or ORNL's Douglas Mashburn, 1987 was a particularly exciting year. The world had become aware of the promise of new ceramics that conduct electricity without resistance at the temperature of liquid nitrogen. He and many other scientists at ORNL and elsewhere worked frantically in their laboratories to make and test these new ceramics. The news media proclaimed that high-temperature superconducting materials would improve the efficiency and lower the cost of underground power transmission lines for cities. Because the electricity they carry isn't partially wasted as heat, superconducting wires are believed to be potentially more economical than conventional copper cables. The appeal of these new materials is that they are superconductive when chilled only by liquid nitrogen, a much less expensive refrigerant than the even colder liquid helium needed to cool low-temperature superconductors such as niobium-titanium alloys. The new materials have also been considered promising for making smaller and more efficient motors, microelectronic circuits that would speed up computers, and compact magnets powerful enough to levitate and propel high-speed trains. Although the timetable for these wonders may have been very optimistic, the possibilities are still very real.
In May 1987, Mashburn, a laser expert then in the Solid State Division, learned that the high-temperature superconducting material yttrium-barium-copper oxide (YBCO) had been made as a thin film using an excimer laser. By September, he and his ORNL associates had demonstrated that indeed thin films of these materials could be made using a process called laser ablation. By establishing a presence in the field of superconducting thin films, ORNL in 1988 became one of three national laboratories designated by the Department of Energy to have a High-Temperature Superconductivity Pilot Center (now called a Technology Center). As a result, under pilot center cooperative agreements, ORNL scientists began collaborating with researchers from industrial firms to develop processes and products of commercial value.
In 1993, the ORNL High-Temperature Superconductivity Technology Center managed by Bob Hawsey boasted its first commercial product--a software-driven, pulsed-laser system for depositing films of a specified composition and thickness. Called the Automated Multilayer Deposition Accessory (AMDA), the system was developed in collaboration with the private firm Neocera, Inc., of College Park, Maryland, in a cooperative research and development agreement (CRADA). ORNL's technical contribution was provided by Mashburn, a researcher now in ORNL's Engineering Technology Division. He helped Neocera researchers incorporate features from two of his inventions into their laser deposition system, resulting in the AMDA device. One invention was a rapid target-switching scheme that he conceived in 1987, and the second was a dual-beam laser arrangement conceived in 1990.
Mashburn has continued to develop his own laser deposition system, similar to AMDA but offering more capabilities. It may solve problems impeding commercial production of high-quality superconducting films, including films thick enough to be made into conductors for magnets and motors. His system offers significant new advances in the deposition of crystalline films by laser. It shows promise for creating defect-free films and improved materials by constructing them an atomic layer at a time. "Perhaps even more exciting," Mashburn says, "the new system could be developed into a 'materials workstation' that would open the door to exploring dozens of new materials."
In the mid-1980s, a Rutgers University Bellcore group led by T. Venkatesan explored excimer laser ablation as a means of depositing materials in the manufacture of integrated circuits. In early 1987, shortly after the discovery of YBCO, Venkatesan's group tried the process on a YBCO pellet and produced a superconducting film. This result came to Mashburn's attention, and he pressed his management to let him proceed with a similar experiment. As luck would have it, Mashburn had brought with him to the Solid State Division an excimer laser and other critical equipment needed to construct the apparatus, and his colleague Lynn Boatner had the special materials needed. Mashburn had salvaged the equipment from the Molecular Laser Isotope Separation (MLIS) Program at the Oak Ridge Gaseous Diffusion Plant, where he had worked on the process from almost its inception until its cancellation in 1982. MLIS was an advanced method for producing enriched uranium for nuclear power stations.
