- Email: [email protected]
Copper Light-induced Plating Contacts with Ni Seed-layer for mc-Si Solar Cells after Laser Ablation of SiNx:H
Available online at www.sciencedirect.com
ScienceDirect Energy Procedia 38 (2013) 713 – 719
SiliconPV: March 25-27, 2013, Hamelin, Germany
Copper light-induced plating contacts with Ni seed-layer for mc-Si solar cells after laser ablation of SiNx:H Alexandru Focsaa,*, Daniele Blanca, Gilles Poulaina, Corina Barbosa, Michel Gauthierb, Nam Le Quangb, Mustapha Lemitia a
Université de Lyon, Institut des Nanotechnologies de Lyon INL-UMR5270, INSA de Lyon, CNRS, Villeurbanne, F-69621, France b EDF ENR PWT (Photowatt), 33 Rue Saint Honoré, 38300 Bourgoin-Jallieu France
Abstract This study focuses on the electrochemical Ni/Cu metallization of multi-crystalline silicon (mc-Si) without alignment steps. Selective ablation of the silicon nitride (SiNx:H) anti-reflection and passivation coating was performed prior to metallization with a frequency tripled Nd:YAG laser. Electroless nickel-phosphorous layers of different thicknesses were deposited as a seed-layer at 95°C on two batches of samples before electrolytic copper thickening. The thickening of the Ni contact by copper was done by light-induced plating (LIP). The influence of laser ablation parameters as well as chemical etching prior to metal deposition was investigated. The morphology of electrochemically deposited Ni and Cu metal layers was investigated by SEM and optical microscope. Laboratory scale solar cells were fabricated to evaluate the electrical properties of the front contacts. A copper thickness of between 9 and 16 μm was necessary in order to optimize the fill factor. The best efficiency measured on 200μm thick p-type mc-Si solar cell with an area of 4.4 cm2 was 15.5%. An average efficiency of 15% over 18 samples has been demonstrated. Such results were obtained without any additional thermal annealing treatment of the Ni seed-layer. The limiting factors as well as possible improvements are discussed.
© 2013 The Authors. Published by Elsevier Ltd. © 2013 Theand/or Authors. Published byunder Elsevier Ltd. Selection peer-review responsibility of the scientific committee of the SiliconPV 2013 Selection and/or peer-review under responsibility of the scientific committee of the SiliconPV 2013 conference conference Keywords: Metallization; Nickel; Copper; Contact; multicrystalline; Silicon; Solar Cells
1. Introduction Advanced front-side architectures can provide significant increases in solar cell efficiency, but most existing approaches need many processing steps and alignment procedures. Laser processing provides a
*
Corresponding author. Tel.: +33 (0)4-72-43-87-38; fax: +(33) (0)4-72-43-85-31. E-mail address: [email protected]
1876-6102 © 2013 The Authors. Published by Elsevier Ltd.
Selection and/or peer-review under responsibility of the scientific committee of the SiliconPV 2013 conference doi:10.1016/j.egypro.2013.07.337
714
Alexandru Focsa et al. / Energy Procedia 38 (2013) 713 – 719
good way to achieve such structures with a minimum number of technological steps [1] and without alignment when coupled with front-side electrochemical metallization. Introduction of advanced metallization is one of the key steps to fabricate solar cells with high performance in a cost-effective way. More than 85% of solar cell manufacturers use thick film screen printing for front side contact formation, but a lot of research has been carried out into alternative metallization schemes [2]. Front-side metallization by electrochemical processes, as suggested by Dubé et al. [3], consisting of selective laser ablation of antireflection layer and plated contacts, can reduce shadowing because of the potential improved aspect ratio and smaller finger width [4,5]. The IMEC team has reported a very good conversion efficiency of 19.6% using Cu-plated contacts instead of silver [6]. In the present paper, the front-side metallization was produced by Ni-P electroless deposition and Cu light-induced plating (Fig.1) on laboratory scale multi-crystalline silicon solar cells. The influence of laser ablation parameters as well as chemical etching prior to metal deposition was investigated. SEM observations and electrical measurements were performed to characterize the solar cells. 2. Experimental The solar cells were made from multi-crystalline p-type silicon wafers (200 μm thick, p type, ~ 1-3 d roughness around 3-5 μm. The emitter was created by thermal phosphorus diffusion at lowwas fully screen-printed with an Al paste and fired to form the back surface field and rear contact. A frequency tripled Nd:YAG laser with a wavelength of 355 nm and a pulse length of 10 ns was used for selective ablation of the SiNx:H anti-reflection coating (Fig.1a). Pulse fluence (energy density) and spot overlap were optimized to limit the surface damage and the debris formation. Laboratory scale solar cells of 2 x 2 cm2 area with laser fluence between 0.2 and 1.1 J/cm2 and a TLM grid were fabricated (Fig. 2b).
