Catalyst deactivation during thiophene HDS: The role of structural sulfur Bas M. Vogelaar a, * , Petr Steiner a , Thomas F. van der Zijden a , A. Dick van Langeveld a , Sonja Eijsbouts b , Jacob A. Moulijn a a Delft University of Technology, Faculty of Applied Sciences, Department of Reactor and Catalysis Engineering, Julianalaan 136, 2628 BL Delft, The Netherlands b Albemarle Catalysts, P.O. Box 37650, 1030 BE Amsterdam, The Netherlands Received 22 July 2006; received in revised form 28 September 2006; accepted 16 October 2006 Available online 15 November 2006 Abstract The deactivation behavior of a Mo/Al 2 O 3 and a NiMo/Al 2 O 3 catalyst during the hydrodesulfurization of thiophene was investigated under gas- phase conditions. Four different deactivation mechanisms were considered: sintering and segregation of the active phase, blocking of the pore structure by coke deposition, poisoning of the active sites by coke and changes in the nature of the active sites induced by differences in pretreatment and reaction conditions. These mechanisms were investigated using quasi in situ XPS, analysis of the pore structure and coke content, and by measuring the HDS activity. During the thiophene HDS reaction, both catalysts show similar deactivation behavior: about 50% of the initial activity was lost during the first hours on stream. The extent of sintering/segregation and pore blocking was marginal and had no significant effect on the activity. A trend between the coke content and the deactivation was observed, but it is concluded that coke does not selectively poison the active sites. The main cause for the observed initial deactivation is the loss of sulfur from the active phase during the reaction. The number of sulfur groups and vacant sites on the catalysts is in equilibrium with the H 2 S/H 2 ratio of the gas phase. Exposure to H 2 S resulted in an increased initial activity or a (partial) reactivation, whereas exposure to H 2 caused a deactivation. It is concluded that the rate-determining step in the HDS of thiophene is catalyzed by structural sulfur; most probably, acidic SH groups facilitate the hydrogenation of the thiophenic ring. # 2006 Elsevier B.V. All rights reserved. Keywords: Deactivation; Hydrodesulfurization; Thiophene 1. Introduction Conventional NiMo and CoMo catalysts are widely used in the hydroprocessing units of oil refineries. Nowadays, catalyst stability becomes more and more important, as heavier feeds need to be processed at higher conversion levels. An important feature of HDS catalysts is their initial deactivation, also called start-of-run deactivation: Within the first days of operation most catalysts lose a significant fraction of their initial activity [1]. Minimizing this initial deactivation may considerably extend the catalyst life in the unit, reducing downtime and costs. To achieve this, further knowledge is needed about the deactivation mechanism of hydroprocessing catalysts. During the gas phase hydrodesulfurization of thiophene the initial deactivation behavior of HDS catalysts is extreme, when the catalyst has been activated in H 2 S/H 2 . Typically about half of the initial activity is lost within the first few hours on stream (for example, see [2]). This reaction and activation procedure is widely used to test the performance of HDS catalysts. However, in most studies the initial deactivation is ignored, and the activity after a number of hours on stream is used as ‘‘steady state’’ activity. Few researchers did investigate the initial deactivation during thiophene HDS. For example Pedraza et al. [3] found that sintering was the major cause of deactivation for unsupported MoS 2 , and Elst et al. [4] attributed the deactivation of supported MoS 2 catalysts to coke deposition. In this study, we further investigated the deactivation behavior of supported HDS catalysts in the gas phase HDS thiophene. This reaction was chosen for its pronounced initial deactivation behavior, its fast response time and ease of operation. This enabled us to monitor rapid changes in catalytic activity and to analyze the spent catalysts without sample www.elsevier.com/locate/apcata Applied Catalysis A: General 318 (2007) 28–36 * Corresponding author. Current address: Albemarle Catalysts, P.O. Box 37650, 1030 BE Amsterdam, The Netherlands. Tel.: +31 20 6347648; fax: +31 20 6347653. E-mail address: Bas.Vogelaar@albemarle.com (B.M. Vogelaar). 0926-860X/$ – see front matter # 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.apcata.2006.10.032