Improved efficiency of multicrystalline silicon solar ...

Materials Science in Semiconductor Processing 16 (2013) 113–117

Contents lists available at SciVerse ScienceDirect

Materials Science in Semiconductor Processing journal homepage: www.elsevier.com/locate/mssp

Improved efficiency of multicrystalline silicon solar cells by TiO2 antireflection coatings derived by APCVD process D. Hocine a,n, M.S. Belkaid a, M. Pasquinelli b, L. Escoubas b, J.J. Simon b, G.A. Rivie re b, A. Moussi c a Laboratory of Advanced Technologies of Electrical Engineering (LATAGE), Faculty of Electrical and Computer Engineering, Mouloud Mammeri University (UMMTO), BP 17 RP, 15000 Tizi-Ouzou, Algeria b Aix-Marseille University, Institut Mate´riaux Microe´lectronique Nanosciences de Provence–IM2NP CNRS UMR 7334, Campus de Saint-Je´rˆ ome, Avenue Escadrille Normandie Niemen–Service 231, F-13397 Marseille Cedex 20, France c Laboratoire des Cellules Photovoltaı¨ques, Unite´ de De´veloppement de la Technologie du Silicium (UDTS), 2 Bd, Frantz Fanon, B.P. 140, 7 Merveilles Alger, Algeria

a r t i c l e i n f o

abstract

Available online 23 June 2012

This paper reports about the adaptation of the chemical vapor deposition (CVD) thin films technology to the fabrication process of multicrystalline silicon solar cells as a simple, low cost and very effective technology for efficiency device improvement by reducing reflection and improving the light-generated current. In this contribution, the higher reflection of a mc-Si solar cell surface is strongly reduced by the deposition of TiO2 antireflection coating (ARC) on the front using the atmospheric pressure chemical vapor deposition method (APCVD). The surface morphology and elemental composition of the TiO2 antireflective layers were revealed using scanning electron microscopy in conjunction with energy dispersive X-ray spectroscopy . The reflectivity was then reduced from 35% to 8.6% leading to the increase of the short circuit current Jsc which was 33.86 mA/cm2 with a benefit of 5.23 mA/cm2 (surface area ¼ 25 cm2) compared to the reference cell (without ARC). This simple and low cost technology induces a 14.26% conversion efficiency which is a gain of þ3% absolute in comparison to the reference cell. The LBIC measurements of a typical multicrystalline cell confirmed the uniformity of the photocurrent distribution throughout the device. These results are encouraging and prove the effectiveness of the APCVD method for efficiency enhancement in silicon solar cells. & 2012 Elsevier Ltd. All rights reserved.

Keywords: Titanium dioxide APCVD Solar cell Antireflection coating Efficiency

1. Introduction There are various of viable renewable energy resources in progress; among them photovoltaics is the most promising as a future clean and sustainable energy technology to replace fossil fuels. The increase in the photovoltaic (PV) market requires the development of high and cost-effective silicon solar cells processing technologies based on the optimization of the ‘‘cost’’ to

n

Corresponding author. Tel.: þ 213 26 21 83 10; fax: þ 213 26 21 01 52. E-mail address: [email protected] (D. Hocine).

1369-8001/$ - see front matter & 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.mssp.2012.06.004

‘‘efficiency’’ ratio for larger scale mass production. Lately, multicrystalline silicon (mc-Si) solar cells predominate the market share and account for 50% of all the PV modules manufactured worldwide [1] because of its lower manufacturing cost at the mass production level [2] combined with relatively high conversion efficiency and high reliability. However, the efficiency of these photovoltaic devices is limited by a number of factors including optical losses caused by the reflection of solar radiations from the cell surface. So, the cost-effective fabrication of an indispensable antireflection coating which should be included at the top surface of the solar cells is still the most important manufacturing step of

