How does EM heating cause motion? Similar to how does heating cause motion, I'm trying to understand how a photon imparts motion to an atom, i.e. adds heat to a gas.
I'm going to hazard a guess, and suggest this occurs something along the lines of: 
An electron bound to a nucleus absorbs a photon (at a compatible frequency), which causes excitation-return to ground state electronic transition. 
I'm guessing* that its this absorption/re-emission(or both) process which imparts momentum. 
Can someone correct me/please explain this process in more detail.
(* I can see some issues with this already, but one thing at a time)
 A: The first point is that a photon carries momentum, so anything that absorbs it acquires that momentum. So after absorbing a photon a gas molecule/atom will have an increased momentum.
If the atom simply re-emits the photon then the momentum of the atom goes back to what it was before and nothing has changed. Heating occurs when the excited atom collides with some other atom/molecule before it can re-emit a photon, and transfers the extra energy to relative motion of the two atoms. The end result is that the energy in the photon gets converted to kinetic energy of gas molecules.
Response to comment:
In a collision the excited atom may relax into the ground state or another, lower energy, excited state. In the latter case a photon of lower energy could be emitted. I don't know the transition probabilities offhand, but I would guess the collision mostly transfers all the energy of the excited state into kinetic energy and no photon would be re-emitted.
Two atoms in the ground state can collide and electronically excite one of the atoms. If the atom then emits a photon the end result is that the kinetic energy is converted back into the energy of a photon. This is how a gas radiates heat and cools.
A: Another way of looking at this is from the thermodynamic equilibrium point of view. Saying that a gas is at a certain temperature implies something called "local thermodynamic equilibrium" (LTE). The populations of atoms in their various energy levels or even ionisation states is determined by this temperature through the Boltzmann factors $\exp(-E/kT)$. The speeds of the gas particles are also given by a Maxwell-Boltzmann distribution at the same characteristic temperature.
If the gas is in equilibrium at this temperature, then all radiative and collisional processes are in detailed balance. That is, the rate at which atomic energy levels are excited/de-excited is the same, the rate of ionisation and recombination is the same and the radiative and collisional processes are in balance separately. So, on average, nothing changes.
Now, your scenario is that you introduce a new radiation field into the picture (e.g. illuminating the gas from one side). If the temperature is to increase then by definition the gas is not in LTE any more and detailed balance does not apply. That is, there are more photon absorptions than re-emissions, the re-emission is (probably) isotropic and the average speed of the atoms is increased as a result and as a result they gain a net momentum in the direction of the incident radiation. Furthermore the ionisation balance in the gas may change so that more atoms become ionised and those electrons can then collisionally interact elsewhere too.
Eventually a new equilibrium state will be reached if all the collisional and radiative processes come into balance again, which will happen at a new, higher temperature.
