Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

SU05

Download as pdf or txt
Download as pdf or txt
You are on page 1of 114

VOLUME 41, NO.

Summer 2005

EDITORIAL _____________
87

The Experts behind G&G: Our Editorial Review Board

Alice S. Keller

FEATURE ARTICLES _____________


88

Characterization and Grading of Natural-Color Yellow Diamonds

John M. King, James E. Shigley, Thomas H. Gelb, Scott S. Guhin,


Matthew Hall, and Wuyi Wang
Using a sample base of more than 24,000, this article explores the color
grading of natural-color yellow diamonds at the GIA Gem Laboratory and
reports on their gemological and spectroscopic properties.

pg. 89

116

Emeralds from the Kafubu Area, Zambia

J. C. (Hanco) Zwaan, Antonn V. Seifert, Stanislav Vrna, Brendan M. Laurs,


Bjrn Anckar, William B. (Skip) Simmons, Alexander U. Falster, Wim J. Lustenhouwer,
Sam Muhlmeister, John I. Koivula, and Hja Garcia-Guillerminet
The geology, production, and gemological properties of emeralds from this,
the second largest emerald-producing country in the world, are described.

150

Mt. Mica: A Renaissance in Maines Gem Tourmaline Production

William B. (Skip) Simmons, Brendan M. Laurs, Alexander U. Falster,


John I. Koivula, and Karen L. Webber
This historic deposit is once again producing fine tourmalines, with both the
mining area and the gems examined in this report.
pg. 146

REGULAR FEATURES _____________________


164

Lab Notes
Two unusual fracture-filled diamonds Large diamond with micro-inclusions of
carbonates and solid CO2 Light blue diamond with type IIb and IIa zones
Natural type Ib diamond with unusually high N content Diamond with unusual
laser drill holes Yellowish orange magnesioaxinite Cultured pearl with cultured-pearl nucleus Dyed golden freshwater cultured pearls More on Cubearing color-change tourmaline from Mozambique

176

Gem News International


Demantoid from northern Pakistan Pyrope-almandine from Ethiopia New
ruby and pink sapphire deposit in Kenya Faceted orange and brown sunstone
from India Staurolite in kyanite from Brazil Obsidian with spessartine inclusions Fluorite inclusions in quartz from Madagascar Bismuthinite inclusions
in rose quartz from Madagascar Yellow hydrothermal synthetic sapphires seen
in India Interesting heated citrine Chocolate Tahitian cultured pearls Lead
glassfilled rubies appear in the Middle East GemmoBasel 2005 report

189

Book Reviews

192

Gemological Abstracts

200

The Last Page: The Other Side of the Lens

pg. 172

pg. 183

The Experts behind Gems & Gemology:

Our Editorial Review Board

hat sets a peer-reviewed journal apart from other publications is its editorial review board. And what makes
Gems & Gemology unique as a professional journal is the intellect that our reviewers bring to the task.
Our review board members, who represent many different aspects of gemological research, independently
evaluate submitted manuscripts for accuracy, research methodology, and usefulness to the journals readership. Every
manuscript that makes it to the pages of G&G has been improved by their comments and suggestions.
The members of this panel are dedicated to the advancement of gemology, working long hours with no remuneration. It
is with great pride and gratitude that I take this opportunity to profile the Gems & Gemology Editorial Review Board.
Shigeru Akamatsu is a leading authority on cultured
pearls. A former manager of the Pearl Research Laboratory at
K. Mikimoto & Co. and currently senior manager of its business administration department, Mr. Akamatsu is also vicepresident of the CIBJO Pearl Commission.
Alan T. Collins is an expert in the study of color centers
and point defects in natural, synthetic, and treated diamonds.
A widely published author and frequent lecturer, Dr. Collins
is professor of physics at Kings College London.
G. Robert Crowningshield, a former vice president of the
GIA Gem Trade Laboratory, is a legend in gemology (see
profile in the Fall 2003 G&G, pp. 184199). Mr. Crowningshield began writing for G&G in 1949 and was the longtime
editor of the Lab Notes section.
John L. Emmett is a leading authority on laser applications
and heat treatment of gem materials. Formerly associate
director of the Lawrence Livermore National Laboratory, Dr.
Emmett is co-founder and president of Crystal Chemistry, a
company specializing in the heat treatment of corundum.
Emmanuel Fritsch has written numerous research articles,
most related to the application of spectroscopy to gemology,
the origin of color in gem materials, and treated and synthetic gems. Formerly with GIA Research, Dr. Fritsch is professor
of physics at the University of Nantes, France.
Henry A. Hnni is a prolific author and lecturer. His
research interests include advanced analytical techniques,
pearl formation, and corundum treatments. Dr. Hnni is
director of the SSEF Swiss Gemmological Institute in Basel
and professor of gemology at the University of Basel.
C. S. Hurlbut Jr. is professor emeritus of mineralogy at
Harvard University. A 1955 Guggenheim fellow, he has written several books on gemology and mineralogy. Dr. Hurlbut
is a past president of the Mineralogical Society of America.
A. J. A. (Bram) Janse, a well-known expert on diamond and
kimberlite occurrences, has been involved in diamond exploration worldwide for nearly 50 years. Dr. Janse is managing
director of Archon Exploration Pty Ltd. in Perth, Australia.
Alan Jobbins has more than 50 years of experience as a
gemologist. A former editor of Journal of Gemmology, he was
the longtime curator of gems and minerals at the Geological
Museum in London. Mr. Jobbins is president of Gem-A, London.
Mary L. Johnson is a frequent contributor to G&G. As manager of research and development at GIA Research in Carlsbad,
Dr. Johnson has helped lead the Institutes groundbreaking
studies on emerald fillers and diamond cut proportions.

EDITORIAL

Anthony R. Kampf conducts research in the areas of descriptive mineralogy, crystal chemistry, and structural crystallography. Dr. Kampf is curator and section head of mineral sciences
at the Natural History Museum of Los Angeles County.
Robert E. Kane is a former manager of identification at the
GIA Gem Trade Laboratory and former director of the
Gbelin Gem Lab in Lucerne, Switzerland. He continues to
write on various gemological topics as president of Fine
Gems International in Helena, Montana.
Alfred A. Levinson is an authority on diamond deposits. A
former editor of Geochimica et Cosmochimica Acta, Dr. Levinson
is professor emeritus of geology at the University of Calgary.
Thomas M. Moses has written a number of important
research articles dealing with gem identification and the
characteristics of natural and synthetic diamonds. Mr. Moses
is vice president of identification and research services at the
GIA Gem Laboratory in New York.
George R. Rossman is professor of mineralogy at the
California Institute of Technology, where he received the
20032004 Richard P. Feynman Prize for Excellence in
Teaching. Among Dr. Rossmans research interests is the study
of how electromagnetic radiation interacts with minerals.
Kenneth Scarratt has more than 30 years of laboratory
experience and is one of the worlds leading authorities on
natural and cultured pearls. Formerly head of the AGTA
Gemological Testing Center, Mr. Scarratt is director of the
new GIA research facility in Bangkok.
James E. Shigley, an important contributor to G&G and
many other journals, is also the editor of the new volume
Gems & Gemology in Review: Synthetic Diamonds. Dr.
Shigley has been director of GIA Research since 1986.
Christopher P. Smith is a widely published gemologist who
has performed research on HPHT-treated diamonds, rubies
and sapphires, and gem treatments. Formerly head of the
Gbelin Gem Lab, Mr. Smith is director of identification services at the GIA Gem Laboratory in New York.

G&G appreciates the selfless contributions of our reviewers. These are truly the experts behind Gems & Gemology.

Alice S. Keller
Editor-in-Chief

GEMS & GEMOLOGY

SUMMER 2005

87

CHARACTERIZATION AND GRADING OF

NATURAL-COLOR YELLOW DIAMONDS


John M. King, James E. Shigley, Thomas H. Gelb, Scott S. Guhin, Matthew Hall, and Wuyi Wang

To better understand the yellow diamonds currently in the marketplace, as well as identify possible
changes in their trends seen over a five-year period, researchers at the GIA Gem Laboratory analyzed gemological data collected on more than 24,000 natural-color yellow diamonds examined
in the calendar years 1998 and 2003. These data included color grade, type of cut, clarity grade,
weight, ultraviolet fluorescence, and UV-visible and infrared spectra. Among natural-color colored
diamonds, those with a yellow hue are some of the most abundant; even so, they are much less
common than the colorless to light yellow diamonds associated with GIAs D-to-Z color grading
scale. Since the yellow color is a continuation of the gradation of color associated with the D-to-Z
scale, there can be misconceptions about the color grading, which involves different procedures
from those used for D-to-Z grading. The grading and appearance aspects, as well as other characteristics of yellow diamonds, are discussed to clarify these differences. The authors have also identified five subgroups of type I yellow diamonds, which (with some overlap) are characterized by representative spectra and color appearances.

oday, yellow diamonds are among the most


widely encountered of the fancy color diamonds (figure 1). From 1998 to the present,
GIA has issued grading reports on more than
100,000 yellow diamonds, by far the most common
of the fancy-color diamonds submitted to our laboratory. In 2003, for example, 58% of the diamonds
submitted for GIA Colored Diamond Grading
Reports or Colored Diamond Identification and
Origin of Color Reports were in the yellow hue.
Nevertheless, these represented only 2.4% of all
diamonds submitted for various grading reports that
year. The dichotomy between being common
among colored diamonds yet relatively rare overall
has created a mix of information and sometimes
erroneous assumptions about yellow diamonds. In
addition, although information about their origin
of color or unusual characteristics (see below) has
been documented over the years, little has been
published about their color appearance and its relationship to color grading (one exception being
Hofer, 1998).

88

YELLOW DIAMONDS

This article presents data on more than 24,000


fancy-color yellow diamonds that were examined
by the GIA Gem Laboratory during the years 1998
and 2003. While our main concern in this study was
gathering information to characterize the gemological and spectroscopic properties of the entire population of samples, we were also interested in identifying any trends in size, color grade, or clarity grade
among the yellow diamonds submitted to us that
might be revealed over time. To that end, we selected a sample population from these two years separated by a five-year span.
Following a brief review of the literature on the
geographic sources, cause of color, and other physical properties of fancy-color yellow diamonds, this
article will focus on expanding the published information about these diamonds by documenting and

See end of article for About the Authors and Acknowledgments.


GEMS & GEMOLOGY, Vol. 41, No. 2, pp. 88115.
2005 Gemological Institute of America

GEMS & GEMOLOGY

SUMMER 2005

Figure 1. While abundant


compared to other colored
diamonds, fancy-color yellow
diamonds represent a small
portion of overall diamond
production. Their beauty and
the depths of color in which
they occur offer a wide range
of possibilities to the jeweler.
The bracelet and the 4.13 ct
oval and 6.20 ct emerald cut
in the rings are courtesy of
Harry Winston Inc.; the 8.70
ct (total weight) StarBurst cuts
in the earrings are courtesy of
Jonathan Doppelt Inc.; the yellow gradation eternity band is
courtesy of N. Smid; and the
two unmounted diamonds
and the crystal are courtesy of
the Scarselli family. Photo by
Harold & Erica Van Pelt.

reporting on their range of color, color grading, clarity grading, and other gemological properties, as
well as their spectroscopic characteristics.
We will also look at a number of aspects of color
observation and appearance (as they apply to these
diamonds) that are known and commonly addressed
by vision scientists but may not be recognized by
the layperson or even the experienced diamond
dealer or retailer. As with the previous GIA studies
of blue and pink diamonds (King et al., 1998, 2002),
this article describes and illustrates some of these
aspects to aid in understanding how they apply to
color grading yellow diamonds.

BACKGROUND
History and Geographic Origin. Yellow diamonds
have long been recognized and prized among collectors (Mawe, 1813; Bauer, 1904; Copeland and
Martin, 1974; Gleason, 1985). For example, in his
description of several famous diamonds he encountered while in India, the French traveler and gem
dealer Jean-Baptiste Tavernier (1676) mentioned
seeing a 137.27 ct yellow diamond that he referred
to variously as the Florentine, the Austrian
Yellow, and the Grand Duke of Tuscany. As
with other colors, yellow diamonds have come to
the publics attention through the interest generated by a number of special stones, such as the historical 128.54 ct Tiffany (figure 2; see Balfour, 2000)

YELLOW DIAMONDS

and 101.29 ct Allnat (see King and Shigley, 2003)


diamonds, or the extraordinary 407.48 ct
Incomparable (illustrated in King et al., 1994, p.
227). Indeed, the first reportedly authenticated diamond found in Africa, which was cut into the
10.73 ct Eureka, is a distinct yellow (Balfour, 2000;
shown in Janse, 1995, p. 231). Its discovery in late
1866/early 1867 helped start the ensuing African
diamond mining rush.
Although there were occasional finds over the
centuries in India, Brazil (Cassedanne, 1989), and
perhaps elsewhere, the first major discovery of
quantities of what today would be considered fancycolor yellow diamonds occurred at several locations
in South Africa in the late 1860s (Janse, 1995; figure
3). The early preponderance of light yellow diamonds from the Cape Province in South Africa led
to their description in the jewelry trade as cape
stones, a usage that continues today (see, e.g., GIA
Diamond Dictionary, 1993). While these light yellow diamonds are not now considered colored diamonds (rather, they fall toward the lower end of
GIAs D-to-Z color scale), their noticeable color distinguished them from those of other deposits that
yielded near-colorless or brown diamonds. Also
found, however, were numerous diamonds that ultimately would be considered yellow. The earliest
recorded find along the banks of the Orange River
was, as noted above, a 21.25 ct (old carat weight)
yellow crystal that was subsequently cut into the

GEMS & GEMOLOGY

SUMMER 2005

89

Noted Auction Sales and Other Publicity. As with


other colored diamonds, yellow diamonds can command high prices and generate much public attention. At its 1988 auction, the Fancy brownish yellow
Internally Flawless Incomparable diamond reached a
bid of $12 million before it was withdrawn without
having met its reserve (Balfour, 1997). An 18.49 ct
Fancy Intense yellow known as the Golden Drop
sold at auction for $203,461 per carat in 1990 (Hofer,
1998), while a 13.83 ct Fancy Vivid yellow sold for
$238,792 per carat in 1997. In February 2005, a 10.02
ct Fancy Vivid yellow diamond sold for $772,848
(more than $77,131 per carat) at Sothebys St. Moritz.
Recently, a 101.28 ct Fancy Vivid yellow cushion
modified brilliant known as the Golden Star was
unveiled to the public in Palm Beach, Florida, by jeweler Laurence Graff, who reported that this diamond
had been cut from a large crystal found at the Finsch
mine in South Africa.

Figure 2. The Tiffany yellow diamond is one of the


best-known diamonds in the world. The 128.54 ct diamond has been in the Tiffany collection since 1879,
and today it is displayed in its own case (as part of
Jean Schlumbergers jewel, Bird on a Rock) at
Tiffany & Co. in New York City. It was graded by
GIA in 1984. Photo courtesy of Tiffany & Co.

Eureka (Janse, 1995). In 1878, a 287.42 ct piece of


rough recovered from a claim on the Kimberley
mine was later fashioned into the Tiffany diamond
(again, see figure 2). In 1964, the Dutoitspan mine
yielded a spectacular 253.70 ct transparent yellow
octahedral crystal known as the Oppenheimer diamond, which now resides in the Smithsonian
Institution.
Today, yellow diamonds are found in the productions of mining operations throughout the
world (Field, 1992, p. 353). The occurrence of fancycolor yellow diamonds is so widespread that no
particular deposit stands out as an important
source (though many of the larger pieces of yellow
rough continue to originate from South Africa).
Some data correlating diamond color and size has
been published for certain deposits (in particular,
the kimberlites in South Africa; see, e.g., Harris et
al., 1979), but the literature contains no such information on yellow diamond abundances for other
occurrences.

90

YELLOW DIAMONDS

Past Studies. Differences noted in crystal form and


physical and spectroscopic properties between (what
were discovered to be) nitrogen-containing diamonds
and those much rarer diamonds with virtually no
nitrogen led early in the 20th century to the recognition by scientists of the type I and type II categories,
respectively (Robertson et al., 1934; also see Anderson,
1963). At the atomic level in type I diamonds, nitrogen
substitutes for carbon as either multiple atoms (aggregates) that occupy adjacent positions in the lattice
(type Ia) or as single isolated atoms (type Ib).
There have been several recent reviews of the
physical properties of diamond, including Field
(1992), Davies (1994), and Wilks and Wilks (1994). A
number of articles have discussed the visible and
infrared spectra of yellow diamonds as they relate to
the discrimination of natural-color diamonds from
those that have been treated to change their color
(Clark et al., 1956a,b; Crowningshield, 19578;
Scarratt, 1979, 1982; Collins, 1978, 1982a,b, 2001;
Woods and Collins, 1982, 1986; Woods, 1984;
Collins et al., 1986; Mita, 1996; Fritsch, 1998; Haske,
2000; Kaminsky and Khachatryan, 2001; De Weerdt
and Van Royen, 2001; Zaitsev, 2001). Additional
articles have discussed the potential use of spectral
measurements to quantitatively evaluate the color
of faceted yellow diamonds (Vendrell-Saz et al.,
1980; Sato and Sunagawa, 1982; Collins, 1984).

* Graded prior to modifications to GIAs colored diamond color grading system in 1995.

GEMS & GEMOLOGY

SUMMER 2005

Gemologists have also recognized that certain


inclusions are characteristic of particular gemstones and, as such, help provide information on
their geologic history and diamond type. For example, Crowningshield (1959) reported that needlelike inclusions, as well as oriented plate-like inclusions, are found on occasion in some yellow diamonds. Such inclusions were later associated with
type Ib yellow diamonds (Crowningshield, 1994).
Color and Color Origin. In general, the presence of
nitrogen atoms gives rise to two kinds of absorption
in the blue region of the visible spectrum; the
remainder of the spectrum is transmitted, leading to
an observed yellow color. In type Ia diamond, aggregates of nitrogen atoms in the form of the N3 center
cause absorption as a sharp line at 415 nm. The N3
center is believed to consist of three nearest-neighbor substitutional nitrogen atoms all bonded around
a common lattice site of a missing carbon atom (a
vacancy). This absorption produces the lighter yellow coloration typical of cape diamonds.
The same N3 center is also responsible for blue
long-wave ultraviolet fluorescence when seen in type
Ia diamonds (Nayar, 1941; Mani, 1944; Dyer and
Matthews, 1957). Based largely on observations with
the prism spectroscope, Anderson (1943a,b, 1962)
noted the general correlation between the intensity of
the 415 nm absorption band and the strength of the
diamonds yellow color (also see Anderson and Payne,
1956). Occasionally, the N3 center is accompanied by
additional broader but weaker lines at 452, 465, and
478 nm (the N2 center; see Clark et al., 1956a; Davies
et al., 1978; Davies, 1981), which also contribute to
the yellow color. These lines are superimposed over a
region of absorption that increases gradually toward
the blue end of the spectrum.
Collins (1980) gave nitrogen concentration values
of up to 3000 ppm for type Ia diamonds. Two additional nitrogen aggregatesdesignated A (a pair of
nearest-neighbor nitrogen atoms) and B (four N
atoms surrounding a vacancy)are usually present
in type Ia diamonds, but neither gives rise to absorption in the visible region (thus, they do not produce
any coloration, but they do produce characteristic
absorption features in the infrared spectrum).
In the much-less-common type Ib diamond, single-substitutional nitrogen atoms cause a broad region
of absorption that increases below about 560 nm (but
no sharp 415 nm absorption band), which in turn produces a more saturated yellow color. In some cases,
type Ib diamonds exhibit a very weak to weak yellow

YELLOW DIAMONDS

or orange short-wave UV fluorescence (see Dyer et al.,


1965; Fritsch, 1998, pp. 2630 and 3637; Hofer, 1998,
pp. 354359). Such diamonds have often been
described as having a canary color (Wade, 1920;
Anderson, 1943a; Liddicoat, 1976). In contrast to type
Ia yellow diamonds, those that are type Ib contain
much smaller amounts of nitrogen, normally well
under 100 ppm (Collins, 1980; Field, 1992).
These subcategories of type I diamonds can be
distinguished on the basis of characteristic features
in their visible and infrared absorption spectra. It
should be mentioned that these subcategories are
not mutually exclusive, since at the atomic level
the diamond lattice can contain different ratios of
aggregated and isolated nitrogen atoms from one
diamond to the next. In addition, infrared spectral
features associated with both isolated and aggregated forms of nitrogen can be recorded from different
portions within the same diamond. For example,

Figure 3. The late 1860s saw the first major discovery of diamonds in South Africa, and with that
came an influx of yellow rough into the market.
The Kimberley and Dutoitspan mines, between the
Orange and Vaal Rivers, were some of the significant early (and continuous) sources of production.
In 1964, the latter mine yielded the spectacular
253.70 ct yellow octahedral crystal known as the
Oppenheimer diamond. Modified from Janse (1995).

GEMS & GEMOLOGY

SUMMER 2005

91

Collins (1980) suggested that many type Ib diamonds also contain A-aggregates (type IaA) of nitrogen. For additional information on diamond types
and optical centers, see Davies (1972); Davies and
Summersgill (1973); Collins (1982a,b, 2001); Bursill
and Glaisher (1985); Fritsch and Scarratt (1992, pp.
3839); Briddon and Jones (1993); Wilks and Wilks
(1994); and De Weerdt and Van Royen (2001).
GIA Color Descriptions for Yellow Diamonds. A
hue has different color appearances depending on its
tone and saturation (King et al., 1994, 2002). The
color description that GIA gives diamonds on grading reports is based on the hue, and on the tone and
saturation of that hue. For example, when a yellow
hue becomes darker in tone and weaker in saturation, it appears increasingly brown, and this is
reflected in the color description (i.e., brownish yellow). In certain darker/weaker areas of the yellow
hue range, the color appears to have both brown and
green components (as compared to similar tones
and saturations of adjacent hues), which results in
descriptions such as brownish greenish yellow and
browngreenish yellow. Figure 4 illustrates these
appearance relationships within and surrounding
yellow on the hue circle. (For a more detailed discussion of the GIA color description system for colored diamonds, please see King et al., 1994.)

MATERIALS AND METHODS


Samples. The 24,668 diamonds reported on in this
article were examined in 1998 (7,213) and 2003
(17,455) at our laboratories in New York and
Carlsbad. The color of each was identified as being
of natural origin (using standard and, where appropriate, advanced gemological testing) and was
described on the grading report as being yellow,
brownish yellow, brownish greenish yellow, or
browngreenish yellow (again, see figure 4). The diamonds ranged in weight from 0.30 to 219.35 ct.
For the purposes of this study, we did not
include diamonds in the narrower greenish yellow,
green-yellow, orangy yellow, or orange-yellow hues.
Not only did these samples represent less than 13%
of the diamonds given predominantly yellow
hues for the two years in which we retrieved data
for this studywith no single hue representing
more than 4%but they are also approached differently in their trading and valuation (K. Ayvazian,
pers. comm., 2004).
Because of time constraints or the type of labo-

92

YELLOW DIAMONDS

ratory service requested by the client (e.g., a less


comprehensive identification and origin report),
the database does not include the same gemological
observations for all the diamonds. For these reasons, we have indicated in the Results if the data
being discussed originate from some smaller subset
of these 24,668 diamonds.
Grading and Testing Methods. We used the GIA
Gem Laboratory standard methodology for color
grading colored diamonds to describe all of the
study samples (see King et al., 1994). Screened and
trained laboratory staff evaluated each of the diamonds using a standardized D65 daylight-equivalent lighting environment (as provided by the
Macbeth Judge II illumination box). Typically, from
three to six staff members independently compared
the overall face-up characteristic color of each diamond to GIA colored diamond color references in
order to determine the appropriate color description
and location in GIAs color space.
Equipment used for the gemological examination included a standard gemological microscope, a
GIA Gem Instruments ultraviolet unit with longwave (365 nm) and short-wave (254 nm) lamps, and
a desk-model prism spectroscope.
Absorption spectra in the ultraviolet to visible
(UV-Vis) range were recorded with a ThermoSpectronic Unicam UV500 spectrophotometer over
the range 250850 nm with a sampling interval of
0.1 nm. The samples were mounted in a cryogenic
cell and cooled using liquid nitrogen (196C).
Infrared absorption spectra were recorded in the
mid-infrared (6000400 cm1, 1.0 cm1 resolution)
range at room temperature with a Thermo-Nicolet
Nexus 670 Fourier-transform infrared (FTIR) spectrometer, equipped with a KBr beam splitter. A 6x
beam condenser focused the incident beam on the
sample, and a total of 1,024 scans (per spectrum)
were collected to improve the signal-to-noise ratio.
Diamond type was determined by infrared and visible spectroscopy. We selected a subset of 10,399
samples in order to study spectral differences within
the two categories (Ia and Ib) of type I yellow diamonds, and to see if there were any relationships
between spectra and color appearance or observations with the microscope.
Absorption coefficients for all spectra were calculated assuming a straight light path through the
faceted gemstone. Using absorption coefficients, we
were able to calculate approximate nitrogen concentrations (Field, 1992).

GEMS & GEMOLOGY

SUMMER 2005

Figure 4. The diamonds within the bold lines of this chart


illustrate some of the color appearances seen among the
samples in the study and the terms associated with them,
which are based on hue, tone, and saturation. As shown
here, diamonds in the yellow hue range often appear to
have brown and green components as they become weaker and/or darker. For the purposes of this illustration, the
transition in color seen from the outer ring (at the top) to
the inner ring (at the bottom) is of weakening saturation
and increasing darkness (as illustrated on the tone/saturation chart to the left). Examples from the adjacent hues of
orangy yellow and greenish yellow are included here for
comparison. Note that brown-yellow (and, if the full range
was shown, yellow-brown and yellowish brown) is in the
orangy yellow hue range, not the yellow range.

YELLOW DIAMONDS

GEMS & GEMOLOGY

SUMMER 2005

93

DATA ANALYSIS AND RESULTS


Weight. A significant portion of our samples
30%weighed 3 ct or more, while 17% weighed 5
ct or more and 5% weighed more than 10 ct. Of the
latter, 23 were larger than 50 ct, and three were larger than 100 ct. In general, there are slight (statistically insignificant) differences in these percentages
for 1998 and 2003 (figure 5).
Cut. Yellow diamonds in the modern marketplace
are most often fashioned as fancy shapes. This trend
was apparent in our data, where 94% of the yellow
diamonds examined were shapes other than round
brilliants. Variations on the radiant cut are included
in the square and rectangular categories that
together represent 52% of all the diamonds studied
(figure 6).
Color Appearance. Figure 7 illustrates the range of
tones and saturations in which these yellow diamonds occur. As these charts illustrate, the colors
of yellow diamonds transition smoothly throughout
the ranges of hue, tone, and saturation (i.e., the
changes in appearance are gradual without abrupt
differences between them). The lighter, most-saturated colors are located toward the upper right corner of the charts (Fancy Intense, Fancy Vivid), while
the darker, less-saturated colors lie toward the opposite corner (Fancy Dark). It is important to note that
this distribution of color appearances differs from
those of pink and blue diamonds (see King et al.,
1998, 2002), where the stronger colors are darker in
tone (i.e., toward the lower right corner of the
chart). Five percent of the samples exhibited a
brownish yellow, brownish greenish yellow, or
browngreenish yellow color appearance.
GIA Fancy Color Grades. The study samples covered
all the GIA fancy grades associated with yellow
diamonds (i.e., Fancy Light, Fancy, Fancy Intense,
Fancy Dark, Fancy Deep, and Fancy Vivid). Unlike
other colored diamonds (except brown), yellow diamonds with grades of Faint, Very Light, and Light are
not considered to be fancy-colored diamonds, but are
part of GIAs D-to-Z color grading scale.
The pie charts in figure 8 show how the yellow
diamonds in our study fell into the fancy grade categories for 1998 and 2003. Colored diamond grading activity at GIA more than doubled from 1998 to
2003, but, as seen in figure 8, the general colorgrade distribution remained consistent during these
two years.

