by Benjamin Rondeau, Emmanuel Fritsch, Francesco Mazzero, Jean-Pierre Gauthier, Bénédicte Cenki-Tok, Eyassu Bekele, and Eloïse Gaillou
from GEMS & GEMOLOGY, Vol. 46, No. 2, pp. 90–105.
© 2010 Gemological Institute of America, Summer 2010
Published on JTV.com: November 2011
See the end of this downloadable PDF for About the Authors, Acknowledgments, and a complete list of References.
A new opal deposit was discovered in 2008 near the village of Wegel Tena, in volcanic rocks of Ethiopia’s Wollo Province. Unlike previous Ethiopian opals, the new material is mostly white, with some brown opal, fire opal, and colorless “crystal” opal. Some of it resembles Australian and Brazilian sedimentary opals, with play-of-color that is often very vivid. However, its properties are consistent with those of opal-CT and most volcanic opals. Inclusions consist of pyrite, bariummanganese oxides, and native carbon. Some samples show “digit patterns”: interpenetrating playof-color and common opal, resembling fingers. The opaque-to-translucent Wegel Tena opals become transparent when soaked in water, showing a remarkable hydrophane character. White opals from this deposit contain an elevated Ba content, which has not been reported so far in opal-CT. The fire and crystal opals are prone to breakage, while the white, opaque-to-translucent opals are remarkably durable. The proportion of gem-quality material in the Wegel Tena deposit seems unusually high, and 1,500 kg have already been extracted using rudimentary mining techniques. The deposit may extend over several kilometers and could become a major source of gem-quality opal.
In early 2008, a new source of play-of-color opal was discovered by farmers near Wegel Tena in northern Ethiopia (Fritsch and Rondeau, 2009; Mazzero et al., 2009; Rondeau et al., 2009). Since January 2009, the deposit has been worked by about 200 local miners. The opals are mostly white, which is uncommon for play-of-color volcanic opal, and may resemble material from Australia or Brazil (figure 1). Some fairly large pieces have been polished (figures 2 and 3). The Wegel Tena opals differ from those found at Mezezo, in Ethiopia’s Shewa Province, or in neighboring Somalia, which are mostly orange to red to brown (e.g., Koivula et al., 1994; Gauthier et al., 2004, and references therein). Two of the authors (FM and EB) traveled to the locality on several occasions. They gathered representative gem material, as well as surrounding rocks, and discussed the opal and its extraction with the miners to develop a better understanding of this promising new deposit.
The opal mining area lies in Wollo Province (also spelled Wolo or Welo), ~550 km north of Addis Ababa and ~200 km north of the Mezezo opal deposit (figure 4). The locality has also been referred to as “Delanta,” which corresponds to a former subdivision (or “awraja”) of Wollo Province. The region containing the deposit is called Tsehay Mewcha, a large area that encompasses scattered farms and a small village, about 17 km northeast of the village of Wegel Tena (figure 4). Tsehay Mewcha is situated on a plateau at an altitude of about 3,200 m. The opal occurs in a horizontal layer that is exposed on a cliff above a canyon tributary of the Blue Nile River. This layer is ~350 m below the top of the plateau (figure 5). Tsehay Mewcha is accessible with a fourwheel-drive vehicle, and the various mine workings are reached by walking down the steep canyon for 30 minutes to more than one hour. Hazardous conditions are created by cliffs in the digging areas, as well as by falling rocks due to mining activities (see www.opalinda.com for more information on access conditions).
The entire region around Wegel Tena consists of a thick (>3,000 m) volcano-sedimentary sequence of alternating layers of basalt and rhyolitic ignimbrite. The layers of basalt or ignimbrite are a few meters to hundreds of meters thick. (Ignimbrite is a volcanic rock of andesitic-to-rhyolitic composition that forms sedimentary-like layers after the volcanic plume collapses and falls to the ground. The particles that form this rock are a heterogeneous mixture of volcanic glass, crystals, ash, and xenoliths.) This volcanic sequence was emplaced with the opening of the East African continental rift during the Oligocene epoch (Cenozoic age), about 30 million years ago (Ayalew and Yirgu, 2003; Ayalew and Gibson, 2009).
