W. M. Keck Cosmochemistry Laboratory - Sample Preparation

This page is intended to give a relatively detailed overview of the requirements for sample preparation that will give you the best quality data. Please take these requirements seriously. If there are problems meeting the requirements, please contact us and discuss your samples. We will work with you to resolve any problems. But do not wait until you arrive at the lab to address sample preparation issues. If we have to remount your samples after you arrive, the time to do this will come directly from your allotted measurement time.

Sample holders and sample dimensions

Figure 1 Standard Cameca sample holder on left and Large-area holder on right. Both take a 1-inch-diameter (2.54 cm) round sample less than 1 centimeter in thickness. [Click for enlargement.]

The standard Cameca ims 1280 sample holder (Figure 1, to right) takes a one-inch-round (2.54 cm) sample with a maximum thickness of ~3/8 inch (1 centimeter). For most work, the best sample is a thin section on glass. We also measure cast epoxy mounts. This mounting technique permits measurements of thicker samples and samples that cannot be easily make into thin sections. All epoxy is not the same; many are not vacuum compatible. Use a high-vacuum epoxy (e.g., Buehler EpoxiCure®). Even with a high-vacuum epoxy, it is desirable to minimize the total amount of epoxy that it used in mounting the sample.

We also have inserts that fit into the standard Cameca sample holders. Many of our standards are mounted in 0.25 inch diameter and 0.25 inch high "bullets" (Figure 2a), which in turn fit into 4-, 5-, or 6-place holders (Figure 2b). These holders permit standards and samples to be mounted in the same holder, eliminating the need for sample changes.

Other inserts can take samples with a variety of shapes and sizes (Figures 3a and 3b). Contact us to discuss special mounting needs.

Figure 2a (above left) : One-fourth-inch diameter by one-fourth-inch high stainless steel "bullet" mounts. The mounts on the left have no sample. The sample is epoxied into the well on the top of the "bullet" and ground and polished flat (two mounts on right). [Click for enlargement.]
Figure 2b (Above right) : The bullet mounts are put into five-place or six-place holders that fit into the standard Cameca sample holder. [Click for enlargement.]

Figure 3a (above left): Other types of sample holders available for our users. [Click for enlargement.]
Figure 3b (above right): Other available types of inserts that fit into standard Cameca sample holders. [Click for enlargement.]

Sample Polish

For best results, samples should be flat and well polished. A standard 1-inch-round thin section polished with 1-micron diamond works well. For very high precision isotopic measurements, relief on the sample surface should be kept to a minimum. Typically, if you grind a section flat with a wheel or other suitable tool and then polish with a series of diamond pastes, stepping down from 15 micron to 9 micron to 6 micron to 3 micron to 1 micron, you should get good results. Do not try to skip the coarser grits and go straight to 1-micron diamond. You will have to polish for a long time, and the longer you polish, the more the soft minerals and epoxy are differentially eroded, generating relief on the sample. Flatten the sample with a course grit and then polish just long enough on each grit to remove the scratches from the coarser grit.

Sample Labeling

Samples should always be labeled. You may know what they are when you bring them, but no one else will know what they are, and you may forget by the time you need to review them a year later. Labels should be permanently engraved into the back of the sample so that cleaning does not remove the label.

Conductive Coating

All samples must either be conductive or will require a conductive coating that is in electrical contact with the sample holder. A metal sample embedded in epoxy must be coated so that charge can be dissipated to the holder. The coating can be either carbon or gold and should be thick enough (generally ~20 to ~30 nm) to show conductivity with a standard multimeter (resistance less than a few megohms). Museum curators typically prefer carbon rather than gold on meteorite thin sections. We can coat samples here at UH.


Image of meteorite sample taken by the navigation camera. Finding the spot you wish to analyze can be very challenging. Document your samples! [Click for enlargement.]

Sample documentation is very important. You have to be able to locate the points that you want to measure in order to make a measurement. The photo on the right shows a view of a meteorite thin section taken by the optical navigation camera on the Cameca ims 1280. The field of view is small and mineral grains often cannot be identified. Good documentation for thin sections consists of backscattered electron images at low magnification and at ~150x (close to the magnification of the navigation camera) that show the minerals you wish to measure and reflected light images (taken after coating) at low magnification and at ~150x showing the surface features on the section. There are navigation tools on the computer to help with finding your spots. A backscattered-electron image of the entire thin section can be useful for the navigation software. But you should not depend on the navigation software. Bring complete documentation. It is possible to mark very small grains using the scanning electron microscope and to locate the marks and position the beam using ion imaging. If you think you will need this capability, contact us to discuss.

Special Considerations

Edges: The electric field that extracts the ions from the sample and introduces them into the mass spectrometer must be homogeneous in order for the ion microprobe to work in a reproducible manner. If the spot of interest is within 2-3 millimeters of the edge of the sample holder or an exposed edge of the sample, the field gets distorted, which affects the results. Particularly for high-precision measurements of isotopic fractionation it is important for the measurements spots to be >3 millimeters from any edges.

Degassing (Epoxy and other considerations: The residual gas pressure in the sample chamber can have adverse effects of several types of measurements. The residual gas in an empty sample chamber consists primarily of H and OH. Even high-vacuum epoxy outgasses into the vacuum system, adding other compounds of H, C, N, and O. Minimizing the amount of epoxy will improve sample-chamber vacuum and will reduce hydride interferences. For H and D/H measurements, mounting the samples without using epoxy is best (mount in indium, for example). Outgassing can be reduced by storing the samples in a vacuum desiccator, or vacuum oven (don't overheat your sample!) for a few weeks, followed by transport to UH in some kind of desiccator box or bag. This kind of outgassing will also help epoxy mounts, but generally this will not be enough to permit H or D/H measurements. It is possible to prepare samples mounts without using epoxy (e.g., a one-inch polished disk). However, in our experience, it is very difficult to impossible to fully degas such samples, making this type of mount unsuitable to H and D/H measurements. Contact us to discuss the details of sample mounting for measurements of volatile elements.

Grain Mounts: Grains less than 5-10 microns in diameter can be measured in an unpolished form. The data will not be of high precision (e.g., <1-2 permil precision), but that is at least partly because there are not enough atoms in your sample. Grains are measured by depositing them on a conductive substrate (gold foil works very well) so that each grain is separated from other materials. Deposition in a drop of liquid with low surface tension (e.g., isopropanol, methanol) will cause grains to disperse across a gold foil or silicon wafer. The grains can also be pressed into the gold. This is desirable for grains larger than about 5 microns in diameter. Smaller grains will stick to the gold by van der Wall's forces. Contact us to discuss details of this type of mounting.

21 April 2017