Design of PM helicon arrays UCLA 1.Optimization of the discharge tube 2.Design of the permanent magnets 3.Design of a multi-tube array 4.Design and construction of a test chamber 5.Antennas and the RF distribution system 6.Experimental results 7.Design of a compact module 8.Ideas for further improvements to be tested
A commercial helicon etcher (PMT MØRI) It required two heavy electromagnets with opposite currents.
Previous experiment with 7 tubes UCLA The “stubby” tube It required a large electromagnet
Plasmas merged; density is uniform UCLA High density and uniformity were achieved
Optimization of discharge tube: HELIC code UCLA D. Arnush, Phys. Plasmas 7, 3042 (2000). Radial profiles are arbitrary, but B and n must be uniform axially. HELIC gives not only the wave fields but also R, the loading resistance.
The HELIC user interface UCLA
The Low-field Peak UCLA
Mechanism of the Low Field Peak UCLA Basic helicon relations
The peak is sensitive to the density profile
The peak depends on the boundary condition
The peak depends on distance from endplate
The peak depends on the type of antenna Single loop: m = 0, bidirectional HH (half-wavelength helical): m = 1, undirectional Nagoya Type III: m = 1, bidirectional
Typical scan of R p vs n, B UCLA Each point requires solving a 4 th order differential equation >100 times. A typical scan takes ~ 3 hours on a PC.
Matrices for optimizing discharge tube UCLA Vary the tube length and diameterVary the RF frequency Vary the pressure and frequency Vary the endplate conductivity
Vary H (endplate distance) for 3” diam UCLA
Vary diam for H = 2” at 100G UCLA
Vary the frequency UCLA
Not much variation with pressure UCLA
Vary the endplate material UCLA Initially, it seems that the conducting endplate is better. However, it is because the phase reversal at the endplate has changed, and the tube length has to be ~1/4 wavelength longer to get constructive interference. By changing H, almost the same R can be achieved.
Relation of R to plasma density UCLA Rp << Rc
Relation of R to plasma density UCLA Rp > Rc
Final design UCLA
The “New Stubby” tube UCLA
Design of PM helicon arrays UCLA 1.Optimization of the discharge tube 2.Design of the permanent magnets 3.Design of a multi-tube array 4.Design and construction of a test chamber 5.Antennas and the RF distribution system 6.Experimental results 7.Design of a compact module 8.Ideas for further improvements to be tested
Characteristics of permanent magnet rings UCLA Internal field External field
The B-field of annular PMs UCLA The field reverses at a stagnation point very close to the magnet. Plasma created inside the rings follows the field lines and cannot be ejected.
Optimization of magnet geometry UCLA Result: Field strength magnet volume Spacing improves uniformity slightly actual
The field of 4 stacked magnet rings UCLA The internal and external fields at various radii. The individual rings can be seen at large radii. Calibration of the calculated field with a gaussmeter.
For the designed tube, B ~ 60G is good UCLA
Proof of principle on 3” diam tube UCLA External field Internal field
Radial density profiles at Z1 and Z2 UCLA Upper probeLower probe x cm -3 Proof of principle: discharge in the external field gives much more plasma downstream.
The final design for 2” tubes UCLA Material: NdFeB Bmax = 12 kG Attractive force between two magnets 2 cm apart: 516 Newtons = 53 kg The magnets are dangerous!
Wooden frame for safe storage UCLA
Single tube, final configuration UCLA Radial B z profiles at various distances below the magnet. Discharge tube
Design of PM helicon arrays UCLA 1.Optimization of the discharge tube 2.Design of the permanent magnets 3.Design of a multi-tube array 4.Design and construction of a test chamber 5.Antennas and the RF distribution system 6.Experimental results 7.Design of a compact module 8.Ideas for further improvements to be tested
Design of array UCLA The density at Z2 is summed over nearest tubes. Radial density profiles at Z1 = 7.4 cm and Z2 = 17.6 cm below discharge.
Computed uniformity n(x) for various y Half-way between rows1/4-way between rows Directly under a rowBeyond both rows
A tube spacing of 7” is chosen UCLA For a single row, a distance L = 17.5 cm between two tubes gives less than 2% ripple in density.
Design of PM helicon arrays UCLA 1.Optimization of the discharge tube 2.Design of the permanent magnets 3.Design of a multi-tube array 4.Design and construction of a test chamber 5.Antennas and the RF distribution system 6.Experimental results 7.Design of a compact module 8.Ideas for further improvements to be tested
An 8-tube linear test array UCLA Top view
Possible applications UCLA Web coaters Flat panel displays Solar cells Optical coatings A web coater
The array source is vertically compact UCLA The magnets can be made in two pieces so that they hold each other on an aluminum sheet. Once placed, the magnets cannot easily be moved, so for testing we use a wooden support. Side view Probe ports
The wooden magnet frame is used in testing UCLA
Water and RF connections UCLA These will be shown in detail later
An 8-tube staggered array in operation UCLA
Design of PM helicon arrays UCLA 1.Optimization of the discharge tube 2.Design of the permanent magnets 3.Design of a multi-tube array 4.Design and construction of a test chamber 5.Antennas and the RF distribution system 6.Experimental results 7.Design of a compact module 8.Ideas for further improvements to be tested
Antennas UCLA The antennas are m = 0 loops made of three turns of 1/8” diam copper tubing. The reason for m = 0 is that m = 1 antennas are too long, and much of the plasma is lost by radial diffusion before exiting the tube. The antenna must be close to the exit aperture and be tightly wound onto the tube. The helicon wave pattern for m = 0 is a peculiar one but theory is straightforward. The wave changes from pure electromagnetic to pure electrostatic in each half cycle.
