skip to main content


Title: Bayesian learning of chemisorption for bridging the complexity of electronic descriptors
Abstract

Building upon thed-band reactivity theory in surface chemistry and catalysis, we develop a Bayesian learning approach to probing chemisorption processes at atomically tailored metal sites. With representative species, e.g., *O and *OH, Bayesian models trained with ab initio adsorption properties of transition metals predict site reactivity at a diverse range of intermetallics and near-surface alloys while naturally providing uncertainty quantification from posterior sampling. More importantly, this conceptual framework sheds light on the orbitalwise nature of chemical bonding at adsorption sites withd-states characteristics ranging from bulk-like semi-elliptic bands to free-atom-like discrete energy levels, bridging the complexity of electronic descriptors for the prediction of novel catalytic materials.

 
more » « less
Award ID(s):
1845531
NSF-PAR ID:
10360533
Author(s) / Creator(s):
; ;
Publisher / Repository:
Nature Publishing Group
Date Published:
Journal Name:
Nature Communications
Volume:
11
Issue:
1
ISSN:
2041-1723
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. Abstract

    Bacillus subtilisendospore‐mediated forsterite dissolution experiments were performed to assess the effects of cell surface reactivity on Mg isotope fractionation during chemical weathering. Endospores present a unique opportunity to study the isolated impact of cell surface reactivity because they exhibit extremely low metabolic activity. In abiotic control assays,24Mg was preferentially released into solution during forsterite dissolution, producing an isotopically light liquid phase (δ26Mg = −0.39 ± 0.06 to −0.26 ± 0.09‰) relative to the initial mineral composition (δ26Mg = −0.24 ± 0.03‰). The presence of endospores did not have an apparent effect on Mg isotope fractionation associated with the release of Mg from the solid into the aqueous phase. However, the endospore surfaces preferentially adsorbed24Mg from the dissolution products, which resulted in relatively heavy aqueous Mg isotope compositions. These aqueous Mg isotope compositions increased proportional to the fraction of dissolved Mg that was adsorbed, with the highest measured δ26Mg (−0.08 ± 0.07‰) corresponding to the highest degree of adsorption (~76%). The Mg isotope composition of the adsorbed fraction was correspondingly light, at an average δ26Mg of −0.49‰. Secondary mineral precipitation and Mg adsorption onto secondary minerals had a minimal effect on Mg isotopes at these experimental conditions. Results demonstrate the isolated effects of cell surface reactivity on Mg isotope fractionation separate from other common biological processes, such as metabolism and organic acid production. With further study, Mg isotopes could be used to elucidate the role of the biosphere on Mg cycling in the environment.

     
    more » « less
  2. Abstract

    The electrochemical ammonia oxidation to dinitrogen as a means for energy and environmental applications is a key technology toward the realization of a sustainable nitrogen cycle. The state-of-the-art metal catalysts including Pt and its bimetallics with Ir show promising activity, albeit suffering from high overpotentials for appreciable current densities and the soaring price of precious metals. Herein, the immense design space of ternary Pt alloy nanostructures is explored by graph neural networks trained on ab initio data for concurrently predicting site reactivity, surface stability, and catalyst synthesizability descriptors. Among a few Ir-free candidates that emerge from the active learning workflow, Pt3Ru-M (M: Fe, Co, or Ni) alloys were successfully synthesized and experimentally verified to be more active toward ammonia oxidation than Pt, Pt3Ir, and Pt3Ru. More importantly, feature attribution analyses using the machine-learned representation of site motifs provide fundamental insights into chemical bonding at metal surfaces and shed light on design strategies for high-performance catalytic systems beyond thed-band center metric of binding sites.

