This. IANAP (...Not A Physicist), but condensing water vapor releases a whole lotta heat. If you could do that "free" - well, setting up a heat engine between your hot condenser and the evaporation-cooled spot where you let the water vaporize again becomes a classic perpetual motion machine.
(Yes, I see that the article claims their system is a mere 10 to 30 times more energy efficient than conventional dehumidifiers. That still sounds far too good to be true.)
Conventional AC and dehumidifiers remove water from air by cooling both until the cooled water drops out of solution. This is supposedly using chemical means to remove hot water from the air so that you don't have to cool it.
Fuel cells and electric motors are vastly more efficient at turning propane into motion than a combustion engine. Its possible they found a dessicant that is particularly easy to reverse at stp. Theymention the absorbing dessicant is exothermic which is unusual I think? And they pump that heat over to regenerate the other dessicant under a slight vacuum. Sounds a bit like the oxygen generators people use for lung conditions, or to weld with.
My understanding is heat from first chamber + external vacuum is used to dry out a second chamber and then the system is reversed when the first chamber is saturated.
Energy is being put into the system but it is less than the AC unit cooling humid air on its own so is a net gain relatively speaking.
From what I understand, the desiccant releases water under slight pressure from vacuum pump, which presumably is less than the energy saved by using the desiccant in the first place.
The release of water from the desiccant occurs at a lower pressure than its absorption. Energy is consumed in maintaining that lower pressure by pumping away the water vapor as it is released (presumably, it is vented outdoors.)
From the patent:
[The vacuum pump] is used to provide modest compression to raise the vapor pressure sufficiently to condense to liquid water. Because the compression work is only done on the water vapor, this minimizes the energy consumption. Lastly, the condensate is pumped up to atmospheric pressure for discharge to a storage vessel (this consumes a trivial amount of additional energy).
Update: The patent also implies that the high efficiency of any given device of this type may only obtain over a limited range of operating conditions:
Sorbents optimized for a 43.degree. C., 60% RH condition are likely to perform poorly and result in much higher power consumption at the more challenging 27.degree. C., 10% RH humidity condition and vice versa.