Advancing tandem solar cells by spectrally selective multilayer intermediate reflectors Andre Hoffmann,1,* Ulrich W. Paetzold,1 Chao Zhang1, Tsvetelina Merdzhanova,1 Andreas Lambertz,1 Carolin Ulbrich,1 Karsten Bittkau,1 and Uwe Rau1 1
IEK5 - Photovoltaik, Forschungszentrum Jülich GmbH, 52425 Jülich, Germany *[email protected]
Abstract: Thin-film silicon tandem solar cells are composed of an amorphous silicon top cell and a microcrystalline silicon bottom cell, stacked and connected in series. In order to match the photocurrents of the top cell and the bottom cell, a proper photon management is required. Up to date, single-layer intermediate reflectors of limited spectral selectivity are applied to match the photocurrents of the top and the bottom cell. In this paper, we design and prototype multilayer intermediate reflectors based on aluminum doped zinc oxide and doped microcrystalline silicon oxide with a spectrally selective reflectance allowing for improved current matching and an overall increase of the charge carrier generation. The intermediate reflectors are successfully integrated into state-of-the-art tandem solar cells resulting in an increase of overall short-circuit current density by 0.7 mA/cm2 in comparison to a tandem solar cell with the standard single-layer intermediate reflector. © 2014 Optical Society of America OCIS codes: (040.5350) Photovoltaic; (350.4238) Nanophotonics and photonic crystals.
References and links 1.
A. Vos, “Detailed balance limit of the efficiency of tandem solar cells,” J. Phys. D Appl. Phys. 13(5), 839–846 (1980). 2. J. Meier, S. Dubail, R. Platz, P. Torres, U. Kroll, J. A. Anna Selvan, N. Pellaton Vaucher, C. Hofa, D. Fischera, H. Keppnera, R. Flückigera, A. Shaha, V. Shklover, and K.-D. Ufert, “Towards high-efficiency thin-film silicon solar cells with the ‘micromorph’ concept,” Sol. Energy Mater. Sol. Cells 49(1–4), 35–44 (1997). 3. C. Ulbrich, C. Zahren, A. Gerber, B. Blank, T. Merdzhanova, A. Gordijn, and U. Rau, “Matching of silicon thinfilm tandem solar cells for maximum power output,” Int. J. Photoenergy 2013, 314097 (2013). 4. J. Meier, S. Dubail, S. Golay, U. Kroll, S. Fay, E. Vallat-Sauvain, J. Feitknecht, J. Dubail, and A. Shah, “Microcrystalline silicon and the impact on micromorph tandem solar cells,” Sol. Energy Mater. Sol. Cells 74(14), 457–467 (2002). 5. B. Rech, T. Repmann, M. N. van den Donker, M. Berginski, T. Kilper, J. Hüpkes, S. Calman, H. Stiebig, and S. Wieder, “Challenges in microcrystalline silicon based solar cell technology,” Thin Solid Films 511–512, 548– 555 (2006). 6. D. L. Staebler and C. R. Wronski, “Reversible conductivity changes in discharge–produced amorphous Si,” Appl. Phys. Lett. 31(4), 292–294 (1977). 7. K. Yamamoto, A. Nakajima, M. Yoshimi, T. Sawada, S. Fukuda, T. Suezaki, M. Ichikawa, Y. Koi, M. Goto, H. Takata, T. Sasaki, and Y. Tawada, “Novel hybrid thin film silicon solar cell and module,” in 3rd World Conf. on PV Energy Conv. (2003). 8. P. Buehlmann, J. Bailat, D. Dominé, A. Billet, F. Meillaud, A. Feltrin, and C. Ballif, “In situ silicon oxide based intermediate reflector for thin–film silicon micromorph solar cells,” Appl. Phys. Lett. 91(14), 143505 (2007). 9. A. Lambertz, T. Grundler, and F. Finger, “Hydrogenated amorphous silicon oxide containing a microcrystalline silicon phase and usage as an intermediate reflector in thin–film silicon solar cells,” J. Appl. Phys. 109(11), 113109 (2011). 10. A. Lambertz, V. Smirnov, T. Merdzhanova, K. Ding, S. Haas, G. Jost, R. E. I. Schropp, F. Finger, and U. Rau, “Microcrystalline silicon–oxygen alloys for application in silicon solar cells and modules,” Sol. Energy Mater. Sol. Cells 119, 134–143 (2013). 11. J. Üpping, A. Bielawny, R. B. Wehrspohn, T. Beckers, R. Carius, U. Rau, S. Fahr, C. Rockstuhl, F. Lederer, M. Kroll, T. Pertsch, L. Steidl, and R. Zentel, “Three-dimensional photonic crystal intermediate reflectors for enhanced light-trapping in tandem solar cells,” Adv. Mater. 23(34), 3896–3900 (2011).
#206526 - $15.00 USD Received 17 Feb 2014; revised 17 Apr 2014; accepted 17 Apr 2014; published 28 Jul 2014 (C) 2014 OSA 25 August 2014 | Vol. 22, No. S5 | DOI:10.1364/OE.22.0A1270 | OPTICS EXPRESS A1270
12. G. Jost, T. Merdzhanova, T. Zimmermann, and J. Hüpkes, “Process monitoring of texture-etched high-rate ZnO:Al front contacts for silicon thin-film solar cells,” Thin Solid Films 532, 66–72 (2013). 13. J. Krc, F. Smole, and M. Topic, “Optical simulation of the role of reflecting interlayers in tandem micromorph silicon solar cells,” Sol. Energy Mater. Sol. Cells 86(4), 537–550 (2005). 14. A. Bielawny, C. Rockstuhl, F. Lederer, and R. B. Wehrspohn, “Intermediate reflectors for enhanced top cell performance in photovoltaic thin-film tandem cells,” Opt. Express 17(10), 8439–8446 (2009). 15. P. G. O’Brien, A. Chutinan, K. Leong, N. P. Kherani, G. A. Ozin, and S. Zukotynski, “Photonic crystal intermediate reflectors for micromorph solar cells: A comparative study,” Opt. Express 18(5), 4478–4490 (2010). 16. P. G. O’Brien, Y. Yang, A. Chutinan, P. Mahtani, K. Leong, D. P. Puzzo, L. D. Bonifacio, C.-W. Lina, G. A. Ozin, and N. P. Kherani, “Selectively transparent and conducting photonic crystal solar spectrum splitters made of alternating sputtered indium–tin oxide and spin–coated silica nanoparticle layers for enhanced photovoltaics,” Sol. Energy Mater. Sol. Cells 102, 173–183 (2012). 17. S. Y. Myong and L. S. Jeon, “Improved light trapping in thin-film silicon solar cells via alternated n-type silicon oxide reflectors,” Sol. Energy Mater. Sol. Cells 119, 77–83 (2013). 18. A. Hoffmann, U. W. Paetzold, T. Merdzhanova, A. Lambertz, O. Höhn, C. Ulbrich, K. Bittkau, and U. Rau, “Spectrally selective intermediate reflectors for tandem thin-film silicon solar cells,” Proc. SPIE 8823, 882305 (2013). 19. W. Theiss, “SCOUT Software Package,” http://www.mtheiss.com. 20. P. Bienstman and R. Baets, “Optical modelling of photonic crystals and VCSELs using eigenmode expansion and perfectly matched layers,” Opt. Quantum Electron. 33(4/5), 327–341 (2001). 21. A. F. Oskooi, D. Roundy, M. Ibanescu, P. Bermel, J. D. Joannopoulos, and S. G. Johnson, “MEEP: A flexible free–software package for electromagnetic simulations by the FDTD method,” Comput. Phys. Commun. 181(3), 687–702 (2010). 22. C. Rockstuhl, F. Lederer, K. Bittkau, T. Beckers, and R. Carius, “The impact of intermediate reflectors on light absorption in tandem solar cells with randomly textured surfaces,” Appl. Phys. Lett. 94(21), 211101 (2009). 23. U. W. Paetzold, E. Moulin, B. E. Pieters, U. Rau, and R. Carius, “Optical simulations of microcrystalline silicon solar cells applying plasmonic reflection grating back contacts,” J. Photon. Ener. 2(1), 027002 (2012). 24. D. Lockau, L. Zschiedrich, S. Burger, F. Schmidt, F. Ruske, and B. Rech, “Rigorous optical simulation of light management in crystalline silicon thin–film solar cells with rough textured interfaces,” Proc. SPIE 7933, 79330M (2011). 25. J. D. Joannopoulos, P. R. Villeneuve, and S. Fan, “Photonic crystals: Putting a new twist on light,” Nature 386(6621), 143–149 (1997). 26. H. A. Macleod, Thin-Film Optical Filters, 3rd ed. (Taylor and Francis, 2001). 27. S. Fahr, C. Rockstuhl, and F. Lederer, “The interplay of intermediate reflectors and randomly textured surfaces in tandem solar cells,” Appl. Phys. Lett. 97(17), 173510 (2010).
