We investigate the interactions of one and two H 2 O molecules at three-coordinated sites of (MgO) 4 and (MgO) 8 clusters. Energies of adsorption are calculated and compared to similar work involving periodic and embedded cluster systems. The cluster model results predict that H 2 O chemidissociates without a barrier directly onto adjacent three-coordinated sites. The chemidissociation energy per water molecule is observed to be greater upon the dissociation of two neighboring water molecules compared to the dissociation of a single water molecule: a positive cooperativity effect. Correlation energy treated at the MP2 level of theory reduces the predicted chemidissociation energy by approximately 10 kcal/mol, and it is found that the MP2 single-point energies evaluated at the RHF-optimized geometries, MP2//RHF, differ from the energy of the MP2-optimized structures by approximately 1 kcal/mol.

An ab initio study is presented concerning the chemisorption of hydrogen on a model of the (100) surface of MgO and Li-doped MgO. The local surface environment was modeled employing cubic and tetragonal clusters composed of 8 and 12 atoms, respectively. The lattice constant for the clusters was fixed at the experimentally determined value for bulk MgO and the geometry of the adsorbate was optimized at the unrestricted Hartree–Fock (UHF) level of theory. Correlation energy was treated at the second-order unrestricted Mo/ller–Plesset (UMP2) level at the UHF optimized geometry. It was found that H2 undergoes heterolytic dissociation at neighboring three-coordinated Mg and O sites (denoted Mg3c and O3c) in MgO with activation energies of 4.2 and 2.4 kcal/mol at the UHF and UMP2 levels, respectively. Li-doped MgO did not support heterolytic dissociation at neighboring Mg and O sites. Instead H2 was found to dissociate homolytically without barrier at two O3c sites and to undergo hydrogen atom abstraction at O3c and O4c sites. At the UHF/UMP2 level, it was found that at O3c sites, abstraction occurs with a 17.9/3.0 kcal/mol barrier, and at O4c sites, abstraction occurs with a 15.7/0.6 kcal/mol activation energy.

No abstract

We discuss issues in developing scalable parallel algorithms and focus on the distribution, as opposed to the replication, of key data structures. Replication of large data structures limits the maximum calculation size by imposing a low ratio of processors to memory. Only applications which distribute both data and computation across processors are truly scalable. The use of shared data structures that may be independently accessed by each process even in a distributed memory environment greatly simplifies development and provides a significant performance enhancement. We describe tools we have developed to support this programming paradigm. These tools are used to develop a highly efficient and scalable algorithm to perform self‐consistent field calculations on molecular systems. A simple and classical strip‐mining algorithm suffices to achieve an efficient and scalable Fock matrix construction in which all matrices are fully distributed. By strip mining over atoms, we also exploit all available sparsity and pave the way to adopting more sophisticated methods for summation of the Coulomb and exchange interactions. © 1996 by John Wiley & Sons, Inc.