Pyrochlore
Pyrochlore is a group name: 12 IMA species end in -pyrochlore, none plain. Cubic niobium oxides, hardness 5–5.5, density 4.45–4.90; the main ore of niobium.
Pyrochlore is a group name, not one species: no plain pyrochlore row is on the September 2026 IMA list, and 12 rows end in -pyrochlore. They are cubic niobium oxides, A2B2X6Y, hardness 5 to 5.5 and density 4.45 to 4.90 on the older Handbook sheet. USGS calls pyrochlore the principal ore mineral for niobium, mostly from carbonatite.
In short
- A group name. A script over the September 2026 IMA-CNMNC Master List finds no plain pyrochlore and 12 rows ending in -pyrochlore: 10 approved since 2011, plus hydropyrochlore and oxycalciopyrochlore, redefined in 2010. Atencio (2021), Frontiers in Chemistry says each name is two prefixes and a root: the dominant anion at Y, then the dominant cation at A, both of the dominant valence; the root is the group, niobium at B.
- The old-label decoder. The Handbook of Mineralogy pyrochlore sheet predates that system. Recast by script, its Oka, Quebec, analysis has calcium leading at A (1.044 plus 0.033 Sr against 0.742 Na) and fluorine at Y (0.899 against at most 0.716 OH): by the rule it reads as fluorcalciopyrochlore. That is our reading, not a published reassignment.
- Documented failure: Hybla. The sheet’s Hybla, Ontario, analysis is calcium-led and niobium-led, but its 11.42% water is not split into OH and H2O and fluorine is only 0.115 of a Y site, so the first prefix cannot be chosen. And on the oxycalciopyrochlore type analysis titanium, 0.89, is the largest B cation: the dominant-valence rule (Nb + Ta 1.08) is what keeps it a pyrochlore.
- Niobium ore. USGS Professional Paper 1802-M (Schulz, Piatak and Papp, 2017) says “pyrochlore is the principal ore mineral for niobium”, with niobium “greatly enriched over tantalum (a ratio of about 200:1)”. USGS Mineral Commodity Summaries 2026, Niobium (Columbium) says most identified niobium resources “occur as pyrochlore in carbonatite” deposits, and puts Brazil at about 93% of world production.
- Magnetic, radioactive? None of the six sheets prints a magnetic property; USGS Open-File Report 99-529, Table 1 drew 2 pyrochlore samples best at 0.90 to above 1.70 A. The old sheet says “May be radioactive”; see is my mineral radioactive and how to store radioactive mineral specimens. This is not safety advice. The tantalum-led group is microlite.
| Analysis (sheet) | B site: 5+ against Ti4+ | A site: leading valence and cation | Y site | Name reached | Density (sheet) | 20.00 g in water |
|---|---|---|---|---|---|---|
| Oka, Quebec (old pyrochlore sheet, 1) | Nb + Ta 1.836 vs Ti 0.120 | 2+: Ca 1.044 (+Sr 0.033) vs Na 0.742 | F 0.899 vs OH at most 0.716 | Fluorcalciopyrochlore (our reading) | 4.33 calc.; 4.45–4.90 if uranoan | 15.38; 15.51–15.92 g |
| Hybla, Ontario (old pyrochlore sheet, 2) | Nb + Ta 1.235 vs Ti 0.546 | 2+: Ca 1.082 vs U 0.305 | F 0.115; the 11.42% H2O is not split into OH and H2O | Calcio-, pyrochlore group; Y prefix undecidable | as above | as above |
| Bayan Obo, China (fluorcalciopyrochlore type) | Nb 1.68 vs Ti + Zr + Sn 0.32 | 2+: Ca 1.14 vs Na 0.74 | F 0.92 vs O 0.08 | Fluorcalciopyrochlore, as named | 4.34 calc. | 15.39 g |
| Maoniuping, China (hydroxycalciopyrochlore type) | Nb + Ta 1.18 vs Ti 0.80 | 2+: Ca 0.74 vs Na 0.58, U 0.40, vacancy 0.21 | OH 1.01, F 0.09 | Hydroxycalciopyrochlore, as named | 5.10 meas. | 16.08 g |
| Věžná, Czech Republic (oxycalciopyrochlore type) | Nb + Ta 1.08 vs Ti 0.89; Ti is the largest single cation | 2+: Ca 1.11 vs Sb3+ 0.69 | O 0.76 vs OH + F 0.24 | Oxycalciopyrochlore, as named | 5.30 meas. (impure) | 16.23 g |
