Multivariate Bayesian Optimization of CoO Nanoparticles for CO2 Hydrogenation Catalysis

Abstract

The hydrogenation of CO₂ holds promise for transforming the production of renewable fuels and chemicals. However, the challenge lies in developing robust and selective catalysts for this process. Transition metal oxide catalysts, particularly cobalt oxide, have shown potential for CO₂ hydrogenation, with performance heavily reliant on crystal phase and morphology. Achieving precise control over these catalyst attributes through colloidal nanoparticle synthesis could pave the way for catalyst and process advancement. Yet, navigating the complexities of colloidal nanoparticle syntheses, governed by numerous input variables, poses a significant challenge in systematically controlling resultant catalyst features. We present a multivariate Bayesian optimization, coupled with a data-driven classifier, to map the synthetic design space for colloidal CoO nanoparticles and simultaneously optimize them for multiple catalytically relevant features within a target crystalline phase. The optimized experimental conditions yielded small, phase-pure rock salt CoO nanoparticles of uniform size and shape. These optimized nanoparticles were then supported on SiO₂ and assessed for thermocatalytic CO₂ hydrogenation against larger, polydisperse CoO nanoparticles on SiO₂ and a conventionally prepared catalyst. The optimized CoO/SiO₂ catalyst consistently exhibited higher activity and CH₄ selectivity (ca. 98%) across various pretreatment reduction temperatures as compared to the other catalysts. This remarkable performance was attributed to particle stability and consistent H* surface coverage, even after undergoing the highest temperature reduction, achieving a more stable catalytic species that resists sintering and carbon occlusion.

Figure 1

Phase map for the colloidal synthesis of rock salt CoO. (a) Importance scores of each experimental variable in determining crystal phase, with the reaction time, reaction temperature, and molar ratio of OA:Co(acac)₂ having the greatest influence on phase determination. (b) Scatter plot of the reactions corresponding to these three most important experimental variables. (c) Predicted phase map after extrapolation corresponding to the three most important experimental variables. The resulting 10 phases or phase combinations are coded by color according to the key, with the target rock salt CoO phase given in green.

Figure 2

Pareto charts for statistical significance of the experimental variables that affect CoO nanoparticle (a) size, (b) polydispersity (σ/d̅), and (c) shape variance and their corresponding response surfaces for the three most important variables in (d), (e), and (f), respectively. The vertical blue lines in the Pareto charts represent the 90% confidence interval (α = 0.10).

Figure 3

(a) Powder XRD patterns of optimized (pink) and unoptimized CoO (yellow) nanoparticles. The reference pattern for rock salt CoO (PDF #00–009–0402) is provided below. TEM images of (b) optimized and (c) unoptimized rock salt CoO nanoparticles. Both scale bars represent 100 nm.

Figure 4

H₂-TPR profile of supported CoO/SiO₂ catalysts, with dashed lines indicating reduction events selected for CO₂ hydrogenation catalytic activity testing.

Figure 5

Catalytic performance of CoO/SiO₂ catalysts in the CO₂ hydrogenation reaction after reductive pretreatments between 300 and 450 °C. Reaction conditions were 300 °C, 3 MPa, WHSV = 1 g CO₂·g·cat^–1·h^–1, and H₂:CO₂ molar ratio of 3. Square symbols are conversion; blue bars are CO C-selectivity; green bars are CH₄ C-selectivity; purple bars are C₂₊ hydrocarbon C-selectivity.

Figure 6

STEM-HAADF images of Opt-CoO/SiO₂, Unopt-CoO/SiO₂, and IWI-CoO/SiO₂ catalysts in the (a–c) as-synthesized forms, (d–f) following reduction at 450 °C, and (g–i) after CO₂ hydrogenation with a 450 °C reductive pretreatment, as well as (j–l) associated particle diameter distributions. All scale bars represent 25 nm.