COLLECTIVE ASPECTS OF STOCHASTIC
NON-EQUILIBRIUM PHENOMENA
AT SURFACES AND INTERFACES

Lorentz Center, Leiden University, 14-25 June 2004

ABSTRACTS:

---

Timothy Halpin-Healy

Barnard College, Columbia University, NY

Within the Realm of KPZ

We review salient features and commonalities linking the statistical mechanics of directed polymers in random media (DPRM), stochastic kinetic roughening phenomena, and the nonequilibrium dynamics of driven lattice gases. Despite nearly two decades of solid work, KPZ-related statistical mechanics remains a very active field of research running the gamut from experimental projects involving, e.g., flameless fire lines, fiber deposition, and billowing cumulus clouds, to the large-scale structure of the universe itself. Compelling theoretical issues have recently been resolved in the beautiful work of Derrida and Lebowitz in calculating the skewed height distribution of 1+1 KPZ, followed up by Prahofer and Spohn, who link the driven lattice gas ASEP (& therefore its blood relations: KPZ, DPRM) to random matrix theory. Continued work on the KPZ triumvirate: kinetic roughening, directed polymer, and driven lattice gases, suggests that this stochastic nonlinear PDE has done, in fell-swoop, for both far-from-equilibrium dynamics & ill-condensed matter, what the 2d Ising Model did years ago critical phenomena.

In this talk the challenges and pitfalls of atomistic materials modeling are briefly reviewed. The principal method to be discussed is the well-known Molecular Dynamics method, which offers atomic scale resolution and insight into full dynamical time evolution. These properties are crucial for the study of nonequilibrium processes, which lie at the heart of the materials behavior. Thereafter a few examples of ongoing work at Delft will be presented: (i) Structure development of a thin film growing on a non-matching substrate, (ii) Dynamics of a moving interface during phase transformation, (iii) Thermal fluctuations acting as key agents in the competition between crack propagation and dislocation emission, and (iv) Erosion of a silicon surface under ion bombardment. The urgent need for a better modeling of silicon emerges from (vi), and the assistance of ab initio calculations will be discussed.

Work in collaboration with M. Dienwiebel, N. Pradeep, K.B. Jinesh1, G.S. Verhoeven1, and J.A. Heimberg

We have constructed a frictional force microscope (FFM) that is able to quantitatively track the forces between a tip and a sample in three dimensions, with a friction force resolution as low as 15 pN, even under normal loads up to several tens of nN. At the heart of the FFM is a dedicated, microfabricated silicon sensor, the "Tribolever"[1].

Our measurements with tungsten tips sliding on graphite surfaces, show a big surprise [2]. We observe familiar, atomic-scale stick-slip motion, in which the tip follows a "least-resistance", zig-zag path over the corrugated graphite surface. However, the amplitude of the friction forces depends strongly on the relative orientation of the tip with respect to the graphite surface. When we rotate the graphite, the average friction force varies between a high and a near-zero value. Our observations support a simple interpretation, in which a small graphite flake intervenes between the tungsten tip and the graphite substrate. The FFM actually records the lateral forces between the flake and the substrate, i.e. the forces between two parallel graphite lattices. By rotating the substrate with respect to the tip, we periodically go through fully aligned and completely misoriented configurations. When the misalignment is sufficiently severe, the lateral forces on the C-atoms in the flake cancel, thereby dramatically reducing the total friction force. This phenomenon has been predicted more than ten years ago, and is referred to as superlubricity [3].

[1] T. Zijlstra et al., Sensors and Actuators: A. Physical 84 (2000) 18.
[2] M. Dienwiebel et al., Phys. Rev. Lett. 92 (2004) 126101.
[3] M. Hirano and K. Shinjo, Phys. Rev. B. 41 (1990) 11837.

In collaboration with A. Delin, A. Dal Corso, A. Smogunov, R.Weht, P. Gava, M. Wierzbowska, M. Fabrizio, G.E. Santoro, and L. De Leo

Nanocontacts that linger between two metal pieces about to break apart and also those that emerge in tip-metal geometries are known to take in some instances the shape of a (segment of a) regular ultra-thin suspended nanowire. I will address here some theoretical issues connected with this phenomenon.

