Impact of Plasma-Induced Surface Damage on the ... - ACS Publications

Article pubs.acs.org/JPCC

Impact of Plasma-Induced Surface Damage on the Photoelectrochemical Properties of GaN Pillars Fabricated by Dry Etching Wei-Jhih Tseng,*,†,‡ D. H. van Dorp,† R. R. Lieten,† P. M. Vereecken,†,§ R. Langer,† and G. Borghs‡ †

IMEC, Kapeldreef 75, 3001 Leuven, Belgium Department of Physics and Astronomy, Katholieke Universiteit Leuven, 3001 Leuven, Belgium § Centre of Surface Chemistry and Catalysis, Katholieke Universiteit Leuven, 3001 Leuven, Belgium ‡

ABSTRACT: n-GaN pillar photoanodes are fabricated by dry etching of a planar GaN epilayer. The increased surface area results in a plateau photocurrent enhancement of 84%. However, surface damage is introduced during dry etching. In this work, the surface damage is controlled by the RF chuck power. The GaN pillars fabricated using the lowest RF power show a similar current onset potential and current−potential slope as the planar GaN. In addition, the damaged GaN surface of the pillars can be removed in NaOH solution, which leads to the plateau current enhancement of 100% and the onset potential shifts −60 mV with respect to planar GaN. A pair of anodic and cathodic peaks is found in the dark cyclic voltammogram of the damaged pillars, which indicates the charging and discharging of the deep-level traps existing at 0.6 eV below the CB edge.

■

INTRODUCTION Water photoelectrolysis using semiconductor photocatalysts is a sustainable approach for the production of hydrogen, a green and renewable chemical fuel.1,2 The solar energy is directly converted and stored in chemical bonds: H2O → H2 + (1/2)O2 (E0 = 1.23 eV). Upon illumination, the semiconductor absorbs the supra-bandgap light and generates photocarriers. After carrier separation in the space charge layer, the photoelectrons and holes migrate to the cathode and anode, respectively, in the photoelectrochemical (PEC) system, performing reduction and oxidation of water at the solution/electrode interface.3,4 Selfdriven photoelectrolysis is desirable considering energy conversion efficiency and system simplicity, which requires conduction band (CB) electrons and valence band (VB) holes with an energy higher than the electrochemical potentials E0(H+/H2) and E0(H2O/O2), respectively. Many semiconductor materials have been investigated extensively with regard to their PEC properties.5−7 GaN-based semiconductors have drawn significant attention for use in photoelectrolysis.8−17 The position of CB edge and VB edge of GaN satisfies the energetic requirement for the self-driven photoelectrolysis.8 The bandgap of the semiconductor can be reduced by alloying it with indium (0.7−3.4 eV), thereby covering the complete solar spectrum.18,19 Surface nanostructuring is found to play an important role in the photoactivity of GaN. The diffusion distance for minority carriers in the charge-neutral region is shortened in the nanostructure. Furthermore, the enlarged depletion area at the nanostructures improves charge separation, leading to a higher photocarrier quantity at the surface.20 Moreover, the abundant © 2014 American Chemical Society

reactive sites at the nanostructured surface facilitate charge transfer at the solution/semiconductor interface.21 Each intermediate state of water oxidation has a characteristic activation energy which is plane dependent.9,10 By revealing particular crystal facets at the GaN nanostructure, the charge transfer efficiency can be improved. Different nanostructuring approaches have been developed at the GaN surface.11−14 Taking advantage of anisotropy in etching, nanostructures of high aspect ratio can be achieved by dry-etching methods.22,23 Using proper mask technology, the dimensions and topography of the nanostructures can be sophistically designed and fabricated after considering the influence of light absorption, charge transfer, and electric resistivity.15,16 In our recent work, we used a simple dry-etching approach to nanoroughen a planar GaN epilayer into a nanopillar array.24,25 The effective surface area of the n-GaN photoanode increases up to 6 times for 1.6 μm-high pillars; the plateau current density increases 84% with respect to the planar GaN. However, the dry-etching induced surface damage promotes surface recombination. Consequently, a higher applied potential is required to prevent charge carrier recombination. In the photoluminescence measurement, the intensity of nearband-edge photoluminescence of GaN drops half after surface roughening, indicating nonradiative recombination is enhanced. Suppressing the formation of plasma damage or removing the damage after dry etching is therefore important in order to Received: April 2, 2014 Revised: May 4, 2014 Published: May 5, 2014 11261

dx.doi.org/10.1021/jp503119n | J. Phys. Chem. C 2014, 118, 11261−11266

The Journal of Physical Chemistry C

Article

improve the photoelectrochemical properties of GaN nanopillars. In this present work, two different strategies for reducing plasma induced surface damage on GaN pillars are presented: plasma bombardment energy control during dry etching and post wet etching to remove surface damage in alkaline solution.

