Volumetric performance for catalyst benchmarking: Uncovering biases in mass‑normalized catalytic metrics

The yields of heterogeneous catalysts are traditionally reported and scored per unit of catalyst mass. However, because volume is the design variable to be optimized when designing fixed-bed reactors, volumetric performance offers a more realistic assessment of a material’s industrial potential. Normalizing yields to catalyst bed volume more appropriately reflects practical reactor constraints and provides a suitable basis for comparison. Moreover, volumetric performance is intrinsically linked to sustainable process design, as reactor volume governs energy use, materials intensity and overall environmental impact. The oxidative dehydrogenation of ethylbenzene to styrene is used as a case study, an alternative to the energy-intensive conventional process. Two catalysts with comparable styrene selectivity were evaluated using both catalyst mass- and catalyst volume-based metrics. While one catalyst delivered a fivefold improvement in mass-based yield, evaluating them volumetrically revealed their performance to be equivalent. These insights demonstrate how mass-based comparisons can introduce significant bias, potentially obscuring the true industrial viability of a catalyst. This mismatch between mass‑normalized activity and the volumetric constraints of fixed‑bed reactors constitutes a methodological gap in current catalyst evaluation practices, which our work addresses by comparing catalysts under equal‑volume conditions. We further show that differences in bulk density, the bulk‑density‑dependent modified residence time (τ*), and the nonlinear dependence of conversion and selectivity on τ* underpin this divergence, introducing systematic benchmarking bias. Using a theoretical model, we found that testing at equal mass and subsequently converting to volumetric metrics can also introduce bias, ultimately due to the non-linear dependence of conversion (and selectivity) on τ*. Bulk density is not a trivial mass‑to‑volume scaling factor; it also exerts implicit effects on the performance, as discussed in the manuscript.