Iván Cabria

Enhancement of hydrogen physisorption on graphene and carbon nanotubes by Li doping

Density-functional calculations of the adsorption of molecular hydrogen on a planar graphene layer and on the external surface of a (4,4) carbon nanotube, undoped and doped with lithium, have been carried out. Hydrogen molecules are physisorbed on pure graphene and on the nanotube with binding energies about 80-90 meV/molecule. However, the binding energies increase to 160-180 meV/molecule for many adsorption configurations of the molecule near a Li atom in the doped systems. A charge-density analysis shows that the origin of the increase in binding energy is the electronic charge transfer from the Li atom to graphene and the nanotube. The results support and explain qualitatively the enhancement of the hydrogen storage capacity observed in some experiments of hydrogen adsorption on carbon nanotubes doped with alkali atoms.

Density functional calculations of hydrogen adsorption on boron nanotubes and boron sheets

Hydrogen adsorption on the recently discovered boron nanotubes, BNTs, and on boron sheets is investigated by density functional calculations. Both molecular physisorption and dissociative atomic chemisorption are considered. The geometric and electronic structures of BNTs and boron sheets have been elucidated. These two novel boron structures present buckled surfaces with alternating up and down rows of B atoms, with a large buckling height of about 0.8 A. The buckled structures are about 0.20 eV/atom more stable than the corresponding flat ones. However, the helicity of some BNTs does not allow for the formation of alternating up and down B rows in the surface and, therefore, these nanotubes have flat surfaces. The buckled and flat nanostructures have different geometric and bonding characteristics, but both are metallic. Molecular hydrogen physisorption energies are about 30–60 meV/molecule on boron sheets and nanotubes, actually lower than in graphene and in carbon nanotubes and far from the energies of 300–400 meV/molecule necessary for efficient hydrogen storage at room temperature and moderate pressures for onboard automotive applications. Chemisorption binding energies on BNTs are about 2.4–2.9 eV/H atom, similar to the ones obtained in CNTs. Finally, the energy barrier from molecular physisorption to dissociative chemisorption of hydrogen is about 1.0 eV /molecule. Therefore, the calculations predict physisorption as the leading adsorption mechanism of hydrogen at moderate temperatures and pressures. The expected hydrogen adsorption capacity of these novel B materials is even smaller than that of CNTs.

Hydrogen storage in pure and Li-doped carbon nanopores: Combined effects of concavity and doping

Density functional calculations are reported for the adsorption of molecular hydrogen on carbon nanopores. Two models for the pores have been considered: (i) The inner walls of (7,7) carbon nanotubes and (ii) the highly curved inner surface of nanotubes capped on one end. The effect of Li doping is investigated in all cases. The hydrogen physisorption energies increase due to the concavity effect inside the clean nanotubes and on the bottom of the capped nanotubes. Li doping also enhances the physisorption energies. The sum of those two effects leads to an increase by a factor of almost 3 with respect to the physisorption in the outer wall of undoped nanotubes and in flat graphene. Application of a quantum-thermodynamical model to clean cylindrical pores of diameter 9.5 A, the diameter of the (7,7) tube, indicates that cylindrical pores of this size can store enough hydrogen to reach the volumetric and gravimetric goals of the Department of Energy at 77 K and low pressures, although not at 300 K. The results are useful to explain the experiments on porous carbons. Optimizations of the pore size, concavity, and doping appear as promising alternatives for achieving the goals at room temperature.

Competition between molecular and dissociative adsorption of hydrogen on palladium clusters deposited on defective graphene

The contribution of Pd doping to enhance the hydrogen storage capacity of porous carbon materials is investigated. Using the Density Functional Formalism, we studied the competition between the molecular adsorption and the dissociative chemisorption of H2 on Pd clusters anchored on graphene vacancies. The molecular adsorption of H2 takes place with energies in the range of 0.7–0.3 eV for adsorption of one to six hydrogen molecules. Six molecules saturate the cluster, and additional hydrogen could only be adsorbed, with much smaller adsorption energies, at farther distances from the cluster. Dissociative chemisorption is the preferred adsorption channel from one to three hydrogen molecules, with adsorption energies in the range of 1.2–0.6 eV. After the first three molecules are dissociatively chemisorbed, three additional hydrogen molecules can be adsorbed non-dissociatively onto the Pd cluster with adsorption energies of 0.5 eV. The desorption of Pd–H complexes is prevented in all cases because the Pd clusters are firmly anchored to graphene vacancies. Our results are very promising and show that Pd clusters anchored on graphene vacancies retain their capacity to adsorb hydrogen and completely prevent the desorption of Pd–H complexes that would spoil the hydrogen releasing step of the cycle.

Simulated porosity and electronic structure of nanoporous carbons

Nanoporous carbon refers to a broad class of materials characterized by nanometer-size pores, densities lower than water, large specific surface areas, and high porosities. These materials find applications in nanocatalysis and gas adsorption, among others. The porosity structure, that determines the properties and functionalities of these materials, is still not characterized in detail. Here, we reveal the detail porosity structure and the electronic properties of a type of nanoporous carbons, the so called carbide derived carbons (CDCs), through a simulation scheme that combines large simulation cells and long time scales at the empirical level with first-principles density functional calculations. We show that the carbon network consists in one layer thick nanographenes interconnected among them. The presence of specific defects in the carbon layers (heptagons and octagons) yields to open pores. These defects are not completely removed through annealing at high temperatures. We also suggest that, in contrast with graphene which is a zero-gap semiconductor, these materials would have a metallic character, since they develop an electronic band around the Fermi level. This band arises from the electronic states localized at the edges of the nanographene layers.

