Search

In this paper we report on galvanostatic pore formation on both n-InP and n-GaAs in acidic liquid ammonia (223 K). Quantitative analyses by atomic absorption have revealed two different dissolution mechanisms (involving 6 or 8 holes) without secondary reaction. Electrochemical behaviours have also shown diverse evolutions. Potential oscillations were observed onto InP while a constant voltage is measured for GaAs. Distinguished pore morphologies observed by scanning electron microscopy have confirmed the different porosification processes.

Porous structures have been anodically grown onto n -InP (100) in HCl. Surface chemistry and pore morphology have been respectively studied by X-ray photoelectron spectroscopy and scanning electron microscopy but in situ electrochemical investigations have been emphasized. Capacitance and photocurrent experiments have been performed while the porous etching occurred. No modifications of the electronic structure (flat band potential remains constant) are observed during the pore formation. However, for high dissolution charges, the photocurrent spectra are sharpened because the porous films behave like electrically “dead layers”. (© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim)

The linear dependency of the n-InP photocurrent on the level of illumination is indepedent of the solution composition, the type of illumination, the exposed crystalline face, or the level of doping. The photodissolution mechanism (for photocurrents between 1.1 and 65 mA cm−2) utilizes six charges to dissolve one molecule. In the absence of an insoluble oxide layer, the ratio between photogenerated holes and injected electrons is independent of the exposed crystalline face, the photon flux, the solution composition and the level of doping. Solutions containing oxygen bearing mineral anions (such as ClO−4, H2PO−2, or SO−24) are predisposed to the formation of an insoluble oxide layer on the surface of the InP under strong illumination.

The use of buffered neutral liquid ammonia medium for the first time provided evidence for a current doubling effect on p-InP during oxygen photoreduction. Unlike results in aqueous medium, a common mechanism of O2 reduction was observed at p-InP and p-GaAs electrodes. When protons were added to the solution, two different oxygen reduction mechanisms occurred at these electrodes. This study emphasizes the important results of the hydrogenated p-InP surface and revealed that liquid ammonia (at 223 K) was perfectly appropriate to understand the mechanism of O2 reduction at InP electrodes.

The influence of the supporting electrolyte on the decomposition mechanisms of an electrode (n-ZnSe) whose behaviour is well-known in aqueous media has been investigated in liquid ammonia (-60 °C) and in liquid ammoniate of sodium iodide (room temperature). In each case, the flat band potential of the ZnSe-electrolyte junction was determined unambiguously giving the following values: −2.3 V/MSE in acidic aqueous medium; −2.2 V/SRE (-60°C) in liquid ammonia and −2.2 V/SRE in neutral ammoniate. For low concentrations of residual water (CH2O ≤ 10−2 M), a cathodic mechanism of decomposition is evidenced. In liquid ammonia, it takes place below −2.0 V/SRE (-60°C) and in neutral ammoniate below −2 V/SRE. In the latter solvent, when CH2O increases, this decomposition is changed for the reduction of water. At positive potentials, in the dark and under illumination, whereas the semiconductor decomposes in aqueous medium, in liquid ammonia and in ammoniate it is the solvent which decomposes, the electrode remaining unmodified.

This work shows the advantages of using a non-aqueous solvent such as ammonia, which can be used at cryogenic temperatures and can be prepared anhydrously, to control the state of a silicon electrode surface in contact with the electrolyte. Depending on the in-situ experimental conditions, it was possible to change the oxidized material surface (native oxide) to an oxide-free surface and conversely. An original technique for pretreatment of the electrodes in a solvated electron solution in ammonia was used to corroborate the results obtained with the usual treatment, HF under an inert atmosphere. The influence of a superficial oxide layer and the growth of this layer under various conditions, was studied carefully by following the variations of two experimental parameters: the flatband potential and the photocurrent onset potential. By changing the pH, in liquid ammonia, the growth of the superficial oxide layer could be completely controlled; moreover, the oxide could be obtained either by a chemical reaction with traces of water or by a photoelectrochemical route.

The influence of the supporting electrolyte on the decomposition mechanisms of an electrode (n-ZnSe) whose behaviour is well-known in aqueous media has been investigated in liquid ammonia (-60 °C) and in liquid ammoniate of sodium iodide (room temperature). In each case, the flat band potential of the ZnSe-electrolyte junction was determined unambiguously giving the following values: −2.3 V/MSE in acidic aqueous medium; −2.2 V/SRE (-60°C) in liquid ammonia and −2.2 V/SRE in neutral ammoniate. For low concentrations of residual water (CH2O ≤ 10−2 M), a cathodic mechanism of decomposition is evidenced. In liquid ammonia, it takes place below −2.0 V/SRE (-60°C) and in neutral ammoniate below −2 V/SRE. In the latter solvent, when CH2O increases, this decomposition is changed for the reduction of water. At positive potentials, in the dark and under illumination, whereas the semiconductor decomposes in aqueous medium, in liquid ammonia and in ammoniate it is the solvent which decomposes, the electrode remaining unmodified.

Band-edge shifts of the n-GaP/electrolyte junction, when the pH and solution redox potential vary, are demonstrated in situ. Large variations (14 pH units and more than 2 V) of these parameters are obtained by using liquid ammonia solutions (neutral and basic) in the presence and absence of solvated electron species. The semiconductor band-edge position was deduced from Mott-Schottky curves in the following ammonia solutions: neutral, solvated electron, basic, basic + solvated electron, basic and basic + solvated electron. The transformation of a solution to the next one was obtained in situ, i.e. without removing the electrode from the solution. Large band-edge shifts could be demonstrated, up to ΔVfb = −1.8 V between the first solution (neutral) and the last one (basic + solvated electron). A reversible band-edge shift was observed between the basic and basic solvated electron solutions. The photoelectrochemical study revealed classical semiconductor/electrolyte behaviour with large photovoltages (up to 1.75 V in the solvated electron solution). The GaP/solvated electron junction behaviour gives some information to the more general investigation concerning the contact of a semiconductor and a redox system, the potential of which is located within the material conduction band.

The influence of the supporting electrolyte on the charge transfer mechanism at the semi-conductor/electrolyte interface is often primordial in the utilization of semiconducting electrodes in solar energy conversion. This work specifies the behaviour of a range of S.C. (SiSe, (III, V) and (II, VI) compounds) in a “non-aqueous” protonic solvent: the liquid ammoniate of sodium iodide (LASI). In each case, the flat band potential was unambiguously determined and the influence of solvated protons, NH+4 and water was studied.

Raman spectra and NH3 vapor pressure measurements indicate that LiClO4·4 NH3 is stable for use as electrolyte in Li batteries below 80 °C. Density, viscosity, and conductivity data are reported for the temp. range 10-80 °C. In this temp. range, the activation energy of ion transport remains mainly constant at 0.26 eV.