The energy landscape of glassy dynamics on the amorphous hafnium diboride surface

Abstract

Direct visualization of the dynamics of structural glasses and amorphous solids on the sub-nanometer scale provides rich information unavailable from bulk or conventional single molecule techniques. We study the surface of hafnium diboride, a conductive ultrahigh temperature ceramic material that can be grown in amorphous films. Our scanning tunneling movies have a second-to-hour dynamic range and single-point current measurements extend that to the millisecond-to-minute time scale. On the a-HfB2 glass surface, two-state hopping of 1-2 nm diameter cooperatively rearranging regions or "clusters" occurs from sub-milliseconds to hours. We characterize individual clusters in detail through high-resolution (<0.5 nm) imaging, scanning tunneling spectroscopy and voltage modulation, ruling out individual atoms, diffusing adsorbates, or pinned charges as the origin of the observed two-state hopping. Smaller clusters are more likely to hop, larger ones are more likely to be immobile. HfB2 has a very high bulk glass transition temperature Tg, and we observe no three-state hopping or sequential two-state hopping previously seen on lower Tg glass surfaces. The electronic density of states of clusters does not change when they hop up or down, allowing us to calibrate an accurate relative z-axis scale. By directly measuring and histogramming single cluster vertical displacements, we can reconstruct the local free energy landscape of individual clusters, complete with activation barrier height, a reaction coordinate in nanometers, and the shape of the free energy landscape basins between which hopping occurs. The experimental images are consistent with the compact shape of α-relaxors predicted by random first order transition theory, whereas the rapid hopping rate, even taking less confined motion at the surface into account, is consistent with β-relaxations. We make a proposal of how "mixed" features can show up in surface dynamics of glasses.

Composition-dependent metallic glass alloys correlate atomic mobility with collective glass surface dynamics

“…Typical large scale images of the surfaces before and after re-melting are shown in Fig. 6,[9][10][11] After one cycle of re-melting, the surfaces of La 60 Ni 15 Al 25 , La 50 Ni 15 Al 25 Cu 10 and La 60 Ni 10 Al 25 Cu 5 crystallized, as evidenced by regular patterns (Fig. All as-cast surfaces (Fig.…”

“…Such hopping is not observed on the other surfaces after re-melting. [6][7][8][9][10][11]20 The cluster size distributions for as-cast, re-melted 1Â and re-melted 2Â surfaces of La 60 Ni 15 Al 15 Cu 10 are shown in Fig. 4g-i).…”

“…[6][7][8][9][10][11] Telegraph-like hopping also occurs on polymer surfaces, although the shape of the cooperatively rearranging regions has not been resolved. [6][7][8][9][10][11] Telegraph-like hopping also occurs on polymer surfaces, although the shape of the cooperatively rearranging regions has not been resolved.…”

Colloids exposed to random potential energy landscapes: From particle number density to particle-potential and particle-particle interactions

“…1 The concept of a PEL is successfully used in many fields of science to determine the properties and behavior of systems ranging from small, large and polymeric molecules, proteins and other biomolecules to clusters, glasses and biological cells. 1 It is also applied to describe the transport over atomic surfaces, [2][3][4] in materials with defects (e.g., ions in zeolites 5 or charge carriers in conductors with impurities 6 ), in inhomogeneous media 7,8 (e.g., porous gels, 9 cell membranes 10 or cells [11][12][13][14] ) or in the presence of fixed obstacles as in a Lorentz gas. 15 They are also used to determine the rates of (bio)chemical reactions, 1,16 the folding of proteins and DNA, [17][18][19][20][21][22][23] as well as the particle dynamics in dense suspensions close to freezing, 24 in glasses 4,[25][26][27][28][29][30][31][32][33][34][35] or, more general, in crowded systems.…”

“…1 It is also applied to describe the transport over atomic surfaces, [2][3][4] in materials with defects (e.g., ions in zeolites 5 or charge carriers in conductors with impurities 6 ), in inhomogeneous media 7,8 (e.g., porous gels, 9 cell membranes 10 or cells [11][12][13][14] ) or in the presence of fixed obstacles as in a Lorentz gas. 15 They are also used to determine the rates of (bio)chemical reactions, 1,16 the folding of proteins and DNA, [17][18][19][20][21][22][23] as well as the particle dynamics in dense suspensions close to freezing, 24 in glasses 4,[25][26][27][28][29][30][31][32][33][34][35] or, more general, in crowded systems. 12 We focus on random potential energy landscapes (rPEL), which have been used in the interpretation of several experimental observations.…”