Titles and Abstracts






Desorption processes in 1D: A mesoscopic approach (2 talks)

by Enrique Abad and Fotis Vikas

Universite Libre de Bruxelles, Belgium


Lattice dynamics of the desorption reaction A + A -> A + S: Role of diffusion (E. Abad)

A Boolean cellular automaton model is developed to describe the 1D lattice dynamics of the irreversible desorption process A + A -> A + S. The underlying classical rate equation predicts a $t^{-1}$ decay to a zero-density state. We study both the no-diffusion limit (immobile reactants) and the diffusion-controlled case on an infinite lattice. In the absence of diffusion, deviations from mean field behavior are observed both in the dynamics and in the steady states. The long time behavior is characterized by relaxation toward a nonvanishing invariant state which depends on the initial conditions. If particles are allowed to diffuse, the density shows an anomalous $t^{-1/2}$ approach to the mean field steady state, in agreement with the off-lattice solution. The early time behavior, however, predicts a slower decay of the density, due to the finite propagation velocity of a local density perturbation, as opposed to the infinite propagation velocity characteristic of diffusion.

A three-state model for reaction controlled cooperative desorption on a one-dimensional lattice (F. Vikas)

We present a master equation approach for the analysis of the dynamics of immobile reactants on a one-dimensional lattice, involving two reactive species and vacant sites. A common feature of all the studied schemes is the strong dependence of the final coverage on the initial conditions, associated with the lack of ergodicity of the invariant state. Our approach leads to full agreement with Monte-Carlo simulations, both asymptotically and dynamically.






A mean-field kinetic lattice gas model of electrodeposition

by Marc-Olivier Bernard and Jean-Francois Gouyet

Laboratoire PMC, Ecole Polytechnique, 91128 Palaiseau cedex, France

Abstract
The arborescent growth of an electrode due to the reduction of an electroactive cation (electrodeposition) was experimentally and theoretically studied in a mesoscopic frame in the early 90's (see [1] as an example).
Our aim is to generalize existing models of Mean Field Kinetic Equations (MFKE, Martin [2] Gouyet [3]), previously applied to various physical systems, such as intercalation [4], dentritic growth and spinodal decomposition in alloys [5], in order to address electrochemical problems.
Electrochemical MFKEs involve solving the Poisson equation to determine the electrostatic potential at any point of the system and to build a plausible microscopic model for reduction at the cathode and dissolution at the anode.
We thus hope to obtain a better understanding of dendritic growth in electrodeposition problems; in particular, the effects of crystal anisotropy, rate of electronic transfer, the mobility of the ions and nearest neighbour interactions.

References:
[1] V. Fleury, Croissance par voie electrochimique d'agregats metalliques ramifies, te?se de l'Ecole polytechnique, 13 decembre 1991.
[2] G. Martin, Phys. Rev. B, 41 2279 (1990).
[3] J.F. Gouyet, Europhys. Lett., 21 [3], pp. 335-341 (1993).
[4] R. Nassif, Y. Boughaleb, A. Hekkouri, J.F. Gouyet and M. Kolb, Eur. Phys. J. B, 1 453-464 (1998).
[5] M. Plapp, Phys. Rev. Lett. 78,26 p. 4970 (1997).





Simple reactions involving structured molecular systems: influence of the internal dynamics

by Alexander Blumen and Gleb Oshanin,

University of Freiburg, Germany

Abstract
As is well known from polymer physics, even in the simplest Rouse-model (involving point-like molecules connected to each other through springs) the dynamics of the whole object may be quite complex. Thus even a single monomer in a linear chain experiences first a sublinear motion, proportional to the square-root of time, followed later by regular diffusion. Such features effect drastically even the simplest decays, such as found for trapping or for target annihilation [1]. In this talk we will focus on such simple A+B->B reactions, when the particles under investigation have an internal structure. Examples will involve, among others, dendrimers and star-like molecules [2,3]. As we will show, this leads to quite complex dynamical patterns, and not in all cases regular scaling laws are obeyed. Furthermore, the introduction of hydrodynamic interactions (which must be accounted for in dilute solutions) adds additional features to the decay laws.

