We have exposed our QW structures to different kinds of plasmas, containing variable mixtures of halogen-based molecules / hydrocarbons / nitrogen and Ar.
Previous studies [1], in particular based on a similar approach of using the QW as an internal probe of the ion – material interaction, have shown that atomic ions with energies in the range of 100 to a few 100’s eV penetrate III-V materials (like GaAs or InP) through a channeling process [2], even if their direction of impingement is not favorable for channeling. The angular dispersion is in general sufficient to produce channeling. The ions channel a certain distance into the crystal, and then stop due to de-channeling mechanisms. As a consequence, a quenching of the PL for the QWs in the region where the de-channeling occurred takes place.
We extended the analysis looking not only at the intensity of the PL lines, but also at their spectral characteristics upon exposure to the ion fluxes. Fig. 1 illustrates a selection of the effects that we were able to identify and explain [3]. On fig. 1-a, a sample was exposed to the particles flux from a plasma discharge in SiCl4/H2/Ar. Under the conditions of this plasma discharge, etching actually occurs at a rate ≈ 250-300 nm/min., and we had to be careful not to etch the whole QW structure immediately. Fig. 1-a shows that:
- the signal for the two QWs at the lower wavelength seems to disappear. These QWs are the ones closest to the sample surface
- the emission bands from the other QWs exhibit energy shifts
Transmission electron microscopy and secondary ion mass spectrometry (SIMS) show that actually the two QWs closest to the top surface have not been etched. SIMS shows the penetration of Cl-species below the surface, which probably play a role in the quenching of the PL for these QWs
The sample on fig. 1-a was undoped. On fig. 1-b, the same experiment is conducted for a sample with the same QW sequence as that of fig. 1-a, but n-type doped such that a large built-in electric field is created across the QWs region. One can see that this changes completely the shape of the PL lines. The reason for this change can be attributed to the quantum confined Stark effect (QCSE), as shown in [4]. Interestingly, as seen on fig. 1-b, the effect of the interaction with the etching ions is significantly more pronounced than for the undoped sample. We have shown that this is caused by a specific channeling interaction involving ions (probably Cl- ions) which strongly screen the QCSE.
Using plasmas containing N2 (which effect is to reduce the etch rate), very significant blue shifts of the PL lines are seen (up to more than 7 nm). These were associated with the incorporation of N deep in the sample, occupying interstitial positions, thus creating local compressive lattice deformation that leads to the blue shifts [3].
All these effects will be discussed in the context of a model that we established to describe the modifications of the luminescence properties of the QW structures under the effect of low-energy ion bombardment. We will also describe experiments on samples where we take benefit of the plasma etching to define nanoscale structures and examine the spatial variation of the PL/CL spectra after interaction with the ion beams.
[1] D. L. Green, E. L. Hu, P. M. Petroff, V. Liberman, M. Nooney, and R. Martin, J. Vac. Sci. Technol., B 11, 2249 (1993).
[2] R. Germann, A. Forchel, M. Bresch, and H. P. Meier, J. Vac. Sci. Technol., B 7, 1475 (1989).
[3] J.P. Landesman, J. Jiménez, C. Levallois, F. Pommereau, C. Frigeri, A. Torres, Y. Léger, A. Beck, and A. Rhallabi, J. Vac. Sci. Technol., A 34, 041304-1 (2016).
[4] L. Viña, E. E. Mendez, W. I. Wang, L. L. Chang and L. Esaki, J. Phys. C: Solid State 20, 2803 (1987).