Hydrothermal synthesis of xonotlite whiskers by ion diffusion

Abstract

Xonotlite, a calcium silicate hydrate, has been industrially produced as a main constituent for heat insulating material [1, 2], building material [1, 3] and artificial wood [1, 4], because of its high stability at high temperature and fibrous crystal form. It is usually prepared from a suspension of silica and calcium hydroxide by hydrothermal treatment [5– 10]. Formation mechanism of xonotlite crystals from the suspension is complicated, because various intermediate reaction products are formed in the reaction processes. For example, calcium hydroxide reacts with low-quartz particles to form calcium-rich CSH (low crystalline calcium silicate hydrates) in the first stage of the reaction. Compositional change in CSH to equivalent Ca=Si ratio leads to crystallization of tobermorite which transforms to xonotlite by further hydrothermal treatment [5]. The size of the xonotlite crystals thus prepared is limited because of formation from these intermediate products. Kunugida et al. [11] pointed out that direct preparation of xonotlite, not through intermediate products such as CSH and tobermorite, was important for preparation of long fibre xonotlite. They obtained long fibre xonotlite crystals with maximum length 2 mm at 350 8C in the continuoustype autoclave by the two processes, in-situ mixing and rapid heating of raw materials. They reported that recrystallization of xonotlite also lengthened the crystals. We consider that continuous crystal growth provides elongation of xonotlite crystals under conditions where xonotlite is stable. We report here the preparation of xonotlite whiskers at 250 8C using a continuous supply of calcium and silicate ions in a multichamber autoclave where silica and calcium hydroxide are separately placed. The multichamber autoclave used in this study is shown in Fig. 1. It was made of stainless steel with total inner volume 50 cm3 and had two separate chambers each with volume 12.5 cm3. Two kinds of silica sources were used; silica gel (Wakogel C-300, Wako Pure Chem. Ind., Japan, 200–300 mesh) and low-quartz powder (SiO2 99.5 wt %, Al2O3 0.2 wt %) consisting of coarse grains from 20–150 im in diameter, produced by crushing silica stone (from Fukushima Prefecture, Japan) and classifying from the water suspension by gravitational sedimentation. Calcium source was a pellet of calcium oxide produced by calcination of calcium carbonate (reagent grade, Wako Pure Chem. Ind., Japan) at 1100 8C for 3 h. After placing the silica and calcium source (0.5 g each) at the bottom of each chamber, decarbonated distilled water (30 cm3) was poured slowly into the autoclave so not to suspend the raw materials. The silica source was brought into contact with the calcium source via the water; thus, ion species dissolved from the two sources diffuse to each other in the autoclave to produce calcium silicate hydrates. Platinum plates were placed in the autoclave (Fig. 1) to collect the reaction products formed by ion diffusion. These operations were performed in a glove box purged of carbon dioxide with nitrogen gas. The autoclave was heated at 250 8C for 2 to 72 h in an electric oven without temperature gradient; it took about 1 h to heat the autoclave to 250 8C. At the conclusion of the hydrothermal treatment, the autoclave was cooled down to room temperature by a cold air blow, and dried in vacuum after removal of the water. The reaction products, precipitated on the platinum plates, were observed by scanning and transmission electron microscope equipped with an energy dispersive X-ray spectrometer and identified by X-ray powder diffraction. When silica gel was used as a silica source, a large quantity of ribbon-like crystals were obtained after 2 h at 250 8C (Fig. 2a). X-ray diffraction showed that they were low crystalline tobermorite