Open Access
Add to BibSonomyAbstract
In this study, we produce InGaN/GaN microcolumn LED (MC-LED) arrays having nonpolar metal sidewall contacts using a top-down method, where the metal contacts only with the sidewall of the columnar LEDs with an open top for transparency. The trapezoidal profile of the as-etched columns was altered to a rectangular profile through KOH treatment, exposing the nonpolar sidewalls. While the MC-LED with no treatment emitted no light because of the etch-damaged region, the MC-LEDs with KOH treatment exhibited much improved the electrical properties with the much higher shunt resistance due to the removal of the etch-damaged region. The optical output power was strongest for the MC-LED with a 5-min treatment indicating an almost complete removal of the damaged region.
© 2013 Optical Society of America
1. Introduction
GaN micro- (or nano-) column structures have been studied in attempt to enhance the performance of light-emitting diodes (LEDs) by improving the light extraction efficiency and reducing the strain due to the large lattice mismatch between GaN and InGaN [1, 2]. One way to produce a microsized column LED is by using the bottom-up method such as selective area growth; another process utilizes the top-down method by etching full-structured conventional LEDs. Though numerous studies have focused on selective area growth, the top-down approach has often been the subject of study since it can realize a columnar structure more easily than when using the bottom-up method [3]. In previous studies, strain relaxation at the surface and sidewall of the GaN columns/pillars has been demonstrated using the top-down approach [4, 5].
The devices from these studies were fabricated using transparent indium-tin-oxide (ITO) for the top electrodes on either filled oxide or merged p-GaN [6, 7]. However, they had problems when using ITO including inferior electrical properties in its resistivity and low work function when compared to metal. Thus, it is preferable to make ohmic contact on m-plane p-GaN (work function of 7 eV) [8] using Ni (5.15 eV)/Au (5.1 eV) layers having ~4 × 10−6 Ω cm2 of resistivity, rather than using ITO (work function of 4.4–4.5 eV), which has ~2 × 10−4 Ω cm2 of resistivity [9, 10]. The thin Ni/Au layer has also been suggested as a top electrode of GaN columnar structures. However, the transmittance when using a metal contact was still lower than when using ITO. As a result, the m-plane p-GaN sidewall with Ni/Au metal contact having an open top was deemed necessary, where the top electrical contact is located such that light can leave the microcolumns with none passing back through the top electrical contact [11]. Therefore, a novel technology using a sidewall metal contact for columnar LEDs is much more beneficial since the metal contact has much closer work function to p-GaN than ITO, better conductivity and higher transparency through an open top.
Another predominant issue in the top-down approach is dry-etch damage resulting from the inductively coupled plasma (ICP) and reactive ion etching (RIE) systems [12, 13]. Several groups have commonly reported the increases in the sheet-resistance of GaN exposed to high density plasmas, along with the decreases in the reverse breakdown voltage and the reductions in the Schottky barrier height in the diodes formed on GaN [14–18]. The roughened sidewall after etching also exhibited the high series resistance in the device and deteriorates the electrical characteristics. Many approaches, including KOH (or NaOH) treatments and photo-enhanced chemical etching (PEC), have been introduced in attempt to remove the plasma-etching damaged region [19–21]. Subsequently, KOH treatment has been found to improve the electrical characteristics through the removal of the damaged region, and by providing the vertical profile in the sidewall with a smooth surface.
The objective of this research is to improve LED performance by utilizing microcolumns having sidewall-contacted arrays fabricated through the top-down approach. Moreover, we also report the control of the microcolumns shape by varying the KOH treatment time and analysis of their electrical and optical characteristics via removal of the etch-damaged region.
2. Experiment
The LEDs discussed in this research have a conventional structure of multiple quantum wells (MQWs) with In0.17Ga0.83N grown by metal-organic chemical vapor deposition (MOCVD) on a 2-inch c-plane sapphire substrate. The LED structures are comprised of a 2-μm thick Si-doped n-type GaN, five periods of InGaN/GaN MQWs, and a 150-nm thick Mg-doped p-type GaN layer.
