Research has shown that self-glazed ceramics with good thermal stability can be developed through the right combination of raw materials and firing temperatures.

It is well known that ceramics with good thermal stability can be achieved through either high thermal conductivity (such as with silicon carbide or carbonaceous/graphitic refractories) or low thermal expansion coefficients (such as with siliceous materials like quartz glass, cordierite ceramic materials or ceramic materials based on lithium silicates).^2-4 In most cases, however, the surface of the ceramic product must be covered with a thin layer of glaze to increase its stability against aggressive agents, to make it impermeable to liquids and gases, to improve the aesthetics of the items or for other reasons. This glazing process typically requires a number of additional operations and can consume a great deal of time, raw materials and energy.

With cordierite ceramic materials, the glazing process is even more difficult. ^2,4 The very low thermal expansion coefficient of these materials, which provides their high thermal stability, ^1,3 doesn’t allow the use of common glazes because they typically have much higher thermal expansion coefficients. Instead, the glaze must be carefully formulated to match the thermal expansion and elasticity of the ceramic body.

But what if an additional glaze weren’t required? What if, through the use of certain raw materials and chemical reactions, the ceramic body could essentially glaze itself?

This “self glazing” phenomenon has been achieved with some technical cordierite ceramics by introducing certain fluxes into the body composition. When fired at high temperatures, these fluxes contribute to the formation of a thin layer of glass fused at the ware surface. This glass film has a very similar composition to the respective ceramic material and prevents the hairline cracks known as “crazing”.^4,5 But is this effect limited only to technical cordierite ceramic bodies?

Experiments

To determine whether this self-glazing effect could be achieved in other types of porcelain bodies, a number of experiments were performed. Based on the previous theoretical considerations, experimental bodies were formulated using the raw materials and chemical compositions listed in Table 1. The formulations were prepared as both slip and plastic bodies. The shaped specimens were dried and then fired at 1300?C in an oxidizing atmosphere in an industrial gas tunnel kiln, with about 20 minutes holding time at the peak temperature and a firing cycle of about 17 hours (cold to cold). The resulting samples had a completely vitrified body of a bright, white-gray color.

The degree of vitrification of the fired samples was determined by measuring their water absorption capacity, accomplished by boiling each sample in water. Their whiteness was determined by measuring their total reflection.

The translucency of the samples was measured both qualitatively and quantitatively. To measure the qualitative translucency, the fired samples were simply examined visually. To measure the quantitative translucency, an intense, focused light beam was transmitted through the samples and was studied in comparison with similar samples of feldspar porcelain. A piece of colorless, completely transparent glass was used as a reference standard.

The lineal thermal expansion coefficient of the samples was determined using a “differential” dilatometer.

The appearance and development of the glaze on the surface of the samples during firing was monitored using a series of test specimens made of an experimental body, which was selected to be representative of the other samples. After drying, firing was conducted in an electric laboratory kiln at different temperatures between 650 and 1350?C. Samples were removed from the kiln at each 50-degree interval so that their surface characteristics could be examined. To ensure that each stage of the self-glaze development was measured accurately, reflection measurements were performed on each sample using a reflectance spectrophotometer with a device for determining the diffuse reflection. A method called “method with indicator,” which uses cobalt chloride as a chemical tracer (tracer agent) in the superficial layer, was also used.^1,8

The chemical composition of the fired samples was determined by X-ray fluorescence, while their mineralogical composition was determined by X-ray diffraction using Ni filter CuKa radiation. The microstructure of the fired samples was examined by scanning electron microscopy (SEM).

Results and Discussion

The chemical composition of the experimental ceramic bodies varied between the following limits (weight %): SiO₂, 50.0-53.0; TiO₂, 0.1-0.2; Fe₂O₃, 0.5-0.7; Al₂O₃, 33.0-37.0; CaO, 0.4-0.6; MgO, 10.0-12.0; Na₂O, 0.6-1.2; K₂O, 1.0-2.5. ^1,6-8

Because of their composition, the bodies could be considered quaternary systems if only those elements present in large quantities are taken into account (K₂O-MgO-Al₂O₃-SiO₂). In reality, however, these bodies are polynary (poly-compound) systems, in which all of their components, even those present in small quantities, have a role in achieving the equilibria phase in each system. The oxides existing in small quantities (TiO₂, Fe₂O₃, CaO and Na₂O) participate mainly in the interactions with the formation of eutectic melts and the developing liquid phases incorporating them; their presence can cause a more advanced degree of vitrification of each body during thermal treatment.

