222 LAB NOTES GEMS & GEMOLOGY FALL 2011

Unusually Large CLINOHUMITEClinohumite, (Mg,Fe2+)9(SiO4)4(F,OH)2,is a fairly rare collectors’ stone withorangy yellow to brownish orange col-oration. Faceted stones typicallyweigh <2 ct, but a few much largerstones have been cut (e.g., 36.56 ct,see L. Massi, “AIGS Bangkok inspectsexceptional clinohumite from Tajiki -stan,” InColor, Summer 2007, pp.30–31). In the present report, we doc-ument the properties of an exception-ally large clinohumite weighing 84.23ct (32.6 × 24.79 × 17.77 mm). Forbackground information on the cut-ting of this stone, see Y. Zhukov,“Clino humite—The mountain fire,”

InColor, Spring 2011, pp. 48–51.The stone was faceted into a pear

brilliant and was brownish orange (fig-ure 1). Internal features consisted ofnumerous two-phase inclusions (fig-

ure 2, left) and “fingerprints,” as wellas distinct angular brownish orangegrowth lines (figure 2, center), as docu-mented previously in clinohumite(e.g., U. Henn et al., “Gem-quality cli-nohumite from Tajikistan and theTaymyr region, northern Siberia,”Journal of Gemmology, Vol. 27, No. 6,2001, pp. 335–339; and Winter 2004GNI, pp. 337–338). Abundant twinplanes were also seen with cross-polarized light (figure 2, right). Thestone was inert to long-wave UV radi-ation, whereas short-wave UV pro-duced moderate orange to strongchalky yellow fluorescence along thegrowth lines. Its Raman spectrum,along with gemological propertiessuch as RI (1.635–1.670) and SG (3.21)values, confirmed that it was clinohu-mite. IR spectroscopy recorded typicalabsorptions for clinohumite, includinga broad band centered at 3560 cm−1 dueto OH stretching and additional bandsat 4100 and 4510 cm−1 caused byMgOH units and Si-OH bonding,

© 2011 Gemological Institute of America

GEMS & GEMOLOGY, Vol. 47, No. 3, pp. 222–233,http://dx.doi.org/10.5741/GEMS.47.3.222.

Editors’ note: All items were written by staffmembers of GIA laboratories.

EDITORS

Thomas M. Moses | Shane F. McClure

Figure 2. Microscopic observation of the clinohumite revealed two-phase (fluid-gas) inclusions,(left, magnified 35×), distinct angular brownish orange growth lines (center, 12×), and numeroustwin planes (right, cross-polarized light, 75×).

Figure 1. This faceted clinohumiteweighs an impressive 84.23 ct.

LAB NOTES GEMS & GEMOLOGY FALL 2011 223

respectively (R. L. Frost et al., “Near-infrared and mid-IR spectroscopy ofselected humite minerals,” Vibra -tional Spectroscopy, Vol. 44, 2007, pp.154–161, http://dx.doi.org/10.1016/j.vibspec.2006.11.002). The FTIR andRaman spectra of this stone are avail-able in the G&G Data Depository(gia.edu/gandg). Energy-dispersive X-ray fluorescence (EDXRF) spectros -copy detected major amounts of Mg,Fe, and Si, and traces of Ti and Mn, asexpected for clinohumite.

There are two main localities thatproduce gem-quality clinohumite—the Pamir Mountains of Tajikistanand the Taymyr region of northernSiberia (Massi, 2007). Unfortunately,the geographic origin of the presentclinohumite is unknown. It is thelargest gem-quality clinohumite evertested at GIA.

Kyaw Soe Moe and Wai Win

DIAMOND

Black Diamond, Colored by StrongPlastic Deformation Black color in diamonds can have var-ious natural or artificial causes, frominclusions (abundant graphite or pin-points, or dense clouds) to heating orextremely strong irradiation. TheNew York laboratory recently exam-ined a black diamond colored byanother mechanism.

This round-cut diamond weighed

0.85 ct (5.70 × 5.78 × 3.87 mm) and wascolor graded as Fancy black (figure 3).Viewed with magnification and verystrong fiber-optic illumination, itrevealed fractures and mineral inclu-sions, as well as a dark brown colorwith banded linear distribution (figure4). Strong plastic deformation and therelated linear banding were also clearlyrevealed in the Diamond View. Its mid-infrared spectrum showed strong N-related absorption in the one-phononregion and a weak H-related feature at3107 cm−1. Also recorded were verystrong absorptions from amber centersat ~4170 and 4070 cm−1, with intensi-ties of 2.0 and 1.1 cm−1, respectively.

The vacancy cluster, a diamondlattice defect closely related to plasticdeformation, usually absorbs visiblelight from ~600 nm to lower wave-lengths. When its concentration isvery high, such as in this diamond, itsabsorption extends across the entirevisible light region and blocks the vastmajority of light. It is very rare to see anatural gem diamond with suchstrong plastic deformation that itcauses a black appearance. This stonedemonstrates yet another cause of nat-ural black color in diamond.

