SUMMER 2020

VOLUME LVI

THE QUARTERLY JOURNAL OF THE GEMOLOGICAL INSTITUTE OF AMERICA

Naturally Colored Yellow and Orange Diamonds

New Technique for Separating Natural from
Laboratory-Grown Diamond
History of the Chivor Emerald Mine, Part II
Optical Whitening and Brightening of Pearls
Summer 2020
VOLUME 56, No. 2
EDITORIAL
193 Yellow and Orange Diamonds, New Identification Techniques, and the
Chronicles of Chivor
Duncan Pay

FEATURE ARTICLES
194 Naturally Colored Yellow and Orange Gem Diamonds: The Nitrogen Factor
Christopher M. Breeding, Sally Eaton-Magaña, and James E. Shigley
Presents the four major groups of defects responsible for the color in nearly all yellow and
p. 195 orange diamonds.
220 Separation of Natural from Laboratory-Grown Diamond Using Time-Gated
Luminescence Imaging
Colin D. McGuinness, Amber M. Wassell, Peter M.P. Lanigan, and Stephen A. Lynch
Researchers from De Beers Group Technology present a technique that expands on the
surface luminescence used in the DiamondView instrument to provide additional imaging
information and quickly distinguish colorless or near-colorless natural diamond from
laboratory-grown diamond.
230 History of the Chivor Emerald Mine, Part II (1924–1970):
Between Insolvency and Viability
p. 255 Karl Schmetzer, Gérard Martayan, and Andrea R. Blake
Chronicles developments at the Chivor emerald mine in Colombia between 1924 and 1970,
based on archival records.
258 Optical Whitening and Brightening of Pearls: A Fluorescence Spectroscopy Study
Chunhui Zhou, Tsung-Han Tsai, Nicholas Sturman, Nanthaporn Nilpetploy,
Areeya Manustrong, and Kwanreun Lawanwong
p. 261
Presents a study that suggests optically brightened pearls can be consistently separated from
non-brightened pearls using fluorescence spectroscopy.
266 Hydrogen and Oxygen Stable Isotope Ratios of Dolomite-Related Nephrite:
Relevance for Its Geographic Origin and Geological Significance
Kong Gao, Ting Fang, Taijin Lu, Yan Lan, Yong Zhang, Yuanyuan Wang, and Yayun Chang
Examines the effectiveness of hydrogen and oxygen isotopes in discriminating dolomite-
p. 269
related nephrites from the four most important occurrences worldwide.

REGULAR FEATURES
281 Lab Notes
Cat’s-eye andradite • Quench-crackled dyed blue chalcedony resembling Larimar • Irradiated
blue diamond with interesting DiamondView image • Multiple radiation stains suggest inter-
p. 293 esting geological residency • Hemimorphite resembling Paraíba tourmaline • Exceptional
purple Montana sapphire • Negative crystal containing a mobile CO2 bubble in blue sapphire
heated with pressure • Pink Mozambique sapphire • HPHT-processed CVD lab-grown
diamonds with low color grades • HPHT lab-grown pink diamond with unstable color
292 G&G Micro-World
Böhmite in corundum • Chabazite in Ethiopian opal • Fluorophlogopite in Burmese spinel
• Unusually shaped rutile crystal inclusions in Russian emerald • Sapphire with negative
crystal containing mobile graphite daughter crystal • Windmills in rare mineral sphalerite
• Trapiche-like pattern in emerald from Pakistan • Quarterly Crystal: Pyrite in quartz
298 Gem News International
Purplish pink diaspore reportedly from Afghanistan • Fossil ivory update with Lee Downey
• The “Fragility of the Eternal” kunzite: Origin, cutting, and identification • Petrified woods
in the Russian Far East: Gemology and origin • Portrait of a Paraíba rough • “Rhodatrolite,” a
new gem material from Indonesia • Field trip to Crater of Diamonds, USA • Greenish blue
p. 312 glass imitating gem silica • Marble imitation of jadeite rough
Editorial Staff
Editor-in-Chief Editors, Lab Notes Assistant Editor
Duncan Pay Thomas M. Moses Erin Hogarth gia.edu/gems-gemology
Managing Editor Shane F. McClure
Contributing Editors
Subscriptions
Stuart D. Overlin Editors, Micro-World James E. Shigley Copies of the current issue may be purchased for
[email protected] Nathan Renfro Raquel Alonso-Perez $29.95 plus shipping. Subscriptions are $79.99 for one
Associate Editor Elise A. Skalwold Donna Beaton year (4 issues) in the U.S. and $99.99 elsewhere. Cana-
John I. Koivula dian subscribers should add GST. Discounts are avail-
Brooke Goedert Editor-in-Chief Emeritus able for renewals, group subscriptions, GIA alumni,
Technical Editors Editors, Gem News Alice S. Keller and current GIA students. To purchase print subscrip-
Tao Z. Hsu Emmanuel Fritsch tions, visit store.gia.edu or contact Customer Service.
Gagan Choudhary Customer Service For institutional rates, contact Customer Service.
[email protected]
Christopher M. Breeding Martha Erickson
Jennifer Stone-Sundberg (760) 603-4502 Database Coverage
[email protected] [email protected] Gems & Gemology’s impact factor is 1.844, accord-
ing to the 2017 Thomson Reuters Journal Citation
Reports (issued July 2018). G&G is abstracted in
Production Staff Thomson Reuters products (Current Contents: Phys-
ical, Chemical & Earth Sciences and Science Cita-
Creative Director Photo/Video Producer Video Production
tion Index—Expanded, including the Web of
Faizah Bhatti Kevin Schumacher Larry Lavitt Knowledge) and other databases. For a complete list
Production and Illustrator Pedro Padua of sources abstracting G&G, go to gia.edu/gems-
Multimedia Specialist Albert Salvato gemology, and click on “Publication Information.”
Russel Samson
Juan Zanahuria
Multimedia Associate Manuscript Submissions
Photographer Christopher Bonine Gems & Gemology, a peer-reviewed journal, welcomes
Robert Weldon the submission of articles on all aspects of the field.
Please see the Author Guidelines at gia.edu/gems-
gemology or contact the Managing Editor. Letters
Editorial Review Board on articles published in G&G are also welcome.
Please note that Field Reports, Lab Notes, Gem News
Ahmadjan Abduriyim Richard W. Hughes George R. Rossman International, Micro-World, Diamonds from the
Tokyo, Japan Bangkok, Thailand Pasadena, California Deep, and Charts are not peer-reviewed sections but
Timothy Adams Jaroslav Hyršl Andy Shen do undergo technical and editorial review.
San Diego, California Prague, Czech Republic Wuhan, China
Copyright and Reprint Permission
Edward W. Boehm Dorrit Jacob Guanghai Shi Abstracting is permitted with credit to the source.
Chattanooga, Tennessee Sydney, Australia Beijing, China Libraries are permitted to photocopy beyond the lim-
James E. Butler A.J.A. (Bram) Janse James E. Shigley its of U.S. copyright law for private use of patrons.
Washington, DC Perth, Australia Carlsbad, California Instructors are permitted to reproduce isolated arti-
cles and photographs/images owned by G&G for
Alan T. Collins Mary L. Johnson Elisabeth Strack noncommercial classroom use without fee. Use of
London, UK San Diego, California Hamburg, Germany photographs/images under copyright by external par-
Sally Eaton-Magaña Stefanos Karampelas Nicholas Sturman ties is prohibited without the express permission of
Carlsbad, California Basel, Switzerland Bangkok, Thailand the photographer or owner of the image, as listed in
the credits. For other copying, reprint, or republica-
John L. Emmett Lore Kiefert Fanus Viljoen tion permission, please contact the Managing Editor.
Brush Prairie, Washington Lucerne, Switzerland Johannesburg,
Gems & Gemology is published quarterly by the
Emmanuel Fritsch Ren Lu South Africa
Gemological Institute of America, a nonprofit educa-
Nantes, France Wuhan, China Wuyi Wang tional organization for the gem and jewelry industry.
Eloïse Gaillou Thomas M. Moses New York, New York Postmaster: Return undeliverable copies of Gems &
Paris, France New York, New York Christopher M. Welbourn Gemology to GIA, The Robert Mouawad Campus,
Reading, UK 5345 Armada Drive, Carlsbad, CA 92008.
Al Gilbertson Aaron Palke
Carlsbad, California Carlsbad, California Chunhui Zhou Our Canadian goods and service registration number
New York, New York is 126142892RT.
Gaston Giuliani Ilene Reinitz
Nancy, France Chicago, Illinois J.C. (Hanco) Zwaan Any opinions expressed in signed articles are under-
Leiden, The Netherlands stood to be opinions of the authors and not of the
Lee A. Groat Nathan Renfro publisher.
Vancouver, Canada Carlsbad, California
Yunbin Guan Benjamin Rondeau
Pasadena, California Nantes, France

About the Cover

Naturally yellow and orange diamonds, both of which are colored by nitrogen defects, are the subject of the lead
article in this issue. Orange diamonds are a rarity of nature, ranging from subtle hues to vivid intensities of color.
The rectangular diamond shown on the cover has a Fancy Vivid orange color and weighs 1.32 ct. The oval is a
Fancy pinkish orange, internally flawless diamond weighing 2.48 ct. Photo by Robert Weldon/GIA; courtesy of
Christian Tse.
Printing is by L+L Printers, Carlsbad, CA.
GIA World Headquarters The Robert Mouawad Campus 5345 Armada Drive Carlsbad, CA 92008 USA
© 2020 Gemological Institute of America All rights reserved. ISSN 0016-626X
Yellow and Orange Diamonds, New
Identification Techniques, and the
Chronicles of Chivor
Welcome to the Summer Gems & Gemology! This issue is packed with informative contents,
including the characteristics of yellow and orange diamonds, new gem identification techniques,
and the colorful history of the Chivor emerald mine.

In the world of fancy-color diamonds, yellow diamonds are by far the most common while those
with an unmodified orange hue are among the very rarest (kudos to Robert Weldon for capturing
them on the cover). Both yellow and
orange diamonds owe their color to “In the world of fancy-color diamonds,
nitrogen-related defects. In the lead yellow diamonds are by far the most
article, a GIA team led by Christopher
M. Breeding and Sally Eaton-Magaña common while those with an unmodified
presents the last in their series orange hue are among the very rarest.”
documenting the gemological and spectroscopic properties
of fancy-color diamonds.

Next, Colin McGuinness and colleagues present a luminescence imaging technique developed by De Beers Group
Technology that can quickly and easily distinguish colorless and near-colorless natural diamond from laboratory-grown
diamond. One marker can identify more than 99% of natural type Ia and type IIa diamond. A pair of newly developed
instruments expand on the technology of the DiamondView instrument introduced by De Beers a quarter of a century ago.

In the third article, Karl Schmetzer and coauthors conclude a sweeping two-part series chronicling the Chivor emerald
mine in Colombia. This article examines events from 1924 to 1970, when the mine was principally owned by an
American firm. While the early era was marked by stock market speculation and scandal, there were also highly produc-
tive periods at the fabled mine.

Optical brightening agents for pearls, which are separate from bleaching and maeshori treatments, are the topic of the
fourth article. A team of GIA researchers led by Chunhui Zhou provides a fluorescence spectroscopy technique for
confidently separating optically brightened pearls from non-brightened pearls.

In the final article, a team of researchers led by Kong Gao of NGTC in China provide a means of discriminating the
geographic origin of dolomite-related nephrite from the four most important sources worldwide. This determination
relies on hydrogen and oxygen stable isotope ratios. Isotope analysis is an emerging method in gemological research.

As always, our regular columns offer an illuminating array of insights. Highlights of the Lab Notes section include an
exceptional purple sapphire from Montana, a vivid blue hemimorphite cabochon resembling Paraíba tourmaline, and
interesting observations on an irradiated blue diamond and a rough diamond crystal with irradiation staining. Micro-
World reveals rare and breathtaking observations from the inner landscapes of gemstones, including what is believed to be
the first reported inclusion of fluorophlogopite in spinel (or any other gem material). In Gem News International, you’ll
find a characterization of purplish pink diaspore reportedly from a new deposit in Afghanistan, a look at the cutting and
identification of the world’s largest faceted kunzite, a study of petrified woods from the Russian Far East, and much more.

On a final note, the G&G Facebook group (www.facebook.com/groups/giagemsgemology) is now more than 10,000
members strong. We thank you for your participation and your continued interest in the journal.

Duncan Pay | Editor-in-Chief | [email protected]

EDITORIAL GEMS & GEMOLOGY SUMMER 2020 193

FEATURE ARTICLES

NATURALLY COLORED YELLOW AND

ORANGE GEM DIAMONDS:
THE NITROGEN FACTOR
Christopher M. Breeding, Sally Eaton-Magaña, and James E. Shigley

Natural yellow gem diamonds are the most common of the fancy-color diamonds, while orange diamonds are among
the rarest when they have unmodified hues. Both categories owe their coloration to atomic-level lattice defects as-
sociated with nitrogen impurities in the diamond structure. Four major groups of defects are responsible for the
color in nearly all yellow and orange diamonds: cape defects (N3 and associated absorptions), isolated nitrogen de-
fects, the 480 nm visible absorption band, and H3 defects. Nitrogen-bearing diamonds are thought to incorporate
isolated nitrogen during growth by substitution for carbon, meaning that natural diamonds start out with yellow to
orange color. However, only the very rare type Ib diamonds maintain that original color. With time at high tem-
peratures deep in the earth, the nitrogen atoms in most diamonds aggregate, resulting in either near-colorless stones
or yellow diamonds colored by cape defects. Yellow and orange diamonds can be grown in a laboratory or created
by color treatments, so a thorough understanding of the defects responsible for color in the natural stones is critical
for identification. Yellow diamonds serve as the best ambassador to the colored diamond world due to their abun-
dance and may be the only colored diamond many people will ever see in a jewelry store.

A
mong fancy-color diamonds, those with sat- in the diamond lattice during growth deep in the
urated blue, green, and red colors are the rarest earth. Nitrogen is the most common impurity in nat-
and generally the most highly valued. Over ural diamond due to the very similar atomic radii of
the last decade, however, diamonds with pure hues nitrogen and carbon atoms (155 and 170 picometer
in these colors have made up less than one-tenth of Van der Waals radii, respectively) as well as the rel-
one percent of all diamonds examined at GIA, mak- ative abundance of nitrogen in the growth environ-
ing them virtually unattainable in the marketplace.
In recent issues of Gems & Gemology, we have doc-
umented the gemological and spectroscopic proper-
ties of the rarest of fancy-color diamonds ranging
In Brief
from pink-to-red, blue, and green to the more un- • Among natural-color diamonds, those with a yellow
usual white and black. This article will address the hue are the most common.
most common colored diamonds, those with yellow • Diamonds with unmodified orange hues are among
hues, while also examining their much rarer orange the rarest of colored diamonds.
cousins (figure 1). This is the last of the fancy color • Yellow and orange diamonds owe their color primarily
groups in this series, and a brief summary of all the to nitrogen impurities.
colored diamond groups is provided at the end of the • Identification of treatment in yellow and orange dia-
article. monds requires careful evaluation of the defects re-
Yellow and orange diamonds owe their color pri- sponsible for the color.
marily to nitrogen impurities that are incorporated

ment. If nitrogen is present when diamond grows, it

See end of article for About the Authors and Acknowledgments.
GEMS & GEMOLOGY, Vol. 56, No. 2, pp. 194–219,
will inevitably be incorporated. Nitrogen is the sixth
http://dx.doi.org/10.5741/GEMS.56.2.194 most abundant element in the universe and accounts
© 2020 Gemological Institute of America for more than 75% of the earth’s atmosphere (Bebout

194 YELLOW AND ORANGE DIAMONDS: THE NITROGEN FACTOR GEMS & GEMOLOGY SUMMER 2020
98% of natural gem diamonds recovered (both color-
less and fancy-color) are type Ia, meaning they con-
tain FTIR-measurable concentrations of nitrogen
impurities, thus proving that nitrogen is present in
most diamond growth environments.
Nitrogen can exist as single isolated atoms substi-
tuting for carbon (i.e., C-centers, single substitutional
nitrogen, or type Ib), aggregated groupings of atoms in
pairs or fours (i.e., A-centers or type IaA, and B-centers
or type IaB), or as more complex defects (e.g., N3, H2,
H3, H4, and NV) (figure 2; see Collins, 1997, 1999,
2001 and Breeding and Shigley, 2009 for more infor-
mation). Some of these nitrogen defects cause dia-
mond to absorb light in the visible range to cause
color (C-centers, N3, H3, and NV) while others pro-
duce no color at all in diamond but exhibit character-
Figure 1. Yellow is the most common of the fancy- istic infrared absorptions (A- and B-centers). Simply
color diamonds, and its close relative orange is having nitrogen present in the diamond lattice is not
among the rarest. Photos by GIA staff. enough to produce a beautiful and rare fancy-color
yellow or orange diamond. The nitrogen must exist
in high enough concentrations and be configured in a
et al., 2013). While the amount of nitrogen in the way that selectively absorbs light toward the blue end
earth’s mantle is significantly lower than that, about of the visible spectrum substantially enough to pro-

Figure 2. Atomic lattice defects cause different colors in diamond. Their configuration makes a significant differ-
ence in the color. Photos by GIA staff.

Type IIa Type IIb

C C C C C C

C C C C B C

C C C C C C

Type Ib (C-centers) Type IaA (A-centers) Type IaB (B-centers)

C C C C C N C N C

C C N C C N N V N

C C C C C C C N C

Time / Temperature

N3 H3

C N C C N C

C V N C V C

C N C C N C

YELLOW AND ORANGE DIAMONDS: THE NITROGEN FACTOR GEMS & GEMOLOGY SUMMER 2020 195
A VISIBLE ABSORPTION SPECTRA

N Cape peaks

ABSORBANCE
Cape defects
(N3 shown in
diagram) N
N

400 450 500 550 600 650 700

B
C-center absorption

Isolated ABSORBANCE
nitrogen N
Figure 3. Yellow and or-
ange diamonds are col-
ored by four main
400 450 500 550 600 650 700 mechanisms: cape de-
fects, isolated nitrogen,
C 480 nm band the 480 nm band, and
H3 defects. Photos by
GIA staff.
ABSORBANCE

Structure
480 nm band not
known

400 450 500 550 600 650 700

D
H3 absorption
H3
N
ABSORBANCE

H3 defects
(N-V-N) N
H3 green
fluorescence
400 450 500 550 600 650 700
WAVELENGTH (nm)

duce yellow or orange color. However, the defects During the last decade, GIA has examined several
that are responsible for yellow and orange color are hundred thousand naturally colored diamonds with
not always the same. Most natural yellow gem dia- yellow or orange hue components, more than two-
monds are colored by either the N3 defect and its as- thirds of which had pure, unmodified yellow hues.
sociated absorptions (collectively termed “cape” In stark contrast to yellow diamonds, which are by
diamonds by gemologists), or by C-centers (figure 3, far the most common fancy color, less than one-
A and B). Other color centers such as the 480 nm vis- tenth of a percent of the stones examined over the
ible absorption band, H3, and H-related defects occa- same period had pure, unmodified orange hues. To
sionally contribute to the color components in yellow our knowledge, data from such a quantity of yellow
diamonds as well (figure 3, C and D). Fancy orange and orange diamonds has never before been compiled
diamonds, however, are nearly always colored by the in a publication. As with the previous articles in this
480 nm visible absorption band or by C-centers (figure series (Breeding et al., 2018; Eaton-Magaña et al.,
3, C and B). Orange color is also created occasionally 2018a, 2018b, 2019), our goal is to provide a detailed
as a primary hue in diamonds colored by the H3 de- account of the gemological and spectroscopic char-
fect in combination with a 550 nm visible absorption acteristics of natural yellow and orange diamonds
band or by nitrogen-vacancy defects. colored by different nitrogen-related defects.

196 YELLOW AND ORANGE DIAMONDS: THE NITROGEN FACTOR GEMS & GEMOLOGY SUMMER 2020
A VISIBLE ABSORPTION SPECTRA
ABSORBANCE

Isolated nitrogen orange

diamond

Figure 4. Each particu-

lar defect absorption
480 band yellow diamond
determines the dia-
mond color. A: If the
400 450 500 550 600 650 700 absorption extends to
WAVELENGTH (nm) ~600 nm, an orange
color is seen; otherwise,
B most nitrogen color
centers produce yellow
color. B: Different con-
centrations of the same
defect can generate yel-
low or orange color.
ABSORBANCE

Higher isolated nitrogen

orange diamond

Lower isolated nitrogen

yellow diamond

400 450 500 550 600 650 700

WAVELENGTH (nm)

For the spectroscopic and abundance evaluations, bodycolors in diamond are produced by a combination
we used a random, representative sampling of 500 of all transmitted visible light above a certain wave-
diamonds, including both yellow and orange stones. length. Yellow color is seen when most of the ni-
For spectra collection details, please see table S-1 in trogen-related absorptions occur primarily in the blue
Breeding et al. (2018). This article is not intended to part of the visible spectrum at wavelengths less than
be a comprehensive review of nitrogen in gem dia- ~510 nm, whereas orange color is seen when the ab-
mond. Instead, it is aimed at helping those in the dia- sorptions extend to ~600 nm (figure 4A). Different de-
mond trade to better understand these beautiful, fects absorb slightly different energy parts of the
more common, fancy yellow-colored gems that are visible spectrum, meaning that the main defects re-
far more likely to cross their desks than many of the sponsible for yellow color are somewhat different than
ultra-rare colored diamonds represented in the pre- those for orange color. Of course, the concentration of
vious articles in this series. each defect can range from low to high, and thus the
absorption band can shift in wavelength location (fig-
CAUSE OF COLOR ure 4B), and other defects due to plastic deformation
Yellow and orange diamonds are closely linked by may impact the color as well (Hainschwang et al.,
their dependence on nitrogen-related defects to pro- 2013). This means that the color produced depends
duce their colors. In fact, the main difference between not only on the defects present but also their abun-
the two colors is a slight shift in the transmission win- dance. For example, low concentrations of nitrogen
dow in the visible spectrum. Both orange and yellow defects can produce pale yellow color, but higher con-

YELLOW AND ORANGE DIAMONDS: THE NITROGEN FACTOR GEMS & GEMOLOGY SUMMER 2020 197
TABLE 1. Mechanisms of yellow and orange color in diamond.

Cause of color Defect(s) responsible Most common colors Geological formation Key gemological observations
Cape defects N3, N2 (478 nm), 451 nm, Yellow, brownish yellow Aggregation of N over time in Strong blue fluorescence; cape lines in
463 nm mantle spectroscope
Isolated nitrogen C-centers Yellow, orangy yellow, orange Incorporated at growth Oriented needle inclusions
480 nm band 480 nm visible absorption band Yellowish orange, orange, yellow Unknown Platy inclusions; yellow or orange
fluorescence
H3 defects H3, 550 nm absorption band Yellow, greenish yellow, Heating in mantle (H3); Localized or dispersed green fluorescence
brownish orange plastic deformation (550 nm) with focused lighting; brown graining

centrations can increase the overall absorption and a vacancy; ZPL [zero phonon line] = 415 nm) and its
thus increase the amount of yellow color and occa- associated absorptions at 451, 463, and 478 nm (N2
sionally shift the transmission window into the or- defect; Davies, 1981) (figure 3A). These absorption
ange region (figure 4B). Here we discuss the color features are associated with “cape” diamonds be-
centers in order of abundance for fancy yellow dia- cause they can be seen as a pattern of sharp absorp-
monds. The order of importance for orange diamonds tion lines when the diamonds are viewed with a
is slightly different and noted in each section below. gemological spectroscope. The term “cape” was orig-
Table 1 provides a summary of each of the major inally used to refer to pale to deeply colored yellow
groups. stones derived from the former Cape Province in
South Africa, which was well known for its yellow
“Cape” Defects. The majority of fancy yellow dia- diamonds (Bruton, 1971; Liddicoat, 1993; King et al.,
monds seen at GIA over the last decade (~74%) are 2005). The term now refers to diamonds from any lo-
colored primarily by a combination of absorptions cation with color saturations typically in the light
from the N3 defect (three nitrogen atoms adjacent to yellow range that are colored by the N3-related ab-

Figure 5. Each of the four groups of color mechanisms for yellow and orange diamonds shows a slightly different
range in hues. Photos by GIA staff.

Increasing brown associated with hydrogen

Cape defects

Increasing orange Increasing brown from plastic deformation

Isolated nitrogen

Increasing orange

480 nm band

Increasing green from H3 fluorescence Increasing orange

H3 defects H3 + 550 nm band

198 YELLOW AND ORANGE DIAMONDS: THE NITROGEN FACTOR GEMS & GEMOLOGY SUMMER 2020
Large cape yellow diamond
Melee-size isolated nitrogen yellow diamond

Path length of light in

diamond is very short, yet
diamond has more color

Path length of light in

diamond is very long, yet
diamond is pale yellow

In simple terms:
Absorbance = Path length of light × concentration of color-causing defect × molar absorption coefficient of the defect

Figure 6. The color of a diamond is controlled by the light that is absorbed and transmitted. Absorbance is a product
of the path length of light, the concentration of defects, and the absorption properties of the defects. C-centers are
much more efficient absorbers than cape defects, and thus most small yellow diamonds are colored by C-centers.

sorptions listed above (figure 5). We will collectively 2 and 3B; C-centers) are the second most common
refer to these N3-related absorptions as cape defects cause of color in natural yellow diamonds, comprising
in this article. The cape defects are also responsible just over 13% of the stones evaluated. C-centers are
for the slight amount of color in most diamonds in very strong absorbers, requiring concentrations of only
the D-to-Z range of the GIA color grading scale (some a few parts per million (ppm) of isolated nitrogen
D-to-Z diamonds have brownish tints from plastic atoms to produce strong yellow color (Collins, 1982,
deformation instead of nitrogen impurities; King et 2001; Hainschwang et al., 2013) (figure 6). These dia-
al., 1994, 2005, 2008). Cape defects are largely as- monds are often referred to as “true canaries” in the
sumed to be relatively weak light absorbers, meaning industry and are popular in the diamond trade (Liddi-
that it takes high concentrations of the defects (or coat, 1976; Collins, 1980; Wang and Poon, 2018).
longer path lengths of light travel, as with larger fac- While the yellow diamonds analyzed for this study
eted diamonds) to produce a saturated yellow color were primarily larger than melee sizes, it should be
(figure 6). For G-color and N-color diamonds of the mentioned that nearly all melee-sized natural dia-
same size and cut, for example, the N color will have monds (<0.20 ct) with Fancy yellow or more intense
significantly more cape defects. Similarly, a fancy color grades are colored by C-centers due to their very
yellow cape diamond of the same size and cut will strong light absorption. The short absorption path
have a higher concentration of the defects. Orange length in a melee stone requires that C-centers be
color is usually not produced by cape defects. present to produce yellow color, as cape defects are sig-
In about 15% of the yellow diamonds colored by nificantly weaker absorbers and cannot produce much
cape defects, we observed significant absorption con- yellow color over very short path lengths (again, see
tributions from hydrogen-related defects (see Breed- figure 6). C-centers are rare in natural diamond, as
ing et al., 2018, for details on H-related defects). they are typically destroyed by combining with other
Often the yellow diamonds with substantial hy- C-centers to form pairs of nitrogen atoms (A-centers)
drogen showed brown or green color components during the normal nitrogen aggregation process that
from the combination of absorptions from nitrogen occurs over geologic time during residence deep in the
and hydrogen-related impurities (figure 5). earth (figure 2; Anderson, 1961; Davies, 1970). While
C-centers produce strong yellow color in diamond, A-
Isolated Nitrogen. Single isolated atoms of nitrogen centers produce no color at all. Higher concentrations
substituting for carbon in the diamond lattice (figures of C-centers can produce deeper yellow colors (though

YELLOW AND ORANGE DIAMONDS: THE NITROGEN FACTOR GEMS & GEMOLOGY SUMMER 2020 199
Figure 7. When gently heated, diamonds colored by a 480 nm band will temporarily change color to a more orangy
hue. Photos by Wuyi Wang.

not always) and orangy components to the diamond chameleon diamonds (Hainschwang et al., 2005;
hue as well (figure 5). C-centers, in conjunction with Breeding et al., 2018), which always exhibit the 480
plastic deformation (Hainschwang et al., 2013), are the nm band (Scarratt, 1984; Fritsch et al., 2007b).
most common cause of primary orange color compo-
nents in diamond. H3 Defects. The H3 defect is the next most common
cause of color in yellow diamonds. H3 is an un-
480 nm Visible Absorption Band. One of the more in- charged defect consisting of two nitrogen atoms ad-
teresting mechanisms for producing yellow color in jacent to a vacancy in the diamond lattice with a ZPL
diamond is a broad absorption band centered at ~480 at 503.2 nm (Collins, 1982, 2001; Shigley and Breed-
nm in the visible spectrum (figure 3C). The makeup ing, 2013b, and references therein; figure 3D). H3 ab-
and structure of the 480 nm band is not well under- sorption can, by itself, produce yellow color and does
stood by scientists. Our observations indicate that it so in about 2% of our spectroscopic sample set (figure
is almost always associated with both A-centers (ni- 5). H3 also has a luminescence component that pro-
trogen pairs) and very low concentrations of C-centers. duces green fluorescence to ultraviolet (UV) and vis-
The defect has been attributed by some authors to ible light (again, see figure 3). As discussed in
substitutional oxygen atoms in the diamond lattice Breeding et al. (2018), if the H3 concentration is high
based primarily on modeling results (Gali et al., 2001; enough and the A-center concentration is relatively
Shiryaev et al., 2010). Based on a decade of GIA data, low (to avoid fluorescence quenching), H3 can pro-
the broad absorption is responsible for the color in duce visibly green diamonds (Collins, 2001; Breeding
~5% of yellow diamonds. Among diamonds with or- et al., 2018). However, when A-center concentrations
ange hue components, the 480 nm band is the second are higher, H3 will produce yellow or greenish yellow
most common cause of color. However, for diamonds diamonds.
with pure (unmodified) orange hues, this feature is re- H3 also frequently combines with plastic defor-
sponsible for the color in ~86% of these ultra-rare and mation–related defects such as the 550 nm absorption
highly valued stones. The absorption band at 480 nm band (discussed in Eaton-Magaña et al., 2018b, as the
also has thermochromic properties, meaning the main cause of color in pink diamonds) to produce
colors are affected by temperature changes. When a more orangy yellow to orange colors, as we see in ~4%
diamond with this feature is heated to 400°–500°C, of the samples surveyed (figures 3 and 5). In fact, this
the absorption band typically broadens and creates a combination of absorptions due to H3 and the 550 nm
temporary color change to more orangy hues (figure band is the only other significantly occurring mecha-
7). This effect also contributes to the color change in nism to produce diamonds with primary orange hues.

200 YELLOW AND ORANGE DIAMONDS: THE NITROGEN FACTOR GEMS & GEMOLOGY SUMMER 2020
OCCURRENCE AND FORMATION The prevalence of nitrogen in the earth, and dia-
Diamonds grow deep in the earth’s mantle, and by mond’s affinity for nitrogen, means that most areas
the time they are in a faceted form they generally of diamond formation have readily available nitrogen
have no remaining unique characteristics from the that will be incorporated during diamond growth.
geographic region (i.e., country of origin) in which The main factor that controls whether nitrogen-bear-
their kimberlites came to the earth’s surface. Non- ing diamonds are yellow tends to be how long they
etheless, certain mines or regions have produced reside at high temperatures deep in the earth (to
many of the blue (Cullinan mine, South Africa), allow nitrogen atoms to aggregate), which has little
green (Brazil and Guyana), and pink to red diamonds to do with geographic regions where kimberlites
(Argyle mine, Australia; Lomonosov mine, Russia; erupt (figure 8). Consequently, fancy-color natural
Smit and Shor, 2017; Breeding et al., 2018; Eaton- yellow and orange diamonds are recovered in nearly
Magaña et al., 2018a, 2018b). This correlation has all mining areas around the world. A few localities
more to do with the unique geological environment are well known for producing high-quality fancy-
required to produce those colors. For example, most color yellow diamonds of different types. Type Ia
blue diamonds require a source of boron (uncommon cape diamonds were named for their occurrences in
in the earth’s mantle) and are thought to grow much South Africa, and the Ellendale mine in Australia has
deeper in the earth than most other diamonds (Smith been a very important source for these types of fancy
et al., 2018). Pink to red diamonds require extensive yellow diamonds (Smit, 2008). Type Ib yellow dia-
plastic deformation, while green diamonds usually monds (containing C-centers) are relatively rare be-
require extensive residence time in near-surface al- cause most diamonds stay at depth in the earth at
luvial settings with the presence of radioactive fluids. high temperatures for prolonged periods of time,

Figure 8. Some color centers are created during diamond growth in the mantle, while others are acquired after long
residence times at high temperatures or during plastic deformation that might occur due to mantle flow or during
initial stages of transport in kimberlite.

Yellow Diamonds Orange Diamonds

H3 defects created Kimberlite H3 defects created

by long residence in transport by long residence in
mantle combined mantle combined
with 550 nm band with 550 nm band
from plastic from plastic
deformation that Plastic deformation deformation that
occurs either during occurs either during
mantle residence or mantle residence or
early in transport early in transport
(type Ia diamond)
Mantle flow (type Ia diamond)

Nitrogen aggregates
N3 and other cape
defects during long 480 nm band creation
residence in mantle. is unknown; possibly at
H3 defects also may growth by incorporation
form (type Ia Isolated nitrogen incorporated during growth of oxygen atoms (type
diamond) (type Ib diamond) Ia diamond)

YELLOW AND ORANGE DIAMONDS: THE NITROGEN FACTOR GEMS & GEMOLOGY SUMMER 2020 201
Figure 9. Famous orange diamonds include The Orange (top, 14.82 ct; courtesy of Christie’s) and the Pumpkin
(bottom, 5.54 ct) photo by Shane F. McClure). Middle row, left to right: Well-known yellows include the Arctic
Sun (30.54 ct; photo by Jae Liao), the Tiffany Yellow (128.54 ct; photo by Carlton Davis), and the Incomparable
(407.78 ct; photo by Jae Liao). Composite image, photos not to scale.

causing preexisting C-centers to aggregate to A- and diamonds are considerably rarer than yellows, par-
B-centers. A few locations, however, such as the ticularly unmodified pure orange hues, but a few fa-
Ekati and Diavik mines in Canada and the Zimmi mous ones exist. The largest is The Orange, a Fancy
deposit in Sierra Leone, are well known for their un- Vivid orange diamond of unknown origin weighing
usually large type Ib yellow diamonds (Shigley and 14.82 ct that sold at auction in 2013 for ~US$2.4 mil-
Breeding, 2013a; Smit et al., 2016, 2018). lion per carat, a record price at the time (DeMarco,
Many very large and famous yellow and orange 2013). Another well-known orange diamond is the
diamonds have been unearthed and sold for strong Pumpkin, a 5.54 ct Fancy Vivid orange stone from
prices at auction (figure 9). One of the largest yellow the Central African Republic purchased the day be-
diamonds is the 407.48 ct Fancy Deep brownish yel- fore Halloween in 1997 for US$1.3 million (Goldberg,
low (and Internally Flawless) Incomparable diamond, 2015).
discovered in the Congo. Other famous yellow dia-
monds include the 128.54 ct Tiffany Yellow (South GEMOLOGICAL OBSERVATIONS
Africa), the 132.27 ct Florentine (India), the 127 ct Fancy yellow and orange diamonds have a few dis-
Sun of Africa (South Africa), the 101.29 ct Allnatt tinctive features that can be seen under magnifica-
(likely South Africa), the 30.54 ct Arctic Sun (Ca- tion with a gemological microscope (table 1 and
nada), and the 4.25 ct Kahn Canary (Arkansas, figure 10). As there are different mechanisms for pro-
United States) (Thompson, 2004; Manutchehr-Danai, ducing yellow and orange colors, the various groups
2013; http://famousdiamonds.tripod.com). Orange have somewhat differing characteristics.

202 YELLOW AND ORANGE DIAMONDS: THE NITROGEN FACTOR GEMS & GEMOLOGY SUMMER 2020
Cape Defects. Cape yellow diamonds tend to have tures than the other groups. Under magnification, it
even color distributions and often contain crystal in- is common to see somewhat dense clouds of particle
clusions, much like colorless and near-colorless dia- inclusions or groupings of oriented needles in dia-
monds. Generally, there is nothing unique about monds colored by isolated nitrogen (figure 10, A and
cape diamonds under magnification. Those with sig- B). These clouds and needles are often in particular
nificant hydrogen-related defects will sometimes zones of yellow or orange color that indicate the dis-
contain finely dispersed particle clouds (figure 10D) tribution of C-centers in a diamond. With the spec-
that may be zoned in a way that creates a cross-like troscope, only a strong general absorption in the blue
appearance (Wang and Mayerson, 2002; Rondeau et end of the spectrum can usually be seen. These dia-
al., 2004). The cape absorption features (415, 451, monds also show rather consistent fluorescence re-
463, and 478 nm) can often be seen as dark lines in a actions. To both long-wave and short-wave UV, type
handheld or desk-model spectroscope, and their flu- Ib diamonds are typically inert. If they show any flu-
orescence tends to be remarkably consistent, with a orescence, it is usually weak orange, yellow, or green.
moderate to strong blue or yellow reaction to long-
wave UV (365 nm) and a weak to moderate yellow to 480 nm Visible Absorption Band. Yellow and orange
orange reaction to short-wave UV (254 nm) light. diamonds colored by the 480 nm band tend to be
variable when viewed in the microscope. Quite often
Isolated Nitrogen. Type Ib yellow and orange dia- the color looks uniform in faceted stones, but there
monds often show more distinctive gemological fea- are subtle zones of yellow or orange intermixed with

A B

Figure 10. Some yellow

and orange diamond
groups show distinctive
inclusions such as color-
zoned particle clouds
(A) and oriented needles
(B) in type Ib diamonds,
oriented platy inclu-
sions in 480 nm band
C D diamonds (C), and pat-
terned clouds in cape
diamonds with hy-
drogen (D). Photomicro-
graphs by GIA staff.
Field of view 3.0 mm
(A), 3.5 mm (B), 2.5 mm
(C), and 5.0 mm (D).

YELLOW AND ORANGE DIAMONDS: THE NITROGEN FACTOR GEMS & GEMOLOGY SUMMER 2020 203
WEIGHT DISTRIBUTION IN YELLOW/ORANGE DIAMONDS SHAPE DISTRIBUTION
20

Other Round
14% 13%
15
% OF DIAMONDS

Pear
12%

10
Cushion
33% Cut-cornered
rectangle
19%
5

Oval
9%

0
0 1 2 3 4 5 6 7 8 9 10 >10

CARAT WEIGHT

Figure 11. Most yellow and orange diamonds seen at GIA have weighed between one and two carats and been fac-
eted into cushion or cut-cornered rectangle shapes.

near-colorless diamond that only show well from flu- brown grainlines caused by plastic deformation with
orescence differences. Diamonds in this group are localized green luminescence from H3 may occa-
commonly very included and sometimes show platy sionally be visible with the microscope. The green
oriented groups of inclusions that are somewhat dis- luminescence is not usually strong or widely distrib-
tinctive (figure 10C). A broad absorption band from uted enough to affect the bodycolor.
450 to 500 nm can often be seen as a dark patch in
the spectroscope. The fluorescence from 480 nm LABORATORY SUBMISSIONS AND GRADING
band diamonds is very distinctive, with moderate to Of the yellow and orange diamonds submitted to
strong yellow or orangy yellow reactions to long- GIA over nearly a decade, 40% were in the one- to
wave UV and weak to moderate yellow reactions to two-carat size range (figure 11, left). The overall size
short-wave UV. These diamonds may phosphoresce distribution included 29% weighing less than one
a weak to moderate yellow after the short-wave UV carat and 28% weighing from two to ten carats.
lamp is turned off. About 3% of the stones exceeded 10 ct. Unlike D-to-
Z diamonds, where round shapes are most common,
H3 Defects. Yellow diamonds colored by the H3 de- the yellow and orange diamonds were most often cut
fect can show a range of gemological properties. into cushions (33%) or cut-cornered rectangles (19%)
There are no characteristic inclusions, and color dis- (figure 11, right). Rounds (13%), pears (12%), and
tribution may be uniform or consist of yellow band- ovals (9%) were the next most common shapes. Clar-
ing. Under fiber-optic illumination in the ity grades covered the entire grading range, and no
microscope, these bands (or random areas within the discernible trends were noted.
stone) may show green luminescence from the H3 The color grade distribution was strongly skewed
defect. With the spectroscope, a broad absorption toward unmodified yellow hues (~79%), followed by
from H3 can be seen at wavelengths < 500 nm and, if greenish yellows with occasional gray or brown com-
the stone is cooled with a spray refrigerant or in liq- ponents (~9%), combinations of orange and yellow
uid nitrogen, the ZPL at 503.2 nm may be discernible with occasional brown components (i.e., orange-yel-
with a spectroscope. The fluorescence reactions of low, yellowish orange, brownish orangy yellow, etc.;
H3-dominated diamonds tend to be green to both ~6%), brownish yellows (~5%), brown-oranges (~1%),
long-wave and short-wave UV. For orange diamonds and finally the extremely rare unmodified orange
colored by a combination of H3 and the 550 nm hues (0.05%). In addition, the color saturation was
band, most of the features are the same. However, strongly skewed toward Fancy Light and greater,

204 YELLOW AND ORANGE DIAMONDS: THE NITROGEN FACTOR GEMS & GEMOLOGY SUMMER 2020
with Fancy the most common. The reason for this yellowish orange” would be primarily orange and
skew is that yellow diamonds with unmodified hues “orange” alone would be termed as an unmodified
in the Faint, Very Light, and Light ranges are classi- hue here) to the distribution of the unmodified yel-
fied on the D-to-Z grading scale and are therefore not low and orange hues. Among the primarily yellow
included in this dataset (King et al., 1994, 2008). diamonds, ~81.5% were colored by cape defects
As mentioned earlier, a random sampling of 500 (sometimes including hydrogen), ~10.5% by isolated
natural yellow and orange diamonds was examined nitrogen (or type Ib), 3.3% by the 480 nm band, and
to assess the causes of color (figure 12). From these ~3% by H3 (with or without the 550 nm band; figure
data we see that cape defects (sometimes including 12). Stones with unmodified yellow hues had a very
hydrogen) accounted for almost three-quarters of the similar distribution of defects, with ~84.5% capes,
stones (73.6%; figure 12). Isolated nitrogen was the ~11% isolated nitrogen, ~3% 480 nm band, ~0.5%
coloring agent in 13.2%, followed by H3 (sometimes H3, and a few oddities (figure 12).
including 550 nm plastic deformation band) at 6.2%, The 500-stone data set was largely composed of
the 480 nm visible absorption band at 5%, and a few yellow diamonds, so in order to similarly evaluate
oddities making up the final 2%. While these data the primary and unmodified orange diamonds, data
describe the various causes of color in the group as a from GIA’s nearly decade-long intake of orange dia-
whole, they are significantly skewed toward the monds (both primary and unmodified, more than
much higher population of yellow diamonds. Inter- 13,000 stones) were screened for color origin. Orange
esting characteristics can be seen by comparing the diamonds, however, show a marked difference be-
causes of color in stones with yellow or orange as the tween those with primary and unmodified orange
primary color component (e.g., “brownish orangy hues. The causes of color in orange diamonds with
yellow” would be primarily yellow, while “brownish additional color components are nearly evenly dis-

Figure 12. Cape defects are by far the most common cause of fancy yellow color in diamond, including unmodified
yellow hues. Isolated nitrogen is the most common producer of orange hue components, though 480 nm bands
dominate the unmodified oranges.

Cause of All Yellow Primary Yellow Unmodified Yellow

and Orange Colors
Other Isolated Other Isolated
Other H3 defects H3 defects
H3 defects 1.8% nitrogen 0.8% nitrogen
2.9% 0.5%
6.2% 2.0% 10.5% 11.1%
480 nm 480 nm
3.3% 3.1%
Isolated
nitrogen
13.2%
5 .0%
m, Cape defects
0n Cape defects
48 81.5% 84.5%

Cape defects
73.6%

Primary Orange Unmodified Orange

Other, 4% Other, 1%

d
te n
Color Saturation Distribution la ge
Iso itro 3%
% OF DIAMONDS

50% n 1
40% Isolated
30% 480 nm band nitrogen
20% 46% 50%
10% 480 nm band
0% 86%
FNT VLT LIT FLT FAN INT VIV FDP FDK

TONE/SATURATION

YELLOW AND ORANGE DIAMONDS: THE NITROGEN FACTOR GEMS & GEMOLOGY SUMMER 2020 205
tributed between isolated nitrogen (~50%) and the diamonds, two additional relatively broad FTIR ab-
480 nm visible absorption band (46%), with the re- sorptions can usually be seen at 1430 and 1450 cm–1;
mainder (4%) consisting mostly of stones with H3 these are always proportional to each other in inten-
and plastic deformation and a few colored by NV sity. The 1450 cm–1 feature has a peak position very
centers (figure 12). Orange diamonds with pure, un- similar to that of the H1a interstitial radiation-re-
modified hues are extremely rare (< 200 seen at GIA lated defect (Collins, 1982; Zaitsev, 2003), but in cape
in the last decade) and strikingly consistent in their diamonds the feature is broader and unrelated to ra-
cause of color, with ~86% from 480 nm bands, ~13% diation exposure. For cape diamonds with hydrogen
from isolated nitrogen, and less than 1% from other impurities, a peak at 3107 cm–1 (three nitrogen atoms
mechanisms (again, see figure 12). From these data, adjacent to a vacancy with a hydrogen atom; Goss et
it is apparent that some unique aspect of the 480 nm al., 2014; figure 13A) and related modes at 1405 and
visible absorption band contributes to a purer, un- 4495 cm–1 are ubiquitous and generally vary in inten-
modified orange hue. sity with hydrogen content.
Yellow and orange diamonds colored by isolated
ABSORPTION SPECTROSCOPY nitrogen always show the related features at 1344
Absorption spectroscopy is a powerful tool for ex- and ~1130 cm–1, but otherwise the FTIR spectra are
amining colored diamonds. This nondestructive somewhat variable. These occur as type Ib diamonds
method directly measures the major impurities in dominated by C-center absorption, type IaA + Ib dia-
the diamond lattice (infrared absorption, measured monds with a combination of A-center and C-center
at room temperature) as well as defects that are re- absorptions (figure 13B), or diamonds dominated by
sponsible for color (visible absorption, measured at A-aggregates with trace concentrations of C-centers.
Because C-centers are such strong absorbers, it only
~77 K; i.e., liquid nitrogen temperature). Each of the
takes a few ppm to produce strong yellow color.
different causes of color in yellow and orange dia-
Consequently, the amount of isolated nitrogen is
monds has characteristic features that can be identi-
often quite low in both types (usually <100 ppma;
fied using absorption spectroscopy. Here we present
concentrations calculated using a modified version
representative spectra from each of the groups to il-
of a deconvolution spreadsheet provided by Dr.
lustrate the differences.
David Fisher of DTC Research Center, Maidenhead,
UK; Boyd et al., 1994, 1995). Very special geological
Infrared Absorption (FTIR). Natural fancy-color yel- conditions that are not completely understood are
low diamonds colored by cape defects have remark- required for C-centers to be preserved at all in natu-
ably consistent FTIR spectra, regardless of color. All ral diamonds. A-center concentration ranges from a
of these diamonds are type Ia with extremely high few ppma to a few thousand ppma and may be
concentrations of nitrogen impurities (often >1000 higher or lower than the C-center concentration in
ppma total nitrogen) that cannot be fully resolved by different samples. Additional FTIR absorptions of
the detectors of FTIR instruments (i.e., the absorp- unknown structure at 1353, 1358, 1363, and 1373
tion peaks exceed the vertical scale of the graph; fig- cm–1 commonly occur in yellow and orange dia-
ure 13). Both A-centers (1282 cm–1) and B-centers monds with isolated nitrogen (figure 13B). Weak
(1175 cm–1) are present, but usually no C-centers H1b (4935 cm–1) absorption also occurs in many type
(1130 and 1344 cm–1) can be detected, which is typi- Ib diamonds. Hydrogen absorptions similar to those
cal of most type Ia diamonds. This is a product of the in cape diamonds sometimes occur, as well, along
natural aggregation process where C-centers (isolated with a series of tiny peaks in the 3000–3500 cm–1
N) aggregate to A-centers (paired N) and finally to B- range that are hydrogen-related and seem to cor-
centers (four N and a vacancy) with enough time at relate with a defect described as the Y-center (Hain-
high temperatures inside the earth (figure 2; Ander- schwang et al., 2012). We have only observed these
son, 1961; Davies, 1970). Cape diamonds also have 3000–3500 cm–1 features in diamonds with isolated
strong platelet peaks (variable position from ~1360 nitrogen. Finally, plastic deformation often occurs
to 1375 cm–1; Allen and Evans, 1981; Breeding and in yellow type Ib diamonds and is represented by
Shigley, 2009), as well as a series of three broad in- “amber center” absorptions (usually occurring at
frared absorptions at ~1490, 1525, and 1550 cm–1 that 4165, 4110, or 4065 cm–1; Massi et al., 2005), espe-
occur almost exclusively in diamonds with cape de- cially in brownish yellow stones or those with an
fects (figure 13A). In most fancy-color yellow cape olive greenish hue.

206 YELLOW AND ORANGE DIAMONDS: THE NITROGEN FACTOR GEMS & GEMOLOGY SUMMER 2020
A
IR ABSORPTION SPECTRA
3107 1550 1525

ABSORBANCE
1490
Cape defects

4000 3500 3000 2500 2000 1500 1000 1600 1560 1520 1480

B C-centers
1344 (C-centers)
A-centers
1358
ABSORBANCE

Amber 1353
center 1373
Isolated nitrogen

1380 1360 1340

4000 3500 3000 2500 2000 1500 1000

C A-centers
Hydrogen peaks
ABSORBANCE

480 nm band

1300 1200 1100 1000

4000 3500 3000 2500 2000 1500 1000

D B-centers

A-centers Amber
center
ABSORBANCE

H3 defects

4350 4250 4150 4050 3950

4000 3500 3000 2500 2000 1500 1000

WAVENUMBER (cm–1)

Figure 13. Each group of color mechanisms for yellow and orange diamonds has a few distinctive IR absorption
spectral features. Spectra are offset vertically for clarity.

