Biomimetic nanostructures have shown to enhance the optical absorption of Ga0.5In0.5P/GaAs/Ge triple junction solar cells due to excellent antireflective (AR) properties that, however, are highly dependent on their geometric dimensions. In practice, it is challenging to control fabrication conditions which produce nanostructures in ideal periodic arrangements and with tapered side-wall profiles, leading to sacrificed AR properties and solar cell performance. In this work, we introduce compound biomimetic nanostructures created by depositing a layer of silicon dioxide (SiO2) on top of titanium dioxide (TiO2) nanostructures for triple junction solar cells. The device exhibits photogenerated current and power conversion efficiency that are enhanced by ~8.9% and ~6.4%, respectively, after deposition due to their improved antireflection characteristics. We further investigate and verify the optical properties of compound structures via a rigorous coupled wave analysis model. The additional SiO2 layer not only improves the geometric profile, but also serves as a double-layer dielectric coating. It is concluded that the compound biomimetic nanostructures exhibit superior AR properties that are relatively insensitive to fabrication constraints. Therefore, the compound approach can be widely adopted for versatile optoelectronic devices and applications.
© 2014 Optical Society of America
Multi-junction III-V solar cells are imperative for concentrator photovoltaics due to their intrinsic material properties, such as high radiation resistance and strong broadband absorption [1,2]. Currently, the monolithically-grown InGaP/GaAs/InGaAs and GaInP/GaAs/ GaInNAs cells are the devices reported to have the highest certified power conversion efficiencies of 37.9% and 44.4% respectively, under AM1.5G one-sun and concentrated illumination . However, designing a broadband antireflective coating (ARC) has been a technological challenge yet to fully exploit the wide absorption range of tandem cells. Conventionally, a quarter-wavelength ARC utilizes destructive interference to reduce surface reflection down to almost zero for specific wavelengths at normal incidence. In order to extend the antireflective function to broadband wavelength ranges, double or multi-layer dielectrics need to be implemented. Previously, we demonstrated biomimetic antireflective structures for triple-junction solar cells [4,5], which were originally inspired by moth’s eyes. For the purpose of camouflage, moth’s eyes have evolved to be composed of sub-wavelength features that result in nearly no reflection at night for broad spectral ranges and wide angle of incidence due to optical diffraction [6–8]. By mimicking the morphology on the surface of solar cells, we can recreate the desired antireflective properties [9–12].
Among available nano- materials and patterning methods [13–15], self-assembled nanosphere lithography has become a promising patterning technique for rapid and large-area manufacturing [16–18]. Previously, we have utilized polystyrene nanosphere lithography, followed by reactive ion etching, to fabricate biomimetic nanostructures on silicon nitride (SiNx) and titanium dioxide (TiO2), achieving broadband absorption and efficiency enhancements for triple-junction solar cells [4, 5, 19]. The surface reflection from the triple-junction solar cells is lower than that from the reference counterpart with a conventional single-layer (SL) ARC. However, the AM1.5G-solar-spectrum-weighted reflectance, <R> is still not comparable to the theoretical value due to the aperiodic arrangement of nanosphere assembly and non-ideal side-wall profiles resulted from dry etch. In particular, we think that the aspect ratio and base filling ratio are two important factors which determine the spatially graded refractive index (GRIN) profile presented in the engineered nanostructures . In this work, we propose compound biomimetic nanostructures by depositing a layer of silicon dioxide (SiO2) on top of the TiO2 nanostructures for triple junction solar cells. The antireflective properties of the compound structures are investigated via a rigorous coupled wave analysis (RCWA) model. It has been observed experimentally that the additional SiO2 layer modifies the geometric profile of nanostructures and also functions as a double-layer dielectric coating. However, unlike the conventional double-layer ARC, the compound biomimetic nanostructures are capable of producing very low reflectance spectra without being sensitive to fabrication constraints, such as dielectric thickness, refractive index, and so on. Therefore, the compound biomimetic nanostructures can be widely adopted for many optoelectronic devices and applications that require superior antireflective properties.
