Pv Magazine, January 2009

Transcript

Photo: Coherent Industry & Suppliers Lasers have found their way into solar cell manufacturing, for instance in laser-fired contacts. Photonic synergy Laser manufacturing: Finlay Colville, Director of Marketing in the solar sector at laser manufacturer Coherent, wonders why the potentials of laser technology are still largely unknown within the solar industry. In his opinion, lasers can improve efficiency and production yield in the next generation of crystalline silicon cell production. “Ah, you are from a laser company? Are lasers used in solar? Why would the solar industry use something so esoteric as a laser?” These are some of the most common questions I get asked when attending solar exhibitions and walking the aisles. But why the disconnect? Why do lasers not have the same prominence within the solar industry as chemical-etching or screen-printing? Moreover, why are cell manufacturers’ sales and marketing per- 54 sonnel not flaunting laser-based processes as a unique selling point? In this article, I will attempt to answer these questions and show that there is a special synergy between lasers and solar that combines these two technologies inextricably. Historically, laser sources have received considerably more attention as precision tooling methods within other high-tech production environments. Prominent examples today are found throughout the materials-processing, microelectronics, and flat panel display market segments. Recently however, laser processes have found their way into solar cell manufacturing, for processes such as edge isolation, laser grooved buried contacts, and wrap-through via drilling. Given the extent to which laser-based tools have become embedded within standardized production lines in the semiconductor and flat panel industries, we can expect 01 / 2009 | www.pv-magazine.com Industry & Suppliers to see lasers rapidly becoming adopted widely in the solar cell production. This migration of technology can already be seen in thin-film fabs with standardized laser patterning tools. and do not require any external water or gas supplies. Nearly all DPSS lasers used within the solar industry operate in a noncontinuous (or ‘pulsed’) regime, wherein the lasers emit short pulses of high-inten- Synergy between lasers and solar Solar panels and laser sources have one important thing in common: photons. While solar panels absorb photons from sunlight and transfer them to electricity (the photovoltaic effect), lasers use electricity to convert electrons to highly intense rays of photons. But lasers allow sub-micron resolution micromachining and ultra-thin (tens of nanometers thickness) selective material ablation on the surface of solar cells. Preferred laser sources today actually use electrically excited diode lasers as ‘initial’ energy sources (or ‘pumps’) for solid-state lasers incorporating neodymium-based gain materials. These diode-pumped solidstate (or DPSS) lasers represent ideal sources for manufacturing environments. They are turn-key operated, run off standard single-phase low-current utilities, Scribing laser grooved buried contacts on the front cell surface. Photo: Coherent sity light – generally in pulses of nanoseconds (10-9 seconds) or picoseconds (10-12 seconds) duration. DPSS lasers typically emit light in the near infra-red spectral region at around Advertisement %NGINEERINGßß0RACTICALß%XPERTISEßFORß2ENEWABLEß%NERGIES K_`e$=`cd@e[ljkip=fild XjXgXikf]G_fkfmfckX`ZjK_`e$=`cdN\\b ))e[$)+k_8gi`c)''0`e8[c\ij_f]9\ic`e #>\idXep 0OLICIES 0RODUCTION 2ESEARCHßß$EVELOPMENT -ARKETSßß&INANCE -ARKETINGßß3ALES 3ITEßVISITSßTOßSTATEßOFßTHEßARTß PRODUCERSßOFßTHIN ½ßLMßTECHNOLOGY :fekXZk 3OLARPRAXISß!' -IRIAMß(EGNER (EADßOFß#ONFERENCEß$EPARTMENT 0HONEßß\ßß\ßß  &AXßß\ßß\ßß  CONFERENCES SOLARPRAXISDE WWWSOLARPRAXISDE `eZffg\iXk`fen`k_1 01 / 2009 | www.pv-magazine.com 55 Photo: Coherent Industry & Suppliers Highly intense laser output beam from a Coherent laser. Applications Generally, laser tools find their niche in processes that require high levels of 56 machining precision or performance at fast speed with high throughput. Or where brittle materials are employed which may be adversely affected by any contact-based technology. So, before we discuss any application details, the rationale for the solar industry’s new fixation with lasers is beginning to fall into place. Indeed, there is a myriad of applications proposed for lasers throughout the solar value-chain. Some of these are ‘researchonly’ investigations: some are fully-integrated high-volume manufacturing steps. Fortunately, understanding them is really quite simple, and we do this in two steps. First, almost all applications for lasers can be categorized by ‘process-type’, or how a laser beam interacts with a wafer during the different stages within a production line: scribe, ablate, drill, or melt. Secondly, lasers are used for processes either on the surface (front or back) or through the cells. In direct comparison to other technologies (plasma or chemical etching, mechanical scribing, etc.) lasers find resonance with solar manufacturers if they satisfy any one of three solar roadmap drivers: r