1. Introduction In tandem solar cells, photocurrent generation for incident light of a broad spectral range is combined with a reduction of intrinsic thermalization losses [1]. In the special case of thinfilm silicon technology, hydrogenated amorphous silicon (a-Si:H) and hydrogenated microcrystalline silicon (µc-Si:H) exhibit appropriate band-gaps for a tandem solar cell [2]. The overall design of the tandem device is restricted by the requirement of current-matching at the maximum power point of both sub cells [3]. The matching requirement is usually achieved by adjusting the thickness of the a-Si:H top cell and µc-Si:H bottom cell. However, the power-matching condition is also affected by the imperfect carrier collection as well as light-induced degradation of a-Si:H which are both inherent to the material [4–6]. Therefore, a central task for advanced optical concepts in tandem thin-film silicon solar cells is the minimization of the physical thickness and a maximization of the optical light path in the aSi:H top cell. Namely, an intermediate reflector (IR) is employed between the a-Si:H top cell and the µc-Si:H bottom cell to reflect non-absorbed photons back to the top cell and, thereby, increase the optical path and absorptance in the top cell [7].
#206526 - $15.00 USD Received 17 Feb 2014; revised 17 Apr 2014; accepted 17 Apr 2014; published 28 Jul 2014 (C) 2014 OSA 25 August 2014 | Vol. 22, No. S5 | DOI:10.1364/OE.22.0A1270 | OPTICS EXPRESS A1271
Fig. 1. (a) Scheme of a tandem thin-film silicon solar cell with intermediate reflector. (b) The external quantum efficiency (EQE) of a tandem thin–film silicon solar cell without (black line) and with intermediate reflector (blue). Full lines give the EQE of the sub cells, the sum of the EQEs is shown in dashed lines.
Figure 1(a) displays a schematic drawing of such an IR embedded in an exemplary layer stack thin-film silicon tandem solar cell. Commonly, thin layers (30–150 nm) of transparent and conductive materials with a refractive index n lower than that of a-Si:H are used as single layer IRs [7–10]. The preferred material is the mixed-phase material microcrystalline silicon oxide (µc-SiOx:H). But also ZnO:Al is researched [2,4]. In contrast to three-dimensional photonic crystal IRs [8], conventional IRs consisting of a single layer provide almost no spectral selectivity. In Fig. 1(b), the external quantum efficiency (EQE) for two tandem solar cells deposited in one run with (blue line) and without IR (black line) is shown. The cells were prepared on a sputtered flat ZnO:Al [12] and SnO:F coated glass from the Asahi Glass Company (AGC) (type VU). By introducing the state-of-the-art 40 nm n-type µc-SiOx:H layer with a refractive index of n = 2.8 at a wavelength of λ = 600 nm between top and bottom cell, a portion of the incident light is reflected back into the top cell increasing the EQEtop of the top cell [10–12]. However, this increase is accompanied by a decrease of the EQEbot of the bottom cell in a wide spectral range (500 nm to 1000 nm), finally reducing the sum EQEsum of EQEtop of the top and EQEbot of the bottom cell [10,13]. In Fig. 1(b), three spectral ranges are indicated by I-III. In the range λ < 520 nm, photons are entirely absorbed before reaching the IR. In range II from 520 < λ < 680nm, an increasing amount of photons reaches the back side of the top cell and, consequently also the IR. Due to the IR, the light path within the top cell and, therefore, the EQE of the top cell is increased. In range III (λ > 680nm), the absorptance of the a-Si:H top-cell is low, therefore, the top cell EQEtop is just slightly increased. The µc-Si:H bottom cell is also highly absorptive in this range. An IR which reflects a large portion of light in this range will prevent the light to reach the bottom cell and, thus, prevent to contribute to charge carrier generation in the bottom cell. This is clearly seen at the maximum of the bottom cell EQEbot (λ = 700 nm) which is reduced by the implementation of single layer IRs (blue line in Fig. 1(b)). In the present work, we design a multilayer IR made from alternating ZnO:Al and µcSiOx:H layers. Multilayer IRs have been proposed and studied by several groups [13–15] and recently fabricated [16,17]. This study bases on our previous work dealing with an optimization of spectral selectivity [18]. The present study demonstrates the design and application of a multilayer IR. The total solar cell short-circuit current density is increased compared to a tandem solar cell with a single layer IR.
