Luận án: Phân tích tính toán giao diện Platin-Nước-Chân không - Timm Lankau, Hamburg 2000

A Computational Analysis of the Platinum-Water-Vacuum Interface Dissertation zur Erlangung des Doktorgrades des Fachbereichs Chemie der Universit at Hamburg vorgelegt von Timm Lankau aus Hamburg Hamburg 2000 Gutachter Prof. Nagorny, Hamburg Dr. Drager, Hamburg Tag der letzten mundlichen Prufung: 14. Dezember 2000 To my mother Die Quanten sind doch eine ho nungslose Schweinerei.

Max Born to Albert Einstein Abstract The simple bilayer model of the platinum-water-vacuum interface can explain most experimental results, but fails on details. Our work suggests that cooperative e ects are very important for the ne-structure of the interface. The platinum-water (chapter 6) and the hydrogen bond (chapter 3) have been found to be very much alike. Intermolecular electron transfer as observed in the water trimer has a strong in uence of the interface structure and can move the water molecules out of their ideal positions (chapter 8).

Our new water-water interaction potential (chapter 4) has been used to explore the potential energy surfaces of the water trimer (chapter 4) and hexamer (chapter 8). The results on the hexamer suggest that a seamless transition between the platinum surface and the ice crystal is not possible. Structures similar to the QLL (Quasi Liquid Layer) have been observed, which can explain the low vertical dipole moment of surface water, while cooperative forces can be used to explain the second desorption peak (165 K) in TDS (Thermal Desorption Spectroscopy) experiments. The weak bond between water and platinum is controlled by two opposing forces: the Coulomb repulsion between the 6s electrons of the platinum cluster and the oxygen atom and the bonding interaction between a platinum 5d orbital and the free electron pair of the water molecule.

A high 6s population, which repells the water molecule, creates at the same time a strong bond among the platinum atoms. It is therefore impossible to create a surface model with strong intermetallic bonds and a strong platinum-water bond at the same time. The quality of the surface model depends strongly on the 6s population and so on the electronic state of the metal cluster. The analysis of the platinum-water bond would have been impossible without a modi cation of the Huckel theory (chapter 5), which was used to select suitable can- didates as surface models and to understand the electronic structure of the platinum cluster and electron movements during the formation of the platinum-water bond.

The intermetallic bond in the platinum cluster is dominated by the 6s orbitals and the 5d orbitals have to be considered in full detail only at active surface atoms. This assumption was nally proofed by the development of a new 1 valence electron ECP (E ective Core Potential) for bulk and passive surface atoms (chapter 7), which can be used to reduce the computational costs for the analysis of larger platinum clusters. Zusammenfassung Ein einfaches Doppelschicht-Modell fur das Grenzsystem Platin-Wasser kann die meisten experi- mentellen Ergebnisse erklaren, versagt aber bei der Interpretation von Details. In dieser Arbeit werden wir nachweisen, da cooperative E ekte sehr wichtig fur die Feinstruktur der Grenzschicht sind.

Die Platin-Wasser Bindung und die Wassersto bruckenbindung sind einander sehr ahnlich (Kapitel 6) und intermolekularer Ladungstransfer, wie er auch im Wassertrimer beobachtet wird, kann einzelne Wassermolekule aus ihrer idealen Position bringen (Kapitel 8). Unser neues Wasser-Wasser-Wechselwirkungspotential (Kapitel 4) wurde benutzt, um die Ener- giehyper ache des Wassertrimers (Kapitel 4) und des Wasserhexamers (Kapitel 8) zu untersuchen. Die Ergebnisse fur das Hexamer lassen vermuten, das ein nahtloser U bergang zwischen der Platin- ober ache und einem Eiskristall wie bisher angenommen nicht moglich ist. Strukturen, ahnlich einer zweidimensionalen Schicht ussigen Wassers auf Eis (engl.

QLL, Quasi Liquid Layer), wur- den an der Grenze zwischen Metall und Eis beobachtet. Solch ein Strukturmodell erklart den kleinen Anteil des Dipolmoments eines Wassermolekuls in der Grenzschicht senkrecht zur Metallo- ber ache, wahrend die oben erwahnten cooperativen Krafte erstmals eine Deutung fur den zweiten Desorptionspeak (165 K) in TDS Experimenten (Thermal Desorption Spectroscopy) bieten. Die schwache Bindung zwischen Platin und Wasser kann mit zwei einander widersprechenden Kraften erklart werden: Die Coulomb-Abstoung zwischen den 6s Elektronen des Platins und dem negativ geladenem Sauersto im Wassermolekul ist die erste Kraft und die zweite bindende folgt aus dem U berlappen eines freien Elektronenpaars des Wassermolekuls mit einem Platin 5d Orbital. Eine hohe 6s Besetzungsdichte, die das Wassermolekul abstot, ezeugt aber gleichzeitig eine starke Platin-Platin Bindung.

