Understanding hydrogen incorporation into palladium requires detailed knowledge of surface and subsurface structure and atomic interactions as surface hydrogen is being embedded. Using density functional theory (DFT), we examine the energies of hydrogen layers of varying coverage adsorbed on Pd(111). We find that H-H and H-Pd interactions promote the formation of the well-known 3×3 phases but also favor an unreported (3 × 3) phase at high H coverages for which we present experimental evidence. We relate the stability of isolated H vacancies of the (3 × 3) phase to the need of H2 molecules to access bare Pd before they can dissociate. Following higher hydrogen dosage, we observe initial steps of hydride formation, starting with small clusters of subsurface hydrogen. The interaction between H and Pd is complicated by the persistent presence of carbon at the surface. X-ray photoelectron spectroscopy experiments show that trace amounts of carbon, emerging from the Pd bulk despite many surface cleaning cycles, become mobile enough to repopulate the C-depleted surface at temperatures above 200 K. When exposed to hydrogen, these surface carbon atoms react to form benzene, as evidenced by scanning tunneling microscopy observations interpreted with DFT.
We study the structure of the threading edge dislocations, or "elbows,"which are an essential component of the well-known herringbone reconstruction of the (111) surface of Au. Previous work had shown that these dislocations can be stabilized by long-range elastic relaxations into the bulk. However, the validity of the harmonic spring model that had been used to estimate the energies of the dislocations is uncertain. To enable a more refined model of the dislocation energetics, we have imaged the atomic structure of these dislocations using scanning tunneling microscopy. We find that the harmonic spring model does not adequately reproduce the observed structure. We are able to reproduce the structure, however, with a two-dimensional Frenkel-Kontorova (FK) model that uses a pairwise Morse potential to describe the interactions between the top layer Au atoms on a rigid substrate. The parameters of the potential were obtained by fitting the energy of uniaxially compressed phases, or "stripes", computed with density functional theory, as a function of surface Au density. Within this model, the formation of the threading dislocations remains unfavorable. However, the large forces on the substrate atoms near the threading-dislocation cores, render the assumption of a completely rigid substrate questionable. Indeed, if the FK parameters are modified to account for the relaxation of just one more atomic layer, threading dislocations can, in principle, become favorable, even without bulk elastic relaxations. Additional evidence for a small elbow energy is that our computed change in the Au(111) surface stress tensor caused by the (3×22) reconstruction is considerably smaller than previous estimates.
We successfully demonstrated the utility of surface science techniques - namely scanning probe microscopy and thermal desorption spectroscopy - on three different material systems: incipient soot formed during fossil fuel combustion, surface oxides passivating polycrystalline nickel hydrogen uptake, and aluminum hydride cluster formation underpinning solid-state hydrogen fuel storage. For all three material systems, surface science techniques haven proven to probe intricate nanoscale phenomena that are critical to macroscale material behavior. This LDRD has gained insight into early-stage pollution formation, the impacts of common contaminants on tritium flow regulation, and the limitations of solid-state hydrogen fuel storage. Our results support the diversification of national energy technologies.
The passivation of polycrystalline nickel surfaces against hydrogen uptake by oxygen is investigated experimentally with low energy ion scattering (LEIS), direct recoil spectroscopy (DRS), and thermal desorption spectroscopy (TDS). These techniques are highly sensitive to surface hydrogen, allowing the change in hydrogen adsorption in response to varying amounts of oxygen exposure to be measured. The chemical composition of a nickel surface during a mixed oxygen and hydrogen exposure was characterized with LEIS and DRS, while the uptake and activation energies of hydrogen on a nickel surface with preadsorbed oxygen were quantified with TDS. By and large, these measurements of how the oxygen and hydrogen surface coverage varied in response to oxygen exposure were found to be consistent with predictions of a simple site-blocking model. This finding suggests that, despite the complexities that arise due to polycrystallinity, the oxygen-induced passivation of a polycrystalline nickel surface against hydrogen uptake can be approximated by a simple site-blocking model.
