Green-luminescing hyalite opal from Zacatecas, Mexico.
Gems coloured mostly by luminescence are truly rare. Therefore, when a new gem material is discovered that is undoubtedly coloured by luminescence, it attracts a great deal of attention. Such is the case with the daylight-fluorescing, vivid green transparent hyalite that was recently unearthed in Zacatecas, Mexico (Fritsch et al., 2014; see Figure 1 and the cover of this issue). The material was first brought to market in October 2013, when five weathered specimens of rough opal showing moderate green daylight luminescence were given to one of the authors (PKMM) by a well-known importer of mineral specimens and lapidary rough from Mexico. Some of the initial specimens were sold at the February 2014 Tucson gem shows, and they attracted much attention because of their noticeable change from nearly colourless to vivid green in indirect sunlight (Moore, 2014). Reaction to the phenomenon by specimen collectors was very positive, so reconnaissance prospecting was undertaken by author PKMM to help determine the quality and quantity of material available. This initial work showed that the strongly luminescent material is sparsely distributed, but a small amount of facetable opal was recovered. Some of the faceted gems were available at the 2015 Tucson shows, and were sold as Electric Opal (trademark pending) because of their vivid green colour in daylight. The largest faceted gem weighed almost 6 ct (see cover of this issue), and hundreds of smaller stones were available in calibrated sizes down to 3 mm.
The mining site lies on the top of a steep 250-m-high mesa, so the use of heavy equipment is impractical and all digging has been manual. The difficult access, sparseness of material and reliance on manual labour will probably limit future supply to a small multiple of the production to date. This article documents in detail the geological occurrence and the luminescence phenomenon that makes this opal such an attractive gem material.
Opal Generalities and Classification of Hyalite
The gem species opal--Si[O.sub.2]xn[H.sub.2]O--is typically associated with play-of-colour material from Australia and more recently Ethiopia, both of which are abundant in today's market. But the term opal more generally refers to a range of hydrated silica materials. In the broadest sense, opal may be divided into microcrystalline and amorphous material (Graetsch, 1994; Gaillou et al., 2008a). The microcrystalline varieties comprise opal-C (rare) and opal-CT (quite common), which are basically poorly crystallized cristobalite (C) with more-or-less tridymite-like (T) stacking. Both cristobalite and tridymite are polymorphs of nominally pure Si[O.sub.2]. They represent the bulk of 'volcanic' opal, such as fire opal from Mexico (Fritsch et al., 2006) and material from Ethiopian deposits (Rondeau et al., 2010). Amorphous opals are divided into two categories (Langer and Florke, 1974): opal-AG and opal-AN. Opal-AG is a gel-like (G) aggregate comprised of regularly ordered hydrated silica spheres, often <1 [micro]m in size, with open space between the spheres (e.g. Stephant et al., 2014). It is quite common, often displays play-of-colour caused by diffraction of light from regularly stacked spheres of similar diameter, and includes material from Australia and Brazil. Opal-AN consists of hydrated silica molecules that are network-forming (N), and therefore it resembles a glass (Florke et al., 1973). Since opal-AN does not have a regular array of spheres or any specific microstructure, it is always 'common' opal (i.e. does not exhibit play-of-colour). Opal-AN is also known as hyalite. It is relatively uncommon in volumes adequate for faceting, and thus is rarely used as a gem material. One of the few examples is botryoidal white opal from Milford, Utah, USA (Johnson and Koivula, 1997).
Most hyalite strongly fluoresces green under both long-wave (LW) and short-wave (SW) UV radiation. In some cases (e.g. from the Erongo Mountains of Namibia, Madagascar and the Thomas Range in Utah, USA; www.mindat.org), hyalite may show a moderate-to-strong yellow-green coloration in daylight, as judged from photographs and unpublished reports, with the green component possibly originating from luminescence (although this has not been confirmed). Strong green fluorescence to visible light has not been reported from these localities, however, making the material described in this article quite unusual.
Crowningshield (1985) described one example of a transparent faceted opal with strong green daylight fluorescence that allegedly originated from Mexico. Its properties, general appearance and microtexture closely match those of the present material. Since there was no evidence of older workings in the currently known mining area, that stone may have come from the surface at the current area or from another incidental find, possibly in a similar geological context elsewhere in Mexico.
Other Daylight-Fluorescing Materials
Luminescence probably contributes to the coloration of several gem varieties, but its influence on visual appearance is hard to quantify (Fritsch and Rossman, 1988; Fritsch and Waychunas, 1993). Nevertheless, there are a few well-documented examples with luminescence induced by visible light, and probably the most famous category comprises certain yellow diamonds, such as a stone owned by Pedro II, the last emperor of Brazil (Moses, 1997; Fritsch, 1998; Shigley and Breeding, 2015). Such yellow to yellow-green gems are commonly referred to as 'green transmitters' (although they should really be called green emitters), and their coloration comes from the well-known H3 colour centre (Fritsch, 1998). More recently, HPHT-treated brown type Ia diamonds have demonstrated the same optical effect. Some green amber also owes its colour to luminescence, although it is rarer than blue amber (which is also coloured by luminescence; Liu et al., 2015).
[Mn.sup.2+]-doped flame-fusion synthetic spinel often exhibits bright green luminescence, stemming from its dopant, Mn2+, in tetrahedral coordination (Kane and Fritsch, 1991). As a man-made material, it has
attracted much less attention for its daylight-induced emission than the present opal. Two other compounds are known for their bright green daylight-induced colour component. The first is 'uranium glass'--also known as 'Uralite glass' (unrelated to the Ural Mountains of Russia)--which usually is a silicate glass doped with traces of uranium. This material was sometimes faceted as a gem, or made into other objects (particularly tableware), and was very popular in the early 20th century. In the USA, its opaque variety is known as 'Vaseline' glass, a popular collectors' item. The second material is 'fluorescein', an organic-based molecule exhibiting strong green daylight luminescence. Fluorescein is used for many monitoring applications, ranging from cell chemistry to tracing underground water flows to highlighting space capsules after splashdown in the ocean. All of these green materials are especially attention-getting because of their daylight fluorescence.
