A schematic view of the stepped structure of Mn-mullite for catalytic conversion of NOx diesel emissions. The red, purple and pale green represent oxygen, manganese and samarium, respectively. Credit: Cho; UTD.

If you want a job done right, ask an oxide.

At least, that is the approach that start-up Nanostellar (Redwood City, Calif.) is taking with regard to catalytic converters for diesel exhaust.

Diesel engines are more fuel-efficient than gasoline engines, but they are spew NOx pollutants and soot. The former are dealt with by means of “selective catalytic reduction,” which reacts the NOx with ammonia to yield benign N2 and water, or by use of a “lean NOx trap,” which adsorbs and reduces NOx to N2. The latter is dealt with by means of a “continuous regenerable trap,” which uses NO2 to oxidize soot. For all of these reactions—tailpipe and soot—a key step is the oxidation of NO to NO2.

The material of choice for catalyzing the NO to NO2 reaction is platinum, however, platinum is expensive, which drives up vehicle cost, and, inconveniently, also attracts thieves who have found they can make a tidy, if shady, living cutting catalytic converters out of cars and trucks.

A paper last week in Science, demonstrated that a manganese-based mixed oxide can provide the same catalytic service as platinum, and maybe even better. One of the coauthors, Kyeongjae (KJ) Cho, professor of materials science and engineering at the University of Texas at Dallas, is a cofounder of Nanostellar.

The group studied (Sm, Gd)Mn2O5, a composition they call “Mn-mullite,” and doped in cerium and strontium. In a phone interview, Cho explained that, although the compositions they studied are not the aluminosilicate mullite familiar to most materials scientists, the atomic structure of the compound is the same, hence the reference.

“We were looking for a conceptual framework and looking for suitable atomic structures in minerals. If the oxygen is too strongly bonded, it is not good, but it’s not good if it is bonded too loosely, either,” Cho said. “The mullite structure is critical to enabling the catalytic reaction. Manganese can accept and give electrons however, aluminum and silicon cannot.”

Similar to mullite’s two-cation nature (3Al2O3–2SiO2), this Mn-mullite compound, is based on a manganese oxide with either samarium or gadolinium. The material was synthesized by coprecipitation methods. Cho says that the coprecipitation route gives better control over the structure of the mixed-phase materials, especially regarding incorporation of the dopants into the atomic structure. The resulting composition was Mn7CeSmSrO14.83, which they labeled MnCe-7:1.

After calcining, samples were hydrothermally aged (10 hours at 820°C in 10 percent steam). “We expect the material to perform at least 10 years or 100,000 miles, so we need to check it under simulated aging conditions,” says Cho.

X-ray diffraction showed that the MnCe-7:1 was a multi-phase mixture comprising 66 weight percent Sm(Ce, Sr)Mn2O5 mullite, eight percent Mn3O4 spinel and 26 percent CeO2 fluorite, with particle sized in the 50-nanometer range. There were also some amorphous regions between the CeO2 and mullite particles.

The group tested the NO conversion effectiveness of several oxides, including MnCe-7:1, against platinum during ramp-up and ramp-down in the temperature regime from 50-350°C. The oxides outperformed the platinum at all temperatures (going up and coming down), and reached a maximum of NO conversion of 45% more than platinum as the maximum ramp-up temperature was approached. Part of this high conversion rate stems from catalytic activity by the spinel Mn3O4 phase at temperatures in the 330-350°C range.

Significantly, the MnCe-7:1 is actively catalyzing at low temperatures (120°C). More pollutants enter the atmosphere during the first five minutes of vehicle operation because the catalytic converter assembly is still cold.

The group turned to density functional theory modeling for a first principles approach to understanding the catalysis reaction mechanism, and also to diffuse reflectance infrared Fourier transform spectroscopy for experimental support. They found that Mn-Mn dimers on a stepped surface of mullite (110) activate the oxygen molecules. (The paper goes into some detail about the energies involved in these reactions.)

Other atomic structures were investigated, Cho said. “We also tried perovskite structures, but they were quite inferior to mullite.”

A UTD press release says that Nanostellar plans to commercialize the oxide catalyst under the trade name Noxicat.

This is not Nanostellar’s first venture into catalytic products and the company’s general efforts in this field have drawn praise from analysts such as Lux Research.

Share/Print