Three types of microcolumn LEDs (MC-LEDs) were fabricated using the sidewall contact architecture; the difference was in duration of the KOH treatment, the times being 0-min (no treatment), 3-min and 5-min. Arrays of 2-μm diameter microcolumns were fabricated in a unit device covering an area of 300 μm × 300 μm. Figure 1 displays a schematic illustration of the fabrication process for the MC-LEDs that were subject to the KOH treatment. First, a 300-nm SiO2 layer was deposited on LEDs using the plasma-enhanced chemical vapor deposition (PECVD) process to form an etching mask. Second, an array of 2-μm diameter nickel (Ni) disks was formed using a lift-off process. Third, the SiO2 layer was dry etched using RIE with CF4(g) to form a microsized mesa mask. Fourth, ~850 nm of the InGaN/GaN layer was dry etched by inductively-coupled plasma reactive-ion etching (ICP-RIE) using Cl2(g), CH4(g), H(g), and Ar(g). In Fig. 2(c), this dry etching process resulted in a trapezoidal columnar shape, which was then altered to produce a vertical shape after the wet etching in the KOH solution (3 mol, 70 °C). Next, the space between the columns was filled using a layer of benzo-cyclo-butene (BCB) and etched by RIE in order to passivate the MQWs, exposing only the p-GaN sidewalls. Here, the height of the microcolumns and thickness of the BCB layer were 850 nm and 1.4 μm, respectively. To expose the 150-nm thick p-GaN, we controlled the etching depth of the BCB layer to 640 ± 50 nm. After that, an Ni (5 nm)/Au (5 nm) current spreading layer was deposited on a 60° tilted sample using an e-beam evaporator in order to improve the p-GaN sidewall contacts. After depositing the Ni/Au layer, the LEDs were annealed under an N2 (10 sccm) ambient at 550 °C for 1 min in a rapid thermal annealing (RTA) chamber to make an ohmic contact with the p-GaN sidewall. Finally, the Ni/Au layer, Ni mask, and SiO2 mask on the top surface were removed using an HF solution (HF:DI = 1:6), and the Cr (30 nm)/Au (100 nm) contact pad was deposited by the e-beam evaporator.
Fig. 2 Results of KOH treatment. (a) Schematics during KOH treatment (0, 1, 3, 5, 10 min.). (b) Time-dependent etching rates and fill factor during KOH treatment. (c) SEM image before treatment. (d) SEM image after 5-min treatment.
A scanning electron microscope (SEM) and optical microscope were used to examine the cross-sectional view and surface. Electroluminescence (EL) was also measured in order to investigate the electrical and optical properties.
3. Results and discussion
During this study, the KOH treatment was the key process for changing the microcolumn shape and was tested over a period from 1 min to 10 min. Figure 2(a) illustrates the profile changes over time during the KOH treatment. Figure 2(b) then displays the plots of the bottom diameter of the microcolumn, decreased diameter per minute (etching rate), and fill factor (FF) of the columns as the etching time is increased. The etching rates turned out to be time-dependent during the KOH treatment [Fig. 2(b)], and decreased over time. The etching rate was fast until the nonpolar plane became exposed after ~5 min [dashed line in Fig. 2(b)], and then seemed to be saturated. Figures 2(c) and 2(d) are side views of the columns before and after the 5-min KOH treatment.
The physical mechanism used to remove the etch-damaged region with KOH treatment uses different etching rates depending on their GaN plane. The etching rate in the Ga-face is very low and the N-face is very high. In our case, the top plane of the GaN microcolumn is a Ga-face, and the sidewall plane (a-plane) of the GaN consists of both Ga and N atoms at the edge, in the c-direction. The N-face of GaN has the higher etching rate than the Ga-face of GaN due to the relatively higher surface energy obtained by exposing N atoms. On the Ga- terminated surface the etching rate is very low even though etching can be carried out at the dislocation in the Ga-face since the OH- ions in the KOH solution cannot attack the Ga-terminated surface [14]. Therefore, the etching rate in the sidewall having both N-polar and Ga-polar is higher than the top of Ga-polar plane.
