Nondestructive Analysis of Current Gain of InP/InGaAs Heterojunction Bipolar Transistor Structures using Photoreflectanc
- PDF / 395,076 Bytes
- 6 Pages / 420.48 x 639 pts Page_size
- 78 Downloads / 205 Views
recombination centers in the emitter layer, which sometimes degrade the current (PL) evaluation at low gain of GaAs-based HBTs, photoluminescence temperatures is known to be effective. 2'- However, the sensitivity of this method is low in regions with a large potential slope such as that near the emitter-base junction, especially at room temperature. In contrast, the sensitivity of photoreflectance (PR) spectroscopy at room temperature is high even in regions with a large potential slope. 4' 5 It has been reported that the electric field profiles of the emitter and the collector regions near the heterojunction of the HBT contactless electroreflectance can be measured by the PR or structure (CER). 61' 8' The electromodulation technique has proven to be effective in detecting structural degradation such as Be dopant diffusion from the base. 9 In spite of its advantages, there have been few reports of the PR method being used for nondestructive characterization of recombination centers in the HBT structure, which are major factors in the current gain reduction. In this paper, we report the correlation between the PR signal intensity from InP/InGaAs HBT wafers and the dc current gain. The measurements of the PR signal intensity reveal the crystal quality as it relates to recombination centers in the emitter and the collector region adjacent to the base layer. Finally, we show that the PR method is eminently suitable for use in the screening of the InP/InGaAs HBT wafers.
257 Mat. Res. Soc. Symp. Proc. Vol. 588 ©2000 Materials Research Society
EXPERIMENT Sample wafers with a HBT structure were grown on 3-inch Fe-doped semiinsulating (SI) InP(001) substrates by metal-organic vapor phase epitaxy (MOVPE). 0 The HBT structure is shown in Fig. 1. The structure was undoped InP buffer layer/n÷-InGaAs subcollector/an undoped collector/ptInGaAs InGaAs
base/n-InP
Thickness Doping (A) (cm- 3 ) n+-lnGaAs Emitter Ca n+-InP n-InP Emitter
500 200 500
3.0xl019 2.OxlO' 9 3.0x10 17
p+-InGaAs Base i-InGaAs Collector
500 3000
3.5x1019
n+-InGaAs
4000
i-InP Buffer
200
5.0x10 18
emitter/n÷-InP cap layer/n'(001) InP:Fe SI Substrate InGaAs cap layer. The n and p dopants were Si and C, Fig. 1. Schematic representation of the InP/InGaAs respectively. From the HBT structure used in this experiment. buffer layer to the base layer, all samples were grown under the same growth pressure, V/III ratio, and growth temperature. The growth temperature for the InP emitter layer and the layers above it was varied between 490 to 530 *C (sample A: 490"C, B: 510"C, C: 530°C). Before the PR measurements, the wafers were annealed in N 2 ambient, as in the reported InP/InGaAs HBT fabrication process, to remove hydrogen from the C-doped InGaAs layer and re-activate C acceptors.' 0 HBTs were fabricated on the sample wafers to measure the current gain. The emitter size was 50x50 ym. Wafers with a Si-doped 2000A n-InP single epilayer were also used as samples. The doping level of the layer was about 3x10' 7 cm-3 , almost the same as that of the HBT emit
Data Loading...
