Si-based complementary metal-oxide-semiconductor (CMOS) is facing physical limitation resulting from the scaling down of devices. In addition, several issues such as short channel effect and increase in the leakage current become critical. In order to solve those problems in the Si-based CMOS, introduction of a new channel material for the transistor such as III-V compound semiconductors and Ge can be one of the solutions due to their high electron mobilities [1, 2]. Among the III-V compound semiconductors, InGaAs is receiving an attention for improving device performance because it has electron mobility as high as 11400 cm²/Vs [3]. However, it is known that III-V semiconductors are not as robust as Si against the wet chemistries. Therefore, in order to introduce InGaAs as a new channel material, surface preparation of InGaAs in wet chemical processes should be controlled. In the current study, we investigated the effect of various wet chemical treatments on the InGaAs surface.

In order to study surface reaction of InGaAs such as oxidation and etching, undoped epitaxial InGaAs (100) on InP:Fe wafers were used. InGaAs samples were dipped in the solutions such as HCl, HCl/H₂O₂ mixture (HPM), NH₄OH, and NH₄OH/H₂O₂ (APM). The chemical state of InGaAs surface was observed by X-ray photoelectron spectroscopy (XPS) and the oxide thickness and etching rate of InGaAs were measured by transmission electron microscopy (TEM) and spectroscopic ellipsometry.

The chemical states (Ind3_5/2, Ga3d and As3d) of InGaAs surface after HCl, HPM, deionized water, NH₄OH and APM are shown in Fig. 1. It is first observed that the oxidation states in In3d, Ga3d and As3d (In₂O₃, Ga₂O₃ and As₂O₃) decreased after HCl treatment as compared to as-received InGaAs surface, while they increased after NH₄OH treatment. On the other hand, significant increases in oxidation states of InGaAs were observed after the treatment of HPM or APM, which suggests that oxidation of InGaAs surface is accelerated by H₂O₂ in aqueous solutions. It is also noted that InGaAs surface was also oxidized after deionized water treatment. Although E-pH diagram is based on the theoretical calculations of thermodynamical equilibrium in a single metal-water system, based on each E-pH diagram of In, Ga and As, it is expected that InGaAs surface can be dissolved in both acidic and basic solution [4]. Therefore, from the fact that the XPS oxidation states of the InGaAs surface after basic treatment were larger than in acidic solution as shown in Fig. 1, it is suggested that dissolution of the oxidized InGaAs layer is dominant in acidic solution, while it is a rate limiting step in the basic solution. Those findings imply that further increase in the oxidant concentration in acidic solution (HCl or HPM) will accelerate overall etching kinetics of InGaAs, while its impact in the basic solution is smaller.

The effect of process sequence on the residual oxide thickness on InGaAs surface was investigated and the results are shown in Fig. 2. The oxide thickness was obtained by ellipsometry and TEM measurements, and the results are relatively well matched each other as shown in Figs. 2(a) and (b). It is here noted that a significant increase in the oxide thickness was observed with 1/1/100 HPM (1HPM) as compared to 1/0.1/100 HPM (0.1HPM) in either HF-last or HPM-last process, which suggests a higher concentration of H₂O₂ in HPM solution produced an increased oxide thickness on InGaAs surface. The amount of material loss of InGaAs epi layer during each sequence is also given in Fig. 2(a). As expected from the observation of XPS spectra, once 1/1/100 HPM (1HPM) is introduced in the wet processes, the loss of InGaAs was much larger due to active and balanced surface oxidation and etching, as compared to 1/0.1/100 HPM (0.1HPM) process. Based on those results, it is concluded that the concentration of H₂O₂ in HPM solution is one of key factors to determine the residual oxidation and overall etching rate of InGaAs layer, which strongly suggests that the concentration of H₂O₂ especially in HPM process should be optimized at the surface preparation of InGaAs semiconductor.

References

[1] J. A. del Alamo, Nature, 479, 317 (2011).

[2] D.H. van Dorp, S. Arnauts, D. Cuypers, J. Rip, F. Holsteyns, S. De Gendt, Solid State Phenom., 219, 56 (2015).

[3] W. Lee and C. G. Fonstad, J. Vac. Sci. Technol., B, 4 (1986).

[4] M. Pourbaix, Atlas of Electrochemical Equilibria in Aqueous Solutions, National Association of Corrosion Engineers: Houston, Texas, (1974).