Is opal a mineral?
Opal has an IMA row, SiO2·nH2O, status G. Its structure varies: of 236 opals X-rayed in a 2019 review, 67 were amorphous opal-A, 161 opal-CT and 4 opal-C.
Yes, with a qualification. Opal has its own row on the September 2026 IMA list, SiO2·nH2O, status G. But opal is not one structure: of 236 opals X-rayed for a 2019 review, 67 were amorphous opal-A, 161 opal-CT and 4 opal-C. GIA gives hardness 5 to 6.5 and up to 20% water.
In short
- The list says yes. The September 2026 IMA-CNMNC Master List gives opal a row: SiO2·nH2O, status G for grandfathered, year printed as a question mark, country unknown. The live crystal vs mineral page lists opal among the historical exceptions the list still carries; this page agrees.
- The structure says it depends. Curtis and co-authors, Minerals 9 (2019), 299 call opal a generic term for a group of amorphous and paracrystalline silica species. X-ray diffraction sorts it three ways: opal-A, broad absorption only, centred on 4.0 Å; opal-CT, peaks at about 4.1 and 2.5 Å; opal-C, peaks at 4.04 and 2.5 Å.
- The A-CT-C split, counted. Of the 236 opals in the review, 67 were opal-A (28.4%), 161 opal-CT (68.2%), 4 opal-C and 4 intermediate. Of its 49 tabled exemplars, a script finds 10 opal-A; 7 of the 13 Australian exemplars are opal-A, against 3 of the 36 from elsewhere.
- Play-of-colour does not decide it. The review says both opal-A and opal-CT can show play-of-colour, and that Ethiopian play-of-colour samples are opal-CT, not opal-A. A precious opal from Coober Pedy and one from Ethiopia can differ in structure while looking alike.
- Why the Handbook has no opal sheet but mentions it. The Handbook of Mineralogy has no opal sheet; its cristobalite sheet lists fibrous or microcrystalline “opal” under cristobalite. Opal-C and opal-CT are named for cristobalite-like and tridymite-like X-ray positions.
| Name | What it is (source) | X-ray pattern (Curtis 2019) | Opals in the 2019 suite (of 236) | Of 49 tabled exemplars | IMA row? | Ordered atomic arrangement? |
|---|---|---|---|---|---|---|
| Opal-A (opal-AG, opal; opal-AN, hyalite) | Amorphous silica | Broad absorption only, centred on 4.0 Å | 67 (28.4%) | 10: 7 Australia, 2 Czech Republic (Valec), 1 Indonesia | No: falls under opal | No |
| Opal-CT | Paracrystalline silica with some structural characteristics of cristobalite and tridymite | Peaks at about 4.1 and 2.5 Å, a further peak near 4.27 Å | 161 (68.2%) | 37, including all 6 Ethiopian and all 4 Virgin Valley, Nevada | No: falls under opal | Partly |
| Opal-C | Cristobalite-like reflections; its exact structure unresolved | Peaks at 4.04 and 2.5 Å | 4 (1.7%) | 1: Iceland | No: falls under opal | Partly |
| Intermediate (CT-C, A-CT) | Between types | Transitional | 4 (1.7%) | 1: Opal Butte mine, Oregon (CT-C) | No | Partly |
| Opal, the name | SiO2·nH2O, up to 20% water (GIA) | Any of the above | 236 | 49 | Yes: G, year ?, country unknown | Not as a whole |
| Cristobalite | SiO2; Handbook sheet: tetragonal, pseudocubic, D 2.32 to 2.36 | Reference for C positions | Not counted | Not counted | Yes: G, 1887, Mexico | Yes |
| Mogánite | SiO2·nH2O on the IMA list; found in chalcedony | Not tested in the review | Not counted | Not counted | Yes: Rn, 1999-035, Spain | Yes: monoclinic (Monico 2024) |
Is opal a mineral? What the IMA list records
Opal is on the list. The September 2026 IMA-CNMNC Master List is the official list of valid mineral species, and it has a row for opal: SiO2·nH2O, status G, meaning a mineral described before the IMA was set up in 1958 and generally considered valid. The year is printed as a question mark and the country as unknown. A script over the list finds the word opal, at a word start, in that row and nowhere else.