Wuyi Wang

Coated Black DiamondGems & Gemology has reported on anumber of interesting black diamondsin recent years (see the Spring 2007,Winter 2007, Fall 2008, and Summer

2010 Lab Notes sections and S. V. Tit -kov et al., “An investigation into thecause of color in natural black dia-monds from Siberia,” Fall 2003, pp.200–209). This color, when natural, istypically caused by dark/intense min-eral or cloud inclusions. It can also beinduced by artificial irradiation orgraphitization along fractures due toheating (naturally or in the laboratory).

Recently, the New York laboratoryreceived a 1.29 ct black pear shape (fig-ure 5) for color origin determination.Infrared absorption spectroscopy con-firmed it was a type Ia diamond withhydrogen impurities. As expected, itwas inert to both long- and short-waveUV radiation. However, microscopicexamination revealed a distinctlylighter color along the facet junctions(figure 6), and these areas showed con-trasting luminescence in the

Figure 3. The black color of this0.85 ct diamond is attributed tovery strong plastic deformation.

Figure 5. This 1.29 ct diamondproved to be coated to enhanceits black color appearance.

Figure 4. With magnification andvery strong lighting, the diamondin figure 3 showed a dark browncolor with banded linear distri-bution. Magnified 70×.

Figure 6. In reflected light, the1.29 ct diamond displays a dis-tinctly lighter color along thefacet junctions. Magnified 30×.

224 LAB NOTES GEMS & GEMOLOGY FALL 2011

Diamond View (figure 7). The girdlealso showed uneven concentrations ofdark color, and there were scratcheson the pavilion that were lighter thanthe black bodycolor; both of thesecould be easily observed with themicroscope. Feathers within the dia-mond were dark colored due to blacknatural inclusions. It was obvious thatthis diamond, though rather dark, hadbeen coated to further enhance itscolor.

This was our first encounter witha coated black diamond. Because ofthe coating, we did not issue a colorgrade. In this case, the black coatingmade the stone appear a deeper, moreeven black. Although coating a blackdiamond can improve its overallappearance or color distribution, inthis case the treatment was detectablewith a microscope.

Erica Emerson

Coated Diamond withSpectroscopic Features of aNatural-Color Pink

Coated diamonds are fairly commonlyencountered, both in the trade and thegrading lab. Most coatings consist of athin film applied to one or more pavil-ion facets, often at or near the girdleedge, though sometimes the coatingwill completely cover the diamond’ssurface. Because of the value of natu-ral-color pink diamonds, pink is acommon choice for such coatings,which are intended to deepen a weakbodycolor or to mask an undesired one(e.g., Winter 2010 Lab Notes, pp.299–300). This treatment can be iden-tified through microscopic observa-tion and UV-Vis spectroscopic fea-tures. In the case of pink-coated dia-monds, the typical absorption band at~520 nm (see A. H. Shen et al.,

“Serenity coated colored diamonds:Detection and durability,” Spring2007 G&G, pp. 16–34, http://dx.doi.org/10.5741/GEMS.43.1.16) is easilydistinguished from the ~550 nm bandin natural-color stones.

The 1.05 ct heart-shaped diamondin figure 8, recently submitted forcolor grading, initially received a gradeof Fancy Light brown-pink. Its UV-Visspectrum showed the broad ~550 nmabsorption band (figure 9) typical ofnaturally colored pink diamonds.With magnification, however, a coat-ing was clearly visible on the pavilionfacets (figure 10). After the coating wasremoved by boiling the stone in acid,the color grade was revised to J (figure8, right).

Because the normally diagnosticabsorption band for a pink coatingwas shifted from 520 to 550 nm,spectroscopy alone would not havedetected this coating. This reinforcesthe essential role of conventionalobservation in gem identification.

Sally Chan

Colorless Untreated Diamondswith High Levels of StrainDiamonds form under enormousheat and pressure deep in the earth,and endure additional stresses duringtheir journey to the surface. Suchstresses can cause irregularities intheir lattice structure, which canimpart color to a diamond in a varietyof ways. For example, brown color intype IIa diamonds is believed to becaused by vacancy clusters, and thedepth of color generally correspondsto the degree of strain (D. Fisher et al.,“The vacancy as a probe of the strainin type IIa diamonds,” Diamond andRelated Materials, Vol. 15, 2006, pp.1636–1642, http://dx.doi.org/10.1016/j.diamond.2006.01.020; D. Fisher,“Brown diamonds and high pressurehigh temperature treatment,” Lithos,Vol. 112S, 2009, pp. 619–624, http://dx.doi.org/10.1016/j.lithos.2009.03.005). Since high-pressure, high-temperature (HPHT) treatment canremove brown color from type IIadiamonds but has a minimal effect

Figure 8. This 1.05 ct diamond owed its apparent Fancy Light brown-pinkcolor (left) to a coating. After the coating was removed by acid cleaning,the diamond was given a color grade of J (right).

Figure 7. In the DiamondView, the coating on the black diamond wasapparent from the different luminescence of the facet junctions,where it had apparently worn off (left). After cleaning with acetoneand corundum powder (with client permission), the coating appearedsmeared and the junctions were less distinct (right).

LAB NOTES GEMS & GEMOLOGY FALL 2011 225

on strain, the presence of strain canbe an important criterion for identi-fying HPHT-annealed stones.

