We demonstrate the importance of London dispersion forces in defining the adsorption capacity within expanded graphite, a simple model of the more complex experimental geometries of activated carbon, using a combination of the non-local correlation functional of Dion et al paired with a recent exchange functional of Cooper (vdW-DF(C09x)) and a classical continuum model. Our results indicate that longer ranged interactions due to dispersion forces increase the volume over which molecules interact with a porous medium. This significantly enhances the adsorption density within a material, and explains recent experimental work showing that the densification of H(2) in carbon nanopores is sensitive to the pore size. Remarkably, our slit pore geometries give adsorption densities of up to 3 wt% at 298 K and 20 MPa which correlates well with experimental values for 9 Å pores-a value that could not be predicted using local density approximation (LDA) calculations. In its entirety, this work presents a powerful approach for assessing molecular uptake in porous media and may have serious impacts on efforts to optimize the properties of these materials.

This paper calculates that the theoretical hydrogen uptake in nanoporous carbons is close to 0.5 wt % at 298 K and 5 MPa, higher than most reported values in activated carbons. The isosteric heats of adsorption for nanoporous carbons and for an expanded graphite model are between 14 and 18 kJ/mol, close to the suitable energy range for practical hydrogen storage (15-40 kJ/mol). Over the density ranges examined, total hydrogen adsorption can be improved by increasing the volume available for adsorption in amorphous carbons. These calculations are performed by using an efficient and accurate method. This method can reproduce previous, more computational intensive calculations in the expanded graphite model yet is readily applicable to more complex geometries. The limitations of this method are discussed carefully; under conditions given above, these limitations are minimal.

This paper uses an efficient and accurate approach to estimate the hydrogen physisorption in various carbon structures. Compared with previous Grand Canonical Monte Carlo (GCMC) and other methods applied to expanded graphite, the method introduced here is shown to be accurate, but the calculation is much faster and more intuitive. Our preliminary results for amorphous carbons show reasonably high hydrogen uptake close to 0.8 wt-% at T=300 K and P=5 MPa.