Species · Oxides

Columbite

Columbite is a root name: columbite-(Fe), -(Mn) and -(Mg) on the IMA list, niobium oxides, hardness 6, density 5.20–6.65. Past 62% Ta2O5 it is tantalite.

IMA status No plain columbite row on the September 2026 list. Three suffixed rows: columbite-(Fe) Fe2+Nb2O6 and columbite-(Mn) Mn2+Nb2O6 (both Rn, 2007 s.p., USA); columbite-(Mg) MgNb2O6 (Rn, 1967 s.p., Tajikistan)Handbook sheets used Ferrocolumbite and manganocolumbite, the sheets’ older names for columbite-(Fe) and -(Mn); no columbite-(Mg) sheet could be fetched; tantalite-(Fe), -(Mn) and -(Mg) as comparatorsCrystal system and habit Orthorhombic, space group Pbcn; short prismatic, equant or flat tabular crystals to 0.75 m, in groups to 18 t; heart-shaped twins on {021} and {023}Hardness, density Hardness 6; density 5.20–6.65 measured; a 20.00 g piece reads 16.15–16.99 g in waterColour, streak, lustre Black to brownish black; streak black to dark brown (Fe), dark red to black (Mn); submetallic to vitreous, commonly tarnished iridescentComposition, worked Ideal FeNb2O6 is 55.03% niobium by mass, FeO 21.28% and Nb2O5 78.72%, the sheet’s ideal columnThe 62% line Niobium leads the B site until Ta2O5 reaches 62.44% of the Nb2O5 + Ta2O5 weight; past that the name is tantaliteMagnetic? Radioactive? Columbite-(Fe) sheet: “Paramagnetic”; USGS separator best 0.40–0.50 A. No sheet prints radioactivity; the IMA formulas hold no U or ThColtan USGS: “columbite-tantalite, also called coltan”

Columbite is a root name, not one species: the September 2026 IMA list has columbite-(Fe), -(Mn) and -(Mg), niobium oxides such as Fe2+Nb2O6. They are orthorhombic, hardness 6, density 5.20 to 6.65, black with a dark streak. Where tantalum outnumbers niobium the mineral is tantalite; USGS calls the series coltan.