Figure 3(a) and 3(b) shows an SEM image of the completed nonpolar sidewall-contacted device. The inset of Fig. 3(b) illustrates that the Ni (5 nm)/Au (5 nm) sidewall contact was well-formed on the microcolumns. In addition, it can also be seen that Ni/Au acted as a current spreading layer on the BCB passivation layer, with ~60% transmission of the emission. Because the top metal layers on the microcolumns were removed using HF solution after the tilted metal deposition process, the metal was well-deposited only on the sidewall and not on the top of the microcolumns. We believe that the removal of the top metal should significantly improve the light extraction efficiency.
Fig. 3 SEM images of nonpolar contact InGaN/GaN LED. (a) A bird eye’s view of the device. (b) Cross-section view of the device.
Figure 4(a) displays the electrical characteristics of the devices. The black line represents the MC-LED with no KOH treatment. The blue line denotes the MC-LED with 3-min KOH treatment, and the red line is with 5-min KOH treatment. The forward voltages for 4 mA of the MC-LEDs with 0-min, 3-min and 5-min of the KOH treatment were 2.7 V, 4.9 V, and 5.5 V, respectively. The MC-LED with no KOH treatment displayed very poor electrical characteristics having a very low breakdown voltage and a very high leakage current because of the dry-etching damage during the ICP etching. The MC-LED having nonpolar faces after the 5-min KOH treatment, however, displayed improved electrical characteristics due to the removal of the dry-etching-induced damage regions. In the lower forward voltage region of the MC-LED after the treatment, a gradual increase instead of an abrupt turn-on was observed, due to most likely the sub-threshold turn-on. Deep levels or surface states in the bulk of the semiconductor may have caused the sub-threshold turn-on and the overall poor outcome is thought to have resulted from the high contact resistance between the metal and the p-GaN sidewall.
A modified equation for the I-V characteristics of a forward-biased p-n junction diode with a shunt resistance Rp (parallel to the ideal diode) and a series resistance Rs (in series with the ideal diode and the shunt) is represented as follows:
The fitted values of the shunt resistances for the MC-LED with 0-min, 3-min and 5-min KOH treatment were 0.027 kΩ, 1.468 kΩ and 851.064 kΩ, respectively. According to the study of Y. B. Hahn et al. regarding the forward turn-on and the reverse breakdown voltages in the plasma induced damage of InGaN / GaN multiple quantum wells in LEDs, together with the etch rate and the surface morphology [15]. The physical degradation of the sidewall along with the rough surface morphology of GaN caused by increased ion scattering induced the deterioration of the forward and the reverse voltages. It was found that the breakdown voltage is strongly affected by the sidewall contamination and that the turn-on voltage is sensitive to the surface roughness of the etched GaN. Moreover, X. A. Cao et al. found that the high ñuxes or energies of the plasma produced p-to-n surface conversion [16]. Therefore, surface treatment for the removal of the damaged region in the sidewall is essential to improve the electrical characteristics.
In the MC-LED with no treatment, most of the current flowed through the very leaky damaged region, which resulted in the very small shunt resistance. The KOH treatment, however, improved electrical characteristics by removing the leaky dry-etch-damaged region. Therefore, through the KOH treatment, the electrical characteristics in the MC-LED improved. As a result, the shunt resistance became much higher than before the treatment, though is yet to be confirmed by a more comprehensive investigation.
No light emission was observed in the MC-LEDs with no treatment, most probably due to the very leaky characteristics. Therefore, optical characterization was conducted only for the MC-LED with 3-min and 5-min treatment. Figure 4(b) presents the EL spectra at 10 mA and 20 mA, indicating that overall intensity is higher for the MC-LED with 5-min treatment than the MC-LED with 3-min treatment even though it has about a 4% lower fill factor. The optical characteristics can be improved by KOH treatment through not only removing the etch-damaged region but reducing the spatial strain [22]. However, we believe that the primary reason for the improved optical characteristics with KOH treatment in this study is the recovered electrical characteristics with smooth surface through removing the etch-damaged region. Therefore, the MC-LED with 3-min treatment may still have had the etch-damaged region on the sidewall, while 5-min treatment almost completely removed the etch-damaged region. As such, we believe that the results obtained from this device confirm its capability for a high performance LEDs. Moreover, the sidewall contact can be used for many other potential applications including separately controlled multi-junction LEDs and solar cells, and novel 3D opto-electronics on ICs to improve the degree of integrations with 3D structure.
