Fulfilling the Multiscale Promise in Materials: Getting Information out of the Atomistic Scale

You can apply to participate and find all the relevant information (speakers, abstracts, program,...) on the event website: https://www.cecam.org/workshop-details/fulfilling-the-multiscale-promise-in-materials-getting-information-out-of-the-atomistic-scale-1283.

Registration is required to attend the full event, take part in the social activities and present a poster at the poster session (if any).  However, the EPFL community is welcome to attend specific lectures without registration if the topic is of interest to their research. Do not hesitate to contact the CECAM Event Manager if you have any question.

Description
The development of new technologies and materials is driven by increasingly challenging requirements: saving energy, minimising waste and the reduction of harmful substances. Atomic-scale simulations are increasingly used in materials design to seek new insights based on mechanistic understanding instead of empiricism. To effectively tackle problems from mechanics, we need multiscale approaches capable of transferring knowledge from the nanoscale to the macroscopic scale [1]. Our workshop will discuss recent developments in this field, new developments made possible by new methodologies, current success stories and limitations.
Our workshop will have an opening session dedicated to our friend and mentor Sandro De Vita. The five-year period since he passed away in 2018 coincides with our last CECAM workshop on this topic and is an appropriate interval over which to assess progress and remaining challenges.
Mechanics
Mechanical properties depend on both microstructure and chemical complexity. The fundamental deformation mechanisms – dislocation motion, twinning, phase transformation, grain boundary sliding – take place at the atomic scale, but the associated microstructural processes span a range of time and space scales that cannot be accommodated by atomistic simulations. A wide variety of scale-bridging methods have attempted to incorporate atomistic information into higher-scale models, either hierarchically (e.g. atomistically informed phase-field models), or concurrently, by embedding quantum mechanics into atomistic simulations [2], or an atomistic region into a continuum domain (e.g. CADD) [3]. These approaches are complemented by recent developments in machine learning [4]. However, design of high-performance industrial materials requires predictive capabilities based on a statistical approach and robust error estimates which are not yet possible.
Thermodynamics
The calculation of phase-diagrams was long-confined to phenomenological approaches like  CALPHAD [5] based on experimental data. However, the last two decades have seen great advances in sampling phase diagrams using electronic structure calculations [6], and nowadays ab-initio-informed thermodynamic databases are routinely used in the industrial development of novel alloys. In upsampled thermodynamic integration [7], a hierarchy of models is fitted to ab initio calculations to compute free energies. Fully unbiased exploration of the phase space is now possible with nested sampling algorithms [8] that efficiently explore the solid configurations and can compute full phase diagrams from atomic-scale information. Recent new developments  include the combined treatment of effects and their thermodynamic state as ‘defect phases’ [9].
Surfaces
Solid/liquid interfaces play a key role in a wide range of materials processes such as corrosion, adhesion and friction. Driven by the pressing need to minimise energy losses, material design is increasingly seeking input from the atomic scale. Energy efficiency is often achieved through reduction of friction, possibly with sustainable lubricants [10]. However, low-viscosity lubricants shift the lubrication mechanism from a purely hydrodynamic to boundary lubrication in which lubricating films at surface asperity contacts can become nanoscale or even disappear [11]. Here, augmenting efficient continuum methods, like Reynolds lubrication equation, a priori or on-the-fly with information calculated at the atomic level (e.g., slip lengths, shear stresses) is a promising way forward. A further example comes in models of hydration and setting of cement, which incorporate interactions between water molecules and ions [12].

References
[1] E. van der Giessen, P. Schultz, N. Bertin, V. Bulatov, W. Cai, G. Csányi, S. Foiles, M. Geers, C. González, M. Hütter, W. Kim, D. Kochmann, J. LLorca, A. Mattsson, J. Rottler, A. Shluger, R. Sills, I. Steinbach, A. Strachan, E. Tadmor, Modelling Simul. Mater. Sci. Eng., 28, 043001 (2020)
[2] J. Kermode, T. Albaret, D. Sherman, N. Bernstein, P. Gumbsch, M. Payne, G. Csányi, A. De Vita, Nature, 455, 1224-1227 (2008)
[3] B. Shiari, R. Miller, Journal of the Mechanics and Physics of Solids, 88, 35-49 (2016)
[4] J. Behler, G. Csányi, Eur. Phys. J. B, 94, 142 (2021)
[5] Z. Liu, J. Phase Equilib. Diffus., 30, 517-534 (2009)
[6] Y. Chang, S. Chen, F. Zhang, X. Yan, F. Xie, R. Schmid-Fetzer, W. Oates, Progress in Materials Science, 49, 313-345 (2004)
[7] B. Grabowski, L. Ismer, T. Hickel, J. Neugebauer, Phys. Rev. B, 79, 134106 (2009)
[8] J. Skilling, Bayesian Anal., 1, (2006)
[9] S. Korte-Kerzel, T. Hickel, L. Huber, D. Raabe, S. Sandlöbes-Haut, M. Todorova, J. Neugebauer, International Materials Reviews, 67, 89-117 (2021)
[10] T. Kuwahara, P. Romero, S. Makowski, V. Weihnacht, G. Moras, M. Moseler, Nat. Commun., 10, 151 (2019)
[11] H. Spikes, Tribol. Lett., 60, 5 (2015)
[12] A. Goyal, I. Palaia, K. Ioannidou, F. Ulm, H. van Damme, R. Pellenq, E. Trizac, E. Del Gado, Sci. Adv., 7, (2021)