NH3on Si(001): Self-organized patterns of adsorbates investigated by a combination of scanning tunneling microscopy experiments and density functional theory calculations

Coadsorption patterns ofNH3molecules on the Si(001) surface observed using scanning tunneling microscopy

“…1 As a prototypical system for examining the selforganization of organic molecules, the ordering of the adsorbed NH 3 molecules has been the subject of many theoretical and experimental investigations. [2][3][4][5][6] In one of the early studies, it has been suggested that the dissociated NH 2 fragments attract each other because of a hydrogen bonding interaction. 2 This attraction leads to the array of the NH 2 species along a dimer row ͑the so-called linear feature͒, and also favors the gauche geometry where two H atoms in the NH 2 group are rotated about the Si-N bond, as drawn in Fig.…”

“…4 However, a recent scanning tunneling microscopy ͑STM͒ study has inferred that both features coexist without any appreciable preference. 6 All the previous studies presumed a unique adsorption geometry of an NH 3 molecule, the so-called on-dimer ͑OD͒ configuration where the dissociated fragments NH 2 and H form bonds with the two Si atoms of a dimer. [7][8][9][10][11][12][13][14][15] However, in a recent publication, 16 we have demonstrated that there exists an additional adsorption configuration which is designated as the inter-dimer ͑ID͒ configuration, where NH 2 and H form bonds with two Si atoms of two adjacent dimers in a dimer row.…”

Co-adsorption patterns of NH3 on Si(001): Comment on “The ordering of the adsorbed NH3 molecules across the Si dimer rows on the Si(001) surface”

“…The most likely interpretation of the high-coverage linear feature is a linear OD structure, as we have concluded. Following from this, we note that the apparent dominance of the zigzag OD feature over the linear OD feature at low coverages reported by Chung et al [1,6] does not manifest at high-coverages where any difference between the two populations is small, with a slight preponderance of linear OD features at room temperature [7]. Interpretation of linear OD features as linear ID features may have affected the low coverage statistics.…”

“…Using an alternative interpretation of the STM appearance of NH 3 fragments [7,8], we find that the correlation across dimer rows for these fragments can be explained by hydrogen bonding interactions between incoming NH 3 molecules and existing NH 2 groups on the surface [8]. We also note that a similar re-interpretation of OD-ID pairing within a dimer row, as reported in their previous work [6], would result in the majority OD-ID pair having adjacent NH 2 groups, providing further evidence of H-bonding interactions.…”

“…There are two possible mechanisms for adsorbate-adsorbate interactions in the ammonia/Si(0 0 1) system: substrate-mediated (via the buckling of adjacent Si dimers) [2]; and direct, H-bonding interactions both between neighbouring NH 2 groups and between incoming NH 3 and existing NH 2 groups [3]. The relative strength of these interactions has been previously studied with STM and DFT calculations both for low coverages of NH 3 as in this paper [4][5][6] and at saturation [7,8]. The crucial point in interpreting STM images is the appearance of the surface after adsorption and decomposition of NH 3 ; we believe that, by not considering one interpretation [8], Chung et al have provided a flawed analysis of their STM data and failed to identify the correct mechanism causing cross-row correlations.…”