Firstly, I will discuss why nanowires arise, what is their stability and their thinning behavior with time, and what determines the formation of "magic" long lived nanowires, endowed with specially stable structures.[1] I will show that magic nanowires may even be monatomic, as observed in some cases. Some predicted features of the thinning process [2] will be checked against recent experiments.[3]

Secondly, I will focus on the 4th and 5th row transition metals, that are nonmagnetic in bulk, and discuss the possible onset of local magnetism in their nanowires. I will show, ed on zero-temperature electronic structure calculations, that magnetism may generally occur in monatomic nanowires of Rh, Ru, and Pd, and also of Pt,Os and of Ir under stress [4]. Even if weak, nanowire magnetism can arise owing to d-band narrowing, sometimes accompanied by some extra emptying of the d-bands in favor of s-bands relative to the bulk metal. In a Pd monatomic wire we find that one-dimensional band edge singularities represent an extra factor that triggers awake the otherwise dormant Hund's rule magnetism.[5]

Analysis of the band structures of nanowires indicates that the onset of magnetism should generally reduce, although not by much, the number of conducting bands crossing the Fermi level. This in turn suggests that ballistic conductance through the wires should also be affected by magnetism, at least at zero temperature.[6] Conductance calculations based on an extension of the complex band structure method to ultrasoft pseudopotentials are presently under way, in order to describe that.[7]

A discussion of the expected transition metal nanowire conductance in connection with experimental data [8,9] will also be given. A novel interpretation of the fractional conductance peaks recently reported will be discussed.

Work sponsored by MIUR COFIN, and MIUR FIRB.

[1] E. Tosatti, S. Prestipino, S. Kostlmeier, A. Dal Corso, and F. Di Tolla, Science, 291, 288 (2001)
[2] E.A. Jagla and E. Tosatti, Phys. Rev. B 64, 205412 (2001)
[3] A.I. Mares, A.F. Otte, R.H.M. Smit, J.M. van Ruitenbeek, cond-mat/0401330
[4] A. Delin and E. Tosatti, Phys. Rev. B {\bf 68}, 144434 (2003)
[5] A. Delin, E. Tosatti, and R. Weht, Phys. Rev. Lett. 92, 057201 (2004)
[6] A. Smogunov, A. dal Corso and E. Tosatti, Surface Science 507, 609 (2002)
[7] A. Smogunov, A. dal Corso and E. Tosatti, Surf. Sci. 532, 549 (2003); and in preparation
[8] V. Rodrigues, J. Bettini, P.C. Silva, and D. Ugarte, Phys. Rev. Lett. 91, 096801 (2003)
[9] C. Untiedt, D.M.T. Dekker, D. Djukic, J.M. van Ruitenbeek, Phys. Rev. B 69, 081401 (2004)

In this presentation I will show that, in contrast to the textbook picture of heterogeneous catalysis, surfaces of catalysts during reactions under "practical" conditions are not static but highly dynamic.

Most surface science studies of catalysis are performed under the well-defined conditions of ultrahigh vacuum or very low gas pressures (<10-9 bar) and there exists a so-called pressure gap of at least nine orders of magnitude between these fundamental studies and the practical conditions of applied catalysis (typically pressures>1 bar and high temperatures). An essential difference is that at elevated pressures the thermodynamic contribution of the gas phase can no longer be neglected and that the presence of the gas phase can even lead to phase transitions of the surface structure and composition. Next to this, the catalytic reaction itself can severely affect the surface structure. Both effects can have dramatic consequences for the activity of the catalyst.

To study model catalysts at practical reaction conditions we used a scanning tunneling microscope, which integrated in a micro-flow reactor, to study the catalytic oxidation of carbon monoxide on platinum [1] and palladium [2] single-crystal surfaces at atmospheric pressures and elevated temperatures. By switching from CO-rich to O2-rich gas flows and vice versa we reversibly oxidized and reduced the surface, as observed in STM movies. The formation of the ultrathin surface oxide had a dramatic (and surprising) positive effect on the CO2 production, which was measured online by means of mass spectrometry. In-situ surface x-ray diffraction experiments, which we performed at the ESRF in Grenoble, confirm this and give structural information of the ultrathin oxides.

