Doing The Impossible
WITH THE PASSAGE OF THE CLEAN AIR Act of 1970, Congress threw down a gauntlet similar in spirit to President John F. Kennedy’s 1961 challenge to put a man on the moon before the end of the decade. Both were bold strokes that placed a burden squarely on the shoulders of the nation’s scientists and engineers. And both looked impossible.
In the race to the moon the United States had only recently developed a rocket that could get off the launch pad with no cargo, let alone make it to the moon with humans aboard. In the race to clean up the air, no one knew if the goal, a 90 percent reduction in automobile emissions from pre1968 levels by the 1975 model year, was even remotely possible. Moreover, Americans loved their thirsty automobiles; the petroleum industry depended on toxic lead in gasoline to increase octane; and automakers didn’t want to add another costly device, especially one that would interfere with the cars’ performance. As Ernest Starkman, vice president in charge of the environmental-activities staff of General Motors, put it, “The cleaner the car is from a pollution standpoint, the harder it is to make it run well.”
Like the race to the moon, this would take a crash program of research and development. The 1975 automobiles would be introduced in September 1974, so there would have to be a factory running at full speed turning out millions of the emissionscleaning devices before then. The car companies would have to start tooling up for manufacturing these models early in 1974, so all specifications would have to be finalized late in 1973, meaning that all development and testing would have to be done by earlier that year. The makers had two or three years to go from basic research to manufacturing. At Corning Glass Works, one of the principal competitors in the race, the time between basic research and mass production of complex products was typically around 17 years.
William Ruckelshaus, who in 1970 was named the first Administrator (i.e., director) of the Environmental Protection Agency, later admitted that the project was presumptuous from the start. “We thought we had technologies that could control pollutants, keeping them below threshold levels at a reasonable cost, and that the only things missing in the equation were national standards and a strong enforcement effort,” he said in 1985. “All of the nation’s early environmental laws reflected these assumptions, and every one of these assumptions is wrong.”
Eugene Houdry, had he lived long enough to see it, would not have been surprised by the Clean Air Act. In fact, he would have welcomed it. Houdry was a French engineer who in the 1920s and 1930s had developed catalytic cracking, revolutionizing the petroleum industry. The Houdry process provided much of the aviation fuel that drove the Allied victory in World War II and the gasoline that powered the automobile boom after the war. In the postwar years he turned his own attention to reducing the damage that gasoline had caused, including (he believed, though the link remains unproven) an increase in lung cancer.
Unburned hydrocarbons from automobile exhaust gases were the primary culprit, he felt, and he had a solution. He began work on the problem in 1948 in Wayne, Pennsylvania, and founded a company called Oxy-Catalyst, Inc., to develop oxidizing catalysts to clean up exhaust gases. The main office building of Oxy-Catalyst was a converted ballroom, and the laboratory was an adjacent stable. Houdry lived in a house about a hundred feet away, so he could experiment with his ideas around the clock. In these humble surroundings he carried out research on catalyst systems that would oxidize carbon monoxide into carbon dioxide and convert unburned hydrocarbons into harmless carbon dioxide and water.
BY 1959 HE HAD DEVELOPED A “CATA lytic muffler” that would fit into the exhaust system of an automobile. It consisted of a rectangular metal box containing 71 porcelain rods arranged in staggered rows among which exhaust gases would pass. The rods were coated with high-surface-area alumina containing tiny islands of platinum, which acted as an oxidation catalyst. “Put them on all cars and watch the lung cancer curve dip,” he boasted.
But there was little pressure yet from consumers to do anything about automobile emissions, and anyway, the lead in gasoline would have “poisoned” the catalyst, forming an alloy with the platinum atoms and preventing oxygen from bonding to them. Houdry was awarded a patent for his muffler in 1962 and died later that year at the age of 70. Rachel Carson published Silent Spring the same year, helping launch a grass-roots environmental movement that would bring about just the kind of pressure that Houdry had needed to create a demand for his product.
