Photoelectrochemical cells for hydrogen production from solar energy - Solar Energy Sciences and Engineering Applications - page 302

hematite's intrinsic electronic properties by varying the deposition method or by dop-
ing the photoelectrode with Si, Ti, Pt, Mo and Cr, among other atoms (Kay et al., 2006;
Satsangi et al., 2010; Glasscock et al., 2007; Hu et al., 2008; Kleiman-Shwarsctein
et al., 2008; Brillet et al., 2010). In an ideal hematite photoanode, the onset is just
anodic of the flat band potential with a photocurrent plateau of 12.6 mA cm − 2 (Tilley
et al., 2010). However, until the present, no research group has been successful in
splitting water by means an -Fe 2 O 3 photoanode without assistance of an external

bias voltage (Krol et al., 2008).
Another interesting material is BiVO 4 semiconductor, with a bandgap of 2.4-
2.5 eV and a reasonable band edge alignment with respect to water redox potentials;
in fact it has the ability to carry out the water photosplitting reaction (Sayama et al.,
2006). Moreover, it has been reported that BiVO 4 is able to show both semiconducting
properties, n- and p-types, (Vinke et al., 1992) as well as a high photon-to-current
conversion efficiencies ( > 40%) at 420 nm (Memming, 2001; Karakitsou and Verykios,
1993). Nevertheless, further improvements on its fundamental electronic structure
and stability are still needed (Sayama et al., 2006).
Over the past few years several efforts have been made in order to find an efficient
harvesting semiconductor under visible light. Cuprous oxide (Cu 2 O), which works
as a photocathode, has an interesting bandgap of 2.0-2.1 eV (Hara et al., 1998).
Theoretical calculations indicate that Cu 2 O can produce up to 14.7 mA cm − 2 , cor-
responding to a light-to-hydrogen conversion efficiency of 18% based on the AM
1.5 spectrum (Paracchino et al., 2011). For solar water-splitting purposes, Cu 2 O has
favorable energy band positions; the conduction band is located 0.7 V negative of the
hydrogen evolution potential with the valence band lying just positive of the oxygen
evolution potential (Paracchino et al., 2011). Since no overpotential is available for
oxygen evolution, the reduction band edge is close to the water reduction potential, the
p-type Cu 2 O can drive half of the water-splitting reaction but an external bias must be
applied to conduct the other half reaction (water oxidation) (Paracchino et al., 2011).
However, the limiting factor of this material is the poor stability in aqueous solutions,
since the redox potentials for the reduction and oxidation of monovalent copper oxide
lie within the bandgap (Hara et al., 1998; Paracchino et al., 2011). The corrosion
sensitivity issue of cuprous oxide under illumination can be addressed by depositing
very thin protective layers by e.g. atomic layer deposition (ALD). In fact, using this
methodology, Grätzel and co-workers have designed, up to now, the best performing
oxide photoelectrode, 7.6 mA cm − 2 at 0 V RHE , using Cu 2 O electrodes protected with
nanolayers of Al-doped zinc oxide and titanium dioxide activated for hydrogen evolu-
tion with electrodeposited platinum nanoparticles, i.e. Cu 2 O was coated with layers of
n-type oxides with structure 5

α

× (4 nm ZnO/0:17 nm Al 2 O 3 )/11 nm TiO 2 (Paracchino

et al., 2011).

10.2.3.2 Non-oxide semiconductor

Non-oxide semiconductors (p-type and/or n-type photoelectrodes) are known to effi-
ciently harvest sunlight, converting it into electricity: amorphous, polycrystalline and
crystalline silicon (a-Si, p-Si and c-Si), gallium arsenide (GaAs), cadmium telluride
(CdTe), gallium phosphide (GaP), indium phosphide (InP), copper indium dise-
lenide (CIS), copper indium gallium diselenide (CIGS) and gallium indium phosphide