SPRING 2023

VOLUME LIX

THE QUARTERLY JOURNAL OF THE GEMOLOGICAL INSTITUTE OF AMERICA

Characterization of Montana Sapphire

Morphology of Colombian Emerald
2023 Tucson Report
 Spring 2023
VOLUME 59, No. 1
EDITORIAL
1 Sapphire from Montana’s Secondary Deposits, Colombian Emerald Morphology,
The 2023 Tucson Shows, and More…
Duncan Pay

FEATURE ARTICLES
2 Gemological Characterization of Montana Sapphire from the Secondary
p. 20 Deposits at Rock Creek, Missouri River, and Dry Cottonwood Creek
Aaron C. Palke, Nathan D. Renfro, Jeffrey R. Hapeman, and Richard B. Berg
Examines the mining history and geological characteristics of the three secondary sapphire
deposits in the state of Montana, as well as the properties of these sapphires (unheated and
heat-treated).
46 Morphology of Colombian Emerald: Some Less Common Cases and
Their Growth and Dissolution History
Karl Schmetzer and Gérard Martayan
Evaluates the growth mechanisms of special morphological features in Colombian emerald,
including possible relationships to trapiche emeralds and samples displaying the gota de
p. 63 aceite effect.

REGULAR FEATURES
72 Lab Notes
p. 74 Purple anhydrite • Pink pyrope garnet • Libyan Desert glass bangle • Atypical bead
cultured pearls • Treated freshwater non-bead cultured pearls with antique appearance
• Heart-shaped golden South Sea cultured pearl • Gastropod shell in a shell blister • Ruby
with synthetic overgrowth • Exceptionally large synthetic rutile

83 The Dr. Edward J. Gübelin Most Valuable Article Award

84 G&G Micro-World
Three-phase inclusion in alexandrite • Apatite in hackmanite • Inclusions in chondrodite
p. 89
• “Flying insect” in diamond • Starry night in diamond • Xenomorphic garnet in diamond
• Arrow-patterned negative crystals in sapphire • Solid inclusions in flame-fusion ruby
• Quarterly Crystal: Columbite(?) in beryl

92 2023 G&G Challenge

94 Diamond Reflections
A look at plastic deformation, one of the most prevalent features in natural diamond.

102 Gem News International

Tucson 2023 • Enormous cat’s-eye aquamarine • Ethiopian orbicular chalcedony • Italian
coral • Exceptional gem acquisitions by the Kreises • Double pseudomorph ikaite-calcite-opal
• Cultured pearl market update • Edison pearls • Natural nacreous pearls • Rare pearls from
ECIJA • Petroleum-included quartz from Madagascar • Orange sapphire from Greenland
• Sweet Home mine rhodochrosite • New tourmaline find in San Diego County • Vibrant
p. 129
green “Transvaal jade” • Misfit Diamonds • Jewelry designs by Sean Hill, Brenda Smith, and
Surbhi • Zeki Karaca luxury pens • Virtu Gem’s ethical supply chain practices • Luminescent
synthetic garnet-like crystals • Exhibits by the GIA Museum and the GIA Library
• Gianmaria Buccellati Foundation Award • Atypical bead cultured pearls from the UAE
• Chilean abalone cultured pearls • Microscopic shells in natural pearls • Color study of
fancy sapphire from Greenland • Liberian ruby • Forsterite in purple Tanzanian spinel
• Grayish blue CVD diamond • Phosphorescence of HPHT-grown diamond • Synthetic color-
change wakefieldite • Heated purplish pink sapphire with monazite inclusions
p. 140 • Chromophores in irradiated pink sapphire
Editorial Staff
Editor-in-Chief Associate Technical Editor Editors, Colored Stones [Link]/gems-gemology
Duncan Pay Si Athena Chen Unearthed
Aaron C. Palke
Managing Editor Editors, Lab Notes Customer Service
James E. Shigley
Stuart D. Overlin Thomas M. Moses (760) 603-4200
soverlin@[Link] Shane F. McClure Editor, Diamond Reflections gandg@[Link]

Editor
Evan M. Smith Subscriptions
Editors, Micro-World
Brooke Goedert Nathan Renfro Contributing Editors Copies of the current issue may be purchased for
James E. Shigley $29.95 plus shipping. Subscriptions are $79.99 for one
Editorial Coordinator Elise A. Skalwold year (4 issues) in the U.S. and $99.99 elsewhere. Cana-
Erica Zaidman John I. Koivula Raquel Alonso-Perez dian subscribers should add GST. Discounts are avail-
Editors, Gem News Editor-in-Chief Emeritus able for renewals, group subscriptions, GIA alumni,
Senior Technical Editor and current GIA students. To purchase print subscrip-
Jennifer Stone-Sundberg Gagan Choudhary Alice S. Keller
tions, visit [Link] or contact Customer Service.
Christopher M. Breeding Assistant Editor For institutional rates, contact Customer Service.
Technical Editor Guanghai Shi
Tao Z. Hsu Erin Hogarth
Database Coverage
Gems & Gemology’s impact factor is 2.045, accord-
Production Staff ing to the 2021 Journal Citation Reports by Clarivate
Analytics (issued June 2022). G&G is abstracted in
Creative Director Photo/Video Producer Multimedia Associate Thomson Reuters products (Current Contents: Phys-
Faizah Bhatti Kevin Schumacher Christopher Bonine ical, Chemical & Earth Sciences and Science Cita-
tion Index—Expanded, including the Web of
Production and Photographer Video Production
Knowledge) and other databases. For a complete list
Multimedia Specialist Robert Weldon Albert Salvato of sources abstracting G&G, go to [Link]/gems-
Michael Creighton gemology, and click on “Publication Information.”

Editorial Review Board Manuscript Submissions

Gems & Gemology, a peer-reviewed journal, welcomes
Ahmadjan Abduriyim Richard W. Hughes Benjamin Rondeau the submission of articles on all aspects of the field.
Tokyo, Japan Bangkok, Thailand Nantes, France Please see the Author Guidelines at [Link]/gems-
Timothy Adams Jaroslav Hyršl George R. Rossman gemology or contact the Managing Editor. Letters
San Diego, California Prague, Czech Republic Pasadena, California on articles published in G&G are also welcome.
Please note that Field Reports, Lab Notes, Gem News
Edward W. Boehm Dorrit Jacob Sudarat Saeseaw International, Micro-World, Colored Stones Un-
Chattanooga, Tennessee Canberra, Australia Bangkok, Thailand earthed, Diamond Reflections, and Charts are not
James E. Butler A.J.A. (Bram) Janse Karl Schmetzer peer-reviewed sections but do undergo technical and
Washington, DC Perth, Australia Petershausen, Germany editorial review.

Alan T. Collins Mary L. Johnson Andy Shen Copyright and Reprint Permission
London, UK San Diego, California Wuhan, China Abstracting is permitted with credit to the source.
Sally Eaton-Magaña Robert E. Kane Guanghai Shi Libraries are permitted to photocopy beyond the lim-
Carlsbad, California Helena, Montana Beijing, China its of U.S. copyright law for private use of patrons.
Instructors are permitted to reproduce isolated arti-
John L. Emmett Stefanos Karampelas James E. Shigley cles and photographs/images owned by G&G for
Brush Prairie, Washington Paris, France Carlsbad, California noncommercial classroom use without fee. Use of
Emmanuel Fritsch Lore Kiefert Elisabeth Strack photographs/images under copyright by external par-
Nantes, France Lucerne, Switzerland Hamburg, Germany ties is prohibited without the express permission of
the photographer or owner of the image, as listed in
Eloïse Gaillou Simon Lawson Nicholas Sturman the credits. For other copying, reprint, or republica-
Paris, France Maidenhead, UK Bangkok, Thailand tion permission, please contact the Managing Editor.
Al Gilbertson Ren Lu D. Brian Thompson Gems & Gemology is published quarterly by the
Carlsbad, California Wuhan, China Florence, Alabama Gemological Institute of America, a nonprofit educa-
Gaston Giuliani Thomas M. Moses Fanus Viljoen tional organization for the gem and jewelry industry.
Nancy, France New York, New York Johannesburg, South Africa Postmaster: Return undeliverable copies of Gems &
Gemology to GIA, The Robert Mouawad Campus,
Lee A. Groat Laura Otter Wuyi Wang
5345 Armada Drive, Carlsbad, CA 92008.
Vancouver, Canada Canberra, Australia New York, New York
Our Canadian goods and service registration number
Yunbin Guan Aaron C. Palke Christopher M. Welbourn is 126142892RT.
Pasadena, California Carlsbad, California Reading, UK
Any opinions expressed in signed articles are under-
George Harlow Ilene Reinitz Chunhui Zhou stood to be opinions of the authors and not of the
New York, New York Chicago, Illinois New York, New York publisher.
Peter Heaney Nathan Renfro J.C. (Hanco) Zwaan
University Park, Pennsylvania Carlsbad, California Leiden, The Netherlands

About the Cover

Designed by Zoltan David, this opal and green tourmaline necklace contains Knightsteel links with 99.42 total
carats of Ethiopian opal beads. The 10.97 ct tourmaline from Afghanistan, faceted by Clay Zava, is set in a bezel
with platinum inlay and 0.22 carats of diamonds. The necklace was one of many stunning pieces David exhibited
at the 2023 Tucson shows, which are reported on in this issue. Photo by Robert Weldon; courtesy of Zoltan David.
Printing is by L+L Printers, Carlsbad, CA.
GIA World Headquarters The Robert Mouawad Campus 5345 Armada Drive Carlsbad, CA 92008 USA
© 2023 Gemological Institute of America All rights reserved. ISSN 0016-626X
 Sapphire from Montana’s Secondary
Deposits, Colombian Emerald Morphology,
the 2023 Tucson Shows, and More…
Welcome to the Spring 2023 issue of Gems & Gemology! A new volume year kicks off
with our recap of the Tucson gem shows, where buyers, sellers, and industry professionals
from all over the globe converge. And with two major feature articles, our regular sections,
and the annual G&G Challenge, this issue is sure to captivate readers.

In our lead article, Dr. Aaron Palke and coauthors present their study of the gemological
characteristics of sapphire from Montana’s secondary deposits. Using Rock Creek,
Missouri River, and
Dry Cottonwood “The article also details the fascinating history of
Creek samples in a sapphire mining in these three Montana mining
variety of colors, districts, which still hold substantial resources.”
the authors provide
typical inclusion scenes and trace element chemistry fingerprints. The article also details the fascinating history of
sapphire mining in these three Montana mining districts, which still hold substantial resources.

Next, Dr. Karl Schmetzer and Gérard Martayan investigate the less common morphological and growth
phenomena of Colombian emerald samples to understand their growth and post-growth history. Based on their
observation of 15 isolated emerald crystals and seven specimens in matrix using optical methods, they offer insight
into the different patterns, with a goal of expanding the database of morphological features in Colombian emerald.

Our regular features provide informative snapshots, beginning with Lab Notes. Highlights from the section
include a rare example of the purple collector gem anhydrite, a heart-shaped golden South Sea cultured pearl
displaying natural shape and color, and the largest sample of synthetic rutile examined at GIA to date. Explore the
inner landscapes of gemstones in our Micro-World section, where you’ll see rare cloud inclusions resembling a
starry night in diamond, a macroscopic mineral inclusion cluster in a Pakistani aquamarine crystal, a three-phase
inclusion in alexandrite, and more. Diamond Reflections returns in this issue, this time exploring plastic deforma-
tion in diamond. Our Gem News International section begins with nearly 30 pages covering the 2023 Tucson gem
shows, ranging from the latest trends and market updates to noteworthy designs, followed by an opulent photo
gallery packed with finds from the shows. The GNI regular features include various pearl discoveries from around
the world and a new source of ruby from Liberia.

Finally, test your gemological knowledge of our 2022 feature articles in the annual G&G Challenge, and find out
which articles received the Dr. Edward J. Gübelin Most Valuable Article Award.

We hope you enjoy the latest issue of Gems & Gemology!

Duncan Pay | Editor-in-Chief | dpay@[Link]

EDITORIAL GEMS & GEMOLOGY SPRING 2023 1

FEATURE ARTICLES

GEMOLOGICAL CHARACTERIZATION OF
MONTANA SAPPHIRE FROM THE SECONDARY
DEPOSITS AT ROCK CREEK, MISSOURI RIVER,
AND DRY COTTONWOOD CREEK
Aaron C. Palke, Nathan D. Renfro, Jeffrey R. Hapeman, and Richard B. Berg

Montana has produced large volumes of sapphire from secondary deposits for more than 150 years. While early
production was mostly for industrial use as watch and instrument bearings, modern heat treatment technology
has enabled the deposits to be worked profitably as gemstone mines, bringing more of this material to the gem
and jewelry trade than ever before. Although pale blue and pale green sapphires make up the bulk of the pro-
duction, fine Montana sapphire comes in nearly every hue, including yellow, orange, purple, and pink, along
with very rare finds of ruby. Secondary Montana sapphire shares some similarities with those from Umba and
Songea in Tanzania, but it is generally straightforward to separate them on the basis of inclusion scenes and
trace element chemistry fingerprints.

N
early 160 years ago, gold prospectors in the oxidizing and reducing environments at high tem-
American West stumbled upon one of North peratures—can be used to either create or intensify
America’s most important gemstone de- a variety of colors (e.g., Emmett and Douthit, 1993).
posits, discovering sapphire in several locations in Whether heated or with a natural, untreated color,
Montana. These were first recovered along the Mis- secondary Montana sapphire occurs in a wide range
souri River near Helena, and additional deposits were
later found at Rock Creek and Dry Cottonwood
Creek. The sapphires recovered from these secondary
deposits generally have overlapping properties and In Brief
are collectively referred to as “Montana sapphire.” • Sapphires have been produced from three secondary
[Note that these are distinct in their gemological deposits in Montana for more than 150 years.
properties from “Yogo sapphires,” which are mined • Most sapphires produced are less than 1 ct and require
from Montana’s only primary deposit, at Yogo Gulch; heat treatment to improve their color. However, large
see Renfro et al. (2017).] Surprisingly, the secondary unheated stones of several carats are available in the
deposits were not initially mined for fine jewelry- market.
grade sapphire but for gold and for industrial-grade • Montana sapphires have characteristic inclusion suites
sapphire to use as watch and instrument bearings. and chemical profiles that allow their origin to be con-
While the deposits produce many stones with ex- fidently identified.
ceptional natural color, the majority require (or at
least benefit from) carefully controlled high-temper-
ature heat treatment to enhance their color for the of sizes and colors including blue, green, yellow, or-
jewelry market. Various heating procedures—in both ange, purple, and pink (figures 1–5). With the rein-
troduction of large-scale mining 25–30 years ago and
the recent acceleration of mining activities at Mis-
See end of article for About the Authors and Acknowledgments.
GEMS & GEMOLOGY, Vol. 59, No. 1, pp. 2–45,
souri River and especially Rock Creek, fine Montana
[Link] sapphire is available in the gem and jewelry market
© 2023 Gemological Institute of America in quantities never seen before. It is also growing in

2 MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023
Figure 1. Top production from Rock Creek from the 2021 mining season, in rough and cut form. The rough stones
range from 10.05 to 67.15 ct and the faceted stones from 3.09 to 12.18 ct. The yellow sapphire at top center was
named the Yellowstone sapphire and is part of the Somewhere in the Rainbow collection. For the faceted stones, the
blue stone on the left and the two blue stones on the bottom are heated, while the others are unheated. Note that
one of the rough stones is missing in the faceted stone layout. Photos by Robert Weldon; courtesy of Potentate Min-
ing, LLC.

popularity with the increasing emphasis on trans- jewelry industry. Demand is also increasing as con-
parency and traceability in supply chains within the sumers become more interested in the less saturated

Figure 2. At 17.04 ct, this natural-color (unheated) Figure 3. The “Bitterroot Flower” brooch by Paula
rough sapphire crystal from Eldorado Bar on the Mis- Crevoshay contains nearly 400 Montana sapphires: 79
souri River is exceptionally large for Montana’s sec- pink (6.03 carats total), 116 yellow (7.46 carats total),
ondary deposits. Photo by Orasa Weldon; courtesy of and 202 additional fancy-color (16.26 carats total).
Pala International. Photo by Robert Weldon; courtesy of Paula Crevoshay.

MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023 3
 stone sapphire, but most were much smaller, from
less than a carat to several carats in size. Seventy-
three Montana sapphires were included in our laser
ablation–inductively coupled plasma–mass spec-
trometry (LA-ICP-MS) trace element measure-
ments, as well as 19 sapphires from Songea,
Tanzania, and 17 from Umba, Tanzania. The Songea
and Umba sapphires were either collected directly
in the field by GIA field gemologists or donated by
Dr. Emmett.

Ultraviolet/Visible/Near-Infrared (UV-Vis-NIR)
Spectroscopy. Spectra were collected with a
PerkinElmer Lambda 950 UV-Vis spectrophotometer
(PE 950) in the 200–1000 nm range, with 1 nm spec-
tral resolution. Spectra presented were collected
from samples oriented as wafers perpendicular to the
optic axis, which allows collection only of the ordi-
Figure 4. This “Fiume” ring features a wide range of nary ray (o-ray) spectrum. Samples used in heat
colors seen in Montana sapphire. The sapphires are all treatment experiments were mounted on aluminum
heat treated, with 2.07 total carat weight. Photo cour- plates with 1 mm apertures during spectroscopy.
tesy of Raintree Jewelry. They were set in place using wax to allow the same
region of the sample to be measured before and after
heat treatment.
blue and teal and fancy-color sapphires that are com-
mon in Montana. Fortunately, these sapphires have Fourier-Transform Infrared (FTIR) Spectroscopy.
unique gemological and chemical characteristics FTIR spectra were collected using a Thermo Fisher
that allow their geographic origin to be confidently Nicolet 6700 FTIR spectrometer equipped with an
determined. XT-KBr beam splitter and a mercury-cadmium-tel-
luride (MCT) detector operating with a 4× beam con-
MATERIALS AND METHODS denser accessory. The resolution was set at 4 cm–1
with 1.928 cm–1 data spacing.
Samples. The Montana sapphires examined in this
study are from a variety of sources. The bulk of the
samples were collected in 2015 by GIA gemologists Raman Spectroscopy. Inclusions were identified,
at the Potentate mine at Rock Creek and from Cass when possible, using Raman spectroscopy with a
Thompson’s mining operation at Eldorado Bar on Renishaw inVia Raman microscope system. The
the Missouri River. Additional samples came from Raman spectra of the inclusions were excited by a
the collections of the authors and were generally Modu-Laser Stellar-REN Ar-ion laser producing
sourced directly from current operations at Rock highly polarized light at 514 nm and collected at a
Creek and Missouri River. Sapphires from the South nominal resolution of 3 cm–1 in the 2000–200 cm–1
Fork of Dry Cottonwood Creek were donated by Dr. range. In many cases, the confocal capabilities of the
John Emmett, director of Crystal Chemistry in Raman system allowed inclusions beneath the sur-
Brush Prairie, Washington, who obtained them from face to be analyzed.
the mining operations of American Gem Corpora-
tion in 1994–1995. Observations described in this LA-ICP-MS. Trace element chemistry was collected
study come primarily from the more than 800 sap- by LA-ICP-MS using a Thermo Fisher iCAP Qc ICP-
phires in the GIA Colored Stone Reference Collec- MS, coupled to an Elemental Scientific Lasers NWR
tion gathered from the sources described above. 213 laser ablation system with a frequency-quintu-
Additional sapphires were provided from the collec- pled Nd:YAG laser (213 nm wavelength with 4 ns
tion of John I. Koivula, GIA’s analytical microscopist. pulse width). Ablation was carried out with 55 μm
The largest sapphire studied was the 67.15 ct Yellow- spot sizes, with a fluence of 8–10 J/cm2 and repetition

4 MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023
Figure 5. Early gold prospecting led to the discovery of Montana sapphire. This gold nugget from Scratchgravel Hills
measures about 4 × 7 cm; courtesy of Bill Larson. The rough sapphires and the necklace containing rough sapphires
and a 3.5 ct lavender spinel cut by Larry Woods are courtesy of Diana Widman. Photo by Robert Weldon.

MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023 5
rates of 20 Hz. The isotope 27Al was used as an inter- they were mainly targeting gold, whose value was
nal standard at 529250 ppmw, and synthetic and nat- much more obvious than the shiny dense pebbles of
ural corundum reference materials were used as sapphire they sometimes found in their jigs and pans.
external standards (Stone-Sundberg et al., 2017). De- When the first sapphires were discovered in 1865 at
tection limits ranged from 0.1–0.3 ppma for magne- the Missouri River, their range of natural colors did
sium, 0.5–2.0 ppma for titanium, 0.03–0.2 ppma for not suit the gem and jewelry industry of that time—
vanadium, 0.06–1.3 ppma for chromium, 1–5 ppma that happened 30 years later, when bright blue sap-
for iron, and 0.03–0.07 ppma for gallium. Trace ele- phires were found at Yogo Gulch. However, when
ment values are reported here in parts per million on sapphires were found in abundance at Rock Creek in
an atomic basis rather than the more typical parts per 1892, another lucrative market opened up for these
million by weight unit used for trace elements in stones as watch and instrument bearings (figure 6).
many geochemical studies. Units of ppma are the In fact, mining efforts until the late 1920s were
standard used in GIA laboratories for corundum, as driven almost entirely by the search for gold and the
they allow a simpler analysis of crystal chemical use of sapphires as watch bearings. The bulk of the
properties and an understanding of the color mecha- sapphires mined in Montana had ended up being
nisms of sapphire and ruby. Conversion factors are used for industrial purposes. The availability and af-
determined by a simple formula that can be found in fordability of synthetic ruby and sapphire (grown by
table 1 of Emmett et al. (2003). the Verneuil process) in the 1920s and 1930s ended
the first great heyday of sapphire mining in Montana
Heat Treatment. Some heat treatment studies were (Berg, 2014). After some small-scale mining to sup-
carried out at the commercial heating facility operated port the war effort in the 1940s, the major operations
by Dale Siegford of the Sapphire Gallery in Philips- in this early period ended shortly after World War II.
burg, Montana. Treatments were claimed to have It was not until the introduction of advanced heat
been at 1400°C, but other details of this commercial treatment procedures that these deposits became
treatment were not disclosed. Additional heat treat- economically viable again.
ment experiments were carried out at GIA’s experi- The following sections will document each of the
mental heating facility using a Thermal Technology three secondary sapphire deposits in more detail. His-
Model 1000-3560-FP24 muffle tube furnace, similar torical information in this section comes primarily
to the system used in Emmett and Douthit (1993). from the works of Berg (2007, 2014) and Berg and
Landry (2018), which are currently the most compre-
hensive historical works on sapphire mining in these
Gemology. Standard gemological instruments were
deposits. These three publications contain original
used to measure refractive index, birefringence, spe-
historical research based on archived historical re-
cific gravity, and pleochroism. Microscopic observa-
sources and interviews with primary historical fig-
tion was performed using either a Nikon Eclipse
ures. When possible, the primary published source of
LV100 compound microscope or Nikon SMZ1500
information is cited; however, when no citation is
binocular microscope outfitted with a Nikon DS-Ri2
provided, it should be understood that the primary
camera for recording images. Differential interfer-
reference is the original historical research in Berg
ence contrast (DIC) images were taken on the LV100.
(2007), Berg (2014), or Berg and Landry (2018) for Dry
Also used were a Nikon SMZ25 stereomicroscope
Cottonwood Creek, Rock Creek, and Missouri River,
equipped with dual fiber-optic illuminators, darkfield
respectively. Additional sources of historical informa-
and brightfield illumination, polarizing filters, and a
tion are Kunz (1890), Pratt (1906), Clabaugh (1952),
Nikon DS-Ri2 camera.
Emmett and Douthit (1993), Garland (2002), Barron
and Boyd (2015), Hsu et al. (2017), and Kane (2020).
HISTORY
Like much of the American West, the history of Missouri River. The first influx of European settlers
Montana is told in the story of gold. The original into the Montana territory began in 1862 with the
prospectors who discovered the state’s sapphire de- discovery of rich gold placers in Grasshopper Creek
posits generally had little in the way of mineralogi- at the town of Bannack. With rumors of gold circu-
cal, much less gemological, knowledge. While these lating, prospectors would seek their fortune up and
prospectors did stumble upon one of the most impor- down nearly every creek and river in the territory, no
tant gem deposits in the North American continent, matter how large or small. Gold in abundance would

6 MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023
Figure 6. Watch bearings fashioned out of Montana sapphires in various stages of completion. The leftmost pile
contains rough sapphires ready for fabrication, and the pile second from the left contains sapphires ground to
wafers of standard thickness to be drilled out to produce bearings. The four piles to the right show finished natural
sapphire watch bearings. Photo by Robert Weldon; courtesy of the Sapphire Gallery.

be found the next year at Virginia City in Alder The earliest workings on these bars involved hor-
Gulch and, importantly, at Last Chance Gulch near izontal drifts cut through the gravel beds, which al-
Helena. lowed miners to avoid the labor of removing the
The first report of sapphire came in 1865, when several meters of overburden that typically overlie
gold prospector Ed Collins found several while work- the sapphire- and gold-bearing gravels (figure 7). Hy-
ing a gravel deposit along the Missouri River near He- draulic mining, which had been used at other gold
lena at what came to be known as Eldorado Bar. The camps, would have been more efficient. Yet here, as
stones were found not in the river itself, but in gravel in much of the American West, this advanced min-
beds sitting atop bedrock just off the banks of the river, ing technique was limited by access to water. De-
sometimes several hundred meters above water level. spite their proximity to the Missouri River, the
Over the years, similar deposits were identified along gravel beds were at a higher elevation than the river
a roughly 21 km stretch of the Missouri River, with level, so the water would have to be pumped up to
eight major gravel bars eventually worked to some ex- the gravel beds or brought in from elsewhere. The
tent. Collins was apparently unsure what to make of first major mining efforts commenced in 1868, when
the gemmy stones he found, so he sent them east to the Eldorado Ditch Company completed a ditch
Tiffany & Co. in New York and other parties overseas from Trout Creek, at a higher elevation than the
(“Eldorado Bar sapphires,” 1867). The gems were iden- Missouri River, to bring water to the upper terraces
tified as sapphire, although the discovery generated lit- of Eldorado Bar (figure 8). This allowed the gravel
tle attention (Browne, 1868, p. 50). These gravel bars beds to be worked by hydraulic mining, in which
sitting alongside the Missouri River were also rich in high-pressure jets of water were employed to wash
gold, so mining activities continued regardless. sapphire-bearing gravels into sluice boxes with riffles

MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023 7
 Figure 7. Mining opera-
tion at Dana’s Bar. The
piles of rock are from the
trenches below. Photo
by Edgar H. Train be-
tween 1870 and 1879,
Lot 026 B3F15.03 Dana’s
Bar, Montana Historical
Society Photograph
Archives.

set up at intervals to capture sapphire and gold (figure Company as a subsidiary of the Sapphire and Ruby
9). This led to the consolidation of smaller claims on Company of Montana, Ltd. to mine Spokane Bar.
Eldorado Bar, creating larger and more efficient oper- Given the worldwide interest being generated
ations. The process of consolidation continued until about the Missouri River deposit, other mining en-
1872, when original discoverer Ed Collins joined terprises came along as well. Notably, Henry Math-
with seven partners to control a 641 hectare claim eson acquired claims on Emerald Bar and brought in
encompassing most of the upper terrace of Eldorado water with the use of a steam pump. Matheson soon
Bar. sold out to a group of investors out of Helena forming
During a lull in mining in the 1880s, brothers the Montana Gold and Gem Company.
Frank D. and Thomas H. Spratt began buying up and Despite the budding success of Missouri River
staking new mining claims, especially on Eldorado sapphires, which were featured at the 1893 Chicago
Bar. In 1887, with the financial backing of a third World’s Fair, the economic Panic of 1893 crippled the
brother, Augustus N. Spratt, they incorporated the market for them. This was compounded, in no small
Trout Creek Mining Company and began actively part, by the discovery in 1895 of sapphires at Yogo
marketing the sapphires on a national and interna- Gulch whose natural cornflower blue color was pre-
tional scale. In 1890, an English syndicate was ferred in the gem market. Shortly thereafter, the Sap-
formed to market Missouri River sapphires. The syn- phire and Ruby Company of Montana, Ltd. went
dicate ended up purchasing properties on Ruby Bar bankrupt and its properties reverted to Augustus N.
owned by Frank and Edgar Langdon, as well as prop- Spratt, who reformed his holdings as the Eldorado
erties belonging to the Spratt brothers at French Bar Gold and Gem Company.
and their Eldorado Bar claims. The English syndicate In 1911, the lowest terraces of Eldorado and
now controlled most of the profitable workings and Spokane Bars were submerged with the completion
was incorporated as the Sapphire and Ruby Company of the Hauser Dam and the creation of Hauser Lake.
of Montana, Ltd. with initial public stock offerings While it may seem a pity to flood precious gem-
of 400,000 shares at £1 (“Montana’s precious stones,” bearing gravels, the advantages of the newly created
1891; “Sapphire and Ruby Company prospectus,” lake did not go unnoticed. Most notably, in 1938, the
1891). Sapphire and gold production commenced Perry-Schroeder Mining Company was formed and
swiftly, with drift mining utilized initially and hy- purchased a Yuba bucket-line dredge. The Perry-
draulic mining techniques applied later. The Spratt Schroeder Mining Company was one of the few op-
brothers also incorporated the Spokane Sapphire erations allowed to continue mining gold throughout

8 MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023
 Figure 8. The flume tak-
ing water from Trout
Creek to the sapphire
and gold mining opera-
tions at the Missouri
River. In this view, the
flume is crossing Trout
Creek itself, with the
water taken from further
upstream. Photo by W.H.
Jackson during the 1869–
1871 Hayden Survey. Lot
032 B5F09.03a Trout
Creek Flume, Montana
Historical Society Photo-
graph Archives.

World War II despite War Production Board Order L- strategic materials. The company was only allowed
208, which shut down gold mining operations across to continue mining gold because it was also produc-
the country in order to focus extraction efforts on ing industrial sapphire, which was in short supply

Figure 9. Hydraulic
mining at French Bar
along the Missouri
River between 1870
and 1879. PAc 97-32.9,
Montana, Montana His-
torical Society Photo-
graph Archives.

MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023 9
given that the synthetic sapphire manufacturers Rock Creek. The discovery of sapphire in the Rock
were all located in Europe at the time. The Yuba Creek district is often attributed to one Emil Meyer,
dredge continued mining for gold and sapphires until a German immigrant who later partnered with Swiss
1947, marking the end of the historic mining activi- native L.J. Moffat to work the deposit (“Sapphire
ties on the Missouri River. In the ensuing years, sev- mined on Rock Creek after 1892 discovery there,”
eral smaller operations continued to mine for 2019). However, some reports attribute the discovery
sapphire along the gravel bars of the Missouri River. to M.H. Bryan Sr. of Philipsburg, who claimed to
Their efforts have been fueled by the development of have been prospecting for gold on the West Fork of
high-temperature heat treatment processes to im- Rock Creek when he noticed several bright stones at
prove the sapphire colors, making them more desir- the bottom of each pan of dirt. He collected the
able in the global gem and jewelry market. Today, the stones and simply gave the choice gems to his grand-
mining is nearly all concentrated on Eldorado Bar, children to play with. Regardless of who truly dis-
where small-scale miners work some virgin gravels covered the deposit, or whether they found it more
but mostly the tailings left over from the Perry- or less contemporaneously, the recorded date of dis-
Schroeder Mining Company in the 1940s. As that covery is 1892. An article from the September 3,
dredging operation was also tailored to recover gold, 1892 edition of The Anaconda Standard (“New sap-
the company’s sapphire recovery was fairly low com- phire fields,” 1892) indicates that local prospectors
pared to modern operations, and many sapphires were already aware of the rich sapphire diggings and
were left behind. were actively preparing to start working the deposit.
Between 1988 and 1993, Sam Speerstra mined six While this deposit is colloquially referred to as
million carats of sapphire from Eldorado Bar (Berg “Rock Creek,” sapphires have largely been mined
and Landry, 2018). Adjacent to American Gem Cor- from several tributary gulches that cut into the moun-
poration’s main lease, Gem Resources Management tain to the north of the West Fork of Rock Creek. The
Corporation mined Eldorado Bar from 1988 to 1992. two most important, especially in the early days of
While American Gem Corporation’s primary mining mining, were Anaconda Gulch and Sapphire Gulch,
activities were focused at Rock Creek (see “Rock but mining has taken place at numerous gulches as
Creek” history section below), their press release well as in placers along the banks of the West Fork of
from September 1994 indicated that consulting geol- Rock Creek, in an area called the Meadow. Also of par-
ogists and engineers had completed bulk sample test- ticular note for this deposit is the relative lack of gold,
ing on 260 of the nearly 1,200 acres (105 of 485 ha) at especially compared to the Missouri River deposit.
Eldorado Bar controlled by the company (O’- What Rock Creek lacks in gold it more than makes
Donoghue, 1995). These tests were conducted to de- up for with an overwhelming abundance of sapphire.
termine occurrence and grades (carats per cubic yard) Rock Creek has produced more than any other deposit
and indicated considerable sapphire reserves at Eldo- in Montana, including Yogo Gulch, likely eclipsing all
rado Bar. American Gem Corporation also conducted of them combined by nearly a factor of 10. All told, at
test sampling across the river channel at Dana’s Bar. least 70 metric tons of sapphire have been mined from
The company never moved beyond the testing the Rock Creek area (Berg, 2014). However, this esti-
phases, electing instead to concentrate its efforts at mate does not include the mid-1990s yield of nearly 4
Rock Creek and Dry Cottonwood Creek. million carats by the American Gem Corporation or
Present-day mining includes operations of the significant recent production from Potentate Mining
Spokane Bar mine and Gold Fever Rock Shop (cur- (Kane, 2020). In fact, Rock Creek is the only secondary
rently the largest operation), Neal Hurni with Lewis deposit in Montana to be worked only for its sapphire
and Clark Sapphires, Blaze Wharton with Blaze-N- content (gold is recovered there as well but has never
Gems, Bruce Scharf and Blue Jewel mine, Don John- been the primary focus of mining).
son with Obsessed Over Gems, and Tim Beard along The early years saw the deposits worked mostly
Gruell’s Bar. Another notable operation in the recent by smaller operations. Figures of note from the first
past is that of Mac Mader, who constructed a suction few years include 400,000 carats of sapphires pro-
dredge in use from 1983 to 1986 to recover gold and duced by William Knuth of Helena and William Mof-
sapphire from Hauser Lake. In 2012, Mader used a fitt of Philipsburg in 1900, of which 25,000 carats
handheld suction dredge operated from a small barge- were fit for cutting. L.J. Moffat and Emil Meyer were
mounted washing plant on Hauser Lake near French two of the important early miners at Rock Creek.
Bar. Moffat’s Swiss heritage may have helped position

10 MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023
 Figure 10. A 1908 map
of the mining claims of
the American Gem Syn-
dicate at Rock Creek.
The various placers are
labeled along with
water-carrying ditches
and streams. Map A-
457, Plat of the claims
of the American Gem
Mining Syndicate, Gran-
ite County, 1908, Mon-
tana Historical Society
Photograph Archives.

these stones in the industrial sapphire market, as in upper Sapphire Gulch, the Star Placer in lower Sap-
most of the production was used by fine watchmak- phire Gulch, and the Anaconda Placer in Anaconda
ers in Switzerland. Their importance in the early Gulch. The company would continue acquiring land
days of mining was recognized with the naming of and patenting placer claims through 1915 (figure 10).
sapphire-bearing gulches. Meyer Gulch was a produc- As with the Missouri River deposit, water has al-
tive tributary of Sapphire Gulch, and Moffat has two ways been the limiting factor in mining operations
gulches named in his honor: the Little Moffat and the at Rock Creek. Water flowing through the gulches
Big Moffat. was insufficient to sustain major mining activity.
The most important era of overall production at With the vast capital it raised, the American Gem
the Rock Creek deposit began in 1901, when articles Mining Syndicate solved this problem at Rock Creek
of incorporation were filed for the American Gem by constructing a series of ditches and flumes. The
Mining Syndicate, based in St. Louis. The company first was Cralle’s Ditch, likely in 1904, which
was incorporated by David Jankower of London and brought water from Stony Lake to the McLure placer
New York; Paul A. Fusz, Moses Rumsey, and Charles at the upper end of Sapphire Gulch, although the
D. McLure of St. Louis; L.S. McLure of San Diego; and water was likely used in Anaconda Gulch as well. In
William Knuth of Helena. The American Gem Min- 1912, the Upper Sapphire Ditch was completed, tak-
ing Syndicate was capitalized with $300,000, with ing water from the North Fork of the West Fork of
Jankower holding 299,996 shares and one share each Rock Creek to the head of North Fork Coal Gulch
distributed to Fusz, Rumsey, A.B. Ewing, and C. and the head of Anaconda Gulch. An additional un-
McLure (“Sapphire mined on Rock Creek after 1892 named ditch was also constructed, bringing water
discovery there,” 2019). In 1902, the American Gem from the West Fork of Rock Creek upstream 1,200 m
Mining Syndicate patented claims at the Ruby Placer to the west to Sapphire Gulch.

MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023 11
 The water from these ditches and flumes sup- In the 1970s, the Thai gem trade developed and per-
ported mining activities during the two most produc- fected high-temperature heat treatment procedures
tive years of 1906 and 1907, which yielded 6,482.3 to dramatically improve the color and clarity of pale
kg and 6,639.9 kg of sapphire, respectively. Mining and milky sapphires, creating deep, rich blue sap-
was carried out by either ground sluicing or hydraulic phires. The original material used was the milky
mining. During these years, mining activities were geuda sapphire from Sri Lanka, but it was soon real-
intense and the mine often operated 24 hours a day. ized that the same heat treatment method could be
From 1903 to 1928, while the American Gem Mining applied to secondary Montana sapphire with out-
Syndicate was working the deposit, 46.4 metric tons standing results. In the early 1980s, a Mr. McCarthy
of sapphire were produced (Berg, 2014). of Helena was reportedly the first person in Montana
Meanwhile, times were changing with increasing to use this high-temperature heat treatment to im-
competition from synthetic ruby and sapphire, espe- prove the color of a Montana sapphire; however, the
cially starting in the 1920s. Although Auguste process was perfected in a technical sense in the
Verneuil grew his first flame-fusion rubies in 1883, 1990s and scrupulously detailed in the work of Em-
the material did not become widely available until mett and Douthit (1993).
the 1910s, when he expanded his enterprise. As more With the ability to transform color through heat
and more flame-fusion ruby and sapphire reached the treatment, the economics of mining and marketing
market, and as the quality of the laboratory-grown Montana sapphire changed for the better. In 1994, a
material improved, Montana sapphire could not new player entered the scene when American Gem
compete. Facing a dire financial situation, the Amer- Corporation purchased much of the gem-producing
ican Gem Mining Syndicate sold most of its claims areas of the Rock Creek district, including the Dann
to Charles Carpp Jr. and J. Walt Kaiser. The two Placer and Anaconda Bench as well as the holdings of
worked the deposit until 1943 and produced about the Gem Mountain Sapphire Corporation. All told,
2.8 metric tons of sapphire, largely through hydraulic American Gem Corporation reported production of
mining (Clabaugh, 1952). They apparently intended more than four million carats of gem-quality rough
to sell the sapphires as gemstones but were unsuc- sapphire, of which more than two million heat-
cessful due to the typically pale colors of the natural treated faceted sapphires to date have been sold into
rough. Therefore, most of the production still went the international market (Kane, 2020). While Ameri-
to the industrial sapphire market. can Gem Corporation, a public company listed on the
After Carpp and Kaiser stopped mining in 1943, Toronto Stock Exchange, carried out a high-profile
the Rock Creek properties passed through several marketing campaign, their efforts never materialized
hands. Carpp and Kaiser sold to Sally and Bill Eaton, into a steady supply of gems, and the company reor-
who operated a fee dig for tourists. The Eatons in ganized as an e-commerce entity in 1999. At this
1966 sold the Meadow property along the West Fork point, Chris Cooney purchased their claims at the
of Rock Creek to Marc Bielenberg and the gulch Anaconda and North Fork Coal gulches, where he
properties to Wilfred Chaussee, who formed the now operates a mining and fee-dig tourism operation
Chaussee Sapphire Corporation and continued run- called Gem Mountain. At the same time, logging
ning the tourist operation. This operation was later company RY Timber acquired the rest of the patented
passed to Wilfred’s daughter and son-in-law, Yvette claims in the Rock Creek deposit.
and Kenneth Clevish. The Clevishes sold in 1980 for The next chapter in the Rock Creek story is being
$1 million to Ted Smith, who renamed the operation written by Potentate Mining. In 2011, Potentate pur-
the Gem Mountain Sapphire Corporation. In the chased Eureka Gulch on the north side of the moun-
meantime, Marc Bielenberg sold the Meadow prop- tain and became the largest landholder on Gem
erty in 1978 to Skalkaho Grazing, Inc., which con- Mountain in 2014 after purchasing most of the south
structed a floating wash plant in 1982 near the side of the mountain from RY Timber. Over the
confluence of the Anaconda and Coal gulches with years, Potentate has actively developed its claims and
the West Fork of Rock Creek. now operates a modern operation utilizing state-of-
Still, mining activities during this period never the-art mining technologies to produce sapphire in
approached the production by the American Gem an environmentally friendly manner while making
Mining Syndicate in the early twentieth century. efficient use of limited water resources. Among the
However, another technological advance on the hori- advances implemented are a new gravel-processing
zon would change the fate of Rock Creek sapphire. facility with a separate gold circuit to maximize gold

12 MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023
recovery, as well as a water clarifier to ensure the re- wood Creek (Kane, 2020). However, the Dry Cotton-
turn of clean water to the pristine mountain streams. wood Creek sapphires did not react as well to heat
In 2020, Potentate finalized the purchase of the treatment as those from Rock Creek (J.L. Emmett,
Meadow property and now controls the bulk of the pers. comm., 2022), and American Gem Corporation
sapphire-bearing areas of the Rock Creek deposit. Po- abandoned operations there in 1996 to focus on Rock
tentate is pursuing an aggressive marketing cam- Creek. In the same year, Gem River Corporation
paign to put Montana sapphire in a prominent place began mining sapphires in the Grand Pre Flats, but
in the world market. mining stopped in 1998 and the area was reclaimed
for nature. No serious mining efforts have taken
Dry Cottonwood Creek. Sapphires were first found place at Dry Cottonwood Creek since then. How-
by gold prospectors in the South Fork of Dry Cotton- ever, the Dry Cottonwood Creek sapphires do react
wood Creek near the city of Butte in 1889 (Kunz, well to heat treatment in an oxidizing environment
1894). Mining commenced shortly thereafter by the to produce variegated yellow and orange colors (R.E.
Northwest Sapphire Company of Butte, which em- Kane, pers. comm., 2022). With changing market be-
ployed hydraulic mining as well as dredging opera- havior and growing interest in such fancy-color sap-
tions at the South Fork of Dry Cottonwood Creek phires, Dry Cottonwood Creek has the potential to
(Struthers and Fisher, 1903). One of the more signif- become a viable mining property.
icant mining enterprises involved the Variegated Sap-
phire Company, which in 1905 installed a
bucket-line dredge at the Grand Pre Flats, an open CHARACTERISTICS OF THE DEPOSITS
area along the South Fork of Dry Cottonwood Creek. Geology. The rugged terrain and landscapes of west-
The dredge worked an area 24–30 m wide, removing ern Montana are the product of both ancient and
12 m of gravel per day and recovering both gold and modern geological forces at work. The very oldest
sapphire. In 1907, ownership of the enterprise trans- rocks encountered in the sapphire-producing areas be-
ferred to Mr. West Dodd of Des Moines, but this op- long to the Proterozoic Belt supergroup, which forms
eration ended by 1911, only to be briefly revived in the bedrock underlying much of the Missouri River
1914 by Nat Simon. However, Simon abandoned it and Rock Creek deposits and also represents much of
after one season of low gold yield and sapphires that the Belt Mountains seen to the east of the Missouri
were pale and not marketable. In 1908, the Consoli- River deposits. The Belt supergroup includes
dated Gold and Sapphire Mining Company of Butte metasedimentary rocks deposited between 1470 to
installed another dredging operation about 1.5 km 1370 Ma in a sedimentary basin formed at the margin
from the Grand Pre Flats, downstream from the of the North American Craton and some other un-
South Fork of Dry Cottonwood Creek. This effort known ancient landmass. The Belt supergroup, an ac-
was short-lived as well due to the low concentration cumulation of sediments an astonishing 15 km thick
of gold and the difficulty of selling the sapphires in at its maximum extent, provides many clues to the
the gem and jewelry industry. ancient movement of the earth’s protocontinents. But
Interest in the Dry Cottonwood Creek deposit re- these primeval geological formations serve as nothing
vived during World War II due to the shortage of in- more than the backdrop of the gemological story. At
dustrial laboratory-grown sapphire coming from the Missouri River, these ancient formations are sim-
Europe, but there was no production in those years. ply the bedrock upon which the sapphire-bearing
Significant mining did not resume until the 1990s, gravels were deposited. The Proterozoic Belt super-
when American Gem Corporation acquired claims group formation here is in the form of an argillite, a
here and at Rock Creek and Missouri River. Ameri- low-grade metamorphosed version of a sedimentary
can Gem Corporation purchased land in 1993 from shale or claystone. At Rock Creek, the Belt super-
Scott Wurster, who had in turn acquired it from Marc group rocks include metasedimentary rocks that un-
Bielenberg, who had previously owned claims in the derlie Cenozoic volcanic formations, which are
Rock Creek district. The advent of high-temperature apparently related to the gem deposits.
heat treatment made Montana sapphire much more The story of sapphire formation begins much later
attractive to the gem and jewelry industry with the in the earth’s history, during the Cenozoic era. While
rich vivid blue and fancy-color gems that could be the sapphires occur in secondary deposits, field ob-
produced. In 1995, American Gem Corporation servations indicate in each case that the original host
mined 400,000 carats of sapphire from Dry Cotton- rock is most likely related to Cenozoic volcanic for-

MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023 13
 MONTANA
N

S m it
UPPER
HOLTER

h R iv
LAKE
American Bar

er
Hauser Dam
Figure 11. Map of sap-
H AU
phire-bearing gravel
Eldorado Bar
bars along the banks of
SER

the Missouri River. El-

LA
K
E dorado Bar is the main
active mining site, but
Dana’s Bar York
there is also small-scale
LAKE
commercial mining at
HELENA Spokane Bar and Gru-
McCune Bar
ell’s Bar. Modified by
Metropolitan Bar
Robert E. Kane from
H AUSER
LAKE
Canyon Berg (2015), with per-
Gruell’s Bar Ferry
Dam mission; courtesy of the
Montana Bureau of
Spokane Bar Mines and Geology.
French Bar

H E L E N A VA L L E Y
R E G U L AT I N G R E SE RVO I R
Emerald Bar
C A N YON
FER RY
LAKE

Sapphire-bearing gravel 0 1.5 3 Miles

Road
Dam 0 1.5 3 Kilometers

mations erupted around 50 Ma. From about 160 to 11). The sapphire-bearing gravel beds are composed
50 Ma, the entire western seaboard of North America predominantly of locally sourced material. This
was a convergent margin, with the ancient Farallon means there are limited options to consider for poten-
plate being subducted beneath the North American tial bedrock sources of the sapphire. Further compli-
continent. Subduction of the Farallon plate led to sig- cating matters is that the Missouri River here has
nificant volcanism throughout the western United been flooded by the construction of the Hauser Dam
States, Canada, and Mexico. The volcanic formations and subsequent creation of Hauser Lake, limiting po-
in the region of the sapphire deposits are related to tential exposures of sapphire-bearing lithologies. The
the late stages of this period, when a flat-lying Faral- Proterozoic Belt supergroup argilites immediately un-
lon slab started to roll back and eventually detach as derlying the gravel beds are not suitable lithologies
it continued its descent into the earth’s mantle. As for corundum formation, and the same is true for the
the sapphire-bearing volcanic formations were Paleozoic and Proterozoic metasedimentary rocks of
weathered and broken down, the sapphires were lib- the nearby Big Belt Mountains and the Paleozoic sed-
erated and entrained in the rich gem gravels being imentary rocks in the Spokane Hills.
mined to this day. The following sections will briefly What little evidence we have comes from one
describe the relevant formations in each deposit. small sapphire-bearing sill found at French Bar.1

Missouri River Geology. There is very little geological

1
context remaining to explain the deposits at Missouri Note that while sapphire-bearing sills have been reported at both
French Bar and Ruby Bar, Berg and Landry (2018) pointed out that
River, where sapphires are recovered from gravel beds historical analysis proved these two bars are the same, and thus it is
sitting atop bedrock terraces above water level (figure likely that only one sapphire-bearing sill has been discovered.

14 MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023
Based on chemical analysis and petrographic obser- tirely. Additionally, many sapphires from the Mis-
vations of thin sections of the volcanic rock, Berg and souri River are found with thin encrustations of
Palke (2016) described it as a basaltic trachyandesite spinel adhering to their surface. This spinel is in-
composed of biotite and augite phenocrysts in a ma- ferred to result from a reaction between the corun-
trix of plagioclase microlites. Rare corundum-bearing dum and magnesium and iron sourced from mafic
xenoliths have been found composed of calcic plagio- magmas that transported the sapphires to the sur-
clase, augite, garnet, margarite, and spinel (Berg and face. This spinel rim is not found on sapphires from
Dahy, 2002). Rock Creek or Dry Cottonwood Creek and is not
Importantly, Berg and Palke (2016) used LA-ICP- even found on all Missouri River sapphires.
MS to analyze the chemistry of sapphires found in
situ in this (sub)volcanic formation, and the results Rock Creek Geology. In the Rock Creek area, sap-
demonstrated that their trace element profiles are phires are found in gravel beds in gulches cutting into
consistent with the gem sapphires found in the Gem Mountain (figure 12). The dominant formations
gravel beds. This indicates that these volcanic forma- in the area are rhyolitic lava flows, dikes, and tuffa-
tions are viable sources of the Missouri River sap- ceous and volcaniclastic formations that overlie
phires. While this single volcanic formation is surely metasedimentary rocks of the Proterozoic Belt super-
too small and has too low a sapphire concentration group. Given that sapphires are found in the gulches
to have supplied gems for the entire Missouri River cutting up into the mountain, their source must have
deposit, it suggests a likely geological origin of the been local. The volcanic formations here are rela-
sapphires, with the implication that other similar tively silica-rich and rhyolitic in composition. Rhy-
sapphire-bearing (sub)volcanic formations must have olite lava flows cap the ridges separating the
been present but were weathered away almost en- sapphire-rich gulches, while the more readily eroded

N N
O
R
T
H
Coal
SA
Gulch PP
HI
Pit RE Ewing
Placer
Dismal
F O R K

Swamp C
H
McLure L
U
Anaconda Placer G
Wild Cat Gulch A
Bench O
R
A NA

R
ulch
U
A
kG Meyer Gulch
Min lch
C O N

e Gu
Pin Queen Gulch
ck
DA

Bla
H
ch

G U L C

G U L C H
Figure 12. Map of sap-
Gul

C OA L
L
G

GU

phire-bearing gulches
U

May

Y
L
C

and placer deposits at

L C

AT
H

GU

Bench
H

Lower Dann Rock Creek. Modified

M O F F

Placer
LC

Placer Bench from Berg (2014).

