Quasi-Ballistic Stable Electron Emission from Porous Silicon Cold Cathodes
- PDF / 880,892 Bytes
- 6 Pages / 414.72 x 648 pts Page_size
- 22 Downloads / 232 Views
Mat. Res. Soc. Symp. Proc. Vol. 509 © 1998 Materials Research Society
graded-band multilayer (G-M) structure. Figures 1 (a) and (b) show the anodization and the corresponding schematic band structure of two experimental PS diodes with a normal structure and a G-M structure, respectively. The conventional normal structured PS was fabricated by a constant anodization current of 50 mA/cm2 for 2.5-5 min, as show in Fig. 1 (a). On the other hand, PS with a graded bandgap was formed by a gradually increasing anodization current from 0 to 100 mA/cm 2, as indicated in Fig. 1 (b). During this process, two low-porosity layers were formed inside the PS layer. The anodization condition for the low-porosity layers was 2.5 mA/cm 2 , 4 sec, with a thickness of -8 nm. This value has been proved to be the optimum thickness for the lowporosity layer in a multilayer structure [5]. Thus, the PS has a G-M structure. The total thickness of the PS layer was 10-20 p.m. Thin Au films with thickness of -10 nm were deposited onto the PS layers as top electrodes. The active area of the PS diodes was 6 mm in diameter.
Graded-Mulitijayer
Normal zW
z
D 0 z
D
w
o z
0 ANOD. TIME
0 ANOD. TIME Evac
Evac
PS (lO-
PS (i0°-20° m)
2O0m)
p~
L-Si Au
n+-Si
Au--I-Au
-8 nm
(b) (a) Fig. 1. Schematic illustration of the anodization and the band structure for (a) normal PS diode and (b) graded-multilayer PS diode. Measurements Cold emission properties were measured in a demountable UHV system (-10-s Pa). In the measurement of the emission properties, particular interests were paid into the relationship between the emission current fluctuation and the output electron energy distribution for the normal and GM structured samples. The fluctuation of diode and emission currents was evaluated by measuring the time dependence of diode current (lps) and emission current (/) under a continuous operation of the samples with a dc bias voltage (Vps). The energy distribution was measured using a conventional ac-retarding-field analyzer consisting of three parallel-plate electrodes, as shown in Fig. 2. The two inner plates were grids with a transmission rate of 50 % each, while the third one served as a collector. Grid G, was kept at a
194
positive potential to maintain a constant emission, while G2 was grounded, such that it is the same potential as the Au electrode. The collector was kept slightly positive with respect to G2 to prevent backscattering of electrons. The slow dc sweeping voltage, e.g. retarding voltage (VR), was modulated with a small ac signal, while the corresponding ac component of the collector current was detected by a lock-in amplifier and was recorded versus VR by a computer. The amplitude and the frequency of a small ac voltage superimposed on VR were 0.2-0.5 V and 123 Hz, respectively. All the grids and collector were coated with lampblack to decrease thgesecondary electron emission from the electrodes. A variable capacitance C was used to eliminate the capacitance of the Au electrode and the collector. This ac method is more d
Data Loading...
