Laboratory Study of Hall Reconnection in Partially Ionized Plasmas
Physical Review Letters
The effects of partial ionization (n i =n n & 1%) on magnetic reconnection in the Hall regime have been studied systematically in the Magnetic Reconnection Experiment. It is shown that, when neutrals are added, the Hall quadrupole field pattern and thus electron flow are unchanged while the ion outflow speed is reduced due to ion-neutral drag. However, in contrast to theoretical predictions, the ion diffusion layer width does not change appreciably. Therefore, the total ion outflow flux and the
... utflow flux and the normalized reconnection rate are reduced. Magnetic reconnection is a fundamental process in many naturally occurring and man-made magnetized plasmas. However, these plasmas are often weakly ionized: The lower solar atmosphere, protostellar and protoplanetary disks, galactic molecular clouds, and tokamak edge plasmas can have typical ionization fractions of 0.1%-1% or lower. Reconnection is of particular importance in the solar atmosphere, as it is thought to be a trigger and/or a driver for eruptive events that originate in the weakly ionized photosphere and chromosphere  but couple to the hot, fully ionized corona  . The effects of weak ionization are often neglected but can sometimes be modeled as single fluid with a modified resistivity or diffusion    . These models depend on the ion-neutral collisional mean free path being shorter than all relevant length scales. However, reconnection can have a large range of scales, so this approximation may not always be valid. For instance, coronal observations  have shown that reconnection features can break down into plasmoids [7, 8] that are much smaller than the overall system size. NASA's soon to be launched Interface Region Imaging Spectrograph (IRIS) mission  will return highly detailed observations of these regions, so an understanding of how reconnection works in partially ionized plasmas is of timely importance. While the smallest of these scales may still be beyond what is currently observable, an understanding of the underlying physics is vital for interpreting and benchmarking the simulation models used to analyze the observations. Collisions with neutrals can have a number of important effects on the reconnection dynamics. Elastic electronneutral collisions increase the plasma resistivity, which is ultimately how the change in magnetic topology occurs, at least in collisional plasmas. The effects of ion-neutral collisions are more subtle. Because of the comparable particle masses, a very strong ion-neutral drag force will act like static friction, and they will behave as a single fluid with an effective mass equal to the total mass density, ¼ n þ i . Swanson et al.  previously observed this effect for hydromagnetic Alfvén waves. A recent theoretical study [11, 12] treated reconnection in partially ionized plasmas and found a similar result. In particular, the Alfvén speed, v A ¼ B= ffiffiffiffiffiffiffiffiffiffiffi 4 i p , and ion inertial length, , are modified to use instead of i , or some fraction of if the coupling is moderate. The ion inertial length is especially relevant to fast, Hall-mediated reconnection, as it sets the length scale on which such a reconnection can occur  . This model is of particular interest since it uses a multifluid approach where ion and neutral dynamics are treated separately instead of relying on a single-fluid Pedersen-Cowling resistivity or ambipolar diffusion. In this Letter, we present the first study of the effects of partial ionization on Hall reconnection in a laboratory plasma. Starting with conditions where neutral effects are nearly unimportant as a base case, we decrease the ionization level while keeping the plasma parameters fixed. We find that the outflow speed is reduced as previously predicted  but we do not observe corresponding changes in the ion layer scale size. Nonetheless, we find these observations to be consistent with the observed reconnection rate scaling and discuss possible reasons for the discrepancies with the model. These experiments were performed at the Magnetic Reconnection Experiment (MRX) facility  . A crosssectional schematic of the cylindrical device near the reconnection region is shown in Fig. 1 . The large gray circles represent the two toroidally shaped flux cores. Each flux core contains two sets of coils: the poloidal field (PF) coils produce the initial X-point geometry, while the toroidal field coils produce the inductive electric field that breaks down the prefilled working gas (helium in these experiments). These experiments are performed in the "pull" reconnection regime, where the plasma is produced while PF current is decreasing and magnetic field lines are pulled towards the flux cores. This produces quasisteady reconnection that lasts for $40 s ($12 A ). In this geometry, plasma and field lines flow into the reconnection region along the R direction and outflows are in the Z direction. The current sheet is then in the toroidal (T) direction.