The problem here is that although your sample may be conductive, it is not conductive enough to conduct away the number of electrons impinging the surface of the sample for such a long period of time. On the normal scan the beam does not sit on any part of the sample long enough to build up charge. On the slow scan you simply put more electrons into the surface quicker than they can dissipate because the beam does not move quickly enough.

Another way to capture this image would be to use a frame integrate and have the beam scan more quickly, but build up the image over a number of frames (for instance 256 frames for a normal scan speed). This relies on having a non-drifting sample.

Probably your SEM image is drifting during slow scan. You'll have to tune your parameters around (acceleration voltage/current/scan time) to get the best image.

If a slow scan is inevitably necessary, I believe there are some software assisted method to produce a drift-corrected image.

Probably your material is not conducting electrons properly. What is the material?

If it is so, you can make a contact from your material to FESEM stub by conducting carbon tape or, you can gold sputter the material.

I try to sum up the upper quotes by asking: Which material do you try to investigate?

The material is conductive. The normal scan is ok. It is just the slow scan that shows this kind of distortion.

The problem here is that although your sample may be conductive, it is not conductive enough to conduct away the number of electrons impinging the surface of the sample for such a long period of time. On the normal scan the beam does not sit on any part of the sample long enough to build up charge. On the slow scan you simply put more electrons into the surface quicker than they can dissipate because the beam does not move quickly enough.

Another way to capture this image would be to use a frame integrate and have the beam scan more quickly, but build up the image over a number of frames (for instance 256 frames for a normal scan speed). This relies on having a non-drifting sample.

I feel that Martins answer is correct. To confirm things try doing the following:

Cover your sample with thin film of gold, do measurement in slow and fast scan. If the problem is over, then Martins answer is accepted.

Hello Sayed

I want to remind you that the scan you show above is distorted as a whole picture, including the scale and text writing themselves. Any way recheck my answer above

Instead of doing a line scan have you tried doing an integrated scan where the SEM program takes multiple images and averages them to produce 1 image?

Integrated image is fine and depositing a thin conductive film on the samples does not change the situation.

This looks like the standard tin balls sample that service engineers use to check resolution, and they are usually quite conductive, and I think already carbon coated. Perhaps one of your scan boards is playing up in your XL microscope?

Sorry I just thought of something else, if you change the integration time of your slow scan (I assume this is "slow scan 3"), does it change the image at all?

These are tin balls. The distortion decreases with decreasing the time of the slow scan but not removed.

It looks rather like a problem with a scan generator, but to be sure you may try to observe piece of metal (a really conductive specimen).

I doubt very much that carbon coating could be better distributed than gold one. However, two coatings will give better (for conductivity) result than one. I should mention that it makes no sense to cover gold with carbon for SEM – gold is too good generator of secondary electron to be shielded from incoming beam. Additionally gold is not the best material for coating; Au-Pd and Pt produce better results. So, to improve coating for specimens with developed topography we can coat it once, then tilt it and coat for second time; even better – sputter coater with rotating stage.

Well if it happens on different samples, then I in my opinion it probably is one of the scanning boards. It could be overheating but more likely has become faulty. You could try unplugging it and reseating it.

The image was taken on a fei (W) instrument? I doubt it is charging. It more appears to be related to the electronics and I agree with Wade, get a technician to have a look at your scan generator board.

First of all, you are using 20 kV. Why? You should always use the LOWEST kV possible with an SEM. For most saples 5 kV is more than enough, gives you better resolution (less spread of the beam in the material) and less charging problems. Then, reduce the spot size as much as possible. Not clear if this is a FEG instrument - if so, try 1 kV. You might be surprised by the image you get!

I think you specimen is conductive. Try to use a copper tape wrapped around the sample and the holder.

One other possibility which is much easier to check, is that although the sample may be conductive, perhaps the grounding wire/s on your stage has popped off and not connected, or not making a good contact. You could check that as well.

Wires are properly attached.

I think there is something wrong with the electron scanning board. The distortion does not seem to be caused simply by charging. Similar to Wade above, please try to check it using a completely conductive thin film sample for example. Switch between fast and slow scan.

If, it were caused by charging (e.g. by the underlying non-conductive polymer board), try to lower the acceleration voltage (to 15 or 20 kV or even less). In order not to reduce resolution too much, you could also try decreasing the working distance.

