17 Jul 2008
Lasers are being exploited in the manufacture of both silicon and thin-film solar cells. Gaetan Rull of Yole Developpement sizes up the opportunity for laser makers.
The most important objective for today's photovoltaic (PV) market is to reduce the ratio of manufacturing cost to power output, and more specifically for solar modules, the manufacturing cost per kilowatt hour (kWh). With this in mind, technical solutions are required to reach grid parity and make PV a competitive energy source. Lasers already have several applications in the PV manufacturing process such as edge isolation or cell structuring. This article will describe the PV industry from the production side and then focus on opportunities for laser manufacturers.
Tracking PV technology
In 2007, crystalline silicon solar cells (c-Si) remained dominant with over 85% of the production, but we observed a strong increase in production capacity in PV thin films (see figure 1). After ULVAC and Oerlikon, the entrance of AMAT changed the future landscape of the PV industry while boosting the production capacity of amorphous silicon (a-Si)and tandem amorphous and crystalline silicon (a-Si/µ-Si) modules.
As reported in Yole's "PV Fab Database 08", we expect the production capacity of a-Si-based thin films to jump from 1.4 GW at the end of 2008 to 2.4 GW in 2009 and 5 GW in 2010. Several fabs reaching a size of 1 GW will be installed in the coming years and we expect the PV industry to follow the same trend.
In order to look at all PV thin-film technologies, we must not forget the modules based on copper, indium, gallium and selenium (CI[G]S) and those based on cadmium and tellurium (CdTe). CIS or CIGS is a promising technology. While the CIGS industry is still fighting to pass an efficiency of 10%, NREL Labs has shown the potential of CIGS technology by reaching an efficiency of 19.9%. We estimate that around 40 CIGS projects are underway globally. The large majority of these are either spin-offs from universities or R&D institutes in the process of raising funds and building plants.
The CIGS industry is fragmented and it is difficult to find two companies that have the same manufacturing process. Considering just the active layer deposition, one can find sputtering, co-evaporation, printing, electro-deposition and molecular beam epitaxy. The lack of manufacturing tools is stalling the ramp-up of fabs.
Machine makers need time to adapt their products to the specific needs of each CIGS project. Veeco provides a co-evaporation deposition system and is planning to launch a sputtering web coater. Centrotherm Photovoltaics offers a turnkey fab for CIGS deposition using sputtering machines.
CdTe is mainly being developed by one player: First Solar. Being the only player able to deposit a CdTe compound with its proprietary vapour transport deposition technology, First Solar has put up strong economic and technological barriers to potential newcomers. The strength of the firm is nothing less than its ability to produce the cheapest PV modules in the world at $1.12/watt peak (source First Solar). First Solar is focusing on large-scale installations such as solar parks or >30 kW roof installations and plans to reach a capacity of 1 GW by the end of 2009.
Another large producer of CdTe is Calyxo, which merged with US firm Solar Fields LLC in November 2007. In the merger, Calyxo (which is owned by Q-Cells) acquired Solar Fields's intellectual property on an atmospheric vapour deposition process. An interesting point is that Solar Fields was founded by Harold McMaster who also founded Solar Cells, formerly First Solar.
Finally, the question of substrate size remains unanswered and we are monitoring current trends. Equipment suppliers that have experience of the flat-panel display industry want to cut the cost while processing Gen5 or Gen8.5 glass sheets. Others, such as Oerlikon, maintain that a substrate size of 1.4 m2 ensures that panels are easier to handle both in production and installation, and can ultimately reduce the cost per kWh.
Coming back to the c-Si segment, cell production capacity is expected to pass 10 GW at the end of 2008. Among this global production capacity (see figure 2), 7.3 GW are located in Asia and 4 GW in China. China is adding more than 2 GW of cell capacity in 2008 only and equipment suppliers are convinced that the situation will remain the same in the following years.
