The mitigation of climate change is one of the major global challenges in the 21st Century. Carbon capture and storage (CCS) is a promising technology to effectively reduce anthropogenic CO2 emissions into the Earth’s atmosphere. There are various candidate materials for CO2 capture but each has its own advatanges and disadvantages. Carbon-based materials are of low-cost and have relatively high cyclicity for CO2 and its porous structure and surface functional groups can be tailored to improve CO2 capture performance. Effective but low-cost carbon precursors need to be explored for potential mass production in the furture. This research aims to explore various polymeric, biomass and graphitic materials as the precursors for the development of effecive carbon sobents for CO2 capture. In addition, the influence of porous structures and chemical dopants on CO2 sorption are also experimentally studied in relation to the porosities and surface chemistry of the sorbents. Five distinct synthesis approaches are explored comparatively to determine the potential of polymeric, biomass and graphitic materials as precursors for effective carbon sorbents. These approaches include a novel method of producing millimetre-sized carbon spheres from poly(acrylonitrile-co-acrylamide)/DMSO solution, chemical activation of London Plane leaves, spruce pine cones and graphite oxide, and ball-milling of graphite. The work on the polymer-derived carbon spheres produced desirable carbon macro-spheres with radially channelled and hierarchically porous structures, via a “one-pot” solvent exchange process. The structure shows excellent CO2 capacity of 16.7 wt% at 25 °C and under 1 bar CO2, enhanced by rich nitrogen doping and microporosities. The biomass-derived carbon sorbents further clarify the influence of metal-dopants, inherited from the biomass precursors, on CO2 adsorption. It was noted that besides nitrogen dopant and ultramicropores (<0.7 nm), residual calcium and magnesium in biomass-derived carbon also enhanced CO2 adsorption on carbon sorbents. The CO2 uptake of a pine cone-derived carbon sorbent (20.9 wt%) has matched the highest CO2 uptake (21.2 wt%) reported in the literature at 25 °C and under 1 bar CO2, though the latter has a relatively large ultramicropore volume. To further clarify the influence of microporous structure and chemical dopants on CO2 uptake, graphite oxide (GO) and ball-milled graphite (BG) were studied as graphitic precursors as these have known chemical structures and their resulting sorbents contain no other chemical dopants. The characterisation results show chemical activation with potassium hydroxide can develop a similar porous structure in GO- and BG-derived carbon, compared with those of polymer- and biomass- derived carbon. However, the former show comparatively lower CO2 capture capacities under the same test conditions (25 °C and 1 bar CO2), which is believed to be due to less well-developed ultramicroporous structure and the absence of chemical dopants. Based on the present experimental data, further analysis reveals that there is a difference between specific surface area calculated by the Brunauer-Emmett-Teller (BET) equation and the Density Functional Theory (DFT) model. The cause of this is the intrinsic difference in the method of calculation, where the BET equation assumes a flat and homogeneous surface, while the DFT model takes the pore shape into consideration. Furthermore, both CO2 uptake and specific CO2 uptake (CO2 uptake/porosity) are plotted against three porosity parameters, namely BET surface area, total pore volume and ultramicropore volume. The plots show those samples with higher nitrogen and metal contents exihibit higher specific CO2 uptakes. To extend the interpretation of results, an Artificial Neural Network (ANN) is adopted as a simulation tool to study the influence of ultramicropore, nitrogen and metal dopants on CO2 uptake. Characteristic results from both the present work and the literature are used as the input data for the simulation. The simulated results show CO2 uptake increases considerably with increasing ultramicropore volume and metal content. However, the nitrogen content has relatively limited influence, compared with the former two, contrary to common belief. Finally, several future lines of work are proposed to further improve the performance of the materials. For the synthesis of carbon spheres, DMSO can be added into the water bath to slow the solvent exchange process. As a result, the macroporous strucure of the sphere can be modified to enhance its mechanical strength. For the study on the biomass-derived carbon, several other leaves can also be used as carbon precursors. The porosities, chemical compositions and CO2 uptakes of the cabon sorbents derived from these leaves can be compared with those of the London Plane leaf-derived carbon, to further clarify the influence of the biological structures and chemical properties of biomass precursors on the resulting carbon. For the work on ANN, the simulation is limited by the available experimental data for metal-doped carbon. The accuracy of prediction by ANN can be further improved when more experimental data are reported in the literature and used for training the network.
|University College London (University of London)
|Electronic Thesis or Dissertation
Page generated in 0.0104 seconds