94

YELLOW DIAMONDS

Figure 5. These three pie charts illustrate the distribution of the 24,668 diamonds in the study in several
weight categories. The bottom two charts show that
there was very little difference seen in the weight categories between 1998 and 2003.

Hue. As noted above, the diamonds in this study


transition smoothly throughout the yellow hue
range. This is consistent with GIAs observations of
yellow diamonds over the past 50 years. At the
extremes, two subtly different color appearances are

GEMS & GEMOLOGY

SUMMER 2005

observed (cooler toward the greenish yellow


boundary and warmer toward orangy yellow) that
are still described as yellow with no modifying terms
(figure 9). In the more saturated grades of Fancy
Intense and Fancy Vivid, the diamond community
sometimes refers to these two appearances as lemony and golden, respectively (N. Livnat, pers.
comm., 2004).
Tone and Saturation. Yellow diamonds occur in
broad ranges of tone and saturation. They reach their
strongest saturation at relatively light tones (again,
see figure 7), with saturation strengths as high as we
have encountered for any other diamond color.
Microscopic Examination. Clarity Grades. Of the
17,152 diamonds examined for clarity, 30% were
VVS or Flawless/Internally Flawless (FL/IF), 50%
were Very Slightly Included (VS), and 20% were
Slightly Included (SI) or Included (I). Some variations
were noted between samples graded in 1998 versus

Figure 6. This pie chart illustrates the percentage


breakdown for the various shapes in which the yellow diamonds in our study were fashioned. Each
shape category may include different cutting styles
(such as step, modified brilliant, and brilliant cuts).
Included in the miscellaneous category are shapes
such as marquise, triangle, shield, kite, and briolette. The cut-cornered or round-cornered rectangular/square sections include variations on the radiant cut, which helped revolutionize the availability
of fancy-color yellow diamonds in the marketplace.

YELLOW DIAMONDS

2003 (figure 10). For example, there was a decrease


in the percentage of FL/IF (from 9% to 5%) and an
increase in those graded SI (from 14% to 19%).
We were interested to see if there might be a
correlation between depth of color and the clarity
of the yellow diamonds in our study, so we
reviewed the overall distribution of clarity grades
for those diamonds graded Fancy Vivid versus those
graded Fancy Light (as these two grades would represent the extremes in strength of color). In general,
the distribution was similar (figure 11), although a
higher percentage of Fancy Vivid diamonds
received grades of Flawless or Internally Flawless
(12% versus 4% for the Fancy Light group); conversely, a higher percentage of Fancy Light diamonds received VS grades (55% versus 45% for the
Fancy Vivid group).
Inclusions. It is common to encounter small clouds
or strings of pinpoint-like solid inclusions in yellow diamonds (figure 12; see also Crowningshield,
1994). Occasionally dark crystals are seen as well
(figure 13). We also noted the presence of small
pinpoint or cloud-like inclusions toward the center
of a number of the large (over 30 ct) diamonds in
our study.
Graphite was the most common mineral inclusion observed, typically occurring in fractures of
various sizes, displaying a flat shape. We occasionally observed euhedral crystals of both peridotitic and
eclogitic minerals. Olivine (colorless), pyrope garnet
(purple), and diopside (green) were observed as inclusions from the peridotitic group, which in general is
more abundant in natural diamonds than the
eclogitic group. From the eclogitic group, we usually found almandine-rich garnet (orange; figure 14),
omphacite (grayish blue), and rutile (dark reddish
orange). These inclusions were generally smaller
than 100 m, but they ranged up to several hundreds of micrometers in longest dimension. In rare
cases, these inclusions displayed octahedral morphology. Over 80% of all natural diamonds are peridotitic (Meyer, 1987). However, in the majority of
the yellow diamonds, it seems that macro inclusions of eclogitic group minerals were as common
as those of peridotitic groups. Euhedral macro inclusions rarely occur in diamonds dominated by isolated nitrogen, and this was also the case for most of
the type Ib diamonds examined in this study. What
we did observe in a few of these type Ib samples,
however, were small, needle-like fractures surrounding solid pinpoint inclusions.

GEMS & GEMOLOGY

SUMMER 2005

95

GIA YELLOW DIAMOND COLOR CHART


WARMER YELLOW

Figure 7. These two tone/saturation charts illustrate the color appearances at two locations in the
yellow hue range: warmer yellow diamonds that lie toward the orangy yellow/yellow hue boundary, and cooler ones that lie toward the greenish yellow/yellow boundary. In each case, the lighter,
more saturated colors are seen in the upper right of the diagram, whereas the darker, weaker colors
are located toward the lower left. On each diagram, the inset chart shows the generalized boundaries
of the GIA Gem Laboratory fancy grades. The shaded areas on the two charts indicate the areas in

96

YELLOW DIAMONDS

GEMS & GEMOLOGY

SUMMER 2005

COOLER YELLOW

the yellow hue range in which diamonds described as predominantly brown (not part of this study)
occur. Note that these two diagrams are for illustrative purposes only; by themselves, they are not
adequate for use in color grading. Because of the inherent difficulties of controlling color in printing
(as well as the instability of inks over time), the colors of the images shown here may differ from the
actual colors of the diamonds. Photos by Elizabeth Schrader and C.D. Mengason.

YELLOW DIAMONDS

GEMS & GEMOLOGY

SUMMER 2005

97

Figure 8. These pie charts show the percentages of


the diamonds in the study in each of the GIA Gem
Laboratory fancy-color grade categories for the
years 1998 and 2003. Note the relatively consistent
distribution between 1998 and 2003, even though
requests for laboratory services more than doubled
over this time period.

Figure 9. These Fancy Intense yellow diamonds illustrate four different positions within this color group.
Members of the diamond trade sometimes describe
those colors that lie near the yellow/greenish yellow
boundary (like the diamond on the far right) as
lemony, and those near the orangy yellow/yellow
boundary (as seen in the diamond on the far left) as
golden. Photos by Elizabeth Schrader.
Figure 10. The diamonds studied were relatively
high in clarity, with approximately 80% of the
studied samples receiving grades of VS or higher.
These percentage distributions were relatively consistent between the sample groups examined in
1998 and 2003.

98

YELLOW DIAMONDS

GEMS & GEMOLOGY

SUMMER 2005

Figure 12. Stringers of numerous tiny inclusions, some


in reflective orientation, create an interesting cloud
formation in this brownish yellow diamond. The presence of this cloud further supports the diamonds natural-color determination. Photomicrograph by John I.
Koivula; magnified 30x.

Color Zoning. Only 8% of the diamonds in our


study exhibited noticeable color zoning. When present, the zoning often took the form of diffused or
concentrated patches of color (an appearance occasionally referred to as scotch and water).
Ultraviolet Fluorescence. We found that 71% of our
samples showed either no or a very faint reaction

Figure 11. These two pie charts show the clarity grade
distributions among those yellow diamonds in our
total sample population with Fancy Vivid and Fancy
Light color grades (which represent the extremes in
yellow color saturation).

Figure 13. Clusters of small black and metallic-looking needles and platelets, such as this triangular
grouping, were seen in some of the yellow diamonds.
These inclusions are probably composed of sulfides
and graphite. The triangular form of this cluster suggests that it might also be a phantom plane.
Photomicrograph by John I. Koivula; magnified 25x.

Graining. When observed under standard diamond


grading conditions, the vast majority of the yellow
diamonds did not exhibit typical forms of internal
graining such as reflective planes or colored banding
(i.e., whitish or brown). However, we found that one
form of internal graining not commonly encountered
in other fancy colors occurs more frequently in yellow
diamonds. In this type of graining, the internal reflective graining plane exhibits a dispersion of color (figure 15) and maintains this appearance through a range
of viewing angles. Because of the dispersive color
appearance, this graining is described as rainbow
graining; in some cases, it may affect the clarity grade.

YELLOW DIAMONDS

GEMS & GEMOLOGY

SUMMER 2005

99

Figure 14. The most recognizable color for an eclogitic


garnet inclusion in a diamond is orange. Such garnets
are mixtures of almandine and pyrope. Their presence
also provides proof of the diamonds natural origin.
Photomicrograph by John I. Koivula; magnified 20x.

(both of which are referred to as none on grading


reports) when exposed to long-wave UV radiation,
while 25% displayed a faint to medium reaction. A

Figure 15. This 5.04 ct yellow diamond exhibits rainbow graining, a type of graining seen more commonly in yellow diamonds than in other colors. When this
dispersion of colors remains visible through a range of
motion, it can affect the clarity grade. Photo by
Vincent Cracco; magnified 12x.

100

YELLOW DIAMONDS

strong or very strong reaction was observed in less


than 4% of the total group (figure 16). Of these latter samples, 92% exhibited blue fluorescence, while
the remainder displayed other colors (including
green, yellow, orange, and white).
Eighty-eight percent of the samples displayed no
reaction or a very faint to faint reaction to shortwave UV radiation. The remainder had medium
(11%) or strong reactions (1%). In terms of their fluorescence colors, 91% were yellow. The remainder
fluoresced various other colors, but predominantly
blue or green.
Spectrometry and Diamond Type Groups. Based on
our subset of 10,399 diamonds, we found that type I
yellow diamonds may be broadly categorized into
five groups based on their UV-visible and infrared
spectroscopy. Representative samples for each
group are illustrated in figure 17.
Group 1. This first group represents the vast majority of yellow diamonds in the subset (92.8%). Based
on the intensity of features in their spectra, they
demonstrated a high concentration of nitrogen and
highly aggregated nitrogen impurities, consistent
with type Ia diamond. Spectral features arising from
isolated nitrogen were rarely seen in the samples in
this group. The UV-visible and infrared spectra for a
representative 1.31 ct Fancy Vivid yellow diamond
are displayed in figure 17A. The UV-visible spectrum is dominated by absorptions due to the N3
center, with zero-phonon-line absorption at 415 nm
and its related sidebands with peaks at 376, 384,
394, and 403 nm. Other absorption features presentat 425, 438, 452, 465, and 478 nmare all
related to the N2 center (Collins, 1982). The absorption features seen between 1400 and 1000 cm1 of
the infrared spectrum suggest high concentrations
of aggregated nitrogen. Due to the completely radiation-absorbing nature of the type Ia nitrogen, spectral fitting could not be performed to calculate exact
concentrations of nitrogen; however, these features
suggest that this diamond has nitrogen concentrations of at least 200 ppm. Other spectral features
characteristic of this group include a strong peak at
about 1370 cm1 associated with plateletsan
extended defect composed of interstitial carbon and,
possibly, nitrogen atoms (Collins, 2001)and weak
absorption peaks at 1547, 1520, and 1495 cm1. At
higher wavenumbers, weak peaks at 3236 and 3106
cm1 are present, which are indicative of the presence of a trace hydrogen impurity (Field, 1992).

GEMS & GEMOLOGY

SUMMER 2005

Group 2. These diamonds comprise approximately


4% of the sample set, and they share many of the
same nitrogen-related infrared and UV-visible spectral features as the diamonds detailed in Group 1.
However, they differ significantly from the Group 1
samples due to strong absorption features related to
hydrogen. This group is represented by a 1.26 ct
Fancy brownish yellow diamond. The infrared spectrum (figure 17B) displays high concentrations of type
Ia nitrogen (similar to Group 1 diamonds) but clearly
differs from them by features at 4495, 4168, 3236,
3106, 2813, 2785, and 1405 cm 1all related to
hydrogen defects (Fritsch and Scarratt, 1992). The
UV-visible spectrum also displays the N3 and N2 features associated with Group 1 diamonds, but it contains hydrogen-related absorptions at 474 and 563
nmand two weak but broad bands at 545 and 555
nmas well. Furthermore, a weak rise from a broad
band resulting from a broad band center around 700
nm to longer wavelengths, possibly hydrogen related,
is present (Fritsch and Scarratt, 1992).
Group 3. We found that 1.7% of the sample set fell
into this group. These diamonds differed from all
the other yellow diamonds because of their green
luminescence to visible light. The 0.40 ct Fancy
Intense yellow diamond seen in figure 17C is an
example of this group. The UV-visible spectrum displays absorption due to the N3 center and weak N2related absorption features situated on a gradual rise
at the blue end of the spectrum (which are collectively responsible for the predominantly yellow
color). The weak green luminescence observed in
these diamonds is a result of the H3 center (a defect
consisting of two nitrogen atoms and a vacancy) at
503.2 nm. The infrared spectrum suggests that
almost all of the infrared active nitrogen is type IaB,
with a calculated total B-aggregate nitrogen concentration of more than 350 ppm.
Group 4. This group encompassed 0.8% of the sample set and is represented by a 2.03 ct Fancy Vivid
yellow diamond. Unlike the other groups, the
infrared spectrum (figure 17D) displays a relatively
low concentration of nitrogen (calculated to be 16
ppm by means of spectral fitting). The nitrogen is
mostly unaggregated (~11 ppm), and these predominantly type Ib diamond spectra exhibit a sharp
absorption line at 1344 cm1, and a broader peak at
~1130 cm1. A small amount (~5 ppm) of aggregated
type IaA nitrogen, represented by an absorption at
1282 cm1, is also present. Notably, the 16001450

YELLOW DIAMONDS

Figure 16. These two pie charts illustrate the strength


of fluorescence of the studied diamonds to long- and
short-wave UV radiation. In general, more than 80%
of the samples displayed no fluorescence or faint fluorescence to both wavelengths. The most common colors observed were blue (92%) for long-wave and yellow (91%) for short-wave UV.

cm1 region lacks any distinctive features. The visible spectrum is virtually featureless aside from a
rise in absorption from 500 nm to shorter wavelengths. However, some diamonds in this group
showed nitrogen-vacancy (NV) related absorption
features at 575 and 637 nm.
Group 5. The fifth group of yellow diamonds made
up 0.7% of the sample set. It is represented by a
1.56 ct Fancy Intense yellow diamond. Absorption
features related to N2 are generally much weaker
than those in Group 1 for comparable color saturations. Similar to the samples in Group 4, the UV-

GEMS & GEMOLOGY

SUMMER 2005

101

102

YELLOW DIAMONDS

GEMS & GEMOLOGY

SUMMER 2005

Figure 17AE. Yellow diamonds may be broadly


divided into five different groups based on their UVvisible and infrared spectra. One set of spectra for
each of the five diamonds illustrated was chosen to
represent the individual groups.

visible spectrum displays a gentle rise in absorption


from 500 nm to shorter wavelengths, but with an
obvious N3 zero-phonon line and its related sidebands. A weak H3 absorption was also observed.
The infrared spectrum, like the spectra for diamonds in Groups 1 and 2, typically consists of
totally absorbing type Ia nitrogen and the platelet
peak. However diamonds of this group commonly
had a broad absorption at 1650 cm 1; three absorptions at ~1480, 1450, and 1430 cm 1 of varying
sharpness; and, in many cases, a very weak absorption at 1344 cm 1 due to isolated (type Ib) nitrogen
(figure 17E).

DISCUSSION
Weight. With nearly one-third of our samples for
both years of our study weighing more than 3 ct, it
appears that the likelihood of encountering comparatively large yellow diamonds is significantly higher
than for other colored diamonds, which typically
occur in small sizes (see, e.g., Hofer, 1998; King et
al., 1998, 2002). To confirm this, we looked at the
average weight for colored diamonds submitted to
the laboratory in the various hues in 2003. The average weight for yellows (3.23 ct) exceeded that for all
other hues and was more than one-third larger than
the average for blues. King et al. (1998) indicated that
14% of the blue diamonds studied were over 5 ct,
compared to 17% for this study. While these percentage differences do not initially appear great, the
largest diamond in the blue study weighed 45 ct,
whereas the present study included 23 diamonds
over 50 ct and three larger than 100 ct. Indeed, from
our experience, no other colored diamond color
occurs so consistently in such large sizes.
Cut. As discussed in previous studies (King et al.,
1998, 2002), the primary goal of the cutter of colored
diamonds is to enhance the final face-up color
appearance. As shown by our data, yellow diamonds
are typically cut as fancy shapes to help attain the
most intense face-up color appearance. For a detailed
discussion on manufacturing aspects of yellow diamonds, see box A.

YELLOW DIAMONDS

Clarity. As with other natural-color diamonds, color


is the primary factor in the valuation of yellow diamonds. Still, their relative abundance causes the diamond trade to seek additional means to differentiate
among them, so the value placed on clarity is higher
than for more rarely encountered colors (N. Livnat,
pers. comm., 2005). This situation is evident in the
shift in requests for specific GIA colored diamond
reports between the two years of our study. Of the
two colored diamond grading services offered by the
laboratory, the Colored Diamond Grading Report
includes clarity grading, while the Identification and
Origin of Color Report (which focuses on the diamonds color information) does not. Requests for
reports that included a clarity grade for the yellow
diamonds described here increased from 57% in 1998
to 75% in 2003. The increased desire for more grading information (as represented by clarity grades)
appears to parallel the increasing interest in yellow
diamonds over this time frame.
Because a higher clarity grade can be a positive
marketing factor, it is not unusual for a manufacturer to try to attain a minimum VS grade when possible (M. Witriol, pers. comm., 2005). Such an
approach is consistent with the finding that 80% of
the diamonds in our study were graded VS or higher. Manufacturers indicate that it is easier to attain
a higher clarity grade with yellow diamonds than
with diamonds in other colors because of the relative purity of some of the larger yellow rough (M.
Witriol, pers. comm., 2005).
Interestingly, our data showed a percentage
decrease for the clarity grade categories of FL/IF and
VVS (from 9% to 5% and 26% to 22%, respectively)
between the two years of our study, but an increase
from 14% to 19% for SI. We do not know if these
changes in clarity grade distribution are a result of
more requests for reports with clarity grades (from
57% to 75%, as noted above), greater effort to
quickly fill market need (i.e., manufacturers are
working toward quicker turn-around rather than
higher clarity), a change in the quality of rough
being used, or a combination of these factors.
The analysis of clarity grade versus strength of
color supported our assumption that there was little
or no relationship between these two characteristics.
As noted in our results, the distribution of clarity
grades for Fancy Light and Fancy Vivid was similar,
with the exception of FL/IF, where there were three
times more diamonds in this category for Fancy
Vivid than Fancy Light. The rarer a diamonds inherent color, the more likely it is that a manufacturer

GEMS & GEMOLOGY

SUMMER 2005

103

BOX A: THE IMPORTANCE OF CUT IN YELLOW DIAMONDS


Manufacturing. Today, yellow diamonds (and colored
diamonds in general) are typically cut as fancy shapes
to enhance their color appearance (see figure A-1). It is
not unusual for yellow rough to occur as octahedra (I.
Wolf, pers. comm., 2005). The best yield and potential
color retention for such crystals is found in square or
near-square shapes such as radiants, emeralds, or cushions (which comprised as many as 61% of the samples
in this study; see again figure 6).
Given the relative lack of inclusions and color zoning in large yellow rough, manufacturers can focus
more on weight retention and color appearance (M.
Witriol, pers. comm., 2004).
Cutting Innovations. Much of the framework on which
todays cutting decisions are based originated with
innovations begun in the 1970s. Previously, diamonds
with light yellow bodycolor were manufactured in the
shapes, cutting styles, and proportions of diamonds on
the D-to-Z scale. Recent personal accounts to one of
the authors (JMK) described how some of these changes
came about (M. Blickman and J. Doppelt, pers. comms.,
2005). By the mid-1970s, New York manufacturers
such as Stanley Doppelt and Henry Grossbard had
begun experimenting with variations on Basil
Watermeyers 1971 Barion cut, a square mixed cut
whose step crown and modified-brilliant pavilion
improved brilliancy and increased yield from the traditional step cut (see Kerr, 1982), in an attempt to hide
imperfections in cheap, square-cut diamonds. The
assumption was that the increased scintillation from
the mixed cut would better disguise inclusions, thus
giving a more pleasing appearance to the eye. From an
early point in this experimentation, Mr. Doppelt (and
later Mr. Grossbard) realized these variations also
enhanced the color of the yellow diamonds they were
recutting, especially light yellow hues. Throughout the
mid- to late 1970s, other experienced cutters experimented with the angles on these diamonds (from offcolor to noticeably yellow) to produce the best face-up
color appearance.
In the late 1970s and early 1980s, the trade became
aware of these potential cutting benefits. Mr. Doppelt,
in conjunction with Louis Glick, introduced the StarBurst cut, which helped gain popularity for yellow diamonds. Mr. Grossbard, for his part, patented the radiant
cut (Overton, 2002). Mr. Grossbards 1976 purchase at
Sothebys Zurich of an off-color 109 ct diamond then
named the Cross of Asia would eventually add to trade
interest as well. He decided to recut that diamond to his
new cut. Industry observers felt that the resultant 79 ct
Flawless, Fancy yellow renamed the Radiant Cut diamond was significantly more valuable than the original
Cross of Asia (M. Blickman, pers. comm., 2005; this
stone was graded prior to the 1995 modifications to the

104

YELLOW DIAMONDS

Figure A-1. This 10.12 ct Fancy Vivid yellow pear shape


illustrates the effect a fancy shape can have on intensifying
the color in a yellow diamond. Courtesy of the Scarselli
family; photo by Elizabeth Schrader.

GIA colored diamond color grading system).


As others in the trade became aware of this recut
diamond and the effectiveness of these new cuts in the
market, more manufacturers began recutting light yellow diamonds into these styles in an effort to achieve
grades of Fancy Light or Fancy (M. Kirschenbaum, pers.
comm., 2004). Key features of these cuts were halfmoon facets on the pavilion, French culets, and a
greater number of facets in general. The most dramatic
differences in face-up color appearance were often seen
when the starting material was a round brilliant (figure
Figure A-2. Some of the most striking differences are seen
when a light yellow round brilliant is recut into a shape
and cutting style, such as the radiant, that can accentuate
the color appearance. For example, the 6+ ct round brilliant
diamond on the left was graded in the W-X range of the Dto-Z scale. When recut as a 4.61 ct radiant (right), it was
graded Fancy yellow. Courtesy of the Scarselli family; composite photo by Elizabeth Schrader.

GEMS & GEMOLOGY

SUMMER 2005

A-2). Over time, the new shapes and cutting styles were
used on diamond rougha practice that resulted not
only in better color grades but also in better weight
retention, so manufacturing light yellow rough became
more profitable (I. Wolf, pers. comm., 2004). Even more
than the opening of new mines or a mines increased
production, these cutting activities were responsible for
the greater number of more intensely colored yellow
diamonds that appeared in the marketplace.

ate a softer, lighter appearance. If the color is near a


grade boundary, such an appearance often results in the
diamond being placed in a lower grade (i.e., with a weaker color; figure A-3). Similarly, small areas of windowing
can mix with the color to create an overall washed out,
weaker (i.e., less saturated) appearance. Overall darkness, or numerous dark areas due to extinction, can
cause a grayed-out, weaker look. Occasionally, these
extinct areas can create a deeper appearance. Again,
these effects can have an impact on color grade if the
diamond is near a grade boundary.
Many diamond cutters are aware of the range of proportions that yield an attractive face-up cut appearance
in colorless to near-colorless diamonds, but those who
also cut colored diamonds often say when it comes to
color, you can throw away the book (I. Wolf, pers.
comm., 2005). An example of this is the rare Fancy
Vivid yellow round brilliant in figure A-4. When
observed by experienced staff and members of the trade,
all agreed it was an exceptional color. But with a total
depth of more than 65% and crown angles greater than
39, it was unorthodox for a round brilliant and, if it
had been near-colorless, would likely be considered less
attractive than many other near-colorless diamonds.
Clearly, the assessment of diamond cut for fancy-color
diamonds is based on very different considerations than
their D-to-Z counterparts.
Finally, cutting can also affect the perceived distribution of color when the diamond is viewed face-up.
When the characteristic color does not predominate in
the face-up position, it is reported as uneven for the
distribution entry on grading reports (see King et al.,
1994). Figure A-5 shows four examples of uneven color
distribution in different shapes and cutting styles of yellow diamonds.

Effects of Cut on Yellow Diamond Color Appearance. As


noted above, the appearance of light-toned colors is easily influenced by the effects of cut. Cutting can create
brightness (white or near-white light return), fire (the
dispersion of various colors), windows (see-through),
and dark areas (extinction). Because colored diamonds
are graded in the face-up position, the overall blend of
sensationssmall areas of brightness, windowing,
and/or extinctionmay intermingle with the yellow
face-up color and affect the overall color appearance. For
example, brightness can visually blend with color to cre-

Figure A-4. With colored diamonds,


cut is evaluated first in terms of its
effect on the color. While this 1.51
ct Fancy Vivid yellow diamond
has proportions that would be
considered unorthodox for a
colorless or near-colorless round
brilliant, here they have produced
a spectacular color appearance.
Photo by Elizabeth Schrader.

Figure A-3. Use of a diamond cut that highlights brightness can cause a light-toned yellow diamond to appear
even lighter. The diamond on the left has numerous
small, bright spots as a result of the manufacturers cutting decisions. The one on the right is the same diamond
with the spots digitally removed and replaced with the
average face-up color. When the small bright spots are
removed, the color appearance is subtly deeper (i.e., it
appears slightly darker and stronger). If such a diamond
were near a grade boundary, this subtle difference in
appearance could move the diamond into a higher grade.
Composite photo by Elizabeth Schrader.

Figure A-5. In some diamonds, the yellow


color can appear limited to a minority of
areas relative to the total area in the faceup position. Such yellow diamonds would
be described as having an uneven color
distribution on GIA grading reports. As
illustrated here, the appearance produced
by this uneven distribution can vary
depending on the cut and proportions used.
Composite photos by Elizabeth Schrader.

