Scanning Electron Microscope (SEM)

Introduction.

The key to producing the best SEM image of a specimen is a full understanding of the effects of the machine variables. Some important variables are; condenser lens current, stigmatism, working distance and accelerating voltage. You can make adjustments by changing these variables and as you will see it often will be a trade-off between multiple variables.

Image Formation using the SEM.

When the electron beam strikes a sample, these electrons will scatter through the sample within a defined area called the interaction volume. During the electron beam-specimen interactions, secondary products like secondary electrons, backscattered electrons, X-rays, heat and light will be formed. Image formation of surface structures using the SEM mainly depends on the production of secondary electrons. Secondary electrons are low energy electrons and when produced deeper within the interaction volume, will be absorbed by the sample. Only secondary electrons close to the surface will be able to escape the specimen (figure 1). The weakly negative secondary electrons will be deflected by a positive pull exerted by the Faraday cage surrounding the secondary electron detector and therefore will contribute to the image formation. Backscattered electrons are also produced deep within the sample but have a much higher energy and because of this, are able to escape from deeper within the interaction volume. Because of their high energy, backscattered electrons will not be deflected by the Faraday Cage and therefore not contribute to the image formation. Only a few backscattered electrons will interfere with the signal for secondary electrons (figure 1).

Figure 1. Interaction volume between the electron beam and the specimen. Notice that the backscattered electrons are able to escape from far deeper in the sample.

Different numbers of secondary electrons produced at different areas of the sample will provide image contrast. If at a certain spot on the sample more secondary electrons are produced, a bright spot will appear on the image. Often many secondary electrons are produced along raised areas of the sample, therefore many electrons will be detected, producing a bright spot on the image. This effect is called the edge effect (figure 2). Some secondary electrons are produced in a valley of the sample and are difficult to deflect by the Faraday Cage. In this case only few electrons will be detected producing a dark spot on the image.

Figure 2. Secondary electrons escape from sample areas close to the surface. The edge effect takes place when many electrons escape from areas with sharp edges or peaks to be deflected by the Faraday Cage (B). Some electrons escape from areas without sharp edges (A and C). Few secondary electrons are detected when generated secondary electrons are only deflected partially by the Faraday Cage due to the location on the sample they are generated (A).

Condenser Lens Current and Resolution.

The current in the condenser lens changes the spot size or diameter of the beam of electrons that scans the sample. More detailed information will be collected when the electron beam scans the same area with a smaller spot size (figure 3). An increased current or a higher number for the condenser lens (CL) setting, will produce a smaller spot size and in general will result in a better resolution (figure 4).

Figure 3. A smaller spot size or smaller diameter of the beam of electrons will resolve more detailed structures of your sample as compared to a beam with a big spot size.

Figure 4. Diatom sample imaged using different CL settings. A small spot size or a higher condenser lens setting (CL = 14) will result in an image with more detail or better resolution (A). A bigger spot size or lower CL setting will resolve less detail (B). Bar is 1µm, Magnification = x 4500, Acceleration Voltage = 5kV, Working Distance = 15mm (A, B).

Condenser Lens Current and Beam Current.

The disadvantage of obtaining a better resolution using a higher condenser lens current is the effect on the beam current or number of electrons available to interact with the specimen. An increase in condenser lens current results in a lower beam current (figure 5). The effect of a lower beam current is that fewer electrons interact with the sample and therefore fewer secondary and back scattered electrons become available to form an image. Electronically this signal can be increased to compensate for the low beam current but this will also result in an increase in electronic noise. Because of the increased noise ('snowy' image on the viewing screen) it will be more difficult to focus the specimen and the final image will start to look grainy as well. Recall that the condenser lens current also effects the spot size (see 'Condenser Lens Current and Resolution'). A correct balance between spot size and beam current needs to be found.

Figure 5. If the current of the condenser lens increases, the focal length of the crossover produced will become shorter resulting in a smaller spot size, increasing the resolution of the final image. A shorter focal length will cause the beam to diverge so that many electrons are not able to enter the final lens resulting in a decreasing beam current as well.

Stigmatism and Resolution.

The stigmator knobs control distortions in the roundness of the spot that is scanned over the sample (stigme means 'mark' or 'spot' in Latin). If the electron beam is not round but elliptical, extraneous information obtained by using this elliptical spot will be collected by the detector and rendered in a round spot on the viewing screen and/or your final image. This phenomenon, called astigmatism, causes a reduction in resolution (figure 6, A-C). When alternating between over and under focus, astigmatism can be recognized by a smeared look in two different directions (figure 6, D and E).

Figure 6. Copy paper imaged to show the effect of stigmatism on resolution. A good image of the copy paper can be obtained when the specimen is in focus with no stigmatism present (A). When stigmatism increases, the resolution will become worse. Little stigmatism present (B). More stigmatism present (C). When stigmatism is present and when the sample is scanned in over or under focus, a smeared image in one or the other direction will show (D and E). Bar is 1µm, Magnification = x 20,000, Acceleration Voltage = 5kV, Working Distance = 8mm, Condenser Lens setting = 15 (A - E).

Working Distance and Resolution.

