The purpose of this study is to reduce the glass substrate reflectivity over a wide spectral range (400-1200nm) without having high reflectivity in the near-infrared region. After making porous SiO2/MgF2 double-layer antireflection (DLAR) thin film structure, the superstrate-type silicon-based tandem cells are added. In comparison to having only silicon-based tandem solar cells, the short-circuit current density has improved by 6.82% when porous SiO2/MgF2 DLAR thin film is applied to silicon-based tandem solar cells. This study has demonstrated that porous SiO2/MgF2 DLAR thin film structure provides antireflection properties over a broad spectral range (400-1200nm) without having high reflectivity at near-infrared wavelengths.
© 2012 OSA
Thin films solar cells on glass are of two configuration types, namely substrate and superstrate . The substrate type involves a glass/Ag grid/front-contact (TCO)/nip (a-Si)/back-contact (TCO)/Ag/glass configuration. The superstrate type involves a glass/front-contact (TCO)/pin (a-Si)/back-contact (TCO)/Ag configuration . Both types are illuminated from the front contact to the back contact. The substrate type of thin film solar cell typically employs encapsulation of the solar cell/module with an additional layer of glass to protect the structure against environmental conditions and physical impact . Unfortunately, the addition of a glass package increases the cost of the substrate type and also results in transmittance loss and decreased conversion efficiency. Compared with the substrate type, superstrate type modules do not require additional glass and thus have lower cost and higher conversion efficiency. Therefore, the superstrate type has been investigated extensively for thin film solar cell fabrication. The superstrate type interface between air and glass reduces light transmittance because of reflection loss in the range of 4%. If the reflectivity loss of 4% is reduced completely, the maximum short-circuit current density (Jsc) in principle is relatively improved about 10.56% . To reduce reflectivity and enhance superstrate solar cell efficiency, one logical solution is to deposit an anti-reflection (AR) coating layer between the glass and the air.
Several methods have been implemented to reduce the reflection of a glass substrate. Approaches mentioned in the literature include addition of silica microspheres or nano-spheres onto the glass [5–7], addition of conically shaped moth-eye structures onto the glass [8,9] and layer-by-layer addition of natural cellulose nano-wires onto the glass . Other antireflection coating techniques have been suggested such as single layer [11–14], double layer [15–17], triple layer [18–20], multi-layer [21,22], graded-index layer  and stacked index layer structures [24,25]. Low reflectance can be obtained from a single-layer antireflection coating only at a specific wavelength. However, for practical solar cell application, low reflectance in both visible and near-infrared wavelengths is required. Low reflectance over a wide wavelength range can be achieved by multi-layer antireflection coatings. For multi-layer antireflection coatings, various materials with different refractive indices are necessary. Unfortunately, multi-layer antireflection coatings can lead to higher reflectivity at near-infrared wavelengths because the multi-layer antireflection thin film structure includes one or more high-refractive-index (n>ns) thin film materials. In a similar study, the MgF2-SiO2 thin films had been prepared by sol-gel process method . Therefore, this paper considers reducing the reflectivity between air and glass by fabricating a porous SiO2/MgF2 DLAR thin film structure. As will be seen, the results give the glass substrate low reflectivity over a wide range of visible light, without high reflectivity in the near-infrared region. Optical characteristics will be presented for the fabricated porous SiO2/MgF2 DLAR thin films. The effects of the fabricated films on solar cell device performance will also be discussed.
2.1. Double-layer antireflection coating
A single layer antireflection coating can be designed to achieve zero reflectance at a single wavelength. On the other hand, a DLAR coating can greatly improve broadband reflectance via the existence of two reflectance minima. The matrix equation for a theoretical DLAR coating of non-absorbing thin films is 
When the nbot is about 1.38 and nglass is about 1.517 from the Eq. (3), the requirement of ntop should be 1.14. However, the material of n = 1.14 cannot be found at present. Hence, we chose n = 1.23 as the top material, due to the n = 1.23 was easily achieved. For the case of DLAR coatings, the required refractive indices determined by Eq. (3) are obtained as ntop = 1.23 and nbot = 1.38. Through Eq. (4), the required thicknesses of the DLAR layers are obtained as dtop = 111.79 nm and dbot = 99.64 nm at the design wavelength of λ0 = 550 nm. Hence, we chose magnesium fluoride (MgF2, nbot = 1.38) as the bottom material and porous silicon dioxide (porous SiO2, ntop = 1.23) as the top material for the antireflection coating materials.
