This is an physical-science example of the "justification problem" -- chemistry isn't computer science, we don't design all out components from the elemental level; we discover and hack molecules, and identify them by interpreting their behaviors (boiling point, what color smoke is produced when mixed with some other stuff, etc) , and then we slap on a label that we hope is correct). It's all a very statistical science, not irrefutable logic.
Given the context, I would have hoped something involving a mass spectrometer or something similar rather than buying something purely on the basis that it is known to have the property of being suitable for cats to urinate into and then hoping that they don't ever change the mixture on any products in the range.