Adlayer structures

Akemann W, Friedrich KA, Stimming U. 2000. Potential-dependence of CO adlayer structures on Pt(l 11) electrodes in acid solution Evidence for a site selective charge transfer. J Chem Phys 113 6864-6874. 

Chang SC, Roth JD, Ho YH, Weaver MJ. 1990. New developments in electrochemical infrared-spectroscopy— Adlayer structures of carbon-monoxide on monocrystalline metal-electrodes. J Electron Spectrosc Relat Phenom 54 1185-1203. 

Zou S, Gomes R, Weaver MJ. 1999. Infrared spectroscopy of carbon monoxide and nitric oxide on palladium(lll) in aqueous solution unexpected adlayer structural differences between electrochemical and ultrahigh-vacuum interfaces. J Electroanal Chem 474 155-166. 

Fig. 17. A 30 x 30 nm image of an ordered adlayer on Cu(lll) at 0.455 V vs. A1 in 55.0 m/o AICI3-EtMelmCl. A schematic of the proposed tetrachloroaluminate adlayer structure is shown. Reproduced from Stafford et al. [104] by permission of The Electrochemical Society.

Fig. 21 Cyclic voltammogram and corresponding in situ STM images of 3 mM 4,4 -bipyridine (44-BP) on Au(lll) in 0.05 M KCIO4, scan rate 10 mV s-1. The sizes of the STM images are 10 x 10 nm2. The following 44-BP adlayer structures have been observed in the potential regions I, II, and III (a) high coverage densely packed phase, (b) striped structure, and (c) rhombohedral phase. The corresponding molecular orientations as derived from in situ IR studies on Au(l 11) are shown in panels (d-f) [303], The pairs of peaks Pl/Pl, P2/P2, and P3/P3 indicate first-order phase transitions between the respective adlayers [304]...

The STM has also been used to follow the evolution of surface-confined reactions such as the oxidation of adsorbed sulfide to form adsorbed Sg and iodide to polyiodide [275,288,289]. The substrate exerts a strong influence on the dimensions and ordering of the adsorbed molecules, particularly the formation of the first monolayer. In a similar manner, studies of the impact of different adlayer structures on the electron transfer kinetics of various soluble redox species have been initiated [290]. 

Fig. 16.14. Ordered structures of Cu layers on Au electrodes, (a) STM image of Cu adlayer on Au(lll). A y/ X y/ )R20° structure is observed. The Cu coverage is 0.33 monolayer, (b) Model of the adlayer structure, (c) STM image of Cu adlayer on Au(lOO). A quasihexagonal structure is observed, (d) Model of the adlayer structure (O - Au atoms in the topmost layer. - Cu adatoms). (Reproduced from Magnussen et al., 1990, with permission.)...

Zhang et al. [197] have studied adsorption of DL-homocysteine and L-homo-cysteine thiolactone on Au(lll) electrode in 0.1 M HCIO4 using CV and STM. Both compounds formed highly ordered adlayer on Au(lll). For both adlayers, structural models have been proposed. 

Shi et al. [68] have studied iodine adsorption on Ag using atomic-resolution electrochemical scanning turmehng microscopy (ECSTM) method. Distinctly different iodine adlayer structures and surface diffusion behavior were observed on mechanically polished pc-Ag in comparison with those obtained on single-crystal electrodes. 

Hanewinkel et al. [84] have studied the change in the surface resistance of an Ag(lOO) electrode caused by the adsorbed bromide. It was suggested that bromide ions adsorbed in the double layer of the hollow sites do not exchange electron with the metal. Shimooka et al. [85] have studied the adlayer structures of Cl and Br, and a growth of bulk AgBr layers on Ag(lOO) electrodes using in situ... 

Ordered Adlayers. Lateral interactions for chemisorbed adsorbates are generally repulsive, although there have been reports of attractive interactions for adsorbates at distances of about twice the distance between neighboring sites.At low coverage repulsive interactions simply keep the adsorbates apart. The adlayer is disordered. At intermediate and high coverages a rich field of possible adlayer structures can be observed even when there is only one type of adsorbate. 

There are various ways in which people have done this. We assume that we model the lateral interactions with pair and multiple-particle interactions as in eqn. (53). This expansion has the advantage that the energy of a system depends linearly on the parameters of our model, and the equations determining these parameters will also be linear and easy to solve. Let s suppose that we have done calculations on N tr different adlayer structures and that we have obtained... 

