Internal conference developed at Institut des Matériaux Jean Rouxel in Nantes (France). Papers included: Bogaerts 1997, Patterson 1999, Bogaerts 2000, Nishikawa 2000, Jackson 2001, Ohta 2002, Jackson 2003, Iordanova 2007, Lotito 2011, Schmidt 2012.

1. On the behavior of argon metastable
Javier García Molleja
In collaboration with:
Axel Ferrec
Pierre-Yves Jouan
Julien Keraudy

2. Three-dimensional density profiles of argon metastable atoms
in a direct current glow discharge: experimental study and
comparison with calculations
 A.Bogaerts, R.D. Guenard, B.W. Smith, J.D.
Winefordner, W.W. Harrison, R. Gijbels
 Spectrochimica Acta Part B 52 (1997) 219-
229

3. Conclusions
 The number density profile is characterized by at least two
peaks, appearing at 2-4 mm and about 12 mm.
 This can be explained by the occurrence of local production
and loss processes which lead to local maxima that cannot
completely be spread out by diffusion.
 The exact position of the peaks and also the absolute value of
the first peak are not yet in complete agreement between
experiment and theory.
 Comparing the absolute values of the metastable density
further away than a few millimeters from the cathode, it can be
concluded that satisfactory agreement is already reached.

4. Relevant figures

5. Optically determined velocity distributions of
metastable argon in the second stage of an inductively
coupled plasma mass spectrometer
 James E. Patterson, Brett S. Duersch, Paul
B. Farnsworth
 Spectrochimica Acta Part B 54 (1999) 537-
544

6. Conclusions
 Laser-induced fluorescence and scanning the
excitation wavelength.
 The velocity profiles of argon metastable are initially
bimodal.
 Is indicative of the presence of a shock wave or
other disturbance at or slightly upstream of the
skimmer orifice.
 Some disturbance to the flow exists near the
skimmer tip.
 The exact nature of the flow disturbance cannot be
determined from our measurements.

7. Relevant figures

8. Description of the argon-excited levels in a
radio-frequency and direct current glow discharge
 Annemie Bogaerts, Renaat Gijbels
 Spectrochimica Acta Part B 55 (2000) 263-
278

9. Conclusions
 A collisional-radiative model, describing the behavior of 64 argon
excited levels, has been developed and applied to argon glow
discharges operated in both the dc and rf mode. The discharge
conditions are typical for a Grimm-type glow discharge.
 The level populations of all excited levels are found to reach a
maximum close to the rf-electrode or cathode, at the boundary
between sheath and bulk plasma, as a result of electron impact
excitation.
 The 4s levels and the low 4p levels are also formed by fast argon ion
and atom impact ionization.
 In the dc discharge, the level populations drop rather quickly in the
bulk plasma. In the rf discharge, on the other hand, the populations are
still considerable in the bulk plasma, due to electron impact excitation
by electrons heated from the fluctuating electric field in the bulk
plasma.

10. Conclusions
 Concerning the relative contributions of various populating and
depopulating processes, we found that direct electron impact excitation
from the ground state and stepwise excitation from lower excited levels
are the most important production processes, as well as radiative
decay and electron impact de-excitation from higher levels.
 Electron impact excitation to higher levels and de-excitation to lower
levels are very important loss processes, together with radiative decay
to lower levels.
 These high energy levels starting from the 4d and 6s levels, are
primarily depopulated by Hornbeck-Molnar associative ionization.
 The intensities of three optical emission lines are plotted as a function
of distance from the electrode. The axial dependence is identical to the
axial level population profiles.

11. Relevant figures

12. Transport mechanisms of ions and neutrals in low-
pressure, high-density plasma etching of high aspect
ratio contact holes
 K. Nishikawa, H. Ootera, S. Tomohisa, T.
Oomori
 Thin Solid Films 374 (2000) 190-207

13. Conclusions
 The transport mechanisms of ions and neutrals in a hole were investigated.
 Etching experiments were performed in electron cyclotron resonance plasmas
with a C4F8/O2 gas mixture.
 Reactive ion etching-lag (RIE-lag) effect was suppressed by increasing the
percentage of O2 addition. The etch rate decreased as the hole diameter
decreased from 0.3 µm.
 The deposition rate of fluorocarbon polymer film on the hole bottom rapidly
increased as the hole diameter decreased and as the percentage of O2 addition
was decreased.
 CF, CF2 and C2 radicals increased by increasing the percentage of O2 addition.
 The ion flux incident on the hole bottom decreased and the ion energy
increased as the aspect ratio of the hole increased.
 The neutral flux decreased as the aspect ratio of the hole increased.
 The etch rate distribution on the hole bottom was estimated by an etching
model that included ion sputtering and ion-assisted etching. Neutrals, as well as
ions, played an important role in hole etching.

