Yijing Zheng, Department of Civil and Environmental Engineering, University of Tennessee, Knoxville, 70 Perkins Hall, Knoxville, TN 37996-2010, Pyoungchung Kim, Civil and Environmental Engineering, University of Tennessee, Knoxville, 70 Perkins Hall, Knoxville, TN 37996-2010, and Sandeep Agnihotri, Civil and Environmental Engineering, University of Tennessee, 73A Perkins Hall, Knoxville, TN 37996-2010.
We have developed a microporosity and surface area estimation method that is specific to single-walled carbon nanotubes (SWNTs). Typically, micropore volume of adsorbents can be determined by the t-plot method, where the statistical thickness, t, of adsorbed layer is a function of the adsorbed amount, and is estimated by several equations such as de Boer equation, Halsey equation and Carbon Black equation. The t-plot method is one of the most common and versatile method. The estimated micropore volume is often specific to a model and to the segment of adsorption isotherm that is fitted to that model. SWNTs are relatively new adsorbents because of which applicability of these models to carbon nanotubes needs to be evaluated. We have developed a technique that compares experimental N2 adsorption isotherm (77 K) of a sample with grand canonical Monte Carlo (GCMC) simulation isotherms of N2 on the external surface of SWNT bundles (10-6 < P/Po < 0.99). In this method, we plot the total experimental adsorption capacity vs. simulated external adsorption capacity from SWNT bundles. The plot is more or less a straight line for adsorption at P/Po Ан 10-3. The slope and intercept of the linear asymptote of the curve are interpreted as the external surface area of the bundles and the available micropore volume, respectively. We are also able to incorporate adsorption from impurities by modeling impurities as planar carbons and adding it to the total non-endohedral contributions. The main advantages of our method is that it is specific to SWNTs and, unlike other methods, it includes the entire experimental adsorption isotherm (10-6 < P/Po < 0.99) in analysis, thus minimizing any experimental errors related to selecting an appropriate segment of an experimental isotherm to fit the data.