https://doi.org/10.3390/min12080975
“Examination of a fly ash derived from the combustion of a low-S, subbituminous Powder River Basin coal by Scanning Electron Microscopy (SEM) and High-resolution Transmission Electron Microscopy (HRTEM), both supplemented by Energy-dispersive X-ray spectroscopy (EDS), showed that the fly ashes were dominated by amorphous phases, Ca-rich plagioclase feldspars, Mg-rich phases, complex Ca-Mg-Al-Si-Ti-Fe grains, and trace amounts of REE-rich particles. Many of the particles were rimmed by a Ca-S, possibly a sulfate. HRTEM-EDS examination of a REE-rich particle proved it to be a mix of light- and heavy-rare earth minerals mixed with amorphous phases.”
“Further imaging and EDS characterizations were performed on the FIB slice noted above on a FEI Talos F200X TEM operating at 200 keV at the University of Kentucky Electron Microscopy Center. The EDS scans of areas of interest (4000–8000 eV) were examined by plotting the data with SigmaPlot version 14.5 and selecting energy (eV) and count ranges for enhancement. The complete EDS data is on Table S1. Fast Fourier transform (FFT) was used to determine the crystalline structure of the sub-micron grains.
3.2.1. REE-Lean Particle: Analysis at Virginia Tech NanoEarth
HRTEM examination of a rounded particle shows that it is cored by an amorphous Ca > Al > >P ≈ Mg ≈ Si > S > Ti mix and rimmed by a largely Ca-S sulfate (?) (Figure 6 and Figure 7). The concentrations of Mg, Si, P, and Ti (Figure 7 and Figure 8) are lower than Ca, Al, and S (Figure 6 and Figure 8), but there is a hint that Mg only mimics the Ca distribution in the particle core, not the rim. Titanium follows a similar distribution to Mg. Both Si and P partially follow the rim’s Ca-S distribution, particularly with the protrusion in the lower-left corner of the Figure 6 image.
Figure 6. TEM-EDS element overlay map showing individual (left column) and combined maps for Al, Ca, and S. Scale = 200 nm. After Hood et al. [15].
Figure 7. Distributions of Mg, Si, P, and Ti accompanying element map on Figure 6. Scale = 200 nm.
3.2.2. REE-Rich Particle: Analysis at University of Kentucky Electron Microscopy Center
Several REE-rich areas were examined in the particle shown on Figure 9. It is emphasized that, along with the particle shown on Figure 5, this is not a common find in this fly ash. The Figure 9 particle was selected following the preliminary SEM-EDS examination based on the promise of a diverse assemblage of REE-rich areas.
Figure 9. Areas 0940, 1213, and 1003 within a Y- and REE-rich fly ash particle. 1213 is within 940. The locations of Figures 17 and 18 are shown. HAADF (high-angle annular dark-field) image.
Areas 1213, 0940, and 1158/1201
Area 1213 is part of a larger area dominated by light REE but with some heavy REE and Y contributions (Figure 9). The high-angle annular dark-field (HAADF) image along with selected element maps is shown on Figure 10. In this example, La serves as a proxy for Ce and the other light REE and Er, along with Y, is a proxy for the heavy REE. The 0- to 10,000-eV range and the 4000- to 7500-eV REE range for area 1 within area 1213 is shown on Figure 11. In this case, the light REE and Gd show significant concentrations. Area 2, towards the upper edge of the particle (HAADF image on Figure 10), did not have significant concentrations of the REE.
Figure 10. Area 1213. Clockwise from upper left: HAADF image; La element map; Er element map; and Y element map.
Figure 11. EDS for area 1 within area 1213. The full counts for the 0 to 10,000 eV range are shown on the (left) and a restricted count range for the 4000–7500 eV range is shown on (right) (dashed box on full-range figure indicates the area of the right figure). The Dy, Ho, and Er “peaks” do not represent significant concentrations of those elements. The Cu peak belongs to the grid holder, not the sample.
