Thermal Analysis of High Melting Ceramics - METTLER TOLEDO

Thermal Analysis of High Melting Ceramics

Introduction 

High-melting ceramics are nowadays employed for very many different purposes. Traditionally, they are used for the construction of melting and sintering furnaces. The production of ceramic materials requires careful control of the starting materials. This article shows how thermal analysis can be used to characterize a typical starting material for the production of the fire-resistant lining of a furnace. The main constituent of the sample is MgO. Besides this it contains organically bound carbon (1.5%), graphite (10%), SiO2 (1.5%), CaO (1.5%), Fe2O3 (1.1%) as well as metallic aluminum (2.1%).

The measurements were performed with a TGA/SDTA851e coupled to an Inficon Thermostar QMS mass spectrometer (mass range 1-300) and also with a DSC822e.

 

Measurements

Ceramic Material (TGA, MS)

Figure 1 displays the results of a TGA-MS analysis. The sample was heated in an alumina crucible at 20 K/min in air (50 ml/min). The TGA curve shows several weight loss steps that can be assigned to different processes. The first step is due to the vaporization of moisture. The quantitative evaluation of this step yields a moisture content of about 0.33%. The next two weight loss steps correspond to the combustion of organically bound carbon and graphite. Organically bound carbon has an appreciably higher specific surface area than graphite and therefore burns at a lower temperature. 

The weight increase above 1000 °C is due to the oxidation of the metallic aluminum. Figure 1 also displays the MS fragment ion curve of m/z 44 (characteristic for CO2). The curve shows that considerable amounts of CO2 are formed in both steps in which combustion of carbon takes place. The ratio of the areas of the two peaks in the MS curve for CO2 (i.e. graphite to organically bound carbon) is expected to be about the same value as for the ratio of the step heights of the corresponding weight loss steps.

In actual fact, a value of 5.7 is obtained for the ratio of the peak areas in the MS curve. The ratio of the corresponding peaks in the TGA curve is however only 3.9. A closer comparison of the MS curve with the TGA curve shows that combustion of the graphite in the TGA curve apparently finishes at 1020 °C (dotted vertical line in figure 1), but that the peak in the MS curve is not complete until about 1060 °C (dotted vertical line 2 in figure 1). The reason for this difference is that oxidation of the aluminum already begins during the combustion of the graphite. 

This means that the step heights determined from the TGA curve for the combustion of the graphite and for the oxidation of the aluminum are both wrong. If one assumes that the first combustion step is due solely to the combustion of organically bound carbon and that the ratio of the CO2 contents measured with the MS represents the true ratio of the content of the two types of carbon, then a relative value of 10.3% is obtained for the content of graphitic carbon

This value agrees well with the information received on the sample. The difference between the graphitic carbon content (10.3%) determined in this way and the weight loss determined with the TGA (7.1%) corresponds to the weight increase that could not be directly measured with the TGA because of the overlap of the combustion of graphite and the oxidation of aluminum. The total step height for the oxidation of aluminum is therefore the sum of this difference (3.2%) and the measured weight increase (1.6%), i.e. a total of 4.8%. Using the stoichiometric ratio for the oxidation of aluminum, this gives a value of about 2.5% for the aluminum content, which is in good agreement with the reference value of 2.1%.

Results and Discussions

Table 1 summarizes the results obtained with the different evaluation methods. The results show that the correct value of the weight loss step can only be determined with the aid of the MS data. Even then, the contents calculated for the two types of carbon still differ slightly from the reference values. The reason for this is that the two combustion processes partially overlap. It would of course be possible to separate the two processes by using a lower heating rate. 

The aluminum content can also be quantitatively estimated from the “calibrated” SDTA signal. The problem here is the overlap of the two exothermic combustion processes and the endothermic melting process of the aluminum. If the sample is measured by DSC in an inert atmosphere, then not only the aluminum content but also the quartz content can be determined.

Thermal Analysis of High Melting Ceramics | Thermal Analysis Application No. UC 184 |  Application published in METTLER TOLEDO Thermal Analysis UserCom 18