The internal friction in glasses of Na₂O-K₂O-Al₂O₃-SiO₂ system was measured with the torsion pendulum technique at a frequency of about 1Hz. The high-temperature peak observed in a single alkali glass exhibited close relation to nonbridging oxygen ions and its height was independent of the alkali mixing ratio. The mixed-cation peak observed in a mixed alkali glass was independent of nonbridging oxygen ions. The high-temperature peak and the mixed-cation peak differed in temperatures at which they occurred. On the basis of above results, it was presumed that the relaxation mechanism of the mixed-cation peak differed from that of the high-temperature peak. The peak appeared at high temperature in a mixed alkali glass containing nonbridging oxygen ions must be a cooperative product of the high-temperature peak and the mixed-cation peak. The following suggestions have been proposed. (1) The high-temperature peak is attributed to coexistence of nonbridging oxygen ions and alkali ions which are strongly affected by nonbridging oxygen ions. (2) The low-temperature peak is due to the movement of alkali ions which are hardly affected by nonbridging oxygen ions. (3) The mixed-cation peak is probably associated with the interaction between different kinds of alkali ions.

View full abstract

A survey has been made on an electrical conduction model of Sb₂O₃-doped SnO₂ semi-conducting glaze by analyzing the relationships between a content of semi-conducting powder (V_e) and a conductivity of a glaze. It has been concluded that a three dimensional connection of a solution layer surrounding a semi-conducting particle formed during a solution process of particles into base glaze seems to play an important role for the formation of conduction paths in the glaze. There has been found a clear separate dependence of conductivity with V_e in the vicinity of V_e=3.88vol%. The effective medium model, e. g. the Maxwell's one for V_e<3.88vol%, and the Scarisbrick's adjacent model (random distribution of conducting particles) for V_e>3.88vol% can respectively satisfy the relationships of V_e vs. conductivity of glazes. An equilibrium thickness has been calculated from a model assuming that semi-conducting cubic particles are arranged in a simple cubic lattice. And moreover, a conductivity distribution in the solution layer has been also roughly estimated on the basis of the Bube's equation which had been derived for a conductivity of such a model system.

View full abstract

The effects of oxides (MgO, Y₂O₃ and lanthanide oxides), carbides (SiC, TiC, ZrC, HfC, TaC and B₄C) and nitrides (TiN, TaN, ZrN and BN) on strength of β-sialon were examined. β-sialon with z=1(+4) (Si_6-zAl_zO_zN_8-z; z=0-4.2, +4; 4eq% excess oxygen over the z=1) from a powder mixture of α-Si₃N₄, α-Al₂O₃ and AlN containing 5wt% of each additives respectively were hot-pressed at 1850°C for 1h under a pressure of 300kg/cm². Three-point bending strength, microstructure and phases of the sintered materials were investigated. Additives were devided into 4 groups.
(1) MgO, Y₂O₃ and lanthanide oxides;
Strength of β-sialon showed a increase (70-90kg/mm²) at room temperature and a decrease (20-30kg/mm²) at 1200°C.
(2) TiO₂, ZrO₂, ZrN, ZrC and HfC;
Addition of these compounds gave a increase in stength of β-sialon at room temperature and decrease at elevated temperature, but not so remarkable as the compounds in Group 1. From the X-ray diffraction patterns of the specimen, X, O′ and TiN in addition to β-sialon were observed in the materials with TiO₂, m-ZrO₂ and HfO₂ were detected in the specimen with ZrN, ZrC and HfC addition.
(3) SiC, NbC, TiC and TiN;
These additives have an effect to increase the strength of β-sialon both at room temperature and 1400°C. They are considered to play a role as a barrier to crack propagation.
(4) BN and TaN;
These additives introduced defects and lower the strength level of β-sialon sintered materials.

View full abstract

Acid decomposition/atomic absorption spectrometry system was applied to the determination of chlorine in silicon nitride powder sample. A teflon pressure vessel was used for the acid decomposition of silicon nitride. To inhibit the formation of HCl which vaporizes from the teflon vessel, a known amount of AgNO₃ was added before the acid decomposition. Silicon nitride (0.5g) was mixed with 10ml of HF, 2ml of HNO₃ and 2ml of AgNO₃ solution, and decomposed in the teflon pressure vessel at 160°C for 16h. After cooling, the sample solution was diluted to about 100ml and stood for at least 2h in the dark. The precipitated AgCL was removed and the filtrate was diluted to 250ml. Chlorine in silicon nitride was indirectly determined by measuring excess Ag in the solution by atomic absorption spectrometry. The matrix components of SiF₆^2- (0.6%), NH₄⁺ (0.1%), HF (1.3%) and HNO₃ (0.2%) in the solution did not affect the absorption of Ag. The recovery of Cl in the acid decomposition procedure was almost 100%; when the sample was decomposed without the addition of AgNO₃, the recovery of Cl was about 70%. Therefore, the addition of AgNO₃ before the acid decomposition was effective for preventing the vaporization of Cl from the teflon vessel. Chlorine in NBS SRM glass samples was determined by this method; the measured values were in good agreement with the recommended values. The detection limit of Cl by this method was 0.002%. This method was applied to the analysis of commercial silicon nitride powder samples. The concentrations of Cl in these samples were 4.5%-0.008%, and the precision (relative standerd deviation) was 1%-25%.