2.2. External quantum efficiency versus short-circuit current density
Generally, the EQE curve of similar solar cell devices without an antireflection coating has the following formula : EQE(λ) = IQE(λ)[1-Rdevice(λ)]. For the DLAR structure employed in this study, the necessary formula corrects to EQE(λ) = IQE(λ)[1-Rdevice(λ) + Rimprove(λ)], where IQE is the internal quantum efficiency, Rdevice is the reflection spectra of the device without the antireflection coating and Rimprove is the reflection spectra of the device with the antireflection coating. The formula relating EQE to short-circuit current density (Jsc) is :
For the bottom layer, a MgF2 thin film was deposited on glass by conventional radio-frequency (13.56 MHz) reactive magnetron sputtering. The substrate temperature was maintained at room temperature while the MgF2 thin film was growing. The radio-frequency magnetron working power and argon gas flow were set at 100 W and 50 sccm, respectively. Working pressures were tested from 3 mtorr to 9 mtorr in order to obtain the required refractive indices. For the top layer, the porous SiO2 thin films were deposited on the MgF2 by the sol gel system. Tetraethyl orthosilicate (TEOS), water (H2O), ethanol (C2H5OH) and ammonia (NH4OH) were set at 0.03 mol, 0.12 mol, 0.36 mol and 0.002 mol respectively. The sol gel temperature and time were maintained at 70°C and 360 min. After dip coating on the MgF2 layer, the coated substrates was spun and then moved into an oven for drying at 80°C and kept there for about 10 min before the next dipping. The samples were baked in a quartz tube with an inner diameter of 10 mm at 500°C for about 120 min. The number of dip/spin coating layers was varied experimentally from 6 to 15 in order to obtain the required thickness.
After adding porous SiO2/MgF2 DLAR thin film structure on one side of glass substrate, fluorine-doped tin oxide (FTO), the top cell, the bottom cell, the Ga-doped zinc oxide (GZO), and the silver are added on the other side of glass substrate in a sequence. Top (a-Si:H) and bottom (uc-Si:H) cell were grown in a plasma enhanced chemical vapor deposition (PECVD) system. The structure of the silicon-based tandem cells is shown in Fig. 1 . The refractive indices and thicknesses of the MgF2 and porous SiO2 thin films were measured by means of automatic scanning spectroscopic ellipsometry (UVISEL, Horiba) in the visible region. Specific surface morphology was observed by scanning electron microscopy (SEM). Fourier-transform infrared spectroscopy (FTIR; Nicolet 5700, Thermo) was used to determine the properties of the porous SiO2 thin films. The resolution was 8 cm−1. An uncoated silicon substrate was used as a reference. The porous SiO2 thin film was deposited on silicon substrate. Solar cell performance was evaluated by measuring current–voltage characteristics under AM1.5G 100mW/cm2 illumination. Quantum efficiency (QE) spectra measurement is performed over the 300–900 nm wavelength range using a Xenon (Xe) lamp. The light transmission and reflection were measured by a UV-Visible-NIR spectrophotometer (U-4100, Hitachi).
4. Results and discussion
The deposition rates and refractive indices of the MgF2 thin films at the various working pressures as obtained by ellipsometer measurement are shown in Fig. 2 . It can be seen that the MgF2 thin film deposition rate increases monotonically as the working pressure increases. This is assumed to be because the number of MgF2 molecules increases at higher working pressures, leading to greater MgF2 thin film thickness. On the other hand, at higher working pressures, the refractive index of the deposited MgF2 thin films declines beyond a working pressure of 5 mtorr, eventually becoming 1.32 at 9 mtorr. From this it is inferred that the MgF2 thin film structure obtained at low pressure deposition is denser than that obtained at higher working pressures. It is assumed that because the deposition rate at lower working pressures slows compared with higher working pressures, the MgF2 molecules at the lower pressures have enough time to form a denser MgF2 thin film structure. For the MgF2 bottom layer, the optimal theoretical refractive index and thickness of the thin film are 1.38 and 99.64 nm, respectively, as determined by Eq. (3). Therefore, by the results of Fig. 2, a working pressure of 5 mtorr is needed to obtain the optimal refractive index of 1.38.
