Transmission Electron Microscopy as a Tool for the Characterization of Soft Materials: Application and Interpretation

Abstract

Transmission electron microscopy (TEM) provides direct structural information on nano-structured materials and is popular as a characterization tool in soft matter and supramolecular chemistry. However, technical aspects of sample preparation are overlooked and erroneous image interpretations are regularly encountered in the literature. There are three most commonly used TEM methods as we derived from literature: drying, staining and cryo-TEM, which are explained here with respect to their application, limitations and interpretation. Since soft matter chemistry relies on a lot of indirect evidence, the role of TEM for the correct evaluation of the nature of an assembly is very large. Mistakes in application and interpretation can therefore have enormous impact on the quality of present and future studies. We provide helpful background information of these three techniques, the information that can and cannot be derived from them and provide assistance in selecting the right technique for soft matter imaging. This essay warns against the use of drying and explains why. In general cryo-TEM is by far the best suited method and many mistakes and over-interpretations can be avoided by the use of this technique.

Keywords: artefacts; cryo‐TEM; sample preparation; self‐assembly; vesicles.

Figure 1

Electron microscopic methods used for the evaluation of “vesicles” and “self‐assembly” structures. Methods of all literature combined (left), and divided into groups of roughly 50 best cited and 50 newest papers of each search (right). Green represents cryo‐TEM, blue staining and red drying, black represents all three of them and grey a different method altogether.

Figure 2

Demonstration of different structures that can be found upon drying of a) a buffer (Hepes, NaCl, α‐Dodecyl Maltoside); various water soluble polymers: b) poly‐(N,N‐dimethylacrylamide (mw 39 kD),35 c and d) polystyrene sulfonate sodium salt (mw 220 kD) and e) polyethylene oxide (mw 97kD); and f) a molecular motor in dichloromethane/acetonitrile.36 These patterns and structures are not representative for these samples in solution, because none of them form aggregates. Different types of drying patterns can be found for different materials, but also on different areas of the grid (c and d), and when varying the polarity of the surface. Black and white scale bars represent 50 nm and 1 µm respectively.

Figure 3

Negative stain EM of phospholipid vesicles (white arrows) mixed with latex spheres (black arrows), negatively stained with a, 2% UAc and b, 2% phospho tungstic acid and schematically visualised with a side‐ and topview (c). Negatively stained objects appear light with a dark halo (arrows). During preparation some sample ended up on the other side of the support grid and was not stained as the UAc solution was only applied to the top of the grid.39 Both liposomes and latex are therefore stained and unstained in this image. The unstained liposomes cannot be seen and may be responsible for the contrast differences in the background. The unstained latex (arrow heads) is amorphous, but appears to have some structure inside, this is in fact the imprint into the carbon support film. A similar effect can be sometimes observed in ice contamination. Stained latex can be interpreted as being hollow, but it is not, which becomes clear from the unstained latex. Notice that the intensity of stained and unstained latex is exactly the same. Uranyl acetate crystals deposited on negatively stained liposome sample (d). In absence of sample the crystals can be mistaken for it. Scale bars represent 100 nm.

Figure 4

Cryo‐electron microscopy of phospholipid vesicles (white arrows) mixed with latex spheres (black arrows) were plunge‐frozen in liquid ethane (a) and schematically visualised with a side‐ and topview (b). In the schemes, the dark grey is the carbon support and the fading layer is the vitreous ice. As ice contamination (black arrowheads) falls on top of the ice‐layer, it creates extra thickness and has more contrast than latex, which sits inside the vitreous ice. In cryo, the difference between solid and hollow objects can be observed. Figures a and b show ice contamination (black arrow head). Radiation damage (figures a and d, white arrowheads) gives light round bells which can be mistaken for the sample (vaguely visible in the background). Scale bars represent 100 nm.

Figure 5

Doxorubicine liposomes (Dox‐NP®) imaged by the most frequently used techniques in soft matter electron microscopy: dried sample without staining (a), UAc stained sample after two minutes of drying (b), negative stained sample (UAc) (c) and cryo‐TEM (d). White scale bars represent 200 nm and black scale bars 50 nm.

Figure 6

End capped amphiphilic overcrowded alkene nanotubes. a) dried without staining, b) dried before staining, c) negatively stained with 2% UAc and d) cryo‐electron microscopy. In the dried sample only the contours of the nanotubes are visible but the details are completely lost (lower panel of a). The black arrow heads indicate ice contamination. Black scale bars represent 1 µm, white scale bars 200 nm.