Unmodeled surface features affecting NADG/NBDG monolayer reconstruction on Au(111)

Determine whether Au adatoms, species introduced during electrospray deposition (such as residual solvent molecules or leached Na/Ca ions), or unit-cell packing effects from neighboring adsorbates are present on Au(111) and account for the observed need to constrain the atoms surrounding the central carboxyl groups at elevated positions during DFT relaxation of the 4-nitrophenyl-alpha-D-galacturonide (NADG) and 4-nitrophenyl-beta-D-galacturonide (NBDG) monolayer models, thereby establishing the specific surface or assembly features responsible for the experimental–theoretical discrepancy.

Background

In reconstructing the 3D atomic-resolution models of NADG and NBDG monolayers on Au(111), the authors used a multi-fidelity workflow combining DFT, Bayesian optimization, and a machine-learning interatomic potential with minima hopping. During final refinement, they found that fully relaxed DFT structures did not match experimental AFM features unless certain atoms near the central carboxyl groups were constrained at higher positions than obtained in unconstrained relaxations.

This mismatch suggests that some physical aspect of the surface or assembly environment was not captured by the computational model. The authors propose several possibilities, including the presence of Au adatoms (potentially mobilized during electrospray deposition and favored near step edges), incorporation of residual solvent or ionic contaminants (e.g., Na/Ca) from electrospray deposition, or small discrepancies in unit-cell dimensions that could miss subtle packing effects from neighboring adsorbates. Clarifying which factor(s) are operative would resolve the origin of the discrepancy and improve the fidelity of monolayer models.

References

This suggests that there could be some feature in the surface which we were unable to capture with our methods. This could for instance be adatoms, an assumption supported by the fact that the monolayers are formed close to step edges, potentially acting as a source of adatoms. It has also been shown that organic molecules deposited onto a surface with solvent, such as methanol, may expedite adatom migration relative to UHV deposition, consistent with the above hypothesis. A second possibility is that the electrospray deposition process itself introduces either solvent molecules or contaminants which could be incorporated into the monolayer structure, such as Na or Ca ions, known to leach from laboratory glassware. A final third option could be that the slightly elongated unit cells of our model structures do not fully capture the packing effects due to neighbouring adsorbates, with the real systems being slightly more crowded than our models.

— Direct imaging of carbohydrate stereochemistry (2410.20897 - Cai et al., 28 Oct 2024) in Computational, Subsection 'Minima hopping for monolayer structures'