Carl Keith of Engelhard Industries (now the Engelhard Corporation), in New Jersey, was a great admirer of Houdry and shared his desire to solve problems with catalysis. In the early 1960s he and Engelhard became involved with 3M in trying to reduce the carbon monoxide emitted from forklifts in enclosed spaces like mines and warehouses, where it could quickly reach toxic levels.
3M had developed a much better material than Houdry’s porcelain rods to hold the platinum catalyst. It was a zirconia-mullite blend that could be made in corrugated sheets and then rolled up into a cylinder. The cylinder in effect formed a honeycomb that the exhaust gas could flow through, with abundant surface area available for catalytic conversion. Keith and his team at Engelhard put a layer of high-surface-area alumina with platinum onto the honeycomb to convert carbon monoxide. He was encouraged by his success, but he knew that the forklift market was not large enough to justify the costs of his research. “I always kept the possibility of developing this technology for the automobile in mind,” he says.
And he still faced the lead problem that had vexed Houdry. In a test in California using leaded gasoline, an experimental catalytic converter lasted for only 12,000 miles. “I was convinced that nobody was going to pay $150 every 12,000 miles to replace a catalytic converter,” he says. So he started testing with unleaded gas and propane and got fantastic results. Still, Engelhard’s main customers were petroleum companies, and the firm didn’t want to push too hard. The pressure would have to come from the government.
In the years since Silent Spring , some halting legislative steps had been taken, but by 1969 the smog over California freeways was so bad that the state instituted its own emissions standards and was the site of two conferences in November to discuss environmental problems, one held by UNESCO in San Francisco and one convened by Gov. Ronald Reagan in Los Angeles. Events then started to move swiftly.
On New Year’s Day 1970, President Nixon signed the National Environmental Policy Act, establishing a new Council on Environmental Quality. In his State of the Union Address in January he heralded the arrival of “a historic period when, by conscious choice, [we] transform our land into what we want it to become.” April 22 was the first Earth Day. The environmental movement had reached a critical mass, and politicians would have to pay attention. On July 9 Nixon submitted a plan to Congress calling for the creation of the Environmental Protection Agency. The agency opened on December 2, and Assistant Attorney General William D. Ruckelshaus became its first Administrator. Nine days later he issued orders to the mayors of Cleveland, Atlanta, and Detroit to clean up their waterways in six months or face court action. He wanted to show that the new agency had teeth. In less than a month Congress passed and Nixon signed into law the Clean Air Act of 1970, giving the EPA the power to establish national air-quality standards. The act took dead aim at automobile emissions, requiring a device that would clean a car’s exhaust gases for at least 50,000 miles, and, as Russell E. Train, who succeeded Ruckelshaus as Administrator of the EPA in 1973, later put it, “leaving it up to the manufacturers to decide how in the world they were going to comply.” It also required the removal of lead from gasoline by 1975.
The manufacturers responded with a chorus of complaints and a flurry of research and engineering activity. Executives predicted that the industry would have to shut down after 1974; engineers agreed. But the government held firm. It was 1975 or bust.
The problem of emissions was attacked from both sides of the engine: before and after. At the inlet end, secondary air supplies to increase the available oxygen, heated precombustion chambers, carburetors that cracked gasoline into smaller hydrocarbons, and dual spark plugs fired in sequence in each cylinder were invented to promote complete combustion. At the outlet end, not only catalytic mufflers but also materials that would absorb the toxic gases were proposed; these materials would have to be changed regularly when the absorbents became saturated. Secondary combustion chambers at the exhaust side of each cylinder and schemes for recirculating exhaust gases for a second pass through the combustion cycle were also patented.
At Engelhard, the catalyst development team that had worked on the forklift project was reborn, this time with the government mandate it had always needed, on February 10, 1970, when President Nixon announced an early version of the upcoming clean-air standards. Carl Keith and John Mooney led a group of 33 people who dusted off their old data and prepared to work on automobiles.
