For example, say a collision between a neutron and a ionized helium atom.
This is a bad example , because 1) it is not a diminution of the macroscopic case of inelastic collisions.2) neutrons are neutral and have very small probability of interaction
Does the energy always go into excitation of the other atom?
No, take the simple case of an electron scattering off a neutral hydrogen atom.
There will be a probability for
scattering elastically in the center of mass, in the lab frame where the hydrogen is assumed at rest kinetic energy will be acqured by the atom . The probability is small since the atom is neutral
Enough energy transferred to the atom for the electron to go to a higher excited state, and decay back to ground state by emitting a photon with the energy difference
the electron loses enough energy for the hydrogen atom to be ionized and its electron leaves with some energy
(this answer of mine has bubble chamber pictures that illustrate part of this)
Collisions between atoms at higher energies than the ones studied classically,may transfer energy as in 2. and 3., replacing the incoming electron of the argument, with another atom. 1. is the usual scattering studied with classical statistical mechanics.
This is at the atomic level, where all interactions are due to the electromagnetic field. Scatterins of neutrons belongs to the study of strong nuclear interactions.