Yellow and orange diamonds colored by a 480 nm light (luminesces) simultaneously. Diamonds colored
visible absorption band are nearly always type Ia and by H3 are always type Ia with variable proportions of
usually have low to moderate concentrations of A- A- and B-centers (figure 13D). For greenish yellow
centers, but sometimes they show very irregular and stones, the total nitrogen, as well as the A-center
distorted one-phonon regions in the infrared spec- concentration, is always relatively low to allow the
trum that are not well understood (figure 13C). In H3 green luminescence to be seen along with the yel-
some samples, trace concentrations of C-centers can low from absorption. In H3-dominated yellow dia-
also be detected. Hydrogen impurities are variable, monds with high A-center concentrations
but when present, they occur with the same peaks (saturated), the green luminescence from H3 is com-
mentioned above for isolated nitrogen yellows. pletely quenched (Collins, 1982, 2001), leaving only
Yellow diamonds colored by H3 defects are quite the yellow color from H3 absorption. Similar to cape
variable in their FTIR spectra. H3 absorbs and emits diamonds, the platelet, 1430 cm–1, and 1450 cm–1 fea-

YELLOW AND ORANGE DIAMONDS: THE NITROGEN FACTOR GEMS & GEMOLOGY SUMMER 2020 207
A VIS-NIR SPECTRA

480 nm band

Isolated nitrogen
ABSORBANCE

Figure 14. Each group of

color mechanisms for
yellow and orange dia-
Hydrogen defects monds has characteris-
N3 tic Vis-NIR absorptions
N2 that produce color. Dia-
Cape defects mond spectra can in-
400 500 600 700 800 900
clude multiple causes
WAVELENGTH (nm) of color, which can
create variations on
B their normally pro-
H3 duced colors. For exam-
ple, spectra with cape
features can include
hydrogen defects and
550 nm band H3 spectra can include
ABSORBANCE

the 550 nm absorption

band. Spectra are offset
vertically for clarity.

H3 fluorescence

400 500 600 700 800 900

WAVELENGTH (nm)

tures as well as occasional H-related features are seen ure 14A). The last three cape peaks listed only appear
in all H3 diamonds. In contrast, brownish yellow and with increasing saturation of the yellow color. The
orange diamonds colored by the combination of H3 observed yellow color comes primarily from the ab-
and a 550 nm visible absorption band are extremely sorptions from N2 and the 451 nm band, with only
consistent as type IaA diamonds with moderate minor contribution from the N3 itself, but all the
amounts of total nitrogen, a weak platelet peak, and cape peaks are proportional in intensity to N3. If hy-
the presence of amber centers (Massi et al., 2005; drogen-related defects are present, two broad asym-
Eaton-Magaña et al., 2018b) that mark the plastic de- metric bands centered at ~730 and ~836 nm often
formation related to the 550 nm band. occur along with a smaller peak at 563 nm (figure
14A), causing the color to have brownish or greenish
Ultraviolet-Visible/Near-Infrared Absorption (UV- components.
Vis-NIR). Much like their FTIR spectra, the UV-Vis- UV-Vis-NIR absorption in yellow and orange dia-
NIR spectra of natural yellow diamonds colored by monds primarily colored by isolated nitrogen is quite
cape defects are extremely consistent, with the N3 different. A very strong absorption by C-centers occurs
ZPL (Collins, 1982, 2001; Zaitsev, 2003; Shigley and in the UV range at ~270 nm (Jones et al., 2009) (not
Breeding, 2013b) at 415 nm, a related broad absorp- shown). Even with a few ppm of C-centers, the absorp-
tion band to shorter wavelengths, as well as the other tion is so strong that the absorption “tail” extends into
N3-related cape absorption features at (listed in order the visible range, creating an absorption continuum to
of intensity) 478 (N2), 451, 463, 435, and 426 nm (fig- ~510 nm in yellow hues and ~600 nm in orange-hued

208 YELLOW AND ORANGE DIAMONDS: THE NITROGEN FACTOR GEMS & GEMOLOGY SUMMER 2020
type Ib diamonds (figure 14A). This gradually increas- analysis is usually performed with the sample at liq-
ing absorption toward the UV is the only feature seen uid nitrogen temperature (~77 K) and can detect dia-
in the UV-Vis-NIR spectra of most of these diamonds. mond defects as low as a few parts per billion (ppb)
Yellow and orange diamonds colored by the 480 (Eaton-Magaña and Breeding, 2016). Each cause of
nm band show the so-named broad absorption cen- color in yellow and orange diamonds shows a few
tered at roughly 480 nm (figure 14A) along with distinctive features in the PL spectra.
another broad absorption at 370 nm and a much Natural yellow cape diamonds consistently ex-
weaker band at 427 nm (not shown). Occasionally a hibit a number of PL peaks including 415 (N3), 496
weak N3 peak at 415 nm can also be seen. The in- (H4), 503.2 (H3), 508, 535, 604, 660, 700, 787, 911,
tensity and, to some extent, the shape of the 480 nm 933, and 952 nm. Besides the strong N3 at 415 nm
band are variable and do not always correlate with that contributes to the cape absorption that causes
the intensity of yellow or orange color. More no- the yellow color, the most prominent tend to be the
tably, the 480 nm absorption extends only to ~550 700 and 787 nm features, which have been attributed
nm in nearly all of these stones, even in the orange to nickel, nitrogen, and likely hydrogen impurities
fancy-color diamonds. Absorption color calculations (Fritsch et al., 2007a; figure 15A). In cape diamonds
indicate that the resulting color should be mostly with abundant hydrogen, these two peaks tend to be
yellow in all cases, strongly suggesting that the lu- very strong, and the NE8 Ni-N defect (Yelisseyev and
minescence discussed in the next section accounts Kanda, 2007) consistently appeared at 793 nm.
for part of the color in the orange stones (Titkov et For yellow and orange diamonds colored by iso-
al., 2015). lated nitrogen, the following PL peaks are often de-
H3-dominated yellow diamonds produce rel- tected: 503.2 (H3), 525, 565.8, 575 (NV0), 578, 604, 637
atively simple UV-Vis-NIR absorptions with the H3 (NV–), 689, 698, 805, 830, 904, 953, and 986 nm (H2).
ZPL at 503.2 nm (Collins, 1982, 2001; Shigley and Of these, the most significant are the NV centers at
Breeding, 2013b) and its related absorption band to 575 and 637 nm (figure 15B) and the H2 defect at 986
shorter wavelengths as their major feature (figure nm. The abundance of C-centers, which are electron
14B). When plastic deformation is also present, a donors, tends to provide extra electrons to adjacent de-
broad 550 nm band and an underlying increase in fects to make them negatively charged (Collins, 1982,
overall absorption toward the UV caused by vacancy 2001). In type Ib diamonds under 514 or 532 nm laser
clusters accompany the H3 absorptions to extend the excitation, the 637 (NV–)/575 (NV0) ratio is almost al-
absorption edge to ~620 nm and produce brownish ways >1, and the H2 defect (two nitrogen atoms adja-
orange hues (figure 14B). Often, weak 535.8 nm and cent to a vacancy in a negative charge state) is
H2 (986 nm) absorption peaks are also seen in these ubiquitous due to the available electrons.
diamonds. PL spectra from 480 nm band stones are notice-
ably different from those exhibited by the other
groups. Between 550 and 700 nm, nearly a hundred
LUMINESCENCE SPECTROSCOPY
small PL peaks are sometimes observed with 488 nm
AND IMAGING laser excitation. Overall, the following peaks are
In addition to absorbing energy, many defects also commonly seen: 589, 592, 616, 619, 678, 683, 693,
emit light (or luminesce) when exposed to certain 709, 723, 799, 807, 817, 883/885, and 904 nm. Of
wavelengths of light. The two major forms of lumi- these, the 799 and 883/885 nm doublet are most
nescence used in gemology are referred to as photo- noteworthy. The 799 nm peak occurs in all 480 nm
luminescence and fluorescence, typically depending band diamonds evaluated, and the 883/885 nm dou-
on whether a laser or UV lamp is used for excitation blet is the well-studied 1.4 eV Ni center (Yelisseyev
and whether the collection occurs at liquid nitrogen and Kanda, 2007). More significant seems to be the
or room temperature. broad luminescence band in the red end of the spec-
trum, centered at ~670–700 nm, when a blue or green
Photoluminescence (PL). PL is a very sensitive, non- laser is used for excitation of some 480 nm band dia-
destructive analytical technique that involves expos- monds (figure 15C; first described in Collins and Mo-
ing a diamond to light of known energy (i.e., with a hammed, 1982). This band signifies red
wavelength in the UV, visible, or infrared range) and luminescence to visible light (from laser excitation
measuring any new light that is given off as a result in this case) that likely plays a role in the color of
of atomic-level interactions within the diamond. PL some orange diamonds. As discussed in the UV-Vis-

YELLOW AND ORANGE DIAMONDS: THE NITROGEN FACTOR GEMS & GEMOLOGY SUMMER 2020 209
A PL SPECTRA
488 nm PL
787
700

Raman
INTENSITY
Cape defects

500 550 600 650 700 750 800 850 900

WAVELENGTH (nm)

B 514 nm PL
637 (NV–)

575
INTENSITY

(NV0)
Raman

Isolated nitrogen

Figure 15. Each group of

color mechanisms for
550 600 650 700 750 800 850 900 950
yellow and orange dia-
WAVELENGTH (nm)
monds has a few dis-
C 514 nm PL
tinctive PL emission
features.

670–700 band
INTENSITY

Raman

480 nm band

550 600 650 700 750 800 850 900 950

WAVELENGTH (nm)

D 503.2 (H3) 488 nm PL

Raman
INTENSITY

H3 defects

495 500 505 510 515 520 525

WAVELENGTH (nm)

NIR section earlier, 480 nm band absorption alone world’s most beautiful orange diamonds (Titkov et
produces only yellow color. However, the combina- al., 2015).
tion of that absorption and this red luminescence is Yellow diamonds colored by H3 and orange dia-
likely the cause of orange color in the majority of the monds colored by the combination of H3 and the 550

210 YELLOW AND ORANGE DIAMONDS: THE NITROGEN FACTOR GEMS & GEMOLOGY SUMMER 2020
A

Cape defects

Isolated nitrogen
Figure 16. Each group of
color mechanisms for
yellow and orange dia-
monds displays differ-
ent aspects of their
fluorescence colors or
C patterns under deep-UV
excitation. Diamond-
View images by GIA
staff.
480 nm band

H3 defects

nm plastic deformation band both show relatively long-wave UV light. Here we further examine the
similar PL spectra, with emissions at 496 (H4), 503.2 deep-UV (<230 nm excitation) fluorescence reactions
(H3), 535.8, 575 (NV0), 637 (NV–), 657, 700, 741 (GR1), of the groups observed using a DiamondView instru-
804, 812, 953, and 986 (H2). Apart from the very ment. Due to the extreme absorption of UV at these
strong H3 PL peak (figure 15D), the only other no- and shorter wavelengths by diamond, this high-
table observation was that the 535.8 nm peak was energy excitation only stimulates fluorescence from
often much larger in stones with the 550 nm visible the surface of a diamond, thus providing a highly re-
absorption band (not shown). solved image of fluorescence distribution within dif-
ferent diamond growth events and sectors (Welbourn
Deep-UV Fluorescence. Earlier in the gemological et al., 1996).
observations section, we discussed the fluorescence Cape diamonds mostly show strong blue fluores-
reactions of each group to standard short-wave and cence to deep UV (due to N3 defects), often with dis-

YELLOW AND ORANGE DIAMONDS: THE NITROGEN FACTOR GEMS & GEMOLOGY SUMMER 2020 211
Natural Lab-Grown Treated Treated
HPHT Irradiation/Annealing HPHT-processed

Figure 17. Orange and yellow diamonds can occur naturally, be grown in a laboratory, or result from the treatment
of off-color natural diamonds. Photos by GIA staff.

tinctive growth bands forming multiple rectangles other fluorescence colors are common as well (figure
within each other from octahedral growth (figure 16C). These textures suggest multiple stages of dia-
16A). Occasionally a thin, bright green band of fluo- mond growth under different conditions with differ-
rescence from localized H3 defects is observed follow- ent defects incorporated. Their irregular patterns are
ing the growth pattern in some areas. Hydrogen markedly similar to chameleon diamonds, which
content has little effect on the deep-UV fluorescence. also exhibit the 480 nm band (Scarratt, 1984; Hain-
Yellow and orange type Ib diamonds often have schwang et al., 2005; Fritsch et al., 2007b), suggesting
irregular deep-UV fluorescence patterns, seemingly a genetic link.
suggesting more chaotic, possibly unstable growth H3-dominated yellow diamonds show very strong
environments recording periods of diamond growth green fluorescence (H3) that appears uniform or is
and resorption. These images show variable patterns isolated to lines of plastic deformation (figure 16D).
of green (H3), orange (NV0), or pink to red (NV–) fluo- Blue fluorescence (N3) is common as well, especially
rescence or some combination of those (figure 16B). in stones with localized green fluorescence. Most of
Oftentimes the red or green fluorescence follows the orange diamonds with H3 and 550 nm bands
lines of plastic deformation within the diamonds, show the green fluorescence associated with H3 very
suggesting that some of the fluorescence was created localized along the plastic deformation slip planes.
after diamond growth during deformation. Overall
fluorescence in type Ib diamonds is relatively low in IDENTIFICATION CONCERNS
intensity to deep UV. Identification of fancy-color diamonds involves deter-
Diamonds colored by the 480 nm band tend to be mining whether a stone grew in the earth or in a la-
even more irregular. Deep-UV imaging shows dark boratory, and whether the color we see is natural or
blue (N3), light blue (unknown defect), greenish yel- has been artificially created by treatment. Yellow dia-
low (correlated with the 480 nm band), and occa- monds (and those with orange hue components) can
sionally green (H3) fluorescence. Inert areas between easily be created in a laboratory using the high-pres-

212 YELLOW AND ORANGE DIAMONDS: THE NITROGEN FACTOR GEMS & GEMOLOGY SUMMER 2020
sure, high-temperature (HPHT) growth method (fig- of color enhancement that is harder to detect is HPHT
ure 17). As mentioned above, nitrogen is extremely treatment of type I diamonds. At pressures and tem-
abundant in the atmosphere and thus readily avail- peratures higher than those in the earth where most
able in an HPHT growth chamber unless strict meas- diamonds grow, HPHT treatment can cause the crea-
ures are taken to remove it. Under normal conditions tion of H3 defects as well as the breakup of aggregated
of growth, most HPHT lab-grown diamonds are yel- nitrogen into isolated nitrogen (Collins, 2001). Both
low to orange in color depending on how much ni- H3 and these newly created C-centers will give a dia-
trogen gets incorporated as C-centers. These mostly mond a yellow color (again, see figure 17). If the start-
type Ib diamonds show typical features related to ing material is a cape diamond, the addition of
their growth in a laboratory such as color zoning that C-centers during treatment can produce an orange
follows cuboctahedral growth patterns, particle color. HPHT treatment may produce graphitized
clouds, metallic flux inclusions, and a lack of birefrin- feathers or inclusions or frosted facets that are not
gence (direct measurement of lattice strain) under properly repolished afterward. Irradiation and anneal-
cross-polarized magnification (Eaton-Magaña et al., ing treatments may leave yellow or orange color zones
2017, and references therein). Natural yellow and or- near the culet. However, it is often impossible to ge-
ange diamonds colored by C-centers have absorption mologically identify these types of treatments, and all
spectra similar to those of their lab-grown counter- fancy-color diamonds should be sent to a laboratory
parts but can usually be separated by the presence of for conclusive determination of color origin.
natural inclusions, uniform to irregular color zoning,
wispy clouds, and a distinct strain pattern. On rare UNUSUAL EXAMPLES
occasions, both HPHT lab-grown and natural yellow
Yellow diamonds are the most common of the fancy-
diamonds can be type Ia with traces of isolated ni-
color diamonds and are dominated by the groups of
trogen, but the same separation criteria apply and lab-
color-causing defects described here. As with anything
grown diamonds generally do not have cape defects
in visible absorption. Chemical vapor deposition in nature, though, there are a few examples of unusual
(CVD) methods can be used to grow yellow diamonds yellow diamonds that do not perfectly fit the cate-
with isolated nitrogen, but it is rarely done and most gories. In our previous colored diamond articles, we
of the products lack the color saturation of HPHT- have presented a few odd samples in this section. Here
grown samples (Kitawaki et al., 2015; Eaton-Magaña we will discuss two other ways to produce yellow
and Shigley, 2016). CVD-grown diamonds can occa- color involving isolated nitrogen: type IaB yellow dia-
sionally have orange hue components if subjected to monds colored by isolated nitrogen and growth-zoned
post-growth irradiation and annealing treatments to yellow diamonds with type Ia cores and type Ib rims.
create relatively low concentrations of NV centers. For most natural diamonds, the aggregation path
Determining whether the yellow or orange color from C-centers to A-centers to B-centers is a one-way
in an earth-grown gem diamond is natural or artifi- route that usually progresses to some point with a
cially produced through treatment is a more challeng- combination of A- and B-centers. The transformation
ing endeavor. Cape-related defects, as a group, cannot from A- to B-centers is the point at which most N3
be added or removed through any known commercial defects are created in nature, and thus most cape dia-
treatment, so their presence is generally a good indi- monds fall somewhere therein. We have already dis-
cator that the diamond naturally exhibited some yel- cussed the unlikelihood of preserving C-centers in
low color. However, it is relatively easy to create H3 natural yellow diamonds, but not in the more ma-
defects through artificial irradiation and annealing turely aggregated type IaB diamonds. Occasionally
treatment that absorbs blue light to create yellow GIA receives a Fancy to Fancy Intense yellow dia-
color (figure 17). This can be done on an off-color mond that is type IaB with extremely weak C-center
brown diamond to make it yellow or even on a pale absorption and no cape defects. These unusual stones
yellow cape diamond to intensify the color. Natural are bothersome, as B-aggregates of nitrogen should
radiation exposure and heating in the earth can have not coexist naturally with C-centers. It is common to
similar effects, making the separation even more dif- disaggregate some nitrogen aggregates during HPHT
ficult. Extreme amounts of irradiation and annealing treatment (~2000°C, 6 GPa; Dobrinets et al., 2013) to
can turn most nitrogen-bearing diamonds to an orange form new C-centers, but these conditions do not
to brown color, but this degree of treatment is easily seem feasible for natural diamond-forming environ-
spotted using absorption spectroscopy. Another form ments, and the presence of B-centers with C-centers

YELLOW AND ORANGE DIAMONDS: THE NITROGEN FACTOR GEMS & GEMOLOGY SUMMER 2020 213
in an FTIR spectrum usually arouses suspicion that are often more indicative of a treated-color diamond
treatment is involved. However, there are a handful due to the mixed sampling. These diamonds repre-
of naturally intense yellow diamonds that are highly sent the larger version of the melee-sized Ekati
aggregated (mostly type IaB), but still have enough stones that have had much of their yellow outer rim
traces of C-centers to produce their saturated colors, removed during faceting. With careful examination,
perhaps indicating very rare, HPHT-like conditions in gemologists can conclusively identify these dia-
the earth. These diamonds usually have small monds as natural, but they are definitely unusual.
feathers on their natural surfaces from movement in
an alluvial environment and contain PL features in- COMPARISON OF ALL FANCY-COLOR
dicating they are naturally colored, despite similar- DIAMONDS
ities with HPHT-treated yellow diamonds. This article completes the fancy-color portion of this
Another unusual group of yellow fancy-color dia- series in Gems & Gemology. The final article will
monds seen at GIA and by Hainschwang et al. (2006) look principally at colorless diamonds and the rest of
displayed FTIR spectra with A-, B-, and C-centers the D-to-Z scale. We wish to conclude this exploration
that normally do not exist naturally. However, the
of fancy-color natural diamonds by providing a brief
diamonds seen at GIA were sourced from the Ekati
summary and comparison of all five diamond color
mine in Canada and were known to have never been
groups: green (Breeding et al., 2018), blue/gray/violet
treated (Lai et al., 2019). After breaking a couple of
(Eaton-Magaña et al., 2018a), pink/purple/red/brown
the melee-sized rough diamonds, we discovered they
(Eaton-Magaña et al., 2018b), white/black (Eaton-
consisted of near-colorless cores and yellow rims
Magaña et al., 2019) and yellow/orange (this article).
from two different episodes of diamond growth (fig-
Table 2 lists the major causes of color in descending
ure 18). FTIR analysis is a bulk measurement tech-
order of abundance within each of the color groups in
nique, meaning the spectrum recorded is an average
the diamonds seen at GIA.
of all of the path lengths within the diamond through
which the infrared beam passes. Thus, a diamond
like this will show both the A- and B-centers from CONCLUSIONS
the core along with the C-centers from the rim in the Yellow and orange are the largest group of fancy-color
same spectrum. Occasionally at GIA, we see larger diamonds and owe their color to four major causes as-
faceted yellow gem diamonds with only small sociated with nitrogen impurities: cape defects, iso-
amounts of yellow color concentrated near the sur- lated nitrogen, the 480 nm visible absorption band, and
face (enough to achieve a Fancy color grade) and the H3 defects (table 1). Cape defects are by far the most
remainder of the stone is colorless. We refer to these important color-producing mechanism for yellow dia-
as “yellow-skin” diamonds, but their FTIR spectra monds, accounting for nearly three-quarters of the dia-

Yellow rim Figure 18. Small yellow

type Ib
diamonds (~0.02 ct
each) from the Ekati
mine in Canada show
colorless type Ia cores,
surrounded by yellow
type Ib rims. IR absorp-
tion spectra from these
stones suggest the si-
multaneous presence of
A-, B-, and C-centers
due to mixing of both
zones during spectra
collection. Photo by
Colorless core
type Ia Mei Yan Lai.

214 YELLOW AND ORANGE DIAMONDS: THE NITROGEN FACTOR GEMS & GEMOLOGY SUMMER 2020
Figure 19. Intensely colored orange and yellow gem diamonds are rare and beautiful, both in appearance and com-
plexity of their origin of color. Left: A 4.08 ct Fancy Vivid orange diamond auctioned by Sotheby’s Hong Kong in
April 2016. Courtesy of Sotheby’s. Right: A yellow and white diamond necklace by Chatila containing a 54.29 ct
cushion-cut Fancy Intense yellow diamond. Photo by Robert Weldon.

monds evaluated, while isolated nitrogen and the 480 H3 or isolated nitrogen defects through HPHT treat-
nm band are most important for orange stones. Yellow ment or artificial irradiation and annealing. Cape de-
and orange diamonds are recovered from almost all dia- fects (as a group) and the 480 nm band, however, are
mond mines worldwide, but South Africa and the Cen- usually not present in lab-grown diamonds. Yellow dia-
tral African Republic have produced some of the most monds are the only natural colored diamonds many
famous and valuable examples. Yellow and yellowish people will ever encounter and serve as an important
orange lab-grown diamonds are easy to produce by the representative for the fancy-color diamond trade, while
HPHT growth method, and diamond color treatments orange diamonds with unmodified hues are among the
routinely create yellow and orange colors by creating rarest of all fancy-color diamonds (figure 19).

ABOUT THE AUTHORS ACKNOWLEDGMENTS

Dr. Breeding is a senior manager of analytics, Dr. Eaton-Magaña The authors thank GIA’s Tom Moses and John King for con-
is a senior manager of diamond identification, and Dr. Shigley is a structive comments on early versions of the article. GIA’s Clau-
distinguished research fellow, at GIA in Carlsbad, California. dia Jannucci and Judy Colbert provided valuable assistance
with images of famous diamonds.

YELLOW AND ORANGE DIAMONDS: THE NITROGEN FACTOR GEMS & GEMOLOGY SUMMER 2020 215
TABLE 2. Causes of color within each major fancy-color diamond group in this G&G series (based on stones submitted to GIA labs).

Percentage of
Representative Representative Vis-NIR
Cause of color diamonds affected
example of colorb spectrum
within color classa

ABSORBANCE
GR1 ZPL
GR1-related
Green (Breeding et al., absorption
GR1 62%
2018)
400 450 500 550 600 650 700 750
WAVELENGTH (nm)

N absorption
H-related

ABSORBANCE
absorption

Hydrogen defects 36%

400 450 500 550 600 650 700 750

WAVELENGTH (nm)

H3 absorption
H3 ZPL

ABSORBANCE
H3 29%
H3 green
luminescence

400 450 500 550 600 650 700 750

WAVELENGTH (nm)

ABSORBANCE
Nickel 1% Ni-related absorption

400 450 500 550 600 650 700 750

WAVELENGTH (nm)

ABSORBANCE
Blue/Gray/Violet (Eaton-
Magaña et al., 2018a) Boron 36%

400 450 500 550 600 650 700 750 800 850
WAVELENGTH (nm)

ABSORBANCE
551 835

Hydrogen defects 31%

400 450 500 550 600 650 700 750 800 850
WAVELENGTH (nm)

N3
GR1
ABSORBANCE

GR1 27%

400 450 500 550 600 650 700 750 800 850
WAVELENGTH (nm)
ABSORBANCE

Micro-inclusions 7%

400 450 500 550 600 650 700 750 800 850
WAVELENGTH (nm)
ABSORBANCE

Pink/Purple/Red/Brown
(Eaton-Magaña et al., 550 nm absorption band 99.5% 550 nm
2018b)
400 450 500 550 600 650 700 750 800
WAVELENGTH (nm)

216 YELLOW AND ORANGE DIAMONDS: THE NITROGEN FACTOR GEMS & GEMOLOGY SUMMER 2020
Percentage of
Representative example Representative Vis-NIR
Cause of color diamonds affected
of colorb spectrum
within color classa

ABSORBANCE
Pink/Purple/Red/Brown Brown absorption
(Eaton-Magaña et al., 31%
2018b) continuum
400 450 500 550 600 650 700 750 800
WAVELENGTH (nm)

NV– (637 nm)

NV0 (575 nm)

ABSORBANCE
NV centers 0.5%

400 450 500 550 600 650 700 750 800

WAVELENGTH (nm)

Fancy White/Black (Eaton- Nano-inclusions to White: 100%

Magaña et al., 2019) Black: 99.2% N/A
micro-inclusions

ABSORBANCE
Very high defect
Black: 0.8%
concentrations
500 550 600 650 700 750 800 850
WAVELENGTH (nm)

Cape peaks

ABSORBANCE
Yellow/Orange (this Cape defects 74%
article)

400 450 500 550 600 650 700

WAVELENGTH (nm)

C-center absorption
ABSORBANCE

Isolated nitrogen 13%

400 450 500 550 600 650 700

WAVELENGTH (nm)

480 nm band
ABSORBANCE

480 nm band 0.5%

400 450 500 550 600 650 700

WAVELENGTH (nm)

H3 absorption
H3 ZPL
ABSORBANCE

H3 defects 6%

400 450 500 550 600 650 700

WAVELENGTH (nm)

a
Some diamonds are affected by more than one cause of color (for example, a green diamond may be colored both by the GR1 defect and H3 green fluorescence);
therefore, percentages within each color class may sum to greater than 100%.
b
Diamonds with relatively saturated color are included here as examples; however, a significant quantity in each class show lighter colors.

YELLOW AND ORANGE DIAMONDS: THE NITROGEN FACTOR GEMS & GEMOLOGY SUMMER 2020 217
REFERENCES
Allen B.P., Evans T. (1981) Aggregation of nitrogen in diamond, in- Band of many colors. G&G, Vol. 54, No. 4, pp. 352–377,
cluding platelet formation. Proceedings of the Royal Society of http://dx.doi.org/10.5741/GEMS.54.2.352
London A, Vol. 375, No. 1760, pp. 93–104, Eaton-Magaña S., Ardon T., Breeding C.M., Shigley J.E. (2019) Nat-
http://dx.doi.org/10.1098/rspa.1981.0041 ural-color fancy white and fancy black diamonds: Where color
Anderson B.W. (1961) Nitrogen in diamond. The Gemmologist, and clarity converge. G&G, Vol. 55, No. 3, pp. 320–337,
Vol. 30, No. 355, pp. 21–22. http://dx.doi.org/10.5741/GEMS.55.3.320
Bebout G., Fogel M., Cartigny P. (2013) Nitrogen: Highly volatile Fritsch E., Hainschwang T., Massi L., Rondeau B. (2007a) Hy-
yet surprisingly compatible. Elements, Vol. 9, pp. 333–338, drogen-related optical centers in natural diamond: An update.
http://dx.doi.org/10.2113/gselements.9.5.333 New Diamond and Frontier Carbon Technology, Vol. 17, No.
Boyd S.R., Kiflawi I., Woods G.S. (1994) The relationship between 2, pp. 63–89.
infrared absorption and the A defect concentration in diamond. Fritsch E., Massi L., Rossman G.R., Hainschwang T., Jobic S.,
Philosophical Magazine B, Vol. 69, No. 6, pp. 1149–1153, Dessapt R. (2007b) Thermochromic and photochromic be-
http://dx.doi.org/10.1080/01418639408240185 havior of “chameleon” diamonds. Diamond and Related
———(1995) Infrared absorption by the B nitrogen aggregate in Materials, Vol. 16, No. 2, pp. 401–408,
diamond. Philosophical Magazine B, Vol. 72, No. 3, pp. 351– https://doi.org/10.1016/j.diamond.2006.08.014
361, http://dx.doi.org/10.1080/13642819508239089 Gali A., Lowther J.E., Deak P. (2001) Defect states of substitutional
Breeding C.M., Shigley J.E. (2009) The “type” classification sys- oxygen in diamond. Journal of Physics: Condensed Matter, Vol.
tem of diamonds and its importance in gemology. G&G, Vol. 13, No. 50, pp. 11607–11613, http://dx.doi.org/10.1088/0953-
45, No. 2, pp. 96–111, http://dx.doi.org/10.5741/GEMS.45.2.96 8984/13/50/319
Breeding C.M., Eaton-Magaña S., Shigley J.E. (2018) Natural-color Goldberg W. (2015) Famous Diamonds – Part 3: The Pumpkin
green gem diamonds: A beautiful conundrum. G&G, Vol. 54, diamond (http://www.williamgoldberg.com/house-of-gold-
No. 1, pp. 2–27, http://dx.doi.org/10.5741/GEMS.54.1.2 berg/2015/11/famous-diamonds-part-3-the-pumpkin-dia-
Bruton E. (1971) Diamonds. Chilton Book Co., Radnor, PA, 273 pp. mond/) [Accessed Feb. 28, 2020]
Collins A.T. (1980) Spectroscopic investigation of a canary yellow Goss J.P., Briddon P.R., Hill V., Jones R., Rayson M.J. (2014) Iden-
diamond. Journal of Gemmology, Vol. 17, No. 4, pp. 213–222. tification of the structure of the 3107 cm–1 H-related defect in
——— (1982) Colour centres in diamond. Journal of Gemmology, diamond. Journal of Physics: Condensed Matter, Vol. 26, No.
Vol. 18, No. 1, pp. 37–75. 14, pp. 1–6, http://dx.doi.org/10.1088/0953-8984/26/14/145801
——— (1997) The electronic and optical properties of diamond. In Hainschwang T., Simic D., Fritsch E., Deljanin B., Woodring S.,
A. Paoletti and A. Tucciarone, Eds., The Physics of Diamond. DelRe N. (2005) A gemological study of a collection of chame-
Course 135, Proceedings of the International School of Physics, leon diamonds. G&G, Vol. 41, No. 1, pp. 20–35,
IOS Press, Amsterdam, p. 273. http://dx.doi.org/10.5741/GEMS.41.1.20
——— (1999) Things we still don’t know about optical centres in Hainschwang T., Notari F., Fritsch E., Massi L. (2006) Natural, un-
diamond. Diamond and Related Materials, Vol. 8, No. 8-9, pp. treated diamonds showing the A, B and C infrared absorptions
1455–1462, http://dx.doi.org/10.1016/S0925-9635(99)00013-8 (“ABC diamonds”), and the H2 absorption. Diamond and Re-
——— (2001) The colour of diamond and how it may be changed. lated Materials, Vol. 15, pp. 1555–1564,
Journal of Gemmology, Vol. 18, No. 1, pp. 341–359. http://dx.doi.org/10.1016/j.diamond.2005.12.029
Collins A.T., Mohammed K. (1982) Optical studies of vibronic Hainschwang T., Fritsch E., Notari F., Rondeau B. (2012) A new de-
bands in yellow luminescing natural diamonds. Journal of fect center in type Ib diamond inducing one phonon infrared ab-
Physics C: Solid State Physics, Vol. 15, No. 1, pp. 147–158, sorption: The Y center. Diamond and Related Materials, Vol. 21,
http://dx.doi.org/10.1088/0022-3719/15/1/012 pp. 120–126, http://dx.doi.org/10.1016/j.diamond.2011.11.002
Davies G. (1970) Aggregation of nitrogen in diamond. Nature, Vol. Hainschwang T., Fritsch E., Notari F., Rondeau B., Katrusha A.
228, No. 5273, p. 758, http://dx.doi.org/10.1038/228758a0 (2013) The origin of color in natural C center bearing diamonds.
——— (1981) The origin of the N2 absorption band in natural yel- Diamond and Related Materials, Vol. 39, pp. 27–40,
low diamonds. Portugaliae Physica, Vol. 13, No. 3/4, pp. 241– http://dx.doi.org/10.1016/j.diamond.2013.07.007
261. Jones R., Goss J.P., Briddon P.R. (2009) Acceptor level of ni-
DeMarco A. (2013) 14.82-carat orange diamond sells for a world trogen in diamond and the 270-nm absorption band. Physi-
record $35.5 million. Forbes.com, https://www.forbes.com/ cal Review B, Vol. 80, No. 3, Article 033205,
sites/anthonydemarco/2013/11/12/14-82-carat-orange-dia- http://dx.doi.org/10.1103/PhysRevB.80.033205
mond-sells-for-a-world-record-35-5-million/#5919e3c61657 King J.M., Moses T.M., Shigley J.E., Liu Y. (1994) Color grading of
Dobrinets I.A., Vins V.G., Zaitsev A.M. (2013) HPHT-Treated Dia- colored diamonds in the GIA Gem Trade Laboratory. G&G, Vol.
monds: Diamonds Forever. Springer, Heidelberg. 30, No. 4, pp. 220–242, http://dx.doi.org/10.5741/GEMS.30.4.220
Eaton-Magaña S.C., Breeding C.M. (2016) An introduction to pho- King J.M., Shigley J.E., Gelb T.H., Guhin S.S., Hall M., Wang W.
toluminescence spectroscopy for diamond and its applications (2005) Characterization and grading of natural-color yellow dia-
in gemology. G&G, Vol. 52, No. 1, pp. 2–17, monds. G&G, Vol. 41, No. 2, pp. 88–115,
http://dx.doi.org/10.5741/GEMS.52.1.2 http://dx.doi.org/10.5741/GEMS.41.2.88
Eaton-Magaña S., Shigley J.E. (2016) Observations on CVD-grown King J.M., Geurts R.H., Gilbertson A.M., Shigley J.E. (2008) Color
synthetic diamonds: A review. G&G, Vol. 52, No. 3, pp. 222– grading “D-to-Z” diamonds at the GIA Laboratory. G&G, Vol.
245, http://dx.doi.org/10.5741/GEMS.52.3.222 44, No. 4, pp. 296–321, http://dx.doi.org/10.5741/GEMS.44.4.296
Eaton-Magaña S., Shigley J.E., Breeding C.M. (2017) Observations Kitawaki H., Hisanaga M., Yamamoto M., Emori K. (2015) Type
on HPHT-grown synthetic diamonds: A review. G&G, Vol. 53, Ib yellow to brownish yellow CVD synthetic diamonds seen at
No. 3, pp. 262–284, http://dx.doi.org/10.5741/GEMS.53.3.262 CGL. Journal of Gemmology, Vol. 34, No. 7, pp. 594–604.
Eaton-Magaña S., Breeding C.M., Shigley J.E. (2018a) Natural- Lai M.Y., Stachel T., Breeding C.M., Stern R. (2019) Yellow dia-
color blue, gray, and violet diamonds: Allure of the deep. monds with colourless cores – evidence for episodic diamond
G&G, Vol. 54, No. 2, pp. 112–131, growth beneath Chidliak and the Ekati mine, Canada. Mineral-
http://dx.doi.org/10.5741/GEMS.54.2.112 ogy & Petrology, Vol. 114, No. 2, pp. 91–103,
Eaton-Magaña S., Ardon T., Smit K.V., Breeding C.M., Shigley J.E. http://dx.doi.org/10.1007/s00710-020-00693-0
(2018b) Natural-color pink, purple, red, and brown diamonds: Liddicoat R.T. (1976) True canary. G&G, Vol. 15, No. 8, p. 235.

218 YELLOW AND ORANGE DIAMONDS: THE NITROGEN FACTOR GEMS & GEMOLOGY SUMMER 2020
Liddicoat R.T., Ed (1993) The GIA Diamond Dictionary. Gemol- natural type Ib-IaA diamonds from Zimmi (West African cra-
ogical Institute of America, Santa Monica, CA, 275 pp. ton). Mineralogy and Petrology, Vol. 112, Supp. 1, pp. 243–257,
Manutchehr-Danai M. (2013) Dictionary of Gems and Gemology. http://doi.org/10.1007/s00710-018-0587-6
Springer Science & Business Media, p. 186. Smit K.V., Shirey S.B., Wang W. (2016) Type Ib diamond formation
Massi L., Fritsch E., Collins A.T., Hainschwang T., Notari F. (2005) and preservation in the West African lithospheric mantle: Re–
The “amber centres” and their relation to the brown colour. Os age constraints from sulphide inclusions in Zimmi dia-
Diamond and Related Materials, Vol. 14, No. 10, pp. 1623– monds. Precambrian Research, Vol. 286, pp. 152–166,
1629, http://dx.doi.org/10.1016/j.diamond.2005.05.003 http://dx.doi.org/10.1016/j.precamres.2016.09.022
Rondeau B., Fritsch E., Guiraud M., Chalain J.-P., Notari F. (2004) Smith E.M., Shirey S.B., Richardson S.H., Nestola F., Bullock E.S.,
Three historical ‘asteriated’ hydrogen-rich diamonds: Growth Wang J., Wang W. (2018) Blue boron-bearing diamonds from
history and sector-dependent impurity incorporation. Dia- Earth’s lower mantle. Nature, Vol. 560, No. 7716, pp. 84–87,
mond and Related Materials, Vol. 13, No. 9, pp. 1658–1673, http://dx.doi.org/10.1038/s41586-018-0334-5
http://dx.doi.org/10.1016/j.diamond.2004.02.002 Thompson R. (2004). The Tiffany Yellow. Famous, Historic and No-
Scarratt K. (1984) “Chameleon diamonds.” Journal of Gemmol- table Diamonds. http://famousdiamonds.tripod.com/tiffanyyel-
ogy, Vol. 19, No. 2, pp. 98–100. lowdiamond.html
Shigley J.E., Breeding C.M. (2013a) Gem News International: Titkov S.V., Mineeva R.M., Zudina N.N., Sergeev A.M., Ryab-
Bright yellow diamonds from Sierra Leone. G&G, Vol. 49, No. chikov I.D., Shiryaev A.A., Speransky A.V., Zhikhareva V.P.
4, pp. 259–260. (2015) The luminescent nature of orange coloration in natu-
——— (2013b) Optical defects in diamond: A quick reference ral diamonds: optical and EPR study. Physics and Chemistry
chart. G&G, Vol. 49, No. 2, pp. 107–111, of Minerals, Vol. 42, No. 2, pp. 131–141,
http://dx.doi.org/10.5741/GEMS.49.2.107 http://dx.doi.org/10.1007/s00269-014-0705-x
Shiryaev A.A, Wiedenbeck M., Hainschwang T. (2010) Oxygen in Wang W., Mayerson W. (2002) Symmetrical clouds in diamond –
bulk monocrystalline diamonds and its correlations with ni- the hydrogen connection. Journal of Gemmology, Vol. 28, No.
trogen. Journal of Physics: Condensed Matter, Vol. 22, No. 4, 3, pp. 143–152.
6 pp., http://dx.doi.org/10.1088/0953-8984/22/4/045801 Wang W., Poon T. (2018) Canary yellow diamonds. G&G, Vol. 54,
Smit K.V. (2008) Diamond formation in craton margin settings: No. 3, pp. 262–263.
Mesoproterozoic age of Ellendale peridotitic diamonds (West- Welbourn C.M., Cooper M., Spear P.M. (1996) De Beers natural ver-
ern Australia). Master’s thesis, University of Cape Town, South sus synthetic diamond verification instruments. G&G, Vol. 32,
Africa. No. 3, pp. 156–169, http://dx.doi.org/10.5741/GEMS.32.3.156
Smit K.V., Shor R. (2017) Geology and development of the Lomo- Yelisseyev A., Kanda H. (2007) Optical centers related to 3d tran-
nosov diamond deposit, northwestern Russia. G&G, Vol. 53, sition metals in diamond. New Diamond and Frontier Carbon
No. 2, pp. 144–167, http://dx.doi.org/10.5741/GEMS.53.2.144 Technology, Vol. 17, pp. 127–178.
Smit K.V., D’Haenens-Johansson U.F.S., Howell D., Loudin L.C., Zaitsev A.M. (2003) Optical Properties of Diamond: A Data
Wang W. (2018) Deformation-related spectroscopic features in Handbook. Springer-Verlag, Berlin.

For online access to all issues of GEMS & GEMOLOGY from 1934 to the present, visit:

gia.edu/gems-gemology

YELLOW AND ORANGE DIAMONDS: THE NITROGEN FACTOR GEMS & GEMOLOGY SUMMER 2020 219
FEATURE ARTICLES

SEPARATION OF NATURAL FROM

LABORATORY-GROWN DIAMOND USING
TIME-GATED LUMINESCENCE IMAGING
Colin D. McGuinness, Amber M. Wassell, Peter M.P. Lanigan, and Stephen A. Lynch

A technique that expands on the surface luminescence imaging used in the DiamondView instrument has been
developed at De Beers Group Technology, Maidenhead, UK. This provides an additional level of imaging information
by way of separating prompt and delayed surface luminescence. The technique has the added benefit of quickly
and easily distinguishing colorless or near-colorless natural diamond from laboratory-grown diamond. It can be
applied when the identification of natural diamond is required in the study of single stones, multiples in batches,
set jewelry, or in a fully automated process. The prompt and delayed luminescence characteristics of natural dia-
mond are compared with a range of chemical vapor deposition (CVD) and high-pressure, high-temperature (HPHT)
synthetic diamonds. Of significant interest are some of the less common CVD synthetic samples that have been
observed in recent years. This article will summarize the luminescence observed in different diamond types, discuss
its spectral characteristics, and serve as a useful reference when interpreting such luminescence images.

T
he task of identifying a diamond will typically signed to authenticate natural diamonds and syn-
fall to a diamond appraiser, grader, or gemologist thetic diamonds. It allows a more sophisticated ob-
while preparing a grading certificate or appraisal. servation to be made by way of a shorter wavelength
Whether a diamond is natural or lab-grown is a key excitation of <225 nm corresponding to the primary
factor in its market value and is of paramount impor- absorption edge and only exciting around 1-micron
tance to the gemologist. Many characteristics can be depth of material, such that observed luminescence
used to distinguish between natural and synthetic dia-
mond, but the inherent variability in the properties of
natural and synthetic diamond can make such a task
difficult.
In Brief
A useful and proven characteristic is the emission • Correctly identifying natural versus laboratory-grown
of luminescence when a diamond is excited by a diamond is important for its market value and main-
taining consumer confidence.
source of ultraviolet energy. Typically, a gemologist
would utilize an ultraviolet lamp with excitation • A time-gated imaging technique has been developed
wavelengths of 365 nm (long-wave ultraviolet, at De Beers Group Technology to allow easy separa-
tion of colorless/near-colorless natural and synthetic
LWUV) or 254 nm (short-wave ultraviolet, SWUV), diamond.
which correspond to the emission lines of a low-pres-
• A blue delayed luminescence can be used as a specific
sure mercury-vapor lamp. In this application, “fluo-
marker for identification of >99% of natural type IIa
rescence” would be observed during ultraviolet and type Ia diamond.
excitation, while “phosphorescence” may be ob-
served when the excitation is removed.
The De Beers DiamondView instrument (Spear
and Welbourn, 1994; Welbourn et al., 1996) was de- can be considered as produced at the surface. Studies
have shown (Martineau et al., 2004) that such lumi-
nescence imaging techniques can provide very useful
See end of article for About the Authors and Acknowledgments.
GEMS & GEMOLOGY, Vol. 56, No. 2, pp. 220–229,
information for identifying all synthetic diamond
http://dx.doi.org/10.5741/GEMS.56.2.220 types. That article reported the identification of all
© 2020 Gemological Institute of America CVD synthetics under study using DiamondView

220 TIME-GATED LUMINESCENCE IMAGING OF DIAMOND GEMS & GEMOLOGY SUMMER 2020
Figure 1. The De Beers SYNTHdetect, an instrument Figure 2. The De Beers AMS2, an instrument for auto-
designed for screening of colorless to near-colorless mated screening of small colorless to near-colorless
diamonds of all cuts mounted in jewelry. Loose diamonds for potential synthetics and simulants at
stones can also be screened with no lower size limit. rates of up to 3,600 samples/hour. The size range is
The dimensions of the instrument are 31.0 cm wide × 0.003–0.20 ct for round brilliants and 0.01–0.20 ct for
34.0 cm deep × 45.8 cm high. A range of jewelry hold- all other cuts. The instrument measures 31.5 cm wide
ers and a flat sample tray are included. Photo by × 40.0 cm deep × 55.0 cm high. Photo by Danny
Danny Bowler/De Beers Group Industry Services. Bowler/De Beers Group Industry Services.

imaging, with techniques such as X-ray topography clude time-gated luminescence (Khong et al., 1994;
providing supporting evidence in more challenging Lindblom et al., 2003; Lipatov et al., 2007), single-
cases. More recently, a set of color emission filters photon counting (Thomaz and Davies, 1978), and
has been included with the DiamondView instru- time-correlated single-photon counting (TCSPC) (Li-
ment to assist with some of the more challenging sil- augaudas et al., 2009; Jones et al., 2020), to name a
icon-doped CVD synthetics that exhibit a degree of few. The method described in this article can be used
luminescence toward the red end of the spectrum, to positively identify a natural type IIa or type Ia col-
normally swamped by blue dislocation luminescence orless diamond by accurately imaging a weak de-
(Martineau, 2017). layed luminescence having peak emission at 455 nm
To help negate the need for further time-consum- and a characteristic unquenched decay constant of
ing and expensive investigation, and to assist less ex- 8.8 ms. An example of a quenched (i.e., a decrease in
perienced gemologists, the imaging technique the luminescence efficiency and a reduction of the
employed by the DiamondView has been extended measured decay time) luminescence decay, which
by way of hardware synchronization of the light has been observed in type IIa and type Ia natural dia-
source and camera (Smith et al., 2017), allowing cam- monds, will be shown for comparison. The absence
era exposures to be controlled relative to the lamp of the blue delayed luminescence, quenched or un-
pulse with microsecond accuracy. As the temporal quenched, would indicate that a sample may be type
characteristics of diamond luminescence can be IIb natural, unusual type Ia natural, synthetic, or sim-
complex and the distinction between fluorescence ulant and would require further testing to verify its
and phosphorescence is not always clear, we use the origin. The De Beers Group screened many millions
term prompt luminescence to describe luminescence of individual colorless diamonds to verify this con-
recorded at the same time as the lamp pulse, with cept, and it is this principle that is used in both the
delayed luminescence recorded after the lamp pulse. De Beers SYNTHdetect jewelry screening instru-
There have been numerous “time-resolved” ment and the second version of the Automated
methods used to study the complex temporal nature Melee Screener (AMS2) (figures 1 and 2) (Martineau
of diamond luminescence. These methods can in- and McGuinness, 2018).

TIME-GATED LUMINESCENCE IMAGING OF DIAMOND GEMS & GEMOLOGY SUMMER 2020 221
TABLE 1. Diamond samples from this study.
More recently, this technique has identified a pre- Sample Description Shape and Cutting Style
viously unknown green luminescence feature at 499 D1 Type IIa natural Round brilliant
nm (Wassell et al., 2018) observable in lab-grown di- D2 Green-fluorescing type Ia natural Round brilliant
amonds obtained from Gemesis Inc. (now Pure
D3 Unusual natural Round brilliant
Grown Diamonds) and in silicon-containing samples
sourced by Element Six that had undergone post- D4 Weak type IIb natural Round brilliant
growth annealing. C1 As-grown nitrogen-containing CVD Round brilliant
C2 CVD of Chinese origin Round brilliant
MATERIALS AND METHODS C3 Gemesis CVD synthetic Modified square brilliant
A selection of colorless/near-colorless natural and C4 Diamond Foundry CVD Round brilliant
lab-grown polished diamond samples were chosen C5 CVD of unknown origin Round brilliant
for this study, the details of which are listed in table S1 Typical HPHT synthetic Round brilliant
1. The samples are not an exhaustive set but were
chosen to demonstrate the wide range of prompt and
delayed luminescence characteristics that can be ob-
served, and to provide a useful reference source for Room-temperature spectral data was collected by
operators interpreting output from the SYNTHdetect an Andor iStar DH320T-18U-E3 intensified CCD
instrument. camera via a Horiba iHR-320 spectrometer, with ex-
Microsecond time-gated imaging was carried out citation from the same source as the imaging setup.
using a Teledyne Dalsa Genie Nano C2050 area-scan Synchronization with the flash lamp and camera was
CMOS camera. Above-diamond band-gap excitation supplied by two externally generated pulsed signals
was provided by a Hamamatsu Photonics L7685 offset in a similar way to the imaging setup. To record
xenon flash lamp spectrally filtered to 190–227 nm prompt luminescence, the intensifier delay was set to
output, with a temporal pulse width of 2.9 μs at full zero, with emission integrated over a 5 μs gate. To
width half maximum. The camera and flash lamp record delayed luminescence, the intensifier delay was
were synchronized using the camera’s internal timer, set to 100 μs or more, with emission integrated over a
with the flash lamp signal offset by 11 μs with respect chosen gate. Here, 40 accumulations were averaged.
to the timer to account for the difference in latency A simplified diagram illustrating the principle of time-
between the two devices. Images shown represent an gated luminescence is shown in figure 3.
average of 20 individual captures. These images are at Selected room-temperature delayed luminescence
1.5× magnification as opposed to 0.25× in SYNTH- decay data was collected by a Horiba Jobin Yvon IBH
detect for the purposes of this publication. TBX-04 thermoelectrically cooled photomultiplier

Figure 3. Simplified dia-

gram of time-gated lumi-
nescence. To acquire a
Lamp delayed spectrum or a de-
layed image, the detector
or camera is activated after
a delay, which ensures that
Prompt luminescence Delayed luminescence
exposure exposure
the prompt luminescence
has decayed almost to zero.
Switching the detector or
camera on and off is analo-
Detector gous to the opening and
or camera
closing of a gate. To detect
Delay time the prompt luminescence,
and reject the delayed lu-
minescence, the detector is
turned on only when the
Time lamp is on.