During the MLIS work, Mashburn was on assignment for several months at a time at Los Alamos National Laboratory. The excimer laser, a high-power ultraviolet pulsed laser, had just been invented at Los Alamos in 1976, when Mashburn joined the team that developed it and adapted it to the MLIS process. During this work he first encountered laser ablation when he unintentionally vaporized optical components for handling high-energy laser beams, damaging expensive parts. Also during this period, he developed his first laser ablation device, a "laser pulse detector" patented in 1981. After cancellation of the MLIS project, Mashburn joined the Atomic Vapor Laser Isotope Separation (AVLIS) team at Lawrence Livermore National Laboratory, where he helped design and install the copper vapor laser system for the AVLIS demonstration facility completed there in 1985. His first big task in ORNL's Solid State Division had been to design and install an excimer laser-assisted chemical vapor deposition system. By 1987, he had accumulated considerable experience on laser applications and insights into the internal workings of lasers.
Boatner, a crystallographer in the Solid State Division who was trying to grow single crystals of YBCO, was excited by Mashburn's proposal to produce superconducting films with laser ablation. Shortly after he heard about it, Boatner saw Venkatesan at a physics meeting and confirmed Mashburn's speculations. Convinced by new information brought back by Boatner, the Solid State Division management finally agreed to limited support of laser ablation experiments. That was all Mashburn needed.
In only 13 days in the summer of 1987, Mashburn and Dave Geohegan, a newly arrived postdoctoral fellow, built a laser ablation chamber as they rushed to compete in the super-conductivity sweepstakes. Then they donned their laser goggles, fired up the excimer laser, and watched as the glass chamber lit up with the ablation phenomenon. On one side of the chamber was a heater holding a strontium titanate substrate supplied by Boatner on which they hoped to form a thin-film YBCO superconductor. In the center of the chamber was the YBCO pellet made by Boatner's group. Coming in from below was the focused excimer laser beam.
In their experiments, pulses of ultraviolet light from the excimer laser entered the chamber from below and within 30 billionths of a second (nanoseconds) heated a spot on the target to over 12,000 degrees C (as they later determined). Atoms on the target surface became so superheated by the laser light that they literally exploded off the surface. As the ejected material traveled from the target to the substrate to make a film, its trail was visible as a glowing plume.
"These atoms are ionized, forming a dense plasma just above the surface of the solid," Mashburn says. "Laser ablation is a unique phenomenon in which a plasma at near-solid density explodes with energy greater than is present in any chemical explosion. The vaporized material deposits on the nearby substrate, which we heated to 500 degrees C to assist our film growth."
The scientists used a single-crystal wafer of strontium titanate as the substrate for the thin films because they suspected it would serve as a good template, forcing the atoms of the YBCO to line up in a neat, orderly pattern--that is, as an oriented crystal. After the film deposit was annealed in high-temperature, high-pressure oxygen, this atomic alignment did, in fact, occur. Their measurements verified that the substrate atoms imposed an order on the film layer just as a waffle iron impresses a grid pattern on waffle batter as it bakes. The results suggested that this technique might be superior to conventional deposition methods such as sputtering or thermal co-evaporation.
Among the many film samples Mashburn and his associates produced were the first superconducting YBCO films at ORNL. He reported his results in a scientific paper and at a meeting of the Materials Research Society. Color photographs of the glowing blue chamber made the covers of the ORNL Review, Mechanical Engineering, and the MRS Bulletin of the Materials Research Society.
Thin films of high-temperature superconducting (HTS) material showed more early promise for applications than the bulk HTS materials. A major use of thin films is expected to be in tiny electronic components not exposed to high magnetic fields. By growing thin films on templates of other crystalline materials, the films' crystal grains can be lined up to form a continuous path for current, thereby avoiding the "weak link," or misaligned grains, problem that limits the amount of electrical current bulk materials can carry.