H

Upper Dann Anaconda Gulch Placer

L E

Placer
L I T T

The Meadow
Sapphire
MONTANA Gulch
S m it
h R iv
er

Big Moffat
Gulch

Sapphire-bearing gulches
and placers
Qal (quaternary alluvium) 0 1 km

MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023 15
tuffs and volcaniclastic units are less well preserved. Origin of the Secondary Montana Sapphires. Given
No other volcanic formations have been identified in the relative lack of hard rock geological evidence in
the immediate area, although it has been postulated these secondary deposits, one must search elsewhere
that there may be hidden plugs of lamprophyres or for clues about the ultimate origin of these sapphires.
other mafic volcanics in the area that have supplied In fact, while the original sapphire-bearing host rocks
the sapphires (Barron and Boyd, 2015). Notably, Berg have mostly weathered away, remnants of the an-
(2014) did identify a basaltic dike exposed along a cient sapphire-producing geological environments
road cut 4 km east of Sapphire Gulch. No sapphires have been preserved within the sapphires as inclu-
have been found in this dike, but to the authors’ sions. Besides hexagonal bands of silk and rutile par-
knowledge it has never been studied thoroughly. ticles, among the most commonly encountered
Still, this basaltic dike is likely too small and too far inclusions are fluid inclusions of an unusual nature.
from the deposit to be a viable source of the sap- These often have the appearance of typical two-phase
phires. While no sapphires have been found in situ fluid inclusions seen in other gems. However, if one
in any volcanic rocks here, Berg (2014) did identify were to cut into one of these fluid inclusions, the
some with what was described as a “rhyolitic” en- fluid material would be frozen in place and not leak
crustation. This encrustation was observed on an- out of the sapphire as expected. These fluid inclu-
other sapphire by Belley (2022), with scanning sions represent blebs of magma and indicate that the
electron microscopy (SEM) images confirming that fluid present during sapphire growth was a molten
the adhering crust was composed of quartz, plagio- silicate rock. Upon transport to the surface in vol-
clase, and potassium feldspar. Importantly, Rock canic events, the sapphires must have cooled quickly
Creek sapphires are never found with the spinel en- enough to quench these melts into glasses.
crustation found on some Missouri River sapphires, The compositions of the melt inclusions are
which is likely formed by interaction with magne- broadly similar across the three deposits, suggesting
sium and iron in the mafic transporting magma at similar conditions of geological formation. The melts
Missouri River. This observation seems to suggest a coexisting during sapphire growth were poor in MgO
mafic source is less likely at Rock Creek. With the and FeO, moderately enriched in Na2O, K2O, and
enormous volume of sapphire produced at Rock CaO, and (surprisingly) relatively rich in SiO2. The
Creek, the original source rock must have been ex- melts can be classified as dacites or rhyolites accord-
tensive, decreasing the likelihood of a hidden source ing to the total alkali silica (TAS) diagram of Le Bas
that was weathered away. Given the prevalence of et al. (1986). Of particular note is that these melts are
rhyolitic volcanics in the area, the evidence suggests all quartz-normative, meaning that if cooled in an
that rhyolitic volcanism is a likely source of the sap- isolated system, they should have precipitated
phires in this deposit. quartz. This seems counterintuitive at first, as it is
well known that corundum and quartz cannot coex-
Dry Cottonwood Creek Geology. Sapphires at Dry ist under ordinary geological conditions. However,
Cottonwood Creek are found in gravels and second- it is possible for corundum to coexist with a quartz-
ary deposits. Bedrock is composed of volcanic for- normative melt in the case of an incongruent (or
mations from the Eocene-age Lowland Creek peritectic) melting reaction. The presence of quartz-
Volcanics. These overlie Cretaceous granitic rocks normative melts then suggests a genetic model in
of the Boulder Batholith, which contacts the Low- which the secondary Montana sapphires grew
land Creek Volcanics near the western boundary of through partial melting of some aluminum-rich rock
the Grand Pre Flats on the South Fork of Dry Cot- (e.g., an anorthosite or some similar plagioclase-rich
tonwood Creek. The Lowland Creek Volcanics are igneous rock). In this model, the magmas that trans-
predominantly composed of welded tuffs and lava ported the sapphires to the surface in volcanic
flows of dacitic composition. No other volcanics or events might have also been responsible for sapphire
other geological formations exist in the immediate formation by initiating partial melting of the pro-
area. While no thorough field studies have been per- posed aluminum-rich protoliths. In this scenario,
formed to identify the source of the sapphires, the the sapphires could still be foreign to the transport-
Lowland Creek Volcanics seem to be a likely origi- ing magmas but would not be accidental xenocrysts.
nal bedrock source, given the likely volcanic origin In fact, based on observations of rhyolitic secondary
of all Montana sapphire and the paucity of other pos- melt inclusions, Palke et al. (2017) suggested the
sible sources. possibility that some sapphires at Rock Creek may

16 MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023
not be xenocrysts at all if the partial melting event
that produced the rhyolites seen at the surface also
produced the sapphires entrained in this magma. Of
course, proving or disproving this hypothesis would
be challenging. At Missouri River, the basaltic tra-
chyandesite magma that likely transported some of
the sapphires is significantly different from the melt
inclusions seen in those stones, confirming their
identity as xenocrysts. The Lowland Creek Volcanics
at Dry Cottonwood Creek have not been thoroughly
studied, and their relation to the melt inclusions seen
in those sapphires is less clear.
The sapphires’ origin from some calcium- and
aluminum-rich protolith such as an anorthosite is
also supported by the prevalence of calcium- and
aluminum-rich mineral inclusions such as anor-
thite-rich feldspar and (clino)zoisite. Other genetic
models have been proposed, mostly based on the use
of trace element chemistry wherein the low gallium
and other trace element patterns were thought to
suggest a metasomatic or metamorphic origin for
the sapphires (Garland, 1992; Zwaan et al., 2015).
Such an origin, however, seems to be contradicted
by the pervasiveness of silicate melt inclusions,
which indicate that the sapphires grew in the pres-
ence of a silicate magma.
Figure 13. Faceted unheated Montana sapphires set in
18K gold. Photo by Orasa Weldon; courtesy of Diana
Widman.
GEMOLOGICAL CHARACTERISTICS OF
SECONDARY MONTANA SAPPHIRE
Physical Properties. Sapphires from Montana’s sec- observed, especially for heated stones with a yellow
ondary deposits yield some unique characteristics or orange core and a blue, green, pink, or violet rim
that make them distinct from other sapphires around
the world. While basic physical properties such as re-
fractive index and specific gravity offer no significant Figure 14. This “Twilight” ring showcases some of the
differences, the inclusion suite and trace element common range of colors seen in Montana sapphire.
chemistry can help to distinguish secondary Mon- The sapphires are unheated, with 3.80 total carat
tana sapphire. However, it should be noted that the weight. Photo courtesy of Earth’s Treasury.
characteristics of individual faceted sapphires from
Montana’s secondary deposits overlap significantly,
and these gems cannot currently be reliably sepa-
rated from each other.
In general, sapphires from Montana’s secondary
sources naturally occur in a wide range of hues, in-
cluding blue, green, yellow, orange, pink, and purple
(figures 13–19). Most of these colors are somewhat
pale, especially when unheated (figures 13 and 14).
Vibrant, saturated colors do occasionally occur nat-
urally, but they are more common in heated stones
(figures 15–17). Montana sapphire has become in-
creasingly popular in nonstandard cutting styles (fig-
ures 18–19). Parti-colored sapphires are frequently

MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023 17
 Figure 15. The “Conchita
Sapphire Butterfly,” a
collaboration between
gem dealer/gemologist
Robert Kane and jew-
elry designer Paula
Crevoshay, highlights
the variety and color
range of heat-treated
secondary sapphires
from Montana. The 18K
yellow gold butterfly is
set with 331 round bril-
liant sapphires and two
cabochon-cut sapphires,
totaling 27.97 carats.
The butterfly was do-
nated by Kane and
Crevoshay to the Smith-
sonian Institution’s Na-
tional Gem Collection
in 2007. Photo by
Harold and Erica Van
Pelt; courtesy of Robert
E. Kane.

(figure 20). Some sapphires exhibit a weak to mod- Robert Kane (Fine Gems International, Helena) un-
erate color change when observed in incandescent covered numerous examples of rubies from color-
to daylight illumination (figure 21). Colors generally graded parcels of pink sapphire. While rare, these
change from shades of pink or orange (incandescent) Montana rubies do exist and may be encountered in
to green or violet (daylight). Rubies are also some- the market. Other examples of Montana ruby are
times found (figure 22). Palke and Hapeman (2019) known, including one particularly large faceted gem
reported only 29 g of rough ruby crystals recovered weighing 1.70 ct (Malaquias, 2018).
from more than 400 kg of mine production at Poten- Sapphire rough from these deposits generally occurs
tate’s operation at Rock Creek. While truly red gem as small crystals measuring a few millimeters in diam-
corundum from Montana’s secondary sapphire de- eter, but larger stones have been found. Notably, the
posits are rare, examination of the collection of unheated blue sapphire known as the “Big Sky” sap-

Figure 16. A suite of heated Montana sapphires, 0.76–1.04 ct, displaying a range of colors. Courtesy of Rachel Hill,
Americut Gems.

18 MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023
 Figure 17. Various colors
of heated Missouri River
sapphire, ranging from
0.16 to 1.65 ct. Photo by
Tino Hammid; courtesy
of Fine Gems Interna-
tional and American
Sapphire Company, ©
Robert E. Kane.

phire, which sold in 2015 for $394,000, weighed 12.54 also recovered a large crystal from Rock Creek in 2018
ct (Kane, 2020). Rough crystals as large as 37.78 ct have weighing 64.14 ct, known as the “Ponderosa” sapphire,
been reported from Eldorado Bar (Berg and Landry, as well as a 67.15 ct yellow sapphire recovered in 2021
2018), while the Rock Creek deposit has produced that cut a 12.18 ct stone called the Yellowstone sap-
stones as large as the faceted 12.62 ct heated stone re- phire. This is currently part of the Somewhere in the
covered by Potentate Mining known as the “Montana Rainbow collection (again, see figure 1). Rock Creek
Queen” (@potentatemining, May 27, 2019). Potentate also produced a notable 39.14 ct stone in 2008 that was

MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023 19
 Figure 18. With their
nonstandard colors,
Montana sapphires
have become popular
for cutters experiment-
ing with nontraditional
cutting techniques.
This 4.17 ct unheated
sapphire cut by Nadine
Marshall was sourced
from the Missouri
River. Courtesy of Ob-
sessed Over Gems.

recovered from the Anaconda Bench at Gem Mountain finest and largest sapphires produced by Potentate dur-
(Berg, 2014). The largest crystal reported from Dry Cot- ing the 2021 mining season in their rough and polished
tonwood Creek weighed 22 ct but had a flat shape and states.
therefore a low yield when cut into a gem (Berg, 2007).
While small stones under a carat are the main produc- Optical Measurements and Specific Gravity. Stan-
tion from Rock Creek, figure 1 shows a suite of the dard gemological testing of Montana sapphire reveals

Figure 19. Montana sapphires in a range of shapes and colors. All are heated, and the faceted stones range from about
0.20 ct for the 3 × 5 mm kite shapes on the left to 2.11 ct for the hexagon on the right. Photo by Dillon Sprague; cour-
tesy of Columbia Gem House.

20 MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023
Figure 20. Parti-colored sapphires are occasionally Figure 22. A suite of nine unheated rubies from the
found in Montana and can combine a wide range of Rock Creek deposit ranging from 0.17 to 0.58 ct.
colors. These examples, ranging from 1.62 to 3.20 ct, Photo by Kevin Schumacher; courtesy of Jeffrey R.
are all heated. Photo by Jeffrey R. Hapeman. Hapeman, Earth’s Treasury, Inc.

properties one would expect for corundum. The av- sorption bands centered around 3000 cm–1 that is
erage refractive index measured is 1.760–1.770, and often referred to as the “Punsiri peaks” but will be re-
the birefringence typically ranges from 0.008 to ferred to here as the acceptor-dominated 3000 cm–1
0.010. The specific gravity ranges from 3.98 to 4.01, series, given that these peaks are related to hydrogen
also consistent with corundum. associated with acceptor ions, in this case Mg2+, in
FTIR spectra of Montana sapphire can show a va- the corundum (Fukatsu et al., 2003; Sangsawong et
riety of features (figure 23). Commonly seen is a weak al., 2016). The acceptor-dominated 3000 cm–1 series
single peak at 3309 cm–1. This peak is believed to be is not an indicator of heat treatment for Montana
related to hydrogen in the corundum structure asso- sapphire, as these peaks sometimes occur naturally.
ciated with titanium substituting for aluminum In fact, this series of peaks is very common in un-
(Moon and Phillips, 1991). Also common in secondary heated Rock Creek sapphire but only occasionally
Montana sapphire is a peak at about 3220 cm–1. A sim- found in Dry Cottonwood Creek sapphire and almost
ilar peak is often seen in Yogo sapphire as well (Renfro never in Missouri River sapphire. The commercial
et al., 2018), but its origin is unknown. Finally, many importance of this finding will be discussed later in
Montana sapphires show the collection of broad ab- the section dealing with heat treatment. The peak at

Figure 21. Some secondary Montana sapphires exhibit a color change between incandescent illumination (left) and
daylight (right). These examples are unheated and range from 0.74 to 0.83 ct. Photos by Robert Weldon; courtesy
of Jeffrey R. Hapeman, Earth’s Treasury, Inc.

MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023 21
 FTIR SPECTRA
0.16 0.40 0.40

A B C
ABSORPTION COEFFICIENT (cm–1)

ABSORPTION COEFFICIENT (cm–1)

ABSORPTION COEFFICIENT (cm–1)

0.14 0.35 0.35

0.12 0.30 0.30

0.10 0.25 0.25

0.08 0.20 0.20

0.06 0.15 0.15

0.04 0.10 0.10

0.02 0.05 0.05

0 0 0
3400 3200 3000 3400 3200 3000 3500 3000 2500
WAVENUMBER (cm ) –1
WAVENUMBER (cm )
–1
WAVENUMBER (cm )
–1

Figure 23. Representative FTIR spectra showing expected features for Montana sapphire, including a single
3309 cm–1 peak (A), a 3220 cm–1 peak (B), and the acceptor-dominated 3000 cm–1 series (known as the “Punsiri
peaks,” C).

3220 cm–1 can be used as an indicator of a secondary less commonly as hexagonal prisms. The rough stones
Montana or primary Yogo origin, but the other FTIR also tend to show a rounded appearance, which might
features such as the 3309 cm–1 peak and the accep- cause one to suspect they have been subjected to
tor-dominated 3000 cm–1 series of peaks are observed rounding by abrasion from water transport. However,
in sapphires from many other origins. Montana sap- close examination with a microscope often reveals
phire from all the secondary deposits may also show pronounced geometric patterns on the surface with ei-
mineral inclusion peaks in their FTIR spectra, espe- ther the hillocky-type texture seen at Rock Creek or
cially for kaolinite or gibbsite. dissolution-type etch features (figures 24–26). In fact,
the sapphire morphology from the three deposits gen-
Shape of Rough Crystals/Morphology. The sapphire erally differs enough to allow identification of the ori-
from Montana’s secondary deposits is generally tabu- gin of parcels of rough sapphires, even if individual
lar or blocky in shape and can also appear somewhat stones cannot be traced back to their source. For in-

Figure 24. SEM image showing the surface of a Rock Figure 25. Hillocky structure on the surface of a rough
Creek sapphire. The hillocks on the surface are typi- sapphire from Rock Creek shown with differential in-
cal of Rock Creek sapphire. Image by Richard Berg terference contrast imaging. Photomicrograph by
(modified from Berg, 2022). Aaron Palke; field of view 2.88 mm.

2 mm

22 MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023
Figure 26. Differential interference contrast imaging of the surface of a rough sapphire from Dry Cottonwood
Creek showing etch features on the basal pinacoid surface (left, perpendicular to the c-axis) and on the prismatic
crystal face (right, parallel to the c-axis). Photomicrographs by Aaron Palke; field of view 0.72 mm.

stance, some 25% of sapphires from Missouri River Figure 30 shows typical shapes for the crystals
have an incomplete or spotty dark green coating of found at all three locations. The two most com-
spinel in depressions on the surface, as seen in the monly encountered shapes are hexagonal prismatic
SEM image in figure 27. This feature is not seen in crystals with slightly developed rhombohedral faces
sapphire from Rock Creek or Dry Cottonwood Creek. (figure 30, A and B). Many show development of the
Dry Cottonwood Creek sapphire shows etching along hexagonal bipyramidal faces as well (figure 30, A and
twinning planes as seen in figure 28, which is less C). Crystals typically exhibit six-fold or three-fold
common in Rock Creek and Missouri River sapphire. symmetry on the basal pinacoid; figure 30B shows a
Finally, Rock Creek sapphire often has growth classic raised trigonal pyramid on the basal pinacoid.
hillocks on the basal surface, some with flat surfaces There are slight variations between the three de-
resembling miniature mesas (figure 24). Examples of posits in the prevalence of the various crystal shapes.
rough from each deposit are shown in figure 29. For example, the majority of the crystals from Dry

Figure 27. SEM image of the surface of a sapphire

from Missouri River showing the spinel coating some- Figure 28. As seen in this SEM image, sapphires from
times seen in stones from this deposit but never in Dry Cottonwood Creek are often dissolved or etched
sapphire from Rock Creek or Dry Cottonwood Creek. along parallel grooves following twin planes. Image
Image by Richard Berg. by Richard Berg (modified from Berg, 2022).

Spinel

300 μm 2 mm

MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023 23
 A B C

Figure 29. Photos of representative lots of unheated rough from each of Montana’s secondary sapphire deposits:
Rock Creek (A), Missouri River (B), and Dry Cottonwood Creek (C). The ruler in each photo is in millimeter scale.
Photos by John Emmett.

Cottonwood Creek are tabular hexagonal crystals and can reveal the structure even in heavily etched
(figure 30C), while elongated hexagonal prisms (fig- crystals. Despite this variation at each locality, crys-
ure 30D) are particularly common at Missouri River. tals of all of these shapes can be found at any of the
At Rock Creek, most of the crystals have been so deposits.
heavily etched/resorbed that most of their crystal
faces are obscured, although color zoning and the pat- Microscopic Features. Secondary sapphires from Mon-
terns of rutile inclusions follow these crystal faces tana contain a wide array of inclusions, and these are
a powerful tool in confirming a Montana origin. In
natural, unheated Montana sapphire, the most com-
Figure 30. Typical shapes seen in rough Montana sap-
mon features are bands of rutile silk, often arranged in
phires. Crystal forms include pinacoidal faces (pc),
hexagonal patterns, either as dense hexagonal clouds
prismatic faces (p), rhombohedral faces (r), and py-
ramidal faces (py). Most common shapes are pris- (figure 31, A and B) or in a loosely packed but overall
matic crystals modified by pyramidal and/or hexagonal pattern (figure 31, C and D). The individual
rhombohedral faces (A and B); flattened, tabular particles making up the clouds may be relatively long,
hexagonal crystals terminated by rhombohedral, py-
ramidal, and pinacoidal faces (C); and elongate pris-
matic crystals (D). Figure 31 (opposite page). Typical inclusion scenes in un-
heated (A–F) and heated (G and H) secondary Montana
sapphires. A: Dense, hexagonal particulate clouds in a
A pc
B Missouri River sapphire. B: Hexagonal particulate clouds
py made up of small to medium-sized rutile needles in an
py
r pc unheated orange sapphire from Rock Creek. C: More
r loosely packed hexagonally arranged silk in an unheated
p p p p sapphire from Rock Creek. D: Fiber-optic illumination re-
p p
veals interference colors from flattened silk particles in
r
this Rock Creek sapphire. E: The rare rubies occasionally
r
py found in Montana have essentially the same inclusions
as the sapphires, such as the flattened rutile silk in this
Rock Creek ruby. F: The flattened platelet-like silk in
sapphires from Missouri River often has unusual mor-
C D pc
phology, with an atoll-like inclusion at the head of a
r
platelet particle. G: Blue color zones leaking out of proto-
pc genetic rutile inclusions in a heat-treated sapphire from
p py Rock Creek. H: Patchy blue color zoning in a Dry Cot-
r p p p p
py
p tonwood Creek sapphire, caused by dissolution of bands
r
of rutile silk into the corundum lattice. Photomicro-
graphs by Nathan Renfro (A, G, H) and Aaron Palke (B–
r F); fields of view 3.31 mm (A), 4.79 mm (B), 4.79 mm (C),
4.57 mm (D), 1.26 mm (E), 1.26 mm (F), 2.19 mm (G),
and 3.31 mm (H).

24 MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023
 A B

C D

E F

G H

MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023 25
slender needles varying in size all the way to very fractures or cleavages perpendicular to the direction
short, stubby, or dusty particles. In some cases, the of elongation. Sapphires from Dry Cottonwood Creek
particles take on a flattened, platelet-like shape and are also more likely to contain zircon inclusions, es-
are highly reflective, often showing iridescent inter- pecially in clusters and galaxies of crystals resembling
ference colors with the use of an intense fiber-optic rice grains (figure 32F). Zircon inclusions are rare in
light due to a thin-film effect (figure 31, D–F). sapphire from Rock Creek and Missouri River.
Given that the majority of Montana sapphires on Mica inclusions may be seen and range from
the market have been heat treated to enhance their nearly colorless to pale shades of yellow to occasional
colors, attention must be paid to the inclusions that dark orange hues (figure 33A). Colorless rounded
can help the gemologist confidently identify this treat- feldspar inclusions are fairly common as well from
ment. The most obvious effect of heat treatment on all three deposits. Note that a number of other min-
Montana sapphire is internal diffusion of blue col- eral inclusions were documented by Zwaan et al.
oration from rutile inclusions and silk. This can be (2015) and compiled in a table of occurrence. While
seen as patchy blue color zones that precisely match many of the sapphires they studied appear to have
the original patterns of the rutile silk. This is most been heated, Zwaan et al. (2015) report a more exten-
easily observed using diffuse transmitted illumina- sive list of minerals identified in their study and from
tion, but fiber-optic illumination can be useful as well. other literature sources as well.
This will lead to patchy blue color zoning surrounding In some rare sapphires from Missouri River, un-
bands of silk that are partially or nearly completely usual multiphase inclusions have a negative crystal
dissolved into the corundum lattice (figure 31G) as morphology in which the shape of the inclusion is
well as dark, inky spots surrounding protogenetic ru- constrained by the crystal forms of the corundum
tile inclusions (figure 31H). More examples will be host and not by the minerals inside the inclusion.
shown in the section on heat treatment below. These multiphase inclusions are composed of mica,
On rare occasions, rutile silk may be dense enough spinel, and some other unidentified phases (figure
to produce a star sapphire when cut in cabochon form 33B). These inclusions are interpreted to have been
(figure 32A). Yet star sapphires from Montana are ex- entrapped initially as a melt phase, which recrystal-
ceedingly rare, and the star is usually poorly devel- lized post-entrapment. In fact, melt inclusions with
oped. Many of them show interesting coloration, with a negative crystal shape are quite common in second-
yellow or orange zones in the core of the crystal, ary Montana sapphire. They often are surrounded by
which is most prominent in heat-treated stones (figure decrepitation halos, which may be partially healed
32B). Twinning with tube-like structures at the inter- and show iridescent interference colors (figure 33C).
section of twin planes is fairly common in sapphires In many cases, these inclusions look like standard
from Dry Cottonwood Creek but may occasionally be two-phase fluid inclusions (figure 33, D and E). How-
seen in sapphires from Rock Creek and Missouri ever, when one of these inclusions is cut into at the
River. These features are observed in sapphires from surface of a stone, the apparently fluid phase does not
other deposits as well (Hughes, 2017) and have been flow away but is frozen in place (figure 33F). These
referred to as “Rose channels” (Notari et al., 2018). inclusions were trapped as a relatively silica-rich
(Note that Rose channels have been referred to erro- melt (Palke et al., 2017) that coexisted with the sap-
neously in the past as “boehmite needles.”) Large, phires as they were growing. After being emplaced at
blocky protogenetic rutile inclusions are common in the earth’s surface, the sapphires must have cooled
stones from all the Montana secondary deposits (figure quickly enough to quench the melt phase into a
32C). In this study, garnets were occasionally encoun- glass. Note that while these inclusions resemble the
tered in sapphires from Dry Cottonwood Creek (figure “melted crystal” inclusions seen in heat-treated sap-
32D). Although Zwaan et al. (2015) observed garnet phires, it is not correct to apply this term to these in-
inclusions in sapphire from Rock Creek and Missouri clusions. These are properly called “melt
River, the present authors have not; these inclusions inclusions,” as they were included as a melt and do
are likely rare compared to material from Dry Cotton- not represent crystalline inclusions melted by artifi-
wood Creek. Garnet inclusions could then be consid- cial heat treatment. Therefore, observation of this
ered a likely indicator of Dry Cottonwood Creek type of inclusion is not, in and of itself, evidence of
origin. Sapphires from all three deposits are likely to heat treatment in Montana sapphire. Identification
contain inclusions of clinozoisite, which takes on an of heat treatment is more reliably carried out by ob-
elongate form (figure 32E), sometimes with multiple servation of the altered crystalline inclusions, the al-

26 MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023
 A B

C D

E F

Figure 32. A: Dense rutile silk in a star sapphire from Rock Creek. B: Yellow core in a heated sapphire from Mis-
souri River. C: Protogenetic rutile inclusion in a Rock Creek sapphire. D: Garnet inclusion in a sapphire from Dry
Cottonwood Creek. E: Clinozoisite inclusion in a Dry Cottonwood Creek sapphire. F: Field of zircon inclusions in
a sapphire from Dry Cottonwood Creek. Photomicrographs by Nathan Renfro (A–E) and Aaron Palke (F); fields of
view 2.56 mm (A), 6.26 mm (B), 1.30 mm (C), 2.82 mm (D), 2.09 mm (E), and 1.26 mm (F).

tered nature of rutile silk particles, and/or color dif- highly lustrous surface with the use of a fiber-optic
fusion around rutile or silk particles. light (figure 34). They are always accompanied by a
Also worth mentioning are sulfides assuming a decrepitation halo where sulfide material has leaked
negative crystal shape. These inclusions are dark out into the surrounding corundum host. The nega-
black in darkfield illumination but may display a tive crystal morphology and observation of a

MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023 27
 A B

C D

E F

Figure 33. Inclusions in natural-color (unheated) secondary Montana sapphire. A: Mica inclusions in a Rock Creek
sapphire. B: Missouri River sapphire with an unusual multiphase inclusion composed of mica, spinel, and other
unidentified phases. C: Two negative crystal inclusions with a partially healed decrepitation halo in a Missouri
River sapphire. D: Field of glassy melt inclusions in a Rock Creek sapphire. E: Single glassy melt inclusion with a
decrepitation halo in a Rock Creek sapphire. F: Glassy melt inclusion breaking the surface of a Missouri River sap-
phire. Photomicrographs by Nathan Renfro (A and B) and Aaron Palke (C–F); fields of view 2.09 mm (A), 1.31 mm
(B), 2.34 mm (C), 0.71 mm (D), 1.26 mm (E), and 0.71 mm (F).

polyphase mineral assemblage by SEM analysis sug- present during sapphire formation in addition to the
gests that these are an additional type of melt inclu- more silica-rich melts described above. In the au-
sion. In this case, a sulfide melt must have been thors’ experience, these sulfide melt inclusions are

28 MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023
 A B

C D

Figure 34. Inclusions in natural-color (unheated) secondary Montana sapphire. A: Sulfide inclusion in a Missouri
River sapphire, shown in darkfield illumination. B: The same sulfide inclusion as in A, but with fiber-optic illumi-
nation. C: Sulfide inclusion in a Missouri River sapphire with a decrepitation halo where molten sulfide material
leaked out into a fracture in the corundum host. D: Field of sulfide inclusions in a Missouri River sapphire. Pho-
tomicrographs by Aaron Palke; fields of view 1.26 mm (A–C) and 4.79 mm (D).

seen almost exclusively in sapphires from Missouri medium-saturation pastel hues of blues, greens, yel-
River and Dry Cottonwood Creek, though they are lows, oranges, purples, and pinks. Beyond the initial
not common even then. Although sulfide inclusions impression from the bodycolor, one may find evi-
have been noted in a few Rock Creek sapphires (J.I. dence to separate these sapphires down the ocular
Koivula, pers. comm., 2022), screening of thousands barrel of a microscope. Umba sapphires especially
of samples by the authors has not identified this in- tend to have distinct inclusion features, the most
clusion in Rock Creek sapphire. common of which is heavy, repeated twinning.
Under cross-polarized light, the variously twinned
Inclusions/Microscopic Observations of Other Non- sectors may create a unique mosaic of colors (figure
Classical Sapphires. The sapphires most likely to be 35A). The boundaries between twinned sectors are
confused with Montana sapphire are those from often prominent and may be filled in with some sec-
Songea in Tanzania. Sapphires from Umba in Tanza- ondary aluminum oxy-hydroxide minerals (figure
nia may also show some generally similar character- 35B). Fields of clustered and single zircon inclusions
istics, but the inclusions are generally distinct are quite common in Umba sapphires (figure 35C).
enough to clearly separate them from secondary Dense aggregations of silk are not as common as in
Montana sapphires. The first thing one might notice Montana sapphires; however, the silk that can be
is the overall similar color range among all these de- seen often takes on a platy but very angular appear-
posits, which tends to include light-toned, low- to ance (figure 35, D–F).

MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023 29
 A B

C D

E F

Figure 35. Typical inclusions in Umba sapphires. A: Multiple twinned sectors in cross-polarized light. B: Secondary
aluminum oxy-hydroxide minerals formed in the intersection of twins. C: Clusters of zircon inclusions. D:
Platelet-like reflective particles with iridescent thin-film interference colors using fiber-optic illumination. E:
Platelet inclusions. F: Silk in an Umba sapphire, from elongate needles to platelet-like inclusions. Photomicro-
graphs by Aaron Palke (A, B, D, and F), Charuwan Khowpong (C), and Ungkhana Atikarnsakul (E); fields of view
3.57 mm (A), 1.42 mm (B), 2.90 mm (C), 2.34 mm (D), 4.64 mm (E), and 2.90 mm (F).

Songea sapphires may have inclusion features blocky to angular or needle-like protogenetic rutile
that could look somewhat similar to secondary Mon- inclusions (figure 36, A and B). The silk in some
tana sapphire inclusion features. Often seen are Songea sapphire can somewhat resemble the silk in

30 MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023
 A B

C D

E F

Figure 36. Representative inclusions in Songea sapphires. A: Protogenetic rutile inclusions with lobe-shaped
fringes. B: Two slender rutile inclusions. C: Typical pattern of rutile silk in a Songea sapphire, including sparsely
distributed medium to long silk and billowy clouds of particles. D: Small rutile dust and needles alongside
unidentified colorless crystals. E. Reflective silk and platelet particles. F. Reflective, scattered silk and particles.
Photomicrographs by Aaron Palke (A–C, E, and F) and Nathan Renfro (D); fields of view 1.26 mm (A and B), 2.34
mm (C), 1.79 mm (D), 1.76 mm (E), and 2.34 mm (F).

Montana sapphire. However, it is generally much (figure 36, C–F). Songea sapphire generally does not
more sparsely distributed and may appear in some- have the well-formed hexagonal bands of silk seen in
what randomly placed patches and as flecks of silk the Montana sapphire.

MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023 31
TABLE 1. Comparison of trace element compositions (in ppma) of sapphires from Montana’s secondary
sources and other deposits.a
Number of
Mg Ti V Cr Fe Ga
samples

Secondary
Montana 31 29 3 6 1600 13
51
(blue, green, (8–111) (3–109) (1–9) (bdl–43) (610–2750) (7–18)
yellow, colorless)

Secondary
Montana 25 29 6 742 1900 15
22
(pink sapphire (11–49) (9–81) (1–27) (32–2400) (620–2960) (11–20)
and ruby)

Songea, Tanzania 17 20 11 106 2580 14

19
(various colors) (9–33) (12–33) (3–21) (37–218) (2130–3650) (13–20)

Umba, Tanzania 23 25 10 19 1330 16

17
(various colors) (bdlb–45) (1–83) (1–23) (bdl–95) (430–2640) (7–30)

Myanmar 41 53 6 31 928 24
219
(various colors) (bdl–1510) (6–1018) (bdl–73) (bdl–1260) (172–3041) (5–82)

Madagascar 30 128 6 7 598 20

263
(various colors) (bdl–167) (bdl–1942) (bdl–43) (bdl–276) (46–2717) (4–51)

Sri Lanka 42 133 8 4 330 24

124
(various colors) (bdl–390) (4–1410) (bdl–49) (bdl–74) (bdl–1070) (3–92)

Detection limits
0.1–0.3 0.5–2.0 0.03–0.2 0.06–1.3 1–5 0.03–0.07
(ppma)

a
Averages are shown first, with the full range given in parentheses.
b
bdl = below detection limit

Trace Element Chemistry. Sapphires from the sec- (2007) as well as the Cr/Ga vs. Fe/Ti diagram of
ondary Montana deposits are classified by gemolo- Sutherland et al. (1998).
gists as “non-classical” sapphires. In essence, this The general trace element characteristics of the
means they are distinct from both the typical meta- secondary Montana sapphire deposits are summa-
morphic sapphires (from Myanmar, Sri Lanka, Mada- rized in table 1, where they are broadly compared
gascar, and Kashmir) and the typical magmatic, with sapphires from other deposits. Note that the
basalt-related sapphires (from Australia, Thailand, trace element chemistry of Montana’s three sec-
Cambodia, Nigeria, and Ethiopia). Their inability to ondary deposits broadly overlaps except for a few
fit into the standard metamorphic/magmatic di- very high-titanium Rock Creek sapphires, so they
chotomy is also seen in their trace element chem- will be considered together here as a single group.
istry. As is typical for non-classical sapphires, the There is some degree of overlap between the Mon-
secondary Montana sapphires have relatively high tana sapphires and a small subset of higher-iron
iron contents, suggesting a magmatic origin, while sapphires from the classical metamorphic deposits
the relatively low-gallium, higher-magnesium con- of Madagascar and Myanmar (figure 37); however,
tents and low-gallium/magnesium ratios suggests a sapphires from these deposits can generally be dis-
metamorphic origin. This discrepancy was pointed tinguished by careful observation of their inclusion
out in Palke et al. (2017), which showed that the characteristics (Palke et al., 2019). The sapphires
Montana sapphires plot in both the metamorphic that are more likely to be confused with the sec-
and magmatic domains in the Fe vs. Ga/Mg and Fe- ondary Montana sapphires are those from Umba
Mg×100-Ti×10 discriminant diagrams of Peucat et al. and Songea in Tanzania. There is broad overlap in

32 MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023
 Fe VS. Ga

19

17

Figure 37. Plot of Fe vs.

15
Ga (in ppma) of the sec-
ondary Montana sap-
phire compared to
Ga (ppma)

13 typical metamorphic
blue sapphires. There is
overlap in the chemistry
11 with some Burmese and
Madagascar sapphires
9
and some minor overlap
with Sri Lanka.
Secondary Montana
Burma (Myanmar)
7
Madagascar
Sri Lanka
5
500 1000 1500 2000 2500 3000
Fe (ppma)

most trace elements for these deposits (figure 38). consistently be used for separating the Montana
Iron contents are higher on average for Songea, and sapphires. But for Montana sapphires occurring in
gallium contents of some Umba sapphires are colorless, green, blue, and yellow ranges, the trace
much higher, but overall these elements cannot element vanadium is a useful identifier. In partic-

Mg VS. Ti
120
Secondary Montana
(blue, green, yellow, colorless)
100 Secondary Montana
(pink, red)
Songea, Tanzania
80 Umba, Tanzania Figure 38. Plot of Mg vs.
Ti (in ppma) of the sec-
ondary Montana sap-
Ti (ppma)

60 phire compared to other

deposits whose stones
may have generally
40 similar gemological
properties.

20

0
0 20 40 60 80 100 120
Mg (ppma)

MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023 33
 V VS. Fe
4000
Secondary Montana
(blue, green, yellow, colorless)
3500
Secondary Montana
(pink, red)
Figure 39. Plot of V vs.
3000 Songea, Tanzania
Fe (in ppma) of the sec-
Umba, Tanzania
ondary Montana sap-
2500 phire compared to other
deposits whose stones
Fe (ppma)

may have generally

2000
similar gemological
properties. The second-
1500 ary Montana sapphires
generally have lower
vanadium contents
1000
than those from Umba
and Songea in Tanzania.
500

0
0 10 20 30 40
V (ppma)

ular, the vanadium vs. iron plot can separate most Emmett and Douthit’s technology was used to heat
Montana sapphires from Umba and Songea sap- the production of the American Gem Corporation
phire (figure 39). The caveat is that this does not in the 1990s and is currently employed by a small
apply to pink sapphire and ruby from Montana, handful of heat treatment facilities in the United
which have higher vanadium concentrations that States.
overlap with sapphire from Songea and Umba. This section chronicles the journey of several sap-
phire wafers through the heat treatment process with
Heat Treatment of Montana Sapphires. The previous careful documentation of their inclusions and spec-
section outlined various internal characteristics with troscopy before and after heating. Additionally, the
an eye toward establishing the provenance of un- samples were cut as oriented wafers with the c-axis
heated Montana sapphire from the Rock Creek, Mis- (or the optic axis) perpendicular to the plane of the
souri River, and Dry Cottonwood Creek deposits. wafer. This allows for only the o-ray spectrum to be
However, the majority of these sapphires entering the collected, while the extraordinary ray (e-ray) spec-
market have been heat treated at high temperatures trum cannot be observed. Note that this section is not
to improve their color, making them more desirable meant to be an exhaustive review of the heat treat-
for use in jewelry. This treatment fundamentally al- ment process for Montana sapphires, but is intended
ters the sapphires’ internal characteristics as well as to give readers an overview of key features that can
their spectroscopic properties. identify heat-treated Montana sapphires and to
It is important to note that, due to their unique demonstrate the effects of heat treatment on stones
chemistry, Montana sapphires require specialized that are commercially available. More technical de-
and carefully controlled heating environments that tails about the heat treatment process can be found
can typically be obtained only in modern resistance- in Emmett and Douthit (1993). The first three exam-
style furnaces. The specific technical requirements ples were heated by Dale Siegford of the Sapphire
for the heat treatment of Montana sapphire were re- Gallery in Philipsburg, Montana, and conditions of
alized in the pioneering scientific work of Emmett heating were not disclosed. The fourth and fifth ex-
and Douthit (1993). These involve careful control of amples were heated at GIA’s experimental heat treat-
not only heating temperatures and duration but also ment facility with carefully controlled conditions
oxidation environments using gas-mixing furnaces. described below.

34 MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023
 Figure 40. The Rock
Creek sapphire in exam-
ple 1, measuring 9.5 ×
8.2 mm, before (left)
and after (right) heat
treatment shows the
blue coloration intro-
duced during treatment.
The gray circle on the
left shows the position
of the UV-Vis spectro-
scopic measurement.
Photos by Aaron Palke.

Example 1: Colorless and Cloudy to Blue. The first bands at 377, 388, and 450 nm related to Fe3+ and a
example illustrates the most advantageous outcome gradual upward slope from the near-infrared to the
for heat treatment of Montana sapphires. The stone ultraviolet region, caused by scattering of light from
was initially colorless but quite cloudy and included the rutile silk. The most obvious change caused by
(figure 40, left). The transmitted light photo of the heating is the creation of broad, intense absorption
unheated stone shows dense, hexagonal bands of ru- bands at 580 and 880 nm. The 580 nm band is re-
tile silk that scatter light, creating brown to yellow- sponsible for the stone’s blue coloration, as it absorbs
ish patches. While some unheated Montana sapphire most of the red light passing through by way of the
can be cut into exceptional colorless or slightly gray- excitation of an intervalence charge transfer (IVCT)
ish gems, this stone’s cloudiness makes it undesir- between Fe2+ and Ti4+ (Dubinsky et al., 2020). This
able for faceting. However, heat treatment causes a band increases in intensity as rutile silk is dissolved
dramatic change in the sapphire’s appearance, creat- into the corundum structure, where the titanium
ing a rich blue coloration where there were once ions introduced can pair locally with iron ions in a
patches of cloudy silk (figure 40, right). charge-balancing substitution of Fe2+ +Ti4+ for two
The change in color is also seen in the UV-Vis spec- ions of Al3+. The origin of the 880 nm band is not well
tra of this sapphire before and after heating (figure 41). understood but may be related to clusters of Fe2+, Fe3+,
The unheated sapphire showed narrow absorption and/or Ti4+ cations in the corundum structure

UV-VIS SPECTRA

14 Unheated
Heated
ABSORPTION COEFFICIENT (cm–1)

12

10 Figure 41. The o-ray UV-

Vis spectra of example 1
8 in its unheated and
heated states demon-
6 strating how blue col-
oration is introduced by
4 heat treatment.

377/388 450 580 880

0
300 400 500 600 700 800 900 1000
WAVELENGTH (nm)

MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023 35
 A B C

Figure 42. Example 1 before heat treatment under fiber-optic illumination (A), after heat treatment under fiber-
optic illumination (B), and after heat treatment under diffuse transmitted illumination (C). Photomicrographs by
Aaron Palke; field of view 4.08 mm.

(Hughes et al., 2017). Of particular note is that while ture of long needles as well as clouds of smaller par-
unheated blue Montana sapphires have a UV-Vis ticles either aligned with the hexagonal growth pat-
spectrum reminiscent of a typical metamorphic blue terns of the corundum or forming less-regular
sapphire, with an 880 nm band that is less intense patterns not ostensibly constrained by the corundum
than the 580 nm band (Palke et al., 2019), this heated lattice (figure 42A). Also seen are some reflective
blue Montana sapphire has a UV-Vis spectrum that platelet inclusions at the top left and bottom center
resembles a magmatic, basalt-related blue sapphire, of figure 42A. The heated stone still shows signifi-
with an 880 nm band that is more intense than the cant scattering of light from these rutile particles
580 nm band. This “reversal” of the UV-Vis spec- using fiber-optic illumination, indicating that the
trum in heated sapphire has been documented by particles have not fully dissolved into the corundum
both Emmett and Douthit (1993) and Hughes and structure. The long rutile needles have been com-
Perkins (2019), and it should provide the basis for pletely destroyed by heat treatment, but the overall
careful application of this classification scheme for pattern and distribution of smaller, cloudy rutile par-
heated blue sapphires. The other notable difference ticles has not changed dramatically (figure 42B).
for the heated sapphire is the decrease in absorption More conclusive evidence of heat treatment comes
in the near-UV and blue/green region of the spectrum from observing the distribution of blue coloration
from about 350 to 500 nm. This is caused by the dis- seen using diffuse transmitted illumination (figure
solution of rutile particles into the corundum struc- 42C) and comparing it to the distribution of incom-
ture and the consequent reduction in light scattering, pletely dissolved rutile inclusions (figure 42B). The
which enhances the blue coloration caused by the blue coloration in this sapphire shows a nearly exact
Fe2+/Ti4+ IVCT. correlation with the incompletely dissolved rutile
Finally, evidence of heat treatment can be found clouds, providing conclusive evidence of high-tem-
through careful microscopic observation. The pris- perature heat treatment. Also notable is that with
tine, unaltered silk in the unheated stone is a mix- the use of intense fiber-optic illumination, the blue

Figure 43. The Rock Creek

sapphire in example 2,
measuring 11.7 × 7.2 mm,
before (left) and after
(right) heat treatment
shows the blue color intro-
duced by heat treatment.
The gray circles on the
right show the positions of
the UV-Vis spectroscopic
measurements. Photos by
Aaron Palke.

36 MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023
 UV-VIS SPECTRA
ABSORPTION COEFFICIENT (cm–1) 20
Unheated, rim Unheated, core
Heated, rim Heated, core
Figure 44. UV-Vis spec-
15 tra of example 2 in its
unheated and heated
state. Two spectra were
10 collected in both states
in the core and rim of
the crystal demonstrat-
5
ing the creation of blue
coloration after heat
treatment.

0
300 400 500 600 700 800 900 1000
WAVELENGTH (nm)

coloration does seem to emanate from the especially heated core also has a pronounced rise in absorption,
dense cloudy regions such as the sector seen in the and its decrease in wavelength from the near-IR to the
top center of the photomicrograph in figure 42B. near-UV is caused by scattering from the densely con-
centrated rutile silk particles. After heating, the 580
Example 2: Brownish Blue to Blue. Our second ex- nm band increases dramatically in the core and slightly
ample involves a sapphire with a fairly well-devel- in the rim. Increased clarity (and decreased scattering)
oped blue coloration in its unheated state. However, is also important for improving the blue color in the
the extremely dense, cloudy nature of this stone, es- core, as there is also a significant decrease in absorption
pecially in its core, would have made it unsuitable in the ~350–500 nm region caused by dissolution of
for faceting. In this stone, the silk lends a brown col- light-scattering silk. Similar to example 1 above, the
oration. As shown in the before/after photos taken heated stone has what would appear to be a magmatic,
with diffuse transmitted light (figure 43), these basalt-related blue sapphire absorption spectrum, with
brownish cloudy regions take on an intensely satu- the 880 nm band more intense than the 580 nm band.
rated blue color after heat treatment. This is in contrast to the ostensibly metamorphic blue
This intensification of blue color can also be seen sapphire spectrum in the unheated stone.
in the UV-Vis spectra collected in two spots in the Finally, microscopic observations are useful for
core and rim of the sample before and after heating. identifying the high-temperature heat treatment of
In the unheated state, both the core and rim have a no- this sapphire. The unheated sapphire shows an ex-
ticeable broad absorption band at 580 nm related to tremely dense concentration of silk, largely com-
Fe2+/Ti4+ IVCT as well as narrow Fe3+-related absorption posed of medium to long needles, especially in the
bands at 377, 388, and 450 nm (figure 44). The un- core (figure 45A). The long needles have become par-

Figure 45. Example 2 before heat treatment with fiber-optic illumination (A), after heat treatment with fiber-optic
illumination (B), and after heat treatment with diffuse transmitted illumination (C). Photomicrographs by Aaron
Palke; field of view 4.79 mm.

A B C

MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023 37
 Figure 46. The Rock Creek
sapphire in example 3,
measuring 9.5 × 9.2 mm,
before (left) and after
(right) heat treatment
shows creation of blue
color by heat treatment.
The gray circle on the left
shows the position of the
UV-Vis spectroscopic
measurement. Photos by
Aaron Palke.

tially dissolved into the corundum in the heated sap- color is, perhaps, less obvious from the UV-Vis spec-
phire (figure 45B). The use of fiber-optic illumination tra (figure 47). The absolute intensity of absorption
reveals that the silk has not completely dissolved and at 580 nm is not significantly different before and
can still scatter light quite intensely. Importantly, in after treatment. The main difference is the decreased
the heated stone, the light scattering off the remnant absorption between ~350–500 nm caused by reduc-
clouds has a distinct blue color as a result of the tion in scattering when the fine rutile silk is dis-
leaching of Ti4+ ions locally into the corundum. The solved into the corundum, allowing more blue light
correlation between dense blue coloration and the to be transmitted. While not as obvious, the 580 nm
remnant rutile clouds can be seen more clearly with band likely becomes more intense as well, although
the use of diffuse transmitted illumination (figure its increased intensity may be masked by the de-
45C), clearly demonstrating heat treatment. crease in absorption related to scattering. Note that
while the 880 nm band does increase after heating,
Example 3: Light Blue to Deeper Blue. The next ex- this change is not significant enough to make it more
ample involves a light blue sapphire with sparse silk intense than the band at 580 nm.
arranged in a hexagonal pattern, which is made ob- Microscopic observation reveals significant alter-
vious in diffuse transmitted light by the brownish ation of the rutile silk upon heating. Before heating,
coloration of the silk. Heat treatment increased the the silk was composed of a mixture of long, medium,
blue coloration by dissolving some of this silk into and short needles and particles arranged in a hexago-
the corundum structure (figure 46). The change in nal pattern (figure 48A). After heating, the long and

UV-VIS SPECTRA
20
Unheated
ABSORPTION COEFFICIENT (cm–1)

Heated

15

Figure 47. UV-Vis spectra

of example 3 in its un-
10 heated and heated state
demonstrating the deep-
ening of blue coloration
by heat treatment.
5

0
300 400 500 600 700 800 900
WAVELENGTH (nm)

38 MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023
 A B C

Figure 48. Example 3 before heat treatment with fiber-optic illumination (A), after heat treatment with fiber-optic
illumination (B), and after heat treatment with diffuse transmitted illumination (C). Photomicrographs by Aaron
Palke; field of view 2.34 mm.

medium silk are completely decomposed and the nificant rutile particles and silk was heated at
hexagonal clouds of particles are made up of very fine, 1700°C for 12 hours in a reducing atmosphere com-
dusty particles (figure 48B). Also notable is the small posed of 0.33 atm partial pressure of H2 and 0.67 atm
inclusion in the core with the concentrically aligned partial pressure of CO2. The sapphire was cooled
decrepitation halo (figure 48A, top center). After heat- slowly at a rate of 4°C/minute. This was done twice
ing, the decrepitation halo has become filled with to ensure the sapphire reached a steady state at these
some residual material, likely from the melting or al- conditions. The same sapphire was heated again at
teration of the originally intact central inclusion. Fi- the same conditions but quenched rapidly by placing
nally, the most conclusive evidence of heat treatment it in a crucible affixed to an alumina rod that can be
comes from comparing the distribution of partially retracted from the bottom of the muffle tube. In this
dissolved silk using a fiber-optic light and blue patches way, the sapphire is cooled from 1700°C to about
of coloration using diffuse transmitted light (figure 1000°C in 1–2 minutes and then further cooled to
48C). The correlation between the remnant silk room temperature in roughly 15 minutes. The spec-
clouds and the blue zones of color allows for easy iden- tra and photos of the sapphire after various stages of
tification of heat treatment in this case. heat treatment demonstrate the deeper blue color of
the rapidly cooled sapphire compared to the slowly
Example 4: Effect of Cooling Rate. While the precise cooled heating (figures 49 and 50). When the sapphire
heating conditions for the above experiments are not is slowly cooled, the titanium dissolved into the
known, additional tests have been carried out at corundum can start to exsolve as rutile (TiO2) as the
GIA’s experimental heat treatment facility. In the solubility of titanium in corundum decreases with
first experiment, a cloudy, pale blue stone with sig- decreasing temperature. As the rutile particles pre-

Figure 49. The sapphire in example 4, measuring 8.1 × 6.5 mm: unheated (A), heated to 1700°C and slowly cooled
(B), and heated to 1700°C and rapidly cooled (C), showing a deeper blue color with rapid cooling. The gray circle
shows the area of analysis. Photos by Aaron Palke.

A B C

MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023 39
 UV-VIS SPECTRA

14 Unheated
Heated, slow cool
Heated, rapid cool
12
ABSORPTION COEFFICIENT (cm–1)

Figure 50. UV-Vis spec-

10 tra of example 4 in its
unheated, heated
(slowly cooled), and
8
heated (rapidly cooled)
states demonstrating
6 the increased Fe2+-Ti4+
IVCT band at 580 nm,
with rapid cooling lead-
4 ing to deeper blue color.

0
350 450 550 650 750 850
WAVELENGTH (nm)

cipitate and grow during cooling, the titanium exits suggests either that rutile was never fully dissolved
the corundum lattice, causing a reduction in the into the corundum lattice at 1700°C or that even
Fe2+-Ti4+ intervalence charge transfer band and a re- with rapid cooling there was still some very rapid
duction in the blue coloration. Rapid cooling exsolution of rutile.
quenches in the high-temperature solubility of tita-
nium in the corundum lattice and retards exsolution Example 5: Creation of Trapped-Hole Coloration.
of rutile upon cooling. Note, however, that rapid For one final example, we consider two sapphires
cooling does not lead to complete dissolution of ru- with entirely different responses to heat treatment.
tile silk and particles in the sapphire, as seen in the Many Montana sapphires are heated to create or in-
photomicrographs in figure 51. The use of intense tensify yellow or orange color by creating what is re-
fiber-optic lighting shows that even the rapidly ferred to as a “trapped hole.” This trapped-hole
cooled sapphire has an abundance of light-scattering chromophore involves a Mg2+ atom substituting for
rutile particles. The presence of these rutile particles Al3+ with a missing electron on a nearby oxygen

Figure 51. Rutile silk and particles in example 4 in the unheated (A), heated (slowly cooled, B), and heated (rapidly
cooled, C) states showing the existence of rutile particles even in the heated and rapidly cooled sapphire. Photomi-
crographs by Aaron Palke; field of view 3.57 mm.

A B C

40 MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023
 A B C

Figure 52. The Rock Creek sapphire in example 5, measuring 0.8 × 0.5 mm: unheated (A), heated to 1200°C in pure
O2 (B), and heated to 1700°C in pure O2 (C). The gray circle shows area of analysis. Photos by Aaron Palke.

anion for charge balance (a trapped hole). This leaves oxygen vacancies instead of the trapped hole on the
the oxygen anion with a charge of –1 and causes it to O– anion.
absorb light at 450 nm, creating intense yellow or or- Figures 52–55 show the results of an oxidizing
ange color (see Dubinsky et al., 2020). Some heaters heat treatment run on two sapphires, one from Rock
who specialize in treating Montana sapphire will Creek (figures 52 and 53) and one from Dry Cotton-
carry out what is referred to as a “fancy burn” in an wood Creek (figures 54 and 55). Both were heated
oxidizing atmosphere to bring out yellow and orange once at 1200°C in a pure O2 atmosphere for 12 hours,
hues. The creation of trapped-hole absorption centers then heated again at 1700°C in a pure O2 atmosphere
by heating Montana sapphires in an oxidizing envi- for 12 hours.
ronment was carefully documented by Emmett and The sapphire from Rock Creek developed an or-
Douthit (1993). ange core after heating in O2 at 1200°C (figure 52).
Many heated fancy-color sapphires will display an The UV-Vis spectrum shows this change, with an in-
intense yellow or orange core where the trapped-hole crease in absorption in the region from 350 to 500 nm
chromophore is developed. This heating needs to be (figure 53). The difference between the heated and un-
carried out in an oxidizing environment in order to heated spectra shows the creation of the trapped-hole
create a trapped hole. Heating in reducing environ- absorption band centered at around 480 nm. The neg-
ments can facilitate the charge balance of Mg2+ by ative dip in the difference spectrum centered at about

UV-VIS SPECTRA
18
Unheated
16 1200°C, 12h
1700°C, 12h
ABSORPTION COEFFICIENT (cm–1)

14 Difference

12
Figure 53. UV-Vis spec-
10 tra of example 5 from
Rock Creek: unheated,
8
heated to 1200°C for 12
6
hours in pure O2, and
heated to 1700°C in
4 pure O2.

-2
350 450 550 650 750 850
WAVELENGTH (nm)

MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023 41
 A B C

Figure 54. The Dry Cottonwood Creek sapphire in example 5, measuring 7.2 × 6.7 mm: unheated (A), heated to
1200°C in pure O2 (B), and heated to 1700°C in pure O2 (C). The gray circle shows area of analysis. Photos by
Aaron Palke.

