Program Speaker
Paul Hlava - The Electron Microprobe and What It Can Do

by Nancy Attaway

Paul Hlava discussed “The Electron Microprobe and What It Can Do”, as presented during July’s meeting. He stated that the full name is the Electron Microprobe X-Ray Analyzer (EMPXA), also known as the electron microprobe analyzer, the electron probe microanaylzer, the electron microprobe, the microprobe, or just the probe. He remarked that he called the remarkable machine sweet names when it worked well and sour names when it did not.

Paul outlined his talk to describe the functions of the instrument, its capabilities and its limitations. He wanted to review the signals of interest created by the machine. Paul also wanted to review the basic design of the instrument and describe the information that it could gather, and he illustrated some examples of that.

Paul began by explaining that the electron microprobe was designed to non-destructively produce quantitative, elemental analyses of micro-volumes of material. The volume depends upon the material examined, as well as the beam conditions. Materials very closely examined were usually metals or ceramics. The instrument can be perverted to look at some polymers and organics. It can also be used qualitatively and semi-quantitatively. The instrument’s high spatial resolution enables photomicrography, such as SE, BSE, X-Rays.

The probe excels when a medium to large number of quantitative analyses are needed over a short distance. The probe performs well when looking across welds, brazes, solder joints, diffusion zones, reaction zones, mineral zoning, etc. Paul showed a slide of the compositional variations in a Kovar Weld, with concentrations relative to the distances. Sampling sizes can vary from a pinpoint to about three inches across by 3/4 of an inch high, the maximum.

Paul remarked on the probe’s limitations. No information could be obtained on crystal structure, valence, or molecules. The instrument does not do hydrogen, helium, or lithium. It has problems with beryllium, boron, carbon, oxygen, fluorine, and neon, because of their very weak x-rays. New pseudo crystals help, but the analyses can become touchy. This is an active area of research that requires special handling. A background problem exists so that trace analyses are difficult. Heat limits some materials.

Paul explained how the probe works with the signals. When a high energy beam interacts with solid materials, as in the probe, a number (9 or 10) of new signals are generated. When using the probe, Paul ignores 1/3 of the signals (auger electrons, cathodoluminescence, etc.). He tolerates or works around 1/3 of the signals (heat, bremsstrahlung, etc.). He uses 1/3 of the signals to his advantage, SE or secondary electrons, BSE or back-scattered electrons, and x-rays or the characteristic of each element. The probe delves into the chemistry by counting the electrons that come off of it.

Paul said that the useful signals are the SE, the BSE, and the X-Rays. Secondary electrons are the valence electrons that are ripped off atoms by primary (beam) electrons. The weak SE signal and creative use of geometry allow for imaging of the topography. Back-scattered electrons are primary electrons repulsed by atomic nuclei. The compositional information is basically related to nucleus size. Characteristic x-rays are formed by the removal of inner shell electrons. The transfer of electrons from higher energy shells yield x-rays. Quantitative analyses are rendered by comparison with known standards. Paul showed examples of BE, BSE, and X-rays. Every signal has a different resolution. The resolution of the SEM runs under 10 nanometers, usually between 5 and 7.

The basic design of the electron microprobe x-ray analyzer includes an electron gun, which has to be stable and accurate, unlike a regular SEM. The basic design also has a column with electromagnetic lenses to focus the electron beam and a coaxial optical microscope to facilitate optical/x-ray focus. The sample stage must be precise. There are two kinds of x-ray spectrometers, WDS and EDS; SE detector, BSE detector, high vacuum system (clean or with cold finger). The instrument’s electronics and data processing equipment accompanies it.

To give an example of a high tech instrument, Paul showed pictures of the JEOL 8900 and the Cameca SX 100. He wanted the audience to note the close-up of the gun, the column, the EDS, the WDS, the sample airlock, and the optics. A light optical microscope is built into the instrument. Paul showed pictures of the SE detector, and he also showed a picture of JEOL stage, along with the sample holders and a universal mount. Paul said that it was imperative that an instrument be able to stabilize the beam position. He remarked on the cost of these machines. A probe can cost between $750,000 to $900,000, and an SEM can cost around $250,000.

Paul discussed the types of analyses available from the probe. His SE images were usually only for documentation. The BSE images obtained were usually only for documentation, as well. The x-ray maps generated were an extremely graphic display of information, and those from the JOEL are semi-quantitative. EDS analyses were usually a guide to WDS. WDS spectrometer scans are sometimes generated to search for trace elements and overlapping x-ray lines. Semi-quantitative WDS analyses allow for assays of individual elements. The probe could also perform spot quantitative and WDS analyses and quantitative WDS traverses. The machines work with solids only and cannot work with liquids or gasses. The usual life of a probe runs about ten years, but some universities have coddled them along for 20 years or more.

Paul’s picture of a secondary electron image showed octahedrons, and he showed a back-scattered electron image. Paul also showed x-ray images. He had a BSE and element maps of a braze joint. Another picture showed corrosion rings around particles.
Paul performed an analysis of a new tourmaline from Nigeria’s Ogbomosho area for GIA. The sample crystal was provided by Nancy Attaway. The crystals exhibited a purplish-pink rim color with an orange core. The two color zones were revealed to be a purplish-pink rim by manganese and an orange core by iron and manganese. Paul showed a plot from the quantitative analyses of Mn and Fe data. He had another plot of Na, Ca, and Bi data that showed an elevated concentration of bismuth from a traverse across the diameter of a Nigerian tourmaline. Paul determined that the Nigerian tourmaline was a liddicoatite. {See Gems and Gemology, Summer 2001, pages 152 and 153.}

Paul recently examined and assayed the contents of vanadium, chrome, and iron as found in Colombian emeralds and Zambian emeralds. He displayed several examples of these emerald species. Thanks, Paul!