Area 0940, the larger area including and adjacent to area 1213 (inset on Figure 12 HAADF image), is generally dominated by light REE, represented by La, with lesser concentrations of Y and heavy REE, the latter represented by Dy (Figure 12). Yttrium and Dy are less dense in the truncated oval La-rich area on the right side, the rectangular particle in the upper-left quadrant, and in the lower left corner of Figure 12. Yttrium and, to a lesser degree, Dy and the other heavy REE, show their most evident concentrations in the region between the higher La concentrations (right side of the lower-left quadrant). On the HAADF image, the latter region shows up as a mottled area, a notably different texture compared to the flanking brighter areas. The mottling might represent few-nm crystals dispersed in the region or it could also be FIB re-deposition or FIB-induced damage, along with small crystallites present. The 0- to 10,000-eV range and the 4000- to 7500-eV REE range for area 3 within area 0940 is shown on Figure 13. Areas 1 and 2 have similar EDS spectra to Figure 13, but their Dy and Er “peaks” are even less significant than the Figure 13 peaks.
Figure 13. EDS for area 3 within area 0940. The full counts for the 0 to 10,000 eV range are shown on the left and a restricted count range for the 4000–7500 eV range is shown on right (dashed box on full-range figure indicates the area of the right figure). The Dy and Er “peaks” do not represent significant concentrations of those elements. The Cu peak belongs to the grid holder, not the sample.
Figure 16. EDS for area 1 within area 1003. The full counts for the 0 to 10,000 eV range are shown on the left and a restricted count range for the 4500–7800 eV range is shown on right (dashed box on full-range figure indicates the area of the right figure). Yttrium appears as a shoulder on the P peak. The “peaks” for Eu, Tb, and Lu are not considered to be significant. The Cu peak belongs to the grid holder, not the sample.
Area 1158 with magnified area 1201 (1158 inset on Figure 12) consists of a mottled region capped by whisker-like fine crystals (Figure 14).
Figure 14. Area 1158 with inset shown for area 1201 (right).
Area 1003
The lower-left corner of the particle shown on Figure 9 contains at least three crystals of a P-Y-HREE-rich mineral (Figure 15). EDS (Figure 16) indicates that the mineral is a REE-rich mineral, possibly xenotime. The EDS spectra is unique in this study in showing significant concentrations of all of the even-number REE along with significant concentrations of most of the heavy REE. The exceptions to the latter trend are the insignificant concentrations of Tb and Lu. While Y has an apparent presence on the element map (Figure 15), the proximity of the Y Lα and Lβ (1.924 and 1.998 keV, respectively) to the P Kα (2.010 keV) means that Y overlaps with P and can only be seen as a shoulder on the low-eV side of the P Kα. For all of the P-rich minerals in this study (xenotime and monazite are possible minerals, but not definitively identified), caution must be applied in interpreting, and not over-interpreting, apparent shows of Y (or any other element). The P- and LREE-rich (with La serving as the proxy for the light REE) “nose” above the latter crystals could not be specifically identified, but, from the chemistry, it would appear to be a LREE phosphate.
Figure 15. Area 1003. Clockwise from upper left: HAADF image of area; HAADF image with xenotime crystals outlined; P element map; Er element map; Y element map; La element map.
Fast Fourier Transform (FFT) Analysis of Mineralogy
Most of the attempts at using FFT in this study to determine the minerals of particles were not successful because the particles were too thick or because multiple nanometer-scale crystals with overlapping orientations did not yield usable results. The crystal in region 0940 (Figure 17; location on Figure 9) is an exception. The lattice interlayer spacings of 3.25 Å and 3.12 Å match those of the (200) crystal plane and (120) plane, respectively, of Cerium orthophosphate monazite, CePO4, with a monoclinic unit cell (JCPDS 32–0199), which is the most thermally stable cerium phosphate phase up to combustion temperatures (>1400 °C) [22,33]. In contrast, the region shown on Figure 18 (location on Figure 9) appears to be amorphous, with no crystalline lattice fringes and no diffraction spots in FFT.
Figure 17. TEM image from STEM (Scanning TEM)-EDS region 0940 at low magnification (a,b) (scales = 200 nm and 20 nm, respectively) and high magnification (c) (scale = 10 nm). HRTEM image and the lattice interlayer space of the area blocked in STEM/EDS Region 0940.
Figure 18. TEM image from STEM-EDS region 1213 at low magnification (a,b) (scales = 1 µm and 100 nm, respectively) and high magnification (c) (scale = 20 nm), shows no crystalline lattice fringes, therefore it is an amorphous phase. The absence of diffraction spots in FFT is shown in the inset of (c).
“