International Journal of Thermophysics最新文献

Silicon-containing compounds are widely used in electronic materials, lubricants, coatings, sealants, refrigerants, and many other industrial applications. However, the limited availability of experimental thermodynamic data and the high computational cost of quantum chemical calculations make accurate predictions of their thermodynamic properties challenging. In this study, an integrated machine learning framework is proposed to predict sigma-profiles directly from molecular structures using a Group Contribution-based Graph Convolution Network (GC-GCN) and to perform thermodynamic calculations using the conductor-like screening model with segment activity coefficient (COSMO-SAC) model. To improve the framework's applicability to silicon-containing chemical systems, the GC-GCN training database was expanded to include 546 silicon-containing compounds, bringing the total to 4102 compounds. In addition, an extended version of the COSMO-SAC model, COSMO-SAC-COOH, was adopted to account for carboxylic acid interactions in thermodynamic calculations explicitly. The predicted sigma-profiles showed good agreement with Density Functional Theory (DFT)-based COSMO descriptors and, through the calculated activity coefficients, reproduced vapor–liquid equilibrium behavior across systems exhibiting diverse phase equilibrium characteristics, with a relative root mean square error of 4.56 % for pressure and a root mean square error of 0.02 for vapor-phase mole fraction relative to experimental data. The proposed framework enables reliable thermodynamic predictions directly without performing quantum chemical calculations or generating molecular structure files, providing a practical and computationally efficient tool for predicting thermodynamic and phase equilibrium properties of a broad range of compounds, including silicon-containing compounds.

New experimental phase-equilibrium data were obtained for the asymmetric binary system carbon dioxide (CO₂) + 2,6,10,14-tetramethylpentadecane (pristane), a heavy branched alkane for which available data remain scarce. Fourteen mixtures with known overall compositions were investigated over the temperature range 293.15 K to 363.15 K using a high-pressure variable-volume visual cell and the synthetic-visual method. Bubble point, dew point, liquid–liquid phase separation, and vapor–liquid–liquid equilibrium pressures were determined by direct visual observation. The system exhibits liquid–liquid immiscibility in the CO₂-rich region at temperatures below approximately 310 K. A three-phase vapor–liquid–liquid equilibrium locus was measured between 292.96 K and 309.95 K and was found to terminate at an upper critical end point located at 309.95 K. Above this temperature, the phase behavior becomes essentially classical, with only a vapor–liquid equilibrium region observed. At 298 K, a barotropic inversion (mass density inversion) between the two immiscible liquid phases in equilibrium was observed. The pressure–temperature phase diagram and the interpolated critical loci indicate that the system belongs to type III according to the conventional classification of binary phase diagrams. The experimental data were modeled using the Peng–Robinson equation of state.

Reliable thermophysical property data requires both accurate measurement and transparent evaluation of uncertainty and source traceability. Although the Korean Thermophysical Properties Databank (KDB) has previously provided critically evaluated data, the evaluation workflow for standard reference data has not been described as a coherent and reproducible procedure. In this study, a structured framework for constructing standard reference data within the KDB is formalized and validated. The procedure integrates traceable data collection, iterative outlier rejection, uncertainty evaluation and propagation, independent cross-checking, and expert committee review into a unified evaluation process. This methodology was applied to six thermophysical properties relevant to process design, safety, and environmental assessment: flash point (FP), octanol–water partition coefficient, autoignition temperature, surface tension, acentric factor, and Hildebrand solubility parameter. Applying the evaluation to 82 488 raw literature data points resulted in 2363 registered standard reference data entries. Cross-checking against independent comparison values provided the evidentiary basis for the registration decisions and revealed errors in supporting database entries. The evaluated data have been incorporated into the publicly accessible KDB platform, together with metadata including uncertainty estimates and quality classifications. This framework establishes a systematic methodology for consolidating heterogeneous literature data into uncertainty-aware standard reference data.

