- Email: [email protected]
Al2O3 antireflection layer between glass and transparent conducting oxide for enhanced light trapping in microcrystalline silicon thin film solar cells
Solar Energy Materials & Solar Cells 101 (2012) 22–25
Contents lists available at SciVerse ScienceDirect
Solar Energy Materials & Solar Cells journal homepage: www.elsevier.com/locate/solmat
Letter
Al2O3 antireflection layer between glass and transparent conducting oxide for enhanced light trapping in microcrystalline silicon thin film solar cells Dong-Won Kang a, Jang-Yeon Kwon b,n, Jenny Shim c, Heon-Min Lee c, Min-Koo Han a a
Department of Electrical Engineering and Computer Science, Seoul National University, 599 Gwanak-ro, Gwanak-gu, Seoul 151-744, Republic of Korea School of Integrated Technology, Yonsei University, 162-1, Songdo-dong, Yeonsu-gu, Incheon 406-840, Republic of Korea c LG Electronics, Seoul 151-744, Republic of Korea b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 29 September 2011 Received in revised form 29 December 2011 Accepted 16 February 2012 Available online 7 March 2012
An Al2O3 antireflection layer was placed between a glass substrate and a transparent conducting oxide layer in order to decrease optical reflection in microcrystalline silicon (mc-Si:H) p–i–n solar cells. Optical simulations showed that reflections were decreased by Al2O3 thin films, these reflections were found to be at a minimum when a 40 nm thick Al2O3 layer was used. Experimental results demonstrated that the measured reflectance of mc-Si:H solar cells was decreased by employing the proposed 40 nm Al2O3 in all wavelength regions and the quantum efficiency was also increased. The short-circuit current was increased from 22.7 to 23.5 mA/cm2 without sacrificing open circuit voltage or fill factor. The average efficiency of devices was improved from 6.02% to 6.32% by introducing 40 nm Al2O3 antireflection layer. & 2012 Elsevier B.V. All rights reserved.
Keywords: Al2O3 Antireflection layer Silicon solar cell Thin film
1. Introduction Silicon thin film solar cells have gained a lot of attention due to their low cost and large-area fabrication potential. Recently, multi-junction structures with different band gaps such as micromorph tandem cells consisting of an amorphous (a-Si) top cell and a microcrystalline (mc-Si:H) bottom cell have been produced and showed improved efficiency over traditional solar cells [1,2]. In the field of silicon thin film solar cells, light trapping to confine incident light in the photo-active layers is one of the essential technologies needed to decrease silicon thickness and increase device stability against light induced degradation [3,4]. Surface texturing of transparent conducting oxide (TCO) increases optical path lengths, which enhances light absorption [5]. TCO texturing also gives antireflection (AR) properties by refractive index grading. However, a silicon deposition on the highly roughened TCO surface can cause a degradation of other solar cell characteristics due to an increase of shunted current paths in the cells produced [6–8]. To mitigate this problem, insertion of a resistive n-type mcSiO:H interlayer was suggested [9]. The insertion of an additional antireflection layer (ARL) to decrease reflection loss has been investigated. The proposed method does not imply a rougher interface at the TCO/silicon interface, therefore permitting to increase current without the
n
Corresponding author. Tel.: þ82 32 749 5837; fax: þ 82 32 818 5801. E-mail address: [email protected] (J.-Y. Kwon).
0927-0248/$ - see front matter & 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.solmat.2012.02.020
above mentioned problems. In typical thin film solar cells with a superstrate configuration (p–i–n type), incident light is reflected at interfaces between mediums with different refractive indices, such as air/glass, glass/TCO, and TCO/Si before it reaches the silicon layer. In order to decrease those reflection losses, several groups investigated a TiO2 (n 2.5) ARL to decrease the reflection at the TCO(n 1.9)/Si(n 3.5) interface [3,10,11]. After the ARL was introduced, an efficiency of approximately 11–12% was achieved in micromorph tandem solar cells [3]. However, further increase of the efficiency is needed to enhance the competitiveness of the silicon thin film solar cells. Thus, additional ARL is required at other interfaces such as the glass/TCO interface. To our knowledge this has been scarcely studied. The suppression of reflection at the glass/TCO interface is crucial for effective light management in silicon thin film p–i–n solar cells. The AR film inserted at the glass/TCO interface should have a refractive index between 1.5 and 1.9. As such, candidate materials are Al2O3, CaCO3, MgO, PbF2 etc. For a suitable ARL at the glass/TCO interface, high optical transmittance in the UV–vis– NIR wavelength regions is required. In addition, magnetron sputtering is the preferred deposition method for processing compatibility, as the TCO (i.e. Al-doped ZnO) layer is typically deposited by magnetron sputtering in silicon thin film solar cells. Given these considerations, Al2O3 or MgO are thought to be the strongest candidates. In this letter, we investigate a magnetron sputtered Al2O3 ARL with the purpose of decreasing the reflection at the glass/TCO interface in mc-Si:H p–i–n thin film solar cells.