Another major player turned out to be Corning Glass Works, in Corning, New York. The company had never delved into catalysis before and had no plans to, but it was involved with the automotive industry on several projects. One was the development of a regenerator wheel for turbine engines, to carry hot exhaust gases back to the inlet side of an engine. The material it was made from was very similar to the 3M honeycomb structure used on forklifts, and Corning engineers met frequently with auto-industry representatives to discuss the progress they were making.
At one meeting the representatives asked if the turbine regenerator’s cellular concept might be used in a catalyst system. The Corning people immediately grasped the magnitude of the question. Answering it in the affirmative would have enormous rewards, but the competition would be fierce. Auto-company engineers were sure they could develop some kind of mechanical engine modification that would solve the problem. As Rodney Bagley, a Corning engineer who became central in the catalytic-converter project, says, “Hanging a chemical reactor under a car was not something the auto companies wanted to do. They thought that it would be a short-term stopgap until they could design an engine that would reduce emissions using a precombustion chamber or some other gizmo.”
The Corning lab involved was led by Bill Armistead, who, Bagley recalls, responded, “We are going to do this, and we are going to commit the people and the money to do it.” Armistead soon had more than half his lab working on the catalytic-converter concept, aiming to have Corning make the entire device.
Ron Lewis, a young mineralogist, was part of a group given the job of finding the best support for a catalytic material. Another group, including Bob Farrauto, who would work for both Corning and Engelhard during the 1970s, investigated platinum, palladium, and some less expensive base metals, seeking the best catalyst to deposit on that support. Irwin Lachman, in the former group, began looking at compositions that would be ideal for the honeycomb substrate. “We went off in every single direction,” he remembers. “There were glass versions, ceramic versions, glass-ceramics.”
They also investigated every manufacturing process they could think of for the cellular structures; they produced sheets of glass with glass nibs sticking out, glass ridges to separate the sheets, stacks of extruded glass tubes in circular and triangular shapes. One engineer extruded alumina in a thin noodle that curled up into a structure like a robin’s nest—definitely flow-through, but very fragile. Through all this Bagley bore in mind that they were going to have to make millions of these devices and make them fast. “I looked at all the techniques they were trying, and I saw that none of them were going to be economically feasible,” he recalls. “I said, ‘Why not extrude them, as a cellular piece?’ Most of the people said it was impossible.”
EXTRUSION INVOLVES PUSHING A SOFT , viscous material, like a clay, through a die to form a desired shape. It is relatively simple when simple shapes are involved but looked dauntingly complex for this job, for many reasons. First, Bagley was talking about extruding a honeycomb structure with hundreds of cells per square inch. Making a metal die of that intricacy using traditional welding methods would require a huge amount of time, and if the honeycomb material was abrasive, it would wear out the die after only a few uses. Second, the walls of the honeycomb would have to be very thin, perhaps 0.01 inch thick. If the walls of the die wore by as little as 0.001 inch, the properties of the honeycomb would change and ruin the catalytic surface. Third, a complex extruded structure might collapse under its own weight, “like a wet newspaper,” in Bagley’s words, as it exited the die, closing all its channels. Fourth, if it miraculously didn’t collapse during extrusion, any attempt to then cut it to a desired length would surely collapse it. Fifth, how would you get enough heat into the middle of such a structure to dry it? And sixth, how could you fire it at high temperatures without its cracking and crumbling?
Undeterred by all these seeming impossibilities, Bagley believed extrusion could work. He was in a meeting with his boss Fred Bickford when the crucial idea popped into his head. “I said, ‘I know how to make a die,’” he recalls, and he jumped up to sketch it on the blackboard. Bickford couldn’t visualize what Bagley was trying to show him from the drawing, so Bagley went out into the lab and came back with a soft refractory brick and a diamond saw. Instead of welding separate pieces together, he would carve the die out of a single block. That was his big breakthrough. He began cutting slots in both ends of the brick until a three-dimensional model of his die emerged. In a few minutes he decided that a combination of thin, intersecting slots forming a honeycomb mesh at the output end and wide, round holes at the input end would be a better design.
