Seeing Inside Liquids in 3D
Electron Tomography on Liquid Suspensions
- Fig. 1: Tomographic stage allowing the acquisition of image tilt series in an ESEM. The sample (d) is mounted on a moving stage (a, b for rotation and translation, respectively) and an annular detector (e) is placed below the sample. Peltier elements (c) are be used to cool down the sample.
- Fig. 2: images of latex particles, supported on a holey carbon membrane, at tilt angle 0°. a) in the dry state. b) in suspension in a water film of thickness around 100 nm. Images taken from .
- Fig. 3: images acquired at different tilt angles (from -60; -30; 0; +30; +60°) and 3D arrangement of the latex particles obtained after volume reconstruction and identification of the spheres. Supplementary information: movie corresponding to the 3D arrangement shown in figure 3.
Water is present everywhere in our everyday life. This paper shows that it is possible to perform 3D observations directly on liquids, with a resolution of a few tens of nanometer. This opens the route to following the materials architecture in 3D during its synthesis or shaping process. The results obtained on water containing latex particles and surfactant molecules, a material used in the construction industry for thermal insulation applications are presented.
Water is essential to life and is the main component of biological cells. In materials science, several types of materials contain or are suspended in water. Water is indeed an environmental-friendly dispersant, as it helps shaping or processing materials into products and can easily be removed at the end. For instance, water-based paints have largely replaced solvent-based paints, which emit volatile organic compounds that can cause health issues. In all cases, analyzing the tiny objects directly in the liquid is crucial to fully understand and design their properties. Electron microscopy turns out to be a key technique for microstructural analysis at the nano- and meso-scales. Yet, conventional electron microscopes operate under vacuum. Specific preparation techniques have been developed to overcome the vacuum limitation. It is possible to freeze extremely rapidly hydrated samples and observe them at low temperatures, or to dry and embed them in a resin. A recent technique consists in encapsulating the sample into a sealed cell formed by two electron-transparent membranes . This development is a real breakthrough, but the processes involved during drying are still unrevealed. Environmental electron microscopes overcome this bottleneck: in these very specific microscopes, the sample is kept under a partial pressure of gas , and can be imaged during dehydration. Here, it is shown that it is even possible to characterize in three dimensions (3D) an aqueous suspension in its native state using an environmental scanning electron microscope (ESEM) operating in transmission.
Materials and Methods
Scanning electron microscopes are conventionally used to image bulk sample surfaces by scanning the sample surface with a focused electron beam.
The image is formed point-by-point and the grey levels in the image either depend on the sample surface topography or chemical composition. In the set-up used for electron tomography (fig. 1), a detector is placed below the sample and collects incident electrons that have been scattered by the sample. As the collected electrons pass through the sample, it is possible to see inside the sample by transparency. The sample stage is also modified: it enables tilting the sample and for each angle, an image is recorded with the detector. In one image pixel, the grey level depends on the sample mass and thickness: a thicker/denser region appears brighter.
To keep the sample liquid in the ESEM microscope chamber, where the partial pressure of water is set to about 1/100th of the ambient pressure, it is necessary to stay on the dew curve and cool down the sample to a few Celsius degrees. This is ensured by Peltier elements placed inside the tomographic stage. Small changes in the water pressure during the experiment allow either water condensation or evaporation.
As the liquid thickness is a key parameter for a successful experiment, the experimental conditions for sample temperature and chamber pressure have to be chosen very carefully. Here the procedure described by A. Bogner et al. is used . Briefly, the dilution of the suspension is adjusted prior to the experiment. Then, a droplet is deposited onto a grid covered with a carbon membrane playing the role of retention basins. Purge conditions are chosen so that the sample does not dry when introduced in the microscope chamber. Then, the liquid thickness is tuned until a suitable contrast is obtained.
The images presented here have been acquired on a suspension of latex particles in water. The latex corresponds to a copolymer derived from styrene and metacrylic acid esters. It forms spheres of diameter around 200 nm . Such materials are involved in paints for instance and are studied in our laboratory to design thermally super-insulating panels for building renovation.