We observed significant hysteresis in these surface reduction-oxidation cycles resulting in a CO-pressure window where both the metallic, low reactive surface and the oxidic surface, with the higher reactivity, were stable. This bistability also led to self-sustained oscillations in the CO2 production [2]. Our observations are in full disagreement with existing models for oscillatory CO oxidation and I will present an alternative model for the self-sustained oscillations observed at atmospheric pressure.

[1] B.L.M. Hendriksen, J.W.M. Frenken, Phys. Rev. Lett. 89 (2002) 046101
[2] B.L.M. Hendriksen, S.C. Bobaru, J.W.M. Frenken, Surf. Sci. 552 (2004) 229
[3] STM movies of catalysts in action can be viewed on our website:
http://www.physics.leidenuniv.nl/sections/cm/ip.

Control of crystal nucleation is important for the design of many materials. For this reason, it is important to have a good understanding of the rate-limiting step in nucleation, i.e. the formation of a critical nucleus. As nucleation is a rare event, it is difficult to obtain direct experimental information about the crucial early stages of nucleation.

With the help of computer simulations, it is now possible to study the pathway for nucleation. In my talk, I shall discuss recent studies of heterogeneous crystal nucleation in colloids and related systems.

Structural fluctuations occur on solid surfaces at temperatures well below the melting point, and can be measured both spatially and temporally using real-space imaging tools such as LEEM, REM and STM. They couple to field gradients to create mass transport that can dramatically change the shape of the interface. For nanoscale structures, fluctuations may represent a ignificant fraction of the structural mass, and thus significantly influence the physical properties (e.g. catalytic, magnetic, electrical, optical) of interest. It is thus of interest to consider structural fluctuations not only from the traditional perspective of correlation functions, but also in terms of their stochastic variability.Ê

We have measured step fluctuations on a number of experimental systems using STM, over a wide temperature variation in some cases. Analysis of the correlation functions shows that all systems fall into one of two classes for step wandering, n=2 (non-conserved noise) or n=4 (conserved noise). The time constant for step wandering is strongly (Arrhenius) temperature dependent, but the correlation time is, surprisingly, temperature independent. This result reveals interesting concepts about the effective system size for measurements of fluctuations. In addition, we have analyzed the First-Passage-related probability distributions, persistence and survival. We find the expected power-laws for persistence, with NO temperature dependence, and exponential decay (at late times) for survival with a time constant (and size issues) directly proportional to the correlation time.

Supported by the NSF-MRSEC, Laboratory for Physical Science, NSF-NIRT and DOE-NNI.

Helium crystals show facets as any usual crystal but no other crystal can grow and melt fast enough to make the propagation of crystallization waves possible at its surface. After nearly two decades of controversies, it is now generally accepted that helium crystals are model systems for the general study of crystal surfaces, but also exceptional systems with unique quantum properties.

In this review, we distinguish what is general to all crystals from what is particular to helium. A central issue among the general properties is the "roughening transition", i.e., the phase transition from a smooth facetted state of the crystal surface at low temperature to a rough fluctuating state at high temperature. We will describe the series of experiments which have significantly improved the understanding of this transition and the related critical phenomena. A particular emphasis is given to the renormalization theory of roughening by Nozieres, which has been carefully compared with experimental measurements. Other general properties of crystal surfaces have also been studied in helium, such as the energy of steps on the facets and their mutual interactions, also several instabilities.

Among the various questions which remain open, an important one is the strength of the coupling of the crystal surface to the underlying lattice. This coupling is rather weak in helium 4 but it might be stronger in helium 3; this is surprising because 3He atoms are lighter , so that larger quantum fluctuations should occur and broaden the liquid-solid interface more in 3He than in 4He. When discussing this issue, we will comment on a related problem, namely the exact way how renormalization is truncated in RG calculations. This truncation problem is a weak point in the theory and we will propose new experiments to solve it, mainly independant measurements of the step energy and of the height-height correlation function.

Experiments show that the friction between mutually incommensurate (IC) surfaces can be anomalously low. Numerical research, using as a very simple toy model an IC version of the sine-Gordon model, claimed that this friction could be strictly zero and one coined the word "super-lubricity" for it.