222 TIME-GATED LUMINESCENCE IMAGING OF DIAMOND GEMS & GEMOLOGY SUMMER 2020
A B C

INTENSITY (ARB. UNITS)

40000
Prompt
35000
Delayed

LUMINESCENCE
30000
25000
20000
15000
10000
5000
0
300 350 400 450 500 550 600 650 700 750
WAVELENGTH (nm)

D E
10000 10000

1000 1000
COUNTS

COUNTS
100 100

10 10

1 1
0 20 40 60 80 100 120 140 160 0 20 40 60 80 100 120 140 160

TIME (ms) TIME (ms)

Figure 4. A: CMOS image of sample D1 showing the spatial distribution of the prompt blue dislocation-related
luminescence signal exhibited by a typical type IIa natural diamond. B: CMOS image showing the delayed lumi-
nescence signal exhibited. Recorded with a delay of 100 μs after the rising edge of the UV pump pulse and inte-
grated for 30 ms. C: Prompt (peak at 435 nm) and delayed (peak at 455 nm) luminescence recorded for sample
D1. Delayed luminescence was recorded with a delay of 100 μs after the rising edge of the UV pump pulse and in-
tegrated for 30 ms. D: Delayed luminescence decay profile of sample D1. In this example, the decay is un-
quenched and can be fitted to a simple monoexponential function with decay constant 8.8 ms. Type Ia natural
diamonds may also show unquenched decays. E: Example of a quenched decay seen in type IIa and type Ia natu-
ral diamonds. It would not be possible to distinguish this from an unquenched example using the imaging setup.
At the time of this writing, it is not fully understood what processes would contribute to a quenching of this de-
layed luminescence.

module via a Horiba iHR320 spectrometer with exci- View instrument. The delayed luminescence (figure
tation as above. Detector and flash lamp were trig- 4B) cannot be observed in DiamondView images due
gered simultaneously, with channels recording both to the weak nature of the luminescence and the fact
prompt and delayed luminescence in each time sweep that it is masked by the prompt dislocation lumines-
of the multichannel analyzer. The time spacing of the cence. It can be seen from the imaging that the blue
channels was selected to suit the decay time being delayed luminescence is not related to the blue dis-
measured. After the data were accumulated, the chan- location prompt luminescence. The spectral data (fig-
nels recording prompt luminescence were rejected. ure 4C) and decay data (figure 4D) also show that it
does not arise from N3 defect fluorescence. Figure 4E
RESULTS is an example of quenched blue delayed lumines-
Natural Diamonds. The luminescence characteris- cence decay and is shown for comparison with the
tics of a typical type IIa natural diamond are shown unquenched decay of sample D1. This blue delayed
in the figure 4 set. The prompt luminescence (figure luminescence is the specific marker used by AMS2
4A) is dominated by blue luminescence from dislo- and SYNTHdetect as a positive identifier of natural
cations and is identical to that seen in the Diamond- diamond, as it is not seen in synthetic diamonds.

TIME-GATED LUMINESCENCE IMAGING OF DIAMOND GEMS & GEMOLOGY SUMMER 2020 223
A B C 50000 45000
45000 Prompt

DELAYED LUMINESCENCE
40000

PROMPT LUMINESCENCE

INTENSITY (ARB. UNITS)

INTENSITY (ARB. UNITS)
40000 Delayed
35000
35000 30000
30000
25000
25000
20000
20000
15000 15000
10000 10000
5000 5000
0 0
300 350 400 450 500 550 600 650 700 750
WAVELENGTH (nm)

Figure 5. A: CMOS image of sample D2 showing the prompt green luminescence related to the H3 defect in type Ia
natural diamond. Underlying prompt blue luminescence attributed to the N3 defect is also observed. B: CMOS
image showing the blue delayed luminescence signal. In this case, structure is observed in the delayed luminescence
signal, similar to the prompt luminescence signal. C: Prompt and delayed (peak at 455 nm) luminescence recorded
for sample D2 showing green prompt luminescence. Delayed luminescence recorded with a delay of 100 μs after the
rising edge of the UV pump pulse and integrated for 30 ms.

This was confirmed in tests by De Beers Group on delay time from the microsecond range to the mil-
over 20 million individual colorless diamonds. In ad- lisecond range reduces the fast-decaying green com-
dition, the Diamond Producers Association ASSURE ponents, allowing the underlying blue
program found that less than 1% of natural diamonds luminescence to be observed—a feature that is
do not exhibit this luminescence and would be re- available to users of the SYNTHdetect instrument.
ferred for further testing (Dupuy and Phillips, 2019). Extensive testing by De Beers indicates that ~0.5%
At the time of writing, the origin of this lumines- of natural diamonds would exhibit luminescence
cence is not fully understood. properties such as this.
The diamond in figure 5A is a predominantly The figure 7 set illustrates a typical example of a
“green-fluorescing” example where the green prompt colorless weak type IIb natural diamond, sample D4.
luminescence is attributed to the H3 defect and the This would show turquoise-colored long-lived de-
blue to the N3 defect, as shown by the spectral data layed luminescence in the DiamondView. In figure
in figure 5C. The spectral data further show that the 7A, the underlying blue dislocation prompt lumines-
delayed luminescence is not related to N3 defect flu- cence can be clearly seen as the long-lived turquoise
orescence, although the delayed luminescence image delayed luminescence shown in figure 7B is gated out.
(figure 5B) shows that it is predominantly in the same This prompt luminescence would be nearly impossi-
spatial region of the stone as the N3 luminescence, ble to observe in the DiamondView due to the unsyn-
which may suggest it is related to higher aggregated chronized nature of the lamp and camera and the
states of nitrogen such as B-centers. strong long-lived delayed luminescence, which would
Diamond D3 is a slightly more unusual natural quickly mask the blue prompt luminescence. Indeed,
diamond in that it exhibits blue prompt lumines- a continual live feed of this prompt luminescence
cence (figure 6A) and a blue/green delayed lumines- image is not possible in the DiamondView. Figure 7C
cence (figure 6B). The green delayed luminescence illustrates the spectral profile, with the peak of the
component is due to the H3 defect (figure 6D). This prompt luminescence at 425 nm and the peak of the
phenomenon has been reported previously and at- delayed luminescence at 480 nm commonly seen in
tributed to the population of triplet states within such diamonds. Due to the high resolution and higher
the defect (Pereira and Monterio, 1991). In this sam- magnification of these images, it is possible to iden-
ple, the spectral data also show a small contribution tify this as natural diamond from the dislocation pat-
of H3 in the prompt luminescence. By extending the terns in the prompt luminescence. However, in a
delay from 100 μs to 2 ms (figure 6C), it can be seen lower-magnification system such as the SYNTH-
that the delayed green luminescence from H3 de- detect, which is designed to look at an ensemble of
cays much faster than the blue delayed lumines- diamonds in jewelry, the dislocation patterns may not
cence used to identify natural diamond. Therefore, be visible and such a diamond would need to be re-
in examples such as these, increasing the camera ferred for further testing. This is due to the potential

224 TIME-GATED LUMINESCENCE IMAGING OF DIAMOND GEMS & GEMOLOGY SUMMER 2020
A B

C D 30000 8000
Prompt

DELAYED LUMINESCENCE INTENSITY

PROMPT LUMINESCENCE INTENSITY
Delayed 7000
25000
6000
20000

(ARB. UNITS)
5000

(ARB. UNITS)
15000 4000

3000
10000
2000
5000
1000

0 0
300 350 400 450 500 550 600 650 700 750
WAVELENGTH (nm)

Figure 6. A: CMOS image of sample D3 showing the prompt luminescence from a slightly more unusual natural dia-
mond. B: CMOS image of sample D3 showing a more unusual blue/green delayed luminescence signal in a natural di-
amond. Recorded with a delay of 100 μs after the rising edge of the UV pump pulse and integrated for 30 ms. C:
CMOS image showing the delayed luminescence recorded with a longer delay of 2 ms after the rising edge of the UV
pump pulse and integrated for 30 ms. The green luminescence component has decayed. D: Prompt and delayed lumi-
nescence recorded for sample D3 showing both blue and green delayed luminescence. Delayed luminescence,
recorded with a delay of 100 μs after the rising edge of the UV pump pulse and integrated for 30 ms, corresponds to
figure 6B. The dotted line at 503 nm indicates the H3 zero-phonon line common to prompt and delayed luminescence.

for certain synthetic types to produce prompt blue lu- CVD Synthetic Diamonds. Sample C1 is a typical
minescence and delayed turquoise luminescence. as-grown CVD synthetic diamond that displays or-

Figure 7. A: CMOS image of sample D4 showing the spatial distribution of the prompt blue dislocation-related lu-
minescence signal exhibited by a typical type IIb natural diamond. B: CMOS image showing the delayed lumines-
cence signal exhibited. Recorded with a delay of 100 μs after the rising edge of the UV pump pulse and integrated
for 30 ms. C: Prompt and delayed (peak at 480 nm) luminescence recorded for sample D4. Delayed luminescence
recorded with a delay of 100 μs after the rising edge of the UV pump pulse and integrated for 30 ms.

A B C 40000
Prompt
INTENSITY (ARB. UNITS)

35000
Delayed
30000
LUMINESCENCE

25000
20000
15000
10000
5000
0
300 350 400 450 500 550 600 650 700 750
WAVELENGTH (nm)

TIME-GATED LUMINESCENCE IMAGING OF DIAMOND GEMS & GEMOLOGY SUMMER 2020 225
A B C 20000
Prompt
9000

PROMPT LUMINESCENCE
18000

DELAYED LUMINESCENCE
INTENSITY (ARB. UNITS)
8000

INTENSITY (ARB. UNITS)

Delayed
16000 7000
14000 6000
12000
5000
10000
4000
8000
6000 3000
4000 2000
2000 1000
0 0
300 350 400 450 500 550 600 650 700 750
WAVELENGTH (nm)

Figure 8. A: CMOS image of sample C1 showing the prompt luminescence from a typical as-grown CVD synthetic
diamond. B: CMOS image showing the delayed luminescence signal exhibited. Recorded with a delay of 100 μs
after the rising edge of the UV pump pulse and integrated for 60 ms. C: Prompt and delayed luminescence
recorded for sample C1. Delayed luminescence recorded with a delay of 100 μs after the rising edge of the UV
pump pulse and integrated for 60 ms.

ange prompt luminescence (figure 8A), primarily good indication that this synthetic diamond has un-
from the NV0 defect (figure 8C). An interesting ef- dergone post-growth annealing, as this delayed lumi-
fect noticed in this type of diamond is the delayed nescence has been previously observed in such
orange luminescence (figure 8B). Spectral analysis samples that have undergone annealing at around
shows that this also originates from the NV0 defect. 1700°C (Wassell et al., 2018).
It can be concluded that the NV0 defect can experi- Sample C3 is a commercially available CVD syn-
ence emission from spin-forbidden transitions in a thetic diamond from Gemesis that has been reported
way similar to the H3 defect (Pereira and Monterio, previously (Wassell et al., 2018). At first glance, the
1991), although this discussion is beyond the scope delayed luminescence CMOS image (figure 10B) ap-
of this article. pears very similar to that of sample C2 in figure 9.
Diamond C2 is a commercially available CVD However, the spectral data highlight that this lumi-
synthetic made in China. Figure 9A shows blue dis- nescence, although similar in color, originates from
location luminescence with discrete narrow bands an entirely different defect, with a zero-phonon line
and a weak general underlying red luminescence. at 499 nm (figure 10C). This defect has been shown
Spectral data in figure 9C shows that the red prompt to be generated by post-growth annealing at around
luminescence is predominantly from the NV0 defect. 1700°C. This sample is of interest in diamond veri-
The delayed green luminescence in this example (fig- fication, as it exhibits blue prompt luminescence
ure 9B) is due to the H3 defect. Delayed lumines- from dislocation patterns in local regions (figure
cence from the H3 defect has been discussed for the 10A) and could be incorrectly identified as a type IIa
natural diamond D3 (again, see figure 6). It is also a natural by a low-magnification system or an instru-

Figure 9. A: CMOS image of sample C2 showing the prompt luminescence from a commercially available CVD
synthetic of Chinese origin. B: CMOS image showing the delayed luminescence signal exhibited. Recorded with a
delay of 100 μs after the rising edge of the UV pump pulse and integrated for 60 ms. C: Prompt and delayed lumi-
nescence recorded for sample C2. Delayed luminescence recorded with a delay of 100 μs after the rising edge of the
UV pump pulse and integrated for 60 ms.

A B C 30000 40000
Prompt
DELAYED LUMINESCENCE
PROMPT LUMINESCENCE

35000
INTENSITY (ARB. UNITS)
INTENSITY (ARB. UNITS)

25000 Delayed
30000
20000
25000
15000 20000
15000
10000
10000
5000
5000
0 0
300 350 400 450 500 550 600 650 700 750
WAVELENGTH (nm)

226 TIME-GATED LUMINESCENCE IMAGING OF DIAMOND GEMS & GEMOLOGY SUMMER 2020
A B

C D 10000
35000 16000

DELAYED LUMINESCENCE INTENSITY

PROMPT LUMINESCENCE INTENSITY

Prompt
Delayed 14000
30000

12000
25000
(ARB. UNITS)

(ARB. UNITS)

COUNTS
10000
20000
1000
8000
15000
6000

10000
4000

5000 2000

0 0 100
300 350 400 450 500 550 600 650 700 750 0 2 4 6 8 10 12

WAVELENGTH (nm) TIME (seconds)

Figure 10. A: CMOS image of sample C3 showing the spatial distribution of the prompt blue luminescence signal
for a commercially available Gemesis CVD synthetic. B: CMOS images of the delayed 499 nm luminescence sig-
nal, with a delay of 100 μs after the rising edge of the UV pump pulse and integrated for 60 ms. C: Prompt and de-
layed (499 nm ZPL) luminescence recorded for sample C3. Delayed luminescence recorded with a delay of 100 μs
after the rising edge of the UV pump pulse and integrated for 60 ms. D: Delayed luminescence decay profile of a
blue-fluorescing Si-containing CVD synthetic. In this example, the decay is complex and non-exponential. An av-
erage decay time of ~1 sec is observed.

ment with a low-resolution camera where blue flu- in the prompt luminescence, consistent with CVD
orescence is used as a natural identifier. Figure 10D synthesis. The spectral data presented in figure 12C
shows the delayed luminescence decay profile for show that this luminescence is predominantly from
the 499 nm feature. It should not be assumed, how- the 499 nm feature with components at 575 nm (the
ever, that this observed decay is the decay profile for NV0 defect) and at 415 nm (the N3 defect), although
the 499 nm feature in general. Rather, it is the decay this level of imaging does not show these defects in
profile of the 499 nm feature in this sample. It was specific localized areas. Interestingly, the delayed lu-
also suggested by Wassell et al. (2018) that prompt minescence seen in figure 12B does not show as
luminescence of 499 nm could not be ruled out due prominent growth band patterns, suggesting the de-
to swamping by blue prompt luminescence.
layed turquoise luminescence does not form these pat-
A commercially available CVD synthetic dia-
terns in the same way as the prompt luminescence.
mond from Diamond Foundry (sample C4) shows
The spectral data show that the delayed luminescence
both a green prompt and green delayed lumines-
cence (figures 11A and 11B). Again, this could be at- is dominated by the broad turquoise feature with only
tributed to the H3 defect, but spectral data (figure a small level of the 499 nm feature visible. This sam-
11C) show that this is a result of the 499 nm feature ple also indicates that the 499 nm feature can experi-
in both prompt and delayed luminescence. ence both prompt and delayed emission.
Sample C5 is a commercially available CVD dia-
mond of unknown origin. Figure 12A shows growth HPHT Synthetic Diamond. Sample S1 is an exam-
bands roughly perpendicular to the growth direction ple of a typical commercially sourced HPHT syn-

TIME-GATED LUMINESCENCE IMAGING OF DIAMOND GEMS & GEMOLOGY SUMMER 2020 227
A B C 40000
Prompt

INTENSITY (ARB. UNITS)

35000
Delayed
30000

LUMINESCENCE
25000
20000
15000
10000
5000
0
300 350 400 450 500 550 600 650 700 750
WAVELENGTH (nm)

Figure 11. A: CMOS image of sample C4 showing prompt green luminescence from the 499 nm feature for a com-
mercially available Diamond Foundry CVD synthetic. B: CMOS image of the delayed 499 nm luminescence sig-
nal, with a delay of 100 μs after the rising edge of the UV pump pulse and integrated for 60 ms. C: Prompt and
delayed luminescence recorded for sample C4. Delayed luminescence recorded with a delay of 100 μs after the ris-
ing edge of the UV pump pulse and integrated for 60 ms.

thetic diamond. The delayed luminescence (figure tifying feature, as this combination of luminescence
13B) is identical to the strong long-lived lumines- features have not been seen in natural diamond.
cence that would be observed in the DiamondView.
This long-lived luminescence is understood and CONCLUDING COMMENTS
has been reported as being interpretable in terms of Diamond screening is becoming increasingly chal-
a donor-acceptor recombination model (Watanabe lenging, and organizations other than gemological lab-
et al., 1997). Of interest here is the prompt lumi- oratories, such as independent jewelers, auction
nescence (figure 13A), which is a weak green/red houses, and pawnbrokers, are charged with this im-
and would normally be obscured by the long-lived portant task. Not only are more sophisticated tech-
luminescence in the DiamondView. This lumines- niques required to assist with this challenge, but the
cence could be due to the presence of metal ions in equipment must not be so complex to use and difficult
the solvent catalyst. However, the prompt lumines- to understand that diamond screening becomes an
cence spectrum shown in figure 13C, along with overly time-consuming and inefficient task. In this ar-
the delayed luminescence, is broad and featureless ticle, a selection of natural and lab-grown polished col-
at room temperature and gives little in the way of orless diamonds has been studied using deep-UV
information. excitation where the recording of the luminescence is
This weak green prompt luminescence and strong synchronized with the excitation source, allowing a
long-lived turquoise luminescence are typical of time-gated measurement. This technique provides a
many HPHT synthetics. They can be used as an iden- useful additional level of verification when screening

Figure 12. A: CMOS image of sample C5 showing the spatial distribution and growth bands perpendicular to the
growth direction in the prompt green luminescence signal exhibited by a commercially available CVD synthetic di-
amond. B: CMOS image showing the delayed turquoise-colored luminescence signal exhibited. The growth patterns
in the prompt luminescence are less prominent here. C: Prompt and delayed luminescence recorded for sample C5.
Delayed luminescence recorded with a delay of 100 μs after the rising edge of the UV pump pulse and integrated for
60 ms. The inset zooms in on the region of the 499 nm ZPL.

A B C
DELAYED LUM. INTENSITY (ARB. UNITS)

30000 40000
PROMPT LUM. INTENSITY (ARB. UNITS)

Prompt 35000
25000 Delayed
490 510 530 550
30000
20000
25000

15000 20000

15000
10000
10000
5000
5000

0 0
300 350 400 450 500 550 600 650 700 750
WAVELENGTH (nm)

228 TIME-GATED LUMINESCENCE IMAGING OF DIAMOND GEMS & GEMOLOGY SUMMER 2020
A B C 20000 25000
Prompt

DELAYED LUMINESCENCE
PROMPT LUMINESCENCE

INTENSITY (ARB. UNITS)

INTENSITY (ARB. UNITS)
Delayed 20000
15000

15000
10000
10000

5000
5000

0 0
300 350 400 450 500 550 600 650 700 750
WAVELENGTH (nm)

Figure 13. A: CMOS image of sample S1 showing a prompt green/red luminescence from a typical colorless HPHT
synthetic diamond. B: CMOS image of sample S1 showing the delayed luminescence from a typical colorless
HPHT synthetic—normally attributed to the presence of boron. C: Prompt (peak ~513 nm) and delayed (peak ~480
nm) luminescence recorded for sample S1. Delayed luminescence recorded with a delay of 100 μs after the rising
edge of the UV pump pulse and integrated for 10 ms.

for challenging synthetics such as blue-fluorescing serve as a useful reference guide and knowledge base
CVD synthetics. Although the examples contained in for all SYNTHdetect users when studying individual
this article are not an exhaustive list, they should samples or screening challenging examples.

ABOUT THE AUTHORS ACKNOWLEDGMENTS

Dr. McGuinness is a senior scientist, and Dr. Lanigan is a research The authors would like to thank Mr. Hugh Leach at De Beers
scientist, at De Beers Group Technology, Maidenhead, Berkshire, Group Technology for mechanical engineering and instrument
United Kingdom. Ms. Wassell is a doctoral student, and Dr. Lynch development effort, and to the team at Element Six whose sam-
is associate dean of research, at the School of Physics & Astrono- ples were used extensively in the development of this technique.
my, Cardiff University, Cardiff, United Kingdom.

REFERENCES
Dupuy H., Phillips J.C. (2019) Selecting a diamond verification in- fidence technical research and diamond verification instru-
strument based on the results of the Assure program: An initial ments. Diamonds – Source to Use Conference 2018, Southern
analysis. Journal of Gemmology, Vol. 36, pp. 606–619. African Institute of Mining and Metallurgy, Johannesburg, June
Jones D.C., Kumar S., Lanigan P.M.P., McGuinness C.D., Dale 11–13, pp. 35–44.
M.W., Twitchen D.J., Fisher D., Martineau P.M., Neil M.A.A., Martineau P.M., Lawson S.C., Taylor A.J., Quinn S.J., Evans D.J.F.,
Dunsby C., French P.M.W. (2020) Multidimensional lumines- Crowder M.J. (2004) Identification of synthetic diamond grown
cence microscope for imaging defect colour centres in diamond. using chemical vapor deposition (CVD). G&G, Vol. 40, No. 1,
Methods and Applications in Fluorescence, Vol. 8, No. 1, pp. 2–25, http://dx.doi.org/10.5741/GEMS.40.1.2
014004, http://dx.doi.org/10.1088/2050-6120/ab4eac Pereira E., Monterio T. (1991) Delayed luminescence of the H3 cen-
Khong Y.L., Collins A.T., Allers L. (1994) Luminescence decay tre in diamond. Journal of Luminescence, Vol. 48-49, pp. 814–
time studies and time-resolved cathodoluminescence spec- 818, http://dx.doi.org/10.1016/0022-2313(91)90248-T
troscopy of CVD diamond. Diamond and Related Materials, Smith J.G.C., McGuinness C.D., Fisher D. (2017) Luminescence
Vol. 3, No. 7, pp. 1023–1027, http://dx.doi.org/10.1016/0925- measurements in diamond. Patent Application WO/2017/001835
9635(94)90112-0 Spear P.M., Welbourn C.M. (1994) Distinguishing natural from
Liaugaudas G., Collins A.T., Suhling K., Davies G., Heintzmann synthetic diamond. Patent Application WO/83/00389A1
R. (2009) Luminescence-lifetime mapping in diamond. Journal Thomaz M.F., Davies G. (1978) The decay time of N3 luminescence
of Physics: Condensed Matter, Vol. 21, No. 36, 364210, in natural diamond. Proceedings of the Royal Society A, Vol. 362,
http://dx.doi.org/10.1088/0953-8984/21/36/364210 No. 1710, pp. 405–419, http://dx.doi.org/10.1098/rspa.1978.0141
Lindblom J., Hölsä J., Papunen H., Häkkänen H., Mutanen J. (2003) Wassell A.M., McGuinness C.D., Hodges C., Lanigan P.M.P.,
Differentiation of natural and synthetic gem-quality diamonds Fisher D., Martineau P.M., Newton M.E., Lynch S.A. (2018)
by luminescence properties. Optical Materials, Vol. 24, No. 1-2, Anomalous green luminescent properties in CVD synthetic di-
pp. 243–251, http://dx.doi.org/10.1016/S0925-3467(03)00130-7 amonds. Physica Status Solidi A, Vol. 215, No. 22, 1800292,
Lipatov E.I., Lisitsyn V.M., Oleshko V.I., Tarasenko V.F. (2007) Spec- http://dx.doi.org/10.1002/pssa.201800292
tral and kinetic characteristics of the pulsed cathodolumines- Watanabe K., Lawson S.C., Isoya J., Kanda H., Sato Y. (1997) Phos-
cence of a natural IIa-type diamond. Russian Physics Journal, phorescence in high-pressure synthetic diamond. Diamond
Vol. 50, No. 1, pp. 52–57, http://dx.doi.org/10.1007/s11182-007- and Related Materials, Vol. 6, No. 1, pp. 99–106,
0005-8 http://dx.doi.org/10.1016/S0925-9635(96)00764-9
Martineau P.M. (2017) The 68th Diamond Conference. University Welbourn C.M, Cooper M., Spear P.M. (1996) De Beers natural ver-
of Warwick, T2.1, July 10–13. sus synthetic diamond verification instruments. G&G, Vol. 32,
Martineau P.M., McGuinness C.D. (2018) De Beers consumer con- No. 3, pp. 156–169, http://dx.doi.org/10.5741/GEMS.32.3.156

TIME-GATED LUMINESCENCE IMAGING OF DIAMOND GEMS & GEMOLOGY SUMMER 2020 229
FEATURE ARTICLES

HISTORY OF THE CHIVOR EMERALD MINE,

PART II (1924–1970):
BETWEEN INSOLVENCY AND VIABILITY
Karl Schmetzer, Gérard Martayan, and Andrea R. Blake

The history of the Chivor emerald mine in Colombia is a saga with countless twists and turns, involving parties
from across the globe. Indigenous people initially exploited the property, followed by the Spanish in the sixteenth
and seventeenth centuries, before abandonment set in for 200 years. The mine was rediscovered by Francisco
Restrepo in the 1880s, and ownership over the ensuing decades passed through several Colombian owners and
eventually to an American company, the Colombian Emerald Syndicate, Ltd., with an intervening but unsuc-
cessful attempt by a German group organized by Fritz Klein to take control. With the Colombian Emerald Syn-
dicate succumbing to bankruptcy in 1923, the property was sold and then transferred in 1924 to another
American firm, the Colombia Emerald Development Corporation. Under the new ownership, stock market spec-
ulation played a far more prominent role in the story than actual mining. Nonetheless, periods of more produc-
tive mining operations did take place under managers Peter W. Rainier and Russell W. Anderton. Yet these were
not enough to prevent the company, renamed Chivor Emerald Mines, Inc. in 1933, from entering insolvency in
1952 and being placed into receivership. Leadership by Willis Frederick Bronkie enabled the firm to regain in-
dependence in 1970 and shortly thereafter to be sold in a series of transactions, with Chivor gradually being re-
turned to Colombian interests.

M
any legends are told about the history of the was involved. Only intermittent operational activ-
Chivor emerald mine.1 The story begins ities took place until 1912, when the German gem
with the sporadic working of the Colom- cutter and merchant Fritz Klein joined Restrepo with
bian mine by indigenous people before being sought an increased focus on the operational side. Early
out by Spanish conquistadores in the first half of the mining by Restrepo and Klein yielded several finds
sixteenth century. The property was exploited by the promising enough for them to travel to Germany to-
Spanish in the sixteenth and seventeenth centuries gether in 1913 to seek investors (figure 1). Further
and then forgotten in the jungle for a period of more work was curtailed one year later, however, when
than two centuries after 1672. Schmetzer et al. (2020) Restrepo died in 1914 and Klein, who had hoped to
chronicled the first part of the modern era commenc- purchase the mine with German funding, was
ing after the 200-year break, covering from 1880 to thwarted by the outbreak of World War I.
1925. During that interval, Colombian miner Fran- When hostilities ended, Klein sought to recom-
cisco Restrepo searched for and rediscovered the mence his efforts to buy the mine in 1919, but an
mine, and mining titles were granted to him and his American corporation, the Colombian Emerald Syn-
associates in 1889. Through a series of transactions, dicate, Ltd., had in the interim obtained an option to
the mining titles and land in the area came under the purchase the mine. That option was exercised, and
ownership of the Compañía de las Minas de Esmeral- the mine was sold in December 1919 to two key rep-
das de Chivor, a Colombian entity in which Restrepo resentatives of the American group, Wilson E. Grif-
fiths and Carl K. MacFadden. On behalf of the
Colombian Emerald Syndicate, mining operations
See end of article for About the Authors and Acknowledgments.
GEMS & GEMOLOGY, Vol. 56, No. 2, pp. 230–257,
http://dx.doi.org/10.5741/GEMS.56.2.230
1
© 2020 Gemological Institute of America See, e.g., Peretti and Falise, 2018.

230 HISTORY OF CHIVOR, PART II: 1924–1970 GEMS & GEMOLOGY SUMMER 2020
1923. The Chivor mine was sold in 1924, and it is
here that the present paper, Part II, picks up the tale.
The following decades of Chivor’s operational his-
tory were dominated by three key mine managers or
administrators, Peter W. Rainier (1890–1945), Russel
W. Anderton (1909–1982), and Willis Frederick Bron-
kie (1912–1979).2 Both Rainier (Green Fire, 1942) and
Anderton (Tic-Polonga, 1953, 1954)3 published books
recounting their adventures, and Bronkie’s successful
management was the subject of several reports, in-
cluding those by Johnson (1959, 1961). These pub-
lications, however, fall short of fully elucidating the
mine’s modern development, and the shortcomings
are analogous to those faced in Part I with Klein’s
work (1941, 1951). For example, Rainier’s account
provided few actual dates, and Anderton likewise of-

Figure 2. Portrait of the German gem merchant Fritz

Klein, who cooperated with Restrepo at Chivor from
1912 to 1914. Klein was subsequently engaged in the
early 1920s to serve as head of mining operations for
the new American owners of the property. Later he
established his own company in Germany, focused
on the emerald trade, and he also worked for the Co-
Figure 1. Portrait of the Colombian miner Francisco lombian government. Photo 1937; courtesy of Carola
Restrepo Escobar (center) in 1913. Restrepo is cred- Kroll.
ited with rediscovery of the Chivor emerald mines in
the 1880s, after two centuries of abandonment. He
later traveled to Germany with Fritz Klein to seek in-
vestors for the project, returning to Colombia in 1913
by ship. Klein followed aboard another ship some-
what later. Photo with passengers and crew
members, most likely of the President Lincoln re-
turning from Germany to Colombia via New York;
courtesy of Eduardo Restrepo Ortega.

were conducted from 1919 to 1924 under the alter-

nating leadership of the English miner and entrepre-
neur Christopher Ernest Dixon, who had lived in
Colombia since the late 1880s, and Klein (figure 2),
who served two terms of several months each under
contracts with the new owners. Although both
Dixon and Klein reported good production, the finan-
cial situation of the company nonetheless deterio-
rated, and an involuntary bankruptcy petition was
filed against the Colombian Emerald Syndicate in

2
See, e.g., Weldon et al., 2016.
3
Tic-Polonga was published in the United States in 1953 and in the
United Kingdom in 1954. The 1954 edition was consulted for purposes
of this study, and references henceforth will be to that publication.

HISTORY OF CHIVOR, PART II: 1924–1970 GEMS & GEMOLOGY SUMMER 2020 231
fered little to contextualize the timeline. Con- Colombian materials were found in the form of
sequently, the present authors sought to fill existing government publications printed, typically anony-
gaps and present a comprehensive survey through re- mously, in the “Diario Oficial (Colombia),” in the
liance on contemporaneous primary documents and “Gaceta Judicial (Colombia)” of the Corte Suprema
references. This second installment focuses on an de Justicia (Supreme Court), and in the reports of
overview of the half-century after the bankruptcy of ministries or other government offices, especially in
the Colombian Emerald Syndicate, from 1924 to the series “Memoria del Ministro de Minas y Petró-
1970. Following brief interim transactions in 1924, leos al Congreso.” These included legislative texts,
Chivor was owned throughout this period by a single decrees, governmental or judicial pronouncements,
American entity, operating first under the name Co- and statements on petitions and complaints. The Ar-
lombia Emerald Development Corporation and from chivo General de la Nación (Colombia) in Bogotá
1933 as Chivor Emerald Mines, Inc. A summary of further proved to be a repository of files collecting
the main events is given in table 1. correspondence, notarial acts (escrituras), and attach-
ments reflecting mine title transfers, sales of shares
SOURCES between various shareholders, and agreements shed-
ding light on the ongoing fluctuations in Chivor
The search for contemporaneous materials spanned
ownership and related events.4
South America, North America, and Europe (i.e., Co-
Governmental collections of a similarly useful
lombia, the United States, and the United Kingdom).
nature were found in the United Kingdom. Particu-
The primary documents found were compared with
larly apposite was correspondence and associated
the known literature on the topic. Full citations for
materials archived at the Royal Academy of Music,
the principal existing publications consulted are
London, where the legacy of Peter Rainier’s sister
Priaulx is preserved.
Relevant information was also found in periodi-
In Brief cals, largely from American publishers, concerning
mining and the mining industry, such as the Engi-
• From 1924 to 1970, the Chivor emerald mine was neering and Mining Journal, Minerals Yearbook, or
under the ownership principally of a single American
firm, operating first as the Colombia Emerald Devel-
Mineral Trade Notes. Financial sector periodicals
opment Corporation and then as Chivor Emerald were equally pertinent, and publications of U.S. uni-
Mines, Inc. versities yielded material on alumni who went on to
• The early years of that period were dominated by stock work as mining engineers or geologists at Chivor.
market speculation and scandal, featuring in particular Even online resources added to the available con-
the manipulations of Frederick Lewisohn and George temporaneous data. Noteworthy in this regard were
Graham Rice. digitized compilations of records detailing transit of
• More productive operational stretches followed with persons, such as passenger lists maintained by the
Peter W. Rainier and Russell W. Anderton managing United States Immigration Office in New York. Those
the mine, but insolvency and receivership nonetheless records detailed information that could incorporate
ensued in 1952. departures and arrivals through national borders, start-
• Leadership by Willis Frederick Bronkie eventually en- ing points, destinations, and, most importantly, spe-
abled the company to exit receivership in 1970 and to cific dates to help establish or corroborate event time
be sold, returning Chivor to Colombian interests. frames. Similarly insightful were digital archives of
newspapers across the United States, which enabled
a view of events as presented to the public.
Finally, personal information obtained from eye-
summarized in the reference list. Other primary doc- witnesses such as Peter Rainier, Jr., Manuel J. Marcial,
uments, such as governmental and judicial reports
and decrees that may appear in broader compen-
diums, certain trade periodicals, archived contracts, 4
Most germane at the Archivo General de la Nación (Colombia) in
business records, newspaper articles, personal corre- Bogotá, were several notarial acts (escrituras) designated herein by es-
spondence, etc., are identified to the extent feasible critura number and date. Likewise instructive was the file “Joaquín
Daza B.,” Volume “Propuestas Minas 99,” Ministerio de Industrias,
in the footnotes, along with brief citations to pub- Departamento de Minas y Petróleos, which contained documents
lications included in the reference list. covering 1929 to 1954 and detailing mine boundaries and history.

232 HISTORY OF CHIVOR, PART II: 1924–1970 GEMS & GEMOLOGY SUMMER 2020
TABLE 1. History of Chivor, 1924–1970.

Date Events in Colombia and the United States

March 1924 The “Chivor Emerald Corporation” is incorporated under the laws of the State of Delaware
August 1924 The mining titles for Chivor 1 and 2 and land in the region are sold to the “Chivor Emerald Corporation” in a bankruptcy sale
November 1924 The “Colombia Emerald Development Corporation” is incorporated in Delaware by three individuals associated with various mining companies
owned by the Lewisohn family; E.J. MacNamara serves as president
Winter 1924 Ownership of Chivor 1 and 2 and the related land is transferred to the “Colombia Emerald Development Corporation” in exchange for stock
1925–1926 A large block of shares in the “Colombia Emerald Development Corporation” is transferred from F. Lewisohn to G.G. Rice and then aggressively
marketed by G.G. Rice
1925–Spring 1926 W. Burns leads mining operations at Chivor
May 1926 C. Mentzel succeeds W. Burns as head of mining operations at Chivor
1926–1928 Legal proceedings implicating F. Lewisohn, E.J. MacNamara, and G.G. Rice in allegations of fraud and stock market manipulation take place in
New York
1927 P.W. Rainier replaces C. Mentzel as the mining engineer overseeing operations at Chivor
December 1929 J. Daza instigates a legal dispute over ownership and control of the Chivor mines, referred to as the Joaquín Daza case
February 1930 P.W. Rainier presents a study describing the Chivor mining operations at a conference in New York
April–August 1930 Geologist V.A. Gilles evaluates the Chivor mine
August 1930 R.E. Sylvester joins P.W. Rainier at Chivor as an assistant
1930 The “Colombia Emerald Development Corporation” sustains a drastic loss for the year on mining operations at Chivor
1931–1957 Attorney F.P. Pace serves the “Colombia Emerald Development Corporation” and “Chivor Emerald Mines, Inc.” in multiple roles, including
president in 1947, subsequent to W.J. Cowan’s tenure
January–February The Chivor mine is inspected in connection with the Joaquín Daza case by Colombian officials together with P.W. Rainier, R.E. Sylvester, and J.
1931 Daza; later security problems arise at the mine
October 1931 R.E. Sylvester returns to New York at the end of his tenure at Chivor; P.W. Rainier travels separately to New York to meet with the mine owners
By early 1932– The Chivor mine is largely dormant after being reported as closed, with little documented operational activity beyond illegal mining and some
1947 work under lease to a Colombian company
September 1933 The name of the “Colombia Emerald Development Corporation” changes to “Chivor Emerald Mines, Inc.”
1936–1940 Records reflect that banker R.E. Henry was serving as president of “Chivor Emerald Mines, Inc.,” with E.J. MacNamara’s term having ended at
some point after 1932
1937–1940/1941 Records reflect that the Chivor mine was being leased to the “Compañía de Esmeraldas de Colombia” and/or being operated with E. Fernández
as the mine administrator
Early to mid-1940s W.J. Cowan holds the position of president for “Chivor Emerald Mines, Inc.”
January 1941 Chivor is evaluated again by Colombian officials in connection with the Joaquín Daza case, with a continued focus on mine boundaries,
leading to a report by N. Rosso
April 1944 P.W. Rainier considers returning to Chivor when approached by the mine owners about a possible consulting position, but no further steps are
taken, and the mine remains inactive
1947 F.P. Pace replaces W.J. Cowan as president. In 1947, 1948, and 1950, Pace travels to Colombia to restart and supervise mining operations on
behalf of “Chivor Emerald Mines, Inc.”; local control in Colombia is placed under a series of managers
Early 1950–Fall R.W. Anderton serves as head of mining operations at Chivor
1950
Late 1950/early A dispute ensues between U.S. and Colombian associates, and the situation at the mine is characterized by lawlessness, with all production
1951 being sold on the black market
By February 1952 Insolvency proceedings commence in Bogotá and result in “Chivor Emerald Mines, Inc.,” being placed in receivership
1952–1970 The Chivor mine is operated under receivership
1952–1953 R.W. Anderton returns to Colombia and again serves in a management capacity at Chivor
January 1955– W. de Freitas serves as receiver and P.P. Patiño as mine inspector at Chivor
September 1957
1957 F.P. Pace dies; O.J. Troster of New York becomes involved, serving as a director and secretary of “Chivor Emerald Mines, Inc.”
1960s (or before) G.D. Besler serves as president of “Chivor Emerald Mines, Inc.”
October 1957– W.F. Bronkie is named receiver of “Chivor Emerald Mines, Inc.,” and simultaneously serves as mining engineer supervising operations at Chivor,
circa 1969/1970 working in conjunction with Spanish mine manager E. Muñoz by the mid-1960s
1960s W.F. Bronkie’s business approach expands to incorporate buying emeralds on the open market in Colombia and selling rough and faceted
stones to international customers, as well as operating jewelry retail shops in Colombia and elsewhere
1968 The press reports that W.F. Bronkie intends to file documentation that will set in motion a process to end the receivership
1970 The receivership is lifted, and as of the year of the transition, American attorney R.S. Pastore is serving as receiver, N.R. Merriam as general
manager, and J. Rodriguez as mine manager
After 1970 Ownership of Chivor is gradually transferred to Colombian stakeholders

HISTORY OF CHIVOR, PART II: 1924–1970 GEMS & GEMOLOGY SUMMER 2020 233
on the report consulted,6 which did not even cover
liabilities in the range of $45,000 of the former
owner, the Colombian Emerald Syndicate (see part
one of this article).7 The Chivor Emerald Corporation
had been incorporated on March 6, 1924, under the
laws of the State of Delaware, by Samuel C. Wood,
Harry C. Hand, and Raymond J. Gorman.8 The stated
corporate purpose was to “mine and prepare for mar-
ket emeralds, diamonds.”9 The three named individ-
uals served as registrants of numerous corporations
across a spectrum of industries in the 1920s (e.g., mo-
tion pictures, radio accessories, petroleum, and
rubber), operating via a business model that acquired
bankrupt or troubled companies (or their assets) and
resold them after improvement in the price.10
Another new entity, the Colombia Emerald Devel-
opment Corporation, was then incorporated on No-
vember 7, 1924, likewise in the state of Delaware, by
Edmund J. MacNamara (1874–?), Ernest W. Brown,
and William B. Anderson.11 The intended business
was “to mine, manufacture and deal in precious
stones, emeralds, rubies, diamonds, jades and gar-
nets.”12 The Chivor claims and land were transferred
to that company in late 1924 in exchange for stock.13
The Chivor Emerald Corporation thereafter served
merely as a holding company, owning approximately
40 percent of the capital stock in the Colombia Emer-
ald Development Corporation but no other assets.14
Figure 3. Personal recollections regarding the history
of the Chivor emerald mine were obtained from eye-
witnesses such as don Alfonso Montenegro. Montene- 5
The Cumulative Daily Digest of Corporation News, No. 3 (1924), pp.
gro, a miner presently residing in the small town of 150, 161. This entity should not be conflated with an unrelated Cana-
Macanal, not far from Chivor, began working at the dian company of the same name that was active at Chivor decades
mine in 1955 at the age of 12 under the receivership later in the 1990s.
6
of Walter de Freitas, and he continued to serve under The Brooklyn Daily Eagle, January 9, 1927, p. 22; The New York
Freitas’s successor Willis Frederick Bronkie. Photo by Times, January 9, 1927; The Bridgeport Telegram, January 10, 1927, p.
5; The Tribune (Coshocton, Ohio), January 10, 1927, p. 4; The New
Gérard Martayan, December 2018. York Times, January 28, 1927; The New York Times, February 20,
1927; Jewelers’ Circular, 94, No. 4 (1927), p. 61; Chicago Tribune,
February 27, 1927, p. 40.
7
Robert D. Fisher Manual of Valuable and Worthless Securities, 6
don Alfonso Montenegro (figure 3), Renata de Jara, (1938), p. 227.
8
Gonzalo Jara, and Robert E. Friedmann, was invalu- State of Delaware, Department of State, Division of Corporations, En-
tity Details; The Morning News, March 7, 1924, p. 11.
able in achieving the aims of this study. 9
The Morning News, March 7, 1924, p. 11.
10
The Morning News, November 15, 1923, p. 9; The Morning News,
CHIVOR AS A SUBJECT OF STOCK MARKET April 14, 1925, p. 11; The Morning News, July 1, 1925, p. 9; Arizona
Republic, March 13, 1929, p. 4.
SPECULATION (1924–1927) 11
State of Delaware, Department of State, Division of Corporations,
To recap the period from 1924 to 1925 bridging Parts Entity Details; The Morning News, November 8, 1924, p. 13; The Phil-
adelphia Inquirer, November 10, 1924, p. 33; The Cumulative Daily
I and II of this study, the mining titles for Chivor 1 Digest of Corporation News, No. 4 (1924), p. 222; Winkler, 1928.
and 2 and the related land ownership were sold in a 12
The Philadelphia Inquirer, November 10, 1924, p. 33.
bankruptcy sale on or about August 23, 1924, to the 13
The Cumulative Daily Digest of Corporation News, No. 4 (1924), pp.
recently formed Chivor Emerald Corporation.5 The 209, 222; Winkler, 1928; Rainier, 1929, 1931.
14
price paid was only US$7,800 to $8,000, depending Robert D. Fisher Manual of Valuable and Worthless Securities, 11
(1946), p. 149.

234 HISTORY OF CHIVOR, PART II: 1924–1970 GEMS & GEMOLOGY SUMMER 2020
Per available evidence, the two firms were headquar- the Santa Fe Gold & Copper Mining Company.23
tered together in New York City and generally con- MacNamara, a mining engineer and entrepreneur
trolled by a common group of individuals,15 with the having wide-ranging participation with Lewisohn
Colombia Emerald Development Corporation appar- Brothers entities, stepped in as president of the Co-
ently established to take over and carry on the emerald lombia Emerald Development Corporation, a posi-
business as the operating enterprise. Registration of tion that he would continue to fill in the years
the two companies in Colombia took place in Feb- ahead.24 After the transfer of assets of the Chivor Em-
ruary 1925 for the Chivor Emerald Corporation16 and erald Corporation to the new company, Frederick
in July 1925 for the Colombia Emerald Development Lewisohn acquired between 40 and 50 percent of the
Corporation.17 The governor of Delaware sub- authorized capital stock of the Colombia Emerald
sequently repealed the charter of the Chivor Emerald Development Corporation for a price in the range of
Corporation in 1961 for failure to pay taxes.18 10 cents per share, or approximately $50,000.25
A number of the principal figures involved in Shares were also being advertised to the American
leadership of the Colombia Emerald Development public by March 25, 1925, for a subscription price of
Corporation were associated with the Lewisohn $2 per share.26 As of May 19, 1926, the company was
group of companies, an American offshoot of the listed on the Boston Curb Exchange at $1⅞ per
Lewisohn family mercantile business based origi- share.27
nally in Hamburg, Germany.19 The brothers Leonard
(1847–1902) and Adolph (1849–1938) Lewisohn had
followed their older brother Julius in the second half 15
For example, the reported addresses for the two companies during
of the 1860s from Hamburg to New York to assist the late 1930s and early 1940s were identical, given as 32 Broadway,
New York City. Robert D. Fisher Manual of Valuable and Worthless Se-
with the American branch. Julius returned to Ham- curities, 6 (1938), p. 205; Robert D. Fisher Manual of Valuable and
burg in 1872, and Leonard and Adolph went on to Worthless Securities (1944), p. 147.
16
create one of the larger copper mining and processing Escritura 344, February 16, 1925, Notaria 1, Archivo General de la
Nación (Colombia).
empires in the United States between 1880 and 1900, 17
Escritura 1607, July 11, 1925, Notaria 1, Archivo General de la
operating as Lewisohn Brothers.20 The German and Nación (Colombia).
American branches of the business formally sep- 18
State of Delaware, Executive Department, Chapter 465 Proclama-
arated in 1887. tion, January 16, 1961.
19
Engineering and Mining Journal, 110, No. 15 (1920), p. 722; Al-
By the time of the 1924 Chivor transaction, the brecht, 2013.
Lewisohn group of companies had already been in- 20
Engineering and Mining Journal, 110, No. 15 (1920), p. 722; Al-
volved in mining activities in Colombia for several brecht, 2013.
21
years. In 1916, the Lewisohn Brothers and Adolph The New York Times, June 13, 1919; The Economist, 89 (1919), p.
39; The Magazine of Wall Street, 25 (1919), p. 57; The New York
Lewisohn & Sons formed the South American Gold Times, July 5, 1959, p. 56.
and Platinum Company. After buying multiple 22
Escritura 3084, December 27, 1919, Notaria 1, Archivo General de
mines in Colombia and merging with various other la Nación (Colombia).
23
entities, this corporation had become the largest pro- The Copper Handbook, 10 (1911), pp. 1528–1529; Trow’s New
York Copartnership and Corporation Directory, 63 (1915), p. 912;
ducer of noble metals (gold and platinum) in that Trow’s New York Copartnership and Corporation Directory, 66 (1919),
country. Frederick Lewisohn (1881–1959), one of p. 1019; The Copper Handbook, 14 (1920), pp. 1256–1257; Moody’s
Leonard Lewisohn’s sons, served as a vice president Manual of Railroads and Corporation Securities, 2, Part 1, Industrial
Section (1921), pp. 1369–1370; Engineering and Mining Journal, 113,
of the company.21 The Lewisohn group likely opted No. 9 (1922), p. 376.
to take advantage of the low price at which the assets 24
The Mines Handbook, 12 (1916), p. 1005; Mineral Resources of
of the bankrupt Colombian Emerald Syndicate could Michigan (1917), p. 53; Engineering and Mining Journal, 122, No. 16
(1926), p. 602; The New York Times, August 7, 1926; The New York
be acquired (as compared to the £46,000, equivalent Times, August 10, 1926; In re Idaho Copper Corporation, Deposition
to US$230,000, previously paid for the Chivor mines of Keyes Winter, Deputy Attorney General, State of New York, Su-
preme Court of New York (Sept. 8, 1926); The New York Times, Oc-
in 1919),22 to establish a new line of business in em- tober 12, 1926; Engineering and Mining Journal, 123 (1927), p. 106;
erald mining. Poor’s Register of Directors of the United States (1928), p. 918; Poor’s
The three listed incorporators of the Colombia Register of Directors of the United States and Canada (1932), p. 1291.
25
Emerald Development Corporation all held roles The Brooklyn Daily Eagle, January 6, 1927, p. 2; Salt Lake Telegram,
January 13, 1927, p. 5; The New York Times, January 28, 1927.
(e.g., president, secretary, or director) in different 26
The News Journal (Wilmington, Delaware), March 28, 1925, p. 17.
mining companies associated with the Lewisohn 27
The Anaconda Standard (Anaconda, Montana), May 19, 1926, p.
group, especially the Seneca Copper Corporation and 12.

HISTORY OF CHIVOR, PART II: 1924–1970 GEMS & GEMOLOGY SUMMER 2020 235
Chivor for the former owner, the Colombian Emerald
Syndicate.29 During his tenure, Burns expended over
$40,000 on the Colombia Emerald Development Cor-
poration’s behalf on the property, but the resultant
emeralds produced were not worth over $4,000.30
Burns later resigned, and he was succeeded by mining
engineer Charles Mentzel (1881–1949), likewise pre-
viously associated with the Lewisohn-owned Santa Fe
Gold & Copper Mining Company and having gained
experience in Colombia in 1916 and 1917.31 Mentzel
arrived at Chivor in late May 1926, and Burns re-
turned to New York on June 13, 1926.32
For operational purposes, materials published by
the Colombia Emerald Development Corporation sug-
gest that the company was only focusing on and/or as-
serting claims over the Chivor 1 and 2 mines,
excluding the other three mines of the Chivor group.33
Meanwhile, as to the stock market machinations,
Frederick Lewisohn had transferred his package of
shares to George Graham Rice (1870–1943, figure 4),
either directly or through Rice’s brother-in-law Frank
J. Silva as a conduit, for a price of 75 cents per share.34
Rice then began to market the shares via his weekly
promotional circular the Wall Street Iconoclast,35
which was sent to a mailing list of 600,000 recip-
Figure 4. George Graham Rice was one of the most in- ients.36 Rice was one of the most infamous financial
famous financial manipulators in the United States manipulators of the era, having already spent time in
during the first half of the twentieth century. In the
prison on multiple occasions for various crimes of
late 1920s, his activities centered on aggressively pro-
fraud and theft.37
moting and selling shares in two entities, the Colom-
bia Emerald Development Corporation, which owned Rice’s tactics are exemplified by two issues of the
the Chivor mining titles, and the Idaho Copper Com- Wall Street Iconoclast we were able to locate from
pany. He worked in cooperation with the president of July and August 1926.38 In those issues, Rice pro-
the Colombia Emerald Development Corporation,
Edmund J. MacNamara, and its largest shareholder,
28
Frederick Lewisohn. The actions of Rice, MacNa- Harvard Alumni Bulletin 20 (1917), p. 440; American Mining & Met-
mara, and Lewisohn became the subject of court pro- allurgical Manual (1920), p. 89.
29
Brock, 1929.
ceedings in New York that began in 1926 and were 30
Ibid.
premised on allegations of fraud and stock market 31
Engineering and Mining Journal, 101, No. 15 (1916), p. 663; Engi-
manipulation. Photo 1911, from The Mining Investor,
neering and Mining Journal, 103, No. 11 (1917), p. 476; Brock, 1929.
Vol. 63, No. 1 (1911). 32
File William Burns, List of United States Citizens Arriving at Port of
New York, June 1926, Ancestry.com.
33
Colombia Emerald Development Corporation, The Story of Emeralds
(Undated), 10 pp. The map included appears to be prepared after Ca-
In the next stage of Chivor’s history, the owning nova (1921) but eliminates the other three mines shown on the origi-
entity became entangled in a stock market scandal nal version.
34
The Brooklyn Daily Eagle, January 6, 1927, p. 2; The New York
while operational activity was minimal. As to the Times, January 6, 1927; The New York Times, January 28, 1927; Engi-
latter, the first mining engineer sent to Chivor by the neering and Mining Journal, 123 (1927), p. 635.
new mine owners was William Burns (1885–1962), 35
For example, The Wall Street Iconoclast, 5, No. 102, July 15, 1926;
who had already been working for the Lewisohn group The Wall Street Iconoclast, 5, No. 105, August 5, 1926.
36
in various functions (e.g., as manager of the Santa Fe The Pittsburgh Press, February 28, 1928, p. 33.
37
Rice, 1913; The Scranton Republican, August 28, 1926, p. 15; Pla-
Gold & Copper Mining Company or as secretary of zak, 2006; Thornton, 2015.
the Rosemont Copper Company).28 Burns reportedly 38
The Wall Street Iconoclast, 5, No. 102, July 15, 1926; The Wall Street
had also performed development work in the 1920s at Iconoclast, 5, No. 105, August 5, 1926.

