Investigating Metals and Metalloids in Soil at Micrometric Scale Using µ-XRF Spectroscopy: A Case Study

Research output: Contribution to journalArticlepeer-review

1 Citation (Scopus)
27 Downloads (Pure)


Micrometric 2D mapping of distinct elements was performed in distinct soil grain-size fractions of a sample using the micro-X-ray Fluorescence (µ-XRF) technique. The sample was collected in the vicinity of São Domingos, an old mine of massive sulphide minerals located in the Portuguese Iberian Pyrite Belt. As expected, elemental high-grade concentrations of distinct metals and metalloids in the dependence of the existent natural geochemical anomaly were detected. Clustering and k-means statistical analysis were developed considering Red–Green–Blue (RGB) pixel proportions in the produced 2D micrometric image maps, allowing for the identification of elemental spatial distributions at 2D. The results evidence how elemental composition varies significantly at the micrometric scale per grain-size class, and how chemical elements present irregular spatial distributions in the direct dependence of distinct mineral spatial distributions. Due to this fact, elemental composition is more differentiated in coarser grain-size classes, whereas griding-milled fraction does not always represent the average of all partial grain-size fractions. Despite the complexity of the performed analysis, the achieved results evidence the suitability of µ-XRF to characterize natural, heterogeneous, granular soils samples at the micrometric scale, being a very promising investigation technique of high resolution.

Original languageEnglish
Pages (from-to)136-150
Number of pages15
Issue number1
Publication statusPublished - 2 Jan 2023


  • soil matrix
  • metal distribution per grain fraction
  • micro-X-ray elemental mapping
  • RGB clustering image analysis
  • k-means


Dive into the research topics of 'Investigating Metals and Metalloids in Soil at Micrometric Scale Using µ-XRF Spectroscopy: A Case Study'. Together they form a unique fingerprint.

Cite this