580 nm indicates that heating also reduced the Fe2+- in this sample by heating at the higher temperature
Ti4+ intervalence charge transfer absorption. Heating of 1700°C in pure O2 (figure 55). This sapphire also
at 1700°C resulted in a negligible change in color, pos- developed blue color in its rim during heating due to
sibly with a slight increase in the trapped-hole absorp- dissolution of rutile particles and introduction of Ti4+
tion band in the UV-Vis spectrum. cations into the corundum structure, leading to cre-
The sapphire from Dry Cottonwood Creek did not ation of Fe2+-Ti4+ IVCT chromophores.
develop the trapped-hole chromophore after heating The fundamental difference in the behavior dur-
at 1200°C (figure 54). The only change at this lower ing the oxidizing heat treatment is in the way excess
temperature was a lightening of the blue color due to Mg2+ cations are charge-balanced in the original un-
a reduction in the Fe2+-Ti4+ IVCT absorption (figure heated sapphire. The FTIR spectrum of the Rock
55). The trapped-hole chromophore is only developed Creek sapphire here contained the acceptor-domi-

UV-VIS SPECTRA
24
Unheated
1200°C, 12h
1700°C, 12h
ABSORPTION COEFFICIENT (cm–1)

19 Difference

Figure 55. UV-Vis spec-

14 tra of example 5 from
Dry Cottonwood Creek:
unheated, heated to
9
1200°C for 12 hours in
pure O2, and heated to
1700°C in pure O2.

-1
350 450 550 650 750 850
WAVELENGTH (nm)

42 MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023
Figure 56. Heat-treated Rock Creek sapphires; the largest rough stone weighs 4.62 ct. Photo by Jeff Scovil; courtesy
of Potentate Mining.

nated 3000 cm–1 series (the “Punsiri band”), which treating Montana sapphires from all three secondary
is caused by H+ cations charge-balancing excess Mg2+ deposits, Dr. Emmett has observed significant differ-
cations. This hydrogen is easily burned away at rel- ences in the way sapphires from these deposits react
atively low temperatures (i.e., 1200°C), leaving the to oxidizing heat treatment. Missouri River sapphires
Mg2+ to be charge-balanced by a trapped hole when do not have the acceptor-dominated 3000 cm–1 series
heated in an oxidizing environment (J.L. Emmett, in the infrared and therefore require high tempera-
pers. comm., 2022). The Dry Cottonwood Creek sap- tures in oxidizing conditions to develop trapped-hole
phire did not have the acceptor-dominated 3000 cm–1 coloration. On the other hand, Rock Creek sapphires
series in its FTIR spectrum, indicating that any ex- often have the acceptor-dominated infrared features
cess Mg2+ would have been charge-balanced by oxy- at 3000 cm–1 (see figure 23) and frequently react well
gen vacancies. Trapped holes in this case can only to an oxidizing fancy burn at temperatures as low as
be created by diffusion of aluminum vacancies into 1200°C. According to Dr. Emmett, a small propor-
the sapphire to annihilate oxygen vacancies, thereby tion of Dry Cottonwood Creek sapphires have the
allowing Mg2+ to be charge-balanced by trapped acceptor-dominated 3000 cm–1 features and will de-
holes instead of the oxygen vacancy (J.L. Emmett, velop yellow or orange color at 1200°C, but more fre-
pers. comm., 2022). But diffusion of aluminum va- quently these require higher temperatures around
cancies requires extreme temperatures and long 1700–1800°C.
heating times, which is why this is only seen in the
sapphire heated at 1700°C for 12 hours and not when CONCLUSIONS
heated at the lower temperature of 1200°C. Only a With large volumes of Montana sapphire being pro-
small number of Montana sapphires have been duced in consistent and reliable quantities (figure 56),
heated this way at GIA’s experimental heat treat- the secondary Montana sapphire deposits are some
ment facility so far. However, in many years of heat of the most important American gem mines in mod-

MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023 43
 Figure 57. A 2.02 ct un-
heated Montana sap-
phire featuring concave
faceting. Courtesy of
Derek Katzenbach.

ern times. The material fits a wide variety of market trace element chemistry, allowing the provenance of
demands, from heated goods under 1 carat size to these stones to be preserved and traced once the
rare, fine, unheated sapphires up to several carats. stones enter the gem and jewelry market. With the
Aggressive marketing by Potentate Mining and its af- wide range of colors produced, Montana sapphire is
filiates as well as by independent artisanal miners, an exceptionally versatile gem, finding a home in
especially through social media, has driven up inter- nearly any style of jewelry (figure 57). With large re-
est in these gemstones both domestically and in the serves remaining at Rock Creek, Missouri River, and
international gem and jewelry market. Identifying Dry Cottonwood Creek and their growing popularity
the origin and treatments for Montana sapphires can with consumers, Montana sapphire is likely to play
be relatively straightforward based on inclusions and an increasingly important role in the market.

ABOUT THE AUTHORS field at Eldorado Bar at Cass Thompson’s operation (Eldo-
Dr. Aaron Palke is senior manager of colored stone research, and rado Sapphire mine) and from Potentate’s operations at
Nathan Renfro is senior manager of colored stone identification, Rock Creek. Some Dry Cottonwood Creek samples were
at GIA in Carlsbad. Jeffrey R. Hapeman is president of Earth’s donated by Dr. John Emmett, who obtained them from the
Treasury (Westtown, Pennsylvania). Dr. Richard Berg is a research mining operations of American Gem Corporation in
geologist emeritus at the Montana Bureau of Mines and Geology 1994–1995. Dale Siegford provided samples for analysis
at Montana Tech of the University of Montana in Butte. and heat treatment services for some of the samples here.
Many thanks are owed to Dr. Keith Barron and Warren
ACKNOWLEDGMENTS Boyd of Potentate Mining for providing samples, informa-
Many samples containing prominent inclusions were sourced tion, and advice for this manuscript. The authors are grate-
from the inclusion collection of John I. Koivula. Samples from ful for thorough reviews by Dr. Emmett as well as two
GIA’s Colored Stone Reference Collection were collected in the anonymous reviewers.

REFERENCES
Barron K.M., Boyd W.F. (2015) The Rock Creek sapphire mine of dian Journal of Mineralogy and Petrology, Vol. 61, pp. 61–85,
Montana – A new era. InColor, No. 28, pp. 2–12. [Link]
Belley P.M. (2023) A method for detrital corundum characteriza- Berg R.B. (2007) Sapphires in the Butte-Deer Lodge area, Montana.
tion in sediments: Case study of the Gem Mountain Mine Montana Bureau of Mines and Geology Bulletin, Vol. 134.
placer sapphire deposit (Rock Creek, Montana, USA). Cana- ——— (2014) Sapphires in the southwestern part of the Rock

44 MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023
 Creek Sapphire District, Granite County, Montana. Montana kali-silica diagram. Journal of Petrology, Vol. 27, No. 3, pp. 745–
Bureau of Mines and Geology Bulletin, Vol. 135. 750, [Link]
——— (2015) Compilation of reported sapphire occurrences in Malaquias C. (2018) Lab Notes: A rare ruby from Montana. G&G,
Montana. Montana Bureau of Mines and Geology Report of In- Vol. 54, No. 4, pp. 434–435.
vestigation 23, 84 pp. Montana’s precious stones (1891) Dalles Daily Chronicle, Dalles,
——— (2022) Morphology of sapphires from secondary deposits, Oregon, October 13.
southwestern Montana. Montana Bureau of Mines and Geol- Moon A.R., Phillips M.R. (1991) Titania precipitation in sapphire
ogy Report of Investigation, Vol. 34. containing iron and titanium. Physics and Chemistry of Min-
Berg R.B., Dahy J.P. (2002) Montana sapphires and speculation on erals, Vol. 18, No. 4, pp. 251–258, [Link]
their origin. In P.W. Scott and C.M. Bristow, Eds., Industrial BF00202577
Minerals and Extractive Industry Geology. Geological Society, New sapphire fields (1892) Anaconda Standard, Anaconda, Mon-
London, pp. 199–204. tana, September 3, p. 8.
Berg R.B., Landry M.T. (2018) Sapphire deposits along the Missouri Notari F., Fritsch E., Caplan C., Hainschwang T. (2018) “Boehmite
River near Helena, Montana. Montana Bureau of Mines and needles” in corundum are Rose channels. G&G, Vol. 54, No.
Geology Bulletin, Vol. 136. 3, p. 257.
Berg R.B., Palke A.C. (2016) Sapphires from an Eocene sill near He- O’Donoghue M. (1995) Montana Sapphires. Unpublished manu-
lena, Montana. Proceedings of the GSA Annual Meeting, Den- script, Sevenoaks, England.
ver, p. 48, [Link] Palke A.C., Hapeman J.R. (2019) Gem News International: Rubies
Berger A.L., Berg R.B. (2006) The Silver Bow sapphire occurrence, from Rock Creek, Montana. G&G, Vol. 55, No. 2, pp. 286–288.
Montana: Evidence for a volcanic bedrock source for Montana’s Palke A.C., Renfro N.D., Berg R.B. (2017) Melt inclusions in allu-
alluvial sapphire deposits. Economic Geology, Vol. 101, No. 3, vial sapphires from Montana, USA: Formation of sapphires as
pp. 679–684, [Link] a restitic component of lower crustal melting? Lithos, Vol. 278,
Clabaugh S.E. (1952) Corundum Deposits of Montana. U.S. Geo- pp. 43–53, [Link]
logical Survey Bulletin 983, Washington, DC. Palke A.C., Saeseaw S., Renfro N.D., Sun Z., McClure S.F. (2019)
Dubinsky E.V., Stone-Sundberg J., Emmett J.L. (2020) A quantita- Geographic origin determination of blue sapphire. G&G, Vol.
tive description of the causes of color in corundum. G&G, Vol. 55, No. 4, pp. 536–579, [Link]
56, No. 1, pp. 2–28, [Link] Peucat J.J., Ruffault P., Fritsch E., Bouhnik-Le Coz M., Simonet
Eldorado Bar sapphires (1867) Tri-Weekly Post (Virginia City, Mon- C., Lasnier B. (2007) Ga/Mg ratio as a new geochemical tool
tana), August 27, p. 3, c. 2. to differentiate magmatic from metamorphic blue sapphires.
Emmett J.L., Douthit R. (1993) Heat treating the sapphires of Rock Lithos, Vol. 98, No. 1-4, pp. 261–274,
Creek, Montana. G&G, Vol. 29, No. 4, pp. 250–272, [Link]
[Link] Pratt J.H. (1906) Corundum and Its Occurrence and Distribution
Emmett J.L., Scarratt K., McClure S.F., Moses T., Douthit T.R., in the United States. U.S. Geological Survey Bulletin 180,
Hughes R., Novak S., Shigley J.E., Wang W., Bordelon O., Kane Washington, DC.
R.E. (2003) Beryllium diffusion of ruby and sapphire. G&G, Vol. Renfro N.D., Palke A.C., Berg R.B. (2018) Gemological character-
39, No. 2, pp. 84–135, [Link] ization of sapphires from Yogo Gulch, Montana. G&G, Vol. 54,
Fukatsu N., Kurita N., Oka Y., Yamamoto S. (2003) Incorporation No. 2, pp. 184–201, [Link]
of hydrogen into magnesium-doped α-alumina. Solid State Ion- Sangsawong S., Pardieu V., Raynaud V., Engniwat S. (2016) Gem
ics, Vol. 162-163, pp. 147–159, [Link] News International: “Punsiri”-type FTIR spectral features in
2738(03)00218-2 natural yellow sapphires. G&G, Vol. 52, No. 3, pp. 325–327.
Garland M.I. (2002) The alluvial sapphire deposits of western Sapphire mined on Rock Creek after 1892 discovery there (2019)
Montana. Ph.D. thesis, University of Toronto. Philipsburg Mail, Philipsburg, Montana, March 7, p. 6.
Hsu T., Lucas A., McClure S., Schumacher K. (2016) Montana’s Sapphire and Ruby Company prospectus (1891) St. James Gazette,
Gem Mountain sapphire mine: Sharing the excitement of London, October 29, p. 16.
mining with the public. GIA Field Report, June 3, Stone-Sundberg J., Thomas T., Sun Z., Guan Y., Cole Z., Equall R.,
[Link] Emmett J.L. (2017) Accurate reporting of key trace elements in
mountain-sapphire-mine ruby and sapphire using matrix-matched standards. G&G, Vol.
Hsu T., Lucas A., Kane R.E., McClure S.F., Renfro N.D. (2017) Big Sky 53, No. 4, pp. 438–451, [Link]
Country sapphire: Visiting Montana’s alluvial deposits. G&G, Vol. Struthers J., Fisher H. (1903) Gems and precious stones. In J.
53, No. 2, pp. 215–227, [Link] Struthers, Ed., The Mineral Industry: Its Statistics, Technology,
Hughes E.B., Perkins R. (2019) Madagascar sapphire: Low-temper- and Trade in the United States and Other Countries to the
ature heat treatment experiments. G&G, Vol. 55, No. 2, pp. End of 1902, Vol. 6, New York, The Engineering and Mining
184–197, [Link] Journal, New York and London, pp. 244–251.
Kane R.E. (2020) America’s royal gem: Montana and Yogo sap- Sutherland F.L., Schwarz D., Jobbins E.A., Coenraads R.R., Webb
phires. InColor, Vol. 45, pp. 30–39. G. (1998) Distinctive gem corundum suites from discrete basalt
Kunz G.F. (1885) Precious stones. In U.S. Geological Survey, Min- fields: A comparative study of Barrington, Australia, and West
eral Resources of the U.S. for 1883-1884, Washington, DC, pp. Pailin, Cambodia, gemfields. Journal of Gemmology, Vol. 26,
723–782. No. 2, pp. 65–85.
——— (1890) Gems and Precious Stones of North America. Sci- Zwaan J.C., Buter E., Mertz-Kraus R., Kane R.E. (2015) Alluvial
entific Publishing Company, New York. sapphires from Montana: Inclusions, geochemistry, and indi-
Le Bas M.J., Maitre R.L., Streckeisen A.L., Zanettin B.A. (1986) A cations of a metasomatic origin. G&G, Vol. 51, No. 4, pp.
chemical classification of volcanic rocks based on the total al- 370–391, [Link]

MONTANA SAPPHIRE FROM SECONDARY DEPOSITS GEMS & GEMOLOGY SPRING 2023 45
FEATURE ARTICLES

MORPHOLOGY OF COLOMBIAN EMERALD:

SOME LESS COMMON CASES AND THEIR
GROWTH AND DISSOLUTION HISTORY
Karl Schmetzer and Gérard Martayan

The morphology of Colombian emerald, including growth and dissolution features of their faces, gives insight
into growth and post-growth history of individual crystals. A collection of 15 isolated crystals and seven emeralds
in matrix was studied by optical methods. The authors observed layered growth of prismatic, dipyramidal, and
basal sectors that leads to crystals with prismatic to columnar or rarely pyramidal habit. In the latter case, pris-
matic growth sectors were not developed. The visual appearance of the crystal faces at the surface was influenced
by etch patterns with the formation of cavities, pits, pointed hillocks, and stepped grooves, followed occasionally
by layered overgrowth. Skeletal growth led to the formation of cavities in crystals, some of them resembling
empty cups with planar bottoms. In these crystals, the basal faces showed indentations or deep cavities, sur-
rounded by shells or rims of emerald, bound on both vertical surfaces by prism or dipyramidal faces. Emeralds
in the form of slightly conical empty tubes also belong to the latter group. Possible relationships to trapiche
emeralds and samples exhibiting the gota de aceite effect are discussed.

I
n general, the morphology of Colombian emerald slightly conical empty tubes (Klein, 1941)—or emer-
originating from different mines is rather simple alds in the form of prismatic, empty “cups” with pla-
and formed by a small number of external crystal nar bottoms (Weldon et al., 2016). In most references,
faces. We observe two dominant planes, the basal
pinacoid and the first-order hexagonal prism, occa-
sionally in combination with small second-order
hexagonal prism faces and first- and/or second-order
In Brief
hexagonal dipyramids (Goldschmidt, 1913; Schwarz • The surface texture of Colombian emerald crystals re-
and Giuliani, 2002; Moore and Wilson, 2016). The in- veals details about growth and post-growth history.
ternal growth pattern of such samples, which nor- • Several crystals show indications of natural etching in
mally show prismatic habit, consists of growth aggressive fluids, and some samples have undergone
planes parallel to the external crystal faces (Kiefert several subsequent growth and corrosion steps.
and Schmetzer, 1991). • In crystals with a conical shape, only basal and
Occasionally, natural emerald and beryl crystals dipyramidal growth sectors were developed.
show etching and dissolution features (an overview • Skeletal growth is observed in emeralds showing the
of the pertinent literature is given in box A). Other form of empty cups or emeralds with indentations on
the basal face.
growth features are due to skeletal and polygonal
growth of beryl crystals (see box B). Both growth fea- • Within the empty cups, polygonal growth of small
tures are related to the observations made in this ar- emerald columns is observed.
ticle for Colombian emerald crystals.
Rarely mentioned are emerald crystals with con-
ical habit (Johnson, 1961a,b)—or vasos in the form of such emeralds with conical habit (figure 1) or emer-
ald vasos or cups (figure 2) have been briefly men-
tioned or have only been pictured without
See end of article for About the Authors and Acknowledgments.
GEMS & GEMOLOGY, Vol. 59, No. 1, pp. 46–71,
considering a possible growth mechanism. In the fol-
[Link] lowing discussion, we will use only the term “cups”
© 2023 Gemological Institute of America for both slightly different variants.

46 MORPHOLOGY OF COLOMBIAN EMERALD GEMS & GEMOLOGY SPRING 2023

 Figure 1. This Colom-
bian emerald (sample
10, 25.0 mm in length)
shows conical habit in
the upper part and pris-
matic habit in the
lower part; the base is
covered with numerous
pointed hillocks. Photo
by G. Martayan.

MORPHOLOGY OF COLOMBIAN EMERALD GEMS & GEMOLOGY SPRING 2023 47

 BOX A: LITERATURE ON DISSOLUTION FEATURES OF BERYL AND EMERALD CRYSTALS

Etch pits on basal, prismatic, and pyramidal faces of might show etch pits, other faces of the same crystal are
Colombian emerald have been described or depicted by completely free of such textures (Petersson, 1889;
a few authors (e.g., Honess, 1917, 1929; Medina et al., Arzruni, 1894; Vrba, 1895; Tschermak, 1897; Kohlmann,
1983; Moore and Wilson, 2016). Considering the min- 1908; Sinkankas, 1981).
eral beryl in general, various forms of tiny cavities or These generally observed features of etch patterns in
pits have been observed with different shapes reflecting beryls from numerous localities are consistent with the
the symmetry of the individual basal, prismatic, or results of etching experiments performed in the labora-
dipyramidal crystal faces (figure A-1). In addition, the tory (Taube, 1895/1896; Feklichev, 1963). Such experi-
shape of etch patterns varies between samples from dif- mental processes were done in various acidic or basic
ferent localities (Kurumathoor and Franz, 2018). It has solutions, but normally only for periods of seconds to
also been mentioned that while some faces of crystals minutes.

m a

Figure A-1. Shape of etch

pits observed optically on
prismatic m and a faces,
on the basal pinacoid c
and on dipyramidal s
c
faces of numerous beryl
crystals from Brazil. After
Kohlmann (1908).

48 MORPHOLOGY OF COLOMBIAN EMERALD GEMS & GEMOLOGY SPRING 2023

 Examination at high magnification by optical and
electron microscopy reveals that the patterns of vari-
ously shaped cavities developed by natural or artificial
etching show stepped surfaces of various forms (Scandale
et al., 1990; Sunagawa and Urano, 1999; Sunagawa, 2003;
Demianets et al., 2006; Dem’yanets and Ivanov-Schitz,
2009; Kurumathoor and Franz, 2018).
All these observations, however, are different from
the effects of heavy dissolution, in which the crystals’
original surfaces are completely or partially dissolved,
forming more or less deep cavities or grooves (Ford, 1906;
Zedlitz, 1941; Bartoshinsky et al., 1969; Koivula, 1981;
Lyckberg et al., 2009; Tempesta et al., 2011). Such heavy
resorption has been mentioned for beryl from specific lo-
calities—e.g., California, Brazil, or Ukraine (figure A-2)—
and it has also been shown for emerald from Colombia
(figure A-3, left).
Furthermore, it should be mentioned that heavy etch-
ing can produce pointed forms consisting of single or mul-
tiple tapering tips representing the residue of the former Figure A-2. Yellow and greenish yellow beryl crystals
as-grown basal plane (figure A-3, right; see Penfield and from the Volodarsk mining area in Ukraine showing
Sperry, 1888; Penfield, 1890; Lacroix, 1896; Sunagawa, heavily dissolved surface patterns due to natural etching
2005; Moore and Wilson, 2016). This feature is also seen and dissolution subsequent to the crystal growth
in the heavily etched Colombian emerald shown in figure process. Crystal length: 61 mm (left) and 43 mm (right).
A-3, left. Photos by Peter Lyckberg.

Figure A-3. Left: A heavily etched Colombian emerald crystal from Muzo measuring 27 mm in length. The basal pinacoid
is completely dissolved, leading to numerous pointed hillocks; the growth sectors related to the a prism faces are optically
reflective. Photo by Jeff Scovil. Right: Etched beryl crystal from Pont de Barost, Haute-Vienne, France, showing a com-
pletely dissolved basal face with numerous pointed hillocks. From Lacroix (1896).

MORPHOLOGY OF COLOMBIAN EMERALD GEMS & GEMOLOGY SPRING 2023 49

 BOX B: LITERATURE ON SKELETAL AND POLYGONAL GROWTH OF BERYL CRYSTALS
From the literature, beryls with hexagonally outlined
rims or walls are known from different pegmatites.
Within these main walls, highly variable, irregularly
shaped to hexagonally outlined areas of beryl are found,
with other pegmatite minerals filling the remaining
space (figure B-1). These beryls, designated as “skeletal”
or “shell” crystals, mostly consist of a rim or shell en-
closing additional beryl shells or rims and other peg-
matite minerals (Hunt, 1892; Shaub, 1937; Johnston,
1945; Norton et al., 1962; Beus, 1966; Sinkankas, 1981).
It is not mentioned in the works cited whether the beryl
inside the dominant surrounding wall represents parts
of a single crystal, connected to the main wall or rim, or
independent individual crystals.
From the viewpoint of crystal growth, in skeletal
crystals with incomplete planes, crystal edges grow in
favor of plane faces (Sunagawa, 1981, 1999). This is of
course dependent on the growth environment. In the
flux synthesis of emerald, for example, it is possible to
grow skeletal crystals with depressions of the m prism
face (Oishi et al., 1994).
On the other hand, the polygonal growth of small
beryl columns on a basal plane of a larger crystal and the
subsequent overgrowth of this structure by later genera-
tions of beryl was described by Sahama (1966) and is
shown in figure B-2.
Visually, a beryl crystal depicted by Hills (1890) and
shown in figure B-3, seems to show both growth features: Figure B-1. Slices of three beryl crystals from Alto Cruzeiro,
skeletal and polygonal growth. In this sample, we ob- Paraíba, Brazil, with skeletal growth structure. Three slices
serve a rim surrounding a larger cavity. Within this cav- were cut from each sample oriented perpendicular to the c-
ity, we observe irregular walls or columns, which end at axis. Within an outer skeletal rim or shell of beryl, further
different heights in the cavity with small crystal faces. beryl crystals and other pegmatite minerals were observed.
From Johnston (1945).

Figure B-2. Sketch of a beryl crystal showing polygonal

growth of small individual pyramidal beryl crystals
ending in basal faces; this morphological structure is Figure B-3. Surface structure in a beryl from Mount An-
seen on the surface and within the crystal. In the latter tero, Colorado, with a rim surrounding a larger cavity
case, this area was overgrown in subsequent growth with irregular walls or columns developed within the
steps. From Sahama (1966). cavity. From Hills (1890).

50 MORPHOLOGY OF COLOMBIAN EMERALD GEMS & GEMOLOGY SPRING 2023

 Figure 2. In sample 17
(26 mm in length), a
base of black shale (not
visible) is covered with
numerous albite, cal-
cite, and pyrite crystals.
In this matrix, five
emerald crystals are
embedded, two of them
developed in the form
of empty cups. The
emerald cup on the
right is 5.5 mm long.
Photo by G. Martayan.

Of commercial interest, in addition to facetable emerald crystal with a horseshoe-shaped appearance

material without special structural properties, are is related to incomplete, nonuniform growth caused
samples with a fixed six-rayed pattern with or with- by mineral inclusions acting as growth obstacles.
out a central core—designated “trapiche” emerald The present paper tries to contribute to the un-
(Bernauer, 1926; Nassau and Jackson, 1970; Pignatelli derstanding of these less common morphological and
et al., 2015; Schmetzer, 2019; Smith, 2021) or sam- growth phenomena of Colombian emerald and offer
ples with an internal growth pattern named gota de descriptions and understanding of different patterns
aceite (Spanish for “drop of oil”) (Gübelin, 1944; mentioned, especially by enlarging the database for
Bosshart, 1991; Ringsrud, 2008; Hainschwang, 2008; such rare samples with remarkable morphological
Schmetzer, 2009; Gao et al., 2017). features. These samples provide a unique opportu-
Some of the morphological features mentioned nity for investigation, but their rarity limits the in-
above, such as the various forms of the trapiche pat- vestigation to nondestructive methods. In this way,
tern, have been studied in detail to evaluate the the paper would further contribute to our under-
growth mechanism. The gota de aceite effect has standing of the growth and post-growth history of
also been properly described, but its formation, either Colombian emerald.
by dissolution (etching) or polygonal growth and, in
both models, by subsequent overgrowth of the sur-
faces of the small hillocks or columnar crystals MATERIALS AND METHODS
formed in the first step, is still a matter of discussion The present study is based on examination of 15 iso-
(Bosshart, 1991). lated emerald crystals and seven specimens with
In a recent publication, Pignatelli et al. (2022) de- emerald crystals in mineral assemblages with albite,
scribed several Colombian emeralds with unusual calcite, dolomite, and pyrite on gray or black shales
habits. According to their results, the growth mecha- (table 1). All samples are from a private collection
nism of two emerald cups, one spongy emerald crystal, and were purchased within the last 20 years in Bo-
and one spear-shaped sample is mainly due to etching gotá or near emerald mines in Colombia. This does
(post-growth dissolution), and the morphology of an not necessarily indicate that all samples were un-

MORPHOLOGY OF COLOMBIAN EMERALD GEMS & GEMOLOGY SPRING 2023 51

TABLE 1. Properties of the examined emerald crystals.
Matrix
Sample no. specimen Emerald crystal Crystal facesc Surface features of
Weightb Remarks Figures
and locality dimensions dimensions (mm)a and habit emerald crystals
(mm)

Crystal on gray L 11.8, Associated minerals:

1: Chivor 22.7 g c, m; columnar Growth steps on m 4
shale, 40 × 28 D 3.5 dolomite, pyrite

Irregular openings in c, Cavities below c, larger

L 18.0, c, m, s;
2: La Pita — 18.61 ct etch pits and grooves than the different 5
D 11.1–12.0 columnar
on m openings

Crystal on c, m, a, p, s; Irregular openings in c,

L 23.6, Associated minerals:
3: Chivor black shale, 21.4 g elongated partly deeply etched m 6
D 5.1–15.5 albite, pyrite, calcite
41 × 30 tabular growth sectors

Partly deeply etched m

L 31.1, c, m, a; growth sectors; in other
4: Coscuez — 11.42 ct — 7
D 7.1–7.6 columnar parts, growth steps on
m

5: La Pita or L 10.0, Growth striations on the

— 0.67 ct c; conical — 8, A and B
Muzo D 4.0 sides of the pyramid

L 8.8, c; conical and Growth striations on

6: La Pita — 0.33 ct — 8A
D 2.0–2.1 prismatic part both parts

L 8.8, c; conical and Growth striations on

7: La Pita — 0.46 ct — 8, A and C
D 2.5–2.8 prismatic part both parts

c, m, a, s, p; Growth steps on
L 9.8, partly conical conical surface and on Groth tubes along the c-
8: Chivor — 1.61 ct 9
D 4.8–5.1 with tabular m faces of the tabular axis in tabular part
part part
Growth steps on m, Residual dark gray
L 20.8, c, m, a, s, p;
9: Muzo (?) — 10.76 ct finer structures on the (carbonaceous) material 10
D 10.8–12.1 conical
conical part in cavities

Growth steps, mainly

parallel to c, p on
L 25.0, m, a; partly Extensions of growth
conical surface; growth
10: La Pita — D (top) 15.0–16.6, 35.87 ct conical with sectors confined to the a 1 and 11
steps on m of the
D (bottom) 9.5–13.0 columnar part prism
columnar part; hillocks
on c

Indentations on c,
Crystal on gray L 15.5, c, m, a, u, p, s; Associated minerals:
11: Chivor 43.2 g growth steps on 12A
shale, 56 × 37 D 3.5–4.1 columnar calcite, pyrite
prismatic faces

aggregate of 11
Crystals on Indentations on c of all
parallel crystals; c, m, a; Associated minerals: 12B and 13,
12: Chivor dark gray shale, 50.0 g crystals, growth steps
largest L 12.5, columnar pyrite, dolomite, albite A and D
51 × 37 on prismatic faces
D 5.0–6.5

earthed within this period or in the last few years— According to the general knowledge of emerald
some of them might originate from Colombian col- formation in the various Colombian deposits, all of
lections, with samples being kept there for several the emeralds grew in cavities and were removed by
years or even decades before they were offered for sale. the miners, either as isolated crystals or in matrix.

52 MORPHOLOGY OF COLOMBIAN EMERALD GEMS & GEMOLOGY SPRING 2023

TABLE 1 (continued). Properties of the examined emerald crystals.
Matrix
Sample no. specimen Emerald crystal Crystal facesc Surface features of
Weightb Remarks Figures
and locality dimensions dimensions (mm)a and habit emerald crystals
(mm)
Three aggregates of
Crystals on nine, eight, or three Indentations on c of all Associated minerals:
c, m, a; 12C and 13,
13: Chivor gray shale, parallel crystals; 31.1 g crystals, growth steps calcite, dolomite, albite,
columnar B and C
44 × 29 largest L 16.5, on prismatic faces pyrite
D 6.5–7.5

Crystal on Deep indentations on c,

L 6.3, c, m, a; Associated minerals:
14: Chivor black shale, 14.6 g growth steps on 14
D 3.8–6.0 columnar calcite, albite, pyrite
49 × 22 prismatic faces

Deep indentations on c,
L 37.4, c, m, a, i, f, k;
15: Chivor — 17.24 ct growth steps on Color zoning 15
D 6.6–7.1 columnar
prismatic faces

Irregularly shaped
L 10.8, c, m, a, s, p; Open cup, growth steps
16: Chivor — 6.37 ct hillocks and columns in 16
D 7.3–8.7 columnar on prismatic faces
cup

cup 1: L 5.5,
D 2.5–3.0
cup 2: D 3.0 Associated minerals:
Five crystals
albite, calcite, pyrite;
on black shale; crystal 1: L 12.0, Open cups, growth
17: Chivor 4.5 g c, m; columnar irregularly shaped 2 and 17
26 long, 11–20 D 2.0 steps on m
hillocks and columns in
wide
crystals 2 and 3 both cups
are smaller
crystal 3: L 3.8

Open cup, openings Irregularly shaped

L 19.6, c, m, a, u, p, s;
18: Chivor — 13.58 ct also in p, growth steps framework of walls in 18
D 9.2–10.0 columnar
on prismatic faces cup

L 3.0 and 7.8, c, m, a, i, u p, s, Internal channels filled

19: Chivor — 2.68 ct Half cup 19
D 7.3–7.5 f, k; columnar with fine-grained beryl

Pyrite at the bottom of

L 13.0,
20: Chivor — 1.96 ct c, m, p Partial cup the cavity, internal 20, left
D 3.5–4.5
channels

L 9.0, Pyrite at the bottom of

21: Chivor — 1.41 ct c, m, a Partial cup 20, right
D 4.0–5.5 the cavity

c, m, a;
Slightly conical tube
L 14.4, elongated Etch pits and grooves
22: Chivor — 1.63 ct along the c-axis through 21 and 22
D 3.4–4.4 columnar, on prismatic faces
the complete crystal
slightly conical
a
L represents the length of the emerald crystal, D represents measurements of diameters between two m prismatic faces.
b
Weights of samples with emerald crystals in matrix are given in grams; weights of isolated emerald crystals are given in carats.
c
Based on a cell with a:c ratio of 1:0.996; basal pinacoid c {0001}, first-order hexagonal prism m {101̄0}, second-order hexagonal prism a {112̄0}, dihexagonal prism
i {213̄0}, first-order hexagonal dipyramids p {101̄2} and u {101̄1}, second-order hexagonal dipyramids s {112̄2} and f {336̄2}, and dihexagonal dipyramid k {213̄1}.

For the evaluation of growth history, it is impor- ally, and crystal faces were determined by goniomet-
tant to look at all morphological features of a sample ric measurements and/or by the measurement of an-
as a summary and not only at isolated crystal faces. gles between crystal faces in the microscope.
The morphology of the samples was examined visu- The samples were studied exclusively by nonde-

MORPHOLOGY OF COLOMBIAN EMERALD GEMS & GEMOLOGY SPRING 2023 53

 c c
A B
s

m a m

Figure 3. Morphology of
c c Colombian emeralds
C D observed in the present
study. A–D: Most sam-
s s ples show dominant m
p u p
prism and basal c faces;
occasionally we also
observed smaller a
m a m a prism faces in combina-
tion with s, p, and u
hexagonal dipyramids.
E: Furthermore, two of
the crystals showed ad-
ditional small i prism
faces in combination
with f and k dipyra-
mids. Drawings by K.
Schmetzer.

c
E

p k s
f

m a i

structive methods, especially by optical microscopy liquid. For such rare materials (isolated emerald crys-
at low magnification (up to 80×). A few transparent tals or matrix specimens), no destructive techniques
samples of appropriate size were also examined in (e.g., slicing or polishing of the crystals) could be ap-
immersion using benzyl benzoate as the immersion plied nor any coating of the samples for an investiga-

54 MORPHOLOGY OF COLOMBIAN EMERALD GEMS & GEMOLOGY SPRING 2023

tion at higher magnification in the scanning electron crystals showed small dihexagonal prism faces i
microscope. ˉ in combination with the second-order hexag-
{2130},
In one sample, the identity of the substances filling ˉ and the dihexagonal dipyra-
onal dipyramid f {3362}
the growth tubes was determined by a combination of ˉ
mid k {2131}. The faces determined on conical or at
X-ray diffraction (using a Bruker D8 Advance Eco X- least partly conical crystals were identical to those
ray diffractometer), energy-dispersive X-ray fluores- observed on all other emeralds.
cence (EDXRF, using a portable Bruker Tracer III-SD So far, the morphology is consistent with the pub-
EDXRF analyzer), and Raman spectroscopy (using an lished literature data above and information deter-
Ahura First Defender portable Raman device). Part of mined from photos of Colombian emerald crystals in
the filling could be removed easily with a needle and numerous publications. The two dipyramids f and k
was made available in that way for examination. are rarely found in Colombian samples, as previously
mentioned by Vrba (1881).
RESULTS AND DISCUSSION
Growth Features on Crystal Faces. Numerous
General Aspects. Some general aspects of crystal
Colombian emerald crystals from the present study
morphology will be described first, which can help
to characterize the individual samples. showed growth features on crystal faces, especially
elongated stepped surface structures on m prism
Morphology. Most isolated emerald crystals were faces (figure 4, A and B). These as-grown surface fea-
slightly distorted (e.g., with different diameters meas- tures represent layered growth in subsequent growth
ured between opposite prisms) but with clearly de- steps and resemble contour lines in topographic
fined faces. Nine of these crystals were columnar to maps. In sample 1 shown in figure 4, different layers
prismatic, and six were conical, some of them with are already seen with the unaided eye, but normally
prismatically developed areas. The emeralds on six the steps between subsequent layers are smaller and
of the seven matrix samples showed columnar to need examination by optical microscopy. To best ob-
prismatic habit, and only one crystal on matrix was serve the layered surface structures with optical mi-
tabular with significant differences of thickness be- croscopy, reflected light is ideal (figure 4C). Similar
tween prism faces in different directions. growth layers on the basal pinacoid of the majority
Idealized crystal drawings are presented in figure of Colombian emeralds, if present, are less pro-
3. Considering the columnar to prismatic crystals nounced and often not observable at the magnifica-
and the tabular crystals in our sample set, dominant tion applied in this study.
crystal faces were the first-order hexagonal prism m These observations are consistent with the exist-
ˉ and the basal pinacoid c {0001}. These were fre-
{1010} ing literature about growth mechanisms of beryl crys-
quently in combination with a smaller second-order tals. Similar stepped growth layers of surfaces of beryl
hexagonal prism a {1120}, ˉ and/or first-order hexago- crystal faces have been described by various authors
ˉ
nal dipyramids p {1012} and u {1011},ˉ and/or a second- (von Kokscharow, 1881; Himmel and Schmidt-Zittel,
ˉ
order hexagonal dipyramid s {1122}. In addition, two 1927; Griffin, 1951a,b; Seager, 1953; Grigor’ev, 1965;

Figure 4. A and B: In sample 1 (11.8 mm in length), stepped surface structures on different prismatic m faces of a
Colombian emerald crystal represent layered growth in subsequent growth steps. C: Best observation of such lay-
ered stepped growth structures is achieved in reflected light. Photos by G. Martayan (A and B) and K. Schmetzer
(C; field of view 6.0 mm).

A B C

MORPHOLOGY OF COLOMBIAN EMERALD GEMS & GEMOLOGY SPRING 2023 55

 A B C

Figure 5. Sample 2 (18.0 mm in length), a columnar emerald that has been heavily corroded by natural etching. A:
The basal plane shows irregularly shaped openings to larger cavities below this face. B: The m prism planes show
etch structures, especially irregularly oriented grooves between hillocks. C: Details of the etch structures with
grooves and hillocks, seen on the m prism face in image B, which are covered by micro-steps as seen in reflected
light. Photos by K. Schmetzer; field of view 3.5 mm (C).

Sinkankas, 1981; Sunagawa and Urano, 1999; Suna- Between the deep irregular grooves, small hillocks are
gawa, 2003; Demianets et al., 2006). observed that show micro-steps on the surface (figure
5C). These micro-steps resemble the pattern pro-
Characteristics of Individual Groups of Samples with duced as fine structure in etch pits by artificial disso-
Respect to Growth or Post-Growth Surface Patterns. lution of beryl crystals (see box A). It is unknown and
Emeralds with Dissolution Features. A group of must be examined in additional samples, which are
three samples with “common” habit (samples 2, 3, not available at the moment, whether this pattern
and 4) showed pronounced dissolution features with and the form of the irregularly shaped grooves are re-
partially dissolved (corroded) crystal surfaces. In the lated to the common inclusion features of Colombian
groups describing samples with extraordinary and emerald (e.g., feathers and healed fractures).
rare habits in the following sections, some emeralds Sample 3 with tabular habit shows a completely
that have undergone dissolution are also mentioned. and irregularly dissolved surface of the m prism, but
To understand the various observations, one general the smaller a prism is not corroded (figure 6A). Only
aspect should be mentioned at the beginning of this small parts of the original surface of the m prism are
section: Some of the crystals examined in this study still present, reflecting under appropriate illumina-
show strong etching and dissolution features only at tion (figure 6B). In this case, we can speak of a pre-
part of the surface. This indicates that only these ferred dissolution of the m prism compared to the a
parts of the emerald crystals were exposed to the ag- prism face. The basal face shows irregularly termi-
gressive dissolution fluid and the other parts of the nated openings to deep cavities, but part of this face
crystals were shielded from the fluid. It can be as- contains no dissolution features (figure 6C), and
sumed that such shielding processes were caused by therefore it might be concluded that this face was
different minerals of the assemblage found in Colom- shielded by minerals from the dissolution fluid.
bian emerald deposits, which were in close contact Sample 4 shows long prismatic habit with first-
with the emeralds’ as-grown surfaces (crystal faces). and second-order m and a prism faces (figure 3A). To-
In the present state, these minerals are at least par- ward one end of the crystal, the prism faces show
tially dissolved or broken away, now exposing the growth steps of an almost undistorted surface (figure
crystal faces of the emeralds for visual examination. 7A). This part represents about one-third of the
Sample 2 (figure 5A) with columnar habit (see fig- length of the crystal. The remaining part is exten-
ure 3B) shows distinct corrosion features. The basal sively etched, deeply corroded on m faces and to a
face reveals several irregularly shaped openings to a much lesser extent on a prism faces (figure 7, A and
cavity below this plane, which is larger than the open- B). In some areas, the m prism faces are completely
ings, extending widely into the crystal. Prism faces dissolved to an area with numerous grooves and
are completely covered by etch structures (figure 5B). hillocks. In contrast, we observe highly reflective

56 MORPHOLOGY OF COLOMBIAN EMERALD GEMS & GEMOLOGY SPRING 2023

 A B C

Figure 6. A: Sample 3 (23.6 mm in length), a heavily corroded emerald crystal with a completely dissolved surface
related to the m prism face, but with a small protruding area related to the a prism that is almost free of corrosion.
B: Only small areas of the original surface of the m prism are still reflective. C: The c basal pinacoid reveals irregu-
larly shaped holes leading to cavities below this face. Photos by G. Martayan (A and C), and K. Schmetzer (B);
field of view 7.6 mm (B).

ledges of the a prism protruding from the dissolved concluded that samples 2–4 have undergone a heavy
m growth sectors (figure 7B). This indicates that, as natural dissolution process in which the original sur-
also observed in sample 3, the a prism sectors were faces of the crystal faces were partially or completely
less intensely corroded. At the end, one basal pina- dissolved by natural etching. In other words, the pat-
coid is only influenced slightly or not at all by corro- terns observed in our Colombian samples are related
sion, while the other basal plane on the other end of to intense natural etching processes causing at least a
the crystal is heavily dissolved (figure 7, C and D). partial dissolution of the as-grown crystal faces and
Evaluation. Comparing these observations with creating deep cavities and irregular grooves between
descriptions from the known literature (box A), it is hillocks with stepped surfaces. All three crystals show

A B Figure 7. Sample 4 (31.1

m
mm in length), a
columnar emerald that
a has been heavily cor-
roded by natural etch-
m
ing. One end of the
crystal (A, upper left)
still shows natural crys-
tal faces, while the rest
of the sample shows
completely dissolved m
prism faces (A and B)
C D and highly reflective
but still undissolved
smaller a prisms (B). At
one end, the basal face
is only slightly corroded
(C), while the basal
face at the other end of
the crystal is heavily
corroded (D). Photos by
K. Schmetzer; field of
view 9.5 mm (C and D).

MORPHOLOGY OF COLOMBIAN EMERALD GEMS & GEMOLOGY SPRING 2023 57

 A

Figure 8. A: Emerald
crystals with pyramidal
or partly pyramidal
habit: sample 5 (left,
10.0 mm in length),
sample 7 (center, 8.8
mm in length), and
sample 6 (right, 8.8 mm
in length). B: Surface
texture of the pyrami-
dal emerald crystal
(sample 5) showing
growth lines related to
the basal pinacoid c
(center) and pyramidal
faces p (left and right).
C: Complex surface tex-
ture of a partly pyrami-
B C dal, partly prismatic
crystal (sample 7)
showing a complex pat-
p tern related to pyrami-
c
dal faces (lower part)
and to basal, pyrami-
p dal, and prism faces
p (upper part). The
growth direction of the
c crystals shown in A–C
is always from the bot-
tom to the top. Photos
by G. Martayan (A) and
K. Schmetzer (B and C);
fields of view 2.4 mm
(B) and 2.4 mm (C).
c
p

various dissolution features on different surfaces. Referring to the observations made in this study
Sample 4, for example, shows heavy dissolution from for samples 5–10, the authors must underscore that
one end with a dissolved basal c face to almost two- with the optical methods applied, we only can de-
thirds of its length, while the other end with basal c scribe the features representing the last stage of
face to about one-third of its length is barely affected. growth and/or dissolution.
In samples 2 and 3, it seems that the basal c faces were Three relatively small emerald crystals, desig-
partially shielded from the dissolution fluid. nated samples 5, 6, and 7, show the habit of a six-
An emerald with pointed forms (sample 10) sided pyramid or the habit of a six-sided pyramid
caused by natural etching will be described in the combined with a six-sided prism (figure 8A). The sur-
next section. face texture of the pyramidal sample 5 shows growth
steps related to basal and pyramidal growth layers
Emerald Crystals with Conical Habit or Conical (figure 8B). The basal growth layers are visible as stri-
Zones. Several samples showed, at least partly, a con- ations perpendicular to the c-axis, which indicate al-
ical habit, comparable to six-sided pyramids. ternating basal and dipyramidal faces. The surface

58 MORPHOLOGY OF COLOMBIAN EMERALD GEMS & GEMOLOGY SPRING 2023

texture of samples 6 and 7 is more complex. In the the m prism faces and s and p dipyramids (see figure
pyramidal parts, growth steps related to basal and py- 3C). Next to this end of the crystal, we observe a con-
ramidal layers are seen; in the prismatic parts, the ical part, followed by a tabular area. The diameter of
crystals show textures related to basal, dipyramidal, this tabular area is slightly larger than the crystal’s di-
and prism faces (figure 8C). These observations were ameter at the upper end. Growth steps are found on
confirmed in immersion with transmitted light (not the surfaces of the conical area (figure 9C) and on the
shown). In the conical parts of all three samples, no tabular part. Only the outer area of the tabular part
growth layers related to prism faces were developed. contains growth channels (hollow tubes) parallel to the
In contrast, the prismatic parts of samples 6 and 7 are c-axis. In transparent areas of the conical part, internal
quite “normal,” with growth zones related to basal, growth planes parallel to the basal pinacoid c and par-
pyramidal, and prismatic faces. allel to two hexagonal dipyramids p and u are observed
Sample 8 consists of a conical and a tabular part (fig- in immersion (figure 9D). These features indicate
ure 9). At the upper (conical) end of the crystal, we ob- growth in subsequent layers, but without development
serve a perfect basal face in combination with part of of prismatic growth sectors in this area of the crystal.

A B

Figure 9. A and B: Sam-

ple 8 (9.8 mm in length),
an emerald crystal with
conical shape and a tab-
ular part, with a pyrite
crystal attached to the
tabular part. C: Growth
steps on the surface of
C D the conical part of the
u crystal. D: Viewed in
immersion, internal
growth planes parallel
to the basal face c and
parallel to the hexago-
c nal dipyramids p and u,
as well as a three-phase
p inclusion with two salt
cubes and one gas bub-
ble in a liquid-filled
u cavity (arrow); the c-
axis runs vertically.
p c
Photos by K. Schmetzer;
fields of view 5.7 mm
(C) and 2.2 mm (D).

MORPHOLOGY OF COLOMBIAN EMERALD GEMS & GEMOLOGY SPRING 2023 59

 Sample 9 is of conical shape. At the upper (wide) where multiple small crystal faces covering different
end of the crystal, we observe a perfect basal face in hillocks reflect light simultaneously (figure 11D).
combination with part of the m prism faces and s and Evaluation. The growth history of natural beryl
p dipyramids (figures 3C and 10A). The m prism faces crystals frequently shows multiple subsequent
reveal growth steps (figure 10B), while the conical growth steps that might be separated by partial dis-
part is covered with finer surface structures. We ob- solution from natural corrosion and etching. The dif-
serve irregularly shaped grooves, hillocks covered by ferent steps of growth history of such samples have
micro-steps, and openings of deep cavities or inden- been characterized by a combination of optical and
tations (figure 10C). Comparing these surface tex- X-ray topography methods, especially for oriented
tures with those of corroded emerald samples 2, 3, slices of the original crystals (Scandale et al., 1990;
and 4, these parts of sample 9 were exposed to an ag- Sunagawa and Urano, 1999; Sunagawa, 2003).
gressive fluid causing dissolution. Within some of In our study, all samples of this group show one
the indentations mentioned, residual dark gray car- common morphological feature: a tapered appear-
bonaceous material is captured (figure 10D). ance, at least in parts of the crystal. This feature is
Sample 10 shows an even more complex mor- caused by a growth process in which no prismatic
phology. The crystal consists of a conical part that is growth sectors were developed during crystal growth.
followed by a prismatic part at the lower end of the This indicates that the conical appearance is caused
crystal (figures 1 and 11A). The prismatic part is nar- by basal and pyramidal sectors growing together,
rower than the conical part at its upper end. The sur- which are seen on the surface of the crystals in re-
faces on both parts show growth steps (figure 11, B flected light or in immersion in transmitted light. In
and C). Protruding from both parts are two areas with the conically developed sample 5, we observe only
extensions of growth sectors confined by the second- such basal and pyramidal growth layers. Samples 6
order hexagonal prism a, which are as thick as the and 7, in addition to a conical part, developed pris-
upper end of the conical part (figure 11A). The basal matic parts in later growth stages (see again figure 8).
pinacoid is covered by tiny pointed hillocks (figures Samples 8, 9, and 10 are emerald crystals with an
1 and 11D). By varying the crystal’s orientation with even more complex growth history. In all three sam-
respect to the light source, positions can be found ples we observe growth stages in which basal and

A B
Figure 10. A: Sample 9
s (20.8 mm in length), an
emerald crystal with coni-
cal shape, with basal pina-
coid c, small m prism
faces, and hexagonal s
dipyramids. B: Growth
steps on the surface of one
of the m prism faces,
viewed almost perpendicu-
lar to the c-axis. C:
Hillocks with fine micro-
C D steps between irregularly
shaped groves on the sur-
face of the conical part. D:
Residual dark gray car-
bonaceous material cap-
tured in indentations on
the surface of the conical
part. Photos by K. Schmet-
zer; fields of view 4.5 mm
(B), 3.5 mm (C), and 7.6
mm (D).

60 MORPHOLOGY OF COLOMBIAN EMERALD GEMS & GEMOLOGY SPRING 2023

 A conical
B
a Figure 11. A: Sample 10
(25.0 mm in length), an
prismatic emerald with conical habit
in the upper left and pris-
matic habit in the lower
right; the base is covered
with numerous hillocks; a
protrusion at the conical
part represents a growth
sector confined to the prism
a. B: Stepped surface texture
C D of the prismatic part. C:
Growth steps at the pris-
matic part (bottom) and the
conical part (top), and the c-
axis runs vertically. D: On
the basal face, the surfaces
of the hillocks consist of nu-
merous small faces that re-
flect light; viewed oblique
to the c-axis. Photos by K.
Schmetzer; fields of view
5.7 mm (B), 9.5 mm (C),
and 5.0 mm (D).

dipyramidal growth sectors were developed, but growth step (see references cited above). Skeletal
without development of related prismatic growth growth (see box B and examples below) with pointed
layers. These growth stages without formation of pyramids and sharp hillocks has never been men-
prismatic growth layers apply at least to part of the tioned in any beryl and is therefore considered very
crystal’s growth history. unlikely for this morphological structure.
In the tabular area of sample 8, we observe growth After corrosion, we observe several subsequent
channels parallel to the c-axis, which were not seen growth steps, in which an area with prismatic growth
in the tapered part of the crystal. The diameter of the layers and the conical zone with basal striations were
tabular area is larger than the diameter of the conical developed. Obviously, in one of these growth steps,
part at its upper end. This might indicate a later par- only basal growth layers without prismatic areas were
tial overgrowth of a crystal with conical habit and, formed. It is not completely understood whether the
in this later growth period, the development of a tab- prismatic m growth sectors in the lower part of the
ular area, on top of the tapered part of the crystal (see crystal were grown before or after the conical zone
again figure 9). with tapered surface of the crystal (see again figure 11).
Sample 9, at its wider end, has a short prismatic
area and a long, tapered cone. This cone shows dis- Emeralds with Incomplete Growth of the Basal
solution features on its surface. These features indi- Face. Several variants of incomplete growth of the
cate an etching process of the conical part after basal pinacoid were observed.
crystal growth and a subsequent growth step with Samples 11, 12, and 13 show incomplete growth
the formation of prismatic layers at the end of the of the basal face c, a pattern that could be described
crystal (see again figure 10). as resembling surface indentations. Sample 11 (figure
Sample 10 has likely undergone several growth 12A) is a single emerald crystal on matrix and reveals
and corrosion processes. The residual parts of pris- well-developed prism and pyramidal faces. Sample
matic a growth zones and the small pyramids or 12 (figure 12B) is an aggregate of approximately 11
hillocks forming the end of the conical part of the crystals, grown parallel to each other with somewhat
crystal indicate strong corrosion after an initial different lengths. All crystals of this aggregate show

MORPHOLOGY OF COLOMBIAN EMERALD GEMS & GEMOLOGY SPRING 2023 61

 A B C

Figure 12. Emeralds with surface indentations on the basal face, developed as either a single crystal (A: sample 11,
4.1 mm in largest diameter) or as aggregates of parallel emerald crystals (B and C: samples 12 and 13 with crystal
length up to 12.5 mm and 16.5 mm, respectively). Photos by G. Martayan.

incompletely developed c faces. Sample 13 (figure or other structures have different heights within the
12C) is similar to sample 12, but with three such ag- rim. In addition, all three samples have a step-like tex-
gregates consisting of approximately nine, eight, and ture on prism faces and no etch features. A represen-
three emerald crystals. All basal faces have a stepped tative example of that pattern is shown in figure 13D.
micro-texture consisting of numerous layers stacked Samples 14 and 15 show somewhat deeper inden-
parallel to the c-axis of the emerald crystal (figure 13, tations on the basal faces. Sample 14 has an almost
A–C). These structures occasionally end at the level continuous rim parallel to the basal face and inside
of the horizontal surface of the rim, but some columns several hillocks within the indentation zone (figure

A B

Figure 13. A–C: The

stepped micro-texture
of emeralds with in-
completely developed
basal planes shows
stacked layers parallel
to the c-axis. D: Growth
steps on the m prism
face of one of the crys-
C D tals in figure 12B.
Shown here are sample
12 (A and D) and sam-
ple 13 (B and C). Photos
by K. Schmetzer; fields
of view 11.5 mm (A),
7.6 mm (B and C), and
4.6 mm (D).