The following results were obtained on the fired samples made of the representative experimental body:

• water absorption capacity – 0.01%;
• whiteness – 66%;
• translucency – qualitatively, normal; quantitatively, as shown in Table 2;
• lineal thermal expansion coefficient, determined in the temperature interval of 20-1000?C (medium value) – a 20-1000?C = 3.64·10-6 ?C-1;
• mineralogical composition (weight %) – cordierite, 51.6; sapphirine, 3.2; corundum, 4.5; a-cristobalite, 1.9; amorphous and crypto-crystalline phase, 38.8.

The appearance and progressive development of the glaze on the surface of the samples, which was analyzed with reflection measurements at 50-degree intervals, is shown in Figure 1.

The SEM micrographs from some of the representative fired samples are shown in Figure 2. The micrographs illustrate a heterogeneous microstructure, which contains some solid phases and some mainly closed pores (air bubbles). These phases appear in different shades of gray (see Figure 2b) and could therefore be separately analyzed.

Each identified phase was chemically analyzed using the X-radiation it emitted, and the results are summarized in Table 3.

Correlating the information provided by the SEM micrographs with the results of the chemical analyses of the identified phases provides the following observations:

• The gray phase is composed mainly of cordierite and, probably, a glass phase formed in a small quantity around the cordierite grains;
• The dark gray crystals are well developed crystals of corundum;
• The light gray phase, which contains a great deal of silicon dioxide and some fluxing oxides, is a glass phase that represents the intergranular binder. Some of the representative fired samples were analyzed to determine their chemical and mineralogical compositions, as well their microstructures. All of the investigations were made on whole samples (marked A-B) and on thin sections that contained one of the exterior faces of each sample (marked A-outer side, and B-inner side), as well as on a zone right around the core of the sample (marked C). The results are presented in Table 4 and Figure 3.
  
  
  

  Table 4
  The chemical and mineralogical data shown in Table 4 denote an increase of potassium oxide and calcium oxide toward the exterior self-glazed faces of the samples (A and B). This is in correlation with the higher content of the amorphous and crypto-crystalline phase in these zones, and compares with the values of the respective characteristics determined for the core zone (C). The amorphous and crypto-crystalline phase is mainly composed of a glass phase, which originated in the liquid phase and was formed in these zones in a larger quantity. This fact is also suggested by the higher content of silicon dioxide determined for these zones.
  
  
  

  Figure 3. SEM micrographs of the experimental fired body.
  The SEM micrographs (Figure 3) show a heterogeneous microstructure, with different crystalline phases incorporated in a glass matrix, where some pores are also present. The thickness of the glaze layer formed at the surface of the sample is in the range of 0.02 to 0.04 mm.
  
  

  Conclusions

  The experimental ceramic bodies were completely vitrified after firing and exhibited a good translucency and whiteness, demonstrating that they were porcelain bodies. They were also multi-phase systems, in which the cordierite and the glass (amorphous and crypto-crystalline) phases are the prevailing phases; therefore, these ceramic bodies can be classified as cordierite porcelains.

  The two phase constituents (cordierite and glass) had a decisive influence on the properties of the ceramic bodies. The glass phase provided the body with compactness and translucency, while the cordierite phase ensured a low thermal expansion of the material. While thermal shock resistance can, in some cases, be positively influenced by porosity, the thermal shock resistance of these ceramic materials could only be explained by the presence of cordierite as a prevailing phase constituent, and not by the presence of pores in the ceramic body.

  During the firing process, self-glazing began after 1150?C. The optimum development of the glaze layer as a continuous, shiny film occurred around 1300?C. The glaze layer that was formed at the surface of the sample by self glazing was around 0.02 to 0.04 mm thick.

  Based on these experiments, it can be determined that self glazing can be achieved on cordierite porcelains by using compounds with low surface tension, such as potassium oxide and sodium oxide. When fired at the correct temperature, these compounds react with silica and form predominant silicated-alkaline melts. The impurities existing in small quantities in the body (TiO₂, Fe₂O₃) cause an additional reduction of the surface tension and also act as fluxes. In the cooling process, this melt becomes stiff, forming the self-glazed layer.

  These heat-resistant, self-glazed porcelain bodies exhibit excellent thermal resistance and can be used to produce items for relatively high temperature (900-1000?C) applications, such as laboratory items and cookware.
  
  

  About the Author

  Dr. Dipl. Eng. Aurica Goleanu is chief of the technical laboratory department at S.C. Apulum S.A., a company that manufactures porcelain products. She is a member of the Romanian Ceramic Society (CEROM), serves as vice president of the CEROM Board, and is also an affiliate member of the Institute of Materials in London. She has been awarded several patents for ceramic inventions. For more information about self-glazed porcelain bodies, contact Dr. Goleanu at S.C. Apulum S.A., Viilor Str., No. 128, 2500, Albia Iulia, Romania; (40) 58-813-803; fax (40) 58-817-112; or e-mail apulum@cristalsoft.ro.