Anodic Photocorrosion Research Articles

Photoelectrochemical evolution of elemental fluorine at titanium dioxide electrodes in anhydrous hydrogen fluoride solutions

Although elemental fluorine has been known for more than 100 years, it is still produced by Moissan's original method, electrolysis of anhydrous HF/alkali fluoride solutions. Only one convenient alternative route for preparing fluorine has been reported to date. Several oxide and oxyfluoride semiconductors have valence band edges that are positive of the F/sup -/, F/sub 2/ standard potential, and holes photogenerated in the valence band of these materials should, in principle, be capable of oxidizing F/sup -/ to F/sub 2/. In aqueous fluoride-containing solutions other processes such as water oxidation and photoanodic corrosion compete effectively with F/sup -/ oxidation, and no fluorine is formed. Even in nonaqueous solutions these wide bandgap semiconductors photocorrode, unless sufficient reductant is present to compete for photogenerated holes. In anhydrous solutions containing sufficient quantities of fluoride ion is reasonable to expect that fluorine might be generated photoelectrochemically at a semiconductor surface which is somehow stabilized against photocorrosion. They report here that elemental fluorine is evolved at n-type TiO/sub 2/ electrodes in anhydrous HF/NaF solutions. Using bandgap illumination (365 nm) gaseous F/sub 2/ is produced with a sizable photovoltage and high power conversion and current efficiencies. While electrode photocorrosion occurs to some extent in all thesemore » experiments, it may be minimized greatly by using fluorine-doped TiO/sub 2/ electrodes.« less

Photochemistry of semiconductor colloids. 22. Electron ejection from illuminated cadmium sulfide into attached titanium and zinc oxide particles

Colloidal solutions of CdS containing colloidal TiO, or ZnO were illuminated with visible light. The fluorescence of CdS (decay time -50 ns) was quenched by Ti02, several Ti02 particles being required per CdS particle. The rate of photoanodic corrosion in aerated solution was drastically increased in the presence of Ti02 In deaerated CdS solutions containing methanol and Cd2+ ions, cadmium metal was formed when Ti02 was present. Methyl viologen was reduced with a quantum yield of close to one, while it reacted about ten times more slowly in the absence of Ti02 These effects are explained in terms of improved charge separation by rapid electron injection from illuminated CdS into the conduction band of attached Ti02 particles. Electron injection into ZnO was less efficient and occurred only in the case of Q-CdS particles (very small particles having a greater band gap). The injected electrons caused a blue shift of the absorption threshold of ZnO. The photocatalytic action of colloidal or suspended semicon- ductor particles is based on the generation of electrons and positive holes which rapidly move to the surface of the particles and initiate redox processes. The efficiency of charge separation is often increased by contacting the semiconductor particle with a metal or another semiconductor. Typical examples are platinized ti- tanium dioxide' and cadmium sulfide2 as well as Ru02-covered Ti02.3 Serpone et al. reported a few years ago that H2 was formed from H2S on CdS powder illuminated with visible light in aqueous solution and that the yield was slightly increased in the presence of Ti02 powder! The effect was explained by an improved charge separation due to electron transfer from the illuminated CdS particles into the conduction band of the Ti02 particles. The increase in yield was only 20%, Le., little above the increase which could be explained by more efficient light absorption of CdS due to the increased internal light scattering by the TiO, additive. In the present paper, experiments with transparent colloidal solutions of CdS containing colloidal Ti02 or ZnO as additives are described. Efficient electron injection from the excited CdS part of the sandwich colloids to the Ti02 or ZnO part was observed with three methods of observation: (1) With use of a CdS colloid that fluoresces with a high quantum yield, the quenching of the fluorescence by added TiO, or ZnO was studied. (2) Redox processes, such as the reduction of excess Cd2+ ions and of methyl viologen and the photoanodic dissolution of CdS, were initiated by visible light illumination and the influence of added Ti02 investigated. (3) In the case of ZnO as additive, the electron injection was accompanied by the typical changes in the absorption spectrum of ZnO which have recently been observed in other experiments on the deposition of excess electrons on small semiconductor particles.s-6