■

EXPERIMENTAL SECTION

A Ga-polar GaN (0001) layer is epitaxially grown on a Si (111) substrate by metal organic chemical vapor deposition (MOCVD).26 The top 2 μm of the n-GaN layer is Si doped with an electron concentration of 5 × 1018 cm−3. The full width at half-maximum (fwhm) of the GaN X-ray diffraction measurement (Panalytical, Philips) is 450 and 550 arc second for the (0002) and (101̅2) reflection, respectively. Fieldemission scanning electron microscopy (Nova, NanoSEM) was used to assess the surface morphology. The GaN nanopillars were fabricated by reactive-ion etching (RIE) dry etching with inductively coupled plasma (ICP) source (Plasma Lab 100, Oxford) using a self-assembled Ni cluster mask.14,25 After dry etching, the Ni mask and residual SiO2 were removed in HNO3 solution and BHF, respectively. GaN pillars of 1.6 μm height were used for this work. The RF chuck power was controlled between 13 to 125 W to introduce different content of surface damages. The rest of the dry-etching parameters remain unchanged during the etching (ICP source power/Cl2 flow/ Ar flow/chamber pressure/temperature = 125 W/10 sccm/5 sccm/5 mTorr/20 °C). After pillar fabrication using 125 W RF power, the samples were immersed into 1 M NaOH aqueous solution in the dark for 1 to 15 min to etch the damaged pillar surface. No solution agitation was applied. The ohmic contact to n-GaN was formed by a metal stack deposition of Ti/Al/Mo/Au at the sample edge, alloying at 700 °C for 1 min in an N2 environment. The GaN sample was mounted on a metal stripe-containing glass slide using a conductive silver paste (Pelco 187, Ted Pella). PVC tape and epoxy (Torr Seal, Varian) were used to define working area of the electrode and to shield other components from exposure to the solution. The PEC measurements were performed using a UV-transparent quartz cell with n-GaN samples as the working electrode (WE) and a 4.5 cm2 Pt mesh as the counter electrode (CE). All potentials are given with respect to a Ag/AgCl reference electrode (Metrohm). A potentiostat (Autolab, Metrohm) was used for current−potential control. All linearscan voltammograms were recorded at a constant scan rate (v) of 100 mV/s from negative to positive potentials, unless otherwise stated. A 0.1 M HBr and 0.2 M Na2SO4 (SigmaAldrich) electrolyte solution was used, with Br− acting as a hole scavenger to avoid photoetching of GaN during the photoelectrolysis. The solution was stirred using a magnetic stirring rod during all PEC measurements. A 150 W Xe lamp (LSB521, L.O.T.-oriel) was used as the light source for the PEC measurements. An air mass 1.5 filter (LSZ289, L.O.T.-oriel) was applied to filter 97% of the UV light (λ < 400 nm) to reduce sample degradation during the photoelectrolysis. The light intensity of the integrated radiance spectrum illuminated at the sample surface was 63 mW/cm2, measured by a thermopile detector (s302c, Thorlabs).

Figure 1. (a) A picture of an as-fabricated nanoroughened GaN pillar array (black) in contrast to the surrounding unetched planar GaN epilayer (light gray). SEM images of the 1.6 μm-high pillars (b) in top view and (c) at 45°. Scale bar: 2 μm.

a great capacity of light absorption, as evidenced by its black color. An SEM top view and 45° side view are shown in Figure 1, parts b and c, respectively. The pillars are homogeneously, distributed with an average density of 3.0 ± 0.4 × 108 cm−2. No specific crystal facets were visually observed at the surface of the GaN pillar structure after dry etching, except the top plane has an orientation of (0001), which is protected by a SiO2 layer during the dry-etching process. Figure 2 shows a photocurrent (jph)−potential (U) plot for 1.6 μm high n-GaN pillars fabricated by dry etching. The chuck

Figure 2. jph−U for the 1.6 μm high n-GaN pillars fabricated by dry etching at different RF powers. The dashed curve is for the jph of the unetched planar GaN. Solution: 0.1 M HBr + 0.2 M Na2SO4; v: 100 mV/s.