After tinkering with this design a little, Bagley sent plans for it to the machine shop, and in a few days he had a prototype die, made from a solid piece of brass and drilled and cut to his specifications. The walls of the die were much too thick for the final product, and the first extruded honeycomb had to be quick-dried in a microwave oven to keep its shape, but at this point he was only trying to prove the concept.
“The first time we used it,” Bagley says, “it worked beautifully. Within a week we had essentially worked out the basics of what is used today.” He applied for a patent on his extrusion method for thinwalled honeycomb structures on November 9, 1971, and was awarded patent number 3,790,654 on February 5, 1974.
Shortly after Bagley’s initial success, Coming’s executives decided to take a huge gamble. They began building a factory to mass-produce monolithic honeycomb structures for catalytic converters in Erwin, a few miles down the road from the research center in Corning. They had no product yet and no guarantees that they would have one; a competitor might knock them out of the game, or the car companies might come up with a preferable mechanical solution. But they started building a factory anyway, with some employees working 100-hour weeks, and pressed on through a tremendous flood in June 1972 that heavily damaged an important testing laboratory.
Why take such a risk? The potential profit was great (first-year sales would total more than $100 million), and they didn’t have the luxury of waiting for a complete product before starting on the factory. They had to be able to deliver devices by the millions in a few short years or face the possibility of shutting down Ford and Chrysler, which were now counting on them. (General Motors was going after a more conventional but also problematic solution, with a catalyst deposited on alumina beads rather than on a honeycomb. GM stuck with this solution for a few years but switched to the honeycomb design because the beads created too much back pressure and tended to disintegrate.)
So they began building a factory flexible enough to accommodate whatever final product they developed. They also installed test engines, automobiles, shaker tables (to study vibration effects), and other test equipment in the lab to cut down on development time. Instead of sending every sample off to Detroit, they basically duplicated the automobile companies’ test capabilities in the little town of Corning.
It was an exhilarating, anxious time. “We lost a lot of sleep worrying about what would happen if all this didn’t work out,” Bagley says. And the honeycomb was just part of it. What catalytically active material should be deposited on the honeycomb?
From Houdry’s pioneering work and Engelhard’s forklift-truck work, elements like platinum and palladium were known to be good oxidizing catalysts, but they were expensive. Base metals like copper and iron could do the job much more cheaply, but no one knew how they would hold up in the high temperature of automobile exhaust.
Research at Corning and Engelhard soon showed that base metals weren’t up to the task. They were easily poisoned by sulfur, and attempts to remove the sulfur were unsuccessful. Moreover, particles of the metals agglomerated and lost surface area in high heat, and they were inefficient at oxidizing hydrocarbons. Basemetal research was quickly dropped in favor of precious metals.
Corning now dropped out of catalyst research to concentrate on the substrate. Engelhard, which had seen its 3M forklift honeycomb crumble after short exposure to exhaust, was happy to leave the substrate work to Corning and take over the catalyst side.
Much work was done to minimize the amount of platinum needed, because of its scarcity and expense. Nobody even knew if there was enough platinum in the ground for all the automobiles in the United States. At the same time, the amount of platinum exposed to the exhaust stream had to be maximized. This meant making the platinum particles as small as possible and distributing them over as large a surface area as possible, so they wouldn’t come into contact and agglomerate.
Keith and Mooney looked at many formulations and were eventually able to reduce the amount of platinum to 0.15 percent of the composition. They precipitated the platinum in fine droplets onto high-surface-area particles of gamma alumina. The gamma alumina was so porous that a single gram of it had a surface area of about 150 square meters, which helped keep the platinum separated. By dipping the honeycomb substrate into a thin slurry containing the platinum-coated alumina particles, they were able to achieve even layers of catalyst loading throughout the honeycomb channels. The porosity of the honeycomb itself was crucial in this respect, but the thermal expansion properties of the catalyst slurry and of the substrate had to match. If they expanded and contracted at different rates, the catalyst would separate from the honeycomb and be blown out the tailpipe. Engelhard worked closely with Corning from the beginning to make sure the materials were compatible.
