Phase Diagrams of Alloys and Their Hydrides via On-Lattice Graph Neural Networks and Limited Training Data.

Efficient prediction of sampling-intensive thermodynamic properties is needed to evaluate material performance and permit high-throughput materials modeling for a diverse array of technology applications. To alleviate the prohibitive computational expense of high-throughput configurational sampling with density functional theory (DFT), surrogate modeling strategies like cluster expansion are many orders of magnitude more efficient but can be difficult to construct in systems with high compositional complexity. We therefore employ minimal-complexity graph neural network models that accurately predict and can even extrapolate to out-of-train distribution formation energies of DFT-relaxed structures from an ideal (unrelaxed) crystallographic representation. This enables the large-scale sampling necessary for various thermodynamic property predictions that may otherwise be intractable and can be achieved with small training data sets. Two exemplars, optimizing the thermodynamic stability of low-density high-entropy alloys and modulating the plateau pressure of hydrogen in metal alloys, demonstrate the power of this approach, which can be extended to a variety of materials discovery and modeling problems.

Patterned Adhesion Layer Enables Rugged Pd-MIS Hydrogen Sensors.

Rugged Pd-metal-insulator-semiconductor (Pd-MIS) hydrogen sensors for detecting charge-exchange particles in fusion reactors have been constructed by utilizing a novel patterned adhesion layer. Poor adhesion at the interface between Pd and SiO2 is a common failure mode for Pd-MIS devices, severely limiting the Pd thickness and their usefulness as hydrogen sensors. The mechanical integrity of the Pd coatings is of particular importance in magnetic fusion energy research where the Pd-MIS diodes are used to measure hydrogen charge-exchange neutral fluence at the wall in tokamaks. In this application, particularly thick Pd contacts are desirable to prevent damage caused by high-energy particles; however, such thick Pd coatings are prone to mechanical failure due to blistering and wire bond detachment during construction or operation. A continuous Ti or Cr adhesion layer is not possible for this application since it would interfere with H uptake at the SiO2 interface, which is essential for the device to generate a response. In this work, we demonstrate that a patterned Cr interlayer substantially improves adhesion while still providing access for hydrogen to reach the SiO2-Pd interface.

Ionicity and hydrogen affinity of water layers on metals.

The ability of metal surfaces to dissociate hydrogen molecules is key to many ways that metals react to their environment. Often, the barrier to dissociation is linked to the formation of surface hydrogen adatoms. Here, we show that hydrogen can be more strongly bound to water-covered surfaces in the form of hydronium ions than as adatoms. Density functional theory reveals that the hydronium binding is proportional to the surface electronic work function. For the case of Pt(111), a particularly high work function surface, the proton affinity of adsorbed water films can be 0.4 eV larger than that of the bare metal surface. This binding is large enough to make the water films susceptible to the formation of hydroxyl and hydronium ion pairs. We present evidence from scanning tunneling microscopy for the existence of hydronium ions in water films on Pt(111). This new insight into the stability of hydronium-containing water layers provides a basis for more realistic models of the chemical reactivity of water films on metals.

Real-time observation of epitaxial graphene domain reorientation.

Graphene films grown by vapour deposition tend to be polycrystalline due to the nucleation and growth of islands with different in-plane orientations. Here, using low-energy electron microscopy, we find that micron-sized graphene islands on Ir(111) rotate to a preferred orientation during thermal annealing. We observe three alignment mechanisms: the simultaneous growth of aligned domains and dissolution of rotated domains, that is, 'ripening'; domain boundary motion within islands; and continuous lattice rotation of entire domains. By measuring the relative growth velocity of domains during ripening, we estimate that the driving force for alignment is on the order of 0.1 meV per C atom and increases with rotation angle. A simple model of the orientation-dependent energy associated with the moiré corrugation of the graphene sheet due to local variations in the graphene-substrate interaction reproduces the results. This work suggests new strategies for improving the van der Waals epitaxy of 2D materials.

Clusters, molecular layers, and 3D crystals of water on Ni(111).

