AVS2009 Session SS2-WeM: Surface Chemistry and Dynamics

The adsorption of xenon was studied on Cu(111), Cu(221), Cu(643), and Cu(653) using temperature programmed desorption (TPD) of xenon and ultraviolet photoemission of adsorbed xenon (PAX). These experiments were performed to study the atomic and electronic structure of step-kinked, chiral copper surfaces. Xenon TPD and PAX were performed in an attempt to distinguish terrace, step-edge, and kink adsorption sites by adsorption energy (TPD) and local work function differences (PAX). The different adsorption sites could not be clearly differentiated using xenon TPD due to the complex behavior of xenon on these surfaces. Comparison of TPD spectra of R-3-methylcyclohexanone (R-3-MCHO) previously performed¹ on stepped and kinked copper surfaces suggests clear differences between Cu(643) and Cu(653) that are not apparent using xenon TPD. However, unique features of xenon adsorption on different copper surfaces were visually and numerically distinguishable using the PAX method. The PAX experiments on the copper surfaces demonstrate local work function differences between kink and step surface sites as well as step and terrace sites. A data fitting model was developed to analyze the PAX data for all four copper surfaces simultaneously. Kink sites were found to have a lower local work function than step sites, and steps, in turn, had a lower local work function than terrace sites. Step/kink adsorption of xenon is favored (for all surfaces studied at 50-70 K) at low coverages of xenon, but these sites do not saturate until monolayer coverage is reached. The results of this research provide several observations regarding the adsorptive behavior of xenon on vicinal copper surfaces.

¹Horvath, J. D.; Koritnik, A.; Kamakoti, P.; Sholl, D. S.; Gellman, A. J., Enantioselective separation on a naturally chiral surface. Journal of the American Chemical Society 2004, 126, (45), 14988-14994.

Various C₂H_x surfaces intermediates formed through the thermal decomposition of acetylene and ethylene on the Pt(111) surface were identified and their stabilities characterized through complementary studies using the techniques of low temperature scanning tunneling microscopy (LT-STM) and reflection absorption infrared spectroscopy (RAIRS). By providing high-resolution vibrational spectra of surface species, RAIRS data is highly sensitive to the chemical identity of those species. However, it does not provide information on the relative coverages of adsorbed molecules and many important stable intermediates are invisible to RAIRS. In contrast, with a LT-STM operated at 4.7 K, individual molecules are observable and their absolute coverages are readily obtained simply by counting, yet their chemical identities cannot be directly determined from the STM images alone. In the case of acetylene, both RAIRS and LT-STM indicate that it adsorbs in a single form at low temperature in which the CC bond is positioned over a three-fold hollow site, and that the interaction with the surface occurs through both di-σ and π bonding. On warming to 250 K in the presence of coadsorbed hydrogen, RAIRS provides strong evidence for the partial hydrogenation of acetylene to form an adsorbed vinyl (CHCH₂) species. The LT-STM images following a 250 K anneal in the presence of coadsorbed hydrogen show a marked increase in the coverage of a species that is identified as vinyl, based on the RAIRS results. Both techniques indicate that adsorbed acetylene is stable up to 300 K. A third species, in addition to adsorbed acetylene and vinyl, is also observed with the LT-STM after a 300 K anneal, which is identified as vinylidene, CCH₂. However, the RAIRS evidence for a vinylidene species is not definitive. In the case of ethylene adsorption at low temperature, RAIRS provides clear evidence for a di-σ bonded form of ethylene at low temperatures that coverts to ethylidyne at 280-420 K. The LT-STM images show that ethylene exists in both π-bonded and di-σ bonded forms at low temperature and that the two forms can be easily interconverted using electron pulses from the STM tip. On the basis of RAIRS experiments, it was found that surface carbon formed through the complete dehydrogenation of acetylene could be hydrogenated to form ethylidyne (CCH₃), from which it was inferred that the surface carbon was in the form of C₂ molecules. A form of surface carbon that can be hydrogenated to ethylidyne was also identified with LT-STM. Other surface species that have been investigated with the two techniques include ethynyl (CCH), ethylidene (CHCH₃), and methylidyne (CH).

Ammonia (NH₃) can be potentially used for hydrogen storage, because the dissociation of ammonia molecules generates hydrogen. Although Ruthenium is the best elementary metallic catalyst for ammonia synthesis, the exact reaction mechanisms at the atomic scale of both synthesis and dissociation over the Ruthenium surface have yet to be understood. Previously, the dehydrogenated products were obtained by annealing the sample and characterized by spectroscopic methods.

We studied the dehydrogenation of single ammonia molecules adsorbed on a Ru(0001) surface, by means of scanning tunneling microscopy (STM) at low temperature. The sudden change in tunneling current during the dissociation process allows us to study the dissociation mechanism and rate of single molecules in detail. We observe the different dehydrogenation steps after selectively applying voltage pulses in the order of a few volts between STM tip and an adsorbed ammonia molecule. The various dissociation products show a distinct imaging contrast. Together with the assignment of the adsorption site, this leads to the identification of the dissociation products. We will discuss to what extent electron induced and electric field driven processes, respectively, are affecting the dissociation mechanism of ammonia and its dissociation products.

[1] Shen et al. J. Phys. Chem. C 112 (2008) 4281

[2] Shen et al. J. Chem. Phys. 130 (2009) 094701

We present detailed theoretical analyses of nominally anionic Iodine and nominally cationic Cs adsorbed on Cu(111) and Pt(111) surfaces. We consider the consequences of the coverage, the choice of on-top or three-fold sites, and the differences between Cu and Pt substrates on the degree of ionicity and on other properties of the interaction. This work extends our earlier studies of I and Cs on Cu(111)¹ and of I/Pt(111).² There we demonstrated that, although both Cs and I were dominantly ionic adsorbates, the properties of the adsorption, including work function changes, ΔΦ, and photoemission binding energy shifts, ΔBE, depended on other factors besides the adsorbate ionicity. However in this earlier work, only a single I or Cs adsorbate was explicitly treated and we neglected both direct and indirect, substrate mediated, adsorbate-adsorbate interactions. Now, we use models where several adsorbates are included, thus explicitly modeling lower and higher coverages. In particular, we examine whether the coverage dependent departures from ideal ionicity are different for cations and anions. Furthermore, we examine how ΔΦ depends on the adsorbate-adsorbate interaction and compare our predictions with measurements of ΔΦ as a function of coverage. This is relevant for the modification of charge transport barriers, which is key for the design of organic electronics.³ The distance of I above Pt(111) is quite different for on-top and three-fold sites,² which has important consequences for ΔΦ and ΔBE. Here, we examine the generality of different heights of ionic adsorbates at these sites and how this difference is affected by the sign of the ionicity and the substrate. Our theoretical methodology uses wavefunctions for cluster models of the surface since this allows us to determine the quantitative importance of the individual physical and chemical mechanisms that contribute to the interaction. Our present work marks the first time that these methods have been applied to study the influence of coverage for halogen and alkali metal adsorbates.

¹P. S. Bagus, D. Käfer, G. Witte, and C.Wöll, Phys. Rev. Lett., 100, 126101 (2008).

²P. S. Bagus, C. Wöll, and A. Wieckowski, Surf. Sci. 603, 273 (2009).

³G. Witte, S. Lukas, P. S. Bagus, and C. Wöll, Appl. Phys. Lett., 87,263502 (2005).