References:
[1] S. Nechaev, G. Oshanin and A. Blumen, J.Stat.Phys.98, 281 (2000).
[2] P. Biswas, R. Kant and A. Blumen, Macromol.Theory Simul.9, 56 (2000).
[3] R. Kant, P. Biswas and A. Blumen, Macromol.Theory Simul., submitted.





Chemical Wave propagation in lattice-gas models for surface reactions: Analysis of the hydrodynamic limit

by Jim Evans

Ames Laboratory and Dept of Mathematics, Iowa State University, USA

Abstract
We develop exact (non-mean-field) reaction-diffusion equations (RDE's) to describe chemical or trigger wave propagation in a simple model for CO-oxidation, and front propagation in a simple model for the reactive removal of unstable mixed NO+CO adlayers. Particular emphasis is placed on the appropriate description of chemical diffusion in mixed adlayers, a key ingredient in these RDE's. Results from the exact RDE's match direct KMC simulations.

References:
[1] M. Tammaro and J.W. Evans, J. Chem. Phys. 108, 762 (1998).
[2] M. Tammaro and J.W. Evans, J. Chem. Phys. 108, 7795 (1998).
[3] M. Tammaro and J.W. Evans, in "Computer Simulation Studies in CMP" D.P. Landau & H.-B. Schuettler, Eds. (Springer, Berlin, 1999).





First-principle electronic structure calculations as a tool for modelling surface reactions

by Fabio Finocchi

Laboratoire de Physique des Solides, Batiment 510, Universite Paris-Sud, Orsay, F-91405, France

Abstract
The Density Functional Theory (DFT) has become a "standard" approach to evaluate, from first principles, the interaction energy between atoms in different physical states (molecules, clusters, surfaces, and solids) without making any "a-priori" approximation on the nature of the chemical bond. We review the theoretical basis of the DFT, with special emphasis on the currently used approximations of the exchange-correlation energy functional. We discuss how energy surfaces can be computed as a function of the ionic coordinates, and stress the limitations and the accuracy of the method. Presently, systems of few hundreds atoms can be studied, and their dynamics followed for some tens of picoseconds. Therefore, energy surfaces computed from first principles can be used as input for other numerical approaches - such as kinetic Monte Carlo - able to cope with larger length scales and longer simulation times. Moreover, ab-initio molecular dynamics provides an unique way to look into atomic scale complex phenomena, such as bond breaking and reforming, often associated to chemical reactions at surfaces. As an illustration, we describe the interaction of water molecules with MgO(100) surfaces, either perfectly planar or containing defects, such as steps, edges, and vacancies.





Fluctuations in catalytic CO oxidation

by Ronald Imbihl

Institut fuer Physikalische Chemie und Elektrochemie, Universitaet Hannover, Germany

Abstract





A Dynamic Monte Carlo Method for Systems of Any Size?

by Tonek Jansen

Schuit Institute of Catalysis, Eindhoven University of Technology, P.O. Box 513, 5600 MB Eindhoven, The Netherlands

Abstract
I will discuss the idea of a Monte Carlo method in which a lattice-gas model of a catalyst's surface is used with each grid point corresponding to a block of sites. The number of adsorbates of each type is known for each block, but not the adlayer structure within a block. By choosing large blocks large systems can be simulated. As diffusion within a block does not change the number of adsorbates in a block, the amount of computer time spent on simulating diffusion can be substantially reduced. The method can be speeded up even more by using a reduced number of effective adsorbates instead of the number of real adsorbates. One problem with the method is how to incorporate the influence of structure of adlayer within the blocks on the kinetics. Another is that with the effective adsorbates fluctuations may not be taken into account correctly.





Mesoscale Dynamics of Reaction-Diffusion Systems

by Raymond Kapral

Chemical Physics Theory Group, Department of Chemistry, University of Toronto, Toronto, Ontario M5S 3H6, Canada

Abstract
A mesoscale Markov chain description of reaction-diffusion media will be used to explore the effects of fluctuations and correlations on chemical oscillations and chemical waves. The range of validity of macroscopic mass-action or reaction-diffusion models and the consequences of their breakdown will be discussed. Simulations and kinetic theory analyses on a number ofchemical models will be used to illustrate the phenomena.