Recently, the Carlsbad laboratoryreceived two colorless type IIa dia-monds for identification: a 0.90 ct E-color marquise and a 0.73 ct D-colorround brilliant. Based on gemologicaland spectroscopic features, these dia-monds were found to be natural anduntreated. However, in the course ofthis determination, we also noticedthat both showed an unusually highdegree of “tatami” strain (figure 11)when viewed with cross-polarized

light. The strong strain patterns weremore consistent with those seen inbrown diamonds. Typically, type IIacolorless diamonds show a tatamipattern, but with a lower concentra-tion of strain laminations.

In addition to visual assessmentusing crossed polarizers, an indirectbut relatively reliable indicator ofstrain is the peak width of certaindefects determined from photolumi-nescence (PL) spectra taken at liquidnitrogen temperature. The PL spec-trum of the 0.90 ct diamond showed

both the neutral (575 nm) and nega-tively charged (637 nm) NV centers;the full width at half maximum(FWHM) for these centers was 0.89and 1.56 nm, respectively. As withthe observed strain, these values weremuch higher than is commonly seenin untreated colorless diamonds. Acompilation of calculated FWHM val-ues for 250 untreated diamonds in theD–E range showed that the NV0 cen-ter had an average width of 0.29 nmwith a standard deviation of 0.07 nm;the highest value was 0.65 nm. Forthe NV– center, the average widthwas 0.30 nm with a standard devia-tion of 0.10 nm; the maximum was0.79 nm. These values are far lowerthan those for the 0.90 ct diamond.

It is unusual to find natural-colordiamonds that endured such a highamount of stress during their historyand yet remained colorless. One pos-sible explanation is that the associat-ed brown coloration was naturallyannealed out during their geologichistory.

Sally Eaton-Magaña

HPHT Treatment for Subtle Color Enhancement of DiamondIt is well known that HPHT anneal-ing can remove brown colorationfrom natural type IIa diamonds, mak-ing them colorless to near-colorless.While a significant number of HPHT-

550

VISIBLE ABSORPTION SPECTRA

WAVELENGTH (nm)

400

0

0.05

0.10

0.15

0.20

450 500 550 600 650 700 750

AB

SOR

BAN

CE

Coated pink diamond

Natural pink diamond

Figure 11. These untreated natural type IIa diamonds (0.90 ct E color, left;0.73 ct D color, right) experienced a high amount of stress, as evidenced bythe extensive “tatami” strain seen in cross-polarized light. They are unusu-al since they do not show any of the brown color that typically correlateswith such strain. (The apparent brown color in these images is an artifact ofthe lighting.)

Figure 9. The coated diamond’s absorption band at ~550 nm was verysimilar to that of a naturally colored pink diamond. Both spectra werecollected at liquid-nitrogen temperature.

Figure 10. A pink coating on thediamond’s pavilion facets wasclearly visible with magnification(here, 40×).

226 LAB NOTES GEMS & GEMOLOGY FALL 2011

treated type IIa diamonds have beenseen in gem laboratories over the pastdecade, little has been published onthe characteristics of the startingmaterials. Recently, however, theNew York laboratory had the oppor-tunity to follow four diamondsthrough the HPHT treatment process.

When they were first received,three of the four round diamonds(1.8–2.6 ct) were color graded in theJ–K range, with the fourth being Mcolor. Photoluminescence spectros -copy confirmed that they were natu-rally colored. Infrared spectroscopyshowed they were very pure type IIa;only one stone had an extremelyweak hydrogen-related absorption, at3107 cm−1. When the diamonds reap-peared in the lab very recently, theircolor ranged from E to F (figure 12).They also showed a 2–5% weightloss, consistent with the necessaryrepolishing after treatment. Their

photoluminescence spectra revealedthe exact changes that were expectedas a result of HPHT annealing.

Though the general impression ofHPHT treatment in the trade is that itis used to turn obviously brown dia-monds colorless, this experience sug-gests that such treatment may also beused to achieve more subtle improve-ments in diamond color.

Wuyi Wang

An Interesting LuminescentCleavage in DiamondAlthough diamond is the hardest natu-ral substance, it still is susceptible tobreakage along its cleavage directions.Often mistakenly described as a “frac-ture,” cleavage is one of the most com-mon clarity characteristics in gem-quality diamonds, and generally is oflittle interest to gemologists. However,

the Carlsbad laboratory recently grad-ed a type Ia colorless round brilliantwith a remarkable cleavage-related fea-ture. The cleavage showed prominentbrown radiation staining, suggestingthat radioactive fluids had been pres-ent in the post-growth environment(figure 13). When the diamond wasviewed face-up in the microscope withdarkfield illumination, the cleavagealso exhibited pronounced green lumi-nescence to visible light (figure 14,left); fiber-optic illumination made thegreen luminescence even more dra-matic. Radiation stains are well docu-mented in the gemological literature(e.g., J. I. Koivula, The Micro World ofDia monds, Gemworld International,North brook, Illinois, 2000).

