Leap frog integrator modifications in highly collisional particle-in-cell codes

Leap frog integration method is a standard, simple, fast, and accurate way to implement velocity and position integration in particle-in-cell codes. Due to the direct solution of kinetics of particles in phase space central to the particle-in-cell procedure, important information can be obtained on particle velocity distributions, and consequently on transport and heating processes. This approach is commonly associated with physical situations where collisional effects are weak, but can also be profitably applied in some highly collisional cases, such as occur in semiconductor devices and gaseous discharges at atmospheric pressure. In this paper, we show that the implementation of the leap frog integration method in these circumstances can violate some of the assumptions central to the accuracy of this scheme. Indeed, without adaptation, the method gives incorrect results. We show here how the method must be modified to deal correctly with highly collisional cases.