Linear Scaling Relationships between Relative Diffraction Peak Intensity and Catalytic Oxidation of Light Alkanes

This study reports a methodology of leveraging highly accessible relative X-ray diffraction peak intensities for the fast identification of reactive facets of polycrystalline catalysts. Specifically, taking polycrystalline α-Fe2O3 as an example, we demonstrate that the ambient-condition X-ray diffraction intensity ratio of (110) to (104) facets (denoted as I(110)/I(104)) of α-Fe2O3 exhibits a strong linear correlation (R2 ≈ 0.9) with catalytic rates across diverse temperatures and flow conditions for methane oxidation. Both linear and nonlinear statistical validations via Pearson correlation, permutation importance, and greedy deletion analyses confirm the exclusive dominance of this feature over conventional physicochemical parameters. Complementary characterizations and mechanistic investigations including X-ray photoelectron spectroscopy (XPS), Raman, electron paramagnetic resonance (EPR), and CH4-TPSR reveal that I(110)/I(104) highly correlates with the preferential exposure of (110) facets vs the competitive (104) facet, as well as enhanced oxygen vacancy formations to facilitate C−H and O2 activation. This finding strongly implies a dominant reactivity of (110) facets among various crystal facets. Critically, this linear relationship was also established in both ethane and propane catalytic lean combustion and can be used to predict the reactivities of unseen structures, indicating a mechanistic consistency across light alkane oxidations dominated by (110) facets. The correlation between the relative diffraction peak intensity and catalytic performance is posited to be generalizable to polycrystalline catalysts for surface-sensitive reactions. Such an experimental quantity will not only be highly accessible for practical activity predictions but more importantly be surface-informative to identify reactive facets via routine XRD measurements, which otherwise require intricate investigations on well-defined model systems through controlled surface sciences and/or theoretical modelings.