Sample Preparation

The laboratory has a wide variety of sample preparation equipment available. While there is some overlap, the equipment is listed below according to its primary use, whether for microscopy or x-ray techniques.

Preparation techniques for Scanning Electron Microscopy
With the advent of variable pressure and environmental microscopes, some of the requirements for sample preparation have been relaxed. However, some preparation is often still required.

Size reduction
The SEM can scan samples up to about 2"x 2" by 1/2" thick. Larger or thicker samples can sometimes be examined. However, users are asked to consult with us about the details.

Samples can be reduced in size using saws. We have a large (18"), high-speed, cut-off saw designed for cutting slabs of concrete from cores that are typically 4" in diameter. We have a small (5"), low-speed, diamond saw for precision cutting or sectioning of samples. We also have a diamond wire saw for precision cutting of fragile samples.


It is often necessary to embed samples in order to prepare a cross section to see internal structure. We have several options available for embedding samples that fall into cold and hot methods.

Cold methods
These methods utilize a liquid resin and hardener mix that is mixed with or poured over the sample, cast into one of a variety of molds, and hardened over time to produce a hard pellet. We have two primary resins which we use.

Epoxy resin is mixed in an 7:1 resin:hardener ratio. The prepared resin has a considerable working time. It flows well to envelop small features and attach well to the sample. It is thus well suited to embedding particulate samples. It normally requires overnight for hardening, although it can be accelerated with modest heating.

Acrylic resin involves a dry powder that is mixed with liquid resin. It partially dissolves, then hardens. It has a short working time (a few minutes); therefore, the samples must all be prepared in their molds before mixing the resin. Flowability is fair. The main advantage is the quick hardening time which requires only a short delay before grinding.

L R White is another acrylic resin. It has very low viscosity in liquid form and is able to flow into very fine pores and cracks. It cures in several hours at 60C, so it is not a true "cold" method. However, that should not be a problem except for the most heat-sensitive samples.

Hot methods
These methods fall into two classes, but both employ a heated, hydraulic press for preparation.

Thermosetting materials flow under heat and pressure and polymerize to their final structure which is stable under subsequent heating. They can be added to a hot mold and ejected in just a few minutes after polymerization is complete and while the mold is still hot. We have bakelite resin and copper- and glass-filled diallyl phthalate resins on hand.

Thermoplastic resins repeatedly melt under elevated temperatures and solidify when the temperatures are lowered. Therefore, the mold must be cooled before the pellet can be ejected. With our air-cooled mold, that means that only two or three of these thermoplastic pellets can be prepared per hour. We have Lucite acrylic resin on hand.

Both types of pellets can be prepared as soon as they are removed from the mold.

Grinding and Polishing
These steps are often necessary to expose a section of interest, to minimize topographic contrast, and to allow the contrast due to the difference in phases to dominate. It is especially necessary if the microstructures to be observed are much smaller than the texture of the rough surface.

We have three grinder/polisher units. Two are variable-speed, 12-inch units while the other is a fixed-speed, 8-inch unit. The first 12-inch unit (Allied) is automated and normally dedicated to the preparation of concrete samples. It uses diamond wheels graded from 270 um down to 6 um. The second 12-inch unit (Leco) and the 8-inch unit (Buehler) use silicon carbide paper ranging from 60-grit to 1200-grit for grinding. The 8-inch unit is used for polishing, when necessary. Polishing is done with alumina powder, diamond paste, or diamond slurry. Several different polishing clothes are available with a variety of textures and naps.

The advent of the variable-pressure microscope means that samples are not required to be either conductive or have a conductive layer applied in order to be examined in the SEM. However, resolution and signal quality are better when samples are coated with gold or other heavy metals. X-ray spatial resolution is also improved when non-conductive samples are coated with carbon and the SEM is operated at high vacuum mode. For this reason, we maintain two coating units in our lab.

An Edwards evaporator is used for coating samples with carbon and occasionally for evaporating metals. Since the Edwards itself requires high vacuum for operation, it takes about 30 minutes or more to run through the first coating cycle.

A Denton or Quorum sputter coater may be used for sputter coating samples with gold or iridium. Sputtering operates at much lower vacuum levels (~0.1 torr) so the unit reaches vacuum much quicker and cycle times are on the order of 6-10 minutes.

Preparation techniques for X-ray methods
X-ray diffraction and fluorescence have quite different preparation requirements than does scanning electron microscopy. The primary concerns for x-ray analysis are sample homogeneity and particle size and random orientation (for diffraction). That requires grinding the sample and sometimes fusing the samples into disks to provide a homogenous composition for analysis. Grinding Homogeneity requires that samples be ground to a size so that the small aliquot chosen for analysis is representative of the original sample. In addition, x-ray diffraction traces are sensitive to particle orientation. Therefore, samples must be finely ground before analysis.

Samples are often ground in a "Shatterbox" which consists of a heavy, hard puck inside a matching mortar. A motor drives the assembly in an orbital motion which spins the puck at high speeds inside the mortar. Samples can often be reduced to 400 mesh in a matter of minutes. Conventional mortar and pestles are available for small quantities of sample, as well as a McCrone Micronizing Mill. Fluxing X-ray fluorescence works best when all elements are uniformly distributed rather than being present as a collection of several phases. Therefore, samples are often fused with a flux at high temperatures into a glass of uniform composition. For this operation we maintain a PANalytical automated fusion system which can process two samples at a time.