The melting and freezing point of various materials is fixed. This is something that may work great at 90F but not so great at 100F.
Unless I'm reading this wrong, this requires the transport of solids or salty liquids. That's going to be a challenge. Salty liquids tend to be corrosive and solids aren't easily pumped places.
Beyond that, the article assumes that all refrigerants have high gwp. The industry is currently in the process of moving to either CO2 or propane. Both of which have fairly small gwp. They do have their own problems, CO2 needs high pressure and propane is explosive. However, they are workable and already in use.
The key reason we use gas is it can cold (or heat) at pretty wide temperature ranges.
Putting salt on roads to prevent ice from forming is a common example of this in action.
A better example: lots of people use salt to force the ice to melt, to chill their beer faster. It's the exact same principle as in the article.
The main issue is entropy - it's hard to remove the salt off the liquid, to make it a solid again, if you want a continuous process. But perhaps if you use a "brine" vs. pure liquid, reverse osmosis could do the trick?
The second issue is that you need to work with a rather specific range of temperatures; make it too high and the liquid won't become solid again, make it too low and it won't melt even if you add the salt. This could be also solved by "bootstrapping" the process with the older HFC-based one, you're still reducing HFC usage this way albeit not completely.
Fuck Science for paywalling research. May they rot in hell.
The article implies they are using electric charge to bring the ions into and out of solution to trigger melting and recrystalization. Does anyone have actual details about the process?
Also, is this paper saying it has a COP of 0.3? That's terrible compared to typical evaporative refrigeration.