| Boziguoer, China (fluornatropyrochlore type) | Nb + Ta 1.94 vs Ti 0.08 | 1+: Na 1.03 vs Pb + Ca 0.51 | F 0.79 | Fluornatropyrochlore, as named | 5.275 calc. | 16.21 g |
| Lueshe, DR Congo (hydropyrochlore, 3) | Nb 1.80 vs Ti 0.20 | H2O 0.99 | H2O 0.86 | Hydropyrochlore, one prefix, as named | 3.40–3.48 meas. | 14.12–14.25 g |
What is pyrochlore? A group of niobium oxides
Not one species. The September 2026 IMA list has no row for plain pyrochlore. Twelve rows end in it: cesiokeno-, fluorcalcio-, fluornatro-, hydrokeno-, hydroxycalcio-, hydroxykeno-, hydroxymangano-, hydroxynatro-, hydroxyplumbo- and oxyplumbopyrochlore, all approved between 2011 (hydroxycalciopyrochlore, 2011-026) and 2024 (oxyplumbopyrochlore, 2024-026), plus hydropyrochlore (Democratic Republic of the Congo) and oxycalciopyrochlore (Czech Republic), both Rd, 2010 s.p. Atencio (2021), Frontiers in Chemistry lists eleven of them; the twelfth, oxyplumbopyrochlore, appears there among “possible new species” and was approved in 2024. The microlite page sets out how these became group names in 2013; the same applies here.
The structure. Atencio gives the general formula A2B2X6Y: A a large cation, a vacancy or H2O; B a high field-strength cation such as Nb, Ta, Ti4+, Sb5+ or W; X typically oxygen; Y an anion, a vacancy, H2O or a very large cation. The pyrochlore group is the one with niobium leading at B and oxygen at X. The old Handbook of Mineralogy pyrochlore sheet gives space group Fd3m, a = 10.35 to 10.47 Å, Z = 8. Atencio adds that synthetic pyrochlores vary far more in composition than natural ones; that literature is outside this page.
What the old sheet records. Cubic, “typically metamict”; typically octahedra modified by other forms, to 7 cm; granular, massive; twinning on {111}, rare. Cleavage {111}, “may be a parting”; fracture subconchoidal to uneven, splintery; brittle; hardness 5 to 5.5. Density 4.45 to 4.90 measured “if uranoan”, 4.33 calculated for CaNaNb2O6F. Translucent to opaque, black to brown, chocolate-brown, reddish brown, amber-orange or red-orange; streak brown; lustre vitreous to resinous. In thin section it is isotropic with weak anomalous anisotropism, n = 1.9 to 2.2. The name: “From the Greek for fire and green, to which color the mineral usually turns on ignition.”
How do you read an old pyrochlore label? The decoder, worked
The rule. Atencio (2021), Frontiers in Chemistry: the first prefix is “the dominant anion (or cation or H2O or vacancy) of the dominant valence at the Y-site”; the second is “the dominant cation of the dominant valence [or H2O or vacancy] at the A-site”; where they are equal only one is used, as in hydropyrochlore. The group is set at B.
Method. The old sheet’s two analyses were recast by script: each oxide divided by its formula weight on standard atomic weights, cations scaled to Nb + Ta + Ti + Zr + Sn + Fe3+ = 2, rare earths counted as Ce2O3. Oka gives Ca 1.044, Na 0.742, Sr 0.033, Nb 1.835, Ti 0.120, F 0.899, close to the printed (Ca1.06Na0.74Sr0.03RE0.06)(Nb1.86Ti0.12…)F0.91(OH)0.25. Its Ta/(Ta + Nb) is 0.000 and Hybla’s 0.070, the figures on the microlite page. The five newer sheets print their own formulas, read as printed.
Result. All five type analyses decode to the name on their sheet, which is the positive control. Oka decodes to fluorcalciopyrochlore: divalent cations 1.078 beat sodium 0.742 at A, and fluorine 0.899 beats any possible hydroxyl, at most 0.716 from the water by difference. Its analysis is strikingly close to the Bayan Obo type material, Ca 1.14, Na 0.74, F 0.92. We call that our reading; no fetched source reassigns Oka material. A specimen still needs its own analysis; see how to label and catalogue a mineral collection.