2. Results and discussion
Since surface reflection at the interface fundamentally arises from photons experiencing an abrupt change of dielectric environments, the GRIN profile is effective in suppressing the reflective loss for broad wavelengths by means of slowly varying dielectric constants in the direction of propagation. Consequently, the antireflective properties have a broad spectral response and are insensitive to the angle of irradiance incidence. To explain in detail, Fig. 1(a) illustrates a GRIN profile associated with an ideal, two-dimensional, closely-packed nano-cone array with a linearly tapered side-wall profile. For simplicity, we assume the same refractive indices for the nanostructures and the underlying semiconductor. As seen in Fig. 1(a), the vertical slice of air volume ratio inside nanostructures varies linearly from 100% at the tip to nearly 0% on the bottom. Therefore, according to the effective medium theory [21,22], the effective refractive index (neff) changes linearly from that of the air (nair) to approximately that of the semiconductor (ns), leading to complete suppression of interface reflection. Moreover, the slope of the neff profile is determined by the aspect ratio of the nanostructures . In contrast, the nanostructures illustrated in Fig. 1(b) have the same pitch and a linearly tapered side-wall profile as in Fig. 1(a), but also have a smaller base area. As a result, the corresponding neff at the interface is smaller than ns, which results in a small refractive-index difference that increases reflection. In this analogy, when the refractive index of nanostructured material is different from that of the underlying semiconductors, the index difference at the interface contributes to finite reflection. Previously, we first demonstrated SiNx-based (n~1.85) biomimetic nanostructures and then replace SiNx with TiO2 (n~2.45) in order to have a higher refractive index that is closer to typical solar cell materials .
While the effective medium theory provides a compelling argument for the antireflective properties of biomimetic nanostructures, their geometric dependence must be revealed by accurate optical modeling. In this section, we employ a commercial implementation of the three-dimensional RCWA method to investigate the geometric dependence of biomimetic nanostructures and compound structures. The simulated structural profile in a unit cell, imposed with periodic boundary conditions, consists of TiO2 nanostructures with parabolic side wall profiles arranged in a triangular lattice, while the SiO2 layer is conformally lying atop. As shown in the inset of Fig. 2(a), both the pitch and height of TiO2 nanostructures are 400 nm. Below the nanostructures, there are a 400-nm-thick TiO2 layer, the 30-nm-thick Al0.5In0.5P window layer, and the Ga0.5In0.5P top-cell material. These dimensions are consistent with the epitaxial structures and the material dispersion of each layer is taken into account for wavelengths between 300 nm and 1000 nm. Furthermore, the electric field is set to be 45-degree linearly polarized to account for the un-polarized solar radiation. The base diameter of TiO2 nanostructures is set to be a fraction of the pitch varying from 1.0 to 0.6 and then converted to the filling ratio (FR), defined as the base area of the nanostructures to the unit cell area.
Figure 2(a) plots the AM1.5G-solar-spectrum-weighted reflectance, <R> as a function of the filling ratio without the SiO2 layer. It can be seen that the lowest <R> is ~1.12%, indeed occurred at closest packing (FR = 0.907), when the base diameter equals the pitch. As discussed previously, when the filling ratio is decreased, <R> is increased. Next, Fig. 2(b) shows <R> as a function of SiO2 layer thickness. It is found that adding a SiO2 layer onto a close-pack nanostructure array (FR = 0.907) can only worsen the antireflective properties, as shown in the black line of Fig. 2(b). On the other hand, when FR = 0.735, adding SiO2 can further lower the weighted reflectance, <R> down to 1.04% and 1.05% with a thickness of 50nm and 100nm, respectively (red line). The calculation results imply that the surface reflectance of the compound structures can be even lower than that of ideal periodic nanostructures. In other words, the combined effect of refractive index matching and GRIN profile has the potential to suppress <R> down to ~1% for complex epitaxial structures underneath. Moreover, it is found that the thickness of SiO2 is not very critical for the FR in the range of 0.7-0.3. This is also different from the interference concept in conventional ARCs. However, when the FR is decreased to around 0.227, the weighted reflectance <R> apparently has an oscillatory dependence on the SiO2 thickness. We think that, in this case, the distribution of nanostructures is sparse across the entire unit cell area. Hence, the SiO2 layer serves more as a double layer dielectric coating and is sensitive to layer thickness. Figure 2(c) plots a complete design map as a function of the filling ratio and the SiO2 layer thickness. The realm of <R> ~1% occurs at FR~0.76 and SiO2 thicknesses between 60and 110nm.
