Ways to Reveal More from your Samples: Ultra-Thin Carbon Films

Application Note for Leica EM ACE600 - Material Research

Much of the battle involved in obtaining good transmission electron microscopy data is the specimen preparation itself. Even though some nanomaterials are already electron transparent (e.g. nanoparticles and proteins) and often do not require further thinning procedures, they need to be dispersed onto thin support films. Carbon films provide a continuous electron transparent support enabling the analysis of specimens without excessive interference from underlying support material. The quality and characteristics of such carbon support films have a major influence on the analysis, whether studying nanomaterials or biological nanostructures.

This application note describes the process of preparing ultra-thin carbon films in a reproducible manner. These support films can be used for a wide variety of nanostructures and allow to obtain a complete insight into the specimens being studied.



Carbon Films: How to define 'quality'?

Carbon films are preferred because of their mechanical strength, conductivity and thermostability. The preparation of ultra-thin carbon films in a straightforward manner can be a laborious task owing to the non-reproducible characteristics of conventional carbon deposition methods. The structure and quality of the carbon filmes are governed by the evaporation characteristics (evaporation mode, vacuum level, evaporation rate, work distance and temperature). Before going into detail on these parameters or on the evaporation of high quality amorphous carbon films, it is essential to define the term 'quality' and hence to discuss which criteria the carbon support films should fulfill in order to be suitable for a broad range of electron microscopy applications.

  • High transparency to electrons: the thickness and density of the support film has an important effect on the contrast and resolution of the image. When mass thickness is comparable to that of the specimen the film can attenuate the intensity of structural details in the image.
  • Adequate strength to withstand electron bombardment
  • Uniform thickness: Film thickness is crucial for analytical investigations, quantitative imaging, or electron tompography.
  • Free of any intrinsic structures, surface irregularities and contaminants
  • Conductive, in order to prevent the accumulation of charges
  • Easy to prepare and reporducible

Nevertheless, obtaining such carbon films is only the first step in the process of making a suitable TEM specimen. In a second step, the sample needs to be applied on the carbon support film. Samples are ususally dispersed in water or a solvent. Unfortunately, dispersions in solvents often contain remaining reaction products (e.g. surfactants, capping agents...) which are hard to eliminate and are a source of contamination. Even though the carbon films were clean prior to the application of the sample, the sample itself introduces contaminants which can diffuse across the carbon surface to the area of interest where they locally decompose and polymerize under the electron beam. This carbon build-up results in poor signal strength. Obviously, the carbon support films should fulfill some additional criteria when used for TEM specimen preparation:

  • If the sample is dispersed in water, the support film should be rendered hydrophilic using a glow discharge or mild plasma treatment. Such treatments are essential in order to disperse nanomaterials or biological structures more evenly and may not damage the carbon film.
  • The support film should be mechanically stable because it can undergo excessive handling during a specific protocol.
  • The support film should withstand other post-treatments e.g. high vacuum heating in order to remove contaminants.

Taking the above mentioned criteria into account, it is clear that a trade-off exists between thickness and stability. Recently graphene is used as a support film, however coating a complete holey carbon film with graphene can be challenging and in terms of stability (e.g. 300 kV measurement, increased acquisition times, excessive grid handling...), ultra-thin carbon films are preferred. To obtain a good stability and charge dissiparion, the ultra-thin films will show local deviations from planarity and can break during electron beam irradiation or handling.

Briefly, two steps in the preparation are crucial:

  1. obtaining a 'high quality' thin carbon film on a substrate and
  2. subsequently transferring this film onto a suitable support structure.

Evaporating ultra-thin films on a substrate

Several approaches are used in order to prepare ultra-thin carbon films. Typically a 5 nm layer of amorphus carbon is evaporated onto a polymeric support film (Formvar, Butvar, Collodion, ...) and subsequently only the polymeric film is dissolved. This method, however, introduces wrinkling of the remaining carbon film.

Additionally, not all of the polymer is dissolved causing the film to be not uniform or clean. The most suitable and thinnest films can be prepared by evaporating a carbon film directly onto freshly cleaved mica. This film is released in water and transferred onto a Quantifoil or holey carbon film. The supporting structure is necessary in order to improve the stability and charge dissipation and limit vibrations during imaging.

