Control of Catalysts’ Structure through Flame Spray Pyrolysis (FSP)
Global energy demand is expected to increase by 47% by 2050. Combustion of fossil fuels such as natural gas, gasoline, and diesel releases pollutant emissions of CO, NOx, SOx, and particulate matter (PM) with the main combustion product of CO2. Methane, the main component of natural gas, has the highest carbon-to-hydrogen ratio among all the hydrocarbons, has a high gravimetric energy density, and has the potential to lower particulate matter (PM) and NOx emissions. However, low engine temperature during the cold start can cause the “slip” of unburned methane in the exhaust, which creates severe environmental concerns as it is 20 times stronger greenhouse effect than CO2. Our group proposed to adopt palladium deposited on CexZr1-xO2 solid solution catalysts synthesized by FSP to remediate the methane slip at low engine temperatures.
We use FSP to control the nanoparticle morphology, crystallite size, crystalline phase, oxygen vacancy concentration, and metal dispersion. We are interested in: 1) Computational modeling of the droplet formation and particle formation in FSP, 2) Synthesis and characterization of atomically dispersed catalysts, and 3) Operando characterization of the catalysts for determining the dynamic environment of active sites.
Predicting Catalysts Properties Using Real-Time Emission Spectroscopy and Machine Learning Models
Many catalysts with various properties can be synthesized by mapping the broad FSP process parameter space. However, post-synthesis characterization for large amounts of candidates retards the discovery speed of desired materials. Here, we combine laser-induced breakdown spectroscopy (LIBS) with machine learning models to create a holistic catalytic synthesis monitoring system to rapidly monitor and predict catalyst quality during FSP. The complex emission spectra of particles are used as descriptors and analyzed by supervised machine learning (ML) models. The monitoring system is applied to predict the phase presence, lattice constant, and oxygen vacancy percentage (OV%) of the Pd-impregnated Ce-Zr-Mn-based catalysts. Our work shows that ML models can optimize FSP process parameters to synthesize solid solution catalysts with desired properties and can improve the efficiency of material discovery in FSP. The project was funded by DOE/SBIR Phase I and Phase II grants and completed in early 2021.
Induction Heating in Catalysis
Production of commodity chemicals, e.g., propylene, ethylene, hydrogen, and ammonia, require high operating temperatures, which generally use natural gas burners to generate the heat needed for the reaction, producing significant greenhouse gas emissions. Besides, heat transfer limitations in natural gas burners can suppress reaction efficiency due to the limited thermal conductivity. Our group overcame this challenge by adopting high-frequency alternating magnetic fields as a primary heat source, in which eddy currents and hysteresis losses are induced in ferromagnetic materials. Functionally designed ferromagnetic nanoparticles provide process heating inside the reactor, improve the reaction’s overall thermal efficiency, and limit GHG emissions compared to conventional heating processes. We work with our industrial partner to create a new generation of catalytic systems to improve energy and reaction efficiencies.
CO2 to Value-added Chemicals
Tri-reforming of methane (TRM) can directly utilize flue gas from a combustion source without requiring gas separation and generate syngas for producing valuable chemicals such as methanol, DME, and light olefins. The TRM combines dry reforming, steam reforming, and partial oxidation of methane. Nickel-based catalysts are widely studied for methane reforming as they can show high activity at a low cost but deactivate due to the sintering and coke formation. To increase the long-term stability of Ni catalysts, we adopted yolk-shell structures for reforming reactions and used their advantageous confinement effect to minimize high carbon deposition and sintering.
Dry reforming of methane (DRM) can directly convert landfill gas consisting of CO2 and methane into syngas. DRM is typically performed at high reaction temperatures (> 750 °C) to minimize coke formation and improve production efficiency. The high-temperature requirement increases the operating costs and causes catalyst sintering, leading to coke formation and catalyst deactivation. We further improved the performance of the yolk-shell catalysts by creating Pt-Ni single-atom alloy catalysts and managed to decrease the operating temperates as low as 500 °C without forming any coke in the catalyst structure and stabilize active metals in the confined morphology.
CO2 Utilization with Tandem Catalysis
Light olefins (C2–C4) can be produced by direct CO2 hydrogenation in a single reactor via the methanol intermediate route, of which CO2 is converted to methanol on metal oxides, followed by a methanol-to-olefins reaction on zeolites. This system can generate higher light olefins selectivity than a modified Fischer–Tropsch synthesis process. Using tandem catalysis, we developed a new catalyst, i.e., indium deposited on ZrO2-based support mixed with SAPO-34 catalyst, and obtained high olefins selectivity and long-time stability. We showed that the primary product of CO from the reverse-water gas shift reaction decreased due to the interaction and proximity of the tandem catalysts.
Funding sources
