Deactivation and regeneration studies in the selective hydrogenation of glycerol to propanol

Abstract

Due to a new EU legislation which entails that a certain percentage of fuels must come from second generation feedstocks, the interest in crude glycerol has grown as it is produced in large amounts as a byproduct during the biodiesel production. The raw glycerol can be converted to propanol via catalytic hydrogenation and then used as an additive for fuel. As the conversion of glycerol to propanol occurs via bi- functional mechanism the catalyst requires an acid component promoting dehydration of glycerol and a metal for hydrogenation. Acid sites can be provided by heteropolyacids, which are strong Brønsted acids. The catalyst that was investigated in this work was a 2 % Pt- 15 % WOx/Zeolite Y catalyst in the form of pellets. Due to harsh reaction conditions, deactivation of the hydrogenation catalyst can pose a problem in industrial catalytic processes because of limited lifetime and expensive catalyst replacement. The most common deactivation mechanisms of catalysts that are used in hydrogenation processes of liquid hydrocarbons are poisoning and fouling. While fouling (e.g. mechanical coke deposition) is mostly observed as a deactivation mechanism for porous materials (e.g. zeolites), poisoning (e.g. sulfur poisoning) is most frequently the reason for deactivation of metals. The aim of this work was the detailed characterization of catalysts used in glycerol hydrogenation and the evaluation of various regeneration methods and their impact on catalyst characteristics. For this, the most important catalyst properties such as metal and overall specific surface area, (crystal) structure, acidity etc. were investigated. The used catalysts were compared to the fresh catalyst. All of the studied characteristics play a big role in the performance of the catalyst and needed to be investigated for fresh, used and regenerated catalysts to determine the origin and extent of catalyst deactivation and to evaluate the success of the recovery treatments and whether the deactivated state of the catalyst could be reversed.In order to reach a good catalytic performance, the active sites, in this case platinum and acid sites, need to be accessible for reaction compounds.Because of deactivation processes, e.g. sintering and coking, active sites may be entrapped in the bulk of the catalyst which may lead to loss of catalyst activity. This is the reason why the characterization of active sites is crucial for catalysis.The regeneration of the used catalysts was studied by two different approaches: thermal treatment and “solvent method”. For the thermal treatment different temperatures, different heating rates and different gas conditions were applied to the catalyst. The temperature range for the thermal treatment was 300 °C – 550 °C and different atmospheres were He, 20 % O2 and 20 % H2. The “solvent method” used different liquids such as water, ethanol and pentane for washing the pellets at room temperature. In addition to these two separate methods, combined regeneration treatments were also performed. According to characterization with CO chemisorption on Pt, N2 physisorption, IR spectroscopy of pyridine adsorption on acid sites and structure analysis by ATR- IR and XRD, the most promising results were achieved with a temperature of 350 °C and 20 % O2. It appears that the organic residues (e.g. due to coking) could only be removed in an oxidizing atmosphere. Higher temperatures also showed no signs of better regeneration. Compared to thermal treatment washing the catalyst pellets with different solvents appeared not as successful regarding the recovery of active sites. Among the four different solvents, pentane appeared to be suited best for carrying out the washing experiments.The results of the combined recovery methods showed no benefit. By all approaches, it was not possible to reach the original characteristics of the fresh catalysts as it appeared that the damage of the catalytic process was too severe.