The most common questions we get from visitors concern sample preparation. These questions are generally in three areas:
The UCLA Cameca IMS 1270 accommodates up to three sample holders that are designed to accept 1" diameter cylindrical samples with a maximum thickness of ~½ cm. The sample is held in place against a machined lip by nonmagnetic springs and backing plates (see below). Samples generally consist either of 1" round, glass-mounted, polished petrographic sections, polished grain mounts potted in epoxy, or composite mounts composed of these materials. Thick petrographic sections are preferred over the standard 30µm variety, especially when the samples are porous and where the mounting medium (epoxy) may interfere with the analysis (e.g. C-isotopes in carbonates). Although preservation issues may dictate that entire 1" round thin sections must be employed, we find that (when possible) it is more efficient to produce a composite mount in which relevant portions of polished thin sections (i.e., those containing planned analysis points) are trimmed down via diamond coring or a wafering saw and juxtaposed together with pre-polished standards. Grain mounts are produced by positioning grains upon double-sided tape and potting in epoxy. Finally, more irregular shaped polished samples can also be accommodated using special sample holders with Ta masks.
Although the OD of the sample mounts are 1" round, it is important to emphasize that the entire upper surface of the sample mount is not accessible for analysis. The sample holder have a 0.1" wide lip machined onto their upper surface to facilitate uniform positioning of the sample. This blocks the outer 0.1" of the mount. In addition, the field inhomogeneity caused by the finite thickness of the sample holder lip effectively requires analysis points to reside in the inner ~0.7" of the sample mount.
In producing epoxy mounts, the resin should be of low viscosity to reduce risk of air bubbles being trapped next to the grains, and should be hard when cured to facilitate polishing. We use 3M® 701DL double-sided tape for mounting and produce air-cured mounts by mixing 5 parts (by weight) Buehler® 20-8130 epoxide resin with 1 part (by weight) Buehler® 20-8132 epoxide hardener and allowing 6-8 hours to cure. A thin coat of Buehler® 20-8185 release agent applied to the confining 1" ID ring prior to pouring the epoxy mixture facilitates later removal. No release agent should applied to the mounting surface. Epoxy out-gasses a significant volume of water vapor in the vacuum of the analysis chamber. For work where hydride interferences are a concern (e.g. 17O measurements) the following steps should be taken to minimize out-gassing. Store samples in a warm oven (50°C), a desiccator or, if possible, a vacuum oven. If samples can be left in the instrument overnight before analysis begins, hydride background will be further reduced.
The primary objective of polishing in producing SIMS mounts is to produce a flat analysis surface. Avoid excessive polishing! Prolonged polishing with sub-micron polishing compound may yield a fine polish but will also tend to "round" grains in grain mounts and excavate trenches around them, hindering analysis near grain boundaries or the analysis of small (<50µm) grains. Diamond, alumina, or silica polishing media may be used. Grain mounts can be effectively sectioned through the use of grinding paper. Typically, a brief, final polish with 1µm diamond yields an adequate polish without compromising flatness. Do not use a polishing medium which will contaminate the sample with the elements to be studied. After polishing, ultrasonically clean the mount with soapy water in clean glassware. Handle the cleaned sample with gloves and rinse with DI water. In the case of zircon U-Pb analysis, ultrasonically clean the sample in 1N HCl for 1 minute to reduce the common Pb contamination (omit this step for monazite). Final cleaning consists of 5 minutes ultrasonic treatment with DI water in clean glassware. Loosely cover the mount with Al foil and place it in a drying oven at 50-75°C for a least one hour.
For almost all work in the geological sciences, samples must have an electrically conductive coating which makes contact with the lip on the sample holder. While either gold or carbon may be employed, gold is preferred since it is more conductive and more easily sputtered than carbon. Gold coats should be 20-40nm. The thickness of carbon coats should be much thinner than that used for electron probe microanalysis (indigo blue on polished brass surface) since C sputters very slowly. Conductivity needs to be tested before the date of analysis by using a multimeter. Au coats should have conductivity not less than approx. 10 Ω. C-coats should yield approx. 10 to 15 MΩ on a 4 cm glass slide with sticky C-tape at both ends to prevent scratching of the coating by the multimeter electrodes. Poorly conductive C-coats can result from bad vacuum conditions during coating: make sure that vacuum during C coating is better than 10-4 Torr. Carbon should be avoided for mounts prepared for U-Th dating because of an isobaric interference on mass 246 (the 230Th-oxide peak). Clearly, carbon coating is also not suitable if C-isotopes are to be measured.