Investigating the World with a Scanning Electron Microscope

Story posted January 08, 2004

Thanks to a grant from the National Science Foundation and the work of Assistant Professor of Geology Rachel Beane, Bowdoin recently became one of the few colleges and universities in the United States to add an Electron Backscatter Diffractometer (EBSD) to a Scanning Electron Microscope (SEM).

The scanning electron microscope magnifies specimens to as much as 300,000 times magnification, and Bowdoin's SEM was already equipped with an Energy Dispersive Spectrometer (EDS), to allow chemical analysis of specimens, but the addition of the EBSD will let Beane and her students investigate the growth of minerals by observing their crystal lattice structures.

To get the images, a filament in the SEM, similar to that in a light bulb but made of tungsten, heats up and shoots electrons down a column to where they hit the specimen. The specimen can be almost anything, but because Beane is observing it in a vacuum, it needs to be conductive. So most specimens, including the minerals that Beane studies, need to be coated with a conductive material. Without the coating, the specimen would appear as a mostly gray spot, but the conductive coating allows a clear image to form under the SEM.

Most often, Beane uses a carbon coating. Because carbon has a low atomic number, it won't interfere with the chemical analysis of the specimen. Occasionally, for specimens for which she wants higher magnification and higher resolution, a coating of gold is used (but gold's atomic number is higher than carbon's, so it is more likely to interfere with a chemical analysis).

A great deal of information can be gleaned from an SEM with an EDS and EBSD. The SEM produces a detailed image of the topography of the specimen. Backscatter electrons, which are the result of collisions of the electron beam with specimen electrons, can be measured to create an image showing in different shades of gray places in the specimen with different atomic numbers.

The EDS analyzes x-rays that are given off by the specimen after it's been hit by electrons. The EDS can then give a qualitative analysis of the specimen, showing what elements are present, and a quantitative analysis, showing the chemical composition of the specimen: For example the chemical composition of a Calcium-rich plagioclase could be 49wt% SiO2, 32wt% Al2O3, 15 wt%CaO, and 3wt%Na2O (wt% is weight percent of the oxides).

When the specimen is an unknown mineral, understanding this composition is an important step in determining what the mineral is.

The EBSD analyzes backscattered electrons, allowing Beane to see the lattice structure of a mineral. The lattice structure is basically its atomic pattern, and every mineral has a distinct lattice structure.

The EBSD cost about $100,000 and consists of a camera, phosphorus screen, detectors, monitor, computer and a software library of all the different lattice structures. The camera and phosphorous screen can pick up the different lattice patterns, and the computer then identifies the mineral by the structure and displays different orientations of the lattice structure in differing shades of gray.

Observing the lattice structures tells Beane how the mineral grew when it was forming: It might have started at one point and grown out, or it might have started growing at several different points which then grew together. At times the lattice structure is even stretched or compressed, which shows the type of force it was growing under.

This kind of analysis gives Bean a powerful tool for both research and teaching.

"This is extremely new in terms of geology," she said.

Though there were methods of obtaining the lattice structures, previously it was very difficult. "What I can get overnight [now], would take at least a month," she said. And there wasn't a method to get the lattice structure of garnets, which is what Beane studies.

The first papers to make use of this technology were published in 1999. Beane immediately wanted to learn the technique, but there was no institution — no university, no small college — in the United States actively using it at the time. Instead Beane traveled to the University of Liverpool for training. Beane was the only one to receive an NSF grant in geology for the technology in 2003.

"It's an extremely big deal that Bowdoin is getting this," she said just before ordering the equipment. "I'm almost giddy, this is so exciting."

Since they were first purchased in 1999 the SEM and EDS have been used in a wide variety of Bowdoin courses.

Beane uses it in her introductory classes as well as her upper-level classes. Associate Professor of Chemistry Beth Stemmler uses it in two of her upper-level courses. Members of the biology department use it in looking at worms. Archaeology professors have used it to look at ancient coins.

Worm's mouth
A worm's mouth under the Scanning Electron Microscope

"It's a great way to get students involved," she said. "In terms of teaching observation, I get real excited by it."

The EBSD will be an important addition to teaching students about lattice structure.

"To be able to see something, rather than just being told it, is huge for the learning part of it," she said.

Much of Beane's work is with garnets, and with analysis from the SEM she can estimate conditions such as temperature, pressure, age, growth patterns and the depth at which they formed a hundred million years ago, or more.

"There's just a huge history of information you can get from a garnet," Beane said. "They aren't found everywhere, but they're fairly common."

At the same time, she's learning that garnets are more complex than was previously thought, which creates challenges for scientists.

"I think there's a lot more to garnets than we really understand," she said.

The larger question is what the forming of particular rocks and minerals indicate about the tectonics of the earth. Right now she's working on discovering at what temperature garnets de-form, meaning they're pulled apart and out of their shape.

In the center of a fault zone, which is where there has been movement of the tectonic plates, rocks are sheared and strained. Beane wants to quantify the strain that occurs across a fault zone.

"I think this technique [EBSD] will help us to do that," she said.

She'll take samples at regular intervals across the Norumbega Fault Zone. No longer an active fault, it runs from southern Maine up through Canada. Maine was once more mountainous than it is now, and as the tectonic plates shifted they altered the topography of the land, so rocks that were underground then, now sit on the surface.

"We're already, in a sense, underground from when that fault occurred," she said. "You're looking at the changes from where that fault occurred, deep in the earth."

Beane is excited that geology courses at Bowdoin are teaching students about the real experience of being a geologist.

"It's a process of discovery. They're doing science," she said. "They don't think of science as a creative endeavor [at first], but I think it is." There's so much missing from what is known about geologic history that creativity is necessary to guide investigation and discover the answers.

« Back | Campus News | Academic Spotlight | | Subscribe to Bowdoin News by Email