Open Access
Add to BibSonomy
Abstract
We investigate the influence of silicon implantation on wet lateral oxidation of AlAs and show that the introduction of n-type doping silicon ions permits the adjustment of the oxidation kinetics and anisotropy. Using mesas with selectively patterned implantation regions, we demonstrate the fabrication of oxide apertures unachievable using the standard process such as oxide lateral gratings whose pitch can range down to 4 µm and crosses with 40°-angle tips. This approach thus constitutes an easy and flexible way to engineer the oxidation process and opens the path to new confinement geometries for lateral confinement patterns in photonics devices and in particular VCSELs.
© 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
The oxidation of AlGaAs is an established process [1,2] by which thin layers of aluminium-rich semiconductor alloys exposed to a hot water atmosphere can be selectively locally transformed into areas of stable insulating aluminium oxide (AlOx). The resulting modifications of the material electrical and optical properties have been used as an efficient lateral refractive index step and carrier confinement element to create a variety of opto-electronic devices of academic and industrial relevance including Vertical-Cavity Surface-Emitting Lasers [3,4], edge-emitting lasers [2], whispering-gallery-mode micro-resonators [5], photonic crystal [6] and nonlinear [7,8] waveguides. In most of these applications, the to-be-oxidized AlGaAs layer composition [9] and the oxidation conditions [10] are chosen to lead to an isotropic transformation since, in that case, the semiconductor/oxide interface(s) delimiting the confinement aperture is free of deformation [11,12] and thus can be inferred from the water-exposed (etched mesa) interfaces by applying a simple scaling rule.
The ability to introduce and control (large) local differences in oxidation rates is however a very attractive feature in, at least, two circumstances. Indeed, such an anisotropy can be beneficially used to slow the oxidation kinetics at specific locations and thereby facilitate the fabrication of devices like VCSELs whose semiconductor/oxide interface shape and dimensions need to be critically and reproducibly obtained [13]. The other and more interesting application of the ability to engineer the oxidation anisotropy is to circumvent the smoothing effects originating from the diffusive nature of the oxidation i.e. to be in position to create devices with sharp oxide/semiconductor interfaces such as grating-based components [14]. To-date, the demonstrated ways to locally modify this semiconductor/oxide transformation are to use arsenic implantation [13], to perform (silica-capped) intermixing [15,16], or to exploit the oxidation of multilayers on non-planar substrates followed by regrowth [14]. Out of these techniques, the use of As implantation appears to be the most straightforward, effective and reliable. Indeed, it has been shown to lead to an increase by a factor of up to four in the oxidation rate beneath the implanted zones compared to beneath the as-grown areas, this difference being attributed to the enhanced diffusion through the gallium-interstitial-vacancies-rich material in and around the implanted areas. In contrast, the (impurity-free) intermixing methods have led to conflicting results whereby the intermixed zone can exhibit weakly enhanced or reduced oxidation rates depending on the sample exact layout (AlAs layer thickness, capping) and oxidation conditions. In this paper, we investigate the behavior of AlAs oxidation upon patterned implantation using n-type doping silicon ions, as a complementary approach to the above-mentioned arsenic-based implantation technique. This choice of implanting Si ions was made because it can tune the in-wafer-plane oxidation kinetics at least through an (indirect) electrochemical effect [17] in a way that not accessible by epitaxial doping and at depths that can be much greater than when implanting with heavier As ions. This study shows the tunable effect of the Si-implantation on the lateral oxidation kinetics enabling the realization of complex and finely controlled lateral confinement schemes.
2. Sample description and processing
The samples used in this study were grown by molecular beam epitaxy on <001>-oriented GaAs substrates and consist of a GaAs buffer and a 50-nm-thick undoped AlAs layer capped by 180 nm of undoped GaAs.
Various SPR700 photoresist patterns were defined by contact lithography on cleaved samples to enable selective implantation. The implantation of Si ions was carried out under an incidence of 7° to avoid channeling effects and using an energy of 180 keV that, according to SRIM calculations, leads to a peak in the Si ion distribution at the bottom interface of the (to-be-oxidized) AlAs layer. The removal of the implantation-protective resist was performed using a combination of solvent and oxygen plasma treatments. Etched mesas were created by another lithography step followed by an isotropic wet etch (H3PO4/H202/DI water with 3/1/25 dilution ratio). Upon resist removal, wet oxidation of the exposed AlAs boundaries was immediately carried out in a reduced pressure (500 mbar) furnace chamber currently commercialized by AlOxtec. The oxidizing atmosphere is composed of water vapor (10 g/l) and N2/H2 (95%/5%) forming gas supplied at a flow rate of 0.6 l/min. The substrate temperature used during the oxidations was selected to be 380°C to permit the oxidation of ∼40µm-side mesas in reasonable time (∼1 h) but also to be in a position to differentiate between processes with weak difference in activation energies. Real-time in-situ micro-reflectometry-based imaging [18] is used to gather the data presented and analyzed hereafter.
