YMnO₃ is known to crystallize in a hexagonal structure [1] and has both ferroelectric and antiferromagnetic properties [2] [3] . The Curry and the Neel temperatures are 914 and 73 K, respectively. And the crystal transition occurred from the space group of P6₃cm to P6₃/mcm 600 to 990 K [4] . Whereas pure YMnO₃ is ferroelectric and insulating, it could be a conducting material with carrier doping. Gutierrez et al. substitute Ni for Mn site of YMnO₃ and observed hopping conduction behavior up to 700˚C [5] .

In this study, sintered bodies of pure YMnO₃ were prepared and the electrical conductivity and the thermoelectric power up to 800˚C were measured. The effect of Fe-substitution for Mn was also examined.

Samples are prepared by ordinary solid state reactions. As starting materials, Y₂O₃ (99.99%), MnO₂ (99.5%), Fe₂O₃ (99.9%) were used. Y₂O₃ were pre-calcined at 600˚C for 10 min for elimination of water. Y₂O₃, Mn₂O₃ and Fe₂O₃ were weighed to prepare Y(Mn_1−xFe_x)O₃ (x = 0, 0.05, 0.1. 0.2, 0.225, 0.275, 0.3, 0.4) and mixed with a pastle in a mortar of Al₂O₃ for 1 h. For YMnO₃, the mixed powders were calcined in crucibles of Al₂O₃ at temperatures among 900˚C and 1300˚C for 2 h. The powders calcined were uniaxially pressed and then isostatically pressed into disks of diameters of about 10 mm or rectangular bar of 4 × 3 × 20 mm. Then, the green bodies were placed in an aluminum-oxide crucible and sintered in air at 1300˚C, 1400˚C or 1500˚C for 4 h.

The phases of the mixed and the calcined powders were analyzed by X-ray diffraction (XRD). The lattice constants of Y(Mn_1−xFe_x)O₃ powders were calculated from the precisely measured XRD patterns. The densities of the sintered bodies were measured by Archimedes' method.

The samples prepared were evaluated by electrical properties, electrical conductivity and Seebeck coefficients. Electrical conductivity was measured using the four-probe method with an alternating current of 1 kHz. As the electrodes, four platinum wires were attached using silver paste. The resistivity was measured in air at temperatures between room temperature and 900˚C at increasing and decreasing temperature rates of 10˚C/min using an infrared image furnace.

Thermoelectric power was measured using an apparatus having double heaters to prepare specimens temperature differences of approximately 10˚C. The temperature of both ends of the specimens were measured using CA thermocouples, and the thermoelectric voltage was measured via two platinum wires (0.1 mm φ) using a computer-controlled mV-meter (HP 3457A). The Seebeck coefficients were determined using a least-squares analysis of five ΔV vs. ΔT data points, where ΔV was the voltage gradient due to the thermoelectric effect of an induced thermal gradient ΔT. The Seebeck coefficient was measured in air at temperature between 100˚C to 800˚C. The absolute (non-relative) values of the Seebeck coefficient were calculated using the measured values and the absolute ones of platinum, which were measured previously by Cusack et al. [6] .

XRD patterns of powders prepared by calcination of mixed powders at temperatures between 900˚C to 1200˚C are shown in Figure 1. As for the mixture of Y₂O₃ and Mn₂O₃, only Y₂O₃ was clearly detected with XRD. When calcined at 900˚C, besides Y₂O₃, YMn₂O₅ was detected. Calcination at temperatures higher than or equal 1000˚C caused YMn₂O₅ react with residual Y₂O₃ resulting YMnO₃ phase. Finally, the single phase of YMnO₃ was prepared to heat at 1100˚C for 2 h.

The relative densities of the sintered bodies fired at 1300˚C, 1400˚C or 1500˚C as a function of calcining temperature of powder preparation are shown in Figure 2. This result indicates that densification of this material needs temperature up to 1500˚C and the best calcining temperature is 1100˚C - 1200˚C. When fired at 1500˚C, color of the bottom of the crucibles changed slightly brown. This shows that firing over that temperature make Mn volatile and the stoichiometry not maintained. In case sintered up to 1500˚C, all the resulting bodies were single phase of YMnO₃.

Figure 3 shows XRD patterns of calcined powders composition of which were Y(Mn_1−xFe_x)O₃ (x = 0, 0.1, 0.2, 0.225, 0.25 and 0.3). When x reached 0.3, the peak reflected from YFeO₃ phase was detected. Figure 4 shows the lattice constants a and c of the Fe doped samples. From these results, the limit of Fe doping for solid solution was said to be 0.25. The relative densities of all of the Fe-doped and single phased samples were around 90%, which showed Fe did not affect to sintering conditions.

The electrical properties of the nominal pure YMnO₃ were those of canonical extrinsic semiconductors. Figure 5 shows electrical conductivity of pure and Fe-doped YMnO₃ sintered bodies. The electrical conductivity of the pure YMnO₃ increased with increasing temperature. However, the plateau was observed from 450˚C to 750˚C, which seems the region of saturation of carriers. Figure 6 shows Seebeck coefficients of the samples. From this graph, it is shown that every sample showed positive Seebeck coefficients, so that the carrier of YMnO₃ was holes. As temperature increased above 700˚C, the Seebeck coefficients of pure YMnO₃ decreased, which

means that the tendency of the intrinsic conduction was detected in this temperature range. This also explained the increase of electrical conductivity above 750˚C.

As Fe substituted for Mn, electrical conductivity around 500˚C increased, and Seebeck coefficient over 700˚C also increased, up to x = 0.1. This showed that Fe substitution made the hole concentration rise and the extrinsic behavior of the conduction became conspicuous. Although Fe can substitute up to x = 0.25, which is concluded from the crystal data, the effect of substitution on the electrical properties was limited up to x = 0.1.

The single phase of YMnO₃ was prepared to heat at 1100˚C for 2 h and densification of this material needs temperature up to 1500˚C and the best calcining temperature was 1100˚C - 1200˚C. The plateau of the electrical conductivity was observed from 350˚C to 750˚C. The limit of Fe doping for solid solution was said to be 0.25 and the effect of substitution on the electrical properties was limited up to x = 0.1.