Deceleration of probe beam by stage bias potential improves resolution of serial block-face scanning electron microscopic images
© The Author(s) 2016
Received: 20 July 2016
Accepted: 9 September 2016
Published: 15 September 2016
Serial block-face scanning electron microscopy (SBEM) is quickly becoming an important imaging tool to explore three-dimensional biological structure across spatial scales. At probe-beam-electron energies of 2.0 keV or lower, the axial resolution should improve, because there is less primary electron penetration into the block face. More specifically, at these lower energies, the interaction volume is much smaller, and therefore, surface detail is more highly resolved. However, the backscattered electron yield for metal contrast agents and the backscattered electron detector sensitivity are both sub-optimal at these lower energies, thus negating the gain in axial resolution. We found that the application of a negative voltage (reversal potential) applied to a modified SBEM stage creates a tunable electric field at the sample. This field can be used to decrease the probe-beam-landing energy and, at the same time, alter the trajectory of the signal to increase the signal collected by the detector. With decelerated low landing-energy electrons, we observed that the probe-beam-electron-penetration depth was reduced to less than 30 nm in epoxy-embedded biological specimens. Concurrently, a large increase in recorded signal occurred due to the re-acceleration of BSEs in the bias field towards the objective pole piece where the detector is located. By tuning the bias field, we were able to manipulate the trajectories of the primary and secondary electrons, enabling the spatial discrimination of these signals using an advanced ring-type BSE detector configuration or a standard monolithic BSE detector coupled with a blocking aperture.
KeywordsSBEM Deceleration SEM Volume reconstruction Serial section Cerebellum Backscatter electron detector
Serial block-face scanning electron microscopy (SBEM) has proved to be a remarkable technique for imaging at moderate lateral and axial resolution (approximately 10 and 40 nm, respectively) and across large fields of view spanning many hundreds of microns of sample. This technique is elucidating processes, where selectively stained cells within large fields of view can be found [1–3] and small details can be followed across multi-scale dimensions. It has been especially useful in neuron tracking across large distances . One SBEM study that tracked mitophagy events from neurons in their neighboring astrocytes , in particular, has fundamentally changed how scientists look at the role played by astrocytes in glaucoma.
The SBEM platform is comprised of an ultramicrotome embedded within the scanning electron microscope (SEM) chamber. Originally conceived by Leighton , use of this imaging technique has become mainstream, because of an automated platform to collect 3D volumes developed by Denk and Horstmann . With SBEM, the embedded microtome removes ~30–80-nm-thick sections from the sample block mounted on a fixed rivet. With each “slice”, the electron-beam raster scans the newly exposed block face, and the BSEs are collected to create an image. A 3D volume is obtained by stacking the images of the block face obtained after each microtomy cycle.
The platform improved by Denk and Horstmann was commercialized by Gatan, Inc., as the 3View system. The latter has been integrated into multiple SEM platforms across manufacturers and features full computer automation. In 2015, a competing SBEM platform, the Teneo VS, was released by FEI Company.
The focused ion beam milling technique (FIB-SEM) is analogous to SBEM, but the block face is vaporized by bombardment by heavy ions to remove subsequent layers of sample [8, 9]. As with SBEM, FIB-SEM backscattered electron images are collected after each bombardment and stacked to form a 3D reconstruction.
In spite of their different methods for removing the exposed block face, SBEM and FIB-SBEM use SEM backscattered electron imaging, and thus, improvements in preparing specimens and detectors are germane to both technologies. In this study, we concentrated on determining the theoretical and experimental factors in BSE collection critical to further improvements, particularly using primary beam reversal potentials to improve detection efficiency and image resolution.
It has long been known that applying a negative potential to the sample reduces beam-landing energies and, thereby, reduces beam damage to the specimen and improves surface detail. This understanding informed the design of the low-energy electron microscope (LEEM) [10, 11]. LEEM has supported complex studies in materials science [12–15]. However, in LEEM, the sample is generally irradiated with a flood beam and uses complex instrumentation of magnetic sector plates to separate electrons leaving the sample from the primary beam. There have been only a few studies of stage bias in a conventional SEM and, specifically, its relevance to imaging biological samples [9, 16, 17]. For imaging biological samples, secondary electron (SE) signal generally provides unsatisfactory contrast. BSE imaging, however, provides a clear contrast between the heavy-metal stain and unstained structures . It was been pointed out that, although beam deceleration in SBEM improves the contrast and resolution of the BSE image, it also leads to severe image artifacts whose cause is unclear . We believe that, to enable more widespread use of the beam-deceleration techniques in SBEM, a comprehensive study of its advantages and pitfalls is needed. This manuscript tries to address these issues.