D.-W. Kang et al. / Solar Energy Materials & Solar Cells 101 (2012) 22–25
23
2. Experiment
3. Results and discussion
Before the fabrication of ARL and mc-Si:H p–i–n thin film solar cells, an optical simulation (ASA, Advanced Semiconductor Analysis) was performed to tentatively investigate the kind of Al2O3 thin film ARL most suitable for this role. Various thicknesses of Al2O3 thin films were considered for use at the glass/TCO interface. After the simulation, sample preparations were carried out to demonstrate experimentally the effects of the inserted ARL. Al2O3 thin films were prepared on corning glass substrates by RF magnetron sputtering. An Al2O3 ceramic target was used and pure Ar sputtering was carried out to deposit the Al2O3 thin films. After the deposition of the Al2O3, Al-doped ZnO (AZO) with a thickness of 1.1 mm was continuously deposited onto the Al2O3 layer by DC magnetron sputtering without a vacuum break. A ceramic AZO (2 wt% Al2O3) target was used to prepare the AZO thin films. The substrate temperature was fixed at 250 1C during the Al2O3 and AZO deposition. For efficient light trapping, the surface texturing of the Al2O3/AZO bi-layer was performed by wet-chemical etching with diluted HCl (0.5%). After, a 30 nm TiO2:Nb (2 wt% Nb2O5) ARL was deposited on the textured AZO at the AZO/Si interface. In order to enhance the Nb-doping effect, the film was annealed in a vacuum at 350 1C [12]. After the annealing, an AZO layer of 10 nm was deposited to protect the TiO2:Nb ARL against hydrogen plasma damage during the Si deposition process [10]. The mcSi:H p–i–n layers were deposited by a multi-chamber RF PECVD (13.56 MHz) system. The final mc-Si:H solar cells have a structure of (glass/Al2O3/AZO/TiO2:Nb/AZO/mc-Si:H p–i–n/AZO/Ag/Al). A cell without the Al2O3 ARL was also fabricated for reference, as shown in Fig. 1. The optical constants of the prepared Al2O3 and AZO thin films were measured by spectroscopic ellipsometry. X-ray diffraction (XRD) was performed to characterize the crystallinity of the Al2O3 and AZO thin films. The electrical resistivities of the AZO and Al2O3/AZO bi-layers were analyzed by Hall measurement with Van der Pauw geometry. The reflectance of the mc-Si:H solar cells was analyzed by a UV–vis spectrophotometer. The external quantum efficiency (EQE) was calculated through the measured spectral response which was measured at 0 V electrical bias over the solar cells.
In order to find whether the Al2O3 thin film is suitable as an ARL, an optical simulation (ASA) was performed. Optical systems with flat interfaces (GENPRO1 mode) were considered in our reflectance simulations. The GENPRO1 is a subroutine that is a part of the ASA optical simulator. It implements calculation of optical properties of the multi-layer system (reflectance, transmittance, and absorptance) with flat interfaces. In this approach, the reflection and transmission at all interfaces and the absorption in all layers of the system are taken into account. In addition, optical constants of the layers are employed in the simulation. Fig. 2 shows the optical reflectance of the mc-Si:H solar cells with various Al2O3 thicknesses. Firstly, fluctuations in reflectance curves were observed due to interference effects between the flat interfaces. The inset graph shows the average reflectance as a function of Al2O3 thickness. The mc-Si:H solar cells with Al2O3 film (20–80 nm) showed lower average reflectance than the cell without the Al2O3 layer. In terms of the finding the optimum Al2O3 thickness, the average reflectance continuously decreased with the increase in Al2O3 thickness from 0 to 40 nm, at Al2O3 thicknesses over 40 nm the average reflectance began to increase from its lowest point. These results imply that the Al2O3 thin films can act as an effective ARL in mc-Si:H solar cells and the film with a thickness of 40 nm exhibits the best AR properties. Based on these simulation results, experiments for mc-Si:H solar cells with an Al2O3 layer at the glass/AZO interface were processed to test the practical effectiveness of the ARL. Fig. 3 shows the optical constants of the deposited Al2O3 and AZO thin films measured by spectroscopic ellipsometry. In the measurement of Al2O3 thin film, a standard Cauchy relationship was used [13]. In addition, the Tauc–Lorentz model has been employed to model the AZO film. Recently, this model has been applied to dielectric function modeling of transparent conductive oxide [14]. The refractive indices of the Al2O3 and AZO films at 550 nm were 1.69 and 1.87, respectively. To minimize reflection loss at the glass/AZO interface, the ARL requires a refractive index of about 1.67, which was calculated by finding the geometric pffiffiffiffiffiffiffiffiffiffi mean of the two surrounding indices (i.e. n1 ¼ n0 n2 where n0 ¼1.5 (glass), n2 ¼1.87 (AZO)). The refractive index of the deposited Al2O3 (1.69) is very close to the optimum value (1.67). The extinction coefficient values were found to be
70
Reflectance (%)
50 40
Average reflectance (%)
60
20
18
Al2O3 (20nm)
17
Al2O3 (40nm)
16
Al2O3 (60nm)
15
Al2O3 (80nm)
14
30
Without Al2O3
19
0
20
40
60
80
Al2O3 thickness (nm)
20 10 0 350
400
450
500
550
600
650
Wavelength (nm) Fig. 1. Structure of the fabricated mc-Si:H p–i–n solar cells (a) without Al2O3 ARL and (b) with Al2O3 ARL. The Al2O3 ARL was inserted at glass/AZO interface to suppress optical reflection of mc-Si:H p–i–n solar cells.
Fig. 2. Simulated optical reflectance of mc-Si:H solar cells as a function of Al2O3 ARL thickness from 20 to 80 nm. The optical system with flat interfaces was considered in the optical simulation. The reflectance was minimized by employing the Al2O3 ARL of 40 nm.