Over the entire volcanic series, only one very thin seam (<1 m thick), hosted by ignimbrite, is mineralized with opal (again, see figure 5). Common opal and play-of-color opal most often cement grains of volcanic debris (figures 6 and 7) or sometimes fill in fractures or cavities in the rock. As a result, the rough gem material often has an irregular shape. Microscopic examination of our samples revealed that, for the most part, the host rock consists of mixed altered material, including clays, common opal, and some minor iron oxy-hydroxides. Some large crystals of alkali feldspar were unaltered, while others were transformed into clays. By comparison, the ignimbrite sampled only a few meters above the opal-bearing layer was unaltered and contained abundant quartz crystals.
The opal is extracted by artisanal miners using homemade tools (figures 8 and 9), as well as picks, hammers, and shovels provided by the Ethiopian government. The mineralized layer extends for hundreds of meters along the flank of the canyon (again, see figures 5 and 8), but the excavations only penetrate 1–2 m into the mountain. Because the diggings are not supported by timbers or other means, mining is very dangerous in some extensively worked places. Tragically, at least 20 miners have died from collapsing rock. The miners are organized in cooperatives that control the distribution of the rough opal sold to gem dealers and eventually cutters in Addis Ababa. Opal production has been significant, with over 1,500 kg of rough extracted to date.
We examined hundreds of rough and polished samples to determine the typical characteristics for this locality. We selected 33 samples to document the gemological properties and note interesting inclusions and growth features, hydrophane character, matrix, and the like (table 1). Eight of these were fashioned as cabochons of various shapes and colors (4.43–18.85 ct). The other 25 samples consisted of rough pieces ranging from ~4 to 965 g (e.g., figure 7), and most contained significant portions of play-ofcolor opal. Some had considerable matrix material. Several of these were sliced or crushed for specific testing, as detailed below.
Selected samples were examined by standard gemological methods to determine their refractive index, hydrostatic specific gravity, and microscopic features (see table 1). Play-of-color was observed with the stone against a dark background (when necessary), and spot illumination placed above, behind (when appropriate), and then perpendicular to the viewing direction (the latter to observe any contra luz effect, i.e., play-of-color revealed by transmitted light). Luminescence was observed with longand short-wave UV radiation using 6-watt lamps.
We tested the hydrophane character of 14 samples by observing the specific gravity, transparency, and play-of-color before and after immersing the samples in water for a few minutes to one hour. For the same 14 stones, the following optical tests were conducted: Polarization behavior was studied using a GIA Instruments polariscope, absorption spectra were observed with an OPL prism-type spectroscope, and the Chelsea filter reaction was determined with illumination from a strong halogen lamp.
Fourier-transform Raman spectra were recorded on all samples using a Bruker RFS 100 spectrometer. Each spectrum included 1,000 scans to increase the signal-to-noise ratio, as opal is a poor Raman scatterer.
The microstructure of five samples (1071, 1073,1074, 1076, and 1101) was investigated with a JEOL7600 scanning electron microscope (SEM) equippedwith a hot cathode/field-effect electron gun, and aHitachi H9000-NAR transmission electron microscope(TEM) operated at 10 kV. TEM samples wereobtained by crushing a small piece of each samplewith a mortar and pestle. The chemical compositionof eight representative samples (1071, 1073,1074, 1076, 1078, 1100, 1111, and 1121), their inclusions,and accompanying minerals (in samples 1100,1106, 1113, and 1122) was measured by energy-dispersivespectroscopy (EDS) with a PrincetonGamma Tech IMIX-PTS detector installed on aJEOL 5800LV SEM. We prepared two thin sections(30 μm thick) of the opal’s host rock for petrographicmicroscopy and analysis of the minerals by EDS.