3. Experimental results
A typical colorized top-view image recorded when oxidizing a sample with both implanted with 1015 ions/cm2 (light blue) and unimplanted zones (reference – in grey) is presented in Fig. 1 together with cross-sectional schematics. On this image, the etched mesa edges appear as black lines, and the oxidized areas are white, the central light grey areas are unoxidized AlAs zones forming the apertures. In the upper part of the image, double mesa structures can be observed in which the larger perimeter delimits the unimplanted areas and whereby it highlights the fact that the etch rate of the GaAs substrate is faster in the implanted areas than in the unimplanted ones. Visual comparison between the upper and lower parts of the image reveals that, in this case and irrespective of the mesa shape and orientation, the oxidation in the implanted areas progresses at reduced rate with respect to unimplanted areas.
Fig. 1. Colorized infrared microscopy image measured in-situ after 27 min of oxidation (at 380°C, 500 mbar) of 42µm-size mesas with (top) unimplanted and (bottom) implanted (at a dose of 1015 ions/cm2) AlAs layer. The blue region highlights the implanted section.
From images such as the one presented in Fig. 1 and acquired at different times during the oxidation process, the oxidation apertures (i.e. the boundaries between the unoxidized and oxidized areas) are extracted using an image edge detection routine and a combination of square/circular fitting [19]. As shown in Fig. 2, the recorded temporal evolutions are found to be linear in all cases, suggesting that the processes were reaction-limited. The oxidation rates are then evaluated by a linear fitting of these curves which include at least 12 data points. It should be pointed out that the oriented oxidation rates that are extracted here are obtained from the analysis of the oxide progression from square mesas in the range where the oxidation fronts remain parallel to the etched facets. As such, the obtained rates do correspond to the ones that would be deduced from the oxidation of (long-edge) linear mesas of identical orientation. As already observed in [19,10–12], the oxidation of the AlAs layer exhibits a crystal-related anisotropy with two principal directions whereby the <100> directions oxidize faster (at ∼0.37 µm/min) than the <110> directions (at ∼0.28 µm/min) and the former controls the oxidation aperture geometry from circular mesas. Additional information provided in Fig. 2 shows that a silicon implantation with dose of 1015 ions/cm2 reduces the oxidation rates respectively to 0.23 and 0.22 µm/min. In these implantation and oxidation conditions, the process turns out to be nearly isotropic, a fact that is clearly confirmed by the images of the bottom part of Fig. 1 and, more specifically, by the close-to-circular shape of the oxide aperture obtained from the oxidation of the circular mesa.
Fig. 2. Kinetics of the lateral oxidation (at 380°C, 500 mbar) for unimplanted (pale blue) and Si implanted (1015 ions/cm2 – dark blue) areas. The round symbols correspond to circular etched mesas, the diamond symbols are for <100>-oriented square etched mesas and the square symbols are for <110>-oriented etched mesas. The quoted dimension is either the inner diameter or square side of the oxide aperture.
Similar analyses were performed for implantation doses ranging from 1012 to 1016 ions/cm2. As shown in Fig. 3, except for the lowest dose, the implantation of silicon ions into the AlAs layer is seen to slow the oxidation. This observation is in agreement with the behavior reported in Ref. [17] where epitaxially n-doped (silicon) AlGaAs layers were also shown to exhibit reduced oxidation rate (compared to p-doped layers) as a result of the electrochemical contribution to the oxidation reaction. As revealed on Fig. 3, the oxidation slowdown factor reaches the extremum value of 1.65 (in the <100> direction) for a dose of 1014 ions/cm2. The observed “bell-shape” relationship can be tentatively assigned to a trade-off between the n-doping oxidation slowdown and the defect-damage induced oxidation enhancement [13], bearing in mind that the defect density exhibits a steadily-rising but nonlinear dependence on the implantation dose [20] and is compounded with a localized strain field [20] which, in turns, can also modify the oxidation kinetics [21]. It is also worth noting that the oxidation fast (<100>-oriented) axis contributions are more significantly modified by the implantation than its slow axis components. This means that upon implantation, the oxidation processes exhibit reduced effective crystal-induced anisotropy, a phenomenon consistent with the introduction of implantation-induced disorder in the AlAs layer. Furthermore, for the highest dose 1016 ions/cm2, there is a sudden jump in the impact of the implantation that may be assigned to the fact that this dose might exceed the threshold of another stage in the AlAs amorphisation [20]. Finally, for the dose of 1012/cm2, the oxidation rate of the implanted layer is seen to be faster than for the unimplanted zone, a behavior believed to be driven by defect density-induced oxidation acceleration and the compensation of the unintentional p-doping of the AlAs layer.