Many gems besides hyalite may have their colour appearance boosted by luminescence emitted in other parts of the visible range. This has been documented for the Tavernier diamond (Liu et al., 1998) and some rare orange diamonds (Titkov et al., 2015). This is also true of ruby, and probably red spinel (Fritsch and Waychunas, 1993). It is likely that the highly sought-after (but poorly defined) 'pigeon-blood' colour of ruby owes part of its attractiveness to daylight-induced red luminescence caused by chromium.
The hyalite deposit is located in western Zacatecas State, central Mexico (Figure 2). The landowner has requested that more specific information on the location be withheld to minimize unauthorized mining. From the nearest municipality, the area is reached by approximately 40 km of paved road followed by 25 km of dirt tracks that wind through bean and squash fields to the base of a 250-m-high mesa (Figure 3). Then, a rough trail leads 1.5 km up a boulder-strewn and cactus-infested slope to the diggings. The miners have established a camp on site, and return to their homes on weekends to process each week's production.
Zacatecas State is largely underlain by the Sierra Madre Occidental Volcanic Province (McDowell and Clabaugh, 1979), and the hyalite is hosted by mid-Tertiary (37-28 million year old) rhyolitic rocks that resulted from repeated volcanic eruptions.
To unravel the geological origin of the opal, it is important to understand the formation of its volcanic host rocks, which were deposited by violent eruptions from 'super-volcanoes' similar to Yellowstone in the USA. Instead of erupting though a central vent like familiar conical volcanoes (cf. Mount Fuji, Mount Rainier or Mount Vesuvius), these erupt around the circumference of the magma chamber (tens to hundreds of kilometres in diameter) through an annular ring-fracture zone (Smith and Bailey, 1968). During eruption, rocks within the ring-fracture zone settle piston-like into the underlying space left by the erupting magma, creating a circular depression called a caldera into which some of the hot ejecta accumulate. Thinner outflow facies extend outward from the intracaldera zone. Much of the erupting lava chills instantly into frothy siliceous volcanic glass, which is shattered into myriad tiny shards and pumice fragments that mix with fine volcanic ash to comprise tuff. As the eruptive pile cools, settles and compacts, escaping heat and volatiles fuse the ejecta into a hard, compact rock known as welded ash-flow tuff or ignimbrite. The most densely welded tuffs are compacted into vitrophere, a crystal-rich volcanic glass containing rock fragments. The thinner and cooler parts of the eruptive pile escape welding, leaving unconsolidated ash mixed with loosely bound crystals (mostly feldspars and quartz) and rock fragments (i.e. poorly welded crystal-lithic tuffs; Smith and Bailey, 1968). In all cases, the volcanic glass that makes up most of the tuff is unstable and decomposes over time. Silica, uranium, iron and other constituents liberated from the decomposing volcanic glass are initially incorporated into vapours emanating from the cooling volcanic pile, which interact with circulating heated groundwaters that emerge as hot springs (Goodell and Waters, 1981; Breit and Hall, 2011). Transformation of glass shards into clays also releases silica and uranium into descending surface waters, and such solutions may redeposit silica as opal, chalcedony or quartz, depending on a number of factors (e.g. Berger et al., 1994; Markusson and Stefansson, 2011). Uranium and other liberated elements are also deposited, either as discrete mineral species or within the framework structure of the chalcedony or opal.
Within the mining area, the daylight-fluorescent hyalite occurs around the base of a 40 x 125 m elongate knob that protrudes from the top of a 1.3-km-long mesa composed of outflows from a 20-km-diameter caldera. The ring-fracture zone lies approximately 450 m from the hyalite occurrence. The mesa consists of a 90-m-thick section of very densely welded rhyolite vitrophere that overlies 160 m of moderately welded ash-flow units (Figure 3). The hyalite occurs in a 10-15 m thick, poorly welded, crystal-rich rhyolite tuff that sits atop the vitrophere and was protected from erosion by an overlying 30 x 100 m remnant of a younger welded tuff unit (Figure 4). The hyalite-bearing tuff probably once covered the entire mesa but it, and any opal it contained, has long-since been eroded way.
The hyalite-bearing layer is composed of 3-50 mm rhyolite rock fragments mixed with 1-4 mm dipyramidal fe-quartz and limpid sanidine phenocrysts in a weakly welded, porous, siliceous, tuffaceous matrix. Groundwater could have readily percolated through this highly permeable unit, leaching silica and uranium from the devitrifying glass, and precipitating opal and the associated species described below. The hyalite occurs sparsely in voids and in fractures cutting the host tuff unit (Figures 5 and 6). The hyalite locally fills the fractures completely, but more commonly it occurs as isolated blebs and botryoidal coatings up to 3 cm in thickness, in most places covering less than 10% of the surface. The intensity of the opal's fluorescence varies widely within and between fractures and pockets, with less than 10% showing the strongest luminescence. The miners report that the daylight fluorescence is strongest in material found within a metre or so of the surface and drops off rapidly below that, so few fractures have been followed any deeper.
A series of uranium-containing minerals is found on the surface of the veins, and the opal locally formed on top of them. They generally consist of yellow sprays or spherulites of fibrous crystals. The most prominent species were identified by Spano et al. (2015) with X-ray diffraction as meta-autunite (Ca[(U[O.sub.2]).sub.2][(P[O.sub.4]).sub.2x6-8[H.sub.2]O), haiweeite (Ca[(U[O.sub.2]).sub.2][[Si.sub.5][O.sub.12][(OH).sub.2]]x6[H.sub.2]O), uranophane (Ca[(U[O.sub.2]).sub.2](Hsi[O.sub.4])x5[H.sub.2]O), and metauranospinite (Ca[(U[O.sub.2]).sub.2][(As[O.sub.4]).sub.2]x8[H.sub.2]O). The silica that formed the opal probably originated from groundwater-based weathering and dissolution of the glass shards in the tuff, as described above.
