Abstract
Measuring the carbon balance of industrial processes is crucial for quantifying carbon emissions in a green economy. While solid products can easily be collected and analysed, sampling gases and volatile compounds that are released to air can be more challenging. Mn oxides from raw materials are reduced in submerged arc furnaces in a reaction with carbon reductant (Eidem et al. 2025). This leads to exhaust hast that typically consists of 70 % CO and 20 % CO2 while the rest is mainly made up by H2 and N2. To measure the carbon emission in exhaust gas from manganese (Mn) production, we compared different methods to reduce the carbon to graphite usable for AMS measurement.
A regular treatment of the exhaust gas would be to combust and separate the resulting gases by freezing out water and CO2 and pumping non-condensable gases. In this case, the combustion was achieved by expanding the sample gas into a closed reactor with MnO2 as oxygen supplier. The reactor was then heated to 550 °C for the reaction (Janovics et al. 2018). The captured CO2 would then undergo reduction in a reduction process with Zn and Fe (Seiler et al. 2019).
As the regular reduction process for CO2 goes through a CO phase, we tried to reduce the gas mixture directly. The gas was expanded into the reactor of the Zn reduction system consisting of two heated tubes (Seiler et al. 2019). One tube is filled with Fe powder for reduction to solid graphite, and the other one with Zn powder for CO formation out of CO2, as well as silver foil for absorption of sulphur compounds. We show the details of these systems and compare the ¹⁴C results from the two methods used to reduce the sample showing that both methods give comparable results.