The attempts to substitute liquid electrolytic contact in photoelectrochemical cells (PEC) for solid electrolyte have been made for the cells with different types of semiconductor electrodes CdS, TiO₂, n-Si-polypyrrole [1] and others. The solid electrolytic contacts used in the noted works as well as in the investigated presented here can be classified into the following three groups:
1) "classical" solid electrolytes;
2) polymeric membrane-based pseudo-solid electrolytes having pores filled with liquid electrolyte;
3) the intermediate type of electrolytes combining the possibility of ion transfer according to the mechanisms inherent in "classic" solid electrolytes with the transport in the network of microchannels containing liquid or pseudo-liquid mobile phase.
In the present work, PECs based on TiO₂ and CdS film electrodes using all the three types of electrolytic contacts have been investigated. With PECs utilizing the contacts of group 1, wide range of solid inorganic electrolytes possessing high ionic conductivity have been tried, the acidic phosphates of divalent metals being of special interest. However it should be pointed out that the noted phosphates may undergo the complicated topochemical transformations [2] in the course of PEC operation due to the thermocycling from -20 to +90oC. This fact not only can be considered as a negative factor, but also as an interesting way to increase the ionic conductivity substantially. In particular, such effects have been observed in crystalline phosphate Mg(H₂PO₄) x4 H₂O and Zn(H₂PO₄) x2 H₂O for which the topochemical processes of hydration-dehydration can be considered as reversible ones as has been demonstrated in our previous works [2,3]. As for the PECs of group 2, film polymeric membranes based on the polyvinyl alcohol, polyvinylpyrrolidone and partially hydrolized polybutyltitanate modified by acidic phosphate compounds have been used in the present work. Such membranes were tested in PECs with counter electrode based on hydride-forming alloys La_0.9Zr_0.1Ni_4.5Al_{0.1 and LaNi3.5Co0.7Al0.8 as a hydrogen accumulated materials.
When studying the PECs of group 3, the emphasis has been done on the electrolytes based on solutions-melts of crystal hydrates like Na2SO4 x10H2O, Na2b4O7 x10H2O, Na3PO4 x12H2O which are used in heat accumulators as a working medium. The characteristic feature of these PECs is the pronounced hysteresis observed on the dependences of photocurrent and capacity on temperature (Fig. 1) and electrode potential.It is seen from Iph vs. to curves that appreciable photocurrent is observed even in the frosen state (t<28 oC), especially when redox couple is present in the system. Interestingly, in the case of TiO2 electrodes, the dynamic increase in temperature (fig.2) cost by the evolution of heat during crystallisation does not bring an adequate rise of photocurrent (see fig.1). At the some time, the opposite behaviour (increase in Iph) has been observed for CdS electrodes in the presence of ferro-ferricianide couple (fig.1c), in spite of the fact that to vs. t dependence is practically identical to that shown in fig.2. The interpretation of the photoelectrochemical properties of the systems considered herein and similar ones is being developed in our laboratory on the basis of the model of percolational ion transfer with allowance made for the dynamics of dielectric constant of the mobile ionic microphase and the theory of critical phenomena in the systems with phase transitions.}

Fig. 1 The temperature dependence of photocurrent Iph (a, b, c) and low frequency (200 Hz) capacitance Css (d) for TiO2 (a, b, d) and CdS (c) electrodes in the freezing (-o-o-o-) and defreezing (-=B7-=B7-=B7-) crystal hydrate Na2SO4 10 H2O with (c) and= without (a, b, d) admixture of 0.05 K3[Fe(CN)6] + 0.05 K4[Fe(CN)6]. Photocurrent is measured at +0.8 V vs. SCE (three-electrode scheme of polarization) using 365-nm line of mercury lamp as the light source. The irradiation intensity in the case b is 20 times as low as in the cases a, c, d. The capacitance is measured at flat-band potential with the use of lock-in technique.

Fig. 2 Temporal dependence of the temperature of the electrolyte (Na2SO4 x10 H2O) during the recording of the curves given in Fig. 1. Point A corresponds to the switching from defreezing to cooling. The jump in the temperature at point B results from the heat release at crystallization.

_References

1.	Kulak A.I. Electrochemistry of Semiconductor Heterostructures. Minsk, 1986 (in Russ.).
2.	Samuskevich V., Lukyanchenko O., Samuskevich L. Ind.Eng.Chem.Res.1997, 36,4791.
3.	Lukyanchenko O., Samuskevich V. Neorganicheskie materialy (in Russ.) 1997, 9, 1127.

_{[BACK] [MAIN] [FURTHER]}