Mott-Schottky measurements

According to measurements of the space charge capacity dependent on the electrode potential (Mott-Schottky measurements, see Section 5.3), any variation of the electrode potential leads usually only to a corresponding change of the potential,... 

The first experiments were reported by Nozik and co-workers, for p-GaP and p-InP liquid junctions [95, 96], In particular, InP was a good candidate, because of its high electron mobility. The authors used p-nitrobenzonitrile (t/redox = -0.86 V (SCE)) as an electron acceptor, because the standard potential of this redox couple occurs 0.44 eV above the conduction band as determined by Mott-Schottky measurements. Photocurrent-potential curves in blank solutions were compared with those of solutions containing nitrobenzonitrile. The observation of increased cathodic photocurrent was reported as evidence for hot electron transfer. 

Figure 6 shows Mott-Schottky measurements in 1 M HF + 1 M HC l solutions with or without 5x10 4 M CTiCli. It is seen that the addition of 1 M HCI has no effect on the position of the Si bandedges. Upon addition of 5xl0"4 M C11CI2 there is also no shift of the flatband potential of 11-type Si. This follows from the fact that the reduction of copper ions in this solution occurs by electron capture from the conduction band and not by hole injection. 

From the Mott-Schottky measurements it follows that the change of the reduction mechanism for copper ions from a valence band to a conduction band mechanism by the addition of 1 M HC l can not be attributed to a shift of the position of the bandedges of the Si. Therefor we suggest that the addition of I M HCI results in the formation of cupric and cuprous chloride species resulting in a shift of the redox Fermi level. 

Obviously, such so-called frequency dispersion hampers a proper determination of Vfb although reliable values are generally obtained at high measuring frequencies (>10kHz). The origin of this nonideal behavior is not well understood [37, 41]. However, to check the rehabihty of Mott-Schottky measurements, the capacitance should be measured in a broad frequency range [42-44]. 

Fig. 3.15 Schematic diagram of a frequency response analyzer (FRA), showing the proper connections to the potentiostat for impedance spectroscopy and Mott-Schottky measurements. The function of the quasi-reference electrode is explained in Sect. 3.6.5...

Under certain conditions it may still be possible to determine the donor density of a porous photoelectrode from a Mott-Schottky measurement. This is the case for high donor densities, when the space charge width is small and is therefore stiU able to track the surface contours. An example of this for nanostructured Si-doped Qt-Fe203 photoanodes is shown in Fig. 3.18, reported by Cesar et al. [57], The actual surface area was estimated by dye absorption experiments, and the estimated donor density for this system was 10 ° cm . The concave shape of the Mott-Schottky plot is consistent with a gradual decrease in the effective surface area (cf. (3.5)) as the potential is increased and the depletion layer progressively penetrates into the bulk [58]. 

A particular class of electrode materials is represented by the transition metal chalcogenides, such as n-WSe2, n-MoSe2, and others, which form layer crystals. As already mentioned in Section 8.1.3, the basal planar surfaces of these electrodes (perpendicular to the c-axis) are relatively stable. In consequence, holes created by light excitation, are not transferred and accumulate at the surface. This leads to a large downward shift of the energy bands, as found by Mott-Schottky measurements [53] and as illustrated in Figure 8.20b (left and middle). The pho-... 

These are important examples because certain photoreduction processes may only be achieved with small particles of a given material. This has been demonstrated for Hj evolution at 50-A PbSe and HgSe colloids in the presence of a hole scavenger such as EDTA or S , which has not been observed with large particles [66]. It should be emphasized, however, that this method is not very accurate compared with Mott-Schottky measurements at compact semiconductor electrodes (see Section 5.2), because the reorganization energy is not known and several kinetic factors may influence the results. 

