Experimental Evidence of a Double Layer in a Large Volume Helicon Reactor

International audienceThe self-consistently generated current-free electric double layer (DL) is shown to scale up with the source tube diameter and appears not to be affected by rf driving frequency and changes in reactor geometry. This Letter presents the first simultaneous measurements of local plasma potential and beam energy as a function of axial position. The DL is shown to be no more than 5 mm thick (20 D lengths) and positioned just downstream of the maximum in the magnetic field gradient. Furthermore, its position relative to the magnetic field is observed to be invariant as the magnetic field is translated axially. Measurements of the potential drop across the DL are presented for pressures down to 0.09 mTorr and the DL strength (DL =T e) is determined to be between 5 and 7. Electric double layers (DLs) are observed in nearly col-lisionless plasmas and are spatially limited changes in potential , which often separate two different types of plasma. They are important as energy dissipation mechanisms and have been associated with the acceleration of charged particles along magnetic fields. Several authors have invoked DLs to explain astrophysical phenomena including reconnection events, solar flares, and aurora [1,2], as these events require the rapid change of electromagnetic energy into particle energy. Numerous satellite based experiments have revealed the existence of strong ''shocklike'' variations in potential and corresponding beams of high energy particles in the magnetosphere [3–5]. More recently, the electric field embedded in the DL has been proposed as a thrust producing mechanism for space propulsion especially in light of the relative simplicity and anticipated reliability of the underlying technology [6]. Large diameter beams of low energy ions are also of considerable interest to the surface treatment and metrol-ogy community for large-scale processing of surfaces. Since in situ measurements are difficult and costly in space, much of the current knowledge on DLs has been guided by laboratory experiments [7–10], theoretical work [8,11], and computer simulation [12 –14]. Historically, experimental regimes have fallen into two main categories, namely, double or triple plasma devices, which are voltage driven [15,16], and current driven discharges with an abrupt change in diameter [17], where the common thread was the requirement of an artificially induced current to drive the DL. However, by carefully adjusting the potentials on immersed grids, it has been shown that it is the flow of ions and electrons rather than the applied potentials in triple plasma devices which create the DL [18]. Previous research on the free expansion of plasmas has also demonstrated the existence of a current-free double layer [19] and explained its physics based on a two-electron population. DLs have also been measured in electronegative plasmas in the absence of a magnetic field [20]. More recently, interest has turned towards the effect of inhomog-enous magnetic fields on the dynamics and stability of DLs [21,22]. In these experiments, the DL is created independently (by some form of current drive) and the magnetic field is used to control its position. In 2003, it was demonstrated that a current-free double layer could be created self-consistently using a helicon source and an expanding magnetic field [6]. This work was significant because it presented the first laboratory demonstration a self-generating current-free stationary electric DL and showed the important role of an expanding magnetic field in the self-consistent formation of the structure. Subsequent studies akin to this one confirmed the presence of an ion beam under similar experimental conditions [23,24] using laser induced fluorescence; however, Sun et al., in particular, reported that it was the neutral pressure more than the magnetic field which had the most influence on the energy of the beam ions and that the DL thickness might be as high as 500 D lengths [23]. This was in stark contrast to Ref. [6], in which, below a certain threshold neutral pressure, the magnetic field was the most sensitive parameter and where the DL thickness was reported to be on the order of 50 D lengths [6,25]. In this Letter we answer three critical questions about the DL first described in Ref. [6]; namely, we make simultaneous measurements of the plasma potential and ion beam, confirming both the presence of an ion beam and the existence of a strong electric structure, we demonstrate that the thickness of the DL is on the order of 20 D lengths, and we show that the DL is tied to the magnetic field at a position close to the maximum in the gradient and half the maximum of the field. In addition, we have extended the study to lower pressures and have demonstrated that the DL scales up with source and diffusion chamber diameter. The large volume helicon diffusion system WOMBAT (waves of magnetized beams and turbulence) is described extensively elsewhere [26], but briefly consists of a 20 cm diameter, 50 cm long Pyrex source tube connected co