Fig. 4.1(d) plots on Ni seeds directly after electroplating in the sulfamate solution, i.e., E = 0 (Fig. 2). The effective barrier height Feff does not show any increase with distance d. Its value ranges from 1 to 2 eV.

3 Results

Typical in-situ STM images are shown in Fig. 2. As expected from XPS measurements, one sees a fissured landscape, making DTS in both valleys and mountain desirable. In the meantime, X-ray photoelectron spectroscopy (XPS) results indicate the existence of different materials on mountains (carbon) and in valleys (Ti02.x), see Fig. 1. As a local spectroscopy method, DTS makes it possible to obtain information about the local chemical nature of the substrate. Figure 3 shows typical Inl(d) on both (a) mountains and (b) valleys. The expected linear dependence of Inl vs. d is followed only for small parts of the plots, indicating a variation in the number of intermediate states active in the tunneling process. The effective tunneling barrier height Feff on mountains increases as tip-substrate distances increase from 0.2 to 4 eV and in valleys from 0.05 to 0.4 eV (Fig. 3).

In order to obtain more detailed information on Ni electroplating, surface morphology changes during Ni electrodeposition were studied by in-situ STM. After large potential pulses of 200 mV had been applied in the scanned Ti02-x region, only slight changes in surface morphology could be observed in Fig. 2(a) not changing much by additional potential pulses. Again, a fissured landscape is observed (Fig. 2(c)).

By increasing the substrate potential range from 0 to 200 mV in steps of 10 mV, the nucleation and growth of small Ni islands, mainly in valleys (Fig. 2) then over thewhole surface, could be imaged. Because the Ni elctrodeposition on Ti02-X is irreversible, DTS on Ni deposits was possible under the same conditions, like DTS measurements on Ti02-x substrates. The Inl(d) curves taken on freshly deposited Ni

distanced [nm] distanced [nm]

Fig.5. 1(d) plots on Si/Ti/Ti02-X/Ni layer measured in air. Compared to measurements in electrolytic solution (Fig. 4), the effective barrier heights was increased. Measurement on a mountain (a) and in a deep valley (b).

nuclei (Fig. 4) do not show any slope increase with increasing distance d. The measured tunneling barrier height ranges between 1 and 2 eV. The result can be compared with ex-situ DTS measurements in Fig. 3. In a distance range between 0 and 0.5 nm the DTS curves show a slow increase of barrier heights from 0.4 to 0.8 eV on mountains, and of 0.07 to 0.11 eV in valleys.

4 Discussion

In Section 2.1 the substrates used and their preparation have been summarized: porous Ti02-x(0H)y/H20 coating the Si wafer, probably with an excess of H2O2 due the oxidation process of the sputtered Ti layer. In Section 3 it was described how ARXPS and STM/DTS confirmed the TiC^-x roughness on a nanometer scale but revealed a stoichiometric inhomogeneity, as well. The carbon mountains may be sputter residue because in Ti sputtering a C sublayer is used for easy removal from the Si wafer [4]. According to Section 3 or [4], the electrodeposition of Ni starts in the pores yielding the NiOx(OH)y interface compound with Ti02-x(OH)y in a thickness range from 0.8 to 2.5 nm. This thick interface oxide-hydroxide is an indication of good wettability and good adhesion of Ni on Ti02-X.. This Ni-interface oxide-hydroxide is much thicker than the natural oxidecoating Ni in a thickness of 0.2 nm shown in Fig. 1. At up about 1 mg/cm2 Ni deposition on carbon, no Ni was found by ARXPS; this leads to the conclusion that carbon inhibits the Ni nucleation, which worsens the adhesion between substrate and Ni overlayer.

Because Fig. 1 gives only a spatial average of the Ni deposition process on TiC>2-x, local detection methods such as STM and DTS were used to identify the surface topography and the corresponding stoichiometry on a nanometer scale. Our STM morphology in Fig. 2 confirms our averaged ARXPS results and shows actual mountains and pores with a height difference of 25 nm on a (200 nm)2 size scale. The potential-controlled Ni deposition in the electrochemical STM cell starts in pores (valleys), forming small nuclei which lead to a more compact Ni deposit shown in Fig. 2. The filling of the pores by such spherical nuclei are a consequence of the large overpotential and the wetting interface Ti02-x(0H)zNi0y.

DTS measurements shown in Figs. 3-5 were made with the constant tunnel voltage Ei = Enp - E = 0.5 V. In all cases step-like decays occurred indicative of changing intermediate states. Measurements on the Ti02-X substrate (Fig. 4) in sulfamate solution showed on the mountains a tunnel barrier height (Feff terminating at 4 eV, and in the pores lower Veff values starting with 0.02 - 0.1 eV. Such low Feff values are typical for semiconductors with varying intermediate states in parallel [7]. This is quite typical for Ti02-X or NiOx(OH)y with O vacancies acting as such states. When the tip leaves Ti02-X for Ad > 1 nm, Veff« 0.4 eV is reached; this is typical for distinct H2O-OH interface. Similar Veff values are observed for Ti02-x-Ni-Ni0x(0H)y surfaces in Fig. 4 in the Ni-covered pores. Thus the tunneling terminates in the intermediate states of the NiOx(OH)y layer related to the rest potential due to handling. Directly after Ni deposition in the sulfamate,.bath, i.e., at E = 0, such an oxide hydroxide layer does not exist on Ni, which is then covered by a polarized 0H-H20 layer only, the compact part of the electrochemical double layer [7]. Such interfaces show Vef{ = 1 eV by one to two intermediate states. The mountains in Fig. 2 show Veg > 0.4 eV increasing to 4 eV under all conditions, indicative of a hydrophobic carbon with its double layer having up to 1 mg/cm2 Ni coverage.

0 0

Post a comment