To receive the next issue by e-mail, visit My Account and confirm you have opted-in to receive e-mail from AE.
AE's inaugural edition of the PV Sun Times® e-newsletter presents solutions for a growing problem in certain PECVD process for solar cell manufacturing: RF electrode crosstalk during inline batch processing of multiple panels. Our Ask the Solar Experts column discusses options for low-temperature TCO processes, the benefits of vacuum-based processing for solar applications, and other advice for the solar industry’s most pressing manufacturing issues.
Increase Throughput with Proper Setup of Multiple RF Electrodes for Inline Batch Systems
In the drive to achieve grid parity, it’s critical to continually increase the number of watts per year that are delivered to consumers via PV devices. For this reason, solar PV factories are always working to increase panel efficiency and factory throughput. A major throughput-boosting trend involves processing multiple panels at one time with inline batch systems containing multiple RF electrodes.
The Challenge: RF Crosstalk
The problem is that as this trend grows, so does a common difficulty associated with it: RF crosstalk caused by unwanted interactions between closely-spaced RF electrodes. This poses rising difficulty as new processes utilize increasing numbers of electrodes within a single chamber. As the number of electrodes grows, the space between them often shrinks, which further increases the likelihood of crosstalk. The damage resulting from RF crosstalk, including substrate arcing and defects, can significantly degrade process and device quality. Other problems that may arise from RF electrode crosstalk include contamination, as well as damage to fixtures, power supplies, and other equipment.
Figure 1. Ideal inline batch process for PECVD processing of multiple substrates
Figure 1 illustrates a PECVD process for solar PV manufacturing in which seven RF electrodes (cathodes and anodes) are closely spaced together in a single chamber. Each powered electrode (cathode) is connected to a separate RF power supply and impedance matching network. Typically, the other electrodes (anodes) are connected to ground. Multiple glass substrates are inserted between these electrodes. At the top of chamber, a manifold supplies gas to the process. This illustration shows the ideal situation for this setup, in which each electrode interacts only with the grounded anodes adjacent to it, the plasma remains localized between the electrodes, and the glass substrates are coated in a uniform manner.
However, in reality, it is extremely common for the plasma to stray into undesired areas because of unwanted interactions between the RF electrodes (crosstalk). Plasma may stray to areas above or below the electrodes, depending upon system and shielding design, as well as partial pressure and total pressure. Arcing may occur at levels significant enough to damage the substrate or process equipment. The most significant issues may occur near components in the chamber such as the shielding or gas manifold, which potentially conduct current and thus may strengthen unwanted electrical connections between the RF electrodes. However, even without the influence of the various chamber components, electrode crosstalk will likely occur, eliminating the potential throughput benefits of this type of process.
The Solution: CEX and Phase Shifting
The problems inherent to this type of process setup are significant, but it is still possible to achieve the throughput enhancements that the multi-RF-electrode inline batch system offers. The following describes key strategies that alleviate electrode crosstalk and the straying plasma associated with it.
CEX
First of all, it is essential to synchronize your RF power supplies. The CEX (common exciter oscillator) feature, provided on the Cesar® RF power supply and other AE products, synchronizes RF output. In Figure 1, the anodes are connected to ground and therefore are at ground potential. Typically, the cathodes are set alternating to one another so that the far left electrode shown in Figure 1 is at 0°, the middle electrode is at 180°, and the one to the right is at 0°. Other systems require a 0° phase shift between all electrodes. In the ideal situation shown in Figure 1, which, generally, can only occur with the use of CEX, the plasma remains isolated between the anodes and cathodes. Without the use of CEX, the plasma potential is random, and unwanted interactions between electrodes can occur.
AE’s Cesar RF power supplies can synchronize as many as six power supplies with CEX. If your process uses more than six RF power supplies, an external CEX connected to all power supplies produces the same results.
