| dc.description.abstract |
In this thesis, the effectiveness of silica-supported cobalt (Co/SiO ) catalysts for Fischer-Tropsch 2 (FT) synthesis, a useful process for the conversion of syngas into clean fuels, was investigated. Emphasis was placed on how the catalyst performance is significantly influenced by the size of silica pores, aiming to decipher the pore size effects on CO conversion and product selectivity in FT synthesis. The Stöber process was used to create the mesoporous silica supports with pores that could be varied from 14 to 28 nm by adjusting the H O: TEOS: NH₄OH: EtOH ratios in the 2 synthesis reaction vessel. The wet impregnation method was used to load 10 wt% cobalt on the silica supports. After loading them with cobalt, 620°C calcination temperature was determined by Thermogravimetric Analysis (TGA) for the calcination of the prepared catalyst samples. The analysis of pore sizes was done using Brunauer–Emmett–Teller (BET), morphology was verified by Scanning Electron Microscopy (SEM), Energy-Dispersive X-ray Spectroscopy (EDS), and X-Ray Diffraction (XRD), the available functional groups were examined by Fourier Transform Infrared Spectroscopy (FTIR), reducibility was tested by Temperature-Programmed Reduction (TPR), and dispersion was determined by using H chemisorption. The catalyst testing on FT 2 synthesis was done, which involved 150-hour runs at 220°C and 20 bar in a fixed-bed reactor, which resulted in CO conversions of 2–23%. The narrative became evident that the reducibility, product distribution, and CO conversion are all affected by the variation of pore size. Mid-sized pores (15–20 nm) were excellent at maximising CO conversion while minimizing light hydrocarbons (C -C ). Larger pores (>20 nm) had the highest selectivity for heavy C chains,
1 4 5+ which was up to 88%. Machine learning was used to enhance these results, training CatBoost on 400 data points from 60 research articles covering 79 variables, including loads and temperatures. It supported our laboratory studies and identified 14–18 nm pores as the optimal zone for the production of long-chain hydrocarbons with a good prediction ability (R² = 0.79). The balance between the catalyst’s stability, selectivity, and activity can be strongly influenced by pore structure. This work provides practical ways for synthesizing optimal performing Co/SiO₂ catalysts by combining rigorous experimental catalyst testing, characterisation, and implementing machine learning. |
en_US |