What was Corning to make the honeycomb of? The materials part of the puzzle would prove to be a particularly hard challenge. Irwin Lachman and Ron Lewis, in Coming’s Ceramic Research Group, would be two of the key players in this part of the project. Lewis joined Lachman after a year of working on the catalyst end of things.
The winning compound would have to have good thermal shock resistance (directly related to thermal expansion) and a high melting point. During a cold start, the exhaust from an automobile goes from ambient temperature to 750 degrees Fahrenheit in 30 seconds; that sudden extreme heat would affect the middle of the honeycomb first, causing it to expand against the still-cool outer layers. This could lead to cracking and failure, so thermal expansion would have to be minimized.
Ultimately the exhaust gases would settle out at temperatures in the general area of 2,000 degrees, depending on the operating efficiency of the engine and whether it was idling or pushing the car at 60 miles an hour. The chosen compound would have to be able to withstand these temperatures—and even higher if a little raw gasoline ever made it through and started to burn in the converter. The honeycomb could not have a chance of melting.
These requirements soon pointed to a naturally occurring magnesium-aluminumsilicate compound called cordierite, with the formula 2MgO·2Al 2 O 3 ·5SiO 2 . It melts at 2,650 degrees and has fairly low thermal expansion. Other ceramic materials have even lower thermal expansion properties, but none has so high a melting point. So cordierite it would be.
CORDIERITE OCCURS NATURALLY, BUT not in any deposit significant enough to make mining it practical; the scientists were going to have to synthesize it. This was fairly easy, if you combined the right proportions of talc (a magnesium silicate material), kaolinite clay (an aluminum silicate material), and alumina. But finding the right sources of those ingredients would take a lot of time and screening.
The first step was to order samples of the three ingredients from every supplier they could find. “We had a basement full of raw materials,” Lachman says, including talcs from Texas, Montana, and France and clays from Georgia and elsewhere. Local ground impurities gave each product a slightly different chemistry. The team had to make hundreds of samples of cordierite using every possible combination of talc, clay, and alumina.
The compounds were mixed and extruded into pencil-shaped rods of cordierite; some of the compound was also put through Bagley’s die to see what kind of honeycomb it would produce. Then the samples went to an analytical lab, where their expansion was monitored as they were heated.
The results for the rods were uniformly disappointing. No matter what materials were used, every rod expanded too much when heated. And the initial trials with the extruded honeycombs showed that they cracked and crumbled after only a few thermal cycles. “There were some very depressing times,” Bagley admits.
But the thermal expansion results for the extruded honeycombs were sometimes more promising. “Every so often we would get results that were unusual,” Lachman says. “We would look at them and say, ‘Wow, this is really low, this is wonderful. It must be a mistake.’”
They were not mistakes. Rechecking the data showed that everything was in order. Yet the results were not reproducible. Another honeycomb made days later from the same formula would return a high thermal expansion number. And rod samples never returned the low numbers obtained by extruded honeycombs from the same mix. What was going on?
Late in 1972 Lewis began thinking like the mineralogist he was trained to be. He had known all along that cordierite has a needlelike structure and that its thermal expansion is low along the length of the needle and much higher along its width. In other words, the needle tends to get fatter when heated, rather than longer. There was no reason to believe that the needles would have some preferred orientation after all the mixing, extrusion, and firing steps, but Lewis couldn’t rule it out. “As soon as I had this idea, I immediately knew all the tests I would have to run to see if it was valid,” he says.
He began slicing up honeycombs and looking at the exposed faces using X-ray diffraction, and he found that most of the needles were aligned parallel to the length of the honeycomb. This configuration was ideal for his purposes. It meant that when heated, the cordierite needles would expand into the open holes of the honeycomb, where they had room to grow without causing damage. Along the honeycomb walls, which was where problems would result, very little expansion occurred. He further confirmed these results by examining the cut faces under a scanning electron microscope and with optical micrography, a technique he had learned as a mineralogist.