#206526 - $15.00 USD Received 17 Feb 2014; revised 17 Apr 2014; accepted 17 Apr 2014; published 28 Jul 2014 (C) 2014 OSA 25 August 2014 | Vol. 22, No. S5 | DOI:10.1364/OE.22.0A1270 | OPTICS EXPRESS A1272
2. Experimental and simulation methods 2.1 Solar cell fabrication The a-Si:H top cell, the µc-Si:H bottom cell, as well as the µc–SiOx:H IR layers were deposited in a plasma–enhanced chemical vapor deposition (PECVD) system on a transparent conductive oxide (TCO) serving as front electrode. Detailed information on the optical and electric properties of n-type µc-SiOx:H films is given elsewhere [10]. As front contact, we use fluorine doped tin dioxide (SnO2:F) coated glass from the Asahi Glass Company (AGC) (type VU) and non-etched, flat ZnO:Al. The ZnO:Al layers were sputtered from a Rotatable Dual Magnetron (RDM) deposition system [12]. 2.2 Characterization Reflectance R and transmittance T of the single layers were measured by a Perkin Elmer UV/Vis spectrophotometer. The optical dispersion of refractive index n(λ) and extinction coefficient k(λ) of these materials was deduced from these measurements based on the Kramer-Kronig relation (KKR) susceptibility model in the software package SCOUT [19]. To achieve the precise sub-bandgap extinction coefficient, photothermal deflection spectroscopy (PDS) measurements were performed and k(λ) is derived. Solar cell reflectance R was measured by an UV/Vis spectrophotometer and the solar cell absorptance is obtained as A = 1-R. The external quantum efficiency EQE was measured using a grating monochromator setup (FWHM of the monochromatic light: λ = 10 nm) measuring in the wavelength range between 300 nm and 1100 nm. 2.3 Electromagnetic simulations A versatile coherent transfer-matrix algorithm [20] was used to calculate reflectance and transmittance of the flat layer stacks. Optical simulations of textured surfaces are much more complex as a statistical, sufficiently large area (4 µm × 4 µm) has to be chosen and Maxwell’s equations have to be solved rigorously [21–24]. We chose a modified open-source finitedifference time-domain (FDTD) solver Meep [21], in order to simulate the 3D distribution of electromagnetic fields [22]. The topography of the surface after a-Si:H deposition is measured by atomic force microscopy (AFM) and used as texture of the IR in the simulation in order to provide sufficient statistics of surface features [24]. The IR was embedded between two nonabsorbing a-Si:H/µc-Si:H half spaces and illuminated by a plane wave. n(λ) and k(λ) of the involved materials were used. Reflectance and transmission was calculated from the Poynting vector in a plane perpendicular to the incident wave propagation direction. 3. Results and discussion 3.1 Design of multilayer stacks Figure 2(a) shows the simulated reflectance of different flat IR configurations into a nonabsorbing a-Si:H half space as a function of the wavelength. For these simulations, we use the transfer matrix method. State-of-the-art IRs consist of a single layer with lower refractive index than the neighboring silicon material. The reflectance into silicon for a ZnO:Al IR (n = 2.0) with a thickness of 114 nm is shown as a blue curve. The single layer IR reflects a portion of the light in the required wavelength range-depending on the refractive index contrast. Still, reflectance at longer wavelengths (region III) is approx. 20% at 680 nm >λ> 1100 nm in case of the ZnO:Al IR with n = 2 and show no spectral selectivity. As the absorption length of a-Si:H is long in region III, reflectance of the IR in this spectral range does not lead to a significant absorption enhancement in the top cell and, furthermore, the back reflected light does not couple into the bottom cell and cannot be absorbed and contribute to charge carrier generation. A Bragg reflector or 1D photonic crystal [14–16]
#206526 - $15.00 USD Received 17 Feb 2014; revised 17 Apr 2014; accepted 17 Apr 2014; published 28 Jul 2014 (C) 2014 OSA 25 August 2014 | Vol. 22, No. S5 | DOI:10.1364/OE.22.0A1270 | OPTICS EXPRESS A1273
provides spectral selectivity by reflecting back just in a certain wavelength range related to its photonic band gap [24].
Fig. 2. (a) Reflectance R into an a-Si:H half space for an increasing number of µc–SiOx:H and ZnO:Al layers. The reflectance of a single layer of 110 nm ZnO:Al (n = 2) is shown in blue. The reflectance of an alternating stack of N = 100 (IR100)/ N = 3 IR3) layers is depicted in black/ red, respectively. The reflectance of the flat IR3 stack (dashed, red line) is compared to an IR3 on rough substrate (full, red) in (b).
In Fig. 2(a), the reflectance of a stack of 100 alternating layers of materials with refractive indices 2 and 2.8 is depicted as black line. Low refractive index λ0/8 layers are integrated on both sides of the stack to maximize the transmittance in range III [26]. As wavelength of design, λ0 = 620 nm is used. A well-pronounced stop band is found between 480 nm and 620 nm, while long-wavelength absorption stays below 10%. A similar behavior is found for a layer stack just consisting of three layers, basically ZnO:Al (n = 2) / µc-SiOx:H (n = 2.8) / ZnO:Al (n = 2). Here the edge is broadened to a lower slope and reflectance in the stop band is around 80% in comparison to 97% in the IR100. Such 100-layer stacks are elaborate and more layers result in an increased series resistance and parasitic absorption of the IR. The simpler 3-layer intermediate reflector (IR3) is composed of a (i) 62 nm thick ZnO:Al layer, (ii) 38 nm thick n-type µc-SiOx:H layer and (iii) a 52 nm thick ZnO:Al layer. In order to integrate the IR3 into state-of-the-art tandem solar cells, it has to be considered that interfaces are textured in order to provide light scattering and, thus, trap the light. Therefore, rigorous optical simulations of the textured layer stack are performed. Figure 2(b) shows the reflectance (red) of the IR3 on flat (dashed line) and Asahi VU-type substrate (full line) into a non-absorbing a-Si:H halfspace over the wavelength. The reflectance of the rough IR shows a similar trend as in the flat case but reflectance is generally reduced by 40% as the symmetry of the filter is disturbed by the roughness [22,27]. A significant portion of the light up to 50% is still reflected for the target spectral range, while above 700 nm, less than 5% are reflected back. It can be concluded that the designed IR3 works as a spectrally selective filter despite its random texture due to the roughness of the front contact. Comparing the IR3 to an IR made from ZnO:Al with a thickness of 114 nm (cyan line) in Fig. 3, the transmittance into the µcSi:H bottom cell in range III obtained by rigorous optical simulations is significantly increased by the IR3 as a consequence of spectral selectivity. The studied IR3 was designed for normally incidence light. The dependence of the reflectance of multilayer intermediate reflectors for oblique angles has been shown in [18]. A shift of the photonic band-gap to higher energies was reported for increasing angles. As light is refracted to lower angles when entering into the high refractive index silicon, normal incidence is a good approximation for typical in-field angles.
#206526 - $15.00 USD Received 17 Feb 2014; revised 17 Apr 2014; accepted 17 Apr 2014; published 28 Jul 2014 (C) 2014 OSA 25 August 2014 | Vol. 22, No. S5 | DOI:10.1364/OE.22.0A1270 | OPTICS EXPRESS A1274
Fig. 3. The transmittance into the non-absorbing µc-Si:H halfspace is shown for the IR3 and ZnO IR simulated by rigorous optical simulations.
