Carbon capture under post-combustion conditions has been the topic of numerous studies in the last decade. Although exhaust gases typically contain different components other than CO2, they are commonly neglected in these studies. The presence of sulfur dioxide, for example, tends to interfere in the carbon capture process through different mechanisms. The present work aimed to evaluate the effects of SO2 on the CO2 retention capacity through the integration of experimental evaluation, column dynamics modeling and molecular simulation (multiscale modeling) for an activated carbon sample under typical conditions found in the post-combustion environment. Results indicate that, under typical post-combustion flue gas conditions, SO2 has little influence on the CO2 retention capacity of the carbon in comparison to other adsorbent materials. The highest concentration of SO2 (5 000 ppmv) led to a decrease of approximately 10 % in the CO2 capture capacity. Significant deactivation (around 40 %) was observed experimentally and by molecular simulation only for very high concentrations of SO2 (50 000 ppmv). Additionally, no evidence of reactions with the material was found and both captured components could be completely desorbed by means of suitable regeneration methods. To access individual pore performance, molecular simulations were implemented for SO2 adsorption using, for the first time, a rigorous heterogeneous pore model (rMD). The results revealed a large deactivation for the 7.0 Å pore, a surprising cooperative effect for the 8.9 Å pore, and indifference for the larger 18.5 and 27.9 Å pores. For SO2 concentrations up to 5 000 ppmv, the use of carbon-based adsorbent could rule out the need of a pre-treatment operation to remove SO2 in carbon capture processes. For higher concentrations, molecular simulation showed that tailoring carbon porosity in the range of 7 Å to 8.9 Å considerably reduces the interference of SO2. Results also point out that the IAST and Langmuir models diverge from the molecular simulation results, particularly at low loadings (up to 5 000 ppmv SO2), indicating the need for caution when applying these models in systems where competitive interactions between molecules are relevant, as is the case with mixtures of CO2 and SO2.
Read full abstract