ARXPS and Stmdts measurements

The interface chemistry up to a depth of about 5 - 10 nm of Ni-TiC>2-x nucleation layers was obtained by ARXPS measurements and is summarized in Fig. 1 [4], Scanning tunneling microscopy (STM) is commonly used for imaging the morphology and atomic structure of solid-state surfaces in contact with ultrahigh vacuum (UHV), gases, or liquids (Fig. 2) [6].

5 nm

carbon

NiOx(OH)y Tt02-x-Substrate

Fig. 1. Models of Ni/Ti02.x films with (a) 0.07, (b) 0.12, and (c) 0.72 mg/cm2 Ni. Those models should be regarded as an average over the whole measurement area. The slow change of XPS intensity decrease at small photo electron take-off angle yields a lateral periodicity length of 40 nm [4].

Fig 2. (a) First Ni nuclei after a current pulse of 600 mV for 20 s (CCM) 200 nm/200 rnn scans show a maximal height variation of 25 nm; (b -c) STM images taken while Ni nuclei were growing. Clearly, one can see the growth while enhancing the electroplating voltage. The CCM height variation is still 25 nm on a 200 nm/200 nm scan.

The STM imaging resolution is mainly determined by the tunneling mechanism, which may be identified by distance tunneling spectroscopy (DTS). DTS is characterized by measuring the tunneling current, It, as a function of the tip-substrate distance, d, at constant tunnel voltage Et. For ex-situ measurements Et is the applied bias between tip and sample. In in-situ experiments Et is given by the difference between the potential of the tip, Etip, and the substrate, E, controlled relative to the electrolyte. Traditionally, the/t (Et,d) dependencies are expressed as Eqs. (1) and (2),

where a is related to the density of states of tip and substrate, depending weakly on d only. The inverse decay length 2k is a function of the tunneling voltage Et and Vea(Et) denotes the effective tunnel barrier height, which decreases linearly with the voltage across the tunnel gap according to V= <X>0 - x/d ■ e Et. Here, ®o is given for a vacuum by the work function 0Vac or by the conduction band of the dielectric causing the tunnel gap. The approximation (<E>o=<I>vac) holds only for the case of a clean metal/vacuum/metal tunnel junction. For in-situ STM, we have to expect adsorbates, especially water adlayers on the surface of both tip and substrate. The dipole resonances in such oriented water molecules or localized states in defective oxides lead to intermediate-state tunneling and, consequently, smaller effective tunnel distances de/f = d!{N\+\). Assuming a uniform density of localized states, the following distance dependencies result [7]:

As one can see, the effective tunnel barrier height decreases with increasing number of intermediate states n and so depends on the oxide or adsórbate coating (Fig.3).

The ex-situ (in air) and in-situ (in Ni sulfamate solution) STM measurements were performed using a Nanoscope 111 (Digital Instruments). In ex-situ experiments the tunneling bias is held constant, whereas in in-situ STM measurements the tip and substrate potentials, Etíp and E, are independently controlled by an external bipotentiostat [8]. The experiments were performed with uninsulated and apiezon-insulated Pt-Ir tips (Digital Instruments), respectively. A superficially oxidized Au wire was used as quasi reference electrode and a Pt wire as counter electrode. The Ni sulfamate solution is the same as used for electroplating our samples (Section 2.1), but was diluted by a factor three with suprapure water and used at 300 K to keep the deposition at a slow, observable rate. Before the STM cell was filled, the solution was deaerated by nitrogen bubbling (Fig. 2, [8]).

The effective barrier height Feff was determined from the slope of the ln(/t) measurements as a function of d at a constant Et of 0.5 V. The It(d) data were acquired while approaching the tip to a minimum distance, do, determined by the setpoint current. The DTS covered a typical range Ad of about 2 nm. Withdrawal and approach of the tip were carried out slowly with a rate of approximately 10"6 cm/s (Fig. 5).

0 0

Post a comment