Electron Microscopy
excerpt from Microcosmos by Jeremy Burgess, Michael Marten and Rosemary Taylor
Copyright 1987, Cambridge University Press, reproduced by permission.

Materials and Inorganic Specimens

The restraints on the TEM specimen tolerance to vacuum, transparency and size - apply with slightly different emphasis to non-biological objects. Such specimens include minerals, metals, engineered parts, crystals, dusts and smokes, and a range of biologically derived materials which are nonetheless fairly inert, such as textiles, paper and timber. It is unnecessary to stabilize these specimens against the vacuum of the microscope column. But the problem of lack of transparency may be greatly increased. Biological objects are made from atoms of low atomic number carbon, oxygen, nitrogen and sulfur. A metal alloy, by contrast, may contain heavy atoms such as lead, tungsten, nickel or uranium. Unless such specimens are very thin indeed, they will scatter the electron beam so much that image intensity will be low and chromatic aberration effects pronounced. As for the limitation on specimen size, it is felt in terms of sampling error. In other words, the small piece of a material selected to produce the specimen may not be truly representative of the bulk.

The preparation of ultrathin specimens can be accomplished in several ways. Many metals can be produced as thin films simply by evaporating them in a vacuum onto the surface of a suitable substrate, which is then dissolved away to leave the thin film specimen. Specimens of metallic compounds can be produced by a similar method, provided that suitable solvents are available. For example, oxides can be studied by the surface oxidation of the parent metal. The layer of oxide is freed from the underlying unreacted material by dissolving the metal in a solvent which does not react with the oxide. Specimens of this type are used to study crystal structure, or mechanical and other properties. The surface relief of materials can also be studied by use of the replica techniques already described for biological specimens.

Of more general application in the production of very thin specimens are methods which begin with a large piece of material and reduce its thickness in stages. Take a rod of alloy steel. The first stage is to produce a slice of it of a size appropriate to the microscope's specimen stage (usually a 3 millimeter diameter cylinder). This can be accomplished by mechanical means (a fine-toothed saw) or chemically by an acid stringsaw. Further stages cannot be carried out mechanically. Methods include chemical thinning, where a fine jet of a dissolving agent is directed onto the center of the specimen until a hole is produced; regions around the hole are then thin enough to be examined by TEM. Or thinning to the point of puncturing the specimen can be achieved electrolytically, by mounting the specimen as an anode in an electrolytic cell. The edges of the sample can be protected by lacquer during the thinning process, or the thinning may be localized by directing a stream of electrolyte onto the center of the sample. Finally, a sample can be thinned by bombarding it with high energy ions at low pressure. This process involves placing the specimen inside a vacuum chamber, and bleeding in a gas (usually argon) which gives rise to ions when a high-voltage discharge is produced within it. Ion-beam thinning is used with metals, minerals, diamond, carbon fibers and other applications. The thinning can be limited to a particular area of the sample by magnetic focusing of the ions; or the sample can be rotated in ^ the ion beam to give overall thinning I on both sides.

With materials such as textile fibers, timber, or asbestos, thin specimens can be produced by sectioning techniques like those described for biological materials. It may only be necessary to embed the sample in a resin suitable for sectioning with diamond knives; chemical fixation and dehydration are superfluous.

A wide range of materials can be placed in the microscope as finely divided fragments. They may be produced mechanically by grinding the original sample (minerals, glasses), or chemically as precipitates. The latter are mounted on specimen grids pre- coated with a thin film of carbon which acts as a support. In the case of ground fragments, only the edges of individual particles are thin enough for TEM examination.

In many cases, the TEM is inappropriate for examination of materials specimens. Failed machine parts, forensic specimens and manufactured goods undergoing quality control, for example, all need to be studied whole. The nature of the information sought means that destructive methods of specimen preparation cannot be used. In such cases the SEM is the instrument of choice. The elaborate procedures needed to prepare biological SEM samples are largely unnecessary, since stability to a high vacuum and conductivity are often properties of the material itself. Metals, textiles, machine parts and microchips can all be examined without any preparation at all, although in most cases it helps to gold-coat the specimen to prevent local charging effects. Even metallic specimens may need such coating for the best results, since they are almost invariably covered with a thin layer of oxidized material of poor conductivity.

Go back to the top of this page.

Go back to the First Page on Electron Microscopes.


Page authored by Paul Perkes and the ACEPT W3 Group
Department of Physics and Astronomy, Arizona State University, Tempe, AZ 85287-1504
Copyright © 1995-2000 Arizona Board of Regents. All rights reserved.