In materials science, a disappearing polymorph is a form of a crystal structure (a morph) that is suddenly unable to be produced, instead transforming into a different crystal structure with the same chemical composition (a polymorph) during nucleation. Sometimes the resulting transformation is extremely hard or impractical to reverse, because the new polymorph may be more stable. That is, they are metastable forms that have been replaced by more stable forms.

It is hypothesized that contact with a single microscopic seed crystal of the new polymorph can be enough to start a chain reaction causing the transformation of a much larger mass of material. Widespread contamination with such microscopic seed crystals may lead to the impression that the original polymorph has "disappeared". In a few cases, such as progesterone and paroxetine hydrochloride, the disappearance gradually spread across the world, and it is suspected that it is because Earth's atmosphere has over time become permeated with tiny seed crystals. It is believed that seeds as small as a few million molecules (about 10⁻¹⁵ grams) is sufficient for converting one morph to another, making unwanted disappearance of morphs particularly difficult to prevent. It is hypothesized that "unintentional seeding" may also be responsible for a related phenomenon, where a previously difficult-to-crystallize compound becomes easier to crystallize over time.

Although it may seem like a so-called disappearing polymorph has disappeared for good, it is believed that it is always possible in principle to reconstruct the original polymorph with a lab that has not been contaminated by the new morph. This was demonstrated in the ranitidine case. However, doing so is usually impractical or uneconomical. In some cases, the original morph can be reconstructed by a different pathway with different chemical kinetics, as in the case of progesterone.

This is of concern to the pharmaceutical industry, where disappearing polymorphs can ruin the effectiveness of their products and make it impossible to manufacture the original product if there is any contamination. There have been cases in which a laboratory that attempted to reproduce crystals of a particular structure instead grew not the original but a new crystal structure. The drug paroxetine was subject to a lawsuit that hinged on such a pair of polymorphs, and multiple life-saving drugs, such as ritonavir, have been recalled due to unexpected polymorphism.

The Gibbs phase rule states that, under most thermodynamic conditions (fixed temperature, pressure, chemical potential, and other intensive thermodynamic properties), for each chemical species, only one phase is thermodynamically stable (i.e. have the lowest Gibbs free energy per volume), except on certain boundaries, such as the coexistence of ice and water right at the freezing point. In particular, since each crystal morph is a phase of matter, this implies that under normal circumstances, there exists only a single crystal morph at thermodynamic equilibrium. However, some phases may be kinetically stable, even if not energetically so.

Disappearing polymorphs occur when there are two morphs of a substance, and one morph has lower Gibbs free energy, but is kinetically slower to form. Thus, when the crystal is first formed, the kinetically faster morph occurs first. Eventually, by accident or catalysis, the other morph occurs, which can then serve as seed crystal. More abstractly stated, disappearing polymorphs are morphs that are kinetically stable but not thermodynamically stable.

In detail, consider the classical nucleation theory of crystallization of water into ice. When liquid water is held just below the freezing point, the relative change in Gibbs free energy for a sphere of ice (relative to an equivalent amount in water) with radius ${\displaystyle r}$ is$${\displaystyle \Delta G={\frac {4\pi }{3}}V_{0}r^{3}+4\pi A_{0}r^{2},}$$where ${\displaystyle V_{0}}$ is the change in free energy per volume, and ${\displaystyle A_{0}}$ is the change in free energy per surface area (the interfacial energy, or surface tension). The term ${\displaystyle A_{0}}$ is usually positive, since there is an energy penalty for the boundary between two different phases of matter. However, as water crosses from above to below freezing point, ${\displaystyle V_{0}}$ turns from negative to positive.

The critical radius, ${\displaystyle r_{crit}}$, satisfies ${\displaystyle {\frac {d}{dr_{crit}}}\Delta G=0}$. A ball of ice with ${\displaystyle r<r_{crit}}$ tends to shrink, but a ball of ice with ${\displaystyle r>r_{crit}}$ tends to grow. A perfectly homogeneous liquid water below the freezing point may thus remain indefinitely liquid, until a single seed crystal of ice appears with ${\displaystyle r>r_{crit}}$, after which it would grow without limit. Similarly, dirt within the water that is attracted to ice would have a negative interfacial energy with ice, which allows an initial seed crystal to form around dirt particles. This competition between kinetics and stability allows the supercooling effect, whereby a clean liquid water without dirt or seed crystals may remain indefinitely in a liquid state. It also allows cloud seeding.