Four scientists spoke about materials science and technology at this symposium, highlighting the remarkable advances made in super-computers and silicon and semiconductor technology, as well as the development of new materials and advanced processing techniques for existing materials. These scientists found it difficult to predict what impact some of these advances will have in developing countries, which often lack the capital and resources required for fabrication of expensive new materials. Other new materials or new technologies are, however, already being used in developing countries. Semiconductor technology, for example, is the basis of satellite-linked instructional networks, and the artificial intelligence techniques used in expert systems applied to medicine and agriculture.

The first speaker, Dr. Edward Starke, dean of the School of Engineering and Applied Sciences at the University of Virginia, summarized the likely progress in the next 20 years in materials science and technology. He indicated that scientists will have to learn to do more with less when developing advanced materials and manufacturing technologies.

Materials are often the controlling factor in the development of new technologies, as, for example, in microelectronics, energy systems, and transportation systems. In the field of microelectronics new or improved materials will be developed for encapsulation of microcircuits, high-resolution etching, microchips, electromagnetic shielding, chemical switching, and optoelectronics. In the energy field, nuclear power systems will require more radiation-tolerant materials, effective systems for waste encapsulation, and materials for fusion reactors. Special challenges will be faced in the transportation sector, where aerospace systems will need lighter weight materials. Improvements in engine technology will be based on ceramics and metal matrix composites, direct interfacing between electronic and mechanical systems, and the increasing use of nonflammable materials. The automotive industry will use polymer composite structures with high productivity' and there will be a move toward greater use of lightweight, corrosion-resistant structures. More efficient automotive engines will run at higher temperatures, calling for ceramic parts in coatings. And, finally, if there is a move toward electric automobiles, lightweight batteries that are much more efficient and longer lasting than those available today must be developed.

Advanced Processing Technologies

According to Starke, "improvements in material performance depend as much on advances in processing as they do on the introduction of new compositions." He believes that industrial engineers should take another look at old processes to evaluate how new advances in processing can make an old technology more efficient and cost effective. Improved material quality and advanced processing technology can improve reliability, he stressed. For example, improved material quality can have a significant impact on semiconductors, electro-optic materials, ferroelectrics, ferromagnetics, and fracture-resistance structural materials. Controlled improvements in processing and careful control of the processing can narrow the "envelope" of various properties of a material. In fact, Starke pointed out that a 10-20 percent shift in the positive direction of the "allowable value" can make as much of an impact as developing a new material.

Advanced processing technologies that have been developed over the last 10-20 years include laser photochemistry, which can be used to make, for example, silicon carbide for reinforced composites, and rapid solidification technology, which offers the opportunity to decrease segregation and to increase supersaturation in materials, and to actually make meta-stable materials that cannot be made using any normal casting technique. Powder metallurgical methods allow one to make a particulate and then rapidly consolidate it into a product that is almost in its final shape, thereby reducing manufacturing and engineering costs. Thermomechanical processing can be used to tailor the microstructure of a material, and therefore the final product, for a specific property - for example, resistance to fatigue cracks. Super-plastic forming is another recent advance. With this technology one can deform metals into shapes in the same way that one can deform plastic.

Other advances have been made with what might be considered old materials, noted Starke. For example, during the energy crisis about 10-15 years ago a big effort was made to lighten transportation vehicles by substituting either epoxy matrix composites or aluminum alloys for steel. Instead, dual-faced, high-strength steels were developed for use with thin gauges. Thus, it is projected that steel will continue to be used by the automotive industry. The amount of steel used by that industry has remained relatively constant over the last few years even though vehicles have become lighter.

During the next 20 years the metal industry needs to develop an on-idle identification system for materials in process. For example, an on-lane surface defect identification system for sheet and plate materials while they are being processed could identify defects in a material so that they can be eliminated or modified during the processing operation. On-line measurement of physical and mechanical properties will allow slight adjustments in the processing to control the properties of a material while it is being processed, instead of waiting to measure the material after production, and possibly scrapping it if it does not meet specifications.

"By using advanced materials and processing technologies," concluded Starke, "one can develop complex engineering structures that will operate over a longer and more reliable lifetime and reduce the quality of scarce materials. Once we understand the role that certain elements have in materials, we can essentially tailor a material's properties by tailoring its microstructure. "

New Materials

Some of the newly developed materials have great promise. For example, lithium, the lightest metal known, can lower the density of an aluminum alloy approximately 3 percent for every weight percent that is added, and it can increase the material's strength. Thus, the substitution of aluminum lithium alloys for the aluminum alloys used currently will mean about 10 percent savings in weight. In the airline industry, for example, this can make a big difference. It has been estimated that the use of aluminum lithium alloys to reduce the weight of an aircraft by about 10 percent will save 15-20 gallons of fuel per year per pound of weight saved (about 35,000 pounds for a Boeing 747). Thus, an airline can save millions of dollars each year in fuel costs.

