The kinetics of homogeneous melting beyond the limit of superheating

Low pressure reversibly driving colossal barocaloric effect in two-dimensional vdW alkylammonium halides

Plastic crystals as barocaloric materials exhibit the large entropy change rivalling freon, however, the limited pressure-sensitivity and large hysteresis of phase transition hinder the colossal barocaloric effect accomplished reversibly at low pressure. Here we report reversible colossal barocaloric effect at low pressure in two-dimensional van-der-Waals alkylammonium halides. Via introducing long carbon chains in ammonium halide plastic crystals, two-dimensional structure forms in (CH3-(CH2)n-1)2NH2X (X: halogen element) with weak interlayer van-der-Waals force, which dictates interlayer expansion as large as 13% and consequently volume change as much as 12% during phase transition. Such anisotropic expansion provides sufficient space for carbon chains to undergo dramatic conformation disordering, which induces colossal entropy change with large pressure-sensitivity and small hysteresis. The record reversible colossal barocaloric effect with entropy change ΔSr ~ 400 J kg-1 K-1 at 0.08 GPa and adiabatic temperature change ΔTr ~ 11 K at 0.1 GPa highlights the design of novel barocaloric materials by engineering the dimensionality of plastic crystals.

Davemaoite as the mantle mineral with the highest melting temperature

Knowledge of high-pressure melting curves of silicate minerals is critical for modeling the thermal-chemical evolution of rocky planets. However, the melting temperature of davemaoite, the third most abundant mineral in Earth's lower mantle, is still controversial. Here, we investigate the melting curves of two minerals, MgSiO3 bridgmanite and CaSiO3 davemaoite, under their stability field in the mantle by performing first-principles molecular dynamics simulations based on the density functional theory. The melting curve of bridgmanite is in excellent agreement with previous studies, confirming a general consensus on its melting temperature. However, we predict a much higher melting curve of davemaoite than almost all previous estimates. Melting temperature of davemaoite at the pressure of core-mantle boundary (~136 gigapascals) is about 7700(150) K, which is approximately 2000 K higher than that of bridgmanite. The ultrarefractory nature of davemaoite is critical to reconsider many models in the deep planetary interior, for instance, solidification of early magma ocean and geodynamical behavior of mantle rocks.

Melting temperature prediction using a graph neural network model: From ancient minerals to new materials

The melting point is a fundamental property that is time-consuming to measure or compute, thus hindering high-throughput analyses of melting relations and phase diagrams over large sets of candidate compounds. To address this, we build a machine learning model, trained on a database of ∼10,000 compounds, that can predict the melting temperature in a fraction of a second. The model, made publicly available online, features graph neural network and residual neural network architectures. We demonstrate the model's usefulness in diverse applications. For the purpose of materials design and discovery, we show that it can quickly discover novel multicomponent materials with high melting points. These predictions are confirmed by density functional theory calculations and experimentally validated. In an application to planetary science and geology, we employ the model to analyze the melting temperatures of ∼4,800 minerals to uncover correlations relevant to the study of mineral evolution.

Superionic iron alloys and their seismic velocities in Earth's inner core

Earth's inner core (IC) is less dense than pure iron, indicating the existence of light elements within it¹. Silicon, sulfur, carbon, oxygen and hydrogen have been suggested to be the candidates^2,3, and the properties of iron-light-element alloys have been studied to constrain the IC composition^4-19. Light elements have a substantial influence on the seismic velocities^4-13, the melting temperatures^14-17 and the thermal conductivities^18,19 of iron alloys. However, the state of the light elements in the IC is rarely considered. Here, using ab initio molecular dynamics simulations, we find that hydrogen, oxygen and carbon in hexagonal close-packed iron transform to a superionic state under the IC conditions, showing high diffusion coefficients like a liquid. This suggests that the IC can be in a superionic state rather than a normal solid state. The liquid-like light elements lead to a substantial reduction in the seismic velocities, which approach the seismological observations of the IC^20,21. The substantial decrease in shear-wave velocity provides an explanation for the soft IC²¹. In addition, the light-element convection has a potential influence on the IC seismological structure and magnetic field.

Colossal and reversible barocaloric effect in liquid-solid-transition materials n-alkanes

Emerging caloric cooling technology provides a green alternative to conventional vapor-compression technology which brings about serious environmental problems. However, the reported caloric materials are much inferior to their traditional counterparts in cooling capability. Here we report the barocaloric (BC) effect associated with the liquid-solid-transition (L-S-T) in n-alkanes. A low-pressure of ~50 MPa reversibly triggers an entropy change of ~700 J kg⁻¹ K⁻¹, comparable to those of the commercial refrigerants in vapor-based compression systems. The Raman study and theoretical calculations reveal that applying pressure to the liquid state suppresses the twisting and random thermal motions of molecular chains, resulting in a lower configurational entropy. When the pressure is strong enough to drive the L-S-T, the configurational entropy will be fully suppressed and induce the colossal BC effect. This work could open a new avenue for exploring the colossal BC effect by evoking L-S-T materials.

