cryo-sem of plants

7MicroscopyandAnalysis | September 2013

Cryo-scanning electron microscopy of plant samples without metal coating, utilizing bulk conductivity Hans-Jürgen Ensikat and Maximilian Weigend Nees Institute for Biodiversity of Plants, University of Bonn, Bonn, Germany

IntroductionCryo scanning electron microscopy (cryo-SEM) is, alongside environmental SEM, the most reliable technique to avoid drying artifacts in biological SEM specimens [1-4]. But due to the very low electrical conductivity of deeply frozen samples, metal coating is conventionally used to increase surface conductivity. This automatically precludes the application of compositional contrast imaging on these samples. In recent years, the availability of advanced preparation equipment and cryo-SEM stages has led to an increased utilization of this technique. However, preparation and handling are not trivial. They require some experience and complex equipment.

The standard preparation protocol includes shock freezing, the transfer of the sample into a vacuum chamber, optionally partial sublimation of ice, sputter coating with metal, and a transfer into the SEM. Shock freezing reduces the disruptive growth of ice crystals in the tissue, and metal coating provides the necessary electrical conductivity to avoid charging. Occasional publications [5] show the feasibility of cryo-preparations with simple equipment and the examination of frozen samples without metal coating, but often at the cost of poor image quality.

In our studies, we observed that the electrical conductivity of uncoated frozen hydrated specimens increases with the temperature. Above -90oC conductivity is often sufficiently high to avoid charging, so that an examination is possible with satisfactory imaging quality in both secondary electron (SE) and backscattered electron (BSE) modes [6-7].

In this article we document the possibilities of cryo-SEM without metal coating on several botanical samples, using simplified preparation methods.

Material and methods

Equipment Scanning electron microscopy was carried out using a Cambridge Stereoscan S200 SEM and a LEO 1450 SEM, both equipped with SE and BSE detectors, an energy-dispersive X-ray (EDX) system and with custom-made cooling holders for

low-temperature examinations (Figure 1). Topographical images with the SE signal were recorded preferentially with 5 to 10 kV accelerating voltage and 10 pA (or less) beam intensity, whereas BSE imaging was performed with 15 to 20 kV and at least 20 pA.

Cryo-preparation and scanning electron microscopy The fresh samples were cooled relatively slowly to below -100oC by placing the entire cooling holder in a liqN2 bath. The initial cooling rate was ca 2 K sec-1 (see diagram Figure 1b). Sample immersion in the coolant was avoided in order to preserve the original specimen surface. Optionally, the frozen samples were freeze-fractured. After closing the sample holder chamber with a cover, the holder was inserted into the SEM. After evacuation of the specimen chamber, the cover of the holder was removed and the sample was examined using very low beam intensities. Due to the thermal insulation, the temperature of the holder rises only

slowly (see diagram Figure 1c). Thus, the samples can be examined for several hours at sufficiently low temperatures.

Freeze drying The cooling holder with the frozen sample was simply left in the SEM overnight, gradually reaching room temperature. In this interval the tissue water sublimated at temperatures far below 0oC. The dried samples were mounted on normal SEM stubs and sputter-coated.

Results The frozen samples were examined, beginning at temperatures far below -100oC, and continuing while the temperatures rose slowly. Image quality gradually improved with increasing temperature. Figure 2 shows the superhydrophobic surface of a leaf of Xanthosoma robustum (‘elephant ear’), which resembles a lotus leaf in its hierarchical structure. The morphology of the papillose epidermis cells is well preserved on the frozen hydrated samples (Figure 2 a-c). At a temperature of -130oC

Figure 1 (a) The custom-made passive cooling holder. Integrated Pt-100 resistors enable temperature measurement and optionally heating. (b) The cooling rate of the holder during ‘slow cooling’ in a LN2 bath. (c) The warm-up rate of the holder in the SEM during examination. Due to the thermal isolation, the useful temperature range from ca -90o to -60oC is retained for at least one hour.

cryo-sem of plants

8 September 2013 | MicroscopyandAnalysis

(Figure 2a), insufficient conductivity of the sample causes bad contrast due to charging. At -70oC, the charging effects have vanished and the contrast is good, even without metal coating. Despite the slow freezing process, the surface structure is not influenced by the formation of ice crystals in the tissue. At higher magnifications, resolution and contrast of fine structures such as the tiny wax crystals appear not optimal (Figure 2c). However, for a high-resolution examination, the following simple freeze-drying produced suitable results. Figures 2 d,e show the freeze-dried sputter-coated sample with only slight shrinkage artifacts and the fine detail of the wax platelets.