The accurate design and performance assessment of energy generation systems that use molten salts as working fluids require accurate characterization of thermal conductivity, with a sufficiently low uncertainty. However, experimental data on several molten salt systems, of interest for various energy generation applications, are generally lacking. Physics-informed models, as well as molecular dynamic (MD) simulations, have the potential to populate gaps in current thermophysical property databases. Existing predictive models using MD simulation (Green–Kubo, non-equilibrium MD) take significantly longer to predict thermal conductivity than classical MD models that otherwise predict salt density with reasonable accuracy. We present an approach that uses the SuperSalt machine learning interatomic potentials (MLIPs) for chloride salts, as well as a structural coherence model (SCM) to predict thermal conductivity of molten salts using volumetric heat capacity, adiabatic speed of sound, and the partial pair distribution function. The SCM uses this function to predict the mean free path of thermal transport and applies generalized kinetic theory to predict thermal conductivity near the melting temperature, achieving results one to two orders of magnitude faster than existing MD methods. By also predicting the isothermal bulk modulus, the temperature dependence of thermal conductivity above the melting temperature can be determined. This integrated SCM-MLIP (ISM) methodology serves as a proof of concept for using MLIPs to predict thermal properties within current experimental uncertainties at a fraction of the computational cost. This enables investigation into thermal conductivity as well as other thermophysical properties across a variety of molten salt compositions. When coupled with targeted experimental studies, the ISM methodology can help contribute to existing databases necessary to safely implement molten salts into power generation systems.

Surface tension data for molten Al–Si alloys are important for modeling high-temperature melt-processing phenomena, but the surface tension values in the literature show considerable scatter, particularly with respect to their composition dependence. In this study, surface tension measurements were conducted for pure liquid Al, pure liquid Si, and molten Al–Si alloys containing 5–80 at % Si using the oscillating drop method with electromagnetic levitation. Oxide-affected heating curves were excluded from the pyrometer calibration, and only oscillation data for which no oxide film was observed on the droplet surface were analyzed. For all examined compositions, the measured surface tension decreased linearly with increasing temperature. Near 1700 K, the surface tension decreased monotonically and nearly linearly with increasing Si concentration, and previously reported nonlinear behaviors, such as relatively high values in the Al-rich region and an increase toward the Si-rich side, were not observed. Uncertainty analysis confirmed that the nearly linear composition dependence was not due to measurement uncertainty. Butler equation calculations were performed using the pure-component surface tension values obtained in this study. The ideal-solution approximation reproduced the measured composition dependence and temperature coefficients more consistently than the subregular-solution calculation using the evaluated excess Gibbs energies. These results indicate that thermodynamic nonideality makes only a limited contribution to the surface tension behavior of molten Al–Si alloys under the present experimental conditions.

We present a new wide-ranging correlation for the viscosity of normal helium (He I) based on critically evaluated experimental data as well as high-accuracy ab initio calculations. The correlation is designed to be used with densities from a high-accuracy Helmholtz equation of state, covering temperatures from 2.1768 K to 1500 K and pressures up to 2000 MPa. The estimated uncertainty at the 95 % confidence level depends on temperature and pressure, spanning from 0.002 % in the dilute gas limit (298.15 K, 0 MPa) to 10 % for the liquid phase at pressures from saturation up to 8 MPa. Finally, we provide comprehensive comparisons with experimental data and Python code to facilitate easy implementation.

Liquid organic hydrogen carrier (LOHC) systems are regarded as a promising option for the safe and efficient storage and transport of hydrogen (H₂). For the LOHC system based on γ-butyrolactone (GBL) and 1,4-butanediol (BDO) that is discussed as a biogenic and biodegradable alternative to fossil-based aromatic/alicyclic LOHC systems, reliable thermophysical property data like liquid dynamic viscosity η_L and interfacial tension σ are still scarce, particularly under process-relevant conditions. This work summarizes the results from recent research activities on the determination of η_L and σ for GBL, BDO, and their binary mixtures with GBL amount fractions of x_GBL = (0.25, 0.50, and 0.75) at temperatures T up to 473 K at a pressure of about 0.1 MPa. Surface light scattering (SLS) was applied for the simultaneous measurement of η_L and σ, complemented by capillary viscometry and pendant drop measurements, while polarization-difference Raman spectroscopy was employed during the SLS experiments to monitor the mixture composition. Observed changes of the latter at elevated T partially restricted a reliable property determination. Within the stable T range, σ decreases continuously with increasing T and increasing x_GBL. η_L decreases strongly with increasing x_GBL, presumably because of the breaking of the strong hydrogen bonds between BDO molecules.

