Ahead by a Century: Discovery of Laves Phases Assisted by Machine Learning.

Laves phases AB2 form the most abundant group of intermetallic compounds, consisting of combinations of larger electropositive metals A with smaller metals B. Many practical applications of Laves phases depend on the ability to tune their physical properties through appropriate substitution of either the A or B component. Although simple geometrical and electronic factors have long been thought to control the formation of Laves phases, no single factor alone can make predictions accurately. Several machine learning models have been developed to discover new Laves phases, including variations caused by solid solubility, using elemental properties solely on the basis of chemical composition. These models were trained on a data set comprising about 3700 entries of experimentally known phases AB2 with Laves and non-Laves structures. Among these models, a decision tree algorithm gave very good performance (average recall of 95%, precision of 94%, and accuracy of 96% on the test set) by using only a small set of descriptors, the most important of which relates to the electron density at the boundary of the Wigner-Seitz cell for the B component. This model provides guidance for new experiments by making predictions on >400000 candidates very quickly. A chemically unintuitive candidate Cd(Cu1-xSbx)2 with a limited solid solubility of Sb for Cu was targeted; it was successfully synthesized and confirmed to adopt a cubic MgCu2-type Laves structure.

Structure and Optical Properties of LixAg1-xGaSe2 and LixAg1-xInSe2.

Complete substitution of Li atoms for Ag atoms in AgGaSe2 and AgInSe2 was achieved, resulting in the solid solutions LixAg1-xGaSe2 and LixAg1-xInSe2. The detailed crystal structures were determined by single-crystal X-ray diffraction and solid-state 7Li nuclear magnetic resonance spectroscopy, which confirm that Li atoms occupy unique sites and disorder only with Ag atoms. The tetragonal CuFeS2-type structure (space group I4̅2d) was retained within the entirety of the Ga-containing solid solution LixAg1-xGaSe2, which is noteworthy because the end-member LiGaSe2 normally adopts the orthorhombic ß-NaFeO2-type structure (space group Pna21). These structures are closely related, being superstructures of the cubic sphalerite and hexagonal wurtzite prototypes adopted by diamond-like semiconductors. For the In-containing solid solution LixAg1-xInSe2, the structure transforms from the tetragonal to orthorhombic forms as the Li content increases past x = 0.50. The optical band gaps increase gradually with higher Li content, from 1.8 to 3.4 eV in LixAg1-xGaSe2 and from 1.2 to 2.5 eV in LixAg1-xInSe2, enabling control to desired values, while the second harmonic generation responses become stronger or are similar to those of benchmark infrared nonlinear optical materials such as AgGaS2. All members of these solid solutions remain congruently melting at accessible temperatures between 800 and 900 °C. Electronic structure calculations support the linear trends seen in the optical band gaps and confirm the mostly ionic character present in Li-Se bonds, in contrast to the more covalent character in Ga-Se or In-Se bonds.