Resolution is INCREASED in most cases by REDUCING the kV. It is a myth (apart from when imaging some dense metals) that high kV gives high resolution on the SEM - Just think about it. When the energy on the beam is increased then the reactive volume (i.e. the volume from which electrons are emitted in the sample) will be increased. The resolution of the SEM image is decided by the effective size of the scanning 'spot' - i.e. the volume from which the emitted electrons are coming from. For most samples this is usually optimised at under 5 kV. For a field emission instrument it can often be around 1 kV. As an added bonus problems such as charging artefacts are also reduced. Of course, you can then reduce the working distance - giving the added bonus of increased resolution. The best combination for resolution in many samples, using an FESEM, is to have an above the lens detector, working distance of 1-2 mm and beam energy of around 1 kV. Just try it!

I agree with a few of the posters above, this looks more like a problem with the scan-generator not charging.

I think it's not a problem of sample conductivity: the magnification is not so high and the disturb is not a real distorsion or drift.

It seems to be a problem of the instrument (probably related to the lens system)

Check if it happens also on other samples, you can try with a very conductive sample; if so you have to call the technical assistance

Try 10-15kv, BSE and another coat of Au-Pd. Tilt towarsd the collecttor in case of SE image.

Sample Preparation is the Key in getting good SEM results.

The sample should be contained within the sample holder (it should not be protruding out)

The sample should be evenly loaded on the sample holder.

The sample should be properly and firmly fixed on the sample stub (vibrations of sample holders can play havoc on SEM output).

Sample need to be sputter coated if non conducting in nature.

Once the SEM is ready for analysis, check the beam shift, astigmatism correction, and film saturation to begin with.

Adjust the spot size, acceleration voltage and line scan.

Select the proper aperture size (low diameter for high resolution )followed by adjusting the accelerating voltage for carrying out the analysis

(Sorry, I might have posted this twice)

Hi, Syed, I had something like that happening to me too and the instrument was fine. Did you solve the problem? There is not much you can do if the sample doesn’t cooperate, so you need to make it do just that. What helped me: variable pressure (if you don’t have it use coating, carbon tape or paint to provide a conductive path); put the region of interest in the direct line of sight with the detector; decrease WD a bit if the sample topography allows it; use frame integral instead of line integral and a scan rate just enough to get a clear image (not too slow; and not too many frames either); use BSE if you have it, helps with charging (if you don’t, you can also negatively bias the SEM detector to mitigate charging); and always degauss your sample, it helps a lot.

As for voltage and spot size, probably you already chose the best you needed for your specimen, so I guess you are stuck with them. If not, as most of the previous comments suggested, try lowering the voltage. At this magnification and for Sn balls, probably you don’t need to mess with a very small spot size.

Hi guys, the image here is stretched out perfectly horizontal. This means it's a scan definition problem not charging. Could be a problem in the scan generator. Charging would not be so perfectly regular in its effect. This can be proven by adjusting the scan rotation; if the distortion remains locked with respect to the scan and not to the sample it cannot be charging. Lewys

Thanks are all the suggestions and comments. The sample is conductive but I will check it again and try different voltages and scan rates. Scan generator, which is more likely the cause of the problem as suggested by researchers, requires attention and will be checked by the service people.

Any other suggestions by the researchers are welcome.

Thanks once again.

It is observed that if the sample is not properly dried during noise reduction these type of problems some time occur,

beam impression also comes.

Even though it is unlikely to be caused by contamination issue, if you do need to clean your microscopes from hydrocarbon contamination, please check out www.piescientific.com. SEMI-KLEEN plasma cleaner can help you maintain a clean electron microscope. It reduces the chance/speed of contamination build up inside electron microscopes. 

First of all, you are using an accelerating voltage of 20 kV. This is, quite frankly, daft. Your 'spot size' also appears to be quite large. You also have a long working distance. Depending on what your electron gun type is, you shouls reduce your accelerating voltage considerably. If it is a thermionic gun, reduce to about 5 kV. If it is a field emission type, then reduce to between 1 and 2 kV - or even less! This will have the immediate effect of not only reducing the charging effect, but will also reduce the interactive volume within the specimen surface, as will reducing the 'spot size' (i.e. that which controls resolution). Lower kV will require a shorter working distance (ultimate resolution needs to work very close with an 'above the lens' detector that can filter out secondary electrons). Unless you are carrying out x-ray analysis you should harly ever use an accelerating voltage above 5 kV, with any instrument. 

Syed,

Did you solve your image distortion problem?  Was it a scanning board?

curious,

Michael Delannoy

The service engineers say that the line generation card needs replacement.

Zameer