I don't think this is a good approach at all, as the molecular weight of the protein is working against you. i.e. you need a very large mass of protein to bind significant amounts of a low molecular weight substance like phosphate.

You don't say why you want to remove phosphate or what the sample is, but you have two far better approaches, one or both of which may suit your sample type (i) desalting, if you have large proteins and you want to change the buffer, or (ii) phosphate binding beads: https://www.abcam.com/inorganic-phosphate-binding-resin-ab270547.html

Thanks Nick Gee! The beads from Abcam are just what I need and I did not know about their existence.

There are cheaper and more efficient methods. You may precipitate the phosphates by adding a counter ion including buffer or anion exchange resins that can be used to get rid of anions. SPE anion exchange or gravity flow anion exchange may be preferred by using a SCX exchanger. Desalting and buffer exchange is a very good option too as Dr.Nick Gee adviced. If you are interested in a couple or a single protein of interest in the solution, it would be more appropriate to use immunoaffinity selective beads to immunocapture your targets. Simultaneously phosphate clean-up can be achieved by applying this approach.