the THz beam has two efficiencies to think about:
1. is electrical efficiency of the THz emitters; this is a nano-klystron array consisting of thousands of cells, focusing on the air in the reaction chamber and expansion tubes;
2. is absorption of THz energy into propellant. we’d want to pick an absorption line in air that is wide enough to absorb well, through delta-v of ±7.8 km/s or more and temperatures from 240 kelvin and up. like other scramjet designs, we’d consider a 5-species model consisting of N2, O2, NO, O+, N+. If absorption tends to be low, long expansion tubes may help.

I tried to keep the temperatures down to under 2000 K, to avoid NOx, but couldn’t.

Using plasma windows on ducting could have an advantage of fewer moving parts.