Transmission electron microscopes (TEM) facilitate analysis of structures in the range of several µm to as small as 0.05 nm . In the biological field, analysis of 3-dimensional structures of cell organelles and macro molecules is of particular interest . To realize in vivo conditions, biological specimen are often being frozen, i.e. vitrified in amorphous ice, and analyzed at temperatures below -140 degrees Celsius. This technique is called cryo-electron-tomography (cryo-TEM). Frozen biological specimens are very sensitive to beam damage. The electron dose has to be very low so that the microscope’s electron detection ability becomes a limiting factor. This leads to very noisy images; the resolution is impaired. In addition to resolution, contrast is of high importance and biological samples show almost no contrast around the focal plane (differently from material science samples). To improve contrast, biological samples can be “stained” with heavy metal salts. If this is not feasible, e.g. for frozen samples, contrast may be improved through strong defocusing. With this method, however, resolution is impaired considerably, as we will see later. Special optical elements, so-called “phase contrast systems” are a new, promising approach to obtain focused images with good contrast. Such systems would not impair the resolution.
What is phase contrast?
Electrons in a TEM can be compared to a light wave. The electrons of the beam interact with the sample; this generates image contrast. An electron wave can be altered in its amplitude and in its phase. In the former case, we talk about amplitude contrast, in the latter about phase contrast. Amplitude contrast is the dominating form of contrast for samples consisting of mainly heavy elements. Such elements scatter the electrons such heavily that they do not contribute to the image generation. The relevant areas of the specimen appear dark. For thin specimen consisting mainly of light elements, e.g. frozen biological material, the amplitude is only insignificantly altered. Structural information is mainly contained in a phase shift of the electron beam; phase contrast is thus the dominating form of contrast. Such samples are called “phase objects”. The phase shift, however, cannot be made visible directly with bright-field microscopy and bright-field images thus contain only very weak contrast (figure 1).
Figure 1: Bright-field TEM image of a liposome vesicle embedded in amorphous ice.
Phase information can be made visible through phase plates.
Phase contrast is generated by interference between scattered and unscattered electrons. If the phase shift is minor, as is the case with phase objects, there is almost no contrast. Figure 2 shows the gradient of the contrast in a TEM in the form of the contrast transfer function (CTF) at different microscope settings. The CTF is different for each TEM and describes the contrast plotted against the object size. The object size is displayed as frequency. For the imaging of biological samples, we need high contrast at low frequencies (0.25nm-1 to 1.5nm-1). The blue graph represents the CTF of a phase object near the focal plane (-10nm defocus). One can see that the contrast transfer at low frequencies is weak. This can be improved if the microscope is put in a strong defocus (500-10,000nm under-focus). The contrast depends on the chosen defocus. The black dotted line in figure 2 shows the CTF at 1,000nm under-focus. Under these conditions, contrast is transferred at low frequencies but resolution is impaired. Furthermore the CTF is highly oscillating with multiple zeros. At those points, no image information is transmitted.
A phase contrast system induces an additional phase shift into the electrons. This alters the interference between unscattered and phase-shifted electrons and increases the contrast (what counts is the absolute value of the CTF, shown in the red graph in figure 2). The transferred contrast is highest at a phase shift of 90 degrees. This principle is used by the phase contrast systems. Figure 2 shows the CTF of a Zernike phase plate. The contrast transfer is highest up to a frequency of about 2nm-1, meaning that high contrast can be achieved for structure sizes up to 0.5nm. The phase plate was invented by Frits Zernike, a Dutch physicist. In 1930 he developed the first phase contrast system for light visible light microscopes (VLM). Initially, the relevance of his invention was greatly underestimated . During WWII, the German Wehrmacht found use for it and built the first phase contrast VLMs. Subsequently, these microscopes found great use in many fields of science, especially in medicine and Zernike was awarded the Nobel price for physics in 1953.
Figure 2: Simulated contrast transfer function (CTF) with and without phase contrast system.
Working principle of a phase plate
Since the 1960s, scientists tried to implement phase contrast systems in TEMs . Key component of such a system is a small and thin plate, the so-called phase plate. Problems with the stability of these phase plates limited the broad acceptance of the method. Only within the last five years a couple of research groups significantly advanced in delivering stable plates and first TEM experiments with biological samples could be performed [6;7]. Nevertheless, even those systems are not technically mature enough to be used for routine work. Figure 3a shows the beam path of a TEM around the objective lens. The scattered (red) and unscattered (blue) electrons generate the magnified image in the image plane. In the back focal plane, scattered and unscattered electrons are separated. In this plane, the phase plate is positioned.
Figure 3: a. Beam path in a TEM around the objective lens. A phase plate is positioned in the back focal plane of the objective lens. b. Section drawing of a TEM. The arrows show the positions of specimen, objective and phase plate.
