Synthetic nano-polycrystalline diamond (NPD)may be one of the most important developments

in synthetic diamond production in recent years.This transparent brownish yellow material is pro-duced not by CVD or traditional HPHT synthesismethods, but rather in a multi-anvil press by a sin-tering process that converts high-purity graphite di-rectly into synthetic diamond. According to thedeveloper of the process, the conversion time aver-ages 10–20 minutes (though can be less than 10 min-utes) at 15 gigapascals and 2,300–2,500°C (T. Irifune,pers. comm., 2012). These pressures and tempera-tures are far higher than those used in the HPHT syn-

thesis of single-crystal diamond. The material con-sists of randomly oriented nanoscale-sized syntheticdiamond crystallites that have been bonded tightlytogether to form what may be thought of as an ultra-hard synthetic diamond ceramic.

A recent article by Skalwold (2012) documented atransparent 5 mm NPD sphere, and served as meansto introduce this new synthetic diamond’s develop-ment and some of its properties. Soon after that arti-cle was published, Dr. Tetsuo Irifune, director ofEhime University’s Geodynamics Research Center(GRC), once again offered one of the authors (EAS) anexclusive opportunity to study this unique material.This time the specimen was a 0.88 ct faceted roundbrilliant (figure 1). In collaboration with researchersat GIA, an analysis of its gemological and spectro-scopic properties was undertaken to establish gemidentification criteria. Further work on this materialwill be conducted by author EAS and by GIA as moresamples become available.

A potential new synthetic gem material in theform of a faceted 0.88 ct brownish yellow nano-polycrystalline diamond (NPD) has undergone afull gemological examination, including detailedspectroscopic analysis, for the first time. While thematerial’s implications for the gem industry arestill unknown, NPD’s optical characteristics alongwith manufacturing improvements make it a po-tentially important synthetic for gemologists to beaware of, even though very few samples havebeen faceted and future production costs andavailability for gem purposes are uncertain. Fea-tures seen with magnification may provide an in-dication of its nano-polycrystalline nature, andthe identification of this material can be con-firmed by spectroscopy and other advanced test-ing methods. NPD is an entirely new kind ofsynthetic diamond, and its development illustratesthe ongoing research on diamond properties fora number of important applications beyond thefield of gemology.

188 NOTES & NEW TECHNIQUES GEMS & GEMOLOGY FALL 2012

CHARACTERIZATION OF A SYNTHETICNANO-POLYCRYSTALLINE DIAMOND GEMSTONEElise A. Skalwold, Nathan Renfro, James E. Shigley, and Christopher M. Breeding

NOTES & NEW TECHNIQUES

See end of article for About the Authors and Acknowledgments.GEMS & GEMOLOGY, Vol. 48, No. 3, pp. 188–192,http://dx.doi.org/10.5741/GEMS.48.3.188.© 2012 Gemological Institute of America

Figure 1. This 0.88 ct synthetic nano-polycrystallinediamond, produced by direct conversion fromgraphite at high temperatures and pressures, repre-sents a completely new type of transparent gem mate-rial. Photo by Robert Weldon.

NEW Skalwold Notes & New Tech fall 2012_Layout 1 9/27/12 11:06 AM Page 188

As initially reported by Skalwold et al. (2012),while this transparent gem material shares somesimilarities with natural and synthetic single-crystaldiamond, it also has properties that are sufficientlydistinctive to be recognizable by gem-testing labora-tories. Given that there are only five known facetedsamples in existence, the opportunity to characterizeit with the assistance of a gemological laboratory wasa singular opportunity.

PRODUCTIONRelatively swift advancements have been made in themanufacturing process of NPD. In less than a decadesince the successful production of this material wasfirst reported (Irifune et al., 2003), improvements haveresulted in relatively large pieces suitable for a numberof applications, most notably for diamond anvils usedin high-pressure research. For this purpose, two NPDcylinders are produced per week, generally measuring6.5–9 mm in diameter and 6.5–8 mm in length (amaximum of 10 mm in each dimension is possible).But if the GRC’s Botchan multi-anvil press were toproduce only NPD, up to 10 cylinders could be madeeach week. The pieces recovered from the press havea slight surface texture, caused by the tantalum cap-

sule in which the graphite-to-NPD conversion takesplace (see figure 4 of Skalwold, 2012). These cylindersare cut by a combination of specialized lasers and pol-ishing techniques. The material requires nontradi-tional cutting and polishing methods because of itspolycrystalline nature and polishing hardness. Unlikenatural and other synthetic diamonds, NPD is uni-formly hard in all directions at the macro scale (al-though hardness varies on the nano scale).