62 MORPHOLOGY OF COLOMBIAN EMERALD GEMS & GEMOLOGY SPRING 2023

 Figure 14. Left: Sample 14
(6.0 mm in largest diame-
ter), a crystal with an al-
most continuous rim
surrounding growth areas
ending in small basal faces,
but at a different height
along the c-axis compared to
the horizontal surface of the
rim. Right: Reflective areas
of the rim and the small
columns inside this rim end-
ing in tiny crystal faces par-
allel to the basal pinacoid.
Photos by G. Martayan (left)
and K. Schmetzer (right).

14, left). Part of these hillocks in the form of tiny first- and second-order prism faces in combination
columns end with faces parallel to the base (figure with first- and second-order hexagonal dipyramids
14, right), but lower than the horizontal surface of (figure 3C and figure 16, A and B). The m prism faces
the rim, which is formed by the incomplete basal are planar; no growth steps or etch features were vis-
face of the crystal. Sample 15, with similar appear- ible in the optical microscope with the magnification
ance, shows an almost colorless region that ends in applied. The wall thickness of the cup varies between
a basal face with indentations (figure 15A). The end 0.8 and 1.4 mm. Irregularly shaped hillocks or
of the crystal with these indentations is separated columns cover the bottom of the cup, which forms a
from the main part by a small zone with intense clear boundary with the other part of the crystal (fig-
green coloration (figure 15B). This demonstrates that ure 16C). In a view parallel to the c-axis, a micro-
at the beginning of the growth of this zone with a structure is observed forming a three-dimensional
length of about 1.5 mm along the c-axis, growth con- framework of walls, columns, and hillocks at the bot-
ditions were different compared to the earlier growth tom of the cavity (figure 16D). Part of the columns
stages. Hollow channels parallel to the c-axis are or irregular walls end in planar faces parallel to the
found in all parts of the crystal. The areas parallel to basal pinacoid (figure 16, A and B).
the basal face show growth steps (figure 15C). Sample 17 is a piece of black shale covered by al-
Several crystals were formed as “emerald cups.” bite, calcite, and pyrite crystals. Embedded in this
Sample 16 is a crystal with columnar habit in the matrix are five emerald crystals (figures 2 and 17A).
form of an open cup. The morphology consists of Two of these are columnar and hollow, forming deep

Figure 15. A: Sample 15 (37.4 mm in length), an emerald crystal with color zoning; at one end, the sample reveals a
continuous rim surrounding growth areas ending in small basal faces. B: Growth and color zoning (arrows) of the
crystal along the c-axis. C: Areas of the rim and small columns inside the rim with planes parallel to the basal pina-
coid appear bright in reflected light, viewed parallel to the c-axis. Photos by K. Schmetzer; field of view 3.6 mm (C).

A B C

MORPHOLOGY OF COLOMBIAN EMERALD GEMS & GEMOLOGY SPRING 2023 63

 A B
Figure 16. A: Sample 16
(10.8 mm in length), an
emerald crystal devel-
oped in the form of a
cup. B: The basal plane
of the cup reflects to-
gether with the basal
faces of small hillocks
or columns in the cup.
C: The bottom of the
cup is located in the
upper half of the crys-
C D tal, as indicated by ar-
rows. D: Framework of
irregular hillocks,
columns, and walls on
the bottom of the cup
in a view parallel to the
c-axis. Photos by G.
Martayan (A) and K.
Schmetzer (B–D); field
of view 4.6 mm (D).

cups with one open end (figure 17B). The cups have space of the cups with irregularly shaped planes.
a wall thickness of approximately 0.5 mm. The bot- Other planes at the ends of the tiny crystals in the
tom of each cup is covered with emerald hillocks or hollow cups reflect together with the residual basal
columns, which frequently end within the hollow face of the rims (figure 17C), indicating that some of

A B Figure 17. A: In sample 17 (26

mm long), a base of black
shale (not visible) is covered
with albite, calcite, and pyrite
crystals. In this matrix, five
emerald crystals are embed-
ded, two of them as deep cups
(center) and three as “nor-
mal” emeralds (lower right
and left). B: Details of the two
emerald cups (5.5 mm long on
the right). C: In reflected
light, the basal plane of one
cup appears bright together
with the basal faces of small
hillocks or columns in the
cup. D: Growth steps on the
C D E surface of one m prism face of
a cup. E: Growth steps on the
surface of the m prism face of
a “normal” emerald crystal.
Photos by K. Schmetzer (A,
C–E) and G. Martayan (B);
field of view 7.6 mm (C–E).

64 MORPHOLOGY OF COLOMBIAN EMERALD GEMS & GEMOLOGY SPRING 2023

 Figure 18. A and B: Sam-
A B
ple 18 (19.6 mm in
length), an emerald crys-
tal developed in the
form of a deep cup. B
and C: Openings at one
end are developed in the
basal pinacoid but also
in the largest p dipyra-
mid (arrows); within the
cup, a framework of ir-
regularly shaped thin
C D E emerald walls is present.
D: The crystal faces at
the inner surface of the
wall are oriented paral-
lel to the outer prism
faces. E: Growth steps
on the surface of an m
prism. Photos by K.
Schmetzer; fields of view
c
11.5 mm (D, vertical)
and 5.1 mm (E).

the hillocks or columns also end in faces parallel to the basal pinacoid, but holes are also developed in the
the basal pinacoid. The m prism faces of the cups largest p dipyramid (figure 18C). The cup’s wall
show growth steps (figure 17D). thickness ranges from 1.5 to 2.0 mm. Within the cup,
The three other emerald crystals are “common” a framework of irregularly shaped thin emerald walls
columnar crystals with prismatic and basal faces (fig- is present (figure 18, B and C). The main walls of the
ure 17A), which also show growth steps on m prism inner surface of the rim are oriented parallel to prism
faces and no etch features (figure 17E). faces (figure 18D). The outer m prism faces are cov-
Sample 18 is the largest crystal, in the form of a ered by growth steps (figure 18E).
deep open cup (figure 18, A and B). Morphologically, Sample 19 is considered to be a half cup. Formally,
the columnar and somewhat distorted crystal shows the sample is described as a crystal formed of two
several prism faces and hexagonal dipyramids (see parts. The first part shows platy habit, with two basal
figure 3, C and D). The main opening at one end is in c faces and m, a, and i prism faces (figure 19A). On

Figure 19. A: Sample 19 (7.8 mm in length), an emerald crystal in the form of a half cup developed with two parts:
as a platy part with lower and upper basal pinacoid (white arrows) and an upper curved rim surrounding about
half of the area of the platy part; this rim is terminated by a small basal face (red arrow). B: Stepped surface of the
basal face on top of the horizontal surface of the rim. C: View of the inner surface of the circular rim on top of the
platy part, in this orientation with reflective steep k hexagonal dipyramids, the elongated cavities are filled with
fine-grained white beryl. Photos by G. Martayan (A) and K. Schmetzer (B and C); field of view 14.5 mm (B).

A B C
k

rim rim
rim
rim

beryl
platy
platy
platy platy

MORPHOLOGY OF COLOMBIAN EMERALD GEMS & GEMOLOGY SPRING 2023 65

the upper side of the first part of the crystal, a curved tified by a combination of X-ray diffraction, EDXRF,
rim is formed that encircles about half of the area and Raman spectroscopy).
formed by the prism faces of the first part. This rim Samples 20 and 21 are also incomplete cups with
is considered to be the second part of the crystal. On only partially developed walls or rims. In both sam-
top of this rim, we observe a small area with a ples, the bottom of the internal cavity is filled with
stepped surface in an orientation parallel to the basal pyrite crystals, surrounded by an incomplete circle
pinacoid (figure 19B). We can consider this to be the of walls (figure 20). In sample 20, the outer m prism
end of the second part of the crystal. faces show growth striations parallel to the c-axis,
The other faces of the outer surface of the rim are and the internal surface of the walls displays open
the already mentioned prism faces m, a, and i in com- channel structures, comparable to the open channels
bination with u, p, s, f, and k dipyramids. Not only are observed in sample 19. The outer and inner surfaces
the f and k dipyramidal faces rarely observed in of the rim in sample 21 are completely flat.
Colombian emerald, but they are also rarely observed
in the mineral beryl as a whole. These faces are steeper Emerald Crystal with an Internal Channel. Sample
than the commonly observed u, p, and s dipyramids 22 is an elongated tube that is open at both ends. From
and are inclined 18.2° and 18.5° to the c-axis. The open top to bottom, the external form of the tube is slightly
structure of the rim allows us to determine the faces conical, and the channel inside it is also wider at one
at the inner surface of this rim, which are the common end, following the external form of the tube (figure 21).
prism faces in combination with the two steep dipyra- The tube was cut at the thinner bottom of the sample,
mids f and k (figure 19C). The open cavities formed at showing a hole with a small diameter (figure 21, left).
the inner surface of this rim, elongated parallel to the This indicates that the original crystal might have
c-axis, are partially filled with a white fine-grained ma- been closed at its end. At the other end with the larger
terial, which was determined to consist mainly of opening, the thickness of the tube walls ranges from
beryl with small admixtures of quartz and albite (iden- about 0.2 to 0.6 mm, with an opening between 2.5 and

Figure 20. Emerald crys-

tals in the form of in-
complete cups with
pyrite crystals at the
bottom of the cavities,
which are surrounded
by incomplete rims.
Sample 20 on the left is
13.0 mm in length, and
sample 21 on the right
is 9.0 mm in length.
Photos by G. Martayan.

66 MORPHOLOGY OF COLOMBIAN EMERALD GEMS & GEMOLOGY SPRING 2023

 Figure 21. Emerald crys-
tal (sample 22, 14.4 mm
in length) in the form of
a slightly conical hol-
low tube (left), with
openings at both ends
(right). The lower end of
the crystal (right, inset)
was cut artificially and
shows a hole (arrow).
Photos by K. Schmetzer.

3.5 mm in diameter. The diameter of the channel The rims or walls end with faces oriented perpen-
through the crystal at the other end is approximately dicular to the c-axis, and the outer surface of the
1.0 mm. Outside, the tube is terminated by m prism rims are formed by various prismatic and pyramidal
faces, which show etch features—i.e., the original sur- faces. The inner surface of the walls or rims are
face of the tube is partially dissolved (figure 22). The formed by the same faces, mainly first- and second-
horizontal part of the wall or rim oriented parallel to order prism faces and occasionally in combination
the basal pinacoid (see figure 21B) reveals small pits, with steep dipyramids. This means that the outer
most likely due to etching. and the inner outlines of the skeletal beryl walls are
Evaluation. Samples 11–21, all originating from parallel to prismatic and steep dipyramidal faces.
Chivor, show the same phenomenon: incomplete Within these depressions or cavities, we observe ir-
growth of the basal face, but with a different degree regular walls or columns, which end at different
of development. In general, our Colombian samples heights in the cavities with small basal faces. In
display an upper rim surrounding indentations (see samples 15 and 16, it is clearly indicated that the
again figures 12 and 13) and somewhat deeper depres- inner bottom of the cavities is confined to a zone of
sions (figures 14 and 15) or somewhat higher walls or strong color zoning, which indicates a change of
rims surrounding deeper cavities (figures 16–20). In growth conditions at this stage of crystal growth (see
other words: The patterns observed in cups of various again figures 15 and 16).
depths in samples 16–21 or in emeralds with deeper Considering the patterns described for beryl with
indentations of basal faces in samples 14 and 15 re- skeletal and polygonal growth (box B), we conclude
flect deeper indentation patterns on basal planes than that the morphology seen in Colombian emerald
those observed in samples 11–13. samples 11–19 can be understood as a combination

Figure 22. The surface of

sample 22 is partially dis-
solved by corrosion—i.e.,
no clear reflective plane is
seen. The same area of the
crystal is shown in trans-
mitted light (left) and in
reflected light (right). Pho-
tos by K. Schmetzer; field
of view 9.5 mm.

MORPHOLOGY OF COLOMBIAN EMERALD GEMS & GEMOLOGY SPRING 2023 67

of both growth phenomena: skeletal growth of the All these features indicate that the described mor-
walls and polygonal growth of columns and walls in- phology of this group of crystals cannot be explained
side the cavities. Or we could describe the pattern simply by etching processes of the basal face. It can-
observed as the growth of a dominant wall and the not be assumed that the different morphological
growth of small columns or pyramids within the area forms described are due to etching, starting at one
surrounded by the rim. Samples 20 and 21 clearly planar basal face and leaving the rest of the crystal,
show inhibited growth of emerald (skeletal growth) especially all prism faces, without any clear sign of
due to pyrite obstacles, forming cavities with incom- etching.
plete walls.
Sample 22, the elongated emerald tube, also origi- Comparison with Properties Observed in Cut Stones.
nates from Chivor. Considering samples 11–21 and It has already been mentioned in various studies that
the heavily corroded samples 2–4, our understanding the heavily included areas between the transparent
of etch patterns in beryl (summarized in box A), and basal and prismatic growth sectors of Colombian
our knowledge about the surface structure of non-cor- trapiche emerald are intensely corroded (e.g.,
roded emerald crystals (sample 1, figure 4, and other Bernauer, 1926; Nassau and Jackson, 1970; Pignatelli
examples in this paper), we interpret the growth his- et al., 2015). In general, Colombian trapiche emeralds
tory of this sample as skeletal growth followed by consist of a basal growth sector in the form of a ta-
etching and partial dissolution of the outer and inner pered core and six prismatic growth sectors. The core
faces of the crystal, i.e. the outer and inner surfaces of and the six prismatic growth sectors and each of the
the walls, but with no regrowth after etching. prismatic sectors are separated from others by areas
Pignatelli et al. (2022) explained the growth of with high concentration of mineral inclusions and
several Colombian emeralds developed as cups with carbonaceous material. The basal growth sector
conical or prismatic shape simply as etching mainly consists of basal and occasionally of pyrami-
processes, without taking skeletal growth into con- dal growth layers, but without any prismatic growth
sideration. While we take into consideration a com- areas.
bination of skeletal growth and post-growth etching Corrosion, in general, commences at crystal de-
for one sample (see description and growth evalua- fects that could be lattice defects or areas at trapped
tion of sample 22), we cannot support the idea that inclusions. Sample 9 described here might represent
all growth structures and morphological patterns of the final stage of such corrosion processes. The general
samples 11–22 are due to “simple” etching processes. appearance is reminiscent of the central tapered core
We conclude that, for samples 11–21, they are not of trapiche emerald (figure 23). Furthermore, we ob-
even partly related to etching processes. served etch structures and residual dark gray carbona-
To support our conclusions with further argu- ceous material. This indicates that the tapered sample
ments, we observed several growth features in this examined in this paper might be the core of a trapiche
group of samples that are not consistent with the emerald that was separated from its prismatic growth
idea of etching, starting at an as-grown basal plane. zones by etching processes. If this interpretation is cor-
In other words: rect, the prismatic growth layers at the upper end of
the crystal (see figure 10B) might be understood as
• Etching would not create internal prism faces,
later overgrowth of the tapered crystal.
which form the dominant pattern of inner sur-
The irregularly shaped hillocks, columns, or walls
face of the circular rims.
within the cavities of the samples with surface in-
• Etching would not stop and form the bottom of dentations on basal faces—i.e., the emerald cups de-
cavities just at the height of a strong color zon- scribed in samples 11–19—resemble the internal
ing along the c-axis of the crystals. growth structure found in the so-called gota de aceite
• Etching would form stepped surfaces and cavi- emeralds (figure 24; e.g., Bosshart, 1991; Ringsrud,
ties, but in our samples we observe instead the 2008; Hainschwang, 2008; Schmetzer, 2009; Gao et
presence of small basal faces at the end of pyr- al., 2017). An overgrowth of the polygonal patterns
amids or columns of different height within the within such a crystal, especially in crystals with shal-
cavities. lower indentations as seen in samples 11–13 (figures
• The layered appearance of the surface of the 12 and 13), in further growth steps would result in
rim and the m prism faces of the crystals does emerald crystals with the characteristic gota de
not reflect etching structures. aceite structural inclusion pattern. It is plausible

68 MORPHOLOGY OF COLOMBIAN EMERALD GEMS & GEMOLOGY SPRING 2023

 SUMMARY AND CONCLUSIONS
Most Colombian emeralds show a columnar to pris-
matic habit with growth steps or growth layers on
prism faces, which can be seen occasionally even
with the unaided eye. Several groups of samples with
specific morphological patterns on the surface have
been studied and described. However, some of the
samples from our study theoretically could be
grouped into several different categories, simply be-
cause they have undergone several subsequent
growth stages and dissolution processes (e.g., sample
10). Because optical observation allows us to exam-
ine only the last result of such multistep growth
processes, not all morphological features of the emer-
ald crystals in the post-grown and post-dissolved
state can be completely understood.
Heavily etched crystals show dissolved m prism
sectors, whereas the a prism zones are less affected by
the dissolution process. The deep cavities or grooves
seen in the post-growth dissolved state have a stepped
microstructure on m prism faces. The basal pinacoid,
in some cases, reveals irregularly shaped openings to
cavities below that face, but this face may also be dis-
solved in a way to show tiny pointed hillocks. Because
the degree of etching varies within a single emerald
Figure 23. Left: Diagram of a trapiche emerald con- crystal, it is concluded that such samples were par-
sisting of two tapered cores representing basal growth
tially shielded from contact with other minerals of the
sectors (green) and six prismatic growth sectors (after
Schmetzer, 2019). Right: The outline of sample 9 con-
natural assemblage from the etching fluid.
forms to the shape of one of the basal growth sectors Conically developed emerald crystals, in general,
in trapiche emerald. show only growth layers related to basal and dipyra-
midal growth faces, but no prismatic growth sectors
were developed in such crystals. The morphology of
that, with changing environmental growth condi- such samples resembles the basal growth sectors of
tions, such further growth steps could take place. A Colombian trapiche emerald.
general scenario with a multistep growth history has Emerald cups showing a skeletal rim with pris-
already been proposed by various authors (e.g., Gü- matic or dipyramidal faces forming its outer and inner
belin and Koivula, 2008; Ringsrud, 2009; Schwarz surface reveal more or less deep indentations on the
and Curti, 2020; Sun and Goa, 2022). basal face. In some cases, the walls or rims form deep

Figure 24. Patterns seen in

faceted Colombian gota de
aceite emeralds display nu-
merous irregular hillocks,
columns, or walls. The left
image is viewed in immer-
sion, parallel to the c-axis.
Photomicrographs by K.
Schmetzer (left) and M.P.H.
Curti, Bellerophon Gemlab
(right); fields of view 1.9 mm
(left) and 1.6 mm (right).

MORPHOLOGY OF COLOMBIAN EMERALD GEMS & GEMOLOGY SPRING 2023 69

cavities of prismatic or conical cups or even elongated face structures are overgrown subsequently with new
central tubes within the crystal. Within the cavities emerald, the final result would show a crystal with
of these cups, starting from a basal plane growth layer, an internal growth structure containing an inclusion
we observe tiny columns or pyramidal crystallites, pattern resembling the pattern described as gota de
which frequently end in small basal faces. If such sur- aceite in Colombian emerald.

ABOUT THE AUTHORS ACKNOWLEDGMENTS

Dr. Karl Schmetzer is an independent researcher living in Peters- The authors are grateful to Professor H.A. Gilg and N. Preisinger
hausen, near Munich. Gérard Martayan is a senior geophysicist for their help with X-ray diffraction, X-ray fluorescence, and
and longtime emerald aficionado residing in Paris. Raman spectroscopy to identify the cavity fillings in one sample.
We are also grateful for numerous helpful comments and sug-
gestions of the three peer reviewers of this manuscript.

REFERENCES
Arzruni A. (1894) Ein Beryllkrystall mit rhomboëdrischer Ausbil- [Link]
dung. Verhandlungen der Russisch-Kaiserlichen Mineralogis- Grigor’ev D.P. (1965) Ontogeny of Minerals. Israel Program for Sci-
chen Gesellschaft zu St. Petersburg, Zweite Serie, Vol. 31, pp. entific Translations Ltd., Jerusalem.
155–160. Gübelin E.J. (1944) Gemstone inclusions. G&G, Vol. 4, No. 12, pp.
Bartoshinsky Z.V., Matkovsky O.I., Srebrodolsky B.I. (1969) Ac- 174–179.
cessory beryl from chambered pegmatites of the Ukraine. Min- Gübelin E.J., Koivula J.I. (2008) Photoatlas of Inclusions in Gem-
eralogical Sbornik (Mineralogicheskii Sbornik), Vol. 23, No. 4, stones, Volume 3. Opinio Publishers, Basel, Switzerland.
pp. 382–397 [in Russian]. Hainschwang T. (2008) Extraordinary “gota de aceite” emerald
Bernauer F. (1926) Die sog. Smaragddrillinge von Muzo und ihre submitted to the lab. GEMLAB Research Newsletter, Vol. 6.
optischen Anomalien. Neues Jahrbuch für Mineralogie, Geolo- Hills R.C. (1890) Etched beryls from Mount Antero, Colorado. Pro-
gie und Paläontologie, Supplemental Volume 54, pp. 205–242. ceedings of the Colorado Scientific Society, Vol. 3, No. 2, pp.
Beus A.A. (1966) Geochemistry of Beryllium and Genetic Types of 191–192.
Beryllium Deposits. W.H. Freeman and Company, San Francisco. Himmel H., Schmidt-Zittel H. (1927) Wachstumsakzessorien am
Bosshart G. (1991) Emeralds from Colombia (Part 2). Journal of Beryll. Zentralblatt für Mineralogie, Geologie und Paläontolo-
Gemmology, Vol. 22, No. 7, pp. 409–425. gie, Abt. A., 1927, pp. 118–125.
Demianets L.N., Ivanov-Shitz A.K., Gainutdinov R.V. (2006) Hy- Honess A.P. (1917) On the etching figures of beryl. American Jour-
drothermal growth of beryl single crystals and morphology of nal of Science, Vol. 43, No. 255, pp. 223–236.
their singular faces. Inorganic Materials, Vol. 42, No. 9, pp. ——— (1929) The theory of crystal etching and its significance in
989–995, [Link] the classification of crystals. Proceedings of the Pennsylvania
Dem’yanets L.N., Ivanov-Schitz A.K. (2009) Beryl: Regeneration Academy of Science, Vol. 3, pp. 52–59.
crystal growth and morphology of regeneration surfaces. Hunt T.S. (1892) Systematic Mineralogy Based on a Natural Clas-
Journal of Surface Investigation. X-ray, Synchrotron and sification, 2nd ed. The Scientific Publishing Co., New York.
Neutron Techniques, Vol. 3, No. 6, pp. 881–887, Johnson P.W. (1961a) The Chivor emerald mine. Journal of Gem-
[Link] mology, Vol. 8, No. 4, pp. 126–152.
Feklichev V.G. (1963) Microcrystalline morphology and investiga- ——— (1961b) All about emeralds – natural or synthetic. Lapidary
tion of the phenomena of solution of beryllium crystals. Trudi Journal, Vol. 15, No. 1, pp. 118–131.
Institut Mineralogii, Geochimii i Kristallochimii Redkich Ele- Johnston W.D. (1945) Beryl-tantalite pegmatites of Northeastern
mentov, Akademija Nauk SSSR, Vol. 18, pp. 85–106 [in Russian]. Brazil. Bulletin of the Geological Society of America, Vol. 56,
Ford W.E. (1906) Some interesting beryl crystals and their associ- No. 11, pp. 1015–1070, [Link]
ations. American Journal of Science, Vol. 22, No. 129, pp. 217– 7606(1945)56[1015:BPONB][Link];2
223. Kiefert L., Schmetzer K. (1991) The microscopic determination of
Gao Y., Ng M., Carmona C., Lin Q. (2017) “Gota de aceite” effect structural properties for the characterization of optical uniaxial
in emerald: Observation and cause. Gems & Technology Con- natural and synthetic gemstones. Part 2: Examples for the ap-
ference, Beijing, pp. 95–97. plicability of structural features for the distinction of natural
Goldschmidt V. (1913) Atlas der Krystallformen, Band I. Carl emerald from flux-grown and hydrothermally-grown synthetic
Winters. Universitätsbuchhandlung, Heidelberg, Germany. emerald. Journal of Gemmology, Vol. 22, No. 7, pp. 427–438.
Griffin L.J. (1951a) LXXIX. Microscopic studies on beryl crystals.– Klein F. (1941) Smaragde unter dem Urwald. Oswald Arnold Ver-
I. Observation of uni-molecular steps. The London, Edinburgh, lag, Berlin.
and Dublin Philosophical Magazine and Journal of Science, Kohlmann H. (1908) Beiträge zur Kenntnis des brasilianischen
Vol. 42, No. 330, pp. 775–786, [Link] Berylls. Neues Jahrbuch für Mineralogie, Geologie und Paläon-
14786445108561306 tologie, Vol. 25 Supplement, pp. 135–181.
——— (1951b) CXXXIII. Microscopic studies on beryl crystals.– Koivula J.I. (1981) Etch figures on beryl. Journal of Gemmology,
II. Dislocations and the growth of {101̄0} prism faces. London, Vol. 21, No. 3, pp. 142–143.
Edinburgh, and Dublin Philosophical Magazine and Journal von Kokscharow N. (Sohn) (1881) Beryll-Krystalle eines neuen
of Science, Vol. 42, No. 335, pp. 1337–1352, Fundortes. Bulletin de l’Académie Impériale des Sciences de

70 MORPHOLOGY OF COLOMBIAN EMERALD GEMS & GEMOLOGY SPRING 2023

 St-Pétersbourg, Vol. 27, No. 1-7, pp. 35–38. ——— (2019) Gem News International: Trapiche emerald from
Kurumathoor R., Franz G. (2018) Etch pits on beryl as indicators of Colombia. G&G, Vol. 55, No. 1, pp. 156–158.
dissolution behaviour. European Journal of Mineralogy, Vol. 30, Schwarz D., Curti M. (2020) Emerald: Modern Gemmology.
No. 1, pp. 107–124, [Link] Bellerophon Gemlab Ltd, Paris and Bangkok.
2703 Schwarz D., Giuliani G. (2002) South America: Colombia. In G. Giu-
Lacroix A. (1896) Minéralogie de la France et de ses colonies. Vol. liani et al., Eds., Emeralds of the World. extraLapis English No.
2, Librairie Polytechnique, Paris, pp. 8–22. 2. Lapis International, East Hampton, Connecticut, pp. 36–45.
Lyckberg P., Chornousenko V., Wilson W.E. (2009) Famous mineral Seager A.F. (1953) The surface structure of crystals. Mineralogical
localities: Volodarsk-Volynski, Zhitomir Oblast, Ukraine. Min- Magazine, Vol. 30, No. 220, pp. 1–25, [Link]
eralogical Record, Vol. 40, No. 6, pp. 473–506. minmag.1953.030.220.02
Medina J.A., Morante M., Leguey S. (1983) Natural etch pits in beryl Shaub B.M. (1937) Contemporaneous crystallization of beryl and
related with the structure. Bulletin de Minéralogie, Vol. 106, No. albite vs. replacement. American Mineralogist, Vol. 22, No. 10,
3, pp. 293–297, [Link] pp. 1045–1051.
Moore T.P., Wilson W.E. (2016) The emerald mines of Colombia. Sinkankas J. (1981) Emerald and Other Beryls. Chilton Book
Mineralogical Record, Vol. 47, No. 1, pp. 5–68. Company, Radnor, Pennsylvania.
Nassau K., Jackson K.A. (1970) Trapiche emeralds from Chivor Smith J.S. (2021) From curio to designer gem…Rise of the trapiche.
and Muzo, Colombia. American Mineralogist, Vol. 55, No. 3- Gems & Jewellery, Vol. 30, No. 4, pp. 26–29.
4, pp. 416–427. Sun X., Gao Y. (2022) Gem Notes: Hexagonal growth structures
Norton J.J., Page L.R., Brobst D.A. (1962) Geology of the Hugo peg- displaying gota de aceite effect in Colombian emerald. Journal
matite Keystone, South Dakota. U.S. Geological Survey Pro- of Gemmology, Vol. 38, No. 1, pp. 11–12.
fessional Paper 297-B, pp. 49–127. Sunagawa I. (1981) Characteristics of crystal growth in nature as
Oishi S., Mochizuki K., Hirano S. (1994) Growth of emerald crys- seen from the morphology of mineral crystals. Bulletin de
tals by the flux evaporation method in MoO3-B2O3 system. Jour- Minéralogie, Vol. 104, No. 2-3, pp. 81–87,
nal of the Ceramic Society of Japan, Vol. 102, No. 1185, pp. [Link]
502–504, [Link] [in Japanese] ——— (1999) Growth and morphology of crystals. Forma, Vol. 14,
Penfield S.L. (1890) Some observations on the beryllium minerals pp. 147–166.
from Mt. Antero, Colorado. American Journal of Science, Vol. ——— (2003) Growth histories of mineral crystals as seen from
40, No. 240, pp. 488–491. their morphological features. In K. Byrappa and T. Ohachi, Eds.,
Penfield S.L., Sperry E.S. (1888) Mineralogical Notes: Beryl. Amer- Crystal Growth Technology. Springer-Verlag, Berlin, Germany,
ican Journal of Science, Vol. 36, No. 215, pp. 317–320. pp. 1–23.
Petersson W. (1889) Om naturliga etsfigurer och andra lösnings- ——— (2005) Minerals formed by vapor growth. In Crystals:
fenomen på beryll från Mursinsk. Bihang till Kongl. Svenska Growth, Morphology, and Perfection. Cambridge University
Vetenskaps-Akademiens Handlingar, Afd. II, Vol. 15, No. 1, Press, Cambridge, UK, pp. 236–250.
pp. 1–38. Sunagawa I., Urano A. (1999) Beryl crystals from pegmatites: Mor-
Pignatelli I., Giuliani G., Ohnenstetter D., Agrosi G., Mathieu S., phology and mechanism of crystal growth. Journal of Gemmol-
Morlot C., Branquet Y. (2015) Colombian trapiche emeralds: Re- ogy, Vol. 26, No. 8, pp. 521–533.
cent advances in understanding their formation. G&G, Vol. 51, Taube H. (1895/1896) Ueber die Aetzfiguren einiger Minerale.
No. 3, pp. 222–259, [Link] Neues Jahrbuch für Mineralogie, Geologie und Paläontologie,
Pignatelli I., Giuliani G., Morlot C., Salsi L., Martayan G. (2022) Supplemental Volume 10, pp. 454–469.
Colombian emerald oddities: Review and formation mecha- Tempesta G., Scandale E., Agrosi G. (2011) Striations and hollow
nisms. Journal of Gemmology, Vol. 38, No. 1, pp. 26–43. channels in rounded beryl crystals. Periodico di Mineralogia,
Ringsrud R. (2008) Gota de aceite: Nomenclature for the finest Vol. 79, No. 1, pp. 75–87, [Link]
Colombian emeralds. G&G, Vol. 44, No. 3, pp. 242–245, Tschermak G. (1897) Lehrbuch der Mineralogie. 5th edition, Al-
[Link] fred Hölder, Vienna, p. 147.
——— (2009) Emeralds: A Passionate Guide. Green View Press, Vrba C. (1881) Smaragd von Sta Fé de Bogota. Zeitschrift für Krys-
Oxnard, California. tallographie und Mineralogie, Vol. 5, No. 5, pp. 430–432.
Sahama T.G. (1966) Polygonal growth of beryl. Bulletin of the Ge- Vrba C. (1895) Beryll von Pisek. Zeitschrift für Krystallographie
ological Society of Finland, Vol. 38, pp. 31–45. und Mineralogie, Vol. 24, No. 1-2, pp. 104–112.
Scandale E., Lucchesi S., Graziani G. (1990) Growth defects Weldon R., Ortiz J.G., Ottaway T. (2016) In Rainier’s footsteps:
and growth marks in pegmatite beryls. European Journal of Journey to the Chivor emerald mine. G&G, Vol. 52, No. 2, pp.
Mineralogy, Vol. 2, No. 3, pp. 305–312, 168–187, [Link]
[Link] Zedlitz O. (1941) Goniometrische Untersuchungen der Lösungs-
Schmetzer K. (2009) A rare ‘gota de aceite’ Colombian emerald und Ätzerscheinungen an einem Beryllkristall von Minas Ger-
which had been treated. Gems & Jewellery, Vol. 18, No. 1, pp. aes, Brasilien. Zentralblatt für Mineralogie, Geologie und
3–4. Paläontologie, Abt. A., pp. 98–108.

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

[Link]/gems-gemology

MORPHOLOGY OF COLOMBIAN EMERALD GEMS & GEMOLOGY SPRING 2023 71

 Lab Notes
Editors
Thomas M. Moses | Shane F. McClure

Purple ANHYDRITE commonly formed by the dehydration

Recently, the Carlsbad laboratory re- of gypsum, which has the chemical
ceived a 2.13 ct transparent purple oc- formula CaSO4·2H2O, and can have
tagonal step cut, measuring 7.32 × multiple colors ranging from colorless
6.34 × 6.01 mm (figure 1), for a colored to pale blue, light pink, and brown that
stone identification service. Micro- is colored by impurities. Occurrences
scopic observation showed twinning of purple anhydrite have been reported
planes, cleavage cracks, needle-like in Switzerland, Iran, and Sri Lanka
inclusions, fluid fingerprints, and (see, respectively, Spring 1998 Gem
strong doubling of many internal fea- News, p. 60; Fall 2000 Gem News, p.
tures, confirming the stone was 262; Fall 1988 Gem News, p. 179).
doubly refractive (figure 2). Faceted gem-quality and single-
Figure 1. This 2.13 ct purple stone crystal anhydrite is rare, and while
Standard gemological testing re-
was identified as the rare collec- the Carlsbad laboratory has examined
vealed a specific gravity of 2.97 and a
tor gem anhydrite. a couple of anhydrite gems, they were
refractive index of 1.570–1.615 with a
birefringence of 0.045. The refractive mostly aggregates bluish in color.
index measurement also indicated the Anhydrite is a calcium sulfate with This is the first example of a transpar-
stone was biaxial positive, and a biax- the chemical formula CaSO4. It is not ent purple faceted single crystal anhy-
ial interference figure was resolved well suited for use in jewelry due to its drite examined at GIA’s Carlsbad
using a conoscope and polarized light. low hardness of 3.5 on the Mohs scale laboratory.
Purple and light purple pleochroism and its perfect cleavage in one direc- Maria Estela Almeida and
was observed using a dichroscope. tion. This orthorhombic mineral is Nathan Renfro
The specimen had no fluorescence re-
action when exposed to long- and
short-wave UV light. These character- Figure 2. Twinning planes and fluid fingerprints were observed throughout
istics were consistent with the rare the faceted purple anhydrite. Field of view 3.89 mm.
collector gem anhydrite.
Raman spectroscopy confirmed
that the stone was anhydrite. Using
polarized ultraviolet/visible/near-in-
frared spectroscopy, we were able to
determine that the purple color re-
sulted from a prominent absorption
band centered at around 550 nm. A
sharp absorption feature was also ob-
served at 307 nm.

Editors’ note: All items were written by staff

members of GIA laboratories.

GEMS & GEMOLOGY, Vol. 59, No. 1, pp. 72–82.

© 2023 Gemological Institute of America

72 LAB NOTES GEMS & GEMOLOGY SPRING 2023

 Chemical analysis revealed that Libyan Desert GLASS Bangle
the 7.26 ct stone was a pyralspite gar- Bracelet
net due to the high aluminum con- Libyan Desert glass is a light green or
centration (values expressed in wt.%: yellow natural glass which is com-
MgO 18.50%, Al2O3 23.05%, SiO2 posed of almost 98% silica. It is a type
41.32%, CaO 1.89%, TiO2 0.05%, of tektite formed by meteorite impact
V2O3 0.02%, Cr2O3 0.04%, MnO in the desert and the fast quenching of
11.93%, Fe2O3 3.17%). The composi- the silica-rich melt produced. It was
tion of this garnet and its gemological first reported by Clayton and Spencer
properties were consistent with py- in 1934 (F. Fröhlich et al., “Libyan De-
rope, which has an RI of 1.73–1.75, an sert Glass: New field and Fourier trans-
SG of 3.78 (+0.009/–0.016), and mag- form infrared data,” Meteoritics and
nesium dominating the X site and Planetary Science, Vol. 48, No. 12,
minor amounts of manganese and 2013, pp. 2517–2530) and is found in
iron. Pyrope garnet, though, has a color Egypt’s Western Desert. Common in-
range of red to reddish orange and col- clusions are bubbles, cristobalite, dark
Figure 3. A 7.26 ct modified
orless. Pink pyrope garnet is incredibly brown streaks, and black iron oxides
round brilliant identified as pink
rare, especially in stones of this size. (J.A. Barrat et al., “Geochemistry and
pyrope garnet. Courtesy of Bill
Pyrope-spessartine can have a origin of Libyan Desert glasses,” Geo-
Vance.
bodycolor similar to pyrope but has chimica et Cosmochimica Acta, Vol.
an RI range of 1.75 to over the limit 61, No. 9, 1997, pp. 1953–1959).
and an SG of 3.78, both higher than Recently, GIA’s Hong Kong labo-
Pink Pyrope GARNET
the values documented for this pink ratory examined a transparent light
The Carlsbad laboratory received a garnet. The fact that garnet can be a
7.26 ct pink modified round brilliant yellow bangle bracelet measuring
mixture of species can make the iden- 66.12 × 13.10 mm and weighing
(figure 3) for a colored stone identifi- tification of these stones challenging.
cation report. Standard gemological 144.67 ct (figure 4). It had a spot re-
This example is an important re- fractive index of 1.44 and revealed
testing revealed that the stone was minder to carefully analyze a garnet’s
singly refractive with a refractive weak yellow fluorescence in short-
gemological properties to accurately wave UV. The chemical composition,
index (RI) of 1.741 and a hydrostatic identify it and how the laboratory can
specific gravity (SG) of 3.77, properties determined by qualitative analysis
use chemical analyses as validation. using energy-dispersive X-ray fluores-
consistent with garnet. Internally the
stone was quite clean, with microsco- Nicole Ahline cence, was nearly pure silica with
pic examination revealing scattered
short needles.
The garnet group is composed of Figure 4. A 144.67 ct Libyan Desert glass bangle measuring 66.12 ×
more than 20 species, all of them 13.10 mm.
sharing the basic chemical formula
X3Y2(SiO4)3. Five of these species are
common within the jewelry indus-
try: almandine, andradite, grossular,
pyrope, and spessartine. Andradite
and grossular are ugrandite garnets;
they have calcium in the X site of
their chemical formula. Almandine,
pyrope, and spessartine are all pyral-
spite garnets containing aluminum
in the Y site. Isomorphous replace-
ment, in which one chemical ele-
ment substitutes for another in a
mineral’s crystal structure, makes it
possible for garnets to be a chemical
mixture of two or more garnet
species. GIA gemologists use a gar-
net’s gemological properties and
chemistry to categorize the stone
into its particular species.

LAB NOTES GEMS & GEMOLOGY SPRING 2023 73

 high-temperature low-pressure silica of cultured pearls from Australian
phase indicates a natural high-tem- Pinctada maxima,” GIA Research
perature formation. From all the gem- News, February 13, 2017). Four
ological, FTIR, and Raman data aBCPs (figure 7) were recently discov-
collected, this bangle was identified as ered in a group of 50 loose pearls sub-
Libyan Desert glass. mitted to GIA’s Mumbai laboratory.
Most Libyan Desert glass is tubu- Externally, the four white to light
lar in shape, and some are approx- cream-colored aBCPs looked similar
imately spherical or rod-like (R.A. to the other pearls submitted in the
Figure 5. Various shapes of gas Weeks et al., “Libyan Desert glass: A lot. When viewed under 40× magnifi-
bubbles were dispersed individ- review,” Journal of Non-Crystalline cation, their surfaces exhibited a typi-
ually and in clusters in the Lib- Solids, Vol. 67, No. 1-3, 1984, pp. 593– cal nacreous surface of overlapping
yan Desert glass bangle. The row 619). Those fragments are usually 2 to aragonite platelets. Energy-dispersive
of gas bubbles in this image re- 63 mm in size, so it is surprising to X-ray fluorescence spectrometry on
sembles peas in a pod. Field of examine this material in the form of all four revealed manganese levels be-
view 2.40 mm. a bangle bracelet. tween 13.30 ppm and 45.60 ppm and
Ching Yin Sin strontium levels between 1064 ppm
and 1822 ppm, characteristic of a salt-
minor amounts of iron, strontium, water environment. Interestingly, op-
potassium, titanium, and zirconium. tical X-ray fluorescence (XRF) of pearl
PEARLS
Based on the differences in molec- 1 revealed a strong yellowish green re-
“Atypical Beads”: Variations of Two action, while pearls 2, 3, and 4 were
ular vibrations detected with Fourier-
Types of Nuclei inert. The ultraviolet/visible reflec-
transform infrared (FTIR) analysis, it
was determined that the molecular Atypical bead cultured pearls (aBCP) tance spectra collected on the four
structure of Libyan Desert glass is sig- are occasionally encountered during pearls showed weak absorption fea-
nificantly different from that of other laboratory testing. The nuclei used tures at around 320–420 nm. Raman
pure silica glasses. It has a higher ratio can take the form of various un- analysis using 514 nm laser excitation
of discontinuities and defects in the drilled, partially drilled, or even showed the expected doublet at
tetrahedral (SiO4) network (F. Fröhlich drilled materials including natural 702/705 cm–1 and peak at 1085 cm–1,
et al., 2013). Microscopic observation pearls (abalone, scallop, turban indicative of aragonite.
revealed common inclusions in glass species, Pteria species, and Pinna Real-time microradiography (RTX)
such as numerous gas bubbles (figure species), freshwater non-bead cul- and X-ray computed microtomography
5) and flow structure. This bangle also tured pearls, coral, plastic, small (μ-CT) analysis revealed a variety of in-
presented abundant white spherulites shells, faceted sapphire beads of var- ternal structures that required interpre-
identified by Raman spectroscopy as ious colors, glass, quartz, and agate tation. RTX imaging of pearl 1 showed
cristobalite (a polymorph of silica). (“Atypical ‘beading’ in the production a thin irregular demarcation close to
These spherulites were small and in-
dividually dispersed throughout the
glass (figure 6). The presence of this Figure 7. Four loose atypical bead cultured pearls weighing 3.30 to 8.53 ct.
Pearls 1–4 are shown from left to right.

Figure 6. White spherulites in the

bangle were identified as cristo-
balite by Raman analysis. Field
of view 5.98 mm.

74 LAB NOTES GEMS & GEMOLOGY SPRING 2023

Sample
Surface image RTX image μ-CT image
details

Pearl 1
Button
5.50 ct

Pearl 2
Drop
3.30 ct

Pearl 3
Oval
8.53 ct

Pearl 4
Button
5.87 ct

Figure 8. Surface appearance and internal structures of the four atypical bead cultured pearls. Fields of view 2 mm,
0.6 mm, 2.86 mm, and 16.3 mm.

the surface which was not continuous. center of the nucleus and a small “or- the culturing process (“Atypical
On viewing the μ-CT, the demarcation ganic tail-like feature” at one end ‘beading’ in the production of cul-
was more discernible and followed the within the boundary separating the tured pearls from Australian Pinctada
irregular outline of the pearl (figure 8, bead from the overgrown cultured maxima,” GIA Research News, Feb-
row 1). An irregularly shaped fresh- nacre (figure 8, row 3). All features ruary 13, 2017). This pearl exhibited
water shell bead was clearly used as the were consistent with a saltwater non- distinct surface scratches that were
nucleus, hence the strong yellowish bead cultured pearl being used as the visible without magnification.
green reaction observed in the optical bead nuclei. The lack of growth arcs Pearl 4 also showed a strong de-
X-ray fluorescence unit (figure 9, left). around the demarcation was possibly marcation feature on the RTX and μ-
Under transmitted light, banding due to rapid nacre deposition during CT images. However, the central area
within the bead nucleus was also ob-
served (figure 9, right); externally, the
pearl had notable indentations on its Figure 9. Left: Pearl 1 revealed a strong yellowish green reaction under
surface. optical X-ray fluorescence. Right: Undulating shell banding visible
Pearl 2 hosted the most interesting within the sample’s shell bead nucleus using transmitted light; field of
nucleus of the four. A rectangular “tis- view 6.75 mm.
sue-box” shaped bead, likely a piece of
cut saltwater shell given the almost
identical radio-opacity to that of its
host, was evident in the RTX and μ-
CT images (figure 8, row 2) (Fall 2022
Gem News International, pp. 378–
380). Externally, the pearl was
smooth, lustrous, and free of any sur-
face blemishes.
The RTX and μ-CT images of
pearl 3 revealed an obvious demar-
cation with a linear structure at the

LAB NOTES GEMS & GEMOLOGY SPRING 2023 75

 1 2 3 4 5 6

7 8 9 10 11 12

Figure 10. A group of 12 pearls ranging from 7.38 × 6.49 × 6.05 mm to 10.61 × 9.05 × 8.05 mm and weighing 44.37
carats total.

differed from the other three pearls, as color saltwater cultured pearls from weighing 44.37 carats total and rang-
it showed a void-like feature in the Pinctada maxima,” Fall 2012 G&G, ing in size from 7.38 × 6.49 × 6.05
middle (figure 8, row 4). The bead nu- pp. 193–197) and the pearls’ internal mm to 10.61 × 9.05 × 8.05 mm. The
cleus used in the process was most structures and external appearance (A. nine drilled pearls appeared to be old,
likely a saltwater non-bead cultured Homkrajae et al., “Internal structures while the three undrilled pearls
pearl based on the fact that the pearl of known Pinctada maxima pearls: looked whiter and much newer over-
was inert to XRF, in keeping with Cultured pearls from operated marine all (figure 10). The shape of the sam-
pearl 3, despite the relatively thin mollusks,” Fall 2021 G&G, pp. 186– ples varied, and the nine drilled pearls
nacre layers that would allow a fresh- 205), it is apparent that all four of possessed a notable “aged” yellow
water bead to react. The pearl was them formed within Pinctada species coloration, which was also evident on
also unusual because it possessed mollusks, most likely Pinctada max- some areas of the three undrilled
three nearly flat “bases” with concave ima. Atypical bead cultured pearls whiter pearls.
features at the center of each. This have always been an interesting and The older-looking pearls possessed
raised questions about its identity sometimes challenging subject. With a dull luster, but the nacre condition
even before X-ray examination, since modern equipment and practical pearl was good. Microscopic examination
the authors have rarely observed nat- testing experience, laboratories such at 70× magnification soon revealed
ural pearls with such features. as GIA aim to remain one step ahead that the color was concentrated on
This is not the first time GIA has of the possible experiments used by the outer layers and within surface-
encountered aBCPs, but to receive cultivators. reaching features, proving that the
four with variations on two types of Rajesh S. Patel, Abeer Al-Alawi, color was not natural and a treatment
nuclei—atypical shell (not typical Lubna Sahani, and Nicholas Sturman had been used to alter their appear-
round shell beads) and non-bead cul- ance (Summer 2017 Gem News Inter-
tured pearls—in one lot was very in- national, pp. 255–256).
teresting. This proves that such pearls Although the color of pearls can
Treated Freshwater Non-Bead
are still circulating in the market and change over time due to various
are being mixed with natural goods in Cultured Pearls with an Antique causes (e.g., care factors such as stor-
an attempt at deception (Fall 2011 Lab Appearance age conditions and contact with
Notes, pp. 229–230). Given the spec- GIA’s Mumbai laboratory recently re- chemicals), these changes tend to be
tral data collected (S. Karampelas, ceived for identification a group of very gradual and develop over many
“Spectral characteristics of natural- nine drilled and three undrilled pearls years. This is partly due to their bio-

76 LAB NOTES GEMS & GEMOLOGY SPRING 2023

 Strikingly, the color was not typical
of natural-color pearls (Summer 2017
Gem News International, pp. 255–
256) and appeared to be restricted to
the surface and around the drill holes
(figure 11). Areas on some pearls also
revealed the original underlying
white color, further proof of treat-
ment (figure 12).
Next, we needed to determine
whether the pearls originated from a
saltwater or freshwater environment,
Figure 11. Two of the samples showing uneven surface coloration and and whether they were natural or cul-
color concentrations around the large drill holes. The worked surfaces tured. Real-time microradiography re-
around the drill holes are also clearly visible. Field of view 6.00 mm. vealed small central twisting
void-like features surrounded by fine
growth lines typical of freshwater
genic composition, which mainly served around the drill holes of all non-bead cultured pearls (figure 13),
consists of calcium carbonate with nine drilled pearls. The surfaces were consistent with those produced by
traces of organic substances, residual also heavily worked around the drill Chinese farms (K. Scarratt et al.,
substances, and water. Dry environ- holes and did not resemble the wear “Characteristics of nuclei in Chinese
ments are generally not advisable as and tear that might be expected on freshwater cultured pearls,” Spring
they may result in surface alterations antique pearls. The surfaces of the 2000 G&G, pp. 98–109). The fresh-
and, in rare cases, weight loss. The nine drilled samples (2–4 and 7–12) water origin was confirmed when the
color can also change over time were also slightly etched, indicating pearls fluoresced a strong yellowish
through wear and tear and the accu- possible exposure to a mild acidic so- green color upon exposure to X-ray
mulation of external contaminants, lution to create an “aging” effect. fluorescence, and further substanti-
so further analysis was required to
prove whether these pearls naturally
discolored over time or were treated Figure 13. The internal structures of four of the pearls showed twisted
to look antique, thereby inflating void-like features surrounded by fine growth arcs characteristic of non-
their value (M.S. Krzemnicki, “Fake bead cultured freshwater pearls. The two wide vertical gray features seen
historic provenance: ‘Aged’ cultured in pearls 3 and 4 are the drill holes.
pearls,” Facette, No. 25, 2019, p. 28).
Visually, it was readily apparent
3 4
that the drill holes were very large in
relation to the size of the pearls.
When viewed with a 10× lens or
under a gemological microscope at
70×, color concentrations were ob-

Figure 12. An area of underlying

white color below the treated sur-
face of one of the samples. Field
of view 5.00 mm.
5 6

LAB NOTES GEMS & GEMOLOGY SPRING 2023 77

ated by manganese values ranging Unlike the traditional bead cul-
from 500 to 1700 ppm and strontium tured pearls typically nucleated with
values from 800 to 1000 ppm, ob- spherical shell bead nuclei, a heart-
tained by energy-dispersive X-ray flu- shaped bead nucleus was used in this
orescence spectroscopy. pearl to achieve its unique shape.
Since Raman spectroscopy is also a Hence, it was described as an “atypi-
valuable analytical tool for differenti- cal bead” cultured pearl (aBCP) ac-
ating natural versus treated color in cording to GIA’s pearl identification
pearls, 514 nm ion-argon laser excita- terminology. Real-time microradiog-
tion was used on the surfaces. The re- raphy (RTX) and X-ray computed mi-
sults revealed a doublet at 702 and 705 crotomography (CT) clearly revealed
cm–1, as well as a peak at 1085 cm–1 in- the internal structure of a heart-
dicative of aragonite, which is seen in shaped bead nucleus (figure 15). The
the majority of pearls. High back- nacre coverage, averaging 0.80 mm
ground fluorescence, often characteris- thick, was evenly distributed all
tic of treated-color pearls, was also Figure 14. A heart-shaped golden around the nucleus, demonstrating
noted in the drilled pearls. The photo- South Sea pearl pendant measur- that the heart shape of the pearl was
luminescence spectra obtained dis- ing 17.78 × 11.65 × 9.07 mm and consistent with the shape of the nu-
played high fluorescence more typical weighing 1.93 g (including the cleus. Magnification revealed no signs
of treated pearls, as well as the ex- 18K yellow metal findings). of surface working or treatment, sug-
pected aragonite peaks. Other than the gesting the pearl was in its original
suspiciously high fluorescence, no form after harvest. It displayed a
other peaks characteristic of dyeing heart-shaped pearl pendant that drew strong orangy yellow color that is
were observed. our attention. The pearl was re- highly sought after among South Sea
All observations and results indi- portedly produced in the Philippines, pearls. Ultraviolet/visible reflectance
cated that the nine drilled pearls had which is a major source of golden and photoluminescence spectroscopy
been treated to make them look an- South Sea cultured pearls. Upon rou- further confirmed its natural color
tique. Pearls 1, 5, and 6 were predomi- tine gemological and advanced in- origin (C. Zhou et al., “Update on the
nantly white and showed only strumental examination—including identification of dye treatment in yel-
minimal evidence of any artificial microradiography, ultraviolet/visible low or ‘golden’ cultured pearls,”
aging attempts. These three showed a spectroscopy, manganese and stron- Winter 2012 G&G, pp. 284–291).
reaction to long-wave UV fluorescence tium contents from energy-dispersive Although this was not the first
more indicative of routine processing X-ray fluorescence, and reaction to time a heart-shaped pearl has been ex-
carried out on freshwater pearls. We long-wave UV—we concluded that it amined in the lab, it is very rare to see
concluded that these freshwater non- was an atypical bead cultured pearl a whole pearl with such a perfect non-
bead cultured pearls were also treated from the Pinctada maxima species, round cultured shape. Previous sub-
to make them appear aged. Hence, the commonly known in the trade as a missions have either been identified
report stated that these were fresh- South Sea cultured pearl (figure 14), as mabe pearls (assembled cultured
water non-bead cultured pearls from and its heart shape and golden color blisters) or have had poorly defined
the Unionidae family and that pearls were entirely natural. heart shapes. Mabe pearls are not con-
2–4 and 7–12 had been color modified.
Since the majority of the surfaces of
pearls 1, 5, and 6 remained white and Figure 15. X-ray images from RTX (left) and CT (right) show the demar-
only showed insignificant areas of dis- cation of a heart-shaped bead nucleus with an average nacre thickness of
coloration, the modified color descrip- 0.80 mm.
tion was not applied.
Andrew Aron, Abeer Al-Alawi, and
Nicholas Sturman

Heart-Shaped Golden South Sea

Cultured Pearl
The heart shape has been a world-
wide symbol of love and romance for
centuries. GIA’s New York laboratory
recently received a golden-colored

78 LAB NOTES GEMS & GEMOLOGY SPRING 2023

sidered whole pearls and often possess
a thin nacre dome top (Fall 2021 Gem
News International, pp. 277–279).
Due to the thin nacre, their shape is
easier to control. Therefore, fancy-
shaped nuclei are commonly used in
mabe pearls.
Many atypical bead cultured pearls
have been studied and examined over
the years at GIA (“Atypical ‘beading’
in the production of cultured pearls
from Australian Pinctada maxima,”
GIA Research News, February 13,
2017; Fall 2022 Gem News Inter-
national, pp. 378–380). However, the Figure 16. The natural shell blister attached to a portion of the shell on
majority of the end products are very which it formed, weighing 2.60 g (13.00 ct) total and measuring 25.54 ×
different from the original shape of the 18.11 × 7.79 mm.
nuclei. This remarkable use of an
atypical bead nucleus to achieve a per-
fect shape in a cultured pearl suggests structure on the shell base (figure 17, X-ray fluorescence spectrometry on
recent improvements in pearl cultur- A and B). The shell was cut and two areas (top and base) revealed man-
ing techniques. worked around the edges, and growth ganese levels below the detection
layers were visible when the cut sur- limit, and strontium levels of 1470
Joyce Wing Yan Ho and
faces were examined. Exposure to X- ppm (top) and 2431 ppm (base) were
Emiko Yazawa
ray fluorescence yielded no reaction consistent with those expected for
for the shell blister. Energy-dispersive saltwater pearls.
A Gastropod Shell in a Unique
Shell Blister
Natural blister pearls and natural Figure 17. Photomicrographs show graduated cellular structures from the
shell blisters have long been a subject natural shell blister examined in the Mumbai laboratory (A and B; fields
of debate for gemologists and can be of view 4 mm and 3 mm, respectively) compared to those from a known
very challenging to definitively iden- Pinnidae shell from GIA’s Bangkok laboratory (C and D; fields of view 4
tify (see G&G Lab Notes from Fall mm and 3 mm, respectively).
1992, Spring 1995, Winter 1996,
Winter 2015, Summer 2016, and A B
Spring 2018). According to the World
Jewellery Confederation (CIBJO), a
natural blister pearl forms when a
natural pearl detaches itself from the
pearl sac and attaches to the inner
wall of the shell, while a natural shell
blister is an internal protuberance
that forms on a shell’s inner surface.
It is usually caused by a foreign object
accidentally finding its way into the
space between the mantle and shell C D
surface.
GIA’s Mumbai laboratory recently
examined an interesting dark greenish
brown baroque-shaped blister at-
tached to a portion of shell weighing
2.60 g (13.00 ct) total and measuring
approximately 25.54 × 18.11 × 7.79
mm (figure 16).
Examination under 40× magnifi-
cation revealed a graduated hexagonal

LAB NOTES GEMS & GEMOLOGY SPRING 2023 79

Figure 18. Top row, left to right: RTX images of the face, the right side, and the length or thickest direction re-
veal a prominent gastropod shell measuring approximately 8.30 × 5.75 mm within the shell blister. Bottom row:
μ-CT images show the structures for the same three directions and provide even clearer details of the entombed
gastropod.