We conducted preliminary analyses of trace-elementcomposition by laser ablation–inductivelycoupled plasma–mass spectrometry (LA-ICP-MS) atthe Institute of Geological Sciences in Bern. Weinvestigated two rough samples representative ofthe marketable opal: one white translucent and theother zoned white translucent and orange transparent,both with play-of-color. (These are not listed intable 1 because they were examined apart from theother samples and only with the LA-ICP-MS technique.)The LA-ICP-MS system consisted of apulsed 193 nm ArF excimer laser with an energyhomogenizedbeam profile coupled with an ElanDRCe quadrupole mass spectrometer. Laser parameterswere set to 16 J/cm2 energy density on thesample, with a pulse duration of 15 ns and a repetitionrate of 10 Hz. Pit sizes were 60 and 90 μm. Thelaser-ablation aerosol was carried to the ICP-MS bya mixed He-H2-Ar carrier gas. Details on the LAICP-MS measurement conditions are available in the G&G Data Depository (gia.edu/gandg).
Stability to crazing was assessed after cutting andpreforming by one of us (FM) who has handlednumerous Wegel Tena opals since their discovery. Afew representative samples from various Wegel Tenaparcels were kept at ambient temperature andhumidity, and visually observed over time. As part ofa proprietary fashioning process, about 50 high-qualityplay-of-color Wegel Tena opals were cycledbetween water and air (at room temperature) for one hour and then longer periods (up to several days).We tested for toughness by dropping five fashionedopals on a concrete surface from a height of~1.5 m, to simulate dropping a stone by accident.
The standard gemological properties are summarizedin tables 1 and 2 and presented below.
From observing hundreds of rough and faceted samples,we determined that most opals from Wegel Tenahave a white bodycolor, while some are pale yellowand a few are darker orange (fire opal) to brownishred (again, see figures 2 and 3). Rare samples have adark “chocolate” brown bodycolor (again, see figure7). Some zoned samples show several layers of contrastingbodycolor and play-of-color (figure 10).
The opals range from opaque to transparent, butmost are translucent. Because the material is turbid,it scatters light efficiently, creating the white bodycolortypical of this deposit. Some of the highestqualityopals are translucent and display a blue scatteringbodycolor (figure 11).
Among the 33 samples tested, we observed thatall opaque-to-translucent samples became moretransparent when immersed in water for a few minutesto one hour, depending on the thickness of thesample. This behavior is typical of hydrophane.There were several degrees of change, the most dramaticbeing a transformation from opaque white totransparent colorless (figure 12). During this process,play-of-color appeared to strengthen. This phenomenonwas fully reversible in one to a few hours,depending on ambient humidity and the thicknessof the gem.
The opals typically displayed a mosaic of purespectral color patches against a translucent whitebodycolor. Generally all spectral colors (from red toviolet) were observed in the play-of-color samples,often with large patches of red and orange. Theintensity of the play-of-color varied on the millimeterscale, from intense to none, even within thesame sample. We did not notice any contra luzmaterial (which is typically transparent).
Play-of-color was commonly distributed alongparallel columns that resembled fingers. We refer tosuch features as digit patterns (figure 13). The playof-color digits were embedded within common opalof slightly different color or transparency. Theircross-section was rounded, or sometimes quitepolygonal when there was little interstitial common opal present. The digits’ tips were often rounded.Some samples showed planar zoning of commonand play-of-color opal that we interpreted as horizontal.When digit patterns were observed in such samples,they were always perpendicular to these planes(figures 10 and 14). Digit patterns were seen in nineof the 33 samples (nos. 1072, 1104, 1105, 1109, 1111,1113, 1114, 1121, and 1122).
Some much rarer play-of-color features alsohave been observed in Wegel Tena opals. Thecabochon in figure 15 (sample 1075) showeddiffraction concentrated in points (not patches)that moved together in a synchronized fashion andchanged color when the stone was tilted, or theintense pinpoint light source moved around. Thisrevealed a perfect organization of the silica spheresthat was distributed throughout the entire cabochon(pseudo single crystal; Fritsch and Rondeau,2009). This “perfect diffraction” of light is seenonly very rarely in natural gem opals. For a videoand more comments on this phenomenon, see www.gemnantes.fr/research/opal/index.php#reciproque.
Another rare optical phenomenon seen in oneopal that we were only able to keep for a short timewas the presence of small, curved rainbows ofdiffraction (figure 16). Usually in play-of-coloropals, each patch of diffraction is homogeneous incolor. Here, the spectral colors within each patchwere diffracted along a small area, ranging from 1 to5 mm. For a video of this phenomenon, seewww.gemnantes.fr/research/opal/index.php#rainbow.
