You mentioned the "Ksp for FePO4 is something like 10^-22". Does this mean if your phosphate is over 1 ppb, it will start binding with iron?? Doesn't this mean I will get FePO4 even if I dose the iron separately, as my tank is consistently at 1 ppm of phosphate?
Mini chemistry lesson: Ksp is called the solubility product. It describes the extent to which a substance dissolves and is used primarily for substances that are barely soluble. In this case, FePO4(s)
. For this system,
Ksp = [Fe][PO4] = 1.3x10^-22
Where [Fe] is the concentration of iron(III) ions in units of molarity. I guessed at the concentration of free phosphate earlier, and I think I overestimated substantially. Your 3.8ppm is 0.00004M PO4, and 0.00025M Fe. If those two solutions are combined, you would expect them to react almost to completion. With some math you find that the final concentration of iron would essentially be 0.00025M and the concentration of phosphate would essentially be zero (5 parts in 10^17, by my math.)
Phosphate and iron will combine at any concentration, and the reaction will get really, really close to completion every time. There is no minimum concentration; they'll just keep on sticking to one another.
The way that we can get around that is by using dilute solutions of both and dosing at different times. The former slows down the precipitation reaction dramatically (and more so with an EDTA complex), so that what might take minutes in concentrate takes days in the tank. The latter gives the plants time to consume the ions before they can react.
Finally, and I'm not sure whether this is true for FePO4 or not, many compounds are difficult to crystallize because the tiny seed crystals they're based on are unstable. These crystals are only stable when there are enough of them to bond into an extended lattice, and with the tiny amounts we're using, that is not likely to occur. Again, I'm not sure whether this is the case for iron(III) phosphate or not. What I do know is that in the concentrations used in aquarium dosing solutions, the concentrations are more than high enough for iron(III) phosphate to precipitate.