Helicon Plasma Sources

The helicon plasma source is a very efficient method for generating high density plasmas using radio waves. Such plasmas can be very useful for basic plasma studies, because there is no large electric current running through the plasma that can disturb the phenomenon you're trying to study. This is especially i important for studies of Alfvén waves. Moreover, the helicon sources on ALESPI and HelCat are unique in that they are steady state (as opposed to pulsed), meaning there are no transient phenomena. Another nice feature of helicon sources is that you can vary the ionization fraction, or the number of neutral particles in the plasma. Neutral particles in the plasma can complicate measurement results, and may negate the effect you're trying to simulate in the lab. On the other hand, there are important effects that neutral particles cause, again in the properties of Alfvén waves.

Helicon sources were originally used in plasma processing (computer chip manufacturing), but have now found their way into laboratory research of both fusion and space plasmas. Typically, helicon sources produce high density (>10¹⁹ m^-3), low temperature (~10 eV = 100,000° C) plasmas with a cross-sectional area of 10-100 cm². They require very little power, typically a few hundred watts, to achieve these high densities. It's not exactly clear why they are so efficient at ionizing a plasma. However, it my be due to Landau damping in the "core" region of the discharge (see the figure below).

Helicon waves are right-hand circularly polarized electromagnetic waves propagating along the magnetic field. They are bounded whistler waves that lie on the fast branch of the cold plasma dispersion relation, propagating in the frequency range w_ci<<w<<w_ce. In cylindrical geometry, for wave propagation along the magnetic field, the solution to the eigenmode equations is an mth order Bessel function of the first kind. For wavelengths much greater than the antenna diameter a the lowest Bessel root for the m=1 mode yields

Here k is the wavenumber, B is the background magnetic field, e is the electron charge, and n_e is the electron density. Although it has been claimed that k is set by the antenna length, our studies indicate that the plasma chooses a particular mode more or less independently of the antenna length. All that changes is the matching circuit tuning, while density remains essentially unchanged.

To create the helicon plasma, one typically uses a special helical twist (m=1) antenna wrapped around a pyrex tube to insulate the antenna from the plasma. In theory the antenna could be in the plasma, though in practice this rarely works. The antenna is connected through a matching network, typically in the pi configuration with the antenna as the inductor, to the radio frequency (RF) amplifier. The ham radio band is ideal for most applications, and at UNM we use frequencies from 5 to 30 MHz, though frequencies near the lower hybrid resonance work best. Argon is the easiest gas to work with, though we have also used helium, and other groups have reported success with hydrogen. The fill pressure in the chamber is a few mTorr. A background magnetic field of several hundred gauss is also required. With a few hundred watts of RF input power the helicon mode can be excited, characterized by a bright blue core in argon, or a pink core in helium. At the right is a view through the viewport of the helicon mode in argon.

Electron Cyclotron Heating of a Helicon Source

We are investigating the efficiency of electron cyclotron heating (ECH) in increasing the electron temperature of the helicon source on ALESPI. Standard ECH cannot work in these discharges because the cyclone frequency is cut off due to the high density of the helicon discharges. So, we are trying a novel technique where we launch whistler waves, which do propagate in the plasma at frequencies up to the electron cyclotron frequency. As discussed with respect to helicons above, these are right-hand electromagnetic waves with a phase velocity maximum for propagation along the magnetic field, which goes to zero for propagation perpendicular to the field. These waves are resonant at the electron cyclotron frequency, suggesting a scenario for heating.

We plan to exploit the very modest magnetic field gradient due to the imperfect solenoid of the confining field; the field decreases slightly both as one moves away from the field coils towards the device axis, and as one moves along the axis from the device center. Right hand waves are launched from the edge, oblique to the field at a frequency below the local cyclotron frequency. These propagate up to the point in the plasma where the magnetic field has dropped to a value such that the waves are resonant, a sort of inverse "magnetic beach". This heating scenario is akin (though opposite) to the heating of fusion plasmas using ion cyclotron waves. The heating technique has been shown effective in another linear device with a non-uniform magnetic field. Because the gradient is slight, we can use a modestly broadband source to heat the plasma over an extended area. Initial modeling suggests that temperatures of a few 10s of eV up to 100 eV are achievable. This depends on the power of the heating sources, and more importantly, on the end losses from the device.

On ALESPI a 200 W ECH system is installed operating at 2.45 GHz - basically a microwave oven. We used an adjustable waveguide system to launch plane polarized waves at angles between 15° and 90° to the axial background magnetic field. In a nutshell, we failed!

First, we confirmed basic physics. At 90° launch angle there is negligible heating of the plasma - the ECH is cutoff. However, at small launch angles we noticed significant cooling. The temperature in the core of the plasma dropped by up to 20%. Any suggestions as to what might be going on?!

Return
to Plasma & Fusion Sciences main page

Return

to Plasma Sciences Laboratory