Adsorbate island formation and dynamics at nanoscale resolution: Experiments and Simulations

by Norbert Kruse*, Christian Voss*, Valentin Medvedev*+, Christopher Bodenstein#, David Hanon#, and Jean-Pierre Boon# (2 talks)

*Chemical Physics at Surfaces and Heterogeneous Catalysis, CP 243, Universit\'e Libre de Bruxelles, 1050--Bruxelles, Belgium
#Center for Nonlinear Phenomena and Complex Systems, CP 231, Universit\'{e} Libre de Bruxelles, 1050--Bruxelles, Belgium
+Department of Chemical Engineering, University of Washington, Seattle, WA 98195-1750, USA

Abstract
Surface catalytic processes produce, under certain conditions, small clusters of adsorbed atoms or groups, called {\em islands} which, after they have been formed, move as individual entities. Here we consider the catalytic reduction of NO with hydrogen on platinum. (i) Using video field ion microscopy, we observe the dynamic motion of small hydroxyl islands on the Pt(001) plane; despite changes in their morphology, the islands dimensions are confined to values corresponding to 10 to 30 Pt atoms suggesting cooperative effects to be in operation. (ii) We construct an automaton (or lattice Monte-Carlo) model on the basis of a set of elementary processes governing the microscopic dynamics. The agreement between the simulation results and the experimental observations suggests a possible mechanism for the formation and dynamics of hydroxyl islands.

Reference:
[1] N. Kruse, C. Voss, V. Medvedev, C. Bodenstein, D. Hanon, and J.-P. Boon, J. Stat. Phys. (to appear June 2000).





New insights for the A+A -> 0 problem in one dimension

by Katja Lindenberg

University of California - San Diego, USA

Abstract
Recent work on the dynamical aspects of the A+A-> 0 reaction (in particular the effect of the distribution of sources) is presented. Our approach is mesoscopic, but the inspiration and results come from a microscopic simulation of a kink-antikink problem.





Ordering, percolation and diffusion in a lattice gas model for co-oxidation with superlattice ordering of oxygen

by Da-Jiang Liu

Ames Laboratory, Iowa State University, USA

Abstract
An order-disorder transition is found in a nonequilibrium surface reaction model for CO-oxidation with superlattice ordering of adsorbed oxygen. This ordering transition strongly influences the percolation properties of superlattice domains of oxygen, which in turn are shown to control the chemical diffusion of coadsorbed CO. The latter constitutes a new type of problem involving transport in disordered media, which is important as CO-diffusion controls spatial pattern formation in the reaction system.

Reference:
[1] D.-J. Liu and J.W. Evans, Phys. Rev. Lett. 84, 955 (2000).





Landau-type mean-field theory of the Langmuir-Hinshelwood reactions between interacting particles on heterogeneous catalytic substrates

by Gleb Oshanin

Laboratoire de Physique Theorique des Liquides, Universite Paris VI, T.16, 4 place Jussieu, 75252 Paris, France

Abstract





Reactive dynamics on low dimensional supports: A Lattice Limit Cycle model

by Astero Provata

National Research Center "Demokritos", Athens, Greece

Abstract
The Lattice Limit Cycle (LLC) model is proposed as a lattice compatible reactive mechanism which in the mean field limit gives rise to sustained limit cycle oscillations. The LLC model concerns an open process and involves reaction, adsorption and desorption mechanisms. The same model, realised by Monte-Carlo simulations on a 2-d lattice gives rise to sustained oscillations of the local concentrations on the lattice. These local oscillations are due to a) local fluctuations and b) competition between domains of different species. The oscillator characteristics depend on the system parameters and on the lattice geometry. Such surface concentration oscillations are observed in catalytic process eg. the CO oxidation on Pt surface.

References:
[1] A. Provata, G. Nicolis and F. Baras, J. Chem. Phys. 110 (1999) 8361.
[2] A. Shabunin, F. Baras and A. Provata (submitted).