D.-W. Kang et al. / Solar Energy Materials & Solar Cells 101 (2012) 22–25
0.5 AZO (n)
2.0
Al2O3 (n) AZO (k)
0.3
1.8 0.2
1.7 1.6
0.1
1.5 1.4
With Al2O3 ARL (40nm)
18
500
600
700
800
900
16 14 12 10 8
0.0 400
Without Al2O3 ARL
20 0.4
Al2O3 (k) 1.9
22
Cell reflectance (%)
2.1
Extinction coefficient k
Refractive index n
24
6
1000
Wavelength (nm)
negligible in all wavelength regions, implying that incident photon loss caused by absorption in the Al2O3 ARL can be neglected. Before solar cell fabrication, the characteristics of the AZO TCO deposited on the Al2O3 ARL were investigated. The variation in physical properties of the AZO is important because it can affect the performance characteristics of the solar cells, such as fill factor (FF). Above all, the crystallinity of the preceding layer influences the subsequent crystalline growth of the AZO film [15,16]. In other words, the polycrystalline growth of the AZO film can be affected by the structural phase of the Al2O3 preceding layer. The crystalline orientation of the precursor film can either enhance or hinder subsequent AZO crystal growth. XRD measurements show that a diffraction peak was not detected in the Al2O3 deposited on glass, implying that the prepared Al2O3 film consisted entirely of amorphous phase materials. Thus, the c-axis crystal growth of the AZO was not affected by the Al2O3 precursor without any specific crystalline orientation. Hall measurement data indicates that the resistivity of the Al2O3/AZO bi-layer film was 3.6 10 4 O cm, this is the same as that of the AZO film grown on the glass substrate. From the results, it is thought that the insertion of the Al2O3 ARL at glass/AZO interface did not deteriorate the c-axis oriented polycrystalline structure and electrical resistivity of the AZO TCO film. Fig. 4 shows the optical reflectance of the fabricated mc-Si:H p–i–n solar cells, which can evaluate the AR effect of the proposed Al2O3. The mc-Si:H solar cells were deposited on textured AZO substrates. The reflectance of the mc-Si:H cell was decreased in all wavelength regions by employing an Al2O3 (40 nm) ARL. This implies that the inserted Al2O3 film works well as an ARL in mcSi:H solar cells. When the experimental results were compared with the simulated data, it was observed that the measured reflectances were smaller than the simulated ones (Fig. 2). This is a result of the surface texturing of the AZO TCO in the experimental unit, as shown in Fig. 1. Light scattering at rough interfaces decreases optical reflectance and increases optical path lengths, which can enhance light absorption in the photo-active layer [4]. The EQE of the mc-Si:H solar cells, deposited, with and without the Al2O3 ARL, was measured to evaluate whether a decrease in reflection leads to an improvement in light absorption, the results are shown in Fig. 5. The EQE of the reference cell was improved by applying the 40 nm Al2O3 ARL, as a result the shortcircuit current (Jsc) was increased from 22.7 to 23.5 mA/cm2. From
400
600 700 800 Wavelength (nm)
500
900
1000
1100
Fig. 4. Reflectance of the fabricated mc-Si:H solar cells without and with Al2O3 ARL measured by spectrophotometer. The mc-Si:H solar cells were fabricated on textured AZO substrates. The cell reflectance was decreased by applying Al2O3 ARL of 40 nm in all wavelength regions.
100 External quantum efficiency (%)
Fig. 3. The refractive index and extinction coefficient of deposited AZO and Al2O3 thin films measured by spectroscopic ellipsometry. The refractive index of AZO and Al2O3 at 550 nm are 1.87 and 1.69, respectively. The refractive index of Al2O3 (1.69) is very close to 1.67 which is estimated value of minimizing the reflection loss at glass/AZO interface.
Without Al2O3 ARL (Jsc=22.7 mA/cm2)
80
With Al2O3 ARL (40nm) (Jsc=23.5 mA/cm2)
60
40
20
0 300
400
500
600
700
800
900
1000
1100
Wavelength (nm) Fig. 5. EQE of the fabricated mc-Si:H solar cells without and with Al2O3 ARL. The EQE was increased by AR effect of Al2O3 ARL, resulted in the increase of Jsc from 22.7 to 23.5 mA/cm2.
these results, we can infer that the optical reflectance was decreased and light absorption was increased by the Al2O3 ARL at the glass/TCO interface. Fig. 6 shows the results of fabricated mc-Si:H solar cells without and with Al2O3 ARL, obtained by I–V measurement. With the Al2O3 ARL, the efficiency of the cells was increased, which was attributed to increase of Jsc rather than Voc or FF. The Voc and FF of the cells showed similar levels (Voc 0.432–0.44 V, FF 0.596– 0.60) regardless of applying the ARL. In case of the TiO2 ARL inserted at the TCO/Si interface, the FF can be sacrificed to a certain extent due to ohmic loss [11]. However, the Al2O3 ARL at glass/TCO interface allows for increased current without a further texturing of the TCO layer, so that current gain is realized with no impact on cell electrical performance. The proposed mc-Si:H solar cells with the 40 nm Al2O3 ARL were found to have average efficiencies between 6.32%, whereas conventional mc-Si:H solar cells without the Al2O3 ARL showed average efficiencies of about 6.02%. As such, the Al2O3 ARL can be said to increase the efficiency of the device by about 0.3%. In highly efficient cells ( 412%), such as tandem or triple junction silicon thin film solar cells, further increases of efficiency even in the small amount (0.3%) is the most
D.-W. Kang et al. / Solar Energy Materials & Solar Cells 101 (2012) 22–25
thin films inserted at the glass/TCO can be an efficient method for improving the efficiency of silicon p–i–n solar cells.
0.55 Jsc
Voc
25
24.5
24.0 0.45
Jsc (mA/cm2)
Voc (V)
0.50
23.5 0.40
Eff.
23.0 6.9
0.65 6.6
6.3
0.60
6.0
Efficiency (%)
FF
This work was supported by the Global Leading Technology Program (no. 2011T100100039) of the Office of Strategic R&D Planning (OSP) funded by the Ministry of Knowledge Economy, Republic of Korea. References
FF
0.55 5.7 0.50 Without Al2O3 Fig. The and The
Acknowledgments
With Al2O3
Without Al2O3
With Al2O3
6. The results of fabricated mc-Si:H solar cells with and without Al2O3 ARL. average values and error bars were employed to exhibit the results. The Voc FF values were similar levels between the cells with and without Al2O3 ARL. efficiency was improved by the dominant increase of Jsc.
challenging issue. Based on our results of 0.3% efficiency gain in about 6% cell, the efficiency gain of highly efficient cell (4 12%) can be greater than 0.3% by applying the proposed method. From this point of view, we believe that additional AR technology such as this Al2O3 ARL at the glass/TCO interface is essential to attain further improvements of solar cell efficiency.