14. Relevant figures

15. Spectral, spatial and temporal characteristics of a
millisecond pulsed glow discharge: metastable argon
atom production
 Glen P. Jackson, Cris L. Lewis, Stephen K.
Doorn, Vahid Majidi, Fred L. King
 Spectrochimica Acta Part B 56 (2001) 2449-
2464

16. Conclusions
 During steady state and plateau conditions the
populations of both 3P2 and 3P0 argon atom states
maximize 1-2 mm above the cathode surface.
 Population processes are likely to be electron impact
and fast atom/ion impact at this distance.
 In the afterglow, the bulk of both the 3P2 and 3P0
metastable states are formed 4-7 mm further from
the cathode surface than during the plateau.
 Argon ion-electron recombination followed by
radiative relaxation is the most probable mechanism
of populating these levels.

17. Relevant figures

18. Non-Maxwellian anisotropic velocity distribution
of metastable argon atoms in a thin discharge
cell
 Yoshimi Ohta, Masahiro Hasuo, Takashi
Fujimoto
 Optics Communications 210 (2002) 245–250

19. Conclusions
 We have measured the absorption spectra of
metastable argon atoms in a 1 mm thick glow
discharge cell.
 At low pressures, narrowing of the linewidth was
observed suggesting an anisotoropic velocity
distribution of the metastable atoms.
 Metastable atoms having large velocities normal to
the wall are lost upon collisions with the wall.
 At low pressures, themalizing collisions of
metastable atoms with ground state atoms are
insufficient to compensate the loss.

20. Relevant figures

21. Probing excitation/ionization processes in millisecond-
pulsed glow discharges in argon through the addition of
nitrogen
 GlenP. Jackson, Fred L. King
 Spectrochimica Acta Part B 58 (2003) 185–
209

22. Conclusions
 The addition of ~1% of nitrogen to a 0.8-torr plasma has
provided insight into the mechanisms of excitation and
ionization in the PGD.
 Optical absorbance and emission measurements demonstrate
the transfer of energy from excited argon atoms to nitrogen
molecules during the voltage-on period, with a subsequent
reduction in the number of metastable states of argon.
 This reduction in metastable atoms reduces the ionization of
sputtered atoms during the voltage-on period, but does not
significantly impact emissions from excited analyte atoms
because the latter are created mostly via collisions with
electrons.
 Attenuation factors for ArH+ and Ar2+, are more than twice as
great as for Ar+.

23. Conclusions
 When the voltage is terminated in the ‘pure’ argon discharge, ToF-MS
data and optical experiments show that argon ion recombination leads
to an increase in metastable states.
 This, in-turn, leads to an increase in the propensity for Penning
ionization. Metal-ions and emissions are observed for several
milliseconds after pulse termination.
 When nitrogen is added it prevents electrons from collisionally cooling
in the afterpeak due to superelastic collisions with vibrationally excited
states of N2, formed during the voltage-on period.
 These superelastic collisions delay the onset of recombination
because fast electrons recombine less-readily than slow electrons.
 During this delayed electron-cooling period, ions and electrons are lost
by diffusion to the walls.
 The afterpeak delay, and magnitude, is shown to be controllable by the
nitrogen partial pressure.

