In this study, the process of metal cation reduction on multi-crystalline silicon in a dilute hydrofluoric acid (HF) matrix is described using Ag(I), Cu(II), Au(III) and Pt(IV). The experimental basis utilized batch tests with various solutions of different metal cation and HF concentrations and multi-crystalline silicon wafers. The metal deposition kinetics and the stoichiometry of metal deposition and silicon dissolution were calculated by means of consecutive sampling and analysis of the solutions. Several reaction mechanisms and reaction steps of the process were discussed by overlaying the results with theoretical considerations. It was deduced that the metal deposition was fastest if the holes formed during metal ion reduction could be transferred to the valence bands of the bulk and surface silicon with hydrogen termination. By contrast, the kinetics were lowest when the redox levels of the metal ion/metal half-cells were weak and the equilibrium potential of the H3O+/H2 half-cells was high. Further minima were identified at the thresholds where H3O+ reduction was inhibited, the valence transfer via valence band mechanism was limited by a Schottky barrier and the dissolution of oxidized silicon was restricted by the activity of the HF species F−, HF2− and H2F3−. The findings of the stoichiometric conditions provided further indications of the involvement of H3O+ and H2O as oxidizing agents in addition to metal ions, and the hydrogen of the surface silicon termination as a reducing agent in addition to the silicon. The H3O+ reduction is the predominant process in dilute metal ion solutions unless it is disabled due to the metal-dependent equilibrium potential of the H3O+/H2 half-cell and the energetic level of the valence bands of the silicon. As silicon is not oxidized up to the oxidation state +IV by the reduction of the metal ions and H3O+, water is suspected of acting as a secondary oxidant. The stoichiometric ratios increased up to a maximum with higher molalities of the metal ions, in the manner of a sigmoidal function. If, owing to the redox level of the metal half-cells and the energetic level of the valence band at the metal–silicon contact, the surface silicon can be oxidized, the hydrogen of the termination is the further reducing agent.

Metal deposition on silicon in hydrofluoric acid (HF) solutions is a well-established process for the surface patterning of silicon. The reactions behind this process, especially the formation or the absence of molecular hydrogen (H2), are controversially discussed in the literature. In this study, several batch experiments with Ag+, Cu2+, AuCl4– and PtCl62– in HF matrix and multicrystalline silicon were performed. The stoichiometric amounts of the metal depositions, the silicon dissolution and the molecular hydrogen formation were determined analytically. Based on these data and theoretical considerations of the valence transfer, four reasons for the formation of H2 could be identified. First, H2 is generated in a consecutive reaction after a monovalent hole transfer (h+) to a Si–Si bond. Second, H2 is produced due to a monovalent hole transfer to the Si–H bonds. Third, H2 occurs if Si–Si back bonds of the hydrogen-terminated silicon are attacked by Cu2+ reduction resulting in the intermediate species HSiF3, which is further degraded to H2 and SiF62–. The fourth H2-forming reaction reduces oxonium ions (H3O+) on the silver/, copper/ and gold/silicon contacts via monovalent hole transfer to silicon. In the case of (cumulative) even-numbered valence transfers to silicon, no H2 is produced. The formation of H2 also fails to appear if the equilibrium potential of the 2H3O+/H2 half-cell does not reach the energetic level of the valence bands of the bulk or hydrogen-terminated silicon. Non-hydrogen-forming reactions in silver, copper and gold deposition always occur with at least one H2-forming process. The PtCl62– reduction to Pt proceeds exclusively via even-numbered valence transfers to silicon. This also applies to the reaction of H3O+ at the platinum/silicon contact. Consequently, no H2 is formed during platinum deposition.

In this study, a lab-scale plant was designed to treat water in continuous flow condition using non-thermal plasma technology. The core was an electrode system with connected high-voltage (HV) pulse generator. Its potentials and limitations were investigated in different experimental series with regard to the high-voltage settings, additions of oxygen-based species, different volume flow rates, and various physical-chemical properties of the process water such as conductivity, pH value, and temperature. Indigo carmine, para-Chlorobenzoic acid, and phenol were chosen as reference substances. The best HV settings was found for the voltage amplitude Û = 30 kV, the pulse repetition rate f = 0.4–0.6 kHz, and the pulse duration tb = 500 ns with an energy yield for 50% degradation G50, which is of 41.8 g∙kWh−1 for indigo carmine, 0.32 g∙kWh−1 for para-Chlorobenzoic acid, and 1.04 g∙kWh−1 for phenol. By adding 1 × 10−3 mol∙L−1 of oxygen, a 50% increase in degradation was achieved for para-Chlorobenzoic acid. Conductivity is the key parameter for degradation efficiency with a negative exponential dependence. The most important species for degradation are hydroxyl radicals (c ≈ 1.4 × 10−8 mol∙L−1) and solvated electrons (c ≈ 1.4 × 10−8 mol∙L−1). The results show that the technology could be upgraded from the small-scale experiments described in the literature to a pilot plant level and has the potential to be used on a large scale for different applications.

No abstract