Is anyone watching the significant SEM image drift (top-down or bottom-up) during the study of cross-sectional views of "conductive thin film/Si" structure? This drift is very undesirable in determining the thickness of the thin film by SEM. The value of the drift varies sometimes from sample to sample. Is this fact associated with our microscope or with our samples? The used equipment - SEM Leo-1450. Sample preparation is simply chipping the Si substrate along the scratched line without additional conductive coating deposition.
Hi Abil,
It shouldn't drift as you describe especially with conductive film on Si (hope it's not HR-silicon). Check if the SEM stage is properly grounded if this happens with several other conductive samples as well.
The drift can be related to either sample or/and the beam. Most common problem is grounding of the sample. Poor ground leads to charge accumulation, which would lead to drift. Your film may be conductive, but unless etched specifically right before film deposition, Si will have an oxide layer, which in some cases can be rather thick and create an insulating barrier to electrons (some Si wafers are even thermally oxidized to make the oxide layer thicker). Double-check that your stage is grounded. Ground the surface of the sample (the film under study). One can use In foil, which typically doesn't leave a trace on the sample, or as an alternative a small droplet of conductive fluid, e.g. silver paint.
Lyuba, Dhamodaran,
thank you for advice and participation!
Silver paint is used for bonding the Si substrate and the grounding of the film.
Si resistivity ~ 50 Om cm, film resistivity
We also sometimes observe this. Charging is one possibility, which is worse than usual in this configuration because of the sharp edge at the top of the sample. A fresh cleave and a well rinsed (rinsed with acetone and ethanol to remove organic contamination) sample help. Grounding, of course, is important but really what I mean by that is just the local contact of the sample to its mount. Is the sample firmly mounted with a conductive adhesive? Thermal gradients can add to the problem. On days when our air conditioning isn't working well, I just have to give up. And the final problem has just been the control software.
Kurt,
I think that non-working air conditioner will affect the "electronic brains" microscope rather than it will contribute to the appearance of the termal gradient on the sample.
???????????
Assuming you are grounding the sample properly and your sample has a decent conductivity: your enemy may be gravity. Often samples are mounted using soft materials such as carbon tape or not so well cured silver paint. When you tilt the SEM stage to observe the cross-section gravity will generate a lateral force component on the sample causing creep in the mounting medium. This typically stabilizes over a time of tens of minutes. Whether creep occurs can be checked by noting the drift direction and checking the behavior when tilting back to the horizontal position.
Drifts can be minimized by loading it on CCC (Carbon conductive cement), this material is like slurry and will be solidifies hard within a minute. you can dip a portion of your film in that CCC which will strongly bounded to your sample stub.
When coming to charging of your oxide, it can be eliminated by taking average scans of 16 at a 0.3 micro sec speed or else you can also decide by fastening the scan speed and observe where the charging is minimized then certain amount of average scans as you desire at that speed. May the images were not good at high S/N but relevantly applies to your work.
- Badges
- Science topic
- Similar topics
- Chemistry
- Nanotechnology
- Nanostructures
- Thin Films