For the porous SiO2 top layer, the thickness and refractive index of the porous SiO2 thin films obtained at various numbers of spin coating layers are shown in Fig. 3 . It can be seen that the refractive index of the porous SiO2 thin film remains between 1.21 and 1.23 even as the number of layers increases. This is because the number of spin coating layers influences only the thickness of the porous SiO2 thin films, not the refractive index. When compared with conventional pure SiO2 thin films, the number of nanopores in our experimental SiO2 thin films is larger. Therefore, the refractive index of the prepared experimental porous SiO2 thin films decreases from 1.46 to 1.23. The very low refractive index of porous SiO2 can be explained with the effective medium model : where VSiO2 is the volume fraction of SiO2 and Vair the volume fraction of the pores filled with air. For general knowledge, the refractive index of SiO2 (nSiO2) and air (nair) is about 1.46 and 1, respectively. When VSiO2 = 1 and Vair = 0, the practical refractive index of SiO2 is about 1.46. When VSiO2 = 0.5 and Vair = 0.5, the practical refractive index of SiO2 is about 1.23. The practical refractive index of SiO2 is similar to our experimental refractive index of porous SiO2.
The SEM images in Fig. 4 show (a) top view and (b) cross-sectional views of the experimentally produced porous SiO2 thin films. It is seen that the produced porous SiO2 thin films have higher nanopore count when compared with pure SiO2 . Figure 5 presents the measured infrared absorbance spectrum of the produced porous SiO2 thin films and also provides evidence of voids. The absorption peaks at 457.8 cm−1, 804 cm−1 and 1065.5 cm−1 are caused by Si-O bond vibration which, due to the appearance of voids, generates an additional open-link absorption peak at 1187.7 cm−1 when compared with pure SiO2 . Therefore, the refractive index of our experimental porous SiO2 thin films decreases from 1.46 to 1.23 with increasing nanopore count. For the top layer, as determined by Eq. (3), the optimal refractive index and thickness of the theoretical porous SiO2 thin film are 1.23 and 111.8 nm, respectively. Therefore, the optimal deposition conditions at the sixth dip/spin layer correspond to a refractive index of 1.23 and a thickness of approximately 111.8nm.
When our experimental porous SiO2/MgF2 DLAR thin film is applied to silicon-based tandem cells, the results can be seen in Fig. 6 and Table 1 . Figure 6 compares the external quantum efficiency (EQE) of the fabricated Si-based tandem cells for the with- and without-porous SiO2/MgF2 DLAR coating conditions. The observations show that the porous SiO2/MgF2 DLAR coating improves the EQE of solar cell at both visible light wavelengths and near-infrared wavelengths. The experimental data shows that the EQE gets better as the Rimprove reduces, presumably because of enhanced transmission as device reflection is reduced, allowing more light to enter the solar cell and be absorbed by the active layer. As seen in Table 1, the short-circuit current density of silicon-based tandem cells has improved from 10.11mA/cm2 to 10.8mA/cm2 when a porous SiO2/MgF2 DLAR coating is added. Jsc improves as the glass surface reflection reduces, confirming net performance enhancement of the solar cell. The enhanced absorption leads to a 6.82% efficiency relative improvement in Jsc. However, as seen in Table 1, it has only a negligible effect to fill factor (FF) and open circuit voltage (Voc) when a porous SiO2/MgF2 DLAR coating is added. The Rimprove result is shown in Fig. 7 . Subtracting the porous SiO2/MgF2 DLAR spectra from the reflective spectra of the uncoated glass gives Rimprove. Figure 7 shows our initial theoretical and our final experimental porous SiO2/MgF2 DLAR spectra, the uncoated glass substrate’s reflective spectra and our theoretical multi-layer antireflection spectra. Light loss via reflection at the glass surface of a solar cell is undesired. Although multi-layer antireflection coatings can reduce reflectivity over a wide range of visible light, conventional AR coatings tend to have high reflectivity at near-infrared wavelengths, causing reduction of the Jsc of tandem cells. The data of Fig. 7 indicate that the prepared porous SiO2/MgF2 DLAR coating has antireflection results over a broad range of visible light wavelengths, without high reflectivity at the near-infrared wavelengths. The AR thin film structure includes one or more high-refractive-index (n>ns) thin film materials, therefore, we can find the high reflectivity at near-infrared wavelengths. However, our experimental porous SiO2/MgF2 DLAR thin film structure is low-refractive-index (n<ns) film materials. The n should change from lower to higher values while the thickness direction is from the top surface to the bottom of the film. Therefore, the high reflectivity at near-infrared wavelengths effectively restrained.
