Synchrotron X-ray Microdiffraction and Fluorescence Imaging of Mineral and Rock Samples

This method can help answer key questions in mineralogy and petrology about grain and phase distribution and mineral physics and chemistry. The main advantage of x-ray microfluorescence and x-ray microdiffraction is to provide correlated elemental and crystallographic analysis at the micron scale while accommodating sample with size up to 10 centimeter and without the need of any vacuum chamber. Samples can be thin sections, they can be embedded in epoxy, or they can even be whole unadulterated rocks, making sample preparation much simpler for this technique compared to other comparable techniques, such as EBSD.

Further, unlike EBSD, this technique allows us to measure natural and synthetic samples that have undergone plastic deformation, such as metamorphic rocks. To begin, attach a crystalline sample to the top half of a kinematic base so that the region of interest is vertically displaced relative to the base by at least 15 millimeters. Mount the top half of the base on the stage in the experimental hutch, and close the hutch.

Open the beamline control software, and initialize the program. Initialize the translation stage and slit controls. Once the beamline control program has been completely turned on, open the x-ray diffraction scan software.

Then, turn on the alignment laser, and open the stage setup menu. Translate the sample stage so that the sample ROI is within approximate visual focus of the rough alignment camera. Now, looking at the fine-focus camera, translate the upper z motor until the laser spot is aligned with the mark on the screen.

Then, open the slit setup menu. Select the appropriate slit size and, if needed, the monochromator energy. Use the mirror jog control to slowly tune the toroidal mirror to maximize the ion chamber count value.

Then, initialize fluorescence mapping, and set the measurement data file path. Provide energy ranges for up to eight elements of interest. Use the upper x and y motors to drive the sample stage to one corner of the area to be mapped.

Mark this as the start position for the x and y axes. Then, drive the stage to the opposite corner, and mark it as the end position for the x and y axes. Fill in the step size and either the velocity or dwell time for the scan.

Confirm that the map covers the sample ROI, and start the scan to map the sample with x-ray fluorescence. Once the x-ray fluorescence scan has finished, provide a folder name and naming scheme for data files in the x-ray diffraction scan window. Ensure that the upper x and y motors are selected.

Fill in the x and y start and end positions, step sizes, and pattern exposure time. Launch the mapping process, and wait for the scan to finish. To begin single crystal microdiffraction analysis, load a single crystal diffraction pattern into the analysis software, and subtract the detector background.

Then, open the calibration parameter menu, and load the appropriate calibration parameter file. Load a standard crystal structure. If necessary, also load a stiffness file with the third-order elastic tensor matrix for the material.

Then, open the peak search tool. Set the desired peak threshold, and run the automatic peak picking process. Manually add or remove peaks as needed.

Initialize the indexing process. If stress will be quantified, also load the stiffness file associated with the structure, and select the appropriate stress parameters. Initialize the strain calculation.

Then, open the automatic analysis procedure window. Set the first file in the map sequence as the image file parameters, and fill in the last file number. Provide a name for the file to be created.

Then, fill in the user directory, the path to the images, the file location where processed files should be saved, and the number of nodes to be used in the calculation. Generate and save the instruction file. Upload the instruction file to the cluster using a file transfer program.

Then, open a terminal window, and run the parameter calculation. Provide the instruction file name when prompted. Once the parameter calculation process has finished, copy the output file to the local computer, and load the file into the analysis program.

Display the map, and select the column that will correspond to the z-values of the 2D plot. Export the data to another plotting program if desired. To begin powder microdiffraction analysis, load a powder diffraction pattern into the analysis software, subtract the detector background, and load the calibration parameter file.

Hover the cursor over pixels in the pattern corresponding to a peak of interest, and note the displayed two-theta and chi values. Then, open the window for integrating the pattern as a function of two-theta, and fill in the two-theta and chi ranges. Select a Gaussian or Lorentzian fit, ensure that the fit is a good visual match, and fit the peak.

Next, open the chi-two-theta analysis window, and select the path to the files to be mapped. Fill in the start and end numbers from the sequence. Name the result file, and run the scan to map the previously fit peak onto each pattern.

Then, integrate one peak across chi, and map it across a 2D map. Export the data to another plotting program if desired. A moissanite sample thought to contain native silicon was first evaluated by x-ray fluorescence.

The intensity range excluded silicon and carbon, allowing the silicon carbide crystal to be distinguished as a low-intensity area compared to the surrounding calcium and iron-rich material. Laue x-ray diffraction patterns were then acquired. Initial indexing of a pattern within the sample body showed a better fit for the 4H silicon carbide polytype than the 6H silicon carbide polytype.

Most of the crystal was easily indexed with 4H silicon carbide. Manual examination of the remaining crystal area showed that this portion was better indexed as 6H silicon carbide. A 6H silicon carbide indexing map of this section had low success in one area.

Within that area, several overlapping diffraction patterns were observed, which indexed as at least three overlapping silicon grains. It is important to manually verify that an automated fit is performed correctly. Here, manual verification allowed us to determine that different parts of the sample actually have different crystal structures and that there's multiple silicon grains present.

Multiple local maxima occurred at the bases of the peaks, indicating multiple subgrains resulting from significant plastic deformation. Powder microdiffraction of an olive snail shell showed that the aragonite pattern was a good match. XRF maps of calcium and iron suggested compositional variation in line with the maps of the aragonite 040 peak width, d-spacing, integrated intensity, and orientation.

The apparent correlation between iron and the d-spacing and orientation was found to be a measurement artifact. Sometimes the measurements must be verified by other analytical tools. For instance, EDS measurements indicate that what we thought was compositional variation was in fact due to diffraction from texture aragonite layers.

After watching this video, you should have a good understanding of how straightforward data collection and processing is for x-ray fluorescence and microdiffraction at ALS 12.3.2. Once mastered, data collection can be very fast, up to 100 pixel per minute. Data analysis can take as little as five minutes when performed on the cluster.

This technique has successfully been used by researchers in order to analyze the strain of quartz and other silica polymorphs to look at the mineralogy of Roman and volcanic concretes, to look at meteorological composition, as well as for analysis of calcite and aragonite in various shells and corals. Complementary method like EDS or EBSD can be performed to answer additional question, such as quantifying elemental distribution. Don't forget that, as with all automated processes, data should be checked manually in order to verify that it was collected and processed correctly.