Efforts are now under way at the GRC to producecolorless NPD. Marketing of the material is still inits infancy, and pricing for industrial use has yet tobe established (T. Irifune, pers. comm., 2012).

MATERIALS AND METHODSA 0.88 ct round brilliant studied for this report meas-ured 6.13 × 6.10 × 3.78 mm. It was graded for color andclarity using standard GIA procedures. Basic gemolog-ical properties were collected (table 1), and additionaltesting was performed with a Presidium GEM Mini-DiamondMaster and a DTC DiamondSure screeningdevice. The sample was examined with a Nikon SMZ-1500 binocular gemological microscope under severallighting configurations. Deep-ultraviolet (~230 nm ex-citation) fluorescence images were captured using aDTC DiamondView instrument.

Visible absorption spectra were recorded with thesample cooled to liquid-nitrogen temperature (~77 K)using a custom-made Ocean Optics high-resolutionspectrometer. Infrared absorption spectra wererecorded with a Thermo-Nicolet Magna IR 760 spec-trometer, while Raman and photoluminescence (PL)spectra were measured with a Renishaw InVia Ramanspectrometer. PL spectra were recorded at liquid-ni-trogen temperature using four excitation wavelengths(325, 488, 514, and 633 nm).

NOTES & NEW TECHNIQUES GEMS & GEMOLOGY FALL 2012 189

TABLE 1. Basic gemological properties of the NPDsample.

Color

Clarity

Appearance

Luster

Refractive index

Specific gravity

UV fluorescence

Long-wave

Short-wave

Phosphorescence

Magnetic attraction

Diamond testers

Presidium

DiamondSure

Magnification

Fancy Deep brownish yellow

Slightly Included (SI1)

Slightly hazy or cloudy

Adamantine

Over the limits of the standard refractometer

3.52

Moderate chalky reddish orange

Weak reddish orange

None observed

None observed

Positive for diamond

Referred for further testing; not “passed” asnatural diamond

Roiled appearance; one black andnumerous pinpoint inclusions; patchyclouds of color zoning

In Brief • Synthetic nano-polycrystalline diamond (NPD) consistsof nanoscale-sized synthetic diamond crystallitesbonded together using an HPHT sintering process.

• A 0.88 ct brownish yellow round brilliant, one of fivefaceted NPD gems currently in existence, was charac-terized for this report.

• Key identifying properties include a hazy, roiled ap-pearance and distinctive mid-IR, visible-range, andphotoluminescence spectra.

• This sample was cut from early production, and fu-ture NPD material may not show the same diagnosticfeatures.

NEW Skalwold Notes & New Tech fall 2012_Layout 1 9/27/12 11:06 AM Page 189

RESULTS AND DISCUSSIONThe GIA color and clarity grades, as well as basicgemological properties, are shown in table 1. Withmagnification, we observed numerous whitish pin-point inclusions along with a single larger black in-clusion, patchy clouds of color zoning, and a roiledappearance (figure 2). This last feature was much likethe effect sometimes seen in the hessonite variety ofgrossular, and it appeared to cause the sample’s slighthaziness when viewed in the face-up position. Theblack inclusion showed a dominant broad Ramanpeak at 1078 cm–1, but we were unable to identify amineral phase from the spectrum. Observation withcrossed polarizing filters showed a cross-hatched ortatami-like strain pattern in several orientations, withno apparent variation in the color, intensity, or distri-bution of the pattern depending on direction (figure3). While the hazy, roiled appearance is quite unusual,visual features alone should not be considered com-pletely diagnostic for NPD, but rather as indicationsthat laboratory testing is needed.

Fluorescence imaging using the DiamondViewshowed orangy red luminescence with an irregularstructure (figure 4). This unusual patchy appearancebears some resemblance to the networks or “webs”often seen in fluorescence images of natural type II di-amonds. The orangy red luminescence seems to arisefrom boundaries between what appear to be larger ag-gregates of the nano-crystallites. This distinctive lumi-

nescence is similar in appearance and color to that gen-erated by N-V optical centers in diamond.

