Skip to content

X-ray Mapping Features in APEX 2.0: Example Analysis of a Geological Sample

Introduction

X-ray mapping allows a visible understanding of elemental distributions within a sampling area. Advanced and smart mapping features have been incorporated into APEX™ 2.0 to give a deeper understanding of the material's chemical nature and ensure ease-of-use. In this application note, a granite sample has been chosen to demonstrate how to employ various mapping features in APEX 2.0 to get high-quality and accurate data as quickly and efficiently as possible. Data was collected using a variable pressure microscope equipped with an EDAX Octane Elite Silicon Drift Detector.

Dynamic Element Mapping

Minimal mapping setup parameters are required before collection. Once the user chooses the map quality, the amp time is automatically selected based on the count rate, and the duration is determined based on average counts per pixel. Users also have the option to set up mapping parameters manually. During mapping data collection, the Dynamic Element Mapping feature gives users the ability to add or remove elements in live mapping mode to show only the elements of interest. Elements, element lines, and user-selected regions of interest can be easily edited during live collection (Figure 1). This feature eliminates the display of element maps that are not needed and gives more representative quant results while collecting with the correct element list.

Elements, element lines, and user-selected regions of interest can be easily added or removed in live mapping mode with Dynamic Element Mapping.
Figure 1. Elements, element lines, and user-selected regions of interest can be easily added or removed in live mapping mode with Dynamic Element Mapping.

Montage Large Area Mapping

Granites are coarsely crystalline igneous rocks. This sample consists of many phenocrysts that are typically over 2 mm in size. Montage Large Area Mapping allows us to map an entire phenocryst in the sample at a user-defined resolution. Low resolution can be used for sample overview, while high resolution captures the sample's full detail. This is done by applying stage movements to collect individual maps through a grid pattern over the phenocryst and stitching them into a montage. The montage maps give a good display of elemental distribution in such a relatively large sampling area, and high-quality spectra can be extracted from small features. An easy-to-use setup wizard with step-by-step instructions guides the user through the entire process.

a) Si K and b) P K montage maps of an entire heart-shaped phenocryst in the granite sample.
Figure 2. a) Si K and b) P K montage maps of an entire heart-shaped phenocryst in the granite sample.

CompoMaps - Live Net Mapping

The montage map of P indicates a large amount of phosphorus-rich grains distributed inside the phenocryst (Figure 2b). These grains are calcium phosphate, and they sometimes coexist with zirconium silicate in igneous rocks. P K and Zr L lines are heavily overlapped with only 29 eV of energy difference, so the region of interest (ROI) maps of these two elements are almost identical if both minerals are present. CompoMaps, the live net mapping feature, can perform background subtraction and peak deconvolution to separate P K and Zr L during mapping for accurate representation. The user can switch between ROI and NET display at any time during live acquisition. To demonstrate this feature, a rock sample of gabbroic nature with both calcium phosphate and zirconium silicate was mapped (Figure 3).

Live ROI maps of a) P K (green) and b) Zr L (red). The two maps are almost identical due to heavily overlapped ROIs. Live NET maps of c) P K and d) Zr L. The two elements are separated out by live background subtraction and peak deconvolution. d) An overlay of P K and Zr L live NET maps show the two elements exist in separate phases.
Figure 3. Live ROI maps of a) P K (green) and b) Zr L (red). The two maps are almost identical due to heavily overlapped ROIs. Live NET maps of c) P K and d) Zr L. The two elements are separated out by live background subtraction and peak deconvolution. e) An overlay of P K and Zr L live NET maps show the two elements exist in separate phases.

Map Rebuild and EDS Quant Maps

If any elements are still missing during live mapping, they can always be added back through map rebuild. Other than rebuilding maps as ROI or NET, quant maps can be created to investigate concentration variations between different minerals in the granite sample. A full quantification routine runs on every pixel, and the maps show in Wt% or At%. With the scales attached to the quant maps, the user can easily estimate the concentration based on color shades. Mineral types can be determined according to the Si concentration change from zero, intermediate to high (Figure 4).

Wt% map of Si K from the center of the phenocryst shown in Figure 1. The brightest areas in the image can be identified as quartz (silicon dioxide) due to the Wt% indicated by the color scale. Si Wt% in stoichiometric quartz is 46.7%.
Figure 4. Wt% map of Si K from the center of the phenocryst shown in Figure 1. The brightest areas in the image can be identified as quartz (silicon dioxide) due to the Wt% indicated by the color scale. Si Wt% in stoichiometric quartz is 46.7%.

