Transition metal and nitrogen doped carbon nanostructures
This article is not included in your organization's subscription. However, you may be able to access this article under your organization's agreement with Elsevier.
Stanislav R. Stoyanova, b, Alexey V. Titova and Petr Krála, ,
aDepartment of Chemistry, University of Illinois at Chicago, Chicago, IL 60607, USA
bNational Institute for Nanotechnology, National Research Council of Canada and Department of Mechanical Engineering, University of Alberta, Edmonton, AB, Canada
Received 21 January 2009;
accepted 11 March 2009.
Available online 24 March 2009.
Abstract
We review our theoretical first-principle studies of carbon nanostructures based on graphene sheets, carbon nanotubes, nanocones and fullerenes that are substitutionally doped with transition metal and nitrogen atoms. The results obtained show that metal doping leads to more stable systems in buckled rather than planar structures. The hybrid structures have low-lying excited states, allowing for catalytic activity, in analogy to metalloporphyrins and metallophthalocyanines, as confirmed in recent experiments with Fe-xN-doped carbon nanotubes. Metal-doped carbon nanocones and nanocapsules based on typical fullerenes manifest remarkable electronic and spin polarizations. Additional doping by boron atoms adjacent to the metals increases their HOMO–LUMO gap, stabilizes their electronic structures and causes that their ground states have higher spin multiplicity, where the spin density is spread over the systems. The metallic sites allow functionalization and potential activation of these nanosystems. The hybrid structures formed can have a broad range of applications in catalysis, molecular electronics, light-harvesting and nanomechanics.
Fig. 1. (top) Nitrogen atom substitution into (a) planar graphene, (b) (7,0) CNT, and (c) (5,5) CNT. (bottom) The Fe–4N substitution sites for (f) graphene sheet and (d,e,g,h) (7,0) CNT clusters. Coordinations of Fe to pyridine-like and pyrrole-like configurations are shown in (d,g) and (e,h), respectively. The configurations are in axial (d,e) and helical orientations (g,h). Terminal hydrogen atoms are not shown.
Fig. 2. Nitrogen substitution energies Eq. (1) for the graphene and CNT clusters. On the horizontal axis, the configurations of the two N substitutions are shown for the zigzag/armchair nanotube clusters, as schematically represented by the six-membered carbon rings on the top (bottom) of the graph for the armchair (zigzag) nanotube clusters. The symmetry axis of the CNTs is oriented horizontally. In the flat graphene, the N configurations are double degenerate.
Fig. 3. Fe binding energies for the planar and curved clusters containing 2 N atoms. The structures are shown schematically on the graph by the six-membered carbon rings and the CNT axis is oriented horizontally.
Fig. 9. Electron spin density (ESD) surface (±0.0004 e/bohr3) of capsules 5 (top left), 23 (top middle), 7 (top right) and and 24 (bottom left). Electrostatic potential (ESP) of capsule 24 in the interval (−0.04,+0.04)e mapped on the 0.0004 e/bohr3 total electron density surface (bottom right).
Fig. 11. Experimental and calculated absorption coefficient functions χ(k) for (a) flat graphene clusters and (b) bent clusters. (c) Linear combination of the Fe–2N and Fe–4N spectra for structures with the lowest substitution energy optimized to fit to the χexp(k). The resulting curve contains 22.26% and 77.74% contributions from the Fe–2N and Fe–4N configurations, respectively.
Fig. 12. (up left) Top view of the Ni-doped cone and (up right) (4,4) CNT. (down) Extended MOs of the Ni-doped cone with a large metallic content. (left to right) The MOs with large composition of the 5s Ni-orbital (E203=2.64 eV), the 11 px Ni-orbital (E193=1.06 eV) and the 6s Ni-orbital (E186=0.24 eV). (last) The HOMO is more spread on the cone and less away from it (E171=−5.55 eV).
Fig. 13. Dependence of the neutral Ni-doped nanocone MO energies on the strength of the electric field . The HOMO and the LUMO are shown by thick lines. The extended Ni-orbitals drop fast with . (inset) The HOMO at electric field of 15.4 V/nm that contains mostly Ni s orbital contribution.
Fig. 14. (Up left to right) Extraction of the electron density from the Ni-doped nanocone in the static electric field , 13.4 and 15.4 V/nm. (down) Extraction of the electron density from the Ni-doped (4,4) CNT in (left) and 20.6 V/nm (right). In all the structures, the electrostatic potential in the interval =(−5.4,+5.4) V is mapped on the surface of the constant electron density 0.0004 e/Bohr3.
Fig. 15. Dependence of the Ni atom Mulliken charge and the average Ni–N distance on the strength of the electric field for the Ni-doped nanocone. (inset) Dependence of the electric field needed for the filling of the extended states on the length of the nanocone (number of CNT’s elementary cells).
Fig. 16. Binding of the Cl− and Na+ ions to the cone. (left) Extraction of Cl− at V/nm and Na+ in V/nm. (right) The Ni–Cl and Ni–Na distances as a function of the field strength.
Fig. 18. (Top) A nanomechanical system formed by two extended capsules coordinated to a bpm ligand. The system can be displaced from its equilibrium bent conformation by bending and twisting. (bottom) The relative energy of the nanomechanical system for twisting and bending. The E and E0 are the total energies of the bent (twisted) and optimized structures, respectively.
(top) Nitrogen substitution energies , Fe binding energies (eV), and Fe–N bond lengths (Å) for the graphene and CNT clusters. The cluster curvature increases from left to right. Pyridine and pyrrole-like coordination sites are denoted as py and pr, respectively
Selected optimized N–N distances for anionic ligands (N–N(L)) and for cones (N–N(co)), Ni–N distances (Å), N–Ni–N angles (°), Mulliken Ni atom charges (q(Ni), e), Ni–N(ligand) total binding energies (Eb(co), eV), and HOMO–LUMO gaps (H–L, eV) for the cones in their ground electronic states (GS). The GS s, d, t, qr and qn are singlet, doublet, triplet, quartet and quintet, respectively. The HOMO–LUMO gaps of open-shell systems are determined relative to LUMOs of the same spin type as the HOMOs. The N–N distances for 1 –6 are for N atoms trans to the Ni atom. The N–N distances for 8 –11 are for N atoms that belong to the same tridentate ligand. The N–N distances for ligands are defined as in the corresponding cones. The N–Ni–N angles for 8 –11 involve N atoms that belong to the same ligand.
Selected optimized metal–N (M–N) and metal-metal (M–M) distances (Å), binding energy (Eb(ca), eV), HOMO–LUMO gaps (H–L, eV) and Mulliken metal atom charges (q, e) Mulliken metal atom spin densities (Spin(M), e) for the capsules in their ground electronic states (GS). The GS types s, t, qn, h, n and ed are singlet, triplet, quintet, heptet, nonet and endecet, respectively. The HOMO–LUMO gaps of open-shell systems are determined relative to LUMOs of the same spin type as the HOMOs. For capsules 10, 10’ and 12 –15, only the diametral M–M distances are listed.