Disconnecting structure and dynamics in glassy thin films

Abstract

Nanometrically thin glassy films depart strikingly from the behavior of their bulk counterparts. We investigate whether the dynamical differences between bulk and thin film glasses can be understood by differences in local microscopic structure. We employ machine-learning methods that have previously identified strong correlations between local structure and particle rearrangement dynamics in bulk systems. We show that these methods completely fail to detect key aspects of thin-film glassy dynamics. Furthermore, we show that no combination of local structural features drawn from a very general set of two-and multi-point functions is able to distinguish between particles at the center of film and those in intermediate layers where the dynamics are strongly perturbed.Confinement of glassy materials to nanometric length scales leads to striking changes to their microscopic dynamics and consequently to their material properties [1]. Direct observations in both experiments and simulations find exponentially more particle rearrangements near the free surface of a film [2][3][4][5]. A key question is whether enhanced dynamics near a free surface (or suppressed dynamics near a substrate) are connected with structural changes. So far, all structural features studied decay too rapidly into the bulk [1] to explain the altered dynamics.Until recently, the study of bulk glassy systems has also been plagued by the inability to connect dynamical changes with structural ones. However, machine learning methods have proven remarkably successful in identifying a local structural quantity, termed "softness" and denoted S i for particle i, that is strongly correlated with particle rearrangements [6][7][8]. Softness is over an order of magnitude more predictive of rearrangements than measures such as the local potential energy or coordination number [9]. The average softness, S , is directly predictive of the relaxation time of a bulk supercooled liquid [7] or aging bulk glass [8], with higher values of S corresponding to shorter relaxation times at higher temperatures. In bulk systems, it is therefore now clear that dynamical slowing down near the glass transition is intimately associated with structural changes. Here we ask whether the enhancement of dynamics near the surface of free glassy films can similarly be understood.The answer is no. Not only does softness fail to predict the enhanced dynamics near the surface of free glassy films, we find that for a very general set of quantities that characterize the local structural environment surrounding a particle, there is no combination of these quantities that can distinguish between parts of the film with very different dynamics. The enhanced dynamics near a free glassy surface therefore appear to be fundamentally different from the enhanced dynamics that result from heating bulk glassy systems. Although we cannot rule out the possibility that structural quantities that we have not considered might tell a different story, our results suggest that near glassy free surfaces, relaxat...

Structure-property relationships from universal signatures of plasticity in disordered solids

“…Using a machine-learning approach akin to linear regression (22), the weighting function is designed to optimize the prediction accuracy for rearrangements (19). In Lennard-Jones glasses (19) and oligomer glasses (38), it has been shown that the energy barrier that must be surmounted for the particle to rearrange decreases linearly with increasing softness. Thus, rearrangements are exponentially more likely to involve particles with high softness.…”

Machine learning characterization of structural defects in amorphous packings of dimers and ellipses

“…We start our analysis of packings made up of the three particle shapes using the established approach and workflow in Refs. [6,[24][25][26][27][28][29][30].…”

“…To define a predictive structural signature, not only for granular materials but for a wide array of disordered systems, researchers have devised and implemented a machine learning technique that computes a single structural parameter that strongly correlates with rearrangement probability [6,[24][25][26][27][28][29][30]. This parameter is known as softness and arises from a support vector machine (SVM) that computes a hyperplane best separating rearranging and non-rearranging particles, given a multitude of structural features ascribed to each particle.…”

Heterogeneous Activation, Local Structure, and Softness in Supercooled Colloidal Liquids

“…Very recent studies have shown that data from molecular dynamics simulations or experiments can be analyzed with machine learning methods to infer a structural order parameter for dynamics in supercooled liquids and disordered solids called "softness" [17][18][19][20][21][22][23]. An analysis based on softness was applied to supercooled liquids, and was shown to simplify conceptual understanding of phenomena such as heterogeneous dynamics and non-exponential relaxation [18], history dependence during aging [19] and dynamics in thin films [20]. Despite these successes, the application of softness to supercooled liquids has been limited to date to simulations.…”