3.2 Experimental results Subsequent to the conceptual design of the IR, the following section shows the implementation of the 3-layer IR (IR3) into state-of-the-art tandem thin-film silicon solar cells. Figure 4(a) shows the EQE and absorptance A = 1-R of a tandem solar cell on a flat ZnO:Al substrate with an a-Si:H i-layer thickness of 330 nm and a µc-Si:H i-layer thickness of 3.2 µm without IR (black) and with IR3 (red). As an effect of the intermediate reflector on the top cell EQEtop, a maximum at λ = 620 nm is found. Here, the absorptance of the solar cell is increased to more than 90%, so nearly the whole portion of back-reflected light is absorbed. With increasing wavelength, less of the back-reflected photons can be absorbed in the top cell and, thus, escape out of the solar cell. This is resembled in the absorptance minimum (reflectance maximum) at λ = 700 nm and, consequently, as a dip in the bottom cell EQEbot. Figure 4(c) shows the EQE ratio path enhancement in the top cell: EQEtop,IR / EQEtop,w/o. A local maximum at λ = 620 nm is found in EQEtop with IR. A decreased EQEbot and absorptance above λ = 800 nm in comparison to the solar cell without IR is found, which is probably due to the substrate-dependent growth of µc-Si:H i-layer. The substrate dependency is visible in the thickness of the i-layer grown on the IR3 that is reduced by 105 nm and the Raman crystallinity IC at a wavelength of λ = 532 nm that is reduced to 47% in comparison to the reference tandem cell without IR (IC = 56%). The IR3 is then integrated into a state-of-theart tandem solar cell on Asahi VU substrate. The photovoltaic parameters of the tandem solar cells with the various studied IRs are shown in Table 1. Figure 4(b) shows the EQE of the solar cell with different IR designs. Without the intermediate reflector (black line), a clear mismatch between top and bottom cell current is seen. Integrating a single layer intermediate reflector leads to an increase of top cell EQEtop between 500 and 750 nm. As mentioned before, the amount of increase depends on the refractive index and thickness of the IR. The ZnO:Al IR (cyan line) reflects a large portion of light which yields a reflectance of around 20% (Fig. 2). The absence of this large fraction of the incoming light leads to the significant losses in the bottom cell. In the case of the µc-SiOx:H IR (blue line), bottom cell losses are lower. Yet, the boosting effect on top cell EQEtop is low as well. The red line illustrates the EQE and absorptance of the IR3. Between λ = 500 nm and 600 nm, the top cell EQEtop exceeds the one of the thick ZnO:Al IR. With increasing wavelength, less photons are absorbed in the top cell resulting in an absorptance dip at around λ = 610 nm.
#206526 - $15.00 USD Received 17 Feb 2014; revised 17 Apr 2014; accepted 17 Apr 2014; published 28 Jul 2014 (C) 2014 OSA 25 August 2014 | Vol. 22, No. S5 | DOI:10.1364/OE.22.0A1270 | OPTICS EXPRESS A1275
Fig. 4. (a) Measured external quantum efficiency EQE and absorptance A = 1-R of a flat tandem solar cell without (black) and with our IR3 (red line). (b) EQE and absorptance A = 1-R for the various IR designs in a textured tandem solar cell on AsahiVU substrate. The EQE ratio path enhancement in the top cell in the flat case EQEtop,IR / EQEtop,w/o is shown in (c) while (d) shows the EQE ratio EQEsum,IR / EQEsum,w/o for the textured tandem solar cells. Table 1. Photovoltaic parameters short circuit current density Jsc of the sub cells and sum, open-circuit voltage Voc, fill factor FF, as well as conversion efficiency η of the solar cells with the studied IR designs measured in a sun simulator at AM1.5 spectrum. Thicknesses of the intrinsic layers were 330 nm (a-Si:H) and a 3.2 µm (µc-Si:H). IR type w/o IR
Jsc,top [mA/cm2] 11.5
Jsc,bot [mA/cm2] 14.2
Jsc,sum [mA/cm2] 25.7
Jsc,meas [mA/cm2] 11.7
Voc [mV] 1.323
FF [%] 75.1
η [%] 11.7
SiOx:H IR
12.0
11.1
23.1
11.0
1.344
71.6
10.8
ZnO:Al IR
13.2
9.4
22.6
9.0
1.352
74.3
9.6
IR3
12.7
11.1
23.8
10.7
1.334
73.8
10.7
The decreasing reflectance of the IR is nicely seen here as the top cell EQEtop remains below the optimized ZnO:Al IR. The bottom cell quantum efficiency of the tandem solar cell with IR3 is higher than the EQEbot of the tandem solar cell with standard µc-SiOx:H IR above a wavelength of 620 nm. Taking the reflectance of the textured IR from Fig. 3(a) into account, this can be attributed to the spectrally selective reflectance of our designed multilayer IR. A meaningful quantity to show this is the ratio EQEsum,IR / EQEsum,w/o as shown in Fig. 4(d). The spectral selectivity is nicely seen in an EQE ratio maximum at λ = 680nm. The minimum at about λ = 720 nm is due to a non-perfect spectral selectivity and could be enlarged by a steeper reflectance edge. The EQEsum ratio of the IR3 is below a ratio of one but superior to the other IR designs in range II and III. As can be seen in Table 1, all IRs turn the top limitation (Jsc,top < Jsc,bot) of the cell without IR into a bottom limitation. For the IR3 reflector the short-circuit current density of the top cell Jsc,top is increased by 1.2 mA/cm2 compared to the configuration without IR and by 0.7 mA/cm2 compared to the standard µc-SiOx:H IR. It can be seen, that the presented solar cells are not current matched. A thinner top-cell or an improved light trapping at the back side would be options to improve the matching. Differences of Voc are small and probably due to the variation of crystallinity of the absorber material. The fill factor FF remains unaltered within the measurement uncertainty not significantly influenced by the incorporation of the IRs.
#206526 - $15.00 USD Received 17 Feb 2014; revised 17 Apr 2014; accepted 17 Apr 2014; published 28 Jul 2014 (C) 2014 OSA 25 August 2014 | Vol. 22, No. S5 | DOI:10.1364/OE.22.0A1270 | OPTICS EXPRESS A1276
4. Conclusions In this paper, a multilayer intermediate reflector composed of ZnO:Al and µc-SiOx:H was designed and experimentally integrated into tandem thin-film silicon solar cells. We have studied the impact of spectral selectivity of various single layer and multilayer intermediate reflectors on the performance of state-of-the-art thin-film silicon tandem solar cells. It was shown by simulations as well as prototypes that single layer IRs lead to an increased reflection for longer wavelengths while multilayer intermediate reflectors provide a much more spectrally selective reflectance minimizing reflection losses. Integrating a multilayer intermediate reflector leads to an increase of top and bottom cell short-circuit current density compared to state-of-the-art single-layer IRs. Such selective filters allows for advanced spectral splitting in future multi-junction solar cells. Acknowledgments We acknowledge J. Kirchoff, H. Siekmann, U. Gerhards and A. Bauer for depositions and laser contacting. R. Carius, S. Lehnen, M. Smeets, M. Meier, M. Zilk and O. Höhn are acknowledged for helpful discussions, as well as M. Ermes and B.E. Pieters for simulation support. We thank Jülich Supercomputing Center (JSC) for simulation on JUROPA cluster through VSR project. The Federal ministry of education and research is acknowledged for funding within the InfraVolt project grant no. 03SF0401C.