236 HISTORY OF CHIVOR, PART II: 1924–1970 GEMS & GEMOLOGY SUMMER 2020
moted the Colombia Emerald Development Corpo- mines.”50 A related investigation by Deputy Attor-
ration, as well as another Lewisohn entity sold on ney General Keyes Winter led him to conclude “that
the Boston Curb Exchange, the Idaho Copper Com- the total value of the output of the mines for two
pany.39 Rice aggressively recommended buying years did not exceed $13,000, and that many of the
shares of the Lewisohn-controlled Colombia Emerald emeralds obtained were in an uncrystallized state
Development Corporation, reporting that the Lewi- and could not be marketed.”51 Such numbers con-
sohn family had acquired the mines in the winter of trasted blatantly with claims proffered to promote
1924/1925. He also identified MacNamara as pres- sale of the stock, which indicated that emeralds of
ident of the company, emphasizing MacNamara’s fabulous quality weighing between 28 and 80 carats,
dual role as vice president of the Lewisohn Explora- with a value between $100 and $500 per carat, had
tion & Development Corporation.40 Rice likewise been found at Chivor.52
played upon the expertise of mining engineers Burns A preliminary injunction restraining promotion,
and Mentzel. Notably, it has been reported that advertising, and sale of Colombia Emerald Devel-
Burns’s resignation was prompted when he “learned opment Corporation stock had been issued by early
that Rice was selling the stock and using his name January 1927. It was then made permanent in a ruling
as bait.” Mentzel, prior to his departure for Colombia handed down on or before February 19, 1927, in the
and without having seen the mine, had already New York Supreme Court (i.e., the trial-level state
written a report in the Wall Street Iconoclast that court).53 Rice and associates were thereby restrained
valued the property at $5,000,000.41 One of Rice’s “from the sale of the stock of the emerald corporation
strategies was thus to use the reputation of an indus- unless complete information about the concern is fur-
trial family well known in the mining business— nished,”54 and not a single share of the stock could be
Lewisohn—to incentivize investment.42 sold without giving a detailed explanation of the ac-
An additional strategy was found in Rice’s glowing tual profit and losses from mining operations.55 Al-
portrayal of emerald production and valuation. The
Wall Street Iconoclast reported extraordinarily rich
finds such as 23,500 carats in June 192643 and cal- 39
The Wall Street Iconoclast, 5, No. 102, July 15, 1926; The Wall Street
culated sensationally high values for the mining prop- Iconoclast, 5, No. 105, August 5, 1926. The Idaho Copper Company
erty and stock, claiming that a share “had a real value was also known as the Idaho Copper Corporation. Thornton, 2015.
40
of between $50 and $75” and that the mines owned Most likely a reference to the Lewisohn Exploration & Mining Co. or
to the General Development Co.
by the company “were valued at $100,000,000 and 41
Brock, 1929.
were capable of producing an annual income for in- 42
Engineering and Mining Journal, 122, No. 16 (1926), p. 602.
vestors of between $2,000,000 and $5,000,000.”44 The 43
The Wall Street Iconoclast, 5, No. 102, July 15, 1926; The Wall Street
stock of the Colombia Emerald Development Corpo- Iconoclast, 5, No. 105, August 5, 1926.
ration rose over a period of 14 weeks from $1 to 44
The New York Times, February 20, 1927; Chicago Tribune, February
27, 1927, p. 40.
$17 5⁄ 8.45 In reality, however, half of the 23,500 carats 45
St. Louis Post-Dispatch, January 6, 1927; The Pittsburgh Press, Feb-
highlighted by Rice were worthless stones, and the ruary 28, 1928.
remainder had an estimated value of 50 cents per 46
Brock, 1929.
carat.46 The company reported a loss of $57,000 for 47
The Cumulative Daily Digest of Corporation News, No. 2 (1927), p.
1926.47 As one mining engineer who worked under 165.
48
the Lewisohns at Chivor would later summarize: “I The Billings Gazette, May 16, 1954, p. 6.
49
The New York Times, August 7, 1926; The New York Times, October
guess they [the Lewisohns] made more money on the 12, 1926.
stock market than mining.”48 50
The Brooklyn Daily Eagle, January 6, 1927, p. 2.
Such discrepancies began to come to light as a re- 51
The New York Times, February 20, 1927; Chicago Tribune, February
sult of actions beginning in 1926 by New York State 27, 1927.
52
Attorney General Albert Ottinger. Under the Martin The New York Times, February 20, 1927; Chicago Tribune, February
27, 1927.
Act (New York General Business Law, Article 23-A, 53
The Brooklyn Daily Eagle, January 6, 1927, p. 2; The Brooklyn Daily
Sections 352–353), he targeted the Wall Street Icon- Eagle, January 9, 1927, p. 22; The New York Times, January 9, 1927;
oclast, Rice, Silva, the Colombia Emerald Devel- The Bridgeport Telegram, January 10, 1927, p. 5; The Tribune (Co-
opment Corporation, MacNamara, and Frederick shocton, Ohio), January 10, 1927, p. 4; The New York Times, January
28, 1927; The New York Times, February 20, 1927; Jewelers’ Circular,
Lewisohn.49 In injunction proceedings, Ottinger 94, No. 4 (1927), p. 61; Chicago Tribune, February 27, 1927, p. 40.
claimed that “not over a spoonful of stones and only 54
Chicago Tribune, February 27, 1927.
two of those salable, have ever been taken from the 55
The New York Times, February 20, 1927.

HISTORY OF CHIVOR, PART II: 1924–1970 GEMS & GEMOLOGY SUMMER 2020 237
Figure 5. By 1927, South
African mining engi-
neer Peter W. Rainier
had been hired to head
operations at the Chi-
vor emerald mine in
Colombia, with the aim
of turning the venture
into a profitable one.
However, a drastic loss
generated in 1930,
when mining costs far
exceeded emerald sales,
led to a directive in
1931 to close the mine.
Photo from the Peter W.
Rainier collection; cour-
tesy of GIA.

though the injunction was later vacated in part to the the alternative course of simply closing the mine in-
extent that it applied to the Colombia Emerald De- advisable. The engineer selected as Mentzel’s succes-
velopment Corporation, Frederick Lewisohn, and sor, apparently under an initial contract of six
MacNamara, on account of insufficient evidence to months,60 was Rainier (figure 5). Born in Transvaal,
establish the actual personal participation of Lewi- South Africa, Rainier had obtained mining experi-
sohn and MacNamara in fraudulent practices, the re- ence in countries such as Nigeria, Southern Rhodesia
straint against the Wall Street Iconoclast, Rice, and (Zimbabwe), and South West Africa (Namibia).61 At
Silva remained in force.56 the time of his engagement, Rainier was living in the
Rice ultimately avoided serious prosecution in United States with his family. Mentzel did not, ho-
connection with his dealings related to the Colombia wever, immediately leave Colombia and remained
Emerald Development Corporation. Still, matters active in the country during 1928 and 1929, presum-
pertaining to the Idaho Copper Company proceeded ably still with the Colombia Emerald Development
to trial and led to him being found guilty in 1928 and Corporation in some capacity.62
sentenced to four years in prison.57 As an anecdotal Rainier began his work in Colombia in 1927, leav-
footnote, while incarcerated at the federal peniten- ing his family behind in the United States when his
tiary in Atlanta on those charges, Rice became a son Peter W. Rainier, Jr. (born August 1926), was only
friend of another prominent inmate, Al Capone.58 a few months old.63 His wife and children would later
By the end of 1928, both the Colombia Emerald join him at Chivor in 1928. Efforts at the mine started
Development Corporation and the Idaho Copper with repairing the infrastructure, after which Rainier
Company had been delisted from the Boston Curb
Exchange.59
56
The New York Times, May 7, 1927.
CHIVOR UNDER PETER W. RAINIER 57
Plazak, 2006.
(1927–1931) 58
Ibid.
59
Star Tribune (Minneapolis), December 18, 1928.
During late 1926 or early 1927, the president and di- 60
Rainier, 1942.
rectors of the Colombia Emerald Development Cor- 61
Weldon et al., 2016. Insofar as Weldon et al. (2016) chronicled
poration decided to hire a new mining engineer to Peter W. Rainier’s account as presented in Green Fire (1942), the fol-
oversee on-site operations at the Chivor mine in Co- lowing discussion will focus more broadly on the chronology as it re-
lombia, replacing Mentzel. The mine had been un- lates to independent documentation.
62
File Charles Mentzel, List or Manifest of United States Citizens
profitable for the preceding two years, but the Arriving at Port of New York, February 1928 and February 1929,
ongoing legal proceedings in New York involving Ancestry.com.
MacNamara and Frederick Lewisohn likely rendered 63
Rainier, 1942; Peter W. Rainier, Jr., pers. comm. 2017.

238 HISTORY OF CHIVOR, PART II: 1924–1970 GEMS & GEMOLOGY SUMMER 2020
was able to improve the yield of operations and the
quantity of facet-grade gem material. At that time, the
traditional open-cast step-cutting technique (figures 6
and 7) continued to be employed. Small terraces or
steps were cut in the extremely steep mountain sur-
face, and the loose debris was washed away by a flood
of previously collected water. Crystals were carefully
removed by hand from the emerald-bearing veins or
cavities within the host rock.
In 1929, Rainier published a study describing his
mining operations at Chivor, and he traveled to New
York to present the paper at a conference of the
American Institute of Mining and Metallurgical
Engineers in February 1930.64 It is also logical to sur-
mise that the trip to New York would have provided
an opportunity to meet with his employer concern-
ing the mining activities.
After Rainier returned to Colombia, he was joined
at Chivor by geologist Verner A. Gilles (1886–1954)
from April to August 1930 for geological mapping
and evaluation of the mine.65 Like many others sent
to Chivor, Gilles had experience working for com-
panies owned by the Lewisohn family.66 His eval-
uation resulted in a report, dated September 20, 1930,
that provided insight into the production and cost of
mining.67 In the period from 1925 to 1929, the mine
yielded 137,000 carats in total, beginning from a low
of 4,000 carats in 1925, the first year of mining oper- Figure 7. The open-cast step-mining operations at
Chivor under Rainier entailed manually excavating
the terraces and then recovering crystals after expo-
Figure 6. In the late 1920s, mining operations at Chi- sure in the host rock. Photo from the Peter W. Rainier
vor under Rainier still used the traditional open-cast collection; courtesy of GIA.
step-cutting technique. Terraces or steps were cut into
the surface of the extremely steep mountain slopes,
and the debris was washed away via water previously ations under the Colombia Emerald Development
collected in tanks. Photo from the Peter W. Rainier Corporation, followed by a large quantity of low-
collection; courtesy of GIA. grade stones in 1926. The costs of mining operations
from 1926 to 1929 aggregated $278,000, and it was
noted that a maximum of 250 workers were on the
payroll of the mine (figure 8). The total selling value
of the emeralds produced from 1925 to 1929 was cal-

64
Rainier, 1929, 1931. The study was initially published in 1929 and
was then reprinted in expanded form in 1931, incorporating com-
ments generated during the discussion that followed oral presentation
of the paper before the American Institute of Mining and Metallurgi-
cal Engineers in February 1930 in New York. A Spanish translation
was published in 1934. Rainier, 1934. See also The Wilkes-Barre
Record, March 21, 1930, p. 2.
65
Verner A. Gilles left Colombia on August 31 and returned to New
York on September 10, 1930. File Verner Gilles, List of United States
Citizens Arriving at Port of New York, September 1930, Ancestry.com.
66
The Billings Gazette, May 16, 1954, p. 6.
67
Gilles, 1930.

HISTORY OF CHIVOR, PART II: 1924–1970 GEMS & GEMOLOGY SUMMER 2020 239
Figure 8. In the late
1920s to early 1930s,
Rainer worked with
several hundred miners
at Chivor. Photo sent
by Rainier to his sister,
with a handwritten
comment reading: “In-
dian miners drinking
their Guarapo ration
Sunday afternoon.”
Priaulx Rainier collec-
tion, Royal Academy of
Music, London.

culated at $274,000, indicating that the average price As these operational activities were taking place,
received per carat was approximately $2. Such figures in the background ownership and control of the
evidenced that the mine was not generally profitable, Chivor mines had become the subject of a legal dis-
with mining costs versus emerald sales value roughly pute, referred to as the Joaquín Daza case. Joaquín
equal at best. The numbers also offered a counter- Daza B. submitted a petition dated December 27,
point to the more optimistic estimation apparently 1929, to the Colombian government, claiming that
held by Rainier, who would later opine that he han- the Chivor mine partly overlapped with his prop-
dled “emeralds worth more than one million dollars” erty and that mining operations were being con-
during his years of responsibility over the mine.68 ducted on his land.74 He proposed that he be
Gilles’s report further addressed geological aspects awarded financial compensation for the infringe-
pertaining to the reserve of emerald-bearing rock at ment. According to Green Fire, Rainier rejected
Chivor, concluding:69 such a proposal after Daza personally visited the
It is almost impossible to give an estimate of the value mine with his demands.75
of this enormous deposit of emerald bearing rock. To In the more formal proceeding and at the request
sample the deposit is out of the question. Even drilling of the government, Rainier responded to Daza’s
would not give an adequate idea of its value, due to the
fact that the ratio between gangue rock and valuable
mineral contents is so great. There is but one way to
68
tell what the deposit is worth, and that is to work it Gilles, 1930.
69
out. Here is where the great hazard of emerald mining Rainier, 1942.
70
must be considered. Working out the emerald bearing Robert Elmer Sylvester received his degree from the School of Mines
formation now in reserve may mean a handsome profit and Metallurgy, Institute of Technology, University of Minnesota, in
and on the other hand it may mean a great loss. After 1927. He married in July 1930 and arrived at Chivor with his wife
all is said and done the question of profit in this case of Josephine at the end of August 1930. Alumni Blue Book, School of
Mines and Metallurgy, Institute of Technology, University of Minne-
mining is largely a matter of luck. sota (1950), 13, p. 135; Rebecca Toov, University of Minnesota Ar-
chives, pers. comm. 2018.
As previously noted, Gilles departed from Colom- 71
The Bismarck Tribune, September 9, 1930, p. 3.
bia in late August 1930, approximately concomitant 72
The Billings Gazette, July 20, 1930, p. 13; Minnesota Chats, January
with the arrival of American mining engineer Robert 2, 1931, p. 3.
Elmer Sylvester (1904–1988, figure 9) to assist Rain- 73
File Robert Sylvester, List or Manifest of Alien Passengers for the
ier.70 Sylvester’s contract had come through the Lew- United States Immigration Officer at the Port of Arrival, New York,
October 1931, MyHeritage.com.
isohn group of companies,71 and a two-year stay in 74
File “Joaquín Daza B.,” Volume “Propuestas Minas 99,” Ministerio
Colombia was planned.72 His tenure would actually de Industrias, Departamento de Minas y Petróleos, Archivo General
end, however, by October 1931 when he returned to de la Nación (Colombia).
75
New York.73 Rainier, 1942.

240 HISTORY OF CHIVOR, PART II: 1924–1970 GEMS & GEMOLOGY SUMMER 2020
claims via a letter dated August 4, 1930.76 Therein,
Rainier explained that he had no exact map of the
mine boundaries, as no such document had been pro-
vided by his employer (figure 10). In connection with
the case, the mine was inspected during the period
from late January to early February 1931 by two gov-
ernment officials, i.e., the lawyer Alfonso Torres Bar-
reto and the engineer Miguel Alvarez Uribe, together
with Daza, Rainier, and Sylvester, and a report was
issued on February 7, 1931. Yet no final resolution
was reached, and ongoing consideration by the gov-
ernment was reflected in documents dated in 1934,
1935, and 1939.77
The consideration apparently intensified in Janu-
ary 1941, when the mine was examined once again
by Colombian officials. A report by engineer Nicolas

Figure 9. American mining engineer Robert Elmer

Sylvester worked with Rainier at Chivor in 1930 and
1931. This photo was sent by Rainier to his sister,
with a handwritten comment reading: “My American
assistant R.E. Sylvester.” Priaulx Rainier collection,
Royal Academy of Music, London.

Figure 10. Letter dated August 4, 1930, written by

Rainier to the Ministerio de Hacienda in Bogotá, indi-
cating that he was not in possession of any maps de-
picting the exact boundaries of the Chivor mining
claims. File “Joaquín Daza B.,” Volume “Propuestas
Minas 99,” Ministerio de Industrias, Departamento
de Minas y Petróleos, Archivo General de la Nación
(Colombia).

Rosso, dated January 29, 1941,78 incorporated a map

showing currently mined sites, estimated mine
boundaries, and relevant infrastructure (figure 11).
Discussions between Rosso and Daza in Guateque
had preceded the inspection, and Daza then reiter-
ated his claims in letters submitted to the govern-
ment dated from August to October 1941, wherein
he also augmented his petition to request considera-
tion of titles to the adjacent mines of Buenavista and
Mundo Nuevo (figure 12; see also figure 25). Even

76
File “Joaquín Daza B.,” Volume “Propuestas Minas 99,” Ministerio
de Industrias, Departamento de Minas y Petróleos, Archivo General
de la Nación (Colombia).
77
Ibid.
78
Ibid.

HISTORY OF CHIVOR, PART II: 1924–1970 GEMS & GEMOLOGY SUMMER 2020 241
Esperanza La Paz
Path to Almeida de Chivor de Chivor
Figure 11. This map, which
has been recreated based
on the original hand-drawn
map, shows the boundaries
of the Chivor 1 and 2
mining claims registered in
1889, as well as the sites of
the Esperanza de Chivor
and La Paz de Chivor
Chivor No. 2 claims registered in 1904.
These mines, referred to as
part of the Chivor group,
Quebrada were all located between
Sinai the Río Sinai and the Río
Rucio (sometimes des-
Water ignated as Quebrada Sinai
Spanish and Quebrada de las
channels
tank
Minas, quebrada meaning
gorge). The map also de-
picts workings, water chan-
Chivor No.1 nels, and tanks existing in
Río
1941. South of the Río
Rucio Rucio, operations at the
Water Buenavista emerald mine
channels
are indicated. The map was
incorporated in a report by
the engineer Nicolas Rosso,
Camp dated January 29, 1941, to
Tanks the Ministerio de Minas y
N
Quebrada
Petróleos. File “Joaquín
Sinai
Daza B.,” Volume “Pro-
puestas Minas 99,” Ministe-
rio de Industrias,
Río Rucio
Departamento de Minas y
Petróleos, Archivo General
Mine Buenavista Río Guavio de la Nación (Colombia).
Path Spanish
channel

well into the 1950s, government documentation Development Corporation. From a personal stand-
showed that the matter remained open, with the last point, Rainier and his wife Margaret acquired a ha-
recorded activity in the case dated to October 1958 cienda known as “Las Cascadas” south of Río Rucio
and with no known definitive solution or final in the Departamento de Cundinamarca. The family
result.79 Such a scenario contrasted markedly with moved there from Chivor in 1930 or 1931, when his
Rainier’s recollection in Green Fire, where he repre- son Peter W. Rainier, Jr., was four.81 Rainier was also
sented that shortly after the 1931 examination of the
mine’s boundaries, he received a telegram from the
79
company’s agent in Bogotá declaring: “Government Diario Oficial (Colombia), 96, No. 30056 (1959), p. 597.
80
Rainier, 1942
decided Joaquín’s claim groundless.”80 81
Peter W. Rainier, Jr., pers. comm. 2017 (“I moved to Chivor when I
Meanwhile, during the 1930 to 1931 period, both was two and my father was already there. I moved away from Chivor
personal circumstances involving Rainier and hostil- when I was four, so assume that the mine was shut down at that time.
ities at the mine property likely impacted operations, That was when my father started Las Cascadas.”). See also the inter-
view of Peter W. Rainier Jr., by GIA staff, 2015, available at
leading to possibly multiple temporary reductions or https://www.gia.edu/gems-gemology/summer-2016-rainier-footsteps-
stoppages of work on behalf of the Colombia Emerald journey-chivor-emerald-mine.

242 HISTORY OF CHIVOR, PART II: 1924–1970 GEMS & GEMOLOGY SUMMER 2020
Figure 12. Letter from Joaquín Daza to the Ministro Figure 13. Christopher Ernest Dixon, business entrepre-
de Minas y Petróleos, dated September 1941, concern- neur and emerald mining expert, worked in the 1890s for
ing his claims that the Chivor mine partly overlapped an English company then operating the Muzo emerald
with his property and that mining operations were mine. In the first decade of the twentieth century, he be-
being conducted on his land. File “Joaquín Daza B.,” came one of the shareholders in the Compañía de las
Volume “Propuestas Minas 99,” Ministerio de Indus- Minas de Esmeraldas de Chivor that owned the mining
trias, Departamento de Minas y Petróleos, Archivo titles for Chivor 1 and 2, and he also co-owned additional
General de la Nación (Colombia). claims in the area. After Chivor was sold in 1919 to the
US-based Colombian Emerald Syndicate, Ltd., he served
as its official representative in Colombia in the early
1920s. Then, in the early 1930s, Dixon assisted Rainier in
involved, at least initially, in establishing operations efforts to combat bandits at Chivor. Photo circa early
at the hacienda. Nonetheless, he remained active as 1930s; courtesy of Catalina and Melissa Dixon.
mine manager over Chivor, assisted by Sylvester,
during the January and February 1931 developments
in the Joaquín Daza case.82 (figure 14)84 and included a handwritten notation on
Regarding hostile conditions at the mine, security the back (presumably by Rainier) that read: “Our gang
concerns would appear to have arisen at some point
after the February 1931 inspection in the Joaquín
Daza case, when the property became the target of re- 82
File “Joaquín Daza B.,” Volume “Propuestas Minas 99,” Ministerio
peated attacks from bandits. Peter W. Rainier, Jr., re- de Industrias, Departamento de Minas y Petróleos, Archivo General
de la Nación (Colombia).
membered: “After he [Peter W. Rainier, Sr.] got Las 83
Interview of Peter W. Rainier, Jr., by GIA staff, 2015, available at
Cascadas in operation then he went back to Chivor https://www.gia.edu/gems-gemology/summer-2016-rainier-footsteps-
and recaptured Chivor [from bandits] with his friends journey-chivor-emerald-mine.
84
Chris Dixon and his sons and a couple of other guys Robert Elmer Sylvester was identified through assistance from Re-
becca Toov, University of Minnesota Archives, pers. comm. 2018, and
and operated the mine for a while.”83 Similarly, a pho- comparison with the photo of Sylvester, dated 1927, published in The
tograph from that era showed Rainier, Christopher Er- Gopher, 40 (1927), p. 443 (electronic p. 463 at https://umedia.lib.
nest Dixon (figure 13), Dixon’s sons, and Sylvester umn.edu/item/p16022coll339:27309/p16022coll339:27197).

HISTORY OF CHIVOR, PART II: 1924–1970 GEMS & GEMOLOGY SUMMER 2020 243
Figure 14. Robert Sylvester was hired in 1930 to assist Rainier in the operations at Chivor. Together with Rain-
ier and Christopher Ernest Dixon, Sylvester was involved in the efforts to combat bandits who had invaded
Chivor in 1931. Left: Detail of a photo from The Gopher, Vol. 40 (1927), p. 443 (electronic page 463 at
https://umedia.lib.umn.edu/item/p16022coll339:27309/p16022coll339:27197); courtesy Rebecca Toov, University
of Minnesota Archives, Minneapolis. Right: Photo with a handwritten comment (not shown) reading: “Our
gang at Chivor during the bandit fighting days.” Shown from the left are Sylvester, MacFadden, Dixon (flanked
by his sons), and Rainier. Photo from the Peter W. Rainier collection, courtesy of GIA.

at Chivor during the bandit fighting days.” That photo a short period due to renewed hostilities from the
must have been taken before the close of October bandits. Rainier’s account of this late 1931 to mid-
1931, as that was when Sylvester left Colombia.85 1932 timeframe, however, has not been corroborated
Final accounting had revealed a drastic loss of by independent evidence, and only the pre-October
$65,000 for the year 1930, calculated based on the ac- 1931 bandit fighting has been verified by the histori-
tual costs of operations reduced by emerald sales cal record.
worth only $48,000.86 Coupled with the negative as- In a report dated November 2, 1931, the United
sessment in the recently prepared report by Gilles, the States Department of Commerce noted, “The Chivor
leaders of the Colombia Emerald Development Cor- emerald mine operated by American interests has
poration instructed Rainier to close the mine and to been closed, owing to depressed prices.”90 It was sub-
return to New York.87 As directed, Rainier traveled to sequently reported that Chivor was closed again in
New York in October 1931 for consultation with the October 1932,91 but no documentation detailing the
mine owners,88 leaving his family at Las Cascadas.
According to Rainier in Green Fire, he returned
85
to Colombia with preliminary permission from the File Robert Sylvester, List or Manifest of Alien Passengers for the
United States Immigration Officer at the Port of Arrival, New York, Oc-
company to continue mining operations at Chivor tober 1931, MyHeritage.com. To the extent that Rainier (1942) is sus-
under his own responsibility.89 He went on to re- ceptible to an interpretation that would place the bandit fighting solely
count that with the help of Dixon and his sons, the after October 1931, such is not supported by the historical record.
86
men “bombed out” the bandits who had overtaken Engineering and Mining Journal, 131 (1931), p. 485.
87
Rainier, 1942.
the mine during Rainier’s absence. Rainier and Dixon 88
File Peter Rainier, List or Manifest of Alien Passengers for the United
then collaborated in mining the deposit for a few States Immigration Officer at the Port of Arrival, New York, October
months, but the Colombia Emerald Development 1931, MyHeritage.com. The travel documentation gave Gachalá (in-
Corporation soon withdrew its permission and sent terestingly not Chivor) as Rainier’s residence in Colombia and indi-
cated that the purpose of the entry was to visit the Colombia Emerald
new staff to run the mine. Rainier and Dixon had to Development Corporation for a planned stay of two months.
hand over operations to the new leaders sent by the 89
Rainier, 1942.
owners and leave Chivor. Yet according to Rainier, 90
Commerce Reports, 4, No. 44 (1931), p. 247.
the new operators again lost control of the mine after 91
Mineral Trade Notes, 2, No. 4 (1936), p. 18.

244 HISTORY OF CHIVOR, PART II: 1924–1970 GEMS & GEMOLOGY SUMMER 2020
28, 1933, the name of the corporate entity was for-
mally changed from Colombia Emerald Devel-
opment Corporation to Chivor Emerald Mines,
Inc.100 The name change was registered in Colombia
in November 1934.101 Corporate headquarters con-
Figure 15. Detail from the final page of a letter dated
April 1944 sent by Rainier to his sister Priaulx Rain-
tinued to be in New York City, and company infor-
ier, discussing a possible return to Chivor as a con- mation under the new name was officially filed with
sulting engineer. Priaulx Rainier collection, Royal the New York State, Department of State, Division
Academy of Music, London. of Corporations, on November 23, 1935.102 Sub-
sequent documentation from 1936 to 1940 showed
that Henry was by then serving as president and Pace
as secretary and/or treasurer.103
mode of operation or the person(s) in charge of the
In a renewed push to regain effective control, a na-
mine in 1932 is available. Rainier remained in Co-
tional police squadron was dispatched to Chivor in
lombia, working as a mining consultant, e.g., for the
1936 to secure the mine pursuant to a contract be-
Colombian government at Muzo and Coscuez in
tween the government and Chivor Emerald Mines,
1933 and 1934.92 After the death of his wife Margaret
in 1937, he left the country.93
Rainier initially settled in Egypt for health rea- 92
Rainier, 1933a, 1933b; Minerals Yearbook (1934), pp. 1095–1096;
sons and then joined the British Eighth Army after Diario Oficial (Colombia), 70, No. 22562 (1934), p. 140. Certain of
the outbreak of World War II, serving from June these projects were undertaken in cooperation with Christopher Dixon.
93
1940 to June 1943. During this period, Rainier was Peter W. Rainier, Jr., pers. comm. 2017.
94
also engaged in writing several books, including Published works included African Hazard (1940), American Hazard
(1942), and Pipeline to Battle: An Engineer’s Adventures with the Brit-
Green Fire (1942), all in the genre of regaling readers ish Army (1943). Additional book manuscripts remained unfinished.
with his adventures.94 Following his retirement Letter by Peter Rainier, March 25, 1943, to his sister Priaulx Rainier,
from the British Army, Rainier traveled in the Priaulx Rainier collection, Royal Academy of Music, London.
95
Letter by Peter Rainier, April 25, 1944, to his sister Priaulx Rainier,
United States, lecturing about his escapades fight- Priaulx Rainier collection, Royal Academy of Music, London.
ing in Africa during the war. Speaking engagements 96
The New York Times, July 7, 1945.
in South America were planned as well, and he even 97
Diario Oficial (Colombia), 69, No. 22208 (1933), p. 353.
considered taking a position as a consulting engi- 98
Rainier, 1942; The Ottawa Journal, November 28, 1942, p. 21. Peter
neer at Chivor when contacted in 1944 by the mine Rainier’s story also provided the basis for a novel written by the Co-
lombian author Torres Neira in 1967, using fictional characters styled
owners about a potential attempt to reopen (figure after the real players (e.g., Rainer became Garnier).
15).95 Rainier died in a hotel fire in Ontario, Canada, 99
Poor’s Financial Records, Industrial Manual, New York (1931), p. 788;
in July 1945.96 Poor’s Register of Directors of the United States and Canada (1932), p.
257. Francis Philip Pace was born in Canada and moved with his par-
ents to Buffalo, New York, in 1880. Pace was naturalized in 1890. Al-
CHIVOR IN DORMANCY AND TRANSITION though his father was a stone cutter, Pace studied law and received a
degree from the New York University School of Law. He went on to
(1932–1947) work in New York as a lawyer and was involved with multiple corpo-
Following closure of the mine, Chivor entered a rate boards. In his engagement with the Colombia Emerald Devel-
period of dormancy from an operational standpoint, opment Corporation and, after its renaming, Chivor Emerald Mines,
Inc., Pace was supported for several years by his son Brice Pace (1915–
at least insofar as concerned official work on behalf 2001), who held a degree in geology from New York University and
of the Colombia Emerald Development Corporation. sought to assist his father and the company as a member of the board
Illegal mining and theft, however, remained a con- of directors in the late 1930s. Tenth Census of the United States, 1880;
The Albany Law Journal, 59 (1899), p. 484; New York Herald, August 5,
stant feature.97 The interest of the press and public 1916, p. 8; Fourteenth Census of the United States, 1920; Gregory,
decreased and was only revived briefly by the release 1995.
100
of Green Fire in 1942.98 State of Delaware, Department of State, Division of Corporations,
Entity Details; The Morning News, September 29, 1933, p. 23; The
From a corporate and administrative perspective, News Journal, September 29, 1933, p. 29.
documentation from the 1931 to 1932 period re- 101
Escritura 2045, November 17, 1934, Notaria 1, Archivo General de
flected that MacNamara remained president of the la Nación (Colombia).
102
Colombia Emerald Development Corporation and New York State Department of State, Division of Corporations, En-
tity Information.
that directors at the time included MacNamara, the 103
The Newark Post, November 26, 1936, p. 2; Who’s Who in New
banker Robert E. Henry, and the attorney Francis York City and State, 10 (1938), p. 599; Poor’s Register of Directors and
Philip Pace (1876–1957, figure 16).99 On September Executives, United States and Canada (1940), p. 960.

HISTORY OF CHIVOR, PART II: 1924–1970 GEMS & GEMOLOGY SUMMER 2020 245
less.”107 Later, as a consequence of the economic and
political situation engendered by World War II, the
mine again became inoperative until 1947,108 with the
exception of the ever-resilient illegal and uncontrolled
mining, and even gun fighting, at the unguarded
site.109 Although the possibility of reopening may have
been broached during the interim, such as by contact-
ing Rainier in 1944 (again, see figure 15), such efforts
went nowhere. After Walter J. Cowan served as cor-
porate president in the early to mid-1940s,110 Pace took
over that position in 1947.111

CHIVOR UNDER RENEWED OPERATIONS

(1947–1951)
After the hostilities of World War II drew to a close,
the Chivor Emerald Mines, Inc., leadership endeav-
ored to restart active mining operations. Pace always
hoped that it would be possible to turn Chivor into a
profitable mining site.112 In service of that aim, and
to address creditor issues, he traveled to Colombia at
least three times, in 1947, 1948, and 1950.113 Opera-
tional control in Colombia was placed under a series
Figure 16. New York attorney Francis Philip Pace be- of managers. James W. Raisbeck Jr., an American,
came associated with the Colombia Emerald Devel- began in July 1947,114 and was succeeded by Luis
opment Corporation by the early 1930s and Salómon, an American, and Jacques Meyer, a Ger-
continued to serve in multiple roles both under that
name and as renamed Chivor Emerald Mines, Inc., in
1933. He initially filled positions as a member of the 104
Diario Oficial (Colombia), 72, No. 23219 (1936), pp. 697–698.
board of directors, treasurer, and secretary, eventually 105
Minerals Yearbook (1938), p. 1295; Minerals Yearbook (1939), p.
becoming one of the largest shareholders as well as 1393; Minerals Yearbook (1940), p. 1461; Minerals Yearbook (1941),
president of the company in 1947. In the late 1940s, p. 1409; Las Industrias Mineras y Manufactureras en Colombia, trans-
Pace traveled to Colombia to restart active mining lation of Mining and Manufacturing Industries, United States Tariff
operations, always in hopes of turning the company Commission, Publication TC-250 (1949), pp. 30–31.
106
into a profitable enterprise. Photo circa 1930s; cour- File “Joaquín Daza B.,” Volume “Propuestas Minas 99,” Ministerio
de Industrias, Departamento de Minas y Petróleos, Archivo General
tesy of Scott Pace. de la Nación (Colombia).
107
Robert D. Fisher Manual of Valuable and Worthless Securities (1944),
p. 147.
108
Minerals Yearbook (1949), p. 1579.
Inc.104 As a result, after several relatively quiet years 109
Mineral Trade Notes, 25, No. 2 (1947), pp. 28–29; Gaceta Judicial
with respect to mine operations, the mine was able (Colombia), 72 (1952), pp. 282–290.
to be leased from 1937 to 1940 (or even later) to the 110
Robert D. Fisher Manual of Valuable and Worthless Securities, 11
Compañía de Esmeraldas de Colombia, which em- (1946), p. 149; Diario Oficial (Colombia), 83, No. 26422 (1947), pp.
454–456.
ployed about 100 workers.105 In January 1941, the 111
Diario Oficial (Colombia), 83, No. 26422 (1947), pp. 454–456;
mine was being operated with Ernesto Fernández as Diario Oficial (Colombia), 83, No. 26436 (1947), p. 657; Diario Ofi-
the mine administrator.106 cial (Colombia), 83, No. 26672 (1948), p. 1132; Morello, 1956;
Nonetheless, by 1943 the outlook of company Gaceta Judicial (Colombia), 97 (1961), pp. 138–145.
112
leadership was not optimistic, with the financial press Scott Pace (grandson of Francis Phillip Pace), pers. comm. 2018.
113
File Francis P. Pace, Passenger Manifests of Airplanes Arriving at
reporting: “Mr. Francis P. Pace, an attorney and treas- Miami, Florida, November 1947 and April 1948, Ancestry.com; Escri-
urer of the company informed us on Oct. 29, 1943, tura 2204, July 12, 1950, Notaria 3a, Archivo General de la Nación
that he has served the company for the last five years, (Colombia); Diario Oficial (Colombia), 87, No. 27372 (1950), pp.
367–368.
and has received nothing for his services. He stated 114
Escritura 1353, April 28, 1947, Notaria 3a, Archivo General de la
that he is the biggest stockholder and biggest creditor Nación (Colombia); Diario Oficial (Colombia), 83, No. 26422 (1947),
of the company. Mr. Pace considers the stock worth- pp. 454–456; Diario Oficial (Colombia), 83, No. 26436 (1947), p. 657.

246 HISTORY OF CHIVOR, PART II: 1924–1970 GEMS & GEMOLOGY SUMMER 2020
man, in December 1947.115 Because of the limited
funds available, the latter two instigated a contractor
system under which contracted foremen hired and
paid their own laborers.116 Salómon and Meyer were
replaced by Eric G. Ramsay and Antonio C. Cosme,
both United States citizens, in July 1950.117
Additional investment and involvement from Co-
lombian parties had also been brought on board by
1950, as reported in the trade press:118
Present-day exploitation of the emerald deposits is di-
rected by the two largest stockholders in the corpora-
tion, one of whom [Pace] has had many years’
association with the firm and whose ceaseless efforts
have salvaged a defunct corporation, with more or less
worthless stock, and has placed it back into production,
with the possibilities steadily increasing that the stock-
holders will at last begin to realize some return on their
investment. The other of those two men [identity un-
known to the authors] is a well-known Bogotá, Colom-
bia, importer and exporter, who is in charge of the
operations in Colombia. This man has himself invested
large sums of money in the property, and, in his position
on the spot, has been able to supervise the mining, ap-
praisal, registration, and exportation of the gem stones.

Such efforts resulted in production of 5,400 carats

in 1947, with yields increasing to 82,370 carats in
1948 and 91,656 carats in 1949.119 Although opera-
Figure 17. The American adventurer, gem dealer, and
tions were impacted in late 1948 due to rioting and
miner Russell W. Anderton took over mining opera-
theft by local workers, the situation was brought tions at Chivor in the early months of 1950. After an
under control after the company sent Carl (Carmine) initial term and a subsequent break, he returned to
Cavallo, an American civilian construction superin- Colombia at the end of 1952. In the second half of the
tendent, to Colombia to restore order at the mine.120 decade, Anderton shifted his attention to other explo-
For a period extending approximately from the ration projects in the Chivor and Gachalá areas.
first to the third quarter of 1950, the New York Photo circa late 1950s; courtesy of Manuel J. Marcial.
owners employed American gem hunter Anderton

(figure 17) to manage on-site operations at the mine.

115
Escritura 4313, December 17, 1947, Notaria 3a, Archivo General Anderton had just returned from gemstone ventures
de la Nación (Colombia); Diario Oficial (Colombia), 83, No. 26672
(1948), p. 1132. in Sri Lanka, having left Colombo on October 25,
116
Wehrle, 1980. 1949, and having arrived in New York on November
117
Escritura 2204, July 12, 1950, Notaria 3a, Archivo General de la 21.121
Nación (Colombia); Diario Oficial (Colombia), 87, No. 27372 (1950), During Anderton’s tenure, mining methods dif-
pp. 367–368.
118
Mineral Trade Notes, 30, No. 1 (1950), pp. 29–38.
fered from the practices under Klein and Rainier. Ac-
119
Minerals Yearbook (1949), p. 1579; Mineral Trade Notes, 31, No. 1 cording to Anderton,122 terraces were cut into the
(1950), pp. 31–32; Minerals Yearbook (1951), p. 551; Minerals Year- hillside only to the extent necessary to reveal the
book (1953), p. 556. structures beneath the overgrowth. If promising veins
120
Tampa Bay Times, December 22, 1952, p. 5. Carl (Carmine) Ca- were found, they were pursued exclusively by tunnel-
vallo apparently remained involved at Chivor in some capacity during
periods in 1951 and 1952, but details of his role are vague. See also ing. The actual mining was still accomplished through
File Carmine Cavallo, Manifesto de Pasajeros, Flight Bogotá – New the contractor system, with foreman hiring and paying
York, December 21, 1951, MyHeritage.com. their own laborers. The contractors identified the tar-
121
File Russell Anderton, List or Manifest of aliens employed on the
vessel as members of crew, November 1949, MyHeritage.com.
get areas they wished to tunnel and, upon approval
122
Anderton, 1950–1951. See also Lentz, 1951; Spence, 1958a, from management, would organize the work. The em-
1958b. eralds found were (theoretically) turned over to the

HISTORY OF CHIVOR, PART II: 1924–1970 GEMS & GEMOLOGY SUMMER 2020 247
mine administration, and the contractors received half of receiver (síndico secuestre or síndico de concurso),
the stones’ value as compensation. Thievery, however, and details of the legal parameters governing the in-
was rampant with this system. teraction between the receiver and company person-
Reported yield for 1950 amounted only to 7,177 nel are equally obscure. Anderton, however, did
carats,123 and by late in the year tensions had arisen apparently return to the mine in a management ca-
between the American and Colombian associates. As pacity in 1952 to 1953, after writing Tic-Polonga.132
Anderton described in a report published in the American press noted in 1952 that “Anderton is in
winter of 1950–1951: “At the present writing Chivor, New York now and plans to return to South America
for all practical purposes, is closed.... The year was soon in another effort to control and operate these
marked by a bitter struggle between the Colombian Mines.”133 After that brief period, Anderton turned
interests and the New York officers for control of the his attention to other prospecting and mining ven-
mine, a situation prevailing at the moment.”124 By tures in the Chivor and Gachalá regions. He re-
that point, in the midst of the mounting difficulties, mained in Colombia, but he gradually moved from
Anderton had been dismissed from his management mining operations to the gem trade in Bogotá.134
role at Chivor.125 From January 1955 to the end of September 1957,
In the last months of 1950 and the first months the position of receiver of Chivor Emerald Mines,
of 1951, the dispute between the American and Co- Inc., in Bogotá was held by Walter de Freitas,135 a Bra-
lombian associates had intensified to the point that zilian lawyer.136 He was assisted by the Swiss min-
trade press reported: “The American group sent eralogist Dr. Jean (Juan) Stouvenel.137 The mine
down a general representative to take charge of the inspector for Chivor throughout this period was
operations. However matters became worse instead Pedro Patiño Patiño (figure 18).138 As receiver, Frei-
of improving, and subsequently chaos and complete tas’s directive was to utilize company property for
lawlessness reigned the mine.”126 The company was the benefit of creditors. Late in his term, however, he
also in dire financial straits, owing more than
700,000 pesos (equivalent to about US$280,000), and
creditors had filed lawsuits. In August 1951, Chivor 123
Mineral Trade Notes, 33, No. 3 (1951), p. 36; Mineral Trade Notes,
Emerald Mines, Inc., hired John M. McGrath to ne- 35, No. 1 (1952), p. 35.
124
gotiate a financial settlement with the United States Anderton, 1950–1951.
125
and Colombian creditors, but no resolution was The byline for the Anderton (1950–1951) article identified the au-
thor as “formerly Manager of Mines at Chivor,” thus indicating a term
achieved. Attempts by the American mine owners that had ended by the winter 1950–1951 quarter.
to sell the mine to the Colombian government were 126
Mineral Trade Notes, 35, No. 1 (1952), p. 35.
rejected as well.127 Instead, the turmoil culminated 127
Domínguez, 1965.
by February 1952 in insolvency proceedings (con- 128
Memoria del Ministro de Minas y Petróleos al Congreso de 1960
(1960), pp. 20–21; Gaceta Judicial (Colombia), 97 (1961), pp. 138–
curso de acreedores) in civil court in Bogotá.128 Ef- 145; Domínguez, 1965. The American mine owners were repre-
forts to sell the mine to the government were also sented in Bogotá by Mauricio Makenzie.
reprised after the proceedings commenced, but the 129
Mineral Trade Notes, 35, No. 1 (1952), p. 35; Minerals Yearbook
offers were rebuffed once again. (1955), p. 440.
130
Minerals Yearbook (1954), p. 611.
During the tumultuous period surrounding the in- 131
Anderton, 1955, 1957.
solvency, the entity apparently suspended mining op- 132
Anderton, 1965.
erations from an official standpoint. Nonetheless, 133
The Tennessean, November 9, 1952, p. 21.
workers resumed some production in 1951, but all 134
Manuel J. Marcial de Gomar, pers. comm. 2017. Marcial de
stones went into the black market.129 With a degree of Gomar, a Florida-based jeweler, began working with Russell W. An-
derton in Colombia in 1955 as an interpreter, and their cooperation
irony, a trade publication commented: “Despite this continued in Anderton’s later enterprises.
trouble, it was reported that a new vein was found at 135
Gaceta Judicial (Colombia), 103–104 (1963), pp. 313–327.
Chivor, and the emeralds produced were said to be the 136
Walter de Freitas studied law at the Universidade Federal de Minas
best quality ever taken from the mine.”130 Gerais in Belo Horizonte, Minas Gerais, Brazil, and received a degree
in 1948. His ensuing career path saw him both working as a mining
attorney and serving as an editor for the Jurisprudência Mineira, a Bra-
CHIVOR UNDER RECEIVERSHIP (1952–1970) zilian publication focused on the mining industry and prepared under
the auspices of the judiciary in Minas Gerais.
Once the insolvency proceedings had been insti- 137
Martínez Fontes and Parodiz, 1949; Gaceta Judicial (Colombia),
gated, Chivor operated under receivership.131 The his- 103–104 (1963), pp. 313–327.
torical record is unclear as to who first filled the role 138
Gaceta Judicial (Colombia), 127 (1968), pp. 5–8.

248 HISTORY OF CHIVOR, PART II: 1924–1970 GEMS & GEMOLOGY SUMMER 2020
Figure 18. Pedro Patiño Patiño worked as a mine in- Figure 19. Chivor was operated under the American
spector at Chivor in the late 1950s, during the period Willis Frederick Bronkie for more than a decade, be-
when Walter de Freitas held the position of receiver of ginning in October 1957. During his tenure, he filled
Chivor Emerald Mines, Inc., in Bogotá. Photo after dual roles as mining engineer over Chivor and re-
2000; courtesy of the Patiño family. ceiver of Chivor Emerald Mines, Inc. Insolvency pro-
ceedings instigated under Colombian law in 1952 led
to a receivership that was eventually lifted in 1970.
Shown on the left of Bronkie is Renata de Jara, who
engaged in a transaction that would become mired worked for him during the 1960s in the Bogotá office.
in legal controversy. Freitas sold a portion of the em- Photo 1967; courtesy of Renata de Jara.
erald production near the end of September 1957 to
Walter Maurer, a German gem merchant. A sub-
sequent lawsuit questioned whether the sale had somewhat disjointed from that point. New York in-
been properly conducted before his final day in office, vestment banker Oliver John Troster (1894–1976)
while he was still authorized as receiver to dispose served as a director and secretary from 1957,142 and by
of company assets on behalf of creditors. The Corte the very early 1960s (or even earlier than that), George
Suprema de Justicia ultimately ruled in August 1961
that the transaction took place before Freitas left the
position at the close of September 1957, and the deci- 139
Gaceta Judicial (Colombia), 103–104 (1963), pp. 313–327.
sion cleared him of wrongdoing.139 140
Ibid.
In October 1957, Freitas was succeeded by Amer- 141
Gaceta Judicial (Colombia), 127 (1968), pp. 5–8. Pedro Patiño Patiño
ican Willis Frederick Bronkie (1912–1979, figure later challenged in a lawsuit whether he had been paid correctly.
142
19).140 Bronkie filled dual roles as mining engineer World Who’s Who in Commerce and Industry, 15th ed., 1967, pub-
lished 1968/69, p. 1391; Who’s Who in Finance and Industry, 16th
over Chivor operations and receiver of Chivor Emer- ed., 1969, published 1970/71, p. 703; Standard & Poor’s Register of
ald Mines, Inc. One week after he took office, on Oc- Corporations, Directors and Executives, 2 (1975), p. 1426; The New
York Times, October 5, 1976. Oliver John Troster had been a partner
tober 8, 1957, Bronkie dismissed Patiño.141 Earlier that in the New York–based company Troster, Singer & Co. since 1920,
year, in February 1957, Pace had died, and records re- and he apparently remained involved with Chivor Emerald Mines,
garding the corporate leadership in New York become Inc., until at least 1975.

