@antinn Thanks for the details.
With pure T density rho of 1 (close enough, easy math) 1g/mm3, 0.3 mg have volume of 0.0003 mm3 = 3 x 10e-4 mm3. Record surface ~ 0.3 m2 = 3 x10e5 mm2
3 x 10e-4 mm3 : 3 x 10e-5 mm2 = 1 x 10e-9 mm = 10e-12 m, which is subatomic. (Angstrom is 10e-10 m). So you have scattered molecules of T on the surface, not even a film. Thickness of impurity is that of the size of the molecule, ~ 1 nm.
As an upshot of this, rinsing with distilled water to get concentration of T down by factor 10–100x is a non-issue. It does not matter. At 100x higher concentration, "film" thickness is still in Angstrom order of magnitude, so still scattered molecules.
Even if my calculations are not precise (record surface is 28 cm2 minus center label area guesstimated 20 cm2), our calculations give different result by factor of 10'000x, which is pretty significant, IMHO. My math skills are not that great, still don't know I passed calculus, but this looks pretty straightforward to me.
I think the problem step is going from weight to thickness, which IMHO is not supported. Going from weight and density to volume, and volume/area to thickness is more direct. my 2c.