Mining and Production
Mining of the hyalite has been done exclusively with pry bars, picks and shovels, and a small gasoline-powered drill. The miners locate and follow iron-oxide stained fractures, and dig until hyalite-bearing voids are encountered. During the rainy season, water collected in earlier excavations is used to wash out the pockets to assist with recognizing the better-coloured opal. During the dry season, the clayey iron-oxide-stained vein fillings fall away to reveal the opal. The voids range from 2 to 20 cm across and each typically produces a few kilograms of rough (including matrix). Weekly production ranges from 100 to 200 kg total of opal-bearing rock, of which less than 5% is specimen or cutting quality (e.g. Figures 7 and 8).
Through early 2015, the production totalled approximately 30 kg of top-quality specimens, 10 kg of facet-grade rough and 30 kg of pieces containing areas suitable for cutting good cabochons. Another 150 kg of moderate-quality and 250 kg of low-quality specimen material also were mined. It is difficult to predict future reserves, in part because the depth and lateral extent of the strongly coloured zone of opal mineralization is unknown. Nevertheless, it appears that there is good potential for another 100-250 kg of top-quality specimen and gem rough.
Cutting and Polishing
Most of the larger gems faceted so far have been cut by one of the authors (MG). The grinding and polishing of this opal have presented no unusual problems compared to transparent opal from other localities. The minor heat sensitivity of opal requires the use of a cold dopping method for attaching the stone to the dop stick. The low RI of this opal (1.45-1.46; see below) requires a pavilion angle of at least 45[degrees], as the critical angle is 43.6[degrees] (based on the lower RI range). The low SG of opal dictates that it takes more volume to attain a given weight as compared to most other gems. For example, a 1 ct standard round brilliant would measure nearly 8 mm in diameter.
The main problems with cutting the material are the various types of inclusions. Fibrous minerals are commonly associated with hollow tubes in the opal, and voids occur near the remnants of the matrix or at the intersection of botryoidal domains. If opened during the cutting process, these voids fill with debris from the wheel, so they must be either left completely enclosed within the finished gem or entirely ground away.
Materials and Methods
All of the rough samples were obtained by one of the authors (PKMM), either directly from the miners or through his Mexican mineral dealer associate. Three polished pieces were prepared by Jacques Le Quere (JackyGems, Auray, France) for this research: one round brilliant (1.32 ct), one flat rectangular cabochon (1.81 ct) and one piece with the botryoidal surface kept on one side and a rounded cabochon polished on the other side (0.83 ct). Five matrix specimens were examined, including two on which green-fluorescing pale yellow hyalite formed in continuity with milky white hyalite showing no daylight luminescence.
All rough and cut samples were examined with a binocular Leica MZ6 microscope equipped with a Nossigem gemmological lighting system, and the three polished opals were tested with a GIA Gem Instruments refractometer and a Mettler Toledo XS104 electronic scale (for mass and hydrostatic SG measurements). UV luminescence observations of all samples were done using a VL-215.LC Vilber-Lourmat filtered UV lamp that contained two ~30-cm-long UV tubes of 15 W each: one for 254 nm SW and the other for nominally 365 nm LW UV radiation. For qualitative intensity descriptions, the sample was placed at a fixed distance of 7 cm from the lamp. Fluorescence colour and intensity were judged compared to two reference stones: a flame-fusion synthetic ruby used as a 'very strong' and 'red' reference for both LW and SW, and a colourless flame-fusion synthetic sapphire representing a 'strong' and 'blue' reference in SW (inert in LW). For this specific case of a visible-light luminescing gem material, we used a 405 nm (InGaN) 0.5 mW laser pointer to stimulate emission under violet radiation. (These devices are often incorrectly referred to as 'blue' lasers, and they may be harmful to the eyes, even operating at such low power.) Two additional faceted stones of large size (12.25 and 12.63 ct), which were cut by author MG to show the diversity of their internal features, were examined by one of us (NR) using a Nikon SMZ1500 trinocular microscope fitted with a Nikon DS-Ri2 camera, and selected inclusions were analysed with a Renishaw InVia Raman spectrometer utilizing a 514 nm argon-ion laser.
The identity of purple blocky crystals found within tubes in the hyalite was established as fluorite using single-crystal X-ray diffraction. Crystals suitable for diffraction were manually isolated, and one was selected for unit-cell determination. A matrix of data was collected using a series of 0.3[degrees] frames in [C] with a collection time of 10 seconds/frame. Data were collected on an Apex II diffractometer with Mo(Xa) X-radiation and a Bruker charge-coupled device (CCD) detector. Unit-cell dimensions were obtained through refinement of 718 reflections using Fast Fourier transform techniques.
Raman spectra of all samples were obtained with a Bruker MultiRAM Fourier-transform Raman spectrometer. Excitation was provided by a 2 W Nd:YAG laser at 1064 nm, in continuous mode, and the data were collected at the standard resolution of 4 [cm.sup.-1], accumulating 1,000 scans.
Imaging of the opal microstructure in one rough sample was performed using a JEOL 7600 scanning electron microscope (SEM). The opal was fragmented using a mortar, and the pieces were etched in hydrofluoric acid (10% HF by volume) for 30 seconds (a standard etch procedure to reveal opal structure; e.g. Gaillou et al., 2008b). The freshly broken and etched surfaces were coated with platinum to ensure electrical conductivity.