4. Conclusion
In this study, we utilized a novel approach for producing nonpolar sidewall-contacted microcolumn LEDs, which can be fabricated using a top-down method. Using a KOH treatment method, the dry-etch damaged region of the microcolumns was removed, effectively decreasing their diameter. Also, the etched profile was transformed from a trapezoidal to a rectangular shape. Three types of LEDs were fabricated and tested: MC- LED with 0-min (no treatment), 3-min treatment and 5-min treatment. While the MC-LED with no treatment had no light emission with a very low shunt resistance Rp because of the etch-damaged region, the electrical properties significantly improved having a much higher shunt resistance Rp through the KOH treatment 0-min to 5-min.
Acknowledgments
This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (NRF-2013R1A1A2A10006632) and Sundiode, Inc., USA.
References and links
1. O. Ambacher, J. Smart, J. R. Shealy, N. G. Weimann, K. Chu, M. Murphy, W. J. Schaff, L. F. Eastman, R. Dimitrov, L. Wittmer, M. Stutzmann, W. Rieger, and J. Hilsenbeck, “Two-dimensional electron gases induced by spontaneous and piezoelectric polarization charges in N- and Ga-face AlGaN/GaN heterostructures,” J. Appl. Phys. 85(6), 3222–3233 (1999). [CrossRef]
2. F. Bernardini, V. Fiorentini, and D. Vanderbilt, “Spontaneous polarization and piezoelectric constants of III-V nitrides,” Phys. Rev. B 56(16), R10024–R10027 (1997). [CrossRef]
3. W. J. Tseng, M. Gonzalez, L. Dillemans, K. Cheng, S. J. Jiang, M. Vereecken, G. Borghs, and R. R. Lieten, “Strain relaxation in GaN nanopillars,” Appl. Phys. Lett. 101(25), 253102 (2012). [CrossRef]
4. S. Keller, C. Schaake, N. A. Fichtenbaum, C. J. Neufeld, Y. Wu, K. McGroddy, A. David, S. P. DenBaars, C. Weisbuch, J. S. Speck, and U. K. Mishra, “Optical and structural properties of GaN nanopillar and nanostripe arrays with embedded InGaN/GaN multi-quantum wells,” J. Appl. Phys. 100(5), 054314 (2006). [CrossRef]
5. E. Y. Xie, Z. Z. Chen, P. R. Edwards, Z. Gong, N. Y. Liu, Y. B. Tao, Y. F. Zhang, Y. J. Chen, I. M. Watson, E. Gu, R. W. Martin, G. Y. Zhang, and M. D. Dawson, “Strain relaxation in InGaN/GaN micro-pillars evidenced by high resolution cathodoluminescence hyperspectral imaging,” J. Appl. Phys. 112(1), 013107 (2012). [CrossRef]
6. B. J. Kim, H. Hung, H. Y. Kim, J. Bang, and J. Kim, “Fabrication of GaN nanorods by inductively coupled plasma etching via SiO2 nanosphere lithography,” Thin Solid Films 517(14), 3859–3861 (2009). [CrossRef]
7. M.-H. Tsai, O. F. Sankey, K. E. Schmidt, and I. S. T. Tsong, “Electronic structures of polar and nonpolar GaN surfaces,” Mater. Sci. Eng. B 88(1), 40–46 (2002). [CrossRef]
8. M. Y. Hsieh, C. Y. Wang, L. Y. Chen, M. Y. Ke, and J. J. Huang, “InGaN/GaN nanorod light emitting arrays fabricated by silica nanomasks,” IEEE J. Quantum Electron. 44(5), 468–472 (2008). [CrossRef]
9. N. Miura, T. Nanjo, M. Suita, T. Oishi, Y. Abe, T. Ozeki, H. Ishikawa, T. Egawa, and T. Jimbo, “Thermal annealing effects on Ni/Au based Schottky contacts on n-GaN and AlGaN/GaN with insertion of high work function metal,” Solid-State Electron. 48(5), 689–695 (2004). [CrossRef]
10. L. Chen, J. Ho, C. Jong, C. C. Chiu, K. Shih, F. Chen, J. Kai, and L. Chang, “Oxidized Ni/Pt and Ni/Au ohmic contacts to p-type GaN,” Appl. Phys. Lett. 76(25), 3703–3705 (2000). [CrossRef]
11. J. C. Kim, S. Ti, and D. E. Mars, “Nanostructure optoelectronic device having sidewall electrical contact,” U.S. Patent 20110297913A1, 8 Dec. 2011.