114

D. Hocine et al. / Materials Science in Semiconductor Processing 16 (2013) 113–117

silicon solar devices. In this context, the adaptation of the CVD thin films technology to the fabrication process of multicrystalline silicon solar cells is the key issue of modern method of photovoltaic power generation. In fact, due to its cost effectiveness, the CVD processing technology becomes an important issue for the large-scale spread of photovoltaic technology across the energy market. CVD is a simple, low cost and very effective technology for efficiency improvement in silicon solar cells by reducing reflection and improving the lightgenerated current. The APCVD or atmospheric pressure chemical vapor deposition is the most promising variant of CVD methods for antireflection coatings manufacturing steps due to its compatibility with industrial requirements such as high throughput deposition [3], high process speed [3], suitability to large area [4] and mass production. In fact, Wong and Wangh [5] demonstrate the capability of the APCVD system of depositing TiO2 onto 580 solar cells per hour. Usually, the most widely used materials as ARC in silicon solar cells are titanium dioxide (TiO2), tin dioxide (SnO2), zinc oxide (ZnO), silicon nitride (SiNx), and zinc sulphide (ZnS) [6–8]. Among them, TiO2 appears as the best choice for this application in industrial use due to its excellent antireflection properties such as a high refractive index [9,10] and excellent transmittance in the visible and near infrared regions of the solar spectrum [11,12]. Since TiO2 is highly transparent, it acts like a window for the transmission of solar radiations. In addition, it presents a very good chemical resistance and an excellent stability at high temperature [13]. From Marshall and Dimova-Malinovska [14], in the solar cell industry, processing techniques and materials are selected to give the maximum cost reduction, while maintaining a relatively good efficiency. Richards [15] has compared the two main candidates for application as antireflection coating on silicon solar cells in the PV industry which are TiO2 and SiNx; he concluded that TiO2 has excellent optical properties and can be deposited by a low cost process whereas SiNx films are deposited by a highly expensive method and exhibit slightly poorer optical properties. In fact, SiNx films are deposited using the PECVD technique at 450 1C [16]; this technique permits a precise control of thickness of the elaborated films but requires high vacuum to generate the plasma and a sophisticated reactor to contain the plasma. Therefore, PECVD is more expensive than APCVD, which is of easier implementation (it requires only a tubular furnace with gas system) avoiding the need for vacuum equipment [3] or specialized coating chambers [4]. Consequently, the PECVD method is unsuitable for mass production techniques in low cost solar cells as reported by Lien et al. [17] who proposed the sol gel process for the deposition of TiO2 antireflective layers. It is true that the sol gel is an alternative, cheap and simple method for the preparation of antireflective layers for low cost solar cells; however the thickness of the produced films which is an important parameter for antireflection coating is not precisely controlled. In the APCVD process, film thickness is directly controlled by deposition time. So APCVD presents two distinctive advantages: uniformity of

the deposited layers which is a key factor for deposition in large area and optimized control of film thickness. This technique is a high throughput inexpensive TiO2 deposition process [18]. Sheel et al.[3] demonstrated the necessity of the APCVD technique for low cost and optimum performance solar cells. H.M. Yates [19] also reported on the high potential of the APCVD method for the efficiency enhancement in the PV cells combining the reduction of manufacturing costs. The use of such TiO2 layers derived by APCVD in multicrystalline silicon solar cell production depends on the possibility of improving the photovoltaic device performance. In this paper we report the results of a study assessing the impact of the TiO2 antireflection coating derived by the APCVD process on the electrical performance (open circuit voltage, short circuit current, fill factor and efficiency) of multicrystalline silicon n þ /p structure solar cells. For this purpose, these solar cells were characterized by current voltage (I–V) measurements in the dark and under illumination (AM 1.5 solar spectrum, 100 mW/cm2, 25 1C) using a sun simulator. We demonstrated that significant improvements in shortcircuit current and thus efficiency of the cells were possible and the related results were compared to those of reference cells (without antireflective layers). The light beam induced current (LBIC) characterization was also applied to examine the uniformity of the photocurrent distribution throughout the typical mc-Si solar cell. Furthermore, the TiO2 antireflective layer deposited on the front surface was characterized by scanning electron microscopy (SEM), and its reflectance was measured by UV–vis reflectance spectrophotometry.

