Zn Redistribution and Volatility in ZnZrOx Catalysts for CO2 Hydrogenation

ZnO–ZrO2 mixed oxide (ZnZrOx) catalysts are widely studied as selective catalysts for CO2 hydrogenation into methanol at high-temperature conditions (300–350 °C) that are preferred for the subsequent in situ zeolite-catalyzed conversion of methanol into hydrocarbons in a tandem process. Zn, a key ingredient of these mixed oxide catalysts, is known to volatilize from ZnO under high-temperature conditions, but little is known about Zn mobility and volatility in mixed oxides. Here, an array of ex situ and in situ characterization techniques (scanning electron microscopy/energy dispersive X-ray spectroscopy (SEM/EDX), transmission electron microscopy (TEM), powder X-ray diffraction (PXRD), X-ray absorption spectroscopy (XAS), X-ray photoelectron spectroscopy (XPS), Infrared (IR)) was used to reveal that Zn2+ species are mobile between the solid solution phase with ZrO2 and segregated and/or embedded ZnO clusters. Upon reductive heat treatments, partially reversible ZnO cluster growth was observed above 250 °C and eventual Zn evaporation above 550 °C. Extensive Zn evaporation leads to catalyst deactivation and methanol selectivity decline in CO2 hydrogenation. These findings extend the fundamental knowledge of Zn-containing mixed oxide catalysts and are highly relevant for the CO2-to-hydrocarbon process optimization.

where FCO and FOX are the molar flow rate in contained C units of CO and oxygenates (OX) at the outlet of the reactor and FinCO2 is the molar flow rate of CO2 at the entrance of the reactor.The selectivity to CO and oxygenates (mainly methanol in our case) was defined as , CO OX 100, CO, OX where Fi is the molar flow rate in contained C units of either CO or oxygenates (OX) at the outlet of the reactor.Space time yields of methanol and CO (STY) has been defined as a function of the space time (τ) where the space time is define in units of mmolCO2 (fed) h gcatalyst-1 or molCO2 (fed) min molZn-1 (in the catalyst after the pretreatment).EXAFS fit at Zn K-edge was not conducted due to limited spectra quality.Indeed, while all the FT-EXAFS where extracted in the 2.5 -10.4 Å -1 k-range, the spectra of ZnZrO-10-700 was extracted until k 8 Å -1 (Figure S7a).This reduced the number of available independent variables from 10 (2.5 -10.4

S2. Supplementary Figures
Å -1 k-range) to 7 (2.5 -8 Å -1 k-range).In both cases the spectra quality was not sufficient to reproduce the fit previously reported.Nevertheless, from a qualitative perspective we observed that ZnZrO-10-700 and ZnZrO-20-700 presented fingerprints of Zn-Zn scattering path from ZnO in both k space and k 2 -weighted FT-EXAFS magnitude and imaginary parts (indicated with arrows in Figure S7), indicating an increase of the ZnO cluster dimension.S1).To estimate the content of Zn, the coordination numbers of Zr-Zr and Zr-Zn scattering paths were refined as 12-Zn and Zn, respectively, where the Zn variable describes the amount of Zn atoms surrounding the absorber Zr atom (Figure S10 and Table S2).However, to reduce the correlation with the scattering path Debye Waller factor contribution, the latter was fixed to the values determined with the Einstein model in our previous work [main text, ref 15].As previously reported, this approach can provide a rough estimation of the Zn content in the material.The best-fit values of Zn ≈ 0 obtained in the Zr K-edge EXAFS fits for ZnZrO-10, ZnZrO-10-700 and ZnZrO-20-700 suggest that the amount of Zn atoms located at the ZnO nano-cluster/ZrO2 matrix interface (and thus contributing to the Zr-Zn scattering path) in these cases is lower than the detection limit of the technique (≈15%, considering the error evaluated from the fit of ZnZrO-20).14) 8( 14) 8( 14) 8( 14) 7( 14)       12000-48000 cm 3 h -1 g ZnZrOx -1 ).Dashed line is the CO 2 -to-methanol equilibrium limitation.Figure S22.Impact of 400-550 ºC H 2 /He pretreatment on the performance of ZnZrO x -20 catalyst (reaction conditions 350 ºC, 30 bar 12000 cm 3 h -1 g ZnZrOx -1 ).

Figure S6 .
Figure S6.Zn K-edge raw absorption data of ZnZrOx-10 reporting pre-edge and post-edge lines Figure S7 Zn K-edge k 2 -weighted EXAFS (a) spectrum and its Fourier Transform magnitude (b) and

Figure S9 Figure S10
Figure S9 Experimental spectra (black line) and best fit (red line) of Zr metal Zr K-edge k 2 -weighted EXAFS
Figure S17.Dark (left) and bright (right) field images of ZnZrO x -10 sample after 400 °C pretreatment Figure S18.Dark (left) and bright (right) field images of ZnZrO x -10 sample after 550 °C pretreatment Figure S19.Dark image (left) and EDX maps of ZnZrO x -10 sample after 440 °C pretreatment in 50 % Figure S20.Dark image (left) and EDX maps of ZnZrO x -10 sample after 550 °C pretreatment in 50 %

Table S2
EXAFS fit results of Zr K-edge FT-EXAFS of the studied samples.FT-EXAFS were extracted in the 2.5-11.2(Å -1 ) k-range and fit performed in the 1 -3.65 Å range.Passive amplitude reduction factor S02 fixed to 0.68 as determined from EXAFS fit of Zr metal reported above.* indicates parameters fixed to values previously evaluated.** The variable 'Zn' indicates the number of Zn atoms replacing the 12 Zr atoms surrounding the absorber Zr in second coordination shell.