Controllably incorporating multiple immiscible metal elements into one single nanostructure has immeasurable technological and scientific potential, but it remains a challenge for the conventional bottom-up synthetic methods. Herein, we presented a general and scalable route to prepare multi-component nanostructured alloys referred to as nanoporous high-entropy alloys (np-HEAs) by combining bulk melting, fast cooling, and dealloying. To demonstrate this concept, we synthesized senary AlNiCuPtPdAu, octonary AlNiCuPtPdAuCoFe, and senary all-non-noble metal AlNiCuMoCoFe np-HEA with ligament sizes of ∼2–3 nm and precisely controlled composition by dealloying the designed precursor alloys. With a naturally formed thin oxide layer of spinel γ-Al₂O₃, AlNiCuPtPdAu np-HEA exhibited greatly enhanced high-temperature stability (up to 600 °C) and CO oxidation activity. Interestingly, with the removal of the surface oxide layer, np-HEA still showed good resistance to coarsening at 200 °C for 10 h due to its intrinsically low diffusivity originating from the multiple-principal-element mixing effect. For electrocatalysis, np-HEA with a low Pt loading amount exhibited 10 times the mass activity of Pt/C for oxygen reduction reaction and maintained 92.5% of its initial activity after 100k electrochemical cycles.

Thermoelectric materials can be potentially applied to waste heat recovery and solid-state cooling because they allow a direct energy conversion between heat and electricity and vice versa. The accelerated materials design based on machine learning has enabled the systematic discovery of promising materials. Herein we proposed a successful strategy to discover and design a series of promising half-Heusler thermoelectric materials through the iterative combination of unsupervised machine learning with the labeled known half-Heusler thermoelectric materials. Subsequently, optimized zT values of ~0.5 at 925 K for p-type Sc_0.7Y_0.3NiSb_0.97Sn_0.03 and ~0.3 at 778 K for n-type Sc_0.65Y_0.3Ti_0.05NiSb were experimentally achieved on the same parent ScNiSb.

Using first-principles calculations accompanied by the transition state theory and an 8-frequency model, we present a comprehensive investigation of the diffusion coefficients of substitutional alloying elements X in dilute α-Ti alloys, where X denotes Al, V, Nb, Ta, Mo, Zr, and Sn. The alloying elements Mo and Al exhibit a maximum and a minimum diffusion rate in dilute α-Ti alloys, respectively. It is found that the nearest-neighbor solute–vacancy binding energies and activation energies are roughly inversely proportional to the volume changes induced by solute atoms. There are two exceptions to this trend: Al and Mo. Besides the physical effect (i.e., solute size), two other key factors governing solute diffusion in dilute α-Ti are clarified: the chemical bonding characteristics and vibrational features of X–Ti pairs. It verifies that the ultrafast diffusivity of Mo arises from the interactions with Ti atoms by metallic bonds and its low-frequency contributions to lattice vibration, while the more covalent bonding nature and the high-frequency contributions to the lattice vibration of Al lead to its ultraslow diffusivity. In addition, the correlation effects of diffusion coefficients are non-negligible for the large solutes Ta, Nb, and Zr, in which the direct solute–vacancy migration barriers are much smaller than the solvent–vacancy migration barriers.