© 2003 The Royal Microscopical Society Journal of Microscopy, Vol. 211, Pt 1 July 2003, pp. 48–53 Received 7 March 2002; accepted 14 March 2003 Blackwell Publishing Ltd. Vitrification of cryoelectron microscopy specimens revealed by high-speed photographic imaging S. KASAS*§, G. DUMAS†, G. DIETLER‡, S. CATSICAS§ & M. ADRIAN¶ *Institut de Biologie Cellulaire et de Morphologie, rue du Bugnon 9, CH-1005 Lausanne, Switzerland †Sociéte Vaudoise d’Astronomie, Ch. des Grandes Roches 8, CH-1004 Lausanne, Switzerland ‡Institut de Physique de la Matière Condensée, and ¶Laboratoire d’Analyse Ultrastructurale, Université de Lausanne, CH-1015 Lausanne, Switzerland §Laboratoire de Neurobiologie Cellulaire, Ecole Polytechnique Fédérale de Lausanne, CH-1015 Lausanne, Switzerland Received 7 March 2002; accepted 14 March 2003 Key words. Cryoelectron microscopy, evaporation, liquid ethane, specimen preparation, vitrification. Summary Cryoelectron microsopy is a widely used technique to observe biological material in an almost physiological, fully hydrated state. The sample is prepared for electron microsopy observation by quickly reducing its temperature to -180 °C. The high- speed cooling induces the formation of vitreous water, which preserves the sample conformation. However, the way vitrification occurs is still poorly understood. In order to better understand the phenomenon, we have used a stroboscopic device to visualize the interaction between the electron microscopy grid and the cryogen. By blocking the free fall of the plunger once the grid has penetrated the coolant by half its diameter, we have elucidated the way in which vitrification propagates. The findings were confirmed by numerical simu- lation. In addition, according to our observations, we now present an alternative way to prepare vitreous specimens. This new method, with the grid parallel to the liquid cryogen surface, decreases evaporation from the sample during its free fall towards the coolant and at the same time achieves a more uniform vitrification over the entire surface of the specimen. Introduction Cryoelectron microscopy was developed in the early 1980s (Adrian et al., 1984; Dubochet et al., 1988). It is widely used today for observing biological materials with high resolution in an almost physiological environment (Harris, 1997; Harris & Adrian, 1999). The main advantage of this method comes from the fact that the sample conserves its native fully hydrated physiological state during electron microscopic observation. To achieve this, the sample is deposited onto a grid and cooled at such a high speed and to such a low temper- ature that the water molecules contained in the sample pass from a liquid to a vitreous state without crystallizing. The method has been successfully applied to observe many kinds of fully hydrated materials that are unfixed and unstained. The vitrification is obtained by rapidly immersing the sample into a liquid cryogen, such as propane or ethane. The cooling speed and the evaporation rate of the liquid before the grid touches the cryogen are believed to be the most sensitive parameters determining the quality of the specimen pre- servation (Cyrklaff et al., 1990; Battesby et al. 1994; Trinick & Cooper, 1990). In order to optimize these parameters, several technical solutions have been proposed (Ballare et al., 1988, 1998; Trachtenberg, 1993; Talmon, 1996; Egelhaaf et al., 2000). However, the exact influence of these parameters is still poorly known. In order to better understand the way in which the sample vitrifies, we have used a stroboscopic system to observe in greater detail the penetration of the grid into the coolant. An apparatus was designed to take pictures of the falling grid using a light flash of a few microseconds duration. Images of the grids touching the surface of the cryogen produced by this apparatus reveal the way the cryogen behaves during immersion. According to these results, we propose an expla- nation of the way vitrification progresses through the specimen and we suggest a new procedure to achieve a more efficient and homogenous vitrification of the sample. Correspondence: Marc Adrian. Tel.: +41 21 692 42 81; e-mail: Marc.Adrian@lau.unil.ch