Backscattered electron imaging explained

By Karl Kersten - Oct 4, 2018

 Backscattered electrons (BSEs) are generated by elastic scattering events. When the electrons in the primary beam travel close to the atom’s nuclei in the specimen, their trajectory is deviated due to the force they feel with the positive charges in the nuclei. Depending on the size of the atom nuclei, the number of backscattered electrons differs. This is the basic principle of BSE image contrast. In this blog we introduce the backscattering coefficient and explain how it is influenced by the inclination of the sample and the primary beam energy.

The backscattering coefficient 

Backscattered electrons are generated by elastic scattering events of the incident electrons in the primary beam and have energy higher than 50eV, as explained in a previous blog. The number of backscattered electrons generated depends on many factors, including the atomic number of the material in the specimen and the acceleration voltage of the primary beam.

The number of backscattered electrons generated from the beam-to-sample interaction is described by a coefficient called backscattering coefficient η and is defined as the ratio between the backscattered current IBSE and the probe current IP:

Schermafbeelding 2018-09-25 om 10.32.57

Where EB is the exit energy of the backscattered electrons. The backscattering coefficient is influenced by the beam acceleration voltage and the atomic number Z, as well as the angle between the sample surface and the incident beam.

The effect of the tilting angle of the sample 

The backscattering coefficient depends on the angle between the incident beam and the sample surface according to the following equation:

Schermafbeelding 2018-09-25 om 10.33.44

Figure 1 shows the angular distribution of the backscattering coefficient for normal incidence, 60° and 80° for different elements (Au, Ag, Cu, Al and Be).
 

Schermafbeelding 2018-09-25 om 10.35.05Schermafbeelding 2018-09-25 om 10.35.23Schermafbeelding 2018-09-25 om 10.35.45

Fig 1: Polar diagram of the angular distribution of the backscattering coefficient, at normal incidence and at 60° and 80° for different elements (Au, Ag, Cu, Al and Be)[1]

In the schematics in Figure 2, the backscattering coefficient in polar coordinates is shown for different elements (in other words, the amount of emitted backscattered electrons for different atomic numbers) together with the SEM geometry, consisting of the pole piece, the BSE detector and the sample.

When the sample surface is perpendicular to the beam, the BSE emission is rotationally symmetric, as shown in Figure 2 on the left. When the sample is tilted, the BSE emission is not symmetric anymore, meaning that, in this case, the signal picked up by the right side of the BSD is less than the signal picked up by the left quadrant.

Schermafbeelding 2018-09-25 om 10.36.21

Fig 2: Schematics of the SEM, consisting of a sample, the BSE detector and the pole piece, together with the angular distribution of the backscattering coefficients of Au, Ag, Cu and Be in polar coordinates for a zero-tilted sample and a tilted sample.

When the sample is inclined, the backscattering coefficient increases, as shown in Figure 3. In this graph, the backscattering coefficient is measured with Monte Carlo simulations for different elements (C, Al, Cu, Ag and Au), at different tilt angles. The coefficient for different elements not only increases with increasing inclination, but also converges.


Schermafbeelding 2018-09-25 om 10.37.06

Fig 3: The backscattering coefficient measured with Monte Carlo simulations for different elements (C, Al, Cu, Ag and Au) for different surface inclination, or tilting angles. [2]

To demonstrate this effect, we imaged with the BSE detector the EDX calibration sample, where three areas made of different materials can be seen: cupper, carbon and aluminum. Figures 4a and 4b show the SEM micrographs of this sample taken at 0° tilting angle and at 45° tilting angle. The histogram of the two images is then extrapolated and shown in figure 4c. The histogram has three peaks related to the composition of the sample. The peaks on the left of the graph are that of the carbon, the darker area in the SEM images, while the peaks on the right are that of cupper, that appears brighter in the BSE image. Most importantly, the three peaks become closer when the tilting angle increases, as predicted in figure 3. 

Schermafbeelding 2018-09-25 om 10.38.19

Fig 4: SEM images of the EDX calibration sample at 0° tilting angle (a) and at 45° tilting angle (b) and histogram of the two images (c), where the three peaks (carbon, aluminum and cupper) are visible.

The effect of the primary beam energy 

The backscattering coefficient is also influenced by the acceleration voltage, as shown in Figure 5. The backscattered coefficient depends on the primary beam energy. This means that the amount of BSE emitted for the same beam current at different primary beam energy will vary.

For lower acceleration voltages, the backscatter coefficient of light elements increases, while the backscatter coefficient of heavy elements decreases. For beam energies higher than 10kV the backscatter coefficient is approximately constant for all elements.

However, in real life, the BSE detector also partly collects SE signals as well. So, if the beam energy increases, the edge effects and the topography information given by the SEs also increases. 

Schermafbeelding 2018-09-25 om 10.39.12

Fig. 5: Backscattering coefficient as a function of the electron energy from 1kV to 30kV for different elements. [1]

When the energy of the primary beam increases, the penetration depth also increases. So thin film layers on the sample surface will appear in BSE images only when using a primary beam with low acceleration voltage.

Figures 6 and 7 show BSE images acquired of a tin ball sample and a flat surface with dirt on it using beam acceleration voltages of 5kV and 15kV. The dirt flakes, on both samples, are mainly made of carbon and appear dark in the low acceleration voltage images.

Since the flakes are very shallow, the beam fully penetrates them when a high acceleration voltage is used, reaching the underlying sample that is then revealed in the images. Hence, at 15kV, the shallow, dirty flakes become more transparent.

Schermafbeelding 2018-09-25 om 10.39.43

Fig 6: BSE images of the tin ball sample taken with 5kV and 15kV beam acceleration voltages. 

Schermafbeelding 2018-09-25 om 10.40.10

Fig 7: BSE images of a dirty surface sample taken with 5kV and 15kV beam acceleration voltages.

References 

[1] Scanning Electron Microscopy, Physics of Image Formation and Microanalysis, L. Reimer, Springer edition, 1998.

[2] Scanning Electron Microscopy and X-Ray Microanalysis, J.I. Goldstein, D.E. Newbury, J.R. Michael, N.W.M. Ritchie, J.H.J. Scott, D.C. Joy.

More on SEM 

In this detailed description, we described how to improve the quality of BSE images when using an SEM. Has this made you curious about the technology behind scanning electron microscopy? And do you want to find out how you can select the right microscope for your research? Then it’s good to know that we have a comprehensive e-guide on How to Choose an SEM available for you.

The e-guide details the working principle of an SEM and enables you to gain a deeper understanding of the microscopy technique that produces high-quality images faster and easier. It helps you to choose the SEM that is most suitable for your research. I highly recommend taking a look at the guide; it’s available here for free.

New call-to-action

 


About the author

Karl Kersten is head of the Application team at Thermo Fisher Scientific, the world leader in serving science. He is passionate about the Thermo Fisher Scientific product and likes converting customer requirements into product or feature specifications so customers can achieve their goals.

Press Room | Privacy Policy | Terms of Use | Sitemap