A Monolithic Photovoltaic-Photoelectrochemical Device for Hydrogen Production via Water Splitting

10.1021/ar00051a007

H. Gerischer in Topics in Applied Physics vol. 31 Solar Energy Conversion B. O. Seraphin Ed. (Springer-Verlag Berlin 1979) pp. 115–172.

The ability of a semiconductor electrode to drive the electrochemical reaction of interest is determined by its band gap (the energy separation between the valence and conduction band edges) and the position of the valence and conduction band edges relative to the vacuum level (or other reference electrode). In contrast to metal electrodes semiconductor electrodes in contact with liquid electrolytes have fixed energies where the charge carriers enter the solution. This fixed energy is given by the energetic position of the semiconductor's valence and conduction bands at the surface (where these bands terminate at the semiconductor/electrolyte interface). The energetic position of these band edges is determined by the chemistry of the semiconductor/electrolyte interface which is controlled by the composition of the semiconductor the nature of the surface and the electrolyte composition. So even though a semiconductor electrode may generate sufficient energy to effect an electrochemical reaction the energetic position of the band edges may prevent it from doing so. For spontaneous water splitting the oxygen and hydrogen reactions must lie between the valence and conduction band edges and this is almost never the case.

G. E. Shakhnazaryan et al. Russ. J. Electrochem. 30 610 (1994).

E. Aharon-Shalom and A. Heller J. Electrochem. Soc. 129 2865 (1982).

10.1126/science.223.4641.1141

Kocha S., Turner J., Nozik A. J., J. Electroanal. Chem.367, 27 (1991).

S. Kocha and J. A. Turner J. Electrochem. Soc. 142 2625 (1995).

H. Yoneyama H. Sakamoto H. Tumura Electrochem. Acta 20 341 (1975).

Nozik A. J., Appl. Phys. Lett.29, 150 (1976).

R. C. Kainthla B. Zelenay J. O'M. Bockris J. Electrochem. Soc. 134 841 (1987).

Sakai Y., et al., Can. J. Chem.66, 1853 (1988).

Lin G., et al., Appl. Phys. Lett.55, 386 (1989).

J. Augustynski G. Calzaferri J. Courvoisier M. Graetzel in Proceedings of the 11th World Hydrogen Energy Conference T. N. Veziroglu C.-J. Winter J. P. Baselt G. Kreysa Eds. (DECHEMA Frankfurt Germany 1996) pp. 2378–2387.

K. A. Bertness et al. App. Phys. Lett. 65 989 (1994).

Bolton J. R., et al., Nature316, 495 (1985).

Kurtz S. R., et al., J. Appl. Phys.68, 1890 (1990).

K. A. Bertness et al. in AIP Conference Proceedings No. 306 R. Noufi and H. Ullal Eds. (American Institute of Physics New York 1994) p. 100.

O. Khaselev and J. A. Turner in preparation.

For electrochemical measurements the wafers were cleaved into samples ∼0.2 cm 2 in area. The samples with gold ohmic contacts were mounted on a Teflon-covered screw electrode with silver epoxy and heated to 80°C for 1 hour. The electrical contact was insulated from the electrolyte by an epoxy coating that also covered the samples' edges. A conventional two-electrode configuration was used with a platinum gauze or foil counter-electrode coated with ruthenium metal. Photoelectrochemical characteristics were measured with an EG&G 263A potentiostat. The electrolyte 3 M H 2 SO 4 was freshly prepared from deionized water having resistivity of 18 megohm/cm. All solutions were made of analytical-grade reagents. Before electrochemical measurements the samples were etched in 1:20:1 HCl:CH 3 COOH:H 2 O 2 solution. A single-compartment cell was used. Although the electrodes were spatially separated no attempt was made to separate the products of the anode and cathode reactions.

To reduce the overvoltage losses associated with the noncatalytic surface of the semiconductor a thin layer of platinum catalyst was electrochemically deposited on the surface of the semiconductor electrodes from a 20 mM H 2 PtCl 6 solution. Photoassisted galvanostatic deposition was performed at a cathodic current density of 1 mA/cm 2 with a platinum quantity corresponding to a charge of 10 mC/cm 2 . A recent publication (28) has shown that the platinum not only acts as a catalyst for hydrogen evolution but also drastically reduces the corrosion reaction of a similar III-V compound indium phosphide. We would expect that platinum would offer a comparable corrosion-inhibiting mechanism here.

We used concentrated light for two reasons: (i) The photocurrent and the amount of gases produced are higher and therefore easier to quantify and (ii) because of their cost the major terrestrial application of the solid-state analog of these tandem cells is in concentrator systems (>100 suns). Therefore any real application using the present solid-state design for photoelectrolysis would also use concentrated light. However the PEC systems will probably be limited to less than 100 suns because that would give rise to a current of about 1A/cm 2 a practical limit for electrolysis. The evolution of gas bubbles on the illuminated surface is also an issue as these bubbles will scatter light reducing the efficiency. Given this limitation as well as problems regarding the amount of catalyst and system design issues a more realistic limit as to the amount of light concentration that can be used for photoelectrolysis is in the range of 10 to 20 suns.

The gas products of the photoelectrolysis were analyzed with a UTI-100C mass spectrometer. A sealed cell was used connected directly to the high-vacuum chamber of the mass spectrometer via a gas inlet system. Before being sealed the cell was purged for several minutes with nitrogen to remove any oxygen. The cell was then sealed and the photoelectrolysis reaction was run for several hours. The external current was monitored with an ammeter. Periodically the gas mixture from the upper part of the cell was sampled by the mass spectrometer system. The same cell was used and the same experimental technique was repeated with the use of two platinum electrodes at the same external current. Within experimental error (∼±20%) the results were the same indicating stoichiometric water splitting.

For 100% electrolysis efficiency this calculation is identical to Bolton's calculation of efficiency (29). For the size of these samples and the amount of hydrogen and oxygen being generated a calculation using external current flow is far more accurate.

M. F. Weber and M. J. Dignam Int. J. Hydrogen Energy 11 225 (1986).

___ J. Electrochem. Soc. 131 1258 (1984).

S. Kocha D. Montgomery M. Peterson J. A. Turner Sol. Energy Mater. Sol. Cells in press.

P. Bogdanoff P. Friebe N. Alonso-Vante J. Electrochem. Soc. 145 576 (1998).

J. R. Bolton Solar Energy 57 37 (1996).

We thank S. Kurtz and J. Olson for the tandem cell samples and A. Dillon for help with the mass spectrometer measurements. This work was supported by the Hydrogen Program of the U.S. Department of Energy.