The Tunneling Junction

This section recalls very briefly the basic concepts necessary for the discussion. Electron tunneling has been established for several decades in vacuum [23] as well as in solid-state structures [24]. Quantum mechanics predicts that electrons can flow between two conductors separated by a distance of the order of 20 A. The energy diagram of a tunneling junction is sketched in Fig. 1 a. From the Bardeen tunneling 

If the tunnel junction of Fig. 1 a is simply immersed in an electrolyte, the polarization between the tip and the sample will promote an electrolysis. A bi-potentiostat is necessary to ensure real tunneling between the sample and the tip. Such a device, classically used in electrochemistry, enables to split the tunnel junction into two sol-id/liquid interfaces, independently polarized against a reference of potential (Fig. 1 b). Using this configuration, also referred to as the four-electrode configuration and introduced very early by several groups, it is possible to avoid any electrochemical transfer between the sample and the tip [25,26]. The reference potential is an electrode whose potential is well defined and constant with respect to the vacuum level. The sample is biased against the reference electrode to monitor reactions at the surface, just as in a classical electrochemical cell. The tip potential is adjusted 

The very recent data of Pan et al. [27] are quite interesting since they underline the influence of dipoles of water molecules in the tunneling gap. Measurements in aqueous solution indeed show that electrons tunneling out of the tip into the substrate electrons do see a tunnel barrier which is nearly 0.5 eV larger than when they tunnel in the reverse direction (for the small tunnel bias used, no difference should be detected). Keeping the tunnel voltage constant, it is observed that the tunnel bar- 

---

The control parameter in an STM, the current in the tunneling junction, is always due to the same physical process. An electron in one lead of the junction has a nonvanishing probability to pass the potential barrier between the two sides and to tunnel into the other lead. However, this process is highly influenced by (i) the distance between the leads, (ii) the chemical composition of the surface and tip, (iii) the electronic structure of both the systems, (iv) the chemical interactions between the surface and the tip atoms, (v) the electrostatic interactions of the sample and tip. The main problem, from a theoretical point of view, is that the order of importance of all these effects depends generally on the distance and therefore on the tunneling conditions [5-8]. 

The useful temperature range of CBT is defined by the number and size of the tunnel junctions. These are realized by vacuum evaporation of 100nm A1 layers (which are... 

T(E) spectrum. When the Fermi level EF is located between the D-HOMO and the A-LUMO resonances, a large rectification effect is observed where T(EF) reaches almost 104. At a low 100 mV bias voltage and in a forward polarity, the tunnel current intensity reached around 1 nA. The T(E) spectrum of Fig. 2b was calculated using the ESQC technique associated with a semiempirical description of the tunnel junction [110]. The full valence MO structure of the junction is taken into account in the calculation. 

The lower trace in Figure 1 shows the results of heating the tunnel junctions (complete with a lead top electrode) in a high pressure cell with hydrogen. It is seen that the CO reacts with the hydrogen to produce hydrocarbons on the rhodium particles. Studies with isotopes and comparison of mode positions with model compounds identify the dominant hydrocarbon as an ethylidene species (12). The importance of this observation is obviously not that CO and hydrogen react on rhodium to produce hydrocarbons, but that they will do so in a tunneling junction in a way so that the reaction can be observed. The hydrocarbon is seen as it forms from the chemisorbed monolayer of CO (verified by isotopes). As... 

Figure 2. Differential spectra of CO chemisorbed on alumina-supported Co particles both before and after heating in hydrogen to 415 K. The chemisorbed CO is seen to react and form hydrocarbons in the tunnel junction. This hydrocarbon species is distinct from that formed on Rh as seen by vibrational modes near 1600...

FIG. 2. The characteristics of the tunnel junction are sensitive to the effects of adsorbates. The effect of one or two Xe atoms on the point of contact resistance is clearly shown. Constant current scans over individual adsorbate indicate that the effective diameter of the Xe atom on Ni(llO) is 0.19 nm. (From Ref. 42.)... 

FIG. 12. A schematic representation of the possible role of water on the potential profile within the tunnel junction. Fast electronic polarization of the solvent diminishes the barrier, while the possibility of forming an intermediate hydrated electron resonant state has also been suggested. (From Ref. 96.)... 

In addition to the physical interactions described above, the tip may also be used to alter the local chemical conditions within the tunnel junction. For example, catalytic rehydrogenation of carbonaceous fragments on Pt(lll) by tip-directed production of atomized hydrogen in vacuum at the Pt-Ir tip has been described [524]. Similar modification schemes may also be envisioned based on limiting the transport of reactants and products into or away from the partly occluded tunnel junction. As noted earlier, such effects may be important in the study of electrodeposition and etching process [126-131]. Nonetheless, much remains to be understood about the detailed physics and chemistry of the immersed tunnel junction. 

The noise of an actual resistance is always higher than the theoretical limit. While for metal resistors the noise level is close to the theoretical limit, the noise level in carbon resistors is much higher. The resistance of the tunneling junction, which is parallel to the feedback resistor, should be taken into account when its value is comparable to that of the feedback resistor. 

Similar to the tunneling junction experiments, the choice of materials is crucial. In the tunneling junction experiment, the existence of an excellent insulator AI2O3 and the ease of growing it on Al surfaces is the key to its... 

In the perturbative "transfer Hamiltonian approach developed by Bardeen 58), the tip and sample are treated as two non-interacting subsystems. Instead of trying to solve the problem of the combined system, each separate component is described by its wave function, i tip and i/zj, respectively. The tunneling current is then calculated by considering the overlap of these in the tunnel junction. This approach has the advantage that the solutions can be found, for many practical systems, at least approximately, by solution of the stationary Schrodinger equation. 

It is not always possible to discern non-elastic tunneling on the background of elastic tunneling. For example, when it proves possible, Fig. 14 shows the dependence of d2J/d V2 on V for the tunnel junction bearing H20 molecules as impurity. Tunnel spectroscopy is, at present, one of the powerful methods of studying the vibrational spectra of molecules in condensed media. 

Finally, a counter electrode (usually Pb) is deposited in a similar way to the base electrode thus completing the tunnel junction. (A fourth dosing method is sometimes used to introduce molecules onto the oxide surface after evaporation of the lead electrode. This method, known as infusion dosing [20-22], is used primarily for compounds that interact weakly with the surface). Figure 3 shows... 

One simple explanation for these results was as follows The electric field at a metal vacuum interface can be >10 times larger than in free space when the conditions required for a surface plasma resonance are met (47). Since the Raman cross-section is proportional to the square of the field, surface plasmons could produce enhancements of >10. This enhancement is probably not large enough to explain the tunneling junction results by itself, but an enhancement in signal of a factor of 100 by the excitation of surface plasmons would increase the Raman intensity from near the limits of detectibility. 

The necessity for a counter-electrode in IETS may be a handicap in some applications, but it can also be used to advantage. McBride and Hall let the surface reaction they were studying proceed to a certain point, quenched it, and then completed the tunneling junctions and ran their spectra. Jaklevic, and Kroeker, Kaska, and Hansma allowed the catalytic reaction to proceed within the junction,... 

---