Using adequate test liquids, it is possible to study, for example, wetting phenomena on superhydrophobic plant surfaces in a frozen state by cryo-SEM. Figures 3 and 4 show a drop of a glycerol/water/lead acetate (5:3:2) solution with suitable wetting properties, that does not crystallize during freezing and can be detected by the BSE signal due to the heavy metal content. Figures 3 a-c show the decreasing charging effects at

Figure 2 Superhydrophobic leaf of Xanthosoma robustum with a hierarchical surface structure of papillose epidermis cells covered with µm-sized wax platelets. (a-c) Cryo-SEM images of the samples without metal coating. In the frozen state, the morphology appears perfectly preserved. (a) At -130oC, insufficient electrical conductivity causes charging effects. (b,c) At -70oC, charging has disappeared. Resolution and contrast of the fine detail are poor on samples without metal coating. (d,e) The same sample after freeze-drying and metal coating shows only slight shrinkage (d) and is well suited for high-magnification examination of the wax crystals (e).

increasing temperatures. At -150oC (Figure 3a) strong charging of the drop causes distortion of the image by deflection of the electron beam in the vicinity of the drop. At -120oC (Figure 3b), contrast is still unsatisfactory due to deflection of the secondary electrons. Finally, at -80oC (Figure 3c) charging disappears, so that the sample can be examined in detail.

Figure 4a shows the point of contact between the drop and the tips of the papillae on the ‘hanging’ or receding side of the drop. (The contacts at the advancing side are hidden under the drop and thus more difficult to observe.) Small droplets on the surrounding papillae tips (Figures 4 b-d) indicate that they have been in contact with the main drop during sample preparation. The colour images were made by combining a SE image (Figure 4c) and a BSE image (Figure 4d), which shows the droplets as bright spots, by the image processing function ‘Combine HSL’.

Compositional contrast imaging using a BSE detector can be used for several applications, as the samples are not metal coated. Accumulation of silica and calcium compounds in plants (biomineralization)

can be examined by compositional contrast imaging and analyzed by EDX. Figures 5 a-c show a stinging hair and other small hairs of a stinging nettle (Urtica bianorii) with both silica and calcium incrustations. Micro-organisms such as bacteria and fungi can be recognized well by staining with osmium tetroxide - the leaf epidermis remains unstained (Figures 5 d,e). In cryo-fractured samples, calcium oxalate crystals or druses (Figure 5f) can be recognized reliably using the BSE image.

Cryo-fractures are valuable for distinguishing aerial from fluid-filled spaces in plant tissues. The druses in the Opuntia stem (Figure 5f) are clearly visible in the aerial space and are covered with a thin transparent membrane, most likely a residue of a dead cell.

The cryo-fracture images of the Xanthosoma leaf (Figures 6 a,b) show the pattern of coarse ice crystals inside the cells, which becomes distinctively visible as the ice begins to sublimate. However, the surface of the inner (parenchymatic) cells appears well preserved, despite their very thin cell wall.

Figure 3 A drop of a non-wetting test liquid on the superhydrophobic Xanthosoma leaf surface at different temperatures. (a) At -150oC, strong charging causes a distortion of the image. (b) At -120oC, charging effects decreased, but the contrast is still bad. (c) At -80oC, the conductivity of the sample is sufficient for a detailed examination.

cryo-sem of plants

9MicroscopyandAnalysis | September 2013

Figure 4 A drop of a test liquid containing lead acetate and drop remnants on top of the leaf papillae at -80oC. (a,b) The colour images were made by combining SE and BSE images. The SE image (c) shows the topography; the corresponding BSE image (d) shows the heavy-metal containing droplets as bright spots due to its compositional sensitivity.

Discussion It is amazing to observe how the image quality of frozen hydrated samples improves and charging effects vanish, when the temperature rises above approximately -100oC. The reason is the bulk conductivity of ice. Ice is known to be a proton semiconductor [8-9], and its conductivity increases strongly with temperature. Proton conductivity is often found in solids such as ice or acidic salts, in contrast to other ions that provide conductivity only in liquids.

At temperatures above -120o to -90oC, a sufficient image quality may be achieved for the plant samples, depending on specimen thickness, beam intensity, magnification, and other factors, presupposing that the samples are completely hydrated, since dry parts would have insufficient conductivity. At temperatures above ca -100oC ice begins to sublimate. As long as solid surfaces are imaged, the incipient sublimation of ice usually does not impair imaging, apart from a slight drift at higher magnifications. On fractured samples, ice sublimation (etching) reveals details of internal cell structures, particularly on shock-frozen samples.

Omitting the metal coating not only simplifies the preparation. It facilitates several applications that use the detection of material contrast, which would be hidden by a metal coating, or the study of the wetting behaviour of liquids. We thus used the cryo-preparation successfully for the study of biomineralization in plants and drop adhesion on superhydrophobic plant surfaces.

Moreover, cryo-SEM is not limited to botanical or biological specimens (Figure 7). The presented experimental setup without temperature stabilization and without metal coating is not dedicated for high resolution imaging and cannot replace high-performance cryo-workstations with their high resolution and stable temperature control for certain applications. But the study of biological samples needs a versatility of preparation and imaging techniques even for imaging at moderate magnifications.

