Machine learning-guided exploration and experimental assessment of unreported compositions in the quaternary Ti-Zr-Cu-Pd biocompatible metallic glass system.

Due to their outstanding elastic limit, biocompatible Ti-based bulk metallic glasses (BMGs) are candidate materials to decrease the size of medical implants and therefore reduce their invasiveness. However, the practical use of classical Ti-BMGs in medical applications is in part hindered by their high copper content: more effort is thus required to design low-copper Ti-BMGs. In this work, in line with current rise in AI-driven tools, machine learning (ML) approaches, a neural-network ML model is used to explore the glass-forming ability (GFA) of unreported low-copper compositions within the biocompatible Ti-Zr-Cu-Pd system. Two types of models are trained and compared: one based on the alloy composition only, and a second based on various features derived from the alloying elements. Contrary to expectation, the predictive power of both models in evaluating GFA is similar. The compositional space identified by ML as promising is experimentally assessed, finding unfortunately low GFA. These results indicate that the ML approach may be premature for specific composition tuning of amorphous metallic materials. We emphasise that the development of ML tools in GFA prediction requires an improvement of the dataset, in terms of homogeneity, size and GFA descriptors, which must be supported by increased reporting of high-quality experimental GFA measurements, both positive and negative. STATEMENT OF SIGNIFICANCE: Biocompatible Ti-based bulk metallic glasses (BMGs) are candidate materials for use in the next generation of minimally invasive dental implants where improved mechanical properties, such as high strength are required. Despite promising in vitro/vivo evaluations, implementation of alloys for practical applications is partly hindered by the presence of copper as the main alloying element. Recent studies have presented AI-guided and machine learning strategies as appealing approaches to understand and describe the glass forming ability (GFA) of BMG-forming compositions. In this work, we employ and evaluate the capacity of a machine-learning model to explore low-copper compositional spaces in the biocompatible Ti-Zr-Cu-Pd system. Our results highlight the limits of such a computational approach and suggest improvements for future designing routes.

Hierarchical Surface Pattern on Ni-Free Ti-Based Bulk Metallic Glass to Control Cell Interactions.

Ni-free Ti-based bulk metallic glasses (BMGs) are exciting materials for biomedical applications because of their outstanding biocompatibility and advantageous mechanical properties. The glassy nature of BMGs allows them to be shaped and patterned via thermoplastic forming (TPF). This work demonstrates the versatility of the TPF technique to create micro- and nano-patterns and hierarchical structures on Ti40 Zr10 Cu34 Pd14 Sn2 BMG. Particularly, a hierarchical structure fabricated by a two-step TPF process integrates 400 nm hexagonal close-packed protrusions on 2.5 µm square protuberances while preserving the advantageous mechanical properties from the as-cast material state. The correlations between thermal history, structure, and mechanical properties are explored. Regarding biocompatibility, Ti40 Zr10 Cu34 Pd14 Sn2 BMGs with four surface topographies (flat, micro-patterned, nano-patterned, and hierarchical-structured surfaces) are investigated using Saos-2 cell lines. Alamar Blue assay and live/dead analysis show that all tested surfaces have good cell proliferation and viability. Patterned surfaces are observed to promote the formation of longer filopodia on the edge of the cytoskeleton, leading to star-shaped and dendritic cell morphologies compared with the flat surface. In addition to potential implant applications, TPF-patterned Ti-BMGs enable a high level of order and design flexibility on the surface topography, expanding the available toolbox for studying cell behavior on rigid and ordered surfaces.

Evolutionary design of machine-learning-predicted bulk metallic glasses.

The size of composition space means even coarse grid-based searches for interesting alloys are infeasible unless heavily constrained, which requires prior knowledge and reduces the possibility of making novel discoveries. Genetic algorithms provide a practical alternative to brute-force searching, by rapidly homing in on fruitful regions and discarding others. Here, we apply the genetic operators of competition, recombination, and mutation to a population of trial alloy compositions, with the goal of evolving towards candidates with excellent glass-forming ability, as predicted by an ensemble neural-network model. Optimization focuses on the maximum casting diameter of a fully glassy rod, D max, the width of the supercooled region, ΔT x, and the price-per-kilogramme, to identify commercially viable novel glass-formers. The genetic algorithm is also applied with specific constraints, to identify novel aluminium-based and copper-zirconium-based glass-forming alloys, and to optimize existing zirconium-based alloys.

Direct Formation of Hard-Magnetic Tetrataenite in Bulk Alloy Castings.

