Socialfusions, I do not know how much you already know, so apologies if this is too basic for you.
Kaolin is found in the form of hexagonal plates with the formula Al2Si2O5(OH)4. At pH <8, the edges of the plates are positively charged, whilst the faces are negative. As pH increases, the polarity of the edges reverses and plate-plate interactions weaken.
By introducing PAA, a polyanion, we are trying to mimic the effect that occurs when you take the pH above 8. There is evidence that suggests the PAA interacts with the Al within the kaolin (PAA adsorption density onto alumina 10x that then on kaolin). We therefore infer that the PAA is adsorping at the plate edges + hinders plate-plate interactions.
When I referred to derivatives, basically I played around with the chain length and end groups. My funding died before I actually got to test any samples in clay, though.