Affordable Access

Activity and selectivity loss modelling on Co based Fischer-Tropsch catalysts

  • Kocic, Stefan
Publication Date
Nov 06, 2019
Kaleidoscope Open Archive
External links


Loss of catalyst activity and selectivity with time-on-stream are one of the greatest limitations for the Fischer-Tropsch (FT) industrial process. There is a lack of consensus about the reasons leading to catalyst deactivation and many different paths towards those inevitable events have been evoked in the literature. Among them, some of the most common hypotheses are oxidation and carburation of active sites. Besides, the diminution of Co catalyst activity during time-on-stream exhibits non-uniform behavior, indicating that this phenomenon might be a result of multiple distinctive events. In this thesis, we concentrate on those paths that concern the active phase only, particularly hydrocarbon-species deposition, active phase oxidation and hydrocarbon-induced surface reconstruction as a function of hydrogen, oxygen and carbon coverage effects. With the aid of periodic Density Functional Theory (DFT) calculations, we determine the Gibbs free energy for a large set of key reactions leading to the formation of CαHβOγ surface species on the Co(111) surface uner FT reaction conditions and we identify intermediates and transition states that may lead to activity and selectivity loss of Co-based catalysts. Hence, we propose here to study how the structure of the cobalt surface evolves as a function of the carbon, hydrogen and oxygen chemical potentials under FT reaction conditions. These calculations allowed us to propose an atomistic structure of some experimentally identified coke precursors and to identify favorable reaction conditions towards their formation. Depending on the (C, H, O) coverages, we identify three structural domains containing surface species related with activity and selectivity trends discussed in the literature so far: firstly, a low C coverage domain, where CHβ monomers are formed, the impact of O atoms is the strongest and leads to adsorbed CO, OH or water as well as to oxidized Co sites; an intermediate C coverage domain, where CαHβ linear oligomers and branched hydrocarbon chains are formed and where reconstruction of Co may take place upon subsurface C migration; and thirdly, a high C coverage domain, where we find the formation of longer branched hydrocarbon chains together with the genesis of a carbon overlayer (graphitic coke-like) that is expected to be the main source of deactivation. For intermediate and high carbon coverages, the impact of O atoms on the surface is weaker and its deposition occurs on top of the carbon overlayer without direct contact with Co sites. With the aid of periodic DFT transition state calculations and microkinetic modeling, we offer some new understandings and ideas related to the mechanism of a carbon induced deactivation phenomenon. Our study shows that surface ethynyl species CCH may be regarded as thermodynamically and kinetically the most plausible deactivation initiators. Moreover, we propose 2+2+2 cycloaddition and some CHβ / CHβ reactions as a mechanism for detrimental coke formation leading to a progressive deactivation by a site-blocking effect. This deactivation mechanism has been integrated to an existing deactivation-free micro-kinetic scheme from the literature. The resulting, two-site deactivation model has been optimized and compared to some experimental observations. Our multiscale (DFT and microkinetic model) reproduces well known experimental trends. Hence, we expect that our work will provide the FT community some valuable insights into this intricate and elusive problem, the kinetics of deactivation, as well as some rational guidelines about how to optimize the catalyst process

Report this publication


Seen <100 times