Molecular dynamics study of phase transitions in Xe

Stochastic Neural Network Approach for Learning High-Dimensional Free Energy Surfaces

The generation of free energy landscapes corresponding to conformational equilibria in complex molecular systems remains a significant computational challenge. Adding to this challenge is the need to represent, store, and manipulate the often high-dimensional surfaces that result from rare-event sampling approaches employed to compute them. In this Letter, we propose the use of artificial neural networks as a solution to these issues. Using specific examples, we discuss network training using enhanced-sampling methods and the use of the networks in the calculation of ensemble averages.

Order-parameter-aided temperature-accelerated sampling for the exploration of crystal polymorphism and solid-liquid phase transitions

The problem of predicting polymorphism in atomic and molecular crystals constitutes a significant challenge both experimentally and theoretically. From the theoretical viewpoint, polymorphism prediction falls into the general class of problems characterized by an underlying rough energy landscape, and consequently, free energy based enhanced sampling approaches can be brought to bear on the problem. In this paper, we build on a scheme previously introduced by two of the authors in which the lengths and angles of the supercell are targeted for enhanced sampling via temperature accelerated adiabatic free energy dynamics [T. Q. Yu and M. E. Tuckerman, Phys. Rev. Lett. 107, 015701 (2011)]. Here, that framework is expanded to include general order parameters that distinguish different crystalline arrangements as target collective variables for enhanced sampling. The resulting free energy surface, being of quite high dimension, is nontrivial to reconstruct, and we discuss one particular strategy for performing the free energy analysis. The method is applied to the study of polymorphism in xenon crystals at high pressure and temperature using the Steinhardt order parameters without and with the supercell included in the set of collective variables. The expected fcc and bcc structures are obtained, and when the supercell parameters are included as collective variables, we also find several new structures, including fcc states with hcp stacking faults. We also apply the new method to the solid-liquid phase transition in copper at 1300 K using the same Steinhardt order parameters. Our method is able to melt and refreeze the system repeatedly, and the free energy profile can be obtained with high efficiency.

Dynamic transition of supercritical hydrogen: defining the boundary between interior and atmosphere in gas giants

Understanding the physics of gas giants requires knowledge about the behavior of hydrogen at extreme pressures and temperatures. Molecular hydrogen in these planets is supercritical, and has been considered as a physically homogeneous state where no differences can be made between a liquid and a gas and where all properties undergo no marked or distinct changes with pressure and temperature, the picture believed to hold below the dissociation and metallization transition. Here, we show that in Jupiter and Saturn, supercritical molecular hydrogen undergoes a dynamic transition around 10 GPa and 3000 K from the "rigid" liquid state to the "nonrigid" gas-like fluid state at the Frenkel line recently proposed, with the accompanying qualitative changes of all major physical properties. The consequences of this finding are discussed, including a physically justified way to demarcate the interior and the atmosphere in gas giants.

An ab initio molecular dynamics study of iron phases at high pressure and temperature

The crystal structure of iron, the major component of the Earth's inner core (IC), is unknown for the IC high pressure (P; 3.3-3.6 Mbar) and temperature (T; 5000-7000 K). There is mounting evidence that the hexagonal close-packed (hcp) phase of iron, stable at the high P of the IC and a low T, might be unstable under the IC conditions due to the impact of high T and impurities. Experiments at the IC P and T are difficult and do not provide a conclusive answer as regards the iron stability at the pressure of the IC and temperatures close to the iron melting curve. Recent theory provides contradictory results regarding the nature of the stable Fe phase. We investigated the possibility of body-centered cubic (bcc) phase stabilization at the P and T in the vicinity of the Fe melting curve by using ab initio molecular dynamics. Thermodynamic calculations, relying on the model of uncorrelated harmonic oscillators, provide nearly identical free energies within the error bars of our calculations. However, direct simulation of iron crystallization demonstrates that liquid iron freezes in the bcc structure at the P of the IC and T = 6000 K. All attempts to grow the hcp phase from the liquid failed. The mechanism of bcc stabilization is explained. This resolves most of the earlier confusion.

Molybdenum at high pressure and temperature: melting from another solid phase

The Gibbs free energies of bcc and fcc Mo are calculated from first principles in the quasiharmonic approximation in the pressure range from 350 to 850 GPa at room temperatures up to 7500 K. It is found that Mo, stable in the bcc phase at low temperatures, has lower free energy in the fcc structure than in the bcc phase at elevated temperatures. Our density-functional-theory-based molecular dynamics simulations demonstrate that fcc melts at higher than bcc temperatures above 1.5 Mbar. Our calculated melting temperatures and bcc-fcc boundary are consistent with the Mo Hugoniot sound speed measurements. We find that melting occurs at temperatures significantly above the bcc-fcc boundary. This suggests an explanation of the recent diamond anvil cell experiments, which find a phase boundary in the vicinity of our extrapolated bcc-fcc boundary.

Xenon melting curve to 80 GPa and 5p-d hybridization

Measurements made in a laser heated diamond-anvil cell are reported that extend the melting curve of Xe to 80 GPa and 3350 K. The steep lowering of the melting slope (dT/dP) that occurs near 17 GPa and 2750 K results from the hybridization of the 5p-like valence and 5p-like conduction states with the formation of clusters in the liquid having icosahedral short-range order (ISRO).

Shock wave propagation in dissociating low-Z liquids: D2

We present direct molecular dynamics simulations of shock wave propagation in liquid deuterium for a wide range of impact velocities. The calculated Hugoniot is in perfect agreement with the gas-gun data as well as with the most recent experimental data. At high impact velocities we observe a smearing of the shock wave front and propagation of fast dissociated molecules well ahead of the compressed region. This smearing occurs due to the fast deuterium dissociation at the shock wave front. The experimental results are discussed in view of this effect.

High-pressure melting of molybdenum

The melting curve of the body-centered cubic (bcc) phase of Mo has been determined for a wide pressure range using both direct ab initio molecular dynamics simulations of melting as well as a phenomenological theory of melting. These two methods show very good agreement. The simulations are based on density functional theory within the generalized gradient approximation. Our calculated equation of state of bcc Mo is in excellent agreement with experimental data. However, our melting curve is substantially higher than the one determined in diamond anvil cell experiments up to a pressure of 100 GPa. An explanation is suggested for this discrepancy.