SG values (before being soaked inwater) ranged from 1.74 to 1.89. After immersion inwater for less than one hour, some samples weighedas much as 10.2% more, resulting in higher SG valuesof 1.90 to 2.00 (table 2). This effect, related to thesamples’ porosity, was fully reversible and repeatable.
Because of the opals’ porosity, wemeasured the refractive index of only five samples.In some instances, reaction with the contact liquid caused white opaque spots to form on the opal,adversely affecting the gem’s transparency and evenits UV luminescence. RI values measured on sample1073 were 1.42 and 1.44. However, we measuredRIs as low as 1.36 to 1.38 on other samples (1071,1076, and 1077). Some indices were easy to read,but most were difficult or simply impossible.Surprisingly, sample 1118a, a plate, showed an RI of1.37 on one face and two indices on the other: 1.42and a very sharp reading at 1.52.
UV luminescence was quite variable,ranging from bluish white to greenish white,yellow, and green. Luminescence intensity rangedfrom inert to strong. All brown samples or zoneswere inert. In long-wave UV, most non-brown sampleshad weak to very weak luminescence that wasfairly turbid, and bluish to greenish white, withweak to very weak greenish white phosphorescencethat lasted a few seconds. The typical short-waveUV reaction was slightly weaker, though the phosphorescencesometimes lasted longer. Some samples—particularly white opals with very good playof-color—luminesced more strongly, with a moderate,mostly white fluorescence. One transparent,near-colorless, play-of-color opal had a strong, puregreen luminescence.
Seven transparent light colorstones, with and without play-of-color, showed noreaction viewed between crossed polarizing filters.Five translucent stones with play-of-color showedcyclic variations in diffraction colors as the stoneswere rotated a full 360°. Two opals (no. 1076 and1109) showed anomalous double refraction (ADR).
No spectrum wasseen in the two lighter-color, transparent stones testedwith the hand spectroscope. The remaining 12milky or orange-to-brown stones showed absorptionin the violet and blue regions, sometimes extendinginto the green. This absorption increased with theamount of light scattered by the stone (from slightlymilky to white), the darkness of the bodycolor (fromyellow to brown), and the thickness of the sample.
Three transparent, light-colored gemsshowed no reaction when viewed through theChelsea color filter. Eleven transparent brown orstrongly diffusing (milky to white) opals appearedorange to bright red.
One sample (no. 1100) contained elongated,cylindrical inclusions measuring approximately800 μm × 1 cm (figure 17). Their surface wasvery irregular. EDS showed that these “tubes” werefilled with silica, which may correspond to chalcedony.The outside surface of the inclusions alsoconsisted of silica. They appeared more difficult to polish than the host opal.
Dispersed micro-inclusions of black, opaquecrystals were abundant in some samples (in particular,no. 1113). EDS analyses revealed iron and sulfur,which suggests they were pyrite (pyrite usuallyappears black in such small dimensions).
Some of the rough opal samples were outlinedby a thin layer (less than 0.1 mm thick) of black,opaque minerals. These were identified by EDS asbarium-manganese oxides (probably hollandite) andnative carbon (probably graphitic carbon). Also pres -ent in such layers were micrometer-sized crystalsthat were identified by EDS as titanium oxides(probably rutile). In rare instances, the black mineralswere included in the body of the opal, filling fissuresor forming dendrites.
We measured the chemicalcomposition of several samples by EDS (both majorand minor elements; see table 3) and LA-ICP-MS(trace elements; see table 4 and the G&G DataDepository). In addition to silica, we detected a significantproportion of Al (0.6–1.9 atomic %) andminor amounts of Ca (0.05–0.6 at.%), Na (up to 0.4at.%), K (0.2–0.5 at.%), and Fe (up to 0.3 at.%). Ironwas not detected in the white samples. These compositionsare typical for opal (Gaillou et al., 2008a).Among the trace elements, white opal containedabundant Ba (140–226 ppm [by weight]), Sr (127–162ppm), and Rb (44–73 ppm). The orange fire opal portionof one sample showed Ba, Sr, and Rb contentsconsistent with those of opal-CT (Gaillou et al.,2008a). For the concentrations of other elements, seetable 4 and the G&G Data Depository.
