Computational NEXAFS Characterization of Molecular Model Systems for 2D Boroxine Frameworks

Abstract

The electronic properties of 2D boroxine networks are computationally investigated by simulating the NEXAFS spectra of a series of molecular models, with or without morphologic defects, with respect to the ideal honeycomb structure. The models represent portions of an irregular 2D boroxine framework obtained experimentally, as supported by the Au(111) surface. The B K-edge NEXAFS spectra are calculated within the transition potential (TP) approximation (DFT-TP). The role of the Au(111) supporting surface on the spectral features has also been investigated by comparing the calculated spectra of a defect-rich model in its free-standing and supported form. The calculated NEXAFS spectra differ from the experimental ones, as the position of the main resonance does not match in the two cases. This finding could suggest the presence of a strong interaction of the 2D boroxine network with the Au substrate, which is not captured in the model calculations. However, good agreement between measured and calculated B K-edge NEXAFS spectra is obtained for a model system, namely, trihydroxy boroxine, in which the B atoms are less screened by the valence electrons compared to the B-B linked boroxine network models considered here. These results suggest catalytic activity in the gold substrate in promoting a weakening or even the breaking of the B-B bond, which is not revealed by calculations.

Keywords: DFT calculations; X-ray absorption spectroscopy; boroxine network.

Figure 1

Ball-and-stick representation of the building blocks models of the THDB network: (A) model M1, portion of an ideal honeycomb boroxine network, (B) model M2, (C) model M3. B atoms in yellow, O atoms in red, H atoms in white.

Figure 2

Calculated B K-edge NEXAFS spectrum of THDB (reported on the right side). (Upper panel): DFT-TP results; (lower panel): ΔSCF results. The stick spectra are broadened by using a Gaussian line shape with FWHM = 0.5 eV. ΔSCF B1s IP (197.28 eV) is reported as a vertical dashed line.

Figure 3

Calculated B1s NEXAFS spectra of M1 (upper panel), M2 (middle panel) and M3 (lower panel) molecular models depicted on the right side. Labels denote the non-equivalent B_i sites (see text for explanation). Transitions from the non-equivalent B_i atoms are reported as vertical colored lines in the spectra. The ΔSCF B1s ionization energies are shown as vertical dashed lines (the IP values are reported in Supplementary Table S1). The stick spectra are broadened by using a Gaussian line shape with FWHM = 0.3 eV.

Figure 4

Comparison between the calculated B1s NEXAFS spectra of free M3 (upper panel) and M3@Au(111) (lower panel). The contributions of the three groups of non-equivalent B_i sites are also shown (colored solid lines). The ΔSCF B1s ionization energies are shown as vertical dashed lines (the IP values are reported in Supplementary Table S1). The stick spectra are broadened by using a Gaussian line shape with FWHM = 0.3 eV. The M3 and M3@Au(111) models employed are reported on the right side.

Figure 5

B1s NEXAFS spectra of boroxinated monolayer ((upper panel), experimental data with permission from ref. [7]), M3@Au(111) (middle panel) and THBoroxine (lower panel) at the two different polarization angles. For the THBoroxine molecule (displayed along the corresponding spectrum), the B1s total spectrum is also reported. The ΔSCF B1s ionization energies are shown as vertical dashed lines (IP mean value for M3@Au, 198.92 eV for THBoroxine). The calculated stick spectra are broadened by using a Gaussian line shape with FWHM = 0.3 eV.