Currently, predominant high-performance permanent magnets contain rare-earth elements. In the search for rare-earth-free alternates, body-centered tetragonal Fe-Ni is notable. The ordering to form this phase from the usual cubic close-packed Fe-Ni is understood to be possible only below a critical temperature, commonly accepted to be 593 K. The ordering is first demonstrated by using neutron irradiation to accelerate atomic diffusion. The tetragonal phase, designated as the mineral tetrataenite, is found in Fe-based meteorites, its formation attributed to ultra-slow cooling. Despite many attempts with diverse approaches, bulk synthesis of tetrataenite has not been reported. Here it is shown that with appropriate alloy compositions, bulk synthesis of tetrataenite is possible, even in conventional casting at cooling rates 11-15 orders of magnitude higher than in meteorites. The barrier to obtaining tetrataenite (slow ordering from cubic close-packed to body-centered tetragonal) is circumvented, opening a processing window for potential rare-earth-free permanent magnets. The formation of tetrataenite on industrially practicable timescales also throws into question the interpretation of its formation in meteorites and their associated cooling rates.

Machine-learning improves understanding of glass formation in metallic systems.

Glass-forming ability (GFA) in metallic systems remains a little-understood property. Experimental work on bulk metallic glasses (BMGs) is guided by many empirical criteria, which are often of limited predictive value. This work uses machine-learning both to produce predictive models for the GFA of alloy compositions, and to reveal insights useful for furthering theoretical understanding of GFA. Our machine-learning models apply a novel neural-network architecture to predict simultaneously the liquidus temperature, glass-transition temperature, crystallization-onset temperature, maximum glassy casting diameter, and probability of glass formation, for any given alloy. Feature permutation is used to identify the features of importance in the black-box neural network, recovering Inoue's empirical rules, and highlighting the effect of discontinuous Wigner-Seitz boundary electron-densities on atomic radii. With certain combinations of elements, atomic radii of different species contract and expand to balance electron-density discontinuities such that the overall difference in atomic radii increases, improving GFA. We calculate adjusted radii via the Thomas-Fermi model and use this insight to propose promising novel glass-forming alloy systems.

Enhancement of Interfacial Hydrogen Interactions with Nanoporous Gold-Containing Metallic Glass.

Contrary to the electrochemical energy storage in Pd nanofilms challenged by diffusion limitations, extensive metal-hydrogen interactions in Pd-based metallic glasses result from their grain-free structure and presence of free volume. This contribution investigates the kinetics of hydrogen-metal interactions in gold-containing Pd-based metallic glass (MG) and crystalline Pd nanofilms for two different pore architectures and nonporous substrates. Fully amorphous MGs obtained by physical vapor deposition (PVD) co-sputtering are electrochemically hydrogenated by chronoamperometry. High-resolution (scanning) transmission electron microscopy and corresponding energy-dispersive X-ray analysis after hydrogenation corroborate the existence of several nanometer-sized crystals homogeneously dispersed throughout the matrix. These nanocrystals are induced by PdHx formation, which was confirmed by depth-resolved X-ray photoelectron spectroscopy, indicating an oxide-free inner layer of the nanofilm. With a larger pore diameter and spacing in the substrate (Pore40), the MG attains a frequency-independent impedance at low frequencies (∼500 Hz) with very high Bode magnitude stability accounting for enhanced ionic diffusion. On the contrary, on a substrate with a smaller pore diameter and spacing (Pore25), the MG shows a larger low-frequency (0.1 Hz) capacitance, linked to enhanced ionic transfer in the near-DC region. Hence, the nanoporosity of amorphous and crystalline metallic materials can be systematically adjusted depending on AC- and DC-type applications.

In situ correlation between metastable phase-transformation mechanism and kinetics in a metallic glass.

A combination of complementary high-energy X-ray diffraction, containerless solidification during electromagnetic levitation and transmission electron microscopy is used to map in situ the phase evolution in a prototype Cu-Zr-Al glass during flash-annealing imposed at a rate ranging from 102 to 103 K s-1 and during cooling from the liquid state. Such a combination of experimental techniques provides hitherto inaccessible insight into the phase-transformation mechanism and its kinetics with high temporal resolution over the entire temperature range of the existence of the supercooled liquid. On flash-annealing, most of the formed phases represent transient (metastable) states - they crystallographically conform to their equilibrium phases but the compositions, revealed by atom probe tomography, are different. It is only the B2 CuZr phase which is represented by its equilibrium composition, and its growth is facilitated by a kinetic mechanism of Al partitioning; Al-rich precipitates of less than 10 nm in a diameter are revealed. In this work, the kinetic and chemical conditions of the high propensity of the glass for the B2 phase formation are formulated, and the multi-technique approach can be applied to map phase transformations in other metallic-glass-forming systems.

Effective electrocatalytic methanol oxidation of Pd-based metallic glass nanofilms.