Phase Shifting
After properly synchronizing your RF power supplies with CEX, if you continue to experience crosstalk and its associated problems, there may be conditions unique to your system that necessitate the additional use a phase shifter to “fine tune” the phase of the RF output. Typically, CEX sets the phase of the RF output at exactly 0 or 180°. This usually localizes the plasma and eliminates problems associated with RF crosstalk. However, each process is unique. Chamber configuration, fixture conductivity, differences in cable length from one power supply to another, and other conditions may create a situation in which a 180° phase shift may not be ideal. In certain processes, instead of 180°, you may need a 179 or 178° shift. In this situation, CEX may be combined with phase shifting to account for any process- or system-specific characteristics and create the best possible plasma localization.
Implementation
Your RF crosstalk problems may be solved solely with the use of CEX, or you may require the two-step CEX-plus-phase-shifter solution in order to reap the full throughput-enhancement potential of your multi-RF-electrode inline batch system. AE application engineers have experience with the implementation of the solutions described above, including determining the ideal phase shift for any process. Further, they are available to answer your questions and provide hands-on help with setup of these solutions in your particular system. Please contact us for more information.
AE’s solar experts, Ken Nauman and Doug Pelleymounter, have a combined 45 years of experience with the processes used in solar PV manufacturing. Here, they offer advice for the solar industry’s most pressing manufacturing issues.
We’d love to hear from you! Please send your solar-related questions and comments to PVSunTimes@aei.com.
- I’m new to AE’s solar offerings. What does AE bring to the table for PV manufacturing?
- Our company is just starting out and needs to order a large volume of equipment at once. Can AE provide all the equipment I need within a tight timeframe?
- Why should I choose a vacuum-based manufacturing process? What advantages does that technique offer over other available PV manufacturing methods, such as printing and evaporation?
- I’m doing a CIGS process and my last layer is a TCO. Do you have any suggestions for controlling the temperature of my TCO process to avoid degrading the active layer underneath it?
- I’m new to AE’s solar offerings. What does AE bring to the table for PV manufacturing?
Answer: Where do we start? AE offers solutions for crystalline silicon, wafer-based solar photovoltaics, as well as for the major thin-film technologies, including amorphous and microcrystalline silicon, CIGS, and CdTe. We have one of the most comprehensive product lines in the industry, which enables us to offer effective solutions for every phase of PV production: power supplies from DC up to 60 MHz, thermal instrumentation, an array of gas, vapor, and pressure control products, and more. These products feature highly developed designs and technologies based on our 25 years of innovating solutions that increase precision, prevent defects, and improve throughput. In fact, for more than 20 years, companies developing and manufacturing photovoltaics have chosen our products. However, what we offer goes beyond our products and technologies. It also includes expert applications support, world-class manufacturing facilities, an established global sales and support infrastructure, and more.
Please see AE’s solar market web page for further details.
Table 1. AE products for solar PV manufacturing
- Our company is just starting out and needs to order a large volume of equipment at once. Can AE provide all the equipment I need within a tight timeframe?
Answer: It’s an exciting time because it’s been many years since a new thin-film process has emerged. One nice thing for the emerging solar market is that it can take advantage of all of the development that has happened for existing, adjacent markets. This development includes technology, as well as the infrastructure for equipment manufacturing and support. AE’s work in markets such as semiconductor, FPD, and industrial coatings has enabled us to develop substantial manufacturing capacity. We already have the processes, facilities, vendors, and other necessary resources in place at our world-class facility in Shenzhen, China in order to efficiently turn around an order of any size, such as a 30 megawatt or larger order for a new solar manufacturing operation.
In addition to equipment, we provide the support you need for the success of your new manufacturing operation. AE application engineers are on call to assist you with process development, setup, optimization, and troubleshooting. They offer valuable insight and expertise based on long-term experience with a host of markets, manufacturing techniques, and process conditions.
With sales and service offices in the major worldwide manufacturing centers, AE also has the global infrastructure to efficiently serve a worldwide industry such as solar. For example, if you are located in Europe, our local office can assist you from a convenient, nearby location. Likewise, if your customer is located in Asia, we have many offices you can work with throughout that continent, as well.
Figure 2. AE’s world-class manufacturing facility in Shenzhen, China can be rapidly scaled to meet the considerable equipment needs of a new solar manufacturing operation
- Why should I choose a vacuum-based manufacturing process? What advantages does that technique offer over other available PV manufacturing methods, such as printing and evaporation?