     
    more » « less
  3. Abstract

    A fundamental understanding of the enantiospecific interactions between chiral adsorbates and understanding of their interactions with chiral surfaces is key to unlocking the origins of enantiospecific surface chemistry. Herein, the adsorption and decomposition of the amino acid proline (Pro) have been studied on the achiral Cu(110) and Cu(111) surfaces and on the chiral Cu(643)R&Ssurfaces. Isotopically labelled 1‐13C‐l‐Pro has been used to probe the Pro decomposition mechanism and to allow mass spectrometric discrimination ofd‐Pro and 1‐13C‐l‐Pro when adsorbed as mixtures. On the Cu(111) surface, X‐ray photoelectron spectroscopy reveals that Pro adsorbs as an anionic species in the monolayer. On the chiral Cu(643)R&Ssurface, adsorbed Pro enantiomers decompose with non‐enantiospecific kinetics. However, the decomposition kinetics were found to be different on the terraces versus the kinked steps. Exposure of the chiral Cu(643)R&Ssurfaces to a racemic gas phase mixture ofd‐Pro and 1‐13C‐l‐Pro resulted in the adsorption of a racemic mixture; i.e., adsorption is not enantiospecific. However, exposure to non‐racemic mixtures ofd‐Pro and 1‐13C‐l‐Pro resulted in amplification of enantiomeric excess on the surface, indicative of homochiral aggregation of adsorbed Pro. During co‐adsorption, this amplification is observed even at very low coverages, quite distinct from the behavior of other amino acids, which begin to exhibit homochiral aggregation only after reaching monolayer coverages. The equilibrium adsorption ofd‐Pro and 1‐13C‐l‐Pro mixtures on achiral Cu(110) did not display any aggregation, consistent with prior scanning tunneling microscopy (STM) observations ofdl‐Pro/Cu(110). This demonstrates convergence between findings from equilibrium adsorption methods and STM experiments and corroborates formation of a 2D random solid solution.

     
    more » « less
  4. Abstract

    Although interfacial solvation plays an important role in determining carbon dioxide reduction (CO2R) kinetics, present understanding of the potential dependent properties of the electrochemical double layer under conditions relevant for CO2R remains limited. This article summarizes the development and recent applications of plasmon‐enhanced vibrational sum frequency generation (VSFG) spectroscopy to study the effects of cation hydration and interfacial solvation on CO2R using CO as a vibrational Stark reporter. Results show that electrolyte cations retain their entire solvation shell upon adsorption to inactive sites, while active sites retain only a single water layer between the gold surface and the cation. Measurements also show that the total interfacial electric field can be separated into two contributions: one from the electrochemical double layer (Stern field) and another from the polar solvation environment (Onsager field). Surprisingly, correlating VSFG spectra with reaction kinetics reveals that it is the solvation‐mediated Onsager field that governs the chemical reactivity at the electrode/electrolyte interface. Measuring the interfacial water spectra during electrocatalysis also provides evidence for the proton source during H2evolution, which competes with CO2R in aqueous electrolyte. These findings highlight the importance of directly probing cation hydration and interfacial solvation, which mediates reaction kinetics at electrochemical interfaces.

     
    more » « less
  5. Abstract

    In situ observations from a 19‐month deployment of current‐ and pressure‐sensor equipped inverted echo sounders (CPIESs) along and across the Gulf Stream near Cape Hatteras capture spatial and temporal variability where this western boundary current separates from the continental margin. Regional hydrographic casts and two temperature cross‐sections spanning the Gulf Stream southeast of Cape Hatteras are used with the CPIESs' records of acoustic travel time to infer changes in thermocline depthDTand Gulf Stream position. Wave‐like Gulf Stream meanders are observed where the Stream approaches the separation location with periods less than 15 days, wavelengths less than 500‐km, and phase speeds between 40 and 70 km d−1. Though meander amplitudes typically decrease by ∼30% on the final approach to Cape Hatteras, some signals are still coherent across the Gulf Stream separation location. Temporal variability in meander intensity may be related to the Loop Current ∼1,400 km upstream. Mesoscale variability is strongest downstream of the separation location where Gulf Stream position is no longer constrained by the steep continental slope. Low frequency transport changes in the Florida Straits are correlated with sea‐surface height gradients along the entire South Atlantic Bight (SAB) and withDTinferred at the CPIES sites. The correlations withDTare likely due to coherent transport anomalies in the Gulf Stream approaching the separation location, which then drive Gulf Stream position changes downstream of the separation location. The patterns of coherent transport anomalies may reflect large‐scale atmospheric forcing patterns or rapid equatorward propagation of barotropic signals along the SAB.

     
    more » « less