Introduction
The scanning electron microscope (SEM) is a powerful tool for taking very high magnification images of a variety of materials. It generates an image by scanning a beam of electrons across a sample’s surface and detecting the electrons generated through sample interactions. This type of imaging requires the sample to be conductive. A non-conductive sample will build up a static surface charge when it is imaged with an electron beam. This poses a major problem, many because non-conductive materials–such as plastics, fabrics, and ceramics–have microstructures that can only be resolved clearly in an electron microscope. The most common way to address this problem is by coating the sample with gold, making it conductive. Gold coating, however, comes with a number of downsides. When gold coating is not an option, imaging with a low voltage electron beam can reduce charging, enabling analysis of non-conductive samples.
Charging on a non-conductive sample
Non-conductors are materials that do not allow electrons to easily move through them. These materials include paper, glass, biological tissues, ceramics, and plastics – some of the most common types of samples imaged with SEMs. Unfortunately, they can also be some of the most challenging samples to image. In contrast to conductive samples, like metals, imaging a non-conductive sample using an SEM can cause beam drift and charging, both of which make it difficult to capture a clear image.
Figure 1 shows a charging piece of paper. The (negatively charged) surface reflects many of the incoming (negatively charged) electrons, which creates very bright areas in the image. The electron beam is also deflected by these regions, causing a blurring effect that can be seen at the edges of the image.
Figure 1: A piece of paper charging badly in the SEM.
Charge balance
When an electron from the beam hits a sample, there are a few possible outcomes that affect the total charge on the sample:
- The electron backscatters, and leaves the sample (no change in charge)
- The electron is absorbed by the sample (negative charging)
- The (primary) electron strike causes emission of secondary electrons (positive charging)
The electron beam is comprised of a huge number of electrons (~109 electrons/second), so all three of these processes occur simultaneously. The relative balance of each of the three possibilities determines the net charge on the sample.
For conducting samples that are electrically grounded, any excess electrons can easily pass through to ground, so the net charge does not change. For non-conducting samples, however, there will be net charging on the sample (unless these processes happen to be perfectly balanced). Figure 2 summarizes these processes.
Figure 2: Electron-sample interactions.
Low voltage imaging
One of the major factors that affects the charge balance is the speed at which that the electrons in the beam strike the sample. This speed is controlled by the instrument’s accelerating voltage setting (measured in kilovolts, kV). At high kV, fast-moving electrons penetrate deeply into the sample, and are less likely to escape. As lower accelerating voltages, an increased share of backscatter + secondary electrons escape the sample, so there is less net charging. Every sample has a specific kV (referred to as the E2 point) where the same number of electrons escape the sample as are injected into it, i.e. electrons in = electrons out. This magic kV point depends on the materials composition, and is also affected by its structure (thickness, roughness), so it must be found experimentally. Typical values are 2 – 5 kV. Once found, imaging at this accelerating voltage causes no net charging, so the sample can be imaged successfully.
Figure 3 shows plastic microparticles imaged at 15 kV and 5 kV. At the higher accelerating voltage, the sample charges badly, and the extreme brightness at the top of the particle obscures the rest of the image. By reducing to 5 kV, the charging is nearly eliminated, and the whole particle can be seen with crisp detail.
Figure 3: Plastic microparticles imaged at (left) 15 kV and (right) 5 kV.
As mentioned above, the charge balance point changes for different materials. 5 kV was sufficient to image the plastic microparticles, but the Teflon film shown in Figure 4 still has significant charge-up at the same accelerating voltage. In this case, imaging the sample required the accelerating voltage to be reduced down to 2 kV. The delicate surface structure of the Teflon can only be seen at this extremely low voltage.
Figure 4: Teflon imaged at (left) 5 kV and (right) 2 kV.
Conclusions
Sample charge-up is a common hindrance to acquiring optimal SEM data on non-conductive samples. By reducing the accelerating voltage of the incident electron beam, the charging is reduced, and the underlying structure is revealed. This method is not without its downsides. First, not all instruments can acquire crisp images at low kV. A field emission system (FE-SEM) is usually preferred for low-kV work, because of the inherently high brightness and good resolution achieved from these advanced instruments.
Charging can also be controlled with two other methods. First is gold sputter coating. A sputter coater deposits a very thin layer of conducting material (most commonly gold) over the surface of the sample. Excess electrons in the sample then have a path to ground, so charging is essentially eliminated. The primary disadvantages to gold coating is that it requires a dedicated instrument. The coating also disrupts EDS measurements, since the metal in the coating will show up as the primary element. However, sputter coated samples are easy to image, so this method is often preferred, especially by non-expert users.
The second method to control charging is low-vacuum imaging. A gas (air or water vapor) is introduced into the sample chamber. The molecules in the gas impact the charged sample surface, become ionized (negatively charged) by collecting an electron, and then discharge that electron at a grounded surface inside the sample chamber. Not all SEMs are capable of low-vacuum imaging, though it is common on modern instruments; all SEMs currently sold by Nanoscience Instruments have this capability.
These three different methods of controlling charge-up—gold coating, low vacuum imaging, and low voltage imaging—each have different advantages and disadvantages. The best solution needs to account for the type of sample, the information required, and the capabilities of the instrument. If you need help, please contact our applications experts for a consultation!