We examined the growth and stability of ice layers on Ni(111) up to ∼7 molecular layers (ML) thick using scanning tunneling microscopy. At low coverage, films were comprised of ∼1 nm wide two-dimensional (2D) clusters. Only above ∼0.5 ML did patches of continuous 2D layers emerge, coexisting with the clusters until the first ML was complete. The structure of the continuous layer is clearly different from that of the 2D clusters. Subsequently, a second molecular layer grew on top of the first. 3D crystallites started to form only after this 2nd ML was complete. 2D clusters re-appeared when thicker films were partially evaporated, implying that these clusters represent the equilibrium configuration at low coverage. Binding energies and image simulations computed with density functional theory suggest that the 2D clusters are partially dissociated and surrounded by H adatoms. The complete 2D layer contains only intact water molecules because of the lack of favorable binding sites for H atoms. We propose molecular structures for the 2D layer that are composed of the same pentagon-heptagon binding motif and water density observed on Pt(111). The similarity of the water structures on Pt and Ni suggests a general prescription for generating low-energy configurations on close-packed metal substrates.

Size-dependent silicon epitaxy at mesoscale dimensions.

New discoveries on collective processes in materials fabrication and performance are emerging in the mesoscopic size regime between the nanoscale, where atomistic effects dominate, and the macroscale, where bulk-like behavior rules. For semiconductor electronics and photonics, dimensional control of the architecture in this regime is the limiting factor for device performance. Epitaxial crystal growth is the major tool enabling simultaneous control of the dimensions and properties of such architectures. Although size-dependent effects have been studied for many small-scale systems, they have not been reported for the epitaxial growth of Si crystalline surfaces. Here, we show a strong dependence of epitaxial growth rates on size for nano to microscale radial wires and planar stripes. A model for this unexpected size-dependent vapor phase epitaxy behavior at small dimensions suggests that these effects are universal and result from an enhanced surface desorption of the silane (SiH4) growth precursor near facet edges. Introducing phosphorus or boron dopants during the silicon epitaxy further decreases the growth rates and, for phosphorus, gives rise to a critical layer thickness for single crystalline epitaxial growth. This previously unknown mesoscopic size-dependent growth effect at mesoscopic dimensions points to a new mechanism in vapor phase growth and promises greater control of advanced device geometries.

Current-induced transition from particle-by-particle to concurrent intercalation in phase-separating battery electrodes.

Many battery electrodes contain ensembles of nanoparticles that phase-separate on (de)intercalation. In such electrodes, the fraction of actively intercalating particles directly impacts cycle life: a vanishing population concentrates the current in a small number of particles, leading to current hotspots. Reports of the active particle population in the phase-separating electrode lithium iron phosphate (LiFePO4; LFP) vary widely, ranging from near 0% (particle-by-particle) to 100% (concurrent intercalation). Using synchrotron-based X-ray microscopy, we probed the individual state-of-charge for over 3,000 LFP particles. We observed that the active population depends strongly on the cycling current, exhibiting particle-by-particle-like behaviour at low rates and increasingly concurrent behaviour at high rates, consistent with our phase-field porous electrode simulations. Contrary to intuition, the current density, or current per active internal surface area, is nearly invariant with the global electrode cycling rate. Rather, the electrode accommodates higher current by increasing the active particle population. This behaviour results from thermodynamic transformation barriers in LFP, and such a phenomenon probably extends to other phase-separating battery materials. We propose that modifying the transformation barrier and exchange current density can increase the active population and thus the current homogeneity. This could introduce new paradigms to enhance the cycle life of phase-separating battery electrodes.

Insight into magnetite's redox catalysis from observing surface morphology during oxidation.

We study how the (100) surface of magnetite undergoes oxidation by monitoring its morphology during exposure to oxygen at ~650 °C. Low-energy electron microscopy reveals that magnetite's surface steps advance continuously. This growth of Fe3O4 crystal occurs by the formation of bulk Fe vacancies. Using Raman spectroscopy, we identify the sinks for these vacancies, inclusions of α-Fe2O3 (hematite). Since the surface remains magnetite during oxidation, it continues to dissociate oxygen readily. At steady state, over one-quarter of impinging oxygen molecules undergo dissociative adsorption and eventual incorporation into magnetite. From the independence of growth rate on local step density, we deduce that the first step of oxidation, dissociative oxygen adsorption, occurs uniformly over magnetite's terraces, not preferentially at its surface steps. Since we directly observe new magnetite forming when it incorporates oxygen, we suggest that catalytic redox cycles on magnetite involve growing and etching crystal.

Intercalation pathway in many-particle LiFePO4 electrode revealed by nanoscale state-of-charge mapping.

