April 1, 2009
More Re-affirming Device Data
Dear Customers and Development Partners,
In recent months, GaN-on-Diamond device data from multiple sources have come in further confirming earlier customer measured data, calculations, and simulations. We now know that if prepared poorly, our GaN-on-Diamond wafers will yield inconsistent and/or poor device results. Because we leave a thin protective layer of SiN atop our GaN-on-Diamond wafers before shipping, care must be taken in etching off the SiN layer prior to device processing. Poor removal of the SiN layer will hurt the 2DEG atop the GaN-on-Diamond wafer.
When the SiN layer is removed well from the GaN-no-Diamond surface, the transistor devices perform very well. Cornell University engineers (led by Professor Lester F. Eastman) recently published results at WOCSEMMAD ‘09 and WOCSDICE ’09 wherein the electrical characteristics (e.g. pinch-off voltage, I-V curves, carrier densities, etc.) of the GaN devices were unchanged after GaN layer transfer from Si to diamond. Record levels of Ft (>70GHz) and Fmax (>90GHz) values were demonstrated. Peak power densities of 4.5-7.7 W/mm, and Power Added Efficiencies (PAE) of 45-60% where reported. Notable passive parameters such as the GaN FET layer’s capacitance and carrier concentration were tracked over a 4-month period and little deviation was detected. The authors pointed out that they expected the device’s performance levels to improve in the weeks and months to come as the fabrication process matures.
We’re pleased at what these data represent. Our goals remain unchanged – to strengthen our capacity and throughput of our wafer production in the weeks and months to come so that more customers and development partners can build the next-generation of RF products. Shown here are selected device data recently published the Cornell engineers.
Warm Regards,
Felix Ejeckam
CEO |
