Dear All My name is Ram. I'm a grad student in Univ of Michigan who is interested in understanding electronic transport in nanoscale systems. I recently started to use TRAN-SIESTA. I have read through the paper on Transiesta [Phys. Rev. B 65, 165401 (2002)]. I have certain questions on the contour integration of Greens function. The green's function of just the Hamiltonian has poles below the real axis and very close to the real axis. This is true because the Hamiltonian is a Hermitian operator. But for the green's function of the device with the self energies, this might not be the case, since the self energies are not hermitian (or in the simplest wide band approximation, they are purely anti hermitian). This leads the greens function to have poles anywhere in the complex plane. So does the idea of complex contour work in this scenario? Further the other question that I had was a-priori error estimation of the integral. Is is possible to calculate the order of error that we could be making in the numerical integration.
It will be of great help if someone could answer me these questions as I could not find answers to the same anywhere in literature. Regards Ram <http://sitemaker.umich.edu/ram/home/>
