22
Langmuir 2008, 24, 22-25
Cryo-Fracture TEM: Direct Imaging of a Random Mesh Phase Sanja Bulut,* Alfredo Gonza´lez-Pe´rez, and Ulf Olsson Physical Chemistry 1, Center for Chemistry and Chemical Engineering, Lund UniVersity, P.O. Box 124, Lund SE-22100, Sweden ReceiVed NoVember 5, 2007. In Final Form: NoVember 22, 2007 A novel and a simple method that allows direct imaging of viscous samples by cryo-TEM (cryo-transmission electron microscopy) is presented. A fracture on the vitrified sample is created in a controlled way. In the fracture, some edges are thin enough to allow direct imaging in transmission mode. The method was used to directly image a nonionic surfactant lamellar phase where a random mesh structure is formed at lower temperatures (98%, was purchased from Nikkol Chemicals Co. Ltd., Tokyo, and used as received. C10E3 samples (40 wt %) were prepared by gently mixing the appropriate amount of the surfactant and Millipore filtered water. Each sample was then divided in two and placed into two sample tubes. One sample tube was stored at 25 °C, and the other, at 5 °C. They were left to equilibrate for 24 h before any cryo-TEM experiment was done. Sample preparation and imaging were performed on five occasions in order to check the reproducibility of the method. The samples were kept in a thermostated bath at a given temperature until the moment they were transferred to the carbon grid. A precooled pipet tip was used for transfer in the case of the sample stored at 5 °C. Specimen Preparation. The preparation of specimens largely followed a standard preparation procedure,3,12,13 using a controlled environment vitrification system (CEVS)14 thermostated to 5 or 25 °C. The temperature control in CEVS is very good, and the only (5) Wennerstro¨m, H.; Evans, D. F. Colloidal Domain: Where Physics, Chemistry, Biology and Technonlogy Meet, 2nd ed.; Wiley-VCH: New York, 1999. (6) Ostro, M. J. Liposomes: From Biophysics to Therapeutics; Marcel Dekker: New York, 1987. (7) Tschierske, C. Curr. Opin. Colloid Interface Sci. 2002, 7, 69-80. (8) Ghosh, S. K.; Ganapathy, R.; Krishnaswamy, R.; Bellare, J.; Raghunathan, V. A.; Sood, A. K. Langmuir 2007, 23, 3606-3614. (9) Baciu, M.; Olsson, U.; Leaver, M. S.; Holmes, M. C. J. Phys. Chem. B 2006, 110, 8184-8187. (10) Medronho, B.; Miguel, M. G.; Olsson, U. Langmuir 2007, 23, 52705274. (11) Baciu, M.; Holmes, M. C.; Leaver, M. S. J. Phys. Chem. B 2007, 111, 909-917. (12) Danino, D.; Bernheim-Groswasser, A.; Talmon, Y. Colloids Surf., A 2001, 183, 113-122. (13) Talmon, Y. Ber. Bunsen-Ges. Phys. Chem. 1996, 100, 364-372.
10.1021/la703443c CCC: $40.75 © 2008 American Chemical Society Published on Web 12/08/2007
Letters
Langmuir, Vol. 24, No. 1, 2008 23
Figure 1. (a) A drop of sample was deposited on the carbon grid, forming a thick layer. (b) The sample was vitrified on liquid ethane (as in conventional cryo-TEM experiments). (c) The carbon grid with the vitrified thick layer was punched with a needle, creating fractures on the thick layer. Finally, the sample was transferred to the electron microscope, keeping T always less than -180 °C where images of the sample could be obtained at the edges of the fractured regions. critical step in the temperature control is the transfer of the sample from the water bath to the CEVS with a pipet. For a 5 °C sample, this step was done using a precooled pipet. A small drop of the viscous sample was placed on the TEM grid covered with a holey carbon film. The drop was then blotted gently to remove excess sample and create as thin a layer of the sample as possible. The grid containing the sample was immediately plunged into liquid ethane at its freezing point (-183 °C) and then stored in liquid nitrogen (-196 °C). Fracturing of the Sample Film. To create fractures in a sample, the following procedure was used. A needle was first cooled in liquid nitrogen. The surface of the vitrified sample was then manually punched with the needle tip, still in liquid nitrogen. In this way, fractures on the order of a few hundred micrometers were created as presented in Figure 1. A typical edge of a fracture contains both thick and thin layers of a sample, the thin ones being possible to image with a cryo-TEM. Sample Transfer. Drying of the wet sample and heating of the vitrified sample were prevented by keeping the sample below -180 °C during each step and continuously, as in the classical cryoTEM protocol for specimen preparation. The vitrified sample was transferred to a liquid-nitrogen-cooled TEM holder using a cryotransfer stage (Oxford Instruments CT-3500) designed for minimal air exposure and heat loss. The setup was a Philips CM120 cryo electron microscope operated at 120 kV. Images were recorded with a GIF 100 (Gatan imaging filter), using a CCD camera in low-dose mode, giving a low electron beam intensity of ca. 10 e-/Å2.15 Imaging can be problematic if the carbon layer is also fractured because the support of the sample is then lost. A large part of a (14) Bellare, J. R.; Davis, H. T.; Scriven, L. E.; Talmon, Y. J. Electron Microsc. Tech. 1988, 10, 87-111. (15) Danino, D.; Gupta, R.; Satyavolu, J.; Talmon, Y. J. Colloid Interface Sci. 2002, 249, 180-186.
