Search

Towards the end of the 19th century and the early 20th century, France was the world's largest producer of antimony, especially due to the rich deposits in the Brioude‐Massiac area. Even though all the mining and smelting activities are long gone, there are still some remains of those activities. The most original of those is the use of roasting slags as aggregates in former plant walls or industrial building. They are macroscopically characterised by a black or red colour with a vesicular aspect similar to natural pozzolanas. Common petrography and mineralogy tools have been used to characterise those slags: optical mineralogy, scanning electron microscopy coupled with energy dispersive spectroscopy (SEM‐EDS) and X‐ray diffraction (XRD). To gain an insight into their mineral complexity, a QEMSCAN map has been produced. The mineralogy is typical of high‐temperature slags: cristobalite, quartz, mullite, fayalite‐forsterite series and spinel. The antimony content is quite high, up to seven percent, under the form of various antimony oxides and native antimony. LAY DESCRIPTION: In the Brioude‐Massiac area, France, antimony has been heavily mined at the end of the 19th century, giving rise to the generation of large amounts of wastes from the metallurgy of the ores. Those wastes (scoria and slags) have been used locally as aggregates to build walls. They consist of high‐temperature phases (SiO2 polymorphs: cristobalite and tridymite, dendritic spinel, glass) and illustrate an early example of the reuse of waste materials, nowadays termed as 'circular economy'. [ABSTRACT FROM AUTHOR]

Within the global trend to valorise various mineral wastes as substituents in Portland cement clinker raw feed, mining wastes are promising candidates. However, they might still contain high levels of metallic elements. Their fate in the kiln is not always understood as well as their incorporation within the various clinker's phases. This is especially the case for antimony. Its in situ microanalysis by the means of energy‐dispersive spectroscopy is tricky since several of its L lines (Lα1 = 3604,72 eV, Lα2 = 3595,32 eV) overlap with calcium lines (Kα1 = 3691,68 eV, Kα2 = 3688,09 eV). Hence, at low concentrations, it is not possible to visualise its characteristics peaks. Increasing the counting rate by increasing the acceleration current results in the generation of spurious sum peaks, rendering the analysis not viable. Wavelength dispersive spectroscopy (i.e. electron microprobe) allows a better spectral resolution and quantification of Sb in the clinker phases. In a Portland cement clinker doped with 1% of Sb‐bearing mining waste, the Sb2O3 content in belite and alite is in the 0.2–0.4 wt.% range as well as for C3A; in C4AF the content is higher, from 1.4 wt.% to 2 wt.%. However, there are microstructural evidence that Sb forms blebs from less than 1 μm in diameter up to 10 μm, included in calcium silicates. Hence, its incorporation in the lattice of calcium silicates is dubious. However, for ferrites and aluminates there is no microstructural evidence for remaining Sb‐bearing phases, suggesting a direct incorporation in crystal lattices. LAY DESCRIPTION: Energy‐dispersive spectroscopy (EDS) is widely used in materials science in order to get microchemical information at a micronic scale. This powerful technique allows the quick estimation of almost the whole periodic table with a single detector. However, some troubles might arise when the X‐ray emission lines of an element overlap with another. This is typically the case of Sb, which overlaps with Ca. The present article exemplifies this fact on the basis of the study of Portland cement clinker made with the incorporation of Sb‐bearing wastes. It shows that for dubious analysis, especially in the case of low concentrations, wavelength dispersive spectroscopy (WDS) should be preferred. [ABSTRACT FROM AUTHOR]