Open Access
Add to BibSonomyAbstract
In order to reduce surface reflection, anti-reflective (AR) coatings are widely used on the surfaces of solar cells to improve the efficiency of photoelectric conversion. This study employed colloidal lithography with a dry etching process to fabricate sub-micron anti-reflection structures on a GaAs substrate. Etching parameters, such as RF power and etching gas were investigated to determine their influence on surface morphology. We fabricated an array of conical structures 550 nm in diameter and 450 nm in height. The average reflectance of a bare GaAs wafer was reduced from 35.0% to 2.3% across a spectral range of 300 nm – 1200 nm. The anti-reflective performance of SWSs was also calculated using Rigorous Coupled Wave Analysis (RCWA) method. Both simulation and experiment results demonstrate a high degree of similarity.
©2012 Optical Society of America
1. Introduction
Gallium arsenide (GaAs) based solar cells (SCs)have been attracting considerable attention due to their high photoelectric conversion efficiency (PCE) and uniform efficiency across a range of temperatures. However, the large difference in the refractive index between air (n = 1) and GaAs (n = 3.8) can lead to considerable Fresnel loss, with a subsequent decrease in PCE. To reduce loss due to reflection, biomimetic sub-wavelength structures (SWSs) have gradually replaced conventional anti-reflective (AR) thin films [1–3] due to their broadband AR properties [4,5].
SWSs arrays can be considered a stack of multi-layer thin films, in which the refractive index of each layer gradually changes due to an alteration in the volume fraction of the structures, which in turn reduces Fresnel reflection [6,7]. Various methods have been proposed to fabricate SWSs, including electron beam (EB) lithography, laser interference lithography, nano-imprint lithography, and colloidal lithography [8–11]. Among these, colloidal lithography is fastest and most cost efficient.
In our previous study of Si materials, we showed how the anti-reflective capability of sub-micron arrays increases with a decrease in the spacing of structures. Among structures of the same height and width, pyramidal structures exhibited lower average reflectance than paraboloidal structures across the spectral range of 300-1200 nm [12]. This study employed colloidal lithography with polystyrene (PS) spheres in conjunction with inductive coupling plasma (ICP) etching to fabricate hexagonal, close-packed, anti-reflective structures on a sub-micron scale. Etching parameters, such as RF power, Argon (Ar), and Chlorine (Cl2) gasses were investigated to determine their influence on surface morphology. Finally, the AR properties of SWSs were measured using a UV-Vis-NIR spectrophotometer and calculated using Rigorous Coupled Wave Analysis (RCWA) method to validate the experiment results.
2. Experiments
2.1 Fabrication of GaAs SWSs
A self-assembled monolayer of PS spheres (600 nm in diameter) was uniformly organized on a GaAs substrate, as shown in Fig. 1(a) . The hexagonal PS spheres were closely-packed with long-range order, as shown in Fig. 1(b).
Fig. 1 (a) Schematic illustration of PS spheres organized on a GaAs substrate; (b) SEM image of close-packed monolayer comprising hexagonal PS spheres 600 nm in diameter
Figure 2 illustrates the fabrication process. A self-assembled monolayer of PS spheres was uniformly sprayed onto a GaAs substrate to act as an etching mask during the dry etching process. The GaAs substrate and PS sphere array were subjected to ICP etching (Sentech SI 500). A mixture of Ar, Cl2, and O2 gasses was introduced at a pressure of 7.5 mTorr. Following the etching process, the remaining PS spheres were removed using O2 plasma etching.
The surface morphology of the sub-micron structures was characterized using a scanning electron microscope (SEM). The reflectance spectra were measured from 300 nm to 1200 nm using a UV-Vis-NIR spectrophotometer at a near-normal incidence of 5°.
2.2 Analysis of AR properties using the RCWA method
To calculate anti-reflection performance, we employed RCWA method to analyze the propagation of electromagnetic plane waves in a GaAs substrate with SWSs [13]. Figure 3 presents a cross-sectional schematic of an SWSs array and the parameters used in the simulation. Parameters P, H, D, and d represent the period, height, bottom diameter, and top diameter of structures, respectively. θ is the incident angle of light, and ns and nm are the refractive indices of the structure and substrate, respectively. One unit cell in the simulated geometrical structure comprised 4*3 SWSs in a hexagonal close-packed array. To facilitate comparison with measured results, all experimental parameters were duplicated in the simulation.
3. Results and discussions
To investigate the influence of RF power on the formation of AR structures, an RF power of (a) 10 W, (b) 20 W, and (c) 50 W was respectively applied in etching for periods of 5 min. ICP power was set at 100 W, and the flow rates of Ar, Cl2, and O2 were 30 sccm, 10 sccm, and 4 sccm, respectively. Figure 4 shows that the height of the structures produced with RF power of 20 W exceeded that of those produced with 10 W. However, the strong ion bombardment resulting from an RF power of 50 W damaged both the PS spheres and resulting structures. Thus, the RF power was fixed at 20 W for all subsequent experiments.