(a)
(b)
Fig. 1. A schematic view of sample geometry (a) and flow chart of solar cells fabrication with electrochemical metallization (b).
715
Alexandru Focsa et al. / Energy Procedia 38 (2013) 713 – 719
Fluence values of the order of 0.40 0.57 J/cm2 were found to be a good compromise between efficient ablation and limited surface damage or debris formation. The grid finger width of cells was 80 μm and the distance between the metal fingers was 1.8 mm (Fig. 2a). The coverage of the cell surface by the front metal contacts was 6.1%. Metallization of the front grid contacts was done by electrochemical processes including Ni electroless seed-layer deposition and Cu thickening by Light-Induced Plating (LIP). Two batches of samples were used: one deoxidized with HF to remove potential debris after ablation and native oxide and the other without HF treatment before the Ni-P plating. 3. Results and discussion 3.1 Ni-P electroless deposition Ni-P electroless deposition is an autocatalytic process involving the immersion of a substrate in a plating bath. Its principle has been described elsewhere [7]. Ni-P was used as a seed-layer before electrolytic copper thickening. Before starting the Ni-P deposition the edges and Al backside screenprinted contacts were protected with an insulation varnish, to avoid Al dissociation and Ni parasitic deposition during the electroless process. The composition of the electroless Ni-P plating bath and the operating conditions are given in [8]. The deposition rate of the Ni-P layer on silicon was 18 μm/h. The thickness of the Ni-P layer was around 150 nm for a process duration of 30 sec and ~ 300nm for 60sec at the operating temperature of 95°C. The wafers with electroless Ni-P seed layers were not subjected to additional thermal annealing. No significant difference was observed between the properties of cells with the two different values of Ni layer thickness.
a)
b)
Fig. 2. a) 86μm wide finger and 380 μm wide busbar after Cu plating; b) Laboratory scale solar cells of 2 x 2 cm2 area with Ni/Cu metallization after laser ablation with different fluence between 0.2 up to 1 J/cm 2.
3.2 Copper thickening The front contact thickening by copper was performed using LIP (light-induced plating) [9-11]. This method uses the photovoltaic effect of the cell to induce a voltage in the p-n junction under illumination without external contacts on the front-side grid: the photo-generated current therefore acts as a plating
716
Alexandru Focsa et al. / Energy Procedia 38 (2013) 713 – 719
current. By choosing the correct potential difference between the anode and the solar cell back-side, the plating of the contacts can be performed without protecting the back-side contacts. In the present case, the solar cell was dipped in an illuminated electroplating copper solu which was kept at room temperature during the process. During the LIP process, the current density was kept constant (~ 1.0 A/dm2). The external applied voltage was kept close to the photovoltage generated by the cell to avoid Cu deposition on the back side of the cell. The Cu deposition time was around 30 min. Fig.3a shows an SEM image of an Ni-P electroless seed layer of ~300nm thickness deposited from an alkaline (pH=8-10) bath on a laser-ablated SiNx:H and Cu layer of ~5 μm thickness plated from an acidic (pH=2) LIP-bath. Fig. 3b shows SEM pictures of a metallic busbar after copper LIP coating of 15μm and illustrates the smooth Cu surface. This demonstrates that a Nielectroless seed layer allows uniform deposition of the Cu layer by LIP.
Cu
SiNx
SiNx Cu
Ni
Ni
a)
SiNx
b)
Fig. 3. Scanning electron microscope image of electroless Ni-P layer deposited on laser-ablated SiNx:H and Cu line performed by LIP (a) and low magnification SEM image of a busbar after Cu plating (b).
3.3 Electrical characterization of solar cells Fig. 4 shows the measured thickness of plated Cu after LIP and the corresponding open-circuit voltage values for 8 cells as a function of the laser fluences used for the selective ablation of SiNx:H. All samples (Fig. 2b) were subjected to the same LIP process for 35 min using a generated LIP current of 1.2 A/dm2. The maximum VOC value of 616 mV corresponds to a laser fluence of 0.44 J/cm2 and an average Cu thickness of 16 μm. For high laser fluence values ( > 0.7 J/cm2) the Voc of the cell is reduced due to damage at the silicon surface and short-circuiting in the emitter (Fig. 4 a). The fill-factor (FF) of the cells also decreases for high laser fluence (Fig. 4 b). The maximum value of FF, 0.75, was obtained for a laser fluence of 0.44 J/cm2 and 16 μm Cu-thickness.
