Ian Szumila
University of Rochester
Earth and Environmental Science Department


In apatite, the mobility of certain elements in increased when under electron microscopy. This effect is also highly anisotropic, dependent on crystal orientation. This effect has been seen in the past for F and Cl in Stormer et al. (1993) For this study, a Durango apatite and an unidentified but likely flourodurango were sectioned along their c and a-b axis. Samples were first mounted and coated with carbon using the machines in the prep lab. The four samples were examined and imaged in SE2, and In-lens detection. Backscatter detection was also used in characterization of samples. X-ray analysis was performed several times on each sample. These analyses were taken several times in the same location for three minutes with times of beam blanking in between. The temporal variation of the anions, fluorine, chlorine, and sulfur were plotted in excel. Other possible variations were investigated. Similar graphs were found for fluorine and chlorine as had been researched before. For sulfur, the data did not resolve into an identifiable trend. The sample was also analyzed for general compositional information using an LA-ICP-MS. Some laser holes made on the sample by that technique were characterized by Atomic Force Microscopy (AFM). Finally, some images were colorized to make them easier to view.


The apatites were sectioned off their a-b and c axes. Then, they were mounted in epoxy. Next, they were polished with 600 grit sand paper, 1 um alumina and finally, colloidal silica. This sample was taken to the SEM prep lab. By using the carbon vaporizer, a very thin layer of carbon was deposited on top of the sample. Conductive tape was placed around the outside of the sample. A small piece of conductive tape was attached to a pin that was then stuck to the bottom of the sample. The outer rim of sample tape and the bottom the sample were also painted with conductive graphite. All these steps create a ground for the sample in the SEM and thus mitigate charging.


The images below were taken using the various detectors on the SEM as well as one with a camera outside the SEM. Using techniques and detectors such as SE2, in-lens detection and backscatter detection, characterization of the surface of the apatite samples was possible.

Fig 1. Picture of sample outside SEM.

Fig 2. Back-scattered Electron image of an Apatite

Fig 3. In-lens images of laser holes

Fig 4. An example of volatization damaging the sample during analyses.

Figure 1 is an image of the samples and sample stage to provide a view as to how the samples were mounted and arranged. Figure 2 shows a backscattered electron image of one of the apatite samples. When imaging with back-scattered electrons, lighter areas represent regions of with a greater density of high Z elements. From this image, the light area is the apatite while the dark area is the epoxy mount. Figure 3 is using in-lens detection the SEM. It also shows two different sets of laser hole. Three large ones are present in the center of the sample. These were the ones used to gather compositional information from this sample via LA-ICP-MS. These holes were created using a 40 um spot, fluence of 11.81 J/cm^2 with 220 shots at a rep rate of 10 per second. As such they are likely too deep for effective AFM imaging. Five much smaller laser holes can be seen in the left of the image. Each of these was created with only very few laser pulses and are the main laser holes that were characterized by AFM. Figure 4 shows beam damage to the sample. It is unknown exactly what this is from. It could be the carbon and silicon coated being melted or blown off during the analyses from the strength of the electron beam or a large migration of volatile. It often appeared at the edge of the area being analyzed midway through.


When an electron hits an atom, it often excites that atom's electrons. When an atom's own electrons move back down towards a lower orbital, they have to expel any energy they have gained. They do this by emitting an X-ray. Since the orbitals can only be traversed discretely, the energies of the x-rays emitted are characteristic of different kinds of atoms. By looking a spectrum of counts vs. the energy of x-rays observed, it is possible to investigate the composition of a sample. This was taken slightly further for the purposes of this research. Several spectra were taken at single locations, to see how the composition of the sample varied with time. By then plotting certain compositions (wt%) vs. beam exposure time, it should be possible to see if and how various elements might be inclined to migrate and escape under the influence of the electron beam.

The accelerating voltage used was 15 kv and, for size, a 3 um box was typically used. Beams were collected using a live time of 6 seconds, although beam dead time would occasionally vary between 30% and 70% with a few seconds of beam blanking in between to save the collected spectra. These conditions are very similar to those used in the Stormer et al. paper. However, the current used in the Stormer et al. paper was typically around 15 nanoAmps. Despite, using the largest aperture, at 120 um, the specimen current monitor on the SEM typically gave a readout of somewhere between 3 and 7 nanoAmps. This may be because the SEM here uses a field emitter, and is not a filament.

Figure 5. A spectrum from the c-axis of one of the apatites.

Above is an example of the kind of spectra received from the apatites via x-ray analysis. The Oxygen and Phosphorus peaks are clearly noticeable. The two peaks near the end of the graph are both representative of apatite. The F peak is definitely noticeable while the sulfur and chlorine peaks are approaching the detection limits of the SEM.