RF powers varied from 13 to 125 W. The dashed jph−U curve is for the planar GaN layer prior to surface roughening. Upon supra-bandgap illumination, excess electron−hole pairs are generated. At potential below the onset potential (Uonset for planar GaN = −0.36 V), carrier recombination is the predominant process; no photocurrent is observed. In the potential range between −0.36 V − 0.06 V, jph of the planar nGaN increases due to partial separation of the electron−hole pairs by the electric field of the space charge. The photogenerated electrons escape to the counter electrode where they cause water reduction, while the photogenerated holes oxidize Br− at the solution/electrode interface. For U > 0.06 V, jph reaches a plateau value. In this potential range, the electric field

■

RESULTS AND DISCUSSION Figure 1a shows an image of a GaN sample with a nanopillar array at the center and the unetched GaN at the edge as a visual reference. In contrast to the planar GaN, the pillar array shows 11262



Article

fabricated pillars and planar GaN, respectively. Both Uonset and Uplateau shift toward negative potential with increasing immersion time. This suggests that the plasma-induced surface damage is effectively removed by the alkaline solution. The 1 min-NaOH-etched pillars show a similar ΔUonset‑plateau and j−U slope to plateau as the pillars of 13 W RF power in Figure 2. The jph in the plateau range (U = 0.5 V) is enhanced by the alkaline treatment: 8.7% and 11.4% for 5 and 15 min of etching, respectively. The morphology of the GaN pillars after immersion in NaOH solution has been assessed by SEM. The as-fabricated pillars are shown in Figure 4, part a with part aa as a

is sufficient to separate the photogenerated charge carriers in the space charge and diffusion layers. All pillar photoanodes have a comparable jph in the plateau range. When the RF chuck power is increased, there is an evident Uonset shift toward more positive potential, and the larger overpotential (ΔUonset‑plateau) required to reach plateau potential (Uplateau). As the ICP discharge is applied during dry etching, activated chlorine species chemically react with gallium at the sample surface and form etch products.27 Meanwhile, Ar and Cl ions are accelerated to the sample surface for an additional physical bombardment to remove the GaCl complexes and constituent atoms, which allows the chemical etching of the succeeding layer to proceed. The physical desorption of the etch products is important for controlling the etch rate and the anisotropy of the etched surface. However, the bombardment of energetic ions creates damage at the surface and near the surface region.28 Specific plasma-induced surface states or traps act as efficient carrier recombination centers,29 which reduce the surface photocarrier concentration for the reagent oxidation at the GaN/solution interface. Therefore, for the most damaged pillars (125 W), see Table 1, a larger ΔUonset‑plateau is required to Table 1. Photoelectrochemical Properties of Different nGaN Pillars sample planar 125W 125W15′ 13W

RFa (W) NaOH etchb (min) Uonsetc (V) ΔUonset‑plateau (V) 125 125 13

15

−0.36 −0.30 −0.42 −0.37

0.42 0.67 0.46 0.41

a

The other parameters of the dry-etching process for all pillar samples: ICP power/Cl2 flow/Ar flow/chamber pressure/temperature: 125 W/ 10 sccm/5 sccm/5 mTorr/20 °C. bThe samples are etched in 1 M NaOH(aq) in the dark without solution agitation. cUplateau is defined as jph reaches 5% lower than the linear jph−U line at the current plateau region. Uonset is defined as the jph reaches 5% of jph (U = Uplateau)

separate the charge carriers for the jph plateau. In contrast, for the low-damaged GaN pillars (13 W), a smaller ΔUonset‑plateau is required or, from another perspective, the j−U slope is larger. Post wet etching using different alkaline solutions is an alternative approach to remove the surface damage of GaN.30−32 In the second experiment of this work, we study the influence of NaOH treatment time on the PEC properties of 1.6 μm-high n-GaN pillars (Figure 3). The pillars were fabricated by dry etching with 125 W RF power. The solid red curve and dashed black curve show the jph−U plot for the as-

Figure 4. SEM images of pillars at 45° before and after the NaOH wet etching: (a), (aa) as-fabricated pillars; pillars after wet etching of 1 M NaOH(aq) in the dark for (b), (bb) 1 min; and (c), (cc) 15 min. Scale bar: 300 nm.

