Geochemical compatibility and discrimination elements of magnetite on the eastern and western beaches of Taiwan

• Published in: Environmental earth sciences 2023-10, Vol.82 (20), p.483, Article 483
• Main Authors: Mitwally, Eslam Mohammed Ali, Yu, Bing-Sheng
• Format: Article
• Language: English
• Subjects: Ablation; Aluminum; Beaches; Biogeosciences; Calcium; Chemical composition; Chromium; Cluster analysis; Clustering; Compatibility; Copper; Earth and Environmental Science; Earth Sciences; Electron microscopy; Elements; Environmental Science and Engineering; Gallium; Geochemistry; Geology; Hydrology/Water Resources; Iron; Laser ablation; Light microscopy; Magnesium; Magnetite; Manganese; Mass spectrometry; Mass spectroscopy; Microscopy; Multivariate analysis; Nickel; Niobium; Optical microscopy; Original Article; Principal components analysis; Scandium; Scanning electron microscopy; Terrestrial Pollution; Titanium; Vanadium; Variables; X-ray diffraction; Zinc
• ISSN: 1866-6280; 1866-6299
• DOI: 10.1007/s12665-023-11165-x

Our primary purposes are to determine the genesis and the main source rocks that provide the magnetite on the beaches, as well as the compatible and discriminate elements of the studied magnetite using the multivariate diagram and principal component analysis (PCA). Optical microscopy, X-ray diffraction (XRD) and scanning electron microscopy (SEM) are primarily used to evaluate the purity of magnetite grains, and laser ablation–inductively coupled plasma–mass spectrometry is primarily used to determine the chemical composition of magnetite. In this study, two types of magnetite were distinguished, namely magmatic magnetite and hydrothermal magnetite. To determine the compatible and discriminant variables, three steps were performed. In the first step, the number of variables was reduced using a multivariate diagram. The results revealed that Zn, P, Ca, Cr, Cu, Na, and K, and Fe, Mg, Al, Ti, V, Mn, Co, Sc, Ni, Ga, Zr, Nb, and Sn served as compatible variables in magmatic and hydrothermal magnetite, respectively. In the second step, box-and-whisker plots, hierarchical clustering, and variable importance in projection (VIP) plots were used to identify discriminant variables based on the aforementioned compatible elements. The results indicated that in both hydrothermal and magmatic magnetite, Fe, Al, Ti, Mg, Mn, V, and Zn served as highly discriminant variables, while Zr, Sn, Sc, and Nb served as poorly discriminant variables. Other elements such as Co, Cr, Cu, Ca, Ga, Ni, K, and Na exhibited varying discriminant values. In the third step, the discriminant variables determined in the second step were arranged using the ranking model of PCA for each study sample. In conclusion, multivariate plots, box-and-whisker plots, VIP plots, and PCA are useful tools for determining the compatible and discriminant elements of magnetite.