There have been recent interesting discussions about root cellars for food storage, which naturally stay below 60F even in the summer.
I have worked a great deal with compressed air in the Navy. Though compressed air (CA) does not use a liquid/gas phase-change like Freon, I observed significant heat/cold cycles resulting in freezing temps.
As air is compressed, heat energy that was spread out becomes concentrated, and the compressor and storage tank became very hot. As the stored CA cooled over time, trapped humidity would condense and pool at the bottom of the storage tank. It was part of my job to regularly open a small drain valve to spray out the moisture.
If this moisture was allowed to accumulate, then the air was used in a sudden high outflow, iceplugs would form at 90-deg piping elbows. Ice could also be observed forming on the outsides of the dump valves as ambient humidity was suddenly cooled. The CA in question was for emergency surfacing of a submarine.
It is also useful to know that saltwater will freeze at a much lower temp than freshwater. I propose to exhaust CA through stainless steel 1/2" hydraulic tubing (easily bent) in a saltwater bath, into which aluminum trays of freshwater have been set.
A recent study uncovered that some off-grid homes have used large out-of-date propane storage tanks (300-PSI capable) as storage for air that had been compressed by a multi-bladed high-solidity wind turbine similar to those found for water-pumping on farms, cattle ranches, and for antique steam railroads.
If the CA cylindrical storage tank were mounted vertically, you could bond aluminum angle to its sides to form vertical fins and wrap the fins to form a shroud for vertical thermosiphon flow to speed ambient air cooling.
Of course the wind compressor turbine and storage tank are mounted above ground. If the cooling coils and ice trays are underground in a heavily insulated root cellar, I believe ice blocks could be made with no electricity and low wind speeds.
If extra cooling of the stored CA was needed to achieve the desired result, pressure-dumping could be scheduled for the coolest part of the summer nights. In addition, water-dripping onto the tank, and perhaps a VAWT-driven fan (blowing ambient air through the shroud) could further cool the tank if needed.
I believe it would be possible to chill the saltwater bath to 20F and for the exhausting air to further cool the airspace by purging air that was warmed by adding ambient-temp foods to the space.
Just an idea, probably very inefficient, but could be easily home-made. I would appreciate any suggestions for improvements or pointing out significant flaws.