Real-time microradiography (RTX), The ultraviolet/visible reflectance G&G, pp. 202–215). When the surface
shown in the top row of figure 18, re- spectra showed a prominent feature at structures were compared to photo-
vealed a pleasant surprise. An obvious around 450 nm and weaker features at micrographs of known Pinna species
and intriguing gastropod shell was 320, 330, 450, and 460 nm proving shell from GIA’s Bangkok laboratory
hidden beneath the overlying layers, that the sample’s color was natural (figure 17, C and D), the results showed
which resulted in further study using and lacked any treatment. Raman some similarity. However, Raman
X-ray computed microtomography (μ- analysis using 514 nm laser excitation spectra indicative of calcite were ob-
CT) (figure 18, bottom row). The mi- showed a series of peaks at 158, 186, tained for the Bangkok research collec-
nute details of a unique marine 204, and 213 cm–1; a weak peak at 280 tion samples, which differed from the
gastropod exoskeleton measuring ap- cm–1; a doublet peak around 701–704 sample studied here. The Atrina vex-
proximately 8.30 × 5.75 mm were ob- cm–1; and a strong peak at 1085 cm–1. illum species may have been respon-
served within a void feature, and the These were indicative of aragonite. sible for producing the natural shell
overlying growth layers on the face The photoluminescence spectra were blister, and this would explain the
side (the surface that grew within the also consistent with the Raman and aragonite-related Raman peaks since
mollusk it formed in) revealed mini- displayed high fluorescence and arag- this mollusk is known to produce
mal growth arcs. Marine gastropods onite-related peaks. These results pearls with an aragonitic structure.
are known to vary in size, and the were unexpected given the cellular- However, since Atrina species pearls
class consists of many thousands of looking structure observed through are often nacreous when aragonite is
species (A. Nutzel., “Larval ecology the microscope, as this type of struc- present and the sample in question
and morphology in fossil gastropods,” ture usually results in Raman spectra was non-nacreous, the contradicting
Paleontology, Vol. 57, No. 3, 2014, pp. indicative of calcite. data still leaves some doubt as to the
479–503). These observations pro- While the lustrous dark color and true identity of the host. Nevertheless,
vided enough evidence to make a the cellular-looking structure indi- this example of a shell blister encom-
clear distinction between a shell cated the host shell was likely a passing a gastropod shell represents a
blister, such as this sample, and a species from the Pinnidae family, it noteworthy phenomenon of nature
shell blister pearl (“Natural shell was not clear given the spectral data that is rarely encountered.
blisters and blister pearls: What’s the collected (N. Sturman et al., “Observa-
difference?” GIA Research News, Au- tions on pearls reportedly from the Jayesh Surve, Abeer Al-Alawi, and
gust 26, 2019). Pinnidae family (pen pearls),” Fall 2014 Nicholas Sturman

80 LAB NOTES GEMS & GEMOLOGY SPRING 2023

Figure 19. A large cavity partially filled with synthetic overgrowth on the pavilion of a heated Burmese ruby in
darkfield lighting (left) and reflected lighting (right). Field of view 7.19 mm.

RUBY with Interesting Synthetic gonally shaped, platy crystals (Fall using polarized light, but careful anal-
Overgrowth 2002 G&G Lab Notes, pp. 255–256). ysis is required because the over-
The Carlsbad laboratory received a When viewed in reflective lighting, growth can completely fill cavities
3.02 ct ruby for a colored stone iden- there is no luster difference between and be inconspicuous.
tification and origin report. Standard the host corundum and the synthetic
overgrowth, as the two materials have Nicole Ahline
gemological testing was consistent
with ruby: a hydrostatic specific the same refractive index. The gaps
gravity of 3.96 and a ruby spectrum that could be observed between the
crystals in this stone showed how Exceptionally Large
in a handheld spectroscope. Inter-
they grew in different orientations. SYNTHETIC RUTILE
nally, the stone displayed white
flakes in a hexagonal formation and Synthetic overgrowth is common The Carlsbad laboratory received three
roiled graining. Partially healed fis- enough that it should always be light yellow modified round brilliants
sures with residue, which result looked for in corundum that has un- weighing 11.13, 11.91, and 110.18 ct
from heating in the presence of a dergone high-temperature heat treat- (figure 20) for identification reports.
flux, were also documented. The in- ment. There are cases where Standard gemological testing revealed
clusion scene, along with the stone’s synthetic overgrowth can be observed the stones to be synthetic rutile based
trace element chemistry collected
using laser ablation–inductively
coupled plasma–mass spectrometry
(LA-ICP-MS), were consistent with Figure 20. Three large synthetic rutile brilliants weighing 11.13, 110.18,
Burmese ruby. This ruby was not and 11.91 ct. Courtesy of Arya Akhavan.
treated by beryllium diffusion, as no
beryllium was detected by LA-ICP-
MS. The curious feature of this
stone was a large cavity on the pa-
vilion that was partially filled with
synthetic overgrowth that was ob-
served in the microscope (figure 19).
Synthetic overgrowth can be a by-
product of heating a piece of corun-
dum at high temperatures, with or
without the presence of a flux. This
creates an environment in the crucible
that allows for the partial dissolution
of corundum and can result in syn-
thetic corundum overgrowth on the
surface and in cavities of the heated
stone. Synthetic overgrowth is typi-
cally an aggregated structure of hexa-

LAB NOTES GEMS & GEMOLOGY SPRING 2023 81

 UV-VIS-NIR SPECTRA
minum in these two samples (~80
110.18 ct synthetic rutile ppm Al compared to ~3 ppm). The
11.13 ct synthetic rutile difference in color was also demon-
strated in the ultraviolet/visible/near-
ABSORBANCE

infrared absorption spectra (figure 21).

This finding aligns with the practice
of growing rutile doped with small
amounts of aluminum to bring the
appearance of the material closer to
nearly colorless (C.H. Moore, “Rutile
400 500 600 700 800 900
boule and method of making the
WAVELENGTH (nm) same,” U.S. Patent 2,715,070, issued
August 9, 1955). Additionally, the
Figure 21. UV-Vis-NIR absorption spectra comparing the 110.18 ct and owner indicated that the two smaller
11.13 ct synthetic rutile studied, here demonstrating the effect of alumi- rutile crystals were grown by Na-
num doping in the smaller stone to produce a nearly colorless appearance. tional Lead Co. between 1961 and
1964, while the larger stone is from a
modern producer.
on over-the-limit refractive index nace to accommodate growth of the Aaron Palke
readings and extreme fire and dou- 110.18 ct crystal. Also notable were
bling, an identification that was sup- the novel facet pattern and precision
ported by Raman spectroscopy. cutting usually only seen in the high-
Of particular note was the aston- end collector market.
ishing size of these stones. The 110.18 Interestingly, the two smaller PHOTO CREDITS
ct synthetic rutile is the largest en- stones had a much lighter yellow Adrianna Gudino—1; Nathan Renfro—2; Tow-
fiq Ahmed—3; Johnny Leung –4; Sze Ling
countered in the GIA laboratory. Even color than the 110.18 ct stone, which
Wong—5, 6; Gaurav Bera—7, 16; Nishka
the smaller 11.13 ct and 11.91 ct was notable even considering the
Vaz—8, 9 (right); Jayesh Surve—10–12, 17 (A
stones are much larger than normal shorter light path length through the and B); Sood Oil (Judy) Chia—14; Nicholas
for this material, which is usually smaller stones. Chemical analysis Sturman—17 (C and D); Nicole Ahline—19;
seen in sizes up to several carats. The using laser ablation–inductively cou- Robert Weldon—20
owner of the stones indicated that the pled plasma–mass spectrometry
manufacturer had to modify their fur- showed a higher concentration of alu-

Join our growing G&G Facebook group of more than 32,000

members, connecting gem enthusiasts from all over the world!

82 LAB NOTES GEMS & GEMOLOGY SPRING 2023

 First Place
METHODS AND CHALLENGES OF ESTABLISHING THE GEOGRAPHIC ORIGIN OF
DIAMONDS
FALL 2022
Evan M. Smith, Karen V. Smit, and Steven B. Shirey
Evan Smith is a senior research scientist at GIA in New York. His primary research focuses on
the geology of gemstones and serves to highlight diamonds as some of the most scientifically
valuable materials on Earth. Dr. Smith holds bachelor’s and master’s degrees in geological engi-
Evan Smith Karen Smit Steven Shirey
neering from Queen’s University and a PhD in geology from the University of British Columbia.
Karen Smit has been conducting research on the origin of natural diamonds and studying dia-
monds from mines worldwide for 15 years, publishing much of her work in scientific journals.
She holds a PhD in geology from the University of Alberta and a master’s in geochemistry from
the University of Cape Town. Formerly a research scientist at GIA in New York, Dr. Smit is cur-
rently a senior lecturer in isotope geochemistry at the University of the Witwatersrand in
Johannesburg. Steven Shirey is a senior staff member in the Earth and Planets Laboratory of the
Carnegie Institution for Science. Dr. Shirey’s main interests include diamonds as the deepest
probe of plate tectonics, the igneous evolution of the earth, and the emergence of the continents.
He holds a PhD from the State University of New York at Stony Brook and is a fellow of the
American Geophysical Union, the Geochemical Society, the Geological Society of America, and
the Mineralogical Society of America (of which he is a former president).

Second Place
GEMS ON CANVAS: PIGMENTS HISTORICALLY SOURCED FROM GEM MATERIALS
FALL 2022
Britni LeCroy
Britni LeCroy is a staff gemologist at GIA in Carlsbad, California. She obtained a bachelor’s
degree in geoscience from the University of Texas at Dallas before receiving gemology diplomas
from GIA and the Gemological Association of Great Britain. A frequent contributor to G&G, her
areas of interest include pearls, organic gem materials, and historical gemology.
Britni LeCroy

Third Place
A CANARY IN THE RUBY MINE: LOW-TEMPERATURE HEAT TREATMENT EXPERIMENTS ON
BURMESE RUBY
WINTER 2022
E. Billie Hughes and Wim Vertriest
E. Billie Hughes is a gemologist and cofounder of Lotus Gemology in Bangkok. An award-winning
photomicrographer and sought-after lecturer, she has delivered talks on gemology around the
world. She is a graduate of the University of California, Los Angeles. Wim Vertriest is manager of
E. Billie Hughes Wim Vertriest
field gemology at GIA in Bangkok, where he curates GIA’s colored stone reference collection. He
obtained a master’s degree in geology (geodynamics and geofluids) from KU Leuven in Belgium.

Many thanks to the members of G&G’s Editorial Review Board for voting this year.

MOST VALUABLE ARTICLE AWARD GEMS & GEMOLOGY SPRING 2023 83

 Editor: Nathan Renfro
Contributing Editors: Elise A. Skalwold and John I. Koivula

Three-Phase Inclusion in Alexandrite Figure 1. Top: Negative crystal in an alexandrite contain-

Natural gemstones form under very intense conditions that ing a fluid and crystals at room temperature. Bottom:
allow for a wide range of interesting phenomena and inclu- Below 31.2°C, multiple gas bubbles appear. Photo-
sions to occur within them. Environmental conditions micrographs by Jamie Price; field of view 1.99 mm.
within the earth’s crust influence the types and amounts of
inclusions that occur. Various categories of inclusions, such
as crystals or structural defects, for example, are widespread
among almost all gem species. On the other hand, there are
less frequent types of inclusions that form only under spe-
cific geological conditions and are found only in certain gem
species. Although less common, it is possible to observe up
to three phases of matter within a single inclusion suite.
At conditions slightly above room temperature (i.e., in
the well light of a gemological microscope), the largest
jagged cavity pictured here contains multiple separate crys-
tals and a transparent colorless fluid (figure 1). At a tem-
perature below 31.2°C, the fluid separates into liquid and
multiple bubbles of carbon dioxide gas are released (figure
2). Differences in environmental conditions during forma-
tion influence the amount of static pressure within the
stone, which explains why some three-phase inclusions are
visible at room temperature while others are only visible at
lower temperatures (see J.I. Koivula, “Carbon dioxide fluid
inclusions as proof of natural-colored corundum,” Fall 1986
G&G, pp. 152–155). Despite the shape and the relief of the

About the banner: Acicular inclusions of the blue mineral dumortierite are
present throughout this quartz crystal from Brazil. Photomicrograph by
Nathan Renfro; field of view 5.68 mm.

GEMS & GEMOLOGY, VOL. 59, NO. 1, pp. 84–91.

© 2023 Gemological Institute of America

84 MICRO-WORLD GEMS & GEMOLOGY SPRING 2023

 Figure 3. A hackmanite with a faded purple color
after storage in the dark (left), illuminated with long-
Figure 2. Close-up view distinctively pointing out wave UV light (center), and after removal from a few
the locations of the crystals (blue arrows) and gas seconds of exposure to the long-wave UV light (right).
bubbles (white arrows). Photomicrograph by Jamie The color has deepened after exposure to UV light.
Price. Photos by Ronnakorn Manorotkul.

crystals present, the overall appearance of this alexandrite’s reaction. After exposure to sunlight or artificial light with
multiphase inclusion is comparable to the typical jagged a UV component, samples can quickly develop a much
three-phase inclusions that are distinctive for Colombian stronger color. One hackmanite sample recently examined
emerald. by the author (figure 3) showed this phenomenon.
Jamie Price Another hackmanite was submitted that displayed an
GIA, Carlsbad interesting elongated white inclusion (figure 4, left) that was
identified by micro-Raman as apatite. As we examined the
sample in the microscope, we tried exposing it to a 6-watt
Apatite in Hackmanite long-wave UV light for approximately one second. Almost
Hackmanite, Na8Al6Si6O24(Cl2,S), is a variety of sodalite no- immediately, a change was observed. The hackmanite de-
table for its tenebrescence. When kept in darkness, samples veloped a more vibrant purple color that is clearly displayed
can fade to pale purple and even gray or translucent to in figure 4 (right).
opaque white. When exposed to long-wave ultraviolet il- E. Billie Hughes
lumination, they can display a strong orange fluorescence Lotus Gemology, Bangkok

Figure 4. Left: In darkfield illumination, a white apatite crystal is visible in the hackmanite. Right: After expo-
sure to long-wave UV light for approximately one second, a dramatic change can be seen. The overall color of the
hackmanite is now a more vivid shade of purple. Photomicrographs by E. Billie Hughes; field of view approx-
imately 5 mm.

MICRO-WORLD GEMS & GEMOLOGY SPRING 2023 85

 Figure 5. This 4.63 ct
chondrodite contained
irregular multiphase
and reflective irides-
cent platelets with a
bright orangy yellow
color. Photo by Adriana
Gudino; courtesy of Bill
Vance.

Inclusions in Chondrodite Chondrodite, Mg5(SiO4)2(F,OH)2 is a member of the hu-

The authors recently examined a 4.63 ct orangy yellow mite group. This rare mineral occurs primarily in meta-
pear modified brilliant (figure 5). The gemological proper- morphosed limestones and dolomites (W.L. Roberts et al.,
ties as well as infrared and Raman spectroscopy identified Encyclopedia of Minerals, 2nd ed., Van Nostrand Reinhold,
the stone as chondrodite. New York, 1990, p. 170). If contact metamorphism occurs
Microscopic observation revealed a lively inclusion in an area with sufficient magnesium, chondrodite and
scene with colorless reflective and iridescent platy inclu- other associated magnesium-bearing minerals such as spi-
sions (figure 6). Dark, highly reflective platelets of graphite, nel, phlogopite, and tremolite may form.
which can be associated with low-relief colorless crystals Chondrodite has been found to originate from Tanza-
of various minerals (figure 7), were also confirmed by nian localities at Mahenge (Winter 2011 Gem News Inter-
Raman analysis.

Figure 7. Dark, highly reflective platelets of graphite as

Figure 6. Platy inclusions were observed throughout well as unidentified low-relief colorless inclusions of mul-
the chondrodite. Photomicrograph by Nathan Renfro; tiple minerals were also observed in the chondrodite. Pho-
field of view 4.11 mm. tomicrograph by Nathan Renfro; field of view 0.91 mm.

86 MICRO-WORLD GEMS & GEMOLOGY SPRING 2023

national, p. 316), Sumbawanga (Winter 2007 Gem News with “flapping wings” and “antennae” caused by fractures
International, pp. 377–379), and Tanga (B. Clark, “Chon- around the inclusion (figure 8). Raman spectroscopy re-
drodite reportedly from Tanga, Tanzania,” Journal of Gem- vealed the inclusion to be pyrope-almandine-grossular gar-
mology, Vol. 34, No. 8, p. 655). It has also been identified net, (Mg,Fe,Ca)3Al2(SiO4)3, a major constituent mineral of
as an inclusion in a spinel from Mogok, Myanmar (Winter eclogite.
2021 G&G Micro-World, p. 383). The gemological proper- Eclogite and peridotite are the major diamond host
ties of this stone were consistent with material from Ma- rocks in the lithospheric mantle. Worldwide, the ratio of
henge, but the origin could not be confirmed. diamonds with eclogitic mineral inclusions to those with
Chondrodite is an uncommon mineral and even more peridotitic inclusions is ~1:2 (T. Stachel and J.W. Harris,
rare as a gemstone. This exceptional example of a bright “The origin of cratonic diamonds – constraints from min-
orangy yellow color, faceted by Bill Vance of Vance Gems, eral inclusions,” Ore Geology Reviews, Vol. 34, No. 1-2,
showcases the types of inclusions that can be found in this 2008, pp. 5–32).
gemstone. Although the presence of inclusions generally reduces
Kendra Carty and Amy Cooper a diamond’s clarity grade, diamonds containing visually
GIA, Carlsbad identifiable mineral inclusions are uncommon. Studies of
some smaller diamonds (~2 mm) from South Africa and
Botswana revealed a low abundance of inclusion-bearing
A “Flying Insect” in Diamond diamonds (about 1% in a total of one million diamonds ex-
The authors recently examined a 0.75 ct type IaA Fancy amined; Stachel and Harris, 2008). For a gem-quality dia-
Light brownish yellow diamond. This diamond, with a mond to contain more than 10 mineral inclusions is
clarity grade of I1, contained at least 10 orange inclusions, therefore notable.
the largest one ~500 μm in its longest dimension. A par- Mei Yan Lai and Taryn Linzmeyer
tially exposed orange inclusion resembled a flying insect, GIA, Carlsbad

Figure 8. This partially exposed orange garnet inclusion with surrounding fractures in a brownish yellow diamond
resembled a flying insect. Photomicrograph by Mei Yan Lai; field of view 1.58 mm.

MICRO-WORLD GEMS & GEMOLOGY SPRING 2023 87

 Figure 9. Rare cloud in-
clusions were observed
in a 0.53 ct D-color dia-
mond. Photomicrograph
by Nathan Renfro; field
of view 2.39 mm.

Starry Night in Diamond rare octahedral morphology was recently encountered in

In astronomy, a star cloud is defined as a group of many the form of a xenomorphic inclusion within diamond.
stars. A diamond recently submitted to GIA’s Surat lab con- Xenomorphism is the process by which a growing mineral
tained cloud inclusions resembling a starry night (figure 9). forces a syngenetic mineral inclusion to adopt a crystal
This diamond was a 0.53 ct D-color natural round brilliant habit familiar to the host (J.I. Koivula, The MicroWorld of
with SI2 clarity. Fourier-transform infrared absorption spec- Diamonds, Gemworld International, Northbrook, Illinois,
troscopy indicated a type IaAB diamond with low nitrogen 2000). Xenomorphic crystals may also be referred to as “an-
content of about 5 ppm. Each of the cloud inclusions had hedral” since they lack their own characteristic crystal
apparently randomly placed dendrites. The star field in this shape in favor of the host’s. However, it should be noted
stone appeared comparable to a diamond described pre- that not all anhedral included mineral crystals are xeno-
viously (Fall 2017 G&G Micro-World, p. 369); however, the morphic. As an octahedron is the most common diamond
individual clouds in that diamond resembled crosses of uni- habit, the garnet inclusion grew into this form, complete
form orientation that corresponded with the cube face. with triangular growth marks visible on one of the octahe-
Photoluminescence (PL) mapping using 455, 532, and dral faces (figure 11). Xenomorphism in diamond can also
633 nm excitations was used to determine whether optical occur with mineral inclusions that do not belong to the
features could be detected within the cloud inclusions, as
that analysis method has proven useful with cloud features
Figure 10. This deep ultraviolet fluorescence
in other diamonds (Fall 2020 Lab Notes, pp. 416–419). Ho-
DiamondView image shows that the cloud inclusion
wever, the PL maps were inconclusive, as no distinctive
features appear to be contained within a single
features were detected within the star-shaped inclusions.
growth zone. Image by Sally Eaton-Magaña.
In deep ultraviolet fluorescence imaging, they appeared to
be contained within the same growth zone (figure 10).
The inclusion suite was an interesting collection of
cloud formations and a delightful discovery in this unusual
diamond.
Sally Eaton-Magaña and Stephanie Shaw
GIA, Carlsbad
Alpesh Vavadiya
GIA, Surat

Xenomorphic Garnet in Diamond

Garnet is a nesosilicate mineral belonging to the cubic
crystal system. The crystal habit most commonly assumed
by this mineral is a dodecahedron. A garnet crystal with

88 MICRO-WORLD GEMS & GEMOLOGY SPRING 2023

 Figure 11. A purplish
pink xenomorphic oc-
tahedral garnet crystal
seen in a diamond.
Note the trigons on the
upper right octahedral
face and stress fractures
around the intersec-
tions of the crystal
faces. Photomicrograph
by Britni LeCroy; field
of view 1.26 mm.

cubic crystal system, and these can be forced into other tern (figure 12). Careful examination revealed a partially
diamond habits. For example, a monoclinic crystal of di- healed fissure intersecting one of the negative crystals, in-
opside may be forced into a cuboctahedral habit. Xeno- dicating the stone had been subjected to heat treatment.
morphic crystals are just one of many unique inclusion Fortunately, this spectacular pattern of negative crys-
types possible within diamond. tals endured the heat treatment process, decorating the in-
Britni LeCroy and Virginia Schwartz ternal micro-world of this sapphire.
GIA, Carlsbad Piradee Siritheerakul and Tao Shiu Hei
GIA, Bangkok

Arrow-Patterned Negative Crystals in Sapphire

Negative crystals are a common inclusion in corundum Unusual Solid Inclusions in Flame-Fusion Ruby
and often rupture upon exposure to moderate heat treat- The author recently examined unusual dark red solid in-
ment. The author recently analyzed a sapphire containing clusions in a flame-fusion laboratory-grown ruby. Standard
a series of negative crystals arranged in an arrow-like pat- gemological testing yielded a refractive index of 1.760–

Figure 12. Negative

crystals arrayed in an
arrow-like pattern.
Photomicrograph by
Suwasan Wongchacree;
field of view 1.07 mm.

MICRO-WORLD GEMS & GEMOLOGY SPRING 2023 89

Figure 13. Numerous natural-looking dark red solid Figure 14. Strong curved growth lines and numerous
inclusions were observed in a flame-fusion labo- gas bubbles and dark solid inclusions were present in
ratory-grown ruby. Photomicrograph by Ezgi Kiyak; the flame-fusion ruby. Photomicrograph by Ezgi
field of view 2.9 mm. Kiyak; field of view 2.9 mm.

1.768, a chromium emission line in red using a handheld temperature, pressure, and chemistry. While some of these
spectroscope, medium red fluorescence under long-wave growth details are submicroscopic and remain hidden dur-
UV, and weak red fluorescence under short-wave UV. ing microscopic observation, other features such as min-
These properties were consistent with ruby. Microscopic eral inclusions and encapsulated fluids are waiting to be
examination using brightfield illumination revealed nu- discovered. In a very real sense, inclusions and their devel-
merous dark red inclusions (figure 13) that resembled crys- opmental sequencing in their host gems and minerals are
tals commonly seen within natural corundum. descriptions of mineralogical genetics.
Furthermore, magnification showed characteristic in- Researchers focused specifically on inclusions usually
ternal features of flame-fusion ruby such as curved striae cannot outbid serious gem and mineral collectors for crys-
and gas bubbles. The strong curved growth lines and nu- tals and fashioned gems with interesting internal targets.
merous gas bubbles (figure 14) suggested that this was an As a result, many such “gems” end up in private collec-
early type of flame-fusion ruby. Although flame-fusion co- tions, and their inclusions are rarely, if ever, fully doc-
rundum is widely available in the marketplace, it is un- umented. This is a tremendous missed opportunity,
usual to encounter a specimen containing natural-looking because inclusions contain geological information that
dark red solid inclusions. These inclusions are possibly un- tells a story about themselves and their host.
melted alumina powder, and the fact that Raman was un- For this issue’s Quarterly Crystal, we recently had the
able to identify something other than alumina supports opportunity to document a well-formed gem-quality, trans-
this (E.J. Gübelin and J.I. Koivula, Photoatlas of Inclusions parent, very light greenish blue terminated hexagonal crys-
in Gemstones, Volume 3, Opinio Publishers, Basel, Switz- tal of aquamarine, with a small amount of light brown
erland, 2008). matrix at the base and a clearly visible inclusion cluster at
Identification of the natural or laboratory-grown origin the near center (figure 15). The aquamarine was obtained
of a stone with this type of natural-looking inclusion can from Muntazir Mehdi of Shad Fine Minerals International
be difficult. Careful observation is needed to correctly iden- in Gilgit-Baltistan, Pakistan. The geographic source was re-
tify such a specimen’s origin. portedly the Kharguluk mine in Baltistan Province.
Ezgi Kiyak At 56.00 ct with corresponding measurements of 26.36
GIA, New York × 16.09 × 15.02 mm, this aquamarine played host to an
easily eye-visible tight cluster of randomly arranged dark
reddish brown to black bladed crystals (figure 16) that re-
Quarterly Crystal: Columbite(?) in Beryl sembled the mineral columbite-(Mn), Mn2+Nb2O6 , as
Those in the gem and mineral community can appreciate shown by E.J. Gübelin and J.I. Koivula (Photoatlas of Inclu-
the near-perfection of a beautifully crystallized mineral sions in Gemstones, Volume 2, Opinio Verlag, Basel, Switz-
specimen or a well-cut gemstone. But these surfaces are erland, 2005, pp. 263, 319).
only the exterior, covering a complex, multilayered story Micro-Raman has significantly reduced the time re-
within. These internal contents often remain unknown quired to instrumentally identify many inclusions. Some
unless we choose to explore below the surface. Somewhat inclusion identifications that once took hours can now be
like trees and their rings, minerals and gems form as con- made in a matter of minutes using this technique. The
centric, typically crystallized layers, recording the devel- Raman instrument can also analyze some subsurface in-
opmental details within each layer through changes in clusions at depths of more than a millimeter in certain fa-

90 MICRO-WORLD GEMS & GEMOLOGY SPRING 2023

 Figure 15. Measuring
26.36 mm in length and
weighing 56.00 ct, this
Pakistani aquamarine
crystal is host to an eye-
visible 7.20 mm min-
eral inclusion cluster.
Photo by Annie Haynes.

vorable instances, which means inclusions do not need to stone. Due to the value of this inclusion specimen, we
be exposed to the surface. opted to forgo destructive analysis and keep the beryl crys-
After several Raman attempts from different directions, tal intact for future gemological exploration.
we realized that destructive analysis would be needed to John I. Koivula and Nathan Renfro
clearly identify the inclusion cluster deep within this GIA, Carlsbad

Figure 16. None of the

analytical techniques at
our disposal were able
to conclusively identify
the mineral making up
the 7.20 mm inclusion
cluster. The distinctive
morphology shown by
these inclusions
strongly suggests that
the cluster is composed
of columbite-(Mn) crys-
tals. Photomicrograph
by Nathan Renfro; field
of view 10.28 mm.

MICRO-WORLD GEMS & GEMOLOGY SPRING 2023 91

 The following 25 questions are from the four 2022 issues of G&G. Please refer to the feature articles in those
issues to find the single best answer for each question.
Visit [Link]/gems-gemology to take the Challenge online. Entries must be received no later than September 1,
2023. All entries will be acknowledged with an email.
Score 75% or better, and you will receive a certificate of completion (PDF file). Earn a perfect score, and your
name also will be listed in the Fall 2023 issue of Gems & Gemology.

1. Nearly half of the diamonds from C. is reliant upon carbon dating. C. has high levels of trace impurities.
Guyana D. is not currently possible. D. rivals Imperial topaz in color.
A. have green spotting or skins.
B. are found as aggregates. 4. Only the most important paintings [Link] higher the silica content detected
C. reach 10 carats in size. were allowed to use this natural in turquoise, the more likely
D. possess D color. pigment during its prime. A. it will command a higher per-carat
A. Ultramarine price.
2. Heating ruby to 600°C may B. Red ochre B. it will have a higher specific gravity.
A. significantly alter the Raman spectra C. Cinnabar C. its bodycolor will be sky blue.
of any spinel inclusions. D. Azurite D. it has been porcelain treated.
B. cause fissures to begin forming
around solid inclusions. 5. Amber from Phu Quoc, Vietnam 8. Chinese sapphire from Muling
C. affect sphene and apatite inclusions A. is mostly opaque. A. is mostly yellow, green, and blue in
the most. B. is almost exclusively pale yellow in color.
D. alter the appearance of any mica color. B. is always heavily weathered in its
inclusions. C. commonly contains fauna. rough form.
D. commonly contains gas bubbles C. shows remarkable optical homo-
3. Geographic origin determination of and spangles. geneity.
diamonds D. can be blue and yellow bicolor.
A. is reliant upon identification of 6. Schneckenstein topaz
inclusions. A. is currently an important commer- 9. Today, the Habachtal emerald mine
B. is reliant upon trace element cial gemstone. A. produces emeralds similar to those
analysis. B. likely formed at low temperatures. from Colombia.

92 G&G CHALLENGE GEMS & GEMOLOGY SPRING 2023

 B. produces a small amount of min- 15. Geographic origin determination of C. They contain high levels of nickel.
eral specimens and facetable diamonds versus colored stones is D. They are generally type IIb or Ib.
rough. complicated by the fact that
C. is one of several Austrian emerald diamonds 21. Low-temperature heat treatment of
mines. A. form in a much less chemically ruby
D. is no longer in operation. diverse environment. A. covers the range of temperatures
from 600–1500°C.
B. can only contain three possible
10. Which gem is considered too B. involves the dissolution of second-
mineral inclusions.
radioactive for daily wear? ary-phase microcrystals.
C. do not contain incompatible ele-
A. Green diamond C. does not always produce the same
ments.
B. Nephrite jade reactions for inclusions of the same
D. have too high a refractive index.
C. Ekanite type.
D. Amber 16. “Natural porcelain” turquoise is D. reliably produces telltale visual fea-
separated from “porcelain-treated” tures.
11. A gem bead cultured pearl
turquoise by 22. Heat treatment of ekanite to improve
A. is of the highest quality if produced
A. infrared and UV-Vis spectroscopy. color
using a freshwater “Edison” pearl.
B. infrared, UV-Vis, and Raman spec- A. is performed on virtually all com-
B. must use a bead with a thermal ex-
troscopy. mercially sold material.
pansion coefficient similar to that
of nacre. C. color, refractive index, and surface B. results in a green similar to peridot.
C. involves incorporating a diamond luster. C. has not been successful.
in a pearl. D. specific gravity, silica content, and D. removes yellow tints.
D. is always under 8 mm in diameter. Raman spectroscopy.
23. In ancient times, white pigment was
12. How do the optical data plots for 17. A possible origin discriminator for created by
Schneckenstein topaz differ using a nephrite may be its A. heating bone in the absence of
wavenumber versus a wavelength A. ratio of thorium to uranium. oxygen.
scale? B. degree of compactness. B. heating bone in the presence of
A. Wavenumber data is directly pro- C. concentration of iron. oxygen.
portional to photon energy. D. depth of green color. C. heating ivory in a closed crucible.
B. Wavelength data is directly propor- D. finely grinding hematite.
tional to photon energy. 18. Blue sapphires from Muling
C. At lower energies, wavelength data A. are mined exclusively from source 24. Jewelry that incorporates near field
is compressed. rocks. communication (NFC) microchips
D. There is no difference. B. all show angular color zoning. A. requires the NFC microchip to be
C. are not all formed by the same ge- placed in a Faraday cage.
13. What happens to pink spinel after ological process. B. must be larger to accommodate a
heating to 1000°C? D. are clearly metamorphic in origin. battery.
A. The full width half maximum C. can easily be found with an app on
(FWHM) of its R- and N-lines de- 19. Copper diffusion treatment of near- your smartphone.
crease. colorless plagioclase feldspars D. cannot be subjected to tempera-
B. Its R-line becomes less dominant A. increases the fluorescence intensity tures greater than 700°F (370°C).
than its N-lines in a PL spectrum. at around 400 nm.
C. The saturation of its pink bodycolor 25. The emeralds from the coronation
B. reduces the concentration of
decreases significantly. Cu+-Cu+ dimers. crown of Napoleon III
D. Its Cr3+ ions transition from non- C. imparts a red color only to A. came from the Habachtal emerald
ideal to ideal classes. labradorite. mine.
D. imparts a red color only to ande- B. contain the gota de aceite optical
14. The red color in plagioclase feldspars effect in many cases.
sine.
A. increases in saturation with in- C. were all cut specifically for this
creasing Cu+ concentration. 20. Why are diamonds from Guyana crown.
B. is only seen in copper-diffused commonly green? D. were clarity enhanced.
material. A. They were damaged by radioactiv-
C. cannot be fully explained today. ity. To take the Challenge online,
D. is due to Cu2+ in the lattice. B. Their color comes from inclusions. please scan the QR code.

G&G CHALLENGE GEMS & GEMOLOGY SPRING 2023 93

 Editor: Evan M. Smith

Plastic Deformation: How and Why

Are Most Diamonds Slightly Distorted?
Extreme hardness is one of the most well-known physical Plastic deformation does not always lead to pink color,
properties of diamond. It is difficult to imagine this same though. More commonly the result is brown color. Our un-
material bending or squishing like putty, but many natural derstanding of brown color in diamond has improved in the
diamonds have evidence of plastic deformation (figure 1). last 20 years, in part due to the need to screen for decol-
This process is even responsible for creating the value be- orizing high-pressure, high-temperature (HPHT) treatment.
hind the most expensive diamonds sold at auction to date, Most mined diamonds have at least some amount of
by both per-carat price and total price. Specifically, the brown color due to plastic deformation, varying from dark
11.15 ct Williamson Pink Star that sold for $57.7 million to barely noticeable (Harris et al., 1979; Fisher, 2009; Do-
and the 59.6 ct Pink Star that sold for $71.2 million owe brinets et al., 2013). In other instances, where the deforma-
their Fancy Vivid pink colors to plastic deformation. tion is less intense, there may be no color imparted. Plastic
deformation is one of the most prevalent features in natu-
ral diamond (Urusovskaya and Orlov, 1964; Harris et al.,
Figure 1. A 4.56 ct rough diamond with plastic defor- 1979). It is a natural phenomenon thought to occur some-
mation lines, highlighted in the sketch on the right. time after a diamond crystallizes, during its residence deep
Deformation has occurred along a set of parallel inter- in Earth’s mantle and/or during its volcanic journey up to
nal slip planes. The planes appear as thin ridges on the surface.
the diamond’s surface, exposed by natural dissolution
of the surface by fluids (i.e., resorption). In fact, the
What Is Plastic Deformation?
whole surface has been resorbed into this overall
dodecahedroid shape, decorated by elongate hillocks When stresses act on a solid, they can cause it to deform, or
and plastic deformation lines, both fine features rep- change shape, in one of three ways. There is elastic defor-
resenting slight interruptions and imperfections in the mation, the kind of deformation that springs are designed
diamond crystal lattice. Images by Evan M. Smith. for. When you remove the stress, it “springs” right back into
its original shape. Brittle deformation is the kind of defor-
mation at play when a water glass falls to the floor and shat-
ters. The material fractures and breaks apart. Finally, there
is plastic deformation, the kind of deformation associated
with putty. The material deforms into a new shape but the
change is lasting, even after removing the stress.

Editor’s note: Questions or topics of interest should be directed to

Evan Smith ([Link]@[Link]).

GEMS & GEMOLOGY, VOL. 59, NO. 1, pp. 94–100.

500 μm
© 2023 Gemological Institute of America

94 DIAMOND REFLECTIONS GEMS & GEMOLOGY SPRING 2023

 Figure 2. Left: Simplified
drawing of a diamond lattice
with a dislocation. The red
dotted circle shows the dislo-
cation where a break in the
regular order of the bonding
structure results in an extra
Slip plane half-plane of carbon atoms,
shown in orange. Stress on
the crystal, indicated by
black arrows, allows stepwise
breaking and reforming of
bonds that causes the dislo-
cation and half-plane to
move to the right along a slip
plane (blue dashed line).
Right: The dislocation has
moved all the way to the
right and out of the crystal so
that it no longer exists. The
Carpet with a wrinkle Moving the wrinkle dislocation and its movement
can be thought of as a carpet
with a wrinkle.

A familiar wire paper clip can serve to demonstrate de- in the kitchen. Although we cannot feel and observe dia-
formation. If you bend it just slightly, it will deform elas- mond deformation in the same way, it involves the same
tically and return to its unbent shape. The elastic behavior basic principles at the atomic scale. During plastic defor-
is what allows a paper clip to hold on to a few sheets of mation, atomic bonds are breaking and re-forming along
paper. If you bend it beyond a certain point, the bend be- imperfections called dislocations. Plastic deformation in
comes permanent. This is plastic deformation. If you bend all crystalline solids, including diamond, is accomplished
it back and forth a few times in one spot, the paper clip will by the creation and movement of dislocations (Nesse,
break. 2017).
Often we see just one of these three mechanisms dom- A dislocation is a linear (one-dimensional) disruption in
inate, whether it is elastic, brittle, or plastic. Deformation the regular, repeating atomic structure. Figure 2 shows a
behavior depends on many factors, such as the tempera- simplified illustration of a dislocation in a crystal lattice.
ture, how much stress is applied, and, of course, what kind The dislocation is shown end-on, and the crystal lattice ap-
of material it is. For diamond, temperature is a key factor. pears to have an extra half-plane of atoms, shown in orange
If you tried to squeeze, twist, or stretch a diamond at room (figure 2), which resembles a half sheet of paper inserted
temperature, the elastic deformation would be minimal. It into an otherwise orderly paper stack. This is an edge dis-
is among the most rigid of all known materials. If it is sub- location, one of two major dislocation types. The second
jected to high stresses, diamond tends to exhibit brittle de- type is a screw dislocation, which has a different geometry.
formation by cracking or chipping. But at higher Dislocations allow a crystal to deform plastically little
temperatures, diamond begins to exhibit plastic deforma- by little as bonds break and re-form. The crystal as a whole
tion. Experiments show that significant plastic deforma- remains fully intact, as only a small number of its bonds
tion requires temperatures above approximately 900°C are broken at any given time. A dislocation is much like a
(DeVries, 1975; Weidner et al., 1994). wrinkle in a carpet (figure 2, bottom). A carpet could be
moved by introducing a wrinkle and working it across the
carpet. Most of the carpet remains in contact with the floor,
How Does Diamond Deform Plastically? but moving the wrinkle shifts the carpet slightly by break-
Many metal objects, including paper clips, deform plasti- ing and reestablishing contact with the floor. Similarly, as
cally under conditions we can easily create and deform to dislocations move through a crystal, they can allow atomic
a degree that is obvious. Our understanding of how crys- layers to move with respect to one another. Systematic
talline solids deform is underpinned by engineering studies movement of many dislocations can allow the whole crys-
of metals, such as steel used for car frames and bridges. tal to deform plastically when subjected to stress.
Aluminum foil is an example of a metal whose plastic de- Given that crystals are orderly structures, it should
formation behavior can be felt as we shape and conform it come as no surprise that dislocation movement is not ran-

DIAMOND REFLECTIONS GEMS & GEMOLOGY SPRING 2023 95

dom. The crystal structure constrains the dislocation move- ally be observed as rows of trigons (figure 3). These negative
ment into certain configurations and directions. In figure 2, trigons are small pits where the diamond has been etched
dislocation movement or “slip” occurs along a plane away preferentially due to the presence of dislocations. Each
(shown as a blue dashed line) that is parallel to the atomic row of trigons marks a separate {111} plane.
layering. In diamond, slip occurs along octahedral or {111} In addition to dislocation movement on {111} planes, a
planes, with the direction of slip within those planes being related plastic deformation mechanism can lead to the
described by <110>, together giving a “slip system” de- crystal lattice being distorted into a totally new orienta-
scribed as {111} <110> (Evans and Wild, 1965). In other tion. This abrupt and localized change in crystal lattice ori-
words, when diamond deforms, the deformation takes place entation is called mechanical twinning or deformation
along internal {111} planes, which is the same orientation twinning. Unlike twins arising from growth, such as
as the faces of an octahedron-shaped diamond crystal. And macles (Harris et al., 2022), mechanical twins in diamond
the three lines defining the edges of that triangular octahe- are strictly from deformation (Titkov et al., 2012; Yu et al.,
dral face have a <110> orientation, which illustrates the 2012). This process is sometimes called glide, and the re-
three possible directions of slip within that octahedral sulting mechanical twin planes are called glide planes
plane. Slip planes will contain many dislocations, not just (Nesse, 2017). In fact, the rows of trigons in figure 3 tech-
one, and deformation can be spread across multiple dislo- nically reveal glide planes, which in this case are associated
cations and multiple slip planes simultaneously. with a pink color. The close-up (figure 3, right) shows that
the trigons are connected by thin pink lines, actually pink
lamellae viewed end-on, that correspond to glide planes.
What Does It Look Like? Pink lamellae like these are typically about 1 micrometer
Plastic deformation involves the creation and movement thick and may consist of multiple closely spaced mechan-
of dislocations through the diamond crystal lattice. Even ical twins (Gaillou et al., 2010).
though this is something that happens at the atomic scale, A more dramatic example of mechanical twinning is
it does leave some visible signs. Figure 1 shows plastic de- shown in figure 4. A sharp step in surface topography runs
formation lines as a series of fine parallel ridges on a dia- all the way around this large rough diamond like a belt and
mond’s surface. These lines trace out where the internal perfectly traces out a {111} plane. This is a ~0.5 mm thick
{111} slip planes meet the outer surface of the diamond. mechanical twin zone, where the diamond has been dis-
They stand out because the diamond surface is resorbed, torted and the lattice reoriented as a result of applied stress.
meaning it has been partially dissolved or etched by natu- Again, the new lattice orientation is not random, but rather
ral fluid or magma (Smit and Shirey, 2020). The slip planes a symmetrical mirror image so that the different lattice ori-
do not etch at the same rate as neighboring regions of the entations match up neatly at the interface between the
crystal, resulting in surface relief and lines visible to the twinned region and the rest of the diamond on either side.
unaided eye. Natural broken cleavage surfaces intersecting the mechan-
Plastic deformation lines (figure 1) are the archetypal fea- ical twin (figure 4, left) demonstrate the change in crystal
ture of plastically deformed diamonds (Harris et al., 2022). orientation by a corresponding change in the cleavage di-
They are frequently but variably observed within some dia- rection within the ~0.5 mm thick twinned zone.
monds from most diamond deposits (Harris et al., 1979). Rough diamonds and their naturally sculpted surfaces
However, only diamonds with resorbed surfaces can show can reveal a lot about their deformation history. Thank-
deformation lines (Gurney, 1989). On primary octahedral fully, since plastic deformation happens internally, it is still
crystal faces, evidence of plastic deformation can occasion- possible to observe it once a diamond is cut and polished.

Figure 3. Left: A 0.31 ct pink

octahedral diamond crystal
with rows of negative
trigons indicating that plas-
tic deformation has taken
place. Right: This close-up
shows the horizontal rows
of trigons marking the loca-
tion of pink {111} glide
planes approximately 1 μm
thick. Photomicrographs by
Evan M. Smith; fields of
view 3.57 mm (left) and
0.71 mm (right).

96 DIAMOND REFLECTIONS GEMS & GEMOLOGY SPRING 2023

Figure 4. Two views of a large rough diamond with a pronounced mechanical twin plane. This 812 ct diamond,
named the Constellation, has both freshly cleaved surfaces (left) and natural resorbed surfaces (right). Both views
show a kink that wraps around the entire surface and defines an internal {111} twin plane where the crystal orien-
tation is reversed with respect to the rest of the diamond. The diamond is 6.6 cm in its longest dimension. Photos
by Jian Xin (Jae) Liao.

Polarizing filters are a simple yet powerful tool for this ap- pattern of anomalous birefringence, such as banding, mot-
plication, whether using a polariscope or a gem microscope tling, or cross-hatching.
(Renfro, 2015). Viewing a diamond using crossed polarizing Plastic deformation can also be observed using cathodo-
filters is an easy way to see anomalous birefringence, luminescence imaging or deep UV fluorescence imaging
which in natural diamond is typically the result of plastic (e.g., DiamondView). These techniques are especially use-
deformation (figure 5). With this technique, a perfect, un- ful for examining polished diamond surfaces, where slip
strained diamond should appear uniformly dark inside. planes can appear as bright or dark lines (figure 6, left). An-
Bending and twisting of the light as it passes through de- other deformation-related pattern sometimes revealed
formed (strained) regions of the diamond will result in a with these methods is dislocation networks (figure 6,

Figure 5. A natural type IIa diamond with evidence of plastic deformation. Examination with a microscope using
crossed polarizing filters (left) reveals internal banding of dark/light and interference colors due to strain-related
anomalous birefringence. In this diamond, the geometry of the banding correlates with the appearance of unusual
whitish graining (right). The hazy whitish graining may be caused by plastic deformation. Photomicrographs by
Evan M. Smith; field of view 4.79 mm.

