Zirconium oxide is used by research scientists to isolate phosphopeptides. Aumina can do the job too. Zirchrom makes specialized columns for isolating phospho proteins. Lewis acids are chemicals that have empty orbitals and can accept electron pairs from Lewis Bases, which has a highly localized HOMO (The Highest Occupied Molecular Orbital) that can donate an electron pair to a Lewis acid.
Here we see a speckling of the zirconium atom already bound to water and hydroxide ions that has some vacancies for the electrons in our phosphate molecule. Wikipedia authors assembled a good piece on the isoelectric point defined as the pH at which a molecule has no net electrical charge.
When I was working with kaolin we bound the DNA to the alumina groups at pH 3-4 and eluted at pH 9. Note the isolectric points of gamma and alpha alumina and very low isolectric point of silica.
The following bentonite / montmorillonite structure is from Scott Labs. “Natural calcium bentonite has a high percentage of exchangeable calcium ions, very few exchangeable sodium ions. Calcium bentonite lattice structures open less but flocculate better, making them well-suited for clarification and lees compaction. Due to their less open lattice structure, calcium bentonites have a lower ability to bind proteins.” This is the link for the kaolin structure.
Kaolin with its sandwich of alumina and silca make it an appealing choice for DNA. In all honesty I found that alumina without the silica performed just as well. Yttria stabilized zirconia performed well. Is bentonite/montmorillonite really the best mineral to bind glyphosate? The water channels may make a good case for aflatoxins. Maybe.