Phase contrast devices are based on two basic concepts: Zernike and Boersch phase contrast systems (figure 4). A Zernike phase plate consists of a thin perforated film. The scattered electrons pass the film. Their phase is shifted. The unscattered electrons pass through the hole. Their phase is not affected. The degree of phase shift is determined by the phase plate’s inner potential, i.e. its thickness and material. Usually, thin amorphous carbon films are used for Zernike phase plates. To achieve a 90 degree phase shift, the thickness of the carbon film needs to be 27nm (200kV TEM) and 35nm (300kV TEM). These carbon film phase plates are not very beam-stable. Their thickness is altered through considerable contamination. The high beam intensity of the unscattered electrons damages the edge of the hole. This limits the lifetime of such a phase plate to about 30 minutes. Boersch phase contrast systems are electrostatic einzel lenses. A ring-shaped lens is placed around the unscattered electron beam. A potential is applied to the lens and the phase is shifted accordingly. The degree of phase shift can be adjusted by altering the voltage. This is an advantage over Zernike phase plates that have a pre-determined phase shift through the film’s material and thickness. On the flipside, the ring lens and its support structures of the Boersch phase plate cut out part of the image information. With state of the art technologies, support structures of about 1µm width can be realized, still big enough to make sample details between 1 and 5nm invisible.
Figure 4: Different phase contrast concepts for TEM. The main technical concepts are Zernike and Boersch phase contrast systems. The KonTEM phase plate combines both concepts.
The KonTEM phase plate links the Zernike with the Boersch concept. It consists of a thin film with a perforation for the unscattered electron beam. The film is thinner than the usual Zernike carbon film and has an increased transmissivity (approx. +15%). This is especially beneficial for cryo-TEM. In addition, the lifetime of the film could be improved. As the inner potential of the KonTEM film is not sufficient to achieve a 90 degree phase shift, a potential is applied to the phase plate to compensate the difference. With this, the phase shift is adjustable within a certain range.
Figure 5 shows a first prototype of the KonTEM phase plate. Each phase plate has a hole milled with a focused ion beam through which the unscattered electrons pass.
Figure 5: KonTEM phase plate chip. Each phase plate contains a hole milled with a focused ion beam microscope through which the unscattered electrons pass.
We have developed a dedicated device for the installation and positioning of the phase plates in the back focal plane of the objective. This includes equipment to apply a voltage to the phase plate in order to adjust the phase shift (figure 6). We could prove the technical concept through experiments in which we used diluted vitrified liposome suspensions as samples. The liposome had an average size of 100nm.
Figure 6: Prototype of a mechanical phase plate holder. The positioning is realized with micrometer screws. It can hold three different phase plates or apertures. Two different voltages can be applied.
Figure 7 shows two images with and without phase plate, all other conditions being equal. The use of a phase plate leads to a significantly better contrast. In the course of our experiments we found that a quick, exact positioning of the phase plate is essential for successful experiments. Therefore, we are currently developing an electric driven system that allows for more exact and quicker positioning of the phase plate in the electron beam.
Figure 7: Comparison between standard bright-field contrast and phase contrast. a. Ice-embedded liposome as bright-field image near focus. b. The same sample in focus with phase plate.
It could be proven that the KonTEM phase plate improves contrast of phase objects. Its stability was significantly higher than that of comparable carbon films. The project is supported by an EXIST Forschungstransfer grant from the Federal Ministry of Economics and Technology, BMWi. Our target is to develop the KonTEM phase plate into a modular and user-friendly system to be installed in new and already installed TEMs.
 Kisielowski, C. et al. (2008) "Detection of Single Atoms and Buried Defects in Three Dimensions by Aberration-Corrected Electron Microscope with 0.5 Å Information Limit" Microsc. Microanal. 14, 469-477
 Baumeister, W. (2002) “Electron Tomography: Towards Visualizing the Molecular Organization of the Cytoplasm” Current Opinion in Structural Biology 12, 679-684
 Danev, R. and Nagayama, K. (2001) “Complex Observation in Electron Microscopy. II. Direct Visualization of Phases and Amplitudes of Exit Wave Functions” Journal of the Physical Society of Japan 70, 696-702
 Zernike, F. (1935) „Das Phasenkontrastverfahren bei der mikroskopischen Beobachtung”Z. techn. Physik 16, 454-457
 Nagayama, K. and Danev, R. (2008) “Phase contrast electron microscopy: development of thinfilm phase plates and biological applications” Philos. Trans. R. Soc. Lond. B Biol. Sci. 363, 2153-2162
 Barton, B., Joos, F., and Schröder, R.R. (2008) “Improved specimen reconstruction by Hilbert phase contrast tomography” J. Struct. Biol. 164, 210-220
 Danev, R. and Nagayama, K. (2008) “Single particle analysis based on Zernike phase contrast transmission electron microscopy” J. Struct. Biol. 161, 211-218