DIAMOND REFLECTIONS GEMS & GEMOLOGY SPRING 2023 97

Figure 6. Deep UV fluorescence images of deformation features. Left: Deformation lines in the pavilion of a round
brilliant diamond, with two distinct sets of parallel lines that trace out octahedral or {111} slip planes. Right: A dislo-
cation network pattern, appearing as a dark mesh against a bright background in the left portion of the diamond but
as a bright mesh against a dark background toward the right. Images by Evan M. Smith; field of view is 4.66 mm.

right), which have a cellular appearance resembling mud multiple dislocations interact or jogged dislocations move
cracks or a fine mesh (De Corte et al., 2006). Dislocation through the crystal (Leipner et al., 2000, 2003). At the high
networks are thought to be the result of not just plastic de- temperatures necessary for plastic behavior, these vacan-
formation, but also a period of recovery at high tempera- cies are mobile and tend to cluster into groups. These small
tures that allows the dislocations to reorganize into a more groups of vacancies can absorb visible light and cause color.
stable configuration (Hanley et al., 1977). This process may Vacancy clusters made up of about 40−60 vacancies are
be similar to hammering or working metals and then heat- thought to be responsible for the common brown color in
ing them up to relax and reconfigure the tangled disloca- diamond (Avalos and Dannefaer, 2003; Hounsome et al.,
tions that have been introduced. Sublithospheric or 2006; Barnes et al., 2007).
superdeep diamonds often have dislocation network pat- The cause of deformation-related pink is believed to be
terns, and it has been suggested that their formation is pro- similar but involves a broad absorption band centered near
moted by low nitrogen content and/or high temperatures 550 nm that is not fully understood (Gaillou et al., 2010). To
(Smith et al., 2019). be fair, not all brown and pink color is attributable to plastic
deformation, but it is the leading cause (Harris et al., 1979;
Fisher, 2009). An excellent review of brown and pink color
How Does Plastic Deformation Cause Color? in diamond was presented in Eaton-Magaña et al. (2018).
In 1999, General Electric and Lazare Kaplan International
announced that HPHT processing could be applied as a com-
mercial treatment to decolorize brown diamonds (Vagarali What Causes the Deformation?
et al., 2004). This spurred research into the cause of brown It is plain to see that many natural diamonds bear evi-
color, which up until that point had been poorly understood. dence of plastic deformation, but it is not intuitively ob-
Visual indications of deformation had long been recognized vious why this should be the case. There must be some
to correlate with the presence of brown and to a lesser extent geological processes that lead to compression or shearing
pink colors in diamond (Urusovskaya and Orlov, 1964; Har- forces that occasionally deform diamonds. The geological
ris et al., 1979). Internally, these colors often concentrate process should be neither too gentle, because the relatively
along slip planes, glide planes, or less distinctly along “grain- soft silicate host rocks should simply flow around the di-
ing,” all of which are deformation features. Theoretical amonds, nor too violent, which would break diamonds
modeling and experiments played important roles in linking rather than plastically deform them. Several theories have
brown color to dislocation movement and the creation of been proposed.
vacancies in the crystal structure (Avalos and Dannefaer, Kimberlites and related eruptions that bring diamonds
2003; Hounsome et al., 2006; Barnes et al., 2007). from the mantle up to the surface could have the power to
A vacancy is a missing atom in the crystal lattice. Va- deform diamonds. The magmatic journey is thought to be
cancies can be created during plastic deformation when relatively violent, however, and it is clear that diamonds

98 DIAMOND REFLECTIONS GEMS & GEMOLOGY SPRING 2023

can break as they ascend within a kimberlite (Harris et al., The deformation histories recorded within natural di-
2022). This may mean that the conditions are too forceful amond are valuable for studying Earth’s interior. In the fu-
and chaotic to develop significant plastic deformation. The ture, it may be possible to constrain the temperature,
early stages leading up to an eruption could also provide pressure, and strain rate (i.e., the speed of deformation) re-
opportunities for diamond deformation. quired to produce the various features we observe. This
Another possibility is that diamond deformation is re- information about deep mantle processes would help paint
lated to the movements at the lithosphere-asthenosphere a clearer picture of how and why rocks deep inside the
boundary, where the base of a rigid continental tectonic earth move around over geologic time.
plate meets underlying weak, mobile rocks at a depth of ap-
proximately 200 km (Stachel et al., 2018). This would in-
volve shearing between the lithosphere, the rigid host rock Diamonds Bear Witness to Our Active Planet
where most diamonds form, and the asthenosphere, the un- The fact that natural diamonds frequently contain evi-
derlying warm flowing rock that is the topmost part of the dence of deformation reminds us that Earth is not a static
convecting mantle. ball of rock. As far as we know, Earth is the only planet
Yet another theory suggests that deformation in some with active plate tectonics, meaning its outer layer is made
settings might be related to the movement of tectonic up of multiple moving plates (Condie and Pease, 2008). The
plates, specifically oceanic lithosphere sinking into the processes of mantle convection and subduction mean
mantle by the process of subduction (Bulanova et al., 2018). Earth’s interior is on the move as well. Measurable effects
Subduction and the motion of tectonic plates could also of these mechanisms exist in the chemical and isotopic
play a role in the deformation of super-deep diamonds. The composition of diamonds and their inclusions. But what
formation of these diamonds has been linked to deep-focus makes a diamond’s plastic deformation features special is
earthquakes (Shirey et al., 2021), which could involve lo- that they are a visible, tangible testament to rock and
calized shearing and faulting, potentially favorable condi- magma churning, shearing, and moving about inside our
tions for plastic deformation of diamonds. active planet.

ACKNOWLEDGMENT
Many thanks to Dr. Steven B. Shirey for discussion and suggestions that helped to improve this column.

REFERENCES
Avalos V., Dannefaer S. (2003) Vacancy-type defects in brown dia- Willems B. (2006) Overview of dislocation networks in natural
monds investigated by positron annihilation. Physica B: Con- type IIa diamonds. G&G, Vol. 42, No. 3, pp. 122–123.
densed Matter, Vol. 340-342, pp. 76–79, DeVries R.C. (1975) Plastic deformation and “work-hardening” of
[Link] diamond. Materials Research Bulletin, Vol. 10, No. 11, pp.
Barnes R., Bangert U., Scott A. (2007) Investigating large vacancy 1193–1199, [Link]
clusters in type IIa diamond with electron energy loss spec- Dobrinets I.A., Vins V.G., Zaitsev A.M. (2013) HPHT-Treated Dia-
troscopy (EELS). Physica Status Solidi (a), Vol. 204, No. 9, pp. monds: Diamonds Forever. Springer Series in Materials Science,
3065–3071, [Link] Springer Berlin, Heidelberg, [Link]
Bulanova G.P., Speich L., Smith C.B., Gaillou E., Kohn S.C., Wib- 642-37490-6
berley E., Chapman J.G., Howell D., Davy A.T. (2018) The Eaton-Magaña S., Ardon T., Smit K.V., Breeding C.M., Shigley J.E.
unique nature of Argyle fancy diamonds: Internal structure, (2018) Natural-color pink, purple, red, and brown diamonds:
paragenesis, and reasons for color. In A.T. Davy et al., Eds., Band of many colors. G&G, Vol. 54, No. 4, pp. 352–377,
Geoscience and Exploration of the Argyle, Bunder, Diavik, [Link]
and Murowa Diamond Deposits, Volume 20. Society of Eco- Evans T., Wild R.K. (1965) Plastic bending of diamond plates. Philo-
nomic Geologists, Littleton, Colorado, pp. 169–190, sophical Magazine: A Journal of Theoretical Experimental and
[Link] Applied Physics, Vol. 12, No. 117, pp. 479–489,
Condie K.C., Pease V., Eds. (2008) When Did Plate Tectonics Begin [Link]
on Planet Earth? Geological Society of America, Special Paper Fisher D. (2009) Brown diamonds and high pressure high tempera-
440, 294 pp. ture treatment. Lithos, Vol. 112, pp. 619–624,
De Corte K., Anthonis A., Van Royen J., Blanchaert M., Barjon J., [Link]

DIAMOND REFLECTIONS GEMS & GEMOLOGY SPRING 2023 99

Gaillou E., Post J.E., Bassim N.D., Zaitsev A.M., Rose T., Fries Nesse W.D. (2017) Introduction to Mineralogy. Oxford University
M.D., Stroud R.M., Steele A., Butler J.E. (2010) Spectroscopic Press, New York.
and microscopic characterizations of color lamellae in natural Renfro N. (2015) Digital photomicrography for gemologists. G&G,
pink diamonds. Diamond and Related Materials, Vol. 19, No. Vol. 51, No. 2, [Link]
10, pp. 1207–1220, [Link] Shirey S.B., Wagner L.S., Walter M.J., Pearson D.G., van Keken
[Link].2010.06.015 P.E. (2021) Slab transport of fluids to deep focus earthquake
Gurney J.J. (1989) Diamonds. In J. Ross, Ed., Kimberlites and Re- depths—Thermal modeling constraints and evidence from
lated Rocks: Proceedings of the Fourth International Kimber- diamonds. AGU Advances, Vol. 2, No. 2,
lite Conference, Volume 2. Geological Society of Australia, [Link]
Sydney, Australia, pp. 935–965. Smit K.V., Shirey S.B. (2020) Diamonds are not forever! Diamond
Hanley P.L., Kiflawi I., Lang A.R. (1977) On topographically iden- dissolution. G&G, Vol. 56, No. 1, pp. 148–155.
tifiable sources of cathodoluminescence in natural diamonds. Smith E.M., Shirey S.B., Richardson S.H., Nestola F., Bullock E.S.,
Philosophical Transactions of the Royal Society of London. Se- Wang J., Wang W. (2019) Reply to: Evidence for two blue (type
ries A, Mathematical and Physical Sciences, Vol. 284, No. 1324, IIb) diamond populations. Nature, Vol. 570, No. 7760, pp. E28–
pp. 329–368, [Link] E29, [Link]
Harris J.W., Hawthorne J.B., Oosterveld M.M. (1979) Regional and Stachel T., Harris J.W., Hunt L., Muehlenbachs K., Kobussen
local variations in the characteristics of diamonds from some A.F., Edinburgh Ion Micro-Probe Facility (2018) Argyle dia-
southern African kimberlites. In H.O.A. Meyer and F.R. Boyd, monds: How subduction along the Kimberley Craton Edge
Eds., Kimberlites, Diatremes, and Diamonds: Their Geology, generated the world’s biggest diamond deposit. In A.T. Davy,
Petrology, and Geochemistry, Vol. 15, American Geophysical et al., Eds. Geoscience and Exploration of the Argyle, Bunder,
Union, Washington, DC, pp. 27–41, Diavik, and Murowa Diamond Deposits, Volume 20. Society
[Link] of Economic Geologists, Littleton, Colorado, pp. 145–168,
Harris J.W., Smit K.V., Fedortchouk Y., Moore M. (2022) Morphol- [Link]
ogy of monocrystalline diamond and its inclusions. Reviews in Titkov S.V., Krivovichev S.V., Organova N.I. (2012) Plastic defor-
Mineralogy and Geochemistry, Vol. 88, No. 1, pp. 119–166, mation of natural diamonds by twinning: Evidence from X-ray
[Link] diffraction studies. Mineralogical Magazine, Vol. 76, No. 1, pp.
Hounsome L.S., Jones R., Martineau P.M., Fisher D., Shaw M.J., 143–149, [Link]
Briddon P.R., Öberg S. (2006) Origin of brown coloration in dia- Urusovskaya A.A., Orlov Y.L. (1964) Nature of plastic deformation
mond. Physical Review B, Vol. 73, No. 12, article no. 125203, of diamond crystals. Doklady Akad. Nauk. SSSR, Vol. 154, pp.
[Link] 112–115.
Leipner H.S., Hübner C.G., Staab T.E.M., Haugk M., Sieck A., Vagarali S., Webb S., Jackson W., Banholzer W., Anthony T., Kaplan
Krause-Rehberg R., Frauenheim T. (2000) Vacancy clusters in G. (2004) High pressure/high temperature production of color-
plastically deformed semiconductors. Journal of Physics: Con- less and fancy-colored diamonds. United States Patent 6692714.
densed Matter, Vol. 12, No. 49, pp. 10071–10078, Weidner D.J., Wang Y., Vaughan M.T. (1994) Strength of diamond.
[Link] Science, Vol. 266, No. 5184, pp. 419–422,
Leipner H.S., Mikhnovich Jr. V.V., Bondarenko. V., Wang Z., Gu H., [Link]
Krause-Rehberg R., Demenet J.L., Rabier J. (2003) Positron an- Yu X., Raterron P., Zhang J., Lin Z., Wang L., Zhao Y. (2012) Con-
nihilation of defects in silicon deformed at different tempera- stitutive law and flow mechanism in diamond deformation.
tures. Physica B: Condensed Matter, Vol. 340-342, pp. 617–621, Scientific Reports, Vol. 2, article no. 876,
[Link] [Link]

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

[Link]/gems-gemology

100 DIAMOND REFLECTIONS GEMS & GEMOLOGY SPRING 2023

Knowledge and
Skills f or Today’s 
Hot Topics

Online and In-Person Seminars Designed

for Gem and Jewelry Professionals
• Natural and Laboratory- • Colored Stones
Grown Diamonds • Pearls
• Jewelry Design • Bench Skills

Visit [Link]/seminars GIA Alumni receive a 10% discount.

for schedules and more information. Seminars are open to all.

C ON T INUING E DUC AT ION B Y

These seminars are avocational and are not intended to prepare individuals for entry-level [Link] are non-credit bearing and do not lead to a GIA academic credential.
They do not fall under the approval of the Accrediting Commission of Career Schools and Colleges (ACCSC) or the Distance Education Accrediting Commission (DEAC).
© 2023 Gemological Institute of America, Inc. All trademarks owned by Gemological Institute of America, Inc. and used under license.
 GEM NEWS INTERNATIONAL
Contributing Editors
Gagan Choudhary, IIGJ-Research & Laboratories Centre, Jaipur, India ([Link]@[Link])
Christopher M. Breeding, GIA, Carlsbad ([Link]@[Link])
Guanghai Shi, School of Gemmology, China University of Geosciences, Beijing (shigh@[Link])

Tahitian, and “Edison” pearls in both venues. Most no-

TUCSON 2023 tably, “Edison” pearls were available in much larger sizes
and with more saturated colors than in years past (figure
2). Non-nacreous pearls were also featured more promi-
This year’s Tucson shows were well attended, and many nently. Ethiopian opal, first introduced in Tucson in 2009,
vendors were relieved by the strong sales. In fact, the only was very popular this year. It has climbed from a novelty
“complaint” from vendors this year was that for many it to a gemstone found in top designer luxury jewels, such as
was too busy to take even a short break away from their the Zoltan David necklace shown on the cover of this issue
booths. Some vendors reported selling less than in 2022, and the pin featuring both Ethiopian opal and conch pearls
but still more than in pre-COVID years.
Supply chain issues resulting from lack of output during
Figure 1. This ring from Graff features a 3.95 ct pink
the pandemic are still being felt. For many sellers demand
sapphire and 7.5 carats of diamond in a spiral design.
often exceeded supply, especially for fine-quality colored
Photo by Robert Weldon; courtesy of Jewelerette Co.
stones, and prices are increasing to reflect that. With this in-
crease expected to continue, the heavy booth traffic could
be due to consumers buying as much as possible at these
price points. This corresponds with comments from Nathan
Renfro on the GIA show services laboratory, which received
a near-record number of submissions this year. As for the
trends in materials submitted to the lab, he indicated a typ-
ical mix of the major stones: mainly ruby, sapphire, emerald,
Paraíba tourmaline, and alexandrite.
At the AGTA GemFair and Gem & Jewelry Exchange
(GJX) shows, sapphire, large freshwater pearls from China,
and Ethiopian opal were among the eye-catching gems.
Sapphire was once again a top seller, with blue and un-
treated colors doing especially well. Many fancy-color sap-
phires, including ones in saturated hues such as magenta,
were available (Pantone’s Color of the Year for 2023 is Viva
Magenta) (figure 1). There were multiple sellers of akoya,

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

trations to Stuart Overlin at soverlin@[Link].
Banner photo of Afghan kunzite by Robert Weldon; courtesy of Dudley
Blauwet, Mountain Minerals International.
GEMS & GEMOLOGY, VOL. 59, NO. 1, pp. 102–162.
© 2023 Gemological Institute of America

102 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023

Figure 2. Left: “Edison” pearls were available in large sizes and attractive colors. Photo by Si Athena Chen; cour-
tesy of A&B Jewelry. Right: A strand of “Edison” pearls of various colors, in sizes from 14 to 16 mm. Photo by
Mimi Travis; courtesy of Yen’s Jewelry & Accessories.

in figure 3. Tourmaline from newly discovered pockets in The demand for one-of-a-kind gems was apparent in all
San Diego County’s Pala District was also on display. aspects—cut, color, and even inclusions. Nontraditional
cuts such as slices, fantasy cuts, rose cuts, kites, and tablets
were readily found in gems ranging from rutilated quartz
Figure 3. This pin features a 63.92 ct Ethiopian opal to diamond (figure 4). Beyond nonstandard cuts, stones
with 23.85 carats of conch pearl, 5.88 carats of natural
pearl, and 15.85 carats of diamond, set in platinum.
Photo by Robert Weldon; courtesy of Pioneer Gems of Figure 4. Earrings featuring diamond slices and rose-
New York. cut rubies. Photo by Mimi Travis.

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023 103

 awareness, which has long been a key differentiator for
some jewelers, is clearly becoming more mainstream as
customers increasingly demand to know more about the
origin and impact of the stones they are purchasing. This
year, jewelers at AGTA and GJX also emphasized the im-
portance of providing an environmentally conscious prod-
uct and the increasing demand from clients for full
chain-of-custody information on stones. Similarly, many
small-scale designers and buyers were seeking out gems
mined in the United States, such as Montana sapphire,
Oregon sunstone, and amethyst from the Four Peaks mine
in Arizona. They noted high demand for these gems, which
seem to satisfy consumer sustainability concerns. In many
cases, the younger generation of jewelers has brought a
focus on ethical and environmentally sound practices to
multigenerational businesses.
We hope you enjoy our coverage of the 2023 Tucson
Figure 5. “Salt and pepper” diamonds containing eye- gem shows, where you’ll find striking pieces we saw, the
visible inclusions. Photo by Jennifer Stone-Sundberg; latest trends, mining updates, and more.
courtesy of Misfit Diamonds. The following contributed to this report: Lisa Neely,
Mimi Travis, Nathan Renfro, Albert Salvato, and Wim
Vertriest.
with obvious inclusions were very popular. The appeal of Jennifer Stone-Sundberg, Si Athena Chen,
these stones centers on their uniqueness. “Salt and pepper” Lisa Kennedy, Robert Weldon, and Erin Hogarth
diamonds with multiple eye-visible white, gray, and black
inclusions were available loose (figure 5) and in jewelry
items throughout the shows. Similarly, nontraditional col- COLORED STONES AND ORGANIC MATERIALS
ors for particular varieties, including color-change stones,
pastel colors (such as “mint” tourmaline), teal, and bicolor Enormous cat’s-eye aquamarine. On the AGTA show’s
stones, were trending (figure 6). Strong demand for these opening day, we were on hand as Gary Bowersox and Kath-
nontraditional or one-of-a-kind stones comes from small- leen Kolt-Bowersox donated an extraordinary cat’s-eye
scale individual designers with the ability to promote to a aquamarine to Dr. Jeffrey Post, mineralogist and curator-
wide audience via social media. in-charge of gems and minerals at the Smithsonian Na-
In 2020, we reported on the first Ethical Gem Fair in tional Museum of Natural History (figure 7).
Tucson. This year, suppliers at the fair were pleasantly sur- The 586.43 ct untreated transparent light greenish blue
prised by the number of large jewelry manufacturers who aquamarine with a sharp cat’s-eye is from the Pech Valley
attended for the first time. The trend toward responsible pegmatite mine in Afghanistan. The rough (figure 8, left)
sourcing, traceability, transparency in practices, and social weighed approximately 7,700 ct and was kept for 15 years

Figure 6. An unusual 6.95 ct cushion-cut bicolor alexandrite from Madagascar displaying a color change from yel-
lowish green and reddish brown in daylight (left) to yellowish orange and orangy brown in incandescent illumina-
tion (right). Photos by Robert Weldon; courtesy of Bryan Lichtenstein, 3090 Gems, LLC.

104 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023

 Figure 7. Left: A 586.43
ct cat’s-eye aquamarine
donated by Gary Bower-
sox and Kathleen Kolt-
Bowersox to the
Smithsonian National
Museum of Natural His-
tory gem collection.
Photo by Robert Weldon;
courtesy of the Smith-
sonian National Mu-
seum of Natural History.
Right: Dr. Jeffrey Post
(left) and Gary Bowersox
with the cat’s-eye aqua-
marine. Photo by Kath-
leen Kolt-Bowersox.

before it was finally cut in 2017 by Rohitha Perera in Sri 9). Processed and brought to market as tumble-polished
Lanka (figure 8, right). The Bowersoxes have several other freeforms, the chalcedony was offered by Orbit Ethiopia Plc.
cut gems from this find, with the next largest weighing in According to managing director Tewodros Sintayehu, the
at 541.96 ct. mine is located in the Gamo Zone region, approximately
Jennifer Stone-Sundberg and Si Athena Chen 600 km from Addis Ababa. The material was first discov-
GIA, Carlsbad ered in the Gerese District and has since been found in the
adjacent districts of Kamba and Garda Marta. Sintayehu
noted that he had a very large quantity of the material
New violet orbicular chalcedony from Ethiopia. At the GJX stockpiled.
show, the author encountered a relatively new find of violet The notable features of this material were its fairly vivid
chalcedony reportedly discovered in Ethiopia in 2019 (figure violet color as well as the multitude of white orbicular struc-

Figure 8. Left: Kathleen

Kolt-Bowersox holding
the 7,700 ct aquamarine
rough before it was cut.
Right: The rough aqua-
marine was cut and pol-
ished in Sri Lanka by
Rohitha Perera. Photos
by Gary Bowersox.

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023 105

Figure 9. This relatively new type of violet chalcedony Figure 10. White orb-like structures were confined to
from Ethiopia (here, 75–198 ct) displays a wonderful a shallow surface layer in the violet chalcedony from
pattern of white orbs. Photo by Annie Haynes; cour- Ethiopia. Photomicrograph by Nathan Renfro; field of
tesy of Orbit Ethiopia Plc. view 15.16 mm.

tures distributed across the surface of the gems (figure 10). tions for harvesting coral are in place today. Harvesting is
Microscopic examination revealed that the white color was limited to coral with trunks having a basal diameter of at
related to a distinct botryoidal structure where a shallow least 7 mm recovered from 50 meters or deeper by licensed
outer layer of the gems contained all of the white areas. Be- scuba divers.
cause the white orbs are confined to a shallow surface layer, Precious coral is the exoskeleton of very tiny marine
this explains why the material was sold as tumble-polished invertebrates, known as coral polyps. When a colony of
nuggets: Cutting the material into cabochons would likely polyps arrives in a particular area, they start to create the
result in the loss of the white orbicular pattern. A few exoskeletons on rocks, growing upward in a branch-like or
unidentified mineral inclusions were also observed with the fan formation. Di Gennaro mentioned that divers do not
microscope, but the cause of the outer layer’s white orb pat- harvest coral with active colonies as these will continue
tern remains unclear. to grow and the future, larger coral will be more valuable.
This new material is quite remarkable for its beautiful Divers therefore tend to harvest the coral where the colony
pattern. Even if it is only made available as freeform tumble- has already left. Unlike other types of corals, such as com-
polished stones, it will be a welcome addition to the chal- mon and reef, these Mediterranean “deep sea” precious
cedony market. corals are not a natural habitat for other marine species.
Nathan Renfro Reef coral species, which are a natural environment to
GIA, Carlsbad many different marine species, are in many cases endan-
gered and must be preserved.
Italian precious coral. Coral from the Mediterranean Sea Di Gennaro explained that the quality of the coral de-
has been harvested and used for adornments for thousands pends on the locality. He showed us a piece of dark red
of years. More than 7,000 species are categorized as pre- coral (figure 11, left), which is the typical color of coral har-
cious, common, or reef coral. Precious coral is defined by vested from northern Sardinia. It is a darker red due to
the World Jewellery Confederation (CIBJO) Coral Commis- cooler water temperatures from strong currents in the area.
sion as “those that are used in jewelry and decoration, Coral from the southern portion of Sardinia (figure 11, cen-
specifically red, pink, and white varieties with porcelain- ter) tends to be lighter in color. Di Gennaro described this
like luster after polishing.” Italian coral harvested from the as a more “classic” color of Mediterranean coral. He also
Corallium rubrum species is considered precious. showed us a piece of reddish orange coral called “Sciacca,”
At the GJX show, the authors spoke with Sergio Di named for a town in southern Sicily where the coral is har-
Gennaro of Fulvio Di Gennaro Srl. (Torre del Greco, vested (figure 11, right). Di Gennaro said that volcanic ac-
Naples) about the Italian coral market. The Mediterranean tivity below the surface results in warmer water
has been precious coral’s major source since it was first col- temperatures, which in turn causes a more orangy color.
lected for ornamentation over 3,000 years ago, followed by Di Gennaro emphasized the importance of sustainability
Taiwan and Japan with smaller quantities. Beginning in the in the coral industry, noting the heritage of the city of Torre
late 1970s, Mediterranean coral was threatened by over- del Greco, which was built on the art of processing coral.
harvesting. Di Gennaro explained that many strict regula- While current restrictions have reduced the amount of coral

106 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023

 Figure 11. Mediter-
ranean coral from dif-
ferent regions. Left:
Dark red fine coral
from the northern coast
of Sardinia. Center two
pieces: “Classic” coral
from the southern coast
of Sardinia. Right:
Orangy coral, termed
“Sciacca,” from the
coast of southern Sicily.
Photo by Lisa Kennedy;
courtesy of Fulvio Di
Gennaro Srl.

being harvested compared to decades past, sustainable pro- change like this one is extraordinary,” he said. “A size of 35
duction will guarantee that we continue to see stunning carats is what makes it outstanding.”
Mediterranean precious coral for decades to come.
Lisa Kennedy and Cristiano Brigida
Figure 12. A 155 g aquamarine crystal containing a liq-
GIA, New York
uid and gas bubble inclusion. The crystal measures
11.9 cm long and the inclusion 4.3 cm long. Photo by
Exceptional gemstone acquisitions by the Kreis family. Robert Weldon; courtesy of Kreis Jewellery GmbH.
Kreis Jewellery (Niederwörresbach, Germany) is known for
master cutting expertise and distinctive jewelry design (see
Spring 2022 GNI, pp. 86–87). At the GJX show, Stefan and
Alexander Kreis showed us three recent gemstone acquisi-
tions.
First was a 155 g aquamarine crystal containing a two-
phase liquid and gas inclusion measuring 4.3 cm long (figure
12; see video at [Link]/gems-gemology/spring-2023-
gemnews-kreis-family-gemstones). The crystal was found
in the 1960s in Marambaia, Minas Gerais, Brazil, and had
been part of a private collection before the Kreises obtained
it in 2022.
“It would even be a cuttable crystal,” Alexander said.
“Of course we would never touch it because of the spe-
cialty of the huge liquid inclusion. A liquid chamber that
size in quartz would already be amazing,” he continued.
“But having it in an aquamarine, which is totally intact—
you have the complete termination. And look at the sheer
size of that crystal.”
Marambaia was known for large quantities of aquama-
rine by 1900 and produced the 110.5 kg (244 lb), 552,500 ct
Papamel, discovered in 1910, the largest known gem-quality
aquamarine crystal. The crystal that was the basis for the
Dom Pedro, the largest cut aquamarine, weighed around 45
kg (100 lb) and was also from Marambaia.
Stefan said they had received a 35.05 ct cat’s-eye alexan-
drite (figure 13) that morning. He noted the size, exception-
ally sharp eye, and superb color change for Sri Lankan
alexandrite, which typically has a weaker color change than
material from Russia and Brazil. Alexander pointed out that
cat’s-eyes are less common in alexandrite than in other
chrysoberyl. “A cat’s-eye alexandrite with a strong color

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023 107

 Figure 13. A 35.05 ct Sri
Lankan cat’s-eye alexan-
drite in daylight (left)
and incandescent light
(right). Photo by Robert
Weldon; courtesy of
Kreis Jewellery GmbH.

The Kreises also showed us a 7.08 ct padparadscha sap- exclusively in White Cliffs and typically occur within
phire (figure 14). “The color mixture of this stone is a real weathering-bleached siltstones and claystones. According
padparadscha color,” Stefan said. Alexander added that a to Dowton, no other type of gem-quality opal was found
range of colors fall into the padparadscha category, but in the same layer where the pineapple opals were mined.
“this is the perfect color.” During mining, it was noticed that two opals had formed
The cat’s-eye alexandrite and padparadscha sapphire will within a few feet of each other but with distinctly different
undoubtedly inspire exquisite jewelry designs from Sonja coloration: One displayed a rich spectrum of play-of-color,
and Vanessa Kreis. Sonja noted, “It is a very intuitive process while the other showed almost no coloration. The 3,510
in which we try to bring out a gem’s natural beauty in such ct Heart of Australia (figure 15) shows rich play-of-color
a way that it blurs the boundaries between jewelry and art.” and is considered one of the finest specimens of this type
Erin Hogarth ever unearthed. In some of Dowton’s samples, the precur-
GIA, Carlsbad sor calcite was not fully substituted by silica gel, and
white calcite crystals can be observed (figure 16).
Ikaite, CaCO3·6H2O, is a rare and metastable hydrated
Rare double pseudomorph ikaite-calcite-opal. At the
carbonate in sedimentary rocks that has only been identified
Pueblo Gem & Mineral Show, Graeme Dowton of Red
in environments ranging from –2° to 7°C in nature (M.L.
Earth Opal (White Cliffs, Australia) exhibited rare double
Vickers et al., “The ikaite to calcite transformation: Impli-
pseudomorph ikaite-calcite-opal gem specimens mined
cations for paleoclimate studies,” Geochimica et Cos-
from the Naatji Nest mine at White Cliffs in New South
mochimica Acta, Vol. 334, 2022, pp. 201–216). The
Wales, Australia. Ikaite was initially substituted with cal-
occurrence of ikaite suggests a period of very cold to near-
cite and then opalized, therefore making these double
freezing paleoclimate conditions in White Cliffs. At ambient
pseudomorph opals. Also called “pineapple opal” in the
market due to the form of their clusters, they are found
Figure 15. The Heart of Australia, a 3,510 ct double
pseudomorph ikaite-calcite-opal gem specimen, is
Figure 14. A 7.08 ct padparadscha sapphire. Photo by considered one of the finest ever unearthed. Courtesy
Robert Weldon; courtesy of Kreis Jewellery GmbH. of Graeme Dowton.

108 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023

 The pearl industry, like others, was adversely affected
by the COVID-19 pandemic. Many dealers noted a supply
shortage of the main types of cultured pearls: South Sea,
Tahitian, akoya, and freshwater. According to dealers,
many pearl farms, especially Tahitian, use Chinese and
Japanese technicians to perform the critical nucleation
step. Due to travel restrictions and border closures during
the pandemic, many experts were unable to travel to these
locations, resulting in smaller harvests since 2020. In some
farming areas, there were not enough employees on-site to
keep the mollusks in good health, and thus the mortality
rate was high. Assuming grafting returned to normal levels
by 2022, a greater influx of supply is expected in 2024,
given the typical 24-month growth periods. Some dealers
Figure 16. The intergrowth of opal and its precursor added that strong demand for cultured pearls from Chinese
calcite in pineapple opal. This specimen is 12 × 10 × 8 and other Asian consumers is contributing to the current
cm and weighs 690 g. Photo by Si Athena Chen; cour- global shortage.
tesy of Graeme Dowton. Small freshwater BC pearls from China (approximately
2–4 mm) were more widely available at the show than in
past years, in near-round to round shapes and in various
temperatures (10°–30°C), ikaite transforms to more stable colors including white, pink, purple, and orange (figure 17).
carbonate polymorphs such as calcite, aragonite, and/or va- Freshwater cultured pearls of this particular size are typi-
terite. The ikaite and its following pseudomorphs act as pa- cally NBC, normally oval (a potato-like shape), and can
leothermometers (D. Shearman and A. Smith, “Ikaite, the look similar to white saltwater Japanese akoya BC pearls
parent mineral of jarrowite-type pseudomorphs,” Proceed- of the same size. However, Gina Latendresse of American
ings of the Geologists Association, Vol. 96, No. 4, 1985, pp. Pearl Company Inc. (Nashville, Tennessee) demonstrated
305–314). When the host sedimentary rocks were weath-
ered, they released silica into groundwater. This silica-bear-
ing groundwater contacted with calcite, which gradually
Figure 17. Near-round to round small freshwater
dissolved and reprecipitated to opal over time (B. Pewkliang
bead cultured pearls (approximately 2–4 mm) were
et al., “The formation of precious opal: Clues from the opal-
widely available at the AGTA show in various colors
ization of bone,” Canadian Mineralogist, Vol. 46, No. 1,
including white, pink, purple, and orange. Photo by
2008, pp. 139–149). However, the detailed formation mech-
Artitaya Homkrajae; courtesy of Betty Sue King,
anism of ikaite-calcite-opal in White Cliffs requires future
King’s Ransom.
investigation.
Si Athena Chen and Jennifer Stone-Sundberg

Cultured pearl market update. At the AGTA show, we

spoke with several pearl dealers to acquire some insight
into the current cultured pearl market. With demand for
pearls trending upward in the secondary market, as re-
ported by The Real Real in its Luxury Consignment Report
in January 2023, it was easy to spot similar trends in other
areas of the pearl market as well.
Bead cultured (BC) pearls are the result of the deliberate
insertion of a bead nucleus along with a small piece of
mantle tissue into the gonad of a living host mollusk to
start the growth of the cultured pearl, a process known as
nucleation or grafting. Non-bead cultured (NBC) pearls
need only a piece of tissue implanted into the gonad or
mantle of a host mollusk to form. A technique originally
developed and successfully operated in Japan before adop-
tion by other global operations including Australia, the
Philippines, Indonesia, French Polynesia, and China, the
nucleation process is intricate and requires expertly trained
technicians.

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023 109

 Figure 18. Left: Small
round freshwater bead
cultured (BC) pearls (top
group of strands) and
small round Japanese
akoya BC pearls (bot-
tom strands) appear very
similar at first glance.
Right: Upon closer in-
spection, the freshwater
pearls more often show a
rounded flat surface fea-
ture. Photos by Artitaya
Homkrajae and Lisa
Kennedy; courtesy of
American Pearl Com-
pany Inc.

that upon closer inspection, the freshwater pearls more tion of colors compared to traditional akoya BC pearls. Some
often show a rounded flat surface feature on an otherwise white, light cream, and cream pearls were also presented as
spherical pearl, a feature usually too small to affect basic unprocessed and untreated—steps normally used to improve
shape (figure 18). This feature is also occasionally observed appearance and alter color. However, many akoya BC pearls
in other types of freshwater cultured pearls. in the market have been processed to generate a classic
Eliko Pearl Company (New York City) displayed numer- white appearance. Vietnam has produced saltwater cultured
ous strands of Vietnamese akoya BC pearls with bodycolors pearls commercially since the 1990s and is expected to pro-
ranging from silver to light gray and light bluish gray with duce approximately 2,000 kg per annum (N. Sturman et al.,
exceptional orient (figure 19). These provide a greater selec- “Vietnam: Shell nuclei, pearl hatcheries, and pearl farming,”
Fall 2020 G&G, pp. 402–415), which is significantly smaller
than Japan’s anticipated annual production of approximately
Figure 19. Vietnamese akoya bead cultured pearls in 20 tons (roughly 18,000 kg) (T. Matsuyama et al., “Mass
various colors ranging from silver to light gray to light mortality of pearl oyster (Pinctada fucata (Gould)) in Japan
bluish gray. The pearls were said to be unprocessed in 2019 and 2020 is caused by an unidentified infectious
and untreated. Photo by Lisa Kennedy; courtesy of agent,” PeerJ, Vol. 9, 2021, article no. 12180).
Eliko Pearl Company. Eric Yen of Yen’s Jewelry & Accessories Inc. (San Fran-
cisco) and Alan Hakimian of Yoko London presented
strands that combined multicolor pearls from different
pearl types, offering variety and high fashion all at once (fig-
ures 20 and 21). Both pointed out that making mixed and
graduated color strands is an efficient use of all the pearls
produced from a harvest, just one example of economic
sustainability in the pearl industry. Producing only per-
fectly matched color stands would lead to more waste.
Overall, pearl dealers are looking forward to the return
of the gem, jewelry, and pearl shows in Hong Kong in 2023,
as these are some of the largest shows for cultured pearls.
Many have not been able to attend since 2019 due to lock-
down restrictions in Hong Kong.
Lisa Kennedy
Artitaya Homkrajae
GIA, Carlsbad

“Edison” pearls: Increasing market presence. Several ven-

dors at both the AGTA and GJX shows carried large and at-
tractively colored “Edison” pearls from China (see C. Zhou

110 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023

Figure 20. This set of ombré pearl strands, made up
of round Tahitian pearls with silver and gold South
Sea pearls ranging from 10 to 10.4 mm, is also known
as shikisai, which means “colors of four seasons” in
Japanese. Photo by Mimi Travis; courtesy of Yen’s
Jewelry & Accessories Inc.
Figure 21. Mixed strand of 79 pearls ranging from 11.8
to 14.8 mm, consisting of pink Chinese freshwater,
black Tahitian, golden Indonesian, and white Aus-
et al., “Detection of color treatment and optical brightening
tralian bead cultured pearls. Also pictured is an ex-
in Chinese freshwater ‘Edison’ pearls,” Summer 2021 G&G,
ceptionally large 20.4 mm South Sea bead cultured
pp. 124–134). These round freshwater bead cultured pearls
pearl, weighing 12 g. Photo by Robert Weldon; cour-
generally ranged in size from 9 to 14 mm. Some baroque
tesy of Yoko London.
pearls were also offered. This year we saw deeper hues and
more purple, bronze, copper, peach, and orangy pink colors
than in the past; lighter colors such as white, cream, and
light pink were more scarce. Popular trends included “rain- “Edison” pearls in sizes over 14 mm are rare and consider-
bow” strands of pearls featuring a mix of colors, strands with ably more costly. That said, she noted that “Edison” pearls
alternating blocks of color, and strands with metallic colors are still a good value, available in an appealing variety of
(figure 22). At GJX, we spoke with Amy Hansen of A&B Jew- colors and sizes with a price point substantially below
elry (Honolulu, Hawaii) to learn about current purchasing akoya and Tahitian saltwater cultured pearls.
trends and the latest developments at the pearl farms. Hansen explained that the cultivation of “Edison”
Hansen indicated that “Edison” pearls are still finding pearls in terms of bead size and culturing times is similar
their place in the market. Her designer clients are seeking to that for saltwater Tahitian pearls. She described im-
larger pearls, and those in the 12–14 mm size range are provements in the process, including Chinese freshwater
doing particularly well, with pearls larger than 13 mm in pearl farmers moving inland to access cleaner rivers and
highest demand. She explained the limits in producing lakes and using more sound environmental practices. The
larger pearls: Larger sizes (particularly greater than 14 mm move to these newer locations has resulted in more colors
in diameter) take much more time to culture, which greatly being achieved, broadening the offerings of these pearls.
limits the yield of top-grade product with respect to shape, Chinese freshwater cultured pearl production has also de-
nacre, surface, and luster. As a result, the highest-quality creased over the past five years as the focus has shifted

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023 111

 At the AGTA show, Alex Vock of ProVockative Gems
Inc. (New York City) shared some of his expertise in natu-
ral saltwater nacreous pearls. With more than 30 years in
the business, he is considered an industry leader specializ-
ing in signed collectible jewelry, natural gemstones, and
natural pearls. He had several pieces on display, including
an attractive pair of natural pearl earrings set with un-
heated Burmese rubies and unoiled emeralds (figure 23).
Vock discussed the global market for natural nacreous
pearls, noting that it was greatly diminished by the 1980s
due to the prevalence of cultured pearls. However, regions
such as many of the Gulf countries and India did not feel
pressure from the cultured pearl market. These countries
have led a resurgence of natural pearl’s popularity, specifi-
cally with Basra pearls. These originate in the Arabian (Per-
sian) Gulf, and their namesake is the city of Basra in Iraq.
Historically, Basra was the center of natural pearl com-
merce, with the greatest selection of natural pearls and the
most informed dealers.
Vock mentioned that the most prized natural nacreous
pearls today are from either the Gulf region or Australia.
Currently, both regions produce what dealers describe as
“new material,” which are freshly harvested natural pearls.
Natural pearls form in the mantle of the pearl mollusk
shells, whereas cultured saltwater pearls grow inside the
gonad of the mollusks. Many natural pearls have a large
flat surface, similar to a button shape, because they grow
against the mantle. For that reason, round, oval, and drop-
shaped pearls are rarest and most valuable.
Figure 22. A graduated strand of “Edison” pearls in
copper and bronze metallic colors, ranging from 9 to Figure 23. A pair of earrings with unoiled emeralds
12 mm. Also shown are three loose pearls: 10 mm weighing about 8 ct total, natural pearls measuring
peach, 11 mm purple, and 11 mm copper. Photo by 9.5–10.0 mm, unheated Burmese rubies weighing just
Robert Weldon. under 5 ct total, and diamonds set in yellow gold.
Photo courtesy of ProVockative Gems Inc.

more to quality over quantity. The wholesale price of fine-

quality “Edison” pearls is still considerably below that of
similarly sized Tahitian and akoya pearls, but we expect
the price gap to narrow somewhat as the quality and pop-
ularity of these beautiful freshwater cultured pearls con-
tinues to increase.
Jennifer Stone-Sundberg and Si Athena Chen

Spotlight on natural nacreous pearls. A natural pearl forms

in the interior of a mollusk within a naturally formed pearl
sac without human intervention. The use of natural pearls
goes back thousands of years, and these biogenic gem ma-
terials have been treasured in jewelry and adornment
throughout human history. In general, pearls can be sepa-
rated into two varieties based on surface structure: nacreous
and non-nacreous. The nacreous pearls are more common
in the market. These are formed by a layered structure of
aragonite platelets together with organic substances as well
as water, and they normally display a pearly luster.

112 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023

Figure 24. Left: The mother shell (Mercenaria mercenaria) and an 18K gold ring featuring a 14.43 × 14.63 mm lilac
quahog pearl. Right: A rare collection of round quahog pearls ranging from 5.65 to 12.33 mm in diameter and in
colors from white and beige to deep purple and tan. Photos by Robert Weldon; courtesy of ECIJA.

When comparing pearls from the two regions, Vock ex- sizes ranging from 5.65 to 12.33 mm in diameter. Quahog
plained, there are some differences. Basra pearls usually pearls are known for their unique range of colors, including
have a warmer bodycolor, while Australian pearls tend to white, beige, pink, lavender, purple, brown, and black. These
be whiter. Also, Basra pearls are often rounder than their 21 quahog pearls showed a wide color range, from white and
Australian counterparts. In fact, the Gulf region has the beige to deep purple and tan (figure 24, right). According to
highest percentage of round natural nacreous pearls of any Jeremy, lilac and deep purple are the most desirable and
producing area. He credited this to the smaller shell of the command the highest price. The color of a quahog pearl is
Gulf region mollusks. determined by various factors, including the clam species
To see Alex Vock explain these common appearances and environmental and geochemical conditions.
with examples from his current collection, go to Another rarity on display was a large (16.44 ct, 14.50 ×
[Link]/gems-gemology/spring-2023-gemnews- 14.36 × 11.95 mm) deep purple, button-shaped non-nacre-
natural-nacreous-pearls. ous pearl showing an obvious flame structure at its surface
Lisa Kennedy and Artitaya Homkrajae (figure 25). This pearl formed from one of the many
Spondylus species, also known as “thorny” or “spiny” oys-
ters, which are harvested on the western coast of the Amer-
Exceptionally rare pearls from ECIJA. At the GJX show,
icas as Spondylus princeps and Spondylus calcifer/limbatus.
ECIJA (Santa Barbara, California) displayed a variety of nat-
ural, wild, nacreous, and non-nacreous pearls, including
conch, oyster, scallop, abalone, and quahog pearls. Only
0.2% of all pearls in the current market are natural (S. Figure 25. This non-nacreous pearl weighing 16.44 ct
Karampelas et al., “Raman spectroscopy of natural and cul- and measuring 14.50 × 14.36 × 11.95 mm is reportedly
tured pearls and pearl producing mollusc shells,” Journal from a Spondylus calcifer/limbatus mollusk from the
of Raman Spectroscopy, Vol. 51, No. 9, 2019, pp. 1813– Sea of Cortez in Baja California. Photo by Robert
1821), and the authors found the non-nacreous collection Weldon; courtesy of ECIJA.
especially notable.
ECIJA owner Aylene Norris and her husband, Jeremy
Norris (Oasis Pearl), have been sourcing for 34 years and spe-
cialize in natural pearls. For the first time in Tucson, ECIJA
displayed a collection of round quahog pearls (figure 24).
Quahog pearls are produced in the bivalve mollusk Merce-
naria mercenaria (figure 24, left), a type of saltwater clam
native to the Atlantic Ocean. These non-nacreous pearls
consist mainly of fibrous aragonite and organic matter. They
are often not perfectly spherical but form irregular shapes
such as button or baroque, making this set of round quahog
pearls extremely rare (figure 24, right). Moreover, they ex-
hibited an attractive and strong porcelaneous surface with

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023 113

Reportedly from calcifer/limbatus from the Sea of Cortez 10 years, whereas pastel colors continue to be a more afford-
in Baja California, the pearl’s large size, attractive purple able solution for designers and collectors alike.
color, and strong flame structure set its value. Previously, Si Athena Chen
GIA has reported that flame structure is usually caused by
subsurface to surface-reaching acicular inclusions (Fall 2016 Chunhui Zhou
Lab Notes, pp 303–304). Although several Spondylus pearls GIA, New York
have been reported in G&G (Summer 2016 Micro-World,
pp. 202–203; Fall 2016 Lab Notes, pp. 303–304), this is one New find of petroleum-included quartz from Madagascar.
of the largest and finest examples. At the Tucson Gem and Mineral Show (TGMS), a new find
ECIJA also featured other nacreous and non-nacreous of quartz with petroleum inclusions was offered for sale by
pearls for designers and buyers, including conch, scallop, Hidden Gem Gallery (Portland, Oregon). According to
and abalone pearls. Each of these species exhibits unique owner G. Moses Samora, the material is from the Andran-
shapes and colors. Conch pearls are typically oval in shape otokana Massif in the Alaotra-Mangoro region of eastern
and exhibit white, beige, brown, yellow, orange, pink, and Madagascar. Petroleum inclusions in quartz have been
red colors. They range in size from 1 mm to over 20 mm. seen for a number of years from many deposits, most no-
Scallop pearls from Nodipecten nodosus, also known as tably from Baluchistan, Pakistan (Spring 2004 Gem News
“Lion’s Paw” or “Mano de Leon,” form button, oval, and International, pp. 79–80). The petroleum in these quartz
baroque shapes that range from white and brown to rare crystals from Madagascar appeared brownish yellow and,
purple or maroon colors. These exhibit beautiful reflective as with many other petroleum inclusions in quartz, reacted
mosaic-like patterns on their surface. Abalone typically to long-wave UV light with yellow to blue fluorescence
produce conical or baroque-shaped pearls exhibiting a com- (figure 26).
bination of vibrant colors and a mirror-like luster. Sym- Microscopic examination revealed numerous complex
metrical abalone pearls, especially in large sizes, are very fluid inclusions containing obvious brownish yellow pe-
rare. troleum, a gas bubble presumed to be methane, dark solids
While GIA has developed seven value factors to evaluate of what was likely asphaltite, and a colorless immiscible
the quality of nacreous pearls (size, shape, color, luster, sur- liquid that was probably water (figure 27). Interestingly, in
face, nacre, and matching; see J.W.Y. Ho and S.C. Shih, some fluid inclusions the methane bubble was mobile pro-
“Pearl classification: The GIA 7 Pearl Value Factors,” Sum- vided there was enough available space within the petro-
mer 2021 G&G, pp. 135–137 and accompanying wall chart), leum-filled cavities. These components are consistent with
it is difficult to set the value for non-nacreous pearls due to petroleum fluid inclusions in quartz from other deposits.
their rarity and individual characteristics. In addition, de- When exposed to long-wave UV light, the petroleum com-
mand dictates the value and price of rare pearls. An increas- ponent strongly fluoresced light yellow (which appeared
ing demand for conch pearls has attracted buyers from all blue in the photomicrographs), as has been previously ob-
around the world. Jeremy indicated that conch pearls are es- served in similar material. The two crystals from this new
pecially popular in the European, U.S., and Asian markets. deposit displayed a scepter morphology, which seems un-
Deep pink to red pearls exhibiting a strong flame pattern are usual, though it remains unclear whether this is diagnostic
the most desirable and have doubled in value over the last for the locality or just unique to these specific examples.

Figure 26. These two quartz crystal scepters containing a multitude of complex petroleum fluid inclusions (left)
that fluoresce yellow to blue to long-wave UV light (right) are from a new deposit in Madagascar. The larger crys-
tal weighs 11.05 ct and measures 22.81 mm in length. Photos by Annie Haynes; courtesy of Hidden Gem Gallery.

114 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023

Figure 27. The new find of quartz from Madagascar contained complex fluid inclusions consisting primarily of pe-
troleum, methane, asphaltite, and water (left). When exposed to long-wave UV light, the petroleum appeared to
fluoresce a strong yellow. However, the digital camera used to capture the image registered the fluorescence as a
blue color (right). Photomicrographs by Nathan Renfro; field of view 2.81 mm.

Microscopic inclusions in quartz are enjoyable to ex- cut into cabochon form. However, transparent material has
amine, and petroleum inclusions are of particular interest always been found in this mine and makes up about 5–
to the collector due to their spectacular UV reaction. The 10% of the production.
introduction of this new quartz find at the Tucson show is At the AGTA show, Greenland Ruby displayed many
an exciting addition to the gem and mineral trade. examples of attractive transparent ruby and pink to orangy
Nathan Renfro and John I. Koivula pink sapphire, including several fancy-color sapphires in
GIA, Carlsbad sizes greater than one carat (figure 28; see also pp. 145–149
of this issue). These fancy colors are achieved by heat treat-
ing nearly colorless to slightly pink transparent material.
Bright orange sapphire from Greenland. The Aappaluttoq A transparent bright orange sapphire weighing 0.89 ct and
mine in Greenland is home to the oldest known ruby- measuring 6.41 × 5.02 mm (figure 29) immediately caught
bearing rocks on Earth. These host rocks are nearly three our eye. Martin Viala, product manager for Greenland
billion years old (A. Polat et al., “New age (ca. 2970 Ma), Ruby, shared with us a video of the original 1.07 g sapphire
mantle source composition and geodynamic constraints on rough (visit [Link]/gems-gemology/spring-2023-
the Archean Fiskenæsset anorthosite complex, SW Green- gemnews-bright-orange-sapphire-greenland). This unusual
land,” Chemical Geology, Vol. 277, No. 1-2, 2010, pp. 1– stone was mined in late 2019 and heated in May 2022.
20). This deposit was initially known for producing Viala noted that it was a textbook example of a chromium
translucent to opaque ruby and pink sapphire, much of it trapped-hole (h•–Cr3+) chromophore (E.V. Dubinsky et al.,

Figure 28. A 1.28 ct heated pink round brilliant sap- Figure 29. A 0.89 ct heated orange oval brilliant sap-
phire. Photo by Robert Weldon; courtesy of Green- phire. Photo by Robert Weldon; courtesy of Green-
land Ruby. land Ruby.

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023 115

 imens on quartz matrix, many of which were sold to muse-
ums in the U.S. and Europe. In 1968, the fragment made its
way from Spray’s widow, Eleanor, through a friend of the
family to geologist and mineral collector Charles Trantham
in California. In 1981, Glenn Vargas, a gem and mineral
dealer and university instructor, expertly cut the stone.
Rhodochrosite often presents challenges for cutting due to
its low hardness (3.5–4.0 on the Mohs scale) and perfect
rhombohedral cleavage.
Rhodochrosite mining at Sweet Home expanded signif-
icantly in 1991—long after silver production ended in the
1960s—when Eleanor Spray’s nephew, F. Leonard Beach,
Figure 30. A 7.94 ct rectangular step-cut rhodochrosite leased the property to a group of investors led by Collec-
from the Sweet Home mine in Colorado’s Alma Min- tor’s Edge Minerals. The group found several new pockets
ing District. Photo by Robert Weldon; courtesy of and in 1992 uncovered the Alma King, a 14.2 × 16.5 cm
Barker & Co. deep cherry red rhombohedron, the world’s largest known
fine rhodochrosite crystal.
The Sweet Home mine closed in 2004, but the group
“A quantitative description of the causes of color in corun- went on to develop the associated Detroit City mine,
dum,” Spring 2020 G&G, pp. 2–28). which has produced some fine specimens since 2019.
Jennifer Stone-Sundberg and Si Athena Chen Erin Hogarth

Sweet Home mine rhodochrosite from 1888. At AGTA, New tourmaline pockets in San Diego County’s Pala Dis-
Ann Barker and Tori Lopez of Barker & Co. (Scottsdale, Ari- trict. At the AGTA show, Bill and Carl Larson (Pala Inter-
zona) exhibited a 7.94 ct rectangular step-cut rhodochrosite national, Fallbrook, California) shared details about new
(figure 30) with a provenance dating back more than a hun- pockets uncovered at the Tourmaline King mine. The first
dred years. The stone is from the Sweet Home mine in Col- was discovered in January 2022—the first major find at the
orado’s Alma Mining District, at an elevation of about 3,300 mine in more than a hundred years, and the first in South-
m in the Rocky Mountains. ern California since the “Big Kahuna” pocket at the Ocean-
Sweet Home was established as a silver mine in 1872. view mine in 2010. “We’ve hit off and on three or four
Miners initially discarded most of the rhodochrosite found pockets, some of which are very good,” Bill said.
in association with the silver ore because it interfered with The Larsons showed us several stones, including a
the amalgamation process used then in silver mining. But 16.96 ct emerald-cut green tourmaline (figure 31) from
the large, highly saturated, and nearly perfectly formed rough found near the first pocket in 2022. Bill said the
crystals later discovered there are some of the world’s finest green tourmaline is relatively uncommon for the Pala Dis-
rhodochrosite specimens. trict. “The green has actually been transparent enough to
Rhodochrosite was first described in the mineralogical
literature in 1813 based on specimens from what is now
Romania. In 1887, George F. Kunz wrote of gem-quality Figure 31. A 16.96 ct emerald-cut green tourmaline
rhodochrosite in Colorado, noting it was the first source from rough found near a new pocket at the Tourma-
with such large and transparent crystals. Most of the line King mine. Photo by Robert Weldon; courtesy of
rhodochrosite available worldwide was light pink, opaque, Pala International.
and often banded—typically used for beads, cabochons, and
carvings. This source introduced pink to red gem-quality
single-crystal rhodochrosite, which is rare. The gemstone
became more widely known in the late 1930s, after a Ger-
man gemologist found large volumes in Argentina’s Cata-
marca Province high in the Andes. (In Argentina, the use
of rhodochrosite dates back to the ancient Incas, who are
said to have believed it was the solidified blood of their
fallen rulers.)
The step-cut stone’s traceable history began in 1888,
when a shaft collapse resulted in a 32.61 ct cleavage frag-
ment. Miner Edward Abbott kept it and in 1925 gave it to
Denver businessman Edwin Spray, who later owned the
mine. The same year, Sweet Home miners found large spec-

116 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023

 them inadequate for cutting and cabbing. But in January of
2023, they discovered a pocket of beautifully colored red
rubellite, too included for faceting but suitable for cabbing
and carving. “We’re hoping that we’ll hit some more
gemmy transparent material,” he said.
“It’s really exciting to have an active mine in Southern
California because of its rich history,” Carl said. “At the
turn of the century we were hitting a lot, and then it was
vacant for so long. So to have some energy coming in with
new material is exciting.”
Erin Hogarth

Vibrant green grossular garnet “Transvaal jade.” At the

Pueblo show, Tom Schneider of TMS Gems (San Diego,
California) offered “Transvaal jade” rough displaying an
exceptionally vivid green color. Also known as “African
jade” or “South African jade,” this is not a true jade
(jadeite or nephrite). The stone is instead a grossular garnet
whose trade name is derived from its massive habit and
Figure 32. Rubellite tourmaline rough (440 g) from one green color, and the type locality. Specimens of this mate-
of the Tourmaline King mine’s new pockets. Photo by rial are typically more translucent and lighter in color (J.
Robert Weldon; courtesy of Pala International. Frankel, “Uvarovite garnet and South African jade (hy-
drogrossular) from the Bushveld Complex, Transvaal,”
American Mineralogist, Vol. 44, No. 5-6, 1959, pp. 565–
facet,” Carl added. “That’s been the best material so far.” 591). Nested in a fine-grained chromite matrix, the opaque
The new pockets have also produced about 3 kg of fine- green portion of the stone in figure 33 had a refractive
color rubellite rough (figure 32). index of 1.731 and was inert to both long-wave and short-
The Tourmaline King mine’s main tunnel produced wave ultraviolet radiation.
eight tons during San Diego County’s massive output of The sample was identified as grossular garnet using
tourmaline in the early nineteenth century, which also in- Raman spectroscopy and further classified as a grossular-
volved the Pala District’s Tourmaline Queen, Pala Chief, andradite garnet using X-ray fluorescence chemical analysis
and Stewart mines, as well as Mesa Grande’s Himalaya and the naming convention proposed in 1995 (M.L. John-
mine. (There was a resurgence of mining at Pala in the son et al., “Gem-quality grossular-andradite: A new garnet
1970s.) from Mali,” Fall 1995 G&G, pp. 152–166). The chemical
Around 1920, mine owner R.M. Wilke and his crew in-
stalled a new tunnel below the main one. “They hit this
very dangerous zone right before they hit the pegmatite,” Figure 33. This 110.4 g partially polished rough
Bill said. Here they encountered a shear zone that caused a grossular garnet “Transvaal jade” and three polished
roof collapse. Wilke abandoned mining efforts around 1922. cabochons (41.95–92.87 ct) display unusually vibrant
Pala International’s mining partner, San Diego Mining green colors between seams of dark chromite matrix.
Company, reclaimed the tunnel several years ago and began Photo by Diego Sanchez; courtesy of TMS Gems.
extending it in 2019. “They got to the dangerous area, and
then they backed up about 10 meters and went around and
hit the pegmatite,” Bill said. “Now they’re in from the front
to the back, where we’re hitting tourmaline.”
Carl recalled his initial look at the first new pocket.
“Kiel Snyder stopped me and said, ‘Everyone else gave up
here,’” he said. “He just believed in it and kept going. Then
they hit a small pocket—nothing of note, but something
that said, ‘There are stones here.’ I think they called that
the ‘Never Give Up’ pocket. Within 10, 15 feet, they hit
that first pocket.”
The crystals were large and well formed, and the min-
ers uncovered a very large terminated quartz crystal, a col-
lector specimen. Carl said the crystals in this pocket lack
the rubellite color and are not typical of the mine, making

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023 117

 Figure 35. “Salt and pepper” diamonds. Left to right:
a 4.30 ct hexagonal step cut, a 5.56 ct octagonal step
cut, a 3.58 natural-color orange cushion cut, and a
3.39 ct pear cut. Photo by Robert Weldon; courtesy of
Misfit Diamonds.