HISTORY OF CHIVOR, PART II: 1924–1970 GEMS & GEMOLOGY SUMMER 2020 249
payroll. The practice of using modern equipment
such as an automatic compressor or bulldozers to re-
move the overburden, which had begun before he
took office, was also continued.148 Initially, work pro-
ceeded through simultaneous use of both step-cut
terracing (i.e., an open-pit technique) and tunneling
(i.e., an underground method following the emerald
veins).149 Step-cut terracing would be particularly
useful in steep areas where a bulldozer was unable to
operate. Tunneling work in the late 1950s involved
eight separately named areas, each with a length of
up to about 100 meters. For example, in April 1958,
96 men and nine contractors were working in two
tunnel areas, one of which yielded gemstones worth
over $100,000 from approximately two meters of the
emerald-bearing vein.150 However, no new tunnels
were dug, and from the mid-1960s to the end of the
decade, exclusively open-pit work was performed.151
During that mid-1960s period, 125 miners were
employed at Chivor,152 with a Spanish mine admin-

143
George Daniel Besler was the son of William George Besler (1865–
1942), the general manager and president of the Central Railroad of
New Jersey. George Daniel Besler received a degree in geology from
Princeton University in 1926, but his subsequent career focused on
industrial technology. He worked for the locomotive firm Davenport-
Besler Corporation, and in the 1930s he and his brother William John
Figure 20. In the 1960s, while Bronkie was leading op- Besler (1904–1985) developed high-pressure steam-powered engines
erations at Chivor in Colombia, George Daniel Besler for cars, locomotives, and the world’s first steam-driven airplane.
served as president of Chivor Emerald Mines, Inc., in Throughout his life, Besler held roles in several companies. Family
members recall that he was president of the company that owned the
New York. Besler held a degree in geology and
Chivor emerald mine in Colombia in the 1960s. Who’s Who in Com-
throughout his career served a number of companies merce and Industry, International Edition, 9 (1955), p. 99; Myers,
in various roles, focused primarily on the devel- 2000; Lynn G. Stewart, pers. comm. 2018.
opment of industrial technologies. Photo 1959; cour- 144
Renata de Jara, pers. comm. 2018. Renata de Jara served as Willis
tesy of Lynn G. Stewart. Frederick Bronkie’s secretary, typing and translating letters, and she
recalls typing correspondence from Bronkie to George Daniel Besler
as president from 1960 or 1961 onward. See also The National Ob-
server, February 19, 1968, p. 8, which refers to George Besler as the
president of Chivor Emerald Mines, Inc.
Daniel Besler (1902–1994, figure 20)143 of New York 145
Gaceta Judicial (Colombia), 97 (1961), pp. 138–145.
had stepped in as president of the company.144 146
Gonzalo Jara, pers. comm. 2018. In addition to improved physical
At the outset of his term, Bronkie faced substan- access, Willis Frederick Bronkie apparently also sought to augment
communications. He was represented in negotiations with the gov-
tial indebtedness to both Colombian and American ernment for a radio transmission license by the Chivor Emerald
creditors, as set forth in an accounting by the Corte Mines, Inc.’s legal agent in Colombia, Hernando Uribe Culla. Diario
Suprema de Justicia in 1961.145 At that time, the Chi- Oficial (Colombia), 98, No. 30571 (1961), p. 286; Diario Oficial (Co-
lombia), 102, No. 31888 (1966), p. 495.
vor mine was still accessible primarily by a one-day 147
Johnson, 1959, 1961.
journey on horseback from Guateque, but a road to 148
Anderton, 1957; Johnson, 1959, 1961.
the mine from the towns Chivor and Almeida was 149
Johnson, 1959, 1961.
constructed during Bronkie’s tenure.146 150
In the late 1950s, the Colombian government also named several
At the beginning, mining under Bronkie con- individuals Inspector of Emeralds in Chivor (Inspector de Esmeraldas
en Chivor), always under renewable three-month contracts. Diario
tinued to employ the system of contractors who Oficial (Colombia), 97, No. 30240 (1960), p. 581; Diario Oficial (Co-
hired and paid their own local workers.147 One in- lombia), 97, No. 30275 (1960), p. 116.
151
spector for each tunnel would be on the company’s Gonzalo Jara, pers. comm. 2018.
152
Anderton, 1965.

250 HISTORY OF CHIVOR, PART II: 1924–1970 GEMS & GEMOLOGY SUMMER 2020
the office in Bogotá and at the mine. He employed
Alfredo Sierra as his so-called second in command
and bodyguard. Life in that era was notoriously dan-
gerous, and once Bronkie was even shot in the back
at his house in Bogotá.157
Between 1963 and 1968, Bronkie’s business ap-
proach gradually transitioned from solely emerald
mining to additional related enterprises. The broader
focus incorporated dealing in and selling rough (fig-
ure 21) and cut (figure 22) stones to international
buyers as well as opening jewelry shops (e.g., in
Bogotá, Cartagena, Panama, Miami, and the Baha-
mas, figure 23).158 The increased need for rough and
especially cut emeralds to supply the jewelry stores
was satisfied by purchasing rough and even faceted
Figure 21. Sorting rough emeralds at the Chivor Emer- stones on the open market in Bogotá. As the 1960s
ald Mines, Inc., office in Bogotá. Photo circa 1965; drew to a close, the merchandising and retail busi-
courtesy of Renata de Jara. ness had overtaken the Chivor mining operations in
prominence for the company.159 By the end of the
decade, the mine employed only 25 workers.160
istrator, Evaristo Muñoz.153 Regardless of the methods Given the reduced contribution from mining to
used, success was extremely sporadic—mining could the company’s revenue-generating capacity, the lead-
go on for months without production of a single em- ership in New York began to contemplate “new
erald crystal of facetable quality.154 The value of the methods to accelerate production.”161 Specifically, in
emerald yield fluctuated on a yearly basis from as 1968 the press reported that Besler, as the company’s
high as $2,000,000 to as low as $200,000.155 In those president, had “sent an engineer to Bogota recently
years, the Colombian government oversaw the re- to convince Mr. Bronkie of the need to abandon even-
ceivership and production largely through the inspec- tually the laborious hand-picking method for a blast-
tor (Interventor de Salinas y Esmeraldas), Rafael A. ing-bulldozing operation that would increase the
Domínguez.156 daily soil turnover by tenfold.”162 Besler was quoted
Bronkie, as both receiver of the company and as saying: “After the receivership ends, we hope things
mining engineer for Chivor, worked alternately in will move along on a more progressive basis.”163

153
Figure 22. Certificate for a faceted emerald from Chi-
Alfonso Montenegro, pers. comm. 2018. Alfonso Montenegro, a
vor, signed in 1964 by Bronkie as Síndico Secuestre
miner, began working at Chivor in 1955 under Walter de Freitas, and
he continued to serve under Willis Frederick Bronkie and mine ad- (receiver). Photo courtesy of Lyon & Turnbull Ltd.,
ministrator Evaristo Muñoz. Fine Art, Antique & Jewellery Auctions.
154
The National Observer, February 19, 1968, p. 8.
155
Cochrane, 1970.
156
Memoria del Ministro de Minas y Petróleos al Congreso de 1961
(1961), pp. 305–308; Memoria del Ministro de Minas y Petróleos al
Congreso de 1963 (1963), pp. 28–31; Memoria del Ministro de Minas
y Petróleos al Congreso de 1964 (1964), pp. 212–214; Memoria del
Ministro de Minas y Petróleos al Congreso de 1965 (1966), pp. 195–
197.
157
The Morning Herald, April 14, 1964, p. 3; Renata de Jara, Gonzalo
Jara, and Robert E. Friedemann, pers. comms. 2018.
158
The National Observer, February 19, 1968, p. 8; Renata de Jara,
pers. comm. 2018; Marcial, 2018.
159
Renata de Jara, pers. comm. 2018.
160
Robert E. Friedemann, pers. comm. 2018.
161
The National Observer, February 19, 1968, p. 8.
162
Ibid.
163
Ibid.

HISTORY OF CHIVOR, PART II: 1924–1970 GEMS & GEMOLOGY SUMMER 2020 251
Figure 23. In the late 1960s, Bronkie’s business approach gradually transitioned from solely mining at Chivor to
entering the jewelry retail business. Flyer advertising shops in Freeport, Bahamas, undated; courtesy of Eduardo
Restrepo Ortega.

The latter point, i.e., the end of the receivership, turning control to the company’s 3,000 American
was apparently becoming a realistic possibility. The stockholders.”164 It was anticipated that when the re-
1968 article noted that Bronkie had “paid the debts ceivership ended, Bronkie would turn over control to
that forced the company into receivership 17 years
ago” and that he expected “soon to file documents,
asking a Colombian court to end the receivership, re- 164
Ibid.

252 HISTORY OF CHIVOR, PART II: 1924–1970 GEMS & GEMOLOGY SUMMER 2020
4
5 6 7 8
1
3 9
2

Figure 24. A view from the campsite of the Buenavista mine, looking in a north-northeast direction to the southern
bank of Río Rucio, shows the vicinity where the early indigenous tunnels were located, as well as the modern
mining areas Los Gavilanes (1), Acuario (2), Calichal (3), San Gregorio (4), Milenio (5), Porvenir (6), Catedral (7),
Klein & Cuatro (8), and El Pulpito (9), each represented by a mining symbol. Photo by G. Martayan, 2018.

a company representative recently sent to Colombia However, the Americans would soon step away from
from New York.165 the Colombian emerald venture. Although the
That anticipation came to fruition two years later. record as to details is beyond the scope of this study,
In 1970, the Colombian court overseeing the insol- Chivor was sold by Bronkie and/or the New York–
vency proceedings ruled that the company’s liabilities, based company representing the American share-
aggregating $330,000, had been satisfied. The receiv- holders in the early 1970s through a series of
ership was lifted, and control of Chivor Emerald transactions.170 Chivor was gradually taken over by
Mines, Inc., was returned to the entity’s New York of- Colombian interests.171
ficials.166 As of the year of the transition, Bronkie had At present, the principal mines in the Chivor
been replaced by Richard S. Pastore, an American law- area are operated by two companies, Soescol Ltda.
yer, as the final receiver; the Colombian attorney Luis and San Francisco C.I.S.A., and/or their controlling
E. Torres was representing the company; Noel R. Mer- shareholders Uvaldo Montenegro and Hernando
riam, a United States citizen, was serving as general Sánchez, respectively.172 Chivor is worked exclu-
manager in Bogotá; and the Colombian Jose Rodriguez sively underground by tunneling, and open-pit em-
was filling the role of mine manager.167 In terms of erald mining in the steep mountain range is no
market value at that juncture, the stock was traded longer performed (figures 24 and 25). Currently,
over the counter at 50 cents per share in 1970.168 Dur-
ing his years in office, Bronkie had been paid, at least
partly, in shares of the company, thereby becoming 165
Ibid.
one of the larger shareholders. 166
The New York Times, May 16, 1970; Cochrane, 1970; Feininger,
1970.
CHIVOR POST-RECEIVERSHIP (1970–PRESENT) 167
Bancroft, 1971; Diario Oficial (Colombia), 108, No. 33382 (1971),
p. 519; Gaceta Judicial (Colombia), 146 (1973), pp. 406–407.
With the termination of the receivership, the New 168
The National Monthly Stock Summary, 117 (1970), p. 228.
York leadership regained control over mining opera- 169
Latin America, 5 (1971), p. 123.
tions at Chivor, and in 1971 the American manage- 170
Gonzalo Jara, pers. comms. 2017, 2018.
ment requested an inspector from the Colombian 171
Keller, 1981; Erazo Heufelder, 2005; Cepeda and Giraldo, 2012.
government to check the emerald production.169 172
Fortaleché et al., 2017.

HISTORY OF CHIVOR, PART II: 1924–1970 GEMS & GEMOLOGY SUMMER 2020 253
1
2
VR
VR La Playa
VR R

Figure 25. A view from the main campsite in Chivor, looking in a southwest direction, shows the steep mountain
range with some small local mines and, on the other side of the valley (VR) of the Río Rucio (R), the mining areas
Buenavista (1) and Mundo Nuevo (2), each represented by a mining symbol. Photo by G. Martayan, 2018.

these primary operators are working numerous tun- quín Daza with whom he was involved in a legal dis-
nels or mines on the different existing conces- pute. Yet Rainier’s recitation contrasts strongly with
sions—including San Pedro, San Gregorio,
Manantial, Oriente, Piedra Chulo, Quebra Negra,
Figure 26. Sign welcoming visitors to the town of Chi-
Gualí, Dixon, and Tesoro.173 Additional galleries
vor. Photo by G. Martayan, 2018.
known within the region are Gavilanes, San Judas,
El Acuario, Mirador, Milenio, Porvenir, Palo Arañ-
ado, Calichal, Camoyo, Klein & Cuatro, Gualí, and
El Pulpito. In the town of Chivor (figure 26), there
is a small informal street market for faceted stones,
specimens in matrix (figure 27), and rough emerald
crystals (figure 28).

LEGENDS AND REALITY: RAINIER, ANDERTON,

AND THE HISTORICAL RECORD
The view of Chivor’s story from 1924 to 1970 has to
date been largely guided by the accounts penned by
Rainier and Anderton in Green Fire and Tic-Polonga,
respectively. Both of those publications, however,
are marked by notable absences. For instance, Rain-
ier never mentioned the involvement of the corpo-
rate leadership in New York in the stock market
manipulation controversy (Lewisohn, MacNamara,
Rice), nor did he name either his predecessors in
mine management (Burns, Mentzel) or his assistants
in the work at Chivor (Gilles, Sylvester). Conversely,
he dedicated a full chapter of his book to “Joaquin
the Bandit,” a character apparently inspired by Joa-

173
Ibid.

254 HISTORY OF CHIVOR, PART II: 1924–1970 GEMS & GEMOLOGY SUMMER 2020
istering or operating the mine from New York,
Bogotá, or Chivor itself left few, if any, traces in pub-
lications or official records.
Nonetheless, by using independent contempora-
neous materials in combination with and in restraint
of the dramatized accounts, it has become possible
to outline a more verifiable chronology of the period.
The almost half-century in which Chivor was
mainly owned and operated by the same American
company saw numerous ups and downs in terms of
productivity, emphasis, and financial fortunes.
In the first years after 1924, profit for the principal
shareholders, in close cooperation with Rice, derived
primarily from the United States stock market.
Mining activities were comparatively minimal, and
public interest centered on the scandals implicating
company leaders Lewisohn and MacNamara. Then,

Figure 28. Rough emerald crystals are also offered by

miners at the local street market in Chivor. Photo by
G. Martayan, 2018.

Figure 27. At the local street market in the town of

Chivor, miners offer emerald specimens in matrix.
Photo by G. Martayan, 2018.

the detailed record found in the Archivo General de

la Nación (Colombia) in Bogotá,174 signaling an in-
tent to offer an adventurous tale as opposed to an en-
tirely factual text. Thus, a comparison of Rainier’s
narration in Green Fire with original documents pre-
served from the period shows that the book em-
ployed real events to establish a rough framework
but that details were supplemented, embellished,
and even rearranged chronologically to suit the
genre. More mundane matters such as the facts from
the late 1920s that led to his engagement by the
mine owners and the colleagues cooperating at Chi-
vor were omitted. With Anderton, the lack of nearly
any dates to frame the events is particularly glaring.
The difficulties are compounded by the fact that an
untold number of individuals involved in admin-

174
File “Joaquín Daza B.,” Volume “Propuestas Minas 99,” Ministerio
de Industrias, Departamento de Minas y Petróleos, Archivo General
de la Nación (Colombia).

HISTORY OF CHIVOR, PART II: 1924–1970 GEMS & GEMOLOGY SUMMER 2020 255
under Rainer’s mine management from the late restore a measure of financial health, and the receiv-
1920s into the beginning of the following decade, the ership was lifted in 1970.
situation shifted to a greater focus on the operational The American shareholders thereby regained full
side. Before long, however, the scenario reversed control over the mine, but within a few years, Chivor
again, with Chivor being closed and the company es- was sold, closing an era and paving the way for the
sentially withdrawing from operating the property to increased local control and productivity existing
any significant degree until after World War II. today. While countless details remain fraught with
Yet another about-face in the late 1940s to early mystery, particularly insofar as concerns the motiva-
1950s brought renewed operational efforts instigated tions incentivizing various decisions and any com-
by Pace on behalf of the company in New York and prehensive accounting of mining costs versus
headed by Anderton at the mine. That era of produc- emerald sales, the tale of Chivor, even as presently
tivity nonetheless shortly fell victim to controversy known, likely has few equals. One thing that is clear,
between American and Colombian associates and however, is that Prof. Robert Scheibe’s conclusion in
mounting debts. The result was insolvency proceed- September 1914 upon examining the mine’s poten-
ings instigated under Colombian law in 1952 and tial—positing a realistic “chance” for a successful
nearly two decades under receivership. Bronkie, as venture but also an extremely high “risk” in buying
receiver and mining engineer, was eventually able to the property—proved a prescient one.

ABOUT THE AUTHORS (grandson of Francisco Restrepo Escobar), Mr. Scott Pace
Dr. Schmetzer is an independent researcher living in Peters- (grandson of Francis Philip Pace), the Patiño family, Ms. Lynn G.
hausen, near Munich, Germany. Mr. Martayan is a senior geo- Stewart (great-niece of George Daniel Besler), and to several indi-
physicist and longtime emerald aficionado residing in Paris, viduals directly involved in events related to the history of Chivor
France. Miss Blake is a gemologist residing in Chevy Chase, such as Mr. Peter W. Rainier, Jr., Mr. Manuel J. Marcial, don Al-
Maryland. fonso Montenegro, Ms. Renata de Jara, Mr. Gonzalo Jara, and
Mr. Robert E. Friedmann. Mr. Dan Plazak and Mr. T.D. Thornton
ACKNOWLEDGMENTS assisted with background concerning George Graham Rice. Ms.
Documents, pictures, biographical data, and other information Rose Tozer and Ms. Judy Colbert at GIA’s Richard T. Liddicoat Li-
concerning the events of the years 1924 to 1970 described here brary in Carlsbad, California, helped with locating materials ar-
have been provided by relatives of the persons involved, who chived in the Institute’s collections. Finally, countless other
have willingly reviewed their private records. Thanks are given unnamed individuals and staff members at various United States
especially to Ms. Catalina Dixon and Ms. Melissa Dixon (great- university archives, libraries, and local administrations should also
granddaughters of Christopher Ernest Dixon), Ms. Carola Kroll be thanked here for numerous details, references, addresses, and
(granddaughter of Fritz Klein), Mr. Eduardo Restrepo Ortega hints that contributed to completing the puzzle.

REFERENCES
Albrecht H. (2013) Adolph Lewisohn, Kupfermagnat im “Gold- Brock F.W. (1929) Investment pitfalls. Oral Hygiene, Vol. 19, pp.
enen Zeitalter”. Hamburg University Press, 148 pp. 789–795.
Anderton R.W. (1950-1951) Report on Chivor emerald mines. Canova L.J. (1921) Legend of the Discovery of Emeralds in the Co-
G&G, Vol. 6, No. 12, pp. 376–377, 379. lombian Andes. Colombian Emerald Syndicate, Ltd., New
——— (1953) Tic-Polonga. Doubleday & Company, Inc., New York, 30 pp.
York, 254 pp. Cepeda I., Javier Giraldo S.J. (2012) Víctor Carranza, alias “el
——— (1954) Tic-Polonga. Museum Press Limited, London, 224 Patrón.” Grijalbo, Random House Mondadori, S.A., Bogotá, 162
pp. pp.
——— (1955) The new Gachalá emerald mine in Colombia. G&G, Cochrane R.S. (1970) U.S. owners regain Colombia gem mine.
Vol. 8, No. 7, pp. 195–196. Lapidary Journal, Vol. 24, No. 5, p. 766.
——— (1957) Emerald outlook in Colombia. G&G, Vol. 9, No. 2, Domínguez R.A. (1965) Historia de las Esmeraldas de Colombia.
pp. 60–61. Gráficas Ducal, Bogotá, 297 pp.
——— (1965) The present status of Colombian emerald mining. Erazo Heufelder J. (2005) Der Smaragdkönig. Piper Verlag,
Lapidary Journal, Vol. 19, No. 3, pp. 374–377. München, 294 pp.
Bancroft P. (1971) The lure of Chivor. Lapidary Journal, Vol. 25, Feininger T. (1970) Emerald mining in Colombia: History & geol-
No. 1, pp. 128–131. ogy. The Mineralogical Record, Vol. 1, No. 4, pp. 142–149.

256 HISTORY OF CHIVOR, PART II: 1924–1970 GEMS & GEMOLOGY SUMMER 2020
Fortaleché D., Lucas A., Muyal J., Hsu T., Padua P. (2017) The Co- lombia. The American Institute of Mining and Metallurgical
lombian emerald industry: Winds of change. G&G, Vol. 53, No. Engineers, Technical Publication No. 258, 21 pp.
3, pp. 332–358, https://dx.doi.org/10.5741/GEMS.53.3.332 ——— (1931) The Chivor-Somondoco emerald mines of Colom-
Gilles V.A. (1930) Geology of the Chivor emerald mine. In O. Buis bia. Transactions of the American Institute of Mining and Met-
and W.F. Bronkie, Eds., Guide Book to the Geology of the Chi- allurgical Engineers AIME, Vol. 96, pp. 204–223.
vor Emerald Mine. Colombian Society of Petroleum Geologists ——— (1933a) Las minas de esmeraldas de Muzo y Coscuez. In-
and Geophysicists, published 1966, pp. 5–18. ternal report, printed 1958 in Boletín de Minas, Vol. 5, Nos. 56-
Gregory E.I. (1995) Incredible Stories of World War II Veterans. 57, pp. 10–19.
Senior Publishers, Vineburg, California, pp. 54–57. ——— (1933b) Informe geológico de las minas de esmeraldas de
Johnson P.W. (1959) The emeralds of Chivor today. Lapidary Jour- Muzo. Internal report, printed 1965 in R.A. Domínguez, His-
nal, Vol. 13, No. 4, pp. 516–524. toria de las esmeraldas de Colombia, Gráficas Ducal, Bogotá,
——— (1961) The Chivor emerald mine. Journal of Gemmology, pp. 237–240.
Vol. 8, No. 4, pp. 126–152. ——— (1934) Formación esmeraldifera de Chivor (Colombia.-Sur
Keller P.C. (1981) Emeralds of Colombia. G&G, Vol. 17, No. 2, pp. América). Minería, Vol. 3, No. 28, pp. 1806–1815.
80–92, https://dx.doi.org/10.5741/GEMS.17.2.80 ——— (1942) Green Fire. Random House, New York, 296 pp.
Klein F. (1941) Smaragde unter dem Urwald. Oswald Arnold Ver- Rice G.G. (1913) My Adventures with Your Money. Richard G.
lag, Berlin, 285 pp. Badger, Boston, 363 pp.
——— (1951) Smaragde unter dem Urwald. Selbstverlag, Idar- Schmetzer K., Martayan G., Ortiz J.G. (2020) History of the Chi-
Oberstein, 291 pp. vor emerald mine, Part I (1880–1925): From rediscovery to
Lentz M. (1951) Emeralds of Chivor. Lapidary Journal, Vol. 5, No. early production. G&G, Vol. 56, No. 1, pp. 66–109,
3, pp. 172–176. https://dx.doi.org/10.5741/GEMS.56.1.66
Marcial M.J. (2018) The Tears of Fura. Outskirts Press, Denver, Spence G.H. (1958a) Emerald mining in Colombia. The Gemmol-
293 pp. ogist, Vol. 27, No. 327, pp. 197–198.
Martínez Fontes E., Parodiz J.J. (1949) Guía de naturalistas sud- ——— (1958b) Emerald mining in Colombia. The Gemmologist,
americanos. Talleres Gráficos Lucania, Buenos Aires, 138 pp. Vol. 27, No. 328, pp. 215–218.
Morello T. (1956) The gem of Colombia. Américas, Vol. 8, No. 10, Thornton T.D. (2015) My Adventures with Your Money: George
pp. 21–24. Graham Rice and the Golden Age of the Con Artist. St. Mar-
Myers W.S. (2000) Prominent Families of New Jersey. Clearfield, tin’s Press, New York, 298 pp.
Baltimore, Vol. 1, pp. 142–143 (reprint, originally published as Torres Neira H. (1967) Una gema y un destino. Bogotá, 178 pp.
The Story of New Jersey, 1945). Wehrle L. (1980) Die Tränen des Mondes. Erlebtes aus Peru, Ecuador
Peretti A., Falise T. (2018) Magnificent Green. Second Edition, und Kolumbien. Meyster Verlag, Wien–München, pp. 205–208.
GRS Gemresearch Swisslab AG, Lucerne, 263 pp. Weldon R., Ortiz J.G., Ottaway T. (2016) In Rainier’s footsteps:
Plazak D. (2006) A Hole in the Ground with a Liar at the Top: Journey to the Chivor emerald mine. G&G, Vol. 52, No. 2, pp.
Fraud and Deceit in the Golden Age of American Mining. The 168–187, https://dx.doi.org/10.5741/GEMS.52.2.168
University of Utah Press, Salt Lake City, Utah, 374 pp. Winkler M. (1928) Investments of United States Capital in Latin
Rainier P.W. (1929) The Chivor-Somondoco emerald mines of Co- America. World Peace Foundation, Boston, p. 127.

For online access to all issues of GEMS & GEMOLOGY from 1934 to the present, visit:

gia.edu/gems-gemology

HISTORY OF CHIVOR, PART II: 1924–1970 GEMS & GEMOLOGY SUMMER 2020 257
FEATURE ARTICLES

OPTICAL WHITENING AND BRIGHTENING

OF PEARLS: A FLUORESCENCE
SPECTROSCOPY STUDY
Chunhui Zhou, Tsung-Han Tsai, Nicholas Sturman, Nanthaporn Nilpetploy,
Areeya Manustrong, and Kwanreun Lawanwong

White pearls and those with very light hues are routinely processed or treated by various methods after harvest,
especially akoya and freshwater cultured pearls. Bleaching is the most prevalent and is often used in addition
to maeshori treatment, an umbrella term for various types of luster enhancements. However, little reference
has been made in the literature to the use of optical brighteners, the focus of this study, within the pearl trade.
In this project, known samples of optically brightened akoya cultured pearls were studied and compared with
their non-brightened counterparts and cultured pearls of various types. In addition, a South Sea cultured pearl
was treated in-house with a commercial whitening product. Our preliminary results indicated that pearls treated
by this method showed strong bluish fluorescence under long-wave ultraviolet (UV) excitation, and can be con-
fidently separated from non-brightened pearls using fluorescence spectroscopy when a distinct fluorescence
maximum at around 430 to 440 nm is observed.

O
ptical brightening agents (OBAs) are chemi- can be generated by exciting the substance via a
cal compounds that can absorb light in the range of energy sources. The molecule in the sub-
ultraviolet and violet region of the electro- stance absorbs the source energy and once excited
magnetic spectrum and emit light in the blue region moves from a lower electronic state to a higher one.
as fluorescence, due to their extended conjugation Immediately after absorbing energy, it loses the en-
and/or aromaticity. They are sometimes called fluo- ergy by emitting a photon; this process of photon
rescent brightening agents or fluorescent whitening emission is called luminescence. Typically, fluores-
agents, and have been frequently used to enhance the
appearance of fabric and paper (Lanter, 1966; Leaver
and Milligan, 1984; Esteves et al., 2004; Bajpai, 2018).
While many types of brighteners are listed in the In Brief
Colour Index (https://colour-index.com), only a • Akoya and freshwater cultured pearls are routinely
handful are commercially important. Some examples processed to improve their appearance.
are shown in figure 1. • Long-wave UV radiation is a quick and useful tool to
Photoluminescence is light emission from any aid in the identification of some treatments.
form of matter after the absorption of photons. Fluo- • Optically brightened pearls appear white with the ad-
rescence is a type of photoluminescence in which a dition of agents, while bleaching makes pearls whiter
molecule dissipates its absorbed energy through the by removing discoloration or pigmentation.
rapid emission of a photon, while phosphorescence • Fluorescence spectroscopy is a quick and effective way
is the emission of radiation in a similar manner to of separating optically brightened pearls from bleached
fluorescence but on a longer timescale, so that emis- and untreated ones.
sion continues after excitation ceases. Fluorescence

cence occurs at a lower energy than the absorbed ra-

diation; this phenomenon is known as the Stokes
See end of article for About the Authors and Acknowledgments.
GEMS & GEMOLOGY, Vol. 56, No. 2, pp. 258–265,
shift (Lakowicz, 2006).
http://dx.doi.org/10.5741/GEMS.56.2.258 Fluorescence occurs widely in nature. For exam-
© 2020 Gemological Institute of America ple, many kinds of gemstones such as rubies, emer-

258 WHITENING AND BRIGHTENING OF PEARLS GEMS & GEMOLOGY SUMMER 2020
Na+ Na+
O O– O– O
S O O S N
N
O
O
Figure 1. Chemical
Disodium 4,4’-bis(2-sulfostyryl)biphenyl 4,4’- Bis(5-methyl-2-benzoxazolyl)stilbene
structures of commonly
used optical brighten-
N N
ing agents.
O
O O
N S N

O
1,4-Bis(2-benzoxazolyl)naphthalene 2,5-Bis(5-tert-butyl-2-benzoxazolyl)thiophene

alds, and diamonds can emit visible light when ex- the sample and visually observe the fluorescence ef-
posed to ultraviolet light. For research purposes, flu- fect (color, intensity, distribution, etc.). Untreated
orescence can be measured by fluorescence white to cream-colored pearls generally show blue to
spectroscopy. The emission signal from the sample greenish blue fluorescence under LWUV excitation.
is collected by a spectrometer to monitor the distri- Akoya and freshwater cultured pearls, which are rou-
bution in fluorescence intensity as a function of tinely processed, typically show strong bluish white
wavelength. Since fluorescence has a short lifetime, fluorescence. Similar to these pearls, OBAs are also
typically decaying within nanoseconds after the ex- highly fluorescent. With exposure to UV radiation, a
citation, the sample must be constantly illuminated bright blue glow is visible from the OBA solution (an
to maintain its fluorescence signal. example is shown in figure 2). The use of fluores-
Fluorescence spectroscopy has been used exten- cence spectroscopy can easily separate the minor dif-
sively as a basic tool in a wide range of research ferences in fluorescence color exhibited by some
fields, including chemistry, biology, mineralogy, and pearls and the OBAs.
gemology (Lakowicz, 2006). For example, structural
defects in minerals usually result in characteristic
fluorescence. A typical application is to use fluores- Figure 2. An example of blue fluorescence emitted by
cence spectroscopy to distinguish natural diamond an OBA solution used in our in-house treatment ex-
from its synthetic counterparts (Tsai and D’Hae- periment under LWUV radiation. Photo by Chunhui
nens-Johansson, 2019). One important feature of Zhou.
fluorescence spectroscopy is its high sensitivity,
which enables it to detect the luminescence signal
from a near-zero background. Modern sensors make
single-photon sensitivity possible. As a result, fluo-
rescence spectroscopy is extremely sensitive, allow-
ing for the detection of weak signals. Pearl, a
biogenically originated material, inherits fluores-
cence characteristics typical of many organic mate-
rials, and the biogenic aragonite that constitutes the
bulk of a pearl is well known for its fluorescence
properties. These facts make fluorescence spec-
troscopy a potentially very useful tool for the iden-
tification of pearls from different species located in
various areas.
Since both pearls and OBAs possess characteristic
fluorescence features, the use of fluorescence spec-
troscopy makes it possible to easily detect pearls
treated by such agents. The traditional way to ob-
serve pearl fluorescence is to use long-wave UV
(LWUV) and short-wave UV (SWUV) lamps to excite

WHITENING AND BRIGHTENING OF PEARLS GEMS & GEMOLOGY SUMMER 2020 259
The use of OBAs in the cultured pearl industry is MATERIALS AND METHODS
not unusual (Song and Fan, 1999; Japan Pearl Promo- Two groups of mostly white-colored akoya bead cul-
tion Society, 2014), as they are sometimes routinely tured pearls ranging from 6.97 × 6.54 mm to 8.14 ×
applied to cultured pearls post-harvest along with 7.48 mm were supplied by Orient Pearl (Bangkok)
other types of processes such as bleaching or maeshori Ltd. One group (139 pearls) was reportedly treated
(which refers to various types of luster enhancement), with unidentified OBAs in addition to the routine
but very little has been written in the gemological lit- bleaching process, while the other group (150 pearls)
erature about the identification of this process (Shor, was only bleached and not treated with OBAs (figure
2007). It is important to differentiate between the 3). In addition, three different types of white cultured
whitening effects due to bleaching and those produced pearls (akoya, freshwater, and South Sea) from GIA’s
by optical brightening, as they depend on two very dif- research collection (figure 4), as well as samples from
ferent mechanisms: bleaching removes, or reduces the an untreated akoya strand reportedly cultured in
concentrations of, the color-causing pigmentations Vietnam, also from GIA’s research collection (figure
within pearls, while optical brighteners add an artifi- 5), were included in the study.
cial optical effect to pearls, making them look less yel- One white South Sea cultured pearl was also
low by increasing the overall amount of blue light treated in-house using a commercial laundry product
emitted due to these chemicals’ blue fluorescence (Rit Whitener & Brightener). According to the label,
properties. The focus of this study is to use fluores- this product contained optical brighteners and
cence spectroscopy as an easy method to detect the sodium chloride. The pearl was immersed in the sat-
presence of OBAs. Such a technique would be of great urated brightener solution (approximately 1 g of
practical interest to the trade and public in order to brightener powder per 100 mL of water) for 3 hours
faithfully differentiate treated from untreated pearls. at a slightly elevated temperature (about 40°C), with

A B

Figure 3. A group of op-

tically brightened and
bleached white akoya
cultured pearls (A) and
a group of bleached-
only (no OBAs) akoya
pearls (B) under normal
daylight (A and B) and
LWUV radiation (C and
C D D). The samples ex-
posed to OBAs exhib-
ited a more intense
bluish fluorescence re-
action (C). Photos by
Nanthaporn Nilpetploy.

260 WHITENING AND BRIGHTENING OF PEARLS GEMS & GEMOLOGY SUMMER 2020
Figure 4. Three white cultured pearls—akoya (left), freshwater (middle), and South Sea (right)—shown in daylight
(left photo) and exhibiting fluorescence under LWUV excitation (right photo). The freshwater pearl exhibited a dis-
tinctive blue fluorescence, which suggested it had been optically brightened. Photos by Sood Oil (Judy) Chia.

occasional stirring. It was subsequently rinsed with and transmitted the collected fluorescence signal
tap water and air-dried before further examination. from the sample to the detector. The spectrometer
All the samples were initially examined using was configured to monitor the spectrum from 400 to
various standard gemological techniques, including 700 nm, which covers the fluorescence range needed
a gemological microscope, ultraviolet-visible (UV- to analyze reactions from untreated pearls and OBAs.
Vis) reflectance spectroscopy, and Fourier-transform Two optical filters were used to isolate the pearl flu-
(FTIR) reflectance spectroscopy. Because these tests orescence from the UV LED in order to avoid the ex-
cannot consistently identify optical brighteners, the citation light overwhelming the sensor, as well as to
results are not shown here. For the purpose of this enhance the sensitivity of the system. A 390 nm
particular project, a fluorescence spectroscopy unit short-pass filter was positioned behind the UV LED,
was designed to measure pearl fluorescence with the and a 410 nm long-pass filter was positioned in front
aim of detecting OBAs. The fluorescence spec- of the detector.
troscopy unit included a 385 nm UV light-emitting In addition to fluorescence spectroscopy, visual
diode (LED) as the excitation light source, a spec- observation of the fluorescence reactions was per-
trometer for monitoring the visible light region as the formed under a conventional 5-watt LWUV (365 nm)
fluorescence sensor, and a bi-fabricated fiber-optic lamp or a GIA-designed LWUV unit incorporating a
probe that guided the excitation light to the sample narrow-band 365 nm UV LED source.

Figure 5. A strand of non-bleached and non-brightened akoya cultured pearls of various colors (left) and their var-
ied fluorescence reactions under LWUV radiation (right). Photos by Sood Oil (Judy) Chia.

WHITENING AND BRIGHTENING OF PEARLS GEMS & GEMOLOGY SUMMER 2020 261
FLUORESCENCE SPECTRA
RESULTS South Sea cultured pearl
6 × 104
The two groups of akoya cultured pearls exhibited Akoya cultured pearl

COUNTS (ARB. UNITS)

Freshwater cultured pearl
similar external appearances in terms of their size, 5 × 104 Commercial office paper

color, shape, luster, and surface condition. The group

4 × 104
that had been bleached and treated with OBAs ap-
peared to be slightly whiter and more lustrous than 3 × 104

the non-brightened group, although this could be due

2 × 104
to the individual differences among the samples. No
systematic before and after comparison on the ap- 1 × 104

pearances of the treated group was available. Under

LWUV radiation, the group treated with OBAs 400 450 500 550 600 650 700
showed a more consistent and stronger bluish reac- WAVELENGTH (nm)
tion (again, see figure 3). Some of the OBA-treated
pearls (especially the cream samples) showed spotty Figure 7. The fluorescence spectra of a white South
and uneven reactions under LWUV, indicating the Sea cultured pearl (green), white akoya cultured pearl
treatment concentrated in surface areas where blem- (purple), white freshwater cultured pearl (blue), and
ishes existed. This was confirmed by microscopy white commercial office paper (red). The freshwater
since the surface condition of the nacre was not per- pearl tested in this study and the white paper show
fect given the lower grade of akoya used in the exper- fluorescence results consistent with those expected
iment. In order to see if OBAs were detectable, for OBAs.
fluorescence spectroscopy was applied to selected
samples from each group, and the results are shown
in figure 6. All bleached-only samples tested showed
broad fluorescence maxima at around 480 to 490 nm, shifted fluorescence maxima are consistent with the
while the bleached and OBA-treated samples showed fluorescence characteristics of commercially avail-
narrower fluorescence maxima at around 430 to 434 able OBAs (Lawson-Wood and Evans, 2018).
nm with an additional shoulder around 470 nm. The The three white cultured pearl samples from GIA’s
collection showed various fluorescence reactions
under LWUV (figure 4, right photo). The South Sea cul-
tured pearl showed a moderate greenish reaction and
Figure 6. Fluorescence spectra of bleached and opti-
the akoya cultured pearl a slightly bluish reaction,
cally brightened (red) and bleached and non-bright-
ened (blue) akoya pearls. The non-brightened pearls
while the freshwater pearl exhibited the most distinc-
show a broad fluorescence maximum at around 480 tive and strongest bluish reaction. These observations
to 490 nm, whereas the OBA-treated samples exhibit indicated that the South Sea cultured pearl was most
a narrower fluorescence maximum at around 430 to likely not treated, the akoya cultured pearl was most
434 nm with a shoulder at around 470 nm. likely traditionally processed with bleaching, and this
particular freshwater pearl was also processed. Its UV
FLUORESCENCE SPECTRA fluorescence reaction looked very similar to the opti-
cally brightened akoya samples we studied. The fluo-
Non-brightened pearls rescence spectra of these three pearls confirmed these
6 × 104
Optically brightened pearls
predictions (figure 7). The South Sea cultured pearl
COUNTS (ARB. UNITS)

5 × 104 showed a fluorescence maximum at around 494 nm,

the akoya cultured pearl at around 482 nm, and the
4 × 104
freshwater pearl at around 432 nm with a shoulder at
3 × 104 around 465 nm. The fluorescence reaction and the
spectrum of the freshwater pearl matched well with
2 × 104
the OBA-treated akoya samples discussed earlier.
1 × 104 Additionally, a piece of standard commercial white
office paper was tested. Its fluorescence reaction under
400 450 500 550 600 650 700 LWUV was found to be strong blue, and its fluores-
WAVELENGTH (nm) cence spectrum exhibited a maximum at around 439
nm with a shoulder at around 474 nm, similar to the

262 WHITENING AND BRIGHTENING OF PEARLS GEMS & GEMOLOGY SUMMER 2020
FLUORESCENCE SPECTRA
hibit this same wide range of bodycolors when rou-
6 × 104
tine processing is not carried out (Otter et al., 2017).
COUNTS (ARB. UNITS)

5 × 104
These pearls showed a variety of weak fluorescence
reactions under LWUV radiation, with the weaker re-
4 × 104
actions the result of fluorescence quenching by their
3 × 104 bodycolors (figure 5, right).
Spectroscopic results showed that in keeping with
2 × 104
bleached akoya pearls, these samples exhibited a flu-
1 × 104 orescence maximum ranging between 480 and 500
nm (figure 8). These results and additional tests (data
not shown) on GIA akoya and South Sea cultured
400 450 500 550 600 650 700
pearls (master samples), as well as on untreated fresh-
WAVELENGTH (nm)
water pearls of various colors from the GIA research
collection, confirmed that non-brightened pearls flu-
Figure 8. Fluorescence spectra of variously colored un-
oresce a weaker blue (cyan) color (480 to 500 nm),
treated akoya pearls, showing fluorescence maxima
ranging from around 480 to 500 nm. The small sharp
while OBA-processed pearls exhibit more intense
peaks shown near 540 and 610 nm are due to the am- blue and violet colors (430 to 440 nm).
bient environment (room light and computer monitor Finally, one white South Sea cultured pearl was
light reflection). treated in-house using a commercial laundry product
(Rit Whitener & Brightener) containing OBAs and
sodium chloride, as indicated on its label. The main
ingredient listed on the manufacturer’s website for
OBA-treated pearls studied. It was no surprise to see this type of product is linear alkylbenzene sulfonate
such features since optical brighteners are routinely (CAS No. 25155-30-0). The simple method described
applied to commercial white paper (Bajpai, 2018). in the Materials and Methods section resulted in an
In order to verify the results, several non-bleached increased bluish fluorescence under LWUV, and also
and non-brightened akoya cultured pearls of various a slightly higher luster and whiter appearance for the
colors that included cream, light yellow, silver, and pearl (figure 9). Its fluorescence changed to shorter
light gray were tested with the same fluorescence wavelengths from around 500 nm to around 443 and
spectroscopy method. Untreated akoya pearls can ex- 478 nm as expected, similar to the fluorescence results

Figure 9. External ap-

pearances and fluores-
cence reactions under
LWUV of a white South
Sea cultured pearl be-
fore (left) and after
(right) in-house treat-
ment using a commer-
cial brightener. Photos
by Sood Oil (Judy)
Chia.

WHITENING AND BRIGHTENING OF PEARLS GEMS & GEMOLOGY SUMMER 2020 263
FLUORESCENCE SPECTRA

Before treatment
sities stronger, OBAs often change the fluorescence
6 × 104
After treatment properties of the pearls more dramatically for both
COUNTS (ARB. UNITS)

OBA solution
5 × 104
the fluorescence hues and intensities. These changes
are consistent with the original fluorescence proper-
4 × 104
ties of the OBAs (shifting the fluorescence maxima
3 × 104 to shorter wavelengths in the blue and violet range).
Since there are many different types of OBAs in the
2 × 104
market and the exact process is unknown, the effects
1 × 104 of OBAs are not within the scope of this study, al-
though we observed a slight increase in luster and
white appearance for the South Sea cultured pearl
400 450 500 550 600 650 700
treated in-house. With more sophisticated methods
WAVELENGTH (nm)
and better chemicals, these effects could be more
prominent. The extent of such a process used in the
Figure 10. Fluorescence spectra of the white South Sea
trade is also not clear; the authors have heard some
cultured pearl before (red) and after (purple) in-house
brightening treatment. The OBA solution’s spectrum
sources in the trade say it is used less often than it
is also shown in blue. was many years ago. Similarly, the stability of such
a process also requires better sample comparison and
prolonged observation, which has not been fully ad-
dressed in this study.
obtained for the OBA solution itself (figure 10). How- In addition to the detection of OBAs in pearls, flu-
ever, the exact OBAs applied for routine pearl treat- orescence spectroscopy could potentially have appli-
ment in the trade are most likely different from the cations in the detection of other color treatments
one applied in the in-house experiment, and the actual used on pearls (Tsai and Zhou, 2020). Furthermore,
methods are probably much more sophisticated and differences in the fluorescence characteristics of var-
time sensitive. This likely explains why the fluores- ious types of pearls may also be used to separate
cence maximum of the pearl treated in-house differs pearls from different host mollusk species, as previ-
from the treated samples obtained from the trade. ously reported (Ju et al., 2011). While the optimal UV
excitation wavelength for pearls has not been fully
DISCUSSION investigated, the excitation source used in this par-
ticular study has proved quite effective. The full po-
The application of OBAs to pearls is not new in the
tential of this advanced technique for pearls and
pearl trade, but it is sometimes not regarded as a
other gemstone identification scenarios warrants
treatment (Japan Pearl Promotion Society, 2014).
greater attention.
OBAs are usually applied in conjunction with bleach-
ing and are thus considered part of routine process-
ing. Little has been discussed in the gemological CONCLUSIONS
literature on the identification of this process, even This study proves that fluorescence spectroscopy, in
though optical brightening and bleaching utilize two combination with visual fluorescence observation,
different mechanisms to enhance the appearances of may provide a quick and easy way to detect the pres-
pearls. In this study, the authors focused on the iden- ence of different optical brighteners that have been
tification of this process using fluorescence spec- applied to various nacreous pearls. The samples ex-
troscopy in combination with visual fluorescence amined generally showed a strong fluorescence peak
observations under LWUV radiation. Both methods at around 430–440 due to OBAs, while non-bright-
proved useful, but the spectroscopic method has the ened pearls usually exhibited a fluorescence peak at
advantage of accurately distinguishing OBA- around 480–500 nm. While cultured pearls may be
processed pearls from routinely processed pearls, re- subject to routine processes involving various types
gardless of their colors, which might affect the visual of chemical agents, it is important for the trade to be
appearance and intensity of their UV fluorescence. aware of this particular component and its identifi-
While bleaching may make their fluorescence inten- cation criteria.

264 WHITENING AND BRIGHTENING OF PEARLS GEMS & GEMOLOGY SUMMER 2020
ABOUT THE AUTHORS ACKNOWLEDGMENTS
Dr. Zhou is a senior manager of pearl identification at GIA in New The authors would like to thank Mr. Devchand Chodhry, manag-
York. Dr. Tsai is a research scientist at GIA in New Jersey. Mr. Stur- ing director of Orient Pearl (Bangkok) Ltd. for providing some
man is a senior manager, Ms. Nilpetploy is a senior staff gemologist, samples in this study to GIA’s Bangkok laboratory.
Ms. Manustrong is a staff gemologist, and Ms. Lawanwong is an
analytics technician, in GIA Bangkok’s pearl department.

REFERENCES
Bajpai P. (2018) Optical properties of paper. Chapter 11 in Bier- ening agents in samples using the FL 6500 fluorescence spec-
mann’s Handbook of Pulp and Paper (3rd ed.). Elsevier, pp. trometer and solid sample holder. Application Brief, Fluores-
237–271, https://doi.org/10.1016/B978-0-12-814238-7.00011-8 cence Spectroscopy, PerkinElmer, Inc., Seer Green, UK.
Esteves M.F., Noronha A.C., Marinho R.M. (2004) Optical bright- Leaver I.H., Milligan B. (1984) Fluorescent whitening agents—a sur-
eners effect on white and coloured textiles. World Textile Con- vey (1974-82). Dyes and Pigments, Vol. 5, No. 2, pp. 109–144.
ference—4th AUTEX Conference, Roubaix, June 22–24. Otter L.M., Agbaje O.B.A., Huong L.T-T., Häger T., Jacob D.E. (2017)
Japan Pearl Promotion Society (2014) Pearl Standard (English ver- Akoya cultured pearl farming in eastern Australia. G&G, Vol.
sion), Revised Version, p 21. http://jp-pearl.com/wp- 53, No. 4, pp. 423–437, http://dx.doi.org/10.5741/GEMS.53.4.423
content/uploads/2017/09/2014版 真 珠 ス タ ン タ ゙ ー ト Shor R. (2007) From single source to global free market: The trans-
-%EF%BC%88英語%EF%BC%89.pdf
゙ formation of the cultured pearl industry. G&G, Vol. 43, No. 3,
Ju M.J., Lee S.J., Kim Y., Shin J.G., Kim H.Y., Lim Y., Yasuno pp. 200–226, http://dx.doi.org/10.5741/GEMS.43.3.200
Y., Lee B.H. (2011) Multimodal analysis of pearls and pearl Song H.C., Fan Y. (1999) Bleaching technique for pearls (in Chi-
treatments by using optical coherence tomography and flu- nese). Journal of Fisheries of China, Vol. 4, pp. 398–402.
orescence spectroscopy. Optics Express, Vol. 19, No. 7, pp. Tsai T.-H., D’Haenens-Johansson U.F.S. (2019) Gemstone screen-
6420–6432. ing and identification using fluorescence spectroscopy, Fron-
Lakowicz J.R. (2006) Principles of Fluorescence Spectroscopy, 3rd tiers in Optics + Laser Science APS/DLS, OSA Technical
ed. Springer US, pp. 5–7. Digest (Optical Society of America, 2019), paper JTu3A.109,
Lanter J. (1966) Properties and evaluation of fluorescent brighten- https://doi.org/10.1364/FIO.2019.JTu3A.109
ing agents. Journal of the Society of Dyers and Colourists. Vol. Tsai T.-H., Zhou C. (2020) Lab Notes: Fluorescence spectroscopy
82, No. 4, pp. 125–132. for colored pearl treatment screening. G&G, Vol. 56, No. 1, pp.
Lawson-Wood K., Evans K. (2018) Measurement of optical bright- 136–137.

For online access to all issues of GEMS & GEMOLOGY from 1934 to the present, visit:

gia.edu/gems-gemology

WHITENING AND BRIGHTENING OF PEARLS GEMS & GEMOLOGY SUMMER 2020 265
FEATURE ARTICLES

HYDROGEN AND OXYGEN STABLE ISOTOPE

RATIOS OF DOLOMITE-RELATED NEPHRITE:
RELEVANCE FOR ITS GEOGRAPHIC ORIGIN
AND GEOLOGICAL SIGNIFICANCE
Kong Gao, Ting Fang, Taijin Lu, Yan Lan, Yong Zhang, Yuanyuan Wang, and Yayun Chang

Hydrogen and oxygen stable isotope ratios of dolomite-related nephrites around the world were studied using
data from the literature (n = 120). These isotope ratios are highly effective for discriminating dolomite-related
nephrites from the four most important origins worldwide. Nephrite from Vitim in Russia has the lowest isotope
ratio values reflected in δ2H and δ18O values, followed by Chuncheon in South Korea and then Xinjiang Uyghur
Autonomous Region in China. Nephrite from Sanchakou in the Qinghai Province of China has the highest values.
Other occurrences are characterized by high δ18O values similar to or higher than those of samples from San-
chakou. The differences are derived mainly from the ore-forming fluids. Vitim and Chuncheon isotope ratio
values were mainly affected by meteoric water (rainwater, lake water, seawater, river water, glacial water, and
shallow groundwater). Xinjiang nephrite-forming fluids were mixtures of magmatic hydrothermal fluids (able
to be modified by metamorphism) and meteoric water. The hydrothermal fluids forming the Qinghai, Luodian,
Dahua, and Xiuyan nephrites underwent some metamorphic alteration or regional metamorphism.

N
ephrite is a near-monomineralic rock composed sources in China: Xiuyan, Tanghe, Dahua, and Luo-
of tremolite-actinolite, Ca2(Mg,Fe)5Si8O22(OH)2. dian (figure 2).
It occurs worldwide (figure 1) and is classified With nephrite jade, a premium is placed on geo-
as dolomite-related or serpentine-related according graphic origin since the gem’s cultural significance
to the different parent rocks and ore-hosting rocks, differs by location. It is possible to have an opinion
and both types form by metasomatism (Yui et al.,
1988; Tang et al., 1994; Yang and Abduriyim, 1994;
Harlow and Sorensen, 2005; Burtseva et al., 2015).
The large and well-known dolomite-related nephrite
In Brief
deposits are distributed in the Xinjiang Uyghur Au- • Geographic origin can have a significant impact on the
value of nephrite.
tonomous Region (hereafter abbreviated as Xinjiang)
of China, Qinghai Province of China, Siberia in Rus- • Hydrogen and oxygen stable isotope ratios, particu-
sia, and Chuncheon in South Korea (figure 1). Data larly the latter, provide a robust tool for origin
determination.
from smaller-scale deposits such as Val Malenco in
Italy and Złoty Stok in Poland are also used in this • Dolomite-related nephrites from Vitim, Chuncheon,
Xinjiang, and Qinghai differ from one another by dis-
study (figure 1). The rest of the data were collected
tinct hydrogen and oxygen stable isotope ratios.
from nephrites produced at multiple small-scale
• Differences in hydrogen and oxygen stable isotope
ratios for nephrite are related to ore-forming fluids.

See end of article for About the Authors and Acknowledgments.

