Bypassing the Carnot Limit
Because DEC converts ordered particle motion rather than random thermal energy, its efficiency ceiling is set by particle optics, not by a temperature ratio.
Ordered versus random energy
Thermodynamics distinguishes ordered energy (a directed flow, fully convertible to work) from random thermal energy (whose convertibility is limited by entropy). A turbine extracts work from a hot gas whose molecules move in every direction; the second law forces it to dump a large fraction as waste heat. The charged products of a fusion reaction, by contrast, stream out along the magnetic field with a well-defined energy — much closer to ordered energy than to a hot gas.
Why Carnot does not apply
The Carnot efficiency, 1 - T_cold/T_hot, is a statement about heat engines cycling a working fluid between two reservoirs. DEC never forms such a reservoir. Instead it lets each ion climb an electrostatic potential hill: an ion of charge q and energy E that is brought to rest against a retarding potential V delivers charge q at voltage V, and if qV is matched to E the work recovered approaches E itself. The ceiling is therefore how well the field is matched to the particle spectrum, not a temperature ratio.
The catch: spectrum spread
Real fusion products are not monoenergetic. The D-3He proton carries about 14.7 MeV and the helium-4 about 3.6 MeV, and each species has a spread from the plasma temperature and from where in the device it was born. A single fixed retarding voltage cannot match a broad spectrum, so a large fraction of energy would be lost as particles either fail to reach the collector or slam into it with leftover energy. This is exactly why the burner uses a staged, multi-modal train rather than one grid.
In short: DEC trades the thermodynamic Carnot penalty for an optics-and-matching problem. Solving that matching problem well is the engineering heart of the DEC train.