Next, we examine the measured and calculated antireflective properties of TiO2 and compound SiO2/TiO2 biomimetic nanostructures fabricated by using polystyrene nanosphere lithography, followed by reactive ion etching, and electron beam evaporation of SiO2. First, a 600-nm-thick titanium dioxide (TiO2) layer was evaporated on a silicon substrate. The refractive index of the deposited TiO2 is around 2.5 with an extinction coefficient close to zero for most of the visible and near infrared wavelengths, characterized by an n&k analyzer (n&k Technology 1200). Next, the device sample was spun-cast with a mixture suspension containing ethanol and polystyrene sub-microspheres (1:1 wt%) with an average diameter of 600 nm. As show in Fig. 3(a), nanospheres were self-assembled into a mono-layer of dense grains with dimensions in several micrometers and within each grain, the polystyrene spheres are packed with finite spacings (Fig. 3(b)). After steady air-drying, the samples were etched by an inductively-coupled-plasma reactive-ion-etching (ICP-RIE) and high-density-plasma reactive-ion-etching (HDP-RIE) systems using a gas mixture of CHF3, O2, BCl3, and Cl2 through individual electronic mass flow controllers. The dry etch generally underwent three steps: First, O2 and CHF3 were used to reduce the size of polystyrene spheres by ICP-RIE. Next, BCl3 was used to increase the etching rate of TiO2 to achieve straight side walls, and lastly, a mixture of BCl3 and Cl2 was used to simultaneously etch nanospheres and TiO2 in order to achieve a tapered side-wall profile by HDP-RIE. We successfully controlled the etching profile by tuning the ratio of gas flow and etching time simultaneously o fabricate the nanostructures shown in Fig. 3(c). The tilted scanning electron microscopic (SEM, JEOL JSM-7000F) images in Fig. 3(c) result from the conditions of O2:CHF3 = 5:5, BCl3 = 40 and BCl3:Cl2 = 35: 35 sccm under 600W RF bias power for etching times of 120, 240 and 60 seconds, respectively.
Figure 4(a) shows the measured reflectance spectrum of the fabricated TiO2 nanostructures shown in Fig. 3(c) at a nearly normal incident angle 5° using an UV/Visible/NIR spectrophotometer (Hitachi U4100), where the reflected light is collected by an integrating sphere. We then establish an optical model, composed of a 3x3 TiO2 nanostructure array in a triangular lattice with ± 25% random position variations, in order to first fit the measured spectrum and then investigate the properties of compound nanostructures with various SiO2 thicknesses. The simulated nanostructure has a base diameter of 450 nm and height of 350nm with a pitch of ~620 nm, which corresponds to a base filling ratio of ~0.68. As shown in Fig. 4(a), the experimental and theoretical curves achieve very good agreement, where the AM1.5G weighted reflectance <R> is 9.94 and 9.29% respectively. With this experimentally validated model, we then calculate <R> as a function of SiO2 layer thickness, as shown in Fig. 4(b). We next deposit a 100- and 200-nm-thick SiO2 layer on top of the fabricated TiO2 nanostructures. The resulting SEM images are shown in Fig. 4(c) and 4(d) respectively. The corresponding <R> is calculated from the measured spectra and the value is very close to the projected, confirming the accuracy of this optical model. As shown in Fig. 4(b), we first note that the values of <R> are higher than the periodic examples even with the same base filling ratio (FR~0.7). This is due to the finite aspect ratio of fabricated structures, particularly the insufficient height, which limits the slope of GRIN profile . Second, the lowest reflectance occurs with a 100-nm-thick SiO2 layer, and the reflectance increases as the SiO2 layer becomes thicker. The dependence of the reflectance on the layer thickness is similar to those observed in Fig. 3(b) for small FRs, indicating the presence of double-layer dielectric coating effect in nanostructures with a low aspect ratio.