Where the decoder fails: Hybla and Věžná
Hybla: water without a split. The Hybla, Ontario, analysis is niobium-led at B (Nb + Ta 1.235 against Ti 0.546) and calcium-led at A (1.082 against 0.305 uranium in two valences). But it reports 11.42% H2O as one figure. Fluorine is 0.115 of a Y site, so it is not a fluor- name; whether the Y site is O, OH, H2O or vacant cannot be read from total water. The approved calcium names left are hydroxycalciopyrochlore and oxycalciopyrochlore; the analysis cannot choose between them.
Věžná: titanium leads, and still a pyrochlore. The oxycalciopyrochlore type analysis prints Ti0.89 Nb0.70 Ta0.38 at B: titanium is the largest single cation. The rule counts the dominant valence first: pentavalent Nb + Ta, 1.08, beat tetravalent Ti, 0.89, and within that pair niobium leads, so the root is pyrochlore. The sheet notes it was formerly ‘stibiobetafite’. A label reading betafite or stibiobetafite needs that arithmetic before it is renamed.
Old names on the sheets. Hydropyrochlore was “Formerly ‘kalipyrochlore’, now redefined”, and Atencio says the 2010 system replaced Hogarth’s of 1977. The old sheet, which cites Hogarth (1977) among its references, defines the group as “(Na + Ca)A > 20%; (Nb + Ta)B > 2TiB; NbB > TaB”, and both Oka and Hybla pass that test too.
What is pyrochlore used for? The main niobium ore
Niobium. USGS Professional Paper 1802-M (Schulz, Piatak and Papp, 2017): “pyrochlore is the principal ore mineral for niobium”, and the pyrochlore group “generally has niobium greatly enriched over tantalum (a ratio of about 200:1)”, typically as a primary mineral in alkaline igneous rocks, “particularly carbonatites (for example, the Araxá carbonatite in Brazil)”. USGS Mineral Commodity Summaries 2026, Niobium (Columbium) says most of the world’s identified niobium resources “occur as pyrochlore in carbonatite” deposits, that Brazil accounts for about 93% of production, and that United States niobium goes about 77% to steels and 23% to superalloys. Columbite was the earlier source; see columbite.
Radioactive? The old sheet says “May be radioactive” and Professional Paper 1802-M that, because it can contain uranium, pyrochlore “tends to be radioactive”; see is my mineral radioactive and how to store radioactive mineral specimens. This page gives no handling or safety advice.
Magnetic? A script finds no magnetic wording on any of the six sheets. USGS Open-File Report 99-529, Table 1 drew 2 pyrochlore samples at 0.10 to above 1.70 A, best 0.90 to above 1.70, marked depends on chemistry; microlite’s best range is above 1.70 A and magnetite’s 0.01. See which minerals are magnetic.
Where is pyrochlore found?
Occurrence, old sheet: “In pegmatites in nepheline syenites and other alkalic rocks; in granite pegmatites and greisens; characteristic in carbonatites; detrital.” Associates: zircon, aegirine, apatite, perovskite and columbite. The fluorcalciopyrochlore sheet adds bastnäsite-(Ce), monazite and magnetite from the Bayan Obo carbonatite.
Distribution, as printed. “Numerous localities.” Fredriksvärn, Larvik and the Langesundsfjord, Norway; Alnö Island, Sweden; near Schelingen, Kaiserstuhl, and the Eifel, Germany; Miass and Vishnevogorsk, Ilmen Mountains, “large crystals from the Tatarkskoye deposit, Yenisei Ridge”, and the Lovozero massif, Russia; San Miguel Island, Azores; Oka, Quebec, and Hybla, Ontario; Narssârssuk, Greenland; the Mbeya carbonatite, Panda Hill, Tanzania; Rouma Isle, Los Islands, Guinea; and the Mt. Weld carbonatite, Western Australia. No British locality is printed.
Type localities of the species, per Atencio: fluorcalciopyrochlore, Bayan Obo, Inner Mongolia; hydroxycalciopyrochlore, the Maoniuping mine, Sichuan; hydropyrochlore, the Lueshe mine, North Kivu; oxycalciopyrochlore, the Věžná I pegmatite, Czech Republic; hydroxykenopyrochlore, the Araxá mine, Minas Gerais. See what is a type locality. More oxides are in the species index. We hold no catalogue and nothing here is offered for sale; the wanted list is the only commercial route.