Small (1 mm × 1 mm × 0.5 mm) pieces of resin-embedded tissues were mounted on aluminum specimen pins (Gatan, Inc., Pleasanton, CA, USA) using cyanoacrylic glue and precision trimmed with a glass knife to a rectangle approximately 0.5 mm × 1.0 mm, so that tissue was exposed on all four sides. Silver paint (Ted Pella, Inc., Redding, CA, USA, http://www.tedpella.com) was used to electrically ground the edges of the tissue block to the aluminum pin, taking care not to get paint on the block face or edges of the embedded tissue to be sectioned. The entire specimen was then sputter coated with a thin layer of gold/palladium. After the top layer of gold/palladium block was removed by the ultramicrotome, the tissue morphology became visible by BSE imaging. The remaining coating on the edges of the block served to reduce charging and did not interfere with imaging.
Implementing the stage bias potential on two SEM/3View platforms
The initial experiments using stage biasing to achieve deceleration were performed on a FEI Quanta 200 FEG scanning electron microscope on loan from FEI. The Quanta FEG was equipped with a high-precision dc power supply to apply a negative-bias potential to the sample. The Quanta FEG was also equipped with a 3View system from Gatan, Inc. The backscatter electron detectors (BSDs) included a monolithic backscatter electron diode detector from Gatan, Inc., as well as a concentric backscatter (CBS) detector from FEI.
These two backscatter detectors are both approximately 9 mm in diameter and include a 1-mm central hole to allow the probe beam to pass through. The two detectors were also positioned to subtend a similar solid angle for BSEs.
The monolithic BSE detector from Gatan, Inc., was read out, though a single output into the amplification and digitization circuitry. The FEI CBS detector, on the other hand, reads signals from the concentric rings of active regions through four individually configurable amplifiers. The geometry of this device allows for the spatial differentiation of signal (i.e., signal reaching the inner vs. the outer rings is read out separately). When used in conjunction with a biasing field, this device is advantageous, as it allows for the ability to spatially discriminate SE from BSE signals (e.g., to eliminate SE signal contamination when performing BSE imaging).
A similar strategy for signal differentiation was implemented on a Zeiss Sigma variable-pressure SEM platform equipped with a Gatan 3View system and a monolithic BSE detector. Gatan, Inc., also provided a high-precision dc power supply to bias the sample potential with a stable and accurate negative voltage. In this case, we used a custom-made aperture to block unwanted SEs from striking the BSE detector and polluting the BSE signal. The SE-blocking aperture was machined to our specifications as determined by the gun-acceleration potential, the negative-bias potential, and the working distance. The blocking aperture was attached to the grounded shielding of the Gatan BSE detector using silver paint (Ted Pella, Inc.) and suspended above the detector by thin wires.
Results and discussion
Factors affecting BSE image resolution in the SBEM
Scattering of incident electrons in the SEM is governed principally by the probe-beam-landing energy and composition of the scattering substrate. The penetration depth of the probe-beam electrons into the sample (electron range), the lateral scattering of the probe beam and BSEs, and the thickness of the ultramicrotome cuts all limit resolution in the SBEM. The electron range for carbon is ~60 and 180 nm for beam energy of 1.5 and 3 keV, respectively, according to the Kanaya and Okayama range equation .
Another, important parameter that determines the ability to resolve fine features in BSE imaging is the lateral-energy spread of the BSEs, i.e., how far the BSEs emerge from the incident-beam impact point and the fraction of the incident beam energy they have. The lateral-energy spread was calculated by dividing the area around the beam impact point into pixels of size 1 and 2 nm for the 1.5- and 3-keV beams, respectively, and the total integrated BSE signal emanating from each pixel was computed as the summation of the energy of all the BSEs from the particular pixel. The lateral-energy spread was measured in nm as the distance from the incident-beam impact point that contains 50 % of the total BSE energy. It was found to be 11 nm and 36 nm for the 1.5- and 3-keV beams, respectively (Fig. 3b, d).