Luminescence emission and excitation spectra as well as 3D block diagrams were obtained for the three fashioned opals and two of the rough samples using a Horiba Fluorolog spectrofluorometer in conjunction with Fluor-Essence software. Two types of detectors were used: a standard photomultiplier and a Synapse CCD detector coupled to an iHR320 imaging spectrometer. A resolution of <1 nm was obtained by juxtaposing many scans to cover a given spectral domain in several sections. The samples were positioned to maximize the measured emission. For comparison, we analysed a reference specimen of whitish hyalite crust with green UV luminescence from Grimes Sexton Quarry, Inyo County, California, USA. Emission spectra were obtained using 365 nm excitation with a 1 nm bandpass, a sampling interval of 1 nm and a 0.2 second integration time. Excitation spectra were obtained for the 524 nm emission using the photomultiplier with a 3 nm spectral bandwidth; the excitation was scanned with a 1 nm bandpass and sampling interval, and a 0.3 second integration time. In addition, emission spectra of a 0.23 ct round brilliant were recorded on a GGTLPL6 Raman/photoluminescence system using 405 nm laser excitation, and a high-resolution echelle spectrograph (average resolution of 0.04 nm) by Catalina Scientific equipped with a Peltier-cooled sCMOS camera. The spectra were recorded with the opal at room temperature and also cooled to 77 K by direct immersion in liquid nitrogen.
UV-visible (UV-Vis) absorption spectra were recorded using a Cary 5G spectrophotometer with a resolution of ~1 nm (sampling and spectral bandwidth were each 1 nm). Each point was collected for 0.1-0.5 seconds. The spectrum selected for illustration in this article was obtained from the 1.81 ct flat cabochon sample, and then double-checked on five other rough samples.
Energy-dispersive X-ray fluorescence (EDXRF) analysis of the three polished samples was performed with a Rigaku NEX CG spectrometer (a 'secondary target' or 'Cartesian geometry' instrument), with analytical parameters optimized for four successive energy ranges of secondary X-rays. This provided enhanced sensitivity, intermediate between that of standard EDXRF and ICP-MS techniques. The chemical analyses are reported as if water is not present, since this technique cannot detect water.
Water content was determined with thermogravimetric (TGA) analysis, using a Netzsch Jupiter STA 449 F3 instrument on a 60 mg sample that was powdered in an alumina crucible. The sample was heated from room temperature to 1,000[degrees]C (1[degrees]C/minute) under a synthetic dry air flow (nitrogen-oxygen mixture with [N.sub.2]:[O.sub.2] = 80:20) of 50 ml/minute.
The radioactive dose rate of the three polished samples (total weight ~0.8 g; Figure 9) was measured for ~1/2 hour with the NaI detector of a Canberra InSpector 1000 instrument. The measurements were done with the samples placed face-down and directly on the detector, which is a geometry comparable to wearing the three hyalites directly on the skin. The nature of the radioisotopes was revealed by gamma spectroscopy of the 1.81 ct flat rectangular cabochon, using a high-purity germanium detector (Canberra XtRa) with low background. The efficiency of the detector was defined for each point of the sample chamber, and the volume of the gem was modelled to increase efficiency and precision (hence the choice of the flat cabochon, which was easiest to model). Because of the low levels measured, a count time of 250,000 seconds (almost three days) was used. These data were collected at a laboratory specialized in the measurement of low radioactivity levels (Subatech-SMART) at the Nantes School of Mines, Nantes, France.
Results and Discussion
Visual Observations and Gemmological Properties
The rough specimens showed a botryoidal or concretion-like morphology (e.g. Figures 7 and 8), which is typically associated with hyalite opal (opal-AN) from many localities. Their diaphaneity ranged from transparent to opaque. The colour appearance of the rough and cut material showed a dramatic change from incandescent or fluorescent lighting (colourless to moderate yellow) to shaded sunlight (vivid greenish yellow to yellowish green; e.g. Figure 1 and the cover image). This change arises from visible-light-induced green luminescence. The phenomenon is apparently enhanced by faceting, which provides a longer optical path length for excitation, as well as reflection of the emitted green colour toward the eye as compared to rough specimens. Shaded sunlight is best for viewing the phenomenon, since the brightness of direct sunlight overwhelms the fluorescence.
Exposure of all the samples to the SW UV lamp excited green luminescence that was even more intense than our 'very strong' luminescence reference stone (e.g. Figure 10, left); it was one of the most intense reactions we have documented in a natural gem. All samples, large or small and with or without zoning of diaphaneity, showed an apparently homogeneous luminescence. The fluorescence was only slightly weaker in LW UV, but still much more intense than our 'very strong' reference stone (Figure 10, right). Excitation with a 405 nm laser pointer, emitting in the violet region of the visible spectrum, also spectacularly induced the green coloration (Figure 11).
RI values ranged from 1.45 to 1.46, and 1.455 was measured from a flat polished surface of one faceted stone. SG values spanned from 2.12 to 2.16. These figures are typical of opal in general. No hydrophane effect was noticed while doing the hydrostatic measurements, suggesting that the samples were not porous. Neither did the stone appear to absorb RI contact liquid.
Viewed with the microscope in transmitted light, the botryoidal nature of the opal was easily observed in the faceted stones. Curved growth zoning resulted from slight differences in the index of refraction between the thin successive layers (Figure 12). Yet, we observed no noticeable colour zoning in diffused lighting; this is consistent with Figure 11 of Crowning-shield (1985). The curved growth zoning showed corresponding anomalous birefringence when observed between crossed polarizers (Figure 13). Such spectacular high-order interference colours reflecting the depositional layering of the opal are characteristic of hyalite (Graetsch, 1994).