In short

  • A root name. A script over the September 2026 IMA-CNMNC Master List finds no row for plain columbite and three suffixed ones: columbite-(Fe), Fe2+Nb2O6, and columbite-(Mn), Mn2+Nb2O6, both renamed in 2007 with the USA as first country; and columbite-(Mg), MgNb2O6, renamed in 1967, Tajikistan. The suffix names the leading A cation. The Handbook of Mineralogy still files the first two as ferrocolumbite and manganocolumbite.
  • The 62% line. Columbite and tantalite share one structure and are split by the B site: niobium-led is columbite, tantalum-led is tantalite. A tantalum atom weighs 1.948 times a niobium atom, so the boundary falls at 62.44% Ta2O5 by weight of the two oxides, not 50%. Recast by script, the eight sheet analyses run from Nb/(Nb + Ta) 0.958 at Ånneröd, Norway, to 0.006 at Morrua, Mozambique.
  • Documented failures. The Herbb #2, Virginia, analysis is named columbite-(Mn) on a manganese lead of 0.510 to 0.461 iron atoms. The sheets’ measured densities meet at 6.65, columbite’s top and tantalite’s floor, so 20.00 g reads 16.99 g in water either way; USGS’s 5.2 against 8.0+ holds only at the ends.
  • Ore and coltan. USGS Professional Paper 1802-M (Schulz, Piatak and Papp, 2017) calls tantalite “the principal ore mineral for tantalum” and pyrochlore the principal one for niobium, and says that until the 1950s niobium was a byproduct of mining columbite-tantalite pegmatites for tantalum. USGS Fact Sheet 2014-3054 names “columbite-tantalite, also called coltan”. See pyrochlore.
  • Magnetic, radioactive, British? The columbite-(Fe) sheet prints “Paramagnetic”; USGS Open-File Report 99-529, Table 1 drew 13 ferromanganocolumbite samples best at 0.40 to 0.50 A, against magnetite’s 0.01. No sheet here prints radioactivity. The columbite-(Mn) sheet names Okehampton, Devon.
The 62% line: the eight analyses on the Handbook of Mineralogy columbite and tantalite sheets, recast by script on standard atomic weights to atoms per 2 B-site cations (Nb + Ta + Ti + Sn + W), with the leading B and A cations and the name they give; IMA names from the September 2026 list (the Handbook sheets call columbite-(Fe) and -(Mn) ferrocolumbite and manganocolumbite); measured density as printed on each sheet; in-water weight of a 20.00 g piece = 20 − 20/D
Analysis (sheet)Nb, Ta per 2 BNb/(Nb + Ta)Ta2O5 share of Nb2O5 + Ta2O5, wtFe, Mn, MgName by the leading cationsSheet density (meas.)20.00 g in water
Ånneröd, Norway (columbite-(Fe) 1)1.901, 0.0830.9586.8%0.731, 0.294, 0Columbite-(Fe)5.20–6.6516.15–16.99 g
Kings Mountain, North Carolina (columbite-(Fe) 2)1.443, 0.4810.75035.7%0.672, 0.279, 0Columbite-(Fe)5.20–6.6516.15–16.99 g
Old Mike mine, South Dakota (columbite-(Mn) 1)1.799, 0.1570.92012.7%0.267, 0.733, 0Columbite-(Mn)5.20–6.6516.15–16.99 g
Herbb #2, Virginia (columbite-(Mn) 2)1.355, 0.5430.71440.0%0.461, 0.510, 0Columbite-(Mn), by 0.049 Mn5.20–6.6516.15–16.99 g
Spittal an der Drau, Austria (tantalite-(Fe) 1)0.868, 1.1000.44167.8%0.772, 0.200, 0Tantalite-(Fe)6.65–7.9516.99–17.48 g
Lipovka, Urals, Russia (tantalite-(Mg) 1)0.780, 1.2000.39471.9%0.400, 0.050, 0.560Tantalite-(Mg)6.717.01 g
Sanarka, Russia (tantalite-(Mn) 1)0.168, 1.8090.08594.7%0.082, 0.980, 0Tantalite-(Mn)6.65–8.0016.99–17.50 g
Morrua mine, Mozambique (tantalite-(Mn) 2)0.011, 1.9880.00699.7%0.003, 0.996, 0Tantalite-(Mn)6.65–8.0016.99–17.50 g
The line itself1, 10.50062.44%—Columbite below, tantalite aboveMeet at 6.6516.99 g

What is columbite? A root name for three niobium oxides

No mineral is called plain columbite on the September 2026 IMA list. A script over the list finds three suffixed rows and no bare one: columbite-(Fe), Fe2+Nb2O6, and columbite-(Mn), Mn2+Nb2O6, both status Rn (renamed) in a 2007 special procedure with the USA as the country of first description; and columbite-(Mg), MgNb2O6, Rn, 1967 s.p., Tajikistan. The suffix is the leading cation on the A site. An old label reading columbite is therefore a root-name label, in the way an old microlite label is a group label, and it can be kept with a note; see how to label and catalogue a mineral collection.

What the sheets record. The Handbook of Mineralogy ferrocolumbite sheet, columbite-(Fe) under its older name, gives orthorhombic crystals, point group 2/m 2/m 2/m, short prismatic, equant or flat tabular, to 0.75 m, and parallel groups to 18 t. Twins are “heart-shaped contact and penetration twins, on {021} and {023}”, which “may produce pseudohexagonal trillings”; see mineral twinning explained. Cleavage distinct on {100}, less distinct on {010}; brittle; hardness 6; density 5.20 to 6.65 measured and [5.43] calculated. Its cell, a = 14.266, b = 5.7321, c = 5.0503 Å, Z = 4, gives 5.431 for FeNb2O6 when worked, which agrees. Black to brownish black, streak black to dark brown, lustre “Submetallic to vitreous, commonly tarnished iridescent”; biaxial (–). The manganocolumbite sheet adds a dark red to black streak and strong red, red-brown and orange pleochroism. No columbite-(Mg) sheet was found at three addresses, so this page quotes no numbers for it.