YELLOW DIAMONDS

GEMS & GEMOLOGY

SUMMER 2005

105

will consider working on other characteristics to further enhance its marketing potential (N. Smid, pers.
comm., 2005). For the less-saturated colors, the manufacturer will factor in additional cutting (loss of
weight), time costs, and current market salability
into the decision to work toward the highest possible
clarity. This approach is followed for colorless to
near-colorless diamonds as well. It is much more
likely for a manufacturer to try, when possible, to
achieve a Flawless grade on a D color diamond
than on a K color (N. Smid, pers. comm., 2005).
Inclusions. Given that the vast majority of the yellow diamonds in our sample group were type Ia, it
is not surprising that their inclusions appeared to be
the same as those encountered in diamonds on the
D-to-Z scale (Koivula, 2000). Since the composition
of these diamonds is essentially the same as those
on the D-to-Z scale (differing only in concentration
of defects), such a continuity of appearance would
be expected.
Over the years, we have tracked the frequency
with which pinpoint or small cloud-like inclusions
occur toward the center of large (over 30 ct) diamonds
submitted to the lab. A brief review of some of our
historic records indicated that they were observed in
more than half of 29 yellow diamonds weighing
between 50 and 60 ct. The reason for this centrally
located feature is not known, although these inclusions may represent growth conditions similar to
those of symmetrical micro-inclusion clouds seen
in other diamonds (see, e.g., Crowningshield, 1965).
In our experience with yellow diamonds, the cloudlike inclusions may be distributed randomly throughout a stone, or they may follow specific growth sectors; in particular, cubic growth sectors {100} tend to
contain more clouds than any other sector (see, e.g.,
Wang and Mayerson, 2002). Occasionally, the clouds
are very obvious and can affect the color appearance
as well as the clarity grade. Although little is known
about the chemistry and phase relations of the tiny
clouds, with few exceptions high levels of infraredactive hydrogen (see below) were commonly detected
by infrared spectroscopy in the regions where these
clouds were concentrated.
Color Zoning. To see if there was a relationship
between our overall findings (8% of the samples
exhibited color zoning) and the diamond type
groups, we examined our subset of 10,399 diamonds
for which type spectra were obtained. Most (92.3%)
of our type Ia yellow study diamonds (Group 1) did

106

YELLOW DIAMONDS

not exhibit noticeable color zoning. However, diamonds with different kinds of nitrogen aggregations
(as evidenced by features in their spectra) had much
higher occurrences of uneven color zoning.
Specifically, 62.3% of diamonds that showed hydrogen absorption lines in their visible spectra (Group
2) had uneven color zoning, usually in the form of
brown growth sectorrelated zones. In addition,
65.1% of diamonds that had green visible luminescence (Group 3) showed uneven zoning, which usually appeared as straight yellow lines paralleling
internal graining. The type Ib diamonds (Group 4)
showed color zoning in 56.8% of its samples, most
commonly in the form of diffused zones or concentrated patches of color, occasionally referred to as
scotch and water. Unlike in other colors, such as
type I pinks and type IIb blues where the zoning
tends to form in discrete bands, the scotch and
water zoning has less effect on manufacturing
decisions with regard to orientation of the zoning in
relation to the face-up appearance (I. Wolf, pers.
comm., 2005). Like Group 1, the diamonds in
Group 5 did not show noticeable color zoning.
Ultraviolet Fluorescence. Of the 10,399 diamonds
studied for diamond type, only 75 showed a shortwave UV reaction that was stronger than the longwave reaction. Perhaps most interesting was that 58
(77.3%) of these diamonds were type Ib. While not a
conclusive test, this observation may be helpful to
the diamantaire who does not have access to sophisticated equipment but wishes to have an indication
of diamond type.
Spectrometry and Diamond Type Groups. As noted
earlier, the yellow color observed in natural diamond is mainly a result of nitrogen, the most common impurity found in this gem. Both aggregated
and isolated forms of nitrogen within the diamonds
lattice are responsible for the unique absorption features in the blue portion of the visible spectrum,
which in turn give rise to the yellow coloration.
Variations in nitrogen aggregation that begin during
diamond formation (as well as the addition of other
trace impurities and other defects) lead to the differences seen in nitrogen-related spectral features.
The complex stages of nitrogen aggregation in
diamond take place over geologic time at high pressures and high temperatures within the earth.
During the initial stages of growth, nitrogen atoms
may substitute for single carbon atoms within the
diamond lattice. Diamonds containing isolated

GEMS & GEMOLOGY

SUMMER 2005

nitrogen are classified as type Ib (here, Group 4);


they owe their unique golden yellow color to a
rise in absorption from 500 nm to lower wavelengths (again, see figure 17D). In general, these yellow diamonds have low total nitrogen, even though
their color is usually strongly saturated. As nitrogen-bearing diamonds continue to develop within
the earth, the nitrogen atoms will migrate through
the diamond lattice and over time may begin to
form the commonly observed A, B, and N3 aggregates. Our spectroscopy results suggest that these
optical centers are represented in the remaining
four groups.
Group 1 diamonds contain a high concentration of N3 centers as well as N2 centers, which are
attributed to a vibronic transition of the N3 center
(Zaitsev, 2001). These are by far the most abundant yellow diamonds. Figure 18 shows overlain
spectra of the Fancy Vivid yellow diamond in figure 17A and a Fancy Light yellow diamond also
from Group 1. These spectra illustrate that the
N2- and N3-related absorptions are stronger in the
Fancy Vivid yellow diamond. Data from our entire
subsample of diamonds on which spectra were
taken suggest that the N2 centers in highly saturated yellow diamonds absorb more strongly than
their less saturated counterparts; however, in
many cases we observed that the N3 center in
intensely colored yellow diamonds had the same
absorbance as in their less intense counterparts.
These data suggest that the N2 center plays a large
role in determining the strength of yellow color
saturation in Group 1 diamonds, as well as contributing to their lemony yellow color.
Group 2 diamonds are closely related to Group
1; however, our representative spectra for this group
indicate high concentrations of hydrogen compared
to Group 1 diamonds. Hydrogen-related absorptions
in the UV-visible spectra may contribute to the
modifying brownish and greenish appearances associated with these diamonds.
Similar to Groups 1 and 2, Group 3 diamonds
contain aggregated N3 and N2 centers that contribute to the overall body color. In addition, however, Group 3 diamonds contain a green luminescence component that is attributed to small concentrations of another form of aggregated nitrogen,
the H3 center, at 503.2 nm. Group 3 diamonds also
tend to contain significant concentrations of highly
aggregated type IaB nitrogen. In fact, our infrared
data suggest that the vast majority of the type Ia
nitrogen in most yellow diamonds with H3-

YELLOW DIAMONDS

Figure 18. These two UV-visible spectra represent the


Fancy Vivid yellow diamond from Group 1 (shown in
figure 17A) and a Fancy Light yellow diamond from
the same group. There are obvious differences in the
strength of the N3 and N2 absorptions between these
two diamonds of very different saturation.

attributed green transmission luminescence is in


the form of B-aggregates.
As diamonds containing aggregated nitrogen
mature in the mantle at elevated temperatures and
pressures, it is possible that defects and nitrogen
aggregates may begin to break up. Group 5 diamonds
have spectral features similar to groups 1, 3, and 4,
and some of these features may be linked to the
destruction of defect centers and the disaggregation
of nitrogen. The presence of the 1480 cm1 peak in
the IR spectrum has been linked to the degrading of
the platelet peak at 1360 cm 1; other features at
1650, 1450, and 1430 cm1 may also be platelet related and increase in intensity as the platelets are
destroyed by increased annealing (Kiflawi et al.,
1998). In addition, Group 5 diamonds typically show
trace amounts of isolated (type Ib) nitrogen, which
suggests the breakdown of aggregated nitrogen after
heating (Fisher and Spits, 2000).
Color Appearances as They Relate to the Diamond
Type Groups. Despite the overlap, we did note different representative color appearances associated
with the five groups. This can be seen in the photos
accompanying the spectra of each group (again, see
figure 17). The chart in figure 19 shows the general
distribution of the diamonds in these groups in relation to the yellow fancy grades. The appearances
associated with this larger distribution of tones and
saturations can be seen when the areas noted in figure 19 are compared to the images in the corresponding locations in figure 7.

GEMS & GEMOLOGY

SUMMER 2005

107

The diamonds associated with Group 1 (more


than 90% of our samples) have the appearance most
commonly associated with yellow diamonds. In the
stronger grades of Fancy Intense or Fancy Vivid yellow, these diamonds are often described as lemony. Group 5 diamonds were similar in appearance
to those in Group 1.
The green luminescence characteristic of Group
3 diamonds can be strong enough to affect the
apparent color. However, the luminescence in our
study samples was not noticeable enough under
standard grading conditions to place the diamonds
outside of the yellow hue range, although they were
located toward the cooler side of that range (i.e.,
toward the yellow/greenish yellow boundary).
Group 2 diamonds often appeared brownish or
had brown and green components, which may be
due to the hydrogen-related absorptions in the UVvisible spectra. It is also likely that the spectral features of Group 4 diamonds (all type Ib) are responsi-

Figure 19. While there are overlapping color appearances among the five spectral groups of yellow
diamonds identified in our study, some areas are
more representative of each particular group than
the others. This generalized tone/saturation chart
(with the fancy grades for yellow diamonds overlaid) illustrates the most common areas in which
the diamonds in the five groups occur.

108

YELLOW DIAMONDS

ble for the strongly saturated, warmer yellow typical of these diamonds. Such golden color appearances are often associated with grades of Fancy
Vivid and Fancy Deep yellow.
Color Appearance Transitions. As mentioned
above, unlike other colors, the initial colored diamond grade associated with yellow is Fancy Light
(again, see figure 7). The grade descriptions of Faint,
Very Light, and Light are part of the D-to-Z scale,
and they correspond to the K-M, N-R, and S-Z letter
grade ranges, respectively (King et al., 1994). The
transition in color appearance from these D-to-Z
scale diamonds to fancy-color yellow diamonds is
smooth, but the grading methodology and philosophy between the two scales changes abruptly.
A fundamental difference between these two
grading scales is the value placed on absence of
color in one (D-to-Z) versus presence of color in
the other (colored diamond). The absence of
color is primarily observed with the diamond in
the table-down position, whereas the presence of
color is only judged in the face-up position (King
et al., 1994).
As the depth of color increases for diamonds on
the D-to-Z scale, the role of face-up observation (in
addition to table-down viewing) also increases in
importance during grading. The transition boundary
between the D-to-Z scale and a fancy-color grid is
the Z grade. At this location, face-up appearance
becomes the determining factor in assigning the
color grade (i.e., a diamond must have a stronger
face-up color appearance than the Z master to be
considered a fancy color regardless of the bodycolor
observed table down; figure 20).
Not only does methodology differ between these
approaches, but the ranges of tone and saturation
associated with a given grade are quite different as
well. Grade ranges for fancy-color diamonds are significantly broader in both tone and saturation
attributes than those on the D-to-Z scale.
In past observations, we have noted that some
yellow color appearance transitions are more common than others, and this was supported by our
data. The most common transition is throughout
the saturation range rather than throughout tone or
both tone and saturation. This is seen in the breakdown of our samples fancy-color grades as well as
their color descriptions. For example, about 97% of
the diamonds in our study (representative of all diamonds in the yellow hue range submitted to the lab
for 1998 and 2003) were Fancy Light, Fancy, Fancy

GEMS & GEMOLOGY

SUMMER 2005

Intense, or Fancy Vivid. All of these grades occur in


a restricted tone range of medium to light, but they
span the entire saturation range for yellow diamonds. Just as fancy grades are defined by related
areas of tone and saturation in color space, some of
the word descriptions (i.e., brownish or greenish)
associated with them are also determined by their
area of tone and saturation within the grades. Many
color order/description systems (e.g., Natural
Bureau of Standards, 1976), including GIAs, arrange
their terms in color space to accommodate increasing appearances of brown or gray that occur as colors darken in tone and weaken in saturation. Since
most of the yellow diamonds in our study were
lighter in tone and of relatively strong saturation,
they did not have a brown or green component.
Understanding this ordering convention, and the
occurrence of our samples in color space, clarifies

why so few of the yellow diamonds in our sample


(5%) had descriptions of brownish yellow, brownish
greenish yellow or browngreenish yellow. It is
understood, of course, that the GIA Gem
Laboratory may not see the full range of yellow diamonds available. Nevertheless, given the large
number of samples examined, this distribution may
serve as a good indication of what goods are prominent in the marketplace.
Grade Distribution Over Time. It was interesting to
note from our data that the distribution of fancy
grades among our sample population was relatively
consistent for the two separate years of our study.
While far from being conclusive, these data give
some indication as to the general distribution of yellow diamonds in that segment of the marketplace
that uses GIA reports.

Figure 20. The transition from the D-to-Z scale to the fancy-color diamond scale for yellow diamonds is
greatly affected by the cutters ability to concentrate and intensify the face-up color, since it becomes the
factor in determining the grade (not the bodycolor seen table down). In the diagram on the left, the X
marks the location of a yellow diamond that lies near the end of the D-to-Z scale. The shaded area illustrates the range of potential tones and saturations and, in this case, grades the diamond could receive
depending on how well the cutter was able to intensify the perceived face-up color. This is illustrated by the
two yellow diamonds on the right. When observed table-down (top), both have a similar bodycolor.
However, when the same diamonds are observed face-up (bottom), the differences in appearance are obvious. While the table-down bodycolor of these two diamonds placed them near the end of the D-to-Z scale,
their face-up appearances resulted in the one on the left being placed on the D-to-Z scale and the other
being graded as a fancy color. Photos by Elizabeth Schrader.

YELLOW DIAMONDS

GEMS & GEMOLOGY

SUMMER 2005

109

important methodology questions: (1) Where and for


what is the grader looking when grading color in a
diamond (i.e., observation)? And (2) how does that
diamond relate to other diamonds in terms of color
appearance (i.e., bracketing)?

Figure 21. The evaluation of color


in a colored diamond requires
that the observer focus on overall
blend (the characteristic color)
rather than detail (individual
components of the color). To see
the overall blend, the observer
uses a technique similar to that
used when viewing a 19th century pointillist painting. The detail
highlighted in this famous
Georges Seurat work, A Sunday
on La Grande Jatte, shows a number of distinct, individual colors; when the observer steps back and looks
at the painting as a whole, these colors blend into a
single color sensation. When grading a colored diamond, one can also pick out colors from a mosaic of
appearances in the diamond, but the grade assigned
is based on the overall blend. 1884, oil on canvas,
81 34 x 12114 in.; image courtesy of The Art Institute
of Chicago, Helen Birch Bartlett Collection.

Observation. As mentioned before, when observing a


colored diamond, the face-up position is the only
view used. However, even when restricted to the
face-up position, differences between observers in the
plane of focus can yield different results. This
plane can be thought of as an imaginary surface parallel to the table facet where the eyes of the observer
are focused to evaluate the color. Depending on the
location of this planenear the table facet, at a depth
near the girdle, or even deeper within the pavilion of
the diamondan observer can reach different conclusions about the color. At the GIA Gem Laboratory,
this plane of focus is located close to the table facet,
not deep inside the diamond, so the grader observes
the overall blend that constitutes the predominant
single color appearance (i.e., the characteristic color).
Figure 22. A thorough understanding of yellow diamond color evaluation requires an awareness of the
full range in which the color occurs. For example,
when the Fancy Deep yellow diamond on the left is
shown next to a Fancy Intense yellow diamond (top
right), the color of the former appears darker and
slightly less saturated, which some might consider less
attractive. However, when the same Fancy Deep yellow diamond is shown next to a Fancy Dark
browngreenish yellow diamond (bottom right), the
attraction of its rich yellow hue is more evident.
Photos by Elizabeth Schrader.

A REVIEW AND ELABORATION OF


FACTORS INVOLVED IN COLOR GRADING
When GIAs colored diamond color grading system
was first documented in this journal (King et al.,
1994), we described our grading methods, environment, and terminology, to help those who use our
reports understand the procedures and factors
involved in assigning a color grade. Since that time,
we have described the grading of blue diamonds and
pink diamonds, specifically (King et al., 1998, 2002).
Here, we would like to use yellow diamonds to
review and elaborate on important aspects of color
grading. From our discussions over the years, we
have found that differences in grading interpretation
can often be resolved by consistently answering two

110

YELLOW DIAMONDS

GEMS & GEMOLOGY

SUMMER 2005

Figure 23. Yellow diamonds that lie near the yellow/greenish yellow boundary may appear less saturated if compared to a yellow diamond of similar
strength that lies near the yellow/orangy yellow
boundary. The two Fancy Intense yellow diamonds
seen here are of similar saturation in GIAs grading
system. To the inexperienced observer, however,
the cooler yellow diamond on the right might
appear weaker than the warmer one on the left.
Photos by Elizabeth Schrader and C.D. Mengason.

Even with consistent lighting, environment, and


viewing geometry, if one grader focuses deep into the
diamond (or changes his/her focus to pick out
highlights at different depths), it is possible to arrive
at a conclusion different from that of another grader
who is focusing on the table plane and observing the
overall blend of face-up appearance.
The blended appearance that results from using
a plane of focus near the table of a colored diamond
can be likened to the merging effect that viewing
distance has on color appearance in a 19th century
pointillist painting such as Georges Seurats A
Sunday on La Grande Jatte (figure 21), a commercially printed page (produced with very small dots
of cyan, yellow, magenta, and black ink), or a
checkerboard pattern. In each instance, more than
one color is visible on close inspection, but they
blend to a single overall color appearance when
viewed at a distance.

those toward the cooler end of the range is


different from those at the warm end, and this difference may be confused with strength of color if the
diamond is not compared and bracketed consistently
(figure 23).
Because yellow diamonds commonly transition
in color appearance through a relatively narrow
range of lighter tones, it is easy to misinterpret
those falling outside that norm. To prevent this,
observers need to be aware of the appearance relationships seen throughout the complete tonal range.
Unless compared to known samples, the saturation
of less commonly encountered colors of darker tone
may be considered stronger than they are (figure 24).
For example, the difference in appearance between a
typical Fancy Light yellow diamond and one that is
Figure 24. If a diamond is near a grade boundary,
darker tone can be confused with increasing saturation if known references are not used. Here, two
examples illustrate this situation. On the far left are
three diamonds near the Y-Z/Fancy Light yellow saturation boundary; the diamonds are of similar saturation and differ only in tone. Comparison to the Fancy
Light yellow reference diamond to their right establishes their grade on the D-to-Z scale. On the right, a
similar situation is seen with three diamonds near the
Fancy Intense/Fancy Vivid boundary; their Fancy
Intense grade is clear when they are compared to the
Fancy Vivid yellow reference diamond to their right.
These tonal differences can be misinterpreted by an
inexperienced observer. Photos by Elizabeth Schrader.

Bracketing. Thorough bracketing of the characteristic color with color references of known location
is crucial to understanding the grading of a colored
diamond. This process allows the observer to
understand the colors location in color space and the
description related to that location. Without that
understanding, proper evaluations may be lacking.
For example, a diamond that appears darker and
weaker when compared to lighter, stronger reference
diamonds (figure 22, top) will appear rich in color
when compared to its darker and weaker counterparts (figure 22, bottom).
Bracketing is also important to accurately grade
yellow diamonds of similar tone and saturation but
at opposite ends of the hue range. The appearance of

YELLOW DIAMONDS

GEMS & GEMOLOGY

SUMMER 2005

111

Figure 25. The light-toned yellow diamond shown in


the center appears to stand out relative to the other
two diamonds. Nonetheless, all three diamonds have
the same Fancy yellow color grade. Photo by
Elizabeth Schrader.

darker toned might lead to the incorrect assumption


that there is a difference in saturation, and hence to
the diamond being graded Fancy yellow. Similarly,
fancy-color yellow diamonds that are light in tone
often appear to stand out from darker diamonds
of similar hue, resulting in the misinterpretation of
a higher grade. Lightness does not necessarily
mean they are stronger or weaker in saturation and
therefore fall into a different grade (figure 25).
Even though saturation-based grade transitions
are relatively common among yellow diamonds, the
observer may not know the full range of appearances within a grade, especially since grade ranges
are quite different from those on GIAs D-to-Z scale.
One can incorrectly assume that two yellow diamonds near the upper and lower boundaries of a
grades saturation range are in different grades if the
entire range is not understood.
A number of the examples above illustrate the
fact that there is a different appearance associated
with each transition step in each one of a colors
three attributes. The change in appearance between
a tonal step (lightness to darkness) is different from
that seen in a saturation (strength of color) step.
Similarly, a transition in hue appears different from
one in tone or saturation. Without understanding
the difference between these appearance transitions,
it is difficult to locate one color in relation to others.
The process is further complicated by the complex
face-up appearance of diamonds. In our experience,
by working consistently with known references
(and, preferably, the same references), a grader can
develop an understanding of the appearance relationships between them and consistently interpret the
transition. Changing references may result in less
consistency because the observer risks misinterpreting an attribute in the changed reference that is not
the determining factor for a grade.
An interesting aspect of grading observation that
occurs at the laboratory but is more common in the

112

YELLOW DIAMONDS

marketplace is the occasional need to make a general assessment of a diamonds color while it is in a
jewelry mounting. The appearance of yellow diamonds may be influenced by jewelry mountings to
various degrees, as the color of the metal (especially
yellow gold) can affect the apparent color of the diamond. It is important to keep this in mind when
making color comparisons. At the laboratory, we
follow our standard grading methodology when
grading colored diamonds mounted in jewelry, but
we assign a more generalized, multiple grade range
for the diamond. Figure 26 shows a loose diamond
that was near the Y-Z, Light yellow/Fancy Light
yellow grade boundary before it was mounted (top)
in an 18K gold ring (bottom). The color appearance
of the diamond when mounted was well within the
Fancy yellow range; depending on the cut, this
effect can be more or less pronounced.
Figure 26. The color of most yellow diamonds may
appear noticeably darker and stronger when set in a
mounting. The diamond on the left in the top image is
near the Y-Z, Light yellow/Fancy Light yellow boundary and is shown next to a Fancy yellow reference diamond. In the bottom image, the larger diamond is
mounted in a ring and placed next to the same reference diamond. While the jewelry designer can use
these effects to enhance the appearance of a yellow
diamond, it is important that the observer be aware of
this difference when evaluating the color of mounted
diamonds. Courtesy of the Scarselli family; photos by
Elizabeth Schrader.

GEMS & GEMOLOGY

SUMMER 2005

SUMMARY AND CONCLUSIONS


Yellow diamonds occur in a wide range of tones and
saturations, but the majority encountered in the
industry fall into the saturation range represented by
the grades Fancy Light, Fancy, Fancy Intense, and
Fancy Vivid. Yellow diamonds occur in some of the
highest levels of saturation we have seen in colored
diamonds to date. The fact that they also occur in
larger sizes than other colored diamonds offers greater
versatility for the jewelry designer (figure 27). With
clarity grades that are commonly VS or higher, the
manufacturer is often able to cut to proportions that
offer maximum weight yield and the most intensified
face-up color appearance without requiring as many
modifications for clarity characteristics as with other
colors. An important manufacturing consideration
with yellow diamonds is that the lighter tones are
more readily affected by cutting decisions than darker,
deeper colors. Brightness, windows, and extinction
can affect the grade if the color is already near a grade
boundary. Understanding the effect of jewelry mountings on the color is also important for the designer or
laboratory/jewelry professional who needs to evaluate
yellow diamonds when mounted.
Many of the routine manufacturing decisions
used for colored diamonds today had their roots in
the cutting innovations first seen with yellow diamonds in the 1970s (again, see box A). Early experiments with cut variations employed to hide inclusions were found to also intensify the face-up appearance of light yellow diamonds. Soon these techniques
were applied to a broader tone and saturation range of
yellow diamonds and, eventually, other colors.
While it is difficult to define groupings within type

ABOUT THE AUTHORS


Mr. King is laboratory projects officer, Mr. Gelb is staff gemologist, Mr. Hall is manager of analytical research services, and Dr.
Wang is research scientist at the GIA Gem Laboratory in New
York. Dr. Shigley is director of research, and Mr. Guhin is grading
lab manager, at the GIA Gem Laboratory in Carlsbad, California.
ACKNOWLEDGMENTS: The authors thank the following individuals for providing information on the geologic occurrences of yellow
diamonds: Dr. Jeff Harris of the Geology Department at the
University of Glasgow, Scotland, and Dr. A.J.A (Bram) Janse of
Archon Exploration Pty Ltd. in Perth, Australia. Kim Cino and
Jacquelyn Haahr at the GIA Gem Laboratory in Carlsbad assisted
with the Horizon computer management system retrieval of data
on yellow diamonds. Tom Moses of the GIA Gem Laboratory in
New York provided insights throughout the development of this
article. Joshua Cohn and Akira Hyatt, both also with the GIA Gem

YELLOW DIAMONDS

Figure 27. Increasing numbers of yellow diamonds are


appearing in haute couture jewelry, as seen in these
Gypsy earrings and the Bulls Eye necklace.
Composite photo courtesy of Graff Diamonds.

I yellow diamonds due to the many overlapping spectral features, we were able to establish five groups with
distinct spectral characteristics. We have shown the
spectral relationships between these representative
groups within yellow type I diamonds and noted a general relationship of these groups to color appearance.
An important goal of this study was to provide a
better understanding of the range of color appearances
associated with yellow diamonds. Understanding this
range will help the growing number of traders in yellow diamonds evaluate these diamonds properly.

Laboratory in New York, provided helpful discussions and coordination of diamond selections for the various images, and assisted
with the selection and appearance relationships of the colored diamond images, respectively. John Koivula, formerly of the GIA Gem
Laboratory in Carlsbad, provided commentary on the inclusion
analysis. The colored diamond color grading staff at GIA in New
York and Carlsbad helped identify several important discussion
points for the article. Numerous members of the diamond trade
gave their time and assistance through the loan of yellow diamonds and helpful discussions. Among them, the authors thank
Isaac Wolf, Mates Witriol, and Lewis Wolf of Lewis Wolf Trading;
Martin Kirschenbaum of M. Kirchenbaum Trading Inc.; Louis Glick
of Louis Glick Diamond Corp.; Kevo Ayvazian of KJA Diamonds
Intl Corp.; Bruno Scarselli of the Scarselli family; Nir Livnat of The
Steinmetz Group; Mace Blickman of Jerry Blickman Inc.; Nancy
Smid of Safdico USA; Jonathon and Alison Doppelt of Jonathon
Doppelt Inc.; and Ara Arslanian of Cora Diamond Corp.

GEMS & GEMOLOGY

SUMMER 2005

113

REFERENCES
Anderson B.W. (1943a) Absorption and luminescence in diamond, Part I. The Gemmologist, Vol. 12, No. 138, pp. 2122.
Anderson B.W. (1943b) Absorption and luminescence in diamond, Part II. The Gemmologist, Vol. 12, No. 139, pp. 2527.
Anderson B.W. (1962) Lines and line systems in the fluorescence
spectra of diamonds. Journal of Gemmology, Vol. 8, No. 5, pp.
193202.
Anderson B.W. (1963) The classification of diamonds on the basis
of their absorption and emission of light. Journal of
Gemmology, Vol. 9, No. 2, pp. 4454.
Anderson B.W., Payne C.J. (1956) The spectroscope and its applications to gemmology. The Gemmologist, Vol. 25, No. 300,
pp. 115119.
Balfour I. (1997) Famous Diamonds. Christie, Manson and Woods
Ltd., London.
Balfour I. (2000) Famous Diamonds, 4th ed. Christie, Manson
and Woods Ltd., London.
Bauer M. (1904) Precious Stones, translated by L.J. Spenser.
Charles Griffin & Company Ltd., London.
Briddon P.R., Jones R. (1993) Theory of impurities in diamond.
Physica B, Vol. 185, pp. 179189.
Bursill L.A., Glaisher R.W. (1985) Aggregation and dissolution of
small and extended defect structures in Type Ia diamond.
American Mineralogist, Vol. 70, pp. 608618.
Cassedanne J.P. (1989) Diamonds in Brazil. Mineralogical Record,
Vol. 20, No. 5, pp. 325336.
Clark C.D., Ditchburn R.W., Dyer H.B. (1956a) The absorption
spectra of natural and irradiated diamonds. Proceedings of the
Royal Society, Vol. A234, pp. 363381.
Clark C.D., Ditchburn R.W., Dyer H.B. (1956b) The absorption
spectra of irradiated diamonds after heat treatment.
Proceedings of the Royal Society, Vol. A237, pp. 7589.
Collins A.T. (1978) Investigating artificially colored diamonds.
Nature, Vol. 273, No. 5664, pp. 654655.
Collins A.T. (1980) Vacancy enhanced aggregation of nitrogen in
diamond. Journal of Physics C: Solid State Physics, Vol. 13,
pp. 26412650.
Collins A.T. (1982a) Colour centres in diamond. Journal of
Gemmology, Vol. 18, No. 1, pp. 3775.
Collins A.T. (1982b) A spectroscopic survey of naturally-occurring vacancy-related colour centres in diamond. Journal of
Physics D: Applied Physics, Vol. 15, pp. 14311438.
Collins A.T. (1984) Pitfalls in color grading diamonds by
machine. Gems & Gemology, Vol. 20, No. 1, pp. 1421.
Collins A.T. (2001) The colour of diamond and how it may be
changed. Journal of Gemmology, Vol. 27, No. 6, pp. 341359.
Collins A.T., Davies G., Woods G.S. (1986) Spectroscopic studies
of the H1b and H1c absorption lines in irradiated, annealed
type-Ia diamonds. Journal of Physics C: Solid State Physics,
Vol. 19, pp. 39333944.
Copeland L.L., Martin J.G.M. (1974) Diamonds: Famous,
Notable and Unique, revised by R.A.P. Gaal and J. Taylor.
Gemological Institute of America, Santa Monica, CA.
Crowningshield G.R. (19578) Spectroscopic recognition of yellow bombarded diamonds and bibliography of diamond treatment. Gems & Gemology, Vol. 9, No. 4, pp. 99104, 117.
Crowningshield G.R. (1959) Highlights at the Gem Trade Lab in
New York. Gems & Gemology, Vol. 9, No. 9, p. 269.
Crowningshield G.R. (1965) Developments and highlights at the
Gem Trade Lab in New York: Clouds in natural-color brown
diamonds. Gems & Gemology, Vol. 11, No. 9, pp. 269270.
Crowningshield G.R. (1994) Gem Trade Lab Notes: Characteristic
inclusions in fancy-color diamonds. Gems & Gemology, Vol.
30, No. 1, pp. 4142.
Davies G. (1972) The effect of nitrogen impurity on the annealing
of radiation damage in diamond. Journal of Physics C: Solid
State Physics, Vol. 5, pp. 25342542.