During the etching process, Cl free radicals were formed through electronic dissociation. These free radicals were adsorbed onto the material and reacted to produce volatile substances, which were removed by cation impact. To evaluate the influence of Cl2 on the etching rate and etch profile, we produced SWSs at various flow rates. Figure 5 presents SEM images of GaAs SWSs etched with Cl2 at flow rates of (a) 10 sccm, (b) 20 sccm, and (c) 30 sccm, for 5 min respectively. ICP power was set at 100 W and RF power was adjusted to 20 W. The flow rates of Ar and O2 were 10 sccm and 4 sccm, respectively.
Under the same etching conditions, the etching rate gradually increased with an increase in the flow rate of Cl2. At a flow rate of 10 sccm, the sidewall of the structures was irregularly shaped. In contrast, when the flow rate was 30 sccm, a tapered profile was generated by the isotropic chemical etching in which Cl2 atoms and Ar ions were reflected from the bottom to the sidewall.
During the dry etching process, Ar ions bombarded the material resulting in defects that caused the reaction gas adsorbed on the material to increase the etching rate. Figure 6 presents SEM images of GaAs SWSs produced at Ar flow rates of (a) 10 sccm, (b) 20 sccm, and (c) 30 sccm for 5 min. ICP power was set to 100 W and RF power was adjusted to 20 W. The flow rates of Cl2 and O2 were 30 sccm and 4 sccm, respectively. The Ar flow rate influenced the surface profile of the SWSs. The etching mechanisms associated with various Ar flow rates are shown in Fig. 7 and Fig. 8 . At a flow rate of 10sccm, chemical etching is the primary mechanism involved in producing a tapered profile. When the Ar flow rate was increased to 20 sccm, stronger ion bombardment resulted in vertical sidewalls near the bottom. At a flow rate of 30 sccm, the impact from an increased number of Ar ions damaged the PS spheres such that PS fragments were sprayed across the structures (Fig. 8(b)), which acts as an etching mask layer to relieve the influence of the ion bombardment on the struture. Continued chemical etching resulted in the elimination of the vertical sidewalls to produce a tapered profile.
Figure 6(c) shows a tapered profile; however, the top of the structures presents a platform shape rather than a continual change across the entire surface. The bottom diameter and height of the structures were 550 nm and 400 nm, respectively. The top diameter was approximately 100 nm. To determine the influence of the flattened cone on reflectivity, we compared the reflectance spectra of structures with various top diameters (0 nm, 100 nm, and 150 nm) in the range of 300 nm-1200 nm using RCWA method, as shown in Fig. 9 . The calculated average reflectance was below 4% for perfectly conical structures (d = 0 nm), which demonstrated the outstanding anti-reflection potential of SWSs. To further decrease reflectance, it was necessary to eliminate the platform shape, as shown in Fig. 6.
Fig. 9 Reflectance spectra of SWSs with a same bottom diameter of 550 nm and height of 400 nm with various top diameters
To avoid producing flat-top structures, we reduced the diameter of the PS spheres to 550 nm through the application of O2 plasma etching for 15s, as shown in Fig. 10 . Following this etching process, the increased spacing between PS spheres allowed the etching gas to react with the GaAs substrate more effectively, which helped to improve lateral etching and reduce etching time.
After reducing the diameter of the PS spheres, ICP etching of the GaAs substrate resulted in conical structures without a flat-top. In the ICP etching process, a low flow rate of O2 can promote the dissociation of Cl ions and increase the free radicals; however, too much O2 dilutes the concentration of Cl2 which reduces the etching rate. In this study, the O2 flow rate was adjusted to 2 sccm. A mixture of Cl2 and Ar gasses was introduced for ICP at flow rates of 30 sccm and 30 sccm, respectively. ICP power and RF power were set at 80W and 20 W, respectively. SEM images of fabricated SWSs are shown in Fig. 11 . The diameter and height of the structures are 550 nm and 450 nm, respectively.
Figure 12 presents the reflectance spectra of a bare GaAs wafer and surface SWSs in the range of 300 nm-1200 nm. To verify the performance of SWSs, the reflectance of the simulation result is also presented for comparison. The SWSs on the surface of the substrate resulted in a significantly reduction in reflectance across the entire spectral range. The average reflectance of bare wafer is 35.0%; the average reflectance of wafers with a conical structure array is 2.3%. Simulation results are very close to those observed in the experiment. Thus, this simulation is reliable as an analytical instrument prior to experimentation.