View full abstract

The strength of sintered silicon nitride and hot-pressed silicon carbide specimens with various kinds of small defects was obtained by four-point flexural tests. The relationship between the strength and equivalent crack length a_e for three kinds of defects, surface roughness, semicircular cracks and edge notches, were almost the same. On the other hand, the fracture toughnesses obtained from the specimens with large defects were higher than those obtained from specimens with small defects. This means that the conventional linear fracture mechanics gives unsafe side estimation for ceramic components with small defects. The relationship between the ratio of the equivalent crack length and the grain size a_e/d, and the ratio of critical stress intensity factors for small and large defects K_C/K_IC was identical for various kinds of ceramics. The strengths of these polycrystalline ceramics with small defects were expressed well with a granular fracture model. The effect of grain diameter on the strength of a smooth specimen and the effect of notch root radius on fracture strength were also explained well with the model.

View full abstract

Porous glasses in the system ZrO₂-SiO₂ containing up to 50mol% ZrO₂ were prepared by the sol-gel process from metal alkoxides and their alkali resistances were investigated. The gels were made by hydrolyzing and drying the solution of Zr-propoxide which was brought into reaction with pre-hydrolyzed Si-ethoxide and were converted into the porous glasses by heating at 500° to ca. 800°C. It was found that the glass-like skeleton was already made up in these temperature regions. The specific surface area was 200-400m²/g and an average pore diameter was about 20Å. Alkali resistance of the porous glasses was examined by measuring the weight loss in 2N NaOH solution at 90°C. The weight loss markedly decreased with increasing ZrO₂ content, that is, the value after 200h corrosion were 80×10^-3, 10×10^-3 and 2.5×10^-3mg/dm² for 10ZrO₂⋅90SiO₂, 30ZrO₂⋅70SiO₂ and 50ZrO₂⋅50SiO₂ porous glasses, respectively. The weight of porous glasses in NaOH solution decreased linearly with time, indicating the corrosion of the porous glasses was the dissolution controlled. The activation energy for the corrosion was about 45kJ/mol, independent of the existence of ZrO₂. Zr-enriched layear was observed to adhere loosely to the glass surface but did not seem to effectively inhibit the dissolution of glasses.

View full abstract

The influence of intergrinding of CaO with cement on the initial hydration process and setting time was investigated. Addition of CaO increased Ca(OH)₂ saturation ratio, the ratio of Ca(OH)₂ concentration to saturated Ca(OH)₂ concentration under the same ionic strength, in the liquid phase, and shortened the peak time of Ca(OH)₂ saturation ratio remarkably. According to this change of Ca(OH)₂ saturation ratio curve, the initiation of the second peak on heat evolution curve of cement, which meant the completion of induction period of alite in cement, also shifted to earlier time. Shortening of setting time of cement by the addition of CaO was proved to be due to such a mechanism. This effect was larger for intergrinding of CaO than mere mixing, therefore the fineness of CaO and homogeneous mixing with cement are considered to be important factors. The effect of CaO addition on the hydration of C₃S or alite is extremely complicated and the effect is remarkably changed with the existence of impurity in C₃S or coexistence materials such as C₃A and/or C₄AF.

View full abstract

When the network-forming trivalent ions are introduced into the sodium silicate glasses, the network structure of glasses gives rise to the characteristic change in the binding state of oxygen. To elucidate the change, the electrical conductivity of 2Na₂O⋅xR₂O₃⋅(8-x)SiO₂ glasses (R: B, Al, Ga and La) has been measured in the wide temperature range from solid to molten states and the change of the binding state in oxygen with addition of a trivalent ion was analyzed through the shift of O1s peak by means of X-ray photoelectron spectroscopy, and the effects of the various trivalent ions on the conductivity were discussed with their structure. In the molten state (above T_g), the electrical conductivity of each system decreased with increasing content of various trivalent ions. Particularly for the network-forming ions such as B³⁺, Al³⁺ and Ga³⁺ ions, the conductivity decreased linearly with increasing [RO₄]^-Na⁺/∑Na⁺ ratio except Al/Na>1 and Ga/Na>1 and the values of conductivity coincided with each other in the same composition ratios. The results show that the network structure of each system retains the similarity of compositional influences on the mobility of Na⁺ ions, even though the different kinds of trivalent ions are introduced. In the solid state (below T_g), the conductivity did not simply depend upon the composition, but upon the glass structure in equilibrium with the frozen temperature (≅T_g) from melt to glass. The opposite behavior of B³⁺ ion and Al³⁺ ion of the conductivity in solid states can be explained by the difference of frozen temperatures. While the conductivity of the glass containing La³⁺ ion as the network-modifier decreased steeply with increasing content of the ion in both the solid and molten states. According to Anderson's model, it was assumed that the binding energy for Na⁺ ion in the activation energy increased in about 4-5kcal/mol with variation from 0 to 1 in R/Na.

View full abstract