Green luminescence in diamondis known to be produced by the H3lattice defect (503.2 nm), which con-sists of two nitrogen atoms separatedby a vacancy (N-V-N)0. When exam-ined in the DiamondView, the stoneshowed clear green fluorescence alongthe cleavage (figure 14, right). Usingthis image as a map, we collected pho-toluminescence data from the cleav-age area. In addition to the H3 defects,analysis also revealed high concentra-tions of the 3H interstitial defect(503.4 nm). An interstitial defectoccurs when a carbon atom is dis-placed from its original lattice posi-tion, typically by radiation. Thus, theoccurrence of 3H with the greenluminescing H3 defects is consistent

Figure 12. Shown here after HPHT treatment (E–F color, 1.74–2.57 ct),these type IIa diamonds originally had color grades of J–M. This demon-strates that HPHT annealing is being used to make subtle color improve-ments in such diamonds.

Figure 13. The brown colorationconfined to the cleavage area ofthis diamond resulted from theinteraction of radioactive fluidsin the post-growth environment.Magnified 20×.

LAB NOTES GEMS & GEMOLOGY FALL 2011 227

with the microscopic evidence ofradiation staining.

This is one of the more unusualexamples of cleavage we have seen inthe Carlsbad laboratory. The dia-mond’s cleavage, radiation staining,and green luminescence all revealimportant information about its his-tory and indicate that radiation expo-sure occurred naturally.

Nathan Renfro and Christopher M. Breeding

A Large HPHT-Annealed Pink Diamond

Unlike irradiation treatment, whichusually adds additional color, anneal-ing a natural diamond under HPHT

conditions can remove preexistingcolors as well as create new ones.When a suitable type IIa brown dia-mond is HPHT annealed at tempera-tures appropriate for removing brownbut not pink coloration, then the pre-existing pink color can be enhanced.Such HPHT-treated type IIa pink dia-monds have been available for morethan 10 years (e.g., Fall 2000 LabNotes, pp. 254–255).

The New York laboratory recentlyreceived the large pink diamond in fig-ure 15 for identification. This mar-quise brilliant (28.15 × 15.41 × 7.71mm) weighed 21.73 ct and was colorgraded Light pink. The color was dis-tributed evenly throughout the stonewith no detectable pink graining, andit showed typical blue fluorescencewhen examined in the DiamondView.The mid-IR spectrum showed type IIafeatures with no detectable hydrogen-related absorption at 3107 cm−1. Aweak band centered at ~550 nm waspresent in the UV-Vis absorption spec-trum at liquid nitrogen temperature.This band is very common in natural-ly colored pink-to-red diamonds andcan be made more visible by removingother unwanted colors. Photo lumi -nescence spectra collected at liquidnitrogen temperature with varyinglaser excitations confirmed that thisdiamond was HPHT annealed.

HPHT annealing has become avery common type of diamond colortreatment. With advances in process-ing techniques, a substantial number

of large diamonds can now be treatedsuccessfully, though these can still beidentified by gem labs. This 21.73 ctpink diamond is one of the largestHPHT-treated pink diamonds identi-fied in GIA’s laboratories.

Wuyi Wang

SYNTHETIC DIAMOND

Gem-Quality CVD SyntheticDiamonds from GemesisIn recent years, limited quantities ofgem-quality synthetic diamonds pro-duced by chemical vapor deposition(CVD) have reached the market. Mostof these have come from ApolloDiamond Inc. (Boston, Massa chusetts).We recently examined similar prod-ucts introduced by Florida-basedGemesis Corp., better known for itsHPHT-grown synthetic diamonds (J. E.Shigley et al., “Gemesis laboratory-cre-ated diamonds,” Winter 2002 G&G,pp. 301–309). In November 2010, thecompany announced plans to marketCVD-grown synthetics.

GIA examined 16 CVD syntheticdiamonds (e.g., figure 16) that werefaceted as round brilliants with theexception of one rectangular cut.They ranged from 0.24 to 0.90 ct, withan average weight of 0.46 ct. Theywere colorless (3 samples), near-color-less (11), and lightly colored (2). Forthe most part, clarity grades fellbetween IF and VVS, due to internalgraining and a few pinpoint inclu-sions. Only four of the 16 samples hadVS clarity.

Infrared absorption spectroscopyrevealed that all of the CVD syntheticdiamonds were type IIa. Unlike typicalas-grown CVD products, no defect-related absorptions were recorded ineither the mid- or near-infraredregions. Photoluminescence analysisof all samples showed moderatelystrong emission from the H3 opticalcenter (zero-phonon line at 503.2 nm),moderately strong emissions from NVcenters (575.0 and 637.0 nm), and mod-erate to strong emissions from the [Si-V]– center (doublet at 736.6 and 736.9

Figure 15. This 21.73 ct Lightpink diamond proved to beHPHT annealed.

Figure 14. Darkfield illumination revealed green luminescence along thecleavage shown in figure 13 (left, see arrows; magnified 20x). In theDiamondView (right), green luminescence from the H3 defect was clear-ly visible. Magnified 30×.

228 LAB NOTES GEMS & GEMOLOGY FALL 2011

nm). In four samples, the [Si-V]– con-centrations were relatively high andcould even be detected in the UV-Visabsorption spectra. A notable feature ofthese synthetic diamonds was theirweak to moderate green fluorescenceto short-wave UV radiation. Most wereinert to long-wave UV radiation, withonly five samples showing very weakgreen fluorescence. In the Diamond -View, they exhibited strong green fluo-rescence and noticeable blue phospho-rescence; characteristic growth stria-tions also were easily seen.