Amorphous carbon films can be obtained through different carbon evaporation modes. Arc-evaporation of pre-shaped graphitic rods is the most widespread technique to obtain amorphous carbon films. Macroparticles associated with the arc process are often seen in the deposited films and result in increased thickness irregularity. Another disadvantage is the high temperature during evaporation (radiant heat), which can damage sensitive samples, result in wrinkled carbon films but more important in larger average sizes of the carbon clusters in the film. The large pressure rise during evaporation even further contributes to the formation of these clusters (increase in granularity).

As a result a severe surface roughness is observed in films with a thickness of 3 to 5 nm. The large heat also affects the frequency of the oscillating quartz crystal, making accurate and reproducible layer deposition difficult. Moreover, the high energetic evaporated carbon species impact the sample, causing a strong adherence which makes releasing the carbon film from the mica substrate difficult. Obviously carbon rod evaporation is not the most optimal method to produce smooth ultrathin 3 nm carbon films in a reproducible manner.

Adaptive carbon thread evaporation: flexible, accurate and reproducible

Adaptive carbon thread evaporation enables a precise and flexible carbon coating of which its applicability far exceeds that of the ultrathin carbon films shown in this application note. Its unparalleled flexibility originates from its unique adaptive process and multi-segment configuration. However, not only the evaporation mode but also the underlying electronics are crucial, especially when evaporating the thinnest films where uniformity and accuracy are of the outmost importance. The ultra-thin carbon films are obtained through multiple evaporation. The multi-segment configuration allows to evaporate carbon from 4 carbon thread segments. This enables the use of thin carbon threads and results in the highest accuracy  as all layers are build up sequentially through adaptive pulsing.

Between each pulse, the thickness of the carbon film and properties of the remaining carbon thread segment are measured in order to adapt the parameters for the next pulse. The high vacuum (5x10-6 mbar) and stable vacuum during evaporation allow the deposition of uniform ultra-thin carbon layers. The deposited films are free of any intrinsic structures, surface irregularities and contaminants. Moreover, radiant heat to the sample is decreased considerably. In the image shown above, consecutive layers were deposited on a filter paper (from left to right: 1, 2, 3, 4 and 5 nm). Carbon thread evaporation is also uniform over a large area (see images below). A round filter paper of approximately 10 cm is shown before (A) and after (B) carbon coating.

Even more important is the combination of flexibility and reproducibility. Properties of amorphous carbon films are strongly dependent on the evaporation characteristics as e.g. evaporation rate, vacuum, mode used etc. Once and evaporation protocol is optimized and stored, reproducible results can be expected coating after coating, without surprises. This reproducibility is crucial, considering the fact that carbon coating is always a part of an elaborate specimen preparation workflow.

Ultra-thin carbon films: procedure

  1. A clean glass petri dish is filled with ultra-pure water.
  2. A filter paper is placed in the bottom of the petri dish and Quantifoils or holey carbon films are carefully placed on the filter paper.
  3. Using the adaptive carbon thread evaporation mode, a 3 nm layer of carbon is deposited onto freshly cleaved mica at a vacuum level of 5x10-6 mbar. Next, the carbon coated mica foil is lowered slowly into the water at an angle of 30 degrees. Water will penetrate the space between the carbon film and mica surface through capillary forces and the film is released.
  4. The filter paper is slowly withdrawn out of the water assuring that the square carbon film is deposited onto the grids.
  5. The filter paper is transferred onto a new filter paper in order to drain most of the water before drying it on a hotplate of 40 degrees for 10 minutes.
  6. the films are ready to be used.

In contrast to the supporting Quantifoil film (see image F) the ultrathin films covering the holes are impeccably clean and uniform. The best procedure to prepare a TEM specimen by applying a dispersion of any sample is to hold the grid with a tweezer and apply a small drop of approximately 5 µl. Next, the excess is blotted away by toughing the edge of the grid on a filter paper. If required the ultra-thin carbon film can be glow discharged or plasma cleaned (low powder and only 5% oxygen for approximately 30 seconds). Below the difference can be seen between a conventional carbon film (left, 15 nm carbon) and an ultra-thin carbon film (right, 3 nm). The lattice of the CdSe quantum plates can be easily observed in the right image.

Sample courtesy of: Daniel Vanmaekelbergh, Debye Institute for Nanomaterials Science, University of Utrecht

Download PDF

Interested to know more?

Talk to our experts. We are happy to answer all your questions and concerns.

Contact Us

Do you prefer personal consulting?