Photochemistry of colloidal semiconductors. 20. Surface modification and stability of strong luminescing CdS particles

The preparation of CdS sols with a mean diameter between 40 and 60 A and a relatively narrow size distribution is described. The colloids could be separated as solids, which in turn could be redissolved to give solutions of some 10/sup -2/ M. Activation of the particles by a cadmium hydroxide precipitate yielded fluorescing samples with a quantum yield exceeding 50%. The blue or green fluorescence occurred close to the band gap energy, which depended on the size of the particles. Violet fluorescing samples of activated ZnS-CdS co-colloids were also prepared. Photoanodic corrosion measurements showed that the activated CdS colloid was 2000 times more stable than the nonactivated one. Under laser illumination the particles became a little larger which is ascribed to an acceleration of Ostwald ripening due to local heating. Extremely intense laser light decomposed the particles into Cd + S.

Preparation of colloidal silicon and preliminary photochemical experiments

A photochemically active powder of silicon is obtained by the slow combustion of silane. The powder contains 33% oxygen and is soluble, to a large extent, in water and alcohols. Its X-ray diffraction pattern indicates the presence of 1.5 nm aggregates of nontetrahedral structure. Illumination of the Si particles in the presence of air leads to photo-anodic corrosion. In deaerated solutions, Ag + and methylviologen ions are reduced, the respective quantum yields in 3 × 10 −3 M solution being 0.1 and 0.02%.

Fundamental problems of water splitting at cadmium sulfide

Oxygen is directly involved in the anodic photocorrosion of CdS leading to sulfate formation according to the overall reaction CdS + 4h + + 2H 2O + O 2 → Cd 2+ + SO 2− 4 + 4H +. Sulfate formation leads to cleaning of the CdS surface from elemental sulfur and therefore improves the electrode behaviour dramatically. Removal of sulfur from the surface shifts the flatband potential of CdS to about −1.7 V(SCE) and the valence-band edge to −0.6 V(SCE). Accordingly holes lose their thermodynamic capability to oxidise water.

Ag + exchange at Cds electrode surfaces

The inhibition effect of Ag + ions on the photoanodic corrosion of CdS semiconductor electrodes has been studied in 0.5 M NaClO 4 + xM AgClO 4 solutions (0 ⩽ x ⩽ 10 −2, pH = 2) by means of potentiostatic current density-potential measurements, ac-impedance and atomic absorption. In the presence of Ag + ions in the solution a thin Ag 2S film of ca. 30–300 nm thickness is formed on the CdS surface due to the heterogeneous cation exchange reaction CdS + 2Ag + = Ag 2S + Cd 2+. The Ag 2S film formation is a transport controlled process which can be described by a parabolic rate law. From kinetic investigations in the temperature range (298 ⩽ T ⩽ 348 K) the activation energy of this process was determined to be about 62 kJ mol −1. Information about the morphology and composition of the surfaces was obtained from optical and scanning electron microscope investigations including EDAX.

Photochemistry of Colloidal Metal Sulfides. 7. Absorption and Fluorescence of Extremely Small ZnS Particles (The World of the Neglected Dimensions)

AbstractExtremely small colloidal ZnS particles (diameter ∼ 1.7 nm) were made by either photo‐degradation of 3 nm particles or rapid precipitation in phosphate containing solution at pH = 7‐−8. The absorption spectra of these particles are different from that of macrocrystalline ZnS, and the changes are regarded as an indication for the transition from semiconductor ZnS to polymolecular ZnS with decreasing particle size. – The dependence of the intensity of the fluorescence on particle size, temperature, photoanodic corrosion and the quenching of fluorescence were investigated for colloids on a silicon dioxide carrier, in phosphate solution, and without a stabilizer. Photo‐anodic corrosion strongly improves the fluorescence properties. – One adsorbed Cd2+ ion per colloidal particle is sufficient for efficient quenching of the fluorescence. However, a new fluorescence band appears which is explained by the formation of a layer of 1:1 co‐colloid at the surface of the ZnS particles. – Methylviologen was also found to be a very efficient quencher. The decay of the fluorescence is wavelength dependent, i.e. the fraction of long‐lived fluorescence is greater at longer wavelengths. – A mechanism is discussed, where the fluorescence centers are anion vacancies, and fluorescence is emitted when electrons trapped in states of different energies and exhibiting different life‐times tunnel to the localized positive holes.