Barocaloric effect, previously reported in solid-solid phase transition materials, offers a green alternative to current cooling technology. Here the authors report colossal BCE in n-alkanes associated with liquid–solid transition.

Uncertainty quantification in classical molecular dynamics

Molecular dynamics simulation is now a widespread approach for understanding complex systems on the atomistic scale. It finds applications from physics and chemistry to engineering, life and medical science. In the last decade, the approach has begun to advance from being a computer-based means of rationalizing experimental observations to producing apparently credible predictions for a number of real-world applications within industrial sectors such as advanced materials and drug discovery. However, key aspects concerning the reproducibility of the method have not kept pace with the speed of its uptake in the scientific community. Here, we present a discussion of uncertainty quantification for molecular dynamics simulation designed to endow the method with better error estimates that will enable it to be used to report actionable results. The approach adopted is a standard one in the field of uncertainty quantification, namely using ensemble methods, in which a sufficiently large number of replicas are run concurrently, from which reliable statistics can be extracted. Indeed, because molecular dynamics is intrinsically chaotic, the need to use ensemble methods is fundamental and holds regardless of the duration of the simulations performed. We discuss the approach and illustrate it in a range of applications from materials science to ligand-protein binding free energy estimation. This article is part of the theme issue 'Reliability and reproducibility in computational science: implementing verification, validation and uncertainty quantification in silico'.

Melting of CaSiO3 Perovskite at High Pressure

Ab initio molecular dynamics simulations predict that CaSiO3 perovskite melts at 5600 K at 136 GPa, and 6400 K at 300 GPa, significantly higher than MgSiO3 perovskite. The entropy of melting (1.8 kB per atom) is much larger than that of many silicates at ambient pressure and of simple liquids and varies little with pressure. The volume of melting decreases rapidly with increasing pressure, to 3 % at 136 GPa, producing a melting slope that diminishes rapidly with pressure. We determine the melting temperature via the ZW method, combining the Z method, for which we clarify the theoretical basis, with a waiting time analysis. The ZW method results are internally confirmed by integrating the Clausius-Clapeyron equation, which also yields our results for the entropy and volume of melting. We find the eutectic composition on the MgSiO3-CaSiO3 join to be x Ca = 0.26 at 136 GPa and that metasilicate melt is denser than coexisting silicates.

Computational Study on Homogeneous Melting of Benzene Phase I

Molecular-dynamics simulations are used for examining the microscopic details of the homogeneous melting of benzene phase I. The equilibrium melting temperatures of our model were initially determined using the direct-coexistence method. Homogeneous melting at a higher temperature is achieved by heating a defect- and surfacefree crystal. The temperature-dependent potential energy and lattice parameters do not indicate a premelting phase even under superheated conditions. Further, statistical analyses using induction times computed from 200 melting trajectories were conducted, denoting that the homogeneous melting of benzene occurs stochastically, and that there is no intermediate transient state between the crystal and liquid phases. Additionally, the critical nucleus size is estimated using the seeding approach, along with the local bond order parameter. We found that the large diffusive motion arising from defect migration or neighbor-molecule swapping is of little importance during nucleation. Instead, the orientational disorder activated using the flipping motion of the benzene plane results in the melting nucleus.

High-pressure/high-temperature phase diagram of zinc

The phase diagram of zinc (Zn) has been explored up to 140 GPa and 6000 K, by combining optical observations, x-ray diffraction, and ab initio calculations. In the pressure range covered by this study, Zn is found to retain a hexagonal close-packed (hcp) crystal symmetry up to the melting temperature. The known decrease of the axial ratio (c/a) of the hcp phase of Zn under compression is observed in x-ray diffraction experiments from 300 K up to the melting temperature. The pressure at which c/a reaches [Formula: see text] (≈10 GPa) is slightly affected by temperature. When this axial ratio is reached, we observed that single crystals of Zn, formed at high temperature, break into multiple poly-crystals. In addition, a noticeable change in the pressure dependence of c/a takes place at the same pressure. Both phenomena could be caused by an isomorphic second-order phase transition induced by pressure in Zn. The reported melt curve extends previous results from 24 to 135 GPa. The pressure dependence obtained for the melting temperature is accurately described up to 135 GPa by using a Simon-Glatzel equation: [Formula: see text], where P is the pressure in GPa. The determined melt curve agrees with previous low-pressure studies and with shock-wave experiments, with a melting temperature of 5060(30) K at 135 GPa. Finally, a thermal equation of state is reported, which at room-temperature agrees with the literature.