4. Conclusions In conclusion, the Al2O3 ARL was proposed in order to decrease reflection loss at the glass/TCO interface in silicon thin film solar cells with a p–i–n configuration (superstrate type). The magnetron sputtered Al2O3 thin films have a refractive index of 1.67 which is close to the estimated optimum of 1.69 for minimizing the reflection at the glass (n ¼1.5)/AZO (n ¼1.89) interface. The optical simulation results employing the optical constants measured for the Al2O3 indicated that the reflectance of the mc-Si:H solar cell is minimized with a 40 nm Al2O3 ARL. Experimental results showed that the reflectance of the fabricated mc-Si:H solar cells was decreased and the EQE was improved by the insertion of the 40 nm Al2O3 ARL. In addition, the Jsc was increased from 22.7 to 23.5 mA/cm2, and it showed the effect of increasing the device efficiency about 0.3%. Our experimental results suggest that Al2O3
¨ ¨ [1] T. Soderstr om, F.-J. Haug, X. Niquille, V. Terrazzoni, C. Ballif, Asymmetric intermediate reflector for tandem micromorph thin film silicon solar cells, Applied Physics Letters 94 (2009) 063501-1–063501-3. [2] M. Despeisse, G. Bugnon, C. Battaglia, M. Boccard, M. Charrie re, P. Cuony, ¨ ¨ ¨ ¨ S. Hanni, L. Lofgren, G. Parascandolo, T. Soderstr om, F. Meillaud, C. Ballif, Optimization of thin film silicon solar cells on highly textured substrates, Physica Status Solidi A: Applications and Materials Science 208 (4) (2011) 1863–1868. [3] C. Das, A. Lambertz, J. Huepkes, W. Reetz, F. Finger, A constructive combination of antireflection and intermediate-reflector layers for a-Si/mc-Si thin film solar cells, Applied Physics Letters 92 (2008) 053509-1–053509-3. [4] J. Muller, B. Rech, J. Springer, M. Vanecek, TCO and light trapping in silicon thin film solar cells, Solar Energy 77 (2004) 917–930. [5] R. Dewan, M. Marinkovic, R. Noriega, S. Phadke, A. Salleo, D. Knipp, Light trapping in thin-film silicon solar cells with submicron surface texture, Optics Express 17 (2009) 23058–23065. [6] U. Malm, M. Edoff, Simulating material inhomogeneities and defects in CIGS thin-film solar cells, Progress In Photovoltaics 17 (2009) 306–314. [7] M. Python, E. Vallat-Sauvain, J. Bailat, D. Domine´, L. Fesquet, A. Shah, C. Ballif, Relation between substrate surface morphology and microcrystalline silicon solar cell performance, Journal of Non-Crystalline Solids 354 (2008) 2258–2262. [8] H. Li, R.H. Franken, J.K. Rath, R.E.I. Schropp, Structural defects caused by a rough substrate and their influence on the performance of hydrogenated nano-crystalline silicon n–i–p solar cells, Solar Energy Materials and Solar Cells 93 (2009) 338–349. [9] M. Despeisse, G. Bugnon, A. Feltrin, M. Stueckelberger, P. Cuony, F. Meillaud, A. Billet, C. Ballif, Resistive interlayer for improved performance of thin film silicon solar cells on highly textured substrate, Applied Physics Letters 96 (2010) 073507-1–073507-3. [10] T. Fujibayashi, Takuya Matsui, M. Kondo, Improvement in quantum efficiency of thin film Si solar cells due to the suppression of optical reflectance at transparent conducting oxide/Si interface by TiO2/ZnO antireflection coating, Applied Physics Letters 88 (2006) 183508-1–183508-3. ¨ [11] M. Berginski, C. Das, A. Doumit, J. Hupkes, B. Rech, M. Wuttig, Properties of TiO2 layers as antireflection coating for amorphous silicon based thin-film solar cells, in: Proceedings of the 22nd European Photovoltaic Solar Energy Conference, 3–7 September 2007, Milan, Italy, pp. 2079–2082. [12] T. Hitosugi, A. Ueda, S. Nakao, N. Yamada, Y. Furubayashi, Y. Hirose, T. Shimada, T. Hasegawa, Fabrication of highly conductive Ti1 xNbxO2 polycrystalline films on glass substrates via crystallization of amorphous phase grown by pulsed laser deposition, Applied Physics Letters 90 (2007) 212106-1–212106-3. [13] E. Langereis, S.B.S. Heil, H.C.M. Knoops, W. Keuning, M.C.M. van de Sanden, W.M.M. Kessels, In situ spectroscopic ellipsometry as a versatile tool for studying atomic layer deposition, Journal of Physics D: Applied Physics 42 (2009) 073001-1–073001-19. [14] H. Fujiwara, M. Kondo, Effects of carrier concentration on the dielectric function of ZnO:Ga and In2O3:Sn studied by spectroscopic ellipsometry: analysis of free-carrier and band-edge absorption, Physical Review B 71 (2005) 075109-1–075109-10. [15] D.-W. Kang, S.-H. Kuk, K.-S. Ji, S.-W. Ahn, M.-K. Han, Highly transparent and high haze bilayer Al-doped ZnO thin film employing oxygen-controlled seed layer, Japanese Journal of Applied Physics 49 (2010) 031101-1–031101-5. [16] X.W. Sun, L.D. Wang, H.S. Kwok, Improved ITO thin films with a thin ZnO buffer layer by sputtering, Thin Solid Films 360 (2000) 75–81.