Other internal features in both the rough and cut stones consisted mostly of small rock fragments, fractures (often nearly planar), colourless-to-yellow mineral inclusions (Figure 14), and very small two-phase inclusions containing liquid and a bubble (Figure 15). The fluid inclusions were trapped between curved botryoidal growth zones and as irregular flat planar fluid bodies between depositional layers. The mineral inclusions consisted of prismatic single crystals with a hexagonal cross-section and crystal clusters that were observed scattered throughout the two large faceted stones. Raman analysis confirmed them to be a phosphate. Not seen in our samples, but documented by Spano et al. (2015), were uranyl phosphate minerals contained within transparent silica tubes, which gave an acicular appearance to the normally platy U-phosphates.
Elongate tubular voids were locally present in our samples, and may represent areas in which U-bearing inclusions were leached after opal deposition. Some of these tubes, hosted by opal protuberances on some rough samples (Figure 16, left), contained minute blocky purple crystals of fluorite (Figure 16, right) that were identified by X-ray diffraction. The tubular inclusions were common in one of the large cut stones examined. They had blocky outlines and were filled with an amorphous solid material that could not be identified with Raman microspectroscopy.
Identification as Hyalite Opal-AN
This Mexican opal does not display play-of-colour, suggesting that its microstructure probably does not consist of the regular stacking of spheres as in opal-AG (such as typical opal from Australia). Instead, visual observations suggest that the samples are opal-AN. Commonly called hyalite, this is probably the rarest variety of gem opal.
Fourier-transform Raman scattering spectroscopy is a reliable method to characterize opal, as it avoids many of the challenges caused by opal's strong luminescence (e.g. Smallwood et al., 1997; Ostrooumov et al., 1999; Rondeau et al., 2004). The specimens yielded spectra typical for opal-A (e.g. Figure 17), showing a distinctive broad band with a maximum at ~435 [cm.sup.-1], and an abrupt drop in Raman scattering at ~500 [cm.sup.-1]. This is different from typical opal-AG, which shows a more Gaussian-like (symmetrical) band centred at ~420 [cm.sup.-1] (Rondeau et al., 2004). Bands also were observed at ~790, 970 and 1075 [cm.sup.-1], as noted in the publications cited above. A relatively broad band terminating in a sharp peak at about 1550 [cm.sup.-1] seems unique to this Mexican hyalite; it has never been observed in the Raman spectra of other opal-AN specimens studied by the authors, and is not documented in the literature. The broad band with an apparent maximum at 3110-3120 [cm.sup.-1] is related to water, but its intensity was not reproducible with our equipment and thus is not a reliable indicator of water content. Raman scattering therefore established that this hyalite is opal-A, but was not sufficient to establish unambiguously that it is opal-AN.
To establish the exact variety of this opal-A, it was necessary to determine its microstructure through SEM imaging. At all magnifications, the samples showed the characteristics of a glass, with a smooth surface containing no holes or spheres (Figure 18). There were only traces of conchoidal fractures and associated patterns, as in glass. These features identified it as opal-AN (cf. Graetsch, 1994).
Luminescence and UV-Vis Absorption Spectroscopy
Green UV luminescence of opal is common and well understood. It is usually stronger in SW than in LW UV excitation, and it originates from [U.sup.6+] in the form of the uranyl [(U[O.sub.2]).sup.2+] ion (cf. Gaillou et al., 2011). What is striking about this hyalite is that the luminescence is also excited in daylight. To understand this behaviour, we first collected the emission spectrum, which is the spectral composition of the luminescence. The green luminescence of the five samples was identical, and showed typical uranyl emission (e.g. Figure 19), with a wide band extending from about 450 nm (end of the violet range) to almost 700 nm (red region). There were five maxima, at ~504, 524, 546, 572 and 604 nm; the most intense was at 524 nm, in the green region, as expected from the colour of the emission. The emission was the same whether obtained using a standard dispersive spectrometer with 365 nm excitation (Figure 19, left) or using a 405 nm laser (Figure 19, right). At liquid-nitrogen temperature, the emission peaks shifted slightly toward shorter wavelengths.
To understand the cause of a given emission, we collected an excitation spectrum across all excitation wavelengths, measuring the intensity of the green fluorescence excited or induced by each one. In this case, excitation spectra were obtained for the 524 nm emission (the maximum of the emission), for the same five samples (e.g. Figure 20, bottom curve). The results showed several intense bands in the UV, increasing gradually toward the SW range. This explains why the green luminescence was more intense under SW UV radiation. However, an additional, less intense pattern was present in the visible range, extending from about 375 nm (still in the UV) to about 445 nm, covering almost all of the violet range. This feature showed several maxima at ~400, 409, 419 and 432 nm. This pattern of excitation in the violet part of the visible range explains why the green luminescence is excited by visible light, and why daylight, which contains violet radiation, so effectively excites the luminescence. It also explains why 405 nm laser light produces such intense fluorescence.
The UV-Vis absorption spectra (e.g. Figure 20, top curve) were very similar regardless of body colour (which ranged from pale to moderate yellow). Absorption in the violet range explains the yellow body colour. This pattern is well known to be due to uranyl absorption (cf. Gaillou et al., 2011; Smith et al., 2013; Faulques et al., 2015 and references therein).
The almost perfect superposition of the excitation and absorption spectra (again, see Figure 20) is a rare occurrence in natural luminescent materials. It suggests that all the energy absorbed by the uranyl molecule is transformed into green light. This makes the system remarkably efficient (similar to a ruby laser). In addition, this means that there is little to no outside influence on the luminescence/ excitation, which is rare in natural systems. Often, for example, one expects [[O.sup.2-][left right arrow] [[Fe.sup.3+] charge transfer to introduce modifications between absorption and actual excitation in natural gem materials (Fritsch and Waychunas, 1993). The broadness of the emission bands in this hyalite is typical of a disordered material, not of a crystalline impurity. This strongly suggests that the uranyl ion is dispersed in the opal matrix.