Most of the powder metallurgy alloys are being developed for either their greater strength or their resistance to high temperatures. The high-temperature titanium alloys and rapid solidified aluminum alloys are 10-20 percent stronger and more temperature resistant than the aluminum alloys available today.

Fibrous Composites

Modern fibrous materials are composed of many kinds of fibers - graphite, boron, silicon carbide, and fiberglass - which designers are able to put together in different forms: chopped fiber, continuous fiber, or prepreg, in which some resin polymer is preimpregnated into the system. Even fabrics can be made of composite materials.

The individual layers of a structure can then be oriented in different directions to optimize the properties for the designed applications - perhaps strength, stiffness, or dimensional stability. The properties of a structural component are all a function of the individual layers used, the kind of fiber used, the matrix material surrounding the fiber (perhaps a polymer or metal), the orientation of the fibers in each of the layers, and the stacking sequence. According to Dr. Carl T. Herakovich¹, professor of engineering science and mechanics at Virginia Polytechnic Institute and State University, the fact that composites are lightweight is probably the biggest reason for the widespread interest in them over the last 25 years. Other engineering advantages of composites are their stiffness, strength, and dimensional stability (the coefficient of thermal expansion of some of these materials is actually zero). Moreover, composites are fatique resistant, crease resistant, and damage tolerant. They also have low friction and, a very important feature, they can be designed.

Some of the economic advantages of composites are their very low maintenance costs (owing to their resistance to corrosion), the low capital investment required (particularly for the manufacturing facility), their small number of parts (especially when using adhesive bonding techniques as opposed to bolts or rivets), and their low fabrication costs. One disadvantage of composites is that there are not enough experienced designers to work with these materials. This situation will change over time, however. The current high per pound cost of composites will also drop.

The applications of composites are wide ranging, from beams and filaments to commercial aircraft (many composites are fire resistant), from helicopters to hulls for America's Cup contenders.

Silicon and Semiconductor Technology

According to Mr. J. Franklin Mayo-Wells, staff assistant for technical coordination/operations at the Center for Electronics and Electrical Engineering, U.S. National Bureau of Standards, the many exciting advances recently made in electronic materials relate to a wide range of components, devices, and systems. For example, new materials are being used for electromagnetic shielding, optical fibers, and electromagnetic antennas. "In each case," he noted, "the new materials provide some significant advantage in reliability, performance, or cost compared to materials now in common use."

Semiconductor technology, which underlies many of the advances in electronics and electrical engineering, is largely silicon semiconductor technology, although other materials are used occasionally. Materials are vitally important in the complex fabrication of a modern integrated circuit, and all must be of the highest quality and purity. Silicon has transformed life in the twentieth century because it is used to construct circuits that provide control and calculation functions. With the control function we can switch electric power to regulate power levels and routing and, perhaps most important, implement the results of the calculation function.

The transistor has remained the heart of most semiconductor devices, including power devices, digital devices such as microprocessors and memory, and microwave (very high-frequency) devices such as amplifiers and receivers. Power transistors, explained Mayo-Wells, implement the control function, while signal transistors and memory, in the biggest and fastest computers, and microprocessors and memory, in all the rest, implement the calculation function. Without the control-microprocessor-memory combination many of the devices that make life easier would not be possible. For example, this combination regulates the ignition and fuel-air flows to automobile engines, regulates the heating and cooling in vehicles and engines, and controls household appliances and power tools. And virtually all modern manufacturing technology depends on it. The microprocessor-memory combination is the heart of personal computers, it plays important roles in communication systems, and it is a building block for many electronic systems.

Although Mayo-Wells finds the future of semiconductor technology "at best cloudy," the pace of semiconductor development has not abated, and memory technology is at the leading edge. The 1 million-bit dynamic random-access memory ([M-bit DRAM) is appearing in the newest personal computers, and attendees at a 1987 conference described fabricated 4M-bit devices and discussed the design of 16M-bit chips. Moreover, some very challenging fabrication requirements are emerging in the form of semiconductor structures for high-speed devices and optical communications.