717
Alexandru Focsa et al. / Energy Procedia 38 (2013) 713 – 719
10
0,590
5
Cu 11-thickening Voc -> (Cu11)
0,580 0,570
0
0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9
1
Laser Fluence (J/cm2)
Copper thickening (μm)
0,600 Voc (V)
Copper thickening (μm)
0,610
15
0,80
20
0,75
15
0,70
0,65
10 5
Cu 11-thickening
0,60
FF (Cu11)
0,55
0
FF (Series Cu11)
0,620
20
0,50 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 Laser Fluence (J/cm2)
a)
1
b)
Fig. 4. Evolution of Cu thickness and I-V parameters of mc-Si solar cells with Ni/Cu metallization (for series Cu11 of cells) as a function of the laser fluence: a) Voc and b) FF.
We studied the influence of deoxidizing of the ablated regions in 2.5% HF before dipping in the Ni-P electroless plating bath. As observed previously, increasing the laser fluence for ablation of SiNx:H decreases the open-circuit voltage of the cells (Fig. 5 a) while the short-circuit current stays roughly constant (Fig. 5 b). The low laser fluences below 0.36 J/cm2 correspond to the ablation threshold of SiNx:H where laser fluctuations cause a large variation of I-V parameters. There are no obvious differences in the values of Voc and Jsc for the two batches of cells. 35
0,610
34
Jsc (mA/cm2)
0,620
Voc (V)
0,600 0,590 0,580
Voc (with HF) Voc (without HF) Poly. (Voc (with HF)) Poly. (Voc (without HF))
0,570 0,3
0,4
0,5
0,6
0,7
0,8
Laser Fluence (J/cm2) a)
32
Jsc (with HF) Jsc (whitout HF) Poly. (Jsc (with HF)) Poly. (Jsc (whitout HF))
31 30
0,560 0,2
33
0,9
1
1,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
0,9
1
1,1
Laser Fluence (J/cm2) b)
Fig. 5. I-V parameters of mc-Si solar cells as a function of laser fluence for series of cells with deoxidation in 2.5% HF and without treatment in HF: a) Voc and b) Jsc. Two Ni thicknesses (Ni~150nm & Ni~300nm) were used.
Table 1 summarizes the photovoltaic parameters of the mc-Si cells produced with optimized laser fluence values between 0.42 and 0.57 J/cm2. The table below gives the open-circuit voltage, short-circuit current and conversion efficiency of tested solar cells considering the total area of 4.4 cm2 for 150nm thick Ni seed-layer and Cu thickening of 9-16 μm. The best cell obtained with a laser fluence of 0.52 J/cm2 exhibits an efficiency of 15.5%.
718
Alexandru Focsa et al. / Energy Procedia 38 (2013) 713 – 719
Table 1. Illuminated I-V parameters of mc-Si cells for the range of laser fluence of 0.42 - 0.57 J/cm2. Efficiency
FF
Voc
Jsc
p.Eff.
p.FF
(%)
(%)
(mV)
(mA/cm2)
(%)
(%)
Average over 18 samples
15.0
73.4
609.4
33.7
15.6
75.8
0.52
340
Best cell
15.5
74.5
615.1
34.0
16.1
77.7
0.56
470
Cells
Rs
Rsh 2
2
0,80
20
0,70
18
0,60
Eff.Cells (%) FF . Poly. (Eff.Cells (%)) Poly. y ((FF .))
16 14 12
0,50 0,40 0,30
10
0,20 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9
1
1,1 1,2
Laser Fluence (J/cm2)
Fig. 6. Evolution of Efficiency and Fill-Factor of mc-Si solar cells with Ni/Cu metallization as ffunction of the laser fluence.
100
FF
22
IQE & Reflectivity (%)
Efficiency (%)
Fig. 6 shows the evolution of fill factor and efficiency of mc-Si solar cells as a function of the laser ablation fluence. The green zone indicates the optimum laser fluence used for the ablation of the SiNx:H layer. The fill factor starts to decrease at a laser ablation fluence of ~ 0.67 J/cm2 due to damage to the emitter surface. As previously noted, below the fluence threshold of 0.36 J/cm2 the ablation of SiNx:H is inconsistent. This gives rise to a highly variable cell FF and efficiency. Figure 7 shows the internal quantum efficiency (IQE) for one of the best cells with Ni/Cu front-side metallization and a laser fluence of 0.48 J/cm2. The calculated value of Jsc from EQE spectra is 34.07 mA/cm2. This is very close to the value of 33.8 mA/cm2 obtained from I-V measurements for the same cell with a ratio of 0.992. The difference of 0.27 mA/cm2 can be explained by the non-optimized process of Ni/Cu plating: parasitic metal micro-particles can be seen on the surface of the SiNx:H layer due to local damage to the coating and this can increase the shadow on the cells. These particles have been observed to contain Cu but are assumed to also contain Ni. The solar-weighted reflectance spectrum from SiNx:H on mc-Si has an average photonic weighted reflectivity of 12.5 13.8% and does not take into account the reflectance from the front grid of the samples.