Contents lists available at SciVerse ScienceDirect
Solar Energy Materials & Solar Cells journal homepage: www.elsevier.com/locate/solmat
Letter
Al2O3 antireflection layer between glass and transparent conducting oxide for enhanced light trapping in microcrystalline silicon thin film solar cells Dong-Won Kang a, Jang-Yeon Kwon b,n, Jenny Shim c, Heon-Min Lee c, Min-Koo Han a a
Department of Electrical Engineering and Computer Science, Seoul National University, 599 Gwanak-ro, Gwanak-gu, Seoul 151-744, Republic of Korea School of Integrated Technology, Yonsei University, 162-1, Songdo-dong, Yeonsu-gu, Incheon 406-840, Republic of Korea c LG Electronics, Seoul 151-744, Republic of Korea b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 29 September 2011 Received in revised form 29 December 2011 Accepted 16 February 2012 Available online 7 March 2012
An Al2O3 antireflection layer was placed between a glass substrate and a transparent conducting oxide layer in order to decrease optical reflection in microcrystalline silicon (mc-Si:H) p–i–n solar cells. Optical simulations showed that reflections were decreased by Al2O3 thin films, these reflections were found to be at a minimum when a 40 nm thick Al2O3 layer was used. Experimental results demonstrated that the measured reflectance of mc-Si:H solar cells was decreased by employing the proposed 40 nm Al2O3 in all wavelength regions and the quantum efficiency was also increased. The short-circuit current was increased from 22.7 to 23.5 mA/cm2 without sacrificing open circuit voltage or fill factor. The average efficiency of devices was improved from 6.02% to 6.32% by introducing 40 nm Al2O3 antireflection layer. & 2012 Elsevier B.V. All rights reserved.
Keywords: Al2O3 Antireflection layer Silicon solar cell Thin film
1. Introduction Silicon thin film solar cells have gained a lot of attention due to their low cost and large-area fabrication potential. Recently, multi-junction structures with different band gaps such as micromorph tandem cells consisting of an amorphous (a-Si) top cell and a microcrystalline (mc-Si:H) bottom cell have been produced and showed improved efficiency over traditional solar cells [1,2]. In the field of silicon thin film solar cells, light trapping to confine incident light in the photo-active layers is one of the essential technologies needed to decrease silicon thickness and increase device stability against light induced degradation [3,4]. Surface texturing of transparent conducting oxide (TCO) increases optical path lengths, which enhances light absorption [5]. TCO texturing also gives antireflection (AR) properties by refractive index grading. However, a silicon deposition on the highly roughened TCO surface can cause a degradation of other solar cell characteristics due to an increase of shunted current paths in the cells produced [6–8]. To mitigate this problem, insertion of a resistive n-type mcSiO:H interlayer was suggested [9]. The insertion of an additional antireflection layer (ARL) to decrease reflection loss has been investigated. The proposed method does not imply a rougher interface at the TCO/silicon interface, therefore permitting to increase current without the
n
Corresponding author. Tel.: þ82 32 749 5837; fax: þ 82 32 818 5801. E-mail address: [email protected] (J.-Y. Kwon).
0927-0248/$ - see front matter & 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.solmat.2012.02.020
above mentioned problems. In typical thin film solar cells with a superstrate configuration (p–i–n type), incident light is reflected at interfaces between mediums with different refractive indices, such as air/glass, glass/TCO, and TCO/Si before it reaches the silicon layer. In order to decrease those reflection losses, several groups investigated a TiO2 (n 2.5) ARL to decrease the reflection at the TCO(n 1.9)/Si(n 3.5) interface [3,10,11]. After the ARL was introduced, an efficiency of approximately 11–12% was achieved in micromorph tandem solar cells [3]. However, further increase of the efficiency is needed to enhance the competitiveness of the silicon thin film solar cells. Thus, additional ARL is required at other interfaces such as the glass/TCO interface. To our knowledge this has been scarcely studied. The suppression of reflection at the glass/TCO interface is crucial for effective light management in silicon thin film p–i–n solar cells. The AR film inserted at the glass/TCO interface should have a refractive index between 1.5 and 1.9. As such, candidate materials are Al2O3, CaCO3, MgO, PbF2 etc. For a suitable ARL at the glass/TCO interface, high optical transmittance in the UV–vis– NIR wavelength regions is required. In addition, magnetron sputtering is the preferred deposition method for processing compatibility, as the TCO (i.e. Al-doped ZnO) layer is typically deposited by magnetron sputtering in silicon thin film solar cells. Given these considerations, Al2O3 or MgO are thought to be the strongest candidates. In this letter, we investigate a magnetron sputtered Al2O3 ARL with the purpose of decreasing the reflection at the glass/TCO interface in mc-Si:H p–i–n thin film solar cells.