According to Mashburn, routine thin films several years ago could carry 5 million amperes per square centimeter and the best films could hold 50 million amperes if cooled to 4 Kelvins. By comparison, the best bulk material could carry only 50,000 amperes per square centimeter. The ability of thin films to carry current was at least 1000 times better than that of the bulk materials. Now thin HTS films are being used in a commercial product introduced in 1992 by Conductus, Inc., of Sunnyvale, California, to detect weak magnetic fields. The technology for making thin films has not yet been perfected, however. Hurdles remain to be overcome to make possible mass production of films of uniform composition and a range of thicknesses for various uses, including superconducting wires. Mashburn says that his and other advances in laser ablation may solve the problems currently blocking achievement of these goals.
In recalling those heady days of 1987, Mashburn says, "When we started ablating these materials, we were exhilarated by the success of our first try. But we soon realized that the films weren't perfect. For better understanding, we took each film to ORNL's Surface Modification and Characterization Collaborative Research Laboratory to determine the composition. In searching for ways to get the best ratios of elements, we changed the laser power, the gas pressure in the vessel, the substrate temperature, and the composition of the target.
"We had to wait a couple of days to get information back on each sample film. After a dozen or so films, we could project that in a brute force approach, it might take 10,000 samples to pin down the optimum conditions for YBCO deposition. At a rate of days per sample, it could take years to figure out how to make perfect films."
Bothered by this situation, Mashburn kept saying to himself, "There has to be a better and faster way to do this exploration." Then he saw the light. The approach, later named LARES, was conceived in the fall of 1987. It drew not only on his HTS film experience but also on his previously gained insights about the internal operation of lasers. However, he had to wait several years for funding to implement his innovative approach.
Meanwhile, numerous groups of American scientists took up the trial-and-error approach to making superconducting films. "Every team in the field," Mashburn says, "was depositing oxygen-deficient films by evaporation, ablation, sputtering, or other methods. Then, films had to be oxidized before they could be characterized. Because there was no reasonable way to monitor the incomplete precursor film in situ, it took days to obtain results for each sample. Over the next few years, the federal government ultimately paid for thousands of samples at a cost of several work days each. It was a long, laborious, expensive way to perfect the process of making superconducting films."
Mashburn envisioned a device that combined a laser, a rotating target drum on which separate oxide targets would be mounted, and a computer-controller that would select (or skip) each target and adjust the laser energy needed to ablate it as it spun by several times each second. Such an automated device would switch targets for pulsed laser ablation much faster than manually operated mechanical shutters. It could not only control stoichiometry but also add buffer layers and contact layers or construct complex multilayered structures with ease. Mashburn began making the drawings and later built his invention, which he called LARES--laser ablation by rapidly exchanged sources.
"LARES gives you control of film composition," says Mashburn, "and you can quickly switch the composition from one type of atom to another within a film." This "atomically abrupt" switch can be made by automatically changing the power of the laser pulse and timing it so that it precisely ablates a material different from the previous target just zapped on the rotating drum.
"This is easy," Mashburn says, "because the time uncertainty between the triggering of the excimer laser pulse and its predicted arrival at the target is just a few nanoseconds--so short that with the proper setup such a laser could tattoo a bullet in flight!"
How could he use LARES to set the composition of a superconducting film?
"First, we must obtain high-purity targets of each element or oxide to be incorporated in the film, mount them on the target drum, and enter their position into the controller. Now, we calibrate the device by temporarily replacing the substrate with a crystal rate monitor," he says. "Then, after chamber evacuation, the laser is programmed to fire several times, at a fixed power level, at only one of the targets. A film of the target material is deposited on the crystal rate monitor. The monitor gives an electrical readout of the mass deposited, which is converted into atomic units or thickness using the composition data. That's why we must use single-element or stable compound targets of predetermined composition. The laser energy and resulting deposition rate are recorded in the computer. The procedure is repeated for a range of energy levels. Then another target is selected and the entire procedure is repeated for it. When all the targets have been calibrated and the information has been stored as a table in the computer, the rate monitor is removed and the substrate is reinserted into the chamber.
