Strain-accelerated HF etching of AlAs for epitaxial lift-off

Abstract

Epitaxial lift-off (ELO) is a process which allows for the separation of a single crystalline III/V thin film or device from the substrate it was deposited on. This process is based on the selective etching of an intermediate AlAs release layer in an aqueous HF solution. The lateral etch rate of the AlAs release layer through a narrow crevice in the weight-induced epitaxial lift-off (WI-ELO) process is much larger than observed for unobstructed planar AlAs layers. It is possible that this increase in etch rate is caused by the tensile strain induced upon the AlAs layer in the WI-ELO setup. In order to verify this assumption, planar AlAs layers, subjected to a controlled curvature, were etched in HF solutions and their etch duration was measured. The applied curvature reduced the already present compressive strain due to lattice mismatch. For large applied bending radii no change in etch rate was observed, because the induced bending is smaller than the already present bending due to the lattice mismatch. Further bending induces a total compressive strain from −0.126% to −0.11%, resulting in an etch rate variation from 0.054 up to 0.066 mm h−1. Measurements on AlAs layers experiencing a tensile strain of +0.286% showed much higher etch rates of 0.134 mm h−1. The present results obtained on etching experiments in the lateral plane are extrapolated to the perpendicular direction so that a combination with the data from previous work becomes feasible. This results in a better microscopic picture of the etch front in the WI-ELO process. It is found that the force exerted by the weight can be projected on an area, limited by the sample width and a depth of approximately 6 µm.

The Influence of InxGa1–xAs and GaAs1–yPy Layers Surrounding the AlAs Release Layer in the Epitaxial Lift-Off Process

“…Therefore, during the last few years, the lateral etch rate and the release layer have been under serious investigation as a function of the process parameters. [7][8][9][10][11][12] The etch rate might be limited either by the reaction kinetics at the etch front or by the diffusion of the etchant (HF) and reaction products toward and from this front. For both situations, it is essential to know the reaction formula of AlAs with an aqueous HF solution.…”

Thin-Film III–V Solar Cells Using Epitaxial Lift-Off

“…Other parameters that have been investigated as strain applied to the wafer during the lift-off process [17], n-or p-type doping levels in the release layer [3], aluminium fraction x (must be >0.6) in the Al x GaAs release layer [3], the addition of a surfactant or anti-foaming agent [22] do not influence the etch rate to the same extent. In summary, it can be concluded that for optimized practically applicable settings an etch rate exceeding 30 mm/hr is feasible, resulting in a lift-off time of around 3 hours for a 4 inch wafer.…”

“…With this goal in mind the work on ELO at the Radboud University Nijmegen was started in the mid 1990s. An ELO setup was developed to allow for a systematic investigation on the basic parameters that determine the etch rate [3,15,16,17,18,19] and a model to describe the process was developed [20,21]. By making some essential modifications to the original ELO method, like the use of a temporary flexible carrier and applying a controlled force to allow the etchant to reach the etch front, the etch rate increased by more than a factor of 100 to rates well above 30 mm/h [16,19,21].…”

Surface depletion mediated control of inter-sub-band absorption in GaAs/AlAs semiconductor quantum well systems

“…6 Etching of the sample was undertaken using a mixture of NH 4 OH:H 2 O 2 :H 2 O (2:1:10), reported to selectively etch GaAs over AlAs, and a 10% mixture of HF, reported to selectively etch AlAs over GaAs. 7 The procedure followed was to place the sample in the NH 4 OH:H 2 O 2 :H 2 O solution for 5s followed by 10s in the 10% HF solution. After rinsing and drying the sample was immediately measured in the FTIR.…”