We obtained similar spectral featuresfor all samples (e.g., figure 18). The apparentmaximum of the strongest Raman band rangedfrom 360 to 335 cm−1. Other, sharper Raman bandswere present at ~3230–3200, 2940, 1660, 1470,1084, 974, and 785 cm−1.
Observing the microstructure of anopal helps us characterize it and understand itsgrowth conditions. Most often, as shown in Gaillouet al. (2008b), two main categories of structures canbe observed: “smooth sphere” structure in opal-A(A for amorphous) or “lepisphere” structure in opal-CT (CT for cristobalite and tridymite; opal-A andopal-CT were originally defined on the basis of theirX-ray diffraction patterns, and later on their Ramanscattering patterns—see Jones and Segnit, 1971;Smallwood et al., 1997). To reveal the internalstructure of an opal, one must first etch the samplein hydrofluoric acid (HF) and then observe the surface by SEM. Gaillou et al. (2008b) showed thatetching in a 10% solution of HF for about 10 secondscan reveal the structure of opal. We encounteredan unexpected reaction, however: Our sampleswere strongly affected by the acid, tending toflake away and develop networks of cracks. Wemodified both the concentration of HF (from 0.01%to 10%) and the duration of acid exposure (from 1second to 3 minutes, with longer times using weakeracid), but we did not observe any organization ofsmooth spheres or lepispheres. Thus, we could notsee any packing of spheres in opals from WegelTena using this technique.
We subsequently studied the structure of theopal using TEM, which revealed a regular networkof spheres ~170 nm in diameter (figure 19). In anotherattempt to explore the structure, we studied thesame sample using SEM with an unusually highvoltage (15 kV) for a sample that was not coated tomake it electrically conductive. Because the opalspecimen was so thin, the electrons were able topass through it to the backscattered electron detector.The similar image that was generated by thissignificantly different technique confirmed theTEM observations. Yet neither method helped usdetermine if the opal’s structure is characterized bylepispheres (as are typical of opal-CT) or smoothspheres (typical of opal-A). Regardless, the regularnetwork of these spheres is responsible for thediffraction of visible light that results in the play-ofcolorshown by the opals.
Opals from certain localities(e.g., Querétaro, Jalisco, and Nayarit States inMexico, and Shewa Province in Ethiopia, both opal-A and -CT) are known to destabilize, or “craze”with time. Cracks develop at the surface, and/or awhite, opaque egg-like structure develops in thecenter of the stone (mostly in opal-CT; Aguilar-Reyes et al., 2005).
In the authors’ experience, ~5% of Wegel Tenaopals develop cracks after initial sawing and preforming.Until now, out of approximately 3,000play-of-color opal cabochons from Wegel Tenareleased into the market during 2008–2010 by oneof us (FM), only three samples were returned after cracking. A parcel of seven opals (including sample1072) thought to be from Mezezo was set aside in2005 because of their unusual appearance; they arenow known to be from Wegel Tena, and all thestones are still intact. Any crazing appears to berestricted to transparent material, in particular paleyellow to orange samples (fire opal) and near-colorless“crystal” opals. A few samples showed spectacular“egg” development (figure 20), as seen in someMexican fire opal. In general, opaque white-to-yellow-to-brown opals from Wegel Tena appear verystable. There is no noticeable difference in crazingbehavior between common and play-of-color opal.
There was no change in appearance (color,diaphaneity, crazing, or play-of-color) in the samplesthat were submitted to alternating periods ofimmersion in water. One customer who wears heropal constantly complained that it became moretransparent when she took a shower, swam, or otherwiseput her hands in water. She recognized, however,that the opal always returned to its originalappearance after some time (depending on the durationof immersion)—which is due to its hydrophanecharacter.
We noticed by accident that Wegel Tena opalscould sustain a fall from 1.5 m onto a concrete floorwith no visible damage, even under the microscope.Repetition of this test on five oval cabochons didnot produce any sign of damage. The same experimentwith five oval cabochons from the Mezezodeposit and three oval cabochons of white opal fromAustralia (including one boulder opal) led to breakageof all samples.
