Factors affecting the behavior of unburned carbon upon steam activation

by 1973- Lu, Zhe

iii The main objective of this study is to investigate the factors that could affect the behavior of unburned carbon samples upon steam activation. Through this work, the relationships among the factors that could influence the carbon-steam reaction with the surface area of the produced activated carbon were explored. Statistical analysis was used to relate the chemical and physical properties of the unburned carbon to the surface area of the activated carbon. Six unburned carbons were selected as feedstocks for activated carbon, and marked as UCA through UCF. The unburned carbons were activated using steam at 850oC for 90 minutes, and the surface areas of their activated counterparts were measured using N2 adsorption isotherms at 77K. For all the unburned carbons investigated in this study, the surface areas of their activated counterparts increased with carbon burn-off levels, and reached a maximum, as a result of the development of porosity by burning out of active carbon structure and opening of the blocked pores, and then decreased as the burn-off level increased due to the pore widening and the destruction and merging of the pore walls between the micropores. The activated carbons produced from different unburned carbon precursors presented different surface areas at similar carbon burn-off levels. Moreover, in different carbon burn-off regions, the sequences for surface area of activated carbons from different unburned carbon samples were different. The factors that may affect the carbon-steam gasification reactions, including the concentration of carbon active sites, the crystallite size of the carbon, the intrinsic porous structure of carbon, and the inorganic impurities, were investigated. All unburned carbons iv investigated in this study were similar in that they showed the very broad (002) and (10) carbon peaks, which are characteristic of highly disordered carbonaceous materials. Unburned carbons were predominantly mesoporous except for sample UCD, which was mainly microporous. It seems that the rapid devolatilization that coals underwent in the combustor promoted the generation of meso- and macropores. In this study, the unburned carbon samples contained about 17-48% of inorganic impurities. Compared to coals, the unburned carbon samples contain a larger amount of inorganic impurities as a result of the burn-off, or at lease part, of the carbon during the combustion process. These inorganic particles were divided into two groups in terms of the way they are associated with carbon particles: free single particles, and particles combined with carbon particles. The analytical data for these factors, as well as those obtained from the author’s previous M.S. work, such as the results of proximate, ultimate, and optical texture analyses, were used for statistical analysis using the main effect plot and fitting line plot methods in this study. The results of the statistical analyses showed that the surface areas of the activated carbons increased with the concentration of active sites, which was measured as the O2 uptakes, the concentration of C-O species, and crystallite diameters, in all carbon burn-off zones for samples UCA, UCD, UCE, and UCF. This indicated that these parameters of unburned carbons played the most significant role on the development of surface areas of the activated carbons. A good correlation was observed between the surface areas of activated carbons and the burn-off levels of the unburned carbons, except for samples UCB and UCC in the burn-off levels of 10-45%, and 45-60%, and this v correlation was good for all unburned carbons when the burn-off was > 60%. This could be attributed to the fact that samples UCB and UCC had the highest concentration of Na+K combined with carbon particles, which can facilitate the development of the surface areas of activated carbons. Moreover, the highest content of isotropic carbon composition that sample UCB had could also lead to the highest increase in surface areas of activated UCB. Furthermore, the regression equations given by fitted line plot method can be used to predict the surface area of activated carbon from the concentration of active sites of the unburned carbon precursors, where high R2 of >90% was obtained, and R2 value increased with carbon burn-off levels. As indicated from the present work, unburned carbons with one of the following properties will produce activated carbons with high surface areas. These properties include: (a) large amount of O2 chemisorption capacity; (b) high concentration of surface C-O complex; and (c) small crystallite diameter; (d) high concentration of Na+K particles that are combined with carbon; (e) high concentration of isotropic carbon.