#206526 - $15.00 USD Received 17 Feb 2014; revised 17 Apr 2014; accepted 17 Apr 2014; published 28 Jul 2014 (C) 2014 OSA 25 August 2014 | Vol. 22, No. S5 | DOI:10.1364/OE.22.0A1270 | OPTICS EXPRESS A1277
IEK5 - Photovoltaik, Forschungszentrum Jülich GmbH, 52425 Jülich, Germany *[email protected]
Abstract: Thin-film silicon tandem solar cells are composed of an amorphous silicon top cell and a microcrystalline silicon bottom cell, stacked and connected in series. In order to match the photocurrents of the top cell and the bottom cell, a proper photon management is required. Up to date, single-layer intermediate reflectors of limited spectral selectivity are applied to match the photocurrents of the top and the bottom cell. In this paper, we design and prototype multilayer intermediate reflectors based on aluminum doped zinc oxide and doped microcrystalline silicon oxide with a spectrally selective reflectance allowing for improved current matching and an overall increase of the charge carrier generation. The intermediate reflectors are successfully integrated into state-of-the-art tandem solar cells resulting in an increase of overall short-circuit current density by 0.7 mA/cm2 in comparison to a tandem solar cell with the standard single-layer intermediate reflector. © 2014 Optical Society of America OCIS codes: (040.5350) Photovoltaic; (350.4238) Nanophotonics and photonic crystals.
References and links 1.
A. Vos, “Detailed balance limit of the efficiency of tandem solar cells,” J. Phys. D Appl. Phys. 13(5), 839–846 (1980). 2. J. Meier, S. Dubail, R. Platz, P. Torres, U. Kroll, J. A. Anna Selvan, N. Pellaton Vaucher, C. Hofa, D. Fischera, H. Keppnera, R. Flückigera, A. Shaha, V. Shklover, and K.-D. Ufert, “Towards high-efficiency thin-film silicon solar cells with the ‘micromorph’ concept,” Sol. Energy Mater. Sol. Cells 49(1–4), 35–44 (1997). 3. C. Ulbrich, C. Zahren, A. Gerber, B. Blank, T. Merdzhanova, A. Gordijn, and U. Rau, “Matching of silicon thinfilm tandem solar cells for maximum power output,” Int. J. Photoenergy 2013, 314097 (2013). 4. J. Meier, S. Dubail, S. Golay, U. Kroll, S. Fay, E. Vallat-Sauvain, J. Feitknecht, J. Dubail, and A. Shah, “Microcrystalline silicon and the impact on micromorph tandem solar cells,” Sol. Energy Mater. Sol. Cells 74(14), 457–467 (2002). 5. B. Rech, T. Repmann, M. N. van den Donker, M. Berginski, T. Kilper, J. Hüpkes, S. Calman, H. Stiebig, and S. Wieder, “Challenges in microcrystalline silicon based solar cell technology,” Thin Solid Films 511–512, 548– 555 (2006). 6. D. L. Staebler and C. R. Wronski, “Reversible conductivity changes in discharge–produced amorphous Si,” Appl. Phys. Lett. 31(4), 292–294 (1977). 7. K. Yamamoto, A. Nakajima, M. Yoshimi, T. Sawada, S. Fukuda, T. Suezaki, M. Ichikawa, Y. Koi, M. Goto, H. Takata, T. Sasaki, and Y. Tawada, “Novel hybrid thin film silicon solar cell and module,” in 3rd World Conf. on PV Energy Conv. (2003). 8. P. Buehlmann, J. Bailat, D. Dominé, A. Billet, F. Meillaud, A. Feltrin, and C. Ballif, “In situ silicon oxide based intermediate reflector for thin–film silicon micromorph solar cells,” Appl. Phys. Lett. 91(14), 143505 (2007). 9. A. Lambertz, T. Grundler, and F. Finger, “Hydrogenated amorphous silicon oxide containing a microcrystalline silicon phase and usage as an intermediate reflector in thin–film silicon solar cells,” J. Appl. Phys. 109(11), 113109 (2011). 10. A. Lambertz, V. Smirnov, T. Merdzhanova, K. Ding, S. Haas, G. Jost, R. E. I. Schropp, F. Finger, and U. Rau, “Microcrystalline silicon–oxygen alloys for application in silicon solar cells and modules,” Sol. Energy Mater. Sol. Cells 119, 134–143 (2013). 11. J. Üpping, A. Bielawny, R. B. Wehrspohn, T. Beckers, R. Carius, U. Rau, S. Fahr, C. Rockstuhl, F. Lederer, M. Kroll, T. Pertsch, L. Steidl, and R. Zentel, “Three-dimensional photonic crystal intermediate reflectors for enhanced light-trapping in tandem solar cells,” Adv. Mater. 23(34), 3896–3900 (2011).
#206526 - $15.00 USD Received 17 Feb 2014; revised 17 Apr 2014; accepted 17 Apr 2014; published 28 Jul 2014 (C) 2014 OSA 25 August 2014 | Vol. 22, No. S5 | DOI:10.1364/OE.22.0A1270 | OPTICS EXPRESS A1270
12. G. Jost, T. Merdzhanova, T. Zimmermann, and J. Hüpkes, “Process monitoring of texture-etched high-rate ZnO:Al front contacts for silicon thin-film solar cells,” Thin Solid Films 532, 66–72 (2013). 13. J. Krc, F. Smole, and M. Topic, “Optical simulation of the role of reflecting interlayers in tandem micromorph silicon solar cells,” Sol. Energy Mater. Sol. Cells 86(4), 537–550 (2005). 14. A. Bielawny, C. Rockstuhl, F. Lederer, and R. B. Wehrspohn, “Intermediate reflectors for enhanced top cell performance in photovoltaic thin-film tandem cells,” Opt. Express 17(10), 8439–8446 (2009). 15. P. G. O’Brien, A. Chutinan, K. Leong, N. P. Kherani, G. A. Ozin, and S. Zukotynski, “Photonic crystal intermediate reflectors for micromorph solar cells: A comparative study,” Opt. Express 18(5), 4478–4490 (2010). 16. P. G. O’Brien, Y. Yang, A. Chutinan, P. Mahtani, K. Leong, D. P. Puzzo, L. D. Bonifacio, C.-W. Lina, G. A. Ozin, and N. P. Kherani, “Selectively transparent and conducting photonic crystal solar spectrum splitters made of alternating sputtered indium–tin oxide and spin–coated silica nanoparticle layers for enhanced photovoltaics,” Sol. Energy Mater. Sol. Cells 102, 173–183 (2012). 17. S. Y. Myong and L. S. Jeon, “Improved light trapping in thin-film silicon solar cells via alternated n-type silicon oxide reflectors,” Sol. Energy Mater. Sol. Cells 119, 77–83 (2013). 18. A. Hoffmann, U. W. Paetzold, T. Merdzhanova, A. Lambertz, O. Höhn, C. Ulbrich, K. Bittkau, and U. Rau, “Spectrally selective intermediate reflectors for tandem thin-film silicon solar cells,” Proc. SPIE 8823, 882305 (2013). 19. W. Theiss, “SCOUT Software Package,” http://www.mtheiss.com. 20. P. Bienstman and R. Baets, “Optical modelling of photonic crystals and VCSELs using eigenmode expansion and perfectly matched layers,” Opt. Quantum Electron. 33(4/5), 327–341 (2001). 21. A. F. Oskooi, D. Roundy, M. Ibanescu, P. Bermel, J. D. Joannopoulos, and S. G. Johnson, “MEEP: A flexible free–software package for electromagnetic simulations by the FDTD method,” Comput. Phys. Commun. 181(3), 687–702 (2010). 22. C. Rockstuhl, F. Lederer, K. Bittkau, T. Beckers, and R. Carius, “The impact of intermediate reflectors on light absorption in tandem solar cells with randomly textured surfaces,” Appl. Phys. Lett. 94(21), 211101 (2009). 23. U. W. Paetzold, E. Moulin, B. E. Pieters, U. Rau, and R. Carius, “Optical simulations of microcrystalline silicon solar cells applying plasmonic reflection grating back contacts,” J. Photon. Ener. 2(1), 027002 (2012). 24. D. Lockau, L. Zschiedrich, S. Burger, F. Schmidt, F. Ruske, and B. Rech, “Rigorous optical simulation of light management in crystalline silicon thin–film solar cells with rough textured interfaces,” Proc. SPIE 7933, 79330M (2011). 25. J. D. Joannopoulos, P. R. Villeneuve, and S. Fan, “Photonic crystals: Putting a new twist on light,” Nature 386(6621), 143–149 (1997). 26. H. A. Macleod, Thin-Film Optical Filters, 3rd ed. (Taylor and Francis, 2001). 27. S. Fahr, C. Rockstuhl, and F. Lederer, “The interplay of intermediate reflectors and randomly textured surfaces in tandem solar cells,” Appl. Phys. Lett. 97(17), 173510 (2010).
