InAs is an important Ⅲ-Ⅴ semiconductors with narrow bandgap,high electron mobility and many applications in micro- and opto-electronic devices [
Journal of Infrared and Millimeter Waves, Volume. 41, Issue 2, 420(2022)
Preparation of epi-ready InAs substrate surface for InAs/GaSb superlattice infrared detectors grown by MOCVD
Total reflection X-ray fluorescence spectroscopy (TXRF) and X-ray photo-electron spectroscopy (XPS) have been used to investigate residual impurities and oxides on polished InAs substrate surface wet cleaned by different solution combination. Metal impurities Si, K and Ca are routinely detected on the cleaned InAs surface and their concentration change with the variation of solution combination. A large quantity of particles (80 nm size) is measured on the InAs substrate surface with higher residual impurity concentration. An effective wet chemical cleaning procedure is presented to prepare InAs substrate surface with less residual impurity, small particle quantity and thin oxide layer, which are beneficial to high quality epitaxial growth.
Introduction
InAs is an important Ⅲ-Ⅴ semiconductors with narrow bandgap,high electron mobility and many applications in micro- and opto-electronic devices [
In this work,NH4OH,HCl and H2O2 based solutions were combined used with an aim to remove particle,residual impurities and leave a thin oxide layer on the substrate surface. Furthermore,an appropriate stoichiometric InAs surface were obtained in order to desorb easily during epitaxial growth process. With the help of TXRF,XPS,Scanning Electron Microscope(SEM)and surface scan measurement,an effective chemical cleaning procedure is presented for the preparation of high quality InAs epi-ready substrate.
1 Experimental
2-inch un-doped InAs wafers with(100)-orientation were sliced from an InAs(100)ingot grown in our laboratory by the liquid-encapsulated Czochralski(LEC)method[
|
Three wet chemical cleaning processes were used to remove wafer contamination including particulate contamination,organic contamination and metallic contamination et al. Residual metal impurities on the wafer surface have been analyzed quantitatively by TXRF. The cleaned wafer surface defects,including particles,scratches,large pits,etc. were detected and classified using a KLA-Tencor Candela surface scanner. Nature oxide layer on the wafer surface was shown by an ellipsometer measurement. XPS is used to check oxide composition. Epitaxial defects morphology and formation are analyzed by SEM and Transmission electron microscopy(TEM),respectively.
2 Results and discussion
2.1 Residual metal impurities and particles
TXRF was used to detect and compare metal contamination on the wafer surface treated with different wet solutions.
|
NH4OH:H2O2:H2O based solution(SC1)are quite effective to remove particles and widely used in the wafer cleaning process of Si and GaAs substrates [
Figure 1.KLA-Tencor Candela particle metrology maps of the three InAs wafer surface treated by different wet cleaning processes(a)sample A,(b)sample B,and(c)sample C
2.2 Nature oxide stoichiometry and thickness
As an epi-ready substrate for high quality epitaxial growth,the residual oxides on the InAs is of considerable importance. The surface of indium arsenide has a large number of indium or arsenic suspension bonds with an unsaturated state after chemical mechanical polishing. The arsenic dangling bond prior to indium dangling bond combines with oxygen atoms from the InAs substrate,forming As2O3 on substrate surface that will continue to react with InAs single crystal surface through chemical
|
Moreover,a substrate wafer with stoichiometric thin oxide layer,less particle and low residual impurity concentration is highly high quality epi-growth. In order to clarify the chemical composition of oxides on the InAs wafer surface from sample B and sample C,XPS was used to analyze the core level and valence-band spectra of the oxides,as shown in
Figure 2.XPS spectra of wafer B and C,details of the binding energy given in Table 4
|
From the measured results,it is able to distinguish the existed oxide from particular shapes of In3s and As3d spectra. Compounds In2O3,As,As2O3 and As2O5 were identified according to the specific binding energy. By calculating the peak intensity area ratio of indium and arsenic,the atomic ration of indium and arsenic are 0.8182 and 0.7061 for sample B and sample C,respectively. This result suggests that sample C has more arsenic rich oxide than sample B on its surface.
2.3 Epitaxial results
On the preparation InAs substrates of sample B and sample C,a 200 nm n+ LWSL bottom contact,an LWSL absorber of 2 µm thickness,a MWSL barrier of 200 nm and another 200 nm-thick p-MWSL contact were deposited in the same run by an Aixtron 2400G3 MOCVD growth system. Details on epitaxial design and growth are described in Ref.[
Figure 3.Normarski microscope(a)sample B,and(b)sample C
Figure 4.Hillocks defect from scanning electron microscopy after SL growth
Figure 5.Transmission electron microscope of hillocks defect
3 Conclusion
HCl-based solution,combined with SC1 is beneficial for the reduction of particles and metals on InAs wafer surface. Lower concentration of H2O2 in the NH4OH mixture produces thinner oxide and arsenic-rich oxide surface. InAs wafer with thin oxide and arsenic-rich oxide surface is good to high quality epitaxial growth. The lower defect density of 140/cm2 from SL by MOCVD is obtained through InAs substrate treatment technique.
[17] Lu X, Zhao Y, Sun W et al. Lattice perfection of GaSb and InAs single crystal substrate[J]. Pan Tao Ti Hsueh Pao/Chinese Journal of Semiconductors, 28, 163-166(2007).
[20] WU Jia, XU Zhi-cheng, Chen JX et al. Wet etching for InAs-based InAs/Ga(As)Sb superlattice long wavelength infrared detectors[J]. Journal of Infrared and Millimeter Waves, 38, 549-553(2019).
Get Citation
Copy Citation Text
Li-Jie LIU, You-Wen ZHAO, Yong HUANG, Yu ZHAO, Jun WANG, Ying-Li WANG, Gui-Ying SHEN, Hui XIE. Preparation of epi-ready InAs substrate surface for InAs/GaSb superlattice infrared detectors grown by MOCVD[J]. Journal of Infrared and Millimeter Waves, 2022, 41(2): 420
Category: Research Articles
Received: Jun. 18, 2021
Accepted: --
Published Online: Jul. 8, 2022
The Author Email: Li-Jie LIU (liulijie@semi.ac.cn)