Dynamics and Mechanism of Carbon Filament Formation during Methane Reforming on Supported Nickel Clusters

CH4–CO2 and CH4–H2O reforming on Ni-based catalysts can lead to the undesired formation of carbonaceous residues. The dynamics of the formation of carbon filaments and encapsulating layers on dispersed Ni nanoparticles (5–11 nm diameter) are determined here using an inertial microbalance to measure mass changes and mass spectrometry to concurrently assess turnover rates under conditions of reforming practice (800–1000 K). The morphology and rate of formation of carbonaceous species were controlled by a ratio of pressures (χ = PCH4PCO/PCO2 or ψ = PCH4PH2/PH2O), which are proportional to each other through the equilibration of water-gas shift) that uniquely determines the thermodynamic activity of carbon at the metal surface ((aC*)s) and the thermodynamic driving force for carbon diffusion and filament formation, based on a reaction-transport model derived from the elementary steps that mediate CH4 reforming. Each sample exhibited three distinct kinetic regimes for carbon formation, which evolved with increasing χ values from undetectable carbon deposition (I) to a constant rate of carbon filament growth without detectable changes in CH4 reforming rates (II) and ultimately to the formation of carbon overlayers with a concurrent decrease in CH4 reforming and carbon formation rates (III). Rates of filament growth in regime II were proportional to χ or ψ values, consistent with a filament growth mechanism limited by carbon diffusion. Such carbon filaments were similar in diameter to the attached Ni nanoparticles. In regime III, the high prevalent carbon activities led to the simultaneous nucleation of several carbon patches, thus precluding the directional diffusion imposed by a single filament and leading to the encapsulation and loss of accessible surface for CH4 turnovers. Filaments formed in regime II were removed when placed under the conditions of regime I via the microscopic reverse of their formation processes. Threshold carbon activities required for the incipient formation of filaments are higher and filament formation rates are lower (for a given χ or ψ) on smaller nanoparticles because of the lower stability (higher thermodynamic carbon activity) of filaments with smaller diameters. Carbon deposition rates decreased with increasing temperature (for a given χ or ψ) because of a corresponding decrease in the lumped kinetic and thermodynamic parameters that relate the surface carbon activity to χ or ψ. The formalism used to describe carbon formation rates, in this study for CH4 reforming rates far from equilibrium and for carbon formation and removal rates that do not disturb the carbon activity set by CH4 reforming turnovers at steady-state, also informs the testing of these assumptions while providing also a framework for the rigorous extension of these reaction-diffusion constructs to more practical conditions, for which these assumptions may no longer apply.