Using AMICS for High Speed Differentiation of Mineral Phases of Similar BSE Intensity

Identification of minerals founded on BSE signal segmentation can be challenging as some minerals have comparable BSE intensities. Advanced Mineral Identification and Characterization System (AMICS) software is a powerful tool for overcoming problems that may arise due to similarities in BSE signal.

The AMICS software uses advanced imaging and analysis algorithms and control to certain scanning electron microscopes (SEM) as well as the QUANTAX energy dispersive X-ray spectrometry (EDS) system manufactured by Bruker.

A major feature of this software is the advanced image analysis to segment particles and guarantee precise mineral phase boundary identification and particle separation.

AMICS uses cutting-edge computer imaging systems to segment the BSE images, utilizing two variable co-dependent parameters to:

  • Control the sensitivity to gray level variation
  • Govern the size of the segmented region

This allows AMICS to detect subtle differences in the BSE image, produced by mineral change or image condition change.

By regulating the size of segmented region, the system is able to simulate X-ray maps (such as GXMAP in MLA) to identify minerals with related BSE gray levels (reduced number of X-ray points). This recent method is able to process big images and is very quick.

The AMICS advanced segmentation circumvents the problems arising from minerals such as Quartz-Albite, Chalcopyrite-Pentlandite and silicates. These minerals have all been examined using this methodology with a high degree of success.

Method

To achieve the measurements in an automated fashion and attain dependable results, an SEM with a backscattered electron (BSE) detector was set up in arrangement with the QUANTAX EDS system and the AMICS software. The system parameters were as follows:

  • SEM: Hitachi S3500
  • HV: 20 kV
  • Beam current: 90,000 cps on copper using EDS  
  • Electron detector: Solid state BSE
  • Magnification: 140x
  • EDS detector: XFlash® 6 | 10 with 126 eV at Mn Kα

Particle Mode

In Particle mode, BSE images are processed and segmented by means of the advanced image processing. Following segmentation each area is then analyzed by X-ray allowing the rapid and independent identification and classification of each segment. The mode allows detailed mapping to deliver output data such as association, liberation and size.

Mapping Mode

In the Mapping mode, the BSE image is used simply to remove the background which is the mounting medium. Each individual particle is then analyzed according to a grid of set X-ray points. Each X-rayed area in the particle is both identified and classified. This approach is a rapid method to attain bulk mineralogy and calculated assay data, but has restrictions in data output, i.e. no liberation or particle sizing.

Samples

The samples examined in this series of proving experiments include: 1) Ground particles from a mineral processing plant, mounted in a 30 mm diameter block of epoxy, polished and carbon-coated in routine sample fashion. 2) Polished thin sections composed of silicate minerals. The analyzed minerals include particles containing Quartz-Albite or Chalcopyrite-Pentlandite.

Quartz-Albite

The BSE image in Fig. 1a was acquired with an image resolution of 1.46 µm/pixel and 2–3 phases can be made out. The advanced particle segmentation using the Particle mode, however, shows numerous variants (Fig. 1b). These disparities are known to occur due to signal variations in beam stability or detector noise, as well as subtle changes in the mineral content or average atomic number.

BSE image of the sample showing little contrast (a), particle segmentation image showing the result of segmentation of fine variations in BSE intensity (b), resulting mineral map showing Quartz, Albite, K-feldspar and Muscovite after Particle mode analysis (c), mineral map showing Quartz, Albite, K-feldspar and Muscovite after Mapping mode analysis with 5 µm step size (d).

Figure 1. BSE image of the sample showing little contrast (a), particle segmentation image showing the result of segmentation of fine variations in BSE intensity (b), resulting mineral map showing Quartz, Albite, K-feldspar and Muscovite after Particle mode analysis (c), mineral map showing Quartz, Albite, K-feldspar and Muscovite after Mapping mode analysis with 5 µm step size (d).

Following X-ray analysis, identification and classification of each segment is performed, resulting in the particle mineral map in Fig. 1c. It shows four main mineral phases, as well as a multi-phase particle. By undertaking the analysis using the Mapping mode, the results obtained confirmed that the AMICS segmentation in Particle mode was successful in distinguishing Quartz and Albite as shown in Fig. 1d.

Chalcopyrite-Pentlandite

As in preceding measurements the Particle mode was applied in the same fashion to the samples of Chalcopyrite-Pentlandite. The BSE image was attained with an image resolution of 1.84 µm/pixel, and this was followed by processing and segmentation, followed by X-ray analysis, and finally mineral identification and classification.

