Flood Gun Effect on Native Oxides of 40 Pure Elements

The thickness of naturally-formed, native oxides on pure Elements is easily measured, and is found to range from 1 nm to 7 nm for most pure elements. A few elements (eg Mg, Ca, Sr, Ba, Pb, V, Tl, and rare earth metals) tend to form native oxides that are >10 nm thick. The alkali metals (Li, Na, K, Rb and Cs) are very reactive and form thick native oxides that are >20 nm thick.

When native oxides are more than 8-9 nm thick the high energy resolution chemical state spectra do not show the pure element (metal) signal that is covered by the native oxide. The ability to detect the pure element under a “thin” naturally formed native oxide depends on the geometry of your electron collection lens and your sample surface.

The native oxide spectra shown here have oxide layers that range from 1-7 nm thickness. Most native oxides (>60%) have native oxides that are 2-5 nm thick. This means that we can record both the pure element peak and the native oxide peak in a spectrum. When we use a 90 degree electron take-off-angle, we can more easily see the pure metal signal. If the electron take-of-angle is <60 degrees, then we might not see the pure metal signal. This is a simple angle effect. The many spectra that follow are these types of spectra. These spectra all show both a pure metal (element) peak and a native oxide peak.

In a few examples the metal oxide peak is small. In a few cases the pure element peak is small. But the important feature is that we can see both the pure element and the native oxide in each spectrum. The feature is sometimes very useful to understand chemical shifts and differential charging effects.

Native Oxide of Aluminum Foil – Sample GROUNDED
and
Native Oxide of Aluminum Foil – Sample FLOATING

Figure: This set of 4 overlay plots (from the same element, but difference conditions) are an example of the plots shown below.

If the lens is “normal” (e- TOA of 90 deg) with respect to the plane of the sample, then your instrument can readily measure the pure element signal that is just under its naturally formed, native oxide. If however, when your collection lens is at a 35 or 45 deg angle with respect to the plane of the sample, then it is not possible to measure or detect the pure element signal that is just under a native oxide that is 5-6 nm thick. (Note: some people reference their angles with the “normal” e-p TOA as a 0 deg angle).

The following sets of spectra, taken in 1988, show that we should turn OFF the flood gun when we analyzed naturally formed Native Oxides of pure elements. However, it is best to float native oxides and then charge correct the spectra with respect to the BE of the pure metal/element. This second method avoids the chance of differential charging, but adds an extra bit of work because you must charge correct the spectra from the samples because the sample is floating (not grounded).

If you leave the Flood Gun turned on when you analyze a Native Oxide, then there will be energy shifts in any native oxide sample that is touching the grounded sample stage that can be misleading or wrong.

If, however, that native oxide sample is forced to float by sitting on an insulator, then there will be almost no “unexpected” energy shifts in the Carbon, Oxygen or Metal signals.

For the following sets of spectra, we used 5 different Flood Gun settings on a series of naturally formed native oxides. In some cases we had to use a 90 deg electron take-off-angle (= 0 deg TOA wrt lens) because the naturally formed native oxide was very thick.

The spectra are high energy resolution chemical state spectra from:

the Main Metal signal,
the C (1s) signal, and
the O (1s) signal.

The instrument was an Surface Science Instruments S-Probe equipped with a manually controlled electron flood gun and monochromatic Aluminum X-rays. The plane of the sample is 35 deg with respect to the electron collection lens.

By tilting, the plane of the sample can be set to a 90 deg angle. (note: different people use the opposing angle to define their lens to sample angle.)

The Electron Flood Gun settings for each sample were varied as follows: