How%20helicons%20started:%201962

1
(No Transcript)
2
How helicons started 1962
3kW 17 MHz 500G
UCLA
3
How helicons started 1970 - 85
1 kW, 1 kG, argon n 1013 cm-3 10X higher than
normal
4
UCLA
In helicon sources, an antenna launches waves in
a dc magnetic field
The RF field of these helical waves ionizes the
gas. The ionization efficiency is much higher
than in ICPs.
5
(No Transcript)
6
A large number of problems arose

• Absorption mechanism and efficiency
• Weak m 1 mode and the Big Blue Mode
• Downstream density peak, axial ion flow
• Non-monotonic axial decay
• Triangular radial profile
• Mass-dependent density limit
• Low-field density peak (30G)
• Density jumps with increasing B0, Prf
• Half-wave antenna better than full-wave
• Endplate charging with small diameters
• High ion temperatures
• Parametric instabilities

UCLA
7
The Landau damping hypothesis
In Landau damping, electrons surf on the wave
The helicons phase velocity is close to that of
an electron near the peak of the ionization cross
section (100eV)
UCLA
8
Landau damping disproved
A fast (RF) energy analyzer was built and
calibrated
RF modulated electron gun for calibration
2-electrode gridded analyzer with RF response
D.D. Blackwell and F.F. Chen, Time-resolved
measurements of the EEDF in a helicon plasma,
Plasma Sources Sci. Technol. 10, 226 (2001)
UCLA
9
Time-resolved EEDFs show no fast electronsabove
a threshold of 10 -4
I-V swept by oscillating Vs
I-V at two RF phases
Loading resistance agrees with calculations w/o
L.D.
Injecting a current causes a beam-plasma
instability
10
The Trivelpiece-Gould mode absorption mechanism

• Helicon waves are whistler waves confined to a
  cylinder.
• Their frequencies are ltlt ?c, so that normally me
  ? 0 is OK.
• However, if me ? 0, the dispersion relation has
  another root.
• The new root is an electron cyclotron wave in a
  cylinder. It is called a Trivelpiece-Gould (TG)
  mode.
• The TG mode exists in a thin layer near the
  surface and is damped rapidly in space, since it
  is slow. The helicon wave has weak damping.
• This mechanism was suggested by Shamrai and
  Taranov of Kiev, Ukraine, in 1995.

UCLA
11
Why are helicon discharges such efficient
ionizers?
The helicon wave couples to an edge cyclotron
mode, which is rapidly absorbed.
12
The H and TG waves differ in k?
k
This axis is essentially k?
UCLA
13
Detection of TG mode was difficult
An RF current probe had to be developed
UCLA
14
A large number of problems arose

• Absorption mechanism and efficiency
• Weak m 1 mode and the Big Blue Mode
• Downstream density peak, axial ion flow
• Non-monotonic axial decay
• Triangular radial profile
• Mass-dependent density limit
• Low-field density peak (30G)
• Density jumps with increasing B0, Prf
• Half-wave antenna better than full-wave
• Effect of endplates and endplate charging
• High ion temperatures
• Parametric instabilities

UCLA
15
Types of antennas
RH and LH helical
Nagoya Type III
Boswell double saddle coil
3-turn m 0
UCLA
16
The m 1 (RH) mode gives much higher density
RH mode
LH mode
UCLA
17
m 1 mode much stronger than m 1
D.D. Blackwell and F.F. Chen, 2D imaging of a
helicon discharge,Plasma Sources Sci. Technol. 6,
569 (1997)
m 1
m 1
18
UCLA
Reason m -1 mode is not easily excited
m 1
m -1
The m -1 mode has a narrower wave pattern
hence, it couples weakly to the TG mode at the
boundary.
19
The dense core (n 1013-14 cm-3) is due to
neutral depletion, allowing Te to increase
The Big Blue Mode
No Faraday shield
With shield
20
A large number of problems arose

• Absorption mechanism and efficiency
• Weak m 1 mode and the Big Blue Mode
• Downstream density peak, axial ion flow
• Non-monotonic axial decay
• Triangular radial profile
• Mass-dependent density limit
• Low-field density peak (30G)
• Density jumps with increasing B0, Prf
• Half-wave antenna better than full-wave
• Endplate charging with small diameters
• High ion temperatures
• Parametric instabilities