In the first part of this talk it is discussed why one could expect super-lubricity to occur in an IC setting. However, subsequently it is shown that it is lost eventually by a mechanism which is the dynamic analogue of the Aubry transition in the static case. In the second part of the talk the long time behavior is discussed and a dynamical renormalization scheme for the sine-Gordon model is sketched to make contact with the KPZ theory.

It is by now well established, both theoretically and experimentally, that growing vicinal surfaces often undergo step meandering instabilities; for recent reviews see [1,2]. Whereas the first meandering mechanism proposed by Bales and Zangwill is a diffusional instability related to transport across terraces, more recently the attention has shifted to instabilities related to the diffusion along the step edges [3]. In either case, however, step meandering is not the end of the story, as the meander morphology often undergoes secondary,/it> instabilities. In the first part of the talk I will describe a recent detailed study [4] of the secondary instability first observed in computer simulations in [5], where the destabilization of the meander pattern can be traced back to the formation of vacancy islands due to the self-crossing of strongly deformed steps.

The second part is motivated by recent experiments on Cu(100), which reveal the coexistence of step meandering and step bunching during growth [1]. It has been suggested that step bunching is caused in this system by a mechanism related to edge diffusion which, because it requires a preexisting meander, can also be viewed as a secondary instability [3] . To put the issue into perspective, I will address more generally the continuum description of step bunching instabilities and ask to what extent it is possible to infer the underlying mechanism from the analysis of scaling properties of step bunches [6].

The talk is based on joint work with Jouni Kallunki, Vesselin Tonchev and Stoyan Stoyanov.

[1] N. Neel, T. Maroutian, L. Douillard, H.-J. Ernst, J. Phys.: Condensed Matter 15, S3227 (2003).
[2] J. Krug, Introduction to step dynamics and step instabilities, cond-mat/0405066.
[3] P. Politi, J. Krug, Surf. Sci. 446, 89 (2000).
[4] J. Kallunki, J. Krug, Europhys. Lett. (in press), cond-mat/0401216.
[5] M. Rost, P. Smilauer, J. Krug, Surf. Sci. 369, 393 (1996).
[6] A. Pimpinelli, V. Tonchev, A. Vidcoq, M. Vladimirova, Phys. Rev. Lett. 88 , 206103 (2002).

In collaboration with N.C. Bartelt, P.J. Feibelman, F.C. Leonard,and G.L. Kellogg.

In this presentation I will discuss measurements of the domain boundary energy in self-assembled Pb/Cu(111) nanostructures. When Pb is deposited on Cu(111) it self assembles into domain patterns of a Pb-rich and a Pb-poor phase. The self-assembly is driven by a stress difference between the two phases. Using low energy electron microscopy (LEEM) we provide direct evidence that the changing periodicity of the domain patterns with temperature is the result of a changing domain boundary energy rather than a changing stress difference between the two phases. Through a capillary wave analysis of the fluctuations of domain boundaries in this two-phase system we directly obtain values for the boundary energy at different temperatures. The interpretation of the measured energies is not straightforward though and I will discuss how the energy values that we obtain from our capillary wave analysis relate to the different interactions that are present in this system.

The work discussed in this presentation was performed at Sandia National Laboratories, a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company, for the U.S. Department of Energy under Contract DE-AC04-94AL85000.

The rapid and multiply branched growth of sparks is well known, and man-made sparks have a history of more than two centuries. However, the initial growth of the discharge channels with velocities of the order of 1000 km/sec becomes experimentally accessible only now. This growth is characterized by a self-induced enhancement of the electric field at the tip of the discharge channel. Gigantic sprite discharges that rise from thunderclouds upwards towards the ionosphere, have become subject of scientific research only recently; they are believd to be of similar nature as the initial stages of small sparks at atmospheric pressure.

I will review recent observations and then explain the state of microscopic modelling, computations and theoretical concepts. Basically, already a single discharge channel has a multiscale structure with a thin ionization front surrounding a rather inert body. I will present computational results with adaptive grids, and I will discuss the properties of ionization fronts, moving boundary approximations for these fronts, and solutions of the moving boundary problem with conformal mapping methods. The result is the prediction that streamers in a sufficiently high potential can branch spontaneously due to a Laplacian instability as is also observed in computations. This quantitative prediction has to be confronted with older phenomenological concepts for spark branching that lead to models of the type of diffusion limited aggregation.

