2550
Control of Pore Depth in GaN Porous Structures Utilizing a Photoabsorption Process Under below-Bandgap Illumination

Tuesday, 15 May 2018
Ballroom 6ABC (Washington State Convention Center)
M. Toguchi, S. Matsumoto, and T. Sato (Hokkaido University)
Gallium Nitride (GaN) and related materials are attracting much attention as effective materials not only for photo-electronic devices but also for energy conversion devices such as solar cells and photocatalytic devices [1]. We have recently showed that the porous structures formed by the photo-electrochemical (PEC) process are very promising for the energy-conversion devices due to their large surface area and low photo-reflectance properties [2]. However, most of the works on the formation of GaN porous structures have been conducted on the high-doped layer having a carrier density of over 1018 cm-3. In view of the device designing, the decrease of carrier density in the absorption layer is the simplest and the most common method to improve the conversion efficiency since the photo-carries contributing to the photocurrents increased with the depletion layer width. In this study, we investigated the formation of the porous structures in the low-doped n-type GaN layer (5×1016 cm-3) grown on the GaN substrate. From a series of the experimental and theoretical results, we have concluded that the illumination of light with a lower energy than GaN-bandgap is very effective to increase the pore depth in the vertical direction.

For the formation of GaN porous structures, the standard electrochemical cell having three electrodes, i.e. a GaN working electrode, a Pt counter electrode and a Ag/AgCl reference electrode, was used with an electrolyte consisting of 1M H2SO4 and 1M H3PO4 solution. A Xe lamp was used for irradiation of monochromatic light whose wavelength was adjusted through a band-path filter. Firstly, the porous structures were formed by the standard PEC process in which an anodic voltage, Va, of 1 V was applied under the front-side-illumination (FSI) with a light wavelength, λ, of 350 nm. Then, the λ was changed to 380 nm corresponding to the photon energy, hv, of 3.26 eV which was below the bandgap energy of GaN (Eg = 3.4 eV).

From the SEM observation on the GaN surface, the high-density array of pores with a diameter of 40-70 nm and a depth of 250 nm was obtained after the PEC process with λ of 350 nm. However, the pore depth did not increase so much with an etching time. This is because the pore wall thinned to breaking point and was removed from the top-surface by the PEC reactions with photo-carriers generated near the surface [3]. On the other hand, the deeper pores were obtained on the sample formed with λ of 380 nm as compared with the sample formed with λ of 350 nm. Furthermore, the growth rate of pores in depth direction increased with the light power. In order to clarify the obtained results, the anodic currents observed with Va of 10 V and λ of 380 nm were compared between the porous and planar samples. The larger photocurrents were observed in the porous sample even under the transparent light with photon energy below the bandgap.

The Frantz-Keldysh (FK) effect is one possible phenomenon to explain the present results. The photons coming from the top-surface penetrated through the GaN but were absorbed by FK effect at the pore tips in which the high-electric field was induced due to the specific sharp structures. In such a situation, the holes are generated only at the pore tips and contribute to the anodization of GaN and pore formation. To clarify the strength of the electric field applied to the GaN porous structures, the potential distribution was calculated by a computer program for solving the 3D Poisson equation. It was clear that a higher electric field was induced at the pore tips and the potential near the surface was more steeply sloped than that of the planar substrate. The result of potential simulation strongly supported the experimental result; namely, the pore depth increased under the illumination of light with λ of 380 nm due to the redshift of the photoabsorption edge arising at the pore tips from the FK effect.

This work was supported by JSPS KAKENHI - JP15K13937,JP16H06421,JP17H03224.

[1] M. Deguchi et al., Jpn. J. Appl. Phys., 52, 08JF07 (2013). [2] Y. Kumazaki et al., J. Electrochem,. Sci., 161, H705 (2014). [3] A. Watanabe et al., ECS Electrochem. Lett., 4, H11 (2015).