As outlined above, electron transfer through the passive film can also be cmcial for passivation and thus for the corrosion behaviour of a metal. Therefore, interest has grown in studies of the electronic properties of passive films. Many passive films are of a semiconductive nature [92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102 and 1031 and therefore can be investigated with teclmiques borrowed from semiconductor electrochemistry—most typically photoelectrochemistry and capacitance measurements of the Mott-Schottky type [104]. Generally it is found that many passive films cannot be described as ideal but rather as amorjDhous or highly defective semiconductors which often exlribit doping levels close to degeneracy [105]. 

C/pB estimated by both electrical (Mott-Schottky) and optical (photocurrent voltammetry) methods in the media studied, for (11 l)-oriented ZnSe electrode surfaces. A different variation was observed for the (110) orientation at pH >6. At pH 0, for both (110) or (11 l)-oriented electrode surface, the flat band potential value was -1.65 V (SHE) and the measured potential stability range (no detected current) was -0.35 to +2.65 V (SHE). A comparison of band levels with the other II-VI compounds as well as decomposition levels of ZnSe is given in Fig. 5.6. 

If VEB is increased, IEB increases and the current density at the electrode eventually becomes equal to JPS. It has been speculated that this first anodic current peak is associated with flat-band condition of the emitter-base junction. However, data of flat-band potential of a silicon electrode determined from Mott-Schottky plots show significant scatter, as shown in Fig. 10.3. However, from C-V measurement it can be concluded that all PS formation occurs under depletion conditions independent of type and density of doping of the Si electrode [Otl]. 

Straight lines plotted in the coordinates iph — < sc or iph —

Mott-Schottky plots C 2 —

depletion layer thickness Lsc on the potential. The straight line segments intersect at a single point. For materials studied in the works cited below, this point coincides, within 0.2 V, with the value of q>tb measured independently by the differential capacity technique. 

Figure 3. Mott-Schottky plots for the illuminated (lll)-face of n-GaP (pH 12.9, modulation frequency 1 kHz). The anodic photocurrents flowing during measurements are shown in units of pAcm 2.

The experiments were performed with single crystal (111) p-Si electrodes with a resistivity of about 5.5 ohm cm non-aqueous electrolytes were used consisting of absolute methanol containing tetramethylammonium chloride (TMAC) or acetonitrile containing tetraethyl ammonium perchlorate (TEAP). The flat-band potentials or p-Si in the two electrolytes were determined from Mott-Schottky plots (in the dark) in the depletion range of the p-Si electrode, from open-circuit photopotential measurements, and from the values of electrode potential at which anodic photocurrent is first observed in n-type Si electrodes. These three methods all yielded consistent flat-band potential values for p-Si of + 0.05V (vs SCE)... 

Electroreflectance spectra were measured for n-CdSe in the liquid junction configuration, and variations of the lineshape as a function of potential were observed. As the potential was reduced below the flatband potential, the electroreflectance signals changed sign. The potential at which this change occurs correlates well with the turn-on potential for light-induced photocurrent and with the intercept of the Mott-Schottky plot. 

Figure 5. Variation of Mott-Schottky intercept with pCl for (100) orientation n-GaAs, 40°C. Circles denote intercept values from automated admittance measurements. Bars signify standard deviation of least-squares straight line.

Fig. 11.12. The results of Hall measurements of mobility are shown in Table 11.3. The Mott-Schottky plot showed a flatband potential of -0.23 V on the NHS. Some electrode kinetic measurements (Miller, 1992) are shown in Fig. 11.13.

The effect of illumination seen in the current/potential behavior is reflected also in capacity measurements as evaluated in the form of Mott/ Schottky-plots (Fig. 2). Illumination leads to a parallel shift of this plot in the same direction and by about the same amount as in I/E curves. The plot is shifted back to its dark position if the appropriate redox couple is added. Other minority carrier acceptors on the other hand are not able to shift the light-plot back onto the plot obtained in the dark. 

Due to the great extension of the space-charge region, almost all the potential drop occurs across it. So we can measure its capacity, Csc, and calculate from the Mott-Schottky relation... 

---