Development of a LIBS spectrometer for real-time chemical analysis of confined mining brines and slurries dataset

Raw data and analysis scripts for the data-derived figures (Figures 6–10) andthe detection-limit table (Table 3) of the paper The mineralogical intestine: development of a LIBS spectrometer for real-time chemical analysis of confined mining brines and slurries Christian Burlet¹, Bob de Waard², Tobias Pinkse³, Giorgia Stasi¹, Yves Vanbrabant¹, Jussi Aaltonen⁴, Tuomas Salomaa⁴, Kalle Hakonen-Milosevic⁴, Kaloyan Nikolov⁵, Claudio Rossi⁶, Eric Pirard⁷ Affiliations: ¹ Geological Survey of Belgium, Royal Belgian Institute of Natural Sciences, Brussels, Belgium ² Spectral Industries, Delft, The Netherlands ³ ReInvent Solutions BV, The Hague, The Netherlands ⁴ Tampere University, Tampere, Finland ⁵ 4DCoders, Tarnovo, Bulgaria ⁶ Centre for Automation and Robotics, Universidad Politécnica de Madrid-CSIC, Madrid, Spain ⁷ University of Liège, Liège, Belgium ## Folder layout Paper_data_and_scripts/ ├── 3D models 3D model of the LIBS slurry tube and 3D printed adapters ├── raw_data/ │ ├── scoria_slurry/ │ │ ├── test_water_.scorie.csv 580 LIBS spectra of the circulating │ │ │ Trooz Zn-smelting scoria slurry │ │ │ (K-UTEC rig, Ibsen UV channel 188–445 nm, │ │ │ 1 Hz, ~10 min; column 0 = wavelength, │ │ │ each further column = one spectrum) │ │ └── water_blank.csv tap-water blank recorded on the same rig │ │ before the scoria was added (single │ │ dark-corrected averaged spectrum, │ │ semicolon-separated: column 0 = │ │ wavelength, column 1 = counts); shown in │ │ gray in the short panel under the │ │ slurry spectrum of Figure 7 │ └── brine/ │ ├── Calibration series 2/ the synthetic KCl/LiCl series of the │ │ paper, 0–82 ppm Li / 0–263 ppm K │ └── Calibration series 3/ the mine-brine titration of the paper, │ 0–737 ppm Li / 0–1400 ppm K │ (one HDF5 file per concentration step; datasets "wavelength" │ [Avantes VIS-NIR, 438–980 nm, 4094 channels] and "intensity" │ [n_shots × 4094 single-shot spectra]; −9999 = dropped frame. │ File names encode the composition and settings, e.g. │ "12K_12Li_5Hz_006.h5", "30brine_5Hz_*.h5", │ "...afterRinse..." / "...afterManualClean..." = window controls. │ Every acquisition used the same ≈1 µs trigger delay, so the │ names carry no delay token. │ The folder names are those of the acquisition archive; an earlier │ low-range synthetic series, 0–10 ppm Li, was dropped from the │ paper as redundant and is not included here) ├── scripts/ │ ├── fig06_line_readout.py → Figure 6 (how a line is measured) │ ├── fig07_scoria_slurry.py → Figure 7 │ ├── fig08_09_10_brine.py → Figures 8, 9 and 10 │ └── lod_table3.py → Table 3 (calibrations and LODs) └── output/ created/filled by the scripts ## How to run pip install -r requirements.txt cd scripts python fig06_line_readout.py python fig07_scoria_slurry.py python fig08_09_10_brine.py python lod_table3.py Each figure is written to `output/` as a vector PDF (journal deliverable) anda 300-dpi PNG preview: ## Requirements Python ≥ 3.9 with `numpy`, `pandas`, `scipy`, `matplotlib`, `h5py`(pinned versions in `requirements.txt`).