D.-W. Kang et al. / Solar Energy Materials & Solar Cells 101 (2012) 22–25
23
2. Experiment
3. Results and discussion
Before the fabrication of ARL and mc-Si:H p–i–n thin film solar cells, an optical simulation (ASA, Advanced Semiconductor Analysis) was performed to tentatively investigate the kind of Al2O3 thin film ARL most suitable for this role. Various thicknesses of Al2O3 thin films were considered for use at the glass/TCO interface. After the simulation, sample preparations were carried out to demonstrate experimentally the effects of the inserted ARL. Al2O3 thin films were prepared on corning glass substrates by RF magnetron sputtering. An Al2O3 ceramic target was used and pure Ar sputtering was carried out to deposit the Al2O3 thin films. After the deposition of the Al2O3, Al-doped ZnO (AZO) with a thickness of 1.1 mm was continuously deposited onto the Al2O3 layer by DC magnetron sputtering without a vacuum break. A ceramic AZO (2 wt% Al2O3) target was used to prepare the AZO thin films. The substrate temperature was fixed at 250 1C during the Al2O3 and AZO deposition. For efficient light trapping, the surface texturing of the Al2O3/AZO bi-layer was performed by wet-chemical etching with diluted HCl (0.5%). After, a 30 nm TiO2:Nb (2 wt% Nb2O5) ARL was deposited on the textured AZO at the AZO/Si interface. In order to enhance the Nb-doping effect, the film was annealed in a vacuum at 350 1C [12]. After the annealing, an AZO layer of 10 nm was deposited to protect the TiO2:Nb ARL against hydrogen plasma damage during the Si deposition process [10]. The mcSi:H p–i–n layers were deposited by a multi-chamber RF PECVD (13.56 MHz) system. The final mc-Si:H solar cells have a structure of (glass/Al2O3/AZO/TiO2:Nb/AZO/mc-Si:H p–i–n/AZO/Ag/Al). A cell without the Al2O3 ARL was also fabricated for reference, as shown in Fig. 1. The optical constants of the prepared Al2O3 and AZO thin films were measured by spectroscopic ellipsometry. X-ray diffraction (XRD) was performed to characterize the crystallinity of the Al2O3 and AZO thin films. The electrical resistivities of the AZO and Al2O3/AZO bi-layers were analyzed by Hall measurement with Van der Pauw geometry. The reflectance of the mc-Si:H solar cells was analyzed by a UV–vis spectrophotometer. The external quantum efficiency (EQE) was calculated through the measured spectral response which was measured at 0 V electrical bias over the solar cells.
In order to find whether the Al2O3 thin film is suitable as an ARL, an optical simulation (ASA) was performed. Optical systems with flat interfaces (GENPRO1 mode) were considered in our reflectance simulations. The GENPRO1 is a subroutine that is a part of the ASA optical simulator. It implements calculation of optical properties of the multi-layer system (reflectance, transmittance, and absorptance) with flat interfaces. In this approach, the reflection and transmission at all interfaces and the absorption in all layers of the system are taken into account. In addition, optical constants of the layers are employed in the simulation. Fig. 2 shows the optical reflectance of the mc-Si:H solar cells with various Al2O3 thicknesses. Firstly, fluctuations in reflectance curves were observed due to interference effects between the flat interfaces. The inset graph shows the average reflectance as a function of Al2O3 thickness. The mc-Si:H solar cells with Al2O3 film (20–80 nm) showed lower average reflectance than the cell without the Al2O3 layer. In terms of the finding the optimum Al2O3 thickness, the average reflectance continuously decreased with the increase in Al2O3 thickness from 0 to 40 nm, at Al2O3 thicknesses over 40 nm the average reflectance began to increase from its lowest point. These results imply that the Al2O3 thin films can act as an effective ARL in mc-Si:H solar cells and the film with a thickness of 40 nm exhibits the best AR properties. Based on these simulation results, experiments for mc-Si:H solar cells with an Al2O3 layer at the glass/AZO interface were processed to test the practical effectiveness of the ARL. Fig. 3 shows the optical constants of the deposited Al2O3 and AZO thin films measured by spectroscopic ellipsometry. In the measurement of Al2O3 thin film, a standard Cauchy relationship was used [13]. In addition, the Tauc–Lorentz model has been employed to model the AZO film. Recently, this model has been applied to dielectric function modeling of transparent conductive oxide [14]. The refractive indices of the Al2O3 and AZO films at 550 nm were 1.69 and 1.87, respectively. To minimize reflection loss at the glass/AZO interface, the ARL requires a refractive index of about 1.67, which was calculated by finding the geometric pffiffiffiffiffiffiffiffiffiffi mean of the two surrounding indices (i.e. n1 ¼ n0 n2 where n0 ¼1.5 (glass), n2 ¼1.87 (AZO)). The refractive index of the deposited Al2O3 (1.69) is very close to the optimum value (1.67). The extinction coefficient values were found to be
70
Reflectance (%)
50 40
Average reflectance (%)
60
20
18
Al2O3 (20nm)
17
Al2O3 (40nm)
16
Al2O3 (60nm)
15
Al2O3 (80nm)
14
30
Without Al2O3
19
0
20
40
60
80
Al2O3 thickness (nm)
20 10 0 350
400
450
500
550
600
650
Wavelength (nm) Fig. 1. Structure of the fabricated mc-Si:H p–i–n solar cells (a) without Al2O3 ARL and (b) with Al2O3 ARL. The Al2O3 ARL was inserted at glass/AZO interface to suppress optical reflection of mc-Si:H p–i–n solar cells.
Fig. 2. Simulated optical reflectance of mc-Si:H solar cells as a function of Al2O3 ARL thickness from 20 to 80 nm. The optical system with flat interfaces was considered in the optical simulation. The reflectance was minimized by employing the Al2O3 ARL of 40 nm.
