And third, the composition of the oxide layer may change with the depth inside the oxide layer. Second, the metal influence via the very small a-parameter might be still there. This cannot be excluded completely, however, we have used a standardized cleaning procedure (see the section on Experimental Details) which allowed avoiding enhanced carbon concentrations in the past. The first one might be imperfect cleaning of the ss. However, why do we observe a change of the WF during sputtering at all? Here, several factors need to be considered. 20 The underlying mechanism of the above mentioned phenomenological correlations 6 is likely the correlation of the phenomenological parameters with the NodE in the ssO. 3 shows the correlation with the d electrons under the assumption that the number of d electrons (NodE) per metal atom in the oxide layer can be determined by the nominal composition of the bulk except the Ni atoms, which are nearly not incorporated in the oxide layer. Theoretically, the number of d electrons will likely play a decisive role in the quantitative explanation of the WF, since they strongly contribute to valence and conduction band of the ssO. Also, the XPS data were consistent with these information. 6 There, EDX results are compared with the bulk composition given by the manufacturer. There are some phenomenological correlations between the chemical composition and these work function values, e.g., with the Mo content or with the Cr equivalent as we have described in our recent work. The determined work function induces an order of the steels which have a very different composition. Table II shows the results quantitatively.

The electrolyte was composed of 0.05M I 2 (Fisher Scientific), 0.6M 1-methyl-3-propylimidazoliumiodide (Sigma Aldrich), 0.2M LiI (Sigma Aldrich), 0.05M guanidinium thiocyanate (Sigma Aldrich), and 0.3M benzimidazole (Sigma Aldrich) in 3-methoxy-propionitrile (Alfa Aesar). The counter-electrode consisted of the same stainless steel mesh and was covered with 1 nm of Pt on both sides. This electrode was then calcined (450 ☌) and covered with the dye N719 (Solaronix, 0.5M in ethanol). Baker), 35.15 g terpineol-ethylcellulose mixture made of 47 g terpineol (Sigma Aldrich), and 3 g ethylcellulose (Sigma Aldrich). The TiO 2 particle coating solution was prepared by 6.75 g anatase nanopowder (Sigma Aldrich), 20 ml ethanol, 5.55 ml mQ water, 1.05 ml acetic acid (J. The solution for the sol-gel process contained 11.7 mM titanium isopropoxide (Acros) and 4.7 mM hydrochloric acid (from 2M aqueous Normadose HCl, Prolabo) dissolved in i-propanol (Merck, LC-MS grade). 5,6 The cells regarded here had working electrodes made of a stainless steel mesh covered with a stack of titania layers, a compact TiO 2 layer made by a sol-gel process and two mesoporous layers of pre-synthesized titania particles. The solar cells discussed in the Introduction section are described in other publications in more detail.

Stainless steel works