Metal oxides have been an attractive option for a range of applications, including hydrogen sensors, microelectronics, and catalysis, due to their reactivity and tunability. The properties of metal oxides can vary greatly on their precise surface structure; however, few surface science techniques can achieve atomistic-level determinations of surface structure, and fewer yet can do so for insulator surfaces. Low energy ion beam analysis offers a potential insulator-compatible solution to characterizing the surface structure of metal oxides. As a feasibility study, we apply low energy ion beam analysis to investigate the surface structure of a magnetite single crystal, Fe3O4(100). We obtain multi-angle maps using both forward-scattering low energy ion scattering (LEIS) and backscattering impact-collision ion scattering spectroscopy (ICISS). Both sets of experimental maps have intensity patterns that reflect the symmetries of the Fe3O4(100) surface structure. However, analytical interpretation of these intensity patterns to extract details of the surface structure is significantly more complex than previous LEIS and ICISS structural studies of one-component metal crystals, which had far more symmetries to exploit. To gain further insight into the surface structure, we model our experimental measurements with ion-trajectory tracing simulations using molecular dynamics. Our simulations provide a qualitative indication that our experimental measurements agree better with a subsurface cation vacancy model than a distorted bulk model.
Using an optical microscopy setup adapted to in-situ studies of ice formation at ambient pressure, we examined a specific multicomponent mineral, microcline, with the ultimate aim of gaining a more realistic understanding of ice nucleation in Earth’s atmosphere. We focused on a perthitic feldspar, microcline, to test the hypothesis that co-existence in some feldspars of K-rich and Na-rich phases are contributing to enhanced ice nucleation. On a sample deliberately chosen to contain lamella, a typical perthitic microstructure, and flat surface regions next to each other, we performed a series of ice formation experiments. We found microcline to promote ice formation, causing a large number of ice nucleation events at around - 27°C. The number of ice nuclei decreased from experimental run to experimental run, indicating surface aging upon repeated exposure to humidity. An analysis of 10 experimental runs of identical conditions did not reveal an obvious enhancement of ice formation at the lamellar microstructure. Instead, we find efficient nucleation at various surface sites that produce orientationally aligned ice crystallites with asymmetric shape. Based on this observation we propose that surface steps running along select directions produce microfacets of an orientation that is favorable to enhanced ice nucleation, similar to previously reported for K-rich feldspars.
We have used scanning tunneling microscopy and density functional theory calculations to study molecular layers of coronene on Cu(111). The structure and stability of these layers is determined by the balance between coronene-substrate and coronene-coronene interactions. Here, we characterize this balance by measuring the maximum coverage before coronene dewets the substrate to form three-dimensional islands. We find that coronene molecules lie parallel to the substrate at the maximum coverage, in contrast to previous observations of tilted coronene on metal surfaces. We attribute this previously reported tilt to a metastability caused by an activation barrier to nucleate three-dimensional islands.
We developed a method for examining ice formation on solid substrates exposed to cloud-like atmospheres. Our experimental approach couples video-rate optical microscopy of ice formation with high-resolution atomic-force microscopy (AFM) of the initial mineral surface. We demonstrate how colocating stitched AFM images with video microscopy can be used to relate the likelihood of ice formation to nanoscale properties of a mineral substrate, e.g., the abundance of surface steps of a certain height. We also discuss the potential of this setup for future iterative investigations of the properties of ice nucleation sites on materials.
Ice in the atmosphere affects Earth's radiative properties and initiates most precipitation. Growing ice often requires a solid surface, either to catalyze freezing of supercooled cloud droplets or to serve as a substrate for ice deposited from water vapor. There is evidence that this surface is typically provided by airborne mineral dust; but how chemistry, structure and morphology interrelate to determine the ice-nucleating ability of mineral surfaces remains elusive. Here, we combine optical microscopy with atomic force microscopy to explore the mechanisms of initial ice growth on alkali feldspar, a mineral proposed to dominate ice nucleation in Earth's atmosphere. When cold air becomes supersaturated with respect to water, we discovered that ice rapidly spreads along steps of a feldspar surface. By measuring how ice propagation depends on surface-step height we establish a scenario where supercooled liquid water condenses at steps without having to overcome a nucleation barrier, and subsequently freezes quickly. Finally, our results imply that steps, which are common even on macroscopically flat feldspar surfaces, can accelerate water condensation followed by freezing, thus promoting glaciation and dehydration of mixed-phase clouds.