Modeling the melting of multicomponent systems: the case of MgSiO3 perovskite under lower mantle conditions

Knowledge of the melting properties of materials, especially at extreme pressure conditions, represents a long-standing scientific challenge. For instance, there is currently considerable uncertainty over the melting temperatures of the high-pressure mantle mineral, bridgmanite (MgSiO3-perovskite), with current estimates of the melting T at the base of the mantle ranging from 4800 K to 8000 K. The difficulty with experimentally measuring high pressure melting temperatures has motivated the use of ab initio methods, however, melting is a complex multi-scale phenomenon and the timescale for melting can be prohibitively long. Here we show that a combination of empirical and ab-initio molecular dynamics calculations can be used to successfully predict the melting point of multicomponent systems, such as MgSiO3 perovskite. We predict the correct low-pressure melting T, and at high-pressure we show that the melting temperature is only 5000 K at 120 GPa, a value lower than nearly all previous estimates. In addition, we believe that this strategy is of general applicability and therefore suitable for any system under physical conditions where simpler models fail.

Melting curve of SiO2 at multimegabar pressures: implications for gas giants and super-Earths

Ultrahigh-pressure phase boundary between solid and liquid SiO₂ is still quite unclear. Here we present predictions of silica melting curve for the multimegabar pressure regime, as obtained from first principles molecular dynamics simulations. We calculate the melting temperatures from three high pressure phases of silica (pyrite-, cotunnite-, and Fe₂P-type SiO₂) at different pressures using the Z method. The computed melting curve is found to rise abruptly around 330 GPa, an increase not previously reported by any melting simulations. This is in close agreement with recent experiments reporting the α-PbO₂–pyrite transition around this pressure. The predicted phase diagram indicates that silica could be one of the dominant components of the rocky cores of gas giants, as it remains solid at the core of our Solar System’s gas giants. These results are also relevant to model the interior structure and evolution of massive super-Earths.

Shock melting method to determine melting curve by molecular dynamics: Cu, Pd, and Al

A melting simulation method, the shock melting (SM) method, is proposed and proved to be able to determine the melting curves of materials accurately and efficiently. The SM method, which is based on the multi-scale shock technique, determines melting curves by preheating and/or prepressurizing materials before shock. This strategy was extensively verified using both classical and ab initio molecular dynamics (MD). First, the SM method yielded the same satisfactory melting curve of Cu with only 360 atoms using classical MD, compared to the results from the Z-method and the two-phase coexistence method. Then, it also produced a satisfactory melting curve of Pd with only 756 atoms. Finally, the SM method combined with ab initio MD cheaply achieved a good melting curve of Al with only 180 atoms, which agrees well with the experimental data and the calculated results from other methods. It turned out that the SM method is an alternative efficient method for calculating the melting curves of materials.

Superionicity and polymorphism in calcium fluoride at high pressure

We present a combined experimental and computational first-principles study of the superionic and structural properties of CaF_{2} at high P-T conditions. We observe an anomalous superionic behavior in the low-P fluorite phase that consists of a decrease of the normal → superionic critical temperature with compression. This unexpected effect can be explained in terms of a P-induced softening of a zone-boundary X phonon that involves exclusively fluorine displacements. Also we find that superionic conductivity is absent in the high-P cotunnite phase. Instead, superionicity develops in a new low-symmetry high-T phase that we identify as monoclinic (space group P2_{1}/c). We discuss the possibility of observing these intriguing phenomena in related isomorphic materials.

First-principles calculation of entropy for liquid metals

We demonstrate the accurate calculation of entropies and free energies for a variety of liquid metals using an extension of the two-phase thermodynamic (2PT) model based on a decomposition of the velocity autocorrelation function into gas-like (hard sphere) and solid-like (harmonic) subsystems. The hard sphere model for the gas-like component is shown to give systematically high entropies for liquid metals as a direct result of the unphysical Lorentzian high-frequency tail. Using a memory function framework we derive a generally applicable velocity autocorrelation and frequency spectrum for the diffusive component which recovers the low-frequency (long-time) behavior of the hard sphere model while providing for realistic short-time coherence and high-frequency tails to the spectrum. This approach provides a significant increase in the accuracy of the calculated entropies for liquid metals and is compared to ambient pressure data for liquid sodium, aluminum, gallium, tin, and iron. The use of this method for the determination of melt boundaries is demonstrated with a calculation of the high-pressure bcc melt boundary for sodium. With the significantly improved accuracy available with the memory function treatment for softer interatomic potentials, the 2PT model for entropy calculations should find broader application in high energy density science, warm dense matter, planetary science, geophysics, and material science.

Modified Z method to calculate melting curve by molecular dynamics

We extend the recently proposed Z method of estimating the melting temperature from a complete liquid and propose a modified Z method to calculate the melting temperature from a solid-liquid coexistence state. With the simulation box of rectangular parallelepiped, an initial structure of perfect lattice can run in the microcanonical ensemble to achieve steady solid-liquid coexistence state. The melting pressure and temperature are estimated from the coexistence state. For the small system with 1280 atoms, the simulation results show that the melting curve of copper has a good agreement with the experiments and is identical in accuracy with the results of the two-phase coexistence method with 24 000 atoms in the literature. Moreover, the method is conceptually simpler than the two-phase coexistence method.