Studies on the stability of RuS2 single crystals and the photo-oxidation of halides

Anodic photocorrosion of n-RuS2 is very slow under moderate conditions but can be accelerated under extreme conditions: high potentials and rapid oxygen evolution. This stability is primarily thermodynamic, but may be kinetic to some extent. Like some other transition metal dichalcogenides, RuS2 shows an onset of photocurrent which is shifted strongly to lower potentials in the series Cl−, Br−, I−. This is related to improved kinetics of the oxidation reaction due to adsorption of the intermediate species.

Studies on the stability of RuS 2 single crystals and the photo-oxidation of halides

Anodic photocorrosion of n-RuS 2 is very slow under moderate conditions but can be accelerated under extreme conditions: high potentials and rapid oxygen evolution. This stability is primarily thermodynamic, but may be kinetic to some extent. Like some other transition metal dichalcogenides, RuS 2 shows an onset of photocurrent which is shifted strongly to lower potentials in the series Cl −, Br −, I −. This is related to improved kinetics of the oxidation reaction due to adsorption of the intermediate species.

Chemically derivatized semiconductor photoelectrodes

Abstract : Highlights of research results from the chemical derivatization of n-type semiconductors with (l,l'-ferrocenediyl)dimethylsilane, (1) and its dichloro analogue, (2) and from the derivatization of p-type semiconductors with (N,N'-bis(3-trimethoxysilyl)propyl)-4,4'-bipyridinium)dibromide, (3) are presented. Research Shows that molecular derivatization with (2) can be used to suppress photoanodic corrosion of n-type Si; derivatization of p-type Si with (3) can be used to improve photoreduction kinetics of horseheart ferricytochrome c; derivatization of p-type Si with (3) followed by incorporation of Pt(O) improves photoelectrochemical H2 production efficiency. Strongly interacting reagents can alter semiconductor/electrolyte interface energetics and surface state distributions as illustrated by n-type WS2/I(-) interactions and by differing etch procedures for n-type CdTe. (Author)

Surface and Redox Reactions at GaAs in Various Electrolytes

The behavior of at open circuit and under controlled potential in various redox electrolytes has been investigated by rotating ring‐disk electrode voltammetry and hydrodynamically modulated disk techniques. These methods established the reactions at the surfaces to be one or more of the following: oxidation or reduction of solution species, corrosion of by charge injection from an oxidant, photoanodic corrosion of , cathodic protection of against dissolution, and the participation of surface film products on in altering the possible reaction paths. In six redox systems in acid and base where surface films do not form, the reaction path at both can be well correlated with the relative position of solution redox potential and the bandedges of at the given pH per Gerischer theory. At intermediate pH, when film formation at the interface is a competitive path, further surface oxidation is suppressed and hole transfer to a soluble reductant (Fe+2‐EDTA) occurs efficiently, although not expected from potential level alone. In the case of , however, film thickness cannot be well enough controlled, either over time or potential range, to be suitable for the demands of photoelectrochemical cells.

Photoanodic properties of an n-type silicon electrode in aqueous solutions containing fluorides

The n-Si electrode shows two distinct characteristic, photoanodic corrosion properties in solutions with fluoride ion or hydrofluoric acid together with a reducing agent. In the presence of a low concentration of fluoride ion or hydrofluoric acid, the photocurrent is observed to decay more rapidly than in the absence of the fluorides. In the presence of a relatively high concentration of hydrofluoric acid such that the oxide dissolves rapidly, current doubling occurs, suggesting that the photoproduced holes all react to corrode the silicon rather than oxidizing the reducing agent. Models to explain the two-way influence of the fluoride ion in the photocorrosion of silicon are presented.

