Boundary conditions for a dense plasma edge

[griffinfarrow]

Hey! I'm trying to use EPOCH to model the edge region of some plasma. I'm happy with an MHD model for the central region, and I'd like to understand a little more how the MHD solution breaks down as the density is lowered at the edge.

I have density and magnetic field profiles like the following:

However, I'd like to only actually model the part between the two dashed lines using PIC.

I'm having problems with the boundary conditions. I see a drop off in magnetic field, strange features in the current density at the edges, and large heating of the plasma

I'm using simple_outflow conditions on the fields, and reflect boundary conditions on the particles
Does anyone have any idea what might be going on/what BCs might be more physical?