operator splitting and time integration schemes

[spacedout]

Good evening

If I decompose a PDE into a sequence of

dF/dt = some homogeneous term
dF/dt = some reaction term

should I use an an implicit or explicit time discretization scheme for each of these simplified PDEs ?

Best to all

[spacedout]

I got astronomical overflows after only 3e-8 sec for explicit euler and after only 6.5e-8 sec for implicit euler

I guess Elmer doesn't like my 3 ion continuity equations