Worked composition. On standard atomic weights FeNb2O6 is 337.65 and 55.03% niobium; as oxides, FeO 21.28% and Nb2O5 78.72%, the ferrocolumbite sheet’s ideal column. MnNb2O6 gives 21.07% and 78.93%, the manganocolumbite column. The name, per the sheet: “For iron (FERRum), and COLUMBia, for America, source of the first specimens.”

Columbite or tantalite? The 62% line

One structure, split by the B site. The three tantalite sheets put tantalite-(Fe), -(Mn) and -(Mg) in the “Columbite group”, in series with the columbites. USGS Professional Paper 1802-M (Schulz, Piatak and Papp, 2017) says columbite (niobium-rich) and tantalite (tantalum-rich) “have very similar properties because they have the same structure and similar chemistries”. The decisive number is the atom ratio Nb/(Nb + Ta).

Method. Each oxide in the eight sheet analyses was divided by its formula weight on standard atomic weights and the cations scaled so that Nb + Ta + Ti + Sn + W = 2, total iron as FeO. Kings Mountain recasts to Nb 1.443, Ta 0.481, Nb/(Nb + Ta) 0.750, the same ratio as the printed Nb1.50 Ta0.50; Spittal an der Drau to Ta 1.100, Nb 0.868 against Ta1.11 Nb0.87; Lipovka to Mg 0.560, Fe 0.400, Ta 1.200, Nb 0.780 against Mg0.56 Fe0.40 Ta1.20 Nb0.78. The table sets out all eight; the leading cations give the sheet’s own name in every case.

Why 62%, not 50%. Ta2O5 weighs 441.89 and Nb2O5 265.81 per formula, so equal numbers of tantalum and niobium atoms put 62.44% of the oxide weight on the tantalum side. A columbite can carry up to that share of Ta2O5 by weight; Herbb #2 carries 40.0% and is still a columbite at 0.714. Reading the name from weight per cent instead of atoms moves the boundary by 12.44 points. The microlite and pyrochlore pair is split by the same pair of elements on a different structure; see microlite.

Can you tell columbite from tantalite by weight?

Only at the ends. USGS Professional Paper 1802-M says tantalite “has a much higher specific gravity (8.0 +) than does columbite (5.2)”. On how to measure specific gravity at home, 20.00 g reads 16.15 g in water at 5.2 and 17.50 g at 8.0, 1.35 g apart. But the sheets print the measured ranges as columbite-(Fe) and -(Mn) 5.20 to 6.65, tantalite-(Fe) 6.65 to 7.95 and tantalite-(Mn) 6.65 to 8.00: they meet at 6.65, 16.99 g in water, where a piece could carry either name.

Documented failure: two analyses on a knife edge. Herbb #2 is named columbite-(Mn) on 0.510 manganese against 0.461 iron, a lead of 0.049 atoms; the sheet’s own microprobe formula, Mn0.52 Fe0.47, agrees. Spittal an der Drau is a tantalite-(Fe) at Nb/(Nb + Ta) 0.441. Neither name can be read from colour, streak or a balance; it takes an analysis. The streaks overlap: black to dark brown on the columbite-(Fe) sheet, dark red to black on columbite-(Mn), black on tantalite-(Fe), red, scarlet to black on tantalite-(Mn), and all print a submetallic to vitreous lustre except tantalite-(Mg), semi-metallic to metallic.

Same formula, other mineral. The IMA list writes tantalite-(Fe) and tapiolite-(Fe) identically, Fe2+Ta2O6, and tantalite-(Mn) and tapiolite-(Mn) as Mn2+Ta2O6; the tantalite sheets call them dimorphous. The ferrotapiolite sheet (tapiolite-(Fe)) is tetragonal and uniaxial, density 7.90 measured, 17.47 g in water for 20.00 g; tantalite is orthorhombic and biaxial.

What is columbite used for? Niobium, tantalum and coltan

An ore of niobium and tantalum, in USGS’s words. USGS Professional Paper 1802-M: “pyrochlore is the principal ore mineral for niobium, and tantalite is the principal ore mineral for tantalum”, and “Until the discovery of pyrochlore-rich deposits in the 1950s, niobium was produced as a byproduct of mining columbite-tantalite-bearing pegmatites for tantalum.” The same paper says columbite and tin slags account for a minor amount of niobium production. USGS Mineral Commodity Summaries 2026, Niobium (Columbium) gives United States niobium use as about 77% steels and 23% superalloys.