The working distance is the distance between the final condenser lens and the specimen. Changing the working distance will have an effect on the spherical aberration of the imaging system and therefore will effect the resolution of the final image. Recall that spherical aberration is due to the geometry of electromagnetic lenses. Electrons passing along the axis of the electron beam refract less than electrons passing through the periphery of the electron beam creating more than one focal point and therefore resulting in an enlarged, unsharp spot (figure 7). When the working distance decreases, this effect of spherical aberration will become less, the spot striking the specimen will become smaller and therefore will improve resolution (figure 8).

Figure 7. Spherical aberration is the failure of the lens system to image central and peripheral electrons at the same focal point. If spherical aberration is present the beam will scan with an enlarged, unsharp spot (A). If spherical aberration is absent, the beam will scan with the sharpest spot possible (B).

Figure 8. A diatom imaged using different working distances. At a longer working distance (WD = 48mm) spherical aberration is present decreasing resolution (A). At a shorter working distance (WD = 8) the effect of spherical aberration is less resulting in an image with improved resolution (B). Bar is 5µm, Magnification = x 3300, Acceleration Voltage = 5kV, Condenser Lens setting = 14 (A and B).

Working Distance and Depth of Field.

The working distance has an effect on the depth in the sample that appears to be in focus or in other words, has an effect on the depth of field. At a short working distance the sample will be scanned with a wide cone of electrons resulting in an image with little depth of field. At a longer working distance the sample will be scanned with a narrow cone of electrons resulting in an image with an increased depth of field (figure 9). Recall that a longer working distance does not give you the optimal resolution (see 'Working Distance and Resolution'). If a sample with large topographical variation needs to be scanned it may be important to use a longer working distance to bring as much of the image into focus as possible; however, some of the resolution will be lost (figure 10). If a relative flat sample is scanned it is possible to benefit from a better resolution using a shorter working distance since depth of field becomes less important (figure 8). Depending on the sample and the contained features, a correct balance between working distance and depth of field needs to be found.

Figure 9. The shape of the cone of electrons (wide versus narrow) depends partly on the working distance and is important in determining the depth of field of a sample. At a short working distance little of your sample appears to be in focus (A). At a long working distance, more of your sample appears to be in focus because of a greater depth of field (B).

Figure 10. A head of a fruit fly imaged using different working distances. At a short working distance (WD = 8) only part of the head is in focus because of a limited depth of field (A). At a longer working distance (WD = 48) more is in focus due to an increase in depth of field (B). Bar is 100µm, Magnification = x 130, Acceleration Voltage = 5kV, Condenser Lens setting = 12 (A and B).

Accelerating Voltage and Resolution.

The acceleration voltage is the high voltage applied to the filament. Together with the application of a small current, it will cause the electrons to leave the filament. The size of accelerating voltages often used in SEM imaging vary between 5kV and 20kV. In general, increasing the accelerating voltage will decrease the spherical aberration of the system and therefore increase the resolution. But, varying the acceleration voltage will also have an effect on the beam-specimen interaction. If a higher accelerating voltage is used, the interaction volume between the beam and the specimen will become bigger because of the greater energy of the beam of electrons (figure 11). This effect is more significant in specimens with low atomic number, such as most biological samples (figure 12).

Figure 11. Effect of the acceleration voltage on the interaction volume. Notice that with an increase in acceleration voltage the interaction volume becomes bigger.

Figure 12. Effect of the acceleration voltage on samples with different atomic number. Notice that the interaction volume of samples with a high atomic number is much smaller as compared to samples with a low atomic number.

Resolution is dependent on the area from which secondary electrons are produced. Normally this area would be defined by the spot size. By increasing the accelerating voltage, backscattered electrons, emitted from a larger area of the sample, interact with the sample on their way out, producing secondary electrons further away from the original spot size (figure 13), thereby reducing the resolution of the image (figure 14). This effect will be much less in a sample with high atomic numbers. In this case, you can benefit from the reduction in spherical aberration when increasing the acceleration voltage because it does not create a large interaction volume and therefore increases the resolution of the final image. In addition, a higher accelerating voltage will increase secondary yield from all parts of the specimen due to a greater beam penetration. This will reduce the edge effect and therefore has a diminishing effect on the contrast (figure 14).

Figure 13. When the interaction volume increases, more backscattered electrons will be able to escape from a bigger volume. These backscattered electrons may generate secondary electrons near the sample surface on their way out, increasing the area from which secondary electrons are produced and therefore reducing the resolution of the final image.

Figure 14. A diatom imaged using different accelerating voltages. Fine detail of a diatom imaged at a low accelerating voltage of 5kV is visible (A). A decrease in resolution and contrast can be observed when a diatom is imaged using a much higher accelerating voltage (20kV) (B). Bar is 1µm, Magnification = x 4000, Working Distance = 8mm, Condenser Lens setting = 15 (A and B).

Charge accumulation.

Specimens need to be conductive when examined using a SEM. When a specimen is nonconductive a negative charge from the electron beam tends to accumulate, thus effecting the final image. Some of the effects are abnormal contrast, appearing of abnormal lines (figure 15B) or shifts/breaks within the image. Biological samples are mainly nonconductive and need to be coated with a thin layer of metal making the surface conductive (figure 15A). Specimens like metals and most semiconductors are conductive and can be examined without coating.

Figure 15. Pollen samples showing the effect of charge accumulation. No effect of charge accumulation is visible (A). Pollen showing the type of charging that produces abnormal lines within the image (B). Bar is 5µm, Magnification = x 2300, Acceleration Voltage = 10kV, Working Distance = 39mm, Condenser Lens setting = 15 (A and B).