Figure 34. Alex Skachkov of Misfit Diamonds. Photo

by Jennifer Stone-Sundberg. designers and consumers. In his words, younger customers
are not necessarily interested in absolutely flawless gems.
Rather, many are looking for gems that charm in their dis-
data show 80.83 mol.% grossular (Ca3Al2Si3O12), 10.50 tinctive and rare “perfect imperfection.” He noted that
mol.% uvarovite (Ca3Cr2Si3O12), 7.11 mol.% andradite these diamonds are also a more budget-friendly option. In
(Ca3Fe2Si3O12), and trace amounts of other garnet. Garnets addition to the “salt and pepper” diamonds, we got a close
in the grossular to andradite range can show 4.64–20.91 look at diamonds and sapphires cut into nontraditional
wt.% Fe2O3 (Johnson et al., 1995) and typically have low shapes, including slices (figure 36).
chromium values (≤0.23 wt.% Cr2O3) (C.M. Stockton and Skachkov mentioned that many of their customers
D.V. Manson, “A proposed new classification for gem-qual- now look for full disclosure of the entire mine-to-market
ity garnets,” Winter 1985 G&G, pp. 205–218). Increased custody chain. Misfit Diamonds provides this information
iron content will impart a yellow, orange, or brown color whenever possible, though at times their diamonds arrive
component, while the most intensely green stones owe in mixed-origin parcels, with stones from a variety of global
their color to chromium and possibly vanadium (Johnson locations. Moving forward, they aim to meet the challenge
et al., 1995). The low iron content (2.49 wt.% Fe2O3) and of reporting full chain of custody, as they predict increasing
elevated chromium content (3.50 wt.% Cr2O3) of this sam- demand for this feature.
ple likely resulted in its remarkable coloration.
Jennifer Stone-Sundberg and Si Athena Chen
The chromium-rich nature of this particular specimen
produced a beautiful vibrant green with a high-polish lus-
ter, making it an outstanding example of this material.
Kendra Carty and Amy Cooper Figure 36. Nontraditional diamond cuts and slices in
GIA, Carlsbad a range of sizes. Photo by Jennifer Stone-Sundberg;
courtesy of Misfit Diamonds.
DIAMONDS

Misfit Diamonds: Beauty in the imperfect. At the AGTA

show, we spoke with Alex Skachkov of Misfit Diamonds
(Vancouver, Canada) about some of their unusual diamonds
(figure 34). When asked what was popular, he showed us
their aptly named “salt and pepper” diamonds (see also fig-
ure 5 of the Tucson 2023 overview), which contained eye-
visible black and white inclusions. These had a very
distinct look and were cut into various nonstandard shapes
(figure 35).
Skachkov described the appeal of these unique and
beautiful “imperfect” diamonds, particularly to younger

118 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023

 Figure 37. The “Seven
Sisters” ring in 18K
gold with 17 bezel-set
diamonds (0.55 carats
total) and 22 sapphires
ranging in size from 2.5
to 4.0 mm in blue, or-
ange, mandarin, pink,
and green (3.15 carats
total). Photo by Robert
Weldon; courtesy of
Sean Hill Designs.

JEWELRY DESIGN A stunning pair of 18K white gold earrings featuring

both diamond slices and faceted diamonds (figure 38) is a
Sean Hill: Exquisite jewelry designs. At the GJX show, case study in filling a plain geometric shape to bring it to
Sean Hill of Tucson presented some striking jewelry de- life. As Dr. Hill said in describing the piece, “A rectangle
signs with masterful control of shape and line. A gold ring is a basic shape, a self-imposed constraint. It is how you
set with diamond, blue sapphire, and fancy-color sapphire choose to fill it that makes it interesting.” Adding to the
(figure 37) drew us into his booth. Dr. Hill, a self-taught design quality, the rectangles are not actually flat but gen-
jeweler, has been active designing and handcrafting jewelry tly curved so that the earrings fit seamlessly on the ear.
since 1998, after a previous career as a university professor A pair of 18K gold earrings with brilliant-cut diamonds
in literature and philosophy. His academic and philosoph- and sapphires (figure 39) illustrates a modern and consid-
ical nature is immediately evident when discussing design, ered take on the classic chandelier earring, with squares
and his meticulous attention to space and volume becomes and rectangles outlining floating round gems on delicate
more apparent the longer one studies his pieces. The strings of gold. Dr. Hill explained some of his jewelry phi-
“Seven Sisters” ring in figure 37 raises each stone to a losophy: “You can always take a stone for its beauty, color,
unique height above the flat base, filling the rectangle with and cut, set it into a finding, and it is all about the stone.
an appealing placement of 17 colorless diamonds inter- But how do you make it into art?”
spersed among 22 sapphires. Jennifer Stone-Sundberg and Si Athena Chen

Figure 38. “Rose Cut” 18K white gold and diamond Figure 39. “Event Horizon” earrings in 18K gold, dia-
earrings. The faceted diamonds are 1.5–2.0 mm in di- mond (0.07 carats total), and orange, mandarin, pink,
ameter. Photo by Robert Weldon; courtesy of Sean green, and blue sapphire (5.2 carats total). Photo by
Hill Designs. Robert Weldon; courtesy of Sean Hill Designs.

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023 119

Figure 40. These handcrafted jewelry pens are created with silver, embellished with 24K gold, and inlaid with dia-
mond, sapphire, and crushed emerald. Left: The peacock design is a miniature painting coated in clear enamel.
Right: A pen with fine metal engraving and dimming just below the 24K gold nib. Photo by Robert Weldon; cour-
tesy of Zeki Karaca.

Zeki Karaca Jewelry: Luxury pens. Some of the most orig- designs and styles, he began making luxury pens and prod-
inal work we saw at the GJX show included bejeweled pens ucts three years ago. He first ventured into these new prod-
made by Turkish designer Zeki Karaca (figure 40). These ucts by crafting silver gemstone handles for Damascus
intricate pens feature exquisite details and fine craftsman- steel knives. After finding success there, he expanded to
ship in metalworking, enameling, engraving, inlay, and meet the demand for luxury pens, letter openers, and mag-
stone setting. In addition to pens, Karaca also creates other nifying glasses.
desk items such as letter openers and magnifying glasses Each pen takes about six months to design and craft.
(figure 41). All production steps are done by hand, making each item
Born in Istanbul, Karaca has 41 years of jewelry-making one of a kind. After designing an item, Karaca creates a sil-
experience, originally designing jewelry to sell in the Grand ver body using fold forming, with the folds created using
Bazaar. To apply his experience and creativity to different emery sheets. Engraving is executed by a master engraver
using different nibbed steel pens, and dimming is used to
emphasize the engraving. Karaka incorporates diamond,
Figure 41. Silver magnifying glass with 24K gold em- sapphire, emerald, and ruby into the designs and uses 24K
bellishments, inlay, carvings, and diamonds. Photo gold for design highlights. Miniature paintings, such as the
by Robert Weldon; courtesy of Zeki Karaca. peacock on the pen cap in figure 40 (left), are created by
hand with a small brush and then coated with a transpar-
ent enamel for protection. For micro-mosaics, gemstones
such as emerald are crushed and added to the pen body, as
seen in both pens in figure 40. The pen parts are welded to-
gether with a laser, and then the 24K gold nibs are added.
With high demand for his products from collectors,
Karaca continues to explore additional high-end objects as
well as custom designs for some clients.
Jennifer Stone-Sundberg and Mimi Travis
GIA, Carlsbad

Brenda Smith Jewelry designs. Brenda Smith Jewelry has

earned four AGTA Spectrum Awards to date for design ex-
cellence. At this year’s AGTA designer showroom, Smith
presented three masterpieces to the authors: two award-
winning rings featuring pearl and blackened gold lace and
an opulent pair of earrings featuring millefiori faces.
Smith’s most prized creation is her Lace Ring design.
She shared her white Lace Ring featuring a large cultured
pearl accented with gold lace, sapphires, and diamonds (fig-
ure 42A), followed by a black Tahitian pearl Lace Ring with

120 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023

 A B C

Figure 42. A: This Lace Ring is crafted in 18K rose gold with blackened gold lace, featuring a cultured freshwater
pearl measuring 16.5 mm in diameter, enclosed by natural-color pink and lavender sapphire round brilliants weigh-
ing 2.86 carats total. The shank of the ring is embellished with 0.06 carats of diamonds. B: This Tahitian pearl Lace
Ring is created in 18K white gold with sandblasted and blackened gold lace. The cultured Tahitian pearl measures
16.0 mm and is surrounded by round brilliant diamonds, 1.35 carats total. The rubies in the shank weigh a total of
0.08 carats. C: A handmade doily crocheted by Smith’s grandmother. Photos courtesy of Brenda Smith.

blackened gold lace, rubies, and diamonds (figure 42B). In- Smith’s one-of-a-kind “Gibson Girl” earrings (figure 43)
spired by her grandmother’s handmade crocheted doily fea- are inspired by the Austrian symbolist painter Gustav
turing a pineapple motif (figure 42C), the rings were created Klimt. Klimt’s paintings combined realistic faces with geo-
as a tribute to her memory. In 2021, the white freshwater metric clothing, and Smith mirrored this style to create
pearl ring won Best Use of Pearls in the AGTA Spectrum this high-end piece of jewelry. The technique used to create
competition and the Tahitian pearl ring won the InDesign the girl faces is millefiori, which means “thousand flow-
Award in the category of Pearl Jewelry Over $5,000. ers” in Italian and produces slices from canes with concen-
tric and colorful patterns. Millefiori techniques can be used
to make jewelry, decorative objects, candy, and art pieces.
Figure 43. These “Gibson Girl” millefiori earrings in Creating these objects requires a skilled artisan who pos-
18K white gold contain Paraíba tourmalines (1.11 sesses a deep understanding of material properties, consid-
carats total), pink tourmalines (1.54 carats total), and erable experience, and patience. American artist Barbara
0.56 carats of aquamarine. Photo courtesy of Brenda McGuire handcrafted the complex face canes for the ear-
Smith. rings by arranging and embellishing polymer clays to create
a lifelike image. Though millefiori objects are commonly
used in decorative pieces, Smith’s design incorporates a
harmonious blend of colored gemstones expertly paired
and crafted to create a lively figure with a unique person-
ality and color scheme, making it an extravagant and ele-
gant piece of jewelry.
Si Athena Chen and Jennifer Stone-Sundberg

YNY Jewels: Designer Surbhi Pandya. A dazzling pair of

carved tanzanite earrings (figure 44) caught our attention at
the GJX show. Designer Surbhi Pandya of YNY Jewels (New
York), who goes by the single name Surbhi, described the
significance of the carving featured on each of the large un-
heated tanzanite crystals. Depicted in violet and green is the
lotus blossom, a flower with ancient cultural significance in
India, rising with its deep roots from the mud without
stains, symbolizing purity and strength. As the blossoms
close up at night and reopen the next day, they also symbol-
ize rebirth. Each carving is immediately surrounded by a
ring of blue Kashmir sapphires and emeralds, followed by
colorless diamonds and bright green tsavorite garnets. Each

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023 121

Figure 44. Surbhi’s “Orient” earrings with tanzanite
carvings surrounded by blue sapphires, emeralds,
white diamonds, and tsavorite garnets with an upper
flower-shaped dangle composed of faceted tanzanites,
emeralds, and pink sapphires, all set in 18K white gold.
The total diamond weight for the pair is 1.60 carats,
and the total colored stone weight is 62.23 carats,
48.32 carats of which are from the two tanzanite carv-
ings. Photo by Robert Weldon; courtesy of YNY Jewels
Figure 46. Designer Surbhi. Photo by Jennifer Stone-
Inc. and Karats Inc.
Sundberg.

upper dangle contains faceted slices of tanzanite and emer-

ald with a central hot pink sapphire. She also shared a pen- containing a lotus carving (figure 45). She explained that the
dant with a paisley-shaped unheated tanzanite similarly paisley shape represents continuity and life.
Surbhi (figure 46) has been making jewelry for 20 years
and was trained by her father, Yogendra Sethi, a renowned
Figure 45. An 18K white gold pendant containing a artist in India. She described designing jewelry as a journey
paisley-shaped unheated tanzanite with a lotus blos- filled with energy. She does not make jewelry to please her-
som carving, surrounded by pavé-set black diamonds self, but out of a drive to create things that speak to others.
and pink and purple princess-cut Sri Lankan sap- Surbhi shared stories of customers’ deeply personal con-
phires. The total weight of the diamonds is 1.00 nections with her jewelry, a profoundly rewarding experi-
carats, and the colored stones total 23.40 carats. ence for her as an artist.
Photo by Robert Weldon; courtesy of YNY Jewels Inc.
Jennifer Stone-Sundberg and Si Athena Chen

RESPONSIBLE PRACTICES

Ethical supply chain practices in Africa with Virtu Gem.

This year marked Tucson’s fourth Ethical Gem Fair, a mar-
ket cooperative of ten responsibly sourced gemstone sup-
pliers (see Spring 2020 GNI, pp. 177–179). The fair was held
January 28–31 at the Scottish Rite Cathedral. Virtu Gem
sells gemstones directly from artisanal mining communi-
ties in Kenya, Malawi, and Zambia (figure 47) and debuted
at the fair in 2022. We spoke with cofounder Susan
Wheeler and Percy Maleta, Virtu Gem’s country exporter
and ambassador in Malawi, about its beginnings, programs,
and impacts.
Wheeler said more new customers come to the Ethical
Gem Fair each year. “All the large companies come by,
too,” she said. “There were a lot this year that everybody
was surprised by. I think that the awareness of what arti-
sanal mining is and what it can be is spreading.”

122 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023

 safety, labor, and environmental conditions at the mines,
all with a focus on women. Their gemstone prices include
a 10% premium that goes toward fulfilling the mining
communities’ various needs. During the height of the
COVID-19 pandemic, the premium was used for food
drives in all three countries and personal protective equip-
ment for Kenyan women miners; currently it goes to the
purchase of safety equipment (figure 48).
Wheeler and cofounders Jessica Hudson and Monica
Gichuhi (figure 49) began selling gemstones from Zambia
online in 2020, after the pandemic forced the cancellation
of a conference they had planned there that would have
provided a market for traders. “There was a call to action
from the Organisation for Economic and Co-operative De-
velopment (OECD) to support all the artisanal miners be-
cause no one was coming to the country to buy their
stones,” Wheeler said. The OECD’s Call to Action for Re-
sponsible Mineral Supply Chains cited COVID-19’s dispro-
portionate impact on artisanal and small-scale miners, who
make up more than 80% of the global mining workforce
and already face inequities in the supply chain. Wheeler
and Hudson are both jewelry designers; Gichuhi is a found-
Figure 47. Some of Virtu Gem’s larger gemstones from
ing member of the Association for Women in Extractives
Kenya, Malawi, and Zambia at the Ethical Gem Fair.
in Kenya (AWEIK). Her connections with mining associa-
Clockwise from top: 45.00 ct citrine, 32.33 ct rutilated
tions in Zambia were a starting point.
quartz, 9.80 ct aquamarine, 21.65 ct citrine, 9.99 ct
Beginning in 2021, Virtu Gem received two grants
aquamarine, 12.32 ct aquamarine, 17.20 ct aquama-
through RJT from the World Bank’s Extractives Global Pro-
rine, and 20.15 ct citrine (center). Photo by Robert
grammatic Support fund. The fund’s purpose is to promote
Weldon; courtesy of Virtu Gem.
sustainable and inclusive mining in developing countries
and thereby reduce poverty. The grants allowed them to ex-
Virtu Gem gives artisanal miners, cutters, and traders pand into Kenya and Malawi, hold virtual cutting work-
formal access to international markets and began as a proj- shops, and implement CRAFT Code in eight mines.
ect of Wheeler’s nonprofit Responsible Jewelry Transfor- (“CRAFT” stands for Code of Risk mitigation for Artisanal
mative (RJT). Virtu Gem’s programs also offer training in and small-scale miners engaging in Formal Trade.) Virtu
cutting and basic gemology and help miners improve Gem also hired experienced international consultants.

Figure 48. Artisanal

emerald miners in
Zambia celebrate a do-
nation by Virtu Gem of
safety equipment and a
jackhammer. Photo
courtesy of Virtu Gem.

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023 123

 Figure 49. Left: Virtu
Gem cofounders Jessica
Hudson, Monica
Gichuhi, and Susan
Wheeler on a recent visit
to Malawi. Right: Percy
Maleta, the organiza-
tion’s country exporter
and ambassador in
Malawi, examines a
rough gemstone. Photos
courtesy of Virtu Gem.

“There’s been a huge amount of progress over the past land at all, especially in Zambia, Kenya, and Malawi,”
two years,” Wheeler said. The funding helped Virtu Gem Maleta said. In Zambia, two of the mines—one emerald and
develop a strong foundation, she said, and grow enough one amethyst—were woman-owned, and in Kenya they
to be able to pass any chain of custody standard for large- worked with a young woman whose family owned a mine.
capacity buyers. One of Virtu Gem’s benefits is helping artisanal miners
While Virtu Gem has had success in all three countries, formalize their mining process. In Malawi, country coordi-
Wheeler said Kenya has seen the most because some in the nator Chiko Manda wanted to focus on rhodolite. Maleta
trade already had Ultra Tec cutting machines and were sell- said women account for roughly 80% of rhodolite produc-
ing to the international market. “After the first few pur- tion there (figure 51). Rhodolite is alluvial and best mined
chases we did, word spread,” she said. “People were
showing up with 50 gems at a time to process.” Miners can
also bring gems to the country coordinators for cutting. Figure 50. Rough tsavorite mined from a site in
Virtu Gem’s gemstones are cut in each country and Kenya. Photo courtesy of Virtu Gem.
tracked with Provenance Proof Blockchain. They include
amethyst, aquamarine, citrine, moonstone, rhodolite and
other garnets, in all three countries; emeralds in Zambia;
tsavorite in Kenya (figure 50); and color-change garnet in
Malawi. The three countries also produce tourmaline in
various colors, including pinkish orange (“sunset”) in
Malawi and Zambia and golden in Kenya and Malawi. Pro-
duction of morganite from newly discovered deposits in
northern Malawi recently began.
Virtu Gem has helped miners implement CRAFT Code
at four mines in Zambia, two in Malawi, and two in Kenya.
CRAFT Code is designed to improve safety, labor, and pay
conditions through education at the mines. Its components
include mine site certification, safety standards and equip-
ment, and environmental planning, including carbon data
tracking and reduction.
CRAFT Code has an emphasis on protecting women,
and Wheeler said they focused on women for the associated
training. Focusing on woman-owned mines was more diffi-
cult, however. “It’s a big challenge for women to own any

124 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023

 Figure 51. Artisanal
rhodolite miners in
Malawi. Photo courtesy
of Virtu Gem.

during the rainy season in Malawi, when the women are been easy,” he said. “But we have people who understand
busy tending their gardens. Virtu Gem worked with a the process, who know that the stone they are leaving with
young couple, Ben and Tamara, who did not have a mining Virtu will be sold, but also understand that if it’s not sold,
certificate. it will be returned. That is trust.”
“The chief of the Chewa people didn’t want the women “I was surprised by the lack of trust at the beginning,”
mining,” Wheeler recalled. “Ben had to go and prove him- Wheeler said. “We have to work hard to earn the trust from
self to the chief, that he wasn’t going to take advantage, everybody in the gemstone community. We couldn’t have
and that the women could still do their agriculture.” Ben done this without our partner, Monica Gichuhi, working
joined Virtu Gem’s calls with miners in Kenya and Zambia within the communities, and our country coordinators.”
who shared advice. “He ended up getting permission to She mentioned Caroline Muchira in Kenya, who is also a
mine on the land and the blessing of the chief,” she said. cutter and a GIA Graduate Gemologist; Manda in Malawi;
“He sent me on WhatsApp a picture that he had gotten his and Pauline Mundia in Zambia, a respected woman leader
certificate. He was so happy.” Maleta said that Ben recently whom everyone calls “Mama Pauline.”
received another land permit and is applying for a second “Living up to expectations is hard every single day,”
mining certificate. “That was a way we could still work Wheeler said. “Tucson is hard because we don’t own any-
with the women,” Wheeler said. thing on display. Those are people sending us their stones
Wheeler mentioned the danger of holding a large that they’ve invested in. Before, they would have just sold
amount of gemstones in an informal market, which Ben them and not invested in separating parcels and cutting.
and Tamara encountered with a quantity of rhodolite. “You That’s a lot of pressure. We’ve earned some of that trust,
have the paradox of the women’s safety,” she said. “Every- and we really have worked hard to earn that.”
body hears about it, and you get all these other dealers and “It’s a dream come true for many to have their stones
traders coming in. How are they going to get a fair price? sold at this level,” Maleta said. “For us, for the country, for
Yet they can’t hold on to it.” She said Ben called his friends the people who are trading, the future is bright.”
and pleaded with them to purchase some of the rhodolite. Erin Hogarth
She would like to see a formal system to address this type
of scenario.
Maleta said the past four or five years have seen an in- SYNTHETICS AND SIMULANTS
flux of Kenyans and Zambians to Malawi to buy gems,
some of whom export them back home and misrepresent Luminescent synthetic garnet-like crystals. At the 22nd
the source country. By exporting directly from Malawi, Street show, the author spoke with Tim Challener of Tur-
Virtu Gem has improved supply chain transparency. tle’s Hoard (Raleigh, North Carolina), who was selling some
Maleta said Virtu Gem pays more than the informal unique rough and faceted luminescent synthetic garnet-like
market, but convincing people to participate can be a chal- crystals, which he referred to as “lumogarnets.” After 12
lenge. “To have someone leave a stone with you and years in the gem business, Turtle’s Hoard began selling only
wait—for maybe two months, three months—it hasn’t laboratory-made materials in March 2022. Challener noted

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023 125

 Challener explained, “They are lab-created, but each one
has its own history beyond its use as a gem. And while there
have been conflicting messages about whether some lab-
grown stones, particularly lab-grown diamonds, are truly
eco-friendly, the materials we work with are an inevitable
byproduct of necessary crystal growth. Our modern lives
rely on these materials being grown—we can’t have cell
phones without lithium niobate or modern PET scanners
without scintillators—so there isn’t really any additional
waste associated with them.”
Turtle’s Hoard offered a variety of “lumogarnets,” includ-
ing rough and faceted luminescent yellow and yellowish
Figure 52. Yellow to greenish yellow faceted cerium- green cerium-doped GAGG (gadolinium aluminum gallium
doped YAG “lumogarnets” ranging from 4 to 12 ct garnet, Gd3Al2Ga3O12), cerium-doped LuAG, and cerium-
apiece. Photo by Lisa Kennedy; courtesy of Turtle’s doped YAG (see again figure 52). In addition, they displayed
Hoard. rough grayish purple neodymium-doped YAG, rough and cut
greenish blue ytterbium-doped YAG (figure 54), rough pink-
ish orange erbium-doped YAG, and rough and faceted green
that their best-selling gemstone is Ce:LuAG, lutetium alu- chromium-doped YAG, a green stone with strong red fluo-
minum garnet (Lu3Al5O12) doped with cerium. This material
was introduced to the optics industry in the late 1990s and
acts as a scintillator, a fluorescent crystal that absorbs high-
Figure 53. The top of a YAG boule doped with cerium
energy radiation such as X-rays or gamma rays and reemits
and neodymium and produced by the Czochralski
them as lower-energy visible light that is much easier to de-
method, measuring approximately 36 mm in dia-
tect. LuAG is closely related to YAG electronically and both
meter and 45 mm tall. Parts of the boule that cannot
have the same garnet structure, yielding a gem with simi-
be used for scientific purposes can be sold to the gem
larly excellent optical and gem properties (see figure 52).
and jewelry trade for faceting. Photo by Lisa
Yttrium aluminum garnet, known as YAG (Y3Al5O12),
Kennedy; courtesy of Turtle’s Hoard.
a lab-grown crystal that crystallizes in the cubic garnet
structure, was one of the most popular diamond simulants
from the 1960s to the mid-1970s due to its relatively good
hardness (about 8.25 on the Mohs scale) and brilliance.
Cartier’s replica of the 69.42 ct Taylor-Burton diamond fa-
mously used YAG. Gadolinium gallium garnet, known as
GGG (Gd3Ga5O12), was the next garnet-like material to be
used as a diamond simulant. GGG is slightly more brilliant
and dispersive than YAG but scratches fairly easily due to
its lower hardness (about 6.5 to 7.5 on the Mohs scale).
Most of these garnet-like crystals are doped with vari-
ous rare earth elements such as neodymium, erbium,
cerium, and ytterbium. These dopants not only increase
the materials’ value in the optics industry but also yield
the variously colored synthetic garnet-like crystals seen in
the gem and jewelry industry today.
These synthetic garnet-like crystals are most com-
monly produced by the Czochralski method, also known
as the pulling method. In this melt process, the boule (a
cylindrical synthetic crystal produced by a melt process)
grows from a thin cylindrical seed crystal that is dipped
into a melt of the desired garnet composition. The seed is
rotated and then lifted very slowly at a controlled rate, al-
lowing the melt to crystallize onto it. The material pro-
duced is extremely pure, with few inclusions or none at all.
However, the material is not actually made for the gem and
jewelry industry; rather, those in the industry can only pur-
chase some of the less-perfect material and off-cuts of the
boules to facet and sell (figure 53).

126 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023

 The exhibit highlighted the “magic ingredient”—
chromium—that gives ruby not only its color but also its
fluorescence. Mogok rubies are renowned for their high
concentration of chromium and low concentration of iron.
The latter element, if present in a high enough concentra-
tion, quenches all or part of the red fluorescence produced
by the chromium. This fluorescence is an important factor
in the high value of Mogok rubies.
TGMS began in 1955 and is one of the longest-running
gem and mineral shows in the world. GIA Museum curator
Terri Ottaway said she was thrilled at the opportunity to
showcase such a gorgeous necklace and include a bit of
chemistry and geology for the gem and mineral enthusiasts.
Figure 54. Greenish blue rough and cut ytterbium- Erin Hogarth
doped YAG. The cabochon in front is 14 × 10 mm.
Photo by Lisa Kennedy; courtesy of Turtle’s Hoard. GIA Library wins exhibit award at TGMS. GIA’s Richard
T. Liddicoat Gemological Library and Information Center
was awarded the Betty Clayton Gibson Memorial Trophy
rescence showing simultaneous flashes of green and red for Best Museum Exhibit for 2023 at the Tucson Gem and
based on lighting, dubbed “Christmas garnet.” Although Mineral Show. The winning exhibit illustrated the impor-
synthetic and possessing a garnet-like structure, YAG, GGG, tance of mercury-free gold mining (figure 56). Historically,
LuAG, and GAGG are not synthetic garnet, since they lack mercury has been used in artisanal and small-scale gold
the chemistry of any garnets known in nature. mining because of its ability to bond with gold to form an
Lisa Kennedy amalgam, but its high toxicity and negative effects on hu-
mans and wildlife have led to increased emphasis on mer-
cury-free options for mining gold.
EXHIBITS The exhibit featured a traditional wooden bowl used
for gold panning, along with many of the library’s books on
GIA Museum’s “Paint the Town Ruby Red” exhibit at ecological jewelry and responsible and sustainable mining
TGMS. At the Tucson Gem and Mineral Show, the GIA and jewelry practices.
Museum exhibited a Burmese ruby and diamond necklace “This exhibit illustrates how mercury-free gold mining
and earrings against a backdrop of the necklace fluorescing can have positive long-term implications for both gold
in ultraviolet light (figure 55). The necklace contains 39 un- miners and the environment,” said library director Robert
treated rubies from Mogok totaling 83.73 carats and 302 Weldon. He noted that among all the deserving exhibits
diamonds totaling 42.62 carats. The largest ruby is 5.00 ct. displayed at TGMS, the GIA Library’s stood out due to the

Figure 55. The GIA Mu-

seum’s “Paint the Town
Ruby Red” exhibit at
TGMS featured a
Burmese ruby and dia-
mond necklace and ear-
rings, courtesy of Mona
Lee Nesseth (Custom
Estate Jewels) and a pri-
vate collector. Photo by
Terri Ottaway.

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023 127

 Figure 56. The GIA Li-
brary’s award-winning
exhibit on mercury-free
mining at the Tucson
Gem and Mineral Show.
Photo by Chris Rogers.

importance of responsible mining and sustainability in the during the AGTA GemFair in Tucson. Guergova, the first
industry. student from the London campus to win the award, designed
In 2021, GIA provided a grant to Mercury Free Mining a stunning bracelet featuring swans to represent together-
(MFM) and the Alliance for Responsible Mining (ARM). ness and loyalty (figure 57).
New methods for concentrating gold continue to be ex- Created in partnership with the Gianmaria Buccellati
plored and tested with the goal of significantly reducing, Foundation in 2018, the award recognizes outstanding talent
or even eliminating, the use of mercury in gold mining. in design among GIA students worldwide. Larry French,
Erica Zaidman chief officer for North America strategies at the foundation,
GIA, Carlsbad said, “On behalf of the Gianmaria Buccellati Foundation,
we want to congratulate Ms. Sara Guergova, this year’s win-
ner, plus all the other finalists whose work so enriched this
ANNOUNCEMENTS year’s competition. We also want to recognize the talented
GIA design instructors who helped guide the students on
Sixth annual Gianmaria Buccellati Foundation Award win- their way to the final judging in Tucson.”
ner. Sara Guergova, a graduate of GIA’s Jewelry Design pro- The 2023 Gianmaria Buccellati Foundation Award for
gram in London, received the sixth annual Gianmaria Excellence in Jewelry Design competition is underway and
Buccellati Foundation Award for Excellence in Jewelry De- open to students in GIA’s Jewelry Design courses who meet
sign. The 12 finalists and winner were announced at the GIA the eligibility requirements. Visit [Link]/buccellati-
Alumni Collective’s “Night at the Museum” event held foundation-award-jewelry-design for more information.

Figure 57. Sara Guer-

gova’s winning design
sketch for the 2022 Gi-
anmaria Buccellati
Foundation Award for
Excellence in Jewelry
Design, featuring gold,
black onyx, ruby, dia-
mond, tourmaline, and
blue sapphire.

128 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023

 2023 Tucson Photo Gallery

The pendant on the right is a miniature recreation of the “Bahia,” a massive rutilated
quartz fashioned by Glenn Lehrer and Lawrence Stoller. Each pendant in the collection
is carved from leftover raw material. The photo on the left is by Harold and Erica Van
Pelt; courtesy of Glenn Lehrer. The photo on the right and the rest of the photos in this
gallery are by Robert Weldon.
This suite of cat’s-eye topaz cabochons of unknown origin features a 135.80 ct center stone. Courtesy of Mayer
& Watt.

This brooch features a 13.27 ct Imperial topaz sur- In this beautifully matched pair of natural pearl ear-
rounded by 8.50 carats of diamonds, mounted in 18K rings, each 10 ct round pearl sits atop an astonish-
yellow gold. Courtesy of Jewelerette & Co. ingly large 32 ct pearl drop. Courtesy of Sima G. Ltd.
A spray of “mango” quartz measuring 92 × 103 mm from the Boyacá Province in Colombia. The yellow coloration
is believed to be due to the mineral halloysite. Courtesy of Cornerstone Minerals.
This naturally formed azurite disc on kaolinite ma- A stunning pair of Alexandre Reza 18K gold earrings.
trix, accompanied by a cluster of malachite, hails In this design, 112 diamonds (totaling 7.86 carats)
from the Malbunka copper mine in Northern Terri- dangle from two emeralds (totaling 2.12 carats).
tory, Australia. It measures 123 × 140 mm. Courtesy Courtesy of Jardin Jewels.
of Fine Art Minerals and Ghulam Mustafa.

Rough and carved hemimorphite, weighing 173.35 ct A vintage Fred of Paris demi-parure of earrings and a
and 112.26 ct, respectively. Courtesy of Evan Caplan. bangle, set with diamonds, rubies, emeralds, and sap-
phires. Courtesy of Jardin Jewels.
This aquamarine spray cluster from Skardu, Pakistan, measures 160 × 102 mm. Courtesy of Fine Art Minerals and
Ghulam Mustafa.
Indicolite from Afghanistan. The 116.7 ct crystal is Earrings and necklace in a jellyfish motif, set in 14K
from Nuristan Province, and the 9.39 ct cut gem is yellow gold. The mabe pearls are from the Sea of
from Kunav Province. Courtesy of Dudley Bauwet Cortez. Courtesy of Columbia Gem House.
Gems and Mountain Minerals International.

Master cutter Mark Gronlund combines natural Oregon A Cartier Les Oiseaux Libérés diamond, emerald, and
sunstone with unheated Nigerian pink tourmaline in this sapphire ring crafted in 18K white gold. Courtesy of
7.35 ct doublet, shown in table and profile views. Cour- Jardin Jewels.
tesy of Desert Sun Mining and Gems.
Aptly named the “Ring of Fire,” this Ethiopian opal in matrix displays a square of dark opal surrounded by a ring
of white opal. Courtesy of Ellie Gem Arts.
This bicolor spodumene is from Nuristan Province, Afghanistan. Terminations of the crystal appear to mimic the
mountain range from which it was mined. Courtesy of Dudley Blauwet Gems and Mountain Minerals International.
Vintage diamond and carved emerald bracelet designed An art nouveau enameled necklace in a floral motif in-
by Alexandre Reza. Courtesy of Jardin Jewels. corporating 20 seed pearls, four natural freshwater
pearls, and a freshwater Mississippi dogtooth pearl.
Courtesy of Bernard Nacht & Co./Under the Crown.

This horn-shaped boulder opal measures 75.5 × 15.9 × 4 mm and weighs 74.20 ct. Courtesy of Dufty Weis Opals Inc.
 Additional RTX and X-ray computed microtomogra-
REGULAR FEATURES phy (μ-CT) analyses were conducted to further study the
internal structures. RTX imaging of pearl A revealed a
non-bead cultured pearl used as a “bead” nucleus (figure
COLORED STONES AND ORGANIC MATERIALS 59). A light gray core surrounded by organic-rich concen-
tric growth structures followed by finer growth arcs was
Pinctada radiata atypical bead cultured pearls from the visible. An obvious demarcation with a small organic tail-
UAE. Pearling is embedded in the culture and traditions of like feature just below the surface overgrown with cul-
everyday life in the United Arab Emirates (UAE), especially tured nacre was also apparent. The lack of growth arcs in
in the trading center of Julfar. Abdulla Al Suwaidi, grand- the nacreous layers was due to rapid growth during the
son of one of the last traditional pearl divers in the UAE, culturing process (“Atypical ‘beading’ in the production
took on the challenge of reviving the country’s pearling in- of cultured pearls from Australian Pinctada maxima,”
dustry. He succeeded in establishing the Suwaidi Pearl GIA Research News, February 13, 2017). Similarly, pearl
Farm in Al Rams, Ras Al Khaimah (RAK), in 2005. B revealed a natural pearl used as a “bead” nucleus (figure
Suwaidi’s cultured pearl farm is recognized as the first of 60), with a very small but clearly visible dark gray core
its kind in the Arabian (Persian) Gulf region. surrounded by faint growth arcs. As with pearl A, a dis-
GIA’s Mumbai laboratory recently examined a quantity tinct demarcation with organic-rich areas and a small or-
of cultured pearls obtained from the farm. The parcel con- ganic tail-like feature were evident just below the surface
tained variously shaped white to cream-colored pearls. Ex- overgrown with cultured nacre. X-ray computed micro-
amination by real-time microradiography (RTX) showed a tomography analysis of both samples revealed clearer im-
variety of different internal structures including bead and ages of the demarcation between the “bead” pearls used
non-bead cultured pearls, but two pearls in particular (fig- as nuclei and the cultured nacre overgrowths.
ure 58) revealed very interesting structures and were con- Optical X-ray fluorescence examination did not show
firmed to be samples from atypical “bead” culturing any fluorescence in either sample. Energy-dispersive X-ray
experiments conducted by the farm. Pearl A was light fluorescence spectrometry revealed manganese levels
cream and near-round, weighing 0.87 ct and measuring below detection limits and strontium levels of 1130 ppm
5.13 × 4.93 mm, while pearl B was cream and button for pearl A and 1446 ppm for pearl B. The results from both
shaped, weighing 0.94 ct and measuring 5.42 × 5.12 × 4.93 testing methods were consistent with a saltwater growth
mm. When viewed under 40× magnification, both pearls environment. The ultraviolet/visible reflectance spectra
exhibited a smooth surface and possessed typical nacreous collected for both pearls showed features around 435 and
overlapping aragonite platelets. 460 nm and an additional weak band at 495 nm. These fea-

Figure 58. Two atypical

bead cultured pearls re-
covered from Pinctada
radiata mollusks from
Abdulla Al Suwaidi’s
farm in the UAE. Pearl
A (left) weighs 0.87 ct,
and pearl B (right)
weighs 0.94 ct. Photo
by Gaurav Bera.

138 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023

 Figure 59. Top: RTX im-
ages in three directions
of the non-bead cultured
pearl used as a “bead”
with a distinct demarca-
tion and an organic tail-
like feature in pearl A.
Bottom: μ-CT scans of
pearl A. The tail feature
is indicated by a yellow
arrow.

tures are similar to spectral observations previously separate from some natural pearls. This would be espe-
recorded for natural Pinctada radiata pearls (A. Al-Alawi cially true of the two samples featured in this report: If ei-
et al., “Saltwater cultured pearls from Pinctada radiata in ther of them were drilled and mixed in strands with
Abu Dhabi (United Arab Emirates),” Journal of Gemmol- natural Gulf pearls, an experienced gemologist would find
ogy, Vol. 37, No. 2, 2020, pp. 164–179). Raman analysis it difficult to separate them from the other pearls. The au-
using 514 nm laser excitation was carried out on the sur- thors have encountered natural pearls with similar dis-
faces of both pearls, and a doublet at 704/705 cm–1 as well tinct boundaries. It should also be noted that the aBCPs
as a peak at 1085 cm–1, indicative of aragonite, were ob- produced at the Suwaidi farm are the result of successful
served. Weak polyenic pigment-related peaks at 1130 and experiments, and they are not commercially available at
1540 cm–1 were only observed for pearl B and were associ- the time of this writing. The farm continues to conduct
ated with its cream coloration. The photoluminescence various experiments and aims to restore the UAE’s place
spectra were also consistent with the Raman results and in the global market by producing the finest Pinctada ra-
displayed high fluorescence together with the aragonite diata cultured pearls available.
peaks, typical of most nacreous pearls.
Abeer Al-Alawi
Using lower-quality pearls (both nacreous and non-
GIA, Global
nacreous), gemstones, and other materials as “bead” nuclei
has been a known practice for the last decade (“Chasing cul- Lubna Sahani
tured pearls at SSEF: Cultured pearls using a natural pearl GIA, Mumbai
as a bead,” 2020, [Link]
Nicholas Sturman
pearls-at-ssef-cultured-pearls-using-a-natural-pearl-as-a-
Bangkok
bead/). Although not seen on a regular basis, atypical
“bead” cultured pearls (aBCPs) are encountered in labora- Chunhui Zhou
tories from time to time and can be very challenging to GIA, New York

Figure 60. Top: RTX im-

ages in three directions
of a natural pearl used
as a “bead” with a
clear demarcation and
an organic tail-like fea-
ture in pearl B. Bottom:
μ-CT scans of pearl B.
The tail feature is indi-
cated by a yellow
arrow.

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023 139

Figure 61. Twelve Chilean abalone cultured pearls and two red abalone (Haliotis rufescens) shells with attached
cultured shell blisters, as well as three 7 mm traditional freshwater shell bead nuclei used to culture the pearls.
Photo by Emily Lane.

Bead cultured abalone pearls from Chile. Abalone is the and various methods have been attempted over the years
common name of the marine mollusk species that belongs (C.Y. Wentzell, “Cultured abalone blister pearls from New
to the Haliotis genus. Various abalone species are distrib- Zealand,” Fall 1998 G&G, pp. 184–200; M. Monteforte and
uted worldwide along cold coastal waters. They live in H. Bervera, “Abalone pearl culture on the west coast of the
rocky habitats, attaching firmly with their muscular foot. Baja California peninsula, Mexico,” World Aquaculture,
Abalone is highly prized for its meat, shell, and pearls. Due 2010, pp. 12–17). However, due to abalone’s hemophilia
to a marked decrease of abalone populations, harvesting and poor tolerance for chemicals and handling, any signif-
wild-caught abalone is heavily regulated. Thus, approxi- icant cut could cause major bleeding, and the nucleation
mately 95% of the global abalone supply comes from farms and implantation procedures could induce infection that
around the world, in countries such as China, South Korea, prohibits the mollusk from producing nacre. Therefore, the
South Africa, Chile, Australia, Taiwan, Japan, the United success rate of bead cultured (BC) abalone pearls was al-
States, and New Zealand (P.A. Cook, “Worldwide abalone ways low and the pearls were of unmarketable quality.
production statistics,” Journal of Shellfish Research, Vol. After 10 years of experimenting, scientists from the Uni-
38, No. 2, 2019, pp. 401–404). versity of Antofagasta, Chile, have successfully created
Abalone’s ear-shaped shells and pearls are known for abalone BC pearls using a patented method for producing
their unique, vibrant iridescent nacre, which is sought after free pearls in abalone. Twelve of their cultured pearls and
in jewelry and decorative items. The iridescence phenome- two abalone shells with a cultured shell blister attached (fig-
non is also known as “orient,” and it is attributed to inter- ure 61) were submitted to GIA for study in May 2022. They
ference and diffraction of light in the multilayered aragonite were the pilot crop of a new pearl culturing technique using
platelet microstructure and organic components. Due to the traditional freshwater shell bead nuclei (5 to 8 mm) inserted
rarity of natural abalone pearls, most abalone products used into grafting channels in the visceral mass of Haliotis
in jewelry today are either abalone shells or assembled cul- rufescens (red abalone) imported into Chile from Mexico in
tured shell blisters (also known as “mabe pearls”). the 1990s (figure 62, left) for abalone farming. The BC pearl
The culturing of whole pearls in abalone mollusks cultivation takes place during the last 24–28 months of the
started around the same time as shell blister culturing for abalone culture process that usually takes 48–52 months,
producing mabe pearl (assembled cultured shell blister), allowing the abalone meat and the pearl to be harvested at

140 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023

Figure 62. Abalone mollusks were grown in an aquaculture recirculation system in eight raceways of 5000 liters.
A local kelp species (shown on the right) was used to feed the mollusks twice per week. Photos courtesy of Jorge
Donoso Mena.

the same time. The facilities include an aquaculture recir- strong iridescence, along with patches and stripes of dark
culation system in eight raceways of 5000 liters, with con- brown organic materials. They exhibited a variety of
tinuous aeration by blowers and a culture water flow of 5000 baroque forms, and four samples possessed horn or tooth
liters per hour (figure 62, right). The local kelp species Lesso- shapes often found in natural abalone pearls. A character-
nia trabeculata obtained from natural, sustainably managed istic underlying botryoidal-like surface structure, usually
sources is used to feed the mollusks twice a week. Despite observed on natural abalone pearls and shells, was also
a high rate of bead nucleus rejection, the survival rate of the present (Fall 2015 Lab Notes, pp. 319–320). Real-time mi-
mollusks after grafting has been 99% and is currently at croradiography (RTX) revealed a round bead nucleus in the
15,000 grafted mollusks per trial, according to the farm. majority of samples, and the nacre layer thickness sur-
The cultured abalone pearls produced using this rounding the bead nuclei ranged from 0.10 to 1.20 mm (fig-
method displayed multicolored nacreous surfaces due to ure 63). One sample lacked a bead nucleus and was

ABL-3 Figure 63. RTX images

ABL-1 reveal 5–8 mm round
shell bead nuclei in a
ABL-5 majority of the abalone
ABL-2 ABL-4 samples (ABL), with the
nacre layer thickness
surrounding the bead
nuclei ranging from
0.10 to 1.20 mm. A
bead nucleus was ab-
sent in sample 2, only
showing layered growth
with void-related fea-
tures. While it is con-
sidered a non-bead
cultured pearl, its inter-
nal structure resembled
the internal structure
ABL-10 commonly found in
ABL-9 natural abalone
ABL-11 ABL-12
ABL-8 baroque pearls. Image
ABL-7
ABL-6 by Amiroh Steen.

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023 141

 RAMAN SPECTRA
2000000
ABL-1
ABL-3
ABL-4
ABL-6 Figure 64. Raman spec-
1500000
1085 tra using 514 nm laser
INTENSITY (counts)

excitation showed high

fluorescence background
and weak peaks of arag-
1000000
onite displayed at 704
and 1085 cm–1. These re-
sults are consistent with
704
10 natural abalone sam-
500000 ples selected for prelimi-
nary comparison study.

0
100 300 500 700 900 1100 1300 1500
RAMAN SHIFT (cm –1)

identified as a non-bead cultured (NBC) pearl. Its layered reflectance and photoluminescence spectra, and Raman
growth with void-related features resembled the internal spectra showed high fluorescence background and weak
structure commonly found in natural abalone baroque peaks of aragonite at 704 and 1085 cm–1 (figure 64). Energy-
pearls. dispersive X-ray fluorescence analysis revealed all the sam-
Ten natural abalone pearls were selected for spectro- ples had very low or below detection limit manganese
scopic and chemical comparison with the cultured pearl content and high levels of strontium ranging from 1100 to
samples, and both types of samples showed similar results. 5000 ppm. High iodine levels were also detected in most
No specific patterns were observed in ultraviolet/visible samples, which is common in abalone. Trace element con-
centrations determined by laser ablation–inductively cou-
pled plasma–mass spectrometry (LA-ICP-MS) were
compositionally similar in both sample types. However,
Figure 65. The cultured and natural abalone pearl
the cultured samples contained lower boron and higher
samples were compositionally similar in trace ele-
potassium contents. All four tested spots of the NBC sam-
ment concentrations determined by LA-ICP-MS.
ples could be clearly separated from natural samples in the
However, the cultured samples contained lower
plot of boron vs. potassium contents (figure 65). Since the
boron and higher potassium than the natural pearls.
one NBC pearl showed an internal structure similar to that
All four tested spots of the NBC samples are clearly
of natural abalone pearls, the plot can potentially be used
separated from natural samples in this plot.
to differentiate NBC pearls from natural samples. How-
ever, further study on both sample types is required. Addi-
B VS. K tionally, all abalone samples showed higher strontium
concentrations than Pinctada species previously studied.
300 The success of abalone BC pearl production could be a
promising addition to the gem and jewelry industry, and
Spot 3
this study provided useful information for future reference.
K (ppm)

200 Artitaya Homkrajae, Amiroh Steen,

Spot 4 Matthew Hardman, and Ziyin Sun
GIA, Carlsbad
100 Cultured Rubén Araya Valencia
Spot 1 Natural Alexander von Humboldt Institute of Natural Sciences,
Spot 2
0 University of Antofagasta
0 25 50 75 100
Jaime Pablo Maturana Zuñiga
B (ppm) Abalone Pearl Technology (APT SpA)
Antofagasta, Chile

142 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023

 uring 3.30 mm in diameter. When viewed under 40× mag-
nification, both exhibited typical nacreous overlapping arag-
onite platelets. Real-time microradiography (RTX) and X-ray
computed microtomography (μ-CT) analyses were carried
out to examine the internal structures in greater detail.
A minute gastropod shell measuring approximately
0.50 × 0.30 mm was observed in pearl A. The shell walls
were very thin, and minimal growth arcs were present in
the nacre surrounding it (figure 67). Marine gastropods are
known to range in size from a few millimeters to more
than a meter, so the size of the shell inside this pearl sug-
gests it was from a juvenile gastropod just beginning to
form its shell. The juvenile shell later forms the proto-
conch or first whorls of an adult gastropod (A. Nutzel, “Lar-
val ecology and morphology in fossil gastropods,”
Paleontology, Vol. 57, Part 3, 2014, pp. 479–503).
RTX imaging of pearl B revealed a minute shell that ap-
peared to be a foraminifera test (shell), measuring approxi-
mately 0.45 × 0.35 mm. Foraminifera are small unicellular
marine organisms found on the sea floor. A thin layer of
Figure 66. Two natural pearls recovered from Pinctada organic matter that appeared darker in the RTX and μ-CT
radiata mollusks by Bahraini divers, weighing 0.06 ct images seemed to envelop the foraminifera test, and a few
(pearl A, left) and 0.25 ct (pearl B, right). Photo by growth arcs were observed within the surrounding nacre
Gaurav Bera. (figure 68). The μ-CT scan was also rendered using special-
ized software (C. Zhou et al., “New 3-D software expands
GIA’s pearl identification capabilities,” GIA Research News,
Microscopic shells in natural pearls from Pinctada radiata. May 13, 2016) to create a three-dimensional image that
For centuries, natural pearl diving was Bahrain’s main in-
dustry. The vast majority of natural pearls fished around
the islands of Bahrain are found in the Pinctada radiata bi- Figure 67. The gastropod shell, measuring approxi-
valve (known in Arabic as mahar). These pearl beds have mately 0.50 × 0.30 mm, observed in pearl A. A: RTX
been known to produce the finest-quality natural pearls in image of the shell. B–D: μ-CT scans of the shell in the
the Arabian (Persian) Gulf region. Given the Gulf’s strate- X, Y, and Z directions, respectively. (The weak white
gic trading location and the richness of its pearling indus- rings on the lower right image are artifacts and not
try, the pearls from this region are coveted by traders and pearl-related structures.)
collectors alike (R. Carter, “The history and prehistory of
pearling in the Persian Gulf,” Journal of the Economic and A B
Social History of the Orient, Vol. 48, No. 2, 2005, pp. 139–
209). GIA’s Mumbai laboratory recently examined a quan-
tity of these pearls obtained from two local Bahraini divers
who claimed they were recovered from wild mollusks liv-
ing in the nutrient-rich shallow waters off the coast of
Sitra, one of Bahrain’s 33 islands.
Natural pearls from Pinctada radiata fished from
Bahrain and the Gulf region in general have a wide range
of shapes and colors. While most of the pearls examined
showed a variety of interesting internal structures, two of
them had a noteworthy feature: a minute shell contained C D
within. Natural pearls are very rare and form inside mol-
lusk shells, and a shell within a pearl within a shell is like
nature’s version of a nesting doll. Such structures have
been encountered by GIA on rare occasion (see Winter
2015 Lab Notes, pp. 434–436).
Both pearls were very small (figure 66). The smaller one
(A) had a strong yellow bodycolor and an oval shape, weigh-
ing 0.06 ct and measuring 2.07 × 1.93 mm. The larger round
one (B) had a light cream color, weighing 0.25 ct and meas-

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023 143

A B
multilocular, or multichambered, with tubular chambers
arranged around a growth axis to form a beautiful milioline
arrangement of chambers (A.R. Loeblich and H. Tappan,
Foraminiferal Genera and Their Classification, Springer,
New York, 1988). The minute size of the shells within these
tiny pearls is a good indicator of their natural origin.
Energy-dispersive X-ray fluorescence spectrometry on
pearls A and B revealed low manganese levels of 39.0 ppm
and 17.8 ppm and high strontium levels of 1768 ppm and
C D 1497 ppm, respectively, which is characteristic of formation
in a saltwater environment. Raman analysis was also car-
ried out using 514 nm laser excitation on the surface of each
pearl. A doublet at 702 and 705 cm–1 as well as a peak at
1085 cm–1 indicative of aragonite were observed, along with
minor polyenic pigment peaks at 1130 and 1530 cm–1. Pho-
toluminescence (PL) spectra were also collected on both
pearls. Pearl A revealed three broad peaks at 620, 650, and
680 nm, characteristic of many naturally colored pearls,
while pearl B showed clear aragonite peaks and low fluo-
Figure 68. The foraminifera test, measuring approxi- rescence. An ultraviolet/visible reflectance spectrum was
mately 0.45 × 0.35 mm, observed in pearl B. A: RTX collected only for pearl B within the 220–850 nm range, as
image of the foraminifera test. B–D: μ-CT scans of the pearl A’s size prevented the detector from obtaining a clear
foraminifera test in the X, Y, and Z directions, respec- result. Faint features at 420 and 495 nm were, like the PL
tively. (Again, the weak white rings on the lower right results, consistent with natural coloration. Similar spectral
image are artifacts.) observations have previously been documented in natural
Pinctada radiata pearls (A. Al-Alawi et al., “Saltwater cul-
tured pearls from Pinctada radiata in Abu Dhabi (United
made it easier to see the external morphology of the shell Arab Emirates),” Journal of Gemmology, Vol. 37, No. 2,
within the pearl (figure 69). The foraminifera test appeared 2020, pp. 164–179).

