There is no “correct” value per se, in the sense that these are not optimized parameters but rather estimates of covalent atomic radii. There are standard tabulations for the main group, which are the default values in Q-Chem and are widely used, but no such standard values exist for transition metals. You have two options here:
(1) Use the radii from the “universal force field” (UFF), which is defined for transition metals but is somewhat less well tested in conjunction with PCM. This is requested with
(along with whatever other options might be in your $pcm input section).
(2) Decide for yourself what a reasonable atomic radius is for Pd and input it as a user-defined parameter. To do this for a single atom (Pd, Z = 46) leaving other radii as the default values, you need to add an input section
where XXX is your choice for the Pd radius, in Angstroms.
Either way, you should be on the lookout for nonsensical results caused by a poor choice of radius.