IEEE ELECTRON DEVICE LETTERS, VOL. 31, NO. 6, JUNE 2010 585
Improved Conversion Efficiency of Textured InGaN
Solar Cells With Interdigitated Imbedded Electrodes
Ray-Hua Horng, Member, IEEE, Mu-Tao Chu, Hung-Ruei Chen, Wen-Yih Liao, Ming-Hsien Wu,
Kuo-Feng Chen, and Dong-Sing Wuu, Member, IEEE
Abstract—Textured n-GaN/i-InGaN/p-GaN solar cells with interdigitated imbedded electrodes (IIEs) eliminating the electrode-shading loss have been investigated. In the absence of the electrode-shading effect, the optimized textured solar cell exhibits a conversion efficiency of 1.03%, which is 78% and 47% higher than those of the conventional structure and the structure with mirror coated on silicon substrate with electrode shading, respectively. The short-circuit current density of this textured IIE device is about 0.65 mA/cm2, which is 71% and 44% higher than those of the two compared structures, respectively.
Index Terms—Electrode shading, interdigitated imbedded
elec-trodes (IIEs), n-GaN/i-InGaN/p-GaN solar cell, textured. I. INTRODUCTION
T
UNABLE InxGa1−xN energy bandgap ranging from 0.7 eV(x = 1) to 3.4 eV (x = 0) and nearly covering the full solar radiation spectrum makes ternary InGaN alloy one of the potential candidates for the application of achieving high-efficiency (50%) multijunction solar cells [1], [2]. However, recent research on InGaN solar cells shows a quite-low con-version efficiency of less than 2% [3], [4], which mainly results from the difficulty in obtaining indium-rich InGaN alloy and the insufficiency of photon absorption in the InGaN layer. A vari-ety of approaches for improving the conversion efficiency via epilayer growth and optical management have been proposed. The techniques of facet-initiated epitaxial lateral overgrowth, pendeo epitaxy, and patterned sapphire substrates [5]–[8] are all milestones making a significant contribution for improving theManuscript received February 23, 2010; revised March 6, 2010. Date of publication May 3, 2010; date of current version May 26, 2010. This work was supported in part by the Industrial Technology Research Institute under Contract 03-981-014 and in part by the Ministry of Education, Taipei, Taiwan, under an ATU program. The review of this letter was arranged by Editor P. K.-L. Yu.
R.-H. Horng is with the Department of Electro-Optical Engineering, National Cheng Kung University, Tainan 701, Taiwan (e-mail: rhhorng@ mail.ncku.edu.tw).
M.-T. Chu and H.-R. Chen are with the Institute of Precision Engineering, National Chung Hsing University, Taichung 402, Taiwan (e-mail: [email protected]; [email protected]).
W.-Y. Liao and M.-H. Wu are with the Electronics and Opto-Electronics Re-search Laboratories, Industrial Technology ReRe-search Institute, Hsinchu 31040, Taiwan (e-mail: [email protected]; [email protected]).
K.-F. Chen is with the Institute of Photonics and Communications, National Kaohsiung University of Applied Sciences, Kaohsiung 807, Taiwan (e-mail: [email protected]).
D.-S. Wuu is with the Department of Materials Science and Engineer-ing, National Chung Hsing University, Taichung 40227, Taiwan (e-mail: [email protected]).
Color versions of one or more of the figures in this letter are available online at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/LED.2010.2046615
epilayer quality. We also optimized the InGaN thickness and the indium proportion in InGaN showing no phase separation and relaxation [9]. However, the nearly optimized InGaN layers, which have no observable defect such as V-shaped pits and dislocations, are typically too thin to absorb sufficient photons in the solar spectrum to maximize the device performance. The back-mirror reflector on silicon substrate [10] is one of the techniques for optical confinement, which is achieved with an increment of 57.6% in short-circuit current density (Jsc). However, there still exists the electrode-shading problem. In this letter, n-side-up devices with interdigitated imbedded elec-trodes (IIEs) for overcoming the electrode-shading problem are provided and fabricated. The effect of the thickness and roughness of the top n-GaN layer on solar cell performance is also studied.
