Thermoelectric Devices Fabricated Using Amorphous Indium Gallium Zinc Oxide

Abstract

Thermoelectric (TE) devices are promising candidates for renewable energy resources. Superior TE materials are achieved by obtaining high power factors ( PF = S 2 s ), where S is the Seebeck coefficient, s is the electrical conductivity. Recently, it is necessary to investigate flexible TE devices for harvesting distributed energy at room temperature. On the other hands, amorphous InGaZnO(a-IGZO) has been already in practical use as a TFT device. The a-IGZO has enormous potential such as transparent, a low temperature process, controllability of the carrier concentration and so on. However, a few studies have been conducted on the TE properties of IGZO. Therefore, I consider a-IGZO can be applied to a transparent flexible TE element by optimizing the TE properties. There is a trade-off between the S and s under variation of the carrier concentration. Optimization of the carrier concentration is important for a high performance TE material. In this study, I optimized the carrier concentration in a-IGZO thin film and evaluated the effects of oxygen ratio on the TE properties. In addition, the obtained TE properties were theoretically analyzed. At first, a-IGZO films (200 nm) were deposited on a glass substrate at 300 K using RF magnetron sputtering. During the sputtering process, the oxygen flow ratio (O 2 /(Ar + O 2 )) was set at values of 0 ~ 6 %. After the sputtering process, the samples were annealed at 300 °C for 2 h in a N 2 atmosphere. After annealing, Mo electrodes were formed. TE properties such as S and s were measured in the range of 100 K ~ 400 K by physical property measurement system (PPMS). Carrier concentration ( n ) was measured by Hall effect method at room temperature. To analyze the measured data, I utilized percolation model that the potential barrier hinders conduction for electron transmission when the Fermi level ( E f ) is located in the tail-states. E f , and other physical parameters were obtained by fitting theoretical equations to the measured S and s . The measured s and S with theoretical results are compared. The s increased and S decreased with decreasing oxygen flow ratio. The fitting was performed in the temperature range from 150 to 350 K. The theoretical results of s fit well with the experimental results. The fitting results of S show good agreement around low temperature and under the higher oxygen flow ratios, the measured S was slightly different from the theoretical result because of nondegenerated behavior. Carrier concentration with dependence of S and s shows the relationship of the trade off between S and s . The carrier concentration increased with decreasing oxygen flow ratio. This result can be adequately explained by the internal defects resulted from the increment of oxygen vacancies. PF showed a gradual increase along with the carrier concentration up to ~ 7.7 × 10 19 cm −3 and then decreased. Therefore, the PF has the maximum value of 0.082 × 10 -3 W / mK 2 . Figure 3 indicates a trend for E f to increase with increasing carrier concentration. It was confirmed that the obtained value of E f by fitting at which PF was maximal was equivalent to the potential barrier height. These finding leads to a conclusion that the TE properties o a-IGZO were controlled by the relationship between E f and the potential barriers. The TE properties of a-IGZO thin films were evaluated with controlled carrier concentrations to obtain better TE properties. By optimizing the carrier concentration, I found that the PF had a maximal value of 0.082 × 10 −3 W/mK 2 at 300 K. This optimized PF was larger than the value of the previous report. Theoretical analysis revealed the relationship of E f with PF .