Willemite
Willemite is Zn2SiO4, hardness 5.5, density 3.89 to 4.19, crystals to 10 cm. It fluoresces yellow-green in both ultraviolet bands, which few species do.
Willemite is zinc silicate, Zn2SiO4, hexagonal, hardness 5.5, measured density 3.89 to 4.19, with prismatic crystals to 10 cm. The Handbook of Mineralogy records it as typically strongly fluorescent yellow-green to yellow-orange under both shortwave and longwave ultraviolet, and it may be phosphorescent. It is a secondary mineral of zinc deposits in limestone.
In short
- It fluoresces in both bands, strongly. Yellow-green to yellow-orange under shortwave and longwave, and it may phosphoresce afterwards. That combination is why willemite, not fluorite, is the species the fluorescent side of the hobby is built on.
- Franklin and Sterling Hill in New Jersey are the reason it is famous, and the pairing there is willemite green against calcite red in the same piece. Both species are recorded by the Handbook as fluorescing under both bands.
- Hardness 5.5 makes it a tough specimen for a fluorescent species — it will not scratch on a fingernail or a coin, and it takes handling far better than the soft secondaries it is often displayed with.
- Beware the two Broken Hills. The Handbook records willemite at Kabwe in Zambia, which was formerly called Broken Hill. That is a different place from Broken Hill in New South Wales, and old labels do not always say which.
- Not all willemite fluoresces. The Handbook's word is typically, not always, and the material that glows brightest carries manganese. A dark brown crystal from Tsumeb may do nothing at all under a lamp.
| Species | Bands named | Colour recorded | Also |
|---|---|---|---|
| Willemite | Shortwave and longwave | Yellow-green to yellow-orange, typically strong | May be phosphorescent |
| Calcite | Shortwave and longwave | Red, blue, yellow and other colours | Phosphorescent, cathodoluminescent, thermoluminescent |
| Aragonite | Longwave and shortwave | Red or yellow | Phosphorescent, cathodoluminescent |
| Adamite | Shortwave and longwave | Lemon-yellow | May phosphoresce |
| Pyromorphite | Longwave and shortwave | Yellow to orange | Piezoelectric if biaxial |
| Scheelite | Shortwave only | Bright bluish white | Also under X-rays |
| Cerussite | Longwave only | Yellow | — |
| Fluorite | Band not stated — just 'under UV' | Blue, violet, green, yellow, red | May be phosphorescent, thermoluminescent, triboluminescent |
| Hydrozincite | Band not stated | Pale blue to lilac | — |
Why willemite is the species the fluorescent hobby is built on
Plenty of minerals fluoresce. Very few do it brightly, in both ultraviolet bands, in a large crystal, at a hardness that survives being handled. Willemite does all four.
The Handbook of Mineralogy's line is worth quoting exactly: typically strongly fluorescent in yellow-green to yellow-orange under SW and LW UV; may be phosphorescent. Compare that with fluorite, where the entry says only that it fluoresces blue, violet, green, yellow or red under UV without naming a band at all — which is honest, because fluorite's response depends entirely on what is substituting into it, but it is no use to somebody deciding which lamp to buy.
The phosphorescence is the part that surprises people the first time. Switch the lamp off and good Franklin willemite continues to glow for a second or two as the excited states decay. It is not a trick of the eye and it is worth doing in a genuinely dark room with adapted vision. Our page on fluorescent minerals covers the physics and choosing a UV lamp covers the instrument.
Franklin and Sterling Hill, and the green-and-red pair
Franklin and Sterling Hill at Ogdensburg, Sussex County, New Jersey, are a zinc orebody in limestone, and the Handbook records large crystals of willemite from both. The associated species listed for Franklin — nasonite, glaucochroite, leucophoenicite — are ones you will meet nowhere else.
What made the district's reputation is a two-colour effect. Willemite fluoresces yellow-green; the calcite it sits in fluoresces red. Both species are recorded by the Handbook as responding under both bands. A single hand specimen under a shortwave lamp therefore shows two sharply distinct colours picking out two minerals that in daylight are a dull greenish-grey mass — which is the clearest demonstration available of what a lamp is actually for.
Two honest caveats. The dramatic Franklin effect is a shortwave phenomenon in practice, and it is much weaker under a longwave lamp than the pictures suggest. And Franklin closed in 1954 and Sterling Hill in 1986, so everything on the market is old stock — which does not make it scarce, since a great deal was collected, but does mean locality data depends entirely on the label. The Handbook's willemite sheet is the reference for the mineralogy; the provenance is a labels question, covered on what ex coll. means on a label.
The two Broken Hills, and other localities
There is a genuine trap in willemite's distribution list and it catches people. The Handbook records willemite from Kabwe in Zambia, and adds the former name in brackets: Broken Hill. Kabwe was called Broken Hill until 1966. It is a lead-zinc deposit in central Zambia and it has nothing to do with Broken Hill in New South Wales, which is a different orebody on a different continent.
An old label reading willemite, Broken Hill is therefore ambiguous, and the ambiguity is not resolvable from the mineral. If the rest of the label is African, or the associated species are the Zambian ones, that settles it; otherwise the honest record is that the locality is uncertain. This is a specific case of a general rule our locality page makes: a place name is only data if it identifies one place.
The rest of the list: the type area at Altenberg near Moresnet in Belgium, which is why the IMA gives Belgium as the country of first description in 1830; Tsumeb and Guchab in Namibia, where willemite occurs with malachite, mimetite, rosasite, duftite and cerussite; Balmat in New York; and in Arizona the Mammoth-St Anthony mine at Tiger and the Red Cloud mine in La Paz County.
Not all of it glows, and what the analyses show
The Handbook says typically strongly fluorescent, and the qualifier is real. Willemite occurs colourless, white, grey, flesh-red, dark brown, honey-yellow, apple-green and blue, and not all of it responds to a lamp.
The two Franklin analyses printed in the Handbook are suggestive. One carries 0.12 per cent MnO; the other carries 8.96 per cent MnO, with the zinc dropping from 72.11 to 61.38 per cent to accommodate it. So manganese substitutes into willemite in quantity, and it does so in the material from the district whose fluorescence is famous. We are not in a position to cite a measurement tying the one to the other on these specimens, so we do not assert the causal link here — only that the manganese is present, in the analyses, in the material that glows.
The practical consequence for a buyer is straightforward and we would rather say it than not. Fluorescence is a property of the individual specimen, not of the species. A Tsumeb willemite crystal is a fine thing and may be inert. If the response is what you want, the specimen has to be tested, and that is a brief to state explicitly on a wanted list rather than an assumption to make from a species name.