Elution from PD-10 column - (Nov/18/2007 )
Try dialysis, or ultrafiltration if you dont have a lot of proteins to work with.
as genehunter-1 said, dialysis may be your best bet for desalting, if you don't mind waiting for a while.
the pd-10 desalting column is filled with sephadex (g-25, i think). you need to maintain an ionic strength of at least 0.15M or you may see interaction of your protein with the matrix, thus the late elution (also it may be eluting with the salt).
if you ultimately require your protein to be in a specific buffer then you should equilibrate and elute the pd-10 column with that buffer (but make sure it is of the proper ionic strength).
this information should be with the literature that came with the columns or you can access it online at the ge healthcare website.
by the way, dialysis also requires a minimum ionic strength to avoid binding of the protein to the membrane.