80 60
Refl._SiN_Cell Cu2 IQE_Cell Cu2
40 20 0 350
450
550
650
750
850
950
1050
Wavelength, nm
Fig. 7. Internal quantum efficiency of Ni/Cu front side metallization mc-Si solar cell and reflectance spectrum.
4. Conclusions The front side metallization of mc-Si solar cells was carried out by electrochemical deposition of Ni/Cu. Laser fluence values of the order of 0.42 0.57 J/cm2, used to locally ablate the SiNx:H antireflection coating, were found to be a good compromise between efficient ablation and limited surface damage or debris formation. Electrical characterization shows reasonably good average cell results with an average efficiency of 15.0 % over 18 laboratory scale solar cells. The best cell gives an efficiency of
Alexandru Focsa et al. / Energy Procedia 38 (2013) 713 – 719
15.5%. Such results were obtained without any additional thermal annealing treatment of the Ni seedlayer. Future work will be focused on the optimization of the process, decreasing the contact finger grid to 40 μm and thermal treatments. Further optimization of the series and shunt resistances are expected when using monocrystalline Si wafers to guarantee the reproducibility of the substrate properties. Acknowledgement The authors thank Stefan Hess from PROLECTRO for providing the copper electrolyte solution used in this investigation and for the fruitful discussion. We gratefully thank K. Fraser from INL for his valuable contribution. This work was supported by the French National Research Agency (ANR) under PROTERRA project. References [1] A. Ogane, K. Hirata, K. Horiuchi, Y. Nishihara, Y. Takahashi, A. Kitiyanan, et al., Laser-doping technique using ultraviolet laser for shallow doping in crystalline silicon solar cell fabrication, Jpn. J. Appl. Phys. 48 (2009) 071201. [2] Guy Beaucarne, Jaap Hoornstra, Gunnar Schubert, Lessons From the 2nd Workshop on Metallization of Crystalline Silicon Solar Cells, Future Photovoltaics , 2010, www.futurepv.com [3] C.E. Dubé, R.C. Gonsiorawski, Improved contact metallization for high efficiency EFG polycrystalline silicon solar cells. Proc. 21th IEEE Photovoltaic Specialists Conference, Florida, 1990, v1, pp. 624-628. [4] A. Knorz. M. Aleman, A. Grohe, R. Preu, S. Glunz, Laser ablation of antireflection coatings for plated contacts yielding solar cell efficiency above 20%, Proc. 24th EU PVSEC, Hamburg, 2009, pp. 1002-1005. [5] J. Bartsch, V. Radke, S Savio, S.W. Glunz, Progress in understanding the current path and deposition mechanisms of lightinduced plating and implications for the process, Proc. of 24th EU PVSEC, Hamburg, 2009, pp. 1469 1474. [6] L. Tous, R. Russel, J. Das, R. Labie, M. Hgamo, J. Horzel et all, Large area copper plated silicon solar cell exceeding 19.5% efficiency, 3rd Workshop on Metallization for Crystalline Silicon Solar cells, Energy Procedia 21 (2012) 58 65. [7] C. Boulord, A. Kaminski, G. Stremsdoerfer, S. Stremsdoerfer, B. Canut, J. F. Chateaux, M. Lemiti, Analysys of wet chemical nickel films for silicon solar cells metallization, 24th EU PVSEC, Hamburg, 2009, pp. 1430-1433. [8] C. Boulord, PhD Thèse, Développement de techniques de métallisation innovantes pour cellules photovoltaïques à haut rendement, 2011, 159p. http://tel.archives-ouvertes.fr/docs/00/67/98/76/PDF/these.pdf [9] A.Mette, C.Schetter, D.Wissen, S.Lust, S.W.Glunz, G.Willeke, Increasing the efficiency of screen-printed silicon solar cells by light-induced silver plating, in: Proceedings of the Fourth World Conference on Photovoltaic Energy Conversion, 2006, pp.1056 1059. [10] M. Aleman, N. Bay, D. Barusha, S.W. Glunz, R. Preu, Front-side metallization of silicon solar cells by nickel plating and light induced silver plating, Galvanotechnik, Photovoltaik, 2, 2009, pp. 412-417. [11] J. Bartsch, V. Radke, C. Schetter, S.W. Glunz, Electrochemical methods to analyse the light-induced plating process, J. Appl. Electrochem. (2010), 40, pp. 757-765.