Figure 69. 3D images

constructed from μ-CT
scans of pearl B illustrate
the morphology of the
foraminifera test. Images
by Emiko Yazawa.

144 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023

 Figure 70. Part of the
suite of sapphires stud-
ied for this project
(weighing 0.23–1.41 ct).
Photo by Sasithorn
Engniwat; courtesy of
Greenland Ruby.

The gemological examination of these Pinctada radi- toq ruby and pink sapphire deposit, located 160 km south
ata pearls proved very rewarding, especially with regard to of Nuuk.
their internal structures. Research into what causes the The corundum extracted at Aappaluttoq is heated in
formation of a pearl in the wild is ongoing. Hence, finding air at temperatures over 1500°C for several hours, with
these minute shells that may be the initiation of growth some flux added. In rare cases, the heating significantly al-
in these two natural pearls is a truly rare circumstance. ters the color of light pink and near-colorless sapphire, re-
Nishka Vaz sulting in fancy-color sapphire. GIA recently studied a
GIA, Mumbai parcel of 22 of these fancy-color sapphires mined and
treated by Greenland Ruby (figure 70).
Nicholas Sturman
A combination of ultraviolet/visible/near-infrared (UV-
Bangkok
Vis-NIR) spectra and trace element chemistry using laser
Abeer Al-Alawi ablation–inductively coupled plasma–mass spectrometry
GIA, Global (LA-ICP-MS) were collected at GIA in Bangkok to identify
the causes of color. For full quantification of the chro-
Color study of fancy sapphire from Greenland. Over the mophores, it is necessary to do this analysis in oriented,
past 70 years, several ruby and pink sapphire localities have clean samples with a known path length. Since most of the
been documented in the region of Greenland’s capital city, Greenland material is polished as cabochons and contains
Nuuk (P.W.U. Appel and M. Ghisler, “Ruby- and sap- some inclusions, it is very challenging to calculate or even
phirine-bearing mineral occurrences in the Fiskenaesset, estimate the absorption coefficients. We limited ourselves
Nuuk and Maniitsoq regions, West Greenland,” GEUS, to a rough interpretation of the chromophores based on the
Geological Survey of Denmark and Greenland, 2014). Sev- shape of the spectra. Orientation of the samples was esti-
eral of these localities are currently exploited on a small mated based on the pleochroism.
scale by the local communities. In 2017, Greenland Ruby Additionally, we could not precisely identify which
A/S started a large-scale mining operation at the Aappalut- area of the stone was analyzed during the spectroscopy, and

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023 145

 UV-VIS-NIR SPECTRA
ABSORBANCE

E-ray Figure 71. A UV-Vis-

O-ray NIR spectrum domi-
nated by Cr3+ features
often results in a pink
to red stone, depending
on the concentration of
300 350 400 450 500 550 600 650 700 750 800 Cr3+ (in ppma). When
WAVELENGTH (nm) the sample is oriented
down the c-axis, the
color often trends to-
Mg Ti V Cr Fe Ga [Ti]-[Mg]
ward reddish purple
S269663 Spot 2 33 36 26 283 500 11 3 (i.e., fancy sapphire).

S269663 Spot 3 34 35 26 287 508 11 1

S269663 Spot 4 37 39 26 268 478 11 2

Detection limits 0.011 0.033 0.004 0.068 0.824 0.002

(ppma)

thus all UV-Vis spectra were bulk measurements and com- Greenland ruby, this comes down to the balance between
bined the color of different zones. In contrast, the LA-ICP- Mg2+ vs. Ti4+, Si4+, and H+. Since the stones have all been
MS analysis was performed at very specific spots, allowing treated in air (oxidizing atmosphere), we can assume that
the analysis of specific zones. H+ has diffused out of them. With the method used to meas-
The literature describes six chromophores responsible ure trace elements (LA-ICP-MS with a quadrupole MS), we
for color in natural corundum (E.V. Dubinsky et al., “A cannot correctly quantify the Si4+ content in these stones.
quantitative description of the causes of color in corun- The possibility that Si4+ plays a role in the color cannot be
dum,” Spring 2020 G&G, pp. 2–28). Four of the chro- excluded. Regardless, there is no indication here that Si4+ is
mophores from that study are responsible for the colors in required to tilt the balance of chromophores when compar-
this suite. ing trace elements vs. observed color. Since we do not aim
Most stones from the Aappaluttoq deposit are colored to quantitatively analyze the color, we ignore any potential
by the Cr3+ chromophore, giving them their typical pink to presence of Si4+. This means that the concentrations of Mg2+
red color (figure 71). The intensity of the color is directly and Ti4+ are the determining factors: Only if the amount of
linked to the concentration of Cr3+, which can go up to sev- Mg2+ is larger than the Ti4+ concentration can trapped holes
eral 1000 ppma, and in exceptional cases even to 15,000 develop in sapphire from Greenland.
ppma. Most of the lighter-colored material has a concen- Two ions in sapphire can pair with trapped holes: Fe3+
tration of a few 100 ppma, resulting in a pink color. and Cr3+. Since the trapped hole preferentially pairs with
Depending on the orientation, it can appear more pur- Cr3+, this is the chromophore observed in these stones. The
ple (when viewed through the c-axis). In some cases, this quantity of trapped holes that can form in corundum is
is so strong that the face-up color of the stone is modified very low, and as such, not all the Cr3+ couples with a
to an extent that it cannot be defined as ruby or pink sap- trapped hole. The final color will be a combination of the
phire and falls into the fancy reddish purple range. Some trapped hole color and the pure Cr3+ chromophore, result-
of the stones develop a more orangy tone after treatment. ing in a combination of pink to red and orange colors.
(A bright orange oval brilliant seen at the 2023 Tucson Trapped holes are very strong coloring agents that can have
shows is highlighted on pp. 115–116 of this issue.) This a major impact even when their concentration is very low.
color can be attributed to the creation of trapped holes in The Fe3+ chromophore is the weakest one encountered
combination with Cr3+ (figure 72). in natural corundum. Very high concentrations of Fe3+ are
To create trapped holes, the chemical balance in the required to create a yellow color, often exceeding several
stone needs to be correct. Only when a stone is acceptor thousand ppma. Since the Greenland rubies have a maxi-
dominated will trapped holes be present. In the case of mum iron concentration under 1500 ppma, this chro-

146 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023

 UV-VIS-NIR SPECTRA
Figure 72. The peaks at
550 and 430 nm in the
o-ray spectrum (black)
suggest the presence of
Cr3+, causing pink to
red in the stone. The
ABSORBANCE

O-ray trace element analysis

Cr 3+ removed (in ppma) shows a clear
excess of Mg2+ com-
pared to the Ti concen-
tration, highlighting
that trapped hole cre-
300 350 400 450 500 550 600 650 700 750 800 ation is very likely.
WAVELENGTH (nm) When we remove the
absorbance spectrum of
Cr3+ from the stone’s
Mg Ti V Cr Fe Ga [Ti]-[Mg]
absorption spectrum
S269596 Spot 1 31 13 15 820 924 12 –18 (blue), it is obvious that
the combination of Cr3+
S269596 Spot 2 31 13 15 804 927 12 –18 with the trapped hole is
S269596 Spot 3 32 13 16 812 971 12 –19 the cause of the remain-
ing color and the strong
Detection limits 0.011 0.033 0.004 0.068 0.824 0.002 orange tint observed in
(ppma)
the stone.

mophore creates only a very weak yellow color (figure 73). common, which makes it very difficult to notice any color
Other common chromophores are much stronger and more contribution by Fe3+. Only when the Cr3+ concentration is

UV-VIS-NIR SPECTRA
Figure 73. This plot
shows the o-ray spec-
trum of a pale yellow
sample (black) and a
modified o-ray spec-
ABSORBANCE

O-ray
trum where some con-
Cr 3+ removed tribution by Cr3+ is
subtracted (blue). When
the Cr3+ spectrum is re-
moved, a pure Fe3+ spec-
trum remains. This
300 350 400 450 500 550 600 650 700 750 800
spectrum is character-
WAVELENGTH (nm) ized by the double peak
at 377–388 nm and one
Mg Ti V Cr Fe Ga [Ti]-[Mg] at 450 nm. The ratio of
the 388 to 377 nm band
S269612 Spot 1 16 14 10 96 1150 12 –2
suggests a relatively
S269612 Spot 2 13 12 9 81 1136 13 –1 low Fe3+ chromophore
concentration, which
S269612 Spot 3 14 12 9 76 1143 14 –2 matches the observed
Detection limits 0.011 0.033 0.004 0.068 0.824 0.002 color and analyzed
(ppma) chemistry (in ppma).

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023 147

extremely low can we detect the pale yellow color from The nature of the treated corundum from Greenland
the Fe3+ chromophore. makes high-quality spectroscopy a challenge. Neverthe-
The Fe2+-Ti4+ chromophore creates a blue color in less, we were able to isolate and identify the chromophores
corundum and is relatively strong (figure 74). When ob- responsible for the fancy colors observed in these sapphires.
served in ruby from Greenland, its effect is usually limited It is important to note that many of these fancy colors
to a slightly increased purplish hue in the chromium-rich developed as a result of heat treatment. The trapped hole–
corundum, occasionally modifying the color to a dominant related chromophores only become visible after H+ diffused
purple hue. Only when the Cr3+ concentrations are ex- out of the stone under oxidizing conditions. These same
tremely low is there a chance for the corundum to have a conditions also allow the iron to oxidize and create the pale
blue color. We still observe some cloudiness after the treat- yellow color by Fe3+ chromophores. The high temperatures
ment in the blue zones. were also required to dissolve rutile particles into the
We assume that this type of rough corundum had a corundum lattice and form Fe-Ti pairs.
milky appearance caused by a high concentration of fine While these fancy-color stones remain a curiosity, they
rutile particles before treatment. During treatment at make for a welcome addition from a mine that is known
temperatures over 1200°C, some of the particles dissolved for its ruby and pink sapphires.
into the lattice, becoming available to pair with Fe2+ that Wim Vertriest, Charuwan Khowpong,
was already present and creating a blue bodycolor. Due to and Polthep Sakpanish
the large amounts of Ti4+ that diffused into the crystal lat- GIA, Bangkok
tice, the iron was forced to remain in a reduced state Martin Viala
(Fe2+). Greenland Ruby

UV-VIS-NIR SPECTRA

E-ray
O-ray
Figure 74. The ex-
tremely low Cr3+ con-
ABSORBANCE

tents allow other

features to be domi-
nant. In this case, Fe2+-
Ti4+ charge transfers
with a large peak at
580–620 nm are respon-
sible for a blue color.
The absorption spec-
trum of the stone also
shows some features re-
300 350 400 450 500 550 600 650 700 750 800 lated to Fe3+, which is a
WAVELENGTH (nm) very weak chromophore
and does not influence
the color.
Mg Ti V Cr Fe Ga [Ti]-[Mg]

S108627 Spot 1 46 120 15 21 544 10 74

S108627 Spot 2 38 54 14 39 548 11 16

S108627 Spot 3 47 52 16 40 613 11 5

Detection limits 0.011 0.033 0.004 0.068 0.824 0.002

(ppma)

148 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023

Figure 75. A handful of gem gravel collected from the Lofa River in Weasua, Liberia, showing the high concentra-
tion of ruby material. Photo courtesy of Diamonds for Peace.

Liberian ruby. A gemological education can often prove quartz, and possible spinel. Samples were purchased from
useful in the most unusual of circumstances. On a recent the miners for analysis, which came as a welcome surprise
trip to rural Liberia with the Japan-based NGO Diamonds since they were unaware that these colored stones could
for Peace (DfP), the author discovered a previously unac- carry any value (figure 76, left).
knowledged source of ruby. The original purpose of the The author subsequently analyzed all the samples at
visit was to deliver a rough diamond grading and valuation the Gemmological Certification Services (GCS) laboratory
workshop to the miners working and living in 17 commu- in London. Using advanced and standard gemological test-
nities across Liberia. During the trip, the author stayed in ing, the identity of all the presumed gem-quality minerals
the town of Weasua, one of the larger mining communi- of the collected specimens was confirmed. Particular focus
ties, to experience the standard of living and the general was placed on the rubies due to their apparent abundance
working conditions of the miners. The poverty is extreme. at the mine site, with the hope of providing another means
Most homes are without electricity, facilities, or running of income for this community.
water, and only three functional water pumps exist for a All ten samples tested were identified as corundum.
population of 5,000, all of which place the health and the Due to the nature of the rough material, standard testing
well-being of this community at risk. DfP has been work- was limited, beyond measuring specific gravity—which,
ing with this community since 2018 and has made signif- averaging 4.0 across all samples, supported corundum. The
icant inroads in supporting and improving the lives of these ultraviolet/visible/infrared absorption spectrum, with its
miners, but there is a long way to go, and strategies to aid chromium-associated features, was also consistent with
their mission are continually being developed. ruby. X-ray fluorescence analysis of the trace element
It was therefore fortuitous that during the visit, the au- chemistry revealed 1029–2377 ppm iron and 721–2915
thor learned that there were other promising gem-quality ppm chromium, with low levels of gallium (54 ppm) and
minerals in the alluvial gravels the miners were collecting vanadium (53 ppm).
in their pursuit of diamond (figure 75). Upon initial inspec- Generally, the color of the alluvial material collected
tion while on-site, the author noted material resembling ranged from pinkish gray to a deep purplish red, with sizes
corundum (primarily ruby), garnet, topaz, rock crystal varying from 6 to 14 mm (1.45–15.90 ct), although the au-

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023 149

 Figure 76. Left: One of
the Liberian diamond
miners displays some of
the rubies she found
while digging. Right: A
ring mounted with a
Liberian ruby. Photos
by Diamonds for Peace
(left) and Alex Herbert
(right).

thor was shown larger pieces on-site. All of the material gemstone reportedly from Tanzania (figure 77). Its refrac-
showed a moderate to weak red fluorescence response in tive index of 1.714 and hydrostatic specific gravity of 3.59
long-wave UV and inert to weak red in short-wave UV,
consistent with high-iron ruby due to the quenching of
Cr3+-associated fluorescence by iron. The majority of the Figure 77. This faceted purple spinel, weighing 0.51 ct
samples showed lamellar twinning, surface-reaching frac- and measuring 5.20 × 4.17 × 3.10 mm, was reportedly
tures, and iron staining. The material was translucent to from Tanzania. Photo by Jinlin Wu.
opaque, meaning that clear observation of any other inclu-
sions was limited.
Although the material is not considered high-quality
ruby, some of the better pieces have been polished to create
attractive cabochons (figure 76, right). Diamonds for Peace
is currently developing a business plan to monetize this
material for the benefit of the community. The GCS labo-
ratory will continue to support Diamonds for Peace as that
organization explores this new potential avenue of income
for the miners of Weasua.
Beth West
Gemmological Certification Services, London

Forsterite in a purple Tanzanian spinel. Purple/violet

spinel from Tanzania is popular in the gem market today.
Recently, author JW encountered an oval-shaped purple

150 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023

 UV-VIS-NIR SPECTRUM
1.0
Fe 2+
Fe 3+
0.9 Cr 3+
V 3+
Mn2+
0.8 Co2+

0.7
Figure 78. The UV-Vis-
ABSORBANCE

NIR spectrum of the

0.6 spinel sample shows
absorption peaks
mainly related to Fe2+
0.5 and Fe3+.

0.4

0.3

0.2
350 400 450 500 550 600 650 700 750 800
WAVELENGTH (nm)

indicated spinel. Its ultraviolet/visible/near-infrared (UV- As a rock-forming mineral, forsterite (Mg2SiO4) itself
Vis-NIR) absorption spectrum was characterized by a se- is common, especially in marble-hosted spinel. However,
ries of absorption bands centered at 555 nm and several it is rarely observed in Tanzanian spinel and has been re-
iron-related small bands centered at around 371, 385, 461, ported only once to our knowledge (E. Gübelin and J.I.
and 478 nm, as well as an extremely weak band at ~625 Koivula, Photoatlas of Inclusions in Gemstones, Vol. 2,
nm for Co2+ (figure 78). This absorption profile revealed Opinio Publishers, Basel, Switzerland, 2005, 829 pp.).
that iron (Fe2+ and Fe3+) was the main coloring element
(P.M. Belley and A.C. Palke, “Purple gem spinel from Viet-
nam and Afghanistan: Comparison of trace element chem- Figure 79. A colorless subhedral forsterite (Fo) inclu-
istry, cause of color, and inclusions,” Fall 2021 G&G, pp. sion and a dolomite (Dol) inclusion with an unknown
228–238). dark material attached. Photomicrograph by Jinlin
Energy-dispersive X-ray fluorescence analysis was per- Wu; field of view 0.66 mm.
formed to determine the sample’s trace element concentra-
tion, which showed 10950 ppmw iron, 963 ppmw zinc, 52
ppmw chromium, 94 ppmw vanadium, and 9 ppmw cobalt.
These results were similar to the chemical features of rep-
resentative purple Tanzanian spinel: extremely high iron
content, medium zinc content, and low chromium and
vanadium contents (G. Giuliani et al., “Pink and red spinels Fo
in marble: Trace elements, oxygen isotopes, and sources,”
Canadian Mineralogist, Vol. 55, No. 4, 2017, pp. 743–761).
Microscopic observation revealed a colorless subhedral
mineral, a dolomite with an unknown dark material at-
tached (figure 79), and oriented short needles (figure 80).
Raman spectroscopy identified the colorless mineral as
forsterite. According to the RRUFF database (B. Lafuente Dol
et al., [Link]
peaks at 822, 855, 918, and 963 cm−1 were consistent with
the main peaks of forsterite (figure 81).

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023 151

 Under high-temperature conditions, spinel and forsterite
are both stable mineral phases in marbles (Giuliani et al.,
2017). Spinel is found in marble or impure marble associ-
ated with calcite, dolomite, forsterite, clinohumite, and
phlogopite in deposits in Myanmar and Vietnam. Thus,
forsterite is commonly seen as an inclusion in spinel from
these sources (G. Giuliani et al., “Pink and red gem
spinels in marble and placers,” InColor, No. 43, 2019, pp.
14–28).
It was noteworthy to see forsterite in the Tanzanian
sample. This finding could indicate a similar formation
environment between spinel in Tanzanian deposits and
those in Myanmar and Vietnam.
Jinlin Wu (wujl@[Link]), Hong Ma,
Figure 80. Oriented short needles in the spinel. Photo- and Huihuang Li
micrograph by Jinlin Wu; field of view 1.10 mm. National Gemstone Testing Center Shenzhen Lab
Shenzhen, China

In Tanzania, the primary deposits of spinel occur in DIAMONDS

marbles and calc-silicate rocks that belong to the Neopro-
terozoic metamorphic Mozambique Belt (G. Giuliani et Grayish blue CVD diamond colored by GR1 and SiV–. Sil-
al., “Les gisements de rubis et de spinelle rouge de la icon is a common impurity in chemical vapor deposition
Ceinture Métamorphique Néoprotérozoıque Mozambi- (CVD) laboratory-grown diamond. The silicon-related SiV–
caine,” Revue de Gemmologie, Vol. 192, 2015, pp. 11–18). center, which has zero-phonon lines at 736.6 and 736.9 nm

RAMAN SPECTRA

823 Forsterite R040018

Colorless mineral

855

Figure 81. Comparison

of Raman spectra of the
325 586
919 959 colorless mineral and
222 302 371 428 541 602 forsterite in the RRUFF
INTENSITY

database. Peaks in the

inclusion spectrum at
312, 405, 664, and 766
cm–1 are from the host
spinel. Spectra are offset
822 vertically for clarity.
855

918 963
225 373 432 543 592

200 400 600 800 1000 1200

RAMAN SHIFT (cm –1)

152 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023

Figure 82. Left: This 1.63 ct grayish blue diamond was identified as a CVD synthetic, with its color caused by both
the GR1 and SiV– centers. Photo by Xiaoyu Zhang. Right: Dark cloud-like non-diamond carbon inclusions. Photo-
micrograph by Meng Li; field of view 0.25 mm.

and is active in absorption and luminescence, is widely National Gemstone Testing Center’s (NGTC) Beijing
considered an identifying feature of CVD diamond. Gen- laboratory recently received a 1.63 ct grayish blue princess
erally, the SiV– center is detected by photoluminescence cut (figure 82, left) for identification. Standard testing iden-
spectroscopy and is not observed using absorption spec- tified it as a CVD synthetic diamond. The sample showed
troscopy for most CVD synthetics due to the low silicon an even color distribution. Microscopic investigation re-
content. The absorption spectra may reveal the presence vealed abundant non-diamond carbon inclusions in the
of the SiV– center in the case of a relatively high silicon form of both crystals and clouds, with the clouds distrib-
content, but rarely in amounts significant enough to affect uted along the growth layer (figure 82, right). The
the color. When the silicon content becomes particularly visible/near-infrared (Vis-NIR) spectrum (figure 83),
high, the absorption of SiV– is in sufficient concentration recorded at liquid nitrogen temperature, included a strong
to influence the bodycolor of a diamond. It has been re- broad SiV– center defect at 737 nm next to an equally
ported that the color of silicon-doped blue CVD diamonds strong GR1 center. The diamond’s grayish blue bodycolor
produced by the PDC Company is attributed to strong ab- was attributed to both the strong GR1 center and the high
sorption of the rather intense SiV– center (A. Peretti et al., intensity of the SiV– center, which produced preferential
“New generation of synthetic diamonds reaches the mar- absorption in the red part of the spectrum. In addition,
ket (Part A): Identification of CVD-grown blue diamonds,” weak absorptions at 830, 856, and 946 nm (the SiV0 center)
Contributions to Gemology, No. 14, 2013, pp. 3–20). were observed, which were also reported in CVD synthetic

VIS-NIR SPECTRUM

737
0.55 741

Figure 83. In the Vis-

NIR region, in addition
ABSORBANCE

to very strong absorp-

0.50 720 730 740 750 tions at 737 nm (SiV–)
and 741 nm (GR1),
weak absorptions at
830, 856, and 946 nm
830
856 (SiV0) were observed.
946
0.45

400 500 600 700 800 900 1000

WAVELENGTH (nm)

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023 153

 thetics, suggesting that it had undergone post-growth high-
temperature annealing before irradiation.
The PL spectrum with 532 nm laser excitation also ex-
hibited a rather large silicon doublet peak at 736.6/736.9 nm
that dwarfs all other PL peaks, including the GR1 in the
spectrum. The intensity ratio between the 736.6/736.9 nm
PL peak and the diamond Raman peak was about 52. The
PL spectrum reaffirmed that the 737 nm absorption peak of
the ultraviolet/visible spectrum was the SiV– defect, and the

PL SPECTRA
503.5
Figure 84. DiamondView imaging showed green fluo- 50000
A 486.3
rescence with blue dislocations. The diamond also

INTENSITY (counts)
revealed a layered growth structure, indicating a 40000
start-stop cycling growth process typical of CVD dia-
mond, as CVD diamond growth takes place layer by 30000 488 490 494
layer on the top surface of the growing crystal. Image
by Xiaoyu Zhang. 20000 486 490 494
Diamond
Raman
10000 503.2 505 530.7 540.7 552.4

diamonds with high silicon content (Peretti et al., 2013; Z. 0

Song et al., “Silicon-doped CVD synthetic diamond with 480 490 500 510 520 530 540 550 560
photochromic effect,” Journal of Gems and Gemmology,
Vol. 18, No. 1, 2016, pp. 1–5).
Infrared spectroscopy showed features typical for type IIa 50000
737
diamond (i.e., no defect-related absorptions were detected). B 736.9
736.6
The diamond was inert to short-wave and long-wave ultra- 40000
INTENSITY (counts)

violet illumination. Under the ultra-short-wave UV radia-

tion of the DiamondView, the sample fluoresced green, with 30000
minor areas showing some blue dislocations. It also showed
subtle growth striations and two parallel planes fluorescing 20000
734 736 738 740
stronger than the bulk of the crystal (figure 84). The sample Sideband of
Diamond
did not phosphoresce. Photoluminescence (PL) spectra taken 10000 Raman 737 nm
573 741 peak
with various laser excitation wavelengths and at liquid ni- 575 637
trogen temperature (figure 85) showed the presence of radi- 0
ation-related features including very strong 3H at 503.5 nm, 550 600 650 700 750 800 850
strong GR1 at 741 nm, and weak 486.3 nm. The PL spectra
also indicated strong nitrogen-vacancy centers at 575 NV0
and 637 NV– nm and multi-nitrogen defects such as H3 and 600
H2, combined with the absence of the unassigned 596/597 C
nm doublet that is normally seen in as-grown CVD syn-
INTENSITY (counts)

986
400

Figure 85. The PL spectrum of the grayish blue CVD

synthetic diamond obtained with 473 nm excitation
200
(A) showed that the H3 center (503.2 nm) and radia-
tion-related features included very strong 3H at 503.5
nm and weak 486.3 nm. B: 532 nm excitation revealed
the NV centers and silicon-vacancy center that are 0
typically observed in CVD-grown diamonds, along 960 965 970 975 980 985 990 995 1000
with strong GR1. C: 785 nm excitation showed the H2 WAVELENGTH (nm)
center at 986 nm.

154 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023

Figure 86. Top: A series of images showing the gradual change (left to right) in the phosphorescence of the 0.71 ct
HPHT-grown colorless diamond after removing it from the LWUV source (left image) in increments of 30 seconds for
150 seconds (right image). Bottom: The gradual change in phosphorescence after exposure to a SWUV source (left
image) in increments of 30 seconds. The greenish blue and orange emissions decay differently, with the blue fading
faster (~90 seconds, fourth image from left) than the orange (~240 seconds, right image). Images by Guy Borenstein.

silicon content of this diamond was so high that it produced Under short-wave ultraviolet (SWUV, 254 nm) radia-
strong absorption in the red range of visible light and affected tion using a standard viewing UV cabinet and a De Beers
the color of the diamond together with the GR1 center. PhosView instrument, the stone displayed a mix of
Wang Yang (wangyang@[Link]), Xiaoyu Zhang, strong greenish blue and medium orange zoned fluores-
and Zhonghua Song cence and phosphorescence (figure 86, bottom). The col-
NGTC, Guangzhou and Beijing ored phosphorescing areas decayed at different rates, with
the greenish blue fading faster than the orange (~90 sec-
onds vs. ~240 seconds, with a half-life of about 25 and 35
HPHT-grown colorless diamond displaying unusual phos-
seconds, respectively).
phorescence effects. Phosphorescence longer than 10 sec-
Deep-UV (<225 nm) imaging demonstrated a strong
onds in reaction to ultraviolet radiation is regularly
greenish blue fluorescence with a clear cuboctahedral struc-
observed in colorless to near-colorless and blue-colored dia-
ture associated with HPHT growth (figure 87). After removal
monds grown by a high-pressure, high-temperature (HPHT)
technique. The gemological laboratory of Stuller Inc. re-
cently tested an HPHT-grown colorless (approximately E
color grade with a blue-gray overtone) oval brilliant dia- Figure 87. Deep-UV imaging of the 0.71 ct colorless
mond weighing 0.71 ct that showed unusual fluorescence diamond showing a cuboctahedral-structured lumi-
and phosphorescence reactions. nescence associated with HPHT growth. Image by
When exposed to a long-wave ultraviolet (LWUV, 365 Guy Borenstein.
nm) source for 60 seconds using a standard UV viewing
cabinet, the stone emitted a gradually increasing strong or-
ange fluorescence, followed by a long-lasting noticeable
phosphorescence (~150 seconds, figure 86, top). The fluo-
rescence and phosphorescence reactions demonstrated a
nonuniform color distribution. Similar results were ob-
served using other light sources, including 395 nm UV
LED, 405 nm blue laser, a tungsten incandescent flashlight,
and a fiber-optic halogen unit.
Given that orange phosphorescence can also be gener-
ated by low- and non-UV light sources, further analysis
was performed using a series of spectral bandpass filters
to identify other wavelengths that stimulate the orange
phosphorescence. Testing demonstrated that medium- to
strong-intensity light centered at around 430 nm produces
the excitation.

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023 155

 FTIR SPECTRUM
2.5

2.0 Figure 88. The infrared

spectrum of the 0.71 ct
ABSORBANCE

HPHT-grown diamond
1.5
revealed it was type IIb
with relatively high lev-
1.0 els of uncompensated
boron (for a colorless
diamond) at 4089, 2929,
0.5
2800, and 2455 cm–1.

0
6000 5000 4000 3000 2000 1000 0
WAVENUMBER (cm–1)

from the excitation source, the stone displayed a phospho- length reinforces the theory that the recorded orange phos-
rescence reaction similar to the reaction under SWUV. phorescence is of a different donor-acceptor pairing.
Infrared spectroscopy revealed a type IIb diamond with Light of 430 nm wavelength can be found in most fluo-
relatively high levels of electrically uncompensated boron rescent and all incandescent indoor light sources, as well as
at 4089, 2929, 2800, and 2455 cm–1 (figure 88). Photolumi- the desk lamps and color grading cabinets found in gemo-
nescence spectroscopy with 532 nm laser excitation at liq- logical laboratories. Therefore, considering the possibility of
uid nitrogen temperature (77K) showed no results, except more diamonds with this property in the market, their phos-
for the diamond’s first- and second-order Raman spectrum phorescence produced from visible light makes it challeng-
peaks, a common tendency in HPHT-grown colorless dia- ing to assign a color in the colorless to near-colorless range
monds (S. Eaton-Magaña et al., “Observations on HPHT- for these diamonds during grading. Even under standard non-
grown synthetic diamonds: A review,” Fall 2017 G&G, pp. or low-UV output 6500K fluorescent bulbs, the orange lu-
262–284). minescence generated by the source becomes noticeable,
The two phosphorescence colors emitted and their rel- adding a yellowish overtone to the blue-gray bodycolor and
atively long duration are well-known properties of HPHT- changing its apparent color grade to approximately F to G.
grown diamonds (S. Eaton-Magaña and R. Lu, Such a change may impair the diamond grader’s ability to
“Phosphorescence in type IIb diamonds,” Diamond and accurately assess the stone’s bodycolor.
Related Materials, Vol. 20, No. 7, 2011, pp. 983–989; Eaton- Guy Borenstein and Sean O’Neal
Magaña et al., 2017), and a blue-orange luminescence com- Stuller Inc.
bination in one gem was also previously reported Lafayette, Louisiana
(Eaton-Magaña and Lu., 2011; B. Deljanin et al., “NDT
breaking the 10 carat barrier: World record faceted and gem-
quality synthetic diamonds investigated,” Contributions to SYNTHETICS AND SIMULANTS
Gemology, Vol. 15, No. 1, 2015) but of weaker intensity.
The laboratory-grown diamond’s property of bicolor, decay- Synthetic color-change wakefieldite. Author AA, a gem-
varying phosphorescence has been previously reported as stone cutter and collector of rare synthetic gems, recently
well (K. Watanabe et al., “Phosphorescence in high-pres- submitted one rough and one faceted stone to GIA for sci-
sure synthetic diamond,” Diamond and Related Materi- entific examination. The material, sourced from RG Crys-
als, Vol. 6, No. 1, 1997, pp. 99–106; Eaton-Magaña and Lu, tals in Bangkok, was stated to be Czochralski-pulled,
2011; Ulrika F.S. D’Haenens-Johansson et al., “Large color- neodymium-doped yttrium orthovanadate, which is com-
less HPHT-grown synthetic gem diamonds from New Dia- monly used in laser applications. It can also form naturally
mond Technology, Russia,” Fall 2015 G&G, pp. 260–279). as the mineral wakefieldite-(Y), with an ideal chemical for-
Considering the HPHT growth method, the long-lasting mula of YVO4. His first observation was a striking color
greenish blue phosphorescence produced by exposure to change from pink or purple to a blue or even green color
SWUV and deep UV was expected. It was previously sug- under various lighting conditions and viewing angles. The
gested to be related to nitrogen-boron donor-acceptor pair finished gemstone (figures 89 and 90) displayed strong fire
recombination (Watanabe et al., 1997). However, the or- due to the material’s very high dispersion.
ange phosphorescence demonstrated a new stimulation Wakefieldite is a rare tetragonal mineral belonging to
source. Its combination with the greenish blue phospho- the xenotime mineral family with an ideal general formula
rescence and the identified visible-light excitation wave- of (REE)VO4, where REE = Y, La, Ce, and Nd among other

156 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023

 Figure 89. Faceted and
rough synthetic wake-
fieldite in LED lighting
with 6400K color tem-
perature. The faceted
stone is 5.04 ct. Photo by
Adriana Gudino; cour-
tesy of Arya Akhavan.

rare earth elements. Its general physical and optical prop- examples of faceted wakefieldite. In fact, this was the first
erties include a hardness of 5, specific gravity of 4.25, uni- faceted synthetic wakefieldite seen in a GIA laboratory.
axial + optic sign with refractive indices of nω = 2.000 and Chemistry and visible absorption spectra were collected to
nε = 2.140, birefringence of 0.140, and dispersion of 0.084. quantitatively study the striking color-change behavior.
With prominent, perfect cleavage and relatively low hard- A polished wafer with 10.6 mm thickness was analyzed.
ness, the material is difficult to polish. Natural wakefield- Laser ablation–inductively coupled plasma–mass spectrom-
ite occurs as very small, non-gem crystals unsuitable for etry revealed the material was composed of mostly yttrium
faceting; Czochralski-pulled materials represent the only and vanadium, with 2.2% neodymium by weight (0.03 per

Figure 90. The color

change of the faceted
synthetic wakefieldite,
observed between day-
light (left) and cool flu-
orescent light F10
(right). Photos by
Aaron Palke; courtesy
of Arya Akhavan.

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023 157

 Nd3+ VISIBLE SPECTRA IN YVO4

4.0 O-ray 100

RELATIVE SPECTRAL POWER DISTRIBUTION

E-ray
F10 90
3.5
D65
Eiko 2800K 80
3.0 610
545
70
ABSORBANCE

2.5
610 60

2.0 50
541
40
1.5
435
30
1.0
20
0.5
10

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

Figure 91. The visible absorption spectra of neodymium-doped YVO4 are shown with the relative power distribution
curves of CIE illuminants F10 and D65 and an Eiko warm white fluorescent bulb with color temperature of 2800K.

formula unit). Since Y3+ does not produce color on its own, the color was determined by the transmission windows in
the color must be derived from the addition of Nd3+ ions sub- the spectrum (440–460, 490–500, 550–570, and 620–670
stituting for Y3+ ions in the crystal lattice. Figure 91 shows nm), which remain constant when switching between these
polarized absorption spectra in the visible range collected illuminants. This was confirmed by observation of the
with an ultraviolet/visible/near-infrared spectrometer for stone in outdoor conditions, which produced the same
both the ordinary ray (o-ray) and extraordinary ray (e-ray). The pinkish purple color seen in incandescent light. The use of
expected color was calculated from the absorption spectra warm fluorescent light (Eiko fluorescent light in figures 91
based on the formula described by Z. Sun et al. (“Quantitative and 92) with color temperature of 2800K also reproduced
definition of strength of chromophores in gemstones and the the greenish blue to blue color seen in cool fluorescent light
impact on color change in pyralspite garnets,” Color Research (F10). It is also worth noting that the absorption features of
and Application, Vol. 47, No. 5, 2022, pp. 1134–1154). Nd3+ ions at 541 and 610 nm were strongly dependent on
As shown in figure 92, the material exhibited a general the polarization of the light, resulting in a strong pleochro-
pink to purple to violet color in most illuminants except ism (grayish purple for the o-ray and green-blue for the e-
the Eiko warm fluorescent light and the o-ray of CIE stan- ray), only observed under F10 illuminant (figure 92).
dard illuminant F10, which represents a typical cool fluo- The color panels under two LED illuminants in GIA’s
rescent light. In fact, the color-change behavior noticed by standard viewing box were also calculated in figure 92. The
author AA could be replicated when comparing illuminant two LED lights, warm and cool, closely imitate the CIE
F10 to any of the other illuminants such as daylight or in- standard illuminants A and D65. The two sets of color pan-
candescent light. Switching between warm incandescent els (A = LED warm pair and D65 = LED cool pair) are ex-
light and actual cool daylight did not produce a dramatic tremely close to each other based on their L, C, and H
change of color. The emission spectrum of F10 fluorescent (lightness, chroma, and hue) color coordinates. This study
also illustrates the consistent results of GIA’s standard
light is mainly composed of sharp peaks at 435, 545, and
viewing box in grading gemstone color.
610 nm (figure 91). Therefore, the color of the synthetic
wakefieldite crystal under illuminant F10 is only deter- Ziyin Sun, Shiyun Jin, Adriana Gudino, and
mined by the small absorption features at ~435, 541, and Aaron C. Palke
610 nm of the Nd3+ ions. Broadband emitters such as incan- Carlsbad, California
descent light and actual daylight (D65 in figure 91), on the Arya A. Akhavan
other hand, do not have these sharp emission features, so Surgical Precision Gems, Toronto

158 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023

 Eiko fluorescent Warm LED Cool LED
Type of illuminant A D65 F10
light 2800K (2700K) (6700K)

E ll c

L 60.95 58.78 57.88 65.36 61.11 58.90

Color
C 32.98 38.38 23.92 18.17 33.40 41.35
coordinates
H 335.97 327.63 285.84 325.25 339.28 329.74

Eiko fluorescent Warm LED Cool LED

Type of illuminant A D65 F10
light 2800K (2700K) (6700K)

E﬩c

L 61.72 61.29 58.90 66.00 61.53 61.30

Color
C 26.68 30.16 27.83 22.54 26.58 32.28
coordinates
H 309.37 308.09 254.63 210.44 311.25 311.94

Figure 92. Color panels of the 10.59 mm wafer in six different lighting conditions were calculated for both the o-
ray and e-ray. In the top row, A and D65 are CIE standard illuminants for incandescent light and daylight. Eiko is
a fluorescent bulb with color temperature of 2800K, while F10 is a CIE standard narrowband fluorescent illumi-
nant. Warm (2700K) and cool (6700K) LEDs represent the illuminants in the GIA standard viewing box.

TREATMENTS 514 nm diode-pumped solid-state (DPSS) laser, with about

10 mW laser power on the sample, 50× long working dis-
Heated purplish pink sapphire from Ilakaka (Madagascar) tance objective lens, confocal mode (100 microns entrance
with colored monazite inclusions. The Laboratoire slit), 1800 lines/mm grating, and about 1.5 cm–1 spectral res-
Français de Gemmologie (LFG) received a 3.02 ct purplish
pink sapphire (figure 93) for identification. Under the mi- Figure 93. A 3.02 ct heated purplish pink sapphire
croscope, zircon and monazite inclusions (figure 94) along measuring 9.57 × 7.74 × 4.55 mm. Photo by Ugo
with tubes (figure 95) were observed. This inclusion scene Hennebois.
was characteristic of similarly colored sapphires from
Ilakaka, Madagascar (W. Wang et al., “The effects of heat
treatment on zircon inclusions in Madagascar sapphires,”
Summer 2006 G&G, pp. 134–150; S. Saeseaw et al., “Low-
temperature heat treatment of pink sapphires from Ilakaka,
Madagascar,” Winter 2020 G&G, pp. 448–457). No indica-
tion of heating was observed. Moreover, the monazite in-
clusions were orange-brown. It was previously observed
that monazite inclusions in pink sapphires turn colorless
when heated above 600°C (see again Saeseaw et al., 2020).
Using Raman spectroscopy, the full width at half maxi-
mum (FWHM) of the band around 1010 cm–1 due to anti-
symmetric stretching vibration SiO4 in the zircon structure
and the band around 975 cm–1 due to PO4 stretching vibra-
tion in the monazite structure can decrease during heating
of pink sapphires from Ilakaka (Wang et al., 2006; Saeseaw
et al., 2020). The zircon and monazite inclusions were ana-
lyzed using a Raman Renishaw inVia spectrometer, with a

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023 159

 main band from 7.1 to 21.7 cm–1 and monazite inclusions
with FWHM of the main band from 14.2 to 18.4 cm–1 (Sae-
seaw et al., 2020; M. Krzemnicki et al., “Zircon inclusions
in unheated pink sapphires from Ilakaka, Madagascar: A
Raman spectroscopic study,” Proceedings of the Online In-
Monazite ternational Gemmological Conference, 2021, pp. 21–23; S.
Karampelas et al., “Détection du traitement thermique à
basse température des corindons,” Revue de Gemmologie
a.f.g., No. 217, 2022, pp. 4–5).
Unpolarized Fourier-transform infrared spectra in dif-
ferent directions were collected using a Thermo Fisher
Magna-IR560 with 4 cm–1 spectral resolution and 500 scans
Zircon
using a diffuse reflectance accessory as beam condenser.
The sample presented a band at 3309 cm–1 and a less intense
band at 3232 cm–1, which in pink sapphire is considered an
indication of heating (Saeseaw et al., 2020). These bands are
Figure 94. A series of colorless zircon inclusions and linked to different forms of titanium associated with hy-
orange-brown monazite inclusions in the heated pur- droxyl defects in corundum (E. Balan, “Theoretical infrared
plish pink sapphire. Photomicrograph by Ugo Hen- spectra of OH defects in corundum (α-Al2O3),” European
nebois; field of view 1.5 mm. Journal of Mineralogy, Vol. 32, No. 5, 2020, pp. 457–467).
In line with previous experiments showing that mon-
azite inclusions can keep their color after heating to 600°C
(Saeseaw et al., 2020), the presence of colored monazite in-
olution. The FWHM of the most intense band of seven mon- clusions alone cannot serve as evidence that a pink sapphire
azite inclusions at around 975 cm–1 ranged from 12.7 to 14.2 from Ilakaka is unheated. On the other hand, the presence
cm–1, and those of 10 zircon inclusions at around 1010 cm–1 of colorless monazite inclusions in pink sapphires from
ranged from 6.4 to 7.8 cm–1. Both appeared to be sharper and Ilakaka could be considered an indication of heat treatment.
less variable compared to FWHM in unheated pink sap-
phires, which present zircon inclusions with FWHM of the Ugo Hennebois, Aurélien Delaunay, and
Stefanos Karampelas ([Link]@[Link])
LFG, Paris

Figure 95. Unaltered tubes in the heated purplish Chromophore behaviors (including the 880 nm absorption
pink sapphire. Photomicrograph by Ugo Hennebois; band) in an irradiated pink sapphire. Irradiation is a known
field of view 2 mm. process for enhancing color in corundum, particularly yel-
low and pink sapphire. Recently, some rubies and pink sap-
phires with bluish tint were reportedly treated with
radiation (“LMHC makes progress on laboratory report har-
monisation, discussed current challenges in detection of
corundum treatments,” LMHC press release, March 14,
2023). However, chromophore behaviors in irradiated
ruby/pink sapphire have not been fully investigated. This
report shows preliminary results of an experiment involving
irradiated pink sapphire. An untreated Madagascar pink sap-
phire with a blue modifier (figure 96) was fabricated as an
optical wafer with two polished faces perpendicular to the
c-axis for ultraviolet/visible/near-infrared (UV-Vis-NIR)
spectroscopic study. Trace element chemistry was analyzed
in the UV-Vis-NIR measurement area using laser
ablation−inductively coupled plasma−mass spectrometry.
This information can be used to link the visual color and
the color-causing trace elements listed in figure 96.
The spectrum of the untreated sample displayed broad
absorption bands at ~400 and ~560 nm together with weak
peaks at ~470 and at 693 nm, demonstrating that the Cr3+
chromophore was responsible for the pink color (blue line
in figure 97; see E.V. Dubinsky et al., “A quantitative de-

160 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023

 Before irradiation After 500 kGy gamma radiation After fade test (6 hrs)

Elements Mg Ti V Cr Fe Ga

Concentration
30 ± 7a 21 ± 4 16 ± 1 137 ± 7 480 ± 27 15 ± 1
(ppma)

Detection limits
0.05 0.14 0.01 0.17 2.2 0.01
(ppma)
a
average ± standard deviation

Figure 96. Color-calibrated photo of a 0.664 ct pink sapphire from Madagascar before and after irradiation, and
after fade test using fiber-optic light for 6 hours. Path length/thickness: 1.514 mm. Photos by Sasithorn Engniwat.

scription of the causes of color in corundum,” Spring 2020 For our radiation experiment, the sample was treated
G&G, pp. 2–28). This sample also showed a weak broad with a gamma irradiation dose of 500 kGy at the Irradia-
band at around 580 nm related to Fe2+-Ti4+ pairs, which pro- tion Center at the Thailand Institute of Nuclear Technol-
duced a blue color, modifying the pink color from Cr3+ to ogy (public organization). After irradiation, the pink
purple-pink. Weak Fe3+-related absorption features were sapphire changed to a padparadscha-like coloration in
also observed, represented by a broad band at ~330 nm and which the subtle blue color was partially removed (figure
narrow peaks at 377, 388, and 450 nm. The iron-related 96). The UV-Vis-NIR spectrum after irradiation showed a
chromophore only causes significant yellow color at rela- significant increase in absorption intensities at wave-
tively high iron concentrations (greater than ~1000 ppma) lengths shorter than ~560 nm or longer than ~660 nm (red
due to Fe3+ chromophores’ weak color strength. line in figure 97). To observe changes in chromophore be-

UV-VIS-NIR SPECTRA
8.5
Before irradiation
7.5 Figure 97. Polarized
ABSORPTION COEFFICIENT (cm–1)

After irradiation
Difference spectrum UV-Vis-NIR spectra
6.5
comparisons of the pink
5.5 Cr 3+ sapphire before and
Cr 3+
410
560 after irradiation with a
4.5 gamma irradiation dose
of 500 kGy. The differ-
3.5
ence spectrum revealed
2.5 a significant increase in
Cr 3+
Trapped hole-Cr 3+ 693 880 trapped hole paired
1.5 ~470
with Cr3+ and the 880
Fe 2+-Ti4+ nm band in the irradi-
0.5 580
0 ated pink sapphire.
-0.5
300 400 500 600 700 800 900 1000 1100
WAVELENGTH (nm)

GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023 161

 UV-VIS-NIR SPECTRA
Figure 98. Polarized
8 After irradiation UV-Vis-NIR spectra
ABSORPTION COEFFICIENT (cm–1)

After fade test comparisons of the pink

7 Difference spectrum
sapphire after irradia-
6 tion with a gamma irra-
diation dose of 500 kGy
5 and after the fade test.
The difference spectrum
4
between after irradia-
3 tion and after fade test
obviously shows the re-
2 duction in the trapped
hole paired with Cr3+
1 chromophore when ex-
posed to intense light
0
300 400 500 600 700 800 900 1000 1100 for several hours.
WAVELENGTH (nm)

havior caused by irradiation, the spectrum after irradiation that has been heat treated, such as heated Madagascar sap-
was subtracted from the pretreatment spectrum. The dif- phire (E.B. Hughes and R. Perkins, “Madagascar sapphire:
ference spectrum (green line in figure 97) revealed that this Low-temperature heat treatment experiments,” Summer
acceptor-dominated corundum with an atomic concentra- 2019 G&G, pp. 184–197) or heated Rock Creek sapphire
tion of magnesium greater than the sum of titanium and (J.L. Emmett and T.R. Douthit, “Heat treating the sap-
silicon concentrations created a strong absorption feature phires of Rock Creek, Montana,” Winter 1993 G&G, pp.
below ~560 nm to the UV region, which matches well 250–272). This feature can also be found in Mozambican
with the trapped hole associated with the Cr3+ spectrum ruby after heat treatment (S. Saeseaw et al., “Update on
for causing orange coloration (Dubinsky et al., 2020). “low-temperature” heat treatment of Mozambican ruby:
Using the absorption cross section information provided A focus on inclusions and FTIR spectroscopy,” GIA Re-
in Dubinsky et al. (2020), the concentration of trapped search News, April 30, 2018). The origin of this band is not
hole-Cr3+ chromophores induced by irradiation was ap- fully understood but may be related to iron clusters.
proximately 0.8 ppma in the sample. This chromophore The 880 nm band was absent in the UV-Vis-NIR spec-
had very high color strength, and therefore a noticeable trum of this sample before treatment. Interestingly, this
color change could be observed with only a tiny concen- pink sapphire developed the 880 nm band after irradiation,
tration of the trapped hole paired with Cr3+ created after with a significant intensity around 1.2 cm–1. Color stability
treatment. The Fe2+-Ti4+ chromophore also reduced by ap- testing is usually performed in gem laboratories for yellow
proximately 0.7 ppma in the treated sample, resulting in to padparadscha sapphires that may contain unstable color
less blue color modification. This slight lightening in blue centers (e.g., A.C. Palke et al., “An update on sapphires
color after irradiation is possibly due to the change in ox- with unstable color,” GIA Research News, December 12,
idation state of iron (which would reduce the number of 2022; Summer 2022 GNI, pp. 259–260). After the color sta-
Fe2+-Ti4+ pairs). In addition, the creation of orange color bility test (figures 96 and 98), the orange coloration intro-
(trapped hole-Cr3+) masked a certain portion of remaining duced by irradiation faded away after exposure to intense
blue color after treatment, as blue and orange are comple- incandescent light due to the dissociation of the trapped
mentary colors. The change in UV-Vis-NIR spectrum cor- hole paired with Cr3+, whereas the 880 nm band remained.
responded with the change in color appearance caused by This preliminary observation suggests that the 880 nm
irradiation. band can be produced by an irradiation process, and this
When studying absorption features in the UV-Vis-NIR feature could provide some interesting evidence of treat-
spectrum of corundum, a broad band centered at around ment, either heat or irradiation, in pink sapphires and ru-
880 nm in the red to near-infrared region is typically asso- bies. However, further studies would be needed to
ciated with basalt-related blue sapphire (e.g., A.C. Palke et understand the development of the 880 nm absorption
al., “Geographic origin determination of blue sapphire,” band.
Winter 2019 G&G, pp. 536–579). This band has little or no Wasura Soonthorntantikul and Wim Vertriest
impact on color. In addition to basalt-related blue sapphires GIA, Bangkok
that are exposed to heat naturally, the 880 nm band can be Aaron Palke
induced in the spectrum of a metamorphic-type sapphire GIA, Carlsbad

162 GEM NEWS INTERNATIONAL GEMS & GEMOLOGY SPRING 2023