The research will provide a fundamental understanding of the factors controlling the adsorption of organic compounds by carbon nanotubes. Quantitative structure-activity relationship (QSAR) models will be developed. There are thousands of organic compounds produced and used worldwide. This research will predict the adsorption of organic compounds by carbon nanotubes in water without the need for performing costly experiments. The approach is to systematically investigate the adsorption and aggregation of carefully selected carbon nanotubes with a wide spectrum of diameter, length, number of walls, and surface functional groups. These will be used with a suite of organic probe molecules with different physicochemical properties in controlled background solution matrices. Carbon nanotubes with a wide range of structural characterists will be extensively characterized and isotherm experiments will be performed to systematically examine the roles of nanotube structural parameters and organic compound physicochemical properties in the adsorption process. Carbon nanotube manufacturers will be able to use these models to develop environmentally sustainable approaches in their production of carbon nanotubes toward different applications.
This study successfully developed predictive quantitative structure-adsorption models for the small molecular weight organic compounds (SOCs) by carbon nanotubes (CNTs). In Figure 1 we provide one example of predicted vs. experimentally determined carbon nanotube adsorption descriptors. The predictive strength and accuracy of generated models were suitable for assessing how well CNTs adsorb SOCs, making them useful in predicting the fate of organic contaminants in natural systems contaminated with carbon nanotubes and for the adsorption of organic chemicals from contaminated waste streams in engineered treatment systems. Testing adsorption of SOCs by CNTs is costly, time consuming and laborious. The combined model we developed for 58 compounds is the most comprehensive model in literature for predicting the adsorption of aromatic contaminants by carbon nanotubes. Models like this are useful in reducing the testing adsorption of organic pollutants via CNTs to accelerate the testing of nearly 70,000 other anthropogenic contaminants. Our models also gave us mechanistic insights into how CNTs adsorb SOCs. In our combined evaluation of aromatic and aliphatic compound modeling, we found that the adsorption of aromatic SOCs by CNTs strongly depended on the hydrophobicity. However, we found that hydrophobic driving forces and possible other interactions such as polarizability contributed to adsorption for aliphatic SOCs. We also investigated the influence of the morphological properties of carbon nanotubes on SOC adsorption. Note the electron microscopy images of the variability of morphological properties of CNTs in Figure 2. Although the specific surface area of our CNTs did control adsorption, especially at higher SOC concentrations, we also observed that multiwall CNTs with larger outer diameters exhibited a higher adsorption capacity for phenanthrene, attributed to the better alignment of benzene rings of the phenantherene molecules with the CNT surfaces. As the phenanthrene concentrated increased the competition for available sorption sites also increases. Therefore at higher concentrations the controlling factor for adsorption shifted from the outer diameter to a specific surface area. This shift clearly indicates the importance of understanding how the morphological features of CNTs affect the adsorption of SOCs. We also compared SOC adsorption of SOCs with CNTs with both graphenes and activated carbons in natural waters. In a study of graphene oxide and graphene nanosheets, we found that natural organic material had less of an effect on SOC adsorption capacity than activated carbon. Note the decrease in percentage (in Figure 3) from competition with natural organic matter molecules. Although the microporous structure of the activated carbon (represented by HD 4000) lost more than 90% of its adsorption capacity from NOM competition, graphene oxide only showed an approximate 20% loss in adsorption capacity. This indicates that the sheet like graphene oxide particles forming a "tree of cards" like structure poses a porous structure with larger openings reducing the competition between small phenantherene molecules and bulky natural organic mater molecules. Although graphene SOC adsorption capacity in the tested natural waters was found to be comparable to those of CNTs and activated carbon for aromatic compounds, graphenes showed lower adsorption capacities for the tested aliphatic SOCs as compared to SWCNTs. Additional research is warranted to further assess the adsorption of SOCs by graphenes.