The only way to physically improve the axial resolution of electron scattering from a theoretical basis is to lower the energy of the probe-beam electrons. However, below 2 keV, beam control is problematic as the electromagnetic fields used to steer the electrons down the column are less stable, and the less-energetic primary beam is more susceptible to external electric and magnetic interference leading to degraded resolution . In addition, at lower accelerating voltages, chromatic aberration can degrade resolution as well. Chromatic aberration is defined as follows:
Decreasing primary beam-landing energy improves axial resolution
We originally used an FEI Quanta 200 ESEM equipped with a Gatan 3View for our SBEM imaging. The optimal performance for this SBEM was achieved at lower landing energies (down to 1.7 keV). At these energies, we achieved a smaller interaction volume in the sample that resulted in improved axial resolution and less energy deposition on the specimen. As a result, we were able to cut finer sections. However, at these lower energies, our probe was also subject increased chromatic aberration effects and we experienced decreased BSE detector sensitivity.
We initially performed landing-energy experiments by placing a 50- and a 30-nm-thick blank section of Durcupan resin cut on a Leica UC EM6 ultramicrotome over a biological sample, then imaging through the layer. As we varied the acceleration voltage, we looked for signal emanating from the layers below. Figure 4 shows a panel of images of the 50-nm-blank Durcupan resin overlays on the left and 30-nm-blank section overlays on the right imaged at 3.0, 2.0, and 1.7 keV. At 3.0 keV, a great deal of signal is observed from below the epoxy layer for both 50- and 30-nm-blank sections. After stepping down to 2.0 keV, the interaction volume is reduced. As expected, we see much less signal coming from the region beneath the 30-nm-blank section and just about no signal coming from the region beneath the 50-nm section. As the acceleration voltage was decreased to 1.7 keV, very few BSEs originating from below 30 nm were detected, demonstrating that we effectively reduced our penetration depth.
However, image quality was degraded due to decreased detector sensitivity. These low-energy-loss BSEs arose from elastically scattered or nearly elastically scattered electrons. Small energy losses are due to interactions with the atomic nuclei of the sample stain, and they backscatter with nearly the same energy as the probe electrons [21, 23]. Since most solid-state backscatter detectors are built from silicon photo-diode circuitry that have a thin-surface passivation layer and metalized electrodes, low-energy BSEs do not deposit much ionization charge in the active region of the diode and are difficult to detect above inherent detector noise.
Testing the impact of beam deceleration on SBEM resolution
To achieve the low electron-landing energies required for thinner penetration while avoiding detector sensitivity issues, we decelerated the probe-beam electrons at the final stage of imaging after the objective lens with the use of a negative-bias voltage applied at the sample stage.
We repeated the same overlay experiments as shown in Fig. 4, but this time, we kept the accelerating voltage fixed at 3.0 keV and adjusted the negative-bias voltage to produce a final landing energy between 3.0 and 1.0 keV. Figure 5 shows the images of the plastic-embedded sample with a 30-nm-blank section overlay across the surface at various landing energies, achieved by adjusting the negative reversal potential. It is clear from the BSE signal-depth retrieval images that we reduced our penetration depth and retained good image quality and signal to noise even at extremely low landing energies. For 30-nm ultramicrotome sectioning, a landing energy of between 1.0 and 1.5 keV was optimum. Electrons striking the BSE detector did so with an energy nearly matching the 3-keV instrument accelerating voltage and with high signal to noise: on the order of 20× greater signal can be achieved at a similar landing energy without the use of the negative-bias voltage.
Repeated imaging of the block face often results in altered viscosity properties and electron-beam-induced damage to the plastic block, causing the block to become difficult to section. By reducing the primary beam-landing energy, we can reduce damage to sample. Moreover, the increase in signal to noise provided by the acceleration of the BSEs reduces the dose per unit area and allows for significantly faster scan rates at higher magnifications. The increase in signal to noise and the ability to operate at lower dose rates should not only benefit SBEM but also other techniques, such as FIB-SEM [8, 24, 25], which is limited by probe-penetration depth and detector sensitivity.
Characterizing detector response as a function of backscatter electron energy
As detailed in the “Methods” section, above, the two backscatter detectors (Gatan BSD and FEI CBS) were configured differently. In the conventional backscatter detection mode without deceleration (Gatan BSD), the signal produced in the silicon photo-diode detector drops off significantly, as the accelerating voltage is lowered. To determine the rate at which the signal in each detector drops off as a function of electron energy, we measured the detector response when the electrons were backscattered from a gold substrate compared with a carbon background as the difference between gold and carbon signals.