A surprising discovery during low temperature (160-250K) growth on Pb/Si(111) is the highly uniform island height distribution observed both with SPA-LEED and STM [1]. The origin of this unusual growth mode is Quantum Size Effects( QSE) i.e. the dependence of the energy of the confined electrons on island height. which leads to height selectivity in a robust and reproducible way. It is possible to control the preferred height by controlling either the growth parameters (coverage and temperature) or the initial substrate reconstruction Si(111)-(7x7) vs Si(111)-Pb-(3x3) [2]. Differences in the electronic structure of preferred vs non-preferred island heights have been measured with STS and confirm that growth is driven by QSE[3] These islands are meta-stable so kinetics also play an important role for their formation. We have measured[4] the key controlling barrier DE=0.32eV at the island edges, which determines the incorporation of atoms from the surrounding region to the island top. Since the QSE energy minima stabilizing the heights are shallow it is important to extend their stability to higher (i.e. room)temperature. This is possible by oxygen adsorption (after the islands are grown at lower temperature) because oxygen increases the edge barrier and therefore suppresses Pb atom diffusion to the island top [5]. Furthermore for better correlation of the measured electronic energy spectra (with either STS or angle resolved photo-emission) to the island dimensions, the thickness of the layer between the island and the substrate is measured to be 1ML which determines the potential well width to be used for the energy spectra calculations[6]. A different type of self-organization in the Pb/Si(111)system is the formation of numerous 2-dimensional equilibrium phases (i.e. 15 phases) in the coverage range between 6/5 and 4/3 ML driven by the long range stress mediated interactions [7]. This is one of the best realizations of the well-known prediction in statistical mechanics, the "Devil's Staircase"(DS) . Surprisingly such a DS can be prepared at low temperature with the phases extending over macroscopic distance( 0.5mm the illuminated area by the beam)[8] as seen both with SPA-LEED and STM[8].

[1]. K. Budde et. al Phys. Rev.B61 Rap. Com, 10602 (2000)
[2]. V. Yeh et al. Phys. Rev.Lett. 85, 5158 (2000)
[3]. M Hupalo M.. Tringides Phys. Rev. B 65, 115406 (2002)
[4]. A. Menzel. et al Phys. Rev. B 67 165314 2003
[5]. S. Stepanovsky et al. Surface Science 515 187 (2002)
[6]. V.Yeh , M.Hupalo, E.H.Conrad , M.C.Tringides Surf. Sci. 551/1-2 23 ( 2004)
[7]. M.Hupalo , et al Phys Rev Lett. 90 216106 (2003)
[8].M Yakes et al Phys. Rev. B( May 2004)

Work supported by Ames Laboratory-DOE in collaboration with M.Hupalo, K.Budde, V. Yeh, M. Yakes, S. Stepanovsky, M. Kammler, A. Menzel, E. Conrad, C. Z. Wang, K.M.Ho, J. Schmalian

Systems with trapped (absorbing) states may exhibit a nonequilibrium phase transition from a noise-free inactive phase into an ever-lasting active phase. First, we briefly review the absorbing critical phenomena based on toy models and universality classes, and discuss over some possible applications and experimental realizations. Connections to chemical reactions, surface roughening, and self-organized criticality will be pointed out.

We discuss the formulation of transition rates in terms of path integral over all the paths that join two adjacent local minima via a saddle point region. Direct solution of the paths via standard Langevin equation is not practical for many systems because of the rare nature of the activation events. We describe how this problem can be circumvented through an analytical transformation of the action . This results in an effective negative friction Langevin equation which spontaneously generates "uphill" path. Simple applications to equilibrium and non-equilibrium problems demonstrates the power of this new approach. Finally, we will discuss the relation of this new approach to the standard "Transition State Theory".

Mechanical instabilities play an important role in the understanding of dynamic friction. High-resolution friction force microscopy allows observing such instabilities on very small scales, reflecting the atomic structure of the solids involved. These atomic stick-slip phenomena fascinate us as a sort of elementary mechanical instabilities, which in their relative simplicity may give some insight into microscopic friction processes. I will discuss experimental results obtained on atomically clean and flat surfaces of ionic crystals. The influence of two parameters is described: The normal load, which changes the lateral potential experienced by the sliding contact, and the sliding velocity, which affects the degree of thermal relaxation in the contact. In a second part of my talk, I will describe how dissipation has been measured by means of dynamic non-contact force microscopy methods. These experiments provide mechanical dissipation maps with atomic resolution, showing enhanced dissipation at low-coordinated sites. Quantification of these exciting results is hampered by the non-linear characteristics of the tip-sample interactions.

