Low-temperature vibrational dynamics of fused silica and binary silicate glasses

Abstract

Inelastic neutron scattering was used to study the vibrational dynamics of fused silica and its mixed binary glasses that were doped with either T iO2 or K2O. The energy transfer were measured from zero to 180 meV where the so-called Boson peaks (BP) at low energy and molecular vibrations at high energy are included. Although most of the vibrational spectra at the high energy resemble those reported in earlier literatures, a defect mode like peak is observed for the doped binary systems near 120 meV. At very low temperature, the BP intensity increases rapidly with temperature and then, at higher temperature, the peak intensity decreases. As a result, a maximum is observed in the temperature dependence of the BP intensity. This maximum was shown in all four samples but the pure SiO2 sample shows the highest intensity peak and the lowest temperature for peak position. Broadband energy spectra reveal a shift of intensity from BP to the more localized modes at higher energy. Temperature evolution of BP and its relationship with heat conduction and thermal expansion are discussed.

Simulation of Inelastic Neutron Scattering Spectra Directly from Molecular Dynamics Trajectories

“…To demonstrate the capability in a representative system, Figure 5 shows an example of silica glass. 43 The simulation of the crystalline phase (α-quartz) with DFT-LD produces a spectrum with more and sharper peaks compared to the experiment. To make a glass with a larger cell, the CHIK potential 44 was used with the LAMMPS 45,46 MD code.…”

“…Compared to the conventional DFT-LD approach, the MD method has obvious advantages in dealing with such complex systems. To demonstrate the capability in a representative system, Figure shows an example of silica glass . The simulation of the crystalline phase (α-quartz) with DFT-LD produces a spectrum with more and sharper peaks compared to the experiment.…”

“…Figure 5. INS spectrum of fused silica measured at VISION,43 in comparison with DFT-LD-simulated spectrum for α-quartz crystal, as well as MD-simulated spectra for silica glasses (made at two different cooling rates) using the CHIK potential.…”

Computing inelastic neutron scattering spectra from molecular dynamics trajectories

“…While there are several experimental techniques to study local dynamics, inelastic neutron scattering (INS) stands apart. INS experiments provide valuable information on the structure and dynamics of molecular systems such as polymers 2 , 4 , metal-organic frameworks 7 , molecular crystals 12 , 13 , and crystalline oxide materials 14 . Unlike optical methods such as FTIR and Raman, INS has no selection rules based on molecular symmetry, meaning all overtones are observable 11 .…”

Predicting Quantum Many-Body Dynamics with Transferable Neural Networks*

“…Introduction -Machine learning (ML) approaches, particularly neural networks (NNs), have achieved great success in solving real-world industrial and social problems [1], such as image recognition [2], high level image synthesis and style transfer [3], human-like raw speech generator [4], producing original melodious MIDI notes [5,6], neural machine translation [7]. Inspired by its widespread applicability, ML was soon adopted by condensed matter physicists in the modeling of quantum many-body behavior and phase transition discovery [8][9][10][11][12][13][14][15][16][17][18]. Compared to so many advances in computer vision [19], speech processing [20], and natural language processing [21], it is natural to ask if recent progress in these more sophisticated ML architectures can benefit or even revolutionize the modeling of quantum systems.…”