Using self-consistent-field molecular-orbital theory, we show that the interaction of hydrogen molecules with a Ni + ion is characteristically different from that with a neutral Ni atom. While hydrogen chemisorbs dissociatively on the neutral metal atom, it is bound to the cation in its molecular form. The atomic bonding is a consequence of the Pauli exclusion principle whereas the bonding of the molecular hydrogen results from an electrostatic interaction. We predict that a Ni + ion can bind at least six hydrogen molecules. PACS numbers: 3l.20.Ej, 35.20.Gs The interaction of hydrogen with metals and metal surfaces has been studied for many years. It is commonly understood that a H2 molecule breaks up into individual atoms at about 0.5 A from the metal surface and atomic chemisorption ensues [1]. No evidence exists, to our knowledge, of molecular chemisorption of hydrogen on the metal surface or inside the bulk metal. However, pairing of hydrogen mediated by a metal atom in rareearth systems [2] and molecular hydrogen in Si [3] have recently been observed.Little is known about the interaction of hydrogen with small metal particles consisting of a few atoms. The first evidence that hydrogen interaction with metal clusters may be fundamentally different from bulk came from recent experiments [4] on hydrogen reactivity and absorption. The reactivity was found to change by orders of magnitude by changing only a few atoms in the cluster. Equally interesting was the observation [5] that metal clusters might be able to absorb as many as eight hydrogen atoms per metal atom even though no corresponding bulk metal hydride exists. Not much is known regarding why clusters behave so differently towards hydrogen than bulk or whether hydrogen is bonded to the cluster in atomic or in molecular form.In this Letter we show that hydrogen interacts the same way with a neutral atom as it does with the atoms on surfaces and in the bulk. However, the interaction of hydrogen with a metal ion is fundamentally different. In the case of the neutral atom, an electron is transferred from the metal atom to the approaching H2 molecule and, in keeping with the Pauli exclusion principle, occupies the antibonding orbital of the molecule. This, in turn, breaks the molecular bond in favor of atomic bonding between the dissociated hydrogen atoms and the metal atom. However, when a H2 molecule approaches a transition-metal ion, it becomes energetically inefficient for the metal atom to donate an electron to H2 since the second ionization potential of the metal atom is rather high. Instead, the ion polarizes the H2 molecule and the bonding between the ion and the H2 molecule is governed by a dipole mechanism. We predict that a single Ni + ion can trap up to ten hydrogen molecules.Our results are based on first-principles calculations of

Articles you may be interested inComputation of interior eigenvalues in electronic structure calculations facilitated by density matrix purification Phosphorus hyperfine structure in the electronic spectrum of the HPCl free radical A fast ab initio model for the calculation of excited electronic states of atoms and molecules in a weakly polarizable environment. I. TheoryEvolution of the electronic structure and properties of neutral and charged aluminum clusters: A comprehensive analysis Ab initio molecular orbital theory based on both density functional formalism and quantum chemical methods has been used to calculate the equilibrium geometries, binding energies, ionization potentials, fragmentation patterns, and electronic structures of neutral and charged boron clusters containing up to six atoms. Calculations have also been performed on restricted geometries for B n X ͑nϭ1,5,12; XϭBe, B, C͒ and B 20 clusters to see if clusters can be designed so as to increase their stability. Energetics of doubly charged B n ϩϩ clusters have also been studied to find the critical size for Coulomb explosion. The results are compared with existing experimental and theoretical data.

No abstract

No abstract

Niu, Rao, and Jena Reply: Manninen and Nieminen [I]have commented as follows: (i) The interaction of a metal ion with a molecule can be understood by a polarization mechanism where the binding energy can be obtained from where a is the polarizability of the molecule, e is the static dielectric constant, and r is the distance between the ion and the molecule, (ii) In the asymptotic limit the average binding energy of the molecule is proportional to 1/A^ and at 7=0 an infinite number of molecules can be bound to the ion. (iii) The asymptotic binding is independent of the ion in question but the binding of the first molecule depends on the chemical identity of the ion.In our Letter [2] we had pointed out that the binding of a large number of H2 molecules with a metal cation is governed by polarization. In Fig. 2 of our Letter the calculated binding energies of up to six H2 molecules were plotted and then extrapolated. It was stated that "the extrapolation would indicate that no more than ten H2 molecules can be bound to one Ni"*" ion.'' This did not imply that in the asymptotic limit the maximum number of H2 that can be bound to the cation is 10. As a matter of fact, had we been able to calculate the binding energy of the tenth molecule it would have been extremely small and its accuracy would be questionable due to the limitations in the Gaussian basis sets as well as the approximations involved in the self-consistent field procedure. We had, therefore, clearly mentioned at the end of the first paragraph on p. 2279 that the maximum number of H2 molecules bound to the Ni"*" ion "is likely to increase if a more extensive basis set is used." We do, however, agree with Manninen and Nieminen that the limiting value of the number of molecules bound to Ni"^ at T^O would indeed be infinite. In reality, however, other factors such as temperature and gas density would severely restrict this number.Regarding the ability of the simple polarization formula in Eq. (1) to quantitatively predict the binding energy of the first molecule, let it suffice to say that in the case of the Ni"*" ion it does a rather poor job. The binding ener-gy is about half of what one calculates using an extensive basis set. The accuracy of Eq. (1) would depend upon a single factor-the distance of the molecule from the ion vis-a-vis the extent of the charge density distribution of the ion. If the molecule is far enough so that the ion appears as a point charge, Eq. (l) would be able to give the binding energy with quantitative accuracy. This is indeed the case when we examine the binding of a Li"^ ion with one H2 molecule. The Li"*" ion has electrons only in the \s^ core and, therefore, is quite compact. The optimized distance of the H2 molecule from the Li"^ ion is 2.01 A. Using this value in Eq. (1) the predicted binding energy is 0.3 eV. Using a very extensive basis set and the method described in our Letter we have calculated this binding energy to be 0.29 eV. These are in excellent agreement. Of course, to calculate the binding energy using Eq. (1...