Thermal expansion analysis is a most widely used method for determining the firing temperature of ancient ceramics, which infers the original firing temperature of ancient ceramics via analyzing the inflection point of the thermal expansion curve. However, for the pottery at an original firing temperature that is lower than the formation temperature of glass phase, the inflection point on the thermal expansion curve is actually the formation temperature of glass phase rather than original firing temperature. To solve a problem of inaccurate determination of low-temperature pottery firing temperature by a thermal expansion method, clay samples fired at different temperatures were prepared with clay as a raw material, and analyzed by X-ray diffraction (XRD), synchronous comprehensive thermal analyzer (TG-DTA), and scanning electron microscopy (SEM). In addition, the first derivative curve corresponding to the top rod thermal expansion curve of clay samples fired at different temperatures fitted to the dehydroxylation expansion peak of kaolinite, and the relationship curve between the dehydroxylation expansion peak of kaolinite and the original firing temperature was also obtained.

A clay as a raw material was processed by grinding, screening, drying, pressing, and sintering at different temperatures. Afterwards, the cylindrical samples with a diameter of 5 mm and a length of 25 mm were prepared with clay by polishing. The thermal behavior of clay raw materials was determined by thermogravimetry-differential thermal analysis (TG-DTA). The chemical composition, phase composition, and microstructure of the raw materials and the samples fired at different temperatures were characterized by X-ray fluorescence spectroscopy (XRF), X-ray diffraction (XRD), and scanning electron microscopy (SEM). The linear expansion or contraction process of the samples during the refiring process at different firing temperatures was tested by a thermal expansion instrument.

The XRF spectra show that the chemical composition of raw materials and samples at different firing temperatures is the similar. The XRD patterns and TG-DTA analysis of raw materials indicate that the dehydroxylation of kaolinite occurs at 300-600 ℃. Based on the XRD patterns, the phases in clay samples fired at different temperatures are consistent with those in clay raw materials. At < 900 ℃, the characteristic diffraction peak intensity of kaolinite in clay samples fired at different temperatures changes with the increase of firing temperature. The SEM images show that the typical layered structure of kaolinite in clay raw material begins to lose its structural hydroxyl groups and gradually transforms into metakaolinite, and this transformation gradually deepens as the firing temperature increases. The results of the thermal expansion test of clay samples fired at different temperatures indicate that there is a linear relationship between the dehydroxylation expansion peak of kaolinite on the first derivative curve of thermal expansion and its original firing temperature. This linear relationship can be expressed by a formula, i.e., A= -2.278 481 7×10^-6T+3.337 870 9×10^-3. To verify the accuracy of the formula, the thermal expansion test of clay samples is conducted using a push rod, and the calculated firing temperature is not significantly different from its original firing temperature, which are 41℃ and 35 ℃, respectively. The push rod thermal expansion method can be thus used to determine the original firing temperature of low-temperature pottery made from clay.

Before the glass phase formation temperature, kaolinite in the sample could lose the hydroxyl group in its structure and gradually transform into metakaolinite as the firing temperature increases, thus causing the expansion of the thermal expansion curve. The intensity of the thermal expansion peak of kaolinite dehydroxylation gradually decreased as the firing temperature increased. The first-order derivatives of the thermal expansion curves of the samples fired at different temperatures were taken, fiting the kaolinite dehydroxylation peak. This linear relationship with the firing temperature could be expressed by a formula, i.e., A= -2.278 481 7×10^-6T+3.337 870 9×10^-3. The clay samples fired at 550 ℃ and 650 ℃ were used for the thermal expansion curves, respectively. The clay samples fired at 550 ℃ and 650 ℃ were used to verify the reasonableness of the fitting formula. It was indicated that the firing temperature of low-temperature pottery made of clay could be determined by the top bar thermal expansion method.