Accurate values of the static dielectric constant of hydrogen are important for both fundamental thermophysical property calculations and emerging dielectric-based sensing applications in hydrogen liquefaction, storage, and processing systems. Previous work has shown that the existing reference correlations for hydrogen exhibit deficiencies at cryogenic conditions, particularly for ortho-rich hydrogen and liquid states. In this work, improved correlations are developed for the static dielectric constant of para- and ortho-hydrogen over temperatures from (1 to 1000) K. The low-density contribution was recalculated using modern ab initio rovibrational energy levels and state-dependent molecular polarizabilities. A new functional representation was developed that reproduces the temperature dependence of the first dielectric virial coefficient over the full temperature range considered, including the strongly nonlinear behavior at cryogenic temperatures. An extensive review of the available dielectric-constant measurements for normal and para-hydrogen was conducted, extending significantly upon the data used by previous dielectric correlations. These data were then used to develop revised correlations for the higher-density dielectric behavior of para- and ortho-hydrogen. The resulting correlations provide substantially improved agreement with available experimental data compared with previous models and better reproduce independent literature measurements that were not used in the fitting process, including recent measurements of the dielectric contrast between hydrogen spin isomers. However, the available experimental database remains limited, and the temperature dependences assumed for the higher-order terms remain only weakly constrained.

The solubility of chrysin in acetone–water binary mixtures (a polar protic–polar aprotic solvent system) was determined across the full acetone mole fraction range (0 to 1) and at temperatures of 293.2 K to 333.2 K. The solubility increases with temperature and shows a pronounced non-monotonic dependence on solvent composition, reaching a maximum at an acetone mole fraction of 0.80, characteristic of co-solvency behavior. A clear solvation transition was observed: acetone preferentially solvates chrysin at intermediate compositions, whereas water dominates in both water-rich and acetone-rich regions. The solubility of chrysin was correlated using 11 thermodynamic models, classified into empirical/semi-empirical models, local-composition activity coefficient models, and COSMO–RS-based quantum chemical models. The Modified Apelblat and van’t Hoff equations showed the best performance. Thermodynamic analysis based on the van’t Hoff equation indicates that the dissolution process is endothermic and non-spontaneous, governed primarily by enthalpy. Integrated molecular-level analysis based on COSMO-RS σ-profiles and ESP further elucidates that macroscopic solvation behavior across the entire composition range arises from microscopic solute–solvent and solvent–solvent interactions.

Yttria-stabilized zirconia (YSZ) is often used as a thermal barrier coating (TBC) material for gas turbines. When calcium-magnesium–aluminum-silicate (CMAS) deposits derived from sand, dust and volcanic ash accumulates on the TBC during operation, the molten CMAS wets and spreads, and reacts with the TBC leading to the formation of Y-depleted ZrO₂, as a result, compromising the thermal protection of the TBC. In this study, the contact angles of CMAS melts on YSZ and Al₂O₃ substrates were measured using the sessile drop method to observe the effects of interfacial reactions on the wetting and spreading behaviors of CMAS melts. To elucidate the thermodynamics of the formation of Y-depleted ZrO₂ in CMAS melts, the saturation solubility of ZrO₂ in four different CMAS melts was measured. Sessile drop measurements and electron probe microanalyzer observations revealed that the contact angles of CMAS melts on the YSZ substrates continuously decreased owing to the progress of interfacial reaction, leading to the formation of Y-depleted ZrO₂. On the other hand, the contact angles of CMAS melts on the Al₂O₃ substrates remained constant because anorthite covered the entire interface and the local equilibrium at the interface was reached. The ZrO₂ solubility measurements revealed that the reprecipitation of Y-depleted ZrO₂ is attributed to the reduction in the saturation solubility of ZrO₂ in the CMAS melts near the interface, resulting from a decrease in the CaO and MgO concentrations and an increase in the Y₂O₃ concentration in the CMAS melt owing to the CMAS/YSZ interfacial reaction.