(Invited) From MRTA to SMRTA: Improvements in Activating Implanted Dopants in GaN
P-type dopant implantation and activation adds additional complexity to the synthesis of p-type GaN. The ability to implant and activate p-type species in GaN is crucial for devices that require selective area doping, such as many vertical devices. The implantation and activation of Mg in GaN is difficult because of the high temperatures required to activate the implanted dopants. Temperatures of at least 1200 ˚C are necessary to achieve Mg activation. At these high temperatures, GaN decomposes, and N vacancies, which are donors, form. Thus, for p-type dopant activation, it is necessary to avoid any compensating decomposition or defects. Advanced annealing processes, such as the Multicycle Rapid Thermal Annealing (MRTA) are required for Mg activation.
In this research, we compare the previously reported MRTA technique with a new method termed Symmetrical Multicycle Rapid Thermal Annealing (SMRTA). The MRTA technique includes a conventional anneal followed by rapid heating and cooling pulses (1). MRTA has previously been shown to activate ~8% of Mg implanted GaN (2). The SMRTA technique differs from the MRTA process by introducing an additional conventional anneal after the rapid heating and cooling pulses to improve the crystal quality and thus repeatability of the process. GaN films grown on sapphire substrates using MOCVD were implanted with a previously described implantation profile using a three-layer AlN capping structure (3, 4). One of the implanted samples was annealed using the MRTA technique, which included a conventional anneal at 1000 ˚C for 30 minutes followed by 40 rapid heating and cooling pulses with a peak temperature of 1350 ˚C. A second sample was annealed using the same MRTA process with the exception of an additional conventional anneal at 1000 ˚C for 30 minutes after the rapid heating and cooling pulses. Characterization performed on the samples includes atomic force microscopy for assessing the surface morphology and Raman spectroscopy for monitoring the crystal quality.
AFM measurements before and after the SMRTA treatment revealed that the surface morphology and surface roughness were unchanged after the annealing process. The RMS roughnesses before and after annealing were 1.55 and 1.65 nm, respectively. Thus, the three-layer cap which has been previously shown to protect the underlying GaN film during the MRTA process also adequately protects the surface during the SMRTA process (3).
To determine crystal structure changes during the annealing processes, the full widths at half maximum (FWHM) of the Raman E2 modes were compared before and after the SMRTA process. After the MRTA process, the E2 Raman mode FWHM decreases from 6.19 cm-1 (as implanted) to 5.4 cm-1 (MRTA annealed). The SMRTA process results in a further decrease in the E2 FWHM to 5.2 cm-1. This decrease in the FWHM indicates an improvement in the crystalline quality of the implanted GaN after the MRTA process and an even further improvement after the SMRTA process. The additional crystalline improvement afforded by the conventional anneal that distinguishes the MRTA and SMRTA processes is attributed to the removal of stresses and associated defects that form in the GaN layer during the rapid heating and cooling pulses. Electrical data for SMRTA annealed devices will also be presented. The additional improvement in crystal quality provided by the SMRTA process will be a key enabling step for future Mg-implanted devices.
1. B. N. Feigelson, T. J. Anderson, M. Abraham, J. A. Freitas, J. K. Hite, C. R. Eddy and F. J. Kub, J. Cryst. Growth, 350, 21 (2012).
2. T. J. Anderson, B. N. Feigelson, F. J. Kub, M. J. Tadjer, K. D. Hobart, M. A. Mastro, J. K. Hite and C. R. E. Jr., Electronics Lett., 50, 197 (2014).
3. J. D. Greenlee, T. J. Anderson, B. N. Feigelson, J. K. Hite, K. M. Bussmann, J. Charles R. Eddy, K. D. Hobart and F. J. Kub, Appl. Phys. Express, 7, 121003 (2014).
4. J. D. Greenlee, B. N. Feigelson, T. J. Anderson, M. J. Tadjer, J. K. Hite, M. A. Mastro, C. R. Eddy, K. D. Hobart and F. J. Kub, J. Appl. Phys., 116 (2014).