114

YELLOW DIAMONDS

Davies G. (1981) The origin of the N2 absorption band in natural yellow diamonds. Portugaliae Physica, Vol. 13, No. 12,
pp. 241261.
Davies G. (1994) Properties and Growth of Diamond. Institution
of Electrical Engineers, London.
Davies G., Summersgill I. (1973) Nitrogen dependent optical
properties of diamond. Diamond Research 1973, pp. 615.
Davies G., Welbourn C.M., Loubser J.H.N. (1978) Optical and
electron paramagnetic effects in yellow Type Ia diamonds.
Diamond Research 1978, pp. 2330.
De Weerdt F., Van Royen J. (2001) Defects in coloured natural
diamonds. Diamond and Related Materials, Vol. 10, pp.
474479.
Dyer H.B., Matthews I.G. (1957) The fluorescence of diamond.
Proceedings of the Royal Society, Vol. A243, pp. 320335.
Dyer H.B., Raal F.A., Du Preez L., Loubser J.H.N. (1965) Optical
absorption features associated with paramagnetic nitrogen in
diamond. Philosophical Magazine, Vol. 11, No. 112, pp.
763773.
Field J.E. (1992) The Properties of Natural and Synthetic
Diamonds. Academic Press, London.
Fisher D., Spits R.A. (2000) Spectroscopic evidence of GE POL
HPHT-treated natural type IIa diamonds. Gems & Gemology,
Vol. 36, No. 1, pp. 4249.
Fritsch E. (1998) The nature of color in diamonds. In G. Harlow,
Ed., The Nature of Diamonds, Cambridge University Press,
Cambridge, UK.
Fritsch E., Scarratt K. (1992) Natural-color nonconductive gray-toblue diamonds. Gems & Gemology, Vol. 28, No. 1, pp. 3542.
GIA Diamond Dictionary (1993) Gemological Institute of
America, Santa Monica, CA.
Gleason B. (1985) Notable Diamonds of the World. Diamond
Promotion Service, New York.
Harris J.W., Hawthorne J.B., Oosterveld M.M. (1979) Regional
and local variations in the characteristics of diamonds from
southern Africa kimberlites. In F.R. Boyd and H.O.A. Meyer,
Eds., Kimberlite, Diatremes and Diamonds: Their Geology,
Petrology, and Geochemistry, Vol. 1, Proceedings of the
Second International Kimberlite Conference, American
Geophysical Union, Washington. D.C., pp. 2741.
Haske M.D. (2000) Yellow diamonds. The Guide, Vol. 19, Issue
5, Part 1, pp. 810.
Hofer S.C. (1998) Collecting and Classifying Coloured
DiamondsAn Illustrated History of the Aurora Collection.
Ashland Press, New York.
Janse A.J.A. (1995) A history of diamond sources in Africa: Part 1.
Gems & Gemology, Vol. 31, No. 4, pp. 228255.
Kaminsky F.V., Khachatryan G.K. (2001) Characteristics of nitrogen and other impurities in diamond, as revealed by infrared
absorption data. The Canadian Mineralogist, Vol. 39, pp.
17331745.
Kerr W.C. (1982) A report on the new Watermeyer split-facet diamond cuts. Gems & Gemology, Vol. 18, No. 3, pp. 154159.
Kiflawi I., Bruley J., Luyten W., VanTendeloo G. (1998) Natural
and man-made platelets in type Ia diamonds. Philosophical
Magazine B, Vol. 78, No. 3, pp. 299314.
King J.M., Moses T.M., Shigley J.E., Liu Y. (1994) Color grading of
colored diamonds at the GIA Gem Trade Laboratory. Gems &
Gemology, Vol. 30, No. 4, pp. 220242.
King J.M., Moses T.M., Shigley J.E., Welbourn C.M., Lawson S.C.,
Cooper M. (1998) Characterizing natural-color type IIb blue
diamonds. Gems & Gemology, Vol. 34, No. 4, pp. 246268.
King J.M., Shigley J.E., Guhin S.S., Gelb T.H., Hall M. (2002)
Characterization and grading of natural-color pink diamonds.
Gems & Gemology, Vol. 38, No. 2, pp. 128147.
King J.M., Shigley J.E. (2003) An important exhibit of seven rare
gem diamonds. Gems & Gemology, Vol. 39, No. 2, pp.
136143.

GEMS & GEMOLOGY

SUMMER 2005

Koivula J.I. (2000) MicroWorld of Diamonds: A Visual Reference.


Gemworld International, Northbrook, Illinois, 157 pp.
Liddicoat R.T. (1976) Developments and Highlights at GIAs Lab
in Santa Monica: A true canary. Gems & Gemology, Vol. 15,
No. 8, p. 235.
Mani A. (1944) The fluorescence and absorption spectrum of diamond in the visible region. Proceedings of the Indian
Academy of Sciences, Vol. A19, pp. 231252, plates 813.
Mawe J. (1813) A Treatise on Diamonds and Precious Stones.
Longman, Hurst, Rees, Orme and Brown, London.
Meyer H.O.A. (1987) Inclusions in diamond. In P.H. Nixon, Ed.,
Mantle Xenoliths, New York, John Wiley and Sons, pp.
501523.
Mita Y. (1996) Change of absorption spectra in type Ib diamond
with heavy neutron irradiation. Physical Review B, Vol. 53,
No. 17, pp. 1136011364.
National Bureau of Standards (1976) Color: Universal Language
and Dictionary of Names. NBS Special Publication 440, S.D.
Catalog No. C13.10:440, U.S. Government Printing Office,
Washington D.C.
Nayar P.G.N. (1941) The luminescence, absorption and scattering
of light in diamonds: Part III. Absorption. Proceedings of the
Indian Academy of Sciences, Vol. A14, pp. 117, plates 13.
Overton T.W. (2002) Legal protection for proprietary diamond
cuts. Gems & Gemology, Vol. 38, No. 4, pp. 310325.
Robertson R., Fox J.J., Martin A.E. (1934) Two types of diamond.
Philosophical Transactions of the Royal Society of London,
Vol. A232, pp. 463535.
Sato K., Sunagawa I. (1982) Quantitative evaluation of colour of
diamonds by spectrophotometric method. Journal of the Gemmological Society of Japan, Vol. 9, No. 4, pp. 317 [in English;

pp. 87101 in Japanese].


Scarratt K. (1979) Investigating the visible spectra of coloured diamonds. Journal of Gemmology, Vol. 16, No. 7, pp. 433447.
Scarratt K. (1982) The identification of artificial coloration in diamond. Gems & Gemology, Vol. 18, No. 2, pp. 7278.
Tavernier J-B. (1676) Travels in India. Translated by V. Ball, W.
Crooke, Ed., republished in 1977 by Oriental Books Reprint
Corp., New Delhi, India.
Vendrell-Saz M., Nogus-Carulla J.M., Mones-Roberdeau L.,
Bosch-Figueroa J.M. (1980) Transmission et absorption des
diamantes taills de la srie cape. Revue de Gemmologie
a.f.g., Vol. 64, pp. 2024.
Wade F.B. (1920) A few words on fancy colored diamonds. The
Jewelers Circular, Vol. 80, No. 1, pp. 187189.
Wang W., Mayerson W. (2002) Symmetrical clouds in diamond:
The hydrogen connection. Journal of Gemmology, Vol. 28,
No. 3, pp. 143152.
Wilks J., Wilks E. (1994) Properties and Applications of Diamond.
Butterworth Heinemann Ltd., Oxford, UK.
Woods G.S. (1984) Infrared absorption studies of the annealing of
irradiated diamonds. Philosophical Magazine B, Vol. 50, pp.
673688.
Woods G.S., Collins A.T. (1982) The 1450 cm1 infrared absorption in annealed, electron-irradiated type Ia diamonds.
Journal of Physics C: Solid State Physics, Vol. 15, pp.
L949L952.
Woods G.S., Collins A.T. (1986) New developments in spectroscopic methods for detecting artificially coloured diamonds.
Journal of Gemmology, Vol. 20, No. 2, pp. 7582.
Zaitsev A.M. (2001) Optical Properties of Diamond: A Data
Handbook. Springer Verlag, Berlin.

More than 30 years of cutting-edge synthetic


diamond research by leading gemological
researchers and producers
Editorial commentary by Dr. James Shigley
of GIA Research
Insights on the past, present, and future of
synthetic diamonds, and their effects on
the gem and jewelry industry

50 separate entries comprising more than


300 pages of material
Includes two wall charts in a sturdy
bound-in pouch: the Separation of
Synthetic and Natural Diamonds (1995),
and Characteristics of HPHT-Grown
Synthetic Diamonds (2004)
Softbound in an attractive slipcase

$49.95
(plus shipping and handling)

To order your copy today, visit


GIA Gem Instruments & Books at
www.gia.edu
Some of the material in SYNTHETIC DIAMONDS is from Gems & Gemology issues that are
long out of print. That means you won't find this information anywhere else.

YELLOW DIAMONDS

GEMS & GEMOLOGY

SUMMER 2005

115

EMERALDS FROM THE


KAFUBU AREA, ZAMBIA
J. C. (Hanco) Zwaan, Antonn V. Seifert, Stanislav Vrna, Brendan M. Laurs, Bjrn Anckar, William B. (Skip) Simmons,
Alexander U. Falster, Wim J. Lustenhouwer, Sam Muhlmeister, John I. Koivula, and Hja Garcia-Guillerminet

Zambia is considered the worlds second most important source of emeralds by value (after
Colombia). The deposits are located near the Kafubu River in the Ndola Rural Restricted Area.
Emeralds have been known from this region since 1928, but significant commercial production
began only in the 1970s. As of mid-2004, most of the emeralds were being mined from large openpit operations at the Kagem, Grizzly, and Chantete concessions. Economic emerald mineralization
is confined almost entirely to phlogopite reaction zones adjacent to Be-bearing quartz-tourmaline
veins that metasomatically altered Cr-bearing metabasite host rocks. Nearly all of the rough is cut
in India and Israel. Zambian emeralds have relatively high R.I. and S.G. values, with inclusions typically consisting of partially healed fissures, as well as actinolite, phlogopite, dravite, fluorapatite,
magnetite, and hematite. They contain moderate amounts of Cr, Mg, and Na, moderate-to-high Fe
contents, and relatively high Cs and Li. Although many features of Zambian emeralds are comparable to those from other commercially important localities, in many cases they may be separated by
a combination of their physical properties, microscopic characteristics, and chemical composition.

ambia is one of the worlds most significant


sources of fine-quality emerald and has been
called the second most important producer
by value after Colombia (see, e.g., Suwa, 1994;
Giuliani et al., 1998). For years, the mines in the
Kafubu area (near the Kafubu River in the Ndola
Rural Restricted Area) have produced emeralds of
uniform color and size, which supported the cutting of vast quantities of calibrated goods. In addition, relatively large, fine stones are occasionally
found (see, e.g., figure 1).
Several articles have been published on
Zambian emeralds, especially the geology of the
Kafubu area, physical properties, and inclusions
(see, e.g., Koivula, 1982; Graziani et al., 1983;
Milisenda et al., 1999; Seifert et al., 2004c). The
present article provides an overview on various
aspects of these emeralds, and includes updated
information on the geology of the area, mining and
production, and gemological properties. Much of

116

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

the historic and geologic information was obtained


during several months of fieldwork in 2001 by two
of the authors (AVS and SV), who studied all the
major Zambian emerald deposits and several minor
ones (Seifert et al., 2004b,c). In July-August 2004,
these authors also examined a new emerald area in
the Musakashi area of north-central Zambia (see
box A). Other authors (JCZ in 1995 and BA since
2002) have visited the deposits, and in September
2004 BA collaborated with authors JCZ, BML, and
WBS on visits to four mines (Chantete, Grizzly,
Pirala, and Twampane). Subsequently, JCZ visited
Ramat Gan, Israel, to learn more about the production and distribution of Zambian emeralds.

See end of article for About the Authors and Acknowledgments.


GEMS & GEMOLOGY, Vol. 41, No. 2, pp. 116148.
2005 Gemological Institute of America

GEMS & GEMOLOGY

SUMMER 2005

Figure 1. Zambia has


been an important
source of emeralds since
the 1980s. Much of the
production is polished
into beads, such as those
in these strands (maximum bead diameter is
16.2 mm). Relatively
large, fine-quality
Zambian emeralds also
are faceted, such as the
center stone (10.42 ct)
in this ring. Courtesy
of Pioneer Gems, New
York; photo Harold &
Erica Van Pelt.

HISTORY
According to Sliwa and Nguluwe (1984), beryl mineralization was first discovered in the Kafubu area (at a
locality that later became known as the Miku mine)
in 1928 by geologists working for the Rhodesia Congo
Border Concession Co. Although initial investigations
did not reveal good-quality gems, Rhokana Co. and
Rio Tinto Mineral Search of Africa continued smallscale exploration work into the 1940s and 50s. In
1966, the claim was passed to Miku Enterprises Ltd.,
and in 1971 the rights to the Miku area were taken
over by Mindeco Ltd., a government-owned company.
The region was subsequently mapped and the Miku
deposit verified by Zambias Geological Survey
Department (Hickman 1972, 1973).
During the 1970s, when local miners discovered
several more deposits, the Kafubu area became a
major producer of good-quality emeralds. Due to the
significant economic potential and extensive illegal
mining, the government established a restricted
zone and forcibly removed the population of this
sparsely inhabited area.
In 1980, a new government-controlled agency, the
Reserved Minerals Corp., took over the major
deposits and prospecting rights to the surrounding
region (Sliwa and Nguluwe, 1984). Kagem Mining
Ltd. (owned 55% by Reserved Minerals and 45% by
Hagura, an Indian-Israeli corporation) was authorized
to conduct exploration and mining in the Kafubu
area. A privatization agreement was signed between
Hagura and the Government of Zambia in May 2001,

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

and the transfer of shares recently was completed by


the government (Govind Gupta, pers. comm., 2005).
Outside the Kagem properties, which lie on the
north side of the Kafubu River, the emerald area has
been subdivided into nearly 500 prospecting plots.
However, many of these claims were established
without the benefit of a thorough geologic evaluation. Small-scale mining currently takes place on
dozens of claims, whereas mechanized activity is
mostly concentrated on the Kagem, Grizzly,
Chantete, and Kamakanga properties.

LOCATION AND ACCESS


The emerald mines are located in the Kafubu area
(also known as the Ndola Rural Restricted Area) of
central Zambia, about 45 km southwest of the town
of Kitwe (figure 2). This region lies in the southern
part of an important copper mining area known as
the Copperbelt (Coats et al., 2001). The emerald
deposits are distributed over ~200 km 2 within
13021311S latitude and 28032811E longitude, on both sides of the Kafubu River.
From Kitwe, the Kafubu area is accessed by a 15
km paved road to Kalulushi, and then by 30 km of
poorly maintained gravel road. The principal mining localities are reached by several dirt tracks, most
of which remain passable throughout the year.
The Kafubu River, a western tributary of the
Kafue River, drains the area together with small
perennial streams. Except for the quartzite ridges in
the southeast, there are no prominent topographic

GEMS & GEMOLOGY

SUMMER 2005

117

Figure 2. The Kafubu


emerald mines are located in north-central
Zambia, 45 km southwest of Kitwe, the nearest town. A relatively
new emerald occurrence
was discovered in the
Musakashi area to the
northwest of Kafubu in
2002.

features and the entire area is typically flat. The average altitude is around 1,200 m above sea level. Over
much of the area, the residual clay-rich soils are yellow-brown or reddish brown with extensive crusts of
laterite, and support relatively thick vegetation.
Access to the area is restricted, although the workers
settlements are located near the producing mines.

GEOLOGY
Regional Geology. The region encompassing the
Zambian Copperbelt and the Kafubu area comprises
a complex assemblage of geologic units (figure 3)
that evolved during three successive orogenies of
mostly Proterozoic age (i.e., mountain-building
events ranging from about 2 billion to 500 million
years ago [My]). The emerald deposits are hosted by
metamorphic rocks of the Muva Supergroup (Daly
and Unrug, 1983) that overlay the basement granite
gneisses along a structural unconformity. The Muva
rocks consist of quartzites, mica schists, and metabasites. Emerald mineralization is hosted by the
metabasites, which consist of talc-chlorite-actinolite magnetite schists (Hickman 1973; Sliwa and
Nguluwe, 1984). These schists are thought to represent metamorphosed volcanic rocks that were dom-

118

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

inated by komatiites (i.e., highly magnesian ultramafic rocks; Seifert et al., 2004c). Their high
chromium content provides a necessary component
for emerald mineralization.
In the Kafubu area, thick layers (up to 200 m) of
the metabasite are intercalated in the mica schist
quartzite sequence. Deposition of the Muva
Supergroup is dated to ~1,700 My. Subsequent folding and metamorphism, which also involved the
basement granite gneisses, took place during the
Irumide orogeny (~1,010 My; De Waele et al., 2002).
Importantly, with the exception of the Kafubu area,
the metabasites are unknown in other portions of the
1,000-km-long Irumide belt of northeastern Zambia
(see, e.g., Daly and Unrug, 1983; De Waele and
Mapani, 2002).
The basement granite gneisses and the Muva
Supergroup were later buried under sediments of
the Katanga Supergroup during the Neoproterozoic
era (i.e., 570900 My). The entire crustal domain
then underwent folding, thrusting, and metamorphism during the Pan-African orogeny, culminating
at ~530 My (John et al., 2004). This tectonic event
produced the most observable deformation and
metamorphic features in the Muva rocks of the
Kafubu area (Hickman, 1973).

GEMS & GEMOLOGY

SUMMER 2005

Figure 3. This simplified geologic


sketch map of the Kafubu emerald
area, modified after Hickman (1973)
and Sliwa and Nguluwe (1984),
shows the distribution of major rock
types and the main emerald deposits
in the area. The extent of the Kafubu
pegmatite field is based on field
observations of mining pits visited by
two of the authors (AVS and SV).

During late stages of the Pan-African orogeny,


rare-element pegmatites and some beryllium-rich
granites intruded various crustal units in central,
eastern, and possibly also northwestern parts of

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

Zambia (see, e.g., Cosi et al., 1992; Parkin, 2000). In


the Kafubu area, field studies at numerous mines
and exploration pits indicate the existence of a
major field of beryllium-bearing pegmatites and

GEMS & GEMOLOGY

SUMMER 2005

119

BOX A: EMERALD FROM THE MUSAKASHI AREA,


NORTH-WESTERN PROVINCE, ZAMBIA
By Antonn V. Seifert, Stanislav Vrna, Bjrn Anckar, and Jaroslav Hyrsl

Figure A-1. The abandoned open-pit Hope mine in the


Musakashi area reportedly was a source of emeralds.
Photo by A. V. Seifert.
Figure A-2. This emerald fragment (8 6 mm) from the
Musakashi area shows good color and transparent
` `cek.
areas between fractures. Photo by V. Z

120

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

A minor emerald rush occurred in 2002 at Chief


Mujimanzovus village in the Musakashi area, Solwezi
district, central Zambia. In August 2002, one of the
authors (BA) saw some high-quality emeralds with some
Senegalese dealers in Kitwe. The color of these emeralds
was significantly different from those of the Kafubu area,
showing an intense bluish green reminiscent of emeralds
from Muzo, Colombia. A field visit to their reported
source was arranged with a Senegalese dealer, but the
miners refused to grant access due to fierce disputes over
the property. Visits to nearby outcrops and artisanal
workings (mined for rock crystal quartz) showed the presence of abundant hydrothermal veins in the area. The
European Unions Mining Sector Diversification
Programme subsequently sponsored the Geological
Survey Department of Zambia to map and explore the
area, with field work undertaken in June 2004. By this
time, it was reported that there was very little activity
and the locals thought the deposit was mined out (Nguni
and Mwamba, 2004). No in-situ emeralds were found, but
small emerald fragments were recovered from eluvial lateritic soil adjacent to quartz veins at two mines (Hope
and Musakashi) in the Chief Mujimanzovu village area.
When two of the authors (AVS and SV) visited the
Musakashi area in July-August 2004, the pits were inactive (see, e.g., figure A-1), and data on the production
and quality of the emeralds proved elusive to obtain.
Nevertheless, a few emerald fragments were obtained
for gemological examination and chemical analyses
(see, e.g., figure A-2).
The gemological properties of three irregular polished
emerald fragments (up to 6 mm in maximum dimension)
were obtained by one of us (JH) using standard instruments and techniques. Refractive indices measured on
two of the samples were 1.5801.587, yielding a birefringence of 0.007. The specific gravity could not be measured accurately, due to abundant fissures in available
stones. The dichroism was blue-green and yellow-green.
The samples were inert to UV radiation. They had a typical chromium-type absorption spectrum and appeared red
with the Chelsea filter. The most interesting characteristic was the presence in all the stones of three-phase inclusions, consisting of a bubble and a cube-shaped crystal in
a liquid (figure A-3), almost identical to those commonly
seen in Colombian emeralds. Note that these inclusions
were very tinyup to 0.1 mm but usually much smaller.
Mineral inclusions in these samples were examined
by AVS and SV, and identified by electron microprobe.
They consisted of minute crystals of sphene (titanite),

GEMS & GEMOLOGY

SUMMER 2005

hydrothermal veins that is nearly 20 km long.


This field overlaps major horizons of metabasites that are enriched in chromium, resulting
in emerald mineralization over a large area
(again, see figure 3). Potassium-argon dating of
muscovite from a pegmatite and an associated
quartz-tourmaline vein gave cooling ages of
447452 My (Seifert et al., 2004c). This corresponds to the approximate time of emerald
mineralization, when the rocks cooled below
350 50C (which is the approximate temperature at which muscovite becomes closed to
argon loss; see Viana et al., 2003).

Figure A-3. All three samples of emerald from the


Musakashi area contained minute three-phase inclusions that consisted of a bubble and a cube-shaped
crystal that were suspended within liquid. Photomicrograph by J. Hyrsl; width of view is 0.5 mm.

Fe-oxides, Na-K feldspar, and quartz, with zeolite aggregates occurring along fissures. In comparison, the Kafubu
emeralds typically contain phlogopite, actinolite, and
apatite. The difference in mineral inclusions is consistent
with field relations that suggest the two emerald areas are
geologically unrelated.
The chemical composition of the three Musakashi
samples is compared to Kafubu emeralds in table A-1.
The Musakashi analyses show significantly higher Cr and
V than those from Kafubu. By comparison, they also contain higher Cr and lower Mg than representative analyses
of dark green emeralds from Colombia.
Although none of the authors are aware of any faceted
emeralds from the Musakashi area, there appears to be
potential for the production of facetable material.
TABLE A-1. Representative electron-microprobe analyses
of Zambian emeralds from the Musakashi and Kafubu areas,
with comparison to Colombian emeralds.
Kafubu
Oxides (wt.%)
SiO261.965.4
TiO2
bdl
Al2O312.517.9
Cr2O3bdl0.84
V2O3bdl0.08
FeOtot0.061.75
MnObdl0.01
MgO0.272.90
CaObdl 0.12
Na2O0.161.99
K2O bdl0.27

Colombiaa

Musakashi

66.58
bdl
15.90
1.31
0.48
0.23
bdl
0.73
0.02
0.65
0.03

66.24
0.01
15.70
1.45
0.47
0.27
0.03
0.72
bdl
0.62
0.02

66.58
0.01
15.64
1.36
0.55
0.31
0.01
0.72
0.01
0.66
bdl

64.93
< 0.02
15.60
0.68
1.87
0.05
< 0.02
1.27
< 0.01
0.68
< 0.10

Representative dark green emeralds from La Pita (left column;


Fritsch et al., 2002) and Somondoco (right column; Kozlowski et
al.,1988). Abbreviation: bdl = below detection limit.

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

64.89
0.03
18.10
0.28
0.12
0.39
0.05
1.24
0.01
1.12
0.07

Local Geology. Emerald miners in the Kafubu


area use a local geologic vernacular that is summarized in the G&G Data Depository
(www.gia.edu/gemsandgemology). Knowledge of
this terminology is critical to understanding their
observations of the geology and emerald mineralization. The emerald mineralization is directly
related to the metasomatic alteration of the Crbearing metabasites by Be-bearing fluids derived
from hydrothermal veins (see, e.g., Coats et al.,
2001; Seifert et al., 2004c). For the most part, economic quantities of emerald are restricted to the
phlogopite reaction zones (typically 0.53 m
wide) between quartz-tourmaline veins and
metabasite. These reaction zones locally contain
aggregates of emerald (figure 4), of which a minor
proportion is gem quality. Only rarely are good
stones found in the quartz-tourmaline veins, or
very exceptionally in less altered, partially phlogopitized host rocks (e.g., at the Grizzly mine).
Localized phlogopite reaction zones in the
metabasites also were caused by emplacement
of simple quartz-feldspar pegmatites, which are
typically 210 m thick and steeply dipping. Field
observations by two of the authors (AVS and SV)
indicate that these pegmatites were emplaced
shortly after the quartz-tourmaline veins. Since
fluids from the pegmatite system contained
some Be, minor emerald mineralization occurs
locally in the phlogopite alteration zones associated with these pegmatites, too. The best emerald mineralization is found in phlogopite schist
near intersections between the quartz-tourmaline veins and the pegmatitesparticularly at
the intersection between steeply dipping pegmatites and flat-lying veins.
The abundance of quartz-tourmaline veins
with associated phlogopite reaction zones in the

GEMS & GEMOLOGY

SUMMER 2005

121

Figure 4. Emerald aggregates are locally found in


phlogopite schist, as shown in this specimen from
Kagems Fibolele mine. Only a small portion of the
emeralds found in these aggregates are gem quality.
` cek.
Photo by V. Z`

Kafubu area suggests a regional influx of


hydrothermal fluids containing Si, B, K, F, and other
elements (Seifert et al., 2004c). The Be-bearing fluids
that altered the metabasites, causing emerald mineralization, are most likely related to a hidden granitic
source. Information from a fluid inclusion study
(Seifert et al., 2004c) is consistent with the results of
the regional geology described above, which indicates
emerald mineralization occurred at temperatures of
360390C and pressures of 400450 MPa.

