Modern clean diesels and other lean-burning engines boast excellent fuel-economy numbers, but ensuring that they meet regulatory limits for nitrogen oxides, carbon monoxide, and hydrocarbon emissions means that car manufacturers must install special exhaust aftertreatment systems that can cost thousands of dollars extra. Conventional “three-way” exhaust catalysts do not work well in the oxygen-rich combustion fumes that lean-burn engines produce.
Automakers have two diesel oxidation catalysts to choose from to control lean-burn NOx emissions. Each has its pros and cons. Selective catalytic reduction (SCR) is established technology, but requires onboard storage of a urea-based reducing agent, which needs protection from winter temperatures and regular replenishment by vehicle owners. The alternative approach, the lean NOx trap (LNT), or the NOx storage and reduction catalyst, works differently.
In an LNT system, the NOx is temporarily stored when it reacts with an alkaline compound on the catalyst surface, and then later released. The multistep process that converts NOx pollutants into benign nitrogen gas requires the use of high-cost platinum and other platinum group metals (PGMs), said Wei Li, Group Manager at General Motors Global R&D’s Chemical Sciences and Materials Systems Lab in Warren, MI. Not only are platinum-based catalysts expensive, but also the metal’s price tends to fluctuate and high temperatures degrade them over time.
Removing the platinum from LNTs and other automotive systems has long been “a Holy Grail for the catalyst industry,” Li said. A GM team, which in addition to Li included Chang Hwan Kim, Gongshin Qi, and Kevin Dahlberg, began studying the problem in 2007.
“Eventually, we found a cheaper replacement, a perovskite oxide (metal oxide) that demonstrates quite good catalytic activity in tests at similar loadings with simulated diesel exhaust,” Li said. Even though the new system still contains some PGMs, “we believe that it’s a big step in right direction.”
GM’s innovation was reported in a paper published recently in Science magazine.
To see how the new catalyst could benefit LNT technology, one must know how the traps work in detail, explained Jim Parks, Group Leader for Emissions and Catalyst Research at Oak Ridge National Laboratory in Tennessee. LNT operation comprises several steps that occur as the engine cycles between lean and rich operating modes. “During the lean-burn phase of the cycle,” he explained, “platinum promotes the oxidation of NO to NO², which is needed to successfully store NOx until the fuel-rich part of the cycle.” The NOx storage medium, usually an alkaline oxide such as barium oxide, is converted to nitrate by reaction with the NOx.
The storage capacity of the oxide is not unlimited, so after the engine runs lean for a couple of minutes, it becomes necessary to “empty” the NOx store. Removing the NOx is accomplished by running the engine rich for 3 to 9 s, which leads to the decomposition of the nitrate and release of the NOx. Another result of running rich is higher levels of carbon monoxide and hydrocarbons in the exhaust stream. These compounds react with the released NOx, reducing it to N². The engine then runs lean again and the cycle repeats.
Parks noted that the platinum catalyst also promotes the oxidization of hydrocarbons to help fully eliminate oxygen at the trap, which must occur if the NOx is to be released. “Finally, during the rich phase, the platinum enables the trap to regenerate by assisting in the NOx release and reduction steps,” he said.
In recent years, platinum prices have swung wildly, driven by the economic downturn, political problems in the nation's operating the mines, and other market disturbances. Platinum’s erratic pricing torments personnel planning to build lean-burn aftertreatment systems several years hence, Li said. The new perovskite catalyst could help free automakers from that uncertainty, saving as much as 70% along the way, the GM researchers claimed.
Li reported that the perovskite oxide catalysts, a family of formulations based on the rare earth element lanthanum, demonstrated “quite good” oxidative activity. Members of the oxide family, which variously contain metals such as cobalt, manganese, or iron, display different active properties.
The team also discovered that doping the lanthanum perovskite oxide compound with strontium to levels of 10% increases the oxidation activity of the material significantly. Finally, the GM researchers determined that the addition of “a little bit of palladium assists the conversion of hydrocarbons and minimizes the poisoning effect of sulfur compounds,” Li said. “Depending on the market, palladium costs four or five times less than platinum.”
The new catalytic system also resists degradation of catalytic action caused by exposure to heat better than does platinum.
“It looks like GM has made a significant stride toward reducing PGMs in LNT catalysts,” Parks observed. He estimated that further development and commercialization should take a minimum of three years as materials specialists fine-tune the material and engineers optimize the manufacturing process. He noted, for example, that the key components of an LNT formulation—the high-surface area support structure, as well as the oxidation catalyst and NOx absorbing sites—need to be precisely engineered so that each functions in close proximity to the others to best promote the desired overall reactions.
The new, lower-cost lean-burn catalysts could in time help speed the adoption of clean diesel and direct-injection gasoline engine technology in the North American car market.