UCLA
21
Machine used for basic studies
r 5 cm L 160 cm
B ? 1kG, Prf ? 2kW _at_ 13-27 MHz, 1-10 MTorr Ar
UCLA
22
Symmetric and asymmetric antennas
The maximum density occurs DOWNSTREAM, while Te
decays. This is due to pressure balance nKTe
constant.
UCLA
23
Line radiation is main loss in Te decay
UCLA
24
Non-monotonic decay of wave downstream
Oscillations are due to beating of radial modes
with different k. Theory fails as density
changes further out.
Average decay rate agrees with collisional
damping.
UCLA
25
Triangular density profiles
Nonlinear diffusion, coupled with a bimodal
ionization source, can explain "triangular"
density profiles.
UCLA
26
A large number of problems arose

• Absorption mechanism and efficiency
• Weak m 1 mode and the Big Blue Mode
• Downstream density peak, axial ion flow
• Non-monotonic axial decay
• Triangular radial profile
• Mass-dependent density limit
• Low-field density peak (30G)
• Density jumps with increasing B0, Prf
• Endplate charging with small diameters
• Half-wave antenna better than full-wave
• High ion temperatures
• Parametric instabilities

UCLA
27
Mass-dependent density limit
As B0 is increased, n rises but saturates at a
value depending on the ion mass. This effect was
first observed by T. Shoji.
UCLA
28
A drift-type instability occurs
M. Light (Ph.D. thesis) found that an instability
occurs at a critical field and causes the density
to saturate.
This is the oscillation spectrum for neon.
He identified the instability as a drift-Kelvin
Helmholtz instability and worked out the theory
for it.
M. Light, F.F. Chen, and P.L. Colestock, Plasma
Phys. 8, 4675 (2001), Plasma Sources Sci.
Technol. 11, 273 (2003)
UCLA
29
Anomalous diffusion results
Outward particle flux was measured with n f
correlations, agreeing with that calculated
quasilinearly from the growth rate.
UCLA
30
Density limit due to neutral depletion
Axial density profile with two 2-kW antennas 1m
apart
UCLA
31
A large number of problems arose

• Absorption mechanism and efficiency
• Weak m 1 mode and the Big Blue Mode
• Downstream density peak, axial ion flow
• Non-monotonic axial decay
• Triangular radial profile
• Mass-dependent density limit
• Low-field density peak (30G)
• Density jumps with increasing B0, Prf
• Endplate charging with small diameters
• Half-wave antenna better than full-wave
• High ion temperatures
• Parametric instabilities

UCLA
32
A density peak occurs at low B-fields
The cause is the constructive interference of the
reflected wave from a bidirectional antenna
UCLA
33
HELIC computations of plasma resistance
Vary the B-field
Vary the endplate distance
Uni- and b-directional antennas
Vary the with endplate conductivity
34
The end coils can also be turned off or reversed
to form a cusped B-field
The field lines then end on the glass tube, which
forms an insulting endplate. An aperture limiter
can also be added.
35
A cusp field or and end block can greatly
increase the density
G. Chevalier and F.F. Chen, Experimental modeling
of inductive discharges, J. Vac. Sci. Technol. A
11, 1165 (1993)
36
A large number of problems arose

• Absorption mechanism and efficiency
• Weak m 1 mode and the Big Blue Mode
• Downstream density peak, axial ion flow
• Non-monotonic axial decay
• Triangular radial profile
• Mass-dependent density limit
• Low-field density peak (30G)
• Density jumps with increasing B0, Prf
• Endplate charging with small diameters
• Half-wave antenna better than full-wave
• High ion temperatures
• Parametric instabilities

UCLA
37
Discharge jumps into helicon modes
n vs. B-field
n vs. RF power
R.W. Boswell, Plasma Phys. Control. Fusion 26,
1147 (1984)
UCLA
38
Transition to helicon mode
A.W. Degeling and R.W. Boswell, Phys. Plasmas 4,
2748 (1997)
UCLA
39
UCLA
A new interpretation of the jumps
The power into the plasma depends on the plasma
loading (Rp) and the circuit losses (Rc)
If Rp is too small, the input power is less than
the losses.
The jump into helicon mode can be computed from
theoretical Rps. The critical power agrees with
experiment.
F.F. Chen and H. Torreblanca, Plasma Sources Sci.
Technol. 16, 593 (2007)
40
A large number of problems arose

• Absorption mechanism and efficiency
• Weak m 1 mode and the Big Blue Mode
• Downstream density peak, axial ion flow
• Non-monotonic axial decay
• Triangular radial profile
• Mass-dependent density limit
• Low-field density peak (30G)
• Density jumps with increasing B0, Prf
• Endplate charging with small diameters
• Half-wave antenna better than full-wave
• High ion temperatures
• Parametric instabilities

UCLA
41
A 1-inch diam helicon discharge
UCLA
42
Critical field is where rLe a
UCLA
43
A large number of problems arose