Meanwhile, the researchers at Engelhard wanted to take their prototypes out of the lab and onto the road. Mooney equipped a Ford Fairlane station wagon with a preliminary version of a catalytic honeycomb in a metal can and ran it on a test track. “Going from forklift trucks to passenger cars was a big deal,” he remembers. “The catalytic converter had to be able to work in Death Valley and in the Arctic—both temperature extremes.” It also had to accommodate wide variations in speed and road conditions. At high temperatures the metal can surrounding the converter expanded while the ceramic substrate did not, leading to a loose substrate rattling around in its shell. At high acceleration the converter would be subjected to vibration from the engine as well as from the road. “Road shock sometimes caused the converters to fall off on a cobblestone test track,” Mooney says. “We had to make sure they didn’t shatter like dinner plates when this happened.”
To help solve all these problems, Ford gave Engelhard access to its Riverside Test Track. Soon 26 cars equipped with catalytic converters were trying to reach the 50,000-mile mark. Eventually 400 cars were involved in the program. The data obtained from the test runs was fed back to research teams at Engelhard. The High Temperature Resistance team had to adapt the catalyst formulation step by step to make it capable of withstanding the most extreme exhaust-gas temperatures; the Low Light-Off team worked to get the conversion reactions to start (“light off”) at lower temperatures, converting gases as soon as possible after the engine was turned on; the Oil Ash team dealt with contamination from burned oil in the exhaust stream. “These were tough years,” Mooney says. “We had to worry about things like what would happen if the converter got completely wet and frozen. Then we had to engineer around them.”
Unbeknown to the researchers, 1973 brought a discovery that almost derailed the whole effort. Health officials at the EPA were confronted with an alarming study showing that a catalytic converter on an internal-combustion engine would combine sulfur contaminant in the gasoline with hydrogen and oxygen to make sulfuric acid. Russell Train, the Administrator of the EPA, presided over an intraagency meeting to decide how to deal with the matter.
“I had two elements of the agency pitted against each other,” Train recalls. “The Mobile Source [i.e., motor-vehicle emissions] people were basically engineers, and the other side of the coin was represented by the health scientists. The latter group argued that catalytic converters would emit a fine aerosol of sulfuric acid, so that anyone standing alongside a Los Angeles freeway would essentially be inhaling a sulfuric acid mist, which was extremely damaging to health. This was a very tough decision to make. I came down on the side of the catalytic converter, which, in hindsight, seems to have been the right decision.” Another worry was the emission of fine particles of platinum and palladium. Extensive testing showed that neither this problem nor sulfuric-acid emission was significant.
Train’s decision saved the project. At this point, with the proper orientation of cordierite needles understood to be key, Lewis and Lachman were being encouraged to apply for a patent, but Lewis wanted to wait until they understood the reasons for the alignment. Apparently, against all odds, the alignment of one of the starting raw materials persisted through all the processing steps. But which starting material? They suspected both clay and talc, the main ingredients. Both are shaped in the form of plates, so both could possibly be stacked in a particular direction in the complex mixture of ingredients.
Lewis began looking at his raw materials in a scanning electron microscope. He learned that the Georgia clay they had settled on for its chemical properties was very highly laminated: It consisted of long stacks of clay plates stuck together like a tower of coins. But that wasn’t the desired configuration. A delaminated clay, with the plates separated, formed the best orientation of cordierite crystals when extruded. The reason the honeycomb die always gave better results than the pencil rods, it turned out, was that the die put more shear force on the compound being extruded, delaminating the stacked plates. Once this was understood, the team ordered a delaminated clay made by the same company and began to get consistent thermal expansion results from their extruded honeycombs.
Lachman had by this time discovered another important property of the cordierite. It tended to form into small crystallites, bunches of needles, that produced a fine pattern of microcracks when the honeycomb cooled after firing. This was good. The cracks gave the honeycomb room to expand without changing its external dimensions when it was heated again. They didn’t even grow and cause a fracture of the honeycomb; they simply expanded and contracted.
WITH MOST OF THE BLANKS FILLED in, Lachman and Lewis applied for a patent on the properly oriented cordierite materials on November 5, 1973. They received patent number 3,885,977 on May 27, 1975. By then the factory was up and running, producing the millions of honeycombs needed to meet the demands of the Clean Air Act.
