Three-band s± Eliashberg theory and the superconducting gaps of iron pnictides

Superconductivity in iron pnictides has been recently proposed to be due to spin-fluctuation mediated coupling between hole and electron bands with order parameters of opposite sign. BCS multiband models give qualitative predictions but cannot simultaneously reproduce the experimental values of critical temperature and energy gaps. We show, instead, that a three-band strong-coupling Eliashberg model can quantitatively reproduce the gaps and their temperature dependence in both the 1111 and 122 families. We also show that this requires small typical boson energies (in agreement with experiments) and high values of the electron-boson coupling constants.