1. Introduction In tandem solar cells, photocurrent generation for incident light of a broad spectral range is combined with a reduction of intrinsic thermalization losses [1]. In the special case of thinfilm silicon technology, hydrogenated amorphous silicon (a-Si:H) and hydrogenated microcrystalline silicon (µc-Si:H) exhibit appropriate band-gaps for a tandem solar cell [2]. The overall design of the tandem device is restricted by the requirement of current-matching at the maximum power point of both sub cells [3]. The matching requirement is usually achieved by adjusting the thickness of the a-Si:H top cell and µc-Si:H bottom cell. However, the power-matching condition is also affected by the imperfect carrier collection as well as light-induced degradation of a-Si:H which are both inherent to the material [4–6]. Therefore, a central task for advanced optical concepts in tandem thin-film silicon solar cells is the minimization of the physical thickness and a maximization of the optical light path in the aSi:H top cell. Namely, an intermediate reflector (IR) is employed between the a-Si:H top cell and the µc-Si:H bottom cell to reflect non-absorbed photons back to the top cell and, thereby, increase the optical path and absorptance in the top cell [7].
#206526 - $15.00 USD Received 17 Feb 2014; revised 17 Apr 2014; accepted 17 Apr 2014; published 28 Jul 2014 (C) 2014 OSA 25 August 2014 | Vol. 22, No. S5 | DOI:10.1364/OE.22.0A1270 | OPTICS EXPRESS A1271
Fig. 1. (a) Scheme of a tandem thin-film silicon solar cell with intermediate reflector. (b) The external quantum efficiency (EQE) of a tandem thin–film silicon solar cell without (black line) and with intermediate reflector (blue). Full lines give the EQE of the sub cells, the sum of the EQEs is shown in dashed lines.
Figure 1(a) displays a schematic drawing of such an IR embedded in an exemplary layer stack thin-film silicon tandem solar cell. Commonly, thin layers (30–150 nm) of transparent and conductive materials with a refractive index n lower than that of a-Si:H are used as single layer IRs [7–10]. The preferred material is the mixed-phase material microcrystalline silicon oxide (µc-SiOx:H). But also ZnO:Al is researched [2,4]. In contrast to three-dimensional photonic crystal IRs [8], conventional IRs consisting of a single layer provide almost no spectral selectivity. In Fig. 1(b), the external quantum efficiency (EQE) for two tandem solar cells deposited in one run with (blue line) and without IR (black line) is shown. The cells were prepared on a sputtered flat ZnO:Al [12] and SnO:F coated glass from the Asahi Glass Company (AGC) (type VU). By introducing the state-of-the-art 40 nm n-type µc-SiOx:H layer with a refractive index of n = 2.8 at a wavelength of λ = 600 nm between top and bottom cell, a portion of the incident light is reflected back into the top cell increasing the EQEtop of the top cell [10–12]. However, this increase is accompanied by a decrease of the EQEbot of the bottom cell in a wide spectral range (500 nm to 1000 nm), finally reducing the sum EQEsum of EQEtop of the top and EQEbot of the bottom cell [10,13]. In Fig. 1(b), three spectral ranges are indicated by I-III. In the range λ < 520 nm, photons are entirely absorbed before reaching the IR. In range II from 520 < λ < 680nm, an increasing amount of photons reaches the back side of the top cell and, consequently also the IR. Due to the IR, the light path within the top cell and, therefore, the EQE of the top cell is increased. In range III (λ > 680nm), the absorptance of the a-Si:H top-cell is low, therefore, the top cell EQEtop is just slightly increased. The µc-Si:H bottom cell is also highly absorptive in this range. An IR which reflects a large portion of light in this range will prevent the light to reach the bottom cell and, thus, prevent to contribute to charge carrier generation in the bottom cell. This is clearly seen at the maximum of the bottom cell EQEbot (λ = 700 nm) which is reduced by the implementation of single layer IRs (blue line in Fig. 1(b)). In the present work, we design a multilayer IR made from alternating ZnO:Al and µcSiOx:H layers. Multilayer IRs have been proposed and studied by several groups [13–15] and recently fabricated [16,17]. This study bases on our previous work dealing with an optimization of spectral selectivity [18]. The present study demonstrates the design and application of a multilayer IR. The total solar cell short-circuit current density is increased compared to a tandem solar cell with a single layer IR.
