Abstract
Thermogravimetric analysis (TGA) is used to estimate calcite content in ore feedstock through the mass loss in the CaCO₃ decarbonation window (700–850 °C). In manganese ores this interval overlaps with the MnO₂ → Mn₂O₃ → Mn₃O₄ reduction chain and silicate dehydroxylation, making unambiguous attribution of the mass loss to carbonate decomposition impossible. X-ray fluorescence (XRF) provides only the bulk Ca content without phase assignment. Reliable calcite quantification is required for the energy balance of pyrometallurgical processing and CO₂ emission estimates at the pre-sintering stage. A cross-validation protocol is proposed to test the carbonate hypothesis. Since CaCO₃ → CaO + CO₂ releases 43.97 wt.% CO₂, the calcite mass fraction from TGA equals the mass loss divided by 0.4397. An independent estimate from XRF uses the molar ratio M(CaCO₃)/M(Ca) = 2.4973. Agreement between the two estimates within the relative discrepancy δ ≤ 5 %, derived from propagated measurement uncertainties, confirms that the mass loss is attributable entirely to decarbonation. The protocol is tested on two contrasting ores from Central Kazakhstan: carbonate-bearing Bogach (positive control) and silicate-bearing Zhaksy (negative control). For Bogach the two estimates converge at 14.33 wt.% (TGA) and 13.58 wt.% (XRF), δ = +5.2 %, confirming a calcite origin. For Zhaksy δ = +70.5 %, clearly violating the hypothesis; two independent checks corroborate this conclusion: the integrated DSC heat effect of 537 J g⁻¹ is sixteen times higher than expected for the XRF-based calcite content, and the dTG–DSC minimum offset of +77 °C (versus +9 °C for Bogach) indicates that the two signals originate from different processes. The protocol uses TGA and XRF and identifies samples for which naive TGA-based calcite estimation would yield an overestimate, providing grounds for directing them to phase-resolved analysis (XRD Rietveld, TG–MS).