2. Experimental details 2.1. TiO2 antireflection coating formation In this work, we present a simplified and fast procedure for increasing the efficiency of mc-Si solar cells. For this study, we performed our experiments on mc-Si solar cells without antireflection coatings (as reference cells) originating from UDTS Institute (Algiers, Algeria). The solar cell of 10 10 cm2 was divided into 5 5 cm2 samples using a diamond cutter. Each sample has a bus bar which is perpendicular to the finger electrodes. Those fingers have a width of 161.6 mm and were aligned with intervals of 2.29 mm on each sample. The cells have a conventional n þ /p structure, a thickness of 300 mm and a junction depth of 0.5 mm. A 60 nm thin film of TiO2 as the antireflective layer was deposited on the front surface of these cells from the titanium tetrachloride (TiCl4) precursor using an APCVD equipment at 450 1C. Our APCVD system consisted of three interconnected parts of: – APCVD reactor consisting of a reaction chamber equipped with a substrate holder, and a heating system with temperature control; – chemical vapor precursor supply system; – gas system consisting of oxygen (O2) and nitrogen (N2).


The formation of the TiO2 antireflection coating proceeds as follows: (a) generation of the TiCl4 vapor precursor delivered to the CVD reactor; (b) transport of the TiCl4 vapor into the reaction chamber by oxygen as a carrier gas; (c) the reaction of the TiCl4 vapor with oxygen gas and TiO2 formation by oxidation taking place by the reaction: TiCl4 ðgÞ þ O2 ðgÞ-TiO2 þ2Cl2 ðgÞ

Table 1 Deposition conditions of APCVD for the TiO2 antireflection coatings. APCVD process parameter

Value

O2 flow rate N2 flow rate TiCl4 heating temperature Deposition temperature Deposition time

1 L min 1 0.5 L min 1 68 1C 450 1C 15 min

115

The flow rates of argon and oxygen were regulated by gas flowmeters integrated to the experimental montage. The chlorine gas (Cl2) released during the reaction was suck from the reactor with the extraction system of a sucking fume cupboard. The TiO2 antireflective layers were deposited in the standard conditions presented in Table 1.

2.2. Solar cell characterization We used the solar simulator to get I–V curves of the solar cells with/without antireflective layers. For a more detailed electrical analysis, we used the LBIC technique for mapping the spatial distribution of the photocurrent of the solar cell. Details regarding the LBIC mapping principle and a schematic drawing of the employed experimental LBIC setup are given by Rivie re et al. [20]. Furthermore, we observed the TiO2 surfaces with a scanning electron microscope which is equipped with an energy dispersive X-ray spectrometer. We also performed the optical reflection measurements in the wavelength range 300–1150 nm by means of ’VARIAN Cary 500 Model ultraviolet visible near infrared (UV–vis-NIR) spectrophotometer’ equipped with an integrating sphere.

3. Results and discussion 3.1. SEM EDS analysis

Fig. 1. SEM image of surface morphology of a typical TiO2 thin film deposited by APCVD.

The surface morphology and elemental composition of the TiO2 antireflective layers were revealed using scanning electron microscopy in conjunction with energy dispersive X-ray spectroscopy. We have taken one film of TiO2 from the deposited films as representative for the SEM study. A typical TiO2 film grown at optimum conditions using TiCl4 is shown in Fig. 1. The TiO2 film shows compact, uniform and dense surface morphology with small grains homogeneous in size. Typically, the corresponding EDS spectrum is presented in Fig. 2, which expressly confirms that the layers contain titanium (Ti) and oxygen (O) elements. A similar EDS pattern was obtained by [21]. No chlorine element was detected in the spectrum, it implies that the TiCl4 precursor has been decomposed completely before the deposition.

Fig. 2. Corresponding EDS spectrum of the TiO2 thin film.