Gemological and spectroscopicobservations strongly suggested thatthese CVD synthetic diamonds wereannealed after their growth, presum-ably to improve their color and trans-parency. This study confirms thatthe quality of CVD synthetic dia-monds continues to improve. Never -theless, gemological and spectro-scopic features can clearly separateGemesis CVD synthetics from natu-ral diamonds.

Wuyi Wang and Thomas M. Moses

Treated Synthetic Diamonds withPink Color Intensified byFluorescenceA combination of irradiation andHPHT annealing can create pink tored colors in both natural and synthet-ic diamonds. We have reported on anumber of these treated stones over

the past few years (see, e.g., Lab Notes:Winter 2005, pp. 341–343; Spring2010, pp. 51–52 and 52–54; andWinter 2010, pp. 300–301). Two moreexamples recently submitted (on sepa-

rate occasions) to the New York labo-ratory showed an unusual fluores-cence effect.

The two round brilliants (0.47 and0.63 ct; figure 17) were color gradedFancy Vivid purplish pink and FancyDeep purple-pink, respectively. Theirmid-IR spectra indicated they weretype Ib, with a very low concentra-tion of isolated nitrogen. However,they did not display the “tatami”strain pattern expected for suchdiamonds, so we tested them usingthe Diamond View. The 0.47 ctround brilliant showed the cubocta-hedral pattern typical of HPHT syn-thetic growth, while the 0.63 ct sam-ple had a CVD growth pattern; bothdisplayed strong orange lumines-cence (figure 17, bottom). Micro-scopic examination revealed strongcolor zoning along growth sectors inthe 0.47 ct synthetic, but this effectwas very subtle in the 0.63 ct sample

Figure 17. These treated-color synthetic diamonds (0.47 and 0.63 ct) havean evenly distributed face-up color. Their Diamond View images (bottom)display the patterns typical of HPHT (left) and CVD growth (right).

Figure 16. These CVD synthetic diamonds were produced by Gemesis.The 0.39 ct round brilliant on the left was graded F color and VVS2clarity; the 0.83 ct sample on the right was graded J-VVS2.

LAB NOTES GEMS & GEMOLOGY FALL 2011 229

(figure 18). The former had only aVVS1-grade feather in the lower gir-dle facet, while the latter containedgraphitized feathers and tiny blackparticles (possibly graphite).

The diamonds’ UV-Vis-NIR ab -sorp tion spectra were collected at liq-uid-nitrogen temperature with broad-band illumination (Avantes AvaLight-HAL-S) and a CCD detector (OceanOptics HR-4000). This light sourcehas significant emission in the visi-ble-light region (>400 nm) but verylittle in the UV region. The 575 nmfeature (zero-phonon line of NV0)appeared as a negative peak (figure19), caused by fluorescence. Strongfluorescence was also present in the~650–825 nm region from the NV–

center (637 nm). In addition, a sharppeak at 595 nm was recorded in theabsorption spectra, and this peak,together with the strong NV featuresat 575 and 637 nm, were also detectedin low-temperature photolumines-cence spectra. These features suggestthat both pink synthetic diamondshad undergone post-growth treat-ments, such as irradiation and anneal-ing. These treatments caused thevacancies to migrate and combinewith preexisting isolated nitrogenatoms, forming the NV centers thatproduce the strong purplish pinkcolor.

The absorption spectra of mostpink diamonds that are naturally col-ored by NV centers show a transmis-sion window in the red region with a

smooth, gentle slope (indicated bythe blue dotted line in figure 19).These samples had a more pro-nounced window (represented by theshaded region) due to the emissionfrom NV– centers. Fluorescence fromthe NV– center, activated by visiblelight, strongly enhances pink to redcolor in diamond. This phenomenonis not as obvious in natural-colorpink diamonds that are colored by

NV centers, due to their much lowerconcentrations.

Kyaw Soe Moe

PEARLCultured Pearls from Pteria SternaWith Plastic Bead Nuclei

It is generally agreed that freshwatermother-of-pearl shell nuclei producethe best results for both saltwater andfreshwater beaded cultured pearls.Occasionally, other materials such assaltwater shell, ceramic, plastic, orwax have been used, but these are notcommon due to high rejection andmortality rates (P. C. Southgate and J.S. Lucas, Eds., The Pearl Oyster, Else -vier, Amster dam, 2008).

Recently, we received two yellow-ish brown samples (both ~8 mm indiameter; e.g., figure 20) for identifica-tion. Microradiography revealed cen-trally positioned beads that werenoticeably more transparent to X-raysthan typical freshwater shell nuclei.The beads appeared to have distinct

595

UV-VIS-NIR ABSORPTION SPECTRUM

WAVELENGTH (nm)

400 500 600

637

575

700 800 900 1000

AB

SOR

BAN

CE

Figure 19. In this UV-Vis-NIR absorption spectrum (here, of the 0.47 ct sam-ple), the unusual shaded region represents the strong emission of NV– cen-ters, which contribute to the pink bodycolor. In most natural-color pinkdiamonds, this region normally shows the absorption indicated by the bluedotted line.

