Photoelectrochemical cells for hydrogen production from solar energy - Solar Energy Sciences and Engineering Applications - page 303

Table 10.2.1 Non-oxide n-type semiconducting materials
with small bandgap (Grimes et al., 2008).

Semiconductor Bandgap

CdSe 1.7 eV

CdTe 1.4 eV

GaP 2.24 eV

GaAs 1.35 eV

InP 1.35 eV

MoS 2 1.75 eV

MoSe 2 1.5 eV

(GaInP 2 ) (Grimes et al. 2008). As mentioned, most of the oxide semiconductors stud-
ied for water-splitting show bandgap and stability issues, thus, particular attention
has been given to non-oxide materials since they have smaller bandgaps enabling the
capture of a larger portion of the solar spectrum energy.
Cadmium sulfide (CdS) has well positioned band edges to efficiently reduce and
oxidize water with an optical absorption of 520 nm (corresponding to a bandgap of
2.4 eV). However, it suffers from anodic photodecomposition by the photogenerated
holes (Grimes et al., 2008). Similar problems of photocorrosion can be found with
other n-type non-oxide semiconductors as the ones presented in Table 10.2.1.
Considering now p-type non-oxide materials, they usually show stable behav-
ior against cathodic photodecomposition since the photoelectrons in excess migrate
towards the semiconductor/electrolyte interface such as the p-Si and p-GaP (Grimes
et al., 2008). However, the flat band position of these materials is unfavorable for the
H 2 O/O 2 redox and a large bias voltage must be applied (Nozik, 1978).
One approach to overcome stability issues is covering the unstable photoelectrodes
with thin films of stable wide bandgap semiconductors with suitable band edges or with
thin metal films (e.g. by chemical vapor deposition (CVD), sputtering or atomic layer
deposition) (Nozik, 1978). Although single-photon systems seem to be the preferable
route to produce hydrogen from solar energy, either with oxide or non-oxide materials,
significant improvements should be accomplished on electronic structure and stability
in order to be used alone in PEC systems - Figure 10.2.2a and 2b.
Clearly, there are three routes that may result in a high efficient system for water-
splitting and without need of an additional bias. These are illustrated in Figures 10.2.2c
and 10.2.3a and 10.2.3c. All strategies share the feature of having two semiconductors
with different bandgaps. This provides a mechanism by which a single electron is
photoexcited twice and, correspondingly generates a larger bias from light. It has
been calculated that this type of system could realistically achieve a solar-to-hydrogen
conversion efficiency of 21.6% (Bolton et al., 1985).
The most compelling approach is however illustrated in Figure 10.2.2c and Fig-
ure 10.2.3a. Here, various combinations of n-type and p-type semiconductors, oxide
and non-oxide, such as n-TiO 2 /p-GaP, n-SrTiO 3 /p-GaP, n-Fe 2 O 3 /p -Fe 2 O 3 have been
used to eliminate the bias needed for water-splitting - Table 10.2.2. Because of the
low performance of the individual electrodes in these dual-photoelectrode devices, the
resulting overall efficiency is low.