GaN growth on Si is very attractive for low-cost optoelectronics and high-frequency, high-power electronics. It also opens a route towards an integration with Si electronics. Early attempts to grow GaN on Si suffered from large lattice and thermal mismatch and the strong chemical reactivity of Ga and Si at elevated temperatures. The latter problem can be easily solved using gallium-free seed layers as nitrided AlAs and AlN. The key problem for device structure growth on Si is the thermal mismatch leading to cracks for layer thicknesses above 1 μm. Meanwhile, several concepts for strain engineering exist as patterning, Al(Ga)N/GaN super-lattices, and low-temperature (LT) AlN interlayers which enable the growth of device- relevant GaN thicknesses. The high dislocation density in the heteropitaxial films can be reduced by several methods which are based on lateral epitaxial overgrowth using ex-situ masking or patterning and by in-situ methods as masking with monolayer thick SiN. With the latter method in combination with strain engineering by LT-AlN interlayers dislocation densities around 109 cm -2 can be achieved for 2.5 μm thick device structures.
|Original language||English (US)|
|Title of host publication||Physica Status Solidi C: Conferences|
|Number of pages||24|
|Edition||6 SPEC. ISS.|
|State||Published - 2003|
ASJC Scopus subject areas
- Condensed Matter Physics