Metatorbernite
Metatorbernite is Cu(UO2)2(PO4)2 with 8 water, hardness 2.5, density 3.52 to 3.70. Torbernite loses 4 of its 12 waters to become it, shrinking about 15 to 19%.
Metatorbernite is the 8-water copper uranyl phosphate, Cu(UO2)2(PO4)2(H2O)4·4H2O, tetragonal, hardness 2.5, density 3.52 to 3.70, and radioactive. It usually forms as a pseudomorph after torbernite by stepped dehydration. Losing 4 of torbernite's 12 waters costs 7.14% of the mass, and the crystal ends up occupying about 15 to 19% less volume.
In short
- Metatorbernite is typically a torbernite crystal that has lost water. The Handbook of Mineralogy describes it as typically pseudomorphous after torbernite by stepped dehydration, so the square plate keeps its outline while the species inside it changes. The same page records it as probably present at every torbernite locality.
- The arithmetic of the change, worked from the IMA formulas: 4 of 12 waters leave, 7.14% of the mass; measured density rises from 3.22 to 3.52 to 3.70; so the same plate occupies roughly 15 to 19% less volume. A change of that size, on a matrix that is not shrinking with it, is consistent with the crazing seen on old plates.
- You cannot reliably tell the two apart by eye. The useful separators are instrumental: refractive index omega 1.618 to 1.631 against torbernite's 1.590 to 1.592, and density. A dull, more opaque green plate is suggestive only.
- Read the Handbook sheet's chemistry block, not its header. The current sheet, revised 30 January 2021, prints the arsenate formula in its title line; its own ideal-composition line, and the IMA list, give the phosphate.
- It is radioactive, and it is 50.77% uranium by mass against torbernite's 47.15%, because what leaves is water. No dose figures and no health advice here; storage is covered on how to store radioactive mineral specimens.
| Species | Formula (IMA) | Water molecules | Uranium by mass | Hardness | Density (measured) | Omega (refractive index) | Fluorescence on the sheet |
|---|---|---|---|---|---|---|---|
| Metatorbernite | Cu(UO2)2(PO4)2(H2O)4·4H2O | 8 | 50.77% | 2.5 | 3.52 to 3.70 | 1.618 to 1.631 | None recorded |
| Torbernite | Cu(UO2)2(PO4)2(H2O)4·8H2O | 12 | 47.15% | 2 to 2.5 | 3.22 | 1.590 to 1.592 | None recorded |
| Autunite | Ca(UO2)2(PO4)2(H2O)11 | 11 | 49.17% | 2 to 2.5 | 3.05 to 3.2 | 1.575 | Strong yellow-green |
| Meta-autunite | Ca(UO2)2(PO4)2·6H2O | 6 | 54.21% | 2 to 2.5 | 3.35 to 3.55 | 1.600 to 1.611 | Yellowish green |
| Saléeite | Mg(UO2)2(PO4)2(H2O)10 | 10 | 50.95% | 2 to 3 | 3.27 | Biaxial; beta 1.570 to 1.582 | Bright lemon-yellow in LW UV |
| Bassetite | Fe2+(UO2)2(PO4)2(H2O)10 | 10 | 49.28% | 2.5 | 3.40 to 3.63 | Biaxial; beta 1.610 | None recorded |
How much does torbernite shrink when it turns into metatorbernite?
This is the calculation behind the flaking, set out so it can be checked. Method: formula masses from standard atomic weights (U 238.029, Cu 63.546, P 30.974, O 15.999, H 1.008) applied to the September 2026 IMA-CNMNC Master List formulas; densities as printed on the Handbook of Mineralogy sheets.
Torbernite, Cu(UO2)2(PO4)2(H2O)4·8H2O, has a formula mass of 1009.72. Metatorbernite, with four fewer waters, is 937.66. The four waters that leave are 7.14% of the starting mass. Both ideal water contents agree with the Handbook's own calculated columns: 21.41% H2O for torbernite with 12 waters and 15.37% for metatorbernite.
Measured density goes the other way: 3.22 for torbernite, 3.52 to 3.70 for metatorbernite. Volume is mass over density, so the ratio of new volume to old is (937.66 ÷ 3.52 to 3.70) ÷ (1009.72 ÷ 3.22) = 0.81 to 0.85. The plate ends up about 15 to 19% smaller.
