You are here: Home Laboratories Pastewka, Lars Research


We use computer simulations to study materials and processes.

Our research revolves around computer modeling of mechanical and chemical properties of material interfaces and their interaction with fluids and other solids. We have a particular focus on tribological phenomena (contact, adhesion, friction, lubrication, wear) that naturally occur at interfaces. Such processes are important in macro- and microsystems and their control is decisive for the lifetime of a device. For example in miniaturized components that have a high surface to volume ratio, interfacial processes can entirely dominate mechanical behavior: At small scales, strength is determined by surface and not bulk defects, the flow of liquids through nanochannels can be controlled by surface topography and chemistry and surface forces such as adhesion and friction can overcome body forces and lead to stiction. Common to these phenomena is the interplay of local chemistry, long-ranged interaction (such as elasticity) and geometrical disorder (such as surface roughness). Models at atomic, mesoscopic and macroscopic scales are therefore required for their understanding.

Contact mechanics

Contact and adhesion of rough surfaces

Contact, friction and wear of natural and engineered surfaces cannot be understood without consideration of surface roughness. Roughness limits the area of intimate atomic contact to isolated areas where summits (asperities) meet. We use large scale contact mechanical calculations at mesoscopic and continuum sclaes to study the influence of surface topography on macroscopic properties such as contact stiffness, adhesion and friction.

Representative publication: PNAS 111, 3298 (2014)

Contact and adhesion of scanning-probe tips

Spherical objects, such as tips on an atomic-force microscope (AFM), are often used as simplifying models for the contact of a single asperity in a rough contact. AFM experiments are carried out at scales that are in principle accessible by molecular simulations. We carry out such calculations to aid the interpretation of experimental AFM friction, topography and adhesion data.

Representative publications: Nano Lett. 14, 7145 (2014)Appl. Phys. Lett. 108, 221601 (2016)

Solid mechanics

Shear-induced melting and solid state amorphization

Mechanically sheared systems can undergo structural transitions. Examples are colloidal crystals that shear-melt or the solid state amorphization of diamond or silicon during tribological loading. Such disordered phases often appear on tribologically loaded interface and it is believe that they constitute a form of solid lubricant. We use molecular dynamics calculations of simple model materials to study the formation of glasses during mechanical shear.

Representative publication: Nature Materials 10, 34 (2011)

Mechanics of amorphous materials

Motivated by the above, we are also interested in the mechanical behavior of traditional bulk amorphous materials. These deform in localized rearrangements of regions of ~100 atoms, so called shear transformations, which coalesce into shear bands on long time scales. We study the deformation of network and metallic glasses in molecular dynamics calculations.

Representative publication: Tribol. Lett. 53, 119 (2014)

Fluid mechanics


Liquids are ubiquitously used to reduce friction and wear in machinery. We use molecular dynamics calculation to study simple lubricants in confined dimensions and the effect of surface topography and surface chemistry on their flow. We are particularly interested in squeeze out of lubricant and the transition from lubricated to solid contact under boundary lubrication conditions.

Representative publication: Science Advances 2, e1501585 (2016)

Mesoscale NMR models

Besides spectroscopy and imaging, nuclear magnetic resonance (NMR) can be used to measure dynamic fluidic properties, such as the diffusion constant or flow velocities. By means of mesoscale particle based simulation with included thermal fluctuations, we can predict echo signals of NMR pulse sequences in arbitrary flow geometries and support the prediction of transport coefficients.

Representative publication: J. Chem. Phys. 144, 244101 (2016)

Microreactor simulation

Chemical micro reactors promise production with high throughput and selectivity. The downside is a more difficult process control with strong limitations for conventional measurement techniques used in macroscopic reactors. Modelling and simulation is therefore mandatory to assist process control. We perform such simulations (among others) for micro-mixers used for continuous hydrothermal synthesis of metal oxide nanoparticles. Here, a cold metal salt stream is mixed with a near- or supercritical water. The later possesses dramatically different solubility properties from conventional water, leading to precipitation of nanoparticles. Beyond computational fluid dynamics, modeling hence requires a correct representation of the water equation of state and all the reactants and reaction products.

Soft matter

Graphene nanoresonators

Constructing nanodevices from single molecules opens the door to sensors and actuators operating on the molecular scale. This allows, e.g., to detect atomistic displacements, the presence of individual molecules, or molecular forces. Conventional all atom simulation of these devices is computationally too expensive due to the large number of atoms and relatively long time scales. Therefore we derive coarse-grained models from the atomistic dynamics by applying the systematic Mori-Zwanzig formalism from statistical mechanics. This gives us a new reduced set of equations with explicit dissipative and noise terms.

Representative publications: J. Chem. Phys. 134, 064106 (2011)J. Chem. Phys. 137, 234103 (2012)

Interaction of cell membranes with their environment

It is known from hydrodynamics that floating objects "see" walls and similar obstacles through hydrodynamic forces, before touching them. How strong is this effect for metabolomic objects near cell membranes? When does the cell "see" for the first time an approaching object? To answer this question we develop mesoscopic models of cell membranes and let them interact with a surrounding medium and nearby nanobeads, in close collaboration with experimental groups at IMTEK.

Personal tools