Video of Electron Bombarded Glass
Ian M Spinelli
Electron Microscope Research Core
University of Rochester Medical Center
OPT307: SEM Practicum
Digital imaging has opened up a lot of possibilities in the EM field. Microscopists can now take endless amounts of images with their greatest concern being the amount of space left on the hard drive. Clearly a minor dilemma compared to spending hours developing film in a dark room. Another benefit digital cameras provide is the ability to record video. This capability is great for putting together instructional videos, or better yet, recording really cool visuals that result from bombarding glass with an electron beam. The latter has proved to be worthy subject matter for an electron microscopy project.
The electron beam in a transmission electron microscope (TEM) carries a lot of energy. With a common TEM, electrons can be emitted from the tip of a filament with an acceleration voltage as high as 200kV. That's over a thousand times larger than the voltage supplied to a typical wall outlet. If specimens are thin enough (<~200nm), the beam can penetrate straight through with little energy transfer. Specimens that are too thick will scatter and absorb a much larger portion of the beam. Upon impact with the material, the kinetic energy of the electrons transforms into heat. Therefore, condensing an electron beam onto particles of glass may heat the structure enough to allow for atomic rearrangement and subsequent crystallization. There should be enough thermal energy to even melt some glass compositions.
1. Sample Preparation
Preparing these samples could not have been any easier. Two types of glass were supplied by the course instructor: phosphate glass and a silver-exchanged sodium aluminosilicate (SESAS). Each was ground to a fine powder with a mortar and pestle and applied to lacy carbon grids by simply dragging the grid though the powder with forceps and lightly blowing off excess powder with an air gun. Below is a colorized TEM image of a phosphate glass particle (purple) on lacy carbon (green).
2. Transmission Electron Microscopy
Transmission electron microscopy was done with a Hitachi 7650 TEM at an acceleration voltage (HV) of 80kV. A relatively low HV setting was chosen to ensure a larger portion of the electron beam interacted with the glass particles, whereas a 200kV beam can transmit through the glass less attenuated, and thus, less energy transfer. A relatively small probe diameter of 2µm was used to increase the amount of energy per unit area traveling through the glass.
3. EDS Mapping
Energy dispersive spectroscopy (EDS) was run with a EDAX detector on a Zeiss Supra 40VP SEM with a HV of 10kV and a working distance (WD) of 16mm. This WD was ideal for providing a large take off angle for x-ray emission, thereby increasing the count yield. Due to time constraints, EDS mapping was acquired at a low resolution of 128x100 with a dwell time of 50ms at each data point.
4. Video Capture
The Hitachi 7650 is equipped with a side-mounted Gatan Erlangshen ES1000W digital camera, which is connected to a Dell PC that runs Gatan's DigitalMicrograph™ software for capturing images. Unfortunately, this software is not capable of capturing video. Luckily it does provide a live digital streaming video (DSV) signal that other video editing software can tap into and record. In this case, VideoStudio™ 7 was used for capturing and editing video.
Initially, capturing video seemed to be an impossibility. As mentioned before, the Gatan camera software only provided a live digital video stream, but lacked the ability to create movie files, such as AVI or MPEG. The software's manual stated that the signal created by the digital streaming video module can be recorded by applications that require streaming video input, such as Windows Movie Maker. Of course, this application was worthless, constantly blabbing about the Sapera acquisition board having an invalid DirectShow configuration. A product manager at Gatan suggested using an application called VideoStudio™ 11 instead, however, the exact same error message haunted the computer screen. This problem was not solved, even after eons of troubleshooting with the Gatan product manager. Finally the course instructor dug up an old CD of VideoStudio™ 7 in hopes that the 11th version was "too advanced". Miraculously, version 7 tapped into the DSV signal of the Gatan camera, even though it too would display the same error prior to capturing video. Below is a two and a half minute clip of a movie put together using the editing tools of VideoStudio™ 7. A number is displayed in the upper right corner of the video for the first few seconds of each clip for referencing.
Scene1 displays a phosphate glass particle that beads up into a sphere as it is melted by the beam. This isn't so hard to believe since phosphate glasses have a relative low melting temperature (~580°C) compared to the more standard soda-lime silicate glasses (~1450°C) that are used for windows, beer bottles, etc (Varshneya 1994). The phosphate glass particle in the Scene 2 does not seem to completely melt, but is heated enough for molecular motion to occur (shifting of bonds angles and lengths), causing the shape of the particle to change as its structure crystallizes. Black spots (possibly crystallites) appear in the right portion of the phosphate glass in Scene 3. A larger chunk of phosphate glass was melted in Scene 4 and then the beam was condensed until there was enough electron density to pass through the particle and visualize the lacy carbon that it rested on.
The particles up to this point appear black, which means the glass is too thick for the electron transmission. The particle in Scene 5 must be much thinner near its edges where it appears transparent. This means a good portion of the electron beam is transmitting through with little specimen interaction, and therefore less energy is transferred into the glass. As a result, the glass seems to maintain a solid state. There is however, a wavy pattern that flows erratically throughout the glass. This effect may be due to the glass thermally expanding as it is heated by the beam. As the thickness of the glass changes, so too does the direction in which the electrons are scattered, generating dark regions that move through the glass.