ConclusionsFrozen hydrated biological samples without metal coating can be imaged and analysed successfully in a conventional SEM at temperatures above -100oC at magnifications up to at least 2000X, due to the electrical conductivity of ice. This is particularly advantageous for the identification of material contrasts, for example, from mineral accumulations or stained structures, with the cryo-preparation as a reliable technique to avoid dehydration artifacts. Omitting the metal coating simplifies the preparation, which can be performed using a self-developed passively cooled cryo-sample holder. Slow cooling of the samples is sufficient, if

Figure 5 Examples for compositional contrast using BSE imaging of frozen samples without metal coating. (a-c) Combined image, BSE image, and EDX element spectrum of a stinging hair of Urtica bianorii. The mineralized structures appear bright in the BSE image (b). (d) Bacteria and (e) fungi, stained with osmium tetroxide, on plant surfaces can be recognized well in the BSE image used for colourisation. (f) Calcium oxalate druses appear as bright stars in the cryo-fractured cactus Opuntia microdasys.

cryo-sem of plants

10 September 2013 | MicroscopyandAnalysis

Figure 7Natural snowflake

solid surfaces are to be examined. These methods open up an easy approach to cryo-SEM, at least when high magnifications are not required.

References1. Pathan, A. K. et al. Sample preparation for scanning electron microscopy of plant surfaces - Horses for courses. Micron 39:1049-1061, 2008.2. Echlin, P. and Taylor, S. E. The preparation and X-ray microanalysis of bulk frozen hydrated vacuolated plant tissue. Journal of Microscopy 141:329-348, 1986.3. Echlin, P. Handbook of sample preparation for scanning electron microscopy and X-ray microanalysis. Springer, New York, USA. 2009. 4. McCully, M. E. et al. Cryo-scanning electron microscopy (CSEM) in the advancement of functional plant biology. Morphological and anatomical applications. Functional Plant Biology 36:97-124, 2009.5. Tang, C. Y. et al. A simple cryo-holder

Figure 6 Cryo-fractured leaf of Xanthosoma robustum at -75oC. The aerial spaces can be clearly recognized. After beginning sublimation of the ice, coarse patterns of ice crystal become visible on the fractured cell faces (b, arrow), but the cell surfaces appear well preserved and undamaged.

facilitates specimen observation under a conventional scanning electron microscope. Microscopy Research and Technique 75:103-111, 2012. 6. Ensikat, H. J. et al. Scanning electron microscopy of plant surfaces: simple but sophisticated methods for preparation and examination. In: Microscopy: Science, Technology, Applications and Education. FORMATEX Microscopy series No. 4, Vol. 1. Eds: Méndez-Vilas, A. and Diaz, J. Formatex Research Center: Badajoz, Spain. pp 248-255, 2010.7. Ensikat, H. J. SEM-based diagnostic tools to determine the quality of superhydrophobic surfaces. In: Proceedings of the 7th International Conference MMT-2012. Eds: Zinigrad M. and Lugovskoy A. Ariel University Center of Samaria, Ariel, Israel. pp 3.1-3.7, 2012.8. Onsager, L. Protonic semiconductors. In: Physics of Ice. Eds: Riehl, N. et al. Plenum, New York, USA. pp 363-368, 1969. 9. Petrenko, V. and Whitworth, R. Physics of Ice. Oxford University Press, 1999, 2001, 2003.

biography Hans-Jürgen Ensikat has worked for 40 years at the University of Bonn as a technician in several EM laboratories. After an apprenticeship in chemistry, he joined the TEM working group ‘Festkörperoberflächen’ in the Mineralogical Institute. Later he relocated into the TEM facility of the Botanical Institute. 1988 he moved to the Barthlott group in the Botanical Institute / Nees Institute, well-known for the “Lotus Effect”, where he is responsible for SEM and AFM. Besides routine microscopy and the education of users, his interest is to improve and develop sample preparation techniques.

abstractA classical problem in scanning electron microscopy is the alteration of specimen surfaces due to sample preparation. Simplified techniques for sample preparation with minimized surface alteration are therefore desirable. For fresh (hydrated) biological material, cryo-SEM after preceding shock-freezing and metal-coating is favoured when the prevention of drying artifacts has priority. A detailed knowledge of the sample properties and behaviour can facilitate new ways to optimize or simplify preparation and microscopy. Above ca. -100oC, ice has a measurable electrical conductivity due to its proton-conducting properties, which is sufficient to prevent frozen hydrated samples from charging during scanning electron microscopy, at least at low beam intensities. The examination of frozen samples without metal coating enhances the flexibility and versatility of cryo-SEM and simplifies sample preparations. We present some applications for the study of botanical samples, including compositional contrast imaging using backscattered electrons (BSE), and the study of wetting phenomena. The capabilities of simplified procedures are here demonstrated, particularly in combination with slow-freezing and the use of suitable test liquids for certain applications. The techniques can be easily adopted for conventional SEM setups.

Corresponding author details Hans-Jürgen Ensikat, Nees Institute for Biodiversity of Plants, University of Bonn,Meckenheimer Allee 170, 53115 Bonn, Germany Tel: +49 (0) 228 733 273Email: ensikat@uni-bonn.de

Microscopy and Analysis 27(6):7-10 (EU), 2013

©2013 John Wiley & Sons, Ltd