Continuous Wall Conditioning in a Toroidal Device without Magnetic Field Using
VHF Discharge
V.E. Moiseenko, A.V. Lozin, V.V. Chechkin, D.I. Baron, L.I. Grigor’eva, Ye.D. Kramskoi, V.B. Korovin, M.M. Kozulya, A.N. Shapoval, V.M. Lystopad,
A. Yu. Krasuk, M.B. Dreval, R.O. Pavlichenko and Uragan-2M team
National Science Center ‘‘Kharkiv Institute of Physics and Technology”,
Akademichna St. 1, Kharkiv, Ukraine
9th IAEA TM SSO, Vienna, 20-23.03.17
IntroductionThe radio -frequency (RF) discharges are used for conditioning of inner walls of
vacuum chambers of fusion devices. They are an essential part of operation of steady-state fusion machines. The aim of the wall conditioning is to remove impurities from the walls which could appear there for different reasons (opening the vacuum chamber to the air, operation of vacuum locks, sputtering of material in the divertor area, etc.)
The wall conditioning is achieved due to interaction of chemically active atoms generated in the discharge with the impurities accumulated at the wall surface.
The wall conditioning discharge can be created by the plasma heating tools, ECRH and ICRH. In this case their operations are usually not very efficient since the load on the generator made by the wall conditioning discharge is too different from the load of heating discharges.
Usage of separate tools for wall conditioning is more efficient. But in this case the most important factor is their compactness.
Continuous VHF discharge
The very high frequency (VHF) discharge for wall conditioning with hydrogen atoms is studied in Uragan-2M device. It is driven by the RF power at frequencies ~140 MHz, higher than usually used in ICRF. For wall conditioning a special small size antenna is designed. The antenna is aimed to excite the slow wave that is damped via electron collisions with neutral gas. The discharge is volumetric: plasma occupies whole confinement volume and even steps out at the edge. The characteristic value of plasma density is 1010 cm-3, electron temperature varies in the range 3-10 eV. The temperature values of probe measurements are compatible with the results of optical diagnostics. Such parameters of discharge are favorable for wall conditioning in hydrogen.
A regime of RF wall conditioning without the magnetic field is also useful for fusion devices. In present report it is studied using the same experimental environment.
Plasma without magnetic field does not slow down the electromagnetic waves. To achieve acceptable damping of the wave, high electron-neural collision frequency is needed. Thus, the RF discharge may be sustained at relatively high neutral gas pressure.
The wall conditioning is achieved due to interaction of atoms generated in the discharge with the impurities accumulated at the wall surface. The intensity of the generation of atoms is proportional to the product of neutral gas pressure and plasma density. Thus, to keep the same rate of atoms generation, at higher neutral gas pressure, the plasma density should be lower.
Low plasma density is a positive factor since the probability of ionization of an impurity desorbed from the wall is small.
Uragan-2M device
Major radius R=1.7 mMinor plasma radius – a 0.24m Toroidal magnetic field - B0 2.4 ТFor K=0.31 (r=0)=0.34, (r=a)=0.47Pumping rate 2x400 dm3/s
As well as in U-3M plasma production and heating in U-2M is performed by the RF heating.
RF generators:
2x200 kW f=5…15 MHz
1MW f=25…40 MHz
1kW f=3…8 MHz
2kW f=130…150 MHz
Antennas and experimental setup
Water cooled small frame antenna for wall conditioning
Small frame antenna for wall conditioning
Experimental arrangements:RF generators P1 kW, f=8.4 MHz,
P2 kW, f=135 MHz,
Diagnostics:
Spectrometry
Probe measurements
Residual gas analysis with mass spectrometer
Cryogenic nitrogen cooled trap
Small antenna position
RF discharges for wall conditioning in weak magnetic field
Two waves exists in cold plasma:
Fast magnetosonic (does not propagate because of low plasma density)
Slow wave can propagate if
or The latter condition provides upper limit to the plasma density.
Slow wave
•Is damped efficiently owing to the electron-neutral collisions
•Can be short-wavelength
V.E. Moiseenko, et al, Nucl. Fusion 54 (2014) 033009
RF discharges for wall conditioning without magnetic field
Two waves exists in cold plasma:
electromagnetic wave
Langmuir (kinetic) wave
and both of them are non-propagative. Thus, plasma is created by the near-field of the antenna.
Continuous VHF discharge without magnetic field
Plasma exists only in vicinity of the small frame antenna.