Reaction fronts in hydrogen oxidation on Pt(111); modelling and quantitative description of STM experiments

by Christian Sachs, M. Hildebrandt, S. Völkening, J. Wintterlin and G. Ertl

Fritz-Haber-Institut der Max-Planck-Gesellschaft, Abt. PC Faradayweg 4-6 14195 Berlin, Germany

Abstract
Recently we suggested an autocatalytic mechanism for the oxidation of hydrogen on Pt(111) below the desorption temperature of H2O [1,2]:
O + H -> OH (I)
OH + H -> H2O (II)
2 H2O + O -> 3 OH + H (III)
This autocatalytic meachanism leads to the formation of reaction fronts on a 100 to 1000 Å scale that were observed by STM over a range of temperatures. Because water is mobile under these conditions the combined reactions II and III represent a reaction diffusion system. The front velocity, obtained from an analytical treatment and also from numeric simulations, depends on the two reaction rates and the diffusion constant of water. LEED and STM experiments were performed in order to measure these parameters. The front velocities evaluated on the basis of these data agree with those determined by the STM experiments. Additionally it is possible to derive their temperature dependence.

References:
[1] S. Völkening, K. Bedürftig, K. Jacobi, J. Wintterlin, and G. Ertl, Phys. Rev. Lett. 83, 2672 (1999).
[2] K. Bedürftig, S. Völkening, Y. Wang, J. Wintterlin, and G. Ertl, J. Chem. Phys. 111, 11147 (1999).





Exact solution for front propagation in a one-dimensional epidemic model

by C. Warren, E. Somfai*, and Leonard Sander

Dept. of Physics, The University of Michigan, Ann Arbor, MI 48109-1120, USA

Abstract
We study the discrete model of the irreversible autocatalytic reaction A + B -> 2A in one dimension, which is a version of the model of J. Mai, I.M. Sokolov and A. Blumen, Phys.Rev.Lett. vol. 77, 4462 (1996), which can also be regarded as the simplest way to look at the spread of an epidemic. We are able to solve the model exactly in one dimension for low concentrations. We find that in the low-concentration limit the average velocity of propagation approaches c/2 where c is the concentration. The front propogation is entirely dominated by fluctuations in the density: the front spends most of its time pinned behind gaps in the density. We have numerical results to indicate that this behavior is preserved in two dimensions, and thus may give some guidance for the behavior of real surface reactions, and even epidemics in populations that are not well mixed.

Supported by DOE grant DEFG-02-95ER-45546.

*Current address: Instituut Lorentz, Leiden, NL-2333





Chemoconvection: A fluid instability driven by a surface chemical reaction

by Francesc Sagues

Departament de Quimica Fisica, Universitat de Barcelona, Spain

Abstract
We will report a systematic experimental study of pattern formation in a thin fluid layer driven by a simple surface chemical reaction. The hydrodyn amic instability arises due to the increase of fluid density in the sub- surface layer caused by the oxidation of glucose to gluconic acid with the methylene blue as a catalyst. The role of different parameters (chemical con centrations, temperature, layer thichness, etc) are analyzed. In addition a theoretical model is developed based on the coupling of a simplified chemical scheme with the Navier-Stokes equation. A linear stability analysis is performed whose results are compared with the experimental observations.





Stochastic simulations of excitable dynamics with applications to heterogeneous catalysis

by Lutz Schimansky-Geier

Institute of Physics, Humboldt-University at Berlin Invalidenstr. 110, D-10115 Berlin, FR Germany

Abstract
It is well known that in systems far from equilibrium an increasing noise may exhibit more ordered state. Typical examples of this 'unusual` behavior are stochastic resonance, noise induced directed motion and noise induced phase transitions.
We apply noise to excitable systems which model a wide class of dynamical behavior in biophysics and chemistry. First we exemplify the ordering role of noise on behalf of a time-dependent excitable dynamics. We calculate the output spectrum of the dynamics and show that an increase of the noise level generates oscillating outputs. If additionally a periodic force drives the excitable dynamics stochastic resonance is supported by this oscillations.
In the second part distributed coupled excitable units under noise are investigated. First we show that the presence of noise sustains moving structures of excitations in the medium. Beginning with a certain noise level the system of coupled units exhibits synchronized collective oscillations. Applications to heterogeneous catalysis is discussed and different simulation algorithms are presented.