Fig. 3. Oxidation slow-down factor induced by implantation for different types of etched mesas. This factor is 1 for unchanged oxidation rates while values greater than 1 (respectively lower than 1) correspond to oxidations of implanted (resp. unimplanted) areas that are slower than in unimplanted (resp. implanted) areas.
Up to this point, we have only considered mesa structures that had been either left as grown or implanted over their entire surface. In the remaining part of the article, we present examples of oxidation of mesas whose surface has been selectively spatially implanted. Indeed, one of the goals of the engineering the oxidation anisotropy is to demonstrate the ability to create oxide-confinement geometries that are not achievable with the standard oxidation technique.
In the first illustration, the implantation-driven oxidation is investigated to make grating structures that could prove useful to implement high-effective-index-contrast distributed feedback or Bragg reflection in waveguide devices such as those demonstrated in [14]. As shown on Fig. 4, the oxidation mesas for these tests are a 350-µm-long 50-µm-high rectangle oriented along the [1–10] directions that are fully implanted but for areas arranged to form gratings whose pitch is 40, 30 and 20 µm respectively and whose fill factor in unimplanted area is ∼0.45. As it can be observed on Fig. 4, the oxidation of these structures leads to periodically modulated oxidation fronts. For sufficiently wide (implanted/unimplanted) areas (topmost image), the progression of the oxidation fronts follows, as expected, the kinetics recorded in the case of spatially homogeneous structures. Near the interface between the unimplanted and implanted regions, it can be observed that the shape of the oxidation fronts are (asymmetric) triangles with extensions of ∼7 µm on the unimplanted side and ∼2 µm on the implanted side. This linear variation of the oxidation profile close to the interface between two materials presenting different oxidation rate is consistent with the observations and the model developed in [22] for the oxidation of multilayer structures with stepped oxidizing (composition) profiles. Furthermore, the triangle summits, which are located at the interface between the implanted and unimplanted regions, have progressed further than in the bulk regions on either side of this boundary. This enhanced localized oxidation is tentatively ascribed to the established lateral p-n junction and afferent oxidation electrochemical reaction since these profiles significantly differ from what has been observed in intermixing-controlled oxidations [16]. As shown in Fig. 5 and consistent with the above-described features, reducing the width and pitch of the unimplanted areas to nominally 2 and 4 µm respectively resulted in the fabrication of a lateral oxidation grating with a 4 µm period and ∼1-µm peak-to-trough corrugation. These results are on par with those obtained by the more complex oxidation of non planar surfaces [14].
Fig. 4. Examples of oxidation gratings (of period = 40, 30 and 20 µm) induced by selectively implanting (vertical) bands with 1016 Si ions/cm2 and performing the oxidation at 380°C for 60 min. The schematic cross-sections taken at the red/yellow dashed lines (top mesa structure) are shown on the right.
Fig. 5. Close-up view of a 4-µm-period oxidation grating created by oxidizing a 50µm-wide 1016-Si-ions/cm2 implanted grating at 380°C for 60 min.
Similarly, as presented on Fig. 6, oxidizing 42-µm-diameter etched mesas implanted with 1016 ions/cm2 except for a 2µm-wide cross oriented along the <110> directions enables the fabrication of oxide apertures with sharp profiles (<40° angles). We note the latter “flower corolla” patterns would be challenging to create using shaped mesas and standard oxidation techniques [23,18] while they might be useful to adjust the modal density of multimode VCSELs [24–26].
Fig. 6. Sequence of images of the oxidation at 380°C of a 42µm-diameter circular mesa implanted with 1016 Si ions/cm2 except for a 2µm-wide cross.