Plasma cloud size increases with the input RF power.
Luminosity decreases by an order of magnitude at the distance of 1 m from the antenna.
Discharge view from upper window
In plasma without magnetic field the electromagnetic wave is evanescent. The kinetic wave (Langmuir wave) is damped shortly if excited. To achieve acceptable damping of the wave, high electron-neural collision frequency is needed. This is provided by relatively high neutral gas pressure.
Plasma discharge spectra for different working gases.
The spectrum of the optical emission for hydrogen, nitrogen and each hydrogen-nitrogen mixture.
Te≈2÷3 eV (estimated
with H2 Fulcher-α band system and Hα intensities) for hydrogen discharge.
N2+ line with the
wavelength 427.8 nm indicates ionization of nitrogen and the hydrogen spectrum indicates hydrogen dissociation processes.
Langmuir Probe Measurements
The Langmuir probe was placed in the small frame antenna cross-section at horizontal midplane at outer part of the torus and moved along radius.
The Langmuir probe potential was varied from -150 to +100 V to take I-V characteristics. Every probe potential was fixed for 2-3 discharges.
All measurements were made in the probe position where the Langmuir probe floating potential was the highest.
Parameters mixture of gases
H2 H2(50%)+N2(50%) H2(<10%)+N2(~90%)
Te, eV 2.7 4.1 1.78
by H2+ if mi=2 a.m.u. ne 5.1∙109 cm-3 1.81∙109 cm-3 ---
by N2+ if mi=28 a.m.u. ne --- 6.77∙109 cm-3 3.23∙109 cm-3
Measurement results are represented as a table:
The pumped gas was condensed on the liquid nitrogen trap inner surface during the RF wall conditioning;
The trap was cut off from the vacuum chamber by two vacuum valves and defrosted;
The closed volume around the trap was filled with the gas evaporated from the trap surface.
Cryogenic trap usage for wall conditioning monitoring
• The pressure increased in the cut off volume as the temperature raised and condensate evaporated from the trap surface.
• The increment of gas pressure Pg in the closed trap volume is proportional to the gas amount pumped from the vacuum chamber and condensed at the liquid nitrogen trap.
• The background level of the vacuum chamber outgasing was determined first, and the amount of condensed gas was measured when there was no wall conditioning and the vacuum chamber was filled with the working gas.
Liquid Nitrogen Trap Measurements
Background pressure in trap volume at the beginning of wall conditioning after 5-minute exposition. No VHF.
1.background level during hydrogen pumping, P
chamber=3∙10-2 Torr.
2.gas mixture: 50% hydrogen + 50% nitrogen, P
chamber=3∙10-2 Torr.
3.background level during nitrogen pumping, P
chamber=2.5∙10-2 Torr.
Pg values comparison for wall conditioning in 1. Hydrogen 2. 50/50% (H2/N2) 3. 10/90% (H2/N2).Exposure time was 5 minutes. Note here that Pg =3.810-2 Torr after exposition of the chamber by hydrogen with pressure 2.9 10-2 Torr and with no VHF. The measurements done in the middle of the wall conditioning period of time.Gas mixtures provide 4 times more intensive volatile substance production than pure hydrogen discharge.
Liquid Nitrogen Trap Measurements
Presidual evolution of residual vacuum chamber
pressure during wall conditioning.
Presidual reached the minimal possible (for certain
conditions) value 6.2∙10-2 Torr in 25 first hours of described wall conditioning regime.
Presidual hasn’t changed until the end of campaign.
Pg measurements during RF wall
conditioning
1. at the beginning of experimental campaign;
2. after 45 hours of RF wall conditioning.
At the beginning Pg value is bigger by an order of magnitude than at the end of wall conditioning campaign.
Summary
The radio-frequency wall conditioning without the magnetic field is performed in continuous regime at the frequency 130 MHz and launched power about 3 kW.
The discharge can exist in high gas pressure 0.1 – 0.01 Torr, is located near the antenna and does not spread around the torus. Its parameters are measured using the Langmuir probe.
The effect of wall conditioning is judged by the amount of substances accumulated at the cryogenic vacuum trap. This amount appears by the order of magnitude higher than without the discharge which indicates obviously the wall conditioning.
Hydrogen, nitrogen and their mixtures had been tried as working gases. The wall conditioning in the mixture 50%/ 50% is selected as the best.
The VHF discharge can be used for efficient wall conditioning both with and without magnetic field.