Abstract
Methane from abundant natural gas reserves represents a valuable potential feedstock for the chemical industry, but liquefying natural gas for storage and transport renders much of the gas present in remote deposits economically inaccessible. A long-standing goal in chemistry has been the development of a catalyst capable of directly converting methane to methanol, a liquid with high energy density that can be easily stored and transported using existing infrastructure. The current industrial process for converting methane to methanol is an inefficient two-step process that results in significant greenhouse gas emissions. A direct process for converting methane to methanol would mitigate these emissions, but because of the stability of methane and the higher relative reactivity of methanol, no industrially-viable catalyst has yet been found that can convert methane to methanol with high yield.
Copper-exchanged zeolite catalysts oxidize methane to methanol with high selectivity, but a low yield per copper has prevented industrial application and hampered active site characterization. Utilizing the proton form of Cu-exchanged MOR and ZSM-5 zeolites instead of the sodium form has been shown to increase methanol productivity, which should facilitate identification of the Cu active sites, but the active site structure has not yet been unambiguously defined. Elucidating the active site structures will facilitate the design of improved catalysts.
Several Cu-exchanged MOR (Cu-MOR) and ZSM-5 (Cu-ZSM-5) catalysts were synthesized via liquid ion exchange, with varying Cu loading and co-cation (Na vs H). The effect of the co-cation on reactivity of the catalysts was investigated for both the cyclic and steady state methane-to-methanol reaction. The Cu speciation in the catalysts was investigated by X-ray absorption spectroscopy (XAS), UV-vis spectroscopy, Raman spectroscopy, and temperature-programmed reduction in H2 (H2-TPR).
Using a methane pressure of 35 bar, the C1 product yield during cyclic operation of a Cu-H-MOR catalyst was 0.42 mol (mol Cu)-1. Linear combination fitting of the Cu K-edge X-ray absorption spectrum (XANES-LCF) of the Cu-H-MOR catalyst showed that 83% of the CuII in the fresh catalyst was reduced to CuI during treatment in helium at 723 K, which is assumed to be the redox-active fraction of Cu for methanol formation. Normalizing the product yield to the redox-active fraction of Cu gave a reaction stoichiometry of 0.50 mol (mol Cu)-1, which is consistent with the presence of dicopper active sites, assuming one turnover per active site in the cyclic reaction. In situ Raman spectroscopy of the O2-activated Cu-H-MOR catalyst revealed no features in the peroxo stretching region, while the deep-UV Raman spectrum showed a feature at 570
cm-1. This feature is consistent with the symmetric Cu-O stretch of a mono-μ-oxo dicopper(II) species, suggesting that this species, and not a μ-1,2-peroxo dicopper(II) species, is the predominant active site in the Cu-H-MOR catalyst.
Substituting Na co-cations for protons decreased the methanol selectivity and increased the CO2 selectivity for both Cu-MOR and Cu-ZSM-5 catalysts. In situ UV-vis spectroscopy revealed a correlation between methanol yield and the prominence of a shoulder at 27,500 cm-1 for the Cu-MOR catalysts, while a correlation between CO2 yield and the intensity of a feature at ~22,000 cm-1 was observed for the Cu-ZSM-5 and Cu-MOR catalysts. These results led to the conclusion that the presence of Na shifts the distribution of Cu from mono-μ-oxo dicopper(II) species (which promote methanol formation) to μ-1,2-peroxo dicopper(II) species (which promote CO2 formation) in Cu-MOR and Cu-ZSM-5. This conclusion was supported by XAS results. A reaction network based on stoichiometry was proposed that describes the formation of intermediates leading to methanol, CO, and CO2 over mono-μ-oxo dicopper(II) and μ-1,2-peroxo dicopper(II) active sites.
