Quantification of Mineralogy with Sigray's XRF Microscopy

Introduction 

Elemental spatial distribution and relative elemental quantification techniques are essential for mineralogy research in the natural resource (O&G and mining) industries. This information is often needed at trace levels and is used to extract as much valuable material as possible and to inform ore processing.

The AttoMap’s automated mineralogy capabilities complement existing approches, including: scanning electron microscope (SEM) based systems, laser ablation inductively coupled mass spectrometry (LAICP-MS), and nanoscale secondary ion mass spectrometry (nanoSIMS). AttoMap provides down to 3-5 µm resolution, sub-ppm sensitivities, and software featuring machine learning algorithms for quantitative segmentation of grains and open-box extensibility.

Current Mineralogical Approaches: MLA and QEMSCAN

Scanning electron microscopes (SEM) based Automated Mineralogy systems have become one of the key methods for characterizing mineralogy. In these systems, an electron beam is stepped across a polished sample surface to excite x-ray fluorescence (XRF), which is recorded as a function of the beam position. This results in a mineralogical map at microns scale resolution as shown in the left-hand side of Figure 1. 

The major limitation of these SEM-based techniques is their sensitivity for trace (e.g. <0.1%) elements due to the large bremsstrahlung background inherent in electron excitation.  

There are some additional challenges to these techniques, including: interference artifacts, matrix dependency, and variability due to conditions (ablation and analytical count times) that can mask or mimic trace element distributions.

Novel Approach with Sigray AttoMap XRF Microscope

AttoMap microXRF was developed from patented x-ray source and optics technologies to enable synchrotron-like microXRF performance in a laboratory system.

For mineralogical investigations, the system features several advantages including:

1. A patented multi-target x-ray source allows users to optimize fluorescence signals of interest and detect trace elements at the sub-ppm level (see Fig. 2).

2. A large sample stage, allowing large, intact specimens to be scanned (up to 300mm travel on the ambient 200 series and 100mm travel on the vacuum 310 series). 

3. Straightforward sample preparation: The large working distance of Sigray’s proprietary optics allows for imaging samples even having some topography, such as powders and particles, and does not require the polished surfaces required by SEM-EDS. No additional preparation is required (e.g., resin embedding, polishing, or carbon coating). 

4. Variable spatial resolutions from 5 to 100 μm, providing flexibility in trading off resolution with FOV/throughput.

5. Mineralogical software analysis tools: Sigray’s software implements k-means clustering (an unsupervised machine learning algorithm) to segment grain boundaries by mineralogy. Grains can be segmented by XRF data alone or by additionally incorporating the correlative optical microscopy images acquired inside the AttoMap. The suite of tools includes an easy-to-use GUI interface and Jupyter python notebooks, allowing easy extension of algorithms and open source collaboration. 

In this blog, we used an AttoMap-200 ambient system to determine the mineralogical composition of a rock sample courtesy of Dr. Sakthi Chinnasamy, IIT Bombay.