Atmospheric ice affects Earth's radiative properties and initiates most precipitation. Growing ice typically requires a particle, often airborne mineral dust, e.g., to catalyze freezing of supercooled cloud droplets. How chemistry, structure and morphology determine the ice - nucleating ability of minerals remains elusive. Not surprisingly, poor understanding of a erosol - cloud interactions is a major source of uncertainty in climate models. In this project, we combine d optical microscopy with atomic force microscopy t o explore the mechanisms of initial ice formation on alkali feldspar, a mineral proposed to dominate ice nucleation in Earth's atmosphere. When cold air becomes supersaturated with respect to water, we discovered that supercooled liquid water condenses at steps without having to overcome a nucleation barrier, and subsequently freezes quickly. Our results imply that steps, common even on macroscopically flat feldspar surfaces, can accelerate water condensation followed by freezing, thus promoting glaciation and dehydration of mixed - phase clouds. Motivated by the fact that current climate simulations do not properly account for feldspar's extreme efficiency to nucleate ice, we modified DOE's climate model, the Energy Exascale Earth System Model (E3SM), to i ncrease the activation of ice nucleation on feldspar dust. This included add ing a new aerosol tracer into the model and updat ing the ice nucleation parameterization, based on Classical Nucleation Theory, for multiple mineral dust tracers. Although t he se m odifications have little impact on global averages , predictions of regional averages can be strongly affected .
HKUST-1 or Cu3BTC2 (BTC = 1,3,5-benzenetricarboxylate) is a prototypical metal-organic framework (MOF) that holds a privileged position among MOFs for device applications, as it can be deposited as thin films on various substrates and surfaces. Recently, new potential applications in electronics have emerged for this material when HKUST-1 was demonstrated to become electrically conductive upon infiltration with 7,7,8,8-tetracyanoquinodimethane (TCNQ). However, the factors that control the morphology and reactivity of the thin films are unknown. Here, we present a study of the thin-film growth process on indium tin oxide and amorphous Si prior to infiltration. From the unusual bimodal, non-log-normal distribution of crystal domain sizes, we conclude that the nucleation of new layers of Cu3BTC2 is greatly enhanced by surface defects and thus difficult to control. We then show that these films can react with methanolic TCNQ solutions to form dense films of the coordination polymer Cu(TCNQ). This chemical conversion is accompanied by dramatic changes in surface morphology, from a surface dominated by truncated octahedra to randomly oriented thin platelets. The change in morphology suggests that the chemical reaction occurs in the liquid phase and is independent of the starting surface morphology. The chemical transformation is accompanied by 10 orders of magnitude change in electrical conductivity, from <10-11 S/cm for the parent Cu3BTC2 material to 10-1 S/cm for the resulting Cu(TCNQ) film. The conversion of Cu3BTC2 films, which can be grown and patterned on a variety of (nonplanar) substrates, to Cu(TCNQ) opens the door for the facile fabrication of more complex electronic devices.
The earliest stages of soot formation in flames are believed to involve the formation of small, nanoscale clusters of polycyclic aromatic hydrocarbon molecules. The structure of these clusters is still highly uncertain, however, impeding the construction of quantitative models of soot inception and growth. To provide insight into the structure of incipient soot, we produced nanoclusters of hydrocarbon molecules by annealing coronene films deposited on Pt(111), and examined them with scanning tunneling microcopy. We find that clusters containing ∼20–100 molecules, are disordered agglomerations of stacks that are ∼5–6 molecules tall. These structures are quite distinct from crystalline coronene, but bear a striking resemblance to recently proposed models for the equilibrium structure of similarly-sized clusters that are assumed to initiate soot formation. In contrast to mature soot, the surfaces of these clusters contain very few molecules with graphitic planes oriented parallel to the surface.