Finally, we compare the photovoltaic performance of triple-junction solar cells with two different antireflective schemes: TiO2 and compound SiO2/TiO2 nanostructures. The device structure and fabrication flow with antireflective nanostructures are described in detail elsewhere [4, 19]. The cell was first characterized under one-sun illumination. Then the bus bar of top electrode was blocked with tape and then the same cell was evaporated with a 70-nm-thick SiO2 layer using an electron-beam evaporator. Figure 5(a) shows the current density-voltage curves and characterization parameters of the device before and after the SiO2 deposition on TiO2 nanostructures. It is also worth noting that the open-circuit voltage (Voc) and the fill factor (FF) for cells with 70-nm-thick SiO2 are slightly deteriorated due to the damaged electric contacts after the deposition of SiO2. However the cell with compound biomimetic nanostructures results in a short-circuit current density (Jsc) of 11.79 mA/cm2 and a power conversion efficiency (η) of 21.85%, which are ~8.9% and ~6.4% increases over the cells with only TiO2 nanostructures, respectively. The clear enhancement in the short-circuit current density results from the improved antireflection, as shown in the reflectivity spectra in Fig. 5(b). The compound nanostructures improve the optical transmission for visible and infrared wavelengths, such that <R> is reduced to 6.7% from 8.7% with the addition of SiO2 in the device. The measured external quantum efficiency (EQE) spectra from the top and middle sub-cells are presented in Fig. 5(b), which confirms that the mid-junction limited photocurrent and the enhancement of EQE correspond well with the reduced reflectance.
In summary, we have successfully analyzed and demonstrated the antireflective properties of compound SiO2/TiO2 biomimetic structures for Ga0.5In0.5P/GaAs/Ge triple-junction solar cells. The device exhibits photogenerated current and power conversion efficiency that are improved by ~8.9% and ~6.4% relative to the reference, respectively. The validated optical model indicates that the additional SiO2 layer not only improves the geometric profile, but also serves as a double-layer dielectric coating. Since the compound biomimetic nanostructures exhibit excellent AR properties and are not sensitive to fabrication constraints, this approach can be widely adopted for many optoelectronic applications.
The authors thank Dr. T. G. Chen and Prof. Keith Barnham at the Imperial College London UK for fruitful discussions. This project was founded by the National Science Council in Taiwan under grant number 100-2628-E-009-020-MY3.
References and links
1. M. Yamaguchi, N. Kojima, A. Khan, T. Takamoto, K. Ando, M. Imaizumi, and T. Sumita, “Radiation-resistant and high-efficiency InGaP/InGaAs/Ge 3-junction solar cells,” International Symposium on Compound Semiconductors, 189–190 (2003). [CrossRef]
2. H. Cotal, C. Fetzer, J. Boisvert, G. Kinsey, R. King, P. Hebert, H. Yoon, and N. Karam, “III-V multijunction solar cells for concentrating photovoltaics,” Energy Environ. Sci. 2(2), 174–192 (2009). [CrossRef]
3. M. A. Green, K. Emery, Y. Hishikawa, W. Warta, and E. D. Dunlop, “Solar cell efficiency tables (version 41),” Prog. Photovolt. Res. Appl. 21(1), 1–11 (2013). [CrossRef]
4. P. Yu, M. Y. Chiu, C. H. Chang, C. Y. Hong, Y. L. Tsai, H. V. Han, and Y. R. Wu, “Towards high-efficiency multi-junction solar cells with biologically inspired nanosurfaces,” Wiley Online Library, Aug. 2, 2012.
