Again, to anticipate complaints, I have first researched books and the internet on this question and not found any answer, and did find plenty of sales pitches cluttering the internet, as usual. The question is: where are the coloring atoms in the structure of the gemstone crystals? For instance is the Cr2O3 that colors the aluminum oxide that is what rubies consist of found as A)substitution for a few atoms of aluminum or for a few atoms of oxygen in the crystal itself; B) floating around and not attached to the aluminum or the oxygen or substituting for it; C) held in a number of the rings formed by the aluminum and the oxygen in the crystal ; or D) other. Or for instance is the iron that is supposed to color silicon dioxide purple all over in a regular substitution for a silicon or an oxygen atom; subsituting just here and there in the structure; or floating around? The books just say a small amount of something colors the gem, but they don’t say where it is or if it is part of the crystal structure. Emerald is beryllium aluminum silicate with chromium, but where is the chromium?
This is recalling back many years, from asking my grandfather (amateur gemologist, but his collections are all in museums now, so he is pretty good at it)…
The compound may be the same, but the crystaline structure of the crystals for each color is (or may be) different (depending?). The coloring compound is not substituting for any part of the source compound (isn’t bumping out the oxygen, say).
That is, you have aluminum oxide = colorant bonds, as well as aluminum oxide = aluminum oxide bonds, and the way the series of molecules line up (the crystaline structure) is somewhat dependant on the molecular shape of the colorant molecule (where it connects affects where others can connect, etc.) - plus, that alignment adds to some of the other properties of the stone (like star sapphires, where you have a VERY orderly arrangement of structure). But IIRC, most stones aren’t one huge orderly set of molecules, but more lots of little chunks of relatively orderly molecules all jammed together willy-nilly. So there is plenty of space in there for both (B) (floating around unattached) and (D) (other, colorant linked in between the other molecules).
Did that make sense? I won’t guarantee it is correct, because I asked that question about 25 years ago, but that is what I remember.
For gemstones like ruby and sapphire, there can be a lot of inclusions of other things, as well as flaw, which can be cracks or places where the structures don’t align in an orderly way - which in rubies/sapphires is a good thing, as the mish-mash of crystaline surfaces all lined up in different ways inside is what gives them a ‘glow’. There are probably plenty of spaces for ‘extra’ colorant molecules ‘floating around’ that way, too. Not to mention really tiny air gaps (some gemstones are quite porous).
You might find more detail if you look for info on creating synthetic rubies and sapphires - they are the same compounds as the ‘real thing’, so there might be something in the ‘how to’ that suggests the ‘what’s going on’.