Photocathodes, p-type

Combinations of a photoanode (n-type) and a photocathode (p-type) of doped hematite have been studied [1, 20, 22], External bias was not needed for photo-oxidation of water. Both H2 and 02 were detected with this PEC diode assembly. However, the overall conversion efficiency was reported as low as 0.05% [1]. 

It is important to determine the conductivity and flat-band potential ( ft) of a photoelectrode before carrying out any photoelectrochemical experiments. These properties help to elucidate the band structure of a semiconductor which ultimately determines its ability to drive efficient water splitting. Photoanodes (n-type conductivity) drive the oxygen evolution reaction (OER) at the electrode-electrolyte interface, while photocathodes (p-type conductivity) drive the hydrogen evolution reaction (HER). The conductivity type is determined from the direction of the shift in the open circuit potential upon illumination. Illuminating the electrode surface will shift the Fermi level of the bulk (measured potential) towards more anodic potentials for a p-type material and towards more cathodic potentials for a n-type material. The conductivity type is also used to determine the potential ranges for three-electrode j-V measurements (see section Three-Electrode J-V and Photocurrent Onset ) and type of suitable electrolyte solutions (see section Cell Setup and Connections for Three- and Two-Electrode Configurations ) used for the electrochemical analyses. 

Near-zero electron affinity for p-type GaAs results from a fortuitous match between the work function of Cs and the bandgap of GaAs (both at 1.4 eV) fora case where the Fermi level in the bulk is located c below the bottom of the conduction band. This was observed prior to fabrication of the first NEA photocathode, p-type GaAs/Cs [5.66] after construction, the experimentally measured quantum efficiency per absorbed photon was in the order of 0.2 for energies above 2eV. A sharp response cuton at 1.4 eV gave a white light sensitivity of about 500 pA/lm - immediately superior to the classic S-20 surface [5.66]. 

The photoelectrochemical reduction of C02 at illuminated p-type semiconductor electrodes is also effective for C02 reduction to highly reduced products. The combination of photocathodes with catalysts for C02 reduction leads to a marked decrease in the apparent overpotential. At present, however, light to chemical energy conversion efficiencies are still very low, and negative in some cases. 

Direct splitting of water can be accomplished by illuminating two interconnected photoelectrodes, a photoanode, and a photocathode as shown in Figure 7.6. Here, Eg(n) and Eg(p) are, respectively, the bandgaps of the n- and p-type semiconductors and AEp(n) and AEF(p) are, respectively, the differences between the Fermi energies and the conduction band-minimum of the n-type semiconductor bulk and valence band-maximum of the p-type semiconductor bulk. lifb(p) and Utb(n) are, respectively, the flat-band potentials of the p- and n-type semiconductors with the electrolyte. In this case, the sum of the potentials of the electron-hole pairs generated in the two photoelectrodes can be approximated by the following expression ... 

In such devices the light-absorbing semiconductor electrode immersed in an electrolyte solution comprises a photosensitive interface where thermodynamically uphill redox processes can be driven with optical energy. Depending on the nature of the photoelectrode, either a reduction or an oxidation half-reaction can be light-driven with the counterelectrode being the site of the accompanying half-reaction. N-type semiconductors are photoanodes, p-type semiconductors are photocathodes, and... 

No naked semiconductor photocathode has been demonstrated to have good kinetics for the evolution of H2, despite the fact that the position of Eqb in many cases has been demonstrated to be more negative than E° (H2O/H2). This means that electrons excited to the conduction band have the reducing power to effect H2 evolution, but the kinetics are too poor to compete with e - h+ recombination. The demonstration that N,N -dimethyl-4,4 -bipyridinium, MV2+, could be efficiently photoreduced at illuminated p-type Si to form MV+ in aqueous solution under conditions where E° (MV2+/+) = E° (H20/H2) when no H2 evolution occurs establishes directly that the thermodynamics are good, but the kinetics are poor, for H2 evolution.(23,47)... 

Scheme V. Representation of the catalytic p-type Si photocathode for Ht evolution prepared by derivatizing the surface first with Reagent III followed by deposition of approximately an equimolar amount of Pd(0) by electrochemical deposition. The Auger/depth profile analysis for Pd, Si, C, and O is typical of such interfaces (49) for coverages of approximately 10 8 mol PQ2 /cm2.

Fig. 3.22 (a) Current - voltage characteristics of a p-type photocathode and that when it is replaced by a platinum shown to illustrate the efficiency calculation using the power saving approach. The shaded area represents the maximum power saving as a result of photoelectrolysis, (b) The efficiency at various photocurrent densities obtained useing the base graph. 