Es gibt entweder eine starke Platin-Platin Bindung im Metallcluster oder eine starke Metall-Wasser Bindung, aber nie beides gleichzeitig. Die 6s Elektronendichte erwies sich als Schlussel zu einem realitatsnahen Ober achenmodell. Eine Modi kation der Huckel-Theorie (Kapitel 5) half uns, die elektronische Struktur des Platin- clusters und die Bewegung der Elektronen wahrend der Wasseradsorption zu verstehen. So war es uns moglich, gezielt nach geeigneten Kandidaten fur die Ober achenmodellen zu suchen.

Die Metall-Metall Bindung im Platincluster wird von den 6s Orbitalen dominiert, wahrend die 5d Orbitale nur wichtig sind fur die Bindung des Wassermolekuls an ein aktives Ober achena- tom. Diese Annahme fuhrte zu der Entwicklung eines funktionierenden 1 Valenzelektronen ECP (E ective Core Potential), da in Zukunft die Untersuchung groer Ober achenmodelle ermogli- chen wird.1 Properties of Water .2 Properties of Platinum .1 Hartree-Fock Calculations .2 Mller-Plesset Perturbation Theory .3 Con guration Interaction and Multicon guration SCF Theory .4 Basis Sets and Basis Set Superposition Error .5 The Morokuma Energy Decomposition Scheme .1 Non Relativistic Core Potentials .2 Relativistic Quantum Mechanics and Core Potentials .7 Interactions between di erent electronic states .1 Intersection of Potential Energy Curves .2 Photoexcitation and Intersystem Crossings .8 Dipole-Dipole Interaction and Polarisation. 40 3 Quantum Chemistry of small Water Clusters 45 3.1 The Water Molecule .1 Single Point Calculations .2 Potential Curve with a Flexible Geometry .3 Potential Curve with a Fixed Geometry .4 Calculation of the BSSE corrected geometry of the water dimer .5 Summary of the Quantum Mechanical Calculations. 81 4 Classical Water-Water Interaction Potentials 85 4.1 Classi cation of the Di erent Potentials .2 Comparison with the Quantum Mechanical Potential Curves .1 An improved model .2 Application of potential N on water trimers .4 Summary of the Calculation with a Classical Potential.

