Sputter cleaning

Tonner et al. have taken scanning XPS microscopies at the Advanced Light Source Synchrotron Radiation Center of Lawrence Berkeley National Laboratory [2.6]. They investigated a polished and sputter-cleaned surface of mineral ilmenite with the nominal composition FeTi03, and used the Fe 3p and Ti 3p lines for imaging. Using synchrotron radiation they demonstrated spatial resolution of approximately 0.25 p,m. 

Good surface preparation is essential, as in all deposition processes, and this can be achieved by chemically cleaning the substrate followed by sputter cleaning with argon just prior to the actual deposition. 

Fig. 4 Survey spectrum of sputter-cleaned Hf metal containing 3% Zr impurity, with all visible core-line peaks labelled, collected using a monochromatic A1 Ka X-ray source (1486.7 eV). The stepped background hatched), found in all XPS spectra, arises from photoelectrons that lose KE by inelastic electron scattering as they travel through the surface and into vacuum...

Figure 6.26. CO stripping voltammetry of UHV sputter-cleaned electrodes in 0.5 M H2S04 on (a) Pt and (b) Ru. Solid curves represent the stripping of CO in the first positive-going sweep dotted lines represent the voltammetric profiles in the absence of CO (adapted from Ref. [153]).

Most of the sputtering data are for pure materials. In fact, most investigators have attempted to avoid surface contamination in order to make measurements truly representative of the substrate material. In a Tokamak reactor the wall will be sputter cleaned and hence these data are relevant. The actual conditions in today s operating Tokamaks are more complex. For example, carbon and other materials layers have been shown to build up on the surface of walls, and hence it is desirable to have sputtering data on samples with impurity layers195. ... 

It turned out that only discharge sputter cleaning provided a sufficiently clean surface. 

Figure 9.15. Influence of sample cleaning on XPS scans taken on a thin-film superconductor, (a) Survey scan from an as-received surface, (b) Survey scan from surface after ion-beam (sputter) cleaning. Note the reduction in the Cls peak after cleaning, (c) Comparative Ba3d scans from both cases. Note the change in shape and size as the surface contaminant layers (probably containing carbonates and hydroxides of Ba in addition to other components) are removed. The peak shapes and intensities of other cations change, too. Initial data represent the composition and chemistry of the contaminant layer, whereas that from sputtered sample represents those of the pure underlying superconductor (possibly with sputter-induced changes that need to be accounted for).

Figure 24 Auger spectrograph of molybdenum silicide film both (a) before and (b) after sputter cleaning.

Lorentz curves with a tail function for the asymmetry of the XPS signals for these transition elements. These sets were kept constant in position, shape and size relative to each other. The XPS spectrum of an actual specimen was described with a least-square fit by variation of the size of the appropriate standard sets [36]. Various standards were prepared and measured for a subsequent data analysis of actual specimens. Only some few examples are mentioned here. Pure metal standards are Ar-sputter cleaned specimens. Fe(III) oxide corresponds to a thick passive layer formed at the positive end of the passive range. For Fe(II), a passive layer formed on Fe5Cr is reduced in 1 M NaOH at ca. = -1.0 V (SHE) [12]. For NiO, oxide grown at 1000 °C on pure Ni in air was used as a standard. For Ni(III)oxyhydroxide, NiOOH was deposited by oxidation of Ni2+ from weakly alkaline solution or formed... 

Previous studies have indicated that adsorption of 2-butanol on a cold metal surface is molecular (there is no adsorption-induced dissociation). To insure that this was the case for 2-butanol adsorbed on a sputter-cleaned permalloy surface, XPS spectra were taken following adsorption on a thick Xe overlayer deposited at 25 K. The inset in Fig. 7 shows a comparison of a C Is core level spectrum of 2-butanol adsorbed on multilayer Xe and compared with a spectrum of the same amount of 2-butanol adsorbed on permalloy. Other than a rigid shift of the binding energies, due to the lower screening of the C Is hole by Xe, the two spectra are very similar. Since Xe is chemically inert it is reasonable to expect that 2-butanol remains intact upon adsorption and, therefore, the same appears to be the case following adsorption on permalloy at 90 K. 

A typical C Is XPS spectrum obtained at the start of a photolysis series is shown in Fig. 4. Using curve-fitting techniques, the spectrum in Fig. 4 can be decomposed into three main peaks. The small one at 282.5 eV is due to atomic carbon bound to a substrate metal atom (C-M). Some of this intensity is due to residual carbon that remained after sputter-cleaning the surface (<0.03 monolayer) and some is due to carbon produced from butanol photolysis during the time it took to acquire the spectrum. The most intense peak at 284.9 eV is due to alkyl carbons (C-H), while the peak at 286.5 eV is due to the chiral C atom bound to the OH group [123]. 

Adhesion Layer RF Sputter Clean As Deposited Anneal 2 In SituAnnealb ... 

Figure 1. SIMS Surface surveys of (a) as-recieved polyimide and (b) after RF sputter cleaning in Ar ion plasma (200W,30min)...

Figure 2. SIMS in depth profile of carbon (polyimide) and Copper introduced during RF sputter cleaning.

---