Figure 18. In diffused light, the 0.47 ct pink synthetic diamond displayedstrong, uneven color zoning along growth sectors (left, magnified 45×).The 0.63 ct sample only showed subtle color zoning (right, 35×).

230 LAB NOTES GEMS & GEMOLOGY FALL 2011

uniform edges and drill holes, althoughthe cultured pearls themselves werenot drilled. One had a slight exteriordepression apparently correspondingto one end of the drill hole.

Further gemological testing (UVfluorescence, UV-Vis reflectance spec-troscopy, and EDXRF analysis) indi-cated that these samples were fromthe saltwater mollusk Pteria sterna, aspecies native to the Sea of Cortez inMexico. Although we have occasion-ally seen cultured pearls with near-X-ray-transparent beads, and reported onsome of them in the Lab Notes section(e.g., Summer 1988, pp. 114–115;Spring 1994, p. 45), we decided toacquire the samples from the client inorder to investigate further and identi-fy the material that was used.

One of the two samples was exam-ined using X-ray computed microto-mography to obtain more detailedimages of its internal structure (figure20, center). The resulting images con-firmed, with more detail, a distinctuniform bead outline and drill hole.The second sample was cut in half forvisual observation and Raman spectro-scopic analysis. The exposed bead wassemitranslucent, white, with a plastic-like appearance (figure 20, right), andwas easily scratched using a metalprobe. Raman spectroscopy showed adominant peak at 997 cm−1, which cor-responded to predominantly aromaticfunctional groups of possible poly-styrene materials.

This is the first time we havereported on plastic beads used in Pteria

sterna mollusks. Whether this atypicalbead material is becoming more preva-lent is unknown.

[Note: Though we have used theterm cultured pearl in this note forsimplicity, we recognize that in thisinstance the usage may not be inagreement with recognized defini-tions, e.g., the CIBJO Pearl Book.]

Chunhui Zhou and Akira Hyatt

Large Conch PearlPearls from the Queen conch, Strom -bus gigas, have long been collectors’items, prized for their unique andattractive color and surface structure.Conch pearls come in a range of col-ors from white to yellow to brown,but the most desirable is pink.

Recently, a large conch pearlmounted in a pendant (figure 21) wassubmitted to the New York laboratoryfor identification. The baroque-shapedlight pink to white specimen measured35.20 × 21.65 × 16.33 mm, weighedapproximately 100 ct, and showed avery fine flame structure (figure 22). X-radiography revealed an unusuallylarge number of layered natural growthstructures, which corresponded to thecontours of the pearl’s surface.

Raman and UV-Vis reflectancespectra were collected from both thelight pink and white regions of thepearl. Raman spectroscopy of the pinkarea with 514 nm excitation showedtwo small peaks at 1520 and 1130 cm−1,which are characteristic of the pig-ments responsible for pink color in

Figure 21. The very large conch pearl in the pendant weighs approximate-ly 100 ct. By comparison, most conch pearls (left, 8.04 ct, 13.53 × 10.65 ×7.95 mm) are much smaller, as are South Sea (second from right, 13.00mm) and akoya cultured pearls (right, 6.75 mm).

Figure 22. The large conch pearlin figure 21 displayed a very fineflame structure. Magnified 100×.

Figure 20. This cultured pearl from Pteria sterna (left, ~8 mm in diameter)proved to have a plastic bead. X-ray computed microtomography of asimilar sample (center) showed the presence of a near-X-ray-transparentbead as well as a drill hole. When the first sample was sawn in half(right), a white plastic bead was revealed inside.

LAB NOTES GEMS & GEMOLOGY FALL 2011 231

conch pearls. These peaks were barelyvisible in the spectrum of the whitearea. The UV-Vis reflec tance spectrumof the pink area also showed strongerabsorption between 480 and 560 nmthan did that of the white area; thisabsorption in the green region of thespectrum is responsible for the pinkcolor. All of these results suggestedthat the pink color was natural and notthe result of treatment.

Though most conch pearls aresmall (see E. Fritsch and E. B. Misi o -rowski, “The history and gemologyof Queen conch ‘pearls,’” Winter1987 G&G, pp. 208–221), large oneshave been reported. The 2005“Allure of Pearls” exhibit at theSmith soni an Museum of NaturalHistory (www.mnh.si.edu/exhibits/Pearls/index2.htm) featured two loosepink conch pearls belonging to collec-tor Susan Hendrickson that weighed17.70 and 22.40 ct, as well as theQueen Mary Brooch set with two pinkconch pearls weighing 24.90 and 28.10ct. We reported on another large conchpearl in 2008 (125.26 ct; Spring 2008Lab Notes, p. 72); though that baroquespecimen was somewhat larger thanthis pearl, the bodycolor was a mix ofpink, orange, and white. Most conchpearls are far smaller (again, see figure21), making this specimen a true rari-ty, in both size and quality.