MRI segmentation fusion for brain tumor detection

Abstract The process of manually generating precise segmentations of brain tumors from magnetic resonance images (MRI) is time-consuming and error-prone. We present a new algorithm, Potential Field Segmentation (PFS), and propose the use of ensemble approaches that combine the results generated by PFS and other methods to achieve a fused segmentation. For the PFS method, we build on our recently proposed clustering algorithm, Potential Field Clustering, which is based on an analogy with the concept of potential field in Physics. We view the intensity of a pixel in an MRI as a “mass” that creates a potential field. Specifically, for each pixel in the MRI, the potential field is computed and, if smaller than an adaptive potential threshold, the pixel is associated with the tumor region. This “small potential” segmentation criterion is intuitively valid because tumor pixels have larger “mass” and thus the potential of surrounding regions is also much larger than in other regions of smaller or no “mass”. We evaluate the performance of the different methods, including the ensemble approaches, on the publicly available Brain Tumor Image Segmentation (BRATS) MRI benchmark database.

Interaction of narrow carbon nanotubes with nitronium tetrafluoroborate salts

Density functional calculations have been performed to investigate the destruction of narrow carbon nanotubes (CNTs) under the attack of nitronium tetrafluoroborate salts. The dissociation of these salts in a solvent produces nitronium and tetrafluoroborate ions which coadsorb on the external surface of the tubes. It is shown that the ions bind strongly to both metallic and semiconducting narrow nanotubes, although stronger to the metallic ones. The nitronium cations bind to the CNTs through a charge transfer mechanism, whereas the tetrafluoroborate anions remain negatively charged upon adsorption on the nanotubes. The surface of the nanotubes gets substantially deformed around the adsorption site of the nitronium ion, but it is hardly changed around the adsorption site of the tetrafluoroborate ion. These results are the theoretical basis to explain the destruction of the narrow CNTs found in the experiments and also to unravel, in agreement with the experimental interpretation, the distinct role played by the nitronium and the tetrafluoroborate ions. The tetrafluoroborate ions contribute to separate the CNTs from the bundles into individual tubes, without affecting the tubes. The nitronium ions, in contrast, modify the electronic and geometrical structures of the narrow tubes leading eventually to their destruction. The implications for the selective removal of intermediate diameter metallic CNTs found in the experiments are also discussed. The adsorption of the neutral nitrogen dioxide molecule is also studied, and the results show that the weak interactions of this molecule with both metallic and semiconducting tubes cannot be used as a model for the strong attack of the nitronium ions to the narrow tubes. The sensor effect of the nanotubes toward adsorption of nitrogen dioxide is also discussed.

A Mean Shift-Based Initialization Method for K-means

Because of its conceptual simplicity, k-means is one of the most commonly used clustering algorithms. However, its performance in terms of global optimality depends heavily on both the selection of k and the selection of the initial cluster centers. On the other hand, Mean Shift clustering does not rely upon a priori knowledge of the number of clusters. Furthermore, it finds the modes of the underlying probability density function of the observations, which would be a good choice of initial cluster centers for k-means. We present a Mean Shift-based initialization method for k-means. A comparative study of the proposed and other initialization methods is performed on two real-life problems with very large amounts of data: Facility Location and Molecular Dynamics. In the study, the proposed initialization method outperforms the other methods in terms of clustering performance.

Searching for DFT-based methods that include dispersion interactions to calculate the physisorption of H2 on benzene and graphene

Simulations of the hydrogen storage capacities of nanoporous carbons require an accurate treatment of the interaction of the hydrogen molecule with the graphite-like surfaces of the carbon pores, which is dominated by the dispersion forces. These interactions are described accurately by high level quantum chemistry methods, like the Coupled Cluster method with single and double excitations and a non-iterative correction for triple excitations (CCSD(T)), but those methods are computationally very expensive for large systems and for massive simulations. Density functional theory (DFT)-based methods that include dispersion interactions at different levels of complexity are less accurate, but computationally less expensive. In order to find DFT-methods that include dispersion interactions to calculate the physisorption of H2 on benzene and graphene, with a reasonable compromise between accuracy and computational cost, CCSD(T), Møller-Plesset second-order perturbation theory method, and several DFT-methods have been used to calculate the interaction energy curves of H2 on benzene and graphene. DFT calculations are compared with CCSD(T) calculations, in the case of H2 on benzene, and with experimental data, in the case of H2 on graphene. Among the DFT methods studied, the B97D, RVV10, and PBE+DCACP methods yield interaction energy curves of H2-benzene in remarkable agreement with the interaction energy curve obtained with the CCSD(T) method. With regards to graphene, the rev-vdW-DF2, PBE-XDM, PBE-D2, and RVV10 methods yield adsorption energies of the lowest level of H2 on graphene, very close to the experimental data.

Automated segmentation of brain tumors in MRI using potential field clustering

We propose potential field clustering, a new algorithm based on an analogy with the concept of potential field in Physics. By viewing the intensity of a pixel in a FLAIR MRI image as a “mass” that creates a potential field, the algorithm is used for tumor localization. The center of the localized tumor cluster is then used as the initial seed in a region growing segmentation algorithm. We evaluate the performance of this segmentation approach on the publicly available brain tumor image segmentation MRI benchmark. The performance of the proposed approach is compared with that of the Force clustering algorithm by Kalantari et al. (2009). Experimental results show that the proposed algorithm is more accurate in localizing tumor centers, which, in turn, results in better segmentations.