GEMS & GEMOLOGY, Vol. 56, No. 2, pp. 266–280,
http://dx.doi.org/10.5741/GEMS.56.2.266 on the origin of a small amount of nephrite by simple
© 2020 Gemological Institute of America visual examination, since some varieties with unique

266 STABLE ISOTOPE RATIOS OF DOLOMITE-RELATED NEPHRITE GEMS & GEMOLOGY SUMMER 2020
Vitim
Xinjiang
Złoty Stok
Val Malenco Chuncheon

Qinghai

Cowell
Dolomite-related nephrite deposit
Serpentine-related nephrite deposit
Nephrite deposit with unknown types

Figure 1. Distribution map of reported nephrite deposits worldwide. The four most important dolomite-related nephrite
sources are Vitim in Russia, Chuncheon in South Korea, and Xinjiang and Qinghai in China. From @naturalearthdata.com.

gemological characteristics such as color, luster, and informative results (Zhong et al., 2013; Luo et al.,
transparency have mainly occurred in specific de- 2015). Unfortunately, rigorous and scientific deter-
posits. In Xinjiang, for example, high-quality white mination of geographic origin is still not available.
primary nephrite occurs in Qiemo County.1 There it Hydrogen and oxygen isotope ratio values (see box
is commonly associated with brown nephrite (figure A), which might vary for the same gemstone from dif-
3A, rough with white core and very thick brown ferent regions due to diverse ore-forming environments
rind). The brown is a color seldom found in nephrite and models, can be used for geographic origin determi-
from other deposits in Xinjiang. The highest-quality
white primary nephrite (figure 3B, white plate)
mostly comes from the Hetian region and Qiemo
TABLE 1. Location names expressed in Chinese pinyin and
County. Placer nephrite (figures 3C and 3D, pendants their English equivalents.
with figures carved out of brownish red skin) occurs
in the Yulongkashi River and Kalakashi River basins. Chinese pinyin English
A considerable quantity of primary nephrite from A’erjinshan Altyn Tagh
Ruoqiang County features a yellow color component Alamasi Alamas
(figure 3E, greenish yellow fish) that is absent from Bayinguoleng region Bayingholin region
other samples. Black nephrite (figure 3F, bangle Hetian Hoten/Hotan
bracelet) colored by graphite, on the other hand,
Kalakashi, Hetian Qaraqash, Hoten
mainly occurs in the Hetian region and has not been
Kashi region Kashkar region
found in Qiemo County or Ruoqiang County. How-
ever, the origin determination of a tremendous Keliya River Keriye/Keriya River
amount of dolomite-related nephrite cannot be Qiemo County, Bayinguoleng Cherchen County, Bayingholin
solved by this simple observation. Previous re- Ruoqiang County, Bayinguoleng Chaqiliq County, Bayingholin
searchers used trace elements combined with appear- Takelamagan Desert Taklimakan Desert
ance to identify geographic origin and obtained some Tashiku’ergan County Taxkorgan/Tashkurghan County
Yecheng County, Kashi Qaghiliq County, Kashkar

1
Yulongkashi, Hetian Yurungqash, Hoten
This paper uses Chinese pinyin to express all Chinese location names
involved, and the corresponding commonly used English names are Yutian County Keriye County, Hoten
listed in table 1.

STABLE ISOTOPE RATIOS OF DOLOMITE-RELATED NEPHRITE GEMS & GEMOLOGY SUMMER 2020 267
A

Xinjiang
ei Liaoning
Heb

Xiuyan
B
Tanghe

Qinghai

Henan

Jian
Sanchakou

gsu
Sichuan

Guizhou
Luodian
Dahua
Primary nephrite deposit Guangxi

Placer nephrite deposit

ver Ruoqiang
o Ri
Qiem
Yulongkashi;
Kalakashi
Yecheng Qiemo
iver

Tashisayi
Keliya R

Hetian
i
Yulongkash

ashi
Kalak
Datong
Alamasi
Agejugai

Figure 2. The distribution of main dolomite-related nephrite deposits of China. Primary and placer nephrite occur
at Xiuyan, but the latter is not plotted in the map.

nation (Giuliani et al., 1998, 2000, 2005, 2007). A mass flight mass spectrometry (LA-ICP-TOF-MS) for stable
spectrometer is needed to determine the isotope ratio isotope analysis of gemstones can be restricted to
values (see box B). The spot produced by secondary ion craters of 10–100 μm in diameter and a few angstroms
mass spectrometry (SIMS), laser ablation–inductively to microns deep (Giuliani et al., 2000, 2005; Abduriyim
coupled plasma–mass spectrometry (LA-ICP-MS), and and Kitawaki, 2006; Wang et al., 2016, 2018). The
laser ablation inductively coupled plasma time-of- craters produced are very small, to the point of not

268 STABLE ISOTOPE RATIOS OF DOLOMITE-RELATED NEPHRITE GEMS & GEMOLOGY SUMMER 2020
A B

Figure 3. Nephrites
from different localities
of Xinjiang showing dis-
tinct appearances. A:
The rough is a typical
piece of white-brown
nephrite from Qiemo
County. B: The white
C D plate is primary
nephrite of the highest
quality, which occurs in
both the Hetian region
and Qiemo County. C
and D: Two pendants
carved from placer
nephrite collected in
the Hetian region fea-
ture brownish red skin
sculpted into different
figures. E: The greenish
yellow fish features a
yellow color component
of primary nephrite
E F from Ruoqiang. F:
Black nephrite like that
of the bangle bracelet
mainly occurs in the
Hetian region. Photos
by Dong He; courtesy of
Elegant China.

being noticeable without magnification. This permits tionship between the characteristics of nephrite and
the method to be applied to gemstones and historical its formation environment.
antiques (Giuliani et al., 2000, 2005).
Geographic origin discrimination of nephrite by WHY DO ISOTOPE RATIOS MATTER TO
isotopes is seldom reported, even though many hy- GEMOLOGISTS?
drogen and oxygen isotope ratio studies on this ma-
terial have been carried out (table 2). By summarizing The application of isotopes has gradually attracted the
and analyzing all available hydrogen and oxygen iso- attention of gemologists (Wang et al., 2016). In addition
topic data of dolomite-related nephrites worldwide to hydrogen and oxygen isotope ratios, which can help
from published references, this study discusses the determine the geographical origins of corundum and
geographic origin discrimination based on the rela- emerald (Giuliani et al., 1998, 2000, 2005, 2007; Wang

STABLE ISOTOPE RATIOS OF DOLOMITE-RELATED NEPHRITE GEMS & GEMOLOGY SUMMER 2020 269
BOX A: INTRODUCTION TO OXYGEN AND HYDROGEN STABLE ISOTOPE RATIOS
Atoms with an equivalent atomic number (i.e., atoms of The isotopic fractionation factor is always a function
the same element) can differ from one another in their of temperature, which can be obtained by theoretical cal-
number of neutrons. For example, 18O has 8 protons and culation or experiment (Graham et al., 1984; Zheng,
10 neutrons, and 16O has 8 protons and 8 neutrons; 2H, 1993, 1995).
also known as deuterium (D), has 1 proton and 1 neu- Both oxygen and hydrogen isotope ratios are also re-
tron, while 1H has 1 proton and no neutrons. Such atoms ported in so-called delta notation given in terms of per
with the same number of protons but different numbers mil (‰). In other words, the delta value
of neutrons are defined as isotopes.
The mass difference inherent from divergent neu- (18O/16O)sample –(18O/16O)standard
δ18O = × 1000‰
trons causes isotopic fractionation, which occurs as the (18O/16O)standard
isotopes of an element are distributed between two sub-
stances or phases in differing ratios in a given system. and
This process can be affected by temperature, equilibrium
or kinetic processes, and other physiochemical (2H/1H)sample –(2H/1H)standard
processes. The isotope fractionation will reach and main- δ 2H = × 1000‰
(2H/1H)standard
tain equilibrium unless conditions change. Therefore,
isotope abundance can be used as a tracer to reveal cer-
tain geochemical processes in geological bodies. in which 18O/16O and 2H/1H are the isotope ratios defined
Isotope ratio, defined as the measured relative abun- above. Values of delta > 0 indicate that relative to the
dance of a heavy isotope to its lighter counterpart (e.g., standard samples, the tested sample has a higher heavy
18
O/16O and 2H/1H), is typically used rather than the iso- isotope abundance, and a negative delta value indicates
tope abundance itself. The isotopic fractionation factor a higher light isotope abundance.
(a) is introduced to represent the extent of fractionation International general isotope standards are issued by
of isotopes between two phases. It is defined as the ratio the International Atomic Energy Agency (IAEA) and the
of isotope ratios in one phase to the other coexisting U.S. National Institute of Standards and Technology
phase. For example, in a system consisting of phase A (NIST). The delta values of hydrogen and oxygen iso-
and phase B, the oxygen isotope fractionation factor can topes are calculated using the value for Standard Mean
be defined as Ocean Water (SMOW), which has 2H/1H of (155.76 ±
0.10) × 10–6, 18O/16O of (2005.20 ± 0.43) × 10–6, and 17O/16O
(18O/16O)A of (373 ± 15) × 10–6. Other hydrogen isotope standard
αA–B =
(18O/16O)B samples include SLAP, GISP, NBS-22, and NBS-30.

et al., 2019), carbon isotopes are considered useful in Thus, relative isotopic abundances are reliable pa-
identifying natural and synthetic diamonds (Wang et rameters for determining geographic origin and offer
al., 2014), and radioactive isotopes have also been used a sound complement to traditional methods.
to determine the ages of gemstones (Link, 2015).
Traditional methods using the parameters of in- DATA AND CALCULATION
clusions, optical characteristics, and trace elements In all, 120 sets of hydrogen and oxygen isotope data
are often not enough to solve the problems of geo- (some lacking hydrogen data) for dolomite-related
graphic origin determination of nephrite. Isotopic nephrites were collected from all known related pub-
analysis has provided geochemical and chronological lished studies, from a variety of researchers (table 2
information for all sorts of geological samples: Stable
and figure 4), to illustrate geographic origin discrimi-
isotopes can be used to study gemstone origin (source
nation with stable isotopic ratios.
materials, formation process, and geographical local-
Hydrogen and oxygen isotope delta values of
ities), whereas radioactive isotopes can be utilized to
determine the formation ages. The stable isotope nephrite can be used to calculate the corresponding
study of dolomite-related nephrite in our works, to- values of its formation fluids. Hydrogen isotope frac-
gether with previous studies on corundum and emer- tionation of tremolite relative to water is not affected
ald (Giuliani et al., 1998, 2000, 2005, 2007; Wang et by temperature in the approximate range of 350° to
al., 2019), show that the geographic origin character- 650°C (Graham et al., 1984), and thus
istics of isotopes in gemstones can be explained from
their formation environment and formation process. 103lnαTr–H = –21.7±2 (1)
2O

270 STABLE ISOTOPE RATIOS OF DOLOMITE-RELATED NEPHRITE GEMS & GEMOLOGY SUMMER 2020
while oxygen isotope fractionation (Zheng, 1993, culated from the delta values of corresponding
1995) can be expressed as nephrite, which is acquired by isotope determination
(see box B).
103lnαTr–H = (3.95 × 106/T2) – (8.28 × 103/T) + 2.38 (2)
2O
GEOGRAPHIC ORIGIN CHARACTERISTICS
In both equations, aTr–H2O is the isotopic fractionation Vitim in Russia, Chuncheon in South Korea, and Xin-
factor (see box A) between the nephrite and its forma- jiang and Qinghai in China are the four most impor-
tion fluid, and T is the absolute temperature (K) of the tant dolomite-related nephrite source areas. The
nephrite-forming system. The nephrite formation relative abundances of the hydrogen and oxygen iso-
temperature is confined to approximately 223°–425°C, topes of nephrites from these regions differ signifi-
especially near 350°C (Tang et al., 1994; Yui and cantly (figure 4). In particular, oxygen isotope δ18O
Kwon, 2002; Chen et al., 2014; Liu et al., 2016), by values (see figure 4 and table 2) range from −20.0‰ to
methods using the homogenization temperatures of −14.6‰, −9.9‰ to −8.2‰, 0.5‰ to 7.9‰, and 11.4‰
tremolite fluid inclusions (Liu et al., 2011a; Chen et to 12.6‰, respectively, without any overlap. Cowell
al., 2014), the combination of the pyrite decrepitation in Australia is considered another large dolomite-re-
temperature and calcite homogenization temperature lated nephrite deposit but is seldom studied. The only
(Wang et al., 2007; Xu and Wang, 2016), the mineral δ2H–δ18O data (see figure 4 and table 2) fall within the
assemblage (Yang, 2013), and isotopes (Yui et al., range of Xinjiang placer nephrite; nevertheless, the δ2H
1988). Thus, the value of 350°C was used to calculate values are significantly higher than those of Xinjiang
the isotopes of fluids from which nephrite forms. primary nephrite.
Both the isotope fractionation factor aTr–H2O and The nephrites from Xinjiang, distributed in a belt
delta values (δ18O, δ2H) are defined after isotope ratios longer than 1300 km, show convergent hydrogen and
(18O/16O, 2H/1H) of nephrite and its formation fluids oxygen isotopic characteristics. The isotope delta
(see box A). Thus, the delta values of the nephrite- values of their primary dolomite-related nephrites
forming fluids (δ18OTr–H2O, δ2HTr–H2O; table 2) can be cal- are covered by placer ones (figure 4).

-40

-60

-80

-100
Figure 4. Hydrogen and
oxygen isotopic compo-
-120
sitions (top), and oxygen
δ2H(‰)

isotopic composition
-140
only (bottom), of
dolomite-related
-160 Alamasi, Xinjiang Sanchakou, Qinghai Kunlun, Xinjiang (placer?) nephrites around the
Xiuyan, Liaoning Chuncheon, Korea Yulongkashi, Xinjiang (placer) world.
Cowell, Australia Złoty Stok, Poland Tanghe, Hebei (secondary/placer)
-180
Dahua, Guangxi Vitim, Russia Kalakashi, Xinjiang (placer)

Xinjiang, n=23 Xiuyan, n=3

Vitim, n=9 Luodian, Guizhou, n=10

-20 -15 -10 -5 0 5 10 15 20

δ O(‰)
18

STABLE ISOTOPE RATIOS OF DOLOMITE-RELATED NEPHRITE GEMS & GEMOLOGY SUMMER 2020 271
BOX B: ANALYTICAL METHODS FOR DETERMINING OXYGEN AND
HYDROGEN ISOTOPE CONCENTRATIONS
Isotope concentrations are commonly meas-
ured with a mass spectrometer operating on
the principle that the degree of deflection of
charged particles in a magnetic field is in- Ionization and
Sample
acceleration
versely proportional to the mass-to-charge
ratio (m/z) (figure B-1). Generally, mass spec-
trometers can be divided into four parts: the Mass separation
(electromagnetic deflection)
sampling system, the ion source, the mass an-
alyzer, and the detector.
Stable isotope analysis has advanced from
macroanalysis to microanalysis and now in-
cludes methods of static mass spectrometry, Software
analysis
laser ablation (multi-collector)–inductively
Detector
coupled plasma–mass spectrometry (LA-
(MC)-ICP-MS), and secondary ion mass spec-
trometry (SIMS). High accuracy and low
sample loss make these technologies suitable Figure B-1. The basic schematic of mass spectrometry. Modified from
for isotopic analysis of gemstones. chem.libretexts.org.

Samples from some relatively small deposits such ids are released from magma (magmatic fluids) or
as Xiuyan in Liaoning Province, Złoty Stok in Poland, metamorphism (metamorphic fluids) due to changes
and Dahua in Guangxi Province (figure 4) show in temperature and pressure. They also can be mete-
slightly higher δ2H values than those of Qinghai oric waters (including rainwater, lake water, seawater,
nephrite and Xinjiang primary nephrite. Their ranges river water, glacial water, and shallow groundwater)
of δ18O values cover that of Sanchakou samples but do heated by geological processes.
not overlap with Xinjiang primary nephrite. Fortu- The original characteristics of the hydrogen and
nately, nephrites from these three regions typically oxygen isotopes of nephrite mainly result from the
have their own distinct appearances. Luodian nephrite ore-forming fluids. The calculated δ2HTr–H2O and
from Guizhou has notably higher δ18O values than the δ18OH2O values of hydrothermal fluids forming the
others (no δ2H value data have been collected). In re- Vitim and Chuncheon nephrites plot near the Craig
cent years, secondary nephrite has been found in the line2 (figure 5), indicating that their predominant ore-
Tanghe River in Hebei Province. It is speculated to be forming fluids were meteoric waters in an environ-
a dolomite-related nephrite according to the regional ment with a high fluid/rock ratio (Yui and Kwon,
geology, field observation, and petrographic analysis 2002; Burtseva et al., 2015).
(Chen et al., 2014). Its hydrogen and oxygen isotope For the Xinjiang nephrite, magmatic fluid, mete-
ratios are completely isolated from others in the plot oric water, and metamorphic water are all possible
of δ2H–δ18O (figure 4) by low δ2H and high δ18O values. candidates for the ore-forming fluids (figure 5), and
a low fluid/rock ratio is indicated (Yui and Kwon,
NEPHRITE-FORMING FLUIDS FROM 2002; Liu et al., 2011a, 2011b, 2016). The δ18OH2O val-
MAGMATIC WATER AND METEORIC WATER ues of the nephrite-forming fluids for Alamasi
nephrite, which occurs in granite-dolomite contact
Fluids containing gases, liquids, and silicate compo-
zones (Liu et al., 2010, 2011a), decrease in the con-
sitions always occur as the most active parts of geo-
logical processes. They are composed mainly of H2O,
CO2, NaCl, metal components, silicate compositions, 2
The Craig line, also referred to as the meteoric water line, represents
and organic matter. The fluids that correspond to the relationship between δ2H and δ18O of meteoric water—i.e., δ2H =
nephrite formation are hydrothermal fluids, which 8δ18O+10 (Craig, 1961). The kaolinite line (Zheng and Chen, 2000)
shown in figure 5 represents the relationship between δ2H and δ18O
refer to gas-liquid two-phase systems having their of kaolinite in weathering profile (i.e., δ2H = 7.5δ18O–220). Most of
own temperatures and pressures. Hydrothermal flu- the soil samples in nature fall on or near the kaolinite line.

272 STABLE ISOTOPE RATIOS OF DOLOMITE-RELATED NEPHRITE GEMS & GEMOLOGY SUMMER 2020
tact zone in the order of granite → nephrite → wall values for green nephrite fluids than for white ones
rock. The δ18OH2O values of magmatic fluids, seldom in the Alamasi deposit (Wan et al., 2002) provide an-
influenced by crustal rocks during intrusion, should other indicator that oxygen isotopes decrease from
equal the high values of the Xinjiang nephrite-form- granite to wall rock. However, the δ18OH2O value
ing fluids (figure 5). The δ18Odol values of wall rock should have increased gradually if water unilaterally
are far lower than those of common carbonates of diffused from the granite to the wall rock, since
sedimentary origin, at only 6.1‰ (Wan et al., 2002). water in equilibrium with nephrite is enriched or
Then, the δ18OH2O value for the water in equilibrium slightly depleted in 18O (depending on the tempera-
with wall rock is 1.6‰ (1000 lnadol–H2O = 3.06 × ture, calculated according to Equation 2 with T
106/T2−3.24 after Zheng and Chen (2000), assuming around 350°C). Considering that the δ2H value of the
that the temperature for the wall rock during Alamasi nephrite is negatively related to the δ18O
nephrite formation was between 252° and 295°C). value (figure 4), the conflict can be explained by du-
This value is lower than those of the fluids in equi- alistic fluid sources. One is post-magmatic hy-
librium with most of the Xinjiang nephrite (figure drothermal fluids provided by the granite forming
5). In addition, considering the characteristics of the the nephrite, while the other must be the meteoric
chemical zoning (Liu et al., 2010), the higher δ18OH2O water from the dolomite marble.

Figure 5. δ2HH2O and

0
δ18OH2O data of dolomite-
related nephrite-forming
ne

-20
fluids at 350°C. Nephrites
Li

3 Metamorphic H 2 O
g

from Alamasi, Yulong-

ai

300-600°C
Cr

-40
kashi, and Kalakashi show
a trend marked by a set of
2
-60 arrows from the bottom
1 right to the top left, which
Primary
-80
magmatic can be explained as dualis-
ne
δ2HH O(‰)

H2O tic fluid sources. Compared

Li
ite

-100 with the primary nephrite

in
2

ol

from Alamasi, the hydro-

Ka

-120 gen isotope of placer

nephrite changes greatly
-140
(indicated by the vertical
arrow) without distinct
-160
δ18OH2O variations, which
are caused by fluid-rock re-
action possibly. The el-
-180
lipses represent hydrogen
and oxygen isotopes of
-200
water from (1) the Qiemo
-20 -15 -10 -5 0 5 10 15 20
River basin (Wang et al.,
δ OH O(‰)
18
2013), covering the A’erjin-
2
shan region; (2) the Hetian
River basin and Keliya
Alamasi, Xinjiang Sanchakou, Qinghai Kunlun, Xinjiang (placer?) River basin, covering the
Xiuyan, Liaoning Chuncheon, Korea Yulongkashi, Xinjiang (placer)
Alamasi, Agejugai, and
Hetian regions; and (3) a
Cowell, Australia Złoty Stok, Poland Tanghe, Hebei (secondary/placer) pond in the Taklimakan
Desert hinterland (Li et al.,
Dahua, Guangxi Kalakashi, Xinjiang (placer) Vitim, Russia
2006).

STABLE ISOTOPE RATIOS OF DOLOMITE-RELATED NEPHRITE GEMS & GEMOLOGY SUMMER 2020 273
TABLE 2. Hydrogen and oxygen isotope delta values of dolomite-related nephrites.

δ18OH2O δ2HH2O
No. δ18O (‰) δ2H (‰) Locality Description Mass spectrometer Reference
(350°C) (‰) (350°C) (‰)
1 3.8 −86.7 4.5 −65 White
2 3.2 −83 3.9 −61.3 White
3 6.1 −93.1 6.8 −71.4 White-green
4 4.6 −89 5.3 −67.3 White-green
5 3.5 −85.1 4.2 −63.4 White-green
6 3.6 −85.9 4.3 −64.2 Alamasi, Yutian County, White-green
MAT-252 Liu et al. (2011a)
7 6.2 −94.7 6.9 −73 Hetian (Xinjiang, China)a
White-green
8 4.1 −90.2 4.8 −68.5 Green
9 3.6 −85 4.3 −63.3 Green
10 4.9 −91.6 5.6 −69.9 Green
11 4.8 −90.4 5.5 −68.7 Green
12 3.8 −86.2 4.5 −64.5 Green
13 3.8 4.5 White
14 3.7 4.4 White-green Wan et al. (2002)
15 3.6 4.3 Green
16 2.3 3.0 Mutton-fat
17 5.8 6.5 Agejugai, Hetian County, White
Wan et al. (2002)
18 5.6 6.3 Hetian (Xinjiang, China) White-green
19 6.5 7.2 Greenb

Yecheng County, Kashi

20 5.3 6.0 Green Wan et al. (2002)
(Xinjiang, China)

Datong, Tashiku’ergan County,

21 4.6 5.3 Green Wan et al. (2002)
Kashi (Xinjiang, China)

Ruoqiang County,
22 4.4 5.1 Bayinguoleng (Xinjiang, China) White-green Wan et al. (2002)

Qiemo County, Bayinguoleng

23 3.4 4.1 White-green Wan et al. (2002)
(Xinjiang, China)

24 3.9 4.6 Light brown

25 4.8 5.5 Brown
26 4.7 5.4 Brown
27 5.6 6.3 White-green
28 4.6 5.3 Green
29 3 3.7 Green
Tashisayi Qiemo County,
30 3.6 4.3 Green MAT-252 Wu (2016)c
Bayinguoleng (Xinjiang,
31 3.7 4.4 China)
32 3.1 3.8
33 3.9 4.6 Green
34 4 4.7 Light brown
35 4.6 5.3
36 4.8 5.5

37 5.2 −71.8 5.9 −50.1 Yulongkashi, Hetian (Xinjiang, White, placer

China) MAT-252 Liu et al. (2011b)
38 3.7 −67.3 4.4 −45.6 White, placer
39 5.6 −72.4 6.3 −50.7 White-green, placer
40 1.1 −55.7 1.8 −34 Kalakashi, Hetian (Xinjiang, White-green, placer
China) MAT-252 Liu et al. (2011b)
41 5 −71.4 5.7 −49.7 White-green, placer
42 2.9 −65.7 3.6 −44 White-green, placer

274 STABLE ISOTOPE RATIOS OF DOLOMITE-RELATED NEPHRITE GEMS & GEMOLOGY SUMMER 2020
δ18OH2O δ2HH2O
No. δ18O (‰) δ2H (‰) Locality Description Mass spectrometer Reference
(350°C) (‰) (350°C) (‰)
43 3.2 −68.7 3.9 −47 Black, placer
44 2.4 −63.3 3.1 −41.6 Black, placer
MAT-252 Liu et al. (2011b)
45 4.5 −69.3 5.2 −47.6 Black, placer
46 3.1 −67.1 3.8 −45.4 Black, placer
47 0.8 −97 1.5 −75.3 Green, placer
48 7.3 −67 8.0 −45.3 Black, placer
49 2.7 −80 3.4 −58.3 Black, placer
50 6.6 −77 7.3 −55.3 Black, placer
Kalakashi, Hetian (Xinjiang,
51 3.6 −87 4.3 −65.3 Green, placer
China)
52 6.7 −58 7.4 −36.3 Black, placer
53 3 −86 3.7 −64.3 Black, placer MAT-252 Liu et al. (2016)d
54 4.9 −85 5.6 −63.3 Green, placer
55 2.2 −100 2.9 −78.3 Black, placer
56 7.9 −79 8.6 −57.3 Black, placer
57 2.5 −88 3.2 −66.3 Green, placer
58 4.3 −109 5.0 −87.3 Green, placer
59 1.6 −93 2.3 −71.3 Green, placer
Primary or placer
60 0.5 to 2.3 −108 to −124 1.24 to 3.04 −86.3 to −102.3 Kunlun (Xinjiang, China) Yui and Kwon (2002)
unknown
61 12.3 −76.9 13.0 −55.2
Dahua (Guangxi, China) Xu and Wang (2016)
62 10.5 −79.8 11.2 −58.1
63 −8.7 −108 −8.0 −86.3
64 −8.4 −114 −7.7 −92.3
65 −9.9 −105 −9.2 −83.3
66 −9 −107 −8.3 −85.3
67 −8.2 −108 −7.5 −86.3 Chuncheon (South Korea) Yui and Kwon (2002)
68 −8.6 −112 −7.9 −90.3
69 −8.9 −109 −8.2 −87.3
70 −9.3 −110 −8.6 −88.3
Yui and Kwon (2002)
71 −9.2 −109 −8.5 −87.3
Cowell, South Australia
72 3.4 −57 4.1 −35.3 Yui and Kwon (2002)
(Australia)

73 −15.52 −119.3 −14.8 −97.6

74 −16.8 −16.1
75 −17.24 −178.5 −16.5 −156.8
76 −15.51 −14.8
Vitim area, Buryatia (Russia) MAT-253 Burtseva et al. (2015)
77 −14.95 −14.2
78 −14.93 −133.2 −14.2 −111.5
79 −15.1 −14.4
80 −14.58 −13.8
81 −18.63 −17.9
Semi-nephrite with
82 −17.33 −16.6 prismatic coarse-
grained tremolite
Semi-nephrite with
83 −20.02 −19.3 Vitim area, Buryatia (Russia) prismatic coarse- MAT-253 Burtseva et al. (2015)
grained tremolite
Semi-nephrite with
84 −17.24 −16.5 prismatic coarse-
grained tremolite

STABLE ISOTOPE RATIOS OF DOLOMITE-RELATED NEPHRITE GEMS & GEMOLOGY SUMMER 2020 275
TABLE 2 (continued). Hydrogen and oxygen isotope delta values of dolomite-related nephrites.
δ18OH2O δ2HH2O
No. δ18O (‰) δ2H (‰) Locality Description Mass spectrometer Reference
(350°C) (‰) (350°C) (‰)
85 10.2 −76.4 (2)e 10.9 −54.7
86 8.3 −76.2 (3) 9.0 −54.5 Złoty Stok, Lower Silesian
MAT-253 Gil et al. (2015a)
87 10.4 −77.2 (3) 11.1 −55.5 (Poland)

88 10.2 −74.6 (3) 10.9 −52.9

Adamo and Bocchio

89 −113 ± 4.8 −91.3 Val Malenco, Sondrio (Italy)
(2013)

90 10 −70 10.7 −48.3

91 9.3 −74 10.0 −52.3
92 8.5 −74 9.2 −52.3
93 8.1 −72 8.8 −50.3
94 13.3 −70 14.0 −48.3
Wang et al. (2007)
95 11.7 −73 12.4 −51.3
96 10.4 −76 11.1 −54.3 Xiuyan (Liaoning, China)
97 10.3 −72 11.0 −50.3
98 9.1 −74 9.8 −52.3
99 9 −70 9.7 −48.3
100 12.4 13.1 White
101 9.3 10.0 Yellow-white Wan et al. (2002)
102 8.7 9.4 Yellow
103 11.4 −86 12.1 −64.3 Green
104 12.3 −87 13.0 −65.3 Green-white
105 12.2 −78 12.9 −56.3 Sanchakou (Qinghai, China) White MAT-251EM Zhou (2006)

106 12.6 −84 13.3 −62.3 “Water line” in white

nephritef
107 7.8 −144 8.5 −122.3 Placer/secondary
108 15.4 −135 16.1 −113.3 Placer/secondary
Tanghe (Hebei, China) Chen et al. (2014)
109 18.9 −109 19.6 −87.3 Placer/secondary
110 11 −143 11.7 −121.3 Placer/secondary
111 15.3 16.0 Green-white
112 14.3 15.0 Green
113 15.6 16.3
114 16.5 17.2
115 14.7 15.4 White
Luodian (Guizhou, China) MAT-251EM Yang (2013)
116 14.5 15.2 White
117 14.1 14.8 White
118 14.6 15.3
119 15.5 16.2 Green-white
120 16.3 17.0
a
The deposits in Hetian and Kashi are counted in the West Kunlun region, while those in Bayinguoleng are counted in the A'erjinshan rengion.
b
The sample is marked as “Bi yu” in Chinese in the original reference, which mostly equates with serpentine-related nephrite. However, we tend to believe
the original authors meant a nephrite with dark green color.
c
Authors Kong Gao, Ting Fang, and Yuanyuan Wang once participated in the research project sponsoring the thesis. Therefore, we can supplement the
content of the original literature, which is not detailed enough.
d
Only those tremolite contents higher than 99 wt.% are chosen from Liu et al. (2016). However, it is not ruled out that individual samples may be
serpentine-related since their Fe contents can be high.
e
The figure in parentheses is the number of samples tested. The number before the parentheses is the average value.
f
A “water line” refers to the band in nephrite that is more transparent than the matrix. It is composed of prismatic coarse-grained tremolite crystals parallel
to each other.

276 STABLE ISOTOPE RATIOS OF DOLOMITE-RELATED NEPHRITE GEMS & GEMOLOGY SUMMER 2020
NEPHRITE-FORMING FLUIDS MODIFIED BY observations, indicate that the hydrothermal fluid
METAMORPHISM OR METASOMATISM forming Luodian nephrite derived from either dia-
The hydrogen and oxygen isotopes of nephrites from base intrusion (Yang et al., 2012; Zhang et al., 2015)
Xiuyan (Duan and Wang, 2002; Wan et al., 2002; or seawater circulation driven by diabase intrusion
Wang et al., 2007) and Złoty Stok (Gil et al., 2015a) (Li et al., 2014). A comparable process occurred at
overlap with each other to some extent (figure 4). Sanchakou: The water in the sediments convected
Their calculated fluid isotopes plot in the regional with magmatic hydrothermal fluids (Zhou, 2006), or
metamorphic water field (figure 5), which is in accor- the acidic magmatic hydrothermal fluids that ex-
dance with their geological environment. The Xi- tracted Mg from gabbro (Yang et al., 2013) reacted
uyan nephrite occurs not far from the famed with wall rocks and formed nephrite. Obviously, the
serpentine jade deposit formed from metamorphic hydrothermal fluids that formed these nephrites
hydrothermal fluids (Wu et al., 2014). Silicon isotope were no longer the original magmatic hydrothermal
studies support the interpretation that the formation fluids, but rather the fluids that had been modified
of the Xiuyan nephrite was related to metamorphic by metasomatism.
fluids (Duan and Wang, 2002; Wu et al., 2014). At
Złoty Stok, some geological bodies related to serpen- XINJIANG PLACER NEPHRITE ISOTOPES AND
tine occur not far from the dolomite-related nephrite FLUID-ROCK REACTION
deposit (Gil et al., 2015a,b). The Xinjiang placer nephrites, which are mainly dug
Like the nephrite-forming fluids of Xiuyan and
out from paleo river beds flowing through the Take-
Złoty Stok, those of Dahua and Sanchakou plot in
lamagan Desert, differ from the primary ones by their
the metamorphic water field (figure 5). The δ18O val-
wide ranges of hydrogen and oxygen isotope ratios,
ues of the Dahua, Sanchakou, and Luodian nephrites
especially δ2H (figure 4). There are four factors poten-
are higher than others (with the exception of
tially influencing this difference:
Tanghe), and these deposits are related to basic ig-
neous rocks of diabase or gabbro (Zhou et al., 2006; • Impurities: Impurities may induce a conspicu-
Yang et al., 2012; Li et al., 2014; Zhang et al., 2015; ously high δ2H value (Liu et al., 2016), as well as
Xu and Wang, 2016), which is distinct from other a wide range of variation.
dolomite-related nephrites. The presence of siliceous • Compositional effect: Most of the placer
components in the wall rocks is another common nephrite tested featured high Fe (Liu et al.,
feature for these three deposits. The wall rock for 2011b, 2016), which can result in a composi-
Dahua nephrite is a suite of interbedded layers of cal- tional effect on hydrogen isotope fractionations
cirudite, calcarenite, and micrite mixed with laminar in a tremolite-H2O system (Vennemann and
siliceous rocks and paramoudra (Xu and Wang, 2016). O’Neil, 1996).
Yang et al. (2013) discussed the relationship between • Complicated derivations: Since several primary
nephrite formation and siliceous veins in the San- deposits occur in the upper reaches of the Yu-
chakou deposit. The country rocks around the Luo- longkashi and Kalakashi Rivers, the placer
dian nephrite are siliceous clayey micrites and cherty nephrite might come from different primary de-
limestones (Yang et al., 2012; Li et al., 2014). These posits, even including serpentine-related
silicalites compensate for the Si shortage during the nephrite (Liu et al., 2016).
formation of nephrite from basic rocks. For Luodian
• Fluid-rock reaction: The δ2H–δ18O trends of
nephrite, this is supported by the δ18O equilibrium
some of the Xinjiang placer nephrites are similar
between quartz and nephrite. The δ18OQz value of the
to those of the Alamasi nephrite (figure 4). The
quartz from the deposit is 22.4‰ (Yang, 2013). Thus,
δ18O value, which is mainly controlled by the
the calculated δ18OTr value for tremolite by the
nephrite itself (Yui et al., 1990), has remained
quartz-tremolite fractionation equation 103lnaQz–Tr =
nearly constant after nephrite formation due to
2.25 × 106/T2 + 0.46 (Zheng, 1995) at 350°C equals
its high closure temperature of 424°C (Brady,
16.15‰, which is in the range of its nephrite δ18O
1995).
value = 14.1‰–16.5‰ (Yang, 2013). The speculation
of compensation is also supported by the Si isotope The closure temperature can be understood as the
accordance between the nephrite and the siliceous lowest temperature of isotope diffusion or loss. That
veins, paramoudra, and silicalites (δ30Si = 1.1‰–1.7‰; is, the δ18O value of the placer nephrite is almost
Yang, 2013). These values, in combination with field equal to that of the primary nephrites. The δ2H value

STABLE ISOTOPE RATIOS OF DOLOMITE-RELATED NEPHRITE GEMS & GEMOLOGY SUMMER 2020 277
of the placer nephrite, however, can be enhanced by lower iron concentration than actinolite, have high
the reaction between meteoric water (desert water chemical reactivity as well (Pacella et al., 2015). Thus,
that has been fractionated by evaporation; figure 5) the hydrogen isotope ratio can re-equilibrate at low
and rock (nephrite). temperature between the placer nephrite and mete-
The hydrogen in hydrous minerals diffuses rap- oric water, enhancing the δ2H value of the former.
idly and shows a closure temperature, below which
it will no longer diffuse and change its composition, CONCLUSIONS
in cooling metamorphic rocks far below the forma- On the basis of formation environment and forma-
tion temperature of the mineral assemblages (Gra- tion process, hydrogen and oxygen isotope ratios of
ham, 1981). The closure temperature (Tc) for nephrites from around the world can be analyzed.
hydrogen isotope volume diffusion can be expressed These isotope ratios, even for oxygen alone, appear
as (Dodson, 1973): to be discrimination criteria for the geographic origin
AτD0 determination of dolomite-related nephrites, espe-
Tc = R/[Eln( )] (3) cially those from Vitim (Russia), Chuncheon (South
a2
Korea) and the Xinjiang Uyghur Autonomous Region
where the time constant is and Qinghai Province of China. However, the
EdT nephrite δ18O values from Xiuyan, Dahua, and Złoty
τ = –RT2/( ) (4) Stok overlap. The isotopic ratio differences are
dt
mainly derived from the ore-forming fluids. The iso-
in which the activation energy for tremolite E = 71.5 topes of dolomite-related nephrites from Russia,
kJ/mol (Graham et al., 1984; Farver, 2010); the gas South Korea, Xinjiang, and Qinghai Province in-
constant R = 8.314 J/mol/K; the anisotropic factor for crease in sequence, and the ore-forming fluids vary
cylinder case A = 27 (Dodson, 1973); the pre-exponen- in the order of meteoric water → mixture of mag-
tial factor in the Arrhenius relationship D0 = 1.21 × matic water and meteoric water → mixed water that
10–8 m2/s, calculated from figure 5 of Graham et al. experienced metamorphism to some extent or is
(1984). Thus, the closure temperature can be as low even dominated by metamorphic fluid. Furthermore,
as 61°C (calculated by grain radius a = 0.5 μm, cooling the hydrogen isotope of the placer nephrite from the
rate dT/dt = –10°C/d) to 123°C (calculated by a = 1 Hetian region of Xinjiang could have been modified
μm, dT/dt = –50°C/d). Since the radius of nephrite by meteoric water when it was buried in paleo river
tremolite can be smaller, the calculated closure tem- beds flowing through the desert.
perature will decrease. Furthermore, an experiment Based on this limited data set, we show that iso-
showed that tremolite can dissolve at a pH of 6.9 at a tope ratio analysis is a new gem origin identification
low temperature of 37°C (Diedrich et al., 2014). tool for gemologists studying nephrite (similar to
Grapes and Sun (2010) suggested that higher porosity what other researchers have shown for emerald and
created by actinolite dissolution results in an expo- corundum). However, we point out with caution that
nential increase in weathering. Tremolite fibers, with more data is needed to optimize our findings.

ABOUT THE AUTHORS ACKNOWLEDGMENTS

Dr. Gao is a research group leader of the National Gems & Jew- The authors wish to express their gratitude for reviewers Dr. Yunbin
ellery Technology Administrative Center (NGTC) and director of Guan at Caltech, Dr. Andy Shen, Dr. George Rossman, and Dr. Ah-
the application R&D department at the Shenzhen Research Insti- madjan Abduriyim for their thorough reviews of this manuscript and
tute, NGTC. Ms. Fang is a postgraduate at the University of Sci- helpful comments and criticism. We thank technical editors Jennifer
ence and Technology Beijing. Dr. Lu is chief scientist of NGTC. Mr. Stone-Sundberg and Tao Z. Hsu for their elaborate work handling
Lan is principal instrument researcher of NGTC, director of Shen- the manuscript. We also appreciate Professor Guanghai Shi for his
zhen Research Institute, and vice director of Shenzhen Lab, constructive suggestions.
NGTC. Mr. Zhang is a research group leader of NGTC and vice di-
rector of Shenzhen Lab, NGTC. Ms. Wang is an analytical techni-
cian, and Mr. Chang is an engineer, at the China Hetian Jade
Product Quality Supervision Testing Center (Xinjiang).

278 STABLE ISOTOPE RATIOS OF DOLOMITE-RELATED NEPHRITE GEMS & GEMOLOGY SUMMER 2020
REFERENCES
Abduriyim A., Kitawaki H. (2006) Applications of laser ablation– 3, pp. 251–270, http://dx.doi.org/10.1007/s00126-006-0105-3
inductively coupled plasma–mass spectrometry (LA-ICP-MS) Graham C.M. (1981) Experimental hydrogen isotope studies III:
to gemology. G&G, Vol. 42, No. 2, pp. 98–118, Diffusion of hydrogen in hydrous minerals, and stable iso-
http://dx.doi.org/10.5741/GEMS.42.2.98 tope exchange in metamorphic rocks. Contributions to Min-
Adamo I., Bocchio R. (2013) Nephrite jade from Val Malenco, Italy: eralogy and Petrology, Vol. 76, No. 2, pp. 216–228,
Review and update. G&G, Vol. 49, No. 2, pp. 2–10, http://dx.doi.org/10.1007/BF00371961
http://dx.doi.org/10.5741/GEMS.49.2.98 Graham C.M., Harmon R.S., Sheppard S.M.F. (1984) Experimental
Brady J.B. (1995) Diffusion Data for Silicate Minerals, Glasses, and hydrogen isotope studies: Hydrogen isotope exchange between
Liquids. American Geophysical Union, Washington, DC. amphibole and water. American Mineralogist, Vol. 69, No. 1–
Burtseva M.V., Ripp G.S., Posokhov V.F., Murzintseva A.E. (2015) 2, pp. 128–138.
Nephrites of East Siberia: Geochemical features and problems Grapes R.H., Yun S.T. (2010) Geochemistry of a New Zealand
of genesis. Russian Geology and Geophysics, Vol. 56, No. 3, nephrite weathering rind. New Zealand Journal of Geology and
pp. 402–410, http://dx.doi.org/10.1016/j.rgg.2015.02.003 Geophysics, Vol. 53, No. 4, pp. 413–426.
Chen C., Yu X.J., Wang S.Q. (2014) Study on gemmological char- Harlow G.E., Sorensen S.S. (2005) Jade (nephrite and jadeitite)
acteristics and ore genesis of nephrite from Tanghe, Hebei and serpentinite: Metasomatic connections. International
Province. Journal of Gems and Gemmology, Vol. 16, No. 3, Geology Review, Vol. 47, No. 2, pp. 113–146,
pp. 1–11 (in Chinese with English abstract), http://dx.doi.org/10.2747/0020-6814.47.2.113
http://dx.doi.org/10.15964/j.cnki.027jgg.2014.03.001 Li K.X., Jiang T.L., Xing L.C., Zhou M.Z., Luo T.Y. (2014) A prelim-
Craig H. (1961) Isotopic variations in meteoric waters. Science, Vol. inary study on mineralogy and ore deposits genetical model of
133, No. 3465, pp. 1702–1703, http://dx.doi.org/10.1126/sci- luodian nephrite jade, Luodian, Guizhou Province, China. Acta
ence.133.3465.1702 Mineralogica Sinica, Vol. 34, No. 2, pp. 223–233 (in Chinese
Diedrich T., Schott J., Oelkers E.H. (2014) An experimental study with English abstract), http://dx.doi.org/10.16461/j.cnki.1000-
of tremolite dissolution rates as a function of pH and tempera- 4734.2014.02.013
ture: Implications for tremolite toxicity and its use in carbon Li W.P., Hao A.B., Zheng Y.J., Liu B., Yu D.S. (2006) Regional envi-
storage. Mineralogical Magazine, Vol. 78, No. 6, pp. 1449–1464, ronmental isotopic features of groundwater and their hydroge-
http://dx.doi.org/10.1180/minmag.2014.078.6.12 ological explanation in the Tarim Basin. Earth Science Frontiers,
Dodson M.H. (1973) Closure temperature in cooling geochrono- Vol. 13, No. 1, pp. 191–198 (in Chinese with English abstract),
logical and petrological systems. Contributions to Mineralogy http://dx.doi.org/10.3321/j.issn:1005-2321.2006.01.025
and Petrology, Vol. 40, No. 3, pp. 259–274, Link K. (2015) Age determination of zircon inclusions in faceted
http://dx.doi.org/10.1007/BF00373790 sapphires. Journal of Gemmology, Vol. 34, No. 8, pp. 692–700.
Duan T.Y., Wang S.Q. (2002) Study on stable isotopes of Xiuyan Liu Y., Deng J., Shi G.H., Lu T.J., He H.Y., Ng Y.N., Shen C.H.,
nephrite (tremolite). Acta Petrologica et Mineralogica, Vol. 21, Yang L.Q., Wang Q.F. (2010) Chemical zone of nephrite in Ala-
No. S1, pp. 115–119 (in Chinese with English abstract). mas, Xinjiang, China. Resource Geology, Vol. 60, No. 3, pp.
Farver J.R. (2010) Oxygen and hydrogen diffusion in minerals. Re- 249–259, http://dx.doi.org/10.1111/j.1751-3928.2010.00135.x
views in Mineralogy and Geochemistry, Vol. 72, No. 1, pp. Liu Y., Deng J., Shi G.H., Yui T.F., Zhang G.B., Abuduwayiti M.,
447–507, http://dx.doi.org/10.2138/rmg.2010.72.10 Yang L.Q., Sun X. (2011a) Geochemistry and petrology of
Gil G., Barnes J.D., Boschi C., Gunia P., Raczyński P., Szak- nephrite from Alamas, Xinjiang, NW China. Journal of Asian
mány G., Bendő Z., Péterdi B. (2015a) Nephrite from Złoty Earth Sciences, Vol. 42, No. 3, pp. 440–451,
Stok (Sudetes, SW Poland): Petrological, geochemical, and http://dx.doi.org/10.1016/j.jseaes.2011.05.012
isotopic evidence for a dolomite-related origin. The Cana- Liu Y., Deng J., Shi G.H., Sun X., Yang L.Q. (2011b) Geochemistry
dian Mineralogist, Vol. 53, No. 3, pp. 533–556, and petrogenesis of placer nephrite from Hetian, Xinjiang,
http://dx.doi.org/10.3749/canmin.1500018 Northwest China. Ore Geology Reviews, Vol. 41, No. 1, pp.
Gil G., Barnes J.D., Boschi C., Gunia P., Szakmány G., Bendö Z., 122–132, http://dx.doi.org/10.1016/j.oregeorev.2011.07.004
Raczyński P., Péterdi B. (2015b) Origin of serpentinite-related Liu Y., Zhang R.Q., Abuduwayiti M., Wang C., Zhang S.P., Shen
nephrite from Jordanów and adjacent areas (SW Poland) and its C.H., Zhang Z.Y., He M.Y., Zhang Y., Yang X.D. (2016) SHRIMP
comparison with selected nephrite occurrences. Geological Quar- U-Pb zircon ages, mineral compositions and geochemistry of
terly, Vol. 59, No. 3, pp. 457–472, http://dx.doi.org/10.7306/gq.1228 placer nephrite in the Yurungkash and Karakash River deposits,
Giuliani G., France-Lanord C., Coget P., Schwarz D., Cheilletz A., West Kunlun, Xinjiang, northwest China: Implication for a
Branquet Y., Giard D., Martin-Izard A., Alexandrov P., Piat D.H. magnesium skarn. Ore Geology Reviews, Vol. 72, pp. 699–727,
(1998) Oxygen isotope systematics of emerald: Relevance for its http://dx.doi.org/10.1016/j.oregeorev.2015.08.023
origin and geological significance. Mineralium Deposita, Vol. 33, Luo Z.M., Yang M.X., Shen A.H. (2015) Origin determination of
No. 5, pp. 513–519, http://dx.doi.org/10.1007/s001260050166 dolomite-related white nephrite through iterative-binary linear
Giuliani G., Chaussidon M., Schubnel H., Piat D.H., Rollion- discriminant analysis. G&G, Vol. 51, No. 3, pp. 300–311,
Bard C., France-Lanord C., Giard D., de Narvaez D., Rondeau http://dx.doi.org/10.5741/GEMS.51.3.300
B. (2000) Oxygen isotopes and emerald trade routes since an- Pacella A., Fantauzzi M., Turci F., Cremisini C., Montereali M.R.,
tiquity. Science, Vol. 287, No. 5453, pp. 631–633, Nardi E., Atzei D., Rossi A., Andreozzi G.B. (2015) Surface al-
http://dx.doi.org/10.1126/science.287.5453.631 teration mechanism and topochemistry of iron in tremolite as-
Giuliani G., Fallick A.E., Garnier V., France-Lanord C., Ohnen- bestos: A step toward understanding the potential hazard of
stetter D., Schwarz D. (2005) Oxygen isotope composition as a amphibole asbestos. Chemical Geology, Vol. 405, pp. 28–38,
tracer for the origins of rubies and sapphires. Geology, Vol. 33, http://dx.doi.org/10.1016/j.chemgeo.2015.03.028
No. 4, pp. 249–252, http://dx.doi.org/10.1130/G21261.1 Tang Y.L., Chen B.Z., Jiang R.H. (1994) Chinese Hetian Nephrite,
Giuliani G., Fallick A., Rakotondrazafy M., Ohnenstetter D., An- Xinjiang. Xinjiang People’s Publishing House, Xinjiang (in
driamamonjy A., Ralantoarison T., Rakotosamizanany S., Tse- Chinese).
heno M.R., Offant Y., Garnier V., Dunaigre C., Schwarz D., Vennemann T.W., O’Neil J.R. (1996) Hydrogen isotope exchange
Mercier A., Ratrimo V., Ralison B. (2007) Oxygen isotope sys- reactions between hydrous minerals and molecular hydrogen:
tematics of gem corundum deposits in Madagascar: Relevance I. A new approach for the determination of hydrogen isotope
for their geological origin. Mineralium Deposita, Vol. 42, No. fractionation at moderate temperatures. Geochimica et Cos-