719
ScienceDirect Energy Procedia 38 (2013) 713 – 719
SiliconPV: March 25-27, 2013, Hamelin, Germany
Copper light-induced plating contacts with Ni seed-layer for mc-Si solar cells after laser ablation of SiNx:H Alexandru Focsaa,*, Daniele Blanca, Gilles Poulaina, Corina Barbosa, Michel Gauthierb, Nam Le Quangb, Mustapha Lemitia a
Université de Lyon, Institut des Nanotechnologies de Lyon INL-UMR5270, INSA de Lyon, CNRS, Villeurbanne, F-69621, France b EDF ENR PWT (Photowatt), 33 Rue Saint Honoré, 38300 Bourgoin-Jallieu France
Abstract This study focuses on the electrochemical Ni/Cu metallization of multi-crystalline silicon (mc-Si) without alignment steps. Selective ablation of the silicon nitride (SiNx:H) anti-reflection and passivation coating was performed prior to metallization with a frequency tripled Nd:YAG laser. Electroless nickel-phosphorous layers of different thicknesses were deposited as a seed-layer at 95°C on two batches of samples before electrolytic copper thickening. The thickening of the Ni contact by copper was done by light-induced plating (LIP). The influence of laser ablation parameters as well as chemical etching prior to metal deposition was investigated. The morphology of electrochemically deposited Ni and Cu metal layers was investigated by SEM and optical microscope. Laboratory scale solar cells were fabricated to evaluate the electrical properties of the front contacts. A copper thickness of between 9 and 16 μm was necessary in order to optimize the fill factor. The best efficiency measured on 200μm thick p-type mc-Si solar cell with an area of 4.4 cm2 was 15.5%. An average efficiency of 15% over 18 samples has been demonstrated. Such results were obtained without any additional thermal annealing treatment of the Ni seed-layer. The limiting factors as well as possible improvements are discussed.
© 2013 The Authors. Published by Elsevier Ltd. © 2013 Theand/or Authors. Published byunder Elsevier Ltd. Selection peer-review responsibility of the scientific committee of the SiliconPV 2013 Selection and/or peer-review under responsibility of the scientific committee of the SiliconPV 2013 conference conference Keywords: Metallization; Nickel; Copper; Contact; multicrystalline; Silicon; Solar Cells
1. Introduction Advanced front-side architectures can provide significant increases in solar cell efficiency, but most existing approaches need many processing steps and alignment procedures. Laser processing provides a
*
Corresponding author. Tel.: +33 (0)4-72-43-87-38; fax: +(33) (0)4-72-43-85-31. E-mail address: [email protected]
1876-6102 © 2013 The Authors. Published by Elsevier Ltd.
Selection and/or peer-review under responsibility of the scientific committee of the SiliconPV 2013 conference doi:10.1016/j.egypro.2013.07.337
714
Alexandru Focsa et al. / Energy Procedia 38 (2013) 713 – 719
good way to achieve such structures with a minimum number of technological steps [1] and without alignment when coupled with front-side electrochemical metallization. Introduction of advanced metallization is one of the key steps to fabricate solar cells with high performance in a cost-effective way. More than 85% of solar cell manufacturers use thick film screen printing for front side contact formation, but a lot of research has been carried out into alternative metallization schemes [2]. Front-side metallization by electrochemical processes, as suggested by Dubé et al. [3], consisting of selective laser ablation of antireflection layer and plated contacts, can reduce shadowing because of the potential improved aspect ratio and smaller finger width [4,5]. The IMEC team has reported a very good conversion efficiency of 19.6% using Cu-plated contacts instead of silver [6]. In the present paper, the front-side metallization was produced by Ni-P electroless deposition and Cu light-induced plating (Fig.1) on laboratory scale multi-crystalline silicon solar cells. The influence of laser ablation parameters as well as chemical etching prior to metal deposition was investigated. SEM observations and electrical measurements were performed to characterize the solar cells. 2. Experimental The solar cells were made from multi-crystalline p-type silicon wafers (200 μm thick, p type, ~ 1-3 d roughness around 3-5 μm. The emitter was created by thermal phosphorus diffusion at lowwas fully screen-printed with an Al paste and fired to form the back surface field and rear contact. A frequency tripled Nd:YAG laser with a wavelength of 355 nm and a pulse length of 10 ns was used for selective ablation of the SiNx:H anti-reflection coating (Fig.1a). Pulse fluence (energy density) and spot overlap were optimized to limit the surface damage and the debris formation. Laboratory scale solar cells of 2 x 2 cm2 area with laser fluence between 0.2 and 1.1 J/cm2 and a TLM grid were fabricated (Fig. 2b).