magnification. Figure 4, parts b and c, shows the GaN pillars in NaOH solution for 1 and 15 min, with parts bb and cc as magnifications, respectively. The Ga-polar (0001) GaN surface is not etched due to its high chemical stability.33,34 In contrast, the defective sidewalls of the pillars are etched by the alkaline solution.35 The {101̅0} planes are gradually revealed, leaving a stepped surface at the pillar wall. After 15 min of etching, a layer of ∼60 nm thick is removed at the bottom of the pillar, as indicated in Figure 4cc. Recently, the dissociative mechanisms of interfacial water molecules at the GaN surface have been studied using first principal calculations.36,37 The GaN surfaces of (0001), (0001)̅ , and (1010̅ ) planes show the capacity for water dissociation, but possess different energy barriers for the formation of intermediate species. It has been experimentally demonstrated that the {101̅0} GaN planes of GaN nanowires

Figure 3. jph−U for the 1.6 μm-high GaN pillars after a postetch wet treatment using 1 M NaOH(aq) for different periods of time. All pillar samples are fabricated by dry etching using an RF power of 125 W. Solution: 0.1 M HBr + 0.2 M Na2SO4; v: 100 mV/s. 11263



Article

are highly photocatalytic active and can perform water splitting spontaneously.11 Following this increased photocatalytic activity of the (101̅0) planes, it is not unlikely that Br− oxidation at the {101̅0} planes of the GaN pillar is enhanced with respect to (0001). This may explain the negative Uonset shift as the {101̅0} planes become more dominant at the pillar structure after etching in NaOH solution. The impact of the dry-etching damage on the PEC property of GaN is shown in Figure 5. A GaN epilayer is blanket etched

Figure 6. (a) j−U for different n-GaN pillars and a unetched planar GaN in the dark. (b) The zoom-in j−U plot of (a). Solution: 0.1 M HBr + 0.2 M Na2SO4; v: 100 mV/s. Figure 5. (a) j−U for 1.6 μm-blanket-etched n-GaN electrode using a RF power of 125 W in the dark and under illumination. Solution: 0.1 M HBr + 0.2 M Na2SO4; v: 1 mV/s.

for better visualization. In the case of the pillar samples, a pair of cathodic and anodic peaks is observed at 0.16 and 0.45 V, respectively. In contrast, this is not the case for the planar GaN. The cathodic peak has a higher peak current and a more symmetric and narrow profile as compared to the anodic peak. The 125 W pillar sample shows the highest peak current and the 13 W pillar sample shows the lowest peak current among three pillars. Figure 7a shows the cyclic voltammogram of the

for 1.6 μm by the dry etching using 125 W RF power without Ni or SiO2 mask. The j−U plot of this blanket-etched GaN in the dark and under illumination are recorded at v = 1 mV/s to minimize the transient current. The dark current remains lower than 10 nA/cm2 at U < 1 V, indicating a low reverse leakage current. However, the PEC property of the blanket-etched GaN is seriously deteriorated with respect to the unetched planar GaN in Figure 1. The jph only increases significantly at U > 0.25 V, and the jph only reaches 0.4 μA/cm2 at U = 1 V. The introduction of surface damage has a clear and distinct impact on the PEC properties. The recombination of charge carriers at the introduced surface states become a predominant process;38 nearly no photocarriers are available for the electrochemical oxidation of Br−. On the basis of the experiments of Figures 3−5, we conclude that the increase of the plateau jph for the pillars etched ≥5 min follows from the removal of damaged GaN at the bottom of the pillar wall. Due to the shape of the self-assembled Ni mask and the different etching rate of the layers, the geometry of the pillars in this work is taper-like. The shape of bottom part of the pillar wall is less vertical and straight than the upper part, as shown in Figure 3a. Consequently, the bottom part of the pillar receives the highest degree of the ion bombardment during the dry-etching process. We therefore propose that most of the photoactivity is lost at the bottom of pillar wall. In Figure 2, all the pillars show a comparable plateau jph. The photoactivity of the GaN at the bottom of the pillar array is believed to be seriously deteriorated even using a weak RF power. Only the less-damaged GaN at the upper part of pillar wall and the top plane can contribute to the jph. Fortunately, the dry-etching damage is normally introduced within a few hundred angstroms from the surface;39,40 the damage decreases significantly with the distance from the surface. After a period of etching in NaOH solution, the damage at the bottom wall is removed. The high quality GaN, which is subsequently exposed (Figure 4cc) can contribute to the photocurrent. Figure 6a shows j−U plots for the GaN pillars and unetched planar GaN in the dark, which is scanned from negative to positive potentials and back. Figure 6b is a zoom-in of Figure 6a

Figure 7. (a) j−U for the sample 125 W at different v (50 to 200 mV/ s) in the dark with the inset showing the cathodic peak current density as a function of v. Solution: 0.1 M HBr + 0.2 M Na2SO4. (b)−(d) Schematic illustrations of the plasma-induced trap levels in the bandgap of n-GaN at the bias 1, 2, and 3 in (a), respectively.