enabling a process to be performed that would not have been possible by using etching or mechanical scribing; r

providing lower manufacturing costs by increased wafer throughput or yield, or by decreased cost-of-ownership; r

increasing the overall cell efficiency. Where lasers are used Applications proposed for lasers throughout the solar industry can be found almost anywhere: for both crystalline silicon (c-Si) and thin-film cell types; raw materials preparation; cell and module stages; academic research, collaborative development, and pilot production lines. However, Photo: Coherent 1064 nanometers. Special crystals are then used to divide this wavelength by a half or a third to either 532 nanometers (green) or 355 nanometers (UV), thereby providing a range of discrete wavelengths for laser based processing. This flexibility in laser output wavelength allows different lasermaterial interactions since every material possesses a unique absorption spectrum which varies with wavelength. Another analogy between solar panels and DPSS lasers is their ‘green’ status. As a ‘renewable’ energy type, the photovoltaic (or ‘PV’) process within a solar cell represents a ‘green’ method of producing electricity. Conversely, DPSS lasers take electricity from the mains and then generate photons (the laser output) without damaging the environment during their operation. This analogy highlights a strong selling point for using lasers within solar manufacturing, at a time when increased scrutiny is being applied to the overall ‘green’ contribution over the lifetime of a solar panel. Selective ablation of silicon nitride layers to create thin grooves for metallization. 01 / 2009 | www.pv-magazine.com Industry & Suppliers Barriers to adoption With so many different options for lasers within next-generation c-Si production lines, how then can the winners and losers be identified? Which processes will remain of research-interest only, and which ones will go mainstream? One way of forecasting this is to identify notable barriers to production entry for laser tools within the industry. Processes that over- 58 come these barriers may then be considered as strong candidates for next generation production line equipment. r

Ease of integration into existing production lines must be evaluated. Here preference will be given to any laser-based processes that are compatible with existing production line handling, in particular those based on standardized turnkey lines. r

Will industrially qualified turn-key laser sources be available for production tools? Understanding which laser sources are suitable for implementation within a 24 hours a day, seven days a week production environment. r

How wide is the process ‘window’? The laser process should be reproducible and repeatable for processing on different cell types: mono or polycrystalline silicon; variable bulk thickness; different dielectric layer stacks and thicknesses. r

Micromachining quality must not be comprised. Quality here includes minimizing laser-induced damage, eliminating bulk microcrack formation, via sidewall finish, or surface material melting or debris. r

