Effect of Hydrogenation on the Electrical and Optical Properties of GaSb
- PDF / 368,252 Bytes
- 6 Pages / 414.72 x 648 pts Page_size
- 86 Downloads / 198 Views
form Photoluminescence Spectrometer with a resolution of 0.5 meV. Deep Level Transient Spectroscopy (DLTS) measurements were carried out to evaluate the extent of deep level passivation by hydrogen. To evaluate the thermal stability of the hydrogenated samples, they were heat treated in ultra-pure hydrogen ambient. Since annealing treatment reduces the overall PL efficiency of the sample (by introducing additional non-radiative centres), the annealing was simultaneously carried out on both the reference and the hydrogenated samples for comparison. RESULTS AND DISCUSSION Properties of surface defects in GaSb created by hydrogen plasma The existence of a layer of defects near the surface was deduced from the abnormal behaviour of the C-V characteristics [8]. During the C-V measurements, capacitance transient was observed which indicated slow emission rate of carriers from the defects. Fig. 1 shows the capacitance transient at 300 K for various reverse bias voltages. The 1/C 2 versus V plot at room temperature under steady state condition (t =oo) exhibits two slopes as shown in Fig. 2. At a reverse bias voltage Vp, the slope changes. The steeper slope corresponds to the donor concentration, ND, while the shallower one to donor plus trap concentration, (ND + NT). From Fig. 2, we evaluated ND and NT as 2 x 1017 and 7 x 1018 cm a respectively. Using the capacitance value at Vp, the thickness of the defect layer was calculated to be • 200 A. Here, it may be pointed out that the thickness of the defective layer is dependent on various experimental parameters like plasma power, substrate temperature, exposure time and doping concentration. For example, the width of the defect layer varied from a few angstroms to 500 A by increasing the exposure time from 30 minutes to 5 hours. The capacitance transients for different temperatures ranging from 77 K to 340 K with a fixed reverse bias of 0.8 V are shown in Fig. 3. The analytical expression for transient capacitance associated with defects whose concentration is comparable to the doping concentration (i.e. ND - NT) can be written as [9] : C(t)
- (1 +exp
/2
()
C(0o) with
NT6 2 NDW 2(0)
(2)
where C(t) is the transient capacitance at time "t", C(oo) is t0 pteady state capacitance, 6 is the thickness of the defect layer, W(oo) is the thickness of the space charge region at t =oco and e,, is the emission rate of the electrons from the #efects. Using the data of Fig. 3 and the above equation, we calculated the e,,'s at different temperatures. Since e, is proportional to a exp(-Ea/kT), the slopes of the e. versus 1/T plots gives the activation energies of the defects Ea's. Here a is the capture cross-section. Fig. 4 shows the plot of ln(rT 2) versus 1000/T where T = 1/e,. It is clear from the figure that the defects have multiple energy levels in the bandgap. The activation energies (Ea's) calculated from Fig. 4 varied from 5.7 to 215 meV. 534
30
ft
2.51
ft
2.0
0 0
10
20 Time (min.)
-02
-o.' -06 -0o8 Voltage (Volts)
30
FIG. 1. Capacitance transients for different
Data Loading...
