Freshly Ion Etched Pure Elements – Capture UHV Gases

forming Metal Carbides, Metal Oxides, or simply physi-sorbing various residual gases.

The following is a series of time based chemical state spectra from 38 pure ion etched elements that were left in a cryo-pumped UHV chamber overnight for >12 hrs.

This first example shows Beryllium metal that was ion etched until all Oxygen and Carbon contamination was removed. The sample was them measured using high energy resolution chemical state spectra. The Be (1s) was measured first, the O (1s) was then measured. The C (1s) was measured last. The total time used to measure changes in clean, ion etched Beryllium was ~15 hrs.

Figure: This set of 3 Montage plots (from Metal, C 1s and O 1s) are an example of the plots shown below.

XPS instruments are usually designed to maintain an UHV level between e-7 mbar to e-10 mbar in the main analysis chamber.

In special instruments, the vacuum can be even greater: e-11 or e-12 mbar. What does that mean? That means that the number of gas molecules inside the main chamber is extremely low. The surface of the walls inside the analysis chamber is normally “fully passivated” which means it has a “stable” (non-reactive) layer of metal oxides which are covered by a layer of physisorbed gas molecules, water, and low molecular weight organic gases.

Even though the surface of the walls are stable, there are still various gases that fly around inside the main chamber. These gases can be: CO, CO2, H2O, O2, H2, CH4, N2, and others. When one gas molecule strikes the chamber wall, various other gas molecules, that are weakly physisorbed (trapped) on the stainless steel (or other material) wall, are suddenly flying around inside the analysis chamber until a different surface can “trap or collect” those various gas molecules. The ability to trap (collect) a gas molecule is related to the “sticking coefficient” of the material that the gas molecule strikes.

A large sticking coefficient means that the material hit by the gas molecules has the potential to trap (collect) a large number of various gas molecules, and vice versa.

After baking an analysis chamber, many of the gas molecules on the surface of the walls are desorbed and removed from the chamber by the vacuum pumps (cryo-pumps, turbomolecular pumps, ion pumps, and dry scroll pumps) or the gas molecules are captured by a “getter” that reacts with various gases. One getter is the Titanium Sublimation Pump (TSP) that is very common in many XPS instruments.

The TSP works by converted solid Titanium rods into “free atomic Ti atoms” or clusters of Ti atoms that float around inside the analysis chamber. Titanium atoms are reactive. Individual titanium atoms react with many of the gas molecules that are flying around inside the analysis chamber. Water will react with Titanium to form TiO2 and H2 gas. Oxygen will form TiO2. CO will react to form Ti Carbide and Titanium Oxide. Methane will react to form Ti Carbide and H2 gas. H2 will react to form Ti Hydrides. N2 is a strong molecule and does not readily react with the Ti atoms.

The process of Argon Ion Etching (using mono-atomic Ar+) sputters (etches) away the top few layers of a material exposing a layer of atoms that can be very reactive. The overlayer was covering these atoms, until the Argon Ions physically forced (destroyed) the overlayer away from their original surface.

The destroyed surface material was converted into gaseous molecules or clusters that fly. Argon ion etching can remove material slowly (1 angstrom/minute), or very quickly (>10 angstroms/second). Speed of removal depends on acceleration, current, beam size, raster size, bond strengths of surface, and the angle of impact.

When a single element material, that has a fully stable, fully passivated surface is scraped in the air or Argon Ion Etched inside the main analysis chamber, the freshly exposed surface atoms have “dangling” bonds, which means they have one or two electrons that are available to react with a gas molecule that strikes it. (Note: these electrons are not the free electrons inside the Fermi sea.)

The reactive dangling bonds on the surface will, if possible, react with any gas molecule that strikes it. That gas molecule will either be physisorbed, chemisorbed, or broken apart to form, if the reaction has enough energy: metal oxide(s), metal hydroxide(s), metal carbide(s), metal-carbonyl complex, or metal-hydride(s).

Some metals (eg Au, Pt, Pd), when Argon Ion etched clean, do not react at all. They can and do form complexes. They can physisorb the organic gases that are inside.

The following series of “Time-Lapse” sets of spectra reveal what happens when a pure element is Argon Ion etched to remove the original overlayer, and expose those “dangling bonds” which are actually unbonded electrons still attached to their atom.

The monochromatic XPS instrument (SSI S-Probe) used for this project was equipped with a CTI Cryo-pump that captures by freezing gas molecules onto a cold plate. The operating vacuum was routinely e-9 mbar. Cryo-pumps are very fast and very efficient to collect gases much faster and more efficiently than ion pumps or turbomolecular pumps, but just like all vacuum pumps, a cryo-pump is not so good to trap CO, or H2.

These Time-Lapse sets were collected by using the Depth Profile spectra collection routine provided by the SSI S-Probe software, in a reverse mode. Depth Profiles usually measure the top 5-10 nm of a surface after removing several nanometers of the surface.

In this case, the surface is strongly Argon ion etched at the start to remove 30-50 nm of the pure metal surface and to produce the least amount of residual Carbon or Oxygen on the surface by watching the, non-scanning “Snap-Shot” oscilloscope and the electron count rate.

When the surface was as clean as possible, the software was started so that it measures the stepwise buildup of oxygen, carbon, and metal compounds that result from the dangling bonds reacting with the residual UHV gases during an overnight run that lasts 10-15 hours. Spectra were collected at the very start and generally show only the pure metal. The next data collection is roughly 30-60 minutes later after the surface has had a chance to collect or react with the UHV gases that remain inside the Cryo-pumped main analysis chamber. (Note: when repeating some of these “Gas Capture” profiles using an ion pump or a turbomolecular pump, the results are different due to difference in UHV gases, and differences in vacuum levels.)

Each “UHV Gas Capture” run include the metal signal, the C (1s) signal, and the O (1s) signal. The spectra at the bottom-front of each Montage plot are the spectra at Time=0 sec. The spectra at the top-back are the spectra from the final measurement that was 10-14 hours later.

For many metals, there is an obvious reaction to form a metal oxide or hydroxide species. Others form metal carbides. The “noble” metals (Ag, Au, Pt) only collect carbon, probably a hydrocarbon.

In many cases, the reaction stops after several hours. Why? The newly formed surface is “stable” or “passivated”. But after exposing that surface to the lab air, the Oxygen and Carbon will most likely increase with time until the surface is “passivated” at 1 atm level.