EXPLORATION
About 2% of the Kafubu area is currently being
mined (i.e., 5 km2), and exploration activities are
limited mostly to the immediate surroundings of
known deposits. The nearly omnipresent cover of
residual soil (210 m thick) is a serious obstacle
to prospecting.
Exploration methods range from witchcraft and
gut feel to the use of advanced geophysical methods, core drilling, and geologic mapping. Kagem has
several highly qualified geologists who methodically monitor exploration and mining activities.
Smaller companies rely on local consultants who
provide geologic interpretation and basic geophysical surveying (described below). Local miners, many
of whom have decades of experience working the
area, also are an important resource. These miners
(locally called sniffers) are experts at locating
emeralds through careful field observations (e.g.,
quartz-rich soil uplifted in tree roots).

122

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

The target ore geology at Kafubu is favorable for


magnetic surveying. This rather simple geophysical
technique has been used in the area for years, and
remains the main tool for emerald prospecting. A
magnetic survey measures variations in the earths
magnetic field intensity over an area of interest (see,
e.g., Cook, 1997). Ferromagnetic minerals and their
contrasting proportions in the earths crust produce
magnetic anomalies that can help define the underlying geology of the target area. However, the expression of magnetic anomalies is highly dependent on
the surrounding geology. For example, low-magnetic
sedimentary rocks together with high-magnetic
igneous or metamorphic rocks will typically result
in distinct anomaly patterns. In the Kafubu area, the
host metabasites are usually the only rocks with
high contents of magnetite, which is the primary
mineral causing magnetic-high readings. The metabasites are contained within generally low-magnetic
quartz-muscovite schists, so the surveys usually produce easily interpreted anomaly patterns.
In addition to magnetic surveying, radiometry
(see Cook, 1997) is often used at Kafubu. With this
method, a gamma ray spectrometer measures U,
Th, and K anomalies to distinguish different rocks
on the basis of their mineralogy; it may also reveal
geologic contacts and major fault structures.
However, the penetration depth of this technique is
very shallow, so only surface or very near-surface
anomalies can be detected. Nevertheless, this technique has successfully revealed underlying pegmatites and hydrothermal veins in the Kafubu area.
Combinations of magnetic highs and radiometric
highs indicate a very favorable geologic setting, and
such geophysical anomalies are tested by digging
pits to verify the presumed occurrence of metabasites and hydrothermal veins. When favorable conditions are encountered, local geologic indicators
may point toward areas with high potential for
emerald mineralization.
Geochemical soil and rock sampling has been
attempted in the area but has not been done systematically. The samples have been analyzed for
Ni, Cr, Co, and similar elements to look for metabasites, as well as for Be, Rb, Cs, and Li as pathfinders
for pegmatites and hydrothermal veins.
Although expensive and time consuming, oriented
drilling exploration programs may reveal important
new emerald accumulations. For example, an extensive drilling survey at a grid spacing of 25 mcarried
out in the southern continuation of Kagem pit D8 in
2003intersected a new and very promising emerald-

GEMS & GEMOLOGY

SUMMER 2005

Figure 5. The Grizzly


mine employs large
earth-moving equipment in a highly mechanized operation. The
excavators used in the
Kafubu emerald mines
typically weigh 2530
tons and have a bucket
capacity of approximately 1.6 m3. Photo
by B. M. Laurs.

bearing zone at a depth of 75 m. A similar survey in


2002 revealed additional emerald mineralization in
the southeastern continuation of the Fwaya-Fwaya
Ext. F10 mine (J. G. Dey, pers. comm., 2003).
So far, emerald mineralization has been mined to
a maximum depth of 5060 m (e.g., at the deepest
Grizzly, Kamakanga, and Kagem mines). However,
structural and lithologic criteria suggest that mineralization continues to deeper levels. Field surveys
and laboratory analyses (Seifert et al., 2004c), as well
as a study of the structural geology (Tembo et al.,
2000), have demonstrated that the potential for substantial reserves and new emerald occurrences in the
Kafubu area remains very high.

MINING
The Kafubu mining area (or Ndola Rural Restricted
Area) has been subdivided into several hundred small
concessions at around 100 hectares each. Most of
these concessions are located in areas with unpromising geology. However, many others are in favorable
areas, and a few of these have been amalgamated into
larger entities such as the one operated by Kagem.
The current license system has many disadvantages
(Zambia cranks up, 2004), as these 100-hectare
concessions are too small to support the investment
needed to start financially viable operations, but there
is no system in place to have such licenses relinquished and offered to capable mining companies.
Furthermore, the area is rife with disputes and
lengthy court cases that delay promising operations.

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

As of August 2004, the main emerald mining


activities took place at the Kagem, Grizzly, and
Chantete concessions. At Kagem, which at 46 km2
is by far the largest license area, producing mines
include Fwaya-Fwaya Ext. F10, Fwaya-Fwaya,
Fibolele, D8, and Dabwisa. The Grizzly mine
recently expanded its operations by acquiring
licenses to additional concessions and investing in
new machinery. The Chantete mine is also an
active producer with modern equipment. Other producers include the Kamakanga, Twampane, and
Akala mines, with sporadic production from the
Pirala, Miku, Ebenezer, and Mitondo mines.
In June 2004, United Kingdombased Gemfields
Resources Plc started systematic exploration at its
Plots 11A-1 and 11A-2 adjacent to the Pirala mine.
The company is now undertaking a full feasibility
study after having carried out drilling, ore body
modeling, and bulk sampling. The initial results are
promising, and the company expects to be in full
operation by late 2005.
All the Kafubu emerald deposits are worked by
open-pit mining. Because of abundant water during
the rainy season (November to March), underground
work is not considered an option; the groundwater
level is too high for sinking shafts, as it would be too
expensive to pump water continuously. Only one
operator, at the Mitondo mine, has tried underground
mining, but they recently ceased these activities.
Mining is done by removing the overburden rock
with bulldozers, excavators, and large dump trucks
(see, e.g., figure 5). At all the big pits, the miners drill

GEMS & GEMOLOGY

SUMMER 2005

123

The smaller operations do not use washing


plants due to a lack of funding and/or security. For
instance, at the Chantete mine, 24-hour-a-day shifts
keep the emerald-bearing rock moving, and all the
emeralds are sorted by hand. This is done throughout the year, except during the rainy season when
the pit fills with water.
Although it is expected that the Kafubu area will
be able to supply a large quantity of emeralds over
the next 20 years (A. Eshed, pers. comm., 2005),
there is growing concern about the expenses
involved in extracting them. For instance, at present
the cost of running the Kagem mines is estimated
to be around US$10,000 per day. The output is
always uncertain; due to the irregular distribution of
the emerald mineralization, it is extremely difficult
to estimate reserves realistically, which makes
emerald mining a high-risk business.

ENVIRONMENTAL IMPACT

Figure 6. At the Grizzly mine, workers drill holes that


will be filled with explosives to open up the area adjacent to an emerald-bearing zone. Photo by B. M. Laurs.

a series of holes (figure 6), so that explosives can be


used to open the areas adjacent to the veins. Once
emerald-bearing schist is exposed, mining is done
manually with hammer and chisel, by so-called
chiselers. The recovered emeralds are put into
cloth sacks or deposited into padlocked metal boxes.
Security is a major problem during the manual
extraction phase. The emerald-bearing zones must
be heavily guarded at night to prevent access by the
numerous illegal miners who become active after
sunset. It is also not uncommon for a chiseler to
cover a newly discovered emerald concentration for
later nighttime excavation.
Standard washing/screening plants are used to
process the ore at some of the mines. A new processing plant at the Grizzly mine was recently put
into operation, with a capacity of 20 tons per hour.
After crushing, the ore is washed and sized, with
larger and smaller pieces separated in a rotating
trommel. Additional vibrating screens further separate the material into specific size ranges for hand
sorting on slow-moving conveyor belts. At Kagem,
the operators rent the equipment used at the washing and sorting plant to keep costs down.

124

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

A detailed study of the environmental impact of


the emerald mines was performed by Seifert et al.
(2004b). The mining activities affect the landscape
and natural mediawater, soil, rocks, and airas
well as human health. Fortunately, there are no
enriched heavy metals or toxic elements in the
Kafubu area.
Various environmental interactions and impacts
are associated with each phase of a mines lifespan

Figure 7. This 40 g emerald crystal fragment from the


Chantete mine is illuminated from behind to show its
transparency and attractive color. Photo by H. Zwaan.

GEMS & GEMOLOGY

SUMMER 2005

sist throughout the year.

PRODUCTION AND DISTRIBUTION


Description of the Rough. The information in this
section was derived from communications with the
mine owners and from the personal experience of the
authors. Much of the emerald rough is recovered as
fragments that show only a few crystal faces (e.g., figure 7), but well-formed hexagonal crystals are sometimes produced (figure 8). Typically, though, these are
broken segments that rarely have natural flat (pinacoidal) terminations. Also seen are emerald clusters
or aggregates (figure 9), which may show well-formed
crystals with perfectly developed terminations, as
well as hexagonal prisms with irregular terminations.
The aggregates may contain several to dozens of individuals within the host schist. Step-growth crystal
surfaces are frequently present, caused by abrupt
changes in diameter. The surface quality of the crystal faces is often rough, but it may be glassy smooth.
The color of the emerald crystals ranges from
light bluish green to dark green. Very pale green or
blue (rarely colorless) beryl also is found. As with
emeralds from other localities, color zoning is common. Larger crystals typically have a core that is
light yellowish green to greenish blue surrounded

Figure 8. This prismatic single crystal of emerald


(slightly more than 8 cm long) was recovered from
the Chantete mine. Photo by Dirk van der Marel,
Naturalis, the Netherlands.

Figure 9. This cluster of well-formed emerald crystals


was found at the contact of a quartz vein with phlogopite schist at the Chantete mine. The width of this
specimen is 9 cm. Photo by Dirk van der Marel,
Naturalis, the Netherlands.

(i.e., prospecting, exploration, mine development,


mining, and closure). The most serious impacts
include deforestation and vegetation removal, land
degradation, increased soil erosion and siltation of
watercourses, habitat loss (resulting in a reduction
of biodiversity), and dangerous sites (e.g., pit walls).
The amount of mined material from the entire
Kafubu area that must be disposed of or stockpiled
is roughly estimated at approximately 25,000
tonnes per day at current levels of activity, with a
total of 70 million tonnes displaced since the
1970s.
The dozens of abandoned mines in the area do
not always have negative consequences. They are
filled by seasonal rainwater, and previously dry
bush can benefit from the presence of a reservoir by
creating new wetland ecological systems that per-

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

GEMS & GEMOLOGY

SUMMER 2005

125

by a deep green rim; other crystals may show both a


core and rim that are deep green, with internal
zones of light and/or medium green.
The crystals typically range from less than 1 mm
to several centimeters long, with pieces exceeding
10 cm encountered occasionally. The largest crystal
observed by the authors was about 14 cm long and
weighed over 3 kg (see, e.g., figure 25 inset in Laurs,
2004). According to a local mine manager, exceptionally large emerald crystals up to 60 cm long
were found in the phlogopite schist at the
Kamakanga Old Pit during the 1980s.
Production of the Rough. Emerald mineralization is
very irregular, with the crystals often aggregated
together (figure 4). These local concentrations may
have grades of several kilograms of emerald per ton
of ore rock. More typical is a dispersed mineralization of a few grams per ton of ore.
Run-of-mine (ROM) emerald typically yields
only a small percentage of material that can be
faceted. Most of it is of bead to cabochon quality.
According to information supplied by T. Schultz
(pers. comm., 2005) and Milisenda et al. (1999), a
typical 10 kg of ROM emerald from a favorable
deposit would yield about 5 g of extra-fine material,
about 100 g of fine material, about 300 g of good
material (in terms of both color and clarity), about
600800 g of material with good color but included,
about 3,000 g of low-quality material, and the
remaining 67 kg of very low commercial value.
At Kagem, an average production of 300 kg
(including low-grade emerald and beryl) every 23
months contains 10% facetable and cab-quality
emeralds, with the remainder usable for beads and
carvings. Typically, the emeralds occur as crystals
up to 60 g each, but occasionally crystals as large as
1 to 3 kg are found. However, these large specimens
usually contain only 12% cuttable material (A.
Eshed, pers. comm., 2005). In terms of quality, F10 is
currently the best-producing Kagem mine, with better (higher clarity) crystals than the former top producer, Fwaya-Fwaya (I. Eliezri, pers. comm., 2005).
The Fibolele and D8 mines generally produce
medium-quality stonesthat is, good colors, but
included, both facetable and cab quality. The
Dabwisa mine mainly produces cabochon material,
some of which is referred to as metallic green
(caused by rusty looking inclusions; A. Eshed,
pers. comm., 2005).
The Grizzly mine produces a quantity similar to
that from the Kagem mines but generally bigger

126

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

pieces (often 50 g to 1 kg), although most are of


medium quality. Large crystals also are found.
Production at the Chantete mine is typically
150300 kg per month, and large crystals are produced occasionally there as well. Only about 10% is
suitable for cutting, and just 1% is facet grade (see,
e.g., figure 7). Stones from this mine tend to be
somewhat lighter in color than those from elsewhere in the Kafubu area.
Production of rough reportedly exceeded US$100
million in the late 1980s (Milisenda et al., 1999).
More recently, the value of emerald production from
the Kafubu area was reported to be about US$20 million annuallyaccording to 2002 statistics from the
Export Board of Zambia for the officially declared
export value of rough emerald. A conservative estimate is that at least another US$5 million worth of
rough is smuggled out of Zambia (Douglas Ngandu,
DN Consulting Associates Ltd., Kitwe, pers. comm.,
2005), although some sources (mainly the Zambian
media) place the number much higher. However, it
is difficult to imagine where such large quantities of
emerald would originate, given that there are only
two large operators and the presence of secret
bonanza mines is not realistic.
Distribution of the Rough. Most of Zambias emeralds are exported to India, mainly for use in that
countrys domestic market, and to Israel for international distribution. According to D. Tank (pers.
comm., 2005), India receives 80% of Zambias
emerald production by weight, and 7075% by
value, with the cutting done in Jaipur. Israeli buyers
usually purchase the higher-quality material.
The larger mining operators have well-established trading arrangements. For example, Kagem
offers their production four times a year at a closedtender auction in Lusaka. The dates are flexible,
depending on when sufficient material is available.
The very small operators and illegal miners rely on
local traders, who are mainly of West African origin
(e.g., Senegal and Mali) and often supply them with
food and other necessities in exchange for emeralds.
The Informal Local Market. Rough emeralds are frequently traded on the streets of Lusaka and Kitwe,
usually as run-of-mine material sold by illegal miners. Occasionally, large top-quality emeralds also
turn up on the informal street market.
Buyers who visit Zambia should be aware of various scams. Green glass (stolen from local traffic
lights) is molded to simulate hexagonal emerald

GEMS & GEMOLOGY

SUMMER 2005

crystals, and quartz crystals are colored by green


marker pens. While these imitations are easy to
identify, there also are more clever imitations such
as pieces of Russian hydrothermal synthetic emerald that are carefully coated in clay and mica.
Production and Sorting of Cut Emeralds. Gemstar
Ltd., based in Ramat Gan, is the largest manufacturer
of Zambian emeralds for the international market; it
consumes most of the higher-quality rough material.
Gemstar produces 2,0003,000 carats per month of
cut emeralds, from approximately 15,000 carats of
rough. As an example of how the company approaches the rough from a commercial viewpoint, the following case was demonstrated to the senior author:
While evaluating a 312 ct piece of rough that was of
very good quality, the manufacturer decided that,
instead of cutting this piece into a few stones weighing 510 ct, he would aim for smaller stones of high
clarity, because there is strong demand for such emeralds. In the preform phase, 105 carats were recovered
from the piece of rough, from which 60 carats of
faceted gems (0.505 ct) were producedfor a yield of
only 20% of the original high-quality crystal. An
example was also provided for medium-quality (i.e.,
more included) rough: Four pieces weighing a total 84
carats had a yield of 20% after preforming and only
10% after faceting (including very small stones with
little commercial value). For cabochon-grade material,
the recovery is sometimes not more than 7%.
As explained by Daniel Madmon, manager of
Gemstars cutting factory, the typical cutting procedure is outlined below:

Figure 10. At Gemstar, after a piece of rough has been


assessed, it is typically sawn into several transparent
pieces for faceting. Photo by H. Zwaan.

Figure 11. The sawn pieces of emerald are evaluated by


the Robogem machine to predict the maximum yield
and preshape the rough. Photo courtesy of E. Eliezri.

1. Using a strong lamp, the cutter assesses how


and where to trim the rough, and the sawing is
done carefully by hand (figure 10).
2. The sawn pieces go to the office for evaluation,
and the cutting style is chosen according to
demand.
3. Next, the sawn pieces are sent to a computerized robot, called Robogem (figure 11).
Manufactured by Sarin Technologies Ltd., this
equipment uses the Sarin DiaExpert system (as
described by Caspi, 1997) to measure each
sawn piece and predict the best possible yield.
The robot is then used to make the girdle of
each stone.
4. Once returned to the factory, the emerald preforms are placed on a dop and faceted by hand
(figure 12), rather than by machine, because a
soft touch is needed for emeralds. First, the

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

GEMS & GEMOLOGY

SUMMER 2005

127

6. Once cutting and polishing is complete, the


dop is gently iced so that the emerald can be
easily removed. Subsequently, the finished
gem is cleaned in an alcohol solution.

Figure 12. After they have been preshaped by


Robogem, the emeralds are cut and polished by hand,
using precise cutting and polishing wheels and equipment. Photo by H. Zwaan.

table and crown facets are cut and (with the


girdle) polished.
5. The cutter then turns the stone on the dop and
cuts/polishes the pavilion side. Taking an
emerald cut as an example, the cutter places a
maximum of three steps on the pavilion to get
the best color, with more steps (up to four or
five) used if greater sparkle is the goal.

Figure 13. At Gemstar, emeralds cut in calibrated


sizes are sorted by color and clarity into numerous
categories. Photo by H. Zwaan.

Each step of the cutting process is monitored by


the master cutter, so that minor adjustments can
be made when necessary. After polishing, stones
with fissures are clarity enhanced by Gemstar
using paraffin oil (i.e., mineral oil), a near-colorless
transparent oil that the company has used for nearly 30 years. Gemstar declares at the time of sale
that its emeralds are enhanced in this way. Resins
like Opticon are never used.
Commercial production is done in calibrated
sizes, such as 5 3, 6 4, 7 5, 8 6, and 10 8
mm. Typically, these are divided equally between
emerald and oval cuts, although other shapes, such
as hearts and pears, also are produced.
Next, the stones are sorted according to color
and clarity into numerous categories (figure 13),
based on master sets. Gemstar keeps an inventory
of about 35,000 carats, so they can repeatedly produce the same colors, makes, and sizes. The steady
availability of calibrated sizes of consistent quality
is extremely important to customers, and the
Zambian emerald is prized for such consistency (A.
Eshed, pers. comm., 2005).
Commercial goods typically weigh up to 5 ct.
Zambian emeralds over 10 ct are rare (again, see figure 1). During a visit to Ramat Gan, the senior author
encountered a rare 14.29 ct medium-dark green stone
(figure 14) that was only slightly included.
Consuming Countries. As mentioned before, most
Zambian emeralds end up in India, where they are
cut and distributed, mainly to the domestic market.
In Israel, Gemstar Ltd. sells 70% of its stones to the
U.S., 15% to the Far East, and roughly 15% to
Europe. It is expected that in the future the growing
markets will be in China, Russia, and eastern
Europe.

MATERIALS AND METHODS


For this study, we examined a total of 127 emeralds
(0.075.69 ct), of which 69 were faceted and 2 were
cabochon cut. Almost all the 56 rough samples
were transparent and suitable for faceting. At least
two windows were polished on each rough sample;
the resulting pieces ranged from 0.46 to 6.98 ct. The
samples were obtained from several mines: 34 from

128

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

GEMS & GEMOLOGY

SUMMER 2005

Figure 14. This exceptional Zambian emerald weighs


14.29 ct. It shows a deep Zambian bluish green
color and is only slightly included. Courtesy of
Gemstar Ltd.; photo by H. Zwaan.

Figure 15. The emeralds studied for this report represented a range of color, as evident in these samples
from the Kagem mines (0.555.69 ct). Photo by Jeroen
Goud, Naturalis, the Netherlands.

the Chantete mine, 51 from the Kagem mines (specific pits not known), and one from the Kamakanga
mine. In addition, a thin section was cut from a
light bluish green emerald from the Mbuwa mine;
this emerald, which was not of gem quality, was
hosted by a quartz-tourmaline vein in phlogopitized
metabasite. The other 40 samples were from
unspecified mines in the Kafubu area.
All of the rough samples and 59 of the cut stones
(see, e.g., figure 15) were examined at the Netherlands Gemmological Laboratory (NGL). Twelve
faceted emeralds (0.224.28 ct) were examined at
GIA in Carlsbad; these were cut from rough produced at the Chantete mine during the 2004 mining
season (see, e.g., figure 16).
Standard gemological properties were obtained
on all the emeralds. We used a GIA GEM
Instruments Duplex II refractometer (at GIA) and a
Rayner refractometer with an yttrium aluminum
garnet prism (at NGL), both with near-sodium
equivalent light sources, to measure the refractive
indices and birefringence. Specific gravity was
determined by the hydrostatic method. We tested
the samples for fluorescence in a darkened room
with four-watt long- and short-wave UV lamps.
Absorption spectra of the stones examined at NGL
were observed using a System Eickhorst Modul 5
spectroscope with a built-in light source. Internal
features were observed with a standard gemological
microscope; in some cases, a polarizing (Leica
DMRP Research) microscope was used as well. In
29 samples, we analyzed inclusions with Raman
spectroscopy using 514 nm laser excitation and two
instruments: a Renishaw 2000 Ramascope (3 sam-

ples at GIA) and a Renishaw Invia (26 samples at


the CCIP French Gemmological Laboratory in
Paris). Polarized absorption spectra of 15 representative samples examined at NGL were taken with a
Unicam UV540 spectrometer, in the range of
280850 nm. Spectroscopy of the 12 samples studied at GIA was performed with a Hitachi U-4001
spectrophotometer (oriented UV-Vis spectra, range
280850 nm) and a Nicolet Magna-IR 760 spectrometer (mid-IR spectra, 400 6000 cm 1 ).
Additional mid-IR spectra were taken of 12 representative samples at NGL using a Thermo Nicolet
Nexus FT-IR-NIR spectrometer.
EDXRF spectroscopy was performed on the 12

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

Figure 16. Many Zambian emeralds are an attractive


saturated medium bluish green, as evident in this selection from the Chantete mine (0.661.35 ct). Courtesy of
Chantete Emerald Ltd.; photo GIA and Jeff Scovil.

GEMS & GEMOLOGY

SUMMER 2005

129

GIA samples with a Kevex Omicron instrument.


The instrument was operated at a voltage of 15 kV
without a filter for low-Z elements, with an aluminum filter at 20 kV for transition metals, and
with an iron filter at 25 kV for heavier elements. An
Eagle Probe EDXRF spectrometer was used at
NGL to analyze the composition of surface-reaching
inclusions on the pavilions of 16 faceted samples.
This instrument has a focus spot of 200 200 m
and a beam diameter of 300 m. A voltage of 25 kV
and a count time of 100 seconds were used for each
measurement.
Quantitative chemical analyses were carried
out on 27 selected emeralds from the Kagem and
Chantete mines at two electron microprobe facilitiesthe Institute of Earth Sciences, Free
University of Amsterdam, the Netherlands (JEOL

model JXA-8800M) and the University of New


Orleans, Louisiana (ARL SEMQ instrument). In
Amsterdam, 78 spot analyses were performed on
the table surfaces of 12 light to dark bluish green
faceted emeralds from the Kagem mines and 3 polished fragments from the Chantete mine. Twenty
analyses were done on surface-reaching inclusions.
Analyses were performed at 15 kV, with a beam
current of 25 nA and a spot size of 3 m. The
count time for the major elements was 25 seconds,
and for the trace elements, 36 seconds (except 50
seconds were counted for V, Cr, F, and Rb). From
two to 16 spots per sample were analyzed. In New
Orleans, the tables of 12 faceted Chantete emeralds were analyzed at 15 kV, with a beam current
of 15 nA and a spot size of 2 m, and a count time
of 45 seconds for each element. Three spots per

TABLE 1. Physical properties of emeralds from the Kafubu area, Zambia.a


Color
Clarity
Refractive indices

Birefringence
Optic character
Specific gravity
Pleochroism
Fluorescence
Chelsea filter reaction
Absorption spectrum
Internal features

Colors range from light to dark green to slightly bluish to bluish green; typically a saturated bluish green with
a medium to medium-dark tone
Very slightly to heavily included
no = 1.5851.599; ne = 1.5781.591
no = 1.5891.597; ne = 1.5811.589b
no = 1.602; ne = 1.592c
0.0060.009 (0.008b, 0.010c)
Uniaxial negative
2.712.78 (except one 0.22 ct sample, which gave a value of 2.81)
2.692.77b
Strong yellowish green (o-ray) and bluish green (e-ray); some stones showed strong greenish yellow (o-ray)
and greenish blue (e-ray)
Usually inert to long- and short-wave UV radiation; sometimes faint green to long-wave
No reaction or light pink to red (deep green samples)
Some absorption between 580 and 630 nm; distinct lines at approximately 636, 662, and 683 nm
Feathers in flat, curved, (rarely) conchoidal forms or undulatory/scalloped shapes
Partially healed fissures with various shapes of two- and three-phase fluid inclusions, but typically equant
or rectangular
Isolated negative crystals containing CO2 and CH4
Parallel oriented decrepitated inclusions appearing as silvery disks or brownish spots, depending on the
lighting
Mineral inclusions (this study): randomly oriented actinolite needles, platelets of phlogopite or rare chlorite,
equant to columnar dravite, fluorapatite, magnetite, hematite, chlorite, quartz, fluorite; carbonates (magnesite/
siderite, ferroan dolomite, ankerite and calcite); niobian rutile, pyrite, talc, zircon, barite, albite, calcite,
sphene (titanite), and beryl
Characteristic inclusions described by other authors: phlogopite/biotite, actinolite-tremolite, and squareand rectangular-shaped fluid inclusions (Milisenda et al., 1999); phlogopite, glauconite, talc, apatite, quartz,
and Fe-Mn and Fe-Cr oxides (Moroz and Eliezri, 1999); apatite, quartz, chrysoberyl, margarite, muscovite,
and rutile or brookite (Graziani et al., 1983); tourmaline, limonite, magnetite, mica, rutile, hematite, and
apatite (Koivula, 1982; 1984), and also chrysotile (Gbelin and Koivula, 1986)
Cavities, representing dissolved columnar mineral inclusions
Either homogeneous color distribution or medium to strong color zoning may occur (as planar zones
oriented parallel to the prism faces)

All data are from the present study unless otherwise noted.
Data from Milisenda et al. (1999), obtained from an unspecified number of samples that they described as representative. Other data on refractive
indices, birefringence, and specific gravity, such as reported by Graziani et al. (1983) on a single sample, and the average values given by Campbell
(1973) for Zambian emeralds, fall within the ranges that are indicated.
c
Data from Schmetzer and Bank (1981) on one dark bluish green sample.
b

130

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

GEMS & GEMOLOGY

SUMMER 2005

sample were analyzed.


In addition, three emerald fragments from the
Musakashi area (box A) were analyzed using a
Cameca SX 100 electron microprobe (15 kV, 40 nA)
at Masaryk University, Brno, Czech Republic.

GEMOLOGICAL PROPERTIES
The gemological properties are summarized in table
1, with details described below.
Visual Appearance. The emeralds showed a wide
variety of colors, ranging from green to slightly
bluish green to bluish green with a light to dark
tone (figure 15). Many were an attractive saturated
bluish green with a medium to medium-dark tone
(figure 16). Typically, the color was evenly distributed, although we saw strong color zoning in some
crystals and polished material. Color zoning in the
faceted stones was best seen from the side, when
looking parallel to the table through the pavilion.
Physical Properties. Roughly 70% of the stones tested showed refractive indices of no = 1.5911.595 and
ne = 1.5831.587. Some light green stones showed
values lower than 1.583, and some dark green
stones had R.I.s above 1.595. The birefringence of
most of the emeralds ranged between 0.007 and
0.008. Only one stone had a birefringence of 0.006,
and only two stones showed 0.009.
Specific gravity values ranged from 2.71 to 2.78.
The majority of the stones (78%) had an S.G.
between 2.72 and 2.76.
The emeralds were typically inert to long- and
short-wave UV radiation, except for a few samples
that fluoresced faint green to long-wave UV. All of
the emeralds with more saturated colors appeared
pink to red under the Chelsea filter; the other samples showed no reaction. Dichroism was strong, in

yellowish green and bluish green; or even stronger,


in greenish yellow and greenish blue. Most of the
emeralds, even the light-colored stones, showed
well-defined absorption spectra when viewed with
the handheld spectroscope. The spectra showed
some absorption between 580 and 630 nm, and distinct lines at approximately 636, 662, and 683 nm.
The violet range (beyond 430 nm) was completely
absorbed. A few light-colored emeralds showed a
weaker spectrum with only a clear line at 683 nm.
Microscopic Characteristics. The stones were very
slightly to heavily included. The most obvious clarity characteristics were fractures, partially healed fissures, fluid inclusions, needles, and occasional
brown flakes.
Fractures and Fluid Inclusions. The most conspicuous inclusions consisted of feathers and partially healed fractures, which both exhibited wide
variations in appearance. Feathers had flat,
curved, or rarely conchoidal forms with mirrorlike reflections, or undulatory/scalloped shapes
with a white appearance caused by rough surfaces.
Less commonly, the feathers were present in parallel formations or in tight clusters with no preferred orientation. One sample from Chantete
contained larger fractures that locally showed
blue flash-effect colors due to filling with a foreign organic substance.
Partially healed fractures were marked by planar
groups of equant, elongated, wispy, or irregularly
shaped fluid inclusions that often showed low relief
and contained relatively small bubbles (figure 17),
indicating that they are probably H2O rich (e.g.,
Samson et al., 2003). Partially healed fractures also
were represented by arrays of pinpoints that formed
parallel lines (see, e.g., figure 18) or fingerprints.

Figure 17. Irregularly


shaped (left) and wispy
(right) fluid inclusions,
such as those present in
these two emeralds from
the Chantete mine, were
widespread in the
Zambian emerald samples. Photomicrographs by
H. Zwaan; magnified 50
and 55, respectively.

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

GEMS & GEMOLOGY

SUMMER 2005

131

Figure 18. Parallel arrays of secondary fluid inclusions


were relatively common in the Chantete emeralds, as
were fractures of various sizes. Photomicrograph by J.
I. Koivula; magnified 10.

Many of the Zambian emeralds had healed fissures consisting of pseudosecondary rectangular to
square fluid inclusions (a hundred or more
microns) that each contained a bubble (figure 19).
These inclusions generally contained either two
phases (liquid and gas, mainly H2O [by inference]
and CO2 [identified by Raman analysis]) or three
phases (liquid, gas, solid), which often looked like
two-phase inclusions because the solids were only
clearly visible with crossed polarizing filters (figure
20). The solids were tentatively identified by optical means as carbonates. In heavily included emeralds, the healed fissures were often accompanied
by minute fluid inclusions, causing a milky
translucency. Fluid-filled tubes oriented parallel to
the c-axis were commonly seen in light-colored
emeralds (figure 21), but they were rare in the saturated medium to dark green stones.
Compared to the abundant pseudosecondary
fluid inclusions in healed fissures, primary inclusions were less common, and were not present in
every stone. They formed isolated negative crystals
with a roughly hexagonal outline and high relief in

Figure 19. Rectangular pseudosecondary two-phase


inclusions, typically present in healed fissures in
many Zambian emeralds, mainly contain H2O (liquid) and CO2 (gas). Photomicrograph by H. Zwaan;
magnified 100.

transmitted light (figure 22). They appeared to be


CO2-rich, each with a large bubble occupying most
of the available space. Raman analysis confirmed
that these inclusions contained CO2 as well as CH4
(an example spectrum is in the G&G Data Depository). Other isolated, irregular-shaped primary
three-phase fluid inclusions, containing an obvious
solid phase, were very rare (figure 23).
Parallel planes of decrepitated inclusions, containing remnants of fluid, were observed in a number of stones. When viewed with oblique illumination, these features only showed up clearly in a specific orientation when the light was properly reflected; in transmitted light, they appeared as parallel
faint brownish spots with a vague hexagonal outline
(figure 24).
Mineral Inclusions. Mineral inclusions were common in the lower-clarity Zambian emeralds.
Colorless needles with diamond-shaped cross sections were visible in many of the stones (figure 25)
and were identified as actinolite in five of the samFigure 20. Many rectangular three-phase inclusions
look like two-phase inclusions (left), because the
solid phase (carbonate)
only became apparent
when viewed with crossed
polarizers (right). Photomicrographs by H. Zwaan;
magnified 100.

132

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

GEMS & GEMOLOGY

SUMMER 2005

Figure 21. Parallel-oriented tubes commonly were present in the light-colored emeralds. In this particular
sample, the tubes are seen only in the light green portion. Photomicrograph by H. Zwaan; magnified 100.

ples by Raman analysis. Two microprobe analyses


of a surface-reaching inclusion gave Mg/(Mg + Fe2+)
ratios of 0.77 and 0.78, indicating a Mg-rich actinolite (per the classification of Leake et al., 1997). This
was similar to the composition of actinolite in the
host metabasites (ratio of 0.780.86; Seifert et al.,
2004c). Some of the actinolite needles were surrounded by material with a fuzzy appearance (probably microfractures; figure 26). The needles were typically straight, although a needle in one emerald
from the Chantete mine was obviously curved.
Also common were pale to moderate brown
platelets (figure 27) that provided Raman spectra
consistent with biotite/phlogopite. Microprobe analyses of eight grains in four different emerald samples
gave phlogopite compositions, which is in agreement with the compositions of micas in the altered
metabasite host rock as documented by Seifert et al.
(2004c) and with the composition of an inclusion in
a bluish green Zambian emerald given by Moroz
and Eliezri (1999). In a few stones that contained a
fair amount of phlogopite, dark green mica also was
found. One such inclusion (figure 28) was identified

Figure 22. Primary inclusions appear as small isolated negative crystals in the Zambian emeralds, with
high relief evident in transmitted light at 50 magnification. The curtain behind the two negative crystals is a partially healed fissure. Only under strong
magnification could a large bubble be observed
inside each cavity (see inset, magnified 250).
Photomicrographs by H. Zwaan.

as chlorite by Raman analysis.


Less abundant than actinolite or phlogopite were
high-relief, dark brown to black, equant to columnar crystals (figures 25 and 29). These inclusions
had a rounded triangular cross section that is typical
of tourmaline. Raman spectra of two such inclusions were consistent with schorl-dravite tourmaline. A semiquantitative EDXRF analysis of the
crystal in figure 29 revealed nearly twice as much
MgO as FeO, and little Ca, identifying it as dravite.
This is consistent with earlier studies of tourmaline
from the Kafubu area by Koivula (1984) and Seifert
et al. (2004c), who found Fe-rich or Ca-Fe rich
dravites, with individual crystals showing significant compositional variations.
Apatite was identified by Raman analysis in four
emeralds from the Kagem mining area. It usually
formed colorless euhedral crystals (figure 30) or
small colorless grains with irregular surfaces. One

Figure 23. This primary


three-phase fluid inclusion
was observed in a Chantete
emerald. Rotating the microscopes polarizer showed the
yellowish green (left) and
greenish blue (right) dichroic
colors of the emerald.
Photomicrographs by J. I.
Koivula; magnified 15.

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

GEMS & GEMOLOGY

SUMMER 2005

133

Figure 24. With oblique illumination, thin parallel planes of


decrepitated inclusions, with
fluid remnants, were commonly seen in the Zambian
emeralds (left). When viewed
in transmitted light, however,
they appeared as faint brownish features with a vague
hexagonal outline (right).
Photomicrographs by H.
Zwaan; magnified 50 (left)
and 200 (right).

heavily included dark bluish green emerald contained numerous hexagonal prisms of apatite, with
two crystals at the surface identified by microprobe
analysis as fluorapatite. Interestingly, this tiny
inclusion itself contained even smaller inclusions of
actinolite and magnetite (figure 31). Minute opaque
black crumbs and larger euhedral-to-anhedral
opaque grains, which formed skeletal features in
parallel planes (figure 32) and had a submetallic luster, were identified as magnetite by both Raman
and microprobe analyses. In a few emeralds, brownred inclusions with similar skeletal shapes were
identified by Raman analysis as hematite (again, see
figure 32). This suggests the presence of martite, a
variety of hematite that is a pseudomorph after
magnetite.
Cavities that appeared to represent casts of dissolved mineral inclusions with a euhedral columnar
habit were seen in a few samples (figure 33).
Figure 25. Colorless needles of actinolite in a Chantete
emerald appear brownish yellow in this image taken
between crossed polarizers. The dark grain between
the two actinolite needles in the center of the image is
probably tourmaline. A network of fractures also is
present in this view. Photomicrograph by J. I. Koivula;
magnified 13.

134

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

Small, slightly rounded, euhedral transparent


crystals were encountered in some emeralds. In most
cases, they were too deep within the stone to be identified by Raman analysis. One sample showed many
of these crystals, both isolated and in clusters. Some
crystals had very low relief and were doubly refractive, whereas others had higher relief and were
isotropic. At the surface, a few crystals of both types
were identified by microprobe as quartz and fluorite,
respectively (figure 34). In one sample, between a
quartz and a fluorite crystal, three carbonates also
were identified: a magnesite/siderite mixture, ferroan
dolomite, and ankerite. Interestingly, these carbonates contained a fair amount of Cr (e.g., up to 1.34
wt.% Cr2O3 in the ankerite). A small beryl inclusion
seen in association with these carbonates contained
less Mg and Fe than the emerald host. Raman analysis confirmed the presence of additional carbonate
inclusions in a few other stones.
Other rare inclusions identified in the Zambian
Figure 26. A fuzzy-looking white coating of micro
cracks and/or fluid inclusions hides the identity of the
rod-shaped solid inclusion inside this Chantete emerald. Similar coatings were seen associated with the
needles that were identified as actinolite in other samples. Photomicrograph by J. I. Koivula; magnified 15.

GEMS & GEMOLOGY

SUMMER 2005

Figure 27. Platelets of phlogopite, as seen here in one


of the Chantete emeralds, are quite common inclusions in Zambian emeralds. Photomicrograph by H.
Zwaan; magnified 80.

Figure 28. A few stones contained chloritelike the


crystal in the center hereas well as phlogopite.
Photomicrograph by H. Zwaan; magnified 80.

emeralds were pyrite and talc, present as fairly large


crystals in one heavily included sample; very small
grains of zircon, which were also observed in some
phlogopite inclusions, where they had caused dark
brown radiation halos; and minute grains of barite,
albite, and calcite. In one cut emerald from the
Kagem mining area, calcite was present in surfacereaching fractures. A small euhedral crystal of
sphene (titanite) was found in one faceted emerald;
its identity was confirmed by Raman analysis. All of

the other inclusions in this paragraph were identified


by a combination of visual appearance and EDXRF
spectroscopy, microprobe, or Raman analysis.
In the thin section of the Mbuwa mine emerald,
some reddish orange inclusions were identified as
niobian rutile (figure 35). Rutile inclusions in
Zambian emerald have previously been described
by various authors (e.g., Graziani et al., 1984;
Gbelin and Koivula, 1986; Seifert et al., 2004c), but
the reddish orange color in the thin section suggested the presence of Fe3+, Nb, and/or Ta. Raman analysis showed a rutile spectrum with fundamental

Figure 29. Equant black-appearing crystals of dravite


(Mg-bearing tourmaline) also were observed, as seen
here in an emerald from the Kagem mining area; the
crystal with the rounded triangular cross section
shows the typical crystal habit of tourmaline. Most of
the tourmaline crystals observed appeared to have
grown where actinolite needles intersect. Photomicrograph by H. Zwaan; magnified 80.

Figure 30. Large and small euhedral apatite crystals


are seen in this emerald from the Kagem mining area.
The apatite inclusions are transparent and show virtually no color. The size (1.0 mm long) of the large crystal is quite unusual; typically apatite seen in Zambian
emerald is smaller, like the crystal near the bottom of
this image, which is about 0.15 mm. Photomicrograph
by H. Zwaan; magnified 80.

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

GEMS & GEMOLOGY

SUMMER 2005

135

vibrations at 442 and 616 cm1, and a qualitative


EDXRF micro-analysis confirmed the abundance of
Ti, with lesser amounts of Nb and Fe, and some Ta.
Growth Features. Widespread in the Chantete
emeralds were parallel growth lines with a fine
lamellar appearance (again, see figure 32). When
observed between crossed polarizers, two samples
showed strain that was oriented parallel to the
growth lines; none of the other samples showed evidence of strain. In many of the samples, the color
was evenly distributed. Others exhibited moderate
to strong narrow zoning of straight, alternating light
green to medium-dark green bands (figure 36),
which was oriented parallel to the prism faces of the
crystals (figure 37). When this color zoning was present in faceted stones, it was often well disguised by
cutting the table of the emerald parallel to one of
the prismatic faces, so that the color zoning could
be observed only when looking at the stone from
the side, through the pavilion. In some cuttable
rough samples, hexagonal color zoning was prominent, often showing a darker green rim and a lighter
or darker green core, with alternating lighter and
darker green zones in between.
Figure 31. This backscattered electron image shows a
minute euhedral inclusion that was identified as fluorapatite by electron-microprobe analysis at the surface
of a heavily included emerald. The fluorapatite itself
contains small laths of actinolite (darker gray) and
magnetite (white). Micrograph by W. J. Lustenhouwer.

136

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

Figure 32. Magnetite in Kafubu emeralds occurs as


tiny opaque inclusions, sometimes in parallel planes
and showing skeletal shapes. The red-brown inclusions are hematite. Note also the fine-scale parallel
growth lines traversing the image. Photomicrograph
by H. Zwaan; magnified 65.

CHEMICAL COMPOSITION
Representative electron-microprobe analyses of
emeralds from the Chantete mine and Kagem mining area are shown in table 2 (see G&G Data
Depository for all analyses). In total, we obtained 78
microprobe analyses of 15 Chantete samples, and 36
analyses of 12 Kagem emeralds. Data from both mining areas were quite similar, although slightly higher

Figure 33. A few of the faceted Zambian emeralds


contained large, open-ended, sharp-edged columnar
cavities that have the appearance of dissolved actinolite inclusions. These cavities were partially filled
with epigenetic debris. Photomicrograph by H.
Zwaan; magnified 65.

GEMS & GEMOLOGY

SUMMER 2005

Figure 34. Inclusions showing


low and higher relief (left) were
identified as quartz and fluorite, respectively. (Photomicrograph by H. Zwaan; magnified
100.) In the backscattered
electron image on the right, the
fluorite crystals appear white,
while the quartz grains look
gray. Carbonates (light gray, see
arrow) also were present. Micrograph by W. J. Lustenhouwer;
width of view 1.0 mm.

Cr and Fe concentrations were measured in the


Kagem stones overall. In addition, the Kagem emeralds showed larger overall variations in the concentration of these elements, as well as in Mg and Na.
The Kagem emeralds were carefully selected to
include the lightest to darkest colors seen in commercially available cut stones (see, e.g., figure 15).
The color range of the faceted Chantete emeralds
tested was more restricted to a typically desirable
medium bluish green (see, e.g., figure 16), although
considerable color variations were present in one of
the rough gem-quality samples of Chantete emerald
(again, see figure 37). The most important chromophore in Zambian emeralds is Cr, which averaged 0.26 wt.%and ranged up to 0.84 wt.%
Cr2O3. Overall, the darkest green stones had significantly higher Cr contents than the lighter and medium green stones, which often showed no straightforward correlation between color intensity and
average Cr content. In contrast, V concentrations

Figure 35. Reddish orange inclusions of niobian rutile


were identified in a thin section of an emerald from
the Mbuwa mine. Photomicrograph by H. Zwaan;
width of view 1.8 mm.

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

were consistently low, averaging just 0.02 wt.%


V2O3 and attaining a maximum of 0.09 wt.% V2O3.
The only other significant chromophoric element
was Fe, which averaged 0.76 wt.% oxide (as FeO)
and showed a maximum of 1.75 wt.% FeO.
The emeralds contained relatively high concentrations of Mg (average of 1.90 wt.% MgO) and
somewhat less Na (average of 1.10 wt.% Na2O). We
detected trace amounts of Ca and K in many of the
samples. In the 12 Chantete emeralds analyzed at
the University of New Orleans, Ti and Mn were
near or below the detection limits in all stones. In
the 15 emeralds analyzed at the Free University of
Amsterdam, traces of Cs were detected in most samples, Sc was documented in some of them, and Rb
was below the detection limit in all the analyses.
Distinct compositional variations were recorded
in analyses of fine-scale color zones in a rough emerald from the Chantete mine (again, see figures 37

Figure 36. Although an even color appearance is very


common in Zambian emeralds, straight narrow color
zoning, as in this heavily included 5.69 ct cabochon,
was also seen in certain orientations.
Photomicrograph by H. Zwaan.

GEMS & GEMOLOGY

SUMMER 2005

137

Figure 37. On the left,


hexagonal color zoning
was observed parallel to
the prismatic crystal
faces in this 4.93 ct cuttable emerald crystal
from the Chantete mine
(no. r29, table 2).
(Photomicrograph by
H. Zwaan; magnified
10). The backscattered
electron image of this
same crystal, on the
right, also reveals finescale growth zoning as
well as the position of
the spots that were analyzed by electron microprobe. The whiteappearing grains are
phlogopite. Composite
micrograph by
W. J. Lustenhouwer.

and 38). The Cr2O3 for this emerald ranged from not
detectable in a narrow colorless zone to 0.46 wt.% in
the darkest green zone, with an average of 0.22
wt.%. As shown in figure 38, there was a good correlation between color and Cr content. Compared to
Cr, similar trends were shown by Fe, Mg, and Na,
and an inverse pattern was found for Al. The Cs content, which ranged from 0.05 to 0.23 wt.% Cs2O,
was highest in the outer portion of the crystal.
In addition to the elements detected by electron
microprobe, EDXRF analysis of the faceted
Chantete emeralds found traces of Ga and Zn in
separate emeralds, with Ga possibly present in an
additional sample. This technique, which is more
sensitive than the electron microprobe, detected
Mg, K, Ca, Cr, Fe, Rb, and Cs in all 12 of the samples analyzed. V also was detected in five samples,
and possibly in two others.

SPECTROSCOPY
Typical UV-Vis-NIR absorption spectra for Zambian emeralds are illustrated in figure 39. The ordinary ray (E | c) showed bands at 372, 440, 478, 610,
637, and 830, as well as a doublet at 680 and 683
nm. The bands at 440, 610, and 830 nm are broad
and the positions were estimated. The extraordinary
ray (E||c) displayed bands at 425 and 650 nm, as well
as absorptions at 632, 662, 684, and 830 nm.

138

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

Though the intensities varied, all the samples


showed the same bands for both the ordinary and
extraordinary rays. The bands at 440 and 610, and at
425 and 650 nm, and the peak at 684 nm, indicate
the presence of Cr3+, which causes the green color
(Wood and Nassau, 1968; Schmetzer et al., 1974).
Additional weaker peaks at 478, 662, 684, and the
doublet at 680 and 683 nm also are caused by the
presence of Cr3+. The shift in absorption from 440 to
425 nm and from 610 to 650 nm causes an absorption minimum at about 512 nm for the ordinary
ray, producing a yellowish green color, and a shift
toward a minimum at about 500 nm for the extraordinary ray, producing a bluish green color. The band
at 830 nm indicates the presence of Fe2+ (Platonov
et al., 1978), whereas the peak at 372 nm denotes
the presence of Fe3+ (Schmetzer et al., 1974). The
exact Fe2+/Fe3+ ratio cannot be determined by this
method.
The FTIR spectra of the emeralds contained the
typical features caused by vibration of H2O molecules (see, e.g., figure 40). The most obvious peaks,
at 7096 and 5265 cm1, are caused by type II H2O
molecules, which are present in the channels of the
crystal structure of beryl. These water molecules
are located adjacent to alkali-metal ions in the channels (Wood and Nassau, 1968), which in the
Zambian emeralds are mainly Na+, Cs+, and Li+.
The H-H direction of the type II H2O molecules is

GEMS & GEMOLOGY

SUMMER 2005

TABLE 2. Representative chemical compositions of Zambian emeralds by electron microprobe.a


Chantete mine
Sample
Weight (ct)
Colorb
Laboratory
Oxides (wt.%)
SiO2
TiO2
Al2O3
Cr2O3
V2O3
Sc2O3
BeO (calc.)
FeOtot
MnO
MgO
CaO
Na2O
K2O
Rb2O
Cs2O
Totale

0.95-2
0.95
Med. bG

0.66-2
0.66
Med. bG

3.47-2
3.47
Med. bG

UNOc

UNO

65.39
bdl
15.42
0.08
0.01
na
13.61
0.37
bdl
2.39
bdl
1.46
bdl
na
na
98.72

Ions based on 18 oxygens


Si
6.001
Ti
bdl
Al
1.668
Cr
0.006
V
0.001
Sc
na
Be
3.001
Fe
0.028
Mn
bdl
Mg
0.327
Ca
bdl
Na
0.260
K
bdl
Rb
na
Cs
na

Kagem mining area

UNO

r29-7
4.93
Colorless
FUAd

r29-12
4.93
Med.
bG
FUA

r29-5
z01r-1 z08b
z08d
z04c
4.93
0.55
0.91
0.91
0.66
Int. med. Lt. bG Med. sl. Med. sl. Med.
bG
bG
bG
bG
FUA
FUA
FUA
FUA
FUA

z03c
1.04
Med.
bG
FUA

z09c
z43a
0.89
0.98
Int. med. Dark bG
sl. bG
FUA
FUA

65.24
bdl
15.33
0.20
0.04
na
13.57
0.34
bdl
2.22
bdl
1.71
bdl
na
na
98.65

64.86
bdl
14.81
0.63
0.02
na
13.52
0.90
0.01
2.47
bdl
1.36
bdl
na
na
98.59

63.52
na
15.81
bdl
0.04
bdl
13.20
0.58
na
1.48
0.02
0.76
0.02
bdl
0.23
95.65

63.21
na
14.97
0.25
0.02
0.01
13.10
0.72
na
1.84
0.02
0.74
0.03
bdl
0.06
94.98

62.29
na
13.58
0.46
0.04
0.03
12.89
1.31
na
2.36
0.02
0.98
0.04
bdl
0.20
94.20

63.40
na
15.96
0.25
0.06
0.03
13.21
0.36
na
1.26
0.03
1.04
0.02
bdl
0.04
95.64

62.44
na
13.65
0.04
0.02
bdl
12.93
1.11
na
2.77
0.03
0.99
0.26
bdl
bdl
94.24

62.14
na
13.62
0.29
0.02
bdl
12.88
1.05
na
2.68
0.06
0.94
0.25
bdl
0.05
93.97

63.17
na
14.09
0.15
0.02
bdl
13.12
1.10
na
2.45
0.06
1.49
0.05
bdl
0.08
95.78

63.08
na
14.39
0.31
0.02
bdl
13.08
0.96
na
2.11
0.05
1.15
0.04
bdl
0.11
95.30

62.71
na
13.63
0.33
0.03
bdl
12.99
0.93
na
2.90
0.07
0.86
0.21
bdl
0.04
94.71

63.35
na
14.62
0.70
0.03
bdl
13.19
1.01
na
1.81
0.04
1.45
0.03
bdl
0.15
96.39

6.000
bdl
1.662
0.014
0.003
na
2.999
0.026
bdl
0.304
bdl
0.305
bdl
na
na

5.990
bdl
1.612
0.046
0.001
na
3.000
0.070
0.001
0.340
bdl
0.244
bdl
na
na

6.010
na
1.763
bdl
0.003
bdl
3.000
0.046
na
0.208
0.002
0.139
0.002
bdl
0.009

6.026
na
1.682
0.019
0.002
0.001
3.000
0.057
na
0.262
0.002
0.137
0.004
bdl
0.002

6.035
na
1.551
0.035
0.003
0.003
3.000
0.106
na
0.340
0.002
0.185
0.005
bdl
0.008

5.996
na
1.779
0.019
0.004
0.002
3.000
0.028
na
0.178
0.003
0.190
0.002
bdl
0.002

6.031
na
1.554
0.003
0.001
bdl
3.000
0.090
na
0.398
0.003
0.186
0.032
bdl
bdl

6.024
na
1.556
0.023
0.001
bdl
3.000
0.085
na
0.387
0.006
0.176
0.031
bdl
0.002

6.013
na
1.581
0.011
0.002
bdl
3.000
0.087
na
0.347
0.006
0.276
0.006
bdl
0.003

6.021
na
1.619
0.023
0.002
bdl
3.000
0.076
na
0.300
0.005
0.212
0.005
bdl
0.005

6.027
na
1.544
0.025
0.002
bdl
3.000
0.075
na
0.415
0.007
0.160
0.026
bdl
0.001

5.997
na
1.631
0.053
0.003
bdl
3.000
0.080
na
0.256
0.004
0.265
0.003
bdl
0.006

BeO was calculated based on an assumed stoichiometry of 3 Be atoms per formula unit. Abbreviations: bdl = below detection limit, bG = bluish green,
Int. = intense, Lt. = light, med. = medium, na = not analyzed, and sl = slightly.
b
Refers to overall color appearance, except for sample r29, in which specific color zones that correspond to each analysis are listed for this gem-quality
crystal (which had an overall color appearance of medium slightly bluish green). The relatively low values for Cr measured in some stones does not
appear to correlate with their bluish green color. Since the electron microprobe is a microbeam technique, the particular points analyzed on each of
those stones may not be representative of the bulk composition of those samples, particularly if color zoning is present.
c
Analyses performed at the University of New Orleans, Louisiana. Background counts were determined by a mean atomic number (MAN) method
(Donovan and Tingle, 1996). Analytical standards included both natural and synthetic materials: albite (Na), adularia (K), quartz and clinopyroxene (Mg,
Ca, Fe, and Ti), chromite (Cr), rhodonite (Mn), V2O5 (V), PbO (Pb), ZnO (Zn), Bi-germanate (Bi), and sillimanite (Si and Al). MAN standards in addition to
the above as appropriate: MgO, hematite, rutile, strontium sulfate, and ZrO2. Detection limits (in wt.%): TiO2 = 0.009, CaO = 0.008, MnO = 0.005, and
K2O = 0.012. Cl was analyzed but not detected.
d
Analyses performed at the Free University of Amsterdam, the Netherlands. Analytical standards included both natural and synthetic materials: diopside
(Si, Mg, and Ca), corundum (Al), fayalite (Fe), Sc2O3 (Sc), jadeite (Na), orthoclase (K), V2O5 (V), Cr2O3 (Cr), RbBr (Rb), Cs2 ReCl6 (Cs), fluorite (F), and
marialite (Cl). Detection limits (in wt.%): Cr2O3 = 0.018, Sc2O3 = 0.012, Cs2O = 0.028, and Rb2O = 0.029. Cl was analyzed but not detected. Low overall totals appear mainly due to low analytical SiO2 data.
e
Analyses do not include H2O. Data on Kafubu emeralds from Hickman (1972) and Banerjee (1995) showed 2.5 wt.% H2O, and 2.61 and 2.69 wt.%
H2O, respectively.

