Germanium redox reactions

The two types of electron transfer in a redox reaction at semiconductors can be distinguished by a number of experimental methods (12,13,14). The mechanisms of some redox reactions at germanium electrodes are summarized In Table I. It is seen that the mechanism of redox reactions with positive normal potentials is associated with the valence band, whereas the mechanism of redox reactions with more negative normal potentials is associated with the conduction band if there is any reaction at all. The situation remains the same when the electrode is moderately polarized in the anodic or cathodic direction. An example is shown in Fig. 11 using a redox system with properties equivalent to those assumed In Fig. 10. 

Redox reactions can be studied at germanium electrodes, undergoing no rapid dissolution or hydrogen evolution, only in a region of polarization over approximately 0.9 volts. Under high currents the assumptions of electronic equilibrium in the interior of the semiconductor are not valid (compare the limited hole current at n-type germanium in the region of dissolution). 

We have not yet studied the Influence of inhibitors on germanium electrodes. The problem is that germanium dissolves in parallel with most redox reactions at germanium electrodes. The dissolution causes a steady cleaning of the surface which should hinder the effect of contamination. The situation is very complicated in the case of cathodic polarization where the surface is covered with hydrogen atoms which act in some respects as inhibitors for others like catalysts. 

When the redox potential was high enough we found no difference between p- and n-type material. Nitric acid is not a very convenient system to study because its redox reaction is in itself very complicated. The reduction scheme for nitric acid contains a step which appears to be catalyzed at the surface. The catalytic activity of a germanium surface is probably influenced by the doping perhaps this could explain your results. We did not find any difference in p- and n-type when we used ferri-cyanide or ceric salts as oxidizers. 

The adsorption of probe molecules followed by different techniques allows one to prove the metal-metal interaction for catalysts prepared by redox reactions. For example, by chemisorption measurements, a decrease in the total amount of adsorbed H2 was observed with an increasing germanium content introduced by catalytic reduction on parent rhodium catalysts [81]. As TEM characterization showed a comparable mean particle size for all the catalysts, such evolution suggests that Ge covers the Rh surface [41]. These results were consistent with those... 

Numerous studies were devoted to the use of bimetallic catalysts, promoted either by an active metal such as rhenium or iridium, or an inactive one such as tin or germanium. These different catalysts were generally prepared by co-impregnation or successive impregnations. Moreover, most of the catalysts prepared by redox reactions were only evaluated in model reactions. For example, Corro et al. [87] prepared Pt-Sn/Al203-Cl catalysts by catalytic reduction or co-impregnation and compared their resistance to coking under cyclopentane feed. Thus, catalysts prepared by the surface redox reaction were less sensitive than the others to deactivation. This result was explained in terms of a more effective interaction between... 

This method for the detection of germanium is only decisive in the absence of certain other materials. Apart from compounds which reduce molybdates directly—e.g., Sn, Fe, As, and Se —arsenic acid, phosphoric acid and silicic acid should not be present, as they also form heteropoly molybdic acids, which enter into the same redox reaction with benzidine. The germanium can, however, be distilled out of hydrochloric acid solution (3.5-4.0 N) as Ge v chloride. The molybdate-benzidine test is then carried out with the distillate. 

On metal electrodes, the transfer coefficients typically approach 0.5. Generally, the transfer coefficients for redox reactions on moderately doped diamond electrodes are smaller than 0.5 their sum a +p, less than 1. We recall that an ideal semiconductor electrode must demonstrate a rectification effect in particular, on p-type semiconductors, reactions proceeding via the valence band have the transfer coefficients a = 0, P = 1, and thus, a +p = 1 [7]. Actually, the ideal behavior is rarely the case even with single crystal semiconductor materials manufactured by use of advanced technologies ( like germanium, silicon, gallium arsenide, etc.). The departure from the ideal semiconductor behavior is likely to be caused by the fact that the interfacial potential drop appears essentially localized, even in part, in the Helmholtz layer, due, e.g., to a high density of surface states, or the surface states directly participate in the electrochemical reactions. As a result, the transfer coefficients a and p have intermediate values, between those characteristic of semiconductors (O or 1) and metals (-0.5). Semiconductor diamond falls in with this peculiarity. However, for heavily doped electrodes, the redox reactions often proceed as reversible, and the transfer coefficients approach 0.5 ( metaMike behavior). 

---