There is a feature in the EDAX software called Quant. This feature examines the compositional variety present in the spectra and does ZAF corrections. These corrections each correspond to a different interference in the returned spectra. The Z correction is a correction for elements of higher atomic density. The A and the F corrections correct for absorption and fluorescence as the x-rays travel through the sample and to the detector. After performing all these corrections and taking into account peak heights, the software will return the wt% of different elements present in the sample.

Figure 6. Durango Flouroapatite, C-axis, F migration.

Figure 7. Durango Flouroapatite, C-axis, Cl migration.

Figure 8. Durango Flouroapatite, C-axis, S migration.

Figure 9. Durango apatite, C-axis, F migration.

Figure 10. Durango apatite, C-axis, Cl migration.

Figure 11. Durango apatite, C-axis, S migration.

Figure 12. Durango Flouroapatite, A-B-axis, F migration

Figure 13. Durango Flouroapatite, A-B-axis, Cl migration

Figure 14. Durango Flouroapatite, A-B-axis, S migration.

Figure 15. Durango apatite, A-B-axis, F migration.

Figure 15. Durango apatite, A-B-axis, Cl migration.

Figure 16. Durango apatite, A-B-axis, S migration.

Excel Plots of Compositions of Various Element Weight Percents vs. Beam Exposure Time

In Figure 6-16, we can see how the different anions in apatite vary with exposure to the electron beam. Figure 6 is very similar to what was found by Stormer et al and demonstrates preferential F migration in apatite along the c-axis (vs. the a-b axis, Fig 12). Similar ideas can also be seen for chlorine in figures 7 and 13 here, when compared to figure 8 in Stormer et al. Unfortunately sulfur does not seem to show any sign of migrating, with the compositional points not seeming to vary in beam exposure time. The exception to this is figure 16 which seems to have a strong downward trend in the last 50 seconds or so of beam exposure. Further research would be needed to confirm if this signal is actually real and not just noise.


Figure 17(a). A single laser pulse on the sample.

Figure 17(b). Simple section across figure x(a)

Figure 18(a). Two laser pulses on the same spot on the sample.

Figure 18(b). Simple section across figure x(a)

Figure 19(a). Three laser pulses on the same spot on the sample.

Figure 19(b). Simple section across figure x(a)

Figure 20(a). Five laser pulses on the same spot on the sample.

Figure 20(b). Simple section across figure x(a)

Figure 21(a). Ten laser pulses on the same spot on the sample.

Figure 21(b). Simple section across figure x(a)

AFM images of laser holes along with simple sections along their surfaces.

Since LA-ICP-MS had been done to study the compositional of the sample, Atomic Force Microscopy was used to examine laser holes made in the sample. Atomic Force Microscopy is a technique that involves moving an atomically thin tip across a sample. The back of the tip has a mirror attached to it that a laser is reflected onto and then into a four quadrant detector. By using the tip in this way as an oscillating cantilever, it is possible to get atomic scale resolution of topography of a sample. The AFM is very sensitive to small amounts of noise and, as such, sits on an anti-vibration table.

Most of the spectra above were collected using a 30 um by 30 um box and at hertz per line of .1000 meaning the SEM took a very long time to generate most of these images. Using the SEM software, it is possible to correct drift using features such subtract surface, subtract sphere and three-points leveling. Some of which has been used on these images. The images on the right were generated by the AFM software using the "simple section" tool which allows you to drag a cross section along the surface of the image and produce a graph of the height found in that cross section.


A similar pattern was seen to fluorine migration as had been seen before in the Stormer et al. paper. The migration of chlorine also seems to be similar to what was found before but the results were much noisier. Since sulfur is an anion, a similar pattern might be expected, although there may be other confounding factors. From the graphs studied, it does not seem likely sulfur was affected by beam exposure but figure 16 hints out the possibility that it could be. AFM was useful for determining how much the laser impacted the apatite grains.


1. Stormer et al. "Variation of F and Cl X-ray intensity due to anisotropic diffusion in apatite during electron microprobe analysis." American Mineralogist, vol. 78, pp. 641-648. 1993.

2. Sha and Chappell "Apatite chemical composition, determined by electron microprobe and laser-ablation inductively coupled plasma mass spectrometry, as a probe into granite petrogenesis" Geochimica et Cosmochimica Acta, Vol 63. Issue 22, pgs. 3861-3881 November, 1999.

Special acknowledgments to Brian McIntyre for teaching OPT 307/407, and Dustin Trail for providing apatite samples.


Please enter any comments, criticisms, questions, etc. below.

Your name:

Email address: