Takehiko Ishikawa

Abstract Miscibility gap alloys (MGAs) are promising candidates for high‑temperature thermal energy storage owing to their high latent heat and intrinsic phase separation. In this study, the liquid–liquid phase separation and subsequent solidification of Fe–Cu alloys were experimentally investigated using an aerodynamic levitator in a reducing atmosphere to suppress oxidation. In situ observations using a high-speed camera revealed that Fe‑rich liquid domains separated first from the undercooled homogeneous liquid, followed by the formation of Cu‑rich liquid domains. These observations are consistent with the asymmetry of the Gibbs free energy of mixing in liquid Fe–Cu alloys. The energy densities of these alloys exceeded the upper range of IRENA’s 2050 target (50–85 kWh m⁻ ³ ) for high-temperature latent-heat storage at Cu concentrations above 40 at. % (Fe60Cu40 and higher), indicating the potential of Fe–Cu alloys as high‑temperature latent heat storage materials. Our results provide insights into the role of microstructural control and, together with favorable thermal properties, offer a promising strategy for the design of MGA‑based thermal energy storage materials produced by casting.

Abstract The La₂O₃–Nb₂O₅ binary system is a unique glass-forming system without conventional network former oxides, exhibiting two distinct glass-forming regions: La₂O₃-rich and Nb₂O₅-rich compositions. To evaluate its glass-forming ability, the temperature dependence of density, viscosity, and surface tension was measured using the electrostatic levitation furnace aboard the International Space Station (ISS–ELF). Melt density showed linear temperature dependence, and thermal expansion coefficients at 2000 K varied from 2.5 × 10⁻⁵ to 4.0 × 10⁻⁵ K⁻¹. Substantial undercooling was observed for glass-forming compositions. Viscosity measurements above the melting point revealed that both La₂O₃-rich and Nb₂O₅-rich melts behave as fragile liquids. Activation energy derived from viscosity data was higher for glass-forming compositions. These results suggest that glass-forming ability can be assessed based on undercooling and activation energy across a wide compositional range, including non-glass-forming melts. The ISS–ELF experiments provide a valuable platform for understanding glass formation in systems inaccessible by terrestrial techniques.

Rare earth aluminum garnets are important materials in optical, dielectric, and thermal barrier applications. To advance the understanding of their melt processing and glass forming ability, we report the atomic structure of molten Yb3Al5O12 over 1770–2630 K, which spans the equilibrium and supercooled liquid regimes. The melt density at Tm = 2283 K is 5.50 g cm−3, measured via silhouette imaging of electrostatically levitated drops over 1010–2420 K. Four separate structure measurements were made with aerodynamically levitated melts using x-ray and neutron diffraction with isotope substitution of Yb (172Yb, 174Yb, or natYb). Empirical potential structure refinement models were developed, which are in excellent agreement with the experiments. Coordination environments for Al–O are predominantly 4- and 5-coordinate, with a mean coordination of nAlO = 4.43(8), while Yb–O environments mostly range from 5- to 8-coordinate, with nYbO = 6.26(8). The cation–oxygen polyhedra are connected primarily by corner-sharing, with edge-sharing constituting up to ∼1/3 of the connectivity among polyhedra with Yb or higher-coordinated Al–O. Structurally, the –Al–O– network in molten Yb3Al5O12 appears conducive to glass formation: nOAl = 1.85(3), there are 1.86 AlOx–AlOx connections per Al atom (e.g., a mixture of Q3 and Q4 units), and the modal ring size is six cations. These characterize a network that is somewhat less constrained compared to SiO2 glass, yet Yb3Al5O12 cannot be quenched into crystal-free glass. Aluminum garnet compositions with larger rare earth cations do form glass, so these characterizations help reveal the structural characteristics corresponding to the limit of glass forming ability in rare earth aluminates.

Abstract Possible existence of dense iron-rich silicate melt layer above Mars’ core is important in understanding the nature and evolution of Mars. However, gravitational stability of iron-rich silicate melt in the Mars’ interior has not been well constrained, due to experimental difficulties in measuring density of iron-rich peridotitic melt. Here we report density measurements of iron-rich peridotitic melts up to 2465 K by using electrostatic levitation furnace at the International Space Station. Our experimentally obtained densities of iron-rich peridotitic melts are markedly higher than those calculated by first principles simulation, and are distinct from those estimated by extrapolating a density model for SiO₂-rich basaltic melts. Our determined density model suggests that peridotitic melt with the Fe/(Mg+Fe) ratio more than 0.4-0.5 has higher density than that at the base of the Mars’ mantle, which indicates gravitational stability of the iron-rich peridotitic melt at the core-mantle boundary in Mars.