Another way to view this hyalite luminescence is to plot a 3D block diagram (Figure 21) showing the emission spectra from 500 to 650 nm (red traces) as a function of excitation wavelength from 248 to 490 nm, in 3 nm intervals. The emission spectra show the same shape (with the five maxima described above, in the same relative proportions) within the excitation range that was investigated, which confirms that only one luminescence centre is involved. Only the intensity of the emission varies with excitation. Within the excitation range there are two regions of maximum emission: one is in the UV and the other is in the visible range (centred within the violet). This is consistent with the excitation spectrum noted above. However, because the 3D scan was not obtained using the same detector, and is uncorrected, the overall shape of the excitation curve is different from that described above, especially where it dips in the UV. Nevertheless, this type of data is useful to visualize the two regions of excitation.
These results were compared to those obtained for our green-luminescing reference hyalite from Inyo County, California. This sample showed the typical behaviour of stronger luminescence in SW than in LW UV radiation, and no emission with daylight illumination. The spectral data were essentially similar in the UV, but there was only a very weak maximum in the visible range. When a 3D block diagram was plotted in a similar fashion to that shown Figure 21, only the UV excitation zone was clearly seen.
Chemical Analysis and Water Content
The EDXRF data should be considered slightly high since the water content of the opal was not taken into account by the instrument. The uranium concentration ranged from approximately 20,600 to 32,400 ppmw [UO.sub.2]. This is consistent with the value of 22,604 ppmw [UO.sub.2] obtained by Spano et al. (2015) using electron microprobe analysis; they also recorded about 94.5% wt.% Si[O.sub.2] and 6,384 ppmw CaO. Calcium was also the only other significant impurity we detected (13,300-35,800 ppmw). Spano et al. (2015) proved that U is incorporated into the opal together with Ca. (All the accompanying uranium minerals also contain Ca.) We detected some Fe (<1,000 ppmw), which is much less than the 3,000 ppm necessary to quench green luminescence (Gaillou et al., 2008a). Possible traces of K were noted in all of the analyses, and one sample showed an unquestionable trace of arsenic (~1,000 ppmw).
The water content of the opal determined by TGA analysis was 2.71 [+ or -] 0.3 wt.%. This is one of the driest opals we have measured. Added to the chemical analysis determined by Spano et al. (2015), the sum of oxides and water does not quite reach 100% (rather ~97%). This discrepancy is not surprising, considering the very different principles of operation of the two instruments used and that these data originate from different samples.
The ambient background level of radioactivity was 0.052-0.065 [mu]pSv/h, and the three opals measured together were in the range of 0.0550.069 [mu]Sv/h. Therefore the radioactivity level of these samples was barely above background. Gamma spectroscopy revealed that only U-related radioisotopes were present above detection limits, despite the long counting time of almost three days. Three decay products of the [sup.238]U family were measured: [sup.234]Th at 19.4 [+ or -] 8.7 Bq/g; [sup.214]Pb at 25.3 [+ or -] 2.8 Bq/g and [sup.210]Pb at 26.5 [+ or -] 3.8 Bq/g. (One becquerel per gram means one disintegration per second and per gram of a particular radioisotope.) Our measurements of such low radioactivity levels over an extended period of time showed that this hyalite is not dangerous to wear (or to cut and polish into a gemstone).
In the experience of one of the authors (PKMM), a significant volume of the bulk opal material will set off radiation detectors at the USA-Mexico border. However, this instrumentation is geared toward the detection of the slightest amount of radioactivity, and has been criticized for its inability to distinguish gamma rays originating from nuclear sources versus those "from a large variety of benign cargo types that naturally emit radioactivity, including cat litter, granite, porcelain, stoneware, banana etc." (http://en.wikipedia.org/wiki/Radiation_Portal_Monitor). Other gem materials that likely contain some amount of U-bearing opal--such as some chalcedony, 'crazy lace' agate (or 'laguna lace' agate) and geodes from Las Choyas, Chihuahua State, Mexico--also are known to set off such detectors.
Origin of Daylight-Induced Luminescence
The main interest in this hyalite is its daylight-induced luminescence (e.g., Figure 22). Why is it observed in this opal and in almost no others? It is apparent from our measurements and the literature (in particular, Gaillou, 2006) that most opals luminescing green to UV radiation show very similar excitation spectra to those of this Mexican hyalite (i.e. uranyl absorption in the violet and blue ranges). And indeed, most opals that fluoresce green to UV radiation also will show green luminescence under a 405 nm laser, which is visible light. However, our eyes do not perceive any green emission when these gems are observed in normal daylight. Why?
Several phenomena compete for the perception of a colour: light transmitted (the body colour), light scattered ('haze' in colorimetric terms) and light reflected (gloss) by an object. The most common result of light scattering is seen in milky gems (e.g. white diamond, jade or opal) or the blue adularescence of moonstone, but luminescence is also a scattering phenomenon. The green colour component of the hyalite described in this article belongs to scattered light encountering the eye. One of the matrix specimens examined by the authors was covered by green-luminescing hyalite that abruptly graded into milky white hyalite that showed no daylight luminescence (Figure 23, left). Under UV radiation (or the 405 nm laser), the green luminescence was equally intense over the entire specimen (Figure 23, right). The emission spectra obtained from all of the opal on the specimen were the same, and surprisingly the excitation spectra also were identical. Therefore the entire specimen also should have displayed the same green luminescence in daylight--if this phenomenon could be fully described by spectroscopy alone. Since the visual perception of the specimen was quite different, the daylight luminescence must be dependent upon the interplay of other factors. First, there must be an absence of any sort of luminescence quencher, such as [Fe.sup.3+] . Second, an absence of other forms of scattering besides luminescence is important: for example if the sample is milky, it is unlikely that daylight luminescence will be seen. This second condition implies that the gem has to be transparent. Third, the presence of uranyl ions is a prerequisite, probably in a certain range of concentration: if there is not enough, the luminescence will be too weak, and if there is too much, the molecules will absorb one another's emission (a process called concentration quenching or self-quenching; Fritsch and Waychunas, 1993), reducing the overall luminescence.