"Now we have a system that can simply be programmed for the desired composition, and the controller will select the proper targets and fire the laser with the correct energy to deliver the needed amount of each target material. With newer lasers, this controller can make the laser power dance up and down in a precise manner hundreds of times per second. This action happens too fast for the eye to follow. Such thorough control of the laser firing, including skipping a target, enables precise regulation of the film composition. The speed and flexibility of this system will allow researchers to do material composition explorations at least 10 times faster than previously."
"We found that as the target wears down, its surface develops microscopic fingers, also called texturing," Mashburn says. "These fingers shift the direction of the plume toward the beam. The first solution to this problem was to slowly scan the laser beam up and down the diameter of a rotating disk target so that the top half of the scan negated the texturing effect created by the bottom half scan. It works well for a single target setup. Unfortunately, this solution does not work for LARES, for here the beam-target drum intersection point must stay fixed."
In 1990 Mashburn conceived a dual-beam optical arrangement to solve the problem of target erosion in LARES. The laser beam coming straight at the target is split into two halves, one going up and one going down. A mirror in each half-beam directs it back toward the target so that the beams intersect at the target after coming from opposite sides of the target. Thus, the texturing tendency of one is simultaneously counteracted by the other, keeping the target smooth.
"The main advantage of such a balanced optical arrangement," says Mashburn, "is that the microscopic fingers do not form in the target. Instead, the target is actually polished. The smoother the pellet, the more likely it is that the desired film composition can be maintained. Additionally, for laser beams that do not have a flat intensity profile, the overlap also improves the beam uniformity."
Because the color and intensity of the laser ablation plumes are intricately related to the type, amount, and condition of material in transit, optical spectroscopy can be employed for precise noninvasive monitoring of the deposition process--a strong potential advantage of the laser ablation technique. "It can determine if the target surface has altered or the delivered laser energy has fallen," says Mashburn, "suggesting that the film composition will not be as expected." However, the spectroscope must always see the same area of the plume to make good comparisons. Because a fixed plume greatly simplifies this task, the dual-beam system makes such monitoring practical in LARES, which can even use the spectroscopy information in a feedback mode. Mashburn notes that he plans to add this capability to his advanced laser ablation system because "it tells you if your process has unintentionally changed and needs correction."
A similar but worse phenomenon affects the fabrication of films of thallium-based superconducting material. It has two volatile components--oxygen and thallium. Mashburn knew about this problem by 1988. It seemed impossible to solve.
"We needed low pressure for good film deposition," he says. "We needed high pressure for good crystal formation in the film. One condition seemed to exclude the other."
It took Mashburn three years to come up with a scheme that allows both low-pressure deposition and high-pressure growth of thin crystalline films to prevent their degradation. He called his invention "individually controlled environments for pulsed addition and crystallization" (ICEPAC). It accomplishes the impossible by taking advantage of the extremely quick nature of pulsed laser ablation. In their 1989 measurements he and Geohegan showed that the intense material pulse in laser ablation lasted only about 20 to 30 microseconds but deposited as much material as an evaporator could in 1 second. Thus, laser ablation is at least 10,000 times faster than evaporation.
"By connecting a laser deposition chamber with a chamber for controlling film growth to prevent the escape of volatile elements, a film can do time sharing of two environments in a very unequal way," Mashburn says. "The film spends 99% of its formative life in the growth environment, visiting the deposition environment only 1% of the time. The heart of this concept is the maintenance of two widely different environments dynamically connected to rapidly and repeatedly exchange the film between them."
In this scheme, Mashburn combines an ICEPAC film-growth chamber with a LARES laser ablation chamber in which dual beams vaporize material from targets on a rotating drum for deposition on a revolving substrate (see figure). The chambers are physically distinct but dynamically connected. The ICEPAC film growth chamber is filled with the volatile material, such as oxygen gas, kept at a desired uniform temperature, pressure, and composition that facilitate film growth. The substrate, on which a film is to be deposited, is placed on a cylindrical rotor so that the surface of the film faces outward and is exposed to the inner wall of the ICEPAC chamber containing the rotor. The revolving substrate passes a small opening between both chambers every revolution.
