An ultracompact picosecond laser meets production demands of the thin film solar cell industry
Ultrashort pulse lasers are used for high-precision structuring of many materials because of their non-thermal or “cold” ablation processing ability. However, compared to conventional laser sources established in industry, these systems are said to suffer from higher investment and operating costs. In the materials processing market picosecond laser sources are garnering more attention because conventional laser sources have reached their limits in ablation quality.
|FIGURE 1. Schematic cross-section of a CIS thin film solar cell.|
Modern picosecond laser systems are now robust enough to be operated in a manufacturing environment. Another important factor for their usability in industry is the cycle time of the manufacturing process. State-of-the-art picosecond sources operate at repetition rates in the 500kHz range where the process speeds needed for industrial applications can be achieved. One of these manufacturing-compatible ultrafast laser sources, the picoREGEN Industrial, from High Q Laser Production GmbH, has an output power of up to 30 W, a pulse duration of 10 ps, and a repetition rate of 500 kHz.
An interesting application area for this laser is thin film processing where the combination of selective ablation with high precision and high speed makes ultrafast lasers unbeatable; especially the structuring of thin film solar cells is today based on nanosecond laser ablation and mechanical scribing. Both of these processes are feasible in principle, but there is still a substantial potential to increase repeatability and cost efficiency in mass production.
|FIGURE 2. Optical microscope image (left) and CIP topography profile (right) of a P1 line structured with a high repetition rate Q-switched ns-laser (Pulse duration approx. 100 ns, repetition rate 100 kHz) with a groove width of approx. 30µm. Edit this paragraph|
Historically solar cells are manufactured utilizing silicon wafer technology. The approach to use low-cost float glass substrates instead of expensive silicon in combination with thin film coatings as functional layers is promising to reduce production cost dramatically. Thin film solar modules can be manufactured in an economical in-line process, consuming considerably less material, thereby having the potential to compete economically with conventional energy sources.
The thin films have to be line structured with galvanic separation without damaging the substrate or other layers. Today the structuring processes integrated switching are based on either nanosecond laser ablation1,2 or mechanical scribing.3,4 Investigations on structuring Mo,5-8 CIS,9 and Zinc oxide (ZnO) or other transparent oxides10 by ultrafast lasers have shown that short laser pulse duration minimizes thermal effects, providing more reproducible results and less damage, but the process speed is still too slow for industrial applications.
|FIGURE 3. CIP topography of a P1 line structure produced with a ps-laser showing galvanic separation over a groove width of approx. 25µm with a resistance > 1 MOhm.|
Recently Copper-Indium-Diselenide (CIS) thin film solar cells have created interest due to their high efficiency compared to other thin film solar cells.12-14 CIS thin film cells consist of basically three functional layers (see Figure 1).
First a 500nm Molybdenum (Mo) layer is sputtered on a few-millimeter-thick glass substrate and then it is structured by nanosecond laser ablation in a so-called Pattern 1 (P1) process. A 1 to 3µm thick CIS layer is now deposited on the structured Mo film that acts as an absorber for light and as p-type semiconductor. In a Pattern 2 (P2) process the CIS is structured and galvanic separated by mechanical scribing. Subsequently, a 1 to 2µm thick transparent and conducting ZnO layer is sputtered on top of the CIS, which forms the n-type side of the solar cell. To improve performance a thin buffer layer (CdS, 100nm) can be added between CIS and ZnO. The Mo layer is often referred to as “back contact” and the ZnO layer as “front contact.” In a final structuring process Pattern 3 (P3) the ZnO- and CIS-layers are structured together on top of the Mo-film by mechanical scribing.
|FIGURE 4. P2 picosecond laser structuring. The figures show a confocal imaging depth profile and a cross-section of the groove, respectively, indicating a separation free of thermal damages.|
Current production employs a nanosecond laser process for the P1 structuring (see Figure 2), which can lead to high rims, causing local shunts and shortcuts through the upper layers, which are comparable to the thickness of the Mo layer. Additionally micro cracks and partial material lift-offs, leading to solar cell lifetime issues, are produced in a range of 5 to 10nm at the sides of the groove. The substrate glass shows melting zones where pulses were overlapping. These effects are drawbacks in the production process. At the melting zones where the pulse overlaps the barrier layer can be damaged. All damages can be traced to thermal effects during nanosecond ablation.
The galvanic separation of the Mo layer with a picosecond laser is shown in Figure 3. Ripple-like contours as well as a slight height asymmetry between upper and lower rim can be observed. The rim formation is less dominant than with nanosecond laser ablation.