Two caveats, stated rather than buried. Torbernite's measured density is for material whose hydration the sheet does not state, and the Handbook's own header gives torbernite 8 to 12 waters, so the true figure for any one specimen could sit outside that bracket. The direction is not in doubt. A thin square plate on a rigid matrix losing that much volume would be expected to craze, lift or split on its perfect {001} cleavage, which is the likely origin of loose green flakes in an old torbernite box. The autunite version of the same arithmetic is on is autunite dangerous.
Can you tell metatorbernite from torbernite by looking at it?
Not reliably. The two species differ in water and structure, not in colour. Torbernite is recorded as emerald-green to apple-green and becoming dull on dehydration to metatorbernite; metatorbernite is pale to dark green with a vitreous to pearly lustre. A plate that has gone opaque and dull is suggestive, and nothing more.
What does separate them is measurable. Refractive index: torbernite omega 1.590 to 1.592, metatorbernite 1.618 to 1.631, a gap an immersion-oil test on a fragment will show. Optical sign: torbernite is uniaxial negative; the metatorbernite sheet gives uniaxial positive or negative, anomalously biaxial in sectors. Density: 3.22 against 3.52 to 3.70, in principle measurable, in practice not on a few milligrams of plates on matrix — our note on measuring specific gravity at home explains why small specimens defeat the method. Powder X-ray diffraction is the definitive test: metatorbernite's strongest line is 8.71 Å on Schneeberg material, torbernite's 10.30 Å on the synthetic 12-water compound.
In practice each of those tests needs a detached fragment, and diffraction needs powder. On a radioactive, friable species that is a decision for someone with the facilities, not a kitchen-table experiment. The honest label for an old, dulled Cornish plate is often “torbernite / metatorbernite, undetermined”, with the date and storage history recorded, as set out on how to label and catalogue a mineral collection.
Is metatorbernite ever primary?
Sometimes, according to its own sheet. The Handbook's occurrence line reads: typically a secondary mineral, a dehydration product of torbernite formed during weathering; formed directly above 75 °C. And its distribution line adds that material from the Gunnislake mine, Calstock, Cornwall, is thought to be primary — that is, grown as metatorbernite rather than dried into it. The analysis printed on the sheet is from Gunnislake.
That is the one locality the sheet singles out. Otherwise it says the species is widespread, probably occurs at all localities for torbernite, and was first described from Schneeberg in Saxony. The torbernite localities themselves — Jáchymov, Schneeberg and Johanngeorgenstadt, a number of Cornish mines including Wheal Basset and Stenna Gwyn, Musonoi and Shinkolobwe in Congo — are listed on the torbernite page rather than repeated here. The Cornish ground is covered on our Cornwall notes.
One disagreement between the two sources, recorded rather than resolved. The Handbook sheet says metatorbernite was first described from Schneeberg; the IMA list gives the United Kingdom as the country of first description, the year as 1916 and the first reference as Mineralogical Magazine 17 (1916), 326. We have not read the 1916 paper and do not adjudicate.
Why the Handbook sheet's header formula is wrong
A small point that will catch anyone who copies formulas. The current Handbook sheet for metatorbernite, revised 30 January 2021, prints Cu(UO2)2(AsO4)2·8H2O in its title line. That is an arsenate, and it is the composition of metazeunerite, which the IMA list writes Cu(UO2)2(AsO4)2(H2O)4·4H2O.
The same sheet's chemistry block gives the ideal composition as Cu(UO2)2(PO4)2·8H2O, its analysis reports P2O5 and no arsenic, and the IMA list gives Cu(UO2)2(PO4)2(H2O)4·4H2O. Three statements against one: metatorbernite is the phosphate. When a reference header and its own analysis disagree, trust the analysis.
For the collector, the practical consequence is on labels. A specimen written up as a copper uranyl arsenate is either zeunerite, metazeunerite or a copied error, and those are separate species on the IMA list. The rest of this reference section is indexed at the species index. There is no catalogue here and nothing here is offered for sale; the wanted list is the only commercial route.