CMP PROCESSING

Learn more about R&D chemical mechanical polishing by requesting our FREE informational CD.

Download our CMP White Paper

 

FREE subscription
Subscribe for free to receive each issue of Semiconductor Today magazine and weekly news brief.

News

12 November 2007

 

SiC electronics market to reach $800m by 2015

With the introduction of a silicon carbide (SiC) switch expected by 2010 expected to drive new developments in the automotive, industrial and IT fields, the annual SiC electronics device market should easily reach $800m by 2015, according to the new report ‘SiC 2007 - Market Analysis of SiC Electronics’ by Yole Développement.

It is now clear that the relative stagnation of the SiC power device market is partially due to a lack of reliable transistor technology, says the market research firm. The power factor correction (PFC) sector is the only one driving SiC device sales, and prospects for higher market penetration are linked mainly to a drop in device cost. In addition, others applications require a complete SiC-based switching cell (i.e. a diode and a transistor).

SiC metal–oxide–semiconductor field-effect transistors (MOSFETs) have been investigated by major SiC R&D teams, but it seems increasingly certain that junction gate field-effect transistor (JFET) or bipolar junction transistor (BJT) technologies may be launched on the market first.

With the introduction last January of 4” diameter SiC substrates coupled with the ZMP (Zero MicroPipe) technology acquired with substrate maker IntrinSiC Semiconductor Corp of Dulles, VI, USA in July 2006, Cree Inc of Durham, NC, USA is now marketing a product able to match the main requirements of power device makers. However, few have yet entered into the production phase and, with the exception of Cree, Infineon and newcomer STMicroelectronics, no other player is commercially active in this segment.

The material cost and small diameter of SiC substrates have always been cited as a constraint to explain the low interest from the big players in the power device sector. So, now that these two issues are being resolved, the prospect of developing a SiC switch is not so far off.

The SiC device business is not yet the most exciting money-making activity, acknowledges Yole . The market research firm estimates that sales of SiC-based power electronic devices generated revenues of about $15m in 2006. The only product that is commercially available is the Schottky barrier diode (SBD), which is now reaching the 1200V and 20A range. This component targets many possible applications, but it is mostly used in high-end PFC systems, where it brings some impressive added value, such as better power oscillation avoidance and the elimination of many passive discrete components. However, the introduction of a SiC switch should definitely boost the market, says Yole Développement.

Diodes and switches are coexisting quite well together but, even if the hybrid approach SiC-silicon approach is an option, full SiC electronics are highly desired for many applications.

To highlight this, hybrid electrical vehicles (HEVs) currently use silicon-based insulating-gate bipolar transistors (IGBTs) and diodes in their inverter module to power the (30–50kW) electric motors. This silicon chipset has to be cooled by a water-based system to maintain a device junction temperature of about 85°C. However, this requires a dedicated water-cooling system (separate to the cooling system for the engine, which can handle higher temperatures). One of the objectives of HEV manufacturers is to simplify this setup by implementing SiC-based electronics that can easily withstand 150°C and more. This would then allow the use of a single water-cooling system rather than two. This way, a money saving of about 15% on the power module could be achieved. However, this approach only applies for a full SiC electronics chipset, and would not be realistic with a hybrid SiC-silicon solution. In terms of the requirements for this particular application, 1200V/100A SiC single chips would be ideal.

Up to now, the main R&D effort has been oriented towards the launch of SiC MOSFETs. Many announcements have been made (e.g. by Rohm, Cree, Acreo, Toshiba etc). However, oxide reliability and poor electronic mobility issues remain partially unsolved. So, even if the MOSFET remains the most studied device type, alternatives such as the JFET and BJT exist and are coming within the scope of some of the most prestigious R&D groups, concludes Yole Développement.

See related items:

Microsemi enters SiC epiwafer supply agreement with SemiSouth

C9 to build fab based on isotopic SiC technology

GaN-based LEDs the main consumer of nitride materials

Article: SiC power devices: if only we had a switch...

Search: Silicon carbide

Visit Bookstore to purchase this report.

Visit: www.yole.fr