Photoelectrochemical cells for hydrogen production from solar energy - Solar Energy Sciences and Engineering Applications - page 309

consists of a solvent with active species to be reduced or oxidized, depending if it is an
alkaline or an acid media. Nevertheless, pure water is not conductive and thus support-
ing ions must be added to ensure the desired charge transfer (Krol et al., 2008). The
photoactive semiconductor immersed in a redox electrolyte is greatly affected by the
solution properties, redox level and stability, interfacial kinetics (adsorption), viscos-
ity, conductivity, ionic activity and transparency within the crucial wavelength region
(Archer and Nozik, 2008). As mentioned, the choice of a suitable electrolyte solution is
very important, mainly in what concerns the redox couple selection. It should improve
the charge-transfer kinetics, the photoelectrode stability, and also should help in pre-
venting undesirable phenomena such as surface recombination and trapping (Archer
and Nozik, 2008). Also, the electrolyte concentration should be sufficiently high to
avoid large ohmic voltage losses. As reported by Roel and co-authors, the voltage
drop is given by V loss

= ×

R E , where I is the total current flowing between the work-
ing electrode and the counter-electrode, and R E is the electrolyte resistance (Krol and
Grätzel, 2012). The electrolyte conductivity strongly depends on the type of ions and
the corresponding concentration value. It is important to add that there is not a linear
relation between conductivity and ion concentration due to incomplete dissociation of
anions and cations and/or ion-solvent interactions. Moreover, deviations from linear-
ity can occur for concentrations above 1 mmol L − 1 ; at high concentrations, i.e. > 1M,
the formation of ion-pairs can result in a decrease of the conductivity with the con-
centration increase. This behavior explains why the conductivity starts to decrease for
concentrations higher than

I

6 M of KOH aqueous solutions. To avoid large ohmic
losses it is then important to guarantee concentrations of at least 0.5 M (Krol and
Grätzel, 2012).
Usually, acid electrolytes such as aqueous H 2 SO 4 or HCl solution (0.5-1 M) are
often used with WO 3 and TiO 2 photoelectrodes (Krol and Grätzel, 2012). Aqueous
NaCl solutions are also used with WO 3 photoanodes, simulating sea water conditions
(Alexander and Augustynski, 2010). For electrodes such as

∼

-Fe 2 O 3 , where alkaline
or neutral electrolyte solutions are preferable, concentrations of 0.5-1 M NaOH or
KOH should be used. Metal oxide semiconductors that are only stable in fairly neutral
environments such as BiVO 4 , 0.5 M Na 2 SO 4 or K 2 SO 4 solutions should be used and
the electrolyte should be buffered (KH 2 PO 4 /K 2 HPO 4 ) to prevent local pH fluctuations
(Krol and Grätzel, 2012).
In photoelectrochemical cells bubbles usually get stacked in the electrode surface
which can generate excessive noise on the photocurrent signal and thus it is necessary
to remove them. This can be done at lab level flashing using a nitrogen or argon stream
or simply by using a magnetic stir bar for stirring the electrolyte (Krol and Grätzel,
2012). Moreover, by using these procedures, the back-reaction of dissolved hydrogen
and oxygen again to water can also be prevented ensuring that the redox potentials
do not change over time. The Portocell with the integrated electrolyte recirculation
systems helps to remove the stacked bubbles, being a solution applicable for both
laboratory and industrial contexts.

α

10.2.5.2 The counter-electrode

As stated before, the preferable configuration is the single-photon system, in which
the photoactive semiconductor works as working-electrode and a metallic material as