D.-W. Kang et al. / Solar Energy Materials & Solar Cells 101 (2012) 22–25
0.5 AZO (n)
2.0
Al2O3 (n) AZO (k)
0.3
1.8 0.2
1.7 1.6
0.1
1.5 1.4
With Al2O3 ARL (40nm)
18
500
600
700
800
900
16 14 12 10 8
0.0 400
Without Al2O3 ARL
20 0.4
Al2O3 (k) 1.9
22
Cell reflectance (%)
2.1
Extinction coefficient k
Refractive index n
24
6
1000
Wavelength (nm)
negligible in all wavelength regions, implying that incident photon loss caused by absorption in the Al2O3 ARL can be neglected. Before solar cell fabrication, the characteristics of the AZO TCO deposited on the Al2O3 ARL were investigated. The variation in physical properties of the AZO is important because it can affect the performance characteristics of the solar cells, such as fill factor (FF). Above all, the crystallinity of the preceding layer influences the subsequent crystalline growth of the AZO film [15,16]. In other words, the polycrystalline growth of the AZO film can be affected by the structural phase of the Al2O3 preceding layer. The crystalline orientation of the precursor film can either enhance or hinder subsequent AZO crystal growth. XRD measurements show that a diffraction peak was not detected in the Al2O3 deposited on glass, implying that the prepared Al2O3 film consisted entirely of amorphous phase materials. Thus, the c-axis crystal growth of the AZO was not affected by the Al2O3 precursor without any specific crystalline orientation. Hall measurement data indicates that the resistivity of the Al2O3/AZO bi-layer film was 3.6 10 4 O cm, this is the same as that of the AZO film grown on the glass substrate. From the results, it is thought that the insertion of the Al2O3 ARL at glass/AZO interface did not deteriorate the c-axis oriented polycrystalline structure and electrical resistivity of the AZO TCO film. Fig. 4 shows the optical reflectance of the fabricated mc-Si:H p–i–n solar cells, which can evaluate the AR effect of the proposed Al2O3. The mc-Si:H solar cells were deposited on textured AZO substrates. The reflectance of the mc-Si:H cell was decreased in all wavelength regions by employing an Al2O3 (40 nm) ARL. This implies that the inserted Al2O3 film works well as an ARL in mcSi:H solar cells. When the experimental results were compared with the simulated data, it was observed that the measured reflectances were smaller than the simulated ones (Fig. 2). This is a result of the surface texturing of the AZO TCO in the experimental unit, as shown in Fig. 1. Light scattering at rough interfaces decreases optical reflectance and increases optical path lengths, which can enhance light absorption in the photo-active layer [4]. The EQE of the mc-Si:H solar cells, deposited, with and without the Al2O3 ARL, was measured to evaluate whether a decrease in reflection leads to an improvement in light absorption, the results are shown in Fig. 5. The EQE of the reference cell was improved by applying the 40 nm Al2O3 ARL, as a result the shortcircuit current (Jsc) was increased from 22.7 to 23.5 mA/cm2. From
400
600 700 800 Wavelength (nm)
500
900
1000
1100
Fig. 4. Reflectance of the fabricated mc-Si:H solar cells without and with Al2O3 ARL measured by spectrophotometer. The mc-Si:H solar cells were fabricated on textured AZO substrates. The cell reflectance was decreased by applying Al2O3 ARL of 40 nm in all wavelength regions.
100 External quantum efficiency (%)
Fig. 3. The refractive index and extinction coefficient of deposited AZO and Al2O3 thin films measured by spectroscopic ellipsometry. The refractive index of AZO and Al2O3 at 550 nm are 1.87 and 1.69, respectively. The refractive index of Al2O3 (1.69) is very close to 1.67 which is estimated value of minimizing the reflection loss at glass/AZO interface.
Without Al2O3 ARL (Jsc=22.7 mA/cm2)
80
With Al2O3 ARL (40nm) (Jsc=23.5 mA/cm2)
60
40
20
0 300
400
500
600
700
800
900
1000
1100
Wavelength (nm) Fig. 5. EQE of the fabricated mc-Si:H solar cells without and with Al2O3 ARL. The EQE was increased by AR effect of Al2O3 ARL, resulted in the increase of Jsc from 22.7 to 23.5 mA/cm2.
these results, we can infer that the optical reflectance was decreased and light absorption was increased by the Al2O3 ARL at the glass/TCO interface. Fig. 6 shows the results of fabricated mc-Si:H solar cells without and with Al2O3 ARL, obtained by I–V measurement. With the Al2O3 ARL, the efficiency of the cells was increased, which was attributed to increase of Jsc rather than Voc or FF. The Voc and FF of the cells showed similar levels (Voc 0.432–0.44 V, FF 0.596– 0.60) regardless of applying the ARL. In case of the TiO2 ARL inserted at the TCO/Si interface, the FF can be sacrificed to a certain extent due to ohmic loss [11]. However, the Al2O3 ARL at glass/TCO interface allows for increased current without a further texturing of the TCO layer, so that current gain is realized with no impact on cell electrical performance. The proposed mc-Si:H solar cells with the 40 nm Al2O3 ARL were found to have average efficiencies between 6.32%, whereas conventional mc-Si:H solar cells without the Al2O3 ARL showed average efficiencies of about 6.02%. As such, the Al2O3 ARL can be said to increase the efficiency of the device by about 0.3%. In highly efficient cells ( 412%), such as tandem or triple junction silicon thin film solar cells, further increases of efficiency even in the small amount (0.3%) is the most
D.-W. Kang et al. / Solar Energy Materials & Solar Cells 101 (2012) 22–25
thin films inserted at the glass/TCO can be an efficient method for improving the efficiency of silicon p–i–n solar cells.
0.55 Jsc
Voc
25
24.5
24.0 0.45
Jsc (mA/cm2)
Voc (V)
0.50
23.5 0.40
Eff.
23.0 6.9
0.65 6.6
6.3
0.60
6.0
Efficiency (%)
FF
This work was supported by the Global Leading Technology Program (no. 2011T100100039) of the Office of Strategic R&D Planning (OSP) funded by the Ministry of Knowledge Economy, Republic of Korea. References
FF
0.55 5.7 0.50 Without Al2O3 Fig. The and The
Acknowledgments
With Al2O3
Without Al2O3
With Al2O3
6. The results of fabricated mc-Si:H solar cells with and without Al2O3 ARL. average values and error bars were employed to exhibit the results. The Voc FF values were similar levels between the cells with and without Al2O3 ARL. efficiency was improved by the dominant increase of Jsc.
challenging issue. Based on our results of 0.3% efficiency gain in about 6% cell, the efficiency gain of highly efficient cell (4 12%) can be greater than 0.3% by applying the proposed method. From this point of view, we believe that additional AR technology such as this Al2O3 ARL at the glass/TCO interface is essential to attain further improvements of solar cell efficiency.