The FEI Quanta FEG 200 microscope was aligned stepwise for accelerating voltages of 5–1-keV accelerating voltage, and a Faraday cup (Ted Pella, Inc.) measured the beam current at a fixed spot size for each voltage. By integrating the backscatter signal difference for the same area of the gold sample and carbon background at each accelerating voltage, we could normalize backscatter signal for beam-current differences between accelerating voltages. In addition, we normalized the signal according to the backscattering coefficient η for gold, where
The results of both detector responses as a function of gun-accelerating voltage are also shown in Fig. 6. To compare the two detectors, we kept the amplifier gains and brightness settings constant for both detectors, and the central ring of the CBS was turned off.
The results show that depending on the gun-acceleration potential and the deceleration voltage, signal level increases of up to 20× and 6× or greater are possible at 1- and 1.5-keV landing energy, respectively, for decelerated probe-beam electrons versus the conventional SBEM. In addition, the signals in the CBS detector used with deceleration are actually higher than those where deceleration was not applied. We attribute this to the fact that the bias fields help collect and collimate the BSE electrons toward the BSD detector.
To demonstrate the increase in signal produced by the acceleration of BSEs when the reversal potential is used, we used the FEI CBS detector with the central ring turned off (the central ring becomes saturated with the SE signals and should be turned off when detecting BSEs with deceleration). The Gatan BSE detector was not used in this deceleration experiment, because the detector becomes saturated from accelerated SEs.
With the FEI CBS detector’s central ring turned off, we explored the BSE signals from the gold sample with 3- and 4-keV gun-accelerating energies and deceleration appropriate to achieve 1- and 1.5-keV landing energies. Although the electrons are decelerated when they travel down the column from the objective pole piece to the sample, electrons leaving the sample as BSEs are re-accelerated by the same amount. Therefore, the BSEs have nearly the same energy as the electron beam at the pole piece, theoretically with some energy loss due to the inelastic collisions with the atoms of the sample. For example, a 3-keV electron will be decelerated to a 1.5-keV landing energy in a −1.5-keV bias potential, and the BSEs will be re-accelerated to energy of almost 3 keV before striking the detector.
We explored a number of column-beam energies and deceleration potentials and found that the best results were obtained with deceleration between −1 and −2 keV and with a gun-accelerating energy of 2.5 keV and higher. For column-beam energies less than 2 keV, the beam current drops off quickly, and the image quality and stability are less than ideal. For deceleration voltages less than −1 keV, the electron re-acceleration was not as dramatic. For deceleration voltages higher than −2.5 keV, focus and astigmatism stability were decreased, and sample movements and distortions became problematic. For this geometry, we obtained our best results with moderate deceleration. However, we feel that, by careful design of the deceleration hardware and geometry, it should be possible to mitigate stability issues, enabling long runs at much higher bias voltages.
Using the biasing field to spatially differentiate SE from BSE signals
The SBEM technique relies primarily on BSE signals for image generation. The image contrast arises from differences in electron scattering from the lighter Z-number atoms (hydrogen, carbon, nitrogen and oxygen) in the tissue and the plastic-embedding media, and the high-Z elements (osmium, uranium and lead) from the stains used in the specimen-preparation protocol. SEs are also created at the sample surface but usually have energies lower than 50 eV and, thus, are not detected in a photo-diode backscatter detector. Once deceleration is applied, these same low-energy secondary electrons are accelerated by the electric fields, striking the detector with sufficient energy to produce a measurable signal.
Enhanced SBEM volume acquisition using deceleration
Adding an aperture to the monolithic BSD enabled the discrimination of BSEs and SEs similar to that of the CBS detector
With the ability to separate BSE and SE signals by turning off the inner ring and using only the outer rings of FEI’s concentric-ring BSE detector in combination with moderate deceleration, we then took the next step to determine if a simplified setup could be used, employing an monolithic Gatan BSE detector with a blocking aperture to remove the large SE signal from our images.
Careful design of the correct blocking aperture diameter for the geometry and the deceleration potential is critical for BSE/SE separation. We found that the most important parameters were the gun-accelerating voltage, decelerating voltage, and landing energy of the probe beam; the distance from the sample to the detector; and the diameter of the SE-blocking aperture.
Weighing these considerations, we built a simple prototype blocking aperture by modifying a standard TEM aperture, which had a 5.0-mm outer diameter ring with a central hole of 0.8-mm diameter. This custom SE-blocking aperture was centered and attached above the diode to the grounded shielding of the Gatan BSE detector. Using this setup, any SEs striking the aperture should be conducted away quickly, leaving only BSEs to be detected on the BSE device.