We reexamined the sample with the microscopeto look for evidence of the nano-crystallite aggregates.Inserting a quarter-wave plate and observing the bire-fringence while rotating the sample showed that most

190 NOTES & NEW TECHNIQUES GEMS & GEMOLOGY FALL 2012

Figure 2. The NPD sam-ple displayed some in-teresting visual featuresattributed in part to thepolycrystalline natureof the material. Smallwhitish pinpoint inclu-sions (A, image width1.7 mm) were scatteredthroughout the speci-men. A dark inclusion(B, image width 0.65mm) was observed butcould not be identifiedwith Raman analysis.Also observed were amottled uneven col-oration (C) and a finelytextured roiled appear-ance (D, image width3.6 mm). Photomicro-graphs by N. Renfro.

Figure 3. When examined between crossed polariz-ing filters, the NPD sample displayed a cross-hatched or tatami-like pattern of low-order (darkgray) interference colors. Unlike similar birefrin-gence seen occasionally in natural diamonds, it re-mained consistent in color and distributionregardless of the sample’s orientation. Photomicro-graph by N. Renfro; image width 4.4 mm.

A B

C D

NEW Skalwold Notes & New Tech fall 2012_Layout 1 9/27/12 11:06 AM Page 190

of the birefringence was due to strain within the in-dividual domains that correspond to those seen withthe DiamondView. The strain in each of the domainswas almost certainly low, causing no more than first-order gray and white interference colors. The samplealso showed evidence of a global strain birefringencesuperimposed on the individual domain birefrin-gence, and it too was first-order gray and white (W. A.Bassett, pers. comm., 2012).

Spectroscopy. The NPD sample’s absorption spectrawere distinct from those of natural diamond. Absorp-tion features corresponding to the mid-infrared one-phonon region of diamond were observed (figure 5),

but nitrogen aggregation state and diamond type couldnot be determined because the spectrum was differentfrom that of type I and type II diamond (Breeding andShigley, 2009). No evidence of hydrogen or boron im-purities was observed in the mid-infrared spectrum.

The visible spectrum displayed increasing absorp-tion toward the blue region, as well as two absorptionbands at approximately 612 and 667 nm (figure 6).This increasing absorption is responsible for the sam-ple’s brownish yellow color. The two absorptionbands are likely associated with the orangy red fluo-rescence, but the assignment of the defect(s) causingthese two bands is unclear. Luminescence features ofsimilar energies reported in CVD synthetic diamond(Dischler, 2012, pp. 108–109) have been attributed to[N-V]0 and [N-V]– centers. Further investigation willbe required to determine if these are indeed the same

NOTES & NEW TECHNIQUES GEMS & GEMOLOGY FALL 2012 191

Figure 5. The NPD sample’s mid-infrared spectrum re-veals an absorption pattern unlike that of natural orsynthetic diamond, especially in the region below1500 cm–1. This portion of the spectrum contains fea-tures that may be due to nitrogen, although diamondtype could not be determined from them. There areno features in this spectrum that suggest the presenceof boron or hydrogen.

WAVENUMBER (cm-1)

AB

SOR

BAN

CE

3500

MID-IR SPECTRUM

2500 2000 1500 10000

2

4

4000 3000

3

1

Figure 4. When examinedin the DiamondView, theNPD sample showed un-evenly distributed orangyred luminescence. Irregulardarker areas of varying sizeare surrounded by the lu-minescence, which seemsto originate along bound-aries that may correspondto aggregated domains ofnano-crystallites. Imagesby David Nelson; field ofview is 3.5 mm (right).

Figure 6. The visible spectrum of the NPD sample dis-plays increasing absorption toward the blue region,which gives rise to its brownish yellow color. Absorp-tion bands at 612 and 667 nm of unclear origin maybe related to the orangy red luminescence observedwith the DiamondView.

WAVELENGTH (nm)

AB

SOR

BAN

CE

VISIBLE-RANGE SPECTRUM

8500

0.30

800750700650600550500450400

0.25

0.20

0.15

0.10

0.05

612

667

NEW Skalwold Notes & New Tech fall 2012_Layout 1 9/27/12 11:06 AM Page 191

optical defects responsible for the fluorescence shownby the NPD material.Photoluminescence spectra recorded with the

four laser excitations displayed weak but similar lu-minescence features at several wavelengths. Thespectrum recorded with 514 nm excitation (figure 7)exhibited the same two peaks at 612 and 667 nm thatwere recorded in the visible absorption spectrum.As reported by Sumiya et al. (2009), the character-

istic brownish yellow color of NPD is thought to becaused by lattice defects within each crystallite due to

plastic deformation resulting from the HPHT condi-tions used for its production. Further investigation isnecessary to confirm the cause of color in this material.

CONCLUSIONSUntil now, gem-quality synthetic diamonds withgood transparency have all reportedly consisted ofsingle-crystal material. This examination of a trans-parent faceted synthetic nano-polycrystalline dia-mond marks a fundamental change. The distinctiveproperties of this material should be readily identifi-able with routine gemological testing. The observa-tion of a roiled appearance within a diamond-likespecimen, though not diagnostic, should alert gemol-ogists that advanced testing is required. The authors stress that the sample reported here

is only one example from a rapidly improving tech-nology. It was faceted from early production and notdestined for applications where clarity, polish, anduniformity are a priority. Therefore, visual observa-tions of this particular sample do not necessarilyapply to all NPD specimens. This underscores theneed for advanced laboratory testing and ongoing re-search as more material becomes available. While NPD has been developed primarily for

high-tech applications that require superior hardnessand toughness, the present sample illustrates its po-tential as a gemstone. In the future, successful effortsto improve or remove color, either during productionor with post-production treatment, may expandNPD’s viability as a gemstone.

192 NOTES & NEW TECHNIQUES GEMS & GEMOLOGY FALL 2012

ABOUT THE AUTHORSMs. Skalwold (elise@nordskip.com) is a gemologist and authorinvolved in research and curating at Cornell University in Ithaca,New York. Mr. Renfro is a staff gemologist, Dr. Shigley is a distin-guished research fellow, and Dr. Breeding is a research scientist,at GIA in Carlsbad.

ACKNOWLEDGMENTSThe authors are grateful to Dr. Tetsuo Irifune, director of the Geo-

dynamics Research Center at Ehime University (Matsuyama,Japan), for generously loaning the faceted NPD sample and for in-sights regarding its nature. For their assistance and helpful discus-sion, we also thank: Dr. William A. Bassett, professor emeritus ofgeology at Cornell University; Dr. Steven D. Jacobsen, associateprofessor of earth and planetary sciences at Northwestern Univer-sity in Evanston, Illinois; David Christie of IMR Test Labs in Lans-ing, New York; and Al Gilbertson, John I. Koivula, David Nelson,and Robert Weldon of GIA in Carlsbad.

Figure 7. This photoluminescence spectrum of theNPD sample, obtained with 514 nm laser excitation,displays the same two spectral features seen in thevisible spectrum at 612 and 667 nm, along with someweaker bands of unknown origin.

WAVELENGTH (nm)

CO

UN

TSPHOTOLUMINESCENCE SPECTRUM

5500

600 650 700 750 800 850

20000

40000

60000

80000

100000

120000

612

667

Breeding C.M., Shigley J.E. (2009) The “type” classification systemof diamonds and its importance in gemology. G&G, Vol. 45,No. 2, pp. 96–111, http://dx.doi.org/10.5741/GEMS.45.2.96.

Dischler B. (2012) Handbook of Spectral Lines in Diamond, Vol-ume 1: Tables and Interpretations. Springer-Verlag, Heidelberg.

Irifune T., Kurio A., Sakamoto S., Inoue T., Sumiya H. (2003) Ul-trahard polycrystalline diamond from graphite. Nature, Vol.421, No. 6923, pp. 599–600, http://dx.doi.org/10.1038/421599b.

Skalwold E.A. (2012) Nano-polycrystalline diamond sphere: Agemologist’s perspective. G&G, Vol. 48, No. 2, pp. 128–131,http://dx.doi.org/10.5741/GEMS.48.2.128.

Skalwold E.A., Renfro N., Breeding C.M., Shigley J.E. (2012) Trans-parent, faceted nano-polycrystalline synthetic diamond. GIANews from Research, www.gia.edu/research-resources/news-from-research/NPD_Diamond.pdf, July 25 [date accessed: July26, 2012].

Sumiya H., Harano K., Arimoto K., Kagi H., Odake S., Irifune T.(2009) Optical characteristics of nano-polycrystalline diamondsynthesized directly from graphite under high pressure andhigh temperature. Japanese Journal of Applied Physics, Vol.48, No. 12, article 120206, http://dx.doi.org/10.1143/JJAP.48.120206.

REFERENCES

NEW Skalwold Notes & New Tech fall 2012_Layout 1 9/27/12 11:06 AM Page 192