A successful experiment depends on two different key points. First, the material studied should be stable over time. On liquids, this means that the right conditions are found to stay at the equilibrium between liquid water and water vapor. Interestingly, the latex spheres can be seen only in the wet state: if the sample dries, the polymer chains diffuse and form a uniform film. Such morphological change cannot be reversed upon rehydration. It is visible in figure 2a, where the interface between two latex particles in contact with each other tends to disappear. Moreover, the grey level along a particle diameter (see yellow line) is relatively constant. This is not the case in the wet state, see figure 2b, where a dark ring can be distinguished. Stability also means that the electron beam used to form images does not induce significant changes in the sample structure. Yet, water can undergo radiolysis (formation of ions, electrons and other free radicals from water molecules) and chemical bonds in polymer chains are also sensitive to electron irradiation. As a consequence, in our protocol, any adjustment of the microscope settings when tilting has been performed out of the region of interest, so as to reduce the electron dose received by the region of interest.
The second key point corresponds to stability during rotation: when tilting the stage, the sample should not move. Actually, small translations of the region of interest can be corrected by aligning the images afterwards, but any movement of internal features will affect the quality of the reconstructed volume. In the case presented here, it couldn’t be seen any significant movement of the latex particles. They are too big to suffer from Brownian motion. They also seem to be immobilized between the holey carbon membrane and the water meniscus. Figures 3a-e shows images at different tilt angles of a water film containing many latex particles. The suspension is so concentrated that the latex particles form quite dense arrangements. The 3D arrangement of the latex particles in one portion of this region of interest is shown in figure 3f (corresponding movie in supporting information). It has been obtained by reconstruction of the volume, identification of the particle center and modeling of the latex particles by spheres. It becomes then possible to measure several morphological parameters in 3D in the liquid state.
As a conclusion, it has been demonstrated that it is possible to study native suspensions containing sub-micrometer sized objects, including polymers, which are known to be electron sensitive and give rise to poor contrasts. It could be shown here the case of latex particles dispersed in water. Because an environmental microscope has been used, where the sample temperature and surrounding water vapor pressure can be finely tuned, it becomes also possible to follow the evolution of 3D morphological parameters upon drying.
Liquids other than water may be introduced, depending on their saturating vapor pressures. This opens the way to investigating in 3D the evolution of materials during in situ reactions.
The authors acknowledge the Consortium Lyon Saint-Etienne de Microscopie (CLYM) for the access to the microscope, the China Scholarship Council (CSC) and Institut Universitaire de France (IUF) for financial support, and BASF for having provided the samples.
 Liquid cell electron microscopy (2016). In F. Ross (Ed.), Liquid Cell Electron Microscopy (Advances in Microscopy and Microanalysis, p.I), Cambridge: Cambridge University Press. doi: 10.1017/9781316337455
 D. Stokes, Principles and Practice of Variable Pressure: Environmental Scanning Electron Microscopy (VP-ESEM), John Wiley & Sons, Ltd (2008) doi: 10.1002/9780470758731
 A. Bogner, G. Thollet, D. Basset, P.H. Jouneau, and C. Gauthier, Wet STEM: a new development in environmental SEM for imaging nano-objets included in a liquid phase, Ultramicrosc. 104: 290-301 (2005) doi: 10.1016/j.ultramic.2005.05.005
 A. Perret, G. Foray, K. Masenelli-Varlot, E. Maire, B. Yrieix, Study of the surfactant role in latex-aerogel systems by scanning transmission electron microscopy on aqueous suspensions, J. Microsc. 269: 3-13 (2018) doi: 10.1111/jmi.12603
 J. Xiao, G. Foray, K. Masenelli-Varlot, Analysis of liquid suspensions using scanning electron microscopy in transmission: estimation of the water film thickness using Monte Carlo simulations, J. Microsc. 269; 151-160 (2018). doi: 10.1111/jmi.12619
Juan Xiao1, Lucian Roiban1, Geneviève Foray1, Karine Masenelli-Varlot1
1 MATEIS, INSA-Lyon, Villeurbanne, France