Search

We use a random forest model to predict the critical cooling rate (RC) for glass formation of various alloys from features of their constituent elements. The random forest model was trained on a database that integrates multiple sources of direct and indirect RC data for metallic glasses to expand the directly measured RC database of less than 100 values to a training set of over 2,000 values. The model error on 5-fold cross validation is 0.66 orders of magnitude in K/s. The error on leave out one group cross validation on alloy system groups is 0.59 log units in K/s when the target alloy constituents appear more than 500 times in training data. Using this model, we make predictions for the set of compositions with melt-spun glasses in the database, and for the full set of quaternary alloys that have constituents which appear more than 500 times in training data. These predictions identify a number of potential new bulk metallic glass (BMG) systems for future study, but the model is most useful for identification of alloy systems likely to contain good glass formers, rather than detailed discovery of bulk glass composition regions within known glassy systems.

Recent discoveries of materials with ultra-low thermal conductivity open a pathway to significant developments in the field of thermoelectricity. Here, we conduct a comparative study of three chemically similar antimonides to establish the root causes of their extraordinarily low thermal conductivity ($0.4-0.6$ Wm$^{-1}$K$^{-1}$ at 525 K). The materials of interest are: the unconventional type-XI clathrate K$_{58}$Zn$_{122}$Sb$_{207}$, the tunnel compound K$_{6.9}$Zn$_{21}$Sb$_{16}$, and the type-I clathrate K$_8$Zn$_{15.5}$Cu$_{2.5}$Sb$_{28}$ discovered herein. Calculations of the phonon dispersions show that the type-XI compound exhibits localized (i.e., rattling) phonon modes with unusually low frequencies that span the entire acoustic regime. In contrast, rattling in the type-I clathrate is observed only at higher frequencies, and no rattling modes are present in the tunnel structure. Modeling reveals that low-frequency rattling modes profoundly limit the acoustic scattering time; the scattering time of the type-XI clathrate is half that of the type-I clathrate and a quarter of the tunnel compound. For all three materials, the thermal conductivities are additionally suppressed by soft framework bonding that lowers the acoustic group velocities, and structural complexity that leads to diffusonic character of the optical modes. Understanding details of thermal transport in structurally complex materials will be crucial for developing the next generation of thermoelectrics.

Corkite, PbFe3(PO4)(SO4)(OH)6, an understudied relative of the jarosite family of Heisenberg antiferromagnets, has been synthesized and its magnetic properties characterized for the first time. Relative to natural samples, synthetic corkite displays signatures in both infrared and Raman spectra of a more ordered arrangement of polyanion groups about the kagom\'e sublattice that retains inversion symmetry. Magnetic susceptibility measurements reveal that dried corkite undergoes a transition to a long-range, antiferromagnetically-ordered state below TN = 48 K, lower than that observed in the majority of jarosite phases, and indicative of further spin frustration. Curie-Weiss fitting of the measured magnetic susceptibility yields an effective magnetic moment of peff = 6.29(1) muB/Fe^3+ and theta_CW = -526.0(1.1) K, analogous to that observed in similar high-spin Fe^3+ systems, and indicative of strong antiferromagnetic coupling. Estimation of the change in magnetic entropy as a function of temperature from T = 0 K to T = 195 K, dS_mag = 14.86 J/mol_Fe^3+ K, is also in good agreement with the dS_mag = Rln(2S+1) = 14.9 J/mol K expected for a S = 5/2 system. In comparison to the pure jarosites, where both structure and magnetism remain largely invariant upon a variety of chemical substitutions, the replacement of one sulfate group per formula unit with a higher-valent phosphate group applies additional steric and electronic pressure on the kagom\'e lattice in corkite, further frustrating the magnetic ground state of the material. Corkite thus represents both an outlier in the known body of jarosite-type materials, and an illustration of how existing structures may be further strained in the development of highly frustrated magnetic systems., Comment: 27 pages, 5 figures, 4 tables

We have searched nearly 40,000 inorganic solids in the Inorganic Crystal Structural Database to identify compounds containing a transition metal or rare earth kagom\'e sublattice, a geometrically magnetically frustrated lattice, ultimately identifying $\sim$500 materials. A broad analysis of the chemical and structural trends of these materials shows three types of kagom\'e sheet stacking and several classes of magnetic complexity. Following the search and classification, we rapidly screen the magnetic properties of a subset of the materials using density functional theory to eliminate those that are unlikely to exhibit magnetic frustration. From the results of our computational screening, we rediscover six materials that have previously been explored for their low temperature magnetic behavior, albeit showing symmetry breaking distortions, spin glass behavior, or magnetic ordering. However, all are materials with antiferromagnetic behavior, which we correctly predict. Finally, we also report three materials that appear to be unexplored for their magnetic properties.