Thus, pointed out Mayo-Wells, "it is very difficult to make more than the most generalized predictions for electronics 10 years down the pike." Silicon technology will probably still dominate, he predicted, given the wealth of existing experience in silicon semiconductor processing technology. Gallium arsenide semiconductor devices are not as prevalent because the practical speed advantage that this material offers over silicon has steadily eroded as the ability to construct increasingly smaller silicon devices with greater control of the electrical and physical properties has grown. Nonetheless, more gallium arsenide and other compound semiconductor devices will probably appear commercially as their processing technology catches up with that of silicon. The speed advantage of gallium arsenide in the context of microwave circuits has already made it the technology of choice for some applications, principally those related to communications and defence. "Josephson-junction technology," noted Mayo-Wells, "formerly restricted to operation at the 4 Kelvin temperature of liquid helium, offers advanced performance with the technologically thrilling development of superconductors that can be cooled to superconducting state above the 77 Kelvin temperature of liquid nitrogen."

Examples of other exciting semiconductor materials that may well be in the marketplace by the turn of the century are the diamond films formed by a chemical vapor deposition process at low temperatures and pressures. "These apparently offer potential for semiconductor devices of high performance, coupled with high-temperature operation and excellent radiation resistance compared to silicon," observed Mayo-Wells. The commercial apparatus for forming these films on a substrate already exists.

Superconductors

Although the enormous publicity that this technology has recently received might lead some people to believe that it was just invented, Dutch scientist Kammerlingh Onnes actually discovered superconductivity in 1911. Superconductors offer no resistance to the passage of electricity, which means that currents can flow through superconducting wires with absolutely no loss. Thus, electricity can be generated and distributed much more efficiently than before.

The materials used currently, however, do not superconduct at anywhere near room temperature, they are not durable, and they are expensive to manufacture. After pointing out these constraints to the widespread use of superconductivity, Dr. Richard E. Harris, leader of the cryoelectronic metrology group at the U.S. National Bureau of Standards (NBS), offered more details. Until 1986, the highest temperature at which superconductivity was observed was 23.5 Kelvin (-417° Fahrenheit). Such a temperature can only be achieved using liquefied helium, which is technically difficult to prepare and absorbs very little heat before boiling away. Furthermore, the wires used today for superconductivity are not very ductile and must be handled carefully to avoid damage. Finally, because superconducting wires must be formed in a copper matrix to work properly, their manufacture is very expensive.

Nevertheless, numerous rather large applications of superconductors have been made, explained Harris. For example, superconducting magnets are used at Fermilab, which has the most powerful accelerator for high-energy physics research in the United States. A superconducting magnet is also used for magnetic resonance imaging (MRI), a new technique for imaging soft tissue in the human body.

Electronic applications of superconductors include the superconducting quantum interference device (SQUID), which is the most sensitive means of measuring magnetic fields. SQUIDs have been used to sense the magnetic fields of the heart and the brain, and they are able to locate problems better than conventional electrocardiograms and electroencephalograms. They have also been used in geological experiments that might lead to new prospecting techniques. Researchers at NBS have already made the first SQUID that works in liquid nitrogen (that is, without liquid helium), and they are now working toward measuring a magnetocardiogram with it. Finally, a new superconducting device for making fast electrical measurements - a time domain reflec- tometer - is also available commercially. It is three times faster than any other completely electrical system.

Although other applications of superconductivity - such as a Japanese test train that floats above a track on superconducting magnets - have captured the imagination of the scientific community and the public, they have not yet made it into production. "Highly efficient generators and power transmission lines have been tested, but none are in use," noted Harris.

One of the difficulties with these unrealized applications is the need for refrigeration. Refrigeration for liquid helium is expensive, making only the largest scale applications economically feasible. Thus, the projections have been for big generators, big computers, and big transmission lines.

As for new developments in superconductivity, the maximum transition temperature for conductivity was at 23.5 Kelvin for over a decade, but in January 1986 researchers at IBM's Zurich research laboratory were able to push this temperature up to about 30 Kelvin using a compound of lanthanum, barium, copper, and oxygen. And in January 1987 workers at the University of Houston changed the chemical elements to yttrium, barium' copper, and oxygen, and achieved a transition temperature of above 90 Kelvin. Thus? in achieving superconductivity, liquid nitrogen could be used for cooling since it makes cooling much easier and since a given amount of liquid nitrogen can absorb about 60 times more heat before boiling away than the same amount of liquid helium.

In concluding, Harris noted that many of the incredible applications reported in the newspapers may come about, but not overnight. It may be that reducing the cooling requirements to those of liquid nitrogen will not be sufficient to make some of the applications economically attractive. And it may be that the best uses of superconductivity have not yet come out of the laboratory. It is hoped that when practical applications do become possible, perhaps in the next 10-20 years, the marginally reduced costs of electricity generation will benefit the developing countries.