Front propagation in one-dimensional autocatalytic systems

by I. M. Sokolov, J. Mai and A. Blumen

University of Freiburg, Germany

Abstract
We consider the front propagation in an autocatalytic reaction between particles which perform random walks on a lattice and react on contact according to the A+B -> 2A reaction scheme. The classical, continuous-medium description of this reaction leads to a Fisher equation whose front solutions can in principle propagate with any velocity $v>v_{c}=2% \sqrt{kDc}$, with $k$ being the local reaction rate, $D$ the diffusion coefficient and $c$ the particles' concentrations. The marginal stability principle requires that for rather general initial conditions the minimal velocity is the one really attained.
The discrete structure of real systems leads to strong departures from this prediction, since it introduces new length-scales into the problem (e.g. the interparticle distance $\lambda =c^{-1/d}$), which determine the local, small-scale structure of the system. In a 1d system (where the classical reaction scheme is absolutely inappropriate) this leads to a stable propagation of the front with a velocity which is fully determined by local ordering phenomena. This situation is modelled numerically under different conditions, e.g. for equal and different diffusion coefficients of reacting species and also taking the finite life-time of $B$-particles into account.
The discussion of the front structure assumes the introduction of a comoving coordinate frame. This can either be taken to move with an average velocity, or can be defined in each realization separately. Since different definitions emphasise different aspects of the local ordering, the average front forms obtained for different coordinate frames differ.
The results of our numerical simulations for the front velocity and of the front forms are discussed within a theoretical framework based on a Smoluchowski-like approximation and on the biased random-walk approach.

References
[1] J. Mai, I.M. Sokolov and A. Blumen, Phys.Rev.Lett. vol. 77, 4462 (1996).
[2] J. Mai, I.M. Sokolov, V.N. Kuzovkov and A. Blumen, Phys. Rev. E, vol. 56, 4130 (1997).
[3] V.N. Kuzovkov, J. Mai, I.M.Sokolov, and A. Blumen, Phys. Rev. E, vol. 59, 2561 (1999).





Molecular Dynamics simulations of chemical reactions in condensed fluids

by Soeren Toxvaerd

University of Copenhagen, Denmark

Abstract
Simulations of three different reaction schemes for the isomerization kinetics in a condensed fluid mixture between two species which only differ marginally in pair energies show that the mixture is bistable and that one of the species dominates at late reaction times. The kinetics are without random elements and are performed on the basis of the steepest descent method in the potential energy. The kinetics establish a domain decomposition with critical fluctuations which ensure the symmetry break. The model(s) offers a possible explanation of origin of biomolecular chirality.





Simulations of the reaction kinetics on nanometer supported catalyst particles

by P. Thormahlen*, H. Persson*, B. Kassemo* and V. P. Zhdanov*#

*Competence Center for Catalysis and Department of Applied Physics, Chalmers University of Technology, 41296 Goteborg, Sweden
# Boreskov Institute of Catalysis, Russian Academy of Sciences, Novosibirsk 630090, Russia

Abstract
Real catalysts frequently consist of nm-sized (1-100 nm) metal particles deposited on the internal surface of a more or less inactive porous support. The complexity of such systems in combination with relatively high (atmospheric) reactant pressures, typical for practical catalysis, hinders the application of surface science methods to their full potential and makes it difficult to interpret the measured reaction kinetics on the basis of kinetic data for single-crystal samples. To bridge these pressure and structure gaps and to form a conceptual basis for the understanding of reactions occurring on supported catalyst, we have performed simulations scrutinizing qualitatively new effects in the reaction kinetics on the nm scale. The attention is focused on the particle size range where the catalyst particles have reached sufficient size to essentially have the electronic properties of bulk crystals, but where they are still small enough that the kinetic effects connected with the particle size are important. We treat in detail such factors inherent for nm chemistry as reactant supply via the support, interplay of the reaction kinetics on different facets of supported particles due to interfacet diffusion, adsorbate-induced reshaping of catalyst particles, and oscillation and chaos on the nm scale. All these factors are demonstrated to be especially important in the case of rapid catalytic reactions occuring far from the adsorption-desorption equilibrium. As a general conclusion, we find that kinetic effects alone (i.e. no special kinetic effects or "active sites") can cause large differences in reaction kinetics on nm particles and single crystals.





In-situ dynamic study of hydrogen oxidation on rhodium

by Thierry Visart de Bocarme

Universite Libre de Bruxelles, 1050 Bruxelles, Belgium

Abstract