Morphologies of Experimentally Shocked Zircons

Martha L. Miller

University of Rochester, Department of Earth and Environmental Sciences

1. Introduction

As part of a larger study looking to identify and quantify the effects shock metamorphism has on the diffusion of radiogenic geochronometers, researchers travelled to Johnson Space Center to utilize the Flat Plate Accelerator (FPA) to simulate a meteorite impact to the samples. This portion of the study is meant to characterize the morphology of the zircons after the experimental shock.

 

Figure 1: The FPA loaded with sample 2 (left). The target, uncapped and without the sample. (center). The target containing sample two after the experiment and the machining required for sample extraction (right).

Two samples were were prepared for our experiment. The first sample was entirely Keuhl Lake zircons, while the second sample was a mixture of a well characterized Bishop Tuff sanidine and Keuhl Lake zircons. The second sample was ~95% sanidine (115 mg) and ~5% zircon (5 mg). Ian Szumila prepared the zircons, used in both samples, by cutting large zircon grains in half and crushing one half of each grain. The remaining halves were saved for future work. Szumila also crushed the sanidine used in sample 2. Each crushed mineral was sieved through a series of mesh sizes and the fractions separated. All grains used in this experiment were from the 150 - 250 micron fractions.

Figure 2: Ian Szmulia preparing the sample for loading into the target.

 

Sample 1 was shocked with 28.4 GPa, whereas sample 2 was shocked with 23.8 GPa. The reduction in pressure, between sample 1 and sample 2, was because the flyer plate of the target housing for sample 1 was blown off during the "shot". This resulted in an almost complete loss of sample 1; however, as you will see below, information was still gained from this sample. Between shots the interior of the FPA is vacuumed out. This meant that the debris from sample 1 could be collected for analysis. Through use of a Light Microscope (LM) and a Scanning Electron Microscope (SEM) it was determined that the debris from sample 1 was composed primarily of melted Teflon (from the projectile used for the impact), metal from the target, and an ash-like substance that had trace amounts of zirconium.

Figure 3: The flyer plate is in the light, whereas the target is still mounted above, and the sample area is exposed (left).

Target 2 after the experiment, while still inside the FPA. Notice that sample 2 is not exposed (right).

Various microscopy techniques were used to study these samples: LM, SEM secondary electron imaging (SE), SEM backscatter electron (BSE) imaging, and energy dispersive x-ray spectroscopy (EDS) on the SEM. Additionally interaction modeling (or electron flight simulation) was used to better understand the interaction volume of Zr for these assorted SEM techniques.

2. Sample Preparation

Sample preparation for LM was was minimal. Sample 1 was simply transferred into a series of petri dishes. This allowed for the debris material, with use of the LM, to be sorted by approximate composition (materials mentioned above). Sample 2 preparation required removing the sample from the target in a manner that would limit the amount of metal from the target, as well as have minimal effects on the exterior geometry of the sample.

To prepare the samples for SEM techniques each stub mount was lined with double sided carbon tape and the samples were sputtered coated with carbon.

Figure 3: Sample 2 in the open target (left), also sample 2(right).

3. Interaction Modeling

Since locating and imaging the shocked zircons was he principle goal of this portion of the study it was important to understand the interaction volume of zirconium for these samples. A modeling program was used to simulate electron flight using the ________ software owned by the Institute of Optics at the University of Rochester. Interaction volume was modeled with 2000 trajectories, a tilt of zero degree, and an acceleration voltage of 20 KeV. The result of the modeling was especially informative for seeing the interaction volume for X-rays generated under these conditions, making collecting an X-ray map more efficient.

Figure 4: Model of Zr interaction volume.

4. Light Microscopy

Light microscopy was the first microscopy technique used to look at the samples. Images were collected to show surface features of formed from the shock to sample 2. Additionally images were collected of sample 1, as aforementioned LM was utilized as a means to begin the preliminary sorting of the collected material. Imaging was done in the Trail Lab at the University of Rochester. The ability to capture images in color with LM, is an asset of the technique, this is particularly useful for seeing where metal from the target may have incorporated into to the samples.

Figure 5: Debris from sample 1, predominantly melted Teflon (top left) melted Teflon from sample 1, note the fibrous morphology (top right), sample 2 in entirety (bottom left), large grain from sample 2 (bottom right).

While the differences in sample 1 and sample 2 are clear in the LM images, this technique does not allow for the clear identification of the zircons.

5. Scanning Electron Microscopy

As mentioned above, three SEM techniques were used: SE, BSE, and EDS. SE detection is useful for analyzing the surface morphology of sample 1 and 2, as well as unshocked Bishop Tuff sanidine and a half of one of the Keuhl Lake zircons that was reserved for later analysis. Having both shocked and unshocked material is critical for understanding the changes to crystal structure.

BSE is utilized to compare relative atomic mass of the components of a sample. The brighter an area on the image the higher the atomic number of the chemical component in that area. This variance in brightness is because elements with high atomic numbers more strongly backscatter electrons than elements with low(er)atomic numbers. BSE was especially useful for this study because the atomic mass of zirconium is much greater than the chemical components found in sanidine, making zircons reasonably conspicuous. An interesting result of utilizing this technique was finding the large areas of inclusions in the Keuhl Lake zircon.

EDS was also implemented to confirm where the zircons had recrystalized. This technique was helpful in distinguishing areas where needlelike zircons and metal from the target had formed in the same region of a crystal. The metal from the target was identified by a combination of Fe and Cr components. EDS was also instrumental in identifying the various components of the inclusions in the Keuhl Lake zircons.

Figure 6: Starting material in SE. Bishop Tuff sanidine surface of a single grain (top left) Bishop Tuff sanidine, single grain (top right), Keuhl Lake zircon (bottom).

The SE image of the zircon makes it evident that there are some inclusions in the grain, but offers little evident to what they may be, because of this BSE and EDS was performed and an elemental map was created. From this map we see that the elemental compositions of the inclusions are potentially a plagioclase feldspar and a potassium feldspar.

.

The backscatter image is of a portion of sample 1 that appeared "ash-like". The image shows some bright point, based on the later elemental maps it appears these could be from zircon and/or from metal originating in the target.

.