125 W pillars in the dark as a function of v. The cathodic peak potential (Upeak) is located at 0.16 V and is independent of v. The cathodic peak current density increases linearly with increasing v (see the inset of Figure 7a). This indicates that the reaction is under kinetic control. In contrast, the anodic Upeak shifts 0.18 to 0.49 V as v increases. The integrated charge of the cathodic current is the same as of the anodic current for all v. The cyclic voltammogram of the 125 W pillars under illumination is shown in Figure 8. The electrolyte was not 11264



Article

observed at 0.16 V. Therefore, we can conclude that the energy of the trap level (Et) is around 0.6 eV, below the CB edge of GaN. When the GaN photoanode is biased to be more positive than 0.6 V below the CB edge, as schematically illustrated in Figure 7b, the Fermi-level is below Et, leaving the trap levels unoccupied. While the n-GaN electrode becomes negatively polarized, the Fermi level moves upward and the surface electron concentration increases. Once Ef is higher than the energy of trap levels, shown in Figure 7c, the traps are rapidly filled, contributing to the cathodic current. The narrow cathodic profile indicates that the charging of the traps is kinetically favorable. When the bias scans positively from −0.2 V, Ef moves downward. The discharging of the traps starts when Ef < Et. Electrons of the traps can be released to CB by multistep tunneling (Figure 7d). Since the detrapping is a complex and slow process, the anodic j−U profile shows a long tail at U > 0.5 V. This peak current of the trapping and detrapping can be an indication of the surface damage for the pillars. The 125 W pillar sample gives the highest peak current due to the strongest damage. After immersion in NaOH solution, most of the damaged layer is removed for the sample 125W15′. However, due to the selective etching, some damage layer, particularly at the bottom between pillars, is not able to be etched. Consequently, the peak currents are reduced but are still higher than the 13 W pillar sample.

Figure 8. j−U of the sample under 125 W illumination. The cathodic peaks located at −0.50 V, −1.06 V, and V > −1.15 V are for the reduction of Br2 oxidized by photoelectrolysis, dissolved oxygen from ambient and proton, respectively. Solution: 0.1 M HBr + 0.2 M Na2SO4; v: 100 mV/s.

purged prior to the measurement and was not agitated during the measurement. For U > −0.3 V, Br− is oxidized by photogenerated holes to Br2 with the assistance of surface traps.41,42 Subsequently, the Br2 at the GaN surface is reduced to Br− in the backward scan at U = −0.5 V. The reduction of dissolved oxygen from ambient contributes to the peak at −1.06 V. Proton reduction occurs for U < −1.15 V. It is clear that no cathodic current is observed at U = 0.16 V. Consequently, the cathodic peak shown in Figure 7a cannot be attributed to the electron charge transfer from conduction band or surface traps. Hence, we ascribe the cathodic peak is ascribed by the electron charging to the damage-induced surface traps. The unetched planar GaN in Figure 6 shows a low current and no anodic peak in the reverse bias regime. We therefore believe the anodic peak of 125 W pillars in Figure 7a is ascribed by the discharging of the occupied traps. The Br− oxidation is a process of photohole capture from the valence band of GaN. The discharging of the traps through electron injection to electrolyte is not possible. Bertoluzzi et al. constructs a model to interpret the cyclic voltammogram of charging and discharging of surface traps.43 When the detrapping of electrons to CB is slower than the charging but is faster than carrier recombination, a broad anodic peak is found, as shown in Figure 4f in ref 40. In this case, the anodic Upeak shifts more positive than the cathodic Upeak and a long tail is found due to the slow discharging. This model confirms to our experimental observation. However, in our case the integrated cathodic charge is the same as of the anodic charge. The trap depletion through carrier recombination should be ignorable. Hole concentration for carrier recombination is expected to be low considering electron concentration (5 × 1018 cm−3) and bandgap of GaN (3.4 eV). Therefore, only charging of traps and discharging of trapped electrons to CB contribute to the cathodic peak and anodic peak, respectively. The trap-assisted tunneling to CB is one of main mechanisms for reverse leakage current at the metal/n-GaN Schottky interface. However, in our case, at the weak surface bandbending, the trapped electrons are not possible to directly tunnel to CB, due to the thick barrier. Multistep tunneling may be involved in the detrapping process.44 Although the traps found in Figure 7 are monoenergetic, the dry etching not only damages the GaN surface, but also creates damage in the near surface region. It is very possible the trapped electrons can tunnel to CB via these plasma-induced traps in the near surface region. The flat-band potential of the 125 W pillars, deduced from the Mott−Schottky relation, is −0.44 V. The cathodic Upeak is

■

CONCLUSIONS In this work, we fabricated n-GaN pillars by ICP dry etching from a planar GaN epilayer. Although, the jph is enhanced after surface roughening, the GaN surface is damaged by the dryetching process. To reduce the surface damage, the plasma bombardment energy is controlled during the dry etching. We found that when the RF power is minimized, the pillar sample shows a similar ΔUonset‑plateau and j−U slope as the planar GaN. Moreover, we demonstrated that the damaged GaN surface can be removed using NaOH(aq) wet etching, which leads to an Uonset shift of 60 mV and a plateau jph, which is 100% higher with respect to the planar GaN. A pair of anodic and cathodic peaks is found in the dark cyclic voltammogram of the damaged pillars, which indicates the charging and discharging of the deep-level traps existing at 0.6 eV below the CB edge.

■

AUTHOR INFORMATION

Corresponding Author

*Phone: +32 (0)16281083; e-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors would like to acknowledge project support from IMEC and funding support from the WaterstofNet and Interreg.

■

REFERENCES

(1) Gratzel, M. Photoelectrochemical Cells. Nature 2001, 414, 338− 344. (2) Turner, J. A. A Realizable Renewable Energy Future. Science 1999, 285, 687−689. (3) van Dorp, D. H.; Hijnen, N.; Di Vece, M.; Kelly, J. SiC: A Photocathode for Water Splitting and Hydrogen Storage. Angew. Chem.-Int. Ed. 2009, 48, 6085−6088. (4) Fujishima, A.; Honda, K. Electrochemical Photolysis of Water at a Semiconductor Electrode. Nature 1972, 238, 37−38. 11265



Article

Nanoroughened GaN by Dry Etching. ECS Electrochem. Lett. 2013, 2, H51−H53. (25) Tseng, W. J.; Gonzalez, M.; Dillemans, L.; Cheng, K.; Jiang, S. J.; Vereecken, P. M.; Borghs, G.; Lieten, R. R. Strain Relaxation in GaN Nanopillars. Appl. Phys. Lett. 2012, 101, 253102. (26) Cheng, K.; Leys, M.; Degroote, S.; Germain, M.; Borghs, G. High Quality GaN Grown on Silicon (111) Using a SixNy Interlayer by Metal-Organic Vapor Phase Epitaxy. Appl. Phys. Lett. 2008, 92, 192111. (27) Pearton, S. J.; Shul, R. J.; Ren, F. A Review of Dry Etching of GaN and Related Materials. MRS Internet J. Nitride Semicond. Res. 2000, 5, 1−38. (28) Lee, J. M.; Chang, K. M.; Kim, S. W.; Huh, C.; Lee, I. H.; Park, S. J. Dry Etch Damage in n-type GaN and its Recovery by Treatment with an N2 plasma. J. Appl. Phys. 2000, 87, 7667−7670. (29) Aspnes, D. E. Recombination at Semiconductor Surfaces and Interfaces. Surf. Sci. 1983, 132, 406−421. (30) Wang, X. H.; Ning, J. Q.; Xu, S. J.; Choi, H. W. Raman and Photoluminescence Characterization of Focused Ion Beam Patterned InGaN/GaN Multi-quantum-wells Nanopillar Array. J. Appl. Phys. 2011, 110. (31) Itoh, M.; Kinoshita, T.; Kawasaki, K.; Takeuchi, M.; Koike, C.; Aoyagi, Y. Fabrication of GaN-based Striped Structures along the ⟨112̅0⟩ Direction by the Combination of RIE Dry-Etching and KOH Wet-Etching Techniques to Recover Dry-Etching Damage. Phys. Stat. Solidi C 2006, 1624−1628. (32) Jeon, D. W.; Jang, L. W.; Jeon, J. W.; Park, J. W.; Song, Y. H.; Jeon, S. R.; Ju, J. W.; Baek, J. H.; Lee, I. H. Improved Photoluminescence Efficiency in UV Nanopillar Light Emitting Diode Structures by Recovery of Dry Etching Damage. J. Nanosci. Nanotechnol. 2013, 13, 3645−3649. (33) Zhuang, D.; Edgar, J. H. Wet Etching of GaN, AIN, and SiC: A Review. Mater. Sci. Eng., R 2005, 48, 1−46. (34) Li, D. S.; Sumiya, M.; Fuke, S.; Yang, D. R.; Que, D. L.; Suzuki, Y.; Fukuda, Y. Selective Etching of GaN Polar Surface in Potassium Hydroxide Solution Studied by X-ray Photoelectron Spectroscopy. J. Appl. Phys. 2001, 90, 4219−4223. (35) Stocker, D. A.; Schubert, E. F.; Redwing, J. M. Crystallographic Wet Chemical Etching of GaN. Appl. Phys. Lett. 1998, 73, 2654−2656. (36) Shen, X.; Small, Y. A.; Wang, J.; Allen, P. B.; Fernandez-Serra, M. V.; Hybertsen, M. S.; Muckerman, J. T. Photocatalytic Water Oxidation at the GaN (101̅0)-Water Interface. J. Phys. Chem. C 2010, 114, 13695−13704. (37) Chen, P.-T.; Sun, C.-L.; Hayashi, M. First-Principles Calculations of Hydrogen Generation Due to Water Splitting on Polar GaN Surfaces. J. Phys. Chem. C 2010, 114, 18228−18232. (38) Li, J.; Peter, L. M. Surface Recombination at Semiconductor Electrodes: Part III. Steady-State and Intensity Modulated Photocurrent Response. J. Electroanal. Chem. 1985, 193, 27−47. (39) Cao, X. A.; Pearton, S. J.; Dang, G. T.; Zhang, A. P.; Ren, F.; Van Hove, J. M. GaN n- and p-Type Schottky Diodes: Effect of Dry Etch Damage. IEEE Trans. Electron Dev. 2000, 47, 1320−1324. (40) Choi, H. W.; Chua, S. J.; Tripathy, S. Morphological and Structural Analyses of Plasma-Induced Damage to n-Type GaN. J. Appl. Phys. 2002, 92, 4381−4385. (41) Schaefer, S.; Koch, A. H. R.; Cavallini, A.; Stutzmann, M.; Sharp, I. D. Charge Transfer across the n-Type GaN−Electrolyte Interface. J. Phys. Chem. C 2012, 116, 22281−22286. (42) Huygens, I. M.; Theuwis, A.; Gomes, W. P.; Strubbe, K. Photoelectrochemical Reactions at the n-GaN Electrode in 1 M H2SO4 and in Acidic Solutions Containing Cl− Ions. Phys. Chem. Chem. Phys. 2002, 4, 2301−2306. (43) Bertoluzzi, L.; Badia-Bou, L.; Fabregat-Santiago, F.; Gimenez, S.; Bisquert, J. Interpretation of Cyclic Voltammetry Measurements of Thin Semiconductor Films for Solar Fuel Applications. J. Phys. Chem. Lett. 2013, 4, 1334−1339. (44) Karmalkar, S.; Sathaiya, D. M.; Shur, M. S. Mechanism of the Reverse Gate Leakage in AlGaN/GaN High Electron Mobility Transistors. Appl. Phys. Lett. 2003, 82, 3976−3978.