Three Chinese companies are in the 2008 top 10 of c-Si production capacity: Suntech, JA Solar and China Sunergy (see figure 3). Suntech plans to reach 1 GW of capacity in 2009 and, if successful, would become the largest c-Si cell fab passing both Q-Cells and Sharp.
Lasers in the solar industry
From a manufacturing point of view, the PV industry is essentially split into two parts: wafer-based and thin-film solar modules. In the wafer-based sector, 6- or 8-inch wafers are processed into solar cells and then assembled into a module ready for installation. Thin-film solar modules are usually monolithically integrated and here a glass substrate (1–5.7 m2) passes through different deposition equipment and becomes a solar module after lamination. Even though these two ways of manufacturing solar modules are different, lasers are used in both of them.
The two main applications for lasers in the PV industry are edge etching for wafer-based solar cells and module patterning for thin-film technologies.
In the wafer-based (c-Si) approach, after the phosphorous doping, edges of the cell have to be etched in order to avoid contact between the front (n-type silicon) and back (p-type silicon) surfaces. This operation, known as edge isolation, reduces the recombination rate and play on cell efficiency. Three technologies can be used, plasma etch, wet bench and laser scribe, each having various pros and cons. Lasers scribers avoid the use of chemicals and provide good-quality etching and can also be integrated into in-line tools and so have a low footprint. Lasers currently have a 50% share in this application and the penetration rate is increasing as fab automation continues. We anticipate a strong increase of technology share for lasers in edge-etching.
The other main application is the patterning of thin-film modules. Pushed by turnkey solution providers, most thin-film fabs produce monolithically integrated modules. Starting from a glass sheet, the transparent conductive oxide electrodes and the active layer are deposited on the whole substrate surface. Between each deposition step laser patterning is required to separate the layer into cells connected in serial. The 10–30 µm width patterns need to be accurate and prevent possible contact between the top and bottom electrodes that would result in a shunt and decrease efficiency of the module.
Lasers are used three times in the manufacture of a-Si and a-Si/µ-Si thin-film technologies (see figure 4). The first scribing is for the front electrode and is done on the layer side using a 1024 nm laser. The two other scribes (for the active layer and the back electrode) are done through the glass by a 532 nm laser.
For CI(G)S technology, due to the use of heat sensitive materials such as molybdenum, only the first scribing can be done using a laser. The two other patterns are done by mechanical scribers (needles) and need a higher level of automation. Many laser suppliers are developing solutions to replace these needles. One solution could be femtosecond lasers although their high price remains a barrier. That said, replacing needles with lasers would reduce maintenance time and simplify the tool.
The laser market for PV applications has a bright future. We expect to see an increase of global thin-film production capacity, a high penetration rate for edge-etching of c-Si cells, and a large number of new applications such as ID marking, defect and shunt repairing and back-contact deposition.
In our next report on equipment and materials for PV, Yole Developpement estimates the market for laser scribing tools (lasers and automation) at $217 m (€137 m) for 2008. The main laser suppliers sharing the PV market are Coherent, Newport and Rofin, and we see Trumpf and LPKF increasing their sales volume. For thin films, lasers are then integrated into automated tools by companies such as Manz or Innolas, or even by the turnkey suppliers themselves such as Oerlikon and AMAT. In the c-Si manufacturing process, lasers for edge etching are integrated into existing tools such as firing furnaces.
Lasers will continue to penetrate and find new applications in the fast growing PV industry. The future seems bright but laser suppliers and integrators still need to succeed in transferring new applications from R&D to production scale whilst keeping the same level of performance.
• Yole Developpement is a strategy and marketing consulting company involved in several fields: photovoltaic, MEMS, compound semiconductor, nanomaterials and life science. Gaetan Rull heads the photovoltaic activity at Yole. He also is the author of the PV Technology, Equipment, Applications and Market report and the PV Fab Database. For more information please contact firstname.lastname@example.org.
• This article originally appeared in the July/August 2008 issue of Optics & Laser Europe magazine.
View pdf of article