The production process for P2 is mechanical scribing of the CIS layer on top of the relatively hard Mo layer and has the disadvantage of line width changes within a range of 60 µm to 120 µm. In Figure 4 the P2 structuring process by picosecond laser ablation is shown. The CIS layer can be removed selectively with high quality and small grooves of approx. 22 µm. The process speed can be scaled up to approximately 3 to 4 m/s with the picoREGEN Industrial laser. The images indicate a selective removal without thermal effects as micro cracks or lift-offs.
The current production process for the separation of a ZnO/CIS double layer on top of the hard Mo layer is similar to the P2 mechanical scribing. The scribed lines display a width of >70 µm and show irregular lift-off, chipping of the ZnO and CIS layers as well as residual material at the line bottom. The speed of scribing is a few cm/s.
By employing a picosecond laser it is possible to structure the ZnO/CIS double layer selectively from the Mo layer without any damage. In Figure 5 double layer structuring of the ZnO/CIS is shown. The selectivity of the structuring process can be clearly seen by the cascaded side walls. The line width of just 18 microns offers higher reliability of the process and a decrease of material loss.
The results show that the picosecond laser at 1064 nm can serve as a universal tool covering all process steps (P1-P3) of structuring CIS thin film solar cells. The high repetition rates of state-of-the-art picosecond lasers enable industry-relevant structuring speeds of a few meters per second and, in addition, the high quality and uniformity of the structured lines show the potential of reducing manufacturing losses through waste and increasing life time of the solar cell modules.
|FIGURE 5. P3 structure performed by picosecond laser ablation: The images show an optical microscope image (left) as well as a cross section (right) of the ZnO/CIS-double layer structured by picosecond laser ablation|
1.A.D. Compaan, I. Matulionis, and S. Nakade, “Laser scribing of polycrystalline thin films,” Optics and Lasers in Engineering, 2000, 34(1): p. 15-45.
2. C. Molpeceres et al., “Microprocessing of ITO and a-Si thin films using ns laser sources.,” Journal of Micromechanics and Microengineering, 2005, 15(6): p. 1271-1278.
3. B. Dimmler and H.W. Schock, “Scaling-up of CIS technology for thin-film solar modules,” Progress in Photovoltaics: Research and Applications, 1996. 4(6): p. 425-433.
4. V. Probst et al., “Advanced stacked elemental layer process for Cu (InGa) Se2 thin film photovoltaic devices,” in Materials Research Society Symposium – Proceedings, 1996, San Francisco, CA, USA: Materials Research Society.
5. J. Hermann et al., “Comparative investigation of solar cell thin film processing using nanosecond and femtosecond lasers,” Journal of Physics D: Applied Physics, 2006, 39(3): p. 453-460.
6. J. Hermann et al., “Selective ablation of thin films with short and ultrashort laser pulses,” Applied Surface Science, 2006, 252(13 SPEC. ISS.): p. 4814-4818.
7. H. Huber et al., “High Repetition Rate Ultrafast Lasers and their Applications in Micro Machining,” in LAMP 2006, International Congress on Laser Advanced Materials Processing. 2006. Kyoto.
8. S. Zoppel, H. Huber, and G.A. Reider, “Selective ablation of thin Mo and TCO films with femtosecond laser pulses for structuring thin film solar cells,” Applied Physics A: Materials Science & Processing, 2007, 89(1): p. 161-163.
9. D. Ruthe, K. Zimmer, and T. Hoeche, “Etching of CuInSe2 thin films – Comparison of femtosecond and picosecond laser ablation,” Applied Surface Science, 2005, 247(1-4): p. 447-452.
10. G. Raciukaitis et al., “Patterning of indium-tin oxide on glass with picosecond lasers,” Applied Surface Science, 2007, 253(15): p. 6570-6574.
11. H.P. Huber et al., “Selective structuring of thin-film solar cells by ultrafast laser ablation,” in Proceedings of SPIE – The International Society for Optical Engineering. 2008. San Jose, CA.
12. V. Probst et al., “Rapid CIS-process for high efficiency PV-modules: Development towards large area processing,” Thin Solid Films, 2001, 387(1-2): p. 262-267.
13. F.H. Karg, “Development and manufacturing of CIS thin film solar modules,” Solar Energy Materials and Solar Cells, 2001, 66(1-4): p. 645-653.
14. M.A. Green et al., “Solar cell efficiency tables (version 29),” Progress in Photovoltaics: Research and Applications, 2007, 15(1): p. 35-40.
Dr. Sandra Zoppel ([email protected]) and Dr. Heinz Huber are with High Q Laser Production GmbH – Hohenems, Austria (www.highqlaser.at).