STABLE ISOTOPE RATIOS OF DOLOMITE-RELATED NEPHRITE GEMS & GEMOLOGY SUMMER 2020 279
mochimica Acta, Vol. 60, No. 13, pp. 2437–2451, Guizhou. Journal of Mineralogy and Petrology, Vol. 32, No.
https://dx.doi.org/10.1016/0016-7037(96)00103-2 2, pp. 12–19 (in Chinese with English abstract),
Wan D.F., Wang H.P., Zou T.R. (2002) Silicon and oxygen isotopic http://dx.doi.org/10.3969/j.issn.1001-6872.2012.02.003
compositions of Hetian jade, Manasi green jade and Xiuyan old Yang T.X., Yang M.X., Liu H.L., Wu Y., Li J. (2013) New under-
jade (tremolite). Acta Petrologica et Mineralogica, Vol. 21, No. standing for Sanchahe nephrite deposit in East Kunlun. Jour-
S1, pp. 110–114 (in Chinese with English abstract). nal of Guilin University of Technology, Vol. 33, No. 2, pp.
Wang H.A.O., Krzemnicki M.S., Chalain J. (2016) Simultaneous 239–245 (in Chinese with English abstract),
high sensitivity trace-element and isotopic analysis of gem- http://dx.doi.org/10.3969/j.issn.1674-9057.2013.02.007
stones using laser ablation inductively coupled plasma time- Yui T.F., Kwon S.T. (2002) Origin of a dolomite-related jade deposit
of-flight mass spectrometry. Journal of Gemmology, Vol. 35, at Chuncheon, Korea. Economic Geology, Vol. 97, No. 3, pp.
No. 3, pp. 212–223. 593–601, http://dx.doi.org/10.2113/97.3.593
Wang H.A.O., Cartier L.E., Baumgartner L.P., et al (2018) A pre- Yui T.F., Yeh H.W., Lee C.W. (1988) Stable isotope studies of
liminary SIMS study using carbon isotopes to separate natural nephrite deposits from Fengtien, Taiwan. Geochimica et Cos-
from synthetic diamonds. Journal of Gemmology, Vol. 36, No. mochimica Acta, Vol. 52, No. 3, pp. 593–602,
1, pp. 38–43. https://dx.doi.org/10.1016/0016-7037(88)90321-3
Wang K., Graham I., Martin L., Voudouris P., Giuliani G., Lay A., ——— (1990) A stable isotope study of serpentinization in the
Harris S., Fallick A. (2019) Fingerprinting Paranesti rubies Fengtien ophiolite, Taiwan. Geochimica et Cosmochimica
through oxygen isotopes. Minerals, Vol. 9, No. 2, pp. 91, Acta, Vol. 54, No. 5, pp. 1417–1426,
http://dx.doi.org/10.3390/min9020091 https://dx.doi.org/10.1016/0016-7037(90)90165-H
Wang S.Q., Zhao C.H., Yu G., Yuan X.M., Duan T.Y. (2007) Xiuyan Zhang Y.D., Yang R.D., Gao J.B., Chen J., Liu Y.N., Zhou Z.R.
Jades in China. Science Press, Beijing (in Chinese). (2015) Geochemical characteristics of nephrite from Luodian
Wang W.X., Wang R.J., Li W.P., Yin X.L., Liu C.L. (2013) Analysis County, Guizhou Province, China. Acta Mineralogica Sinica,
of stable isotopes and hydrochemistry of rivers in Tarim Vol. 35, No. 1, pp. 56–64 (in Chinese with English abstract),
Basin. Hydrogeology & Engineering Geology, Vol. 40, No. 4, http://dx.doi.org/10.16461/j.cnki.1000-4734.2015.01.009
pp. 29–35 (in Chinese with English abstract), Zheng Y.F. (1993) Calculation of oxygen isotope fractionation in
http://dx.doi.org/10.16030/j.cnki.issn.1000-3665.2013.04.020 hydroxyl-bearing silicates. Earth and Planetary Science Letters,
Wang W.Y., D’Haenens-Johansson U., Smit K., Breeding C.M., Vol. 120, No. 3–4, pp. 247–263, http://dx.doi.org/10.1016/0012-
Stern R. (2014) Carbon isotope analysis of CVD synthetic gem 821X(93)90243-3
diamonds. In 2014 GSA Annual Meeting in Vancouver, British ——— (1995) Oxygen isotope fractionation in amphiboles. Science
Columbia. Geologica Sinica, Vol. 30, No. 1, pp. 1–11 (in Chinese with Eng-
Wu L.J. (2016) Study on gemological and mineralogical character- lish abstract).
istics and genesis of nephrite in Tiantai, Qiemo, Xinjiang. Mas- Zheng Y.F., Chen J.F. (2000) Stable Isotope Geochemistry. Science
ter’s thesis. China University of Geosciences, Beijing. Press, Beijing (in Chinese).
Wu Z.Y., Wang S.Q., Ling X.X. (2014) Characteristics and origin Zhong Y.P., Qiu Z.L., Li L.F., Gu X.Z., Luo H., Chen Y., Jiang
of nephrite from Sangpiyu, Xiuyan County, Lioaning Province. Q.Y. (2013) REE composition of nephrite jades from major
Acta Petrologica et Mineralogica, Vol. 33, No. S2, pp. 15–24 (in mines in China and their significance for indicating origin.
Chinese with English abstract). Journal of the Chinese Society of Rare Earths, Vol. 31, No.
Xu L.G., Wang S.Q. (2016) Gemological characteristics and genesis 6, pp. 738–748 (in Chinese with English abstract).
of Dahua nephrite. Acta Petrologica et Mineralogica, Vol. 35, Zhou Z.Y. (2006) Study on ore-forming tectonic setting and mech-
No. S1, pp. 1–11 (in Chinese with English abstract). anism of Sanchakou nephrite (tremolite Jade), East Kunlun.
Yang F.X., Abduriyim A. (1994) Honten jade and its marketing. Ph.D. thesis, Tongji University, Shanghai.
Journal of China Gemstone, Vol. 1, pp. 81–84. Zhou Z.Y., Liao Z.T., Ma T.T., Yuan Y. (2006) Study on the genetic
Yang L. (2013) Study on petro-mineral features and genetic mech- mechanism and material source of Sanchakou nephrite deposit
anism of Luodian jade, Guizhou Province. Ph.D. thesis. in East Kunlun. Contributions to Geology and Mineral Re-
Chengdu University of Technology, Chengdu. sources Research, Vol. 21, No. 3, pp. 195–198, 202 (in Chinese
Yang L., Lin J.H., Wang L., Tan J., Wang B. (2012) Petrochemical with English abstract), http://dx.doi.org/10.3969/j.issn.1001-
characteristics and genesic significance of Luodian jade from 1412.2006.03.010

280 STABLE ISOTOPE RATIOS OF DOLOMITE-RELATED NEPHRITE GEMS & GEMOLOGY SUMMER 2020
Lab Notes
Editors
Thomas M. Moses | Shane F. McClure

Cat’s-Eye ANDRADITE semitransparent to opaque and

The Carlsbad laboratory received a showed blue to greenish blue body-
49.65 ct semitranslucent to opaque colors with a network-like structure
brownish yellow oval cabochon (fig- of whitish zones (figures 3 and 4).
ure 1) for identification. The stone’s These beads were acquired at an orna-
most notable feature was its display mental gem material shop in Tokyo
of cat’s-eye phenomenon across the and were sold as “sea-blue chal-
dome. It had a refractive index over cedony,” but the resemblance to Lari-
the limit of the refractometer and a mar led to our investigation.
hydrostatic specific gravity of 3.74. The beads were identified as chal-
Raman spectroscopy was used to cedony based on spot RI readings of
assist with the identification, and the 1.53 to 1.54, SG values ranging from
Figure 1. A 49.65 ct brownish yel- 2.48 to 2.65, and microscopic features
data collected (figure 2) were matched
low andradite cabochon with such as parallel curved bands (figure
with andradite garnet from the GIA
chatoyancy across the dome. 4B). Raman spectroscopic features (fig-
reference database. Analysis by en-
ergy-dispersive X-ray fluorescence ure 5) also suggested that the beads
Quench-Crackled Dyed Blue were mainly composed of quartz crys-
(EDXRF) showed a composition con-
sistent with andradite garnet, CHALCEDONY Resembling Larimar tals and matched chalcedony (e.g., D.
Ca3Fe2(SiO4)3, further proof of the Recently, GIA’s Tokyo laboratory ex- Pop et al., “Raman spectroscopy on
stone’s identity. amined five round beads that were gem-quality microcrystalline and
Internally, the stone is host to
very fine and tightly packed needle-
Figure 2. Raman spectrum of the cat’s-eye andradite compared to the
like inclusions running parallel to the
known reference. Spectra are offset vertically for clarity.
base. With the stone cut as an oval
cabochon, light is reflected off the in-
clusions, creating a rolling eye across RAMAN SPECTRA
the dome. This phenomenon is com- 80000
mon in many gemstone species, and Andradite cat’s-eye
the demantoid variety of andradite Andradite reference
garnet has been known to display
INTENSITY (COUNTS)

60000
cat’s-eye phenomenon (Spring 2018
Lab Notes, pp. 58–59). This stone is a
larger example of cat’s-eye andradite
40000
garnet.
Nicole Ahline
20000

Editors’ note: All items were written by staff

members of GIA laboratories. 0

GEMS & GEMOLOGY, Vol. 56, No. 2, pp. 281–291.

2000 1800 1600 1400 1200 1000 800 600 400 200

© 2020 Gemological Institute of America RAMAN SHIFT (cm–1)

LAB NOTES GEMS & GEMOLOGY SUMMER 2020 281

RAMAN SPECTRA
25000
Quench-crackled dyed chalcedony
Larimar

20000

INTENSITY (COUNTS)
15000

10000

5000

0
0 200 400 600 800 1000 1200 1400
Figure 3. Two of five round beads RAMAN SHIFT (cm–1)
of quench-crackled dyed chal-
cedony and a pear-shaped Lari-
Figure 5. Raman spectra of quench-crackled dyed blue chalcedony (red
mar cabochon. The bottom bead
trace) and Larimar (blue trace). The dyed chalcedony Raman peaks
is 10.25 mm in diameter.
match those of chalcedony and not Larimar.

amorphous silica varieties from Roma- touched with a hot point. Under a and appearance. Larimar is a rare blue
nia,” Studia Universitatis Babeș- gemological microscope, the stones re- variety of pectolite with the ideal
Bolyai, Geologia, Vol. 49, 2004, pp. vealed whitish zones following frac- chemical formula of NaCa2Si3O8(OH)
41–52). Raman spectra of the beads tures. In order to consider the overall that displays white, green, and pale to
showed no characteristics of chryso- structure of such fractures, we cut sky blue colors. Blue color seen in
colla (Spring 2020 Gem News Interna- them in half. Cross sections of the Larimar is believed to be caused by
tional, pp. 188–189). Qualitative beads showed that the fractures with the presence of small amounts of Cu2+
EDXRF analysis detected peaks related whitish zones were concentrated on within its structure (e.g., R.E.
to Si, Fe, and Cu, indicating that Cu the surface (figure 6). Such a structure Woodruff and E. Fritsch, “Blue pecto-
was likely the color-causing element. is a known characteristic of quench- lite from the Dominican Republic,”
Ultraviolet-visible (UV-Vis) spectra crackled dyed chalcedony and was re-
also showed a broad band from 500 to ported previously (see Winter 2009 Lab
1000 nm that corresponded with Cu. Notes, p. 288). These beads probably Figure 6. Cross sections of two
There was no polymer detected by represent a new color variety. round beads of quench-crackled
Fourier-transform infrared (FTIR) spec- These quench-crackled dyed chal- dyed chalcedony, 10.25 mm in
troscopy and no reaction when cedonies resemble Larimar in color diameter. Note that the white
color is concentrated on the sur-
face and along internal fractures.
Figure 4. A: A quench-crackled dyed chalcedony bead containing
whitish zones following fractures; field of view 11.80 mm. B: Another A
bead showing a curved banding structure; field of view 12.80 mm. C: A
pear-shaped Larimar cabochon (also shown in figure 3); field of view
16.20 mm. Note that it has a mottled blue and white color and contains
other mineral inclusions.

A B C
B

282 LAB NOTES GEMS & GEMOLOGY SUMMER 2020

Winter 1989 G&G, pp. 216–225). We
compared these beads with Larimar
(the pear cabochon on the right in fig-
ure 3) to reveal their differences. The
Larimar cabochon showed a mottled
blue and white color with network-
like whitish zones and other natural
mineral inclusions (figure 4C) (again,
see Woodruff and Fritsch, 1989). It had
a spot RI reading of 1.60 and an SG of
around 2.80. On the other hand, the
quench-crackled dyed blue chal-
cedony beads had whitish zones fol- Figure 7. This 1.00 ct Fancy Deep greenish blue diamond (left) owes its
lowing a network of fractures (figure color to artificial irradiation, which is also evidenced by the color concen-
4B) and no other mineral inclusions. tration at the culet (right).
This material can easily be sepa-
rated from Larimar using standard
gemological testing such as RI and dence of NV centers forming from quent annealing conditions created
SG. Advanced testing is useful in con- that low-temperature heating. an unusual treated diamond for scien-
firming the cause of blue color and Here we see a combination of tific study and an interesting reinter-
the presence of polymer treatment. treatment conditions that generated pretation of a “nailhead” diamond—a
Makoto Miura an NV-related fluorescence on the term often applied to a poorly cut
pavilion and the crown; however, the stone that is dark on the table. Here,
original blue fluorescence of the nat- this effect can only be revealed under
DIAMOND ural, pre-irradiated diamond persists UV illumination.
on the table, which was “shielded”
Irradiated Blue Diamond with Sally Eaton-Magaña
during the irradiation (figure 8, right).
Interesting DiamondView Image The table experienced a lower radia-
We often receive lab-irradiated dia- tion dose, fewer NV centers formed as
monds at the laboratory that have ei- a result, and we observe a transition Multiple Radiation Stains Suggest
ther natural or laboratory-grown from the orange NV fluorescence to Interesting Geological Residency
origins (e.g., Summer 2018 Lab the intrinsic blue fluorescence. Thus, A rough diamond crystal weighing
Notes, pp. 215–216). However, we re- the DiamondView fluorescence col- 4.05 ct was observed with radiation
cently received a 1.00 ct Fancy Deep lected from the table creates an in- staining on its surface. Radiation
greenish blue irradiated diamond (fig- triguing image. staining is thought to occur when ra-
ure 7) with some unusual gemologi- The combination of the intrinsic dioactive fluids or minerals are adja-
cal features. The diamond’s spectra defect concentrations along with the cent to a diamond crystal in the earth.
and imaging showed that it had re- specific irradiation dose and subse- The radiation imparts damage to the
ceived artificial irradiation likely fol-
lowed by annealing, which created
unusually high concentrations of ni-
Figure 8. The DiamondView fluorescence images of the irradiated dia-
trogen vacancy (NV) centers.
mond collected from the pavilion (left) show a high concentration of NV
As with most lab-irradiated dia-
centers creating the orange fluorescence. On the table (right), the NV con-
monds, it was irradiated with the
centration has decreased sufficiently so that the NV-related orange fluo-
table side down; thus, the culet facing
rescence transitions to the diamond’s intrinsic blue fluorescence.
the beam received the highest dose of
irradiation while much of the table
facet was comparatively protected.
The irradiated pavilion shows orange
fluorescence due to a high concentra-
tion of NV centers (figure 8, left). Irra-
diated diamonds are often subjected
to some low-temperature annealing
after irradiation in order to “stabilize”
the defects (e.g., Spring 2018 Gem
News International, pp. 105–107);
however, we generally do not see evi-

LAB NOTES GEMS & GEMOLOGY SUMMER 2020 283

are subjected to heat, the stains will
turn to a brownish color. This partic-
ular diamond had both green and
brown radiation stains on the surface.
Nearly all of the stains existed in
pairs, one brown and one green, and
were of the same shape. This has been
reported previously (C.M. Breeding et
al., “Natural-color green diamonds: A
beautiful conundrum,” Spring 2018
G&G, pp. 2–27), and the proposed
mechanism is that the diamond crys-
tal is adjacent to a radioactive sub-
stance that imparts a green stain.
When the diamond crystal is shifted
slightly, the radioactive minerals cre-
ate a new stain of the same shape but
slightly displaced. The diamond was
heated before the second set of stains
had formed, which turned the initial
Figure 9. Radiation staining is visible on opposite sides of the rough dia- set of stains brown.
mond crystal. On the upper surface the green stain is closer to the bottom In this rough crystal, the stain
of the image, whereas on the lower surface the green stain is closer to the pairs are on opposite sides and in op-
top of the image. The arrows indicate the direction of displacement, posite directions (figure 9). The shift
which are in opposite directions on opposite sides of the rough crystal. must have occurred in the direction
Field of view 2 mm. from brown to green, which means
this crystal most likely rotated with
diamond lattice, leaving behind va- When initially formed, these stains respect to its environment. The other
cancies and interstitial carbon atoms. are green or greenish in color. If they explanation is that the host material
on either side shifted in opposite di-
rections by roughly the same amount.
Figure 10. A portion of one side of the 4.05 ct rough diamond (left; field of To examine the spectroscopic dif-
view 3 mm) shows the relative placement of several radiation stains that are ferences between the green and brown
detailed with PL mapping at right. A: The photographed image of a very dark radiation stains, we collected photo-
radiation stain and the corresponding PL map of diamond Raman intensity. luminescence (PL) maps using 532 and
B: The image of a brown radiation stain separated from its green counterpart 455 nm excitation in confocal mode.
and the corresponding PL map of its diamond Raman intensity. C: An image We collected data on several of the ra-
of a light radiation stain. Its PL map plots the TR12 intensity, instead of diation stains but will focus this dis-
Raman intensity as in A and B, which was much higher within the radiation cussion on three of them (figure 10).
stain than within the colorless section. The PL maps in A and B were col- Figure 10A has dark, almost black col-
lected with 532 nm excitation and plot the diamond Raman peak area at 573 oration in the center with dark green
nm, while the map in C with 455 nm excitation plots the TR12 peak area at and brown color around the periphery.
470 nm and normalized to the diamond Raman peak area. Figure 10B shows green and brown ra-
diation stains but with a colorless sec-
A 80k tion between them, and figure 10C
B has very light radiation staining.
Raman Intensity

In the radiation stain imaged in

Diamond

figure 10A, the diamond Raman peak

was quite broad and distorted within
B
the radiation stain compared to the
surrounding diamond (figure 11,
inset). The broadened and distorted
0
Raman peaks are indicators of the ra-
C 1.0
diation damage brought upon these
C A
TR12

areas and are consistent with prior ob-

servations of other very dark radiation
0 stains (e.g., S. Eaton-Magaña and K.S.

284 LAB NOTES GEMS & GEMOLOGY SUMMER 2020

PL SPECTRA
this is consistent with prior estimates
Diamond Raman of alpha radiation penetration (Eaton-
Diamond Raman
80000 Magaña and Moe, 2016).
In the radiation stains shown in
figure 10, the brown portions of the
INTENSITY (COUNTS)

60000 radiation stains displayed features

closer to the intrinsic diamond (lower
diamond Raman width of the radia-
40000 570 572 574 576 578
tion stain shown in figure 10A and its
corresponding spectra shown in figure
Colorless 11, along with higher intensities of
Radiation-related Green radiation stain GR1
Raman peak at 583 nm Brown radiation stain
H3 and NV– of radiations stains
20000
(1640 cm–1) shown in figures 10B and 10C; spectra
NV– not shown). These features suggest
that the time and temperature that
0
created the transition from green to
600 650 700 750 brown as the diamond shifted to a
WAVELENGTH (nm) new position also brought some
“healing” from the localized radiation
effects detected within the green radi-
Figure 11. These 532 nm excitation spectra show differences in features
ation stains. This sample was inter-
between the brown and green radiation stains imaged in figure 10A and a
esting scientifically, as it allowed
nearby colorless portion of the diamond. In the colorless diamond, the
some direct comparison of radiation
Raman peak is much larger and shows the NV– peak and a weak GR1.
stain features created by the same
The brown and green stains show much lower diamond Raman intensity,
point sources.
and the peak width is greater within the green stain (inset). Both the green
and brown radiation stains also show a radiation-related Raman peak at Troy Ardon and Sally Eaton-Magaña
1640 cm–1 (583 nm; Eaton-Magaña and Moe, 2016) and comparatively pro-
nounced GR1 peaks. Other unknown and intrinsic features are also pres-
ent in these spectra. HEMIMORPHITE Resembling
Paraíba Tourmaline
A 5.61 ct semitransparent blue cabo-
Moe, “Temperature effects on radia- shows the area between the green and chon (figure 12) was submitted to the
tion stains in natural diamonds,” Dia- brown radiation stains approaching New York laboratory for a Paraíba
mond and Related Materials, Vol. 64, the features of the surrounding color- tourmaline origin report because of
2016, pp. 130–142). The average less diamond. Both figures 10B and its electric blue color and its similar
Raman peak FWHM (full width at 10C show indications of less radiation RI and internal features. Microscopic
half maximum) within the green por- damage compared to figure 10A. The observation revealed parallel tubes,
tion of the radiation stain was ~1.6 TR12, H3, and NV– features were de-
nm, while the average width within tected in both the green and brown ra-
the brown portion of the radiation diation stains, and the natural Figure 12. This 5.61 ct electric
stain was ~1.1 nm. For comparison, diamond features of H3 and NV– blue cabochon was identified as
the Raman width in the colorless sec- showed higher intensity in the brown. hemimorphite.
tions was ~0.5 nm. The detected The GR1 and TR12 were slightly
Raman intensity was also much higher in the green radiation stains
lower than that of the surrounding than in their brown counterparts and
colorless diamond (figures 10A and greater still than in the surrounding
11). Natural diamond and irradiation- colorless diamond (figure 10C). For
related features such as TR12 (470 the radiation stains pictured in figures
nm), H3 (503.2 nm), and NV– (637 nm) 10B and 10C, the diamond Raman
were not detected within this heavily widths were generally equivalent to
damaged radiation stain. The GR1 those of the surrounding diamond.
peak was detected, though with The PL maps also demonstrated that
higher intensity within the green ra- the boundary of the high GR1 inten-
diation stain (figure 11). sity extended laterally ~30 μm beyond
In figure 10B, the PL map is con- the colored radiation stain and into
sistent with the visual image and the surrounding colorless diamond;

LAB NOTES GEMS & GEMOLOGY SUMMER 2020 285

RAMAN SPECTRA
16000
Unknown
RRUFF Hemimorphite
14000
930

12000
INTENSITY (COUNTS)

683
10000

8000

6000 400

332 452
4000

2000
Figure 14. This 10.47 ct sapphire
is notable as a large, unheated
0
purple sapphire from Montana.
150 250 350 450 550 650 750 850 950 1050 1150
RAMAN SHIFT (cm–1)
14) for an identification and origin re-
port. Standard gemological testing
Figure 13. Raman spectra of a 5.61 ct semitransparent blue cabochon that
gave a 1.762 to 1.770 refractive index,
matched with hemimorphite in the RRUFF database (no. R070536).
indicating corundum, and a hydro-
static specific gravity (SG) of 4.00. The
stone displayed no fluorescence under
long-wave and short-wave UV.
Microscopic examination showed
fluid inclusions, and white granular SAPPHIRE
an interesting combination of inclu-
flake-like inclusions. The stone’s Exceptional Purple Montana sions, including an iridescent healed
optic character could not be deter- Sapphire fissure surrounding a crystal, particu-
mined due to the cabochon shape and
The Carlsbad laboratory recently re- late clouds with intact and unaltered
abundance of inclusions, and only an
ceived a 10.47 ct purple octagonal rutile needles/silk, flaky stringers, and
approximate RI of 1.61–1.63 with a
modified brilliant-cut sapphire (figure twinning with intersection tubules
birefringence of 0.02 could be ob-
tained. Specific gravity was 3.45,
which is inconsistent with tourma-
line, and the fluorescence reaction Figure 15. Flaky particles and stringers of colorful rutile silk seen in the
was medium blue under short-wave 10.47 ct purple Montana sapphire. Field of view 1.58 mm.
and inert under long-wave UV radia-
tion. We confirmed its identity by
Raman spectroscopy, which showed
a match with hemimorphite
(https://rruff.info/Hemimorphite/R0
70536), as shown in figure 13.
Hemimorphite is a zinc silicate,
Zn4Si2O7(OH)2·H2O; the purest form
is white or colorless. Impurities cause
different colors, such as copper (Cu2+)
for a bluish and greenish tint, ferrous
iron (Fe2+) for green, and ferric iron
(Fe3+) for brown. Laser ablation–induc-
tively coupled plasma–mass spec-
trometry (LA-ICP-MS) composition
analysis revealed that this stone con-
tained no iron but did contain copper,
which produced the electric blue
color.
Sudarat Saeseaw

286 LAB NOTES GEMS & GEMOLOGY SUMMER 2020

FTIR SPECTRUM
(figure 15). This inclusion scene is 4.0

consistent with unheated corundum

from Montana (Winter 2018 Lab

ABSORPTION COEFFICIENT (cm–1)

3.5
Notes, pp. 434–435).
Laser ablation–inductively cou-
3047
pled plasma–mass spectrometry (LA- 3.0

ICP-MS) was used to conclusively

determine the stone’s trace element 2.5
chemistry, and the results were com-
pared to corundum samples from
GIA’s colored stone reference collec- 2.0

tion. Trace element measurements

indicated ranges of 13.6–15.0 ppma 1.5
Mg, 13.1–15.0 ppma Ti, 7.76–8.36
ppma V, 38.3–39.4 ppma Cr, 1560–
1600 ppma Fe, and 16.5–17.0 ppma 1.0
3500 3300 3100 2900 2700 2500
Ga. The chemistry matched well
WAVENUMBER (cm–1)
with reference stones GIA has col-
lected from Montana’s secondary de-
posits. With a combination of
microscopic observation, advanced Figure 16. The blue sapphire’s FTIR spectrum shows a strong absorption
testing, and GIA’s reference collec- band centered around 3047 cm–1, suggesting treatment with heat and
tion, we were able to confirm the ge- pressure.
ographic origin of this sapphire.
Of all North American corundum
localities, Montana reigns supreme.
cavities, or negative crystals. Nega- gently heated by the microscope bulb,
While rubies are only rarely found,
tive crystals in sapphire are usually the gas bubble became smaller and
facet-quality sapphire is mined at a
filled with liquid and gaseous carbon disappeared as the gas and liquid ho-
number of different locations. Fine
dioxide (CO2) and may also contain mogenized. The presence of CO2 fluid
blue and purple gems have been
solids such as diaspore or graphite. inclusions in negative crystals is im-
mined at Yogo Gulch in central Mon-
At GIA’s Bangkok laboratory, the portant evidence of natural origin and
tana, but cut gems over one carat are
author recently examined a 6.41 ct the absence of thermal enhancement
rare. Fancy sapphires are found asso-
transparent blue faceted sapphire. It (J.I. Koivula, “Carbon dioxide fluid in-
ciated with the placers of Missouri
had a refractive index of 1.760–1.768, clusions as proof of natural-colored
River and at Dry Cottonwood Creek
and its fluorescence reaction was corundum,” Fall 1986 G&G, pp. 152–
and Rock Creek. These deposits pro-
inert under long- and short-wave UV 155). Interestingly, this sample exhib-
duce larger material, but the colors
radiation. Fourier-transform infrared ited a rounded bubble in a negative
tend to be pale. The successful devel-
opment of heat treatment technology (FTIR) spectroscopy is a useful tech-
there has renewed interest in Mon- nique to determine heat treatment in
tana sapphire. corundum. The sample’s FTIR spec- Figure 17. When the sapphire was
This particular gem is consistent trum (figure 16) showed a broad band viewed under cross-polarized
with those found in Montana’s sec- centered around 3047 cm–1 that has light, a plane of small birefrin-
ondary deposits. It is a particularly been reported in sapphires treated gent crystals was observed in a
fine example due to its saturated pur- with heat and pressure (M.S. Krzem- partially healed fissure. Field of
ple color, large size, and absence of nicki et al., “Sapphires heated with view 4.10 mm.
heat treatment. pressure – A research update,” Spring
2019 InColor, pp. 86–90).
Maryam Mastery Salimi and
Microscopic observation revealed
Nathan Renfro
altered growth tubes, dissolved parti-
cles, melted crystals with altered fin-
gerprints, and small birefringent
Negative Crystal Containing a crystals in a low-relief partially healed
Mobile CO2 Bubble in Blue fissure (figure 17), proof of high-tem-
Sapphire Heated with Pressure perature heat treatment with pressure
Among the common inclusions seen (Winter 2018 Gem News Interna-
in almost every mineral are primary tional, p. 458). When the sapphire was

LAB NOTES GEMS & GEMOLOGY SUMMER 2020 287

Figure 18. A negative crystal with a mobile CO2 bubble in the blue sapphire heated with pressure. Field of view 2 mm.

crystal that moved and was still ob- Pink Mozambique Sapphire Advanced testing confirmed that
served when gently heated with a hot Recently, GIA’s Carlsbad lab received the sapphire was unheated and from
point (figure 18). Raman spectroscopy an 11.16 ct pink sapphire (figure 19) Mozambique. Infrared spectroscopy
confirmed the bubble as carbon diox- for a colored stone identification and produced a spectrum consistent with
ide (CO2). To our knowledge, a mobile origin report. Standard gemological unheated sapphire. The spectra showed
CO2 bubble in a negative crystal has testing was consistent with sapphire, a single peak at 3309 cm–1 with no sec-
not been reported in a sapphire heated giving a refractive index of 1.759– ondary peak at 3232 cm–1 (Summer
with pressure. 1.769 and a hydrostatic specific grav- 2019 Gem News International, pp.
Nattida Ng-Pooresatien ity of 4.00. 290–291). Laser ablation–inductively
The sapphire contained various coupled plasma–mass spectrometry
inclusions (figure 20) such as negative (LA-ICP-MS) confirmed the stone’s
crystals and transparent colorless chemical composition of a higher
Figure 19. An 11.16 ct pink sap-
crystals. These inclusions showed no iron and lower vanadium content,
phire from Mozambique.
indication of thermal alteration, sug- which is typical of Mozambique
gesting the stone was unheated. Bands sapphires.
of silk arranged in a hexagonal forma- Mozambique is known for rubies
tion with irregularly shaped thin that are saturated in color. With its
films intermixed with particle clouds 11.16 ct size and pure pink hue, this
could be seen throughout the stone. unheated Mozambique sapphire was
This suite of inclusions was consis- a unique encounter.
tent with a Mozambique origin. Nicole Ahline

Figure 20. A negative crystal (left; field of view 1.79 mm) and a cluster of
silk (right; field of view 2.57 mm) that could be seen throughout the pink
sapphire.

288 LAB NOTES GEMS & GEMOLOGY SUMMER 2020

ible absorption spectra taken at liquid
nitrogen temperature revealed the
presence of the SiV– center, a defect
commonly observed in CVD syn-
thetic diamonds. Photoluminescence
(PL) spectra taken with various exci-
tation wavelengths and at liquid ni-
trogen temperature showed the
presence of a strong SiV– doublet
(736.6/736.9 nm), strong NV0/– (575
and 637 nm) centers, numerous peaks
in the 520–550 nm region, and a
strong H3 (503.2 nm) center. Further-
more, all the spectra showed an ab-
sence of the SiV0 (946 nm) and
596/597 nm centers. This last center
is commonly observed in as-grown
CVD synthetic diamonds but is re-
Figure 21. Face-up image of one of the HPHT-processed CVD laboratory- moved by post-growth HPHT treat-
grown diamonds submitted to the Antwerp lab. This 1.02 ct round brilliant ment. A weak H2 center (986 nm) was
synthetic diamond has J-equivalent color and VS2 clarity. observed in six stones. Weak Ni-re-
lated absorption peaks were detected
in 830 nm PL spectra of all the sam-
ples, but at higher wavelengths than
SYNTHETIC DIAMOND cific NVH0 band at 3123 cm–1 (P.M. expected (884–885 nm, potentially at-
HPHT-Processed CVD Laboratory- Martineau et al., “Identification of tributed to high strain). Microscopic
Grown Diamonds with Low Color synthetic diamond grown using analysis using crossed polarizing
chemical vapor deposition [CVD],” plates confirmed the presence of
Grades
Spring 2004 G&G, pp. 2–25). UV-vis- strain throughout the stones.
When CVD-produced diamonds are
grown with nitrogen added to the gas
mixture, or when the growth rate is
relatively high, the as-grown crystal Figure 22. DiamondView image of one of the 11 CVD samples shows
ends up having an undesirable brown green fluorescence and clear linear growth striations. The layers indicate
color. In order to improve this color, stop-start growth.
CVD-grown diamonds are often
treated at high-pressure, high-temper-
ature (HPHT) conditions after their
growth. The Antwerp laboratory re-
cently received 11 CVD laboratory-
grown diamonds that had been
subjected to HPHT treatment after
growth but still had a relatively uni-
form low color grade (J-equivalent,
figure 21).
The samples were round brilliants
with weights ranging from 1.02 ct to
1.33 ct and clarity grades in the VS
range. Commonly observed clarity
characteristics were crystals, needles,
and pinpoints along with feathers.
They were classified as type IIa dia-
monds with weak additional absorp-
tion peaks at 1332, 1344, 2947, and
3031 cm–1. Four showed a weak ab-
sorption peak at 3107 cm–1 (N3VH).
None of them showed the CVD-spe-

LAB NOTES GEMS & GEMOLOGY SUMMER 2020 289

Figure 24. Left: Strong color zoning observed under natural lighting condi-
tions. Right: The fluorescence pattern follows this zoning, as observed in
the DiamondView image (strong short-wave UV).

yellowish green to orangy yellow indicative of post-growth treatment to

when heated or concealed in the dark create the desirable pink color.
Figure 23. This 3.02 ct HPHT- for long periods of time. This diamond was grown with a
grown diamond with color equiv- Recently submitted to GIA’s New very carefully controlled nitrogen
alent to Fancy Vivid pink had an York laboratory for examination was a content that produced approximately
unstable color. 3.02 ct laboratory-grown diamond 1.6 ppm of type Ib single substituted
with the equivalent color grade of nitrogen (post treatment). The nitro-
Fancy Vivid pink (figure 23). Typical of gen content was calculated from a
an HPHT-grown laboratory-grown di- normalized mid-infrared spectrum.
DiamondView images showed
amond, it exhibited strong color zon- Post-growth treatment of irradia-
green luminescence (linked to the
ing (figure 24, left) related to its typical tion/annealing generated very strong
strong H3 center revealed by PL spec-
hourglass synthetic growth structure. absorption from the nitrogen vacancy
tra) and a weak bluish green phospho-
Of note, it exhibited very strong or- (NV) centers that are responsible for
rescence. All samples showed clear
ange fluorescence when exposed to both the pink bodycolor and the very
linear growth striations. Some re-
ultra-short-wave ultraviolet radiation strong orange-red fluorescence (see
vealed a multilayer stop-start growth.
(figure 24, right). This is of note as it is the PL spectrum in figure 25).
The orientation of the growth layers
gives an indication of the angle at
which the stone is cut from the rough
plate (figure 22). Figure 25. 514 nm (green) photoluminescence spectrum showing strong
NV (nitrogen vacancy) centers at 575 and 637 nm. These are responsible
Based on all the observations, we
for the very strong bodycolor and orange-red fluorescence.
conclude that the analyzed CVD lab-
oratory-grown diamonds underwent
HPHT treatment after growth. Their PL SPECTRUM
relatively low color grades are believed
to be linked to the duration of the 575
treatment. More efficient treatment
would likely result in a higher color
grade.
INTENSITY

Ellen Barrie

HPHT Laboratory-Grown Pink 637

Diamond with Unstable Color
Raman

While rare, it is not uncommon for

certain types of diamond to change
color when exposed to different light- 560 580 600 620 640
ing or thermal conditions. Consider,
WAVELENGTH (nm)
for example, chameleon-type dia-
monds, which can change color from

290 LAB NOTES GEMS & GEMOLOGY SUMMER 2020

Figure 26. A dramatic hue shift from pink to orange was observed when the synthetic diamond was exposed to
short-wave UV radiation. Color swatches (bottom) were generated from recorded LCH values both before and
after UV bleaching. The two images (top) were taken using standard lighting in the GIA colorimeter, which
recorded the LCH values.

When excited by short-wave UV ure 26, bottom) were generated that

radiation, the nitrogen vacancy color- accurately represent the color change PHOTO CREDITS
causing centers are bleached and de- observed. The images in figure 26 Diego Sanchez—1, 7 (left), 14; Shunsuke
saturated, resulting in a hue shift to (top) were captured using a GIA-built Nagai—3, 6; Makoto Miura—4; Sally Eaton-
orange (figure 26). This is a tempo- colorimeter with standard lighting Magaña—7 (right), 8, 10; Troy Ardon—9; Nut-
rary color change, and within min- conditions. tapol Kitdee—12; Aaron Palke—15; Nattida
utes the diamond will revert to its With the recent influx of labora- Ng-Pooresatien—17, 18; Robison McMurtry—
tory-grown fancy-color diamond into 19; Nathan Renfro—20; Ellen Barrie—21;
stable pink color under normal day-
the market, these phenomena are of Michaela Harinová—22; Jian Xin (Jae) Liao—
light conditions. 23; Paul Johnson—24 (right)
Using LCH (lightness, chroma, note to consumers.
and hue) values, color swatches (fig- Kaitlyn Mack and Paul Johnson

LAB NOTES GEMS & GEMOLOGY SUMMER 2020 291

Editor
Nathan Renfro

Contributing Editors
Elise A. Skalwold and John I. Koivula

Böhmite in Corundum
Gemologists at Bangkok’s Lotus Gemology laboratory re-
cently examined a large lot of rubies believed to originate
from the Montepuez area of Mozambique. Microscopic ex-
amination confirmed that assessment. Many of these
stones displayed features in their infrared spectra that are
typical for böhmite, with a characteristic “twin peak”
spectrum at ~3309 and 3089 cm–1 (figure 1).
One of the stones showed a large area cut through on a
pavilion facet that appeared to be a foreign substance. A sim-
ilar substance was seen in many of the fissures in the stone
(figure 2). Judicious use of the hot point under the micro-

Figure 1. “Twin peak” infrared spectrum of böhmite Figure 2. A large area of a foreign substance cut
in the Mozambique ruby, with prominent peaks at through on the surface of a ruby from the Montepuez
~3309 and 3089 cm–1. area of Mozambique. This substance infilled not only
this cavity but also many narrow fissures in the stone.
IR SPECTRUM Micro-Raman analysis identified it as böhmite. Dark-
0.53
field + diffuse overhead illumination. Photomicro-
graph by Richard W. Hughes; field of view 3 mm.
3089
ABSORPTION COEFFICIENT (cm–1)

3309
0.52
scope produced no reaction. We then moved to analyze the
substance via micro-Raman. The result was böhmite, which
neatly confirmed the results from the infrared spectrum.

0.51

About the banner: This plate of muscovite mica from Brazil contains an in-
teresting frond of tourmaline that shows vibrant color using polarized light.
0.50
Photomicrograph by Nathan Renfro; field of view 15.67 mm.
Editors’ note: Interested contributors should contact Nathan Renfro at
[email protected] and Stuart Overlin at [email protected] for submission
information.
0.49
3500 3400 3300 3200 3100 3000 2900 2800 2700 GEMS & GEMOLOGY, VOL. 56, NO. 2, pp. 292–297.
WAVENUMBER (cm ) –1 © 2020 Gemological Institute of America

292 MICRO-WORLD GEMS & GEMOLOGY SUMMER 2020

The drilled bead weighed 12.02 ct and measured 13.4 mm
in diameter. It contained small black inclusions of pyrite
but also a crystalline inclusion in the form of aggregated
pseudocubes (figure 3). This inclusion was identified as
chabazite via Raman analysis. The pseudocubes were in
fact near-cubic rhombohedra.
While this mineral has already been documented in
veins and opal nodules from Ethiopia, it is rarely seen in cut
stones (B. Chauviré, “Genèse de silice supergène sur Terre
et implications sur Mars,” PhD thesis, University of
Nantes, 2015; N.D. Renfro and S.F. McClure, “Dyed purple
hydrophane opal,” Winter 2011 G&G, pp. 260–270).
Chabazite, CaAl2Si4O12· 6H2O, is a member of the zeolite
family. This mineral is formed in gaseous cavities of extru-
Figure 3. Chabazite inclusion in an Ethiopian opal. sive volcanic rocks, which is similar to how opals from
Photomicrograph by Aurélien Delaunay; field of view Ethiopia form. This inclusion is also proof of the Ethiopian
1.42 mm. geographic origin of this gem, as it has not been described
in opal from other localities.
Böhmite, AlO(OH), is a polymorph of diaspore. In Aurélien Delaunay and Ugo Hennebois
corundum it is generally a secondary mineral, infilling sur- Laboratoire Français de Gemmologie
face-reaching openings after the host corundum has grown. Emmanuel Fritsch
Richard W. Hughes Université de Nantes, France
Lotus Gemology, Bangkok
Fluorophlogopite in Burmese Spinel
Rare Inclusion of Chabazite in a Precious Opal Lotus Gemology in Bangkok received a large parcel of dark
From Ethiopia red spinels for identification. Their colors, inclusion fea-
An opal received for analysis at the Laboratoire Français de tures, and trace element chemistry suggested that they
Gemmologie (LFG) in Paris was classified as a light opal, originated from Myanmar’s Mogok Stone Tract.
with a bodycolor that matched to neutral gray N8 on the In one faceted specimen, a crystal was found breaking
Munsell scale. The play-of-color was divided into sections the surface (figure 4). The crystal’s surface luster suggested
with well-defined outlines that fit together or were rounded a refractive index below that of the host spinel, and under-
off in the form of digits (B. Rondeau et al., “On the origin of cutting suggested it was also significantly lower in hard-
digit patterns in gem opal,” Fall 2013 G&G, pp. 138–146). ness. The inclusion’s interior also displayed signs of

Figure 4. This crystal broke the surface of a dark red Figure 5. In darkfield illumination, cleavage was visi-
spinel believed to be from Mogok. It had a lower luster ble inside the crystal. Micro-Raman analysis showed
than the spinel, and undercutting indicated a lower it to be fluorophlogopite. Photomicrograph by
hardness. Photomicrograph by Richard W. Hughes, Richard W. Hughes, darkfield illumination; field of
diffuse overhead illumination; field of view 2.5 mm. view 2.5 mm.

MICRO-WORLD GEMS & GEMOLOGY SUMMER 2020 293

cleavage (figure 5). Thinking it was possibly a carbonate,
we proceeded to do micro-Raman analysis on the crystal.
To our surprise, the result was fluorophlogopite.
Fluorophlogopite, KMg3(AlSi3O10)(F,OH)2, is a species in
the mica group and a member of the trioctahedral mica sub-
group. To the best of our knowledge, this is the first reported
instance of this species in spinel. Nor are we aware of it hav-
ing been found included in any gem mineral to date.
Richard W. Hughes

Unusually Shaped Rutile Crystal Inclusions in

Russian Emerald
Russian emeralds have been mined since the mid-nine-
teenth century near Yekaterinburg in the Ural Moun- Figure 6. This long yellowish brown rod was identified
tains. They are classified as schist-related emeralds, by Raman spectroscopy as rutile. Photomicrograph by
meaning they formed in the reaction zone of a pegmatite Charuwan Khowpong; field of view 1.75 mm.
intruding into schists. Similar emeralds are found in
Zambia, Brazil, and Ethiopia. All of these deposits can
show a similar inclusion scene (S. Saeseaw et al., “Geo-
graphic origin determination of emerald,” Winter 2019 dense clouds of needles indicated the stone had not been
G&G, pp. 614–646). heat treated (J.I. Koivula, “Internal diffusion,” Journal of
A 1.489 ct Russian emerald crystal collected by GIA at Gemmology, Vol. 20, No. 7/8, 1987, pp. 474–477). Notably,
the Mariinsky Priisk mine showed several interesting in- this stone had one large negative crystal containing an
cluded crystals. The crystals appeared as long yellowish opaque hexagonal crystal (figure 8). The hexagonal daugh-
brown rods (figure 6). In some cases, they appeared as a ter crystal was free to move within the negative crystal
group of disordered coarse needles (figure 7). They were when the stone was rocked and tilted. Careful infrared
identified by Raman spectroscopy as rutile. Although rutile spectroscopic measurements suggested that this negative
crystal inclusions are common in schist-related emeralds, crystal was mainly filled with water and CO2 based on
this crystal morphology is unusual and has not previously water-related peaks at 3706 and 3605 cm–1 and a large CO2
been mentioned in the literature. Therefore, this inclusion broad band at 2340 cm–1. However, the opaque daughter
might be unique and different from the other schist-related crystal was not identified by Raman spectroscopy because
the negative crystal containing this daughter phase was lo-
emerald sources. These unusual patterns could potentially
cated in a deeper section of the stone. Its hexagonal crystal
be an indicator of Russian origin.
shape and appearance suggested that the mobile hexagonal
Charuwan Khowpong
GIA, Bangkok

Figure 7. The group of disordered coarse needles of ru-

A Sapphire with a Negative Crystal Containing a tile. Photomicrograph by Charuwan Khowpong; field
Mobile Graphite Daughter Crystal of view 3.65 mm.
Mineral and fluid phases within negative crystals in meta-
morphic sapphires provide us with general information on
the sapphire and the trapped fluid itself. Such inclusions
indicate a stone has not been heated when intact negative
crystals are observed containing carbon dioxide (e.g., Fall
1986 Lab Notes, pp. 152–155). The author recently had an
opportunity to identify a natural star sapphire containing
a unique daughter crystal phase within a negative crystal.
This stone, which measured 22.20 × 18.00 × 14.05 mm
with an estimated weight of 62 ct, was set in a ring. It dis-
played a six-rayed asterism due to light reflection from nu-
merous long intersecting needles. Gemological
observation and properties confirmed that this stone was
a natural sapphire. The presence of unaltered mineral crys-
tals and the chromophore cannibalization pattern around

294 MICRO-WORLD GEMS & GEMOLOGY SUMMER 2020

Figure 8. This opaque
hexagonal daughter
crystal, of what is likely
graphite, moves within
the large negative crys-
tal in the metamorphic
star sapphire. Photo-
micrographs by
Makoto Miura; field of
view 4.0 mm.

daughter crystal was graphite, which has been previously Trapiche-Like Pattern in an Emerald from Pakistan
documented (E.J. Gübelin and J.I. Koivula, Photoatlas of GIA’s colored stone research collection contains many
Inclusions in Gemstones, Vol. 3, Opinio Verlag, Basel, unique samples. The emerald sample shown in figure 10
Switzerland, 2008, p. 268). This is the first example of a was mined at Swat Valley in the Khyber Pakhtunkhwa re-
negative crystal with a mobile solid phase in corundum gion of Pakistan and fabricated by polishing two parallel
that the author has encountered. windows perpendicular to the c-axis of the hexagonal crystal
Makoto Miura in order to show its inclusions. The sample weighed 0.47 ct
GIA, Tokyo prior to fabrication and 0.36 ct after processing.

Windmills in Rare Mineral Sphalerite Figure 9. A group of windmill-like inclusions with or-
ange color concentrations observed in sphalerite.
Sphalerite is a rare gem mineral that crystallizes in the
Photomicrograph by Ungkhana Atikarnsakul; field of
cubic crystal system and consists largely of zinc sulfide
view 3.1 mm.
formed with variable iron, with the chemical formula of
(Zn,Fe)S. Its color is usually yellow, brown, red, or gray to
gray-black with high dispersion and high adamantine lus-
ter. The author recently examined a high-quality faceted
orange sphalerite weighing 47.56 ct that revealed three-
ray structures reminiscent of windmills with each pro-
peller arm separated by 120°. Additionally, the orange
color was concentrated in the immediate region of the
three radial arms of each propeller-like structure (figure 9).
Other typical internal characteristics of sphalerite such as
strong angular internal growth with orange color zoning,
particulate clouds, tiny crystals, and fingerprints, were
also observed. This distinctive structure resembling wind-
mills is the first such feature the author has encountered
in sphalerite.
Ungkhana Atikarnsakul
GIA, Bangkok

MICRO-WORLD GEMS & GEMOLOGY SUMMER 2020 295

trapiche structure previously documented in some emer-
alds from Pakistan (Fall 2019 Gem News International,
“Trapiche emerald from Swat Valley, Pakistan,” pp. 441–
442). In this case, the distinct growth feature consisted of
a colorless hexagonal core from which six-rayed fibrous
arms extended. Lighting was an important consideration
in maximizing the color in this stone, and the features
were fully observed using brightfield illumination. This
was a wonderful example of a pattern rarely encountered
in emeralds.
Piradee Siritheerakul and Charuwan (Gibb) Khowpong
GIA, Bangkok

Quarterly Crystal: Pyrite in Quartz

Figure 10. A distinct near-colorless hexagonal core
Pyrite and quartz are two relatively common minerals.
with radiating fibrous arms is evident in this Paki-
Transparent colorless quartz, known as rock crystal, is host
stani emerald from Swat Valley. Photomicrograph by
to a wide variety of mineral and fluid inclusions, more so
Charuwan Khowpong; field of view 1.40 mm.
than any other gem mineral. While pyrite and rock crystal
are both common, finding a good example of crystalline
pyrite inside of rock crystal quartz is a gem mineral collec-
At first glance, nothing unusual was noted in this sam- tor’s prize.
ple, but when correctly illuminated under the microscope Known for its fine emeralds, the Chivor mine in
it revealed a pattern that could easily pass for trapiche. Colombia is also a mineralogical source of both pyrite and
The stone showed radiating zones of turbid growth with colorless quartz. On occasion, crystals of Colombian
six-fold symmetry. This makes it a trapiche-like stone quartz are known to host pyrite. As shown in figure 11 (op-
rather than a true trapiche, which has distinct arms that posite page), we recently had the opportunity to examine
separate the growth zones. While classic trapiche patterns one such specimen, which hosted numerous bright metal-
are mostly associated with Colombian emeralds, the Swat lic brassy yellow modified octahedrons of pyrite (figure 12).
region also produces emeralds with radiating black inter- The inclusions were not randomly scattered throughout
growths. Still, this structure was very different from the the 466.27 ct quartz, but instead were situated neatly just

Figure 12. These modi-

fied octahedrons of
pyrite were deposited
along one prism face of
their quartz host.
Photomicrograph by
Nathan Renfro; field of
view 7.62 mm.