Design of a 13% efficient n-GaAs1−xPx semiconductor–liquid junction solar cell

We report here the design of the most efficient non-aqueous semiconductor–liquid junction solar cell studied to date. Our approach involves the use of ternary semiconductor electrodes made from solid solutions of a large hand gap material, GaP, and a small band gap material, GaAs. We demonstrate here that photoanodes consisting of such materials are capable of simultaneously yielding high open circuit voltages and favourable wavelength response to the solar spectrum. A few n-type semiconductor–liquid junction solar cells in aqueous solutions have been reported to yield high (>10%) solar-to-electrical conversion efficiencies1–3. However, for most materials, rapid photoanodic corrosion dominates the interfacial photochemistry4–8. Non-aqueous solvent systems can suppress electrode decay due to corrosion4,7,8; but modest (<6%) conversion efficiencies have been observed for all photoanodes studied in solar irradiation conditions9–13. The photoanodes used here yield over 13% solar-to-electrical conversion efficiencies, or more than double the efficiency of any other non-aqueous semiconductor–liquid junction solar cell previously reported.

Photoelectrochemical reaction behaviour of platinum disulphide with water and reducing agents

PtS 2 is a layer-type semiconductor (Δ E G ≅ 0.95 eV), similar to MoS 2 or WS 2, in which photoexcitation also produces holes in an energy band derived from transition metal d-states. There is consequently no anodic photo-corrosion leading to molecular sulphur. Since platinum cannot easily leave the electrode surface as a positive ion, the formation of metal sulphate is also inhibited. The main photo oxidation product in contact with an aqueous electrolyte is therefore oxygen, allowing the formation of platinum complexes as a possible side reaction, which occurs in presence of suitable complexing agents. The investigation reveals that, during anodic polarization, the PtS 2 surface passes through different oxidation states which affect the potential drop in the Helmhotz-layer and thus the flat band potential of the electrode. The photooxidation of redox systems also involves considerable changes in the electrode surface which are reflected in or relatively systematic shift of photocurrents in dependence of the redox potential of the electrolyte. Valuable new scientific informations on photoelectrochemical mechanisms can be obtained from this system.

Photoelectrochemical behavior of n-gallium arsenide electrodes in ambient-temperature molten-salt electrolytes

Photoelectrochemical (PEC) characterization of n-GaAs electrodes was carried out in room-temperature molten-salt electrolytes by using the aluminum chloride-butylpyridinium chloride (AlCl/sub 3/-BPC) system as a representative example. The working potential limits for the above electrodes in the melts, containing varying ratios of AlCl/sub 3/ and BPC, were established by cyclic voltammetry. Flat-band potential (V/sub fb/) measurements on n-GaAs in the same melts enabled location of the semiconductor band edge positions relative to the melt stability windows. In electrolytes containing AlCl/sub 3/ and BPC in the 1:1 molar ratio, the available range of potential was wide enough to probe the entire band-gap region. On the other hand, the potentials corresponding to the conduction band edges of n-GaAs were beyond the cathodic stability limit of both the 2:1 and 0.75:1 AlCl/sub 3/-BPC compositions. The electrode dissolution behavior of illuminated n-GaAs electrodes was investigated by cyclic voltammetry in melts of varying composition containing no intentionally added electroactive species. The onset of photoanodic corrosion currents was significantly positive of the values observed in aqueous electrolytes. The redox behavior of ferrocene-ferricenium ion couple (Fe(Cp)/sub 2//Fe(Cp)/sub 2//sup +/) was studied by cyclic voltammetry on vitreous carbon electrodes in the 2:1, 0.75:1, and 1:1 AlCl/sub 3/-BPC electrolytes. Underpotentialsmore » developed for the photoanodic process on n-GaAs relative to the reversible (dark) thermodynamic values on vitreous carbon were direct evidence for the sustained conversion of light energy to electrical energy.« less