In collaboration with Per Hyldgaard, Chalmers U. of Technology, Goteborg, Sweden.

Since metallic surface states on (111) noble metals are free-electron like, their propagators can be evaluated analytically [1-3]. Since they are well-screened, one can use simple tight-binding formalism [4] to study their effects. (Applications to metallic surface states on semiconductors may also be fruitful [3]) The needed phase shifts can be extracted from experiment [1-3]. Hence, one can now make quantitative predictions of these slowly-decaying, oscillatory indirect interactions. For the (isotropic!) pair interactions (which decay as the inverse square of adatom-adatom separation), remarkable agreement has been obtained with experiments by two groups [1,5]. We have extended the formalism to consider the full indirect ("triple") interaction of 3 adsorbates, which is the sum of the 3 constituent pair interactions plus the non-pairwise "trio" contribution, which tends to decay with the 5/2 power of perimeter [4]. While some evidence exists that the pair interaction alone is inadequate at non-asymptotic separations [5], there has not yet been a comparable experimental confirmation; trio interactions between adatoms and dimers [6] are likely to be dwarfed by direct-interaction effects in the dimer, but other effects can be envisioned [3].

Here, we concentrate on interactions due to ordered overlayers and to linear defects, relating the latter to the interactions of (n x 1) ordered overlayers and both to the constituent pair and trio interactions. We compare with experimental studies [7] of interactions of adatoms with adchains and of consequent 1D motion of adatoms trapped between two such parallel chains. We discuss implications for step-step interactions (on vicinal surfaces), with attention to the modification of the surface state itself for small terrace widths [8]. We finally comment on extracting pair interactions from first-principles calculations of ordered overlayers or from analogous experiments.

PH was supported by ATOMICS, financed by the Swedish Foundation for Strategic Research;
TLE supported by NSF Grants MRSEC DMR 00-80008 and EEC-0085604.
We are grateful to many of the cited experimentalists for enlightening interchanges.

[1] J. Repp, F. Moresco, G. Meyer, and K.-H. Rieder, Phys. Rev. Lett. 85, 2981 (2000).
[2] P. Hyldgaard and M. Persson, J. Phys.: Condens. Matt. 12, L13 (2000).
[3] P. Hyldgaard and T.L. Einstein, Europhys. Lett. 59, 265 (2002); Surf. Sci. 532-5C, 219, 219 (2003); Appl. Surf. Sci. 212-3, 856 (2003).
[4] T.L. Einstein, in Handbook of Surface Science, vol. 1, ed. W.N. Unertl (Elsevier, Amsterdam, 1996), chap. 11.
[5] N. Knorr, H. Brune, M. Epple, A. Hirstein, M. A. Schneider, and K. Kern, Phys. Rev. B 65, 115420 (2002).
[6] K. Morgenstern, K.-F. Braun, and K.-H. Rieder, submitted preprint.
[7] Jascha Repp, Ph.D. thesis, Freie Universitaet, Berlin (2002), unpublished.
[8] K. Morgenstern, K.-F. Braun, K.-H. Rieder, PRL 89, 226801 (2002) & refs therein.

The friction force on the atomic scale, as seen in Friction Force Microscopy experiments, is for the first time analyzed theoretically in the entire range of scanning velocities and temperatures. Possible regimes of atomic friction are discussed, with an emphasis to a new one - the thermal drift regime - as an important alternative to the known stick-slip case. For this regime, a rigorous analytical solution of the problem is found. Thermally activated jumps are shown to lead to a substantial decrease in the mean friction force with decreasing velocity or increasing temperature: temperature acts as a lubricant. Friction is proven to vanish in the limiting case of zero velocity, in contrast to the suggestions of earlier simulations. A new phenomenon of "thermolubricity" is predicted to be an important alternative to the known "mechanistic" superlubricity: friction can be negligibly small in the cases when the surface corrugation is large enough to produce sizable friction in the framework of known mechanistic models.