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

GEMS & GEMOLOGY

SUMMER 2005

139

for which spectra were taken, but the intensity varied between samples. Similar UV-Vis-NIR and
FTIR spectra were described by Milisenda et al.
(1999), and UV-Vis spectra of a light green emerald
from the Miku mine were illustrated by Schmetzer
and Bank (1981).
FTIR spectroscopy also is helpful to identify possible fillers used for clarity enhancement (see e.g.,
Johnson et al., 1999; Kiefert et al., 1999). None of the
spectra we obtained indicated the presence of an artificial resin in any of the stones examined. Also, we
did not observe any yellow or blue flash effects (a possible indication that an artificial resin might be present) in the fissures, except for an apparent local blue
flash effect observed in one stone from the Chantete

Figure 39. These polarized absorption spectra of an


emerald from the Chantete mine show bands that are
representative of all the Kafubu samples, and indicate
the presence of not only Cr3+ and Fe2+, but also Fe3+.
The intensities of the bands varied somewhat
between samples, but no straightforward relationship
could be established between band intensities and Cr
and Fe concentrations (or FeO/Cr2O3 ratios) measured
by electron microprobe. The spectra shown here were
collected on a 4.27 ct emerald cut that contained an
average of 0.17 wt.% Cr2O3 and 0.36 wt.% FeOtot.

Figure 38. Electron-microprobe analyses of the


Chantete emerald in figure 37 revealed systematic
variations in the composition of the various growth
zones. There was a good correlation between color
and Cr content, with similar trends shown by Fe, Mg,
and Na; the trends for all these elements were opposite that of Al. A significant increase in Cs was evident in the outer portion of the crystal.

perpendicular to the c-axis; the intensity of these


peaks was very strong in all the spectra we
obtained. Wood and Nassau (1967) demonstrated
that their intensity increases as the amount of alkalis increases. The broad band between roughly 3900
and 3400 cm1 is caused by type I and type II H2O
molecules. Type I H2O molecules are freely present
in the channels, without being linked to other ions,
and their H-H direction is parallel to the c-axis of
the beryl crystal. The peak at 2357 cm1 is caused
by CO2. This peak was present in all the samples

140

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

GEMS & GEMOLOGY

SUMMER 2005

mine, as described in the Microscopic Characteristics


section. Stones with sufficient fractures containing a
near-colorless filler showed FTIR spectra that are typical for an oil (figure 41). Although the type of oil cannot be established with this method, at the Chantete
mine we were told that Johnsons baby oil (mineral
oil) is used on the rough. Likewise, mineral oil is
reportedly used by emerald cutters in Israel. No green
fillers were seen in any of the samples.

DISCUSSION

Figure 40. This representative FTIR spectrum of a


Kafubu emerald shows peaks at 7096 and 5265 cm1
that are caused by the vibrations of water molecules
adjacent to alkali-metal ions in the channels of the
beryl structure. The broad band between 3900 and
3400 cm1 also is caused by water molecules in this
position, and by free water molecules. The peak at
2357 cm1 is due to CO2. The intensity of this peak
varied substantially between samples.

Figure 41. Peaks in the 31002800 cm1 range were


seen in the FTIR spectra of both faceted and cuttable
rough specimens of Kafubu emerald, indicating the
presence of oil. No artificial resin was found in the
stones that were tested. Spectra have been offset vertically for clarity.

Physical Properties. The measured refractive indices


and specific gravity values (table 1) are fairly consistent with those presented by Milisenda et al. (1999).
An unusual dark bluish green emerald described by
Schmetzer and Bank (1981) showed extreme values
for refractive indices, no = 1.602 and ne = 1.592, with
a birefringence of 0.010.
Inclusions in Zambian emeralds have been
described by Campbell (1973), Koivula (1982, 1984),
Graziani et al. (1984), Gbelin and Koivula (1986),
Milisenda et al. (1999), and Moroz and Eliezri
(1999). These studies largely agree with and support
our findings (table 1). However, in the present
study, we did not encounter glauconite, chrysoberyl, margarite, muscovite, or chrysotile, which
suggests these mineral inclusions are rare. It should
also be noted that the chrysoberyl, margarite, and
muscovite were identified in a heavily included
crystal fragment (Graziani et al., 1983), which may
not be representative of gem-quality material. The
presence of isolated CO2-CH4bearing negative
crystals has not been reported previously in
Zambian emerald.
Chemical Properties. Our chemical analyses
showed a wider range of trace-element concentrations than was indicated by the seven analyses presented by Milisenda et al. (1999). Our Cr content
was fairly consistent with their results but was
slightly higher in some stones; V, Fe, Mg, Ca, Na,
and K also showed a much wider range. In addition,
our study indicated the presence of Cs and occasionally Sc. The values we measured for the main
trace elements all fell within the ranges that were
presented earlier by Graziani et al. (1984) on one
emerald, Hnni (1982) on two emeralds, Schwarz
and Henn (1992) on 11 emeralds from the Kamakanga mine (mean concentrations only), and Moroz
et al. (1998) and Moroz and Eliezri (1998) on three
samples. Our data also were consistent with the

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

GEMS & GEMOLOGY

SUMMER 2005

141

analyses of noncuttable rough Zambian emeralds


reported by Seifert et al. (2004c), except those rough
samples had slightly higher Na. Moreover, we
found a wider range of Mg and Fe concentrations,
with lower values in light green cut emeralds.
Using Schwarzs empirical subdivision of low,
medium, and high concentrations of elements in
emerald (see, e.g., Schwarz, 1990a,b; Schwarz and
Henn, 1992), the Zambian stones generally show a
low V content, a moderate amount of Cr, Mg, and
Na, and a moderate-to-high Fe content. Notable is
the relatively high Cs content in many of our samples. As observed by Bakakin and Belov (1962), Cs is
typically present in Li-rich beryl. Li could not be analyzed by electron microprobe, but PIXE/PIGE analyses of 11 Zambian crystal fragments by Calligaro et
al. (2000) indicated an average Li content of 580 ppm,
along with a high average Cs content of 1,150 ppm.
Examination of the growth zoning in the
Chantete emerald crystal (figure 38) showed that as
Cr, Fe, and Mg increase, Al decreases. This is
expected from the known substitutions in emerald
(Aurisicchio et al., 1988), namely of Al3+ in the octahedral site by Fe2+ and Mg2+, plus Fe3+, Cr3+, and V3+.
The presence of Cs+, which is located in the channel
of the beryl structure, is coupled with the substitution of Li+ for Be2+ in the tetrahedral site and clearly
follows a different trend. Na appears to reflect mainly the trend of Cr, Fe, and Mg, but also seems to follow the general trend of Cs. This could be explained
by the fact that Na+ can be involved in coupled substitutions for Al3+ in the octahedral site, but also for
Be 2+ in the tetrahedral site (compare, e.g., to
Aurisicchio et al., 1988; Barton and Young, 2002).
From the trend shown in figure 38, it is clear that
the main substitution in the Chantete emerald crystal took place in the Al3+ octahedral site.
The analyses of the faceted emeralds did not show
a consistent correlation between color and Cr content (see table 2 and the G&G Data Depository).
However, analysis of the various color zones in the
Chantete emerald described above provided a good
opportunity to explore the relationship between color
and trace-element content within a single crystal.
When the data from this crystal are graphed according to the atomic ratios that were plotted by Barton
and Young (2002), a clear trend is revealed that confirms Cr as the main chromophore (figure 42). Also
plotted in figure 42 are two of the darkest green emeralds analyzed for this study (from Kagem), which
extend the trend shown by the zoned Chantete sample. This trend resembles the one diagramed by

142

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

Figure 42. Data from the color-zoned Chantete emerald crystal in figures 37 and 38, together with two of
the darkest green emeralds analyzed for this study,
are plotted on this beryl composition diagram from
Barton and Young (2002) to reveal the relationship
between composition and color. The data points are
labeled according to the growth zones shown in figure
37, and plotted by color. A clear correlation is
revealed between the depth of emerald coloration and
Cr content. The arrow (from Barton and Young, 2002)
indicates the general trend of beryls from Somondoco
(Colombia) and Khaltaro (Pakistan) that range from
colorless or pale blue to deep green.

Barton and Young (2002) for pale blue to dark green


beryls from Somondoco, Colombia and Khaltaro,
Pakistan. The non-emerald analyses from those localities contained very little or no Cr and plotted along
the X-axis, like the colorless zone (analysis no. 7) of
the Chantete emerald in this study.
Identification. Separation from Synthetics. The higher R.I. values of Zambian emeralds (table 1) make
them easy to distinguish from their synthetic counterparts. Synthetic emeralds typically have lower
refractive indicesroughly between 1.56 and 1.58
and lower specific gravities of 2.652.70 (see, e.g.,
Schrader, 1983; Liddicoat, 1989; Webster, 1994;
Schmetzer et al., 1997), although some Russian
hydrothermal synthetic emeralds have shown R.I.s
up to 1.584 and S.G.s up to 2.73 (Webster, 1994;

GEMS & GEMOLOGY

SUMMER 2005

Koivula et al., 1996). Russian hydrothermal synthetic


emeralds can be readily separated from natural emeralds by their distinctive chevron growth zoning or, if
present, by unusual internal characteristics such as
tiny red-brown particles. Only in the absence of these
features would infrared or EDXRF spectroscopy be
needed to identify this synthetic (Koivula et al., 1996).
Separation from Emeralds of Different Geographic
Origin. The commercially important Colombian
emeralds typically show lower refractive indices of
1.5691.580 (e.g., Ringsrud, 1986; Gbelin, 1989;
Boehm, 2002), although Bosshart (1991) gave a
wider R.I. range of 1.5651.588. Colombian emeralds can be further distinguished by their distinctive
inclusion scenery, including abundant three-phase
inclusions containing halite cubes.
A comparison of other commercially available
emeralds with R.I., birefringence, and S.G. values
similar to those from Kafubu is shown in table 3 and
discussed below. As previously mentioned, roughly
70% of the Zambian stones tested showed no =
1.5911.595 and ne = 1.5831.587. Only some light
green stones had R.I. values below 1.583, and dark
green stones may show R.I.s above 1.595. According
to these data, accurate measurement of the R.I. can
distinguish most of the commercial medium and
dark green Zambian emeralds from the emeralds of
Itabira, Brazil. In addition, the R.I.s of most
Zambian emeralds appear to show only a slight overlap with those of emeralds from the Ural Mountains.
The internal features seen in emeralds from the
various localities in table 3 show many similarities,
so their use for determining geographic origin is
rather limited (for further discussion, see Schwarz,
1998). This overlap is due to the similar geologic
environment in which these emeralds form (various
types of metamorphic schist). Nevertheless, some
distinguishing internal features are listed in table 3.
When present, these characteristics can be used to
clearly separate Kafubu emeralds from those of
Santa Terezinha de Gois (Brazil), Sandawana
(Zimbabwe), and the Swat Valley (Pakistan).
The UV-Vis spectra can only help distinguish
between the various emerald occurrences to a limited extent, by indicating differences in the valence
state of iron. Only the UV-Vis spectra of Sandawana
and, possibly, Ural emeralds show distinctive features. Interestingly, this method can also be used
effectively to distinguish between emeralds from
Nigeria (containing Fe2+ and Fe3+) and Colombia
(virtually no iron and no Fe3+; Henn and Bank, 1991;

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

Moroz et al., 1999), which may contain very similar


three-phase inclusions. Note that Nigerian emeralds also fall outside the scope of our discussion due
to their lower optical properties.
The chemical composition of the various emeralds in table 3 shows considerable overlap. However,
trace elements may help eliminate some localities
for a specific stone. For instance, relatively high K
contents have been documented in emeralds from
Kafubu and Mananjary.
PIXE/PIGE analysis, which is a highly sensitive
technique that is capable of measuring light elements such as Li, has revealed interesting trends for
Cs, Li, and Rb (Calligaro et al., 2000; table 3).
Compared to Kafubu, emeralds from Mananjary
showed lower Li content, and those from Habachtal
had smaller amounts of Rb. Enriched Cs is indicative of emeralds from Kafubu, Mananjary, and
Sandawana. The validity of these trends should be
tested further with additional analyses of representative samples from the various deposits.
Measuring stable oxygen isotope ratios is another
(not entirely nondestructive) technique that can provide additional information. This technique was first
applied to Zambian emeralds by Eliezri and Kolodny
(1997), and has seen wider application to emeralds
from various deposits by Giuliani et al. (1998, 2000).
The oxygen isotope data are expressed as 18O,
which is the relative difference between the 18O/16O
ratio of the sample and that of SMOW (standard
mean ocean water), expressed in per mil (). The
18O values of emeralds from some of the localities
in table 3 show considerable overlap (figure 43). Data
for Kafubu emeralds exhibit a range similar to that
reported for emeralds from the Ural Mountains.
This brief overview shows that even when optical properties, internal characteristics, and chemical
data are carefully combined and evaluated, it may
not always be possible to differentiate Kafubu emeralds from those of other localities. It appears, however, that in many cases the use of chemical analysis is helpful. Even so, more analyses of representative emeralds from various deposits are needed to
better characterize their variations in chemical
composition.

SUMMARY AND CONCLUSION


The Kafubu area became a significant producer of
good-quality emeralds in the 1970s. As of August
2004, the main production took place at the Kagem,
Grizzly, and Chantete concessions, which are all

GEMS & GEMOLOGY

SUMMER 2005

143

TABLE 3. Properties of emeralds with R.I. and S.G. values similar to those from Kafubu, Zambia.
Properties

Kafubu area,
Zambia

Physical and optical properties


References
This study, Milisenda et al. (1999)

R.I.
no
ne
Birefringence
S.G.
Distinguishing
internal features

Iron peaks in
UV-VIS spectra

PIXE/PIGE datac
No. samples
Trace elements (ppm)
Li
Cs
Rb

Itabira district,
Brazil

Ural Mountains,
Russia

Mananjary region,
Madagascar

Lind et al. (1986),


Schwarz (1990a;
2001)

Hnni et al. (1987),


Schwarz et al. (1988),
Epstein (1989), Henn
and Bank (1991);
Gbelin (1989),
Schwarz (1998),
Zwaan (2001)

Gbelin (1974),
Hnni and Klein (1982),
Sinkankas (1981),
Schwarz and Henn
Bank (1982),
(1992), Schwarz (1994)
Mumme (1982),
Schmetzer et al.
(1991), Gbelin and
Koivula (1986), Laskovenkov and Zhernakov (1995)

1.5851.599
1.5781.591
0.0060.009
2.69 2.78
Skeletal magnetite
and hematite (if
present)

1.5921.600
1.5841.593
0.0060.010
2.752.77
Abundant opaque
inclusions, such as
black spinel octahedrons; pale brown
to colorless carbonate
rhombohedra; fluid inclusions are very small
and rare

1.580 1.590
1.5741.583
0.0040.009
2.722.74
Numerous parallel
growth tubes (with
remarkable variation
of fluid inclusions)

1.5811.591
1.5751.584
0.007
2.722.75
Irregular color distribution (if present)

1.5881.591
1.5801.585
0.0060.009
2.682.73
Elongate quartz crystals,
parallel to the c-axis,
often associated with
growth tubes; fibrous
aggregates of talc (if
present); rhombohedral
carbonate crystals (rare)

Fe2+ and Fe3+


features

Fe2+ and Fe3+


features

Fe2+ and Fe3+


features

Fe2+ features only,


or Fe2+ and Fe3+
features

Fe2+ and Fe3+


features

Schwarz (1990a)

Schwarz (1990b),
Zwaan (2001)

Schmetzer et al.
(1991), Schwarz
(1991)
27

Hnni and Klein (1982),


Schwarz and Henn
(1992)
7

Electron-microprobe dataa
References
This studyb

No. analyses
Oxide (wt.%)
SiO2
TiO2
Al2O3
Cr2O3
V2O3
FeOtot
MnO
MgO
CaO
Na2O
K2O
Cs2O
Rb2O
Sc2O3
Mo2O3

Santa Terezinha de
Gois, Brazil

114

15

71

61.9 65.4
bdl
12.517.9
bdl0.84
bdl0.08
0.061.75
bdl0.01
0.272.90
bdl0.12
0.161.99
bdl0.27
bdl
bdl0.23
bdl0.07

63.866.5

12.214.3
0.061.54
0.08
0.771.82
0.02
2.483.09

1.461.73
0.03

63.367.0
0.04
13.916.9
0.06 1.42
0.07
0.411.30
0.08
1.392.64
0.10
0.791.93
0.08
bdl
bdl

0.06

64.666.9
0.05
14.218.3
0.01 0.50
0.04
0.101.16
0.03
0.292.23
0.03
0.611.72
0.07

63.365.0
bdl
12.814.6
0.080.34
0.03
0.911.46

1.713.00

1.282.16
0.050.21

11

20

12

580 (230)
1150 (540)
140 (60)

170 (90)
350 (300)
15 (2)

720 (260)
360 (230)
40 (60)

130 (70)
610 (580)
190 (70)

Abbreviations: bdl = below detection limit; = no data.


The range of composition reported in the literature for Zambian emeralds falls within the analyses obtained for this study.
c Data from Calligaro et al. (2000), average elemental content from three point analyses per sample in ppm by weight. Standard deviations are given in
parentheses. Detection limits (in ppm) for Rb=2, Li=50, and Cs=100.
a

144

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

GEMS & GEMOLOGY

SUMMER 2005

Sandawana,
Zimbabwe

Swat Valley,
Pakistan

Habachtal,
Austria

Gbelin (1958),
Zwaan et al. (1997)

Henn (1988),
Gbelin (1989)

Gbelin (1956),
Morteani and
Grundmann (1977),
Nwe and Grundmann
(1990), Schwarz
(1991), Gbelin and
Koivula (1986),
Grundmann (2001)

1.5901.594
1.5841.587
0.0060.007
2.742.77
Abundant thin
needles and curved
fibers of amphibole;
extremely small fluid
inclusions (if present);
black grains of chromian ilmenorutile
(very rare); phlogopite
virtually absent
Fe2+ features only

1.5841.600
1.5781.591
0.0060.009
2.702.78
Many two-and threephase inclusions
(similar to Colombia),
but with no rectangular to square shapes;
common black chromite and rhombohedral dolomite

1.5821.597
1.5741.590
0.0050.007
2.702.77
Commonly show
patchy color distribution; most stones are
heavily included and
fractured (rarely cuttable and only in
small sizes)

Fe2+ and Fe3+


features

Fe2+ and Fe3+


features

Zwaan et al. (1997)

Henn (1988), Hammarstrom (1989)

Schwarz (1991)

41

176

47

62.663.2
0.05
13.013.7
0.611.33
0.040.07
0.450.82
bdl
2.522.75

2.072.41
0.030.06
0.060.10

bdl
bdl

62.2 65.8
0.02
13.115.0
0.152.05
0.010.06
0.202.51
bdl
2.213.10
bdl
1.082.11
0.02

0.19

64.666.1
0.03
13.314.5
0.010.44
0.04
0.611.87
0.05
2.33 2.92
0.02 0.04
1.542.24
0.010.10

0.04

22

13

800 (310)
710 (320)
350 (200)

350 (130)
120 (80)
6 (3)

190 (210)
370 (390)
18 (13)

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

Figure 43. Significant overlap is evident in the oxygen


isotopic values of emeralds reported from some of the
localities in table 3. The range of 18O values for
Kafubu emeralds is similar to that for emeralds from
the Ural Mountains. Data from Eliezri and Kolodny
(1997), Giuliani et al. (1998), Cheilletz et al. (2001),
Gavrilenko (2001), and Zwaan et al. (2004).

open-pit mines. Research and exploration activities


(e.g., Tembo et al., 2000; Seifert et al., 2004c)
demonstrate that the potential for new emerald
occurrences in the Kafubu area remains very high.
During a late stage of the Pan-African orogeny,
probably around 450 million years ago, emerald mineralization at Kafubu was caused by metasomatic
alteration of Cr-bearing metabasites, which were
invaded by Be-bearing fluids derived from hydrothermal veins. Economic emerald concentrations are
almost entirely confined to phlogopite reaction
zones between quartz-tourmaline veins and the
metabasites.
Kafubu emeralds show large variations in physical
properties, with refractive indices and specific gravities higher than those of emeralds from Colombia,
but comparable to emeralds from many other commercially important localities. Kafubu emeralds are
characterized by partially healed fractures with various shapes of fluid inclusions. Characteristic mineral
inclusions are randomly oriented actinolite needles
and phlogopite platelets, and dravite, fluorapatite,
magnetite, and hematite. Skeletal magnetite and
hematite have not been described as inclusions in
emeralds from other occurrences. Inclusions such as
dravite, fluorapatite, and fluorite, as well as niobian
rutile identified in one stone, confirm a metasomatic
origin related to the intrusion of a rare-element pegmatite-hydrothermal vein system.
The chemical composition of the Kafubu emeralds is characterized by a wide range of trace-ele-

GEMS & GEMOLOGY

SUMMER 2005

145

ment contents, with generally moderate amounts of


Cr, Mg, and Na, and a moderate-to-high iron content as both Fe2+ and Fe3+. Notable is a relatively
high content of Cs and Li in many samples. The V
content is low, and Sc may be detected.
Most of Zambias emeralds are exported to
India, mainly for use in the domestic market, and
to Israel for international distribution. Commercial
production of emeralds for the international market is done in calibrated sizes; commercial faceted
goods typically weigh up to 5 ct, are slightly bluish
green to bluish green with medium to dark tones,
Figure 44. Zambia is a recognized source of
attractive emeralds that are readily available
in calibrated sizes, but occasionally exceptional stones are produced, as shown by the 15.32
ct emerald in this platinum pendant that is set
with diamonds. The emerald is unusually
large and clean for a Zambian stone, and a
laboratory report has verified that it has not
been treated in any way. Courtesy of Kothari
and Company Inc., Los Angeles, CA; photo
Harold and Erica Van Pelt.

ABOUT THE AUTHORS


Dr. Zwaan is director of the Netherlands Gemmological
Laboratory and curator of minerals and gems at the National
Museum of Natural History Naturalis, Leiden, the
Netherlands. Drs. Seifert and Vrna are geologists at the
Czech Geological Survey, Prague, Czech Republic. Mr. Anckar
is a gemologist and geologist with the European Unionfunded
Mining Sector Diversification Programme (EU-MSDP), Lusaka,
Zambia. Mr. Laurs is editor of Gems & Gemology, and Mr.
Muhlmeister is senior research associate in the GIA Gem
Laboratory, Carlsbad. Mr. Koivula is chief gemologist at the
AGTA Gemological Testing Center laboratory in Carlsbad. Dr.
Simmons is professor of mineralogy, and Mr. Falster is analytical instrumentation manager, in the Department of Geology
and Geophysics at the University of New Orleans, Louisiana.
Mr. Lustenhouwer is chemical analyst in the Faculty of Earth
Sciences at the Free University of Amsterdam, the
Netherlands. Mr. Garcia-Guillerminet is director of the CCIP
French Gemmological Laboratory, Paris, France.
ACKNOWLEDGMENTS: Chantete Emerald Ltd., Kitwe,
Zambia, is thanked for providing valuable information on
the mining operation, granting full access to the deposit,
and supplying samples for examination. The authors are
grateful to Enrico Storti (E&M Storti Mining Ltd., Kitwe,
Zambia) for his kind assistance with visiting the Pirala and
Twampane mines, and Siradiou Diop (Grizzly Mining Ltd.,
Kitwe) for accommodating them at the Grizzly mine.
Elizabeth P. Quinn, staff gemologist at the GIA Gem
Laboratory, Carlsbad, is thanked for providing gemological
properties of some emeralds from the Chantete mine. Dr.

146

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

Andy Hsi-Tien collected the UV-Vis spectra at the GIA Gem


Laboratory in Carlsbad.
Avraham Eshed, president of Gemstar Ltd., Ramat Gan,
Israel, is thanked for valuable information on mining, production, and distribution; for providing photographs; and for giving
access to his cutting factory and the opportunity to inspect a
large number of commercial goods that are entering the market. Israel Eliezri (Colgem Ltd., Ramat Gan, Israel) is thanked
for loaning samples for examination, and for supplying additional photographs and information. Ashok Sancheti (Pioneer
Gems, New York) and Lalit Kothari (Kothari and Company Inc.,
Los Angeles, California) are thanked for loaning jewelry for photography. Trevor Schultz (Schultz Mining Ltd., Kitwe, Zambia)
offered feedback on the run-of-mine production of emerald
from the Kafubu area. Govind Gupta (Kagem Mining Ltd.,
Kitwe, Zambia) gave updated information on the privatization
agreement between Hagura and the Government of Zambia.
Dharmendra Tank (Heeralal Chhaganlal Tank Manufacturing
Jewelers, Jaipur, India) provided information on the international distribution of Zambian emeralds. J. G. Dey, chief geologist
at Kagem (Kitwe, Zambia), described the successful drilling
operation at the Fwaya-Fwaya Ext. F10 mine. Dirk van der
Marel and Jeroen Goud (National Museum of Natural History
Naturalis, Leiden) are thanked for taking photographs, and
Dirk also for general assistance.
Facilities for electron-microprobe analyses were provided
by the Free University of Amsterdam and by NWO, the
Netherlands Organization for Scientific Research. Access to
Raman facilities was provided by the CCIP French
Gemmological Laboratory, Paris. Electron-microprobe analyses of the samples in box A were performed at Masaryk
University in Brno, Czech Republic, by P. Sulovsky .

