Reduced Al-Ga interdiffusion in GaAs/AlGaAs multiple quantum well structure by introducing low hydrogen content SiNx cap
- PDF / 317,144 Bytes
- 6 Pages / 414.72 x 648 pts Page_size
- 50 Downloads / 164 Views
achieved with the same dielectric capping material. In this study, in order to investigate the dependence of DCQWD on the film growing condition, reactant gas ratio was varied during the growth of SiNx capping layers on MQW substrate by PECVD method. The dependence of DCQWD on the quality of SiNx capping layer was characterized by the hydrogen content in the capping layer. EXPERIMENT The GaAs/A1GaAs MQW semi-insulating GaAs substrate of MQW structure is shown in 7 nm thick GaAs wells with 10
structure used in this study was grown by MBE on without any intentional doping. The schematic diagram Fig. 1. As one can see, MQW structure consists of four nm thick A1O. 2Gao.8As barrier.
GaAs 100nm AIo.4Gao.6As 300nm AIo.2Gao.8As 40nm 4xGaAs(7nm)/AIo.2Gao.eAs(1 Onm) AIo.2Gao.aAs 40nm AIo.4Gao.6As 300nm
S.1 GaAs Sub. Fig. 1 MQW substrate structure. In order to find the dependence of DCQWD on the SiN. film growth condition, reactant gas ratio was changed during SiNg growth by PECVD method. NH3 gas flow rates were changed from 0 sccm to 40 sccm at fixed SiH4 gas flow rate of 20 sccm. During the growth, 30 W RF power was applied and the base pressure was kept at 0.9 Torr by adding N 2 gas. In this case, diluted SiH4 (5% in nitrogen) gas was used. The growth temperature and growth time were 300 °C and 20 min., respectively. After SiNx growth on MQW sample, thermal treatment of the samples was accomplished by RTA at 950 'C for 30 sec. Disordering of the MQW samples was observed by PL spectra at 9 K after removing the SiN. capping layer. The characterization of SiNx capping layers on the MQW sample was carried out using an ellipsometer, a surface profiler, Auger electron spectroscopic method and elastic recoil detection (ERD) method of Rutherford Back Scattering (RBS). RESULTS AND DISCUSSION As shown in Fig. 2, the thickness of SiNx film decreases with the increase of the
420
NH3 flow rate from 270 nm to 160 nm. The refractive index of SiNx film also decreases slightly with the increase of the NH3 flow rate from 1.85 to 1.83. The refractive indices indicate that these SiNx films are N-rich films. The data from Auger electron spectroscopy also showed that the atomic percent of nitrogen of the SiNx film is almost 67 % for all SiNx films as seen in Fig. 3, although nitrogen content of stoichiometric Si3N 4 is 57 %. However oxigen was not detected. As reported by Dun et al.'s [9], N2 gas can be used as a nitrogen source of the SiNs film grown by PECVD method. Since N 2 gas was used not only as a carrier gas but also as a nitrogen source in our experimental condition, all SiN. films could exhibit nearly same composition. This N-rich SiNg might be the result from the lack of SiH 4 in our experimental condition. However, although all SiNx films exhibit the same composition, the data from ERD reveals different characteristics in hydrogen content of the film. The hydrogen 72
30 W RF power 300 oC substrate Temp. growing 20 m1n.
250
0 sccm NH 20 sccm NH . ............ ---- 40scem NH3
......
200-
. 4....N.
64-
I-
150
0
10 20
Data Loading...