The addition to Gatan’s microscopy suite of a new autofocus and autostigmation routines should enable stable imaging for weeks at a time using this technique. Future development of deceleration-based detection would also benefit from fabrication of a variety of blocking apertures to better optimize aperture geometry to the deceleration potential employed.
This study demonstrates that using beam deceleration in the SBEM results in higher quality images. Thinner axial-penetration depth achieved with probe-beam deceleration improved axial resolution. A remarkable increase in total signal collection of 20-fold or higher was produced with the use of relatively low-deceleration potentials as compared with the conventional SEM imaging using the same landing energies. This increase in signal allowed for much lower interrogation currents on the sample, resulting in improved sample-sectioning properties.
Here, we demonstrated a procedure that mitigates the degradation of the image quality resulting from low-energy SEs when using deceleration. By removing SEs from the images, using the FEI concentric backscatter detector or a simple SE-blocking aperture in a monolithic detector, beam deceleration becomes a feasible approach to achieve limited penetration depths with sufficient sample and imaging stability to produce large volumes of data. Additional improvements in sample preparation, particularly the use of conductive epoxy resins formulated for SBEM, should improve the technique of beam deceleration by producing uniform electric fields at the sample surface and reducing surface-charging effects.
We believe that this study demonstrates the first of many improvements that can be made to SBEM imaging. We observed drift in image position, astigmatism, and focus changes, so these artifacts still represent challenges that need to be addressed to optimize volume quality. We found, however, that these were not significant obstacles at lower deceleration voltages, and we now believe that, with improved system geometry and the addition of autofocus software now available from Gatan, these instabilities can be addressed easily, enabling significant improvements in volumetric reconstructions.
serial block-face scanning electron microscopy
backscattered electron detector
JB, SP, and ME designed the experiments. JB performed the experiments and analyzed the data. TD and EB helped with sample preparation. RR performed the Monte Carlo simulations. VA helped with data processing. JB, RR, SP and ME wrote the paper. All authors read and approved the final manuscript.
This work was supported by a grant from the NIH National Institute of General Medical Sciences under award number P41GM103412 to Mark H. Ellisman, to operate the National Center for Microscopy and Imaging Research. The authors would like to acknowledge Gatan, FEI Company, and Carl Zeiss for loan of SBEM equipment used in this study.
The authors declare that they have no competing interests.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
- Shu, X., Lev-Ram, V., Deerinck, T.J., Qi, Y., Ramko, E.B., Davidson, M.W., Jin, Y., Ellisman, M.H., Tsien, R.Y.: A genetically encoded tag for correlated light and electron microscopy of intact cells, tissues, and organisms. PLoS Biol 9(4), e1001041 (2011)View ArticleGoogle Scholar
- West, J.B., Fu, Z., Deerinck, T.J., Mackey, M.R., Obayashi, J.T., Ellisman, M.H.: Structure-function studies of blood and air capillaries in chicken lung using 3D electron microscopy. Respir Physiol Neurobiol 170(2), 202–209 (2010)View ArticleGoogle Scholar
- Williams, M.E., Wilke, S.A., Daggett, A., Davis, E., Otto, S., Ravi, D., Ripley, B., Bushong, E.A., Ellisman, M.H., Klein, G., Ghosh, A.: Cadherin-9 regulates synapse-specific differentiation in the developing hippocampus. Neuron 71(4), 640–655 (2011)View ArticleGoogle Scholar
- Jurrus, E., Hardy, M., Tasdizen, T., Fletcher, P.T., Koshevoy, P., Chien, C.B., Denk, W., Whitaker, R.: Axon tracking in serial block-face scanning electron microscopy. Med Image Anal 13(1), 180–188 (2009)View ArticleGoogle Scholar