II. EXPERIMENT
The p-GaN: Mg/i-In0.1Ga0.9N/n-GaN : Si epitaxial layers of the conventional structure of high quality [9] are grown on undoped-GaN (u-GaN)/c-plane (0001) sapphire substrates by metalorganic chemical vapor deposition. The sandwich p-i-n structure consists of a 150-nm Mg-doped top p-type GaN layer, a 150-nm intrinsic absorption InGaN layer, and a 3-μm Si-doped bottom n-type GaN layer. The indium composition of InGaN was confirmed by X-ray rocking curve and PL measurements. The emission wavelength (measured by PL) of In0.1Ga0.9N is about 393 nm. Each chip with an area of 1× 1 mm2 is fabricated. The electrode contact pads are patterned with Cr/Au (25 nm/200 nm) by thermal evaporation. The schematic cross section of the conventional p-i-n structure is shown in Fig. 1(a). Undoped GaN and indium tin oxide act as an epitaxial buffer layer and a transparent conducting layer for current spreading, respectively. The process flow of the GaN/InGaN solar cells with the IIE structure is schematically shown in Fig. 1(b)–(e). First, the top side (p side) of the conventional device is upside down bonded onto a thin-film-coated silicon substrate, which consists of SiO2 (900 nm) and Ti/Al (15 nm/250 nm), using epoxy glue bonding, as shown in Fig. 1(b). The laser lift-off process shown in Fig. 1(c) is then employed for sapphire removal. Subsequently, the pattern of the electrode contact pads is defined by photolithography, which is followed by a highly anisotropic etching process, as shown in Fig. 1(d), using an inductively coupled plasma reactive-ion etching to expose the p- and n-contact pads. The IIE device is consequently finished with the wet processing of H3PO4 and NaOH solution for reducing the thickness and roughening the
HORNG et al.: IMPROVED CONVERSION EFFICIENCY OF TEXTURED InGaN SOLAR CELLS 587
Fig. 4. Cross-sectional SEM images of D3 with n-GaN etching times of (a) 0, (b) 3, and (c) 6 min.
(0–6 min) of the window layer, are shown in Fig. 3. The curve ofVoc keeps nearly constant, which is comprehensible since the optical management is supposed to be independent of the electrical potential. The whole curve ofJscfor D3 with various n-GaN etching times is greater than 0.59 mA/cm2, and the optimal value occurs at an etching time of 3 min, which is denoted as “D3-3” in the x-axis. The FFs of D3, which are merely 3%–4% higher than those of D1 and D2, seem to be dramatically unaffected by the etching time. Theη in the figure has a quite similar tendency to Jsc. The optimal η of 1.03% for D3 also occurring at an etching time of 3 min is 78% and 47% higher than those for D1 and D2. Thus, the improvement of the efficiency for D3 is apparently attributed to the increment of the photocurrent from employing effective light-absorbing treatments, including the back-mirror reflector, the surface texture, and the IIE structure.
It is well known that the doped window layer is typically highly defective and light absorbent [11]. Thus, the too-thick window layer will prohibit the incoming photons from ef-fectively entering the absorbing layer, leading to insufficient absorption for the photovoltaic effect. The reduction and opti-mization of the thickness of the window layer have also been at-tempted for the IIE structure. The cross-sectional SEM images of D3 with n-GaN etching times of 0, 3, and 6 min are shown in Fig. 4(a)–(c), respectively. The structure of D3 in Fig. 4(a), corresponding to Fig. 1(e), can be clearly observed and con-trasted for each layer except the undistinguishable boundary of the p-i-n layers. The intact surface of the window layer and the thickness of the epitaxial layers (∼7.5 μm) can be seen in Fig. 4(a) right after the up-side-down pattern transfer and sapphire removal. After the H3PO4etching time of 3 min and surface roughening, a 5-μm epitaxial layer with a pyramidal textured surface is shown in the inset of Fig. 4(b). However, a 6-min etching time instead leads to the partial damage of the pyramidal texture, as shown in the inset of Fig. 4(c). As a result, some local planar regions are supposed to cause the undesired reflection and deteriorate the optical confinement. Therefore, the evidence from Fig. 4 shows that the mechanism of attaining the optimized efficiency may not be the optimized thickness of the window layer but most likely the integrity of the surface texture.