(a)
(b)
Fig. 1. A schematic view of sample geometry (a) and flow chart of solar cells fabrication with electrochemical metallization (b).
715
Alexandru Focsa et al. / Energy Procedia 38 (2013) 713 – 719
Fluence values of the order of 0.40 0.57 J/cm2 were found to be a good compromise between efficient ablation and limited surface damage or debris formation. The grid finger width of cells was 80 μm and the distance between the metal fingers was 1.8 mm (Fig. 2a). The coverage of the cell surface by the front metal contacts was 6.1%. Metallization of the front grid contacts was done by electrochemical processes including Ni electroless seed-layer deposition and Cu thickening by Light-Induced Plating (LIP). Two batches of samples were used: one deoxidized with HF to remove potential debris after ablation and native oxide and the other without HF treatment before the Ni-P plating. 3. Results and discussion 3.1 Ni-P electroless deposition Ni-P electroless deposition is an autocatalytic process involving the immersion of a substrate in a plating bath. Its principle has been described elsewhere [7]. Ni-P was used as a seed-layer before electrolytic copper thickening. Before starting the Ni-P deposition the edges and Al backside screenprinted contacts were protected with an insulation varnish, to avoid Al dissociation and Ni parasitic deposition during the electroless process. The composition of the electroless Ni-P plating bath and the operating conditions are given in [8]. The deposition rate of the Ni-P layer on silicon was 18 μm/h. The thickness of the Ni-P layer was around 150 nm for a process duration of 30 sec and ~ 300nm for 60sec at the operating temperature of 95°C. The wafers with electroless Ni-P seed layers were not subjected to additional thermal annealing. No significant difference was observed between the properties of cells with the two different values of Ni layer thickness.
a)
b)
Fig. 2. a) 86μm wide finger and 380 μm wide busbar after Cu plating; b) Laboratory scale solar cells of 2 x 2 cm2 area with Ni/Cu metallization after laser ablation with different fluence between 0.2 up to 1 J/cm 2.
3.2 Copper thickening The front contact thickening by copper was performed using LIP (light-induced plating) [9-11]. This method uses the photovoltaic effect of the cell to induce a voltage in the p-n junction under illumination without external contacts on the front-side grid: the photo-generated current therefore acts as a plating
716
Alexandru Focsa et al. / Energy Procedia 38 (2013) 713 – 719
current. By choosing the correct potential difference between the anode and the solar cell back-side, the plating of the contacts can be performed without protecting the back-side contacts. In the present case, the solar cell was dipped in an illuminated electroplating copper solu which was kept at room temperature during the process. During the LIP process, the current density was kept constant (~ 1.0 A/dm2). The external applied voltage was kept close to the photovoltage generated by the cell to avoid Cu deposition on the back side of the cell. The Cu deposition time was around 30 min. Fig.3a shows an SEM image of an Ni-P electroless seed layer of ~300nm thickness deposited from an alkaline (pH=8-10) bath on a laser-ablated SiNx:H and Cu layer of ~5 μm thickness plated from an acidic (pH=2) LIP-bath. Fig. 3b shows SEM pictures of a metallic busbar after copper LIP coating of 15μm and illustrates the smooth Cu surface. This demonstrates that a Nielectroless seed layer allows uniform deposition of the Cu layer by LIP.
Cu
SiNx
SiNx Cu
Ni
Ni
a)
SiNx
b)
Fig. 3. Scanning electron microscope image of electroless Ni-P layer deposited on laser-ablated SiNx:H and Cu line performed by LIP (a) and low magnification SEM image of a busbar after Cu plating (b).
3.3 Electrical characterization of solar cells Fig. 4 shows the measured thickness of plated Cu after LIP and the corresponding open-circuit voltage values for 8 cells as a function of the laser fluences used for the selective ablation of SiNx:H. All samples (Fig. 2b) were subjected to the same LIP process for 35 min using a generated LIP current of 1.2 A/dm2. The maximum VOC value of 616 mV corresponds to a laser fluence of 0.44 J/cm2 and an average Cu thickness of 16 μm. For high laser fluence values ( > 0.7 J/cm2) the Voc of the cell is reduced due to damage at the silicon surface and short-circuiting in the emitter (Fig. 4 a). The fill-factor (FF) of the cells also decreases for high laser fluence (Fig. 4 b). The maximum value of FF, 0.75, was obtained for a laser fluence of 0.44 J/cm2 and 16 μm Cu-thickness.