4. Conclusions
In this paper, the influence of silicon implantation on the characteristics of wet lateral oxidation of AlAs has been reported. It has been shown that the introduction of n-type doping silicon ions can reduce the oxidation rate by a factor of up to 1.65 while simultaneously reducing the crystalline anisotropy as result of an interplay between electrochemical and damage-induced effects. Using mesas with selectively patterned implantation regions, we have demonstrated that the fabrication of oxide apertures with sharp interfaces such as oxide lateral gratings whose pitch can range down to 4 µm and crosses with 40°-angle tips. This approach thus constitutes way to engineer the oxidation process to implement either complex or finely-controlled lateral confinement schemes and useful complements the previously-reported techniques based on the oxidation of intermixed or non-planar regions.
Funding
Horizon 2020 Framework Programme (M-ERA.NET 2 - Travel (project #7404)); Centre National de la Recherche Scientifique (Renatech, the French network of cleanroom facilities).
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. J. M. Dallesasse, N. El-Zein, N. Holonyak, K. C. Hsieh, R. D. Burnham, and R. D. Dupuis, “Environmental degradation of Al x Ga 1− x As-GaAs quantum-well heterostructures,” J. Appl. Phys. 68(5), 2235–2238 (1990). [CrossRef]
2. J. M. Dallesasse and N. Holonyak, “Oxidation of Al-bearing III-V materials: A review of key progress,” J. Appl. Phys. 113(5), 051101 (2013). [CrossRef]
3. D. L. Huffaker, D. G. Deppe, K. Kumar, and T. J. Rogers, “Native-oxide defined ring contact for low threshold vertical-cavity lasers,” Appl. Phys. Lett. 65(1), 97–99 (1994). [CrossRef]
4. J. M. Dallesasse and D. G. Deppe, “III-V oxidation: discoveries and applications in vertical-cavity surface-emitting lasers,” Proc. IEEE 101(10), 2234–2242 (2013). [CrossRef]
5. S. Calvez, G. Lafleur, A. Larrue, P.-F. Calmon, A. Arnoult, G. Almuneau, and O. Gauthier-Lafaye, “Vertically coupled microdisk resonators using AlGaAs/AlOx technology,” IEEE Photonics Technol. Lett. 27(9), 982–985 (2015). [CrossRef]
6. K. Welna, M. Hugues, C. P. Reardon, L. O’Faolain, M. Hopkinson, and T. F. Krauss, “Photonic crystal nanocavities in GaAs/AlGaAs with oxidised bottom cladding,” Photonics and Nanostructures - Fundamentals and Applications 11(2), 139–144 (2013). [CrossRef]
7. A. Fiore, V. Berger, E. Rosencher, P. Bravetti, and J. Nagle, “Phase matching using an isotropic nonlinear optical material,” Nature 391(6666), 463–466 (1998). [CrossRef]
8. M. Savanier, C. Ozanam, L. Lanco, X. Lafosse, A. Andronico, I. Favero, S. Ducci, and G. Leo, “Near-infrared optical parametric oscillator in a III-V semiconductor waveguide,” Appl. Phys. Lett. 103(26), 261105 (2013). [CrossRef]
9. K. D. Choquette, K. M. Geib, C. I. Ashby, R. D. Twesten, O. Blum, H. Q. Hou, D. M. Follstaedt, B. E. Hammons, D. Mathes, and R. Hull, “Advances in selective wet oxidation of AlGaAs alloys,” IEEE J. Select. Topics Quantum Electron. 3(3), 916–926 (1997). [CrossRef]
10. G. Lafleur, G. Almuneau, A. Arnoult, S. Calvez, and H. Camon, “Anisotropy in the wet thermal oxidation of AlGaAs : influences of process parameters,” Opt. Mater. Express 8(7), 1788–1795 (2018). [CrossRef]
11. S. Calvez, G. Lafleur, A. Arnoult, A. Monmayrant, H. Camon, and G. Almuneau, “Modelling anisotropic lateral oxidation from circular mesas,” Opt. Mater. Express 8(7), 1762–1773 (2018). [CrossRef]
