Steven Dutch, Professor Emeritus, Natural and Applied Sciences, University of Wisconsin - Green Bay
Silica possesses a remarkable variety of polymorphs, summarized in the tablebelow:
polymorph | Density | Crystal Class | Stability |
Cristobalite | 2.33 | Cubic | Above 1470 C |
Tridymite | 2.28 | Hexagonal | Above 870 C |
Quartz (High) | 2.53 | Hexagonal | Above 570 C |
Quartz (Low) | 2.65 | Rhombohedral | Surface Conditions |
Coesite | 2.93 | Monoclinic | Above 20 kb |
Stishovite | 4.30 | Tetragonal | Above 80 kb |
Cristobalite and tridymite also have high and low forms. The high forms are shown in the diagrams. Low tridymite is orthorhombic and pseudohexagonal, low cristobalite is tetragonal and pseudo-cubic.
See the discussion of tridymite for an alternative view of the tridymite stability field.
Although the formation temperature of tridymite, 870 C, is well within the normal magmatic temperature range, tridymite generally forms in silica-rich rocks of lower temperature, and must therefore form metastably in most cases.
Many mineralogists, particularly in Europe, follow the lead of German mineralogist Otto Floerke and regard tridymite as a polymorph that requires impurities (like water) to catalyze its formation, and therefore not a true member of the phase diagram above. They point to its anomalously low density and the fact that inversion of pure quartz or cristobalite to tridymite is not observed in the absence of impurities.
![]() |
We can regard tridymite as made of layers of paired tetrahedra, with successive layers alternating in stacking. |
No terrestrial magma gets as hot as the equilibrium formation temperature of cristobalite (1470 C), and since it is common in cavities in low-temperature volcanic rocks like obsidian, as well as some sedimentary settings, it must form metastably. It probably forms in situations where rapid growth kinetically favors a highly open structure, and possibly also by recrystallization of opaline silica.
Cristobalite also consists of layers of paired tetrahedra, except these alternate in an ABC cubic pattern.
![]() |
Here we see the relationship of the tetrahedra to the cubic unit cell of cristobalite. This is a very open structure. The oxygen layers are alternately 75 and 25 per cent filled compared to cubic close packing. |
Coesite consists of four-membered rings in chains, which in turn are cross-linked by other chains. Two layers are shown below.
In the view below, parts of four layers are shown. This view is slightly oblique for visibility. In a view directly down the two-fold axis and perpendicular to the glide planes, alternate layers would be directly superposed. The green and dark blue layers would be superposed, as would the light blue and yellow layers.
Stishovite is isostructural with rutile. It forms at pressures so great that silica tetrahedra break down and silicon assumes six-fold coordination. In the view below, blue atoms are oxygen, red are silicon.
Silica consists of silicon tetrahedrally coordinated to four oxygen atoms, with silicon atoms linked by intervening oxygen atoms. Ice consists of oxygen tetrahedrally coordinated to four hydrogen atoms, with oxygen atoms linked by intervening hydrogen atoms. So it should not be surprising that there is a close- indeed remarkable - similarity between the polymorphs of silica and the polymorphs of ice. The similarity is purely geometrical, driven by the need to find successively closer packings while maintaining tetrahedral coordination.
Tridymite and cristobalite are related in exactly the same manner as the low-pressure polymorphs of ice, Ice Ih and Ice Ic, and even have identical space groups. Ice II is rhombohedral like quartz, but differs in lacking screw axes in the centers of the six-membered tetrahedral rings.
Silica polymorph | Space Group | Ice polymorph | Space Group |
Quartz (low) | R3121 | Ice II | R3 |
Tridymite (high) | P63/mmc | Ice Ih | P63/mmc |
Cristobalite (high) | Fd3m | Ice Ic | Fd3m |
Coesite | C2/c | Ice V | C2/c |
Stishovite | P42/mnm | Ice VI | P42/mmc |
Return to Mineralogy-Petrology Index
Return to Thin-Section Index
Return to Crystals and Light Index
Return to Crystal Structures Index
Return to Mineral Identification Tables
Return to Professor Dutch's Home Page
Created 13 March, 2002, Last Update