Two Oak Ridge technologies may make superconducting wire cost-competitive.
Thanks to two ORNL technologies, the second-generation, high-temperature superconducting (HTS) wire that American Superconductor Corporation plans to produce at its pilot plant in 2006 may prove to be competitive in cost with conventional copper wire. Twenty years after high-temperature superconductivity was discovered, HTS wires may be used for underground transmission cables, oil-free transformers, superconducting magnetic-energy storage units, fault-current limiters, high-efficiency motors, and compact generators.
The first generation of high-temperature oxide superconductors
relied on bismuth strontium calcium copper oxide in
a silver tube. ORNL has worked with Southwire of Georgia, to
demonstrate the use of 1G wires in the company's test power
cables. The second generation, or 2G HTS wire, uses yttrium
barium copper oxide (YBCO).
For the substrate of the 2G wire, American Superconductor has chosen a textured nickel-tungsten alloy fabricated by ORNL's rolling-assisted biaxially textured substrates (RABiTS) process. In this process developed by a team in ORNL's Metals and Ceramics Division, low-cost nickel replaces the high-purity silver of 1G wire, enabling the fabrication of 2G wires at a lower cost. The cost of 1G wire is 30 times that of copper wire. RABiTS templates provide a texture that is transferred through buffer layers to the YBCO superconductor. American Superconductor currently uses a reactive sputtering system to deposit three buffer layers on RABiTS—yttrium oxide, yttrium-stabilized zirconium, and cerium oxide—followed by the superconductor.
Nickel can poison the YBCO layer, destroying the superconductive properties. To transfer the texture from the template to the superconductor while preventing the diffusion of nickel metal to the YBCO film, buffer layers are needed. These insulating layers also reduce both alternating current losses and the thermal expansion mismatch between the crystal lattices of the substrate and the superconductor.
American Superconductor uses this architecture in the 4-centimeter (cm)-wide tapes that the company produces and slits into 4.4-millimeter (mm)-wide, 2G HTS wires. The wires are then topped with a thin layer of silver and sandwiched between protective copper layers, the only parts of the wire exposed to the liquid nitrogen coolant.
Parans Paranthaman, a materials chemist in ORNL's Chemical Sciences Division, has led the development of a wet chemical process to replace reactive sputtering in the deposition of the buffer layers on 2G HTS wire. American Superconductor has adopted ORNL's patented metal-organic deposition (MOD) technique to scale up their YBCO superconductor activities. They will switch to MOD buffers when Paranthaman's team makes a wire that achieves 250 amperes/cm. In 2005, the ORNL chemists used MOD to develop a superconducting wire with a current density of 200 amps/cm.
In the ORNL process, a nickel-tungsten tape on a spool is pulled through a chemical bath, much like a plastic substrate used to make movie film. The metal tape is dipped in or slot-die coated with alcohol-based solutions of lanthanum zirconium oxide, then cerium oxide, and finally a blue YBCO solution. When the tape is heated in a furnace between each coating, the alcohol evaporates, the organic compounds decompose, and a superconducting tape with only two buffer layers emerges.
"We use only 10 to 15 milliliters of solution to coat a 4- cm tape and recover all the leftover chemicals," Paranthaman says. "In reactive sputtering, only 10 to 20 percent of the material is used and the rest is deposited throughout the vacuum chamber. A reactive sputtering system for making hundreds of meters of wire costs $5 million, whereas an annealing furnace for the chemical coating process costs only about $250,000."
Using both RABiTS and MOD, American Superconductor has the potential to make a low-cost 2G HTS wire that outperforms copper wire. Noting that 20 years were required for optical fibers to move from concept to commercialization, Paranthaman states the obvious, "We have come a long way."
Web site provided by Oak Ridge National Laboratory's Communications and External Relations