If A str > par + 1 then we have more information than we strictly need. This information can be used to get some idea of how well the lateral interactions can be determined. This can be done by taking A paj. + 1 adlayer structures to compute the parameters, substitute these in E the other adlayer structures, and compare this with One can also use all adlayer structures and... 

The adsorption energy E 2 is obviously a different parameter from the lateral interactions It is not uncommon to see people determine this parameter separately. This can be done by including a calculation of an adlayer structure for which one assumes that there are no lateral interactions. Once E 2 is known, the lateral interactions can then be determined as described above with the other adlayer structures. 

Table 1 All adlayer structures for which the adsorption energy of the oxygen atoms have been determined. Also given are the coverage 6 and the adsorption energy per oxygen atom Eads (in kJjmol). Data is reproduced from references. ...

First note that when we have just enough adlayer structures to determine all parameters (Astr = Apar + 1) that we cannot say anything about the errors we have. We get per definition an exact match between and one should have at least > Apai + 1. Suppose then that this is the case, and that we have calculated adsorption energies per adsorbate but the real values are We hope that the errors... 

We see that the terms that differ for various adlayer structures do not depend on the systematic error a. These terms are also the ones that contain the lateral interactions. This means that the systematic error does not affect the lateral interactions. They depend only on the smaller random errors p . This means that lateral interactions can be determined better than one might suppose having some idea of the accuracy with which one can compute adsorption energies with DFT. The systematic errors cause only a shift in l s-... 

Looking at errors as above also shows that singling out one adlayer structure, say = 1, to determine l s is not a good idea. Because then = l s and one will set = -EI s and work with alc to determine the lateral interactions. However... 

If we treat all adlayer structures equally and we have Nstr > A par + 1 > then if is natural to use a least-squares procedure i.e., we minimize... 

NO/Rh(lll). - This system is a good example of how in a relatively simple system subtle combinations of lateral interactions and various sites with different adsorption energies can lead to a number of surprising adlayer structures. Experimentally there are two well-defined adlayer structures. At ML there is a c(4 x 2)-2NO structure with two NO molecules per unit cell. It has taken some time to determine the sites for these NO molecules, but it is now generally accepted that one NO is at an hep site, but the other is at an fee site. At ML there is a (2 x 2)-3NO structure with three NO molecules per unit cell 50,183 structure there are equal numbers of NO molecules at hep, fee,... 

If we allow the lateral interactions to vary over a too large range during the optimization, then we occasionally get adlayer structures in the kMC simulations that differ from those found experimentally. This does not mean necessarily that very different temperature-programmed desorption spectra are... 

Figure 15 shows how the leave-one-out error changes with the model of the lateral interactions. It is clear that the nearest-neighbor and next-nearest-neighbor interactions are the most important. These two interactions already provide a good description of all the adlayer structures. The next-next-nearest-neighbor interaction does not improve the model, but remarkably the linear... 

The root-mean-square error of the model with only nearest-neighbor and next-nearest-neighbor interactions is 3.9 kJ/mol. The one of the model that also includes the linear 3-particle interaction is 2.9 kJ/mol. Using these errors as estimates for the errors in DFT calculations gives the error estimates of the lateral interactions in table 2. These should be regarded as lower estimates. If we would use a larger estimate for the errors in DFT then the errors of the lateral interactions would increase proportionally. Note that, because the term is always the same in the expressions for independent of the adlayer structure, possible systematic errors in DFT (cr in Section 3.4.2) only affect E 2 and not the lateral interactions. 

Moreover, as the system of Eq. (24) is a starting point for the cluster-type approach and the MC method, therefore in principle the two methods can be used in combination. It has been proposed by Hood et al. [297] to write down the kinetic equations for describing variations in the occupancy state of each lattice site, i.e., to abandon consideration of the lattice ensemble, and to solve the system of the equations with the dimension equal to the number of sites. The system of the connected equations has been solved numerically. In each time interval the desorption probability for a given molecule is determined by the random sampling and then the general adsorbate change found. The combination approach allows to trace the adlayer structure and to construct a correlation between the structural and the kinetic behavior of the process. Such an approach has been applied to the N2/Ru(001) system to obtain a qualitative agreement with experiment [298]. 

---