The excitation spectrum of the hyalite basically indicates that its luminescence may be excited by visible light, but other conditions, probably not limited to those cited above, are necessary for this rare phenomenon to occur. Obviously, further research is needed to fully understand why certain materials show a luminescence colour in daylight. A comprehensive study of such materials compared to their counterparts that do not have such an emission colour component would be necessary, and should include an investigation into the details of luminescence perception by the human eye.
Green daylight-luminescing opal from Zacatecas, Mexico (e.g. Figure 24), has been identified as hyalite by a combination of its opal-A Raman signal and glassy texture in the scanning electron microscope. It occurs in a volcanic environment, as is usually the case for hyalite. Its luminescence is due to dispersed uranyl ions, with a maximum emission at 524 nm, and excitation in both the UV and the visible range. Although the opal may contain up to 0.3% wt.% U[O.sub.2], detailed radiation measurements recorded only very low levels of emitted radioactivity, so there is no danger in wearing or cutting this gem. Several factors contribute to the daylight luminescence of this opal--in particular its diaphaneity or ability to scatter light--but more research is needed to fully elucidate the factors governing the perception of daylight luminescence.
Perhaps 20 kg of medium-to-top-quality rough material remains to be faceted, and prospecting for more opal sites is in progress at the time of this writing. Whether more such bright green fluorescent hyalite will be discovered remains to be seen, but the geology of the deposit is reasonably well understood, and only the immediate areas around the known deposit are prospective as elsewhere the host unit has been eroded away.
The authors thank three anonymous reviewers for their constructive comments. Florian Massuyeau, luminescence engineer at the IMN, Nantes, France, was very helpful with obtaining the spectra and preparing some of the figures.
Berger G., Claparols C., Guy C. and Daux V., 1994. Dissolution rate of a basalt glass in silica-rich solutions: Implications for long-term alteration. Geochimica et Cosmochimica Acta, 58(22), 4875-4886, http://dx.doi.org/10.1016/0016-7037(94)90218-6.
Breit G.N. and Hall S.M., 2011. Deposit model for volcanogenic uranium deposits. United States Geological Survey Open File Report 2011-1255, 5 pp.
Crowningshield R.C., 1985. Gem Trade Lab Notes: Yellow green opal. Gems & Gemology, 21(2), 110-111.
Faulques E., Massuyeau F., Kalashnyk N., Perry D.L., 2015. Application of Raman and photoluminescence spectroscopy for identification of uranium minerals in the environment. Spectroscopy Europe, 27(1), 14-17 and 25, www.spectroscopyeurope.com/images/stories/ArticlePDfs/Perry-27_1.pdf.
Florke O.W., Jones J.B. and Segnit E.R., 1973. The genesis of hyalite. Neues Jahrbuch fur Mineralogie, Monatshefte, 110(2), 82-89.
Fritsch E., 1998. The nature of color in diamonds. In G.E. Harlow Ed., The Nature of Diamonds, American Museum of Natural History & Cambridge University Press, Cambridge, 23-47.
Fritsch E. and Rossman G.R., 1988. An update on colors in ge78ms. Part 3: Colors caused by band gaps and physical phenomena. Gems & Gemology, 24(2), 81-102, http://dx.doi.org/10.5741/gems.24.2.81.
Fritsch E. and Waychunas G., 1993. Chapter 15: Gemstones. In M. Robbins, Ed., Fluorescence: Gems and Minerals under Ultraviolet Light, Geoscience Press, Phoenix, Arizona, USA, 149-174.
Fritsch E., Gaillou E., Rondeau B., Barreau A., Albertini D. and Ostroumov M., 2006. The nanostructure of fire opal. Journal of Non-Crystalline Solids, 352, 3957-3960, http://dx.doi.org/10.1016/).jnoncrysol.2006.08.005.
Fritsch E., Spano-Franco T. and Megaw P., 2014. Gem Notes: Green daylight-fluorescent hyalite opal from Mexico. Journal of Gemmology, 34(4), 294-296.
Gaillou E., 2006. Relations entre Nanostructure, Proprietes Physiques et Mode de Formation des Opales A et CT. PhD thesis, University of Nantes, France, 310 pp.
Gaillou E., Delaunay A., Rondeau B., Bouhnik-Le Coz M., Fritsch E., Cornen G. and Monnier C., 2008a. The geochemistry of opals as evidence of their origin. Ore Geology Reviews, 34, 113-126, http://dx.doi.org/10.1016/j.oregeorev.2007.07.004.
Gaillou E., Fritsch E., Aguilar-Reyes B., Rondeau B., Post J., Barreau A. and Ostroumov M., 2008b. Common gem opal: An investigation of micro- to nano-structure. American Mineralogist, 93, 1865-1873, http://dx.doi.org/10.2138/am.2008.2518.
Gaillou E., Fritsch E. and Massuyeau F., 2011. Luminescence of gem opals: A review of intrinsic and extrinsic emission. Australian Gemmologist, 24(8), 200-201.
Goodell P.C. and Waters A.C., 1981. Uranium in Volcanic and Volcaniclastic Rocks. American Association of Petroleum Geologists, Studies in Geology, 13, 275 pp.
Graetsch H., 1994. Chapter 6: Structural characteristics of opaline and microcrystalline silica minerals. In PJ. Heaney et al., Eds., Silica: Physical Behavior, Geochemistry and Materials Applications, Mineralogical Society of America, Washington DC, USA, 209-232.
Johnson M.L. and Koivula J.I., 1997. Gem News: Botryoidal white opal from Milford, Utah. Gems & Gemology, 33(1), 63-64.
Kane R.E. and Fritsch E., 1991. Gem Trade Lab Notes: Spinel, with unusual green fluorescence. Gems & Gemology, 27(2), 112-113.