The RF system UCLA The critical elements are the junction box and the transmission lines.
Antenna connections (1) UCLA The antennas must be connected in parallel with cables of equal length. The first trial was to use standard RG/8U cables and N connectors. These arced and overheated.
Reason that RF connectors don’t work UCLA High voltages and currents occur when there is NO PLASMA; that is, before breakdown or when the plasma disrupts. Then, if the RF power on a tube is set for 400W, the peak-to-peak voltage can be 5kV, and the rms current can be 12A. Once the discharge is on, the RF power goes into the plasma rather than into the cables and connectors.
Antenna connections (2) UCLA In the second try, all connections were solidly soldered, and RG/393 teflon- insulated cable was used. This method works for CW operation in experiment but may be marginal for industrial use.
Connections (3): a rectangular transmission line UCLA For w >> h, we find
Impedance for various pipe diameters UCLA For Z 0 = 50 , h ~ 3/4”, but exactly 50 is not necessary
Transmission line construction (1) UCLA
Transmission line construction (2) UCLA
Transmission line construction (3) UCLA
Transmission line (4): water connections UCLA No high voltage is applied along a water line.
Pictures (1) UCLA
Pictures (2) UCLA
Pictures (3) UCLA
Design of the matching circuit UCLA “Standard” circuit“Alternate” circuit Analytic formulas from Chen. The important part is that the impedance changes with cable length. F.F. Chen, Capacitor tuning circuits for inductive loads, UCLA Rept. PPG-1401 (unpublished) (1992); F.F. Chen, Helicon Plasma Sources, in "High Density Plasma Sources", ed. by Oleg A. Popov (Noyes Publications, Park Ridge, NJ), Chap. 1 (1995)
Adapted to N tubes in parallel UCLA The problem with array sources is that the cable lengths cannot be short. The match circuit cannot be close to all the tubes.
C1, C2 for N=8, L = 0.8 H, Z1 = 110 cm, Z2 = 90 cm UCLA
Variation with the number of tubes N UCLA Note that it is not possible to match to 1 or 2 tubes with the same length cables used for 8 tubes.
Design of PM helicon arrays UCLA 1.Optimization of the discharge tube 2.Design of the permanent magnets 3.Design of a multi-tube array 4.Design and construction of a test chamber 5.Antennas and the RF distribution system 6.Experimental results 7.Design of a compact module 8.Ideas for further improvements to be tested
Experimental layout UCLA Staggered configuration Compact configuration Four probe positions
Effect of B-field strength (magnet height D) UCLA Variation of density with D Variation of loading resistance with D
Variation with RF power and Ar pressure UCLA Variation of density with RF power Variation of density argon pressure
Density jump inside the tube UCLA compared with theory for various circuit resistances R c
Deployment of movable probe array UCLA
An linear array of 15 probes UCLA H. Torreblanca, Multitube helicon source with permanent magnets, Thesis, UCLA (2008).
Density profiles across the chamber (1) UCLA 03.5 Staggered configuration, 3kW Side Langmuir probe
UCLA 03.5 Compact configuration, 3kW Side Langmuir probe Density profiles across the chamber (2)
UCLA Density profiles across the chamber (3) Staggered configuration, 3kW Bottom probe array
UCLA Density profiles across the chamber (4) Compact configuration, 3kW Bottom probe array
UCLA Density profiles along the chamber (1) Staggered configuration, 3kW Bottom probe array
UCLA Density profiles along the chamber (2) Compact configuration, 3kW Bottom probe array
Calibration of the collector array UCLA
Design of PM helicon arrays UCLA 1.Optimization of the discharge tube 2.Design of the permanent magnets 3.Design of a multi-tube array 4.Design and construction of a test chamber 5.Antennas and the RF distribution system 6.Experimental results 7.Design of a compact module 8.Ideas for further improvements to be tested
A compact, 8-tube module UCLA
Stacked modules for large-area coverage UCLA
Match circuit fits on top of module UCLA
Design of PM helicon arrays UCLA 1.Optimization of the discharge tube 2.Design of the permanent magnets 3.Design of a multi-tube array 4.Design and construction of a test chamber 5.Antennas and the RF distribution system 6.Experimental results 7.Design of a compact module 8.Ideas for further improvements to be tested
Ferrites for better coupling UCLA The RF energy outside the antenna is wasted. Perhaps it can be captured with a ferrite cover.
Untested ideas UCLA One-piece ceramic tube Ferrite transformer coupling
Varying the magnet shapes and spacings UCLA Varying the ID and OD of PMs shows that B depends mainly on total volume of magnet. This shows that not much uniformity is lost if the magnet spacing is zero