Figure 2. Lamellar phase formed by C10E3 in water imaged at 25 °C. The black arrows on the images show the edges of individual bilayers. sample can then start moving and even tear off from the rest during imaging. Great care should thus be taken to choose a suitable area of the sample that has good support from the carbon grid. The fracturing should also be performed as gently as possible.
3. Results and Discussion The cryo-fracture method was applied to a lamellar phase of the nonionic surfactant of 40 wt % C10E3 in water. At 25 °C, the system forms a regular lamellar phase16 without perforations, and upon decreasing the temperature to 5 °C, a random mesh phase is obtained.10 Repeated experiments were performed at both temperatures. When a fracture was made on a vitrified sample, the stacked bilayers were fractured, creating edges with a lower thickness that could be imaged. Mainly bilayers parallel to the carbon grid were observed. This could be a consequence of the thin film and orientation from the interface but may also (16) Ali, A. A.; Mulley, B. A. J. Pharm. Pharmacol. 1978, 30, 205-213.
24 Langmuir, Vol. 24, No. 1, 2008
Letters
Figure 3. Cryo-TEM images of 40 wt % C10E3 at 5 °C. The white arrows show the presence of perforations on the mesh phase. The black arrows show different bilayers on the edges.
Figure 4. Test of the beam sensitivity effect on the lamellar phase at 25 °C for 40 wt % C10E3. Images were captured in low-dose mode (ca. 10 e-/Å2). Upper image, 1 s exposure time; lower image, 5 s exposure time.
be due to the shear forces induced by the blotting procedure. Occasionally, perpendicular orientations were observed, showing lamellar stacking. Lamellar Phase. The protocol was first applied on the regular lamellar structure, vitrified from 25 °C. Typical images are shown in Figure 2. The stacked bilayer structure can be observed at the edges with no perforations visible. Mesh Phase. Images of experiments performed at 5 °C are shown in Figure 3. In this sample, we clearly see perforations in the bilayer. The perforations appear to be randomly distributed. They sometimes reach through several bilayers, whereas in other cases the perforated domains are observed in only one bilayer. The diameters of the perforations are on the order of few hundred nanometers. Both the lamellar and mesh phases have been imaged in several experiments in our laboratory, always showing the same behavior on the fractures, hence validating the reproducibility of the
presented protocol. Because both samples (25 and 5 °C) were prepared in exactly the same way, perforations cannot result from punching the surface of the vitrified sample. Beam Sensitivity. Samples imaged in cryo-TEM experiments can be damaged by the radiation.13 Therefore, it is very important to verify that the random mesh structure obtained at 5 °C is not an artifact of radiation damage. All of the images were captured using a low-dose-mode setup (ca. 10 e-/Å2) in order to minimize the risk of beam damage. In Figure 4, we show the test of two different beam exposure times operating in lowdose mode. The upper image was obtained with a short exposure time (1 s) whereas the lower image was obtained with a longer beam exposure time (5 s). The 5 s sample is clearly radiation-damaged (lower image). Note that the small holes resulting from the radiation damage are very different in shape and size compared to the perforations presented in the random mesh phase in Figure
Letters
3. All of the images used for further evaluation were thus obtained with exposure times of e1 s. To summarize, a simple method to directly image viscous samples by cryo-TEM has been presented. The method was successfully applied to study a classical lamellar structure and a random mesh phase with perforated bilayers. The mesh structure was imaged here for the first time. This new simple protocol appears to be useful for imaging various viscous systems such as liquid crystals or gels using standard cryo-TEM equipment. Its range of applicability is currently under investigation.
Langmuir, Vol. 24, No. 1, 2008 25
Acknowledgment. The authors gratefully acknowledge Gunnel Karlsson for the invaluable help with cryo-TEM imaging experiments and fruitful discussions of the implementation of the present protocol. A.G.-P. is grateful to the EU Research Training Network, CIPSNAC (contract no. MRTN-CT-2003504932). This work was also financed by the Swedish Research Council (VR) and by Stiftelsen for Strategisk Forskning (SSF) through the Colloid and Interface Technology (COLINTECH) project. LA703443C