12. K. Alfaro-Bittner, R. G. Rojas, G. Lafleur, S. Calvez, G. Almuneau, M. G. Clerc, and S. Barbay, “Modeling the lateral wet oxidation of AlxGa1−x As into arbitrary mesa geometries,” Phys. Rev. Applied 11(4), 044067 (2019). [CrossRef]
13. L.-H. Laih, H. C. Kuo, G.-R. Lin, L.-W. Laih, and S. C. Wang, “As+-implanted AlGaAs oxide-confined VCSEL with enhanced oxidation rate and high performance uniformity,” IEEE Photonics Technol. Lett. 16(6), 1423–1425 (2004). [CrossRef]
14. P.-C. Ku, J. Hernandez, and C. Chang-Hasnain, “Buried selectively-oxidized AlGaAs structures grown on nonplanar substrates,” Opt. Express 10(19), 1003 (2002). [CrossRef]
15. A. R. Massengale, T. Ueda, J. S. Harris, C. Y. Tai, M. D. Deal, J. D. Plummer, and R. Fernandez, “Localised impurity induced layer disordering for lithographic control of the lateral oxidation of AlAs,” Electron. Lett. 33(12), 1087 (1997). [CrossRef]
16. T. Gi and T. Miyamoto, “Shape control of AlGaAs selective oxidation by intermixing method,” Jpn. J. Appl. Phys. 55(11), 110303 (2016). [CrossRef]
17. F. A. Kish, S. A. Maranowski, G. E. Höfler, N. Holonyak, S. J. Caracci, J. M. Dallesasse, and K. C. Hsieh, “Dependence on doping type (p / n) of the water vapor oxidation of high-gap Alx Ga1− x As,” Appl. Phys. Lett. 60(25), 3165–3167 (1992). [CrossRef]
18. G. Almuneau, R. Bossuyt, P. Collière, L. Bouscayrol, M. Condé, I. Suarez, V. Bardinal, and C. Fontaine, “Real-time in situ monitoring of wet thermal oxidation for precise confinement in VCSELs,” Semicond. Sci. Technol. 23(10), 105021 (2008). [CrossRef]
19. S. Calvez, A. Arnoult, A. Monmayrant, H. Camon, and G. Almuneau, “Anisotropic lateral oxidation of Al-III–V semiconductors: inverse problem and circular aperture fabrication,” Semicond. Sci. Technol. 34(1), 015014 (2019). [CrossRef]
20. S. Creutzburg, E. Schmidt, P. Kutza, R. Loetzsch, I. Uschmann, A. Undisz, M. Rettenmayr, F. Gala, G. Zollo, A. Boulle, A. Debelle, and E. Wendler, “Defects and mechanical properties in weakly damaged Si ion implanted GaAs,” Phys. Rev. B 99(24), 245205 (2019). [CrossRef]
21. F. Chouchane, G. Almuneau, N. Cherkashin, A. Arnoult, G. Lacoste, and C. Fontaine, “Local stress-induced effects on AlGaAs/AlOx oxidation front shape,” Appl. Phys. Lett. 105(4), 041909 (2014). [CrossRef]
22. A. Fiore, Y. A. Akulova, J. Ko, E. R. Hegblom, and L. A. Coldren, “Postgrowth tuning of semiconductor vertical cavities for multiple-wavelength laser arrays,” IEEE J. Quantum Electron. 35(4), 616–623 (1999). [CrossRef]
23. M. Moser, S. Eitel, and W. Kaiser, “VCSEL with non-circular mesa and current confinement for higher order lateral mode emission,” U.S. patent US 8,355,423 B2 (January 15, 2013).
24. J. Lavrencik, S. Kota Pavan, V. A. Thomas, and S. E. Ralph, “Noise in VCSEL-based links: direct measurement of VCSEL transverse mode correlations and implications for MPN and RIN,” J. Lightwave Technol. 35(4), 698–705 (2017). [CrossRef]
25. T. Aoki, R. Kubota, S. Yoshimoto, M. Yanagisawa, T. Ishizuka, and H. Shoji, “High-speed VCSEL arrays for 400 Gbit/s data center interconnects,” SEI Technical Review (90), 71–75 (2020).
26. T. G. Czyszanowski, A. Brejnak, M. Gębski, A. K. Sokół, M. Marciniak, E. Pruszyńska-Karbownik, M. Wasiak, J. Muszalski, J. A. Lott, and I. Fischer, “Enhancing optical output power by breaking VCSEL circular symmetry,” in Vertical-Cavity Surface-Emitting Lasers XXV, K. D. Choquette and C. Lei, eds. (SPIE, 2021), p. 10.