4. Conclusions
This study fabricated sub-micron conical structures on a GaAs substrate by combining PS sphere lithography with ICP etching. In order to fabricate SWSs with a tapered profile, we systematically investigated how etching parameters (RF power, Ar, and Cl2) influence the surface morphology. Conical structures without a flat-top were fabricated by employing O2 plasma etching to narrow the diameter of PS spheres followed by slight adjustments to the parameters in the ICP etching process. Experimental results demonstrate that SWSs suppress the average reflectance to below 2.3% across the spectral range of 300 nm to 1200 nm. The simulation model of SWSs was developed. Both simulation and experiment results for anti-reflective performance demonstrate a high degree of similarity.
References and links
1. S. Y. Lien, D. S. Wuu, W. C. Yeh, and J. C. Liu, “Tri-layer antireflection coatings (SiO2/SiO2–TiO2/TiO2) for silicon solar cells using a sol–gel technique,” Sol. Energy Mater. Sol. Cells 90(16), 2710–2719 (2006). [CrossRef]
2. B. Y. Su, Y. K. Su, Z. L. Tseng, M. F. Shih, C. Y. Cheng, Z. H. Wu, C. S. Wu, J. J. Yeh, P. Y. Ho, Y. D. Juang, and S. Y. Chu, “Antireflective and radiation resistant ZnO thin films for the efficiency enhancement of GaAs photovoltaics,” J. Electrochem. Soc. 158(3), H267–H270 (2011). [CrossRef]
3. S. M. Jung, Y. H. Kim, S. I. Kim, and S. I. Yoo, “Design and fabrication of multi-layer antireflection coating for III-V solar cell,” Curr. Appl. Phys. 11(3), 538–541 (2011). [CrossRef]
4. H. Xu, N. Lu, D. Qi, L. Gao, J. Hao, Y. Wang, and L. Chi, “Broadband antireflective Si nanopillar arrays produced by nanosphere lithography,” Microelectron. Eng. 86(4–6), 850–852 (2009). [CrossRef]
5. P. C. Tseng, P. Yu, H. C. Chen, Y. L. Tsai, H. W. Han, M. A. Tsai, C. H. Chang, and H. C. Kuo, “Angle-resolved characteristics of silicon photovoltaics with passivated conical-frustum nanostructures,” Sol. Energy Mater. Sol. Cells 95(9), 2610–2615 (2011). [CrossRef]
6. K. Yamada, M. Umetani, T. Tamura, Y. Tanaka, H. Kasa, and J. Nishii, “Antireflective structure imprinted on the surface of optical glass by SiC mold,” Appl. Surf. Sci. 255(7), 4267–4270 (2009). [CrossRef]
7. J. Y. Chen and K. W. Sun, “Enhancement of the light conversion efficiency of silicon solar cells by using nanoimprint anti-reflection layer,” Sol. Energy Mater. Sol. Cells 94(3), 629–633 (2010). [CrossRef]
8. Q. Xie, M. H. Hong, H. L. Tan, G. X. Chen, L. P. Shi, and T. C. Chong, “Fabrication of nanostructures with laser interference lithography,” J. Alloy. Comp. 449(1–2), 261–264 (2008). [CrossRef]
9. K. S. Cho, P. Mandal, K. Kim, I. H. Baek, S. Lee, H. Lim, D. J. Cho, S. Kim, J. Lee, and F. Rotermund, “Improved efficiency in GaAs solar cells by 1D and 2D nanopatterns fabricated by laser interference lithography,” Opt. Commun. 284(10–11), 2608–2612 (2011). [CrossRef]
10. S. H. Hong, B. J. Bae, J. Y. Hwang, S. Y. Hwang, and H. Lee, “Replication of high ordered nano-sphere array by nanoimprint lithography,” Microelectron. Eng. 86(12), 2423–2426 (2009). [CrossRef]
11. H. L. Chen, S. Y. Chuang, C. H. Lin, and Y. H. Lin, “Using colloidal lithography to fabricate and optimize sub-wavelength pyramidal and honeycomb structures in solar cells,” Opt. Express 15(22), 14793–14803 (2007). [CrossRef] [PubMed]
12. Y. C. Lee, Y. Y. Chou, and C. C. Chang, “Fabrication of broadband anti-reflective sub-micron biomimetic structures by polystyrene sphere lithography on a Si substrate,” (submitted) J. Adv. Eng. (2012).
13. N. Yamada, O. N. Kim, T. Tokimitsu, Y. Nakai, and H. Masuda, “Optimization of anti-reflection moth-eye structures for use in crystalline silicon solar cells,” Prog. Photovolt. Res. Appl. 19(2), 134–140 (2011). [CrossRef]
