Photoevaporation due to high-energy stellar photons is thought to be one of the main drivers of protoplanetary disk dispersal .
The fully or partially ionized disk surface is expected to produce free-free continuum emission at centimeter ( cm ) wavelengths that can be routinely detected with interferometers such as the upgraded Very Large Array ( VLA ) .
We use deep ( rms noise down to 8 \mu Jy beam ^ { -1 } in the field of view center ) 3.5 cm maps of the nearby ( 130 pc ) Corona Australis ( CrA ) star formation ( SF ) region to constrain disk photoevaporation models .
We find that the radio emission from disk sources in CrA is surprisingly faint .
Only 3 out of 10 sources within the field of view are detected , with flux densities of order 10 ^ { 2 } \mu Jy .
However , a significant fraction of their emission is non-thermal .
Typical upper limits for non-detections are 3 \sigma \sim 60 ~ { } \mu Jy beam ^ { -1 } .
Assuming analytic expressions for the free-free emission from extreme-UV ( EUV ) irradiation , we derive stringent upper limits to the ionizing photon luminosity impinging on the disk surface \Phi _ { \mathrm { EUV } } < 1 - 4 \times 10 ^ { 41 } s ^ { -1 } .
These limits constrain \Phi _ { \mathrm { EUV } } to the low end of the values needed by EUV-driven photoevaporation models to clear protoplanetary disks in the observed few Myr timescale .
Therefore , at least in CrA , EUV-driven photoevaporation is not likely to be the main agent of disk dispersal .
We also compare the observed X-ray luminosities L _ { X } of disk sources with models in which photoevaporation is driven by such photons .
Although predictions are less specific than for the EUV case , most of the observed fluxes ( upper limits ) are roughly consistent with the ( scaled ) predictions .
Deeper observations , as well as predictions spanning a wider parameter space , are needed to properly test X-ray driven photoevaporation .