114 5 Huckel calculations for the Analysis of Ptn 117 5.1 Theory of the Huckel-approximation for the platinum 6s electrons .2 Huckel calculations for Pt3 .3 Analysis of the Pt5 pyramid .1 Huckel calcualtions fot Pt5 .2 The interface between di erent ECPs .3 5d-6s interaction in Pt5 .4 The rotational barrier in Pt5 H2O .5 Summary on Pt5 .4 Huckel Calculations for Pt9 Cluster .5 Two Slabs Pt17 Cluster .6 The Second Next Neighbour .7 Quantitative Analysis of the HMO Calculations .1 Results for the Platinum Dimer .2 Results for Platinum Pentamer (Pyramid) .8 Summary and Conclusions from the 6s Huckel calculations. 142 6 Platinum atom calculations involving 18 Valence Electrons 145 6.1 One platinum atom .1 The electronic states of platinum .2 Platinum and a single water molecule .3 The In uence of the Pseudo Potential .4 E ect of the BSSE on the platinum-water interaction .5 The Platinum-Hydrogen interaction .6 Summary of the results for single platinum atom .1 The electronic structure of the platinum dimer .2 Platinum Dimer and Water .1 In uence of the geometry .2 The water metal interaction in Pt2 H2O .3 In uence of the electronic state of the platinum dimer .4 Movement of water on the surface .3 Summary of the results for the platinum dimer .3 The platinum trimer .1 Electronic structure of the equilateral Pt3-cluster .2 The interaction of water with Pt3-cluster .1 Water bound to the hollow site on the cluster .2 Water bound on top .3 Pt3 surface model for the Pt(100) surface .3 Summary of the results for Pt3 .4 The Pt5-pyramid .1 Electronic structure of the Pt5 -pyramid .2 The interaction of the Pt5-pyramid with water .3 Summary of the results for the platinum pyramid .5 The Pt9-cluster .6 The electronic structure of the metal cluster as a function of its size .7 EHT calculations on the Ptn H2O system .8 Summary of the calculations with a 18 valence electrons platinum atom. 199 7 Calculations with 1-Valence Electron per Platinum 203 7.1 Numerical properties for the new 1-electron ECP .1 Numerical Results for the 6s orbital .3 Simpli cation of the Radial 6s Electron Density Function .4 The quest for the new 6s Wavefunction .2 Numerical Results for the Pt 6p Orbital .3 Numerical Results for the Pt 6d orbital .4 Summary and compilation of all 1 electron properties .2 Principle questions about a 1 electron ECP .1 Is it possible to use the method from Hay and Wadt? .2 Which conditions has a gaussian type wavefunction to ful ll? .3 Is a positive eigenvalue for the 6d orbital physically reasonable? .4 How does Gaussian 94 calculate the energy of the orbitals? .5 Under which conditions is global minimum possible? .6 How does Gaussian calculate the ECP? .7 What happens if the ECP vanishes as r becomes in nite? .8 Which form has an ECP with a local energy minimum for d-electrons? .9 Is it possible to create a local minimum with two or more functions? .3 How describe the 6d electron? .1 What problems are connected with the ECP by Zurita et al.3 How does Ucore L controll the dimer's properties? .4 How strong is the in uence of Ucore L on the PtH bond? .5 How does UL change the electronic structure of Pt5? .6 What happens if two di erent ECPs interact with each other? .7 Is the LanL1MB cluster a suitable surface model? .4 Conlusions from the analysis of the 1 valence electron ECP. 247 iii 8 Water Clusters on the Platinum Surface 251 8.3 Results and Discussion .2 (H2O)6 on a Virtual Metal Surface .2 The free water hexamer .3 The constrained hexamer .4 Variation of the surface lattice constant.

284 9 Final Conclusions and Further Proceedings 285 10 Bibliography 295 10.2 Programs used for this work .2 Basis set and ECPS used with in this work .3 Selected Water Monomer Data .4 Results from Single Point Calculations for the Water Dimer .5 In uence of the BSSE on the Monomer's Geometry .6 Changing W3 and Bending the Hydrogen Bond .7 Energies of the Morokuma Energy Decompostion for the Water Dimer .8 Di erent water-water interaction potentials from the literature .9 Geometry optimisation of water clusters within potential N .2 Conversion of rotational into cartesian coordinates .10 The C++ Gaussian 94 interface .11 Huckel calculations for Pt9 .12 Huckel calculations for Pt17 .13 Construction of gure 7. 357 iv List of Figures 1.1 Structure of ice 1h.2 Ideal bilayer structure.3 Scienti c environment of this work.1 In uence of the BSSE on the interaction energy.2 Formation of a trimer.3 Interaction and mixing of the monomer's orbitals.4 Dipole-dipole interaction.1 Sketch of the molecular orbitals of water.2 Geometry of the water dimer.3 Variables of the dimer.4 Composition of the water dimer's MOs.5 In uence of the electrostatic interactions on the orbital energies.6 MO 5a' of the water dimer.7 MO 6a' of the water dimer.8 MO 7a' of the water dimer.9 MO 8a' of the water dimer.10 Electron density in the symmetry plane.11 Formation of the water dimer.12 Interaction energy and BSSE.13 Minima in gure 3.14 Energy composition, not BSSE corrected.15 Changing of rOH with dOO.16 Changing of charges during dimerisation.17 Correlation energy vs.18 Geometry of a possible ion pair.19 Energy during the proton transfer.20 Proton charge during proton transfer.21 Chargre transfer
q during proton transfer.22 Interaction energy with rigid geometries.23 Details from gure 3.24 Comparison of the charge transfer.25 Di erence between a exible and a rigid monomer geometry.26 Details from gure 3.29 Repulsion between water molecules.30 Oxygen-oxygen repulsion aOO.31 Hydrogen-Hydrogen repulsion aHH.32 Local minimum of H-H repulsion.33 Arti cial BSSE minimum of aHH.34 BSSE corrected minimum geometry.35 Geometry of trimer I.36 Geometry of trimer II.37 Geometry of trimer III.38 Geometry of trimer IV.39 Three possible conformers of trimer III with their MP2 energies in Hartree.40 3a' orbital of trimer I.