Frustrations Of A Ceramicist

While nuclear magnetic resonance (NMR) equipment represents the biggest market for superconductivity technology, more than 60% of the current technology is applied there, the flurry of activity in the new high-temperature superconductors has not yet translated to state-of-the art changes in NMR. In fact, researchers face three major hurdles in the race to realize the potential of these promising ceramics. For most of the past 15 months, attention has centered on YBa2Cu3O7 usually referred to as "1-2-3," a ceramic that becomes superconducting below 93 K. Exciting as the research has been, the material poses considerable impediments to commercial users. First, 1-2-3 in most bulk samples has a relatively low capacity to carry current (critical current density). It is difficult to create and, perhaps worst of all, shows low strength and little ductility. The low critical current density comes from a phenomenon known as “weak-link” behavior. In a polycrystalline sample, individual grains or single crystals are superconducting with a relatively high critical current density, more than a million amps per cm2. But the grain boundary creates a barrier to the transport of supercurrent. Much of the present processing research looks toward eliminating weak-link behavior, and the first step has been to try to identify its source. Researchers have suggested a number of theories on the origin of weak-link behavior, misorientation of crystals, different chemistry at the grain barriers or foreign atoms or unreacted components of the material, such as copper oxide, concentrated there. Misorientation is a popular theory because researchers already know that each crystal assumes a different orientation when the ceramic is heated. If the grains in a polycrystalline sample are randomly oriented, a misalignment at the grain boundaries forces electrons to change direction and limits the transfer of charge between grains. A number of laboratories have been trying to align the grains by mechanical or other means. Many laboratories prepare bulk ceramics by traditional techniques. A researcher starts with the basic fine powder, mills it, adds a binder and presses it into a "green" compact. At this stage, material goes into a kiln for firing. In the intense heat, the loosely packed particles in the green compact move closer together, or "densify," yielding a material almost twice as dense as porcelain or china. Smooth, black with a dull luster, the ceramic feels unexpectedly heavy. In the 1-2-3 system, when the material densifies, a tiny amount of it liquifies at the grain boundaries. The liquid may be a mixture of barium cuprate and unreacted copper oxide. At the sintering temperature, usually about 900 degrees C, the liquid may completely coat all of the individual particles. The liquid, which is not superconducting, greatly reduces the ability of the material to carry current. Some research suggests that by working with a small particle size, sintering temperatures can be reduced and the extent of the liquid phase may be minimized. As the 1-2-3 material cools from the firing temperature, its crystal structure changes. This change in the lattice creates large internal stresses in the material, which may lead to microcracking and even gross failure. Researchers are finding that by using slow cooling and, again, small grain size, they may help to reduce internal stress. These new superconductors are brittle ceramics and generally do not bend under pressure; They cannot be drawn the way a metal wire is formed. As with most other ceramics, they must be formed in the green state and fired to final shape and dimensions. The 1-2-3 material cannot be cofired with the usual organic insulators because the insulation simply cannot stand the heat. At firing temperature of 900 degrees C, the insulation degrades excessively. Furthermore, this ceramic cannot be fired in the presence of metallic copper—the stabilizer used in most applications, because metallic copper robs the superconductor of oxygen. The YBa2Cu3O7 is no longer superconducting when it loses even a fraction of the oxygen. The 1-2-3 compound does appear relatively stable with respect to noble metals, and a number of groups have reported success in coprocessing with silver. However, while silver is less expensive than other noble metals, its cost is probably prohibitive for most large-scale applications. Individual layers and monofilaments of superconducting 1-2-3 have been formed by a number of methods including tape casting and extrusion. In the case of extrusion, the viscosity is high, resulting in a plastic or putty-like mass which can be formed in a filament under-pressure in a die. In tape casting, the viscosity is somewhat lower and the mix can be spread like paint. Either of these can be laminated with a ceramic insulator be- fore firing, as long as the shrinkage of the insulation and the superconductor are matched to avoid fracturing the material. The alternative to green forming and firing to final shape is the creation of a thin wire that can be flexed after firing. Wire diameters of a few micrometers may be required. At present, most ceramic-forming processes are limited to diameters greater than many tens of microns. In recent weeks, two new systems have been discovered. Both are still mixed oxide systems with copper-oxygen planes. One of these contains bismuth and one contains thallium. Each shows promise for improving properties over the 1-2-3 material. The bismuth system may be less brittle than the 1-2-3, but has been more difficult to fabricate. In the thallium system, higher critical current densities may be reached in bulk samples, but the toxicity of thallium has intimidated many investigators. While the 1-2-3 material may be a long way from many practical applications, significant improve ment in properties has been achieved over the past year. Critical current density in bulk ceramic samples has risen from less than 100 amps per cm2 to over 1,000. amps per cm2, with a few investigators reporting critical current density in excess of 10,000 amps per cm2. We have not yet given up hope for 1-2-3. But as we work to overcome its limitations, we continue the search for other systems with better properties. In the meantime, NMR just has. to wait. Roger Poeppel is manager of the ceramics section in the materials and components technology division at Argonne National Laboratory.

Frustrations Of A Ceramicist

Interested in reading more?

Become a Member of