JOURNAL OF MATERIALS SCIENCE LETTERS 19, 2 0 0 0, 1987 – 1990 Compositional dependence of crystallization in the glass-ceramics system Bi 2 O 3 -In 2 O 3 -MnO 2 -B 2 O 3 ZHIGANG ZOU ∗ National Institute of Materials and Chemical Research, 1-1 Higashi, Tsukuba, Ibaraki 305, Japan E-mail: zou@nimc.go.jp JINHUA YE, HIRONORI ARAKAWA National Research Institute for Metals, 1-2-1 Sengen, Tsukuba, Ibaraki 305, Japan The discovery of a semiconducting property in vana- dium oxide [1] has led to a search for other non-alkali oxide glasses compounds with semiconducting prop- erties. It is well known that the properties of glass- ceramics can be changed by substituting some group-III elements for group-V elements. For this purpose, a va- riety of systems including bismuth glasses have been proposed [2–5]. As described in a previous paper [6], a new Bi- system Bi-In-Cu-B-O was prepared by a super-rapid quenching method using a halogen arc-imaging furnace and a twin roller apparatus. The compositional depen- dence of the crystallization in the glass-ceramics sys- tem Bi 2 O 3 -In 2 O 3 -CuO-B 2 O 3 has been understood. The glass transition and crystallization temperature of these glasses decrease with increasing In 2 O 3 content. For the present study, we chose the Bi 2 O 3 -In 2 O 3 -MnO 2 -B 2 O 3 system [7]. The characteristic behavior of the glass- ceramic system Bi 2 O 3 -In 2 O 3 -CuO-B 2 O 3 was investi- gated and the crystallization tendency of the system was determined by DTA and X-ray powder diffraction methods. Glass-ceramics in the system Bi 2 O 3 -In 2 O 3 -MnO 2 - B 2 O 3 were prepared by a super-rapid quenching method using a halogen arc-imaging furnace and twin roller apparatus; the roller moving rate can be var- ied from 0 to 6000 rpm [6]. Commercial powders of Bi 2 O 3 (99.9% Wako Pure Chemical Industries. LTD.), In 2 O 3 (99.9% Wako Pure Chemical Industries. LTD.), MnO 2 (99.9% Wako Pure Chemical Indus- tries. LTD.) and B 2 O 3 (99.9% Wako Pure Chemical Industries. LTD.) were used as starting materials. A base body of the glass-ceramic (70%Bi 2 O 3 -30%In 2 O 3 (mol%)) was first synthesized. Before use, powders of Bi 2 O 3 and In 2 O 3 were dried at 500 ◦ C and 600 ◦ C, respectively. The base body was sintered in a Pt cru- cible for 40 h at 1000 ◦ C using a vertical quench fur- nace. The base body of the glass-ceramics and addi- tions of MnO 2 and B 2 O 3 were mixed with the powder. Then, the mixed powder was pressed at 300 MPa into a thin body and melted [6]. After quenching through a twin-roller apparatus, glass flakes were obtained. The glassy state in the quenched samples and the crystalline phases in the system were examined by X-ray pow- der diffraction (XRD) analyses at room temperature ∗ Author to whom all correspondence should be addressed. Figure 1 X-ray powder diffraction pattern of the base body.  Bi 2 In 60 O 93 ;  In 2 O 3 ;  unknown phase. using Cu K α (λ = 1.5405 ˚ A). All phases investigated in this study were analyzed by means of scanning elec- tron microscopy-X-ray energy dispersion spectroscopy (SEM-EDS) with an accelerating voltage of 25 kV. Dif- ferential thermal analysis (DTA) of the samples was carried out between room temperature and 1300 ◦ C in air, with α-Al 2 O 3 as the standard material and a heat- ing rate of 10 K/min. Fig. 1 shows a typical X-ray diffraction pattern of the base body sintered at 1000 ◦ C. Most of the diffraction peaks were successfully indexed on the base of the tetragonal system with a primitive lattice, a = 0.7718 nm, c = 0.5647 nm and the P42 1 C(114) space group. Other peaks were indexed on the basis of In 2 O 3 with the hexagonal system, R3C(167) space group with a = 0.5487 nm, c = 1.4510 nm. A few re- flections could not be indexed within the above systems. The unknown peaks probably arise from stable inter- mediate phases with compositions between Bi 60 In 2 O 93 [8] and In 2 O 3 . DTA and thermogravimetric analysis (TGA) mea- surements were carried out on the base body. The peak positions were determined from the temperatures at which the peak height became a maximum. The peak positions are the average of the measured values. Fig. 2 shows the DTA and TG curves of the base body. DTA results show two peaks at 963 ◦ C and 1105 ◦ C. In order to elucidate the nature of these peaks, a second run in- volved heating a sample to 1200 ◦ C, cooling to 300 ◦ C, and subsequently taking the data up to 1300 ◦ C. The 0261–8028 C  2000 Kluwer Academic Publishers 1987