116


3.2. Reflectance measurements The mc-Si solar cell suffers from a high natural reflectivity at the surface which reduces the short circuit current and thus the efficiency of the device. Therefore, the reduction of optical losses consisting in reflection of the part of incident light from the front surface of the cell is extremely important for an efficient solar device. The UV–vis reflectance spectra for the TiO2 thin films in the wavelength range from 300 nm to 1150 nm are shown in Fig. 3. The reflection of the bare multicrystalline silicon substrate is about 35%, and the value was reduced to 8.61% as the minimum average reflectance after the deposition of the TiO2 antireflection coating. A similar value of reflectance was obtained by Richards [15] for TiO2 as the antireflection coating.

reference cell (without ARC). This consequently gives a maximum cell efficiency of 14.26% which is a gain of þ3% absolute in comparison to the reference cell. Thus, for a given reference photovoltaic cell with an efficiency of Z%, the APCVD deposited TiO2 antireflection coating allows an enhancement of this efficiency by DZ ¼3%. According to these results, the improvement in conversion efficiency of solar cells is mainly attributed to the increase in the short circuit current due to the reduction of reflectivity from the front surface and the enhancement of light transmission by the TiO2 antireflective layer. From [17], an increase in efficiency by DZ ¼3.13% was achieved on their silicon solar cells due to the deposition of a single-layer TiO2 ARC using the solgel method. The industrial solar cells fabricated from multicrystalline silicon have conversion efficiency around 14% [22]. A similar photovoltaic parameter result were obtained by [23] for multicrystalline silicon solar cells.

3.3. Solar cell characterization results 3.4. LBIC characterization The light I–V characteristics of mc-Si solar cells with/ without antireflection coatings were measured; the typical experimental results are summarized in Table 2, and the resulting curves are shown in Fig. 4 for optimum conditions. The most important electrical parameter for the solar cell industry is efficiency Z because it incorporates the open circuit voltage (Voc), short circuit current (Jsc) and fill factor (FF) as given by

Z ¼ V oc Isc FF=Pin

ð1Þ

where Pin is the incident light intensity. From the data given in Table 2, the solar cell with antireflection coating has shown the best photovoltaic performance, in particular, a short circuit current Jsc of 33.86 mA/ cm2 which increases by DJsc ¼5.23 mA/cm2 compared to the

50

Fig. 5 shows a typical LBIC two-dimensional mapping of the distribution of photocurrent across the solar cell. The LBIC measurements of a typical multicrystalline cell confirm the good homogeneity of the local photocurrent across the device. In addition, the almost uniform photocurrent shown in this figure proves that the effect of recombining defects inside the poly-Si grains and grain boundaries on the current collection is negligibly small. 4. Conclusion The adaptation of the APCVD technique to the fabrication process of the multicrystalline silicon solar cells permits further reduction of manufacturing cost (comparatively to

TiO2 on Si

Reflectance (%)

40 30 20 10 0 400

600 800 Wavelength (nm)

1000

Fig. 3. UV–vis reflectance spectra of the APCVD deposited TiO2 thin film.

Fig. 4. Light I V characteristics of the typical mc-Si solar cells with and without TiO2 antireflection coating (ARC).

Table 2 Evolution of the experimental electrical parameters for mc-Si solar cells with and without antireflection coating. Cell

Jsc (mA/cm2)

Voc (mV)

FF

Z (%)

mc-Si solar cell without ARC(reference cell) mc-Si solar cell with TiO2 ARC

28.63 33.86

561 585

0.70 0.72

11.24 14.26


117

Algerian Ministry of Higher Education and Scientific Research and French Egide. References

Fig. 5. A typical LBIC two-dimensional mapping of the distribution of photocurrent across the solar cell.

the PECVD process), while maintaining its relatively good efficiency. The use of the APCVD technique for the formation of the TiO2 antireflective layer on the front surface of mc-Si solar cell can enhance its power conversion efficiency by 3% in absolute value (surface area¼25 cm2). The improvement in efficiency of mc-Si solar cell from 11.24% to 14.26% is mainly attributed to the increase in short circuit current density from 28.63 mA/cm2 to 33.86 mA/cm2 with a benefit of DJsc ¼5.23 mA/cm2 due to the reduction of reflectivity from the front surface and the enhancement of light transmission by the TiO2 antireflection layer. The LBIC characterization confirmed the uniformity of the photocurrent distribution throughout the device. The results confirm the suitability of the CVD technology for efficiency improvement in mc-Si silicon solar cells which is compatible with industrial requirements. Thanks to cost reduction and improved efficiency, the CVD technology offers a way to increase the competitiveness of mc-Si solar cells and has a great potential for large commercial scale production.

