Improve Image Quality of Serial Block-Face Scanning Electron Microscopy

The automated technique, SBFSEM, is used to obtain serial images in an SEM for 3D reconstruction designed by Denk and Horstman¹ and commercialized by Gatan as the 3View^® system. The large volumetric datasets thus generated can be quantitatively assessed at a high axial and lateral resolution, just like the examination of mitochondrial networks in mouse pancreatic islets of Langerhans made by Leapman et al.² In addition, SBFSEM holds huge potential in materials science applications to create 3D images of bulk materials and interfaces with nanometer resolution through large volumes³.

Datasets can contain hundreds or thousands of images and therefore, loss of image quality caused by contamination and charging can spoil an experiment. The BSE backscatter electron detector must be cleaned and replaced frequently because of deteriorating contrast from hydrocarbon contamination.

Many electron microscopists employ remote plasma cleaning with the Evactron^® plasma cleaner to acquire the best possible images from their instruments by reducing or removing contamination. The oxygen radicals that are produced in the plasma oxidize hydrocarbon compounds, yielding species such as CO₂, CO, and H₂O which are removed by the vacuum system. At XEI Scientific, several studies were carried out that showed that the Evactron can effectively remove hydrocarbons and this was established with quantifiable results using previously contaminated quartz crystal microbalances to determine the cleaning rates.⁴ The benefits of plasma cleaning for the optimized generation of 3View datasets and cleaning of backscatter detectors is addressed by the current study.

Theory and Background

An SBFSEM system consists of an SEM, a diamond-knife microtome mounted on the interior wall of the microscope’s chamber door, and software and hardware to control the acquisition process (Denk and Horstmann, 2004). To produce serial images without user intervention, a resin-embedded specimen is cut and imaged thousands of times. Figure 1a shows an example of a standard laboratory configuration with the dedicated 3View stage on the FEI Quanta 250 at the Institute of Biotechnology of the University of Helsinki. Figure 1b shows the close-up view of the 3View microtome which includes the knife in the clear position on the left and the illuminated sample holder on the right. The compact Evactron EP plasma cleaner (Figure 1c) can be fitted at various positions on SEM chambers based on port availability, considering the port furthest away from the vacuum outlet as the optimal position. Shown in Figure 1d is an example of a hydrocarbon-contaminated backscatter detector which showed reduced signal to noise and poor contrast.

Figure 1. The hardware utilized for SBFSEM experiments: a) a typical SBFSEM system, b) details of the Gatan 3View microtome, c) the Evactron EP plasma cleaner mounted on the Zeiss Sigma SEM chamber, d) a contaminated backscatter detector.

Materials and Methods

A Sigma VP SEM (Carl Zeiss Inc.), fitted with a 3View SBFSEM System (Gatan Inc.) and an XEI Scientific Evactron EP decontaminator (Figure. 1c), was used to conduct the experiment. First, a resin-embedded mouse heart muscle specimen was imaged at 1-3 kV, 2,000X magnification. Then the sample was irradiated at 30 kV for 10 minutes and subsequently imaged to check for loss of contrast on areas of colloidal silver and the formation of contaminated scanning rectangles on the machined aluminum SBFSEM stub. Using Gatan GMS 3.2 software, histograms of image contrast on regions of colloidal silver paint and sections of Helicobacter pylori on Caco-2 cells before and after cleaning were generated. In order to eliminate hydrocarbon contamination from the sample, backscatter detector, and chamber, a cleaning recipe of 5 minutes at 20 W and 2 minutes off X5 was used.

Results

Figure 2 depicts contamination artifacts (a, b) on the aluminum SBFSEM stub which are not detected after a single recipe of plasma cleaning (c, d). In Figure 3, a comparison of adjacent areas of colloidal silver dag revealed a 14% increase in BSE contrast following a single plasma cleaning cycle. The measurements were repeated thrice to confirm the consistency of the results. To ensure that the backscattered electrons are more sensitive to a layer of contamination on the detector, a low accelerating voltage of 1.2 kV was used. When lower accelerating voltages are used, there is less charging and beam damage to sample surfaces.

