Carbon formation on Rh-substituted pyrochlore catalysts during partial oxidation of liquid hydrocarbons

Identification of three different reducible sites in Rh-substituted pyrochlore catalyst via temperature programmed reduction. •Rh-substituted pyrochlores were tested for partial oxidation of n-tetradecane.•XRD revealed a secondary crystal phase (LaRhO3) at high Rh substitution.•LaRhO3 decomposed at reduction temperatures equal to reaction conditions.•TPO revealed two types of carbon formed during POX: low and high temperature.•LTC reached a steady state quantity within 2h; HTC accumulated over time. Rh-substituted pyrochlore catalysts were prepared and tested for the partial oxidation (POX) of n-tetradecane (TD), a diesel fuel surrogate. The catalysts were substituted lanthanum zirconates with the chemical formula La1.89Ca0.11Zr1.75−xRhxY0.25O7−y (with either 0, 1, 2, or 3wt% of Rh loading). The catalysts underwent a variety of pre- and post-reaction characterization with a focus on deposited carbon, the primary cause of catalyst deactivation. To assess the nature of potential catalytic sites (typically attributed to Rh), each substituted catalyst produced three temperature programmed reduction peaks, with an increase in the most reducible sites at higher Rh loadings. In additions, the catalysts with 2 and 3wt% Rh substitution displayed a secondary crystal phase in their XRD spectra; however, this phase disappeared when the sample was reduced up to 900°C, representative of reaction conditions. To assess carbon deposition after reaction testing, a temperature programmed oxidation (TPO) was conducted on each sample from 250 to 750°C. Each sample produced three CO2 peaks between 250 and 550°C, referred to as low temperature carbon (LTC) peaks, and two peaks at higher temperatures (>600°C). The 2wt% catalyst was run for different times under POX of TD, and the TPO results showed that the amount of LTC reached a steady state within 2h and did not increase even up to 18h under reaction conditions. This indicates that the LTC is not likely responsible for catalyst deactivation. In contrast, the quantity of carbon associated with the higher temperature peaks under TPO increased with reaction time. The carbon associated with the high temperature peaks was highly graphitic, and is typically attributed to deposits on the catalyst inert surfaces and/or inert bed material. After a partial burn-off of the LTC, Raman analysis showed no significant change in the surface carbon. These results suggest that improvement in catalyst performance may be accomplished by modificat