4. Conclusions In conclusion, the Al2O3 ARL was proposed in order to decrease reflection loss at the glass/TCO interface in silicon thin film solar cells with a p–i–n configuration (superstrate type). The magnetron sputtered Al2O3 thin films have a refractive index of 1.67 which is close to the estimated optimum of 1.69 for minimizing the reflection at the glass (n ¼1.5)/AZO (n ¼1.89) interface. The optical simulation results employing the optical constants measured for the Al2O3 indicated that the reflectance of the mc-Si:H solar cell is minimized with a 40 nm Al2O3 ARL. Experimental results showed that the reflectance of the fabricated mc-Si:H solar cells was decreased and the EQE was improved by the insertion of the 40 nm Al2O3 ARL. In addition, the Jsc was increased from 22.7 to 23.5 mA/cm2, and it showed the effect of increasing the device efficiency about 0.3%. Our experimental results suggest that Al2O3
¨ ¨ [1] T. Soderstr om, F.-J. Haug, X. Niquille, V. Terrazzoni, C. Ballif, Asymmetric intermediate reflector for tandem micromorph thin film silicon solar cells, Applied Physics Letters 94 (2009) 063501-1–063501-3. [2] M. Despeisse, G. Bugnon, C. Battaglia, M. Boccard, M. Charrie re, P. Cuony, ¨ ¨ ¨ ¨ S. Hanni, L. Lofgren, G. Parascandolo, T. Soderstr om, F. Meillaud, C. Ballif, Optimization of thin film silicon solar cells on highly textured substrates, Physica Status Solidi A: Applications and Materials Science 208 (4) (2011) 1863–1868. [3] C. Das, A. Lambertz, J. Huepkes, W. Reetz, F. Finger, A constructive combination of antireflection and intermediate-reflector layers for a-Si/mc-Si thin film solar cells, Applied Physics Letters 92 (2008) 053509-1–053509-3. [4] J. Muller, B. Rech, J. Springer, M. Vanecek, TCO and light trapping in silicon thin film solar cells, Solar Energy 77 (2004) 917–930. [5] R. Dewan, M. Marinkovic, R. Noriega, S. Phadke, A. Salleo, D. Knipp, Light trapping in thin-film silicon solar cells with submicron surface texture, Optics Express 17 (2009) 23058–23065. [6] U. Malm, M. Edoff, Simulating material inhomogeneities and defects in CIGS thin-film solar cells, Progress In Photovoltaics 17 (2009) 306–314. [7] M. Python, E. Vallat-Sauvain, J. Bailat, D. Domine´, L. Fesquet, A. Shah, C. Ballif, Relation between substrate surface morphology and microcrystalline silicon solar cell performance, Journal of Non-Crystalline Solids 354 (2008) 2258–2262. [8] H. Li, R.H. Franken, J.K. Rath, R.E.I. Schropp, Structural defects caused by a rough substrate and their influence on the performance of hydrogenated nano-crystalline silicon n–i–p solar cells, Solar Energy Materials and Solar Cells 93 (2009) 338–349. [9] M. Despeisse, G. Bugnon, A. Feltrin, M. Stueckelberger, P. Cuony, F. Meillaud, A. Billet, C. Ballif, Resistive interlayer for improved performance of thin film silicon solar cells on highly textured substrate, Applied Physics Letters 96 (2010) 073507-1–073507-3. [10] T. Fujibayashi, Takuya Matsui, M. Kondo, Improvement in quantum efficiency of thin film Si solar cells due to the suppression of optical reflectance at transparent conducting oxide/Si interface by TiO2/ZnO antireflection coating, Applied Physics Letters 88 (2006) 183508-1–183508-3. ¨ [11] M. Berginski, C. Das, A. Doumit, J. Hupkes, B. Rech, M. Wuttig, Properties of TiO2 layers as antireflection coating for amorphous silicon based thin-film solar cells, in: Proceedings of the 22nd European Photovoltaic Solar Energy Conference, 3–7 September 2007, Milan, Italy, pp. 2079–2082. [12] T. Hitosugi, A. Ueda, S. Nakao, N. Yamada, Y. Furubayashi, Y. Hirose, T. Shimada, T. Hasegawa, Fabrication of highly conductive Ti1 xNbxO2 polycrystalline films on glass substrates via crystallization of amorphous phase grown by pulsed laser deposition, Applied Physics Letters 90 (2007) 212106-1–212106-3. [13] E. Langereis, S.B.S. Heil, H.C.M. Knoops, W. Keuning, M.C.M. van de Sanden, W.M.M. Kessels, In situ spectroscopic ellipsometry as a versatile tool for studying atomic layer deposition, Journal of Physics D: Applied Physics 42 (2009) 073001-1–073001-19. [14] H. Fujiwara, M. Kondo, Effects of carrier concentration on the dielectric function of ZnO:Ga and In2O3:Sn studied by spectroscopic ellipsometry: analysis of free-carrier and band-edge absorption, Physical Review B 71 (2005) 075109-1–075109-10. [15] D.-W. Kang, S.-H. Kuk, K.-S. Ji, S.-W. Ahn, M.-K. Han, Highly transparent and high haze bilayer Al-doped ZnO thin film employing oxygen-controlled seed layer, Japanese Journal of Applied Physics 49 (2010) 031101-1–031101-5. [16] X.W. Sun, L.D. Wang, H.S. Kwok, Improved ITO thin films with a thin ZnO buffer layer by sputtering, Thin Solid Films 360 (2000) 75–81.