5. M. Y. Chiu, C. H. Chang, M. A. Tsai, F. Y. Chang, and P. Yu, “Improved optical transmission and current matching of a triple-junction solar cell utilizing sub-wavelength structures,” Opt. Express 18(S3Suppl 3), A308–A313 (2010). [CrossRef] [PubMed]
7. S. J. Wilson and M. C. Hutley, “The optical properties of ‘moth eye’ antireflection surfaces,” Opt. Acta Int. J. of Opt. 29(7), 993–1009 (1982). [CrossRef]
9. J. Y. Chen, W. L. Chang, C. K. Huang, and K. W. Sun, “Biomimetic nanostructured antireflection coating and its application on crystalline silicon solar cells,” Opt. Express 19(15), 14411–14419 (2011). [CrossRef] [PubMed]
10. S. Jeong, E. C. Garnett, S. Wang, Z. Yu, S. Fan, M. L. Brongersma, M. D. McGehee, and Y. Cui, “Hybrid Silicon Nanocone-Polymer Solar Cells,” Nano Lett. 12(6), 2971–2976 (2012). [CrossRef] [PubMed]
11. D. S. Kim, M. S. Park, and J. H. Jang, “Fabrication of cone-shaped subwavelength structures by utilizing a confined convective self-assembly technique and inductively coupled-plasma reactive-ion etching,” J. Vac. Sci. Technol. B 29(2), 020602 (2011). [CrossRef]
12. J. W. Leem, Y. M. Song, and J. S. Yu, “Broadband wide-angle antireflection enhancement in AZO/Si shell/core subwavelength grating structures with hydrophobic surface for Si-based solar cells,” Opt. Express 19(S5Suppl 5), A1155–A1164 (2011). [CrossRef] [PubMed]
13. J. Tommila, V. Polojärvi, A. Aho, A. Tukiainen, J. Viheriälä, J. Salmi, A. Schramm, J. M. Kontio, A. Turtiainen, T. Niemi, and M. Guina, “Nanostructured broadband antireflection coatings on AlInP fabricated by nanoimprint lithography,” Sol. Energy Mater. Sol. Cells 94(10), 1845–1848 (2010). [CrossRef]
14. Y. F. Huang, S. Chattopadhyay, Y. J. Jen, C. Y. Peng, T. A. Liu, Y. K. Hsu, C. L. Pan, H. C. Lo, C. H. Hsu, Y. H. Chang, C. S. Lee, K. H. Chen, and L. C. Chen, “Improved broadband and quasi-omnidirectional anti-reflection properties with biomimetic silicon nanostructures,” Nat. Nanotechnol. 2(12), 770–774 (2007). [CrossRef] [PubMed]
15. T. Lohmüller, M. Helgert, M. Sundermann, R. Brunner, and J. P. Spatz, “Biomimetic interfaces for high-performance optics in the deep-UV light range,” Nano Lett. 8(5), 1429–1433 (2008). [CrossRef] [PubMed]
16. 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]
17. A. D. Ormonde, E. C. M. Hicks, J. Castillo, and R. P. Van Duyne, “Nanosphere lithography: Fabrication of large-area Ag nanoparticle arrays by convective self-assembly and their characterization by scanning UV-visible extinction spectroscopy,” Langmuir 20(16), 6927–6931 (2004). [CrossRef] [PubMed]
19. K. H. Hung, T. G. Chen, T. T. Yang, P. Yu, C. Y. Hong, Y. R. Wu, and G. C. Chi, “Antireflective scheme for InGaP/InGaAs/Ge triple junction solar cells based on TiO2 biomimetic structures,” in Proceedings of 38th IEEE Conference on Photovoltaic Specialists Conference(2012), pp.003322–003324.
20. C. H. Chiu, P. Yu, H. C. Kuo, C. C. Chen, T. C. Lu, S. C. Wang, S. H. Hsu, Y. J. Cheng, and Y. C. Chang, “Broadband and omnidirectional antireflection employing disordered GaN nanopillars,” Opt. Express 16(12), 8748–8754 (2008). [CrossRef] [PubMed]
21. Z. Xiong, F. Zhao, J. Yang, and X. Hu, “Comparison of optical absorption in Si nanowire and nanoporous Si structures for photovoltaic applications,” Appl. Phys. Lett. 96(18), 181903 (2010). [CrossRef]
22. X. Hu, C. T. Chan, J. Zi, M. Li, and K. M. Ho, “Diamagnetic response of metallic photonic crystals at infrared and visible frequencies,” Phys. Rev. Lett. 96(22), 223901 (2006). [CrossRef] [PubMed]