717
Alexandru Focsa et al. / Energy Procedia 38 (2013) 713 – 719
10
0,590
5
Cu 11-thickening Voc -> (Cu11)
0,580 0,570
0
0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9
1
Laser Fluence (J/cm2)
Copper thickening (μm)
0,600 Voc (V)
Copper thickening (μm)
0,610
15
0,80
20
0,75
15
0,70
0,65
10 5
Cu 11-thickening
0,60
FF (Cu11)
0,55
0
FF (Series Cu11)
0,620
20
0,50 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 Laser Fluence (J/cm2)
a)
1
b)
Fig. 4. Evolution of Cu thickness and I-V parameters of mc-Si solar cells with Ni/Cu metallization (for series Cu11 of cells) as a function of the laser fluence: a) Voc and b) FF.
We studied the influence of deoxidizing of the ablated regions in 2.5% HF before dipping in the Ni-P electroless plating bath. As observed previously, increasing the laser fluence for ablation of SiNx:H decreases the open-circuit voltage of the cells (Fig. 5 a) while the short-circuit current stays roughly constant (Fig. 5 b). The low laser fluences below 0.36 J/cm2 correspond to the ablation threshold of SiNx:H where laser fluctuations cause a large variation of I-V parameters. There are no obvious differences in the values of Voc and Jsc for the two batches of cells. 35
0,610
34
Jsc (mA/cm2)
0,620
Voc (V)
0,600 0,590 0,580
Voc (with HF) Voc (without HF) Poly. (Voc (with HF)) Poly. (Voc (without HF))
0,570 0,3
0,4
0,5
0,6
0,7
0,8
Laser Fluence (J/cm2) a)
32
Jsc (with HF) Jsc (whitout HF) Poly. (Jsc (with HF)) Poly. (Jsc (whitout HF))
31 30
0,560 0,2
33
0,9
1
1,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
0,9
1
1,1
Laser Fluence (J/cm2) b)
Fig. 5. I-V parameters of mc-Si solar cells as a function of laser fluence for series of cells with deoxidation in 2.5% HF and without treatment in HF: a) Voc and b) Jsc. Two Ni thicknesses (Ni~150nm & Ni~300nm) were used.
Table 1 summarizes the photovoltaic parameters of the mc-Si cells produced with optimized laser fluence values between 0.42 and 0.57 J/cm2. The table below gives the open-circuit voltage, short-circuit current and conversion efficiency of tested solar cells considering the total area of 4.4 cm2 for 150nm thick Ni seed-layer and Cu thickening of 9-16 μm. The best cell obtained with a laser fluence of 0.52 J/cm2 exhibits an efficiency of 15.5%.
718
Alexandru Focsa et al. / Energy Procedia 38 (2013) 713 – 719
Table 1. Illuminated I-V parameters of mc-Si cells for the range of laser fluence of 0.42 - 0.57 J/cm2. Efficiency
FF
Voc
Jsc
p.Eff.
p.FF
(%)
(%)
(mV)
(mA/cm2)
(%)
(%)
Average over 18 samples
15.0
73.4
609.4
33.7
15.6
75.8
0.52
340
Best cell
15.5
74.5
615.1
34.0
16.1
77.7
0.56
470
Cells
Rs
Rsh 2
2
0,80
20
0,70
18
0,60
Eff.Cells (%) FF . Poly. (Eff.Cells (%)) Poly. y ((FF .))
16 14 12
0,50 0,40 0,30
10
0,20 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9
1
1,1 1,2
Laser Fluence (J/cm2)
Fig. 6. Evolution of Efficiency and Fill-Factor of mc-Si solar cells with Ni/Cu metallization as ffunction of the laser fluence.
100
FF
22
IQE & Reflectivity (%)
Efficiency (%)
Fig. 6 shows the evolution of fill factor and efficiency of mc-Si solar cells as a function of the laser ablation fluence. The green zone indicates the optimum laser fluence used for the ablation of the SiNx:H layer. The fill factor starts to decrease at a laser ablation fluence of ~ 0.67 J/cm2 due to damage to the emitter surface. As previously noted, below the fluence threshold of 0.36 J/cm2 the ablation of SiNx:H is inconsistent. This gives rise to a highly variable cell FF and efficiency. Figure 7 shows the internal quantum efficiency (IQE) for one of the best cells with Ni/Cu front-side metallization and a laser fluence of 0.48 J/cm2. The calculated value of Jsc from EQE spectra is 34.07 mA/cm2. This is very close to the value of 33.8 mA/cm2 obtained from I-V measurements for the same cell with a ratio of 0.992. The difference of 0.27 mA/cm2 can be explained by the non-optimized process of Ni/Cu plating: parasitic metal micro-particles can be seen on the surface of the SiNx:H layer due to local damage to the coating and this can increase the shadow on the cells. These particles have been observed to contain Cu but are assumed to also contain Ni. The solar-weighted reflectance spectrum from SiNx:H on mc-Si has an average photonic weighted reflectivity of 12.5 13.8% and does not take into account the reflectance from the front grid of the samples.