Coming’s engineers added methylcellulose to the cordierite mixture to produce a plastic material that flowed easily during extrusion but stiffened when exiting the die, eliminating the danger that the honeycomb would collapse like a wet newspaper and allowing it to be cut without caving in its channels. Drying the honeycomb uniformly was made possible by dielectric heating technology, applying varying electric fields to metal plates surrounding the honeycomb. And after careful study of how the ceramic body reacted to heating, the engineers devised a firing schedule that prevented the honeycomb from crumbling. The rate of heating would be sped up, slowed down, or stopped at certain temperatures to allow reactions to take place so as not to threaten the integrity of the structure.
Bagley still considers the cordierite honeycomb a miracle material. “Having a major breakthrough is very rare in any company,” he says. “In the catalytic converter we had two major breakthroughs: a new process and new materials that didn’t exist before.”
Engelhard performed a miracle of its own to meet a new set of EPA emissions standards, which required a significant reduction in NO and NO 2 (collectively referred to as NOx) gases. NOx reacts with oxygen and sunlight to form ground-level ozone, a major component of smog, and with water and unburned hydrocarbons to form nitric acid and other nasty chemicals. The EPA had left NOx standards out of the 1975 requirements so researchers could focus on carbon monoxide and hydrocarbons, but they were scheduled to take effect the following year. During the 1973-74 gas crisis, the EPA extended them to 1978, and in 1977 they were extended again, to 1981. When they took effect, the same catalyst system that added oxygen to carbon monoxide and hydrocarbons would have to remove oxygen from NOx, to leave nitrogen (N 2 ). Two catalytic converters could be used in series, the first one oxidizing, the second reducing, but this would add significantly to the bulk and cost of the equipment.
“John Mooney and I sat down with a piece of paper and started looking at the air-fuel ratios needed for the oxidation and reduction reactions,” Keith says. “We realized that if we could run the reactions near stoichiometric conditions”—meaning at the ideal ratios specified by the chemical reactions—“we had a tiny window of opportunity.” There was a small range of air-fuel conditions where all three reactions (carbon monoxide, hydrocarbon, and NOx) could take place, and even so a new catalyst recipe would be needed.
So they began adding small amounts of reducing catalysts to the platinum formulation. As before, they tried to minimize the amount of catalyst to keep costs down. After much experimenting, they found that inserting a small amount of rhodium into the platinum catalyst could do the trick. To ensure that there was always some oxygen available for the oxidizing, they added ceria, a compound that would act as an oxygen reservoir and release oxygen as needed by switching back and forth between its CeO 2 and Ce 2 O 3 forms. Now both oxidation and reduction could take place at the same time in the same catalytic converter. Such threeway catalysts are still being used and improved on today. Keith and Mooney have obtained nearly a dozen patents on aspects of the device over the years.
This kind of control over chemical reactions wouldn’t have been possible without the computing and sensing technology that was just emerging. The German firm Bosch developed zirconia sensors that could monitor the amount of oxygen in the combustion stream, improving the combustion process to diminish the cleanup work to be done by the catalytic converter. And oxygen sensors in the exhaust stream kept the air-fuel ratio in the small range where three-way catalysis could take place.
The catalytic converter is one of the great environmental success stories of the last century. When each goal was reached, the government raised the standards another notch, and the scientists kept meeting them. The result was automobiles that ran more efficiently with better gas mileage, confounding fears that environmentally friendly cars would be sluggish cars. The current challenge is to produce a zero-emission vehicle.
For their efforts in developing the emissions-cleaning device that has now been used on approximately half a billion cars, Bagley, Lachman, and Lewis were inducted into the National Inventors Hall of Fame in 2002. Keith and Mooney received the Walter Ahlstrom Prize, an international engineering award, in 2001 for their invention of the three-way catalytic converter. Together these inventors have kept an estimated 800 million tons of pollutants out of the atmosphere since the 1970s.