The intercalation pathway of lithium iron phosphate (LFP) in the positive electrode of a lithium-ion battery was probed at the ∼40 nm length scale using oxidation-state-sensitive X-ray microscopy. Combined with morphological observations of the same exact locations using transmission electron microscopy, we quantified the local state-of-charge of approximately 450 individual LFP particles over nearly the entire thickness of the porous electrode. With the electrode charged to 50% state-of-charge in 0.5 h, we observed that the overwhelming majority of particles were either almost completely delithiated or lithiated. Specifically, only ∼2% of individual particles were at an intermediate state-of-charge. From this small fraction of particles that were actively undergoing delithiation, we conclude that the time needed to charge a particle is ∼1/50 the time needed to charge the entire particle ensemble. Surprisingly, we observed a very weak correlation between the sequence of delithiation and the particle size, contrary to the common expectation that smaller particles delithiate before larger ones. Our quantitative results unambiguously confirm the mosaic (particle-by-particle) pathway of intercalation and suggest that the rate-limiting process of charging is initiating the phase transformation by, for example, a nucleation-like event. Therefore, strategies for further enhancing the performance of LFP electrodes should not focus on increasing the phase-boundary velocity but on the rate of phase-transformation initiation.

Growth from below: graphene bilayers on Ir(111).

We elucidate how graphene bilayers form on Ir(111). Low-energy electron diffraction (LEED) reveals that the two graphene layers are not always rotationally aligned. Monitoring this misalignment during growth shows that second-layer islands nucleate between the existing layer and the substrate. This mechanism occurs both when C segregates from the Ir and when elemental C is deposited from above. Low-energy electron microscopy (LEEM) and angle-resolved photoemission spectroscopy (ARPES) show that second-layer nucleation occurs preferentially under the first-layer rotational variants that are more weakly bound to the substrate. New-layer nucleation tends to occur inhomogeneously at substrate defects. Thus new-layer nucleation should be rapid on substrates that weakly bind graphene, making growth unstable toward mound formation initiated at substrate defects. In contrast, stronger binding permits layer-by-layer growth, as for Ru(0001). ARPES shows that bilayer graphene has two slightly p-doped π-bands. The work function of bilayer graphene is dominated by the orientation of the bottom layer.

Graphene Islands on Cu foils: the interplay between shape, orientation, and defects.

We have observed the growth of monolayer graphene on Cu foils using low-energy electron microscopy. On the (100)-textured surface of the foils, four-lobed, 4-fold-symmetric islands nucleate and grow. The graphene in each of the four lobes has a different crystallographic alignment with respect to the underlying Cu substrate. These "polycrystalline" islands arise from complex heterogeneous nucleation events at surface imperfections. The shape evolution of the lobes is well explained by an angularly dependent growth velocity. Well-ordered graphene forms only above ∼790 °C. Sublimation-induced motion of Cu steps during growth at this temperature creates a rough surface, where large Cu mounds form under the graphene islands. Strategies for improving the quality of monolayer graphene grown on Cu foils must address these fundamental defect-generating processes.

Towards better 3-D reconstructions by combining electron tomography and atom-probe tomography.

Scanning transmission electron microscope tomography and atom-probe tomography are both three-dimensional techniques on the nanoscale. We demonstrate here the combination of the techniques by analyzing the very same volume of an Al-Ag alloy specimen. This comparison allows us to directly visualize the theoretically known artifacts of each technique experimentally, providing insight into the optimal parameters to use for reconstructions and assessing the quality of each reconstruction. The combination of the techniques for accurate morphology and compositional information in three dimensions at the nanoscale provides a route for a new level of materials characterization and understanding.

Nucleation-limited dewetting of ice films on Pt(111).

Quantifying dewetting phenomena at the microscopic level is the key to deciphering how a balance between kinetic and equilibrium effects determines ice-film morphology on Pt(111). Overcoming the difficulty of imaging nominally insulating ice multilayers with scanning tunneling microscopy allowed us to track the dewetting process. The results show that the rate at which new layers nucleate, and not surface diffusion, determines how fast individual crystallite shapes equilibrate. Applying nucleation theory to measured growth rates versus crystallite size, we obtain new bounds on the energetics both of step formation on ice and of the Pt-ice interface.

Surface self-organization caused by dislocation networks.

We report a new mechanism of self-organization that can lead to robust surface ordering. We have quantitatively analyzed the thermal motion of holes created by sulfur atoms in a silver monolayer on a ruthenium surface, which we observed in real time with scanning tunneling microscopy. We find that the stability of the array of holes is determined by the arrangement and structure of misfit dislocations in the film.