This paper demonstrates a low-cost and simple method for enhancing the efficiency of superstrate-type silicon-based tandem cells. The technique involves adding a porous SiO2/MgF2 DLAR thin film coatings between air and glass. By adjusting working pressures, MgF2 thin films were obtained with the theoretically optimized refractive index of n = 1.38 and thickness of d = 99.64 nm. The optimal deposition conditions were found to be a working pressure of 5 mtorr, a deposition time of 30 min and a power of 100 W, at room temperature. It was found that the refractive index of the experimentally produced porous SiO2 thin films remained between 1.21 and 1.23 regardless of the number of spin coating layers. For the porous SiO2 top layer, the optimal deposition conditions at six layers produced the theoretically optimal refractive index (n = 1.23) and thickness (d = 111.8nm). A refractive index of 1.23 for the produced porous SiO2 thin films was demonstrated by SEM and FTIR methods. When the produced porous SiO2/MgF2 DLAR film was applied to a silicon-based superstrate type tandem solarcell, it was found that the short-circuit current density (Jsc) was relatively improved by 6.82% when compared with the conventional sample. Consequently, the porous SiO2/MgF2 DLAR film relatively improved the efficiency of the silicon-based superstrate type tandem solar cell by 7.14%. The experimental data for the best-result superstrate tandem solar cells with a porous SiO2/MgF2 DLAR film yielded the following results: fill factor (FF = 72%), open-circuit voltage (VOC = 1351.7 mV), short-circuit current density (JSC = 10.8 mA/cm2), efficiency (η = 10.5%). Through theory and experiment, this paper has demonstrated an experimental porous SiO2/MgF2 DLAR coating and the theory for enhancing the JSC of superstrate silicon-based tandem cells. The produced porous SiO2/MgF2 DLAR thin film clearly demonstrates good antireflection properties over a wide spectral range (400-1200nm), without high reflectivity in the near-infrared wavelengths. Since the SiO2/MgF2 DLAR film has a gradual decrease of refractive index from the bottom to top of the film, it shows excellent broadband AR covering from visible light to near infrared wavelengths.
This work was supported by the National Science Council of Taiwan under contract numbers NSC-100-2221-E-006-043-MY2 and NSC-100-2221-E-230-008- and by the Research Center for Energy Technology and Strategy, National Cheng Kung University.
References and links
1. F. J. Haug, D. Rudmann, G. Bilger, H. Zogg, and A. N. Tiwari, “Comparison of structural and electrical properties of Cu(In,Ga)Se2 for substrate and superstrate solar cells,” Thin Solid Films 403-404, 293–296 (2002). [CrossRef]
2. T. Brammer, W. Reetz, N. Senoussaoui, O. Vetterl, O. Kluth, B. Rech, H. Stiebig, and H. Wagner, “Optical properties of silicon-based thin-film solar cells in substrate and superstrate configuration,” Sol. Energy Mater. Sol. Cells 74(1-4), 469–478 (2002). [CrossRef]
3. T. Brammer, W. Reetz, N. Senoussaoui, O. Vetterl, O. Kluth, B. Rech, H. Stiebig, and H. Wagner, “Optical properties of silicon-based thin-film solar cells in substrate and superstrate configuration,” Sol. Energy Mater. Sol. Cells 74(1-4), 469–478 (2002). [CrossRef]
4. K. Orgassa, U. Rau, Q. Nguyen, H. W. Schock, and J. H. Werner, “Role of the CdS buffer layer as an active optical element in Cu(In,Ga)Se2 thin-film solar cells,” Prog. Photovolt. Res. Appl. 10(7), 457–463 (2002). [CrossRef]
5. M. Tao, W. Zhou, H. Yang, and L. Chen, “Surface texturing by solution deposition for omnidirectional antireflection,” Appl. Phys. Lett. 91(8), 081118 (2007). [CrossRef]
6. K. M. Yeung, W. C. Luk, K. C. Tam, C. Y. Kwong, M. A. Tsai, H. C. Kuo, A. M. C. Ng, and A. B. Djurisic, “2-Step self-assembly method to fabricate broadband omnidirectional antireflection coating in large scale,” Sol. Energy Mater. Sol. Cells 95(2), 699–703 (2011). [CrossRef]