JaeWon Chang and Akira Hyatt

SAPPHIRE

Natural and Synthetic GreenSapphires with Similar Color butRemarkably Different ChromophoresRecently, two green gems were submit-ted to the New York laboratory. Onestone (12.80 × 10.90 × 8.75 mm; figure23, left), mounted in a ring with numer-ous transparent near-colorless stones,was easily identified as a natural sap-phire by its inclusions and trace-ele-ment content measured by LA-ICP-MS.The other gem (2.07 ct; 7.77 × 6.09 ×4.93 mm) had no readily apparent inclu-sions (figure 23, center). Careful exami-nation of the latter gem in immersionrevealed faint curved color banding,

indicative of a melt-grown synthetic.LA-ICP-MS analysis confirmed that itwas synthetic sapphire, as cobalt wasthe only trace element detected (aver-age 25 ppma). Previous reports oncobalt-doped melt-grown synthetic sap-phires were published in Spring 1996Lab Notes (p. 51), Spring 2001 GemNews International (pp. 75–77), andSpring 2008 Lab Notes (pp. 72–73).

Sapphires—whether natural orsynthetic—with such a distinct greencoloration are rarely seen in the lab,and it is quite interesting that twovery different compositions producedsuch a similar coloration. UV-Vis-NIRspectroscopy of the natural sapphireshowed an absorption spectrum typi-cal of corundum with a high iron con-tent: strong lines at 450, 388, and 377nm from Fe3+, and broad bands at 580and 700 nm due to Fe2+-Ti4+ interva-lence charge transfer (figure 24). Inaddition, the 1050–1075 nm band (dueto Fe3+) was surprisingly strong. It iswell known that a high content of Fe3+

induces a yellow coloration in sap-phire, and even small concentrationsof Fe2+-Ti4+ pairs contribute a bluecolor component. The resulting greencolor of this sapphire, which had anaverage Fe content of 1155 ppma, isquite consistent with its composition.

Polarized UV-Vis-NIR spectra of

the synthetic sapphire showed onlytwo broad bands centered at 450 and660 nm, which have been attributedto Co3+ substituting for octahedralAl3+ (e.g., R. Müller and Hs. H. Gün -thard, “Spectroscopic study of thereduction of nickel and cobalt ions insapphire,” Journal of Chemi calPhysics, Vol. 44, No. 1, 1966, pp.365–373, http://dx.doi.org/10.1063/1.1726471). The two band positionswere virtually identical for the ordi-nary and extraordinary rays, but theintensity of the 660 nm band wasgreater for the ordinary ray. Further -more, a weak sharp line at approxi-mately 691 nm was visible in theordinary ray spectrum. A transmis-sion window near 500 nm in the spec-tra of both samples is largely responsi-ble for their similar green coloration.

Color coordinate calculationsusing the spectra reported in Müllerand Günthard (1966) for 2-mm-thickwafers gave L*a*b* coordinates of81/51/20 and 86/–34/36 for the ordi-nary ray and extraordinary ray, respec-tively (again, see figure 23). These cor-respond to a coloration that is remark-ably similar to the synthetic sapphirereported here, supporting our conclu-sion that its color is due to Co3+.

Emily V. Dubinsky

Figure 23. The 12.80-mm-long sapphire on the left was identified as natu-ral, with its color due to a high iron content. The 2.07 ct gem on the rightis a cobalt-bearing synthetic sapphire. For comparison, color coordinates(far right) were calculated for optical spectra reported in Müller andGünthard (1966) for Verneuil synthetic sapphire containing Co3+

(ordinary ray, top; extraordinary ray, bottom).

232 LAB NOTES GEMS & GEMOLOGY FALL 2011

Unusual Be-Diffused PinkSapphire . . . A Cautionary NoteThe Carlsbad laboratory recentlyexamined a 1.40 ct pink sapphire withunusual orange color zones that wassubmitted for a sapphire report. Initialmicroscopic observation revealednumerous thermally altered light-col-ored crystals, some of which were sur-rounded by discoid “fingerprints” (fig-

ure 25), an inclusion scene consistentwith high-temperature heat treat-ment. Immersion revealed a planarorange zone near the culet and a shal-low orange zone parallel to the tablefacet (figure 26). Orange color zones inheat-treated pink sapphires may beindicative of beryllium-diffusion treat -ment. However, when the stone wastested by laser–ablation inductivelycoupled plasma–mass spectrometry

(LA-ICP-MS) in the girdle area (as isGIA’s standard practice, to avoid plac-ing an ablation spot in an area thatcould affect the stone’s appearance),no Be was detected. This was not sur-prising because there was no orangecolor near the girdle.

The orientation of the orange zon-ing in this sapphire was perpendicularto the c-axis. This is a common orien-tation for natural orange zones in pad-paradscha sapphire that are caused bytrapped hole defects induced by anexcess of natural magnesium (J. L.Emmett et al., “Beryllium diffusion ofruby and sapphire,” Summer 2003G&G, pp. 84–135). To rule out suchorange zoning from naturally incorpo-rated trapped hole centers, it was nec-essary to check for the presence of Bedirectly in one of the orange colorzones. After obtaining permissionfrom the client, LA-ICP-MS measure-ments revealed more than 35 ppma Bein an orange area, which is enough toimpart significant orange color in apink sapphire.

It was obvious from the shape andorientation of the color zones that thestone was beryllium-diffusion treatedas a tabular preform or piece of roughthat had sufficient sacrificial materialto allow all traces of Be to be removedfrom the girdle area during cutting. It isdifficult to say whether this was adeliberate attempt to defeat the stan-dard practice of testing for beryllium inthe girdle. Nevertheless, careful gemo-logical observation in immersion andan understanding of color modificationin corundum prevented this Be-dif-fused sapphire from being misidenti-fied as a standard heated sapphire.