IV. CONCLUSION
In summary, GaN/InGaN double-heterojunction solar cells, in the absence of electrode-shading loss, have been demon-strated with an improved conversion efficiency of 1.03%, which is 78% and 47% higher than those of the conventional struc-ture and the strucstruc-ture with mirror coated on silicon substrate, respectively. The tendency of the conversion efficiency of the proposed structure is in well agreement with the short-circuit current density. It reveals that the improvement of conver-sion efficiency is attributed to the increased photocurrent from enhancing light absorption, including the surface texture, the back-mirror reflector, and the IIE structure. To verify the opti-mized thickness of the window layer, the consistent textured surfaces for different etching conditions need to be used for future work.
REFERENCES
[1] A. De Vos, Endoreversible Thermodynamics of Solar Energy Conversion. Oxford, U.K.: Oxford Univ. Press, 1992.
[2] A. Barnett, D. Kirkpatrick, C. Honsberg, D. Moore, M. Wanlass, K. Emery, R. Schwartz, D. Carlson, S. Bowden, D. Aiken, A. Gray, S. Kurtz, L. Kazmerski, T. Moriarty, M. Steiner, J. Gray, T. Davenport, R. Buelow, L. Takacs, N. Shatz, J. Bortz, O. Jani, K. Goossen, F. Kiamilev, A. Doolittle, I. Ferguson, B. Unger, G. Schmidt, E. Christensen, and D. Salzman, “Milestones toward 50% efficient solar cell modules,” in
Proc. 22nd Eur. PV Solar Energy Conf., Milan, Italy, Sep. 2007.
[3] J.-K. Sheu, C.-C. Yang, S.-J. Tu, K.-H. Chang, M.-L. Lee, W.-C. Lai, and L.-C. Peng, “Demonstration of GaN-based solar cells with GaN/InGaN superlattice absorption layers,” IEEE Electron Device Lett., vol. 30, no. 3, pp. 225–227, Mar. 2009.
[4] B. R. Jampana, A. G. Melton, M. Jamil, N. N. Faleev, R. L. Opila, I. T. Ferguson, and C. B. Honsberg, “Design and realization of wide-band-gap(∼2.67 eV) InGaN p-n junction solar cell,” IEEE Electron Device
Lett., vol. 31, no. 1, pp. 32–34, Jan. 2010.
[5] B. Beaumont, P. Vennegues, and P. Gibart, “Epitaxial lateral overgrowth of GaN,” Phys. Stat. Sol. (B), vol. 227, no. 1, pp. 1–43, Sep. 2001. [6] A. Usui, H. Sunakawa, K. Kobayashi, H. Watanabe, and M. Mizuta,
“Characteristics of FIELO-GaN grown by hydride vapor phase epitaxy,” in Proc. Mater. Res. Soc. Symp., Nov. 2001, vol. 639, pp. G5.6.1–G5.6.10. [7] T. S. Zheleva, S. A. Smith, D. B. Thomson, K. J. Linthicum, P. Rajagopal, and R. P. Davis, “Pendeo-epitaxy: A new approach for lateral growth of gallium nitride films,” J. Electron. Mater., vol. 28, no. 4, pp. L5–L8, Apr. 1999.
[8] K. Tadatomo, H. Okagawa, Y. Ohuchi, T. Tsunekawa, T. Jyouichi, Y. Imada, M. Kato, H. Kudo, and T. Taguchi, “High output power InGaN ultraviolet light-emitting diodes fabricated on patterned substrates using metalorganic vapor phase epitaxy,” Phys. Stat. Sol. (A), vol. 188, no. 1, pp. 121–125, Nov. 2001.
[9] X. Zheng, R. H. Horng, D. S. Wuu, M. T. Chu, W. Y. Liao, M. H. Wu, R. M. Lin, and Y. C. Lu, “High-quality InGaN/GaN heterojunctions and their photovoltaic effects,” Appl. Phys. Lett., vol. 93, no. 26, pp. 261 108-1–261 108-3, Dec. 2008.
[10] R. H. Horng, S. T. Lin, Y. L. Tsai, M. T. Chu, W. Y. Liao, M. H. Wu, R. M. Lin, and Y. C. Lu, “Improved conversion efficiency of GaN/InGaN thin-film solar cells,” IEEE Electron Device Lett., vol. 30, no. 7, pp. 724– 726, Jul. 2009.
[11] C. H. Qiu, C. Hoggatt, W. Melton, M. W. Leksono, and J. I. Pankove, “Study of defect states in GaN films by photoconductivity measurement,”
