M. Niimura1), M. Lamoureux2), A. Goto1), and Y. Yano1)
1)Cyclotron Laboratory, RIKEN, 2-1 Hirosawa, Wako City, Saitama 351-0198 Japan.
2)University of P. & M. Curie, 4 place Jussieu, 75252 Paris, Cedex 05, France.
ECR ion source (ECRIS) has established its reputation in the long-term stability. As an injector for cyclotron and RFQ accelerators, ECRIS can deliver a beam of the highly charged ion (HCI) perpetually on the order of 10 days without a problem. For recent years, however, we haveperceived a short-term (on the order of 1 sec) instability in the HCI beam intensity and diamagnetic signal. Recently, an instability at the frequency of a few Hz was observed with a max-B ECRIS when operated in pure nitrogen . Even under a gas-mixing condition, Lamoureux and coworkers have observed a similar oscillation on the order of 1 Hz in various ions using Quadrumafios min-B ECRIS, and discovered an increase in the thresholdmicrowave (rf) power above which HCI intensity starts oscillating under a fixed gas pressure . A linear relationship was evident for the threshold rf-power with respect to the gas pressure; but different slope for different ion. Further, they have shown that the extracted-ion beam current is maximum whenever the rf-power becomes marginal to the threshold of theinstability.
In this contribution we introduce a rational model, which could explain the instability and all related phenomena mentioned above. We also show the ways to stabilize as well as to optimize the system remotely before and after the instability: possibility of a remote control. Mostsignificantly, however, we can declare that the observed instability [1,2] is the interchange instability recorded for the first time in ECRIS plasma, and that this observation is the first direct evidence to prove the presence of a hot-electron ring (HER) in the min-B as well as max-B ECRIS. The consequence is rather serious: it implies a defect in the stochasticheating (STH) model to explain the HER in the min-B ECRIS. The STH model was thought to have explained the HER in the max-B ECRIS because the second-harmonic resonance exists around the midplain; but no such resonance exists in the min-B ECRIS. This validates the cross-field heating (CFH) model , role of which was found complementary for STH model.
High rf-power and high-vacuum are favorable conditions for HCI production, but have a risk to bring the system marginal to the interchange instability that promotes a chaotic radial transport (or loss) of energetic (or hot) electrons. Fortunately, however, this is not a destructive instability for the case of ECRIS plasma, because the system can return to the stable regime, whenever the hot-electrons decrease its concentration by spitting themselves out of the HER or if the wall-sputtering and/or gas-mixing can increase the concentration of neutrals (n), so that the condition given by eq. (49) of ref.  is satisfied. The stability condition predicts that the hot electron density (or rf-power) has to be increased linearly with the gas pressure (~ n) and with mass-to-charge ratio (M/Z). This agrees with experimental results , because the M/Z of Xe30+, Ar6+, and Kr7+ increase as 4.5, 6.7, and 12.3, respectively. This is a proof of the interchange instability that assumes presence of a HER. Thus, a stable HER (cohabitant with high-Z ions) can form only at a higher pressure. Of course the HCI component tends to decrease, while the total beam current increases, with pressure. This can explain what were observed with the ECRIS programmed to study the second-Bernstein (B2) mode heating,or CFH, scheme .
 D. Meyer et al., Proc. 13th int. workshop on ECR ion sources, Texas USA(1997) p.72.
 M. Lamoureux et al., ibid, p.67 and p.211.
 M. Niimura et al., ibid, p.52.
 E. Tojyo et al., BEAMS1998, Nov. 25-26, Tokyo (1998).
Presenting Author : Niimura M.
Presentation : Oral