Answer: Today’s methods for PV manufacturing include sputtering (PVD), PECVD, printing, evaporation, and more. However, vacuum-based processes such as PVD and PECVD offer definite benefits that the other methods simply can’t deliver. Specifically, PVD and PECVD provide atomic-level control that enables you to more precisely determine film characteristics, such as stoichiometry, crystallinity, and uniformity across the substrate. PVD and PECVD also produce fewer defects than other methods. This high level of control culminates in two critical benefits for today’s solar panel manufacturers: greater PV efficiency and increased throughput.
Figure 3. Simplified representation of a sputtering (PVD) process—Other PV manufacturing methods can’t match the precision of vacuum-based processes, which work on the atomic level
Figure 3 illustrates the atomic-level behavior of a sputtering process. In the first step of this process (left), argon atoms are ionized. An accelerated electron strikes an atom in an inelastic collision that removes an electron from the atom, creating an Ar+ ion. Next, during the sputtering step (middle), the Ar+ ion is accelerated toward the negative cathode surface. It strikes with enough energy to remove target material. In the final phase (right), the target material reaches the substrate surface, where it is deposited as a thin film. For more information on sputtering, please see our Sputter Spotlight® E-newsletter.
Another benefit of using a vacuum-based process is the fact that within the areas of PVD and PECVD, a great deal of expertise and technological development has been amassed that can be applied directly to PV manufacturing. AE offers over 25 years of experience, as well as a comprehensive and highly developed product portfolio that enables an exceptional level of control over film properties compared to other subsystem manufacturers. For example, our products enable a lower defect rate, which not only increases solar cell efficiency, but allows higher-power operation as well, resulting in increased throughput. Higher-power operation also enables successful coating of large-area substrates. Our Crystal® AC power supply has a long track record of success in achieving the power levels required for architectural glass applications (including low-E coatings for the passive solar market), which also makes it ideal for the increasing substrate sizes in the PV industry. Please see our Design Aspects of Large-Area Coating Supplies white paper for more information.
In fact, AE’s expertise in large-area coating for industries such as FPD and architectural glass has direct application to large-area PV manufacturing. We’ve honed our products, technologies, and expertise in these adjacent markets, as well as the semiconductor industry, which, of course, is the original silicon-wafer application. You could say that AE cut its teeth in the semiconductor industry, an industry that requires extreme manufacturing precision and allows little or no margin for error. In fact, semi has the smallest process window of any industry. Therefore, our products and technologies are designed around the concept of precision, a fact that benefits solar in the form of increased cell efficiency and process throughput.
- I’m doing a CIGS process and my last layer is a TCO. Do you have any suggestions for controlling the temperature of my TCO process to avoid degrading the active layer underneath it?
Answer: Absolutely! Thermal budgeting is a pressing issue for many manufacturing applications today. A little background for our readers: most PV manufacturing processes deposit the TCO layer first, before any other layers. However, for CIGS (and some thin-film Si) solar cells, the TCO is the last to go down. Unlike metal layers, which can be deposited with cold processes because their electrical conductivity is relatively unaffected by temperature, the conductivity of TCOs is highly affected by heat. To produce sufficient electrical conductivity, conventional TCO processes are performed at high temperatures. The problem is that for CIGS processes, which deposit the TCO last, this may exceed the thermal budget of all preceding layers. Excessive temperature can cause diffusion of the dopant within the active layers underneath the TCO, resulting in significant PV performance degradation. Further, if the substrate is temperature-sensitive, it can actually melt under the temperatures of conventional TCO deposition processes. This is a particular issue for flexible polymer substrates.
Figure 4. For a CIGS solar panel, the last layer deposited is the TCO, while on a-Si and CdTe panels, the first layer deposited is the TCO. This poses special heat-related challenges for CIGS manufacturing
So, what is the answer to this seemingly dire situation? Power methods exist that can be performed in a temperature range that will not cause diffusion of the active layers or substrate melting, while producing good TCO conductivity. These are standard methods with records of success for other processes requiring temperature control, such as for electrodes for FPD color filters, and transparent conductors for touch panel processes. Please contact us for details on an effective solution for your temperature-sensitive process.