Fig. 2 shows the BSE image (a) and the segmentation by small discrepancies within each particle (b). Each segment is again individually analyzed to provide a thorough identification and classification of the three mineral phases Quartz, Chalcopyrite and Pentlandite (c). The mapping results in Fig. 2d endorse the fact that the AMICS segmentation was able to positively distinguish Chalcopyrite and Pentlandite.

BSE image (a), segmentation image (b) and resulting mineral maps showing Chalcopyrite, Pentlandite and Quartz in Particle mode (c) and Mapping mode (d).

Figure 2. BSE image (a), segmentation image (b) and resulting mineral maps showing Chalcopyrite, Pentlandite and Quartz in Particle mode (c) and Mapping mode (d).

Fig. 3 shows further cases of the mineral phases Chalcopyrite, Pentlandite and Pyrrhotite.

BSE images (a, d), segmentation images (b, e) and mineral maps using the Particle mode (c, f) for two different particles in the sample.

Figure 3. BSE images (a, d), segmentation images (b, e) and mineral maps using the Particle mode (c, f) for two different particles in the sample.

Silicate minerals in a polished thin section

The instances in Figs. 4 and 5 validate the distinction between silicates such as Quartz and Feldspar or Pyroxene and Amphibole in solid rock sections. Fig. 4 illustrates the analysis of a single frame at 0.71 µm/pixel, which illustrates the variation in the Plagioclase composition across the sample.

BSE image (a), segmentation image (b) and resulting mineral maps after Particle mode (c) of a silicate containing thin section. A single frame was measured at 0.71 µm/pixel.

Figure 4. BSE image (a), segmentation image (b) and resulting mineral maps after Particle mode (c) of a silicate containing thin section. A single frame was measured at 0.71 µm/pixel.

Fig. 5 displays the analysis of 12 frames with an image resolution of 3.3 µm/pixel, with a total measurement time of 9 minutes 42 seconds. It proves the utility of this method to detect numerous phases that are not immediately distinguishable using just simple BSE analysis methods.

BSE image (a), segmentation image (b) and resulting mineral maps after Particle mode (c) of silicate minerals. 12 image frames were measured at 3.3 µm/pixel.

Figure 5. BSE image (a), segmentation image (b) and resulting mineral maps after Particle mode (c) of silicate minerals. 12 image frames were measured at 3.3 µm/pixel.

Conclusion

Minerals with comparable BSE intensities like Quartz-Albite and Chalcopyrite-Pentlandite can be successfully distinguished using AMICS’s unique, advanced segmentation methodology in Particle mode.

Similarly, silicate minerals also have similar BSE intensities but phases such as Quartz, Albite and different Plagioclase minerals can often be reliably discerned by using AMICS’s advanced segmentation technology.

By using AMICS software, the measurement of solid samples such as polished thin sections can be realized within a realistic time frame and with continuous joining of individual images.

Author:  Gerda Gloy, Application Scientist AMICS, Bruker Pty. Ltd.

This information has been sourced, reviewed and adapted from materials provided by Bruker Nano Analytics.

For more information on this source, please visit Bruker Nano Analytics.

Citations

Please use one of the following formats to cite this article in your essay, paper or report:

  • APA

    Bruker Nano Analytics. (2020, February 20). Using AMICS for High Speed Differentiation of Mineral Phases of Similar BSE Intensity. AZoM. Retrieved on May 30, 2020 from https://www.azom.com/article.aspx?ArticleID=13155.

  • MLA

    Bruker Nano Analytics. "Using AMICS for High Speed Differentiation of Mineral Phases of Similar BSE Intensity". AZoM. 30 May 2020. <https://www.azom.com/article.aspx?ArticleID=13155>.

  • Chicago

    Bruker Nano Analytics. "Using AMICS for High Speed Differentiation of Mineral Phases of Similar BSE Intensity". AZoM. https://www.azom.com/article.aspx?ArticleID=13155. (accessed May 30, 2020).

  • Harvard

    Bruker Nano Analytics. 2020. Using AMICS for High Speed Differentiation of Mineral Phases of Similar BSE Intensity. AZoM, viewed 30 May 2020, https://www.azom.com/article.aspx?ArticleID=13155.

Ask A Question

Do you have a question you'd like to ask regarding this article?

Leave your feedback
Submit