Kelvin probe in-situ TEM

First some background:

When a nanowire is put into resonance with the in-situ TEM-SPM method by applying a alternating electric field with the resonance frequency of the nanowire between the two tips a offset voltage can be used to reduce the netcharges between the nanowire and the opposing tip. This offset voltage should be dependent on the work-function of the tip material and the nanowire itself, when the netcharges are removed by applying the offset voltage it is visualized in TEM as a reduction of the resonance amplitude of the wire. So by plotting the resonance amplitude as function of offset voltage one can in turn measure the work function of the nonmaterial. This is called the kelvin probe method.

The results:

I made some test on this, and the first result was as can bee seen below, a plot with a nice linear behavior with a minimum, indicating that the difference in work-function of the Mo6S3I6 nanomaterial and the gold-tip is 0.76 eV, -reasonable?

I tried to repeat the experiment but got inconsistent results it appeared that the minimum amplitude due to the offset voltage was dependent on the distance between the nanowire and the opposing tip. Therefore a made a experimental serie on distance between nanowire and tip, as can be seen below. The length of the nanowire is about 6 um and the tested range of distances were 0.4 um to 9.9 um. As seen in the plot only short distances produces a minimum amplitude dependence on offset voltage, and the minimum shifts in offset voltage for different nanowire tip distances. This is strange if the minimum in resonance amplitude is said to only be dependent on the work-function of the nanowire and opposing tip. There also exist a non-linear term ter in the Kelvin probe equation, perhaps this is influencing the results? Short range effects due to electrostatic forces might also have an influence, but if this was the case one should expect the oposite effect but here at longer distances the effect of a minimum amplitude dependence on offset voltage disappear.

Electrostatic charging is a common problem in TEM, in which the electron beam introduces charges in the sample. I also tested this dependecy in the Kelvin probe experiment. As can be seen in the last graph above, several beam intensities in the range from 0.18mA/cm² to 2.5mA/cm² were tested and indead it do have a influence on the resonance amplitude minimum. The minimum shifts roughly 1 Volt at the lowest beam setting relative to the highest. The intensity ranges tested here are considered to be in the normal intensity range, no extreme setting that is.

So as for now, no good result were achieved, proper tip distance setting and how to reduce beam influence needs to be solved.

How to measure the beam current density in TEM

Here I will explain for you how to measure the beam current in the TEM by a alternative method to the standard Faraday cup approach

This is a convenient method if you have a in-situ TEM-SPM sample holder which enables you to directly measure the electric current of the electron beam. Below you find a list of the things I use:

1. Transmission electron microscope (JEOL-2000FX).
2. in-situ TEM-SPM holder modified Nanofactory version.
3. PXI digital multimeter card (PXI-4071) with picoamps range.
4. Copper wire 0.25mm in diameter.
   

You replace the sample wire in the sample holder as the sketch below shows. Use of light weight metal wire is crucial to minimize electron back scattering. For copper it is around 30% of the electrons that are scattered and therefore will not contribute to the measurement so therefore you need to compensate for this. Besides this it should be straight forward from here on.

The highest measured beam-current was in my case 909 pA at 30kX magnification, spot size 1, condenser aperture 120 um.

Current is depending on filament power, spot-size (C1), magnification, and condenser aperture.

The CCD camera operates best in the beam current range of 20 pA (dark-screen) to 136 pA (bright-screen), while 100 pA is the normal value. This corresponds to a beam current density of 0.5 to 3.3 mA/cm2 and normal is 2.5 mA/cm2

(normal operating conditions) Below is the measurement data table.

Schematics: using the in-situ TEM probe for beam-current measurement.

Some references for you to read regarding this method

Ref. Williams, Carter

p.78

"You can measure the beam current with a picoammeter in the earth line"

"You can calibrate the Faraday cup measurement against the TEM screen exposure meter"

"The beam current is usually in the range of nanoamps to picoamps."

p.112

"If you don't have a Faraday cup, it is possible to get a approximate reading on the current by just measuring the current through a insulated line from a bulk region of the specimen and correcting for electron backscatter. Backscattering is independent of accelerating voltage and aproximately linear with atomic number up to about Z = 30. For example backscatter coeficent for Cu is about 0.3 and for Al it is about 0.15."

-I used the PXI-4071, soft-panel version at 1uA range =1pA resolution. Copper wire 0.25mm

p.79 "To measure the beam size in TEM/STEM you must form a image of the beam on the TEM viewing screen under conditions where you know, or can calibrate the magnification."

-I used the CCD-camera, it's been calibrated previously, to image the beam. I don't considerer the radial intensity distribution of the beam.

Ref. Reimer

p.193

"A fraction n of the incident electrons can leave the specimen as backscattered electrons (BSE)...n is known as the backscattering coefficient."