80 60
Refl._SiN_Cell Cu2 IQE_Cell Cu2
40 20 0 350
450
550
650
750
850
950
1050
Wavelength, nm
Fig. 7. Internal quantum efficiency of Ni/Cu front side metallization mc-Si solar cell and reflectance spectrum.
4. Conclusions The front side metallization of mc-Si solar cells was carried out by electrochemical deposition of Ni/Cu. Laser fluence values of the order of 0.42 0.57 J/cm2, used to locally ablate the SiNx:H antireflection coating, were found to be a good compromise between efficient ablation and limited surface damage or debris formation. Electrical characterization shows reasonably good average cell results with an average efficiency of 15.0 % over 18 laboratory scale solar cells. The best cell gives an efficiency of
Alexandru Focsa et al. / Energy Procedia 38 (2013) 713 – 719
15.5%. Such results were obtained without any additional thermal annealing treatment of the Ni seedlayer. Future work will be focused on the optimization of the process, decreasing the contact finger grid to 40 μm and thermal treatments. Further optimization of the series and shunt resistances are expected when using monocrystalline Si wafers to guarantee the reproducibility of the substrate properties. Acknowledgement The authors thank Stefan Hess from PROLECTRO for providing the copper electrolyte solution used in this investigation and for the fruitful discussion. We gratefully thank K. Fraser from INL for his valuable contribution. This work was supported by the French National Research Agency (ANR) under PROTERRA project. References [1] A. Ogane, K. Hirata, K. Horiuchi, Y. Nishihara, Y. Takahashi, A. Kitiyanan, et al., Laser-doping technique using ultraviolet laser for shallow doping in crystalline silicon solar cell fabrication, Jpn. J. Appl. Phys. 48 (2009) 071201. [2] Guy Beaucarne, Jaap Hoornstra, Gunnar Schubert, Lessons From the 2nd Workshop on Metallization of Crystalline Silicon Solar Cells, Future Photovoltaics , 2010, www.futurepv.com [3] C.E. Dubé, R.C. Gonsiorawski, Improved contact metallization for high efficiency EFG polycrystalline silicon solar cells. Proc. 21th IEEE Photovoltaic Specialists Conference, Florida, 1990, v1, pp. 624-628. [4] A. Knorz. M. Aleman, A. Grohe, R. Preu, S. Glunz, Laser ablation of antireflection coatings for plated contacts yielding solar cell efficiency above 20%, Proc. 24th EU PVSEC, Hamburg, 2009, pp. 1002-1005. [5] J. Bartsch, V. Radke, S Savio, S.W. Glunz, Progress in understanding the current path and deposition mechanisms of lightinduced plating and implications for the process, Proc. of 24th EU PVSEC, Hamburg, 2009, pp. 1469 1474. [6] L. Tous, R. Russel, J. Das, R. Labie, M. Hgamo, J. Horzel et all, Large area copper plated silicon solar cell exceeding 19.5% efficiency, 3rd Workshop on Metallization for Crystalline Silicon Solar cells, Energy Procedia 21 (2012) 58 65. [7] C. Boulord, A. Kaminski, G. Stremsdoerfer, S. Stremsdoerfer, B. Canut, J. F. Chateaux, M. Lemiti, Analysys of wet chemical nickel films for silicon solar cells metallization, 24th EU PVSEC, Hamburg, 2009, pp. 1430-1433. [8] C. Boulord, PhD Thèse, Développement de techniques de métallisation innovantes pour cellules photovoltaïques à haut rendement, 2011, 159p. http://tel.archives-ouvertes.fr/docs/00/67/98/76/PDF/these.pdf [9] A.Mette, C.Schetter, D.Wissen, S.Lust, S.W.Glunz, G.Willeke, Increasing the efficiency of screen-printed silicon solar cells by light-induced silver plating, in: Proceedings of the Fourth World Conference on Photovoltaic Energy Conversion, 2006, pp.1056 1059. [10] M. Aleman, N. Bay, D. Barusha, S.W. Glunz, R. Preu, Front-side metallization of silicon solar cells by nickel plating and light induced silver plating, Galvanotechnik, Photovoltaik, 2, 2009, pp. 412-417. [11] J. Bartsch, V. Radke, C. Schetter, S.W. Glunz, Electrochemical methods to analyse the light-induced plating process, J. Appl. Electrochem. (2010), 40, pp. 757-765.
719