Here AVmax = Vsave(max) represents the difference in voltages at the semiconductor electrode and metal electrode at the maximum power conversion point. For example, in their experiment using a p-type InP photocathode, Heller and Vadimsky [120] obtained a current 23.5 mA/cm at maximum power point. A voltage of 0.1 IV vs SCE was applied in the case of InP electrode and -0.33V vs SCE in the case of platinum electrode, to obtain this current. Thus, the maximum saved voltage AVmax= 0.11( 0.33) V= 0.43V. Therefore, Psaved=0.43 V X 23.5 mA/cm = lO.lmW/cm. As they used a solar illumination of 84.7 mW/cm, the efficiency is 11.9%. 

Like other non-oxidic semiconductors in aqueous solutions, surface oxidized and photocorrosive InP is a poor photoelectrode for water decomposition [19,27,32,33], To enhance properties several efforts have focused on coupling of the semiconductor with discontinuous noble metal layers of island-like topology. For example, rhodium, ruthenium and platinum thin films, less than 10 nm in thickness, have been electrodeposited onto p-type InP followed by a brief etching treatment to achieve an island-like topology on the surface [27,28]. In combination with a Pt counter electrode, under AM 1.5 illumination of 87 mW/cm the metal (Pt, Rh, Ru) functionalized p-InP photocathodes [27] see a reduction in the threshold voltage for water electrolysis from 1.23 V to 0.64 V, and in aqueous HCl solution a photocurrent density of 24 mA/cm with a photoconversion efficiency of 12% [27]. 

Dominey RN, Lewis NS, Bruce JA, Bookbinder DC, Wrighton MS (1982) Improvement of photoelectrochemical hydrogen generation by surface modification of p-type silicon semiconductor photocathodes. J Am Chem Soc 104 467-482... 

At the n-type interface, the electric field generated causes photogenerated conduction band electrons to move into the bulk of the semiconductor, to the back metal contact, and into the external circuit. The valence band holes access the semiconductor interface due to the influence of the interfacial electric field (Fig. 28.2). Thus, redox species can be oxidized by the excited n-type semiconductor. These materials act as photoanodes. On the other hand, the electric field in a p-type material is reversed in potential gradient therefore, excited electrons move to the semiconductor surface, while holes move through the semiconductor to the external circuit (Fig. 28.2). These materials are photocathodes. The presence of an electric field at the semiconductor-electrolyte interface is usually depicted by a bending of the band edges as shown in Figure 28.2. Elec-... 

Takahashi and co-workers (69,70,71) reported both cathodic and anodic photocurrents in addition to corresponding positive and negative photovoltages at solvent-evaporated films of a Chl-oxidant mixture and a Chl-reductant mixture, respectively, on platinum electrodes. Various redox species were examined, respectively, as a donor or acceptor added in an aqueous electrolyte (69). In a typical experiment (71), NAD and Fe(CN)g, each dissolved in a neutral electrolyte solution, were employed as an acceptor for a photocathode and a donor for a photoanode, respectively, and the photoreduction of NAD at a Chl-naphthoquinone-coated cathode and the photooxidation of Fe(CN)J at a Chl-anthrahydroquinone-coated anode were performed under either short circuit conditions or potentiostatic conditions. The reduction of NAD at the photocathode was demonstrated as a model for the photosynthetic system I. In their studies, the photoactive species was attributed to the composite of Chl-oxidant or -reductant (70). A p-type semiconductor model was proposed as the mechanism for photocurrent generation at the Chi photocathode (71). 

Taking a general view of the above studies, we note that Chl-coated metal (platinum) electrodes commonly function as photocathodes in acidic solutions, although the photocurrent effcien-cies tend to be lower compared to systems employing semiconductors. This cathodic photoresponse may arise from a p-type photoconduc-tive nature of a solid Chi layer and/or formation of a contact barrier at the metal-Chl interface which contributes to light-induced carrier separation and leads to photocurrent generation. 

Bokris and Uosaki (1) have studied transient photo-assisted electrolysis current for systems including a p-type semiconductor photocathode and dark Pt anode. A set of current vs. time scans taken with a ZnTe photocathode system is shown in Figure 6. 

A review of photo-assisted electrolysis studies performed with p-type semiconductor photocathode/dark Pt anode systems suggests that a complementary phenomena arising from the presence of OH ions produced during the reduction half-cell reaction,... 

---