SG was in the reportedrange for opal (Webster, 1975). However, some samplesshowed large weight gains during immersion inwater, up to 10.2%. This is probably related to thehigh porosity of these samples, detectable simply bytouching a sample with the tongue to test its “stickiness.”RI ranged from as low as 1.36 to 1.43, withone “secondary” reading of 1.52. Values as low as1.36 have been previously measured in hydrophanefrom Slovakia (Reusch, 1865) and in opals fromMexico (Spencer et al., 1992). The RI sometimes variedstrongly even within a single sample, dependingon the orientation. Similar effects were seen inShewa opals (Johnson et al., 1996). They are probablydue to local physical or chemical heterogeneities, ascommonly observed in opals.
The large patches of red and orange seen insome of the play-of-color Wegel Tena opals are notcommon in Brazilian and Australian opals. Wefound the digit patterns to be very common inWegel Tena opals, though in some cases they wereonly visible with a microscope. Digit patterns weredescribed previously only in opals from Mezezo(see figure 14 in Johnson et al., 1996; Gauthier etal., 2004). We know of only one non-Ethiopian opalwith digit patterns; it was seen in Australia by oneof the authors (EF). Also, Choud hary (2008)described another such opal from an unspecifiedsource. In contrast, the planar zoning of commonand play-of-color opal is often observed in opalsfrom other deposits, volcanic or otherwise.
As expected, the spectroscope spectrum and color filter reaction were not useful for identification.The red color filter reaction, as well as theabsorption of the violet and blue regions in moststones, stems from two possible factors that maycombine in a given stone. Intense light scattering(the corresponding stones were milky or white)attenuates violet and blue wavelengths, as they arescattered preferentially at an angle. Also, the yellow-to-brown bodycolor is due to a continuum ofabsorption, increasing from the red toward the violetregion. Hence it will also contribute to blockingthe violet-to-blue (and even some green) wavelengths.The resulting color in rectilinearly transmittedlight in both cases is orange or red, as seenwith the spectroscope or the Chelsea filter. Notethat this is a good example of a bright red Chelseafilter reaction that has nothing to do with the presenceof chromium in the stone.
The UV luminescence properties were typicalfor opal. The fluorescence is a mix of intrinsic silicasurface-related violetish blue emissions andextrinsic uranium-related green emission. The latteris often more visible with short-wave UV(Fritsch et al., 2001; Gaillou, 2006; Gaillou et al.,2008a).
The observed hydrophane behavior is alreadyknown for some rare opals (Webster, 1975), particularlyfrom Opal Butte, Oregon (Smith, 1988). In thesamples studied here, there was no relationshipbetween the capacity of opal to absorb water andthe change in its visual appearance (compare figure11 and table 2).
The main Raman peaks at~1084, 974, 785, and 357 cm−1 are typical for opal-CT and distinctly differ from those of opal-A, forwhich the Raman spectrum displays a main bandaround 420 cm−1 (Smallwood et al., 1997; Fritschet al., 1999; Ostrooumov et al., 1999). The bands at~3230 and 1660 cm−1 are attributed to the presenceof water (Ostrooumov et al., 1999; Zarubin, 2001),and the band at 2940 cm−1 to cristobalitic water(Gaillou et al., 2004; Aguilar-Reyes et al., 2005).
The occurrence of digit patterns remainssomething of a mystery, but they seem to have beenvertical at the time of their formation. To date, digitpatterns have been observed almost exclusively inmaterial from the two Ethiopian deposits, WegelTena and Mezezo. It follows, then, that the geologicconditions under which opals formed at these areasshare common characteristics. However, the dispositionof opal in its host rock differs between thetwo localities. At Wegel Tena, opal most oftencements volcanic grains, while at Mezezo it usuallyforms nodules filling the cavities in volcanic rock.The latter is the case for most volcanic opal localities,such as Mexico and Oregon. Also, the inclusionscenes observed in opals from Wegel Tena aredifferent from those observed in opal from any otherdeposit (e.g., Gübelin and Koivula, 1986).