24. Relevant figures

25. Optical emission spectroscopy diagnostics of
inductively-driven plasmas in argon gas at low
pressures
 S.Iordanova, I. Koleva
 Spectrochimica Acta Part B 62 (2007) 344–
356

26. Conclusions
 An optical emission spectroscopy method for simultaneous
determination of electron temperature and density is developed and
applied to argon inductively-driven tandem plasma source at low
pressures.
 The method is based on CR model, suggested in the study, for Te, ne
determination.
 The so-called cross-point method is applied to define Te and ne, through
comparison of theoretical and experimental results for line-intensity
ratios.
 The gas temperature is obtained through Doppler broadening
measurements of two argon spectral lines.
 The plasma parameters reached in the first chamber determine to a
great extent those in the second chamber.
 The calculations indicate that the metastable levels play an important
role in the 2p level populations and the spontaneous transitions from
these levels determine the main radiative processes in the argon
plasma of the inductively-driven discharges at low pressure and
applied power.

27. Relevant figures

28. Characterization of argon metastable species
as function of time, space, and current
of a pulsed dc glow discharge
 G.Lotito, T. Nelis, Ph. Guillot, D. Günther
 Spectrochimica Acta Part B 66 (2011) 619–
626

29. Conclusions
 Optical emission and absorption measurements have been performed
on an iCCD camera using a pulsed GD source in order to investigate
the spatial, temporal and power dependent properties of the Ar* states
at (3P2) 1s5.
 Both types of images have shown during the plateau regime that the
emission intensity and the argon metastable density are highest in the
proximity of the cathode, the negative glow region.
 During the afterglow temporal regime, this maximum is shifted in a
region 6–8 mm away from the cathode. The increase in this spatial
region during the afterglow compared to the plateau region is 50 %.
This holds true for power densities of 0.08 W mm−2 to 0.15 W mm−2.
 When the power is decreased, the overall intensity decreases. the
emission peak observed during the plateau regime splits into two
separate peaks, one moving away from the cathode and the other
remaining in the proximity of the cathode. At the same time the
maximum in the afterglow moves towards the cathode.

30. Conclusions
 At power densities below 0.02 W mm−2 a distinct afterglow peak cannot be
observed anymore. A distinct afterglow is observed only for power densities
above 0.08 W mm−2.
 For LAGD-TOFMS the area of optimal overlap between the plasma plume and
the discharge is about 6 mm away from the cathode, if ion production through
Penning ionization should be favored.
 At the same time the power density in the pulsed discharge should be
maintained relatively high.
 Mass spectrometric measurements have revealed that the analyte ion detection
is strongly enhanced after pulse termination, regardless of the absence of an
increase in the metastable density during the “afterglow”.
 The increase of the Ar* in the afterglow and the characteristic increase of the
signal of the cathode material observed in MS is not so straightforward.
 Although Penning ionization certainly plays an important role in the generation
of analyte ions throughout the discharge pulse, and particular during the
afterglow when direct electron impact ionization becomes less likely due to a
decrease of the electron density, it cannot explain the manifold increase of the
analyte ion signal observed by MS during the afterglow.
 Possible role of Ar2 dimers and dimer ions in the formation of the afterglow.

31. Relevant figures

32. Ion mass spectrometry investigations of the discharge during
reactive high power pulsed and direct current magnetron sputtering
of carbon in Ar and Ar/N2
 S. Schmidt, Zs. Czigány, G. Greczynski, J.
Jensen, and L. Hultman
 JOURNAL OF APPLIED PHYSICS 112,
013305 (2012)

33. Conclusions
 Reactive DCMS and HiPIMS of graphite in an Ar/N2 atmosphere yield a
discharge that contains mainly C+, N+, N2+, as well as Ar+ and to a lower
extent CN+ and C2N+ ions.
 The ion species in HiPIMS processes exhibit higher ion energies and a
lower amount of fully thermalized ions compared to DCMS.
 As the ion-flux at 50 % of nitrogen in the process gas in HiPIMS the
flux is composed of 26 % C+ and 47 % N+ in contrast to the DCMS
plasma, which contains 8 % of C+ and 70 % of N+ at equivalent process
settings.
 The comparison on the ion-flux shows further a rising amount of CxNy
(x, y 2) species in the HiPIMS plasma.
 The evaluation of the target current confirms that chemical sputtering
is active at the substrate and at target as soon as nitrogen is added to
the sputter process.
 The most pronounced FL structure is obtained for CNx films sputtered
in HiPIMS mode at 430 °C.

34. Relevant figures

35. Relevant figures