Compared to their conventional polycrystalline Pd counterparts, Pd79Au9Si12 (at%) - metallic glass (MG) nanofilm (NF) electrocatalysts offer better methanol oxidation reaction (MOR) in alkaline medium, CO poisoning tolerance and catalyst stability even at high scan rates or high methanol concentrations owing to their amorphous structure without grain boundaries. This study evaluates the influence of scan rate and methanol concentration by cyclic voltammetry, frequency-dependent electrochemical impedance spectroscopy and a related equivalent circuit model at different potentials in Pd-Au-Si amorphous NFs. Structural and compositional differences for the NFs are assessed by high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM), energy dispersive X-ray (EDX) mapping and X-ray diffraction (XRD). The ratio of the forward to reverse peak current density ipf/ipb for the MG NFs is ∼2.2 times higher than for polycrystalline Pd NFs, evidencing better oxidation of methanol to carbon dioxide in the forward scan and less poisoning of the electrocatalysts by carbonaceous (e.g. CO, HCO) species. Moreover, the electrochemical circuit model obtained from EIS measurements reveals that the MOR occurring around -100 mV increases the capacitance without any significant change in oxidation resistance, whereas CO2 formation towards lower potentials results in a sharp increase in the capacitance of the Faradaic MOR at the catalyst interface and a slight decrease in the corresponding resistance. These results, together with the high ipf/ipb = 3.37 yielding the minimum amount of carbonaceous species deposited on the thin film during cyclic voltammetry and stability in the alkaline environment, can potentially make these amorphous thin films potential candidates for fuel-cell applications.

Metallic Glass Films with Nanostructured Periodic Density Fluctuations Supported on Si/SiO2 as an Efficient Hydrogen Sorber.

Nanostructured metallic glass films (NMGF) can exhibit surface and intrinsic effects that give rise to unique physical and chemical properties. Here, a facile synthesis and electrochemical, structural, and morphologic characterization of Pd-Au-Si based MGs of approximately 50 nm thickness supported on Si/SiO2 is reported. Impressively, the maximum total hydrogen charge stored in the Pd-Au-Si nanofilm is equal to that in polycrystalline Pd films with 1 µm thickness in 0.1 m H2 SO4 electrolyte. The same NMGF has a volumetric desorption charge that is more than eight times and 25 % higher than that of polycrystalline PdNF and Pd-Cu-Si NMGF with the same thickness supported on Si/SiO2 , respectively. A significant number of nanovoids originating from PdHx crystals, and an increase in the average interatomic spacing is detected in Pd-Au-Si NMGF by high-resolution TEM. Such a high amount of hydrogen sorption is linked to the unique density fluctuations without any chemical segregation exclusively observed for this NMGF.

Local microstructure evolution at shear bands in metallic glasses with nanoscale phase separation.

At room temperature, plastic flow of metallic glasses (MGs) is sharply localized in shear bands, which are a key feature of the plastic deformation in MGs. Despite their clear importance and decades of study, the conditions for formation of shear bands, their structural evolution and multiplication mechanism are still under debate. In this work, we investigate the local conditions at shear bands in new phase-separated bulk MGs containing glassy nanospheres and exhibiting exceptional plasticity under compression. It is found that the glassy nanospheres within the shear band dissolve through mechanical mixing driven by the sharp strain localization there, while those nearby in the matrix coarsen by Ostwald ripening due to the increased atomic mobility. The experimental evidence demonstrates that there exists an affected zone around the shear band. This zone may arise from low-strain plastic deformation in the matrix between the bands. These results suggest that measured property changes originate not only from the shear bands themselves, but also from the affected zones in the adjacent matrix. This work sheds light on direct visualization of deformation-related effects, in particular increased atomic mobility, in the region around shear bands.

Atomic and vibrational origins of mechanical toughness in bioactive cement during setting.

Bioactive glass ionomer cements (GICs) have been in widespread use for ∼40 years in dentistry and medicine. However, these composites fall short of the toughness needed for permanent implants. Significant impediment to improvement has been the requisite use of conventional destructive mechanical testing, which is necessarily retrospective. Here we show quantitatively, through the novel use of calorimetry, terahertz (THz) spectroscopy and neutron scattering, how GIC's developing fracture toughness during setting is related to interfacial THz dynamics, changing atomic cohesion and fluctuating interfacial configurations. Contrary to convention, we find setting is non-monotonic, characterized by abrupt features not previously detected, including a glass-polymer coupling point, an early setting point, where decreasing toughness unexpectedly recovers, followed by stress-induced weakening of interfaces. Subsequently, toughness declines asymptotically to long-term fracture test values. We expect the insight afforded by these in situ non-destructive techniques will assist in raising understanding of the setting mechanisms and associated dynamics of cementitious materials.

Sub-micrometer soft lithography of a bulk chalcogenide glass.

We demonstrate, for the first time, time- and cost-effective replication of sub-micrometer features from a soft PDMS mold onto a bulk chalcogenide glass over a large surface area. A periodic array of sub-micrometer lines (diffraction grating) with period 625 nm, amplitude 45 nm and surface roughness 3 nm was imprinted onto the surface of the chalcogenide AsSe(2) bulk glass at temperature 225°C, i.e. 5°C below the softening point of the glass. Sub-micrometer soft lithography into chalcogenide bulk glasses shows good reliability, reproducibility and promise for feasible fabrication of various dispersive optical elements, anti-reflection surfaces, 2D photonic structures and nano-structured surfaces for enhanced photonic properties and chemical sensing.