Scaling relationships and theory for vibrational frequencies of adsorbates on transition metal surfaces

Abstract

Adsorbate vibrational excitations are an important fingerprint of molecule/surface interactions, affecting temperature contributions to the free energy and impacting reaction rate and equilibrium constants. Furthermore, vibrational spectra aid in identifying species and adsorption sites present in experimental studies. Despite their importance, knowledge of how adsorbate frequencies scale across materials is lacking. Here, by combining previously reported experimental data and our own density-functional theory calculations, we reveal linear correlations between vibrational frequencies of adsorbates on transition metal surfaces. Through effective-medium theory, linear muffin-tin orbital theory, and the d-band model, we rationalize the squares of the frequencies to be fundamentally linear in their scaling across transition metal surfaces. We identify the adsorbate-binding energy as a descriptor for certain molecular vibrations and rigorously relate errors in frequencies to errors in adsorption energies. We also discuss the impact of scaling on surface thermochemistry and adsorbate coverage.

Infrared spectroscopy data- and physics-driven machine learning for characterizing surface microstructure of complex materials

“…The dataset is located in Supplementary Tables 1 and 2. In addition to describing chemisorbed CO frequencies well, the relative intensities for chemisorbed CO spectra can be computed reliably from DFT despite significant errors in DFTcalculated CO adsorption energies 28,30,31 . CO frequencies do not correlate with adsorption energy as the governing physics for each is different (see Supplementary Fig.…”

“…CO is extensively used as a probe molecule for IR studies because of its well-defined peaks. It is also ideal for computations because its frequencies can be computed accurately 28 , its distinctive C-O stretch frequency depends on both site type 28 and coordination 50 , and its vibrational modes are driven by interaction with the metal surface 31 . We extend the model using NO as a probe molecule in the SI because NO binds in a similar manner to CO.…”

“…We selected the RPBE density functional 52 with D3 dispersion corrections 53 because it has been shown to work well for calculating properties of adsorbates 54 and accurately reproduces the Pt lattice constant (3.935 Å computed vs 3.912 Å experimental 55 ). Importantly, it estimates accurate frequencies for chemisorbates 31 . For nanoparticle systems involving spin-polarized calculations, the initial magnetic moments for each atom were set to the VASP default.…”

Property–Activity Relations for Methane Activation by Dual‐Metal Cu–Oxo Trimers in ZSM‐5 Zeolite

“…Accordingly, we calculated the pyrrole adsorption energies [Δ E ads (C 4 H 5 N)] at active (μ‐O) sites for all the models and the corresponding redshifts of NH stretching vibrations [∆ v NH (C 4 H 5 N)], and explored the dependency of the Δ E int ≠ on these parameters (Table , Figure d,e). Lansford et al demonstrated that the following empirical relation exists for the frequency dependency on adsorption energywhere α and β are constants. Therefore, the [∆ v NH (C 4 H 5 N)] 2 was used as a linear function of Δ E int ≠ .…”

“…The development of scaling relations represents an important advance, which link the activation barriers with the adsorption properties of chemical species across different catalyst surfaces . This allows the description of the catalytic activity in terms of structural properties such as adsorption energies or vibrational frequencies of reaction intermediates or guest probe molecules . A better understanding of the property–reactivity relationships enables an accelerated predication of catalyst performance and an improved efficiency of catalyst screening and design.…”