Coltan. USGS Fact Sheet 2014-3054 lists African countries that “mine for tantalum minerals (such as columbite-tantalite, also called coltan) through artisanal mining”, and Professional Paper 1802-M names Burundi, the Democratic Republic of the Congo, Nigeria, Rwanda and Uganda. Coltan is a trade name for the series, not an IMA row. Columbium, per the Fact Sheet, was the name Charles Hatchett gave in 1801 to the element he found in a columbite sample from a Massachusetts mine, sent to the British Museum in 1734; the name niobium became official in 1949.

Is columbite a metal? No: an oxide mineral, 55.03% niobium by mass in the ideal iron end. The metals are niobium and tantalum; the distinction is set out on is pyrite a metal.

Where is columbite found? Pegmatites, placers, and Devon

Occurrence, per the sheet: “An accessory constituent of granite pegmatites; rarely in carbonatites; a detrital mineral in placer deposits.” Associates: albite, microcline, beryl, lepidolite, muscovite, tourmaline, spodumene, lithiophilite, triphylite, amblygonite, triplite, apatite, samarskite, microlite and cassiterite.

Distribution, as printed. Columbite-(Fe): Middletown, Portland and Haddam, Connecticut; the Beecher Lode near Custer and around Keystone, South Dakota; Spruce Pine and the Foote mine, Kings Mountain, North Carolina; Ånneröd and Tveit, Norway; Miass, Ilmen Mountains, Russia; Craveggia, Italy; Ambatofotsikely and Antsirabe, Madagascar; Greens Well, Wodgina and Greenbushes, Western Australia; the Ishikawa district, Japan. Columbite-(Mn): the Globe and Harding pegmatites, New Mexico; Cañon City, Colorado; Elk Creek and the Old Mike mine, South Dakota; Mineral Hill, Pennsylvania; Herbb #2, Virginia; Tanco, Manitoba; Varuträsk and Utö, Sweden; “Around Okehampton, Devon, England”; San Piero in Campo, Elba. Tantalite-(Mn) adds Glenbuchat, Aberdeenshire, and Salinas, Minas Gerais.

Nigeria and Jos? No columbite or tantalite sheet here names a Nigerian locality. USGS names Nigeria among the artisanal coltan countries, and USGS Mineral Commodity Summaries 2026, Tantalum lists Nigeria among tantalum producers. Professional Paper 1802-M lists the Jos Plateau, Nigeria, as an alkaline granite and syenite niobium–tantalum deposit and mentions “tin (and tantalum-niobium) mining in the Jos area”, without naming the mineral mined, so this page does not say which one it is. Neither sheet marks a type locality; see what is a type locality. Radioactive? The sheets print nothing and the IMA formulas hold no uranium or thorium; for associates that do, see is my mineral radioactive. This is not safety advice. 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.

Questions

Is columbite a mineral?
Columbite is a root name: the September 2026 IMA list has columbite-(Fe), -(Mn) and -(Mg) as the species. The Handbook columbite-(Fe) sheet gives hardness 6 and density 5.20 to 6.65.
What is the difference between columbite and tantalite?
Which of niobium or tantalum leads the B site. By weight of the two oxides the line falls at 62.44% Ta2O5, worked; USGS Professional Paper 1802-M (Schulz, Piatak and Papp, 2017) gives tantalite a much higher specific gravity.
Is columbite magnetic?
Its Handbook sheet says paramagnetic. USGS Open-File Report 99-529, Table 1 drew 13 ferromanganocolumbite samples at 0.10 to 0.70 A, best 0.40 to 0.50; magnetite’s best is 0.01.
Is columbite radioactive?
None of the columbite or tantalite sheets prints radioactivity and the IMA formulas hold no uranium or thorium; see is my mineral radioactive. This is not safety advice.
What is coltan?
USGS Fact Sheet 2014-3054 calls columbite-tantalite “also called coltan”; it is a trade name for the series, not an IMA species.
Is columbite found in Nigeria?
No Handbook sheet here names a Nigerian locality. USGS Professional Paper 1802-M (Schulz, Piatak and Papp, 2017) names Nigeria among countries with artisanal columbite-tantalite mining and lists the Jos Plateau as a niobium–tantalum deposit.