Cacoxenite
Cacoxenite, Fe3+24AlO6(PO4)17(OH)12(H2O)75: hexagonal, golden radial sprays to 8 mm, hardness 3 to 4, density 2.2 to 2.6. The formula gives its density, 2.217.
Cacoxenite is a hydrated iron aluminium phosphate, Fe3+24AlO6(PO4)17(OH)12(H2O)75 on the IMA list (G, 1826, Czech Republic): hexagonal, in golden yellow to orange radiating acicular aggregates to 8 mm, hardness 3 to 4, density 2.2 to 2.6, silky lustre. Worked through, that formula gives the Handbook’s ideal analysis and its calculated density, 2.217, exactly.
In short
- The species. The Handbook of Mineralogy sheet gives a hexagonal mineral, acicular along [0001], typically in compact concentric spherical to radial aggregates to 8 mm, also randomly fibrous, as coatings or an intergranular cement; hardness 3 to 4; density 2.2 to 2.6; yellow, golden yellow, brownish yellow, reddish or deep orange, silky. The September 2026 IMA-CNMNC Master List writes Fe3+24AlO6(PO4)17(OH)12(H2O)75, status G, 1826, Czech Republic.
- The 2.217 check. Worked on standard atomic weights, the IMA formula weighs 4632.95 and comes out at 41.36% Fe2O3, 1.10% Al2O3, 26.04% P2O5 and 31.50% H2O: the sheet’s ideal column exactly. With the sheet’s cell (a 27.559, c 10.550 Å, Z 2) it gives a density of 2.217, the sheet’s calculated figure, at the light end of the measured 2.2 to 2.6. The charges balance at 75 each way.
- Documented failure: the aluminium. The formula carries one aluminium for every 17 phosphorus. The sheet’s analysis from the Hrbek mine, Svatá Dobrotivá, Czech Republic, has no Al2O3 at all; the Eleonore mine, Germany, analysis works out at 2.61 Al per 17 P. The ideal is a midpoint, not a measured specimen.
- Where, and with what. A common accessory mineral in oxidized magnetite and “limonite” iron ores, in Fe,Mn-bearing novaculites, and rarely in iron-rich sediments and soils. Associates: dufrénite, rockbridgeite, beraunite, strengite, wavellite, magnetite and “limonite”. Of 22 places on the strengite sheet, 6 are on cacoxenite’s too, by script.
- Cacoxenite in amethyst. The sheet names neither quartz nor amethyst nor any Brazilian locality. The identity of yellow-brown sprays inside amethyst is a separate question with its own evidence, set out on what is cacoxenite in amethyst.
| Source | Fe2O3 wt% | Al2O3 wt% | P2O5 wt% | H2O wt% | Fe:P (molar) | Al per 17 P | Density |
|---|---|---|---|---|---|---|---|
| IMA formula, worked | 41.36 | 1.10 | 26.04 | 31.50 | 1.412 (24:17) | 1.00 | 2.217 from the cell |
| Sheet column (3), ideal | 41.36 | 1.10 | 26.04 | 31.50 | 1.412 | 1.00 | 2.217 calculated |
| Sheet column (1), Svatá Dobrotivá (St. Benigna), Czech Republic | 41.46 | none | 25.71 | 32.81 | 1.433 | 0.00 | 2.2 to 2.6 measured (sheet range) |
| Sheet column (2), Eleonore mine, Germany | 40.37 | 2.89 | 26.18 | 30.59 | 1.371 | 2.61 | 2.2 to 2.6 measured (sheet range) |
What is cacoxenite? A hexagonal iron phosphate
Cacoxenite is a ferric iron phosphate with a little aluminium and a great deal of water, crystallising in the hexagonal system. The Handbook of Mineralogy cacoxenite sheet gives point group 6/m and space group P63/m, a 27.559, c 10.550 Å, Z 2. Crystals may be crudely hexagonal and acicular along [0001]; the usual habit is compact concentric spherical to radial aggregates to 8 mm, or bundles, randomly fibrous, or coatings and an intergranular cement. At that size most crystals are a lens subject; see what is a micromount.
Physical properties. Hardness 3 to 4; measured density 2.2 to 2.6, calculated 2.217. The colour runs yellow to brownish yellow or reddish orange, golden yellow, deep orange, rarely greenish; yellow in transmitted light; silky lustre; semitransparent. Optically uniaxial (+), ω 1.575 to 1.585, ε 1.635 to 1.656, so the birefringence, worked, is 0.050 to 0.081; pleochroic, O pale yellow, E canary-yellow to yellow-orange. The sheet prints no cleavage and no streak, and this page supplies neither.
