Methods to stabilize scanning tunneling microscope(STM)

[Potaeto]

When the STM becomes unstable(large variations in the current), what procedures should I take in order to fix it?

I am new to the STM, but after a few month`s experience I notice that one of the main problems is stabilizing it.

When the tunneling current becomes unstable, I`ve tried to vary the bias voltage, current, gain, time constant, and applying voltage pulses. I find that applying pulses much higher than the bias voltage seems to be most effective.

However, there are times when nothing works, and the topographic image lines can vary in orders of tenths of nanometers, with no similarity in the left and right scan directions.

How much does sample preparation affect the STM measurement? I have to prepare my samples on HOPG in ambient conditions. To avoid contamination, I heat the sample after preparation of HOPG, then I dry in a vacuum dessicator. If there are more efficient methods, I`m willing to hear all advices.

 

[ZapperZ]

Have you tried varying the tip-to-sample distance?

Zz.

 

[Potaeto]

Yes, I`ve tried that a few times, but it did not work.

I suspect that it has something to do with sample preparations, because samples prepared in ultra high vacuum by evaporation seem to exhibit nice images with roughness less than 10 pm.
My samples are carbon nanotubes, thus they cannot be evaporated, all I can do is deposit them onto a substrate in ambient conditions.
Once in a while, I have been able to observe high resolution, but upon scanning twice in the same area, the images became unstable.

 

[Potaeto]

Yesterday I found one method that worked.

The left-right scan image lines bared no resemblance and the roughness was in orders of tenths of nanometers.
At the time I was observing at 0.7 nA and 1.8 V. I increased the voltage to 3.00 V and the current to 2.0 nA. After scanning for sometime, part of the tip drops onto the substrate and the tip tracks it much more easily. I then slowly decrease the current, then the voltage back to normal scanning conditions, and moved to observe a clean area.

The system was more stable with height variations below 200 pm. I tried this a few more times when it became unstable, and although it did not yield the best results everytime, the system always became more stable.