12. X. A. Cao, S. J. Pearton, G. T. Dang, A. P. Zhang, F. Ren, and J. M. Van Hove, “GaN n- and p-type Schottky diodes: Effect of dry etch damage,” IEEE J. Trans. Electron Devices. 47(7), 1320–1324 (2000). [CrossRef]
13. D. G. Kent, K. P. Lee, A. P. Zhang, B. Luo, M. E. Overberg, C. R. Abernathy, F. Ren, K. D. MacKenzie, S. J. Pearton, and Y. Nakagawa, “Electrical effects of N2 plasma exposure on dry-etch damage in p- and n-GaN Schottky diodes,” Solid-State Electron. 45(10), 1837–1842 (2001). [CrossRef]
14. D. Li, M. Sumiya, S. Fuke, D. Yang, D. Que, Y. Suzuki, and Y. Fukuda, “Selective etching of GaN polar surface in potassium hydroxide solution studied by x-ray photoelectron spectroscopy,” J. Appl. Phys. 90(8), 4219–4223 (2001). [CrossRef]
15. Y. B. Hahn, R. J. Choi, J. H. Hong, H. J. Park, C. S. Choi, and H. J. Lee, “High-density plasma-induced etch damage of InGaN/GaN multiple quantum well light-emitting diodes,” J. Appl. Phys. 92(3), 1189–1194 (2002). [CrossRef]
16. X. A. Cao, S. J. Pearton, A. P. Zhang, G. T. Dang, F. Ren, R. H. Shul, L. Zhang, R. Hickman, and J. M. Van Hove, “Electrical effects of plasma damage in p-GaN,” Appl. Phys. Lett. 75(17), 2569–2571 (1999). [CrossRef]
17. X. A. Cao, H. Cho, S. J. Pearton, G. T. Dang, A. P. Zhang, F. Ren, R. J. Shul, L. Zhang, R. Hickman, and J. M. Van Hove, “Depth and thermal stability of dry etch damage in GaN Schottky diodes,” Appl. Phys. Lett. 75(2), 232–234 (1999). [CrossRef]
18. J. S. Jang, S. J. Park, and T. Y. Seong, “Effects of surface treatment on the electrical properties of ohmic contacts to (In)GaN for high performance optical device,” Phys. Status Solidi A 194(2), 576–582 (2002). [CrossRef]
19. D. A. Stocker, E. F. Schubert, and J. M. Redwing, “Crystallographic wet chemical etching of GaN,” Appl. Phys. Lett. 73(18), 2654–2656 (1998). [CrossRef]
20. J. A. Bardwell, I. G. Foulds, J. B. Webb, H. Tang, J. Fraser, S. Moisa, and S. J. Rolfe, “A simple wet etch for GaN,” J. Electron. Mater. 28(10), L24–L26 (1999). [CrossRef]
21. A. C. Tamboli, M. C. Schmidt, S. Rajan, J. S. Speck, U. K. Mishra, S. P. DenBaars, and E. L. Hu, “Smooth top-down photoelectrochemical etching of m-plane GaN,” J. Electrochem. Soc. 156(1), H47–H51 (2009). [CrossRef]
22. S. Y. Bae, D. J. Kong, J. Y. Lee, D. J. Seo, and D. S. Lee, “Size-controlled InGaN/GaN nanorod array fabrication and optical characterization,” Opt. Express 21(14), 16854–16862 (2013). [CrossRef] [PubMed]