• Absorption mechanism and efficiency
• Weak m 1 mode and the Big Blue Mode
• Downstream density peak, axial ion flow
• Non-monotonic axial decay
• Triangular radial profile
• Mass-dependent density limit
• Low-field density peak (30G)
• Density jumps with increasing B0, Prf
• Endplate charging with small diameters
• Half-wave antenna better than full-wave
• High ion temperatures
• Parametric instabilities

UCLA
44
Half wavelength helical antennas are betterthan
full wavelength antennas
L. Porte, S.M. Yun, F.F. Chen, and D. Arnush,
Superiority of half-wavelength helicon antennas,
LTP-110 (Oct. 2001)
45
A large number of problems arose

• Absorption mechanism and efficiency
• Weak m 1 mode and the Big Blue Mode
• Downstream density peak, axial ion flow
• Non-monotonic axial decay
• Triangular radial profile
• Mass-dependent density limit
• Low-field density peak (30G)
• Density jumps with increasing B0, Prf
• Endplate charging with small diameters
• Half-wave antenna better than full-wave
• High ion temperatures
• Parametric instabilities

UCLA
46
Anomalously high ion temperatures
Unusually high Tis are observed by laser induced
fluorescence. This happens near lower hybrid
resonance, but no special heating is expected
there.
J.L. Kline, E.E. Scime, R.F. Boivin, A.M. Keesee,
and X. Sun, Phys. Rev. Lett. 88, 195002 (2002).
47
A large number of problems arose

• Absorption mechanism and efficiency
• Weak m 1 mode and the Big Blue Mode
• Downstream density peak, axial ion flow
• Non-monotonic axial decay
• Triangular radial profile
• Mass-dependent density limit
• Low-field density peak (30G)
• Density jumps with increasing B0, Prf
• Endplate charging with small diameters
• Half-wave antenna better than full-wave
• High ion temperatures
• Parametric instabilities

UCLA
48
The energy absorption mechanism near the antenna
may be nonlinear, involving parametric decay of
the TG wave into ion acoustic waves
Lorenz, Krämer, Selenin, and Aliev used
1. Test waves in a pre-formed plasma 2. An
electrostatic probe array for ion oscillations 3.
Capacitive probes for potential oscillations 4.
Microwave backscatter on fluctuations 5.
Correlation techniques to bring data out of noise
B. Lorenz, M. Krämer, V.L. Selenin, and Yu.M.
Aliev, Plasma Sources Sci.Technol. 14, 623 (2005).
49
A helicon wave at one instant of time
Note that the scales are very different!
UCLA
50
Damping rate in the helicon afterglow
The damping rate increases with Prf, showing the
existence of a nonlinear damping mechanism.
UCLA
51
Excitation of a low-frequency wave
The LF wave is larger with the e.s. probe than
with the capacitive probe, showing that the wave
is electrostatic.
As Prf is raised, the sidebands get larger due to
the growth of the LF wave.
UCLA
52
Oscillations are localized in radius and B-field
The fluctuation power and the helicon damping
rate both increase nonlinearly with rf power.
UCLA
53
Proposed parametric matching conditions
k2
k1
k0
k0 helicon wave, k1 ion acoustic wave k2
Trivelpiece-Gould mode This was verified
experimentally.
UCLA
54
Evidence for m 1 ion acoustic wave
The cross phase between two azimuthal probes
reverses on opposite sides of the plasma.
kq is larger than kr, and both increase linearly
with frequency.
From the slope one can calculate the ion acoustic
velocity, which yields Te 2.8 eV, agreeing
with 3 eV from probe measurements.
UCLA
55
With a test pulse, the growth rate can be seen
directly
From probe data
From mwave backscatter
Growth rate vs. power
Growth rate vs. power
56
Conclusion on parametric instabilities
Kramer et al. showed definitively that damping
of helicon waves by parametric decay occurs near
the axis. They identified the decay waves,
checked the energy balance, and even checked the
calculated instability threshold and growth
rate. However, this process is too small to be
the major source of energy transfer from the
antenna to the plasma. It is still unknown what
happens under the antenna, where it is difficult
to measure. It could be that the waves observed
were actually created under the antenna but
measured downstream.
UCLA
57
Many problems have been solved, but some still
remain!

• Absorption mechanism and efficiency
• Weak m 1 mode and the Big Blue Mode
• Downstream density peak, axial ion flow
• Non-monotonic axial decay
• Triangular radial profile
• Mass-dependent density limit
• Low-field density peak (30G)
• Density jumps with increasing B0, Prf
• Endplate charging with small diameters
• Half-wave antenna better than full-wave
• High ion temperatures
• Parametric instabilities

Current project Commercialization of an
industrially viable helicon source using
permanent magnets for the dc field, and multiple
sources for large-area coverage.
UCLA
58
Title here