Work supported by the Foundation for Fundamental Research on Matter (FOM research project 03PR2272)

The intermittent avalanche dynamics of systems, whether experimental or theoretical, that exhibit "self-organized criticality" bears great resemblance to another class of non-equilibrium dynamics: depinning/pinning transitions. It turns out that this connection can be established more firmly, and the focus of this talk is in explaining how the mapping between SOC and depinning works, and its consequences. These concern several pertinent issues in "classical" SOC discussion: universality classes, the upper critical dimension, and the role of the particular ensemble in which the phenomenon is studied. Finally, connections to the general picture of absorbing state phase transitions are discussed, together with some in particular experimental suggestions for further work.

First-Principles calculations are able to provide accurate total energies. The widespread availability of efficient codes, and efficient machines, means that such calculations are becoming increasingly routine, and have been undoubtedly illuminating, no less in surface-science applications as in other areas of condensed-matter physics and chemistry. Yet, there are many physical questions for which the standard total energy calculations, alone, even if we knew the exact exchange-correlation functional, are not sufficient to answer. In this talk, I will outline some, and efforts in my group to make progress on them.

(1) Dynamics of protons. Measurements of isotope effects associated with diffusion of hydrogen/deutetrium/tritium in metals reveal some extremely interesting effects. For example, in Pd, deuterium diffuses faster than hydrogen, but tritium less. Hydrogen diffusion in BCC metals show marked non-Arrehenius behaviour at low temperature. Can they be explained on the basis of a Erying-type transition state theory? How can one use total energy calculations to develop a theory for proton diffusion going beyond the Born-Oppenheimer approximation?

(2) At finite temperatures, it is the free-energy, and not total energy, which determines the stability of a given system. The former requires one to measure relative entropies, which are however, difficult to access. How important are they in practice? We discuss a specific example of this, concentrating on the evaluation of the residual entropy of ice bi-layers on metal surfaces.

(3) Metal surfaces in strong applied electric fields. We apply a recently developed method (Lozovoi and Alavi, PRB Phys. Rev. B 68, 245416, (2003)) for applying strong fields in the context of total-energy slab calculations to study field-evaporation of Al adtoms from Al surfaces.

In collaboration with Meessoon Ha and Jussi Timonen

Stochastic driven flow along a channel is one of the simplest non-equilibrium statistical processes. It undergoes a dynamic queuing phase transition with novel scaling dimensions. Below the transition, a traffic jam develops, which is macroscopic in the sense that the length of the queue scales linearly with system size. Above the transition, at weak obstructions, only a power-law shaped queue remains. Fast bonds create only power-law shaped depletion queues. The implications of these results to faceting of growing interfaces and localization of directed polymers in random media, both in the presence of a columnar defect are pointed out as well. Phys.Rev.E, 68, 056122 (2003).Ê

Research in collaboration with K.A. Lorincz, M. Welling and C.M. Aegerter.

We study experimentally a 3-dimensional pile of rice with a 1x1 square meters sized floor area. We find that the avalanche distribution is a powerlaw and that finite size scaling (FSS) is obeyed. From a comparison of the FSS exponents with the growth and roughness exponents of the rice surface, we find [1] that for this experimental system two exponent scaling relations are obeyed that were previously derived on theoretical ground by Packzuski, Maslow and Bak [2] (PMB). From the waiting times between avalanches, first- and all-return times are obtained, which obey two other PMB exponent scaling relations. The avalanche distribution exponent in the steady state is compared with the exponent of the gap-equation in the transient state and again found to be in agreement with a PMB relation [3]. In conclusion: our rice pile obeys many of the exponent scaling relations predicted for Self-Organized Criticality (SOC) and thus is a unique experimental system for the study of SOC and for schemes to keep a system from becoming SOC (as is desired for e.g. avalanche prevention). As physically very different system with analogous behaviour we study the vortex landscape in superconductors where e.g. also FFS is observed. In addition, we study in superconducting niobium thin films the effect of changing static disorder. By adding small amounts of hydrogen, which is absorbed in the niobium, we locally destroy superconductivity. Front motion is studied in the same sample as a function of increasing hydrogen concentration and hence increasing disorder. A transition from smooth to rough to fractal flux penetration is observed [4].

