Use this URL to cite or link to this record in EThOS:
Title: Catalytic mild hydrogenation of pygas
Author: Sutherland, Luke Malcolm
ISNI:       0000 0004 6493 9914
Awarding Body: University of Glasgow
Current Institution: University of Glasgow
Date of Award: 2018
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Thesis embargoed until 22 Jan 2022
Access from Institution:
In the production of ethylene and propylene via steam cracking, one of the major by-products is pygas. This mixture has a high octane number, owing to the large quantity of aromatics, diolefins, and olefins present within this mixture. This stream contains, in particular, a significant quantity of benzene and toluene. This is a waste product, but is currently heavily utilised for the extraction of benzene, mainly in order to produce styrene, cumene, and cyclohexane. While the demand for these compounds will continue to increase, the usage of pygas products will decrease in terms of its other main use, as a petrol additive, due to the increasingly strict regulation regarding the total allowable content of aromatics in fuel. When attempting to refine pygas for use as an aromatic chemical feedstock, the most common method of purification is to perform a hydrocracking reaction. Reactions will usually be carried out in two steps; the first is to hydrogenate the diolefins and styrene over, typically, a Pd-alumina catalyst under mild conditions. The second stage involves hydrogenating olefins, removing sulfur compounds by converting them into H2S, acid-catalysed ring opening of napthenics, and cracking paraffins. The aim of this research was to perform an analysis of the effect of reaction mixture on the retention of benzene when passed over an industrially-used bifunctional metal/zeolite hydrocracking catalyst, and to analyse the carbon laydown and other relevant effects produced by altering the reaction mixture. Previous work looking at pygas has mostly been carried out using only one, or a few, reaction compounds for simplicity of analysis. A number of studies use only styrene as a model for pygas. This is an excessively simplistic model of the reaction, and neglects the interactions between the various components of pygas present in a real reaction setting. Therefore, within this research, a mixture of alkanes, cycloalkanes, and aromatics are used to make a model reaction feed. These were reacted over the hydrocracking metal/zeolite catalyst, and an as-prepared zeolite catalyst. This reaction mixture model is more comprehensive in scope than most research performed, without also including olefins, which would accelerate coking of the catalyst, therefore obscuring the more basic interactions between aromatic and saturated paraffin compounds. The efficacy of each reaction mixture was measured by running a model feed based on common pygas compositions in industry, then running reactions in which a single one of four of the the six feed components was removed from the mixture. When the as-prepared zeolite support was used as a catalyst it was found to crack more of the aromatics, benzene and toluene, along with producing significantly more xylenes due to disproportionation, than the metal/zeolite catalyst. One of the main causes of catalyst deactivation in hydrocracking catalysts is coking, due to carbon deposition. The effect of coking was analysed using thermogravimetric analysis (TGA) ex-situ, which was also verified using CHN analysis. The as-prepared zeolite support was found to produce more carbon laydown than the metal/zeolite catalyst, although it is unclear if the difference would have a significant effect during longer reactions. Within the two groups of catalyst, metal/zeolite and zeolite, the total quantity of coke detected for each reaction mixture was found to show some variation, but would be similar. Some of the reaction mixtures showed a difference in product composition in the offline GC results, although the cause of this is unclear. The results in these reactions cannot, therefore, be anticipated based upon a simple addition-subtraction model in terms of feed components, as there appears to be a complex interdependence between these components that influences the final retention of products observed, that was not assumed to be present before this work was carried out. Analysis of the coke deposit by Raman spectroscopy revealed that the graphite platelets in all reactions were between 10 and 13 nm. Therefore, it appears coke deposition was observed only on the external surface of the catalyst.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
Keywords: QD Chemistry