628
M. Ohta et al. / Journal of Alloys and Compounds 451 (2008) 627–631
2. Experimental
of 1.50. In addition, all the powders are contaminated with the
carbon and oxygen impurities. The carbon impurity is due to
the thermal dissociation of CS2. Although the synthesis route
which involves the reaction of rare-earth oxides with CS2 gas
offers advantage of the low-temperature formation of Ln2S3, it
The samples comprising the single-phase ␥-Ln2S3 were
obtained after pressure-assisted sintering at temperature above
1473 K. The composition of sintered samples verified by chem-
ical analysis is listed in Tables 1 and 2. For ␥-Gd2S3, the atomic
ratio of sulfur to rare-earth metal is essentially unchanged over
this sintering temperature range. On the other hand, the sul-
fur content of ␥-Tb2S3 decreases as the sintering temperature
increases. When the compound ␥-Tb2S3 was annealed under
vacuum at 1853 K for 5 h, the formation of monoclinic Tb5S7
phase was confirmed [14]. It seems that the ␥-Tb2S3 phase is
not stable at high temperature.
Fig. 1 shows the SEM micrographs of fracture surfaces of
Gd2S3 prepared by pressure-assisted sintering. The sample sin-
tered at 1473 K consists of small particles and contains several
pores. The SEM micrographs reveal that the samples are fully
compacted by sintering at temperatures of not less than 1773 K.
The grain size ranges between 5 and 20 m. These results
are very similarly to results on Tb2S3 sintered samples. For
Tb2S3, the dense samples with the grain size ranging from 5
to 20 m were obtained by sintering at high temperature of not
less than 1673 K. The examination of the sample sintered at
The synthesis route which involves the reaction of rare-earth oxides with
CS2 gas was used to prepare the rare-earth sulfides Ln2S3 (Ln = Gd and Tb). The
meanparticlesizeofGd2O3 (99.9%, Shin-EtsuChemicalCo., Japan)is1 mand
that of Tb4O7 (99.9%, Shin-Etsu Chemical Co., Japan) is 3 m. The rare-earth
oxide powders, loaded on alumina boat, were placed into the reaction tube. The
reaction tube was purged with Ar gas. CS2 gas was introduced into the reaction
tube by bubbling Ar carrier gas through liquid CS2 as soon as the samples had
reached 1248 K. The flow rate of Ar gas was fixed at 50 ml/min. After being
held for 8–10 h, the samples were gradually cooled to room temperature under
Ar atmosphere.
The synthetic powders were then consolidated by pressure-assisted sintering
in order to achieve densities close to the theoretical values. The powders were
placed in a graphite die and the die was introduced into the sintering apparatus
(SPS-511S, Sumitomo Coal Mining Co., Japan). The chamber was evacuated
down to 7 × 10−3 Pa. The powders were heated at rate of 25 K/min up to the
desired temperature. After being held for 1 h, the powders were cooled at rate of
50 K/min to 873 K, followed by slow cooling to room temperature. An uniaxial
pressure of 50 MPa in the die was applied during the heating and holding.
The crystal phase of samples was studied by X-ray diffractometry using
Cu K␣ radiation (Rint-Ultima+, Rigaku Co.). The rare-earth content was deter-
mined by chelatometric titration at pH 5.5. Xylenol orange was used as an
indicator. The sulfur, oxygen, and carbon contents were determined by using
LECO-TC436 nitrogen/oxygen and LECO-CS444 carbon/sulfur determinators.
The atomic ratios of sulfur to rare-earth metal x was accurate to 0.01. The
electrical resistivity and thermal conductivity measurements were made using
a four-probe dc technique and a laser-flash method, respectively. These mea-
surements were carried out in a vacuum chamber evacuated to less than 1.0 Pa.
The Seebeck coefficient was determined from a slope obtained by a plot of the
Seebeck voltage versus the temperature difference. The Seebeck voltage was
measured under a helium atmosphere. The temperature difference between the
ends of samples was varied from 0 to 10 K.
The thermoelectric power factor (S2/ρ) was calculated from
the measured electrical resistivity and Seebeck coefficient data.
The thermoelectric power factor at 873 K versus sintering tem-
perature is shown in Fig. 2. For Gd2S3, the thermoelectric power
factor increases as the sintering temperature increases up to
1773 K, and then it remains constant. The result suggests that
the improvement with respect to the thermoelectric properties is
realized by fabricating the fully dense sample. In Tb2S3, a dra-
3. Results and discussion
After reactions of the rare-earth oxides with CS2 gas at
1248 K, all the products are comprised of an orthorhombic ␣-
phase. The results of chemical analysis of the synthetic powders
are summarized in Tables 1 and 2. The atomic ratios of sulfur
to rare-earth metal are close to the normal stoichiometric ratio
Table 1
Chemical analysis of the synthetic Gd2S3 powder and the sintered samples
Sample
Chemical composition (at.%)
S/Gd (atomic ratio)
Gd
S
O
C
Synthetic powder
38.3
38.2
37.6
38.4
38.2
57.6
57.0
56.1
57.5
57.1
1.24
1.88
1.75
2.04
1.78
2.84
2.90
4.54
2.12
2.90
1.50
1.49
1.49
1.50
1.49
Sample sintered at 1473 K
Sample sintered at 1673 K
Sample sintered at 1773 K
Sample sintered at 1873 K
Table 2
Chemical analysis of the synthetic Tb2S3 powder and the sintered samples
Sample
Chemical composition (at.%)
S/Tb (atomic ratio)
Tb
S
O
C
Synthetic powder
38.4
38.4
38.3
38.4
38.9
56.5
56.6
55.8
54.9
55.8
2.59
2.84
2.93
3.64
3.44
2.59
2.19
3.24
3.19
1.62
1.47
1.47
1.46
1.43
1.43
Sample sintered at 1423 K
Sample sintered at 1573 K
Sample sintered at 1673 K
Sample sintered at 1713 K