The aim of this study was to test the possibilities of using microsilica, B2O3, talc and borosilicate glass as antioxidants in magnesia carbon refractories, instead of the most common B4C, Al and Si due to cost reasons. The results of these investigations are compared to an antioxidant free reference sample, whose recipe is already in use by a company. To test if the assumed carbide phases between carbon and the antioxidants, which are one of the reasons for the use of antioxidants in magnesia carbon refractories, are formed, pretests with smaller samples have been done. Therefore, small tablets consisting only of carbon black and one of the antioxidants to be tested have been fired at different temperatures. These burned samples have been tested according to the phase composition using reflected light microscopy, XRD and SEM. As a result of these investigations, as well as the calculations done using the software FactSage, the carbide formation has been confirmed and the firing temperature for the following laboratory tests have been fixed to 1500°C and 1600°C.
To test the applicability of these antioxidants (talc, borosilicate glass, B2O3 and microsilica) in an already in use refractory MgO-C recipe, refractory cylinders using the given recipe however, with different antioxidants have been produced. The number of added antioxidants has been calculated according to the amount of metallic Al used in a given reference recipe. They have been chosen, that with each oxide addition an equivalent amount of metallic Si or B was added to the recipe. Due to a pressing process with reduced degassing, significant cracks running parallel to the horizontal axis of the cylinder have formed. To avoid these cracks, the content of the binder has been reduced in the sample containing B2O3 as an additive. As this reduction did not lead to a decrease of the crack formation, the original binder content was used again in all the other samples pressed. The pressed cylinders have been dried for 10 hours at 150°C, and afterwards they have been burned using reducing conditions (embedding in coal breeze) and a specific temperature program. The burned samples have been tested according to their mechanical properties, as well as their phase formation. Compared to the antioxidant free reference sample, all the samples showed a lower density and a higher porosity. Talc and borosilicate glass containing samples showed the lowest decrease of the density, approximately 0,1 g/cm3 or even less. B2O3 addition led to the highest decrease of density of around 0,6 g/cm3. The volume expansion of microsilica containing samples raised from 0,35% at 1500°C firing temperature to 2,8% at 1600°C. All the other additives showed similar values of the thermal expansion at both firing temperatures. The volume expansion of the B2O3 containing cylinder was 12,5%, which was also visible due to the strong crack formation. Each of the samples had a higher Youngs modulus at 1500°C firing temperature than the reference sample, whereas borosilicate glass reached the highest values. At 1600°C talc, microsilica and B2O3 showed similar values as the reference sample, borosilicate reaches more than double of those values.
The investigations concerning the phases formed during the firing process have been done by using a reflected light microscope, XRD and SEM. To identify the phases formed in the sample, EDS has been used at the SEM. A formation of Mg3B2O6 (M3B) within the boron containing cylinders has been investigated at the SEM, which has also been calculated in advance using the software FactSage. This phase formation has only been detected in cylinders with B2O3 as the antioxidant, using XRD. Forsterite has been calculated and detected in all silica containing samples (talc, microsilica and borosilicate glass). A formation of SiC could only be observed with a microsilica addition. Impurities of other elements were inserted in the samples either via the magnesia or the carbon fraction, as they could also be detected in the antioxidant free reference sample using SEM.