This is the reason why most fusion approaches rely on thermal systems. In a thermal system, the ions have a bell-shaped distribution of energies and undergo many collisions before they leave the region in which they are confined and their energy leaves the system.
To achieve net gain, the temperature, density and energy confinement time must be above a certain threshold. If the system is non thermal, like a stream of hydrogen ions where the distribution of energies is a spike, the energy in the hydrogen ions that are deflected by glancing blows must be recaptured somehow.
This is what's wrong. The energy required to accelerate the ions is much higher than the energy which can be harvested this way.
The concept under discussion substitutes a simpler setup that accelerates particles using laser induced plasmas from a very small table top laser. The thing could "almost" be battery powered.
- Fire a stream of hydrogen ions with a particle accelerator at a chunk of Boron 11. - The hydrogen and boron combine and release heat and helium. - Use the heat from that to run a turbine and keep running your particle accelerator.
It seems like you would be ending up with lower-energy collection of atoms. Does that work but it is just not efficient enough to keep running the accelerator, or what?