On the difference between gas- and liquid-phase hydrotreating test reactions Original Research Article

In industrial practice, hydrotreating of oil fractions is carried out in either a gas-phase process or a trickle flow process. We previously noticed that a remarkable difference exists between the relative activity of mixed sulfide catalysts in gas-phase and liquid-phase hydrodesulfurization (HDS) reactions. In the literature, however, no satisfying explanation with respect to the possible fundamental differences between these reactions can be found. In this paper, we report an elaborate investigation on the effect of reaction conditions, type of reactant and type of the catalyst on the occurrence of differences between the relative activity, i.e. ranking, of mixed sulfide catalysts in gas- and liquid-phase reactions.Striking differences were observed between the ranking of nitrilo-triacetic acid (NTA) and conventionally prepared NiMo catalysts in thiophene gas-phase HDS and liquid-phase dibenzothiophene (DBT) HDS. Importantly, these differences did not depend on the nature of the reacting sulfur-containing compound. This allows the generalisation that NTA-based Ni(Mo) catalysts are relatively more active in gas-phase HDS reactions, whereas conventionally prepared NiMo catalysts are relatively more active in liquid-phase HDS reactions. An analogous behaviour was observed for low- and high-temperature sulfided NiW/γ-Al2O3 catalysts, of which the latter is much more active in gas-phase HDS reactions and the former is more active in liquid-phase HDS reactions. It is concluded that this so-called ‘gas–liquid-phase controversy’ is a generic phenomenon in hydrotreating reactions over metal sulfide catalysts. It was verified that mass transfer limitations do not play a role in this matter.