At Wegel Tena, the opal-mineralized layer isconcordant with volcano-sedimentary deposits thatextend for hundreds of kilometers. No systematicprospecting for opal has been conducted, but a localminer reported to one of us (EB) that opal sampleshave been found in the same geologic unit and at asimilar depth on the flank of the Great Rift Valley,40 km from the present workings (YapatsuPurpikole, pers. comm., 2009). We believe, then,that the extent of the opal-bearing layer is probablymuch greater than what is known today.
From our preliminary measurementsof two samples, trace-element compositionwas comparable to opals from Mezezo (Gaillouet al., 2008a), with the following notable exceptions:Y (0.07–0.3 ppm), Nb (0.3–2.8 ppm), and Th (<0.2ppm) were lower in the Wegel Tena opals, whereasSc (1.5–2.3 ppm), Rb (39–73 ppm), Sr (72–162 ppm),and Ba (82–226 ppm) were higher. The enriched Baconcentration in opals from Wegel Tena is surprising.Looking at opals from elsewhere (Gaillou et al.,2008a), we see that those that formed in a volcanicenvironment always have low Ba (<100 ppm), whereasthose from a sedimentary environment contain100–300 ppm Ba. However, we note that the relativelyhigh Ba contents in some Wegel Tena opalsare consistent with the geologic environment that also resulted in the presence of Ba-Mn oxides in fissuresmentioned above.
Areas of different composition within singlesamples raise the question of their petrogenesis.The orange opal is richer in Fe and presents a typicalopal-CT trace-element composition (comparewith Gaillou et al., 2008a), whereas the white opalhas unusually high contents of Ba, Sr, and Br comparedto volcanic opals.
Digit patterns are typical of Ethiopianopal, from either Mezezo or Wegel Tena, regardlessof the bodycolor. The digit patterns somewhatresemble the columnar structure observed in syntheticopals (such as Gilson), but they also have significantdifferences. Digit patterns are made of twomaterials: one transparent with play-of-color and theother more turbid, often without play-of-color. Theyform rounded columns, whereas the columns in syntheticsare far more regular and polygonal. Hence,careful observation is sufficient to distinguish thenatural from synthetic opal.
To our knowledge, white opals from Wegel Tenarepresent the only example of opal-CT with a Bacontent greater than 100 ppm. Therefore, the combinationof a satisfactory Raman spectrum withtrace-element analysis makes it possible to identifywhite opals from this locality. The fire opals displayeda chemical composition comparable to thatof opal-CT from other localities.
Translucent opals from Wegel Tena resisthydration/dehydration cycles much better thanopals from Mezezo, which crack and break moreeasily during such tests. A stabilization process hasbeen developed to prevent crazing of Ethiopian opal(Filin and Puzynin, 2009), but in our experience thisappears unnecessary for translucent opals fromWegel Tena.
Wegel Tena, Ethiopia, is a source of significantquantities of high-quality play-of-color opal (e.g., figure21), with a mostly white bodycolor that is moremarket-friendly than the mostly brown materialfrom Mezezo, ~200 km to the south. The opals arefound in a specific layer of a thick volcanic sequenceof alternating basalt and ignimbrite of Oligoceneage. Systematic prospecting is needed to assess theextent of the opal-containing layer and the productionpotential, which appears quite significant.
Many of these new opals showed what wedescribe as “digit patterns,” a common feature inEthiopian opals. The Wegel Tena samples typicallyshowed the Raman spectrum of opal-CT, as domost opals that formed in a volcanic environment.Among the inclusions, we noted cylinders of silica(probably chalcedony), black microcrystals (probablypyrite), and coatings/fissures filled with bariummanganeseoxides and graphite-like carbon. Thecombination of an enriched Ba content in the whiteopals (>100 ppm), a Raman spectrum typical of opal-CT, and the presence of digit patterns characterizesopals from Wegel Tena. Further trace-element analysesare necessary to strengthen the chemical criterionfor Ba. The opaque-to-translucent opals fromWegel Tena are much more stable than those fromMezezo, which have a tendency to craze. Also,physically they are surprisingly tough. The WegelTena area has the potential to become a leadingsupplier of high-quality white play-of-color opal.
Published Feb 2011
Published August 2011
Published August 2011
Published July 2011
Modified April 2011
Published December 2011