Imaging with the "dirty" dark field method was attempted to try and mix things up. Scene 6 shows the switch from bright field to dark field by simply moving the objective aperture into the path of the zero-order beam. Melting a phosphate particle did not seem any more interesting in dark field. Scene 7 is a little more intriguing. Another relatively thin piece of glass was located, which exhibited the same wavy patterns as the particle in Scene 5. The visual display in dark field consisted of bright beams (diffracted electrons) shooting out from the edges of the particle. The edge of the glass particle may have in fact crystallized if the bright emissions are indeed electrons that have diffracted off lattice planes. Amorphous materials do not diffract electrons since they lack lattice planes.
The last two scenes show footage of the SESAS glass. There appears to be solid black blobs that eject from the surface of the particles. These objects could be silver being rejected by the glass as the electron beam heats up the structure and allows for atomic mobility of the silver. EDS analysis of these black bodies to verify their composition would have been a good idea, however, the EDS detector on the TEM was out for repair.
2. Beam-Specimen Interaction Simulation
Electron Flight Simulation™ software was used to create a visual representation of how the electron beam interacts with glasses of varying thicknesses. Below are two interaction volume schematics. The vertical red line represents the electron beam and the blue lines represent the paths taken by the electrons through the specimen (shaded gray). Unfortunately, the simulation software doesn't show the path of the electrons after the pass through the specimen. Both schematics were generated by directing the software to simulate an interaction volume with 32,000 electrons accelerated at 80kV into amorphous silica with thicknesses of 200nm and 2µm. Clearly the interaction volume is much larger (i.e., more energy per unit volume) in the thicker specimen, whereas the electron beam shoots right through the 200nm specimen. This data supports why the thicker glass particles were more prone to melting under the beam.
3. EDS Mapping
Below is a SEM image of the surface of SESAS glass. This glass underwent a procedure known as ion-exchange. In this process, glass is submerged in a molten salt (AgNO3 in this case) at a temperature high enough for ionic mobility, but low enough to prevent the glass from softening. The ionic gradient at the interface between the salt and glass causes silver to go into the glass and sodium to come out. As the glass returns to room temperature, compressive stresses develop at the glass surface since silver is larger than sodium, thereby preventing the glass structure from contracting to its original volume. The brighter protuberances in the micrograph below were suspected to be silver that had migrated out of the glass due to the compressive forces. In order to confirm this hypothesis, EDS analysis was executed on the area to create an elemental map. Below the SEM image are maps for sodium (green), silicon (blue), oxygen (red), and silver (yellow).
The pattern of the silver map matches up nicely with the lumps suspected to be silver. The surrounding surface appears to contain sodium, silicon and oxygen, which corresponds with the sodium aluminosilicate glass composition. The silver appears to be in metallic form since the oxygen map has gaps where the silver is present.
The maps above demonstrate two examples of how surface topography can attenuate x-ray detection. In looking at the SEM image, notice the dip in the top-center portion of the specimen surface. The signal intensity of this area of oxygen map is lower. X-rays emitted from a dip in the surface are more difficult to detect since they are more likely to be reabsorbed by the rising portion of the surface between the emission site and the detector. This effect is not as prominent in the silicon and sodium maps since the x-rays formed by these elements are of higher energy, thus are not absorbed as readily. The second example results from a similar effect. Note how the empty areas that represent the silver in the silicon map are slightly stretched northwest by southeast, rather than round like the silver protuberances. This is most likely a shadowing effect, where the humps of silver absorb the silicon x-rays before they can reach the detector. Based on the direction the holes are stretched, the detector is probably located in the upper left region relative to the sample. These attenuation effects can be minimized by increasing the WD and/or tilting the specimen so the surface is more normal to the detector.
Recording video of electron microscopy can be a very useful tool, not only for teaching, but for displaying dynamic beam-specimen interactions. It would be very convenient if future software applications for EM digital cameras were capable of acquiring video on their own, rather than having to rely on third party software. Based on the first few scenes of the movie, it is apparent that it is possible to melt phosphate glass via electron bombardment. Whether or not the thinner piece of phosphate glass in Scene 5 was actually crystallized is less definite. In future work, electron diffraction could be employed to determine if the particle was indeed crystalline. The black spheres that ejected from the glass particles in the last two scenes are likely silver. Additional future work could involve EDS analysis of the black bodies to verify their composition, as was done with the SESAS glass in the SEM.
Thank you to Brian McIntyre for his light-hearted yet austere instruction, Bill Mollon from Gatan for troubleshooting my video capture software woes, Karen Bentley for virtually unlimited access to her TEM, and Gayle Schneider for being someone to vent with throughout this arduous project.
Egerton, Ray F. Physical Principles of Electron Microscopy: An Introduction to TEM, SEM, and AEM. New York: Springer, 2007.
Varshneya, Arun K. Fundamentals of Inorganic Glasses. San Diego: Academic Press, Inc., 1994.