The definition is the harder test. Tropper’s EOLSS chapter defines a mineral as a naturally occurring homogeneous solid with a definite, but not fixed, composition and a highly ordered atomic arrangement. The clause-by-clause test is worked for quartz on is quartz a mineral and for natural glass on is obsidian a mineral; it is not repeated here. For opal, the clause that matters is the ordered arrangement, and the answer to it depends on which opal is in your hand.
Mineraloid. The word is sometimes applied to opal, but none of the sources fetched for this page uses it, so this page does not rely on it. The collecting guides cover the other borderline names.
Is opal amorphous? Opal-A, opal-CT and opal-C
Three types by X-ray diffraction. Curtis and co-authors, Minerals 9 (2019), 299, from the South Australian Museum and Flinders University, review the classification Jones and Segnit proposed in 1971, which they say has been widely adopted for nearly 50 years. Opal-A, divided into opal-AG and opal-AN (hyalite), gives broad absorption only, centred on 4.0 Å: no crystal reflections. Opal-CT gives two prominent peaks at about 4.1 and 2.5 Å, and opal-C prominent peaks at 4.04 and 2.5 Å.
What the letters mean. A is for amorphous; C and T record the similarity of the reflections to cristobalite and tridymite. The review says the exact atomic structures of opal-CT and opal-C remain largely unresolved, and that opal-CT appears to be not just an intergrowth of the two minerals but a paracrystalline form of silica with some of their structural characteristics.
The quotable line: opal-A fails the ordered-arrangement test; opal-CT and opal-C pass it in part. That is why the honest answer is yes on the list and it depends on the structure.
Which opal is crystalline? The A-CT-C split, counted
The whole suite. The review X-rayed over 230 opals from the South Australian Museum, Flinders University, the University of Adelaide, Museum Victoria, the Smithsonian and recent acquisitions: 67 opal-A, 161 opal-CT, 4 opal-C and 4 intermediate, 236 in all. Worked, 28.4% were amorphous and 68.2% opal-CT.
The 49 exemplars. Its Table 1 lists the samples examined in detail. Typed by script and deduplicated by sample number, it holds 10 opal-A, 37 opal-CT, 1 opal-C from Iceland and 1 CT-C from the Opal Butte mine, Oregon. Australia supplies 13: 7 opal-A, from White Cliffs, Coober Pedy, Lightning Ridge, Springsure and one unlocalised sample, and 6 opal-CT. The 36 from elsewhere include 3 opal-A: two from Valec, Bohemia, which the review treats as hyalite (opal-AN), and one from Mamuju, West Sulawesi. Every Ethiopian, Turkish, Tanzanian and Nevada exemplar is opal-CT.
Play-of-colour does not tell you. The review says both opal-A and opal-CT can show play-of-colour, though the silica in their spheres differs, and that its Ethiopian play-of-colour samples were opal-CT, not opal-A. Properties of opal as a gem are on opal.
When the answer changes: hyalite, cristobalite and quartz
Hyalite is opal-AN in the review’s scheme: amorphous, with no IMA row of its own, so it is opal on the list and fails the ordered-arrangement clause. Cristobalite is a separate species with its own row, G, 1887, Mexico; the Handbook of Mineralogy cristobalite sheet writes fibrous or microcrystalline material as “opal”, in quotation marks. A label reading opal on such material is not wrong by the list, and a label reading cristobalite needs an X-ray.
Chalcedony is not opal. It is quartz, with mogánite, and a different density: the gem society gives chalcedony 2.55 to 2.91 against GIA’s opal 2.15. The full comparison is on is chalcedony quartz. Organic and glassy look-alikes are on is amber a mineral and is obsidian a mineral.
Synthetic opal is laboratory-made, so it fails the naturally occurring clause whatever its structure. Nothing on this site is offered for sale; the wanted list is the only commercial route.