Spill-in terms in negf

[qm1984]

Dear ATK team,

In the NEGF  section in the manual  it is shortly stated that the density matrices D_LL,D_LC, ... are included correctly. I there any reference how does is done ?

Best,
Marius

[kstokbro]

We have not published this yet. ATK use an LCAO description, and the challenge is that many basis orbitals in a device have tails outside the central region, or basis orbitals in the electrode has tails inside the central region. All these tails must be included correctly in the DFT implementation, and this is what we call spill in.

ATK is one of the only codes where this is done properly, and this is why we can handle very efficiently heterogenous systems, like the interface between Ni-graphene or metal-organic interface

K. Stokbro, M. Engelund and A. Blom, Atomic-scale model for the contact resistance of the nickel-graphene interface, Phys. Rev. B 85, 165442 (2012).

K. Stokbro and S. Smidstrup, Electron transport across a metal-organic interface: Simulations using nonequilibrium Green's function and density functional theory,  Phys. Rev. B 88, 075317 (2013).

[qm1984]

Thank you for your reply. To bad that there is no reference so far.  So what you are saying is that there is a fine contribution to the charge density n(r) in C due to basis functions in L and R ?  This additionally complication basically originates in the non-orthogonal basis set used. Is this correct ?