[1] C.M. Aegerter, R. Gunther and R.J. Wijngaarden, Phys. Rev. E 67 (2003) 051306
[2] M. Packzuski, S. Maslow and P. Bak, Phys. Rev. E 53 (1996) 441
[3] C.M. Aegerter, K.A. Lorincz, M.S. Welling and R.J. Wijngaarden, Phys. Rev. Lett. 92 (2004) 058702
[4] M.S. Welling, C.M. Aegerter, R.J. Westerwaal, S. Enache,R.J. Wijngaarden, and R. Griessen, Physica C 406 (2004) 100

Adhesion, friction and wear are major problems in many modern high-tech applications. In particular they limit both the fabrication yield and operation lifetime of many microelectromechanical (MEMS) devices. Thus, a good understanding of microscale contact mechanics and adhesion in MEMS is fundamental for the design of such systems.

The van der Waals interaction is the weakest attractive force between solids. But even this interaction is so strong that under ideal conditions the force of order 10000 Newton (the weight of a car) would be necessary to separate two bodies if the contact area between them would be of order 1 square-centimeter. Thus the fundamental question related to adhesion is not why it is sometimes observed but rather why it is usually not observed. In this talk I will address this adhesion paradox.

I have developed a theory of contact mechanics and adhesion between an elastic solid and a hard randomly rough substrate. The theory takes into account that partial contact may occur between the solids on all length scales. I present numerical results for the case where the substrate surface is self affine fractal. When the fractal dimension is close to 2, complete contact typically occur in the macro asperity contact areas, while when the fractal dimension is larger than 2.5, the area of (apparent) contact decreases continuously when the magnification is increased. An important result is that even when the surface roughness is so high that no adhesion can be detected in a pull-off experiment, the area of real contact (when adhesion is included) may still be several times larger than when the adhesion is neglected. Since it is the area of real contact which determines the sliding friction force, the adhesion interaction may strongly affect the friction force even when no adhesion can be detected in a pull-off experiment.

I also briefly consider adhesion relevant to biological systems, e.g., flies, crickets and lizards, where the adhesive microstructures consist of arrays of thin fibers. The effective elastic modulus of the fiber arrays can be very small which is of fundamental importance for adhesion on smooth and rough substrates. I show how the adhesion depend on the substrate roughness amplitude and apply the theoretical results to lizards.

Slowly sheared granular matter does not flow homogeneously like a liquid. Instead, granulates form rigid, solid-like regions separated by narrow shear bands where the material yields and flows. Despite their crucial importance, granular shear flows are still poorly understood, in part because shear localization itself remains enigmatic. The formation of shear bands, which have a typical thickness of five to ten grain diameters is a prominent manifestation of heterogeneity, a crucial thread connecting many of the unusual properties of granular matter. In fact, similar heterogeneities occur in many athermal systems, and part of the interest in granular media stems from the fact that they can easily be manipulated to exhibit an extremely wide range of complex behavior. The shear-heterogeneities are difficult to capture by continuum theories, and a nuisance from a practical point of view. They also hinder the testing of ideas related to granular temperature and jamming.

In this talk, experiments in which very wide shear zones can be created are introduced. Depending on the geometry of the system, we find universal and wide bulk shear zones, extremely delocalized creep flows or narrow wall-localized shear bands. The main properties of these flows will be discussed, emphasizing a few properties that are universal, i.e. particle independent. In the final part of the talk, recent ideas on jamming and athermal systems will be discussed, and an attempt will be made to connect these to our experiments

D. Fenistein and MvH, Nature 425, 256 (2003);
D. Fenistein, J.W. van de Meent and MvH, PRL 92, 094391 (2004).

We discuss a systematic formalism to derive the equations of motion for meniscus and contact line dynamics of fluids in confined geometry. The equations are derived from bulk phase-field formulation, which can be derived from a more microscopic density functional approach. The method is based on a variational approach applied to Rayleigh dissipation and free energy functionals. Through successive projections, equations of motion are obtained for 2D meniscus and 1D contact line. We discuss applications of the formalism to contact line motion in capillary tube with spatial inhomogenieties in the walls. In this case, the contact line fulfills a highly nonlinear equation with quenched stochastic noise. We present results of analysis of such equations for selected cases.