RESUMO
Electrochemical oxidation of CH4 is known to be inefficient in aqueous electrolytes. The lower activity of methane oxidation reaction (MOR) is primarily attributed to the dominant oxygen evolution reaction (OER) and the higher barrier for CH4 activation on transition metal oxides (TMOs). However, a satisfactory explanation for the origins of such lower activity of MOR on TMOs, along with the enabling strategies to partially oxidize CH4 to CH3OH, have not been developed yet. We report here the activation of CH4 is governed by a previously unrecognized consequence of electrostatic (or Madelung) potential of metal atom in TMOs. The measured binding energies of CH4 on 12 different TMOs scale linearly with the Madelung potentials of the metal in the TMOs. The MOR active TMOs are the ones with higher CH4 binding energy and lower Madelung potential. Out of 12 TMOs studied here, only TiO2, IrO2, PbO2, and PtO2 are active for MOR, where the stable active site is the O on top of the metal in TMOs. The reaction pathway for MOR proceeds primarily through *CH x intermediates at lower potentials and through *CH3OH intermediates at higher potentials. The key MOR intermediate *CH3OH is identified on TiO2 under operando conditions at higher potential using transient open-circuit potential measurement. To minimize the overoxidation of *CH3OH, a bimetallic Cu2O3 on TiO2 catalysts is developed, in which Cu reduces the barrier for the reaction of *CH3 and *OH and facilitates the desorption of *CH3OH. The highest faradaic efficiency of 6% is obtained using Cu-Ti bimetallic TMO.
RESUMO
Machine-learning models for predicting adsorption energies on metallic surfaces often rely on basic elemental properties and electronic and geometric descriptors. Here, we apply categorical entity embedding, a featurization method inspired by natural language processing techniques, to predict adsorption energies on bimetallic alloy surfaces using categorical descriptors. Using this method, we develop a machine-learned representation from categorical descriptors (e.g., surface composition, adsorbate type, and site type) of the slab/adsorbate complex. By combining this representation with numerical features (e.g., slab metal stoichiometric ratios), we create the CatEmbed representation. Remarkably, decision tree models trained using CatEmbed, which includes no explicit geometric information, achieve a Mean Absolute Error (MAE) of 0.12 eV. Additionally, we extend this technique to predict reaction energies on bimetallic surfaces, creating the CatEmbed-React representation, which achieves an MAE of 0.08 eV. These findings highlight the effectiveness of categorical entity embedding for predicting adsorption and reaction energies on bimetallic alloy surfaces.
RESUMO
Materials that combine magnetic order with other desirable physical attributes could find transformative applications in spintronics, quantum sensing, low-density magnets and gas separations. Among potential multifunctional magnetic materials, metal-organic frameworks, in particular, bear structures that offer intrinsic porosity, vast chemical and structural programmability, and the tunability of electronic properties. Nevertheless, magnetic order within metal-organic frameworks has generally been limited to low temperatures, owing largely to challenges in creating a strong magnetic exchange. Here we employ the phenomenon of itinerant ferromagnetism to realize magnetic ordering at TC = 225 K in a mixed-valence chromium(II/III) triazolate compound, which represents the highest ferromagnetic ordering temperature yet observed in a metal-organic framework. The itinerant ferromagnetism proceeds through a double-exchange mechanism, which results in a barrierless charge transport below the Curie temperature and a large negative magnetoresistance of 23% at 5 K. These observations suggest applications for double-exchange-based coordination solids in the emergent fields of magnetoelectrics and spintronics.