Nathan Renfro and Amy Cooper

Unusually High Beryllium in Three Blue SapphiresTrace amounts of natural berylliumare sometimes detected in untreatedsapphires, such as those from Mada -gascar. Typically, natural concentra-tions range from <1 ppma to about4–5 ppma Be. Higher concentrations(>15 ppma) are usually associatedwith beryllium diffusion treatment.

UV-VIS-NIR ABSORPTION SPECTRA

WAVELENGTH (nm)

300

0

0.5

1.5

2.0

1.0

2.5

3.0

3.5

4.0

4.5

500 700 900 1100 1300

AB

SOR

BAN

CE

Fe3+

Fe3+

Co3+ Co3+

Fe2+-Ti4+

Natural, mixed ray

Synthetic, ordinary ray

Synthetic, extraordinary ray

Figure 24. In these UV-Vis-NIR spectra, the natural green sapphire showsstrong Fe-related features that are absent from the synthetic sample,which has bands attributed to Co3+. The approximate path lengths of thebeam are 9 mm for the natural stone and 5 mm for the synthetic sample.

Figure 25. The altered appear-ance of these crystals, which areprobably zircon, is consistentwith the high-temperature heattreatment necessary for Be diffu-sion of the host sapphire.Magnified 70×.

Figure 26. The orange color zonesalong the culet and table of thispink sapphire, as seen immersedin methylene iodide, were causedby Be-diffusion treatment. Magni -fied 10×.

LAB NOTES GEMS & GEMOLOGY FALL 2011 233

In addition, naturally occurring beryl-lium in corundum is often presentwith traces (usually <1 ppma) of nio-bium, tantalum, light rare-earth ele-ments, and thorium (see A. Shen etal., “From the GIA Laboratory: Beryl -lium in corundum: The consequencesfor blue sapphire,” GIA Insider, Vol.9, No. 2, Jan. 26, 2007).

During routine testing, three bluesapphires recently submitted for iden-tification were found to contain rela-tively high levels of beryllium. Thethree consisted of one ring-mountedstone and two loose sapphires weigh-

ing 3.17 and 3.21 ct (e.g., figure 27).Two of the samples showed micro-scopic evidence of heating, while theother sapphire (3.17 ct) had inclusionfeatures that proved it was unheated.The 3.21 ct stone showed a chalkyblue reaction to short-wave UV radia-tion (consistent with its low iron con-tent) and strong blue bands whenviewed in immersion, while the othertwo sapphires were inert to long- andshort-wave UV radiation due to theirrelatively high iron contents. Allthree stones contained milky clouds.

Trace-element analysis of all threesapphires with LA-ICP-MS showed awide range of Be concentrations, fromvirtually absent up to 33 ppma (figure28). The concentrations of Be were, ingeneral, positively correlated to varioustransition metals and light rare-earthelements, including niobium, tanta-lum, tungsten, lanthanum, cerium,hafnium, and thorium (e.g., figure 29).Higher concentrations of all these ele-ments, including Be, were often associ-ated with the clouds, as documentedpreviously in Be-bearing untreated sap-phires. Magnesium also showed a posi-tive correlation with Be, but with large

variations. Titanium showed both pos-itive and negative correlation with Be,possibly due to Ti concentration varia-tions in the blue zones.

The amount of Be in these sap-phires is similar to or higher than thatrecorded in some Be-diffused stones.Their properties suggest that theycame from at least two differentdeposits or deposit types, but their spe-cific origins could not be determined.This is another example showing thatthorough gemo logical and chemicalanalysis is necessary to identify beryl-lium-diffusion treatment.

Ren Lu and Andy H. Shen

PHOTO CREDITSJian Xin (Jae) Liao—1, 3, 5, 8, 12, 15, 16,20 (left), 21, 23, and 27 (left); Kyaw SoeMoe—2, 17 (bottom), and 18; WuyiWang—4; Erica Emerson—6 and 7; Jason Darley—10; David Nelson—11;Nathan Renfro—13, 14, 25, and 26; Sood Oil (Judy) Chia—17 (top) and 23(right); Nicholas Sturman—20 (center);Chunhui Zhou—20 (right); JaeWonChang—22; Ren Lu—27 and 28.

Figure 28. Higher concentrationsof Be in this blue sapphire (val-ues in ppma are shown next toeach spot) were often associatedwith the presence of milkyclouds. Magnified ~75×.

Nb AND Ta VS. Be

Be (ppma)

00

40

20

30

10

50

5 10 15 20 25 30 35 N

b or

Ta

(ppm

a)

Figure 29. The concentrations of niobium (Nb, circles) and tantalum (Ta,triangles) showed an overall positive correlation to Be levels in the threesapphires (red = 3.17 ct, blue = 3.21 ct, and green = mounted sapphire). Toconvert to units of ppmw, the ppma values are multiplied by 0.44 for Be,4.56 for Nb, and 8.87 for Ta.

Figure 27. This 3.21 ct sapphirecontained relatively high levelsof naturally occurring beryllium.