what keep Palmour motivated. “Revolutionary switch technologies seem to come along once every 15-20 years and I think we’re on the verge of something big. I believe we’ve only begun to discover what’s possible for this technology and I want to see it come to fruition.” //KU