Max Pixel Spectrum

The sum spectrum is beneficial for identifying the primary elements in the sample. Still, it tends to hide minor elements in inclusions that only make up a small contribution relative to the primary matrix. A maximum pixel spectrum can be built to identify these features by finding the pixel with the highest number of counts for each point on the energy axis and using that as the intensity for the processed, artificial spectrum. Figure 5a shows that the Mn Kα peak is not visible in the montage map's sum spectrum but sticks up in the maximum pixel spectrum. With this information, Mn can be included in the rebuilt montage elemental maps (Figure 5b). Spectra extracted from these small inclusions indicates they are iron-titanium oxides with approximately 5 Wt% of Mn. Because Fe is abundant and Ti is spatially widely distributed in the phenocryst, these tiny oxide minerals are not obvious in Ti and Fe's montage maps. The maximum pixel spectrum is noisy since each data point is based on a single pixel, but it highlights elements below the signal-to-noise ratio in the sum spectrum.

a) A max pixel spectrum (cyan) of the montage map shows a tiny Mn Kα peak that is not visible in the sum spectrum (red outline). b) A rebuilt montage map of Mn K indicates small inclusions with minor Mn inside the phenocryst. c) A montage map of Ti K. The iron-titanium oxides with minor Mn are not obvious in this map.
Figure 5. a) A maximum pixel spectrum (cyan) of the montage map shows a tiny Mn Kα peak that is not visible in the sum spectrum (red outline). b) A rebuilt montage map of Mn K indicates small inclusions with minor Mn inside the phenocryst. c) A montage map of Ti K. The iron-titanium oxides with minor Mn are not obvious in this map.

Smart Phase Mapping

The montage map of Si K displays a wide range of color shades, indicating different types of silicates and other non-silicate minerals inside the phenocryst (Figure 2a). Using the Smart Phase Mapping feature in APEX 2.0, the generic phase maps can be quickly created along with live acquisition for fast characterization of minerals. The phases can be renamed or combined during collection or post-processing, if necessary (Figure 6). The user can control phase separation tolerance. During live acquisition, it is possible to identify the phase further using either the quantification of the extracted phase spectrum or a pre-defined phase library. Users also have the option of using Spectrum Match during post-processing. Minimal operator interaction is required from beginning to end. These phase analysis methods also work in other applications, such as basic inclusion analysis in metals and most other phase-related applications.

Phase map and elemental maps of the blue phase from a sampling area inside the phenocryst. The blue phase has been renamed quartz because it is composed of Si and O and confirmed by extracted spectra.
Figure 6. Phase map and elemental maps of the blue phase from a sampling area inside the phenocryst. The blue phase has been renamed quartz because it is composed of Si and O and confirmed by extracted spectra.

CPS Mapping and Normalization

For this sample and other geological samples, cracks may be created in the sample by geological processes. Also, surface topography may be introduced due to the different polishing resistance of minerals. The quality of X-ray maps is often affected by these conditions that can result in count rate changes. Previously, it has been left to the user to decide and interpret the effects of topography on the sample analysis. The Counts Per Second (CPS) map feature in APEX 2.0 will now assist in these often difficult to understand interpretations. The CPS map provides a visual representation of the X-ray count rate at every pixel in the dataset. The brightest pixels indicate the highest count areas, and dark or black areas indicate little or no X-ray counts. It is a quick and easy check of the variations in count rate across the sample area. For example, the black regions in Figure 7a reveal a lack of counts due to topography. These regions can be compensated for by applying a CPS normalization (Figures 7b and 7c), another benefit of the CPS smart feature.

a) A CPS map of the center of the phenocryst. The black triangle region in the center is a crack on the surface. The dark region along the left boundary of the oblate grain in the top is due to surface topography created by different polishing resistance. b) The original ROI map of O K. c) A ROI map of O K after CPS normalization. The variations in count rate due to topography are compensated.
Figure 7. a) A CPS map of the center of the phenocryst. The black triangle region in the center is a crack on the surface. b) The original ROI map of O K. c) A ROI map of O K after CPS normalization. The variations in count rate due to topography are compensated.

Conclusion

In conclusion, this kind of analysis applies to a wide range of real-world materials and applications. The ease-of-use, flexibility, and customization of the mapping features in APEX 2.0 give users a variety of powerful tools to better understand their samples and showcase their data.