# Can a crystal structure of a substance appear or disappear?

I read somewhere before that some substances no longer crystallize naturally in some specific structures, although it do that readily before. Instead, new structures that never happen before become widespread.

Edit: @Buttonwood get the correct name for this phenomenon: disappearing polymorphs.

So:

1. Is it true?
2. If it is true, what is this phenomenon called?
3. If it is true, what is some high profile event of this phenomenon.
• "I read somewhere before" gives others very little to go on. Please give a specific references. There is nothing like data and a good references. – Hal Dec 3 '20 at 15:06
• @Hal I read it in an answer in worldbuilding stack. I cannot find the answer again, so I think I should ask here. – DTN Dec 3 '20 at 15:10
• I read somewhere about an army of spear-wielding tribesmen mounted on bears. Wait, I think it was on worldbuilding.SE too. Seriously, no. – Ivan Neretin Dec 3 '20 at 16:34
• @IvanNeretin It may sound far-fetched. But in my interpretation of this vague question lacking a lot of detail, it is about disappearing polymorphs; which is in some fields a considerable bump in the road to success. – Buttonwood Dec 3 '20 at 16:37
• Does not the question imply behaviour of substance is time dependent, what would be breach of time translation symmetry and therefore breach of mass and energy conservation law ? – Poutnik Dec 3 '20 at 16:44

In short: Yes.

But it is not possible to offer an exhaustive answer to all aspects intervening here. But scratching barely the surface of the large topics of crystallography / materials science / pharmacokinetics let's try this one:

• Substances may solidify in a regular pattern, i.e., from a given point within the bulk, you may then predict both local structure in proximity (a few Angstroms, $$\pu{1 Å} = \pu{10^{-10}m}$$) as well as in far distance (e.g., for a macroscopic crystal). This regular pattern may be described by a model, the crystal structure.

• When solidifying, atoms, ions, molecules intend to do so minimizing the energy. Especially for flexible molecules (think e.g., about alkyl chains, phenyl rings), this may be tricky. Maybe you recall the Newman projections of simple, flexible alkanes

(credit)

where the energy about an individual molecule is plot: often, there are multiple energetic minima, but only one global energetic minimum. For the solidification of a material, the situation is even more complex, because minimization of the energetic minimum comprises both the molecule in question, as well as its interactions with nearest molecules around. If the rate of crystallization is too high (e.g., just crushing out a compound as a precipitate), it may happen that the molecules in the solid state again found an energetic minimum, but not not the global one. Like maybe an alkyl group still is in gauch orientation to an other.

• The possibility that the same molecules arrange in space to yield different pattern is called polymorphism. It is not something limited to organic molecules, though; it is equally observed for minerals ($$\ce{TiO2}$$ as rutile or anatase) and elements (e.g., sulfur) and obtaining one, or an other polymorph (to be read as «particular arrangement of molecules in the solid state») may depend on parameters like temperature, pressure, concentration of other substances during the crystallization.

• There are attempts to predict how molecules crystallize (e.g., IUCr's contests about small molecules example, a new one is in preparation; which is not the same as CASP14 about protein folding like AlphaFold2), but the reliability of these computational predictions depend on multiple parameters.

• There are multiple properties which may alter greatly by a different arrangement of the same molecules in the solid state, in different polymorphs. For solid pharmaceuticals, for example, the rate these active ingredients dissolve by ingestion, and thus enter the anabolism of the patient may vary a lot. Companies are very interested to obtain not only the same, chemically pure compounds for the pills, but equally to obtain again and again the same polymorph with the tested properties. Equally because in some jurisdictions both a compound, as well as the polymorph may be protected by a patent (an entry into the topic).

• Finally, yes it already happened to large businesses that they were unable to crystallize the molecules in the intended polymorph, a.k.a. disappearing polymorphs. As you may imagine, consequences for for the company and for customers / patients may be significant. PubMed allows you to read a survey publication about the topic without a paywall.

• This is really just a reflection of our inability to completely control and accurately reproduce experimental conditions. – Andrew Dec 4 '20 at 19:51