Fuel Cell, LED Makers Look to Materials, Equipment Suppliers to Help Cut Production Costs
Paula Doe, SEMI, San Jose -- Semiconductor International, 7/13/2007
Fuel cells or photovoltaics (PV), alternative energy all boils down to the cost of scalable production, and the semiconductor industry is the poster child for that. PV makers have very little margin to work with, since PV suppliers sell their cells for, at best, twice the cost of the silicon wafers, putting immense cost pressure on the manufacturing process. The sector has already converted to wire saws to slice its silicon with less kerf loss, and is now using 200-µm-thick wafers while experimenting with 100 µm ones. But the PV industry doesn’t have the cost structure to treat thin wafers as carefully as the IC industry does. That is why there is so much innovation in printing thin films and roll-to-roll processes, which in theory can bring the cost of PV down to $1.00/W with throughput of 100 ft/min, and could potentially bring solar generation costs down to 10 cents/W, with efficient manufacturing of more light-absorbent films. Clearly, there is a lot of work yet to be able to achieve these types of levels in production.
Conventional LCD-like thin film on glass equipment makes sense in the world as it is today, but not in the world as it is becoming — unless you think printing is not going to work. “The PV industry, as it’s evolving, is a big opportunity for equipment, but the types of things the semiconductor industry has to sell cost too much, don’t fit my needs too well, and I don’t know how to use them. The clock is ticking until the PV industry goes off and finds other suppliers and the semiconductor industry misses the boat,” warned a cell maker pursuing roll-to-roll manufacturing techniques.
With fuel cells, the main problem is high costs, though good progress is being made in improving efficiencies. There are numerous ways more sophisticated nanomaterials and manufacturing processes could help. Hydrogen storage is probably the biggest barrier, and nanostructures would be likely to allow more storage with less weight. There is also room for improvement in the proton conducting membrane, which is still typically DuPont Nafion, borrowed from another industry back in the 1960s. Nanocomposite electrolytes could potentially improve mechanical stability, reduce fuel crossover and improve water retention, a key issue because water is the limiting factor. The catalyst is now screen printed onto the membrane, but better control of the catalyst particles’ shape and arrangement would improve efficiency. Designing nanostructures to create as much surface area as possible for the reaction would also help. MEMS processes can be used to make an array of small membrane columns, but reducing them further pushes up manufacturing cost. Another big need is better gas seals at the interfaces, because the main failure mechanism is seal failure. A less gas permeable material is a big need.
| Source: S. Litser, G. McClean, PEM Fuel Cell Electrodes. J. Power Sources 2004 |
Solid-state lighting could also bring big cuts in energy usage and carbon emissions if the manufacturing technology is improved. Gallium nitride (GaN) light emitting diodes (LEDs) passed fluorescent lights as the world’s most efficient white light source, recently achieving output of 130 lumens per watt. That means a 5-7 mm 7 W LED gives out as much light as a conventional 60 W bulb, or as much as a 15 W compact fluorescent tube. Of course, the white LED costs $60, but that’s down from $100, and is likely to drop further to $20 in two years, which would cut the payback period down to only about a year, potentially saving billions in energy usage and carbon emissions. Progress has been made on control of deposition, doping and lattice mismatch, but the big 2-3× step up in efficiencies comes from nanostructuring the surface, so more light can escape without being reflected back. The ability to fabricate at the nano level is giving a great increase in light output. Various approaches include a saw-toothed surface of microcones, a ZnO megacone bonded to each device, or imprinted diffraction gratings.
Bringing down costs depends on solving a bunch of manufacturing problems. It needs a scale up of the epi. Yields are terrible. Colors and power are really variable. The phosphor manufacturing is not under control. The epi is barely under control. And the chip fabrication varies; devices have different power levels and all sorts of white ranging from cold to warm in the same batch. Suppliers are ramping production cautiously to try to avoid overcapacity, which prevents equipment suppliers from being willing to devote too much attention to their issues. "Unless people say what the volume is going to be,” noted one supplier, “we’re not going to be able to spend much time to help them try to scale up."
Fuel Cell Stack Welding
03/01/2007Energy Efficiency – The New Killer App
07/19/2007Fuel Cell Technology Gets Practical
07/13/2007Profitless Prosperity
09/01/2008