Figure 2. Contamination artifacts on the surface of the aluminum stub (a,b) are absent after plasma cleaning (c,d).

Figure 3. A 14% increase in contrast was seen after plasma cleaning a region of colloidal silver dag surrounding the SBFSEM specimen.

Figure 4. The block face of mouse heart tissue was scanned 20X and imaged before and after plasma cleaning.

These results match satisfactorily with the data published in Joubert⁵ illustrated in Figure 5. In this case, the analysis of contrast levels in both images of resin-embedded cells measured an increase of 15% following 3 cycles of cleaning with 6 minutes for each cycle. Such types of cleaning protocols help obtain accurate 3D modeling and comparative morphometric analyses of SBFSEM datasets by enhancing image contrast and SNR, and removing imaging artifacts and charging.

Figure 5. Analysis of images in Joubert⁵ indicates a 15% increase in contrast of thin sections of Helicobacter pylori on cACO-2 cells after plasma cleaning.

The length of the pump downtime not only extends the lifetime of backscatter detectors but also shows dependence on hydrocarbon contamination levels in FIBs and SEMs. Hence, this time could be used as an indicator of the cleanliness of the vacuum system. The following data shows that Evactron plasma cleaners considerably decrease the pump downtime of the SEMs and FIBs as well as hydrocarbon contamination, and thus help increase sample processing throughput without affecting the quality of the analysis.

In downstream or remote plasma cleaning, oxygen radicals are formed which fill the vacuum chamber in the form of excited plasma. Air is used as the process gas in Evactron plasma cleaning. At low pressures, the excited metastable nitrogen molecules exhibit a flowing UV afterglow with a characteristic pink/violet color. The neutral radicals’ concentration in the plasma flowing UV afterglow is a function of the production rate in the plasma and the loss rate in the neutral afterglow.

Conclusions

Including plasma cleaning as a routine operation during SBFSEM experiments can reduce contamination artifacts and charging, enhance image quality, speed pumpdown, and maintain a pristine vacuum system during extended data collection. With the increasing need for extended operation in scanning electron microscopes, the necessity to maintain clean conditions in the vacuum chamber also increases. It is important that SBFSEM systems are operational 24/7 and preferably maintained in uncontaminated condition with uncompromised image quality.

When large blockface resin-embedded specimens are imaged frequently, hydrocarbons are released into the vacuum chamber, resulting in decreased detector efficiency. As an example, to assure the best conditions for rapid imaging, the Zeiss MultiSEM 505 comes with two plasma cleaners to remove adventitious hydrocarbons in the load lock and the main chamber.

The present generation of Evactron Turbo Plasma™ De-Contaminators employs a gentle, down-stream plasma afterglow process to eliminate hydrocarbon (HC) contamination from FIBs, SEMs, and other analytical tools. Evactron cleaning becomes faster and spreads throughout the chamber at turbopump pressures. This is attributed to longer mean-free-paths that cause less recombination of oxygen radicals in the required three-body collisions and decreased scattering to chamber walls. In the majority of cases, short plasma cleaning cycles are enough to eliminate contamination and considerably decrease pump downtime, enabling high throughput of sample processing and analysis.

The Evactron range of plasma cleaners provides fast, powerful, and effective cleaning over an extensive range of pressures, allowing artifact-free, high-quality images and increased efficiency of sample analysis. It is possible to maintain optimal detector performance if plasma cleaning is included in routine maintenance protocols.

References and Further Reading

1. Denk, W. and H. Horstmann (2004) PLoS Biol. 2, 1900.
2. Leapman, R. et al., (2016) Microscopy and Microanalysis 22 (Suppl. 3), 1104.
3. Hashimoto, T. et al., (2016) Ultramicroscopy 163, 6.
4. Vane, R and E. Kosmowska (2016) Microscopy and Microanalysis 22 (Suppl. 3), 46.
5. Joubert, L. M. (2013) Microscopy and Analysis May Issue, 15.

This information has been sourced, reviewed and adapted from materials provided by XEI Scientific.

For more information on this source, please visit XEI Scientific.