- Nguyen, J.V., Soto, I., Kim, K.Y., Bushong, E.A., Oglesby, E., Valiente-Soriano, F.J., Yang, Z., Davis, C.H., Bedont, J.L., Son, J.L., Wei, J.O., Buchman, V.L., Zack, D.J., Vidal-Sanz, M., Ellisman, M.H., Marsh-Armstrong, N.: Myelination transition zone astrocytes are constitutively phagocytic and have synuclein dependent reactivity in glaucoma. Proc Natl Acad Sci USA 108(3), 1176–1181 (2011)View ArticleGoogle Scholar
- Leighton, S.B.: SEM images of block faces, cut by a miniature microtome within the SEM—A technical note. Scan. Electron Microsc. 2, 73–76 (1981)Google Scholar
- Denk, W., Horstmann, H.: Serial block-face scanning electron microscopy to reconstruct three-dimensional tissue nanostructure. PLoS Biol 2(11), 1900–1909 (2004)View ArticleGoogle Scholar
- Knott, G., Marchman, H., Wall, D., Lich, B.: Serial section scanning electron microscopy of adult brain tissue using focused ion beam milling. J. Neurosci 28(12), 2959–2964 (2008)View ArticleGoogle Scholar
- Ohta, K., Sadayama, S., Togo, A., Higashi, R., Tanoue, R., Nakamura, K.: Beam deceleration for block-face scanning electron microscopy of embedded biological tissue. Micron 43(5), 612–620 (2012)View ArticleGoogle Scholar
- Bauer, E.: The resolution of the low-energy electron reflection microscope. Ultramicroscopy 17(1), 51–56 (1985)View ArticleGoogle Scholar
- Telieps, W., Bauer, E.: An analytical reflection and emission Uhv surface electron-microscope. Ultramicroscopy 17(1), 57–65 (1985)View ArticleGoogle Scholar
- Frank, L., Mullerova, I.: Strategies for low- and very-low-energy SEM. J Electron Microsc 48(3), 205–219 (1999)View ArticleGoogle Scholar
- Matsuda, K., Ikeno, S., Mullerova, I., Frank, L.: The potential of the scanning low energy electron microscopy for the examination of aluminum based alloys and composites. J Electron Microsc 54(2), 109–117 (2005)View ArticleGoogle Scholar
- Mullerova, I.: Imaging of specimens at optimized low and very low energies in scanning electron microscopes. Scanning 23(6), 379–394 (2001)View ArticleGoogle Scholar
- Khursheed, A., Osterberg, M.: A spectroscopic scanning electron microscope design. Scanning 26(6), 296–306 (2004)View ArticleGoogle Scholar
- Pluk, H., Stokes, D., Lich, B., Wieringa, B., Fransen, J.: Advantages of indium-tin oxide-coated glass slides in correlative scanning electron microscopy applications of uncoated cultured cells. J Microsc 233(3), 353–363 (2009)View ArticleGoogle Scholar
- Titze, B., Denk, W.: Automated in-chamber specimen coating for serial block-face electron microscopy. J Microsc 250(2), 101–110 (2013)View ArticleGoogle Scholar
- Kanaya, K., Okayama, S.: Penetration and energy-loss theory of electrons in solid targets. J Phys D Appl Phys 5(1), 43 (1972)View ArticleGoogle Scholar
- Joy, D.C.: Monte carlo modeling for electron microscopy and microanalysis. Oxford University Press, Oxford (1995)Google Scholar
- Goldstein, J.I., Newbury, D.E., Echlin, P., Joy, D.C., Lyman, C.E., Lifshin, E., Sawyer, L., Michael, J.R.: Scanning electron microscopy and microanalysis. Springer, Berlin (2003)View ArticleGoogle Scholar
- Reimer, L.: Scanning electron microscopy—physics of image formation and microanalysis. In: Hawkes, P.W. (ed.) Springer series in optical sciences, vol. 45. Springer, Berlin (1998)Google Scholar
- Agar, A.W., Alderson, R.H., Chescoe, D.: Principles and practice of electron microscope operation, In: Glauert, A.M. (ed.) American Elsevier Publishing Co, New York (1974)
- Joy, D.C., Joy, C.S.: Low voltage scanning electron microscopy. Micron 27(3–4), 247–263 (1996)View ArticleGoogle Scholar
- Helmstaedter, M., Briggman, K.L., Denk, W.: 3D structural imaging of the brain with photons and electrons. Curr Opin Neurobiol 18(6), 633–641 (2008)View ArticleGoogle Scholar
- Schroeder-Reiter, E., Perez-Willard, F., Zeile, U., Wanner, G.: Focused ion beam (FIB) combined with high resolution scanning electron microscopy: a promising tool for 3D analysis of chromosome architecture. J Struct Biol 165(2), 97–106 (2009)View ArticleGoogle Scholar
- Assa’d, A.M.D., El Gomati, M.M.: Backscattering coefficients for low energy electrons. Scanning Microsc 12(1), 185–192 (1998)Google Scholar
- Phifer, D., Tuma, L., Vystavel, T., Wandrol, P., Young, R.J.: Improving SEM imaging performance using beam deceleration. Microsc Today 17(4), 40–49 (2009)View ArticleGoogle Scholar