The trion: two electrons plus one hole versus one electron plus one exciton

Abstract

We first show that, for problems dealing with trions, it is totally hopeless to use the standard many-body description in terms of electrons and holes and its associated Feynman diagrams. We then show how, by using the description of a trion as an electron interacting with an exciton, we can obtain the trion absorption through far simpler diagrams, written with electrons and \emph{excitons}. These diagrams are quite novel because, for excitons being not exact bosons, we cannot use standard procedures designed to deal with interacting true fermions or true bosons. A new many-body formalism is necessary to establish the validity of these electron-exciton diagrams and to derive their specific rules. It relies on the ``commutation technique'' we recently developed to treat interacting close-to-bosons. This technique generates a scattering associated to direct Coulomb processes between electrons and excitons and a dimensionless ``scattering'' associated to electron exchange inside the electron-exciton pairs -- this ``scattering'' being the original part of our many-body theory. It turns out that, although exchange is crucial to differentiate singlet from triplet trions, this ``scattering'' enters the absorption explicitly when the photocreated electron and the initial electron have the same spin -- \emph{i}. \emph{e}., when triplet trions are the only ones created -- \emph{but not} when the two spins are different, although triplet trions are also created in this case. The physical reason for this rather surprising result will be given.