7. Y. Wang, L. Chen, H. Yang, Q. Guo, W. Zhou, and M. Tao, “Spherical antireflection coatings by large-area convective assembly of monolayer silica microspheres,” Sol. Energy Mater. Sol. Cells 93(1), 85–91 (2009). [CrossRef]
8. S. H. Hong, B. J. Bae, K. S. Han, E. J. Hong, H. Lee, and K. W. Choi, “Imprinted moth-eye antireflection patterns on glass substrate,” Electron. Mater. Lett. 5(1), 39–42 (2009). [CrossRef]
9. N. C. Linn, C. H. Sun, P. Jiang, and B. Jiang, “Self-assembled biomimetic antireflection coatings,” Appl. Phys. Lett. 91(10), 101108 (2007). [CrossRef]
10. P. Podsiadlo, L. Sui, Y. Elkasabi, P. Burgardt, J. Lee, A. Miryala, W. Kusumaatmaja, M. R. Carman, M. Shtein, J. Kieffer, J. Lahann, and N. A. Kotov, “Layer-by-layer assembled films of cellulose nanowires with antireflective properties,” Langmuir 23(15), 7901–7906 (2007). [CrossRef]
11. M. Chigane, Y. Hatanaka, and T. Shinagawa, “Enhanced antireflection properties of silica thin films via redox deposition and hot-water treatment,” Sol. Energy Mater. Sol. Cells 94(6), 1055–1058 (2010). [CrossRef]
12. A. Jonsson, A. Roos, and E. K. Jonson, “The effect on transparency and light scattering of dip coated antireflection coatings on window glass and electrochromic foil,” Sol. Energy Mater. Sol. Cells 94(6), 992–997 (2010). [CrossRef]
13. C. Ballif, J. Dicker, D. Borchert, and T. Hofmann, “Solar glass with industrial porous SiO2 antireflection coating: measurements of photovoltaic module properties improvement and modelling of yearly energy yield gain,” Sol. Energy Mater. Sol. Cells 82(3), 331–344 (2004). [CrossRef]
14. G. Wu, J. Wang, J. Shen, T. Yang, Q. Zhang, B. Zhou, Z. Deng, B. Fan, D. Zhou, and F. Zhang, “A novel route to control refractive index of sol-gel derived nano-porous silica films used as broadband antireflective coatings,” Mater. Sci. Eng. B 78(2-3), 135–139 (2000). [CrossRef]
15. Z. Liu, X. Zhang, T. Murakami, and A. Fujishima, “Sol-gel SiO2/TiO2 bilayer films with self-cleaning and antireflection properties,” Sol. Energy Mater. Sol. Cells 92(11), 1434–1438 (2008). [CrossRef]
16. H. Nagel, A. Metz, and R. Hezel, “Porous SiO2 flms prepared by remote plasma enhanced chemical vapour deposition - a novel antireflection coating technology for photovoltaic modules,” Sol. Energy Mater. Sol. Cells 65(1-4), 71–77 (2001). [CrossRef]
18. S. W. Kim, D. S. Bae, and H. Shin, “Zinc-embedded silica nanoparticle layer in a multilayer coating on a glass substrate achieves broadband antireflection and high transparency,” J. Appl. Phys. 96(11), 6766–6771 (2004). [CrossRef]
19. J. T. Cox, G. Hass, and A. Thelen, “Triple-layer antireflection coatings on glass for the visible and near infrared,” J. Opt. Soc. Am. 52(9), 965–969 (1962). [CrossRef]
20. M. H. Asghar, M. B. Khan, S. Naseem, and Z. A. Khan, “Design and preparation of antireflection films on glass substrate,” Turk. J. Phys. 29, 43–53 (2005).
23. W. H. Lowdermilk and D. Milam, “Graded-index antireflection surfaces for high-power laser applications,” Appl. Phys. Lett. 36(11), 891–893 (1980). [CrossRef]
24. Y. Y. Liou, C. C. Liu, C. C. Kuo, W. C. Liu, and C. C. Jaing, “Design of universal broadband visible antireflection coating for commonly used glass substrates,” Jpn. J. Appl. Phys. 46(8A), 5143–5147 (2007). [CrossRef]
25. O. Duyar and H. Z. Durusoy, “Design and preparation of antireflection and reflection optical coatings,” Turk. J. Phys. 28, 139–144 (2004).
27. H. Nagel, A. G. Aberle, and R. Hezel, “Optimised antireflection coatings for planar silicon solar cells using remote PECVD silicon nitride and porous silicon dioxide,” Prog. Photovolt. Res. Appl. 7(4), 245–260 (1999). [CrossRef]
28. I. Pereyra and M. I. Alayo, “High quality low temperature DPECVD silicon dioxide,” J. Non-Cryst. Solids 212(2-3), 225–231 (1997). [CrossRef]
29. Y. Liu, W. Ren, L. Zhang, and X. Yao, “New method for making porous SiO2 thin films,” Thin Solid Films 353(1-2), 124–128 (1999). [CrossRef]