Wafer throughput (speed) of the laser tool should be aligned with current state-of-the-art wafer-per-hour and not represent any bottleneck to production lines. Currently, state-of-the-art lasers can be integrated within tools as part of some turn-key production lines, largely overcoming the first two barriers noted above. The process ‘windows’ are somewhat different for each laser based technique, ranging from ‘wide’ as in the case of Edge Isolation to somewhat ‘narrower’ for emerging applications such as Laser Fired Contacts. Throughput remains an area for improvement however; for example, the ability to implement some laser processes, such as front surface dielectric ablation of thin lines for subsequent metallization of the front contacts or Emitter Wrap Though, remains a challenge for a single laser-based tool in production. Currently multiple lasers and scan heads, with parallel wafer processing, present a capital cost barrier that will only be overcome with next-generation lasers with higher average power levels. Of course, capital and running costs are extremely important. However, this merits a different type of discussion. And this brings me onto another common miscon- ception that often comes from laser suppliers or laser equipment tooling integrators, when trying to break into the equipment supply-chain. Often I hear people saying: “Solar energy is a dollar-towatt cost-driven industry. Therefore, this means that only very low cost lasers will enable the decreased dollar-to-watt roadmap demands of the solar industry!” Well, the first part of this is true! But decreased dollar-to-watt module costs come primarily from increasing the efficiency and yield at the cell manufacturing process, in addition to lowering raw material costs in particular silicon. Laser tooling is a onetime capital expenditure hit for new production lines, and has different cost-benefit criteria to consider. We can identify a sequence of steps to explain this better: first, get the quality optimized with no compromise; then provide the throughput with this level of quality; then assess the capital-cost performance-enhancement ratio; finally, ensure that the running costs are minimized by using robust and proven long-lifetime lasers only. Dielectric ablation To illustrate the above issues, we outline now one of the most exciting applications for lasers in next-generation c-Si production lines – Dielectric Ablation. This process involves the selective removal of silicon nitride or silicon oxide dielectric layers, typically used for antireflection coatings (front surface) and passivation (front and back surfaces). This process has also been described as Selective or Dielectric Removal, Contact Opening, and even (mini) Via-Drilling. It forms an integral part of different next-generation high-efficiency c-Si designs, and represents a technique that is exclusively laserenabling. Traditionally, current-carrying metal lines, or conductors, are fabricated using a screen printing process in which a silverPhoto: Xsil/Joe Callaghan to forecast the applications where lasers are poised to play their most enabling role, a few questions must be asked: 1. Which areas form the bulk of the laser research in solar within leading research laboratories worldwide? Today, research in advanced c-Si manufacturing is concerned with new cell concepts, such as the Laser Fired Contacts originated at the Fraunhofer-ISE, or the i-PERC process (industrial passivated emitter and rear cell) promoted extensively at IMEC in Belgium. 2. Which part of the value chain depends most on technology differentiation? Lasers are becoming increasingly used within the wafer-to-cell stage of the value-chain. Here, any increase in cell efficiency, decrease in manufacturing costs (due to thinner silicon wafers) or reduction in operating costs all result in the largest factors contributing to overall lower dollar-to-watt panel costs downstream. 3. Where is manufacturing done today with technologies which will struggle to meet the demands imposed by the solar industry’s roadmap? Highlighted here are any contact-based processes (such as screenprinting) which are anticipated to struggle in maintaining adequate yield levels as cell thicknesses decrease steadily from around 200 micrometers to less than 160 micrometers over the next few years. 4. What cell type is set to continue dominating production output for the next decade at least? Here the emphasis is on modifying the standard c-Si cell type in production today, typically using screenprinted front contacts with a full aluminum deposited back-surface-field (AlBSF). Incremental changes to production line equipment are therefore anticipated to have the highest impact for lasers, in contrast to altogether ‘novel’ cell concepts which may require new process equipment across the whole cell production line stage. Drilled vias through wafers are used in Metal and Emitter Wrap-Through processes. 01 / 2009 | www.pv-magazine.com Industry & Suppliers 01 / 2009 | www.pv-magazine.com mize the thermal diffusion depth, or Heat Affected Zone, around the ablated region. Any material damage has a direct impact on carrier recombination rates, and consequently the overall cell efficiency. Finally, in order to control the ablation accurately, material removal is generally performed through single pulse ablation, necessitating high energy densities (or fluences) at the focused laser spot. Such a laser with high-energy picosecond pulses in the UV was, until the past few months, only available from research-grade lasers with very low throughput and unsuitable for implementation within any production line. The transition from ‘research-grade’ to ‘production-ready’ laser performance required the use of highly intenseintensestable laser pumping sources and robust optical designs that allow turn-key operation over tens of thousands of