At the instant when the film is in front of the opening, the laser is fired and a plume of nonvolatile material shoots through the opening to add a layer of atoms to the film. By adjustment of the ablation conditions, each additional layer can be varied in thickness from one atom to many atoms. The laser-deposited layers on the substrate react with gas in the ICEPAC chamber as the sample is carried around. The film sample is at the opening long enough for deposition of nonvolatile material; however, its presence there is too brief (1% of the time) for a significant amount of the film's volatile elements to escape. The film sample is repeatedly and sequentially exposed at predetermined times to the material from both LARES and ICEPAC.
The dual-beam system is essential to operation of ICEPAC, Mashburn says, because it helps keep the plume aimed at the opening. Because the opening represents 1% of the ICEPAC chamber circumference, a fixed plume position is needed to ensure that the ablated material is deposited on the substrate.
"Current laser ablation practice uses thousands of pulses to grow typical films of a few hundred nanometers in thickness," Mashburn says. "Thus, there is ample opportunity to add evenly to the film volatile elements such as thallium or oxygen from their vapors in the film-growth chamber. A longer time for layer crystallization can be allowed by skipping the laser firing on the next revolution, or series of revolutions, of the rotor. This technique would permit the growth time for a 300-nanometer film to be varied from 10 seconds to many hours for research studies on film growth mechanisms. The ratio of volatile elements in the film can be changed by adjusting the partial pressures of those elements in the film-growth chamber."
Manufacturers of advanced flat-panel displays for notebook computers, high-definition television, and other uses have begun exploring ways to employ Mashburn's laser ablation technology. OSRAM Sylvania and Planar Systems have asked him to conduct laser ablation experiments on phosphors to make them more suitable for application to flat-panel displays. "In one case," he says, "we plan to use LARES to construct a brand new phosphor material--an artificial material made by multilayering. If this fabrication method is successful, it could revolutionize the lighting industry as well as the display industry." Two CRADA proposals in this area have been submitted and more are in process.
A small startup company, which makes super-sensitive charge-coupled device (CCD) sensors for astronomical telescopes, is considering using LARES to fabricate its CCD sensors. Finally, an established semiconductor fabrication equipment supplier is also considering the potential of LARES to meet its needs.
Some 4000 known inorganic materials exist in the world. Most of these are made by wet chemistry or some variation of mixing elements together and heating them up. The problem with the latter approach is that some materials that might be possible are never recovered because certain elements are segregated out during the heating process.
Mashburn has conceived of a "materials workstation" that would combine LARES, the dual-beam scheme, and ICEPAC. "Such a setup" he claims, "will open the doors for the exploration and discovery of many new materials, perhaps as many as 100,000." These materials could be made one layer of atoms at a time. Such artificially layered materials could include new multilayered structures called superlattices, which have special properties for electronic and photonic applications.
One of the goals of ORNL in carrying out DOE's high-performance computing and communications initiative is to make the massive number of complex calculations required to design new alloys and other materials. "If the researchers at the Laboratory's Center for Computer Sciences calculate that a new, presumably complex, material would be stable once formed and might have desirable properties, such as superconductivity," Mashburn says, "we could use such a materials workstation to rather quickly construct the material as a thin film."
For such a sophisticated materials workstation, Mashburn would incorporate noninvasive technology (such as Raman spectroscopy) to directly monitor the film as it grows. He might also add an artificial intelligence program that would "learn" the compositional and structural features that make a good film from the spectroscopic results. The program could guide the researchers in producing new stable materials that possess the desired properties.
Already, Mashburn has met the challenge of turning his concepts into actual working devices, refined versions of which may be designed and built. If sufficient funding is available, the world may someday learn of promising new materials that exceed our dreams and are made in Mashburn's laser machine.
[ Next article | Search | Mail | Contents | Review Home Page | ORNL Home Page ]