Analysis of Metastable Species During Plasma Processing Using Mass Spectroscopy

Mass spectrometers can now be operated at pressures in the proximity of those applied in most plasma processing systems due to the availability of particle detectors capable of being used at pressures up to 4x10^-4Torr. As a result, sampling of both neutral and ionized species from plasma reactors can be improved.

It is possible to operate the Hiden Analytical quadrupole mass spectrometer (QMS) in a mode, wherein the electron energy released by the ionization source is variable. This mode is known as Threshold Ionization Mass Spectrometry, or TIMS.

A specific ionization energy is required for different elements to have an orbiting electron removed from their shells. This energy varies with the electron orbital, for instance, the ionization energies of outer shell electrons are typically weaker due to lower electrostatic forces and greater distance from the positively charged nucleus. This results in electron impact ionization efficiency curves (Figure 1).

Figure 1. The electron impact ionization efficiency curves

The onset of ionization process of neutral particles takes place at minimum (threshold) energy of the impacting electrons, which is dependent and distinct to any species available in a gas matrix.

This leads to a spectral “identifier” or “fingerprint” for all molecular or atomic species. For instance, helium/ deuterium (He/D2) ratios can be accurately determined for neutral species during plasma fusion, using the TIMS technique.

The by-product of plasma fusion is helium ash. This quantification is generally excluded while employing a QMS in conventional mass spectral modes because of the overlapping convoluted mass spectral signatures of both D2 and He at 4amu.

Here, the actual mass separation is only 0.02amu. Figure 2 depicts the Hiden Analytical QMS running in TIMS mode, showing the electron energy spectra for D2 and He with ionization onsets at 15.4 and 24.5eV, respectively.

Figure 2. The electron energy spectra for D2 and He with ionization onsets at 15.4 and 24.5eV, respectively.

Figure 3 presents the resulting electron energy spectrum during the concurrent presence of both of these gases, clearly showing a deconvolution of the two species in the TIMS spectra.

This shows the possibility of detecting the presence of D2 in helium down to ppm detection levels. The Joint European Torus experimental nuclear fusion facility in Oxford, UK, uses Hiden Analytical TIMS equipped mass spectrometers for routine analyses.

Figure 3. The electron energy spectrum acquired when D2 and He present simultaneously

Experimental Results

Helium has an ionization potential of 24.6eV. The section AB of the curve in Figure 4 represents the metastable He*_m atoms, which are resilient against spontaneous decay. Upon impacting the detector, they have adequate energy to produce pulse counts.

The section BC of the curve in Figure 4 represents the metastable and ionized helium contributions for electron energies beyond 24.6eV. Similar data were acquired in other analyses for argon, krypton, and neon. Figure 4 includes the data for krypton.

Figure 4. Experimental results

The form of the curves depicted in Figure 4 may be known by reference to Figure 5. Figure 6 shows the schematic view of the system used to obtain the data shown in Figure 4.

Figure 5. This image shows the reference to understand the form of the curves shown in Figure 4.

Figure 6. The schematic view of the system used to obtain the data shown in Figure 4.

It is possible to maintain RF plasma in the reactor, between an electrode and the sampling orifice of the mass spectrometer. The ions entering the Hiden mass spectrometer from the reactor could be controlled using the electrodes behind the orifice. By operating the particle detector at pressures of 4.10^-4Torr, gases were introduced into the reactor or directly into the mass spectrometer’s source.

Here, the plasma was operated in helium, but the mass spectrometer’s internal ionization source was turned off and its sampling system was set to reject all plasma ions. Under these conditions, the detector recorded the arrival of He*_m generated in the plasma. The metastable signal varied in proportion to the plasma power and to the gas pressure in the reactor (Figure 7).

Figure 7. The metastable signal was proportional to the plasma power and to the gas pressure in the reactor

The replacement of the helium plasma with the oxygen plasma showed no energetic particles in the plasma because the energy of the metastable oxygen species, though have long lifetime, is inadequate to be recorded by the detector.

Figure 8 shows the recorded signals for a mixture of helium and oxygen by setting the sampling system again to reject plasma ions and operating both plasma and mass spectrometer source.

Figure 8. The recorded signals for a mixture of helium and oxygen with both plasma and mass spectrometer source operating, and the sampling system again set to reject plasma ions.

In Figure 8, the section BC of the helium curve shows ions produced from ground-state helium sampled from the reactor, whereas section AB presents ions produced from sampled metastable helium.

Metastable helium atoms produced in the source between 20 and 25 eV have a small contribution. The threshold energy (not shown) is predicted to be roughly 5eV (Figure 5). Penning ionization was not observed in the internal source for oxygen.

Figure 9 presents the data for pure oxygen under the conditions of 15W plasma at 30mTorr and a mass spectrometer source pressure of 2.10^-4Torr. The area below 16eV seems to have two components with onset potentials differing by roughly 1eV. This is as anticipated when the sampled oxygen consists of metastable a ¹Δ_g oxygen, or when the latter was generated in the mass spectrometer source.

Figure 9. The results for pure oxygen under the conditions of 5W plasma at 30mTorr and a mass spectrometer source pressure of 2.10^-4Torr.

Conclusion

Metastable species generated in the plasma can be directly detected when the pressure differential between a plasma reactor and an attached mass spectrometer is reduced, provided the species have adequate internal energy and long lifetimes.

Additionally, the detection of metastable species and other plasma products with lower energy and longer lifetimes is also simplified. These measurements may be applicable in view of the role of energetic neutral species in the plasma processing of surfaces.

This information has been sourced, reviewed and adapted from materials provided by Hiden Analytical.

For more information on this source, please visit Hiden Analytical.