296 MICRO-WORLD GEMS & GEMOLOGY SUMMER 2020

Figure 11. This example of Colombian pyrite-containing rock crystal weighs 466.27 ct and measures 61.90 mm in
length. Photo by Diego Sanchez.

under the surface of one of the prism faces. Another inter- cally (at the same time) from a directional fluid rich in
esting feature was that all of the pyrite inclusions were pyrite forming iron sulfide.
present in a range of sizes with well-defined crystal faces, John I. Koivula and Nathan Renfro
suggesting that the pyrite crystals precipitated syngeneti- GIA, Carlsbad

MICRO-WORLD GEMS & GEMOLOGY SUMMER 2020 297

Gem News International
Contributing Editors
Emmanuel Fritsch, University of Nantes, CNRS, Team 6502, Institut des Matériaux Jean Rouxel (IMN),
Nantes, France ([email protected])
Gagan Choudhary, Gem Testing Laboratory, Jaipur, India
([email protected])
Christopher M. Breeding, GIA, Carlsbad ([email protected])

COLORED STONES AND ORGANIC MATERIALS Milas District, Muğla Province, Turkey (M. Hatipoğlu and
S. Chamberlain, “A gem diaspore occurrence near Pinarcik,
Purplish pink diaspore reportedly from Afghanistan. Dia- Mugla, Turkey,” Rocks & Minerals, Vol. 86, No. 3, pp. 242–
spore, an orthorhombic aluminum oxide hydroxide mineral 249). Gem diaspore from Turkey has been reported since the
with the formula AlO(OH), is a relative newcomer to the late 1970s (K. Scarratt, “Faceted diaspore,” Journal of Gem-
jewelry scene, with some stones possessing a captivating mology, Vol. 17, No. 3, 1980, pp. 145–148), but it was not
ability to change color between pale green, yellow, pink, until commercial mining began in 2005 that a consistent
and purple when viewed in different sources of light (C. supply became available and was subsequently marketed
Shen and R. Lu, “The color origin of gem diaspore: Correla- under various trade names including Zultanite and Csarite
tion to corundum,” Winter 2018 G&G, pp. 394–403). Here (S. Kotlowski and L. Rosen, “A short history of diaspore and
we document a new find of purplish pink gem diaspore re- its trade names Zultanite and Csarite,” International Gem
portedly from Nangarhar Province, Afghanistan (figure 1). Society, www.gemsociety.org/article/short-history-diaspore-
Previously, transparent crystals of diaspore in sizes suit- trade-names-zultanite-csarite). Additionally, alluvial de-
able for jewelry purposes were only recovered from mining posits in Myanmar near Mong Hsu have yielded limited
operations on Ilbir Mountain near the village of Pinarcik, amounts of chromium-colored gem-quality crystals in small
sizes since 2004 (U Hla Kyi and K.K. Win, “A new deposit
of gem quality colour-change diaspore from Möng Hsu,
Figure 1. This purplish pink diaspore is reportedly Myanmar,” Australian Gemmologist, Vol. 22, No. 4, pp.
from a new deposit in Afghanistan, discovered in Feb- 169–170).
ruary 2020. The faceted stones weigh 0.57 to 1.60 ct; Dealers Salman Khan (ARSAA Gems & Minerals) and
courtesy of Donald Hofler. The rough stones range Noshad (Noshad Gems Enterprises), both based in Pesha-
from 4.15 to 17.52 ct; courtesy of Salman Khan. Photo war, Pakistan, reported that in March 2020, a purplish pink
by Diego Sanchez. diaspore from a new deposit had reached the gem market.
The new material was reportedly coming from the Goshta
district of Nangarhar Province near the village of Ragha.
The authors obtained several samples from Salman Khan
and Donald Hofler, a gem dealer in Texas, to perform gem-
ological and advanced testing to characterize this material.
These samples would be classified as F-type samples (col-
lected in the international market) according to the classi-

Editors’ note: Interested contributors should send information and illustra-

tions to Stuart Overlin at [email protected] or GIA, The Robert Mouawad
Campus, 5345 Armada Drive, Carlsbad, CA 92008.

GEMS & GEMOLOGY, VOL. 56, NO. 2, pp. 298–316.

© 2020 Gemological Institute of America

298 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SUMMER 2020

A B candescent lighting, but no distinct color change of the
dominant hue was observed in the faceted samples.
Advanced spectroscopic instrumentation was used to
further characterize the new diaspore material. One sample
was optically oriented into a polished cube with parallel
faces polished perpendicular to the crystallographic axes
(T. Thomas et al., “Device and method of optically orient-
ing biaxial crystals for sample preparation,” Review of Sci-
C D
entific Instruments, Vol. 85, No. 9, 2014, 093105). The
ultraviolet/visible/near-infrared (UV-Vis-NIR) absorption
spectra of the polished cube revealed broad absorption
bands centered in the range of 560 and 580 nm for alpha,
beta, and gamma polarizations (figure 3). Under daylight-
equivalent lighting, the saturated purple-pink color was
consistent with the alpha vibration direction, pale greenish
Figure 2. Microscopic examination of the diaspore re- yellow for the beta direction, and light purplish pink for
vealed numerous elongate fluid inclusions (A) and the gamma direction. Under incandescent light, the ple-
planes of fluid inclusions (B). Also observed in two ochroic colors were saturated purple-pink for the alpha
rough samples were fine black rutile needles (C), and color, orange-pink for the beta color, and light pink for the
one stone contained an unidentified brown hexagonal gamma color. Color swatches were calculated using the
crystal (D). Photomicrographs by Nathan Renfro; measured spectra to produce a representative color for each
field of view 1.88 mm (A), 1.44 mm (B), 4.47 mm (C), pleochroic color and also to show how the colors mix rel-
and 1.41 mm (D). ative to their crystallographic axes (figure 4).
Observations under the microscope showed the cleav-
age direction in the faceted stones to be nearly parallel to
fication scheme used by GIA’s field gemology department the table (it is inclined slightly, presumably to avoid cleav-
(W. Vertriest et al., “Field gemology: Building a research ing when polishing the table facet). This orientation places
collection and understanding the development of gem de- the b-axis approximately perpendicular to the table facet,
posits,” Winter 2019 G&G, pp. 490–511). It was also re- which will facilitate the best color face-up, composed pri-
ported by the Peshawar-based dealers that preliminary marily of a mixture of the alpha and gamma rays.
experiments with gamma irradiation resulted in samples
becoming an undesirable dark brownish orange color, but
this was not tested further by the authors.
Microscopic examination showed that the dominant Figure 3. UV-Vis-NIR spectra were collected from an
type of micro feature in the material is fluid inclusions, optically oriented sample to resolve the alpha, beta,
typically consisting of a gas bubble, liquid, and occa- and gamma vibrational directions in the Afghan dia-
sionally a solid phase. Fine curved black rutile needles spore. All three spectra showed a broad absorption
were also observed in two of the samples (confirmed by band centered between 560 and 580 nm resulting pri-
Raman analysis), while one diaspore crystal contained a marily from vanadium and chromium. In this sam-
brown hexagonal crystal that the authors were unable to ple, the average concentrations were 57 ppma
identify (figure 2). Brown epigenetic staining and cleavage vanadium, 20 ppma chromium, 205 ppma iron, and
cracks were observed in several samples. Strong doubling 84 ppma titanium.
from birefringence was also characteristic of the material.
Gemological testing revealed a biaxial positive mineral UV-VIS -NIR SPECTRA
ABSORPTION COEFFICIENT (cm–1)

with refractive indices of nα = 1.700, nβ = 1.721, and nγ =

1.749, and a specific gravity of 3.38. The maximum bire- 2.5
α = E || a
β = E || b
fringence observed was 0.049. Examination with a dichro-
γ = E || c
scope showed three distinct pleochroic colors, consistent 2.0

with a biaxial mineral: dark purplish pink, pale yellow, and 1.5
pale pink. The diaspore showed a very weak red fluores-
cence under long-wave ultraviolet light and an irregular 1.0

chalky yellow fluorescence reaction under short-wave UV.

0.5
When the samples were examined with daylight-equiv-
alent illumination and compared to incandescent illumi- 0
nation, a slight color shift was observed. The samples 350 450 550 650 750 850 950 1050

appeared less purplish, going from purplish pink in day- WAVELENGTH (nm)
light-equivalent lighting to a nearly pure pink hue in in-

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SUMMER 2020 299

Figure 4. One diaspore
D65 Illumination sample was fabricated
into a crystallographi-
b
cally oriented block in
α γ
order to measure the po-
larized spectra, which

Op t
Unpolarized were used to analyze the
light ray R2

ic a
color. This figure shows

xis
Unpolarized α = E ll a the relationship of the
light ray R1
2V β = E ll b polarized spectra meas-
Optic axis γ = E ll c ured parallel to each
Optic axis
2V = 85°
crystal axis and the re-
β sulting colors produced
γ 2V when the polarized spec-
α c
a
tra are mixed together
β under daylight (top) and

Op t
incandescent (bottom)
ic a conditions, which is
xis
what the viewer would
observe using unpolar-
Unpolarized light ray R3
ized light transmitted
through the sample nor-
malized to a 1 cm path
length. Under daylight-
Incandescent Light equivalent lighting, the
b
alpha color was sat-
α γ
urated purple-pink, the
beta color was light
greenish yellow, and the
Op t

Unpolarized
light ray R2 gamma color was light
ic a

pink. Under incandes-

xis

Unpolarized α = E ll a
light ray R1 cent light, the alpha
2V β = E ll b color was a saturated
Optic axis γ = E ll c purple-pink, the beta
Optic axis
2V = 85° color was orange-pink,
β and the gamma color
γ 2V was light pink. The
α c
a acute 2V angle, or angle
β between the two optic
Op t

axes, was measured as

ic a

85° in this sample. Both

xis

axes lie in the optic

Unpolarized light ray R3
plane that is perpendicu-
lar to the b crystal axis.

Raman analysis confirmed the material’s identity as di- This exciting discovery of purplish pink diaspore
aspore (figure 5, left). Fourier-transform infrared (FTIR) spec- from Afghanistan may prove to be a significant new de-
troscopy showed absorption features at approximately 2115 posit. While production volume is currently unknown,
and 1990 cm–1 and broad absorption bands at approximately several kilograms of material have been reported and
4080 and 3000 cm–1, all of which are consistent with dia- large faceted stones of nearly 50 ct have been described
spore (figure 5, right). Chemical analysis of seven samples on social media (Mark Smith, @thailankagems). The
showed a bulk composition consistent with that of diaspore, color of this diaspore can be attributed mainly to trace
and average notable trace element concentrations showed element concentrations of chromium and vanadium. The
27 ppma chromium (ranging from 14 to 48 ppma), 31 ppma concentrations of these elements are generally different
vanadium (from 14 to 59 ppma), 213 ppma titanium (from from those in Turkish diaspore, which has significantly
67 to 428 ppma), and 252 ppma iron (from 137 to 350 ppma). lower vanadium, and those in Burmese material, which

300 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SUMMER 2020

R AMAN SPECTRUM FTIR SPECTRUM
100000
449
INTENSITY (COUNTS)

95000

ABSORBANCE
90000

85000 500
669
334
80000

75000
200 400 600 800 1000 1200 1400 5500 5000 4500 4000 3500 3000 2500 2000 1500

RAMAN SHIFT (cm–1) WAVENUMBER (cm–1)

Figure 5. The purplish pink gem material from Afghanistan showed a Raman spectrum (left) that matched dia-
spore, with a dominant peak at 449 cm–1 and smaller peaks at 334, 500, and 669 cm–1. Its FTIR spectrum (right)
was consistent with diaspore, with absorption features at approximately 2115 and 1990 cm–1 and broad absorption
bands at approximately 4080 and 3000 cm–1.

generally has much higher chromium. This new attrac- ivory. But legislative measures banning all ivory in re-
tive color of gem-quality diaspore will certainly be a wel- sponse to the poaching of African elephants have dimin-
come addition to the gem trade. ished ivory demand.
Ian Nicastro Downey got his start in silver and turquoise and estab-
San Diego, California lished Artifactual in the late seventies. Since then, he has
lived and worked in an ivory carving village in Bali with
Wim Vertriest
artisans he calls “the most talented carvers in the world.”
GIA, Bangkok
When he first arrived in Bali, they were carving bone and
Nathan Renfro, Ziyin Sun, Aaron Palke, and Paul Mattlin coconut shell. His task was to carve animals and animal
GIA, Carlsbad skulls out of fossil ivory. “We created a trade,” he said. He
showed us two of Artifactual’s current offerings (figure 6).
Fossil ivory update with Lee Downey. Lee Downey, owner “But now the material has become flat-out illegal in eight
of Artifactual (Tucson, Arizona), spoke with us in early states.”
2020 about his 35 years in the fossil ivory carving trade and Since 2014, California, Hawaii, Illinois, Nevada, New
the impact of ivory bans on his business. Until recently Hampshire, New Jersey, New York, and the District of Co-
Downey carved fossil ivory from mammoth and walrus lumbia have enacted prohibitions against trade in ivory.
tusks thousands of years old; he has never used elephant (Two other states, Oregon and Washington, have banned

Figure 6. Fossil walrus

ivory carvings by Arti-
factual. The base of the
octopus carving is chal-
cosiderite. Photos by
Kevin Schumacher.

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SUMMER 2020 301

Figure 8. A selection from Artifactual’s inventory of
fossil ivory. Clockwise from top left: belt buckle with
mammoth ivory, knife with mammoth tooth handle,
belt buckle with mammoth ivory, and frog carving of
fossil walrus ivory. Photo by Kevin Schumacher.

never have. It just keeps weathering out of the dirt.”

Downey added that this ivory is at least 10,000 years old.
Mammoth ivory can be differentiated from elephant
ivory by a simple field test to observe its natural grain
lines. But there is no test to determine whether a spe-
cimen of walrus ivory comes from a fossil or a modern
animal. According to Downey, this is a problem in the
trade.
Artifactual uses fossil walrus ivory from St. Lawrence
Figure 7. A mammoth tusk carved with sea life by
Island, Alaska. This ivory is 500 to 15,000 years old and
Lee Downey. Photo by Kevin Schumacher.
dug by the native Yupik people. “It’s part of their income,
and they own the island,” he said. “For us it’s a legitimate
and ethical trade.”
only ivory from living elephants.) Downey said that Ha- Downey added that the legislation on ivory also pro-
waii’s ban in particular “tanked” the trade. Customers hibits mammoth tooth because it comes from an ivory-
even in states where ivory remains legal are reluctant to bearing animal. He showed us a knife with a carved
purchase fossil ivory, he said. mammoth tooth handle (figure 8). Knife makers are some
“I understand why they’re doing it,” Downey told us. of the biggest buyers of mammoth tooth, he said, and the
“I’d rather save all wildlife if there was a way.” But he said bans have dramatically impacted their business.
the effect of the bans on his business has forced him to As a result of the decreased demand for ivory, Artif-
prepare to exit the fossil ivory trade. He is selling his cur- actual is moving its business to other materials, including
rent ivory inventory at lower prices and not buying ad- jet, natural American turquoise and variscite, steel (figure
ditional material. 9, left), and meteorite. “Gaia” (figure 9, right) is carved
Downey showed us a mammoth tusk carved with sea from a piece of the Gibeon iron-nickel meteorite, which
life (figure 7) as well as two belt buckles featuring mam- landed in prehistoric times in Namibia and is believed to
moth ivory (figure 8). All of them use material sourced be around four billion years old. The company also does a
from Alaska. “This washes up on beaches and comes out lot of business in silver.
of gold mines,” he said. “It’s highly sought after by the Downey has also sought out materials similar to ivory
knife makers. It’s got colors that elephant ivories will but without the ethical and legal challenges. He has begun

302 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SUMMER 2020

Figure 9. Left: “Alpha,” a stainless steel foundry-cast skull by Artifactual, one of six in a series. Right: “Gaia,” the
second skull Downey carved from a piece of the Gibeon meteorite, has a tridymite crystal inclusion in the fore-
head. The geometric pattern on its surface was caused by the slow process of cooling as the meteor moved through
space. Photos by Kevin Schumacher.

carving moose antlers naturally shed by the animals (figure for the ivory,” he said. He also recently began carving vin-
10). “The moose antler has been a really good substitute tage billiard balls. “We’ll carve anything that’s carvable,”
he said.
For our video interview with Lee Downey, visit
Figure 10. A moose antler carving of a horned toad by https://www.gia.edu/gems-gemology/summer-2020-gem-
Lee Downey. Photo by Duncan Pay. news-fossil-ivory-update.
Erin Hogarth and Duncan Pay
GIA, Carlsbad

The “Fragility of the Eternal” kunzite: Origin, cutting,

and identification. Lapidary, the art of colored stone fash-
ioning, is an art that requires a keen eye, talent, and great
patience. Different gem materials have their particular
quirks that must be considered when planning the var-
ious steps required to reach the ultimate goal of obtaining
the most aesthetic stone while usually also optimizing
weight yield.
Spodumene is known to challenge even the most ex-
perienced cutter owing to its high fragility, perfect cleavage
planes, and unpredictability in different cutting directions.
This is further magnified when the rough is exceptionally
large and the cutter’s goal is to produce the largest faceted
kunzite in the world.
This was author VT’s aim as he prepared to work on a
piece of rough spodumene that weighed 2,950 g. The un-
treated crystal (figure 11) originated from Kunar Province
in Afghanistan. Although a wide crack was visible at one
end and about 30% of the opposite end contained numer-
ous inclusions, the removal of both sections still left more
than 1.5 kg of relatively clean rough after 20 hours of wire
sawing (the safest method under the circumstances). Since
the rough was oversized, some of the routine cutting ac-

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SUMMER 2020 303

Figure 11. The pronounced pleochroism of the rough spodumene crystal weighing 2,950 grams, seen here in two
orientations, was one of the properties that had to be considered during fashioning. Photos by Arjuna Irsutti.

cessories such as dops and transfer holders, as well as the with custom-made dops provided by his friend and former
cutting tactics, had to be adapted for the job. student Danial Hu. The pre-faceting step took considerable
Prior to the start of the process came the design ele- time to complete because the goal was to create an impor-
ment. No journey can be completed without knowing the tant piece of art and each facet needed to be cut in a pattern
destination, and so rather than playing it safe and opting that matched the stained glass window in the Notre Dame
for a simple design, the artist, VT, chose a truly challenging cathedral with a precision of tenths of a millimeter. Per-
facet pattern of his own design that incorporated 914 fac- haps the most complex part of the process was creating 16
ets. The “Fragility of the Eternal” design is based on the mains and 16 additional lines extending from the center
stained glass window of the iconic Notre Dame cathedral (culet) to the girdle, each containing eight (main) and five
in Paris. The stone is the fifth in a series of six cut designs (additional) parallel-sided facets, continuous with one
that fall under VT’s “World Heritage” project. another. An equally challenging task was ensuring the pro-
VT would also have to revise his usual strategy due to portions were correct while cutting 18 rows of facets of dif-
the size. So instead of cutting the pavilion first, followed by ferent sizes and shapes, where one often does not depend
the girdle, crown, and finally the table facet, it was necessary on another, and finally making the last row of facets cor-
to start with the table facet for this masterpiece. This pre- respond exactly with the girdle, which was already pre-
sented its own challenges since the faceting required spe- shaped. It was impossible to change the width of the final
cialized equipment belonging to a friend in Moscow. The facets without breaking the pattern, which meant that the
choice of equipment was also dictated by the material’s fra- pavilion had to finish precisely where the girdle started. In
gility and perfect cleavage, which prevented the use of coarse other words, the width of all 18 rows had to be exactly pro-
laps. This combined to make the work time-consuming, portional to the diagram. Another look at figure 12 is
especially since large facets were the order of the day. enough to show that the girdle does not simply separate
The next step required the rough to be rounded in order the crown from pavilion as in traditional faceted stones: It
to achieve the basic outline of the stone in its face-up po- is an essential element of the design.
sition. This meant that sufficient pressure had to be applied Many chances for something to go awry existed during
to the 1.5 kg preform using both hands. Eventually the cor- the cutting process. Only a slight error in judgment or cal-
rect shape resulted. The diameter measured approximately culation would result in disaster. For example, if the rows
85 mm, while the table facet came in at 65 mm. This al- of facets were just a bit too wide, there would not have
lowed the final weight to be approximated using formulas been enough space for the design and the pavilion would
and the facet plan shown in figure 12. The result, between have required a complete recut. If some were too narrow,
2,500 and 3,000 ct, indicated that it would supersede the the girdle would have needed recutting, resulting in a
largest known faceted spodumene, an 1,800 ct green spo- smaller stone, which would also have meant recutting the
dumene cut by John Sinkankas in 1959 that was destined table facet. Fortunately, the final facet positions aligned
for the Royal Ontario Museum in Toronto. perfectly around the girdle, and the stone’s final diameter
Although the rest of the preforming could not be done (83.7 mm) was off by just over 1.0 mm from the initial
traditionally by hand, VT managed to use his faceting unit calculations.

304 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SUMMER 2020

bination of wax and superglue. The wax is placed on the
hot dop, which is softened by the heat. The cold stone is
then placed, pavilion first, in contact with the wax, and an
impression of the surface is made in it. One drop of super-
glue is enough to secure the stone firmly in position, and
it can be removed by reapplying a little heat to warm the
wax again.
After three months of work, the final polishing step
was completed. The gem’s removal from the dop was an
exciting moment, and it was wonderful to finally see the
end product in all its glory (figure 14). The finished piece
weighed 3,051 ct, very close to the maximum estimated
weight of 3,000 ct.
The final step in the stone’s journey was its submission
to GIA’s Bangkok laboratory for examination and a report.
The gemologists were excited to study such a large kun-
zite, although the size, 83.75–83.79 × 65.00 mm, presented
challenges that limited the testing equipment that could
be used. Hence, basic gemological techniques came to the
fore, with Raman spectroscopy the only advanced tech-
nique used as a final confirmation. The stone was found to
be anisotropic with a loupe and microscopic observation

Figure 13. The polishing step on the stone’s pavilion

begins to show the intricate optical effects of the com-
pleted piece. Photo by Victor Tuzlukov.

Figure 12. A working diagram showing the facet posi-

tions. These helped estimate the final weight of the
finished stone before work began on the rough. Photo
by Victor Tuzlukov.

The next steps, pre-polishing and polishing of the pa-

vilion facets (figure 13), were straightforward in compari-
son. Yet, as mentioned previously, the sheer size meant
that the vertical pressure had to be applied equally over the
whole stone during pre-polishing, requiring modification
of the technique used for routine work on smaller stones.
If too much force was applied to the left or right side, or
back and forth, the facets would become too large and the
pattern would be broken. To overcome this, the stone’s
own weight provided sufficient force on the rotating lap so
that the pressure was always equally distributed.
After more than two months of daily work, the pavilion
was polished. The next step was transferring the stone to
a new dop to polish the crown. Unlike traditional lapidary
procedures, directly heating the stone as part of the transfer
was not an option, as the process could crack the stone
along the twin plane in the direction it needed to be
dopped, and because the size was once again a factor when
it came to how evenly the heat warmed the stone. This re-
sulted in the application of a technique that the late Jeff
Ford from Michigan shared with VT in 2006, using a com-

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SUMMER 2020 305

Figure 14. The 3,051 ct “Fragility of the Eternal,” con- Figure 15. Marble polar bear with silver fish on a pet-
taining 914 facets, is believed to be the world’s largest rified opalized wood stand. Photo by Svetlana Kul-
cut kunzite. Photo by Adisorn Wattanavanich. tenko; courtesy of Anton Akulenko.

up to 70× magnification, and this property was also appar- Local jewelers started using the wood fossils as stands
ent in terms of the strong pleochroism that could be easily for souvenirs (figure 15) and polished collectible samples.
seen as the stone was turned. The same purple and green The petrified wood appeared in regional jewelry stores and
colors are also clearly evident in figure 11. attracted Chinese tourists’ attention, creating demand for
The RI readings of 1.660–1.676 (birefringence of 0.016) the raw material from China’s market.
together with the trichroic colors seen through a dichro- More than a hundred samples were studied at the An-
scope showed it was optically biaxial. Inclusions such as alytical Center of the Far Eastern Geological Institute of
growth tubes, transparent crystals, negative crystal finger- the Far Eastern Branch of the Russian Academy of Sciences
prints, and a “natural” retained on the girdle all proved the (FEGI FEB RAS). We examined the mineral composition,
stone was natural. Combined with the weak and moderate structure, and gemological characteristics using standard
orange zoned long-wave ultraviolet fluorescence reaction, gemological equipment, a Nikon E100 POL optical micro-
the data were consistent with spodumene, kunzite variety. scope, and a MiniFlex2 X-ray diffractometer (XRD). Several
Since the stone was cut and submitted by VT with a examples of petrified wood were found: white, yellowish,
series of photographs documenting this fact, the GIA report marble-like, chalcedony-like, banded, banded chalcedony-
(dated May 27, 2020) also mentioned that it was represented
as cut by him. Since it was the largest known faceted kun-
zite examined by GIA at the time, and indeed the largest fac-
Figure 16. Cabochons of petrified wood from the Rus-
eted example of its kind known to exist, a GIA Notable
sian Far East. Photo by Dmitrii G. Fedoseev.
Letter was issued to accompany the report.
Victor Tuzlukov
Bangkok
Patcharee Kaewchaijaroenkit and Nicholas Sturman
GIA, Bangkok

Occurrence of petrified woods in the Russian Far East:

Gemology and origin. Petrified woods are used all over the
world as an excellent material for souvenirs, jewelry, and
spectacular collectible pieces. A new occurrence of petri-
fied wood was discovered in 2014 in the Primorsky Krai re-
gion of the Russian Far East, near the village of Kiparisovo.
The fossils were found in the northwestern part of a sand
and gravel quarry, at the contact of rhyolitic volcanic ash
tuffs and basalts overlapping them. The sample sizes varied
from several centimeters to two meters.

306 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SUMMER 2020

Figure 17. Top: Sche-
AR RD A matic cross section of a
conifer’s trunk showing
WR the transection (A), ra-
dial section (B), tangen-
tial section (C), annual
WR ring (AR), resin ducts
ML (RD), wood rays (WR),
BP
bordered pits (BP), and
middle lamellae (ML).
Bottom left: Transec-
B
tion of petrified wood
with wood rays (hori-
zontal lines), resin
ducts, and annual ring;
field of view 5.5 mm.
Bottom center: Radial
section of petrified
C wood with wood rays
(horizontal lines) and
tracheids with bordered
pits (round-shaped);
field of view 2.7 mm.
Bottom right: Tangen-
tial section of petrified
wood with chains of
middle lamellae; field
of view 5.5 mm. Photo-
micrographs by Dmitrii
G. Fedoseev.

like with brownish growth rings, partially carbonized, and The geology of the area allowed us to imagine the for-
black coalified (figure 16). mation conditions. As a result of catastrophic volcanic
The refractive indices of polished plates ranged from eruptions of rhyolitic magmas as glowing ash clouds, and
1.40 to 1.54 in different areas and corresponded to chalced- late effusions of mantle magmas, thick layers of volcanic
ony, opal, and quartz. Luminescence was weak bluish or ash and overlapping basalt flows were formed. The basalt
greenish under short-wave UV; most samples were inert lava flows outpoured into the water basin with a tempera-
under long-wave UV. Samples had a fibrous cellular struc- ture exceeding 1000°C. Lavas overlapped the flooded trunks
ture. The shape of the cells (tracheids) was angular, rectan- and thick bottom sediments of ash silts. When the basalt
gular, and sometimes subsquare (figure 17, bottom left). melt came into contact with water, the surface of the lava
The transections of the samples showed very narrow sin- flow instantly quenched with the formation of pillow lavas.
gle-row horizontal wood rays. Vertical resin ducts that The space between the pillow lavas was filled with clastic
looked like white dots were located in the latewood area glassy rocks known as hyaloclastites.
of the annual rings (vertical dark band) (figure 17, bottom Thus, the flooded trees were under the hyaloclastites
left). On the radial sections of the samples, we observed and pillow lavas. Some trees, under the weight of a lava
wood rays (horizontal lines) and bordered pits (round- flow, had taken a vertical position. Some trunks that sunk
shaped pores) (figure 17, bottom center). Middle lamellae in hyaloclasts were charred, while trunks in ash silts re-
(chains-like cells) were seen on a tangential section of pet- mained unchanged. During silicification, the charred parts
rified wood (figure 17, bottom right). of the trunks acquired a black color and the uncharred
The absence of vessels, the unique type of wood rays parts became white, light yellow, to brownish (figure 18).
and tracheids with bordered pits, the character of the mid- Buried trees underwent strong deformation with flattening
dle lamellae, and the presence of vertical resin ducts indi- of the trunks, splitting, and fragmentation of wood.
cated that these samples belonged to coniferous plants. The main source of silica when replacing the cells of
X-ray diffraction analysis showed that all the varieties of trees by quartz or opal was ash silts. This was promoted by
petrified wood had an opal-cristobalite-tridymite compo- a low-alkaline water-saturated volcanic ash with a high
sition. SiO2 content (over 72 wt.%). The silicon-containing

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SUMMER 2020 307

suddenly unable to perform their vital role in the gem-
stone supply chain. Without a constant flow of new
rough material, the wheels that drive the entire industry
stop turning.
When a crisis like this happens, people are forced to
turn to other sources. In the trading center of Chanthaburi,
Thailand, the most obvious source of rough is easily over-
looked: the back of the safe. Over the years, rough parcels
of varying qualities have been kept for a rainy day, and in
some cases material has been held back for years to keep
prices buoyant via artificial scarcity.
Difficult times have seen many established businesses
releasing some of these old parcels, sometimes of material
not seen in volume for years (or even decades), such as Ken-
yan rubies from the John Saul mine or spessartine “Fanta”
garnet from Nigeria.
Both authors had the privilege to handle one of the
unique pieces that came from the back of someone’s safe
into the market.
When SBL first saw this piece, his immediate reaction
was that it was tanzanite. Several other colleagues and
knowledgeable friends had the same first impression. It
was difficult to imagine the true identity of this crystal,
which was the size of a duck’s egg.
This was an 85 g, 427 ct piece of unheated cuprian or
Paraíba-type tourmaline rough from Mozambique (figure 19;
Figure 18. Left: Part of a petrified wood. Right: Opal- see also the video at www.gia.edu/gems-gemology/summer-
ized petrified wood sample with play-of-color. Photos
by Dmitrii G. Fedoseev.
Figure 19. This 85 g Mozambican cuprian tourma-
line was released in the Chanthaburi market in May
aqueous solutions penetrated woods that, under anaerobic 2020. While everyone can agree that this is a unique
conditions, represented local geochemical barriers where and valuable piece, there is no agreement on the
precipitation of free silica replacing plant cells occurred. price that can be put on it. Photo by Simon Bruce-
Depending on the concentration of silicon in the solution Lockhart.
in plant cells, the deposition of opal, chalcedony, or quartz
occurred. The foregoing description allows us to imagine a
majestic picture of how Nature created this wonderful
combination of the worlds of plants and stones.
Vera Pakhomova, Valentina Solyanik, Dmitrii Fedoseev,
Svetlana Y. Kultenko ([email protected]), and
Vitaliya B. Tishkina
Far East Geological Institute (FEGI FEB RAS)
Vladivostok, Russia
Valeriya S. Gusarova
Far Eastern Federal University (FEFU), Vladivostok

Portrait of a Paraíba rough: Challenging gems in hard

times. The impact of COVID-19 on the global gem trade
is unquestionable. Traditional sales and services in most
consuming countries came to a grinding halt. But the ef-
fects of lockdowns are also deeply felt in distant trade
centers that rely on a steady supply of rough gemstones to
feed their cutting, treating, and manufacturing industries.
Many of the brokers that take care of the transfer
from producing countries to manufacturing hubs were

308 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SUMMER 2020

UV-VIS -NIR SPECTRUM
Furthermore, heat treatment of any stone is unpredict-
able, but tourmaline is a notoriously fickle gem to heat.
Would you simply remove the frosted surface and heat it
before sawing and preforming? This carries the risk that
the gem will shatter uncontrollably during heating, but
ABSORBANCE

also the possibility of producing a single Paraíba tourma-

line of spectacular size and a coveted deeper color. Or
would you play it safe by cutting and grinding it into sev-
eral smaller preforms and heating the many pieces? This
mitigates some of the risks, but sacrifices the potential for
capitalizing on the size of the rough.
Even if you play it safe, there is still no guarantee that
minor unintended cracking will not require the resizing
400 500 600 700 800 900 1000
and re-orientating of preforms before faceting, which could
WAVELENGTH (nm) significantly reduce the final color and weight yet again.
The transformative yet risky process of slicing, pre-
Figure 20. The UV-Vis-NIR spectrum shows clear ab- forming, heating, and faceting is in some ways a gamble.
sorption peaks at 660 and 900 nm that are typical for The reality is that neither buyer nor vendor actually knows
cuprian tourmaline. what will happen, and there is an element of chance lay-
ered on top of the manufacturer’s skill. The coveted
“Paraíba” hue and saturation after heating are not guaran-
teed, and cases of “overcooking” cuprian tourmalines until
2020-gemnews-mozambique-paraiba-rough). The stone was only a pale pink color remains are not unheard of. These
shown in the rough gem trading circuit of Chanthaburi near many variables create unknowable outcomes.
the end of May 2020. The GIA laboratory in Bangkok was The most prudent buyers calculate their offers based
able to do some limited testing on the stone, confirming that on the risk of as many as six or more paler pieces, and
it was indeed a cuprian tourmaline (figure 20). offer much lower than expected when seeing this magnif-
Large crystals of copper-bearing tourmaline were seen icent piece of rough. The vendors optimistically hold to
a decade ago at the advent of the Mavuco mines, but noth- their dream of a higher value based upon two or three
ing of this size and quality has been seen on the open mar- larger, well-saturated pieces—and with such a wide dis-
ket for some time. For more on this material, see B.M. parity in value-outcome consensus, rough gems like this
Laurs et al., “Copper-bearing (Paraíba-type) tourmaline take time to sell. In the case of this piece, no agreement
from Mozambique,” Spring 2008 G&G, pp. 4–30. was reached after several rounds of back-and-forth be-
There is little doubt that such a large, prestigious piece tween the different parties, and the stone was returned to
attracts attention. After that, the consensus stops. While its owner.
it is undeniable that this stone is very valuable, no one can While this story is not unique, it highlights many as-
agree on precisely how valuable it is. pects of the gem trade that most people throughout the
The decisions of how to heat and cut it to open up that supply chain are unfamiliar with. In these tough COVID
coveted blue-green color produced by copper are critical to times, it becomes especially important to realize how
the value-adding process. The results of heat treating Mo- many hands have handled the gems we are seeing and
zambican cuprian tourmaline can be seen in figure 22 of which challenges every single one faces.
Laurs et al. (2008).
Simon Bruce-Lockhart
The unheated Paraíba gem was taken to the handful of
Chanthaburi, Thailand
factories in Chanthaburi and Bangkok capable of buying
and value-adding such a piece. Factoring in cracks, internal Wim Vertriest
stress, and the orientation of some inclusions, all the fac- GIA, Bangkok
tories determined that the stone should be cut into several
pieces. A new deposit of pink natrolite from Indonesia. In early
But everyone had a different opinion of exactly how to 2020, a parcel of stones representing some unique material
manufacture this piece. And this is where the dreams of found in early 2020 from the island of Nusa Kambangan,
Paraíba profits get smashed on the anvil of market reality. Central Java, Indonesia, was sent to Taiwan Union Lab of
A well-saturated large crystal will, when cut into nu- Gem Research (TULAB) for certification service. At the be-
merous smaller pieces, exhibit a paler saturation of the ginning, wholesalers claimed that the gemstone was thom-
original hue. Given that the market will pay much more sonite. Because its texture was very similar to the
per carat for larger Paraíbas of intense color than for smaller green-blue pectolite variety Larimar produced in the Dom-
paler Paraíbas, the range of potential value outcomes re- inican Republic, the stone was also misrepresented with
garding the number of pieces to cut is dizzying. the trade name “pink Larimar.”

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SUMMER 2020 309

Figure 21. Rough stones
and polished cabo-
chons of pink natrolite.
The heart-shaped cabo-
chon is 35.6 mm wide.
Photo by Yu-Ho Li.

By observing the rough stones with host rocks provided spectra of natrolite (Na2Al2Si3O10·2H2O) instead of thom-
by the supplier, the occurrence was found to be vein-filling sonite (NaCa2Al5Si5O20·6H2O) (figure 22). The UV-Vis
or cavity-filling within basalt. These stones had orange- spectra revealed that the pink natrolite stones had a wide
pink to brownish pink color with white zoning and showed absorption band at 400–570 nm (figure 23). The EDXRF
a botryoidal or radial crystal habit (figure 21). results also indicated that the pink gemstone was natro-
The average specific gravity of this parcel of gem- lite and contained a trace iron component. On the basis
stones was 2.24, and the spot RI was 1.49. With the of EDXRF and UV-Vis results, the orange-pink to brown-
owner’s consent, the parcel of polycrystalline material ish pink color appeared to have been caused by Fe3+; how-
was tested with a Vickers hardness tester. The values ever, this presumption still needs further verification.
were converted to Mohs hardness and ranged from 4.4 to This natrolite from Indonesia has a unique rose pink
4.6. The samples’ Raman spectra were analyzed and com- color and Larimar-like texture on polished surfaces, which
pared to the RRUFF mineral spectral database; unex- is not common in natrolite from other localities. Although
pectedly, the results were consistent with published initially misrepresented by the merchant as thomsonite or

R AMAN SPECTRA
30000
Pink natrolite from Indonesia
Natrolite: RRUFFID R040112
25000 Thomsonite: RRUFFID R050091
Figure 22. Stacked
Raman spectra of the
INTENSITY (COUNTS)

20000
pink natrolite from
Indonesia compared to
15000 those of natrolite and
thomsonite published
in the RRUFF data-
10000
base; all spectra are
normalized and base-
5000 line corrected.

0
150 650 1150 1650 2150 2650 3150 3650

RAMAN SHIFT (cm–1)

310 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SUMMER 2020

VISIBLE SPECTRUM
100

90

80
Figure 23. The normal-
70 ized visible-light trans-
60
mission spectrum of
pink natrolite shows a
%T

50
wide absorption band
40 at 400–570 nm that re-
sults in a brownish
30
pink to orange-pink
20 color.
10

0
400 450 500 550 600 650 700

WAVELENGTH (nm)

pink Larimar, its beautiful appearance and durability are and learn about mantle-derived magmas and associated
indeed comparable to Larimar. Subsequently, a new trade volcanism. As a side quest, we wanted to try prospecting.
name “rhodatrolite,” meaning “rose natrolite,” was devel- The park’s main attraction is diamonds, but there are
oped by the merchant for marketing this gemstone because also olivine, pyrope, almandine, and amethyst to be found.
of its color and texture. With enough mine production, Most of the diamonds recovered there are small (approx-
pink natrolite from Indonesia has the potential to become imately 0.20 ct or less); see figure 24. Rarely, prospectors
a popular gemstone on the market. have found stones larger than 1 carat, including the famous
Yu-Ho Li Uncle Sam, a whopping 40.23 ct. The park’s diamond-bear-
Institute of Earth Sciences ing rocks are comprised of hypabyssal olivine lamproite
National Taiwan Ocean University, Keelung and phlogopite-rich tuffs and breccias, characterized by
Huei-Fen Chen
Institute of Earth Sciences and
Figure 24. These diamonds from Crater of Diamonds
Center of Excellence for Oceans
State Park were mined previously by other prospec-
National Taiwan Ocean University
tors. They range up to approximately 0.06 ct. Photo
Shu-Hong Lin by Roy Bassoo.
Institute of Earth Sciences
National Taiwan Ocean University
Taiwan Union Lab of Gem Research, Taipei

DIAMONDS

Finders, keepers: Field trip to Crater of Diamonds, USA.

There’s only one place on earth where the general public
can prospect for diamonds directly from a primary kimber-
lite pipe: Crater of Diamonds State Park. This park is nes-
tled among the pines, 100 miles off the interstate near the
town of Murfreesboro, Arkansas. It boasts a network of sce-
nic walking trails, picnic sites, and campsites. At its Dia-
mond Discovery Center, visitors can learn about the local
geology. Staff are also on hand to identify any minerals that
are taken home, per their “finders, keepers” policy. The
park is an ideal spot for a field trip. The state also hosts
many other unusual igneous rocks, including carbonatite,
lamprophyre, and lamproite. With this in mind, our uni-
4 mm
versity petrology class piled into a van to visit Arkansas

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SUMMER 2020 311

Figure 25. Prospecting for diamonds at the Mauney mine, now Crater of Diamonds State Park, circa 1908. Photo
courtesy of the Crater of Diamonds Archives.

high Mg# values (atomic ratio of Mg to Fe in an igneous 25). After a string of ownership changes and failed business
rock) and K2O content (E. Walker, “Petrogenesis of the Prai- ventures, visions of a South African–style diamond district
rie Creek, Arkansas, diamondiferous olivine lamproite,” never materialized. In 1972 the state of Arkansas pur-
PhD thesis, University of Western Ontario, Ontario, Ca- chased the land and opened it to public prospecting. Since
nada, 1991). The deposit grades between 0.01 and 1.25 ca- then, visitors have found more than 33,000 diamonds
rats per 100 tons (D.P. Dunn, “Xenolith mineralogy and (www.arkansasstateparks.com).
geology of the Prairie Creek lamproite province, Arkan- After a couple of days learning about the igneous rocks
sas,” PhD thesis, University of Texas, Austin, 2002). In the in the area, it was time to test our luck at treasure hunting.
early twentieth century, when the richest material was We gathered buckets, shovels, and sieves from park head-
mined, there was a brief diamond rush in the area (figure quarters for a small rental fee and headed out to 15 hectares

Figure 26. A local prospector digging for diamond ore and demonstrating sieving technique. Photos courtesy of
Glenn Worthington.

312 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SUMMER 2020

0.5 mm 0.5 mm 0.5 mm

Figure 27. Three of the diamonds the authors collected from Crater of Diamonds State Park (approximately 0.01–
0.05 ct). Note the irregular and fragmented morphology and brown to yellow colors. Photos by Roy Bassoo.

of routinely plowed “blue ground.” We chose a spot, shov- octahedral with uncommon dodecahedrons and macles.
eled soil into buckets, and carried these heavy parcels of Resorption pits and terraced surfaces were pervasive. The
dirt to the wash station for sieving (figure 26). The trick is diamonds were also fragmented, likely due to the explosive
to shake the sieve back and forth just beneath the water in nature of the diatreme crater.
the horizontal plane, then rotate and repeat. This motion Treasure in hand, we then turned to the science. Back
should concentrate the heavy minerals in the center and in our lab, we ground and polished the diamonds into
bottom of the sieve. Then, like a master chef, the prospec- wafers to examine their composition and crystal interiors.
tor flips the entire mineral concentrate upside down and Raman spectroscopy determined that inclusions of rutile
examines the heavies that sink to the bottom. and diopside were common (figure 28). N contents, meas-
In three hours, we found a small collection of dia- ured with Fourier-transform infrared spectroscopy, ranged
monds! Granted, they were also very small in size, ranging from trace amounts (type IIa) to 1250 ppm (type IaAB).
from 0.01 to 0.05 ct (figure 27). Most prospectors at Crater Cathodoluminescence imagery showed blue to turquoise
of Diamonds miss stones this small, but as trained min- response colors, typical of diamonds mined from primary
eralogists we were attentive to small crystals. The dia- deposits worldwide. They also feature fascinating patterns
monds were predominantly colorless, although a couple of growth banding and resorption truncations, which indi-
were yellow to brown. Their dominant morphology was cate a complex mantle crystallization history (figure 29).

Figure 28. Terraced dissolution surface textures and inclusions of rutile and diopside embedded within the dia-
mond. The diopside inclusions are aligned along a growth band. Photomicrograph by Roy Bassoo.

Rutile Terraces

Resorption Diopside
pit

0.15 mm

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SUMMER 2020 313

At the same time, the current lack of production in Taiwan
Growth bands is driving prices substantially higher. Thus, imitations and
treatments are gradually emerging in the marketplace, con-
fusing customers. Recently we obtained three polished spe-
cimens from a gift shop in Taidong County, Taiwan (near
the locality of the gem silica mine), that appeared to have
Resorption good quality of color and transparency (figure 30). The mer-
truncation chant claimed they were gem silica mined in Taiwan, but
they were subsequently identified as glass imitations.
Standard gemological tests showed the following prop-
erties of the three samples: color—uniform greenish blue;
diaphaneity—translucent; spot RI—1.48; SG—2.48 to ap-
proximately 2.57; fluorescence—inert to long-wave UV ra-
diation, weak greenish white to short-wave UV radiation.
Actual gem silica, meanwhile, has the following properties:
color—commonly uneven; spot RI—1.54; SG—approx-
imately 2.63. Examination with a gemological microscope
revealed numerous obvious gas bubbles of different sizes (fig-
ure 31, left). In the case of the marquise sample, numerous
gas bubbles led to a relatively low specific gravity. When
0.25 mm viewed with oblique incident light, many small dimples
were visible on the polished surface (figure 31, right). In ad-
Figure 29. Cathodoluminescence image of a polished dition, there were some small conchoidal fractures along the
diamond wafer, showing complex banding and trun- bottom edges of samples. In order to gain a more thorough
cations that highlight different diamond growth con- understanding, further advanced analysis was carried out.
ditions. Image by Roy Bassoo. FTIR reflectance spectroscopy showed a main reflected
band at approximately 1065 cm–1 (figure 32), consistent with
the characteristic spectrum of glass (see Fall 2019 GNI, p.
443–445; T.B. Wang et al., “Relationship between the
We found Crater of Diamonds to be a thoroughly enjoy- frequency of the main LO mode of silica glass and angle of
able treasure hunting adventure. Our visit gave us hands- incidence,” Journal of Chemical Physics, Vol. 119, No. 1,
on experience with diamond-bearing igneous rocks and 2003, pp. 505–508). Chemical analysis by LA-ICP-MS de-
challenged us to imagine how diamonds formed beneath tected multiple elements: Si (averaging 617000 ppmw), Na
the North American craton. We have new appreciation for (averaging 151000 ppmw), Al (averaging 101000 ppmw), Ca
the back-aching labor required by the artisanal miners (averaging 24900 ppmw), K (averaging 19600 ppmw), Ba (aver-
worldwide who are the foundation of the gem trade. Scien-
tific understanding and trade awareness were the substance
of our visit, highlighted by moments of joyful discovery. We
encourage readers to put Crater of Diamonds State Park on Figure 30. Two oval and one marquise cabochon of
their gemology “bucket list” to hunt for their very own greenish blue glass samples ranging from 2.17 to 2.60
American diamond. ct. Photo by Min Ye.
Roy Bassoo and Kenneth S. Befus
Department of Geosciences, Baylor University
Waco, Texas

SYNTHETICS AND SIMULANTS

Greenish blue glass imitating gem silica. Gem silica, also

known as “chrysocolla chalcedony,” is considered the most
valuable variety of chalcedony. Its attractive blue to bluish
green color is generated by finely disseminated minute in-
clusions of chrysocolla. The main sources include Taiwan,
the United States, Mexico, Peru, and Indonesia. For years,
gem silica has been especially popular in Taiwan. Due to the
pleasing saturated blue color of Taiwanese material, which
is among the world’s finest, it commands a premium price.

314 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SUMMER 2020

Figure 31. Left: Numerous
gas bubbles of various sizes
Bubbles were typical internal fea-
tures of the three samples;
field of view 1.28 mm.
Dimple Right: Many small dimples
could be found on the pol-
ished surface; field of view
3.70 mm. Photomicrographs
by Min Ye in overhead light-
ing (left) and oblique illumi-
nation (right).

aging 17800 ppmw), Pb (averaging 3840 ppmw), Mg (averaging Combining all the evidence, the three greenish blue
2580 ppmw), Fe (averaging 1620 ppmw), Cu (averaging 5940 samples were confirmed to be artificial glass colored by an
ppmw), and Cr (averaging 716 ppmw). This composition iden- unknown copper additive. Gas bubble inclusions were the
tified the three samples as silicate glass, which is decidedly diagnostic feature for identification, and FTIR spectra were
different from gem silica (whose majority component is silica also helpful. This case is a reminder that customers need
and chrysocolla, a copper silicate). UV-Vis-NIR spectra of the to be cautious when purchasing gemstones, even near the
specimens presented a strong broad absorption band around geographic source.
760 nm (figure 33), revealing Cu2+ as the coloring agent. In ad- Min Ye and Andy H. Shen ([email protected])
dition, two weak absorption bands at 424 and 439 nm might Gemmological Institute
have been related to trace amounts of Fe3+, while the 690 nm China University of Geosciences, Wuhan
band was possibly due to the existence of Cr3+ (see W. Thiem-
sorn et al., “Redox ratio and optical absorption of polyvalent
ions in industrial glasses,” Bulletin of Materials Science, Vol.
30, 2007, pp. 487–495; V. Vercamer, “Spectroscopic and struc-
Figure 33. Typical UV-Vis-NIR absorption spectra of
tural properties of iron in silicate glasses,” PhD thesis, Uni-
the three samples (green trace) were compared with
versité Pierre et Marie Curie-Paris VI, 2016, pp. 97–145).
those of Cu2+-doped glass (blue trace) and natural
greenish blue gem silica (red trace, collected by the au-
thor). In the typical plot for the glass samples, the ab-
Figure 32. Representative infrared reflectance spectra sorption band around 760 nm is assigned to Cu2+. Two
of the three samples (green trace) were compared with weak absorption bands in the glass imitation at 424
the reference spectra for glass (blue trace) and natural and 439 nm might be related to trace amounts of iron,
gem silica (red trace, collected by the author). The while the 690 nm band is possibly due to the presence
diagnostic reflected peak of the glass sample is near of chromium. The spectra are offset for clarity.
1065 cm–1, quite different from the gem silica peaks.
The spectra are offset for clarity.
UV-VIS -NIR SPECTRA
IR SPECTRA 760 Glass sample
690
Cu2+-doped glass
Glass sample 1065 Gem silica
Glass reference 465
Gem silica 424
439 755
1058
REFLECTANCE (%)

ABSORBANCE

720

1415

1600 1400 1200 1000 800 600 400 400 600 800 1000 1200 1400 1600

WAVENUMBER (cm–1) WAVELENGTH (nm)

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SUMMER 2020 315

Marble imitation of jadeite rough. In recent decades, jadeite
prices have risen dramatically. Driven by profit, a variety
of imitations are found in the jadeite jewelry market.
Meanwhile, imitations such as quartzite are fixtures in the
rough jadeite market. Recently, our research group received
for testing a 30 kg stone that resembled jadeite rough and
was submitted as such. The stone had a yellow weathered
skin with a grainy texture, similar to jadeite. There was a
narrow “window” (figure 34) that showed the green color
of the material underneath the skin. For identification pur-
poses, the authors exposed another area on the surface of
the rough and then ground it for testing.
Due to the weight, some standard gemological testing
methods could not be used, such as measurement of re-
fractive index and specific gravity. Therefore, the sample Figure 35. The newly exposed part of the rough stone
was studied by visual and 10× loupe observation, hard- was identified as marble. Photo by Huang Jing.

Figure 34. This 30 kg stone was submitted as a piece ness testing, chemical reagent detection, and infrared
of jadeite rough. A narrow window (in the red circle) spectroscopy.
opened on the surface of the rough reveals a green Scratch testing revealed a Mohs hardness between 2 and
material underneath the skin. Photo by Huang Jing. 5, lower than jadeite’s hardness of 6–7. The surface was yel-
low, just like the weathered skin of jadeite rough, and the
narrow window in figure 34 revealed green color below the
skin. Scrubbing the window with alcohol removed the
green color, indicating that the window had been dyed. To
test the material, diluted hydrochloric acid was dropped on
the newly exposed portion in figure 35. The acid foamed
continuously, showing that the sample contained carbonate
(Z. Hanli et al., “Study on acid polishing of carbonate white
jade,” Journal of Gems and Gemmology, Vol. 5, No. 4,
2003, pp. 24–27). The acid testing combined with visible
observations and hardness testing identified the sample as
marble. The jadeite granular structure came from the mar-
ble’s cleavage planes.
Testing of the exposed part by infrared spectrometry re-
vealed 1525, 1432, 1080, 883, 670, 553, and 483 cm–1 ab-
sorbance peaks, characteristic of serpentinized marble.
The serpentinized marble’s yellow weathered skin and
grainy texture could have easily been mistaken for jadeite
rough. This serves as an alert to the industry to be careful
even when examining rough, as imitations are showing up.
Huang Jing, Yan Shuyu, and Chen Shuxiang (cor-
responding author, [email protected])
School of Materials Science and Engineering
Qilu University of Technology
(Shandong Academy of Sciences)
Shandong, China
Cheng Youfa
National Gold and Diamond Testing Centre of China
Shandong

316 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SUMMER 2020