hours. Thankfully this type of laser system has been on the laser industry roadmap for some time, and last year saw the first release of a turn-key industrial platform to meet the application quality demands. While the roadmap for lasers in the solar industry certainly involves increased output powers, shorter wavelengths in the UV, and shorter pulsewidths at the picosecond level, there are other drivers for laser systems. These include the use of socalled ‘Quasi-CW’ lasers with very high repetition rates at the 100 MHz level. Here, laser-material interactions are no longer ‘single-pulse’ processes. Rather they are based on the use of multiple pulses on the sample in a very short time period. At the other end of the pulse-spectrum, longerpulse lasers are required for some melting applications such as Laser Fired Contacts. Next-generation tools As the solar industry prepares to enter another growth phase over the next decade, next generation tooling will be required to maintain key c-Si technology roadmap objectives. Laser-based tooling can be expected to play an enabling role within new production lines, using next-generation state-of-the-art DPSS lasers with performance envelopes driven by opportunities within the solar industry. This transition will help redress the imbalance, or ‘disconnect’, referred to at the beginning of the article. Increasing visibility of laser tools within standardized cell production lines in the future will ensure that they receive the same level of attention previously only afforded to etching and screen printing technologies. In the future then, I look forward to eating humble pie and walking the aisles once more at the solar trade-shows. Then it will be my turn to ask the question: “Talk to me about which tools use lasers as standardized processes in your production lines today, and which ones you are hoping to introduce in the next twelve months!” V Finlay Colville Photo: Coherent containing paste is deposited in a desired pattern and subsequently fired in a high temperature sintering process through the silicon nitride antireflection coating. This procedure comes with a couple of drawbacks. First, the screen printing process is by default contact-based. Second, high temperature firing limits the types of materials that can be used for the front contacts themselves. A range of novel schemes has been proposed to address these issues, partly in anticipation of thinner wafers below the 180 micrometers mark. These approaches often involve very precisely controlled laser cut grooves, subsequently filled by an electroless plating method. The laser step here represents one of the most challenging applications for lasers to date in c-Si cell manufacturing. And as such, the right choice of laser output parameters is essential to obtain the processing quality outlined earlier in this article. For the laser beam is required to selectively remove only material from the dielectric layer(s), whether for front or back contact formation, while causing the minimum of damage to the underlying c-Si material. For front-surface Dielectric Ablation, typically we have a silicon nitride thickness of around just 70 to 90 nanometers. The true challenge of ‘Selective’ Ablation could not be more evident. In actual fact, the challenge here for equipment manufacturers is two-fold. Both on the laser process and the subsequent metallization. The requirement to remove the dielectric layer with ultrahigh precision and not cause any excessive damage to the underlying n-type doped region has been the subject of considerable research over the past couple of years. And this is one of the first laser based processes in c-Si manufacturing that can only be done by the very latest types of industrially-qualified pulsed laser sources. This is where solar cell production equipment roadmaps meet laser source roadmaps. The critical requirements for Selective Ablation and minimized peripheral damage to the underlying c-Si demand the use of lasers operating with very short wavelengths and very short pulsewidths. Short wavelengths in the UV region (355 nanometers is most common) allow stronger absorption within the materials: short pulsewidths at the sub-nanosecond level (called picosecond) are essential to both increase the ablation quality and to mini- Finlay Colville, Director of Marketing in the solar sector at Coherent. Coherent, based in Santa Clara, California, USA, is one of the world's largest laser manufacturers. A GREEN ALTERNATIVE In the crystalline silicon production lines, lasers are to date only used at the Edge Isolation stage. They are used for niche applications to drill the vias through the wafers for both Metal and Emitter WrapThrough processes, and for a small quantity of production lines that use Laser Grooved Buried Contacts. Laser processes are mainly for new and enabling processes, like Wrap-Through, Dielectric Ablation, Laser Fired Contacts, and Laser Dopant Diffusion. It is only in the Edge Isolation stage that they compete directly with wet chemical processes. They offer a green alternative because lasers do not require any special utility provisions. No toxic chemicals, no external plant water supply. As a result, lasers have a near-zero carbon footprint. Lasers provide a highly precise means of removing or altering materials with sub-micron resolution on the surfaces of a c-Si solar cell. They are able to remove very thin layers of less than 100 nanometers without altering the materials above or below the layer being ablated. This is important because these features are integral to next generation front and back contact processes coming into production for use with ultra-thin wafers below 160 microns thickness. The processes involve thermal laser-material interaction, either as direct heating, melting, or absorption within a substrate as a means to ablate the materials. 59