Low-temperature anomalies of a vapor deposited glass

Abstract

We investigate the low temperature properties of two-dimensional Lennard-Jones glass films, prepared in silico both by liquid cooling and by physical vapor deposition. We identify deep in the solid phase a crossover temperature T * , at which slow dynamics and enhanced heterogeneity emerge. Around T * , localized defects become visible, leading to vibrational anomalies as compared to standard solids. We find that on average, T * decreases in samples with lower inherent structure energy, suggesting that such anomalies will be suppressed in ultra-stable glass films, prepared both by very slow liquid cooling and vapor deposition.Low-temperature crystalline solids are usually described in terms of harmonic vibrations around a perfect periodic lattice (phonons). Within this framework, defects such as vacancies and dislocations can be treated as small perturbations. This description breaks down for amorphous solids such as glasses, foams, emulsions, plastics, colloids, granular materials, bacterial colonies, and tissues [1][2][3][4][5][6][7]. In these systems, the identification of "defects" becomes challenging because the solid ground state is strongly disordered. As a consequence, amorphous solids display many universal anomalies with respect to crystals. Examples are the so-called Boson Peak, an excess of low-energy vibrational modes [8]; the anomalous scaling of heat capacity and thermal conductivity with temperature [9,10]; the irreversible plastic response to arbitrarily small perturbations [1,3,4,11]; and highly cooperative relaxation dynamics, contributing to the socalled β-processes [12][13][14].These anomalies have been widely reported in amorphous solids of very different nature. Interestingly, recent experimental work has shown that by preparing glasses through a process of physical vapor deposition, one can produce ultra-stable states that lie deep in the free energy landscape [15]. Compared to their liquid-cooled counterparts, vapor-deposited glasses show higher density [16] and kinetic stability [15,17]. When these ultra-stable glasses are studied at very low temperatures, it is found that the anomalies characteristic of amorphous solids are strongly supressed [18][19][20][21][22].Many theoretical approaches to this problem are based on the study of the potential energy landscape of glassforming particle systems [13,[23][24][25][26][27][28]. These studies have suggested that glass anomalies can be interpreted in terms of glass states being not well-defined energy minima, but structured metabasins containing a collection of sub-basins separated by barriers of variable size [29,30], see Fig. 1. In particular, recent work [31][32][33] has identified a set of simple observables (the mean square displacement between identical "clones" of the original system) that allows one to detect easily the development of a structure of sub-basins inside a glass metabasin.In this work, using the methods of [31-33], we explore in silico the potential energy landscape of binary Lennard-Jones glass films prepared through tw...

Hierarchical Landscape of Hard Disk Glasses

“…This complex organisation of the landscape contrasts strongly with the behaviour of stable glasses formed using pair potentials with no cutoff, where no jamming transition is present [32,37]. Therefore our findings suggest that systems approaching the jamming transition have a landscape that is more complex than ordinary amorphous solids.…”

“…Remarkably, this aging dynamics is spatially correlated over a relatively large length scale, comparable to the system size L/2 ≈ 15, and involves a large number of particles. Thus, we conclude that the sub-basins structure that we detect for 2d hard disks, is qualitatively distinct from the highly localised defects found in both nearly 1d hard disks and 3d and 2d soft spheres [20,32,37].…”

Universal non-mean-field scaling in the density of states of amorphous solids

“…Near the jamming transition point (p ≪ 1), one can conclude that the Gardner transition plays the dominant role in generating quartic scaling considering the numerical and experimental evidence for the Gardner transition [54][55][56] and the consistency between the mean-field and numerical results [30,57]. However, this scenario may not hold apart from jamming where amorphous solids do not show the strong signature of the Gardner transition [58,59]. In this region, the MCT transition would be the main cause of the quartic scaling.…”

Gardner Transition in Physical Dimensions

“…In this paper we argue that at least the Gardner transition is not a real transition in physical dimensions, d ≤ 3. The Gardner transition is the transition associated with the emergence of a complex free-energy landscape composed of many marginally stable sub-basins contained within a larger glass metabasin [2]. It is thus similar to a state of a spin glass in a phase with broken replica symmetry [3].…”

“…(1), y i (t+t W ) and y i (t W ) take the same sign; similarly, y A,i (t) and y B,i (t) take the same sign in Eq. (2). Then, on the plateau, ∆(t, t W ) and ∆ AB (t) can be approximated by…”