Trials for oxide photo-thermoelectrics

Ichiro Terasaki 0 Ryuji Okazaki 0 Partha Sarathi Mondal 0 Yu-Chin Hsieh 0 0 I. Terasaki (&) R. Okazaki P. S. Mondal Y.-C. Hsieh Department of Physics, Nagoya University , Nagoya 464-8602, Japan Thermoelectrics is an energy conversion technology from heat into electricity, and vice versa, through the thermoelectric phenomena in solids, while photovoltaics is an energy conversion technology from solar photon energy into electricity using the photo-excitations in solids. We are trying to find a way to combine thermoelectrics with photovoltaics to establish a new method to generate renewable energy with high efficiency. In this article, we show two approaches for this purpose using oxide materials: thermoelectric energy conversion by photo-excited carriers and the thermoelectric power generation using a focused light as a heat source. - No one can deny that our modern society is based on vast consumption of electric/chemical energies. Since all the developing countries have the right to enjoy life as comfortable as the advanced countries do, energy demands are increasing year by year in spite of serious shortage of petroleum. Thus, a search for sufficient energy resources is a responsibility of researchers in all areas of science and technology. Best energy resources are of no doubt This work was partially supported by ALCA, Japan Science and Technology Agency, and by The Mitsubishi Foundation. renewable energies, which preliminarily come from the solar energy. The energy conversion technique using the solar energy is classified into two; The one is photovoltaics in which an electronhole pair created by an incident photon is separated by an internal electric field at the pn junction [27]. This technology is now commercially available as solar battery cells. The other is solar-thermal energy conversion, where heat generated by focused sunlight vaporizes water to rotate a gas turbine [34]. Although these two techniques are matured, there remain issues to be addressed. In photovoltaics, the conversion efficiency is close to a theoretical limit, and raw materials of silicon of high quality are about to run out. Of course alternative materials are being developed, but the cost and natural abundance are still issues. In the case of solar-thermal conversion, the conversion efficiency is not satisfactory except for some areas around the equators. We have studied thermoelectric energy conversion using oxide materials, which are superior at high temperatures in air [9, 13]. A serious drawback of thermoelectrics is poor efficiency [28]. A good thermoelectric material requires high electrical conductivity, large Seebeck coefficient, and low thermal conductivity at the same time, which is very difficult to be realized in real materials. In fact, reliable calculations of materials parameters do not give promising results. Since such calculations are done near equilibrium states, we hope that the thermoelectric performance may go beyond theoretical limitations in non-equilibrium states [40, 41]. As such, we are trying to find ways to break through the poor efficiency by focusing non-equilibrium states. In this article, we show our preliminary results for two types of photo-excited thermoelectrics. One is the thermoelectrics using the photo-Seebeck effects (Fig. 1a), and the other is the thermoelectric energy conversion in a Fig. 1 Schematics of (a) power generation using photo-Seebeck effect and (b) power generation from focused light large temperature gradient using solar light as a heat source (Fig. 1b). One may associate the two with photovoltaic and solar-thermal energy conversions. To emphasize our originality, we will briefly summarize the preceding works. The photo-Seebeck effect was first reported by Tauc [31] in 1955, and was later examined in conventional semiconductors [7, 15]. Note that the word photo-thermoelectric is confusing; it stood for the photoSeebeck effect before the 80s, but is now used as photothermal energy conversion through the Seebeck effect (for example, see [2, 12]). To our knowledge, our work is the first trial for the thermoelectric energy conversion using the photo-Seebeck effect. In contrast, the second trial shown in Fig. 1b has been examined by many groups. Originally Telkes [32] examined the concept of solar thermoelectric generator in 1954. Naito et al. [18] designed a power converter using concentrated solar light and achieved a high temperature of 2,200 K in vacuum. Suter et al. [30] fabricated the solar thermoelectric generator and analyzed the efficiency and the maximum power. Fan et al. [3] reported an efficiency of 3 % using commercially available Peltier modules. Thus, our originality lies on the fact that our device is made of single crystals of transition metal oxide, which can work at 800 K in air. Photo-Seebeck effect in oxide single crystals Although the first observation of the photo-Seebeck effect was reported in the mid 50s [31], there have been very few reports on the photo-Seebeck measurements since the 80s. Thus, we had to newly establish measurement procedure with recently available equipments such as a light-emitting diode (LED). Here, we elaborate on the measurement and analysis details for the photo-Seebeck coefficient, which was not included in the published papers [17, 20]. Commercially available substrates were used as ZnO single crystals, and flux method was employed for making single crystals of PbO. The photoconductivity of ZnO and PbO single crystals was measured with a two-probe method. The thermoelectric voltage of ZnO was measured with two-probe technique with several temperature difference and several photon intensities [20]. The resistance of PbO was too high to use the same measurement setup as in the case of ZnO. Instead, the thermoelectric current was measured with several temperature difference and several photon intensities [17]. The photo-Seebeck effect is a change in the Seebeck coefficient with light illumination, which can be evaluated by comparing the thermoelectric voltage before and after illumination. However, the light illumination affects various properties at the same time; It causes the photovoltaic voltage at the contacts, increases the sample temperature, and changes the temperature difference. To be more quantitative, the measured thermoelectric voltage VTE can be expressed by temperature difference DT at dark as where S is the Seebeck coefficient at dark and V0 is the offset voltage. When illuminated, the voltage can be written as VTE S dSDT dT V0 dV0; where dS is the photo-induced change in the Seebeck coefficient. dT and dV0 are the photo-induced temperature difference and photovoltaic component, respectively. Note that we measured the voltage in a cryostat (Quantum Design PPMS), where the sample temperature was strictly controlled. The best way to measure the photo-Seebeck coefficient is to measure VTE as a function of DT dT. When VTE is found to be a linear function of DT dT (...truncated)