What the IMA list records. The September 2026 list writes Fe3+24AlO6(PO4)17(OH)12(H2O)75, status G (grandfathered), 1826, Czech Republic, first reference Archiv für die Gesammte Naturlehre 8 (1826), 446. The Handbook writes the water as nH2O with n about 75. A script finds one row beginning Cacox and no row for amethyst. The name, per the sheet, is from the Greek for a bad guest, for the phosphorus content that degrades the quality of iron made from the host ores.
Cacoxenite formula, worked: the 2.217 check
The IMA formula reproduces both the sheet’s ideal analysis and its calculated density. Counted out, Fe3+24AlO6(PO4)17(OH)12(H2O)75 holds 24 Fe, 1 Al, 17 P, 161 O and 162 H. The 72 positive charges of the ferric iron and the 3 of the aluminium are balanced by 12 on the six oxide oxygens, 51 on the phosphate groups and 12 on the hydroxyls: 75 each way. On standard atomic weights the molar mass is 4632.95.
As oxides. Twelve Fe2O3, half an Al2O3, eight and a half P2O5 and 81 H2O give 41.36, 1.10, 26.04 and 31.50 wt%, totalling 100.00: the sheet’s column (3) to the hundredth. By element that is 28.93% iron, 11.37% phosphorus and 0.58% aluminium, an Fe:P ratio of 24:17, or 1.412. With the sheet’s cell, V = (√3/2)a²c = 6939.2 Å3 and Z = 2, the same formula gives a density of 2.217, the figure the sheet prints as calculated.
On a balance. At the measured 2.2 to 2.6 a 20.00 g solid piece would read 10.91 to 12.31 g in water; at 2.217, 10.98 g. The sheet’s aggregates reach 8 mm, so the figure is for comparison with other minerals, not a test anyone can run on a spray. The method is on how to measure specific gravity at home.
Does cacoxenite contain aluminium? The documented failure
The formula says one aluminium per 17 phosphorus; the sheet’s two analyses say none and 2.61. Method: each chemistry column was converted to moles of Fe, Al and P from its oxide weights, by script. The analysis from the Hrbek mine at Svatá Dobrotivá (St. Benigna), Czech Republic, prints P2O5 25.71, Fe2O3 41.46 and H2O 32.81, with no Al2O3: Fe:P 1.433. The Eleonore mine, Germany, analysis prints 2.89 Al2O3 and 0.14 insoluble: Fe:P 1.371, but (Fe+Al):P 1.524, with 2.61 Al per 17 P.
What that shows. The ideal column is the formula, not a specimen; real material scatters on both sides of it. Iron plus aluminium against phosphorus runs from 1.433 to 1.524 across the two analyses, against 1.471 in the formula. A single analysis, from either locality, would not reproduce the IMA formula, and neither the sheet nor the IMA list says how much aluminium a cacoxenite may carry.
Its iron-phosphate neighbours. On the same scale, Fe:P is 1.000 in strengite, 1.500 in beraunite and dufrénite and 1.667 in rockbridgeite, each worked from its IMA formula; cacoxenite’s 1.412 sits between. That ladder is set out on the strengite page. vivianite, the ferrous phosphate, is 1.500 with no ferric iron at all.
Where is cacoxenite found? Iron ores, not gem pockets
Where the sheet places it. Widespread in small amounts. In the Czech Republic, the Hrbek mine, Svatá Dobrotivá (St. Benigna), Třenice, Zbirov and Cerhovice; in Germany, the Eleonore and Rotläufchen mines near Giessen, Hesse, and Hühnerkobel near Zwiesel, Bavaria; Rochefort-en-Terre, Morbihan, France; Shanagolden, Co. Limerick, Ireland; Kiruna, Sweden. In the USA, the Palermo #1 mine, New Hampshire; Hellertown, Moore’s Mill and Noble’s mine, Pennsylvania; Indian Mountain, Cherokee Co., Alabama, with groups large for the species; the Wilson Springs (Potash Sulphur Springs) mine, the Avant claim near Shady and Three Oak Gap, Arkansas; the Vanleer mine, Tennessee; and the Palmetto mine, Florida.
The shared-place count. Method: 22 place names were taken by hand from the strengite sheet’s Distribution field and matched against cacoxenite’s by script, accent-folded and on word boundaries. 6 are on both: the Eleonore and Rotläufchen mines, Kiruna, the Palermo #1 mine, Indian Mountain and Three Oak Gap. Each sheet also names the other as an associate.
What the sheet does not say. It names no quartz, no amethyst and no Brazilian locality; its Occurrence line is oxidized iron ores, novaculites, sediments and soils. Anything said about cacoxenite inside quartz has to come from elsewhere, and the evidence we could fetch is weighed on what is cacoxenite in amethyst. The reference section continues at the species index. We hold no catalogue and nothing here is offered for sale; the wanted list is the only commercial route.