GaN based Blue Laser diode ECE 355 presentation Seiyon Kim Dept. of ECE and Microelectronics Lab
Outline -Motivation: Why we need BLUE lasers? -Material property: GaN vs. other wide bandgap materials -Growth Techniques -Dislocations -Blue LD: InGaN multi-quantum well structure -Future research -Conclusion
Why BLUE Laser? -laser printing (smaller optics, larger depth of field) -high density optical storage (630-650nm, 4.7GB 400-430nm, 20-50GB) -projection display (replace halide lamp) -spectroscopy -sensing
Why GaN devices? SiC, GaP : indirect bandgap II-VI : short lifetime (1~100hrs.) GaN : high efficiency, in spite of much more dislocations than II-VI(~10 10 >>10 3 cm -2 )
How to grow GaN? No GaN substrate - large lattice mismatch, bad quality Substrate: Sapphire(~13%), SiC(~3.5%), Spinel(Mg 2 Al 4 O ~9.5%), Si, GaAs, ZnO Equipment: MOVPE (TMGa, NH 3 ), MBE, HVPE Method: two-step growth (buffer growth+epitaxial layer) ELOG (epitaxially laterally overgrown GaN) PE(pendeoepitaxy)
ELOG Growth -fewer dislocation (~10 7 cm -2 ) -LD lifetime enhanced (~2500hrs)
Bulk GaN (TDI, October, 2001) First bulk GaN How to make? -they never tell us... 1.5 inch diameter Lower dislocation density than conventionally grown GaN 3~4 inch commercial wafer expected in 6 months
Dislocation One-, two dimensional structural defect High density in GaN related materials (~10 7-10 10 cm -2 ) Non-radiative recombination center - Sugahara et al. (CL/TEM) Electrically active, path of current, generate heat -cause to fail LD Characterization is not easy- generally studied by TEM, AFM, STM
InGaN MQW Laser
Future research - prolong life time (present: blue/violet ~10000hrs, blue~500hrs) high power LD for DVD (~30-40mW) and projector (~1W) reduce cost(~$5000 now!) - reduce defect density, develop better p-type ohmic contact - growth : GaN or AlN bulk, p-type doping, various substrates other than sapphire and SiC
Conclusion - GaN/InGaN/AlGaN system makes highly efficient blue/uv emitting laser diodes - There are several growth techniques to grow GaN layer with low defect density to overcome short lifetime of current LDs. (ELOG) - Dislocations act as non-radiative centers to degrade device performance
Questions?
Characterization by Selective Etching Crystalline region removed left dislocations New, easy way to count and characterize dislocations - high resolution spectroscopy(cl, NSOM)
Dislocation as nonradiative center (a) SEM image bright spot~whisker (b) CL image dark spots~nonradiative centers
Cross-sectional CL Smaller bandgap emission, larger YL in whisker region