Acknowledgments The authors gratefully acknowledge the financial support of the CMEP–PHC–Tassili Project ‘‘09 mdu 775’’ of the

[1] Yuang-Tung Cheng, Jyh-Jier Ho, Song-Yeu Tsai, Solar Energy 85 (2011) 87–94. [2] A. Poullikkas, International Energy and Environment Foundation (2010) 617–634. [3] D .W. Sheel, H.M. Yates, P. Evans, Thin Solid Films 517 (2009) 3061–3065. [4] D.W. Sheel, J.M. Gaskell, Thin Solid Films (2011), doi:10.1016/j.tsf. 2011.04.206. [5] D.C. Wong, A. Waugh, Materials Research Symposium Proceedings 426 (1996) 503–511. [6] D.S. Ginley, C. Bright, MRS Bulletin 25 (2000) 15. [7] C. Kilic, A. Zunger, Physical Review Letters 88 (2002). [8] H.P. Deshmukh, P.S. Shinde, P.S. Patil, Materials Science and Engineering B 130 (2006) 220–227. [9] M. Laube, F. Rauch, C. Ottermann, O. Anderson, K. Bange, Nuclear Instruments and Methods in Physics Research B 113 (1996) 288–292. [10] B.S. Richards, S.F. Rowlands, A. Ueranatasun, Solar Energy 76 (2004) 269–276. [11] P.S. Shinde, S.B. Sadale, Solar Energy Materials & Solar Cells 92 (2008) 283–290. [12] R. Zallen, M.P. Moret, Solid State Communications 137 (2006) 154–157. [13] B.S Richards, J.E. Cotter, C.B. Honsberg, S.R. Wenham, Procedings of the 28th IEEE Photovoltaic Specialists Conference, NJ, 2000,IEEE, pp. 375–378,. [14] J.M. Marshall and D. Dimova-Malinovska, Photovoltaic and Photoactive Materials–Properties, Technology and Applications, vol. 80, p. 117. London, 2002. ISBN 1-4020-0823-6. [15] B.S. Richards, Ph.D. Thesis, University of New South Wales, Australia, 2002. [16] AG. Aberle, Solar Energy Materials & Solar Cells 65 (2001) 239–248. [17] Shui-Yang Lien, Dong-Sing Wuu, Wen-Chang Yeh, Jun-Chin Liu, Solar Energy Materials & Solar Cells 90 (2006) 2710–2719. [18] Andrew F. Thomson, Keith R. McIntosh, Bryce S. Richards, Proceedings of the 22nd European Photovoltaic Solar Energy Conference, 3 7 September 2007, Milan, Italy. [19] H.M. Yates, P. Evans, D.W. Sheel, Z. Remes, M. Vanecek, Thin Solid Films 519 (2010) 1334–1340. [20] Guillaume A. Rivie re, Jean-Jacques Simon, Ludovic Escoubas, Wilfried Vervisch, Marcel Pasquinelli, Solar Energy Materials and Solar Cells (2012), doi:10.1016/j.solmat.2012.01.030. [21] Zhen-Hong Hua-Yan Si, Xu Sun, Wei-Wei Kang, Hao-Li Zi, Zhang, Microporous and Mesoporous Materials 119 (2009) 75–81. [22] Erwann Fourmond, Gilles Dennler, Re´mi Monna, Solar Energy Materials & Solar Cells 65 (2001) 297–301. [23] W. Neu, A. Kress, W. Jooss, P. Fath, E. Bucher, Solar Energy Materials & Solar Cells 74 (2002) 139–146.