Publikationsrepositorium - Helmholtz-Zentrum Dresden-Rossendorf

Modern SEM-EDS-based automated mineralogy such as Mineral Liberation Analysis (MLA) is a method in which BSE-image analysis and EDS analysis are combined.
Mineral Liberation Analysis is used for a rapid, spatially resolved, automatic, petrographic analysis of solid samples, often in applied mineralogy and metallurgical processing. Amongst other applications, this system can help to determine the chemical composition, mineral mode and micro textures in various sample types. Despite its fast acquisition time (6-12h for scanning of a full 4.5x2.5 cm sample) and the high-resolution nature of BSE imaging combined with the mineral identification capabilities of SEM-based automated mineralogy, it has rarely been applied to volcanic samples [1,2,3,4].
We present here work demonstrating the advantages of using MLA in volcanological studies, especially for fine-grained samples. We applied MLA technique to volcanic samples from Salina Island (Italy). The Salina Island, located in the centre of the Aeolian archipelago, had a rich eruptive history during the past ca. 245 ka that is divided in six eruptive Epochs [5]. Our research focuses on the last eruptive epoch, especially on the eruptive products of the Pollara tuff ring, namely the Punta Fontanelle Formation (Lower Pollara) and the Vallone del Pozzo Formation (Upper Pollara). The pyroclastic eruption from the Upper Pollara formation produced stratified deposits with dark basalt to andesite scoriae in the lower part and light coloured andesite to rhyolite in the upper part. The presence, in the Upper Pollara pyroclastic deposit, of white and grey-banded pumices of sub-alkaline basalt to rhyolite composition are the evidence of mingling/mixing processes between basaltic andesite and rhyolitic magma batches. Analysis of the pumice with SEM-EDS-based MLA (Fig. 1) provides significant information: discrimination of melts with different chemical compositions (rhyolitic in orange, andesitic in red and basaltic in blue in Fig. 1b), proportion of each melt, micro and macro textures between the different melts, mineral mode, mineral association, grain and vesicle geometry, mineral orientation, internal zonation in phenocrysts, reaction rims, etc.
These valuable data, combined with microprobe analyses of the volcanic glass and minerals provide clues on the mixing/mingling processes and the eruption dynamics. In conclusion, the application of SEM-based automated mineralogy e.g. MLA can add important information contributing to the understanding of the pretrogenetic and formation processes of volcanic rocks (and their micro textures).
References:
[ 1] Potter-McIntyre S L et al. 2014 J. Sediment Res. 84 875-892
[ 2] Rukhlov A S et al. 2013 Chem. Geol. 353 280-302
[ 3] Neave D A et al. 2014 J. Pet. 55 2311-2346
[ 4] Ayling B et al. 2011 GRC Transactions 35 301-305
[ 5] Lucchi F et al. 2014 Geol. Soc. London Memoirs 37 155-21