**5. Conclusions**

SEM-AM is a combined analytical tool of upgrading an SEM. It was initially developed and adopted for the fast and effective characterisation of metal-bearing ores and their processing products. A main driving force of SEM-AM development was to control and improve the effectivity of mineral processing.

The SEM-AM is organised in several successive steps (1–3) which can be controlled and adopted by the users. A measurement begins by the collection of a BSE image and its processing by various image analysis software routines (step 1). In step 2, the electron beam is focussed to induce EDS spectra at selected points according to the adjustments of the BSE image analysis. Step 3 is the classification of the gained sample EDS spectra against a list of approved reference EDS spectra. Different algorithms of the EDS spectra classification are realised by the various SEM-AM software providers.

Combinations of steps 1 and 2 lead to the principal SEM-AM measurement routines (1) point counting modal analysis; (2) particle analysis; (3) sparse phase search, and (4) EDS spectral mapping. These can be applied for the characterisation of natural and artificial electron-beam stable particulate, granular and solid massive materials. The routine (1) follows the traditional point counting method for estimation of the modal contents of different phases in a sample and provides no further information on particle properties and geometries. The particle analysis (2) measurement routines are designed for the analyses of grain mounts with 104–106 particles for addressing the special tasks in mineral processing and mining. This measurement routines allows the quantification of modal mineralogy, calculated assay, sizes and geometries of particles and grains as well as association and liberation of particles with di fferent chemical compositions within a bulk sample and without previous phase separations. This is not possible by conventional sedimentological particle size and size distribution analysis methods. The sparse phase search measurement routines (3) were developed to find rare phases like small grains of gold and platinum group minerals which is hard and time-consuming in a manual search. BSE triggers allow for focusing the SEM-AM sparse phase search on other interesting phases. The EDS spectral mapping methods appear as the most versatile routines to resolve intra-particle details, chemical zonation and phase relationships in combination with the full particle sizes and geometries.

The application of SEM-AM requires a high-quality preparation of the samples. Grain mounts with epoxy raisins in blocks and on glass, epoxy-embedded hard material fragments and epoxy-glued thin (30 μm) and thick (100 μm) sections on glass are the mostly used compounds. The detection and selection of a suitable epoxy type and mixture which does not evaporate under high vacuum conditions and remains stable under the electron beam after hardening is one of the principal challenges in sample preparation. The other di fficulty is the development of sample-specific multi-step polishing procedures to produce relief-free planar surfaces in samples containing phases of very di fferent hardness.

As demonstrated by several case studies in this contribution, the EDS spectral mapping measurement methods appear to have the most promising potential for novel applications apart from metal-bearing ores and mineral processing. This especially concerns the SEM-AM applications to artificial materials such as ashes, slags and ceramics. It has to be clarified that the complete identification of minerals and their denotation through the classification of their EDS spectra remains incomplete. For this, a full and reliable mineral identification requires additional XRD data. It also appears that XRD data identifies phases which appear in microcrystalline aggregates and not in larger grains. Many of the materials which potentially could be characterised by SEM-AM consist of microcrystalline, amorphous and glassy phases. In such cases, the generic labelling of reference EDS spectra and their subsequent target component grouping allow an interesting utilisation of the SEM-AM for novel studies not feasible by other analytical methods.

**Author Contributions:** Conceptualisation, B.S., D.S.; methodology, D.S., S.G.; investigation, B.S., D.S., S.G.; resources, B.S.; data curation, B.S., DS, S.G.; writing—original draft preparation, B.S., D.S.; writing—review and editing, B.S., D.S., S.G.; visualisation, B.S., D.S.; project administration, B.S., D.S., S.G. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the Helmholtz Institute Freiberg for Resource Technology, the Deutsche Forschungsgemeinschaft (DFG) Grant SCHU676/20 and through contract work for numerous enterprises. The Open Access Funding and APC was funded by the Publication Fund of the TU Bergakademie Freiberg.

**Acknowledgments:** The authors acknowledge the grea<sup>t</sup> expertise of Andreas Bartzsch, Roland Würkert and Michael Stoll in preparation of numerous standard and special geomaterials at Helmholtz Institute Freiberg for Resource Technology. The constructive comments of three anonymous reviewers considerably contributed to improvement of the manuscript.

**Conflicts of Interest:** The authors declare no conflict of interest.
