Energy and Mass Spectrometry of Processing Plasmas

Plasmas consist of complicated mixtures of neutral and charged species. It is certainly critical for the design and control of any surface engineering process that the user has as much information as possible about the components of a given plasma, about how the behavior of these components changes as the electrical conditions of the plasma are varied, and about the interactions that take place between the plasma components and the substrate and target surfaces that are exposed to them. This article outlines the application of a mass/energy analyzer in diagnostics of a range of processing systems.

The measurements to be described were made using a Hiden mass/energy spectrometer, the most common form of which is represented in Figure 1. The following are the main parts of the instrument: the internal ionization source (which is turned off when sampling ions from a plasma), its ion sampling region, the quadrupole mass selector, the cylindrical, electrostatic energy analyzer, and the channeltron ion detector and pulse counting circuitry. Certain other forms of the spectrometer are also used, the most recent being that in which the analyzer is incorporated in the Hiden HPR60 instrument which is designed for sampling from plasmas operated at atmospheric pressures.

Figure 1. Hiden mass/energy analyzer.

Experimental Investigations

a) Neutral Species

The mass-identification of the presence of neutral species, firstly in the absence of a plasma and then with the plasma operating, is the most simple measurement for any reactor. Generation of new species in the plasma or through its interaction with the walls of the reactor or substrate and target surfaces and the fragmentation of the working gas are effortlessly identified. To enhance the accuracy of the measurements, it is occasionally recommended to arrange for the sampling to be done via a doubly or triply differential pumping geometry incorporating a ‘chopper’ arrangement to improve the discrimination between the required plasma signals and competing background effects (1, 2).

It is possible to add considerably to the information derived about the neutral species generated in a plasma if the energy of the electrons used in the source of the mass spectrometer can be altered. Then, it is feasible to study, generally near the threshold ionization energy, changes in the populations of specific ions produced in the source as the sampled neutral population is modified by changes in the plasma conditions. Figure 2 displays typical data from a Hiden HPR60 instrument used to analyze a plasma needle discharge operated in a mixture of air and helium (2).

Figure 2. Ion signals as function of electron energy in mass spectrometer source.

Several of the processing gases that are used particularly for the deposition of DLC and optical coatings and semi-conductor etching are electro-negative; the examples of such electro-negative gases include oxygen, sulphur hexafluoride, carbon dioxide, and methane. Hence it is generally beneficial to operate the ionization source of the mass spectrometer to generate negative ions (3, 4, 5). Even if the dominant negative ion formed from two different parents is the same, the variations in production of the ion with electron energy are very unique. This is evident as in the case of carbon dioxide and nitrous oxide, both of which give O^- as their dominant negative ion.

b) Ion Species

The ionization source of the Hiden analyzer is turned off and the plasma ions are directly sampled to facilitate the identification of the positive and negative ions produced in the plasma and to arrive at the surface of a substrate. For ions that are mass-selected, it is then plausible to ascertain their energy distributions, which are infamous to greatly impact the hardness, adhesion, and other characteristics of deposited films such as diamond-like carbon, titanium oxide, titanium nitride, and other compounds. It is often helpful if the measurements of the relative abundance and energies are carried out on ions sampled at a point which is directly equivalent to one on the substrate being processed. This can be done by including the sampling orifice of the mass/energy analyzer in the surface of the substrate holder, or in the surface of another electrode if this is in a directly equivalent electrical condition. To show an example, Figure 3 illustrates the energy distributions measured for N⁺ ions arriving at the cathode of a DC plasma when it was maintained at potentials of between –600 and –750 V.

Figure 3. Energy distributions of N⁺ from a nitrogen discharge for cathode voltages of 650, 700, 750 and 800 V, obtained at a pressure of 50 Torr.

Variations in the relative abundance of specific ions in a reactor and their energies with modifications in the conditions of the plasma can be compared with the emergent changes in the characteristics of a deposited film. For instance, data for a titanium nitride deposition system employing a combination of an inductively-coupled (ICP) source and a D.C. magnetron, used to increase the amounts of N⁺ and N₂⁺, represented (6) that the addition of a small amount of helium to the I.C.P source increased the nitrogen ion signals appreciably and gave a positive correlation with improvements in the tribological characteristics of the TiN films. Similar experiments for the growth of optical-quality titanium dioxide films were also recorded.

It is progressively common to run magnetron deposition reactors in a variety of pulsed modes. Lin et al. (7) recently published measurements for the deposition of CrAlN films by using a Hiden mass/energy analyzer attached to a dual magnetron system. These measurements have shown how the ion energy distributions in such systems are highly influenced by the exact waveforms of the pulsed voltages applied to the two magnetrons. A sample of the energy distributions measured is depicted in Figure 4. In the paper titled TF1-2 by Lin et al., the distributions and their correlation with the characteristics of the deposited films are explained in detail at this meeting. Bradley and his collaborators (8) have explained other work which includes time-resolved measurements with a Hiden instrument for a pulsed magnetron reactor.

Figure 4. Ion energy distributions for A1⁺ as a function of probe position.

In magnetron deposition, high power impulse magnetron sputtering (HIPIMS) systems are the most recent development. In HIPIMS, the emphasis is on generating large numbers of ionized metal ions in the vicinity of the magnetron. Several groups are again examining the energy distributions of the ion populations and the corresponding ion populations with Hiden mass/energy analyzers. Ehiasarian et al. (9, 10) compared their HIPIMS reactor with a more traditional magnetron system having energies around three times higher. The results showed that the HIPIMS reactor can release between four and five times more reactive titanium and nitrogen ions. Vicek et al (11) have found the ion flux to their substrate to contain up to 92% copper ions with energies up to 45 eV relative to ground potential. The voltages that are applied to the magnetron electrodes produces waveforms. These waveforms are then compared with the energy distributions. This comparison aids remarkably in the design and interpretation of the operating conditions.

Conclusions

It is evident that the measurement of mass and energy distribution spectra for the neutral species and ions present in a broad range of processing plasmas is of substantial help in enhancing the operation and design of reactors. The Hiden mass/energy instrument has found extensive use for such measurements and, along with a molecular beam sampling system, is now being applied to plasmas operated at atmospheric pressures.

References

1. H Singh, J W Coburn and D B Graves J Vac sci Technol A17, 2447, 1999
2. Y Aranda Gonzalvo, T D Whitmore, J A Rees, D L Seymour and E Stoffels J Vac Sci TechnolA24 ,550, 2006
3. J A Rees, D L Seymour, C L Greenwood and A Scott. Nucl. Inst. & Methods B 134,73, 1998.
4. W Stoffels, E Stoffels and K. Tachibana Jpn. J. Appl. Phys. 36, 4638, 1997
5. K Teii, M Hori, M Ito, T Goto and N Ishii J. Vac. Sci. Technol. A18,1, 2000
6. C Muratore J J Moore and J A Rees Surf Coat Tech 12, 163, 2003
7. J Lin J J Moore B Mishra W D Sproul and J A Rees Surf Coat Technol 201,4640 2006
8. J W Bradley, H Bäcker, Y Aranda Gonzalvo, P J Kelly and R D Arnell Plasma Sources Sci. Technol- 11, 165, 2002
9. E Ehiasanian, Y Aranda Gonzalvo and T D Whitmore Plasma Proc & Polymers 2007 (to be published)
10. J Bohlmark, M Lattemann, J T Gudundsson, A P Ehiasarian, Y Aranda Gonzalvo, N Brenning and U Helmersson Thin Solid Films 515, 1522, 2006
11. J Vicek P Kudlacek K Burcalova and J Musil J Vac Sci Technol A25, 42, 2007

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

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