#206526 - $15.00 USD Received 17 Feb 2014; revised 17 Apr 2014; accepted 17 Apr 2014; published 28 Jul 2014 (C) 2014 OSA 25 August 2014 | Vol. 22, No. S5 | DOI:10.1364/OE.22.0A1270 | OPTICS EXPRESS A1272
2. Experimental and simulation methods 2.1 Solar cell fabrication The a-Si:H top cell, the µc-Si:H bottom cell, as well as the µc–SiOx:H IR layers were deposited in a plasma–enhanced chemical vapor deposition (PECVD) system on a transparent conductive oxide (TCO) serving as front electrode. Detailed information on the optical and electric properties of n-type µc-SiOx:H films is given elsewhere [10]. As front contact, we use fluorine doped tin dioxide (SnO2:F) coated glass from the Asahi Glass Company (AGC) (type VU) and non-etched, flat ZnO:Al. The ZnO:Al layers were sputtered from a Rotatable Dual Magnetron (RDM) deposition system [12]. 2.2 Characterization Reflectance R and transmittance T of the single layers were measured by a Perkin Elmer UV/Vis spectrophotometer. The optical dispersion of refractive index n(λ) and extinction coefficient k(λ) of these materials was deduced from these measurements based on the Kramer-Kronig relation (KKR) susceptibility model in the software package SCOUT [19]. To achieve the precise sub-bandgap extinction coefficient, photothermal deflection spectroscopy (PDS) measurements were performed and k(λ) is derived. Solar cell reflectance R was measured by an UV/Vis spectrophotometer and the solar cell absorptance is obtained as A = 1-R. The external quantum efficiency EQE was measured using a grating monochromator setup (FWHM of the monochromatic light: λ = 10 nm) measuring in the wavelength range between 300 nm and 1100 nm. 2.3 Electromagnetic simulations A versatile coherent transfer-matrix algorithm [20] was used to calculate reflectance and transmittance of the flat layer stacks. Optical simulations of textured surfaces are much more complex as a statistical, sufficiently large area (4 µm × 4 µm) has to be chosen and Maxwell’s equations have to be solved rigorously [21–24]. We chose a modified open-source finitedifference time-domain (FDTD) solver Meep [21], in order to simulate the 3D distribution of electromagnetic fields [22]. The topography of the surface after a-Si:H deposition is measured by atomic force microscopy (AFM) and used as texture of the IR in the simulation in order to provide sufficient statistics of surface features [24]. The IR was embedded between two nonabsorbing a-Si:H/µc-Si:H half spaces and illuminated by a plane wave. n(λ) and k(λ) of the involved materials were used. Reflectance and transmission was calculated from the Poynting vector in a plane perpendicular to the incident wave propagation direction. 3. Results and discussion 3.1 Design of multilayer stacks Figure 2(a) shows the simulated reflectance of different flat IR configurations into a nonabsorbing a-Si:H half space as a function of the wavelength. For these simulations, we use the transfer matrix method. State-of-the-art IRs consist of a single layer with lower refractive index than the neighboring silicon material. The reflectance into silicon for a ZnO:Al IR (n = 2.0) with a thickness of 114 nm is shown as a blue curve. The single layer IR reflects a portion of the light in the required wavelength range-depending on the refractive index contrast. Still, reflectance at longer wavelengths (region III) is approx. 20% at 680 nm >λ> 1100 nm in case of the ZnO:Al IR with n = 2 and show no spectral selectivity. As the absorption length of a-Si:H is long in region III, reflectance of the IR in this spectral range does not lead to a significant absorption enhancement in the top cell and, furthermore, the back reflected light does not couple into the bottom cell and cannot be absorbed and contribute to charge carrier generation. A Bragg reflector or 1D photonic crystal [14–16]
#206526 - $15.00 USD Received 17 Feb 2014; revised 17 Apr 2014; accepted 17 Apr 2014; published 28 Jul 2014 (C) 2014 OSA 25 August 2014 | Vol. 22, No. S5 | DOI:10.1364/OE.22.0A1270 | OPTICS EXPRESS A1273
provides spectral selectivity by reflecting back just in a certain wavelength range related to its photonic band gap [24].
Fig. 2. (a) Reflectance R into an a-Si:H half space for an increasing number of µc–SiOx:H and ZnO:Al layers. The reflectance of a single layer of 110 nm ZnO:Al (n = 2) is shown in blue. The reflectance of an alternating stack of N = 100 (IR100)/ N = 3 IR3) layers is depicted in black/ red, respectively. The reflectance of the flat IR3 stack (dashed, red line) is compared to an IR3 on rough substrate (full, red) in (b).
In Fig. 2(a), the reflectance of a stack of 100 alternating layers of materials with refractive indices 2 and 2.8 is depicted as black line. Low refractive index λ0/8 layers are integrated on both sides of the stack to maximize the transmittance in range III [26]. As wavelength of design, λ0 = 620 nm is used. A well-pronounced stop band is found between 480 nm and 620 nm, while long-wavelength absorption stays below 10%. A similar behavior is found for a layer stack just consisting of three layers, basically ZnO:Al (n = 2) / µc-SiOx:H (n = 2.8) / ZnO:Al (n = 2). Here the edge is broadened to a lower slope and reflectance in the stop band is around 80% in comparison to 97% in the IR100. Such 100-layer stacks are elaborate and more layers result in an increased series resistance and parasitic absorption of the IR. The simpler 3-layer intermediate reflector (IR3) is composed of a (i) 62 nm thick ZnO:Al layer, (ii) 38 nm thick n-type µc-SiOx:H layer and (iii) a 52 nm thick ZnO:Al layer. In order to integrate the IR3 into state-of-the-art tandem solar cells, it has to be considered that interfaces are textured in order to provide light scattering and, thus, trap the light. Therefore, rigorous optical simulations of the textured layer stack are performed. Figure 2(b) shows the reflectance (red) of the IR3 on flat (dashed line) and Asahi VU-type substrate (full line) into a non-absorbing a-Si:H halfspace over the wavelength. The reflectance of the rough IR shows a similar trend as in the flat case but reflectance is generally reduced by 40% as the symmetry of the filter is disturbed by the roughness [22,27]. A significant portion of the light up to 50% is still reflected for the target spectral range, while above 700 nm, less than 5% are reflected back. It can be concluded that the designed IR3 works as a spectrally selective filter despite its random texture due to the roughness of the front contact. Comparing the IR3 to an IR made from ZnO:Al with a thickness of 114 nm (cyan line) in Fig. 3, the transmittance into the µcSi:H bottom cell in range III obtained by rigorous optical simulations is significantly increased by the IR3 as a consequence of spectral selectivity. The studied IR3 was designed for normally incidence light. The dependence of the reflectance of multilayer intermediate reflectors for oblique angles has been shown in [18]. A shift of the photonic band-gap to higher energies was reported for increasing angles. As light is refracted to lower angles when entering into the high refractive index silicon, normal incidence is a good approximation for typical in-field angles.
#206526 - $15.00 USD Received 17 Feb 2014; revised 17 Apr 2014; accepted 17 Apr 2014; published 28 Jul 2014 (C) 2014 OSA 25 August 2014 | Vol. 22, No. S5 | DOI:10.1364/OE.22.0A1270 | OPTICS EXPRESS A1274
Fig. 3. The transmittance into the non-absorbing µc-Si:H halfspace is shown for the IR3 and ZnO IR simulated by rigorous optical simulations.
3.2 Experimental results Subsequent to the conceptual design of the IR, the following section shows the implementation of the 3-layer IR (IR3) into state-of-the-art tandem thin-film silicon solar cells. Figure 4(a) shows the EQE and absorptance A = 1-R of a tandem solar cell on a flat ZnO:Al substrate with an a-Si:H i-layer thickness of 330 nm and a µc-Si:H i-layer thickness of 3.2 µm without IR (black) and with IR3 (red). As an effect of the intermediate reflector on the top cell EQEtop, a maximum at λ = 620 nm is found. Here, the absorptance of the solar cell is increased to more than 90%, so nearly the whole portion of back-reflected light is absorbed. With increasing wavelength, less of the back-reflected photons can be absorbed in the top cell and, thus, escape out of the solar cell. This is resembled in the absorptance minimum (reflectance maximum) at λ = 700 nm and, consequently, as a dip in the bottom cell EQEbot. Figure 4(c) shows the EQE ratio path enhancement in the top cell: EQEtop,IR / EQEtop,w/o. A local maximum at λ = 620 nm is found in EQEtop with IR. A decreased EQEbot and absorptance above λ = 800 nm in comparison to the solar cell without IR is found, which is probably due to the substrate-dependent growth of µc-Si:H i-layer. The substrate dependency is visible in the thickness of the i-layer grown on the IR3 that is reduced by 105 nm and the Raman crystallinity IC at a wavelength of λ = 532 nm that is reduced to 47% in comparison to the reference tandem cell without IR (IC = 56%). The IR3 is then integrated into a state-of-theart tandem solar cell on Asahi VU substrate. The photovoltaic parameters of the tandem solar cells with the various studied IRs are shown in Table 1. Figure 4(b) shows the EQE of the solar cell with different IR designs. Without the intermediate reflector (black line), a clear mismatch between top and bottom cell current is seen. Integrating a single layer intermediate reflector leads to an increase of top cell EQEtop between 500 and 750 nm. As mentioned before, the amount of increase depends on the refractive index and thickness of the IR. The ZnO:Al IR (cyan line) reflects a large portion of light which yields a reflectance of around 20% (Fig. 2). The absence of this large fraction of the incoming light leads to the significant losses in the bottom cell. In the case of the µc-SiOx:H IR (blue line), bottom cell losses are lower. Yet, the boosting effect on top cell EQEtop is low as well. The red line illustrates the EQE and absorptance of the IR3. Between λ = 500 nm and 600 nm, the top cell EQEtop exceeds the one of the thick ZnO:Al IR. With increasing wavelength, less photons are absorbed in the top cell resulting in an absorptance dip at around λ = 610 nm.
#206526 - $15.00 USD Received 17 Feb 2014; revised 17 Apr 2014; accepted 17 Apr 2014; published 28 Jul 2014 (C) 2014 OSA 25 August 2014 | Vol. 22, No. S5 | DOI:10.1364/OE.22.0A1270 | OPTICS EXPRESS A1275
Fig. 4. (a) Measured external quantum efficiency EQE and absorptance A = 1-R of a flat tandem solar cell without (black) and with our IR3 (red line). (b) EQE and absorptance A = 1-R for the various IR designs in a textured tandem solar cell on AsahiVU substrate. The EQE ratio path enhancement in the top cell in the flat case EQEtop,IR / EQEtop,w/o is shown in (c) while (d) shows the EQE ratio EQEsum,IR / EQEsum,w/o for the textured tandem solar cells. Table 1. Photovoltaic parameters short circuit current density Jsc of the sub cells and sum, open-circuit voltage Voc, fill factor FF, as well as conversion efficiency η of the solar cells with the studied IR designs measured in a sun simulator at AM1.5 spectrum. Thicknesses of the intrinsic layers were 330 nm (a-Si:H) and a 3.2 µm (µc-Si:H). IR type w/o IR
Jsc,top [mA/cm2] 11.5
Jsc,bot [mA/cm2] 14.2
Jsc,sum [mA/cm2] 25.7
Jsc,meas [mA/cm2] 11.7
Voc [mV] 1.323
FF [%] 75.1
η [%] 11.7
SiOx:H IR
12.0
11.1
23.1
11.0
1.344
71.6
10.8
ZnO:Al IR
13.2
9.4
22.6
9.0
1.352
74.3
9.6
IR3
12.7
11.1
23.8
10.7
1.334
73.8
10.7
The decreasing reflectance of the IR is nicely seen here as the top cell EQEtop remains below the optimized ZnO:Al IR. The bottom cell quantum efficiency of the tandem solar cell with IR3 is higher than the EQEbot of the tandem solar cell with standard µc-SiOx:H IR above a wavelength of 620 nm. Taking the reflectance of the textured IR from Fig. 3(a) into account, this can be attributed to the spectrally selective reflectance of our designed multilayer IR. A meaningful quantity to show this is the ratio EQEsum,IR / EQEsum,w/o as shown in Fig. 4(d). The spectral selectivity is nicely seen in an EQE ratio maximum at λ = 680nm. The minimum at about λ = 720 nm is due to a non-perfect spectral selectivity and could be enlarged by a steeper reflectance edge. The EQEsum ratio of the IR3 is below a ratio of one but superior to the other IR designs in range II and III. As can be seen in Table 1, all IRs turn the top limitation (Jsc,top < Jsc,bot) of the cell without IR into a bottom limitation. For the IR3 reflector the short-circuit current density of the top cell Jsc,top is increased by 1.2 mA/cm2 compared to the configuration without IR and by 0.7 mA/cm2 compared to the standard µc-SiOx:H IR. It can be seen, that the presented solar cells are not current matched. A thinner top-cell or an improved light trapping at the back side would be options to improve the matching. Differences of Voc are small and probably due to the variation of crystallinity of the absorber material. The fill factor FF remains unaltered within the measurement uncertainty not significantly influenced by the incorporation of the IRs.
#206526 - $15.00 USD Received 17 Feb 2014; revised 17 Apr 2014; accepted 17 Apr 2014; published 28 Jul 2014 (C) 2014 OSA 25 August 2014 | Vol. 22, No. S5 | DOI:10.1364/OE.22.0A1270 | OPTICS EXPRESS A1276
4. Conclusions In this paper, a multilayer intermediate reflector composed of ZnO:Al and µc-SiOx:H was designed and experimentally integrated into tandem thin-film silicon solar cells. We have studied the impact of spectral selectivity of various single layer and multilayer intermediate reflectors on the performance of state-of-the-art thin-film silicon tandem solar cells. It was shown by simulations as well as prototypes that single layer IRs lead to an increased reflection for longer wavelengths while multilayer intermediate reflectors provide a much more spectrally selective reflectance minimizing reflection losses. Integrating a multilayer intermediate reflector leads to an increase of top and bottom cell short-circuit current density compared to state-of-the-art single-layer IRs. Such selective filters allows for advanced spectral splitting in future multi-junction solar cells. Acknowledgments We acknowledge J. Kirchoff, H. Siekmann, U. Gerhards and A. Bauer for depositions and laser contacting. R. Carius, S. Lehnen, M. Smeets, M. Meier, M. Zilk and O. Höhn are acknowledged for helpful discussions, as well as M. Ermes and B.E. Pieters for simulation support. We thank Jülich Supercomputing Center (JSC) for simulation on JUROPA cluster through VSR project. The Federal ministry of education and research is acknowledged for funding within the InfraVolt project grant no. 03SF0401C.
#206526 - $15.00 USD Received 17 Feb 2014; revised 17 Apr 2014; accepted 17 Apr 2014; published 28 Jul 2014 (C) 2014 OSA 25 August 2014 | Vol. 22, No. S5 | DOI:10.1364/OE.22.0A1270 | OPTICS EXPRESS A1277