Langer K. and Florke O.W., 1974. Near infrared absorption spectra (4000-9000 [cm.sup.-1]) of opal and the role of water in these Si[O.sub.2]xn[H.sub.2]O minerals. Fortschritte der Mineralogie, 52(1), 17-51.
Liu Y., Shigley J.E., Moses T. and Reinitz I., 1998. The alexandrite effect of the Tavernier diamond caused by fluorescence under daylight. Color Research and Application, 23(5), 323-327, http://dx.doi.org/10.1002/(sici)1520-6378(199810)23:5<323::aid-col8>3.0.co;2-y.
Liu Y., Shi G. and Wang S., 2015. Color phenomena of blue amber. Gems & Gemology, 50(2), 134-140, http://dx.doi.org/10.5741/gems.50.2.134.
Markusson S.H. and Stefansson A., 2011. Geothermal surface alteration of basalts, Krysuvik Iceland--Alteration mineralogy, water chemistry and the effects of acid supply on the alteration process. Journal of Volcanology and Geothermal Research, 206, 46-59, http://dx.doi.org/10.1016/].jvolgeores.2011.05.007.
McDowell F.W. and Clabaugh S.E., 1979. Ignimbrites of the Sierra Madre Occidental and their Relation to the Tectonic Evolution of Western Mexico. Geological Society of America Special Publication 180, 113-124, http://dx.doi.org/10.1130/SPE180p113.
Moore T.P., 2014. What's new: Tucson show 2014. Mineralogical Record, 45(3), 345-375.
Moses T., 1997. Gem Trade Lab Notes: Diamond--Two noteworthy stones from the Americas. Gems & Gemology, 33(1), 54-55.
Ostrooumov M., Fritsch E., Lasnier B. and Lefrant S., 1999. Spectres Raman des opales: Aspect diagnostique et aide a la classification. European Journal of Mineralogy, 11, 899-908, http://dx.doi.org/10.1127/ejm/11/5/0899.
Rondeau B., Fritsch E., Guiraud M. and Renac C., 2004. Opals from Slovakia ("Hungarian" opals): A re-assessment of the conditions of formation. European Journal of Mineralogy, 16, 789-799, http://dx.doi.org/10.1127/0935-1221/2004/00160-789.
Rondeau B., Fritsch E., Mazzero F., Gauthier J.-P., Cenki-Tok B., Bekele E. and Gaillou E., 2010. Playof-color opal from Wegel Tena, Wollo Province, Ethiopia. Gems & Gemology, 46(2), 90-105, http://dx.doi.org/10.5741/gems.46.2.90.
Shigley J.E. and Breeding C.M., 2015. Visible absorption spectra of colored diamonds. Gems & Gemology, 51(1), 41-43, http://dx.doi.org/10.5741/gems.51.1.41.
Smallwood A., Thomas P.S. and Ray A.S., 1997. Characterization of sedimentary opals by Fourier transform Raman spectroscopy. Spectrochimica Acta A, 53, 2341-2345, http://dx.doi.org/10.1016/s1386-1425(97)00174-1.
Smith N.A., Cerefice G.S. and Czerwinski K.R., 2013. Fluorescence and absorbance spectroscopy of the uranyl ion in nitric acid for process monitoring applications. Journal of Radioanalysis and Nuclides Chemistry, 295, 1553-1560, http://dx.doi.org/10.1007/s10967-012-1942-4.
Smith R.L. and Bailey R.A., 1968. Resurgent Calderas. Geological Society of America Memoir 116, 613662.
Spano T.L., Burd A.M., Kovacks J.D. and Burns P.C., 2015. Distribution of uranium and uranyl minerals near and within hyalite opal. Poster presentation at 249th ACS National Meeting & Exposition, Denver, Colorado, USA, 22-26 March.
Stephant N., Rondeau B., Gauthier J.-P., Cody J. and Fritsch E., 2014. Investigation of hidden periodic structures on SEM images of opal-like materials using FFT and IFFT. Scanning, 36(5), 487-500, http://dx.doi.org/10.1002/sca.21144.
Titkov S.V., Mineeva R.M., Zudina N.N., Sergeev A.M., Ryabchikov I.D., Shiryaev A. A., Speransky A.V. and Zhikhareva V.P., 2015. The luminescent nature of orange coloration in natural diamond: Optical and EPR study. Physics and Chemistry of Minerals, 42, 131-141, http://dx.doi.org/10.1007/s00269-014-0705-x.
Dr Emmanuel Fritsch
Institut des Materiaux Jean Rouxel (IMN) CNRS, Team 6502 and University of Nantes, BP 32229, F-44322 Nantes Cedex 3, France. Email: firstname.lastname@example.org
Dr Peter K. M. Megaw
IMDEX Inc., P.O. Box 65538, Tucson, Arizona, 85728 USA
Tyler L. Spano
Department of Civil and Environmental Engineering and Earth Sciences, University of Notre Dame, Indiana, 46556, USA
Boris Chauvire and Dr Benjamin Rondeau
LPGN CNRS, Team 6112 and University of Nantes, BP 92208, F-44322 Nantes Cedex 3, France
Coast-to-Coast Rare Stones, P.O. Box 647, Mendocino, California, 95460 USA
Dr Thomas Hainschwang
GGTL Laboratories, Gnetsch 42, LI-9496
Gemological Institute of America, 5345
Armada Drive, Carlsbad, California, 92008
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Feature Article|
|Author:||Fritsch, Emmanuel; Megaw, Peter K.M.; Spano, Tyler L.; Chauvire, Boris; Rondeau, Benjamin; Gray, Mic|
|Publication:||The Journal of Gemmology|
|Date:||Jun 1, 2015|
|Previous Article:||Lead-glass-filled yellow sapphires.|
|Next Article:||Raman spectroscopy of ancient beads from Devin Castle near Bratislava and of four intaglios from other archaeological finds in Slovakia.|