(5) Heller, A. Conversion of Sunlight into Electrical Power and Photoassisted Electrolysis of Water in Photoelectrochemical Cells. Acc. Chem. Res. 1981, 14, 154−162. (6) Walter, M. G.; Warren, E. L.; McKone, J. R.; Boettcher, S. W.; Mi, Q.; Santori, E. A.; Lewis, N. S. Solar Water Splitting Cells. Chem. Rev. 2010, 110, 6446−6473. (7) Chen, X. B.; Shen, S. H.; Guo, L. J.; Mao, S. S. Semiconductorbased Photocatalytic Hydrogen Generation. Chem. Rev. 2010, 110, 6503−6570. (8) Beach, J. D.; Collins, R. T.; Turner, J. A. Band-edge Potentials of n-type and p-type GaN. J. Electrochem. Soc. 2003, 150, A899−A904. (9) Chen, P.-T.; Sun, C.-L.; Hayashi, M. First-Principles Calculations of Hydrogen Generation Due to Water Splitting on Polar GaN Surfaces. J. Phys. Chem. C 2010, 114, 18228−18232. (10) Fujii, K.; Iwaki, Y.; Masui, H.; Baker, T. J.; Iza, M.; Sato, H.; Kaeding, J.; Yao, T.; Speck, J. S.; Denbaars, S. P.; et al. Photoelectrochemical Properties of Nonpolar and Semipolar GaN. Jpn. J. Appl. Phys. 2007, 46, 6573−6578. (11) Wang, D. F.; Pierre, A.; Kibria, M. G.; Cui, K.; Han, X. G.; Bevan, K. H.; Guo, H.; Paradis, S.; Hakima, A. R.; Mi, Z. T. WaferLevel Photocatalytic Water Splitting on GaN Nanowire Arrays Grown by Molecular Beam Epitaxy. Nano Lett. 2011, 11, 2353−2357. (12) Hwang, Y. J.; Wu, C. H.; Hahn, C.; Jeong, H. E.; Yang, P. D. Si/ InGaN Core/Shell Hierarchical Nanowire Arrays and Their Photoelectrochemical Properties. Nano Lett. 2012, 12, 1678−1682. (13) Basilio, A. M.; Hsu, Y.-K.; Tu, W.-H.; Yen, C.-H.; Hsu, G.-M.; Chyan, O.; Chyan, Y.; Hwang, J.-S.; Chen, Y.-T.; Chen, L.-C.; et al. Enhancement of the Energy Photoconversion Efficiency through Crystallographic Etching of a c-plane GaN Thin Film. J. Mater. Chem. 2010, 20, 8118−8125. (14) Benton, J.; Bai, J.; Wang, T. Enhancement in Solar Hydrogen Generation Efficiency using a GaN-based Nanorod Structure. Appl. Phys. Lett. 2013, 102, 173905−173905. (15) Richter, T.; Lueth, H.; Meijers, R.; Calarco, R.; Marso, M. Doping Concentration of GaN Nanowires Determined by OptoElectrical Measurements. Nano Lett. 2008, 8, 3056−3059. (16) Wallys, J.; Teubert, J.; Furtmayr, F.; Hofmann, D. M.; Eickhoff, M. Bias-Enhanced Optical pH Response of Group III-Nitride Nanowires. Nano Lett. 2012, 12, 6180−6186. (17) van Dorp, D. H.; Weyher, J. L.; Kooijman, M. R.; Kelly, J. J. Photoetching Mechanisms of GaN in Alkaline S2O82− Solution. J. Electrochem. Soc. 2009, 156, D371−D376. (18) Hahn, C.; Zhang, Z.; Fu, A.; Wu, C. H.; Hwang, Y. J.; Gargas, D. J.; Yang, P. Epitaxial Growth of InGaN Nanowire Arrays for Light Emitting Diodes. ACS Nano 2011, 5, 3970−3976. (19) Yam, F. K.; Hassan, Z. InGaN: An Overview of the Growth Kinetics, Physical Properties and Emission Mechanisms. Superlattices Microstruct. 2008, 43, 1−23. (20) Spurgeon, J. M.; Atwater, H. A.; Lewis, N. S. A Comparison Between the Behavior of Nanorod Array and Planar Cd(Se, Te) Photoelectrodes. J. Phys. Chem. C 2008, 112, 6186−6193. (21) Kato, H.; Asakura, K.; Kudo, A. Highly Efficient Water Splitting into H2 and O2 over Lanthanum-Doped NaTaO3 Photocatalysts with High Crystallinity and Surface Nanostructure. J. Am. Chem. Soc. 2003, 125, 3082−3089. (22) de Boer, M. J.; Gardeniers, J. G. E.; Jansen, H. V.; Smulders, E.; Gilde, M. J.; Roelofs, G.; Sasserath, J. N.; Elwenspoek, M. Guidelines for Etching Silicon MEMS Structures using Fluorine High-Density Plasmas at Cryogenic Temperatures. J. Microelectromech. Syst. 2002, 11, 385−401. (23) Jansen, H. V.; de Boer, M. J.; Unnikrishnan, S.; Louwerse, M. C.; Elwenspoek, M. C. Black Silicon Method X: A Review on High Speed and Selective Plasma Etching of Silicon with Profile Control: An In-Depth Comparison between Bosch and Cryostat DRIE Processes as a Roadmap to Next Generation Equipment. J. Micromech. Microeng. 2009, 19, 033001−033042. (24) Tseng, W. J.; van Dorp, D. H.; Lieten, R. R.; Mehta, B.; Vereecken, P. M.; Borghs, G. Enhanced Photocatalytic Activity of 11266


Impact of Plasma-Induced Surface Damage on the ... - ACS Publications

Recommend Documents