GEMS & GEMOLOGY

SUMMER 2005

REFERENCES
Aurisicchio C., Fioravanti G., Grubessi O., Zanazzi P.F. (1988)
Reappraisal of the crystal chemistry of beryl. American
Mineralogist, Vol. 73, No. 78, pp. 826837.
Bakakin V.V., Belov N.V. (1962) Crystal chemistry of beryl.
Geokhimiya, Vol. 5, pp. 420433.
Banerjee A. (1995) Investigation of fluid inclusion in emeralds of different geological origins by microchemical analysis and IR-reflexion-spectroscopy. Boletn de la Sociedad Espaola de Mineraloga, Vol. 18-1, pp. 1819.
Bank H. (1982) Geschliffene zonar gebaute Smaragde aus dem Gebiet
der Tokowaya, Ural, UdSSR. Zeitschrift der Deutschen
Gemmologischen Gesellschaft, Vol. 31, No. 3, pp. 193194.
Barton M.D., Young S. (2002) Non-pegmatitic deposits of beryllium:
Mineralogy, geology, phase equilibria and origin. In E.S. Grew, Ed.,
Beryllium: Mineralogy, Petrology and Geochemistry, Reviews in
Mineralogy and Geochemistry, Vol. 50, Chapter 14, pp. 591691.
Boehm E. (2002) Gem News International: New emerald find from
the Chivor region, Colombia. Gems & Gemology, Vol. 38, No. 4,
pp. 350351.
Bosshart G. (1991) Emeralds from Colombia, part 2. Journal of
Gemmology, Vol. 22, No. 7, pp. 409425.
Calligaro T., Dran J.C., Poirot J.P., Querr G., Salomon J., Zwaan J.C.
(2000) PIXE/PIGE characterisation of emeralds using an external
micro-beam. Nuclear Instruments and Methods in Physics
Research B, Vol. 161163, pp. 769774.
Campbell I.C.C. (1973) Emeralds reputed to be of Zambian origin.
Journal of Gemmology, Vol. 13, No. 5, pp. 169179.
Caspi A. (1997) Modern diamond cutting and polishing. Gems &
Gemology, Vol. 33, No. 2, pp. 102121.
Cheilletz A., Sabot B., Marchand P., de Donato P., Taylor B.,
Archibald D., Barrs O., Andrianjaffy J. (2001) Emerald deposits in
Madagascar: Two different types for one mineralising event. EUG
XI, Strasbourg, France, April 812, p. 547.
Coats J.S., Mosley P.N., Mankelow J.M., Mwale M., Chikambwe
E.M., Muibeya B., Ndhlovu K.C., Nzabara F. (2001) The Geology
and Mineral Resources of Zambia. Geological Survey Department
Memoir 6, Ministry of Mines and Minerals Development, Lusaka,
Zambia.
Cook F.A. (1997) Applications of geophysics in gemstone exploration.
Gems & Gemology, Vol. 33, No. 1, pp. 423.
Cosi M., de Bonis A., Gosso G., Hunziker J., Martinotti G., Moratto S.,
Robert J.P., Ruhlman F. (1992) Late Proterozoic thrust tectonics,
high-pressure metamorphism and uranium mineralization in the
Domes area, Lufilian Arc, northwestern Zambia. Precambrian
Research, Vol. 58, No. 14, pp. 215240.
Daly M.C., Unrug R. (1983) The Muva Supergroup of northern
Zambia; A craton to mobile belt sedimentary sequence.
Transactions of the Geological Society of South Africa, Vol. 85,
No. 3, pp. 155165.
de Waele B., Mapani B. (2002) Geology and correlation of the central
Irumide belt. Journal of African Earth Sciences, Vol. 35, No. 3,
pp. 385397.
de Waele B., Wingate M.T.D., Fitzsimons I.C.W., Mapani B. (2002)
Untying the Kibaran knot: A re-assessment of Mesoproterozoic
correlations in southern Africa based on SHRIMP U-Pb data from
the Irumide belt. Geology, Vol. 31, No. 6, pp. 509512.
Donovan J.J., Tingle T.N. (1996) An improved mean atomic number
correction for quantitative microanalysis. Journal of Microscopy,
Vol. 2, No. 1, pp. 17.
Eliezri I.Z., Kolodny Y. (1997) The isotopic composition of oxygen in
emeralds as an indicator of origin. 26th International Gemmological
Conference, Idar-Oberstein, Germany, p. 14.
Epstein D.S. (1989) The Capoeirana emerald deposit near Nova Era,
Minas Gerais, Brazil. Gems & Gemology, Vol. 25, No. 3, pp.
150158.
Fritsch E., Rondeau B., Notari F., Michelou J.-C., Devouard B., Peucat
J.-J., Chalain J.-P., Lulzac Y., de Narvaez D., Arboleda C. (2002) Les
nouvelles mines dmeraude de La Pita (Colombie) 2e partie.
Revue de Gemmologie, No. 144, 2002, pp. 1321.
Gavrilenko E. (2001) Emeralds of the Ural mountains (Russia):
Geology, fluid inclusions and oxygen isotopes. 28th International
Gemmological Conference, Madrid, Spain, pp. 3640.

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

Giuliani G., France-Lanord C., Coget P., Schwarz D., Cheilletz A.,
Branquet Y., Girard D., Martin-Izard A., Alexandrov P., Piat D.H.
(1998) Oxygen isotope systematics of emerald: Relevance for its
origin and geological significance. Mineralium Deposita, Vol. 33,
No. 5, pp. 513519.
Giuliani G., Chaussidon M., Schubnel H.J., Piat D.H., Rollion-Bard
C., France-Lanord C., Giard D., Narvaez D.D., Rondeau B. (2000)
Oxygen isotopes and emerald trade routes since antiquity.
Science, Vol. 287, No. 5453, pp. 631633.
Graziani G., Gbelin E., Lucchesi S. (1983) The genesis of an emerald
from the Kitwe District, Zambia. Neues Jahrbuch fr
Mineralogie, Monatshefte, Vol. 4, pp. 75186.
Graziani G., Gbelin E., Lucchesi S. (1984) Report on the investigation of an emerald from the Kitwe District, Zambia. Australian
Gemmologist, Vol. 15, No. 7, pp. 227234.
Grundmann G. (2001) Die Smaragde der Welt. In D. Schwarz and R.
Hochleitner, Eds., Smaragdder kostbarste Beryll, der teuerste
Edelstein, extraLapis No. 21, pp. 2637.
Gbelin E.J. (1958) Emeralds from Sandawana. Journal of Gemmology,
Vol. 6, No. 8, pp. 340354.
Gbelin E.J. (1956) The emerald from Habachtal. Gems & Gemology,
Vol. 8, No. 10, pp. 295309.
Gbelin E.J. (1974) Die Innenwelt der Edelsteine. Hanns Reich
Verlag, Dsseldorf, Germany.
Gbelin E.J. (1989) Gemological characteristics of Pakistani emeralds.
In A.H. Kazmi and L.W. Snee, Eds., Emeralds of Pakistan:
Geology, Gemology and Genesis, Van Nostrand Reinhold, New
York, pp. 7592.
Gbelin E., Koivula J.I. (1986) Photoatlas of Inclusions in Gemstones.
ABC Edition, Zrich, Switzerland.
Hammarstrom J.M. (1989) Mineral chemistry of emeralds and some
associated minerals from Pakistan and Afghanistan: An electron
microprobe study. In A.H. Kazmi and L.W. Snee, Eds., Emeralds
of Pakistan: Geology, Gemology and Genesis, Van Nostrand
Reinhold, New York, pp. 125150.
Hnni H. (1982) A contribution to the separability of natural and synthetic emeralds. Journal of Gemmology, Vol. 18, No. 2, pp. 138144.
Hnni H.A., Klein H.H. (1982) Ein Smaragdvorkommen in
Madagaskar. Zeitschrift der Deutschen Gemmologischen
Gesellschaft, Vol. 31, No. 1/2, pp. 7177.
Hnni H.A., Schwarz D., Fischer M. (1987) The emeralds of the
Belmont mine, Minas Gerais, Brazil. Journal of Gemmology, Vol.
20, No. 7/8, pp. 446456.
Henn U. (1988) Untersuchungen an Smaragden aus dem Swat-Tal,
Pakistan. Zeitschrift der Deutschen Gemmologischen Gesellschaft, Vol. 37, No. 3/4, pp. 121127.
Henn U., Bank H. (1991) Auergewhnliche Smaragde aus Nigeria.
Zeitschrift der Deutschen Gemmologischen Gesellschaft, Vol. 40,
No. 4, pp. 181187.
Hickman A.C.J. (1972) The Miku Emerald Deposit. Geological
Survey Department Economic Report No. 27, Ministry of Mines
and Mining Development, Republic of Zambia, 35 pp.
Hickman A.C.J. (1973) The Geology of the Luanshya Area. Report of
the Geological Survey No. 46, Ministry of Mines, Republic of
Zambia, 32 pp.
John T., Schenk V., Mezger K., Tembo F. (2004) Timing and PT evolution of whiteschist metamorphism in the Lufilian ArcZambezi
Belt Orogen (Zambia): Implications for the assembly of
Gondwana. Journal of Geology, Vol. 112, No. 1, pp. 7190.
Johnson M.L., Elen S., Muhlmeister S. (1999) On the identification of
various emerald filling substances. Gems & Gemology, Vol. 35,
No. 2, pp. 82107.
Kiefert L., Hnni H.A., Chalain J.-P., Weber W. (1999) Identification
of filler substances in emeralds by infrared and Raman spectroscopy. Journal of Gemmology, Vol. 26, No. 8, pp. 487500.
Koivula J.I. (1982) Tourmaline as an inclusion in Zambian emeralds.
Gems & Gemology, Vol. 18, No. 4, pp. 225227.
Koivula J.I. (1984) Mineral inclusions in Zambian emeralds.
Australian Gemmologist, Vol. 15, No. 7, pp. 235239.
Koivula J.I., Kammerling R.C., DeGhionno D., Reinitz I., Fritsch E.,
Johnson M.L. (1996) Gemological investigation of a new type of
Russian hydrothermal synthetic emerald. Gems & Gemology,
Vol. 32, No. 1, pp. 3239.

GEMS & GEMOLOGY

SUMMER 2005

147

Kozlowski A., Metz P., Estrada Jaramillo H.A. (1988) Emeralds from
Somondoco, Colombia: Chemical composition, fluid inclusions
and origin. Neues Jahrbuch fr Mineralogie, Abhandlungen, Vol.
159, No. 1, pp. 2349.
Laskovenkov A.F., Zhernakov V.I. (1995) An update on the Ural
emerald mines. Gems & Gemology, Vol. 31, No. 2, pp. 106113.
Laurs B. (2004) Gem News International: Update on several gem
localities in Zambia and Malawi. Gems & Gemology, Vol. 40,
No. 4, pp. 347350.
Leake B.E., Woolley A.R., Arps C.E.S., Birch W.D., Gilbert M.C.,
Grice J.D., Hawthorne F.C., Kato A., Kisch H.J., Krivovichev V.G.,
Linthout K., Laird J., Mandarino J., Maresch W.V., Nickel E.H.,
Rock N.M.S., Schumacher J.C., Smith D.C., Stephenson N.C.N.,
Ungaretti L., Whittaker E.J.W., Youzhi G. (1997) Nomenclature of
amphiboles: Report of the subcommittee on amphiboles of the
International Mineralogical Association Commission on New
Minerals and Mineral Names. Mineralogical Magazine, Vol. 61,
No. 2, pp. 295321.
Liddicoat R.T. (1989) Handbook of Gem Identification, 12th ed., 2nd
rev. printing. Gemological Institute of America, Santa Monica, CA.
Lind T., Henn U., Bank H. (1986) Smaragde von Sta. Terezinha de
Goias, Brasilien, mit relativ hoher Lichtbrechung. Zeitschrift der
Deutschen Gemmologischen Gesellschaft, Vol. 35, No. 4, pp.
186187.
Milisenda C.C., Malango V., Taupitz K.C. (1999) Edelsteine aus
SambiaTeil 1: Smaragd. Zeitschrift der Deutschen Gemmologischen Gesellschaft, Vol. 48, No. 1, pp. 928.
Moroz I., Eliezri I.Z. (1998) Emerald chemistry from different
deposits. Australian Gemmologist, Vol. 20, No. 2, pp. 6469.
Moroz I., Eliezri I.Z. (1999) Mineral inclusions in emeralds from different sources. Journal of Gemmology, Vol. 26, No. 6, pp.
357363.
Moroz I., Panczer G., Roth M. (1998) Laser-induced luminescence of
emeralds from different sources. Journal of Gemmology, Vol. 26,
No. 5, pp. 316320.
Moroz I., Roth M.L., Deich V.B. (1999) The visible absorption spectroscopy of emeralds from different deposits. Australian
Gemmologist, Vol. 20, No. 8, pp. 315320.
Morteani G., Grundmann G. (1977) The emerald porphyroblasts in
the penninic rocks of the central Tauern Window. Neues
Jahrbuch fr Mineralogie, Mitteilungen, Vol. 11, pp. 509516.
Mumme I. (1982) The Emerald. Mumme Publications, Port Hacking,
New South Wales, Australia.
Nguni B., Mwamba K. (2004) Evaluation of the Potential and
Authentication of Emerald Mineralization in North-Western
Province, Musakashi Area, Chief MujimanzovuSolwezi District.
Draft Economic Report, Geological Survey Department, Lusaka,
Zambia.
Nwe Y.Y., Grundmann G. (1990) Evolution of metamorphic fluids in
shear zones: The record from the emeralds of Habachtal, Tauern
Window, Austria. Lithos, Vol. 25, No. 4, pp. 281304.
Parkin J. (2000) Geology of the St. Francis Mission and Musakashi
River Areas. Explanation of Degree Sheet 1226 NE and SE
Quarters and Part of 1126 SE Quarter. Report No. 80, Geological
Survey Department, Lusaka, Zambia, 49 pp.
Platonov A.N., Taran M.N., Minko O.E., Polshyn E.V. (1978) Optical
absorption spectra and nature of color of iron-containing beryls.
Physics and Chemistry of Minerals, Vol. 3, No. 1, pp. 8788.
Ringsrud R. (1986) The Coscuez mine: A major source of Colombian
emeralds. Gems & Gemology, Vol. 22, No. 2, pp. 6779.
Samson I., Anderson A., Marshall D. (2003) Fluid Inclusions,
Analysis and Interpretation. Mineralogical Association of Canada
Short Course Series, Vol. 32, 374 pp.
Schmetzer K., Bank H. (1981) An unusual pleochroism in Zambian
emeralds. Journal of Gemmology, Vol. 17, No. 7, pp. 443445.
Schmetzer K., Berdesinski W., Bank H. (1974) ber die Mineralart
Beryll, ihre Farben und Absorptionsspektren. Zeitschrift der
Deutschen Gemmologischen Gesellschaft, Vol. 23, No. 1, pp.
539.
Schmetzer K., Bernhardt H., Biehler R. (1991) Emeralds from the Ural
Mountains, USSR. Gems & Gemology, Vol. 27, No. 2, pp. 8699.
Schmetzer K., Kiefert L., Bernhardt H.-J., Zhang B. (1997)
Characterization of Chinese hydrothermal synthetic emerald.
Gems & Gemology, Vol. 33, No. 4, pp. 276291.

148

EMERALDS FROM THE KAFUBU AREA, ZAMBIA

Schrader H.W. (1983) Contributions to the study of the distinction of


natural and synthetic emeralds. Journal of Gemmology, Vol. 18,
No. 6, pp. 530543.
Schwarz D. (1990a) Die brasilianischen Smaragde und ihre Vorkommen:
Santa Terezinha de Gois/Go. Zeitschrift der Deutschen
Gemmologischen Gesellschaft, Vol. 39, No. 1, pp. 1344.
Schwarz D. (1990b) Die chemische Eigenschaften der Smaragde I.
Brasilien. Zeitschrift der Deutschen Gemmologischen Gesellschaft, Vol. 39, No. 4, pp. 233272.
Schwarz D. (1991) Die chemische Eigenschaften der Smaragde III.
Habachtal/sterreich und Uralgebirge/UdSSR. Zeitschrift der
Deutschen Gemmologischen Gesellschaft, Vol. 40, No. 2/3, pp.
103143.
Schwarz D. (1994) Emeralds from the Mananjary region, Madagascar:
Internal features. Gems & Gemology, Vol. 30, No. 2, pp. 88101.
Schwarz D. (1998) The importance of solid and fluid inclusions for the
characterization of natural and synthetic emeralds. In D. Giard, Ed.,
Lmeraude, Association Franaise de Gemmologie, Paris, pp. 7180.
Schwarz D. (2001) Gemmologie des SmaragdsIst die Herkunftsbestimmung von Smaragden mglich? In D. Schwarz and R.
Hochleitner, Eds., Smaragd, der kostbarste Beryll, der teuerste
Edelstein, extraLapis No. 21, pp. 6873.
Schwarz D., Henn U. (1992) Emeralds from Madagascar. Journal of
Gemmology, Vol. 23, No. 3, pp. 140149.
Schwarz D., Bank H., Henn U. (1988) Neues Smaragdvorkommen in
Brasilien entdeckt: Capoeirana bei Nova Era, Minas Gerais.
Zeitschrift der Deutschen Gemmologischen Gesellschaft, Vol. 37,
No. 3/4, pp. 146147.
Seifert A.V., Vrna S., Lombe D.K. (2004a) Evaluation of
Mineralization in the Solwezi area, Northwestern Province,
Zambia. Final Report for the Year 2004. Czech Geological
Survey, Record Office File Report No. 2/2004, Prague, Czech
Republic, 20 pp.
Seifert A., Vrna S., Lombe D.K., Mwale M. (2004b) Impact
Assessment of Mining of Beryl Mineralization on the
Environment in the Kafubu Area; Problems and Solutions. Final
Report for the Year 2003. Czech Geological Survey, Record Office
File Report No. 1/2004, Prague, Czech Republic, 63 pp.
` cek V., Vrna S., Pecina V., Zachari`s J., Zwaan J.C.
Seifert A.V., Z`
(2004c) Emerald mineralization in the Kafubu area, Zambia.
Bulletin of Geosciences, Vol. 79, No. 1, pp. 140.
Sinkankas J. (1981) Emerald and Other Beryls. Chilton Book Co.,
Radnor, PA.
Sliwa A.S., Nguluwe C.A. (1984) Geological setting of Zambian emerald deposits. Precambrian Research, Vol. 25, pp. 213228.
Suwa Y. (1994) Gemstones: Quality and Value, 8th ed. Gemological
Institute of America and Suwa & Son Inc., Santa Monica, CA.
Tembo F., Kambani S., Katongo C., Simasiku S. (2000) Draft Report
on the Geological and Structural Control of Emerald
Mineralization in the Ndola Rural Area, Zambia. University of
Zambia, School of Mines, Lusaka, 72 pp.
Viana R.R., Mnttri I., Kunst H., Jordt-Evangelista H. (2003) Age of
pegmatites from eastern Brazil and implications of mica intergrowths on cooling rates and age calculations. Journal of South
American Earth Sciences, Vol. 16, pp. 493501.
Webster R. (1994) Gems: Their Sources, Descriptions and
Identification, 5th ed. Revised by P.G. Read, ButterworthHeinemann, Oxford, England.
Wood D.L., Nassau K. (1967) Infrared spectra of foreign molecules in
beryl. Journal of Chemical Physics, Vol. 47, No. 7, pp. 22202228.
Wood D.L., Nassau K. (1968) The characterization of beryl and emerald by visible and infrared absorption spectroscopy. American
Mineralogist, Vol. 53, No. 56, pp. 777799.
Zambia cranks up the emerald sector (2004) ICA Gazette, June, pp. 1,
1011.
Zwaan J.C. (2001) Preliminary study of emeralds from the Piteiras
emerald mine, Minas Gerais, Brazil. 28th International
Gemmological Conference, Madrid, Sprain, pp. 106109.
Zwaan J.C., Kanis J., Petsch E.J. (1997) Update on emeralds from the
Sandawana mines, Zimbabwe. Gems & Gemology, Vol. 33, No.
2, pp. 80100.
Zwaan J.C., Cheilletz A., Taylor B.E. (2004) Tracing the emerald origin by oxygen isotope data: The case of Sandawana, Zimbabwe.
Comptes Rendus Geoscience, Vol. 336, pp. 4148.

GEMS & GEMOLOGY

SUMMER 2005

MT. MICA:
A RENAISSANCE IN MAINES
GEM TOURMALINE PRODUCTION
William B. (Skip) Simmons, Brendan M. Laurs, Alexander U. Falster, John I. Koivula, and Karen L. Webber

The Mt. Mica area in southwestern Maine has been mined for tourmaline and other pegmatite
gems since the 1820s. Most tourmaline production occurred during the late 1800s to the 1910s,
with occasional finds made from the 1960s to 1990s. Since May 2004, a new mining venture has
produced gem- and specimen-quality tourmaline in a variety of colors. The faceted stones typically are yellowish green to greenish blue, although pink and bicolored or tricolored stones have
been cut. Their gemological properties are typical for gem tourmaline. Chemical analysis shows
that the tourmaline mined from pockets at Mt. Mica is mostly elbaite, with lesser amounts of
schorl, rossmanite, and foitite.

he Mt. Mica deposit has been famous for 185


years as a producer (and the original source) of
gem tourmaline in Maine. Tourmaline crystals and cut gems from this historic locality are
found in museums and private collections worldwide, and they have been documented in classic
works of literature (Hamlin 1873, 1895). After years
of no or little activity, recent mining by Coromoto
Minerals (Gary and Mary Freeman) has yielded large
tourmaline crystals and a modest amount of gem
rough. Several stones have been cut, in a wide range
of colors (see, e.g., figure 1). Prior to this venture,
from 1998 to 2003, Coromoto Minerals successfully
mined the Orchard pegmatite in Maine for aquamarine and heliodor.
With the cooperation of Coromoto Minerals,
two of the authors (WBS and KLW) recently documented some of the gem pockets as they were excavated, and chemically characterized the tourmalines that were produced. This article first reviews
the history of Mt. Mica and then examines the
gemological and chemical properties of tourmaline
from this historic locality.

150

MT. MICA TOURMALINE

HISTORY
Mt. Mica is the site of the first reported occurrence
of gem tourmaline in the U.S. (Hamlin, 1895). Since
tourmaline was discovered there in 1820 by Elijah
L. Hamlin and Ezekiel Holmes (Hamlin, 1873), the
pegmatite deposit has been worked by numerous
ventures. Some of these activities, as they relate to
tourmaline production, are summarized in table 1.
An important development occurred in 1886,
when a large pocket found by Augustus C. Hamlin
and mine superintendent Samuel Carter yielded a
great number of specimens, including a 24 x 5 cm
green tourmaline crystal that was broken into four
pieces. Faceting of this tourmaline produced the 34.25
ct center stone for the famous Hamlin necklace
(Perham, 1987). This necklace, commissioned by A. C.
Hamlin, featured 70 cut stones from Mt. Mica with a
total weight of 228.12 carats, which included pink,

See end of article for About the Authors and Acknowledgments.


GEMS & GEMOLOGY, Vol. 41, No. 2, pp. 150163.
2005 Gemological Institute of America

GEMS & GEMOLOGY

SUMMER 2005

Figure 1. Recent mining


at the historic Mt. Mica
pegmatite has produced
gem tourmaline in a
variety of colors. Shown
here are some of the
samples that were characterized for this report
(0.7811.72 ct). The
stones were faceted by
Dennis Creaser (Creaser
Jewelers, South Paris,
Maine) and are courtesy
of Coromoto Minerals
Inc.; photo Jeff Scovil.

blue-green, blue, and green tourmaline, as well as colorless tourmaline and beryl (figure 2). It was donated to
the Harvard Mineralogical Museum in 1934 and is
considered one of the most significant pieces of North
American jewelry ever produced (Fales, 1995).
Using simple drilling and blasting techniques, the
early miners at Mt. Mica uncovered numerous pockets from shallow workings (figure 3). As these excavations progressed to deeper levels of the pegmatite, the
miners began employing more innovative techniques. During the period from 1890 to 1913, when
Loren B. Merrill and L. Kimball Stone had the miner-

al exploration rights at Mt. Mica, they used a large


derrick to remove boulders from a deep trench (figure
4). The location of the stacked boulders that formed
one side of the trench is shown in figure 5, along with
the sites of other workings and discoveries.
Mining activity at Mt. Mica was eventually
affected by the fabulous new tourmaline finds in
southern California that began in the early 1900s.
The abundance of tourmaline from California contributed to the decline in mining at Mt. Mica, where
activity was sporadic at best from the 1920s until
recently.

TABLE 1. Chronology of gem tourmaline production at Mt. Mica, prior to mining by Coromoto Minerals.
Year
1822
~1866

18681890

1891
1899

1904

1926
196465
1979
1990s

Event

Reference

Cyrus and Hannibal Hamlin produce red and green tourmaline.


Ordesser Marion Bowkera, owner of the farm that encompasses Mt. Mica, opens a large
pocket of tourmaline. This creates a renewed interest in the mine by Augustus and
Elijah Hamlin.
Augustus and Elijah Hamlin produce numerous fine tourmaline specimens and gem material
from several pockets; much of the top-quality material is acquired by Harvard University and
Tiffany & Co. in New York.
Loren B. Merrill and L. Kimball Stone recover exceptional blue tourmalines.
Merrill and Stone find a 411 ct blue-green tourmaline gem nodule that was part of a crystal
over 20 cm long. A second 584 ct gem nodule found in a later pocket is now in the Harvard
Mineralogical Museum collection.
Merrill and Stone open a large pocket that produces over 75 pounds (34 kg) of tourmaline
crystals, including near-colorless nodules and a single multicolored tourmaline crystal over
30 pounds (13.6 kg).
Howard Irish purchases Mt. Mica, but the deposit lays idle until the 1940s. In 1949 he leases
the deposit to the United Feldspar Corp.
Frank Perham produces green and bicolored tourmaline.
Plumbago Mining Corp. excavates the large Dagenais pocket.
Specimen- and gem-quality tourmalines are occasionally produced by Plumbago Mining Corp.

Hamlin (1873)
Hamlin (1895)

Hamlin (1895);
Perham (1987)
Hamlin (1895)
Perham (1987)

Perham (1987)

King (2000)
King (2000)
Francis (1985)
R. Naftule, pers.
comm. (2005)

Although the spelling is commonly indicated as Odessa Bowker in the literature, independent research by R. Sprague (e.g., 1870 U.S. Federal
Census, State of Maine) indicated that the farm owner and miner of the tourmaline pocket was a man named Ordesser.

MT. MICA TOURMALINE

GEMS & GEMOLOGY

SUMMER 2005

151

In 196465, Frank Perham mined the old Merrill


and Stone diggings and produced some notable tourmalines, from which an eye-clean 25 ct green stone
and a flawless 59.59 ct blue-green stone were cut
(King, 2000). The latter stone and the original crystal
were documented by Crowningshield (1966a,b).
Perham continued to work the property after it was
purchased by Plumbago Mining Corp. in 1973. In
1979, the large Dagenais pocket (4 x 5.5 x 16 m)
was found and required two months to excavate
(Francis, 1985). Later, after a brief period of inactivity,
Plumbago and private investors re-opened Mt. Mica
in 1989. Although tourmaline was recovered sporadically, the results were not considered economically
viable, and mining ceased in the late 1990s.
Coromoto Minerals acquired the Mt. Mica property in 2003 and soon started to systematically
remove portions of the entire pegmatite. This
approach has proven highly successful, with 43
pockets found in the first two years of mining. Two
of the pockets have produced large gemmy crystals
of tourmaline, which rival the best material that
has come from Mt. Mica in its 185-year history (see
Mining and Production section).
Figure 2. The Hamlin necklace is considered one of
the most significant pieces of North American jewelry. It was commissioned by Augustus C. Hamlin after
he and Samuel Carter found a large pocket of tourmaline at Mt. Mica in 1886. The necklace contains
228.12 carats of colored tourmaline, as well as colorless tourmaline and beryl. Courtesy of Harvard
Mineralogical Museum; photo Tino Hammid.

Figure 3. The yellow flags mark the location of gem


pockets in the early days of mining at Mt. Mica,
circa 1890. Shown are L. Kimball Stone (left) and
Loren B. Merrill (right); modified from Bastin
(1911, plate 12, p. 84).

152

MT. MICA TOURMALINE

LOCATION AND ACCESS


The Mt. Mica mine, which is closed to the public, is
located about 6 km northeast of the small town of
South Paris in Oxford County, southwestern Maine
(figure 6). The mine is situated on a small hill, at an
elevation of 295 m (970 feet), in forested terrain.
Due to typically severe winter conditions, most of
the pegmatite mines in this area are operated from
March through October. The Mt. Mica mine is
operated year round, weather permitting.

GEOLOGY
The Mt. Mica tourmaline deposit is a large, pocketbearing granitic pegmatite that belongs to the
Oxford pegmatite field of southwestern Maine (Wise
and Francis, 1992). The Oxford field is spatially related to the Sebago batholith. Pegmatites are concentrated within and around the northeastern margin of
this batholith, and they are therefore inferred to be
genetically related. Uranium-lead isotopic data indicate that the age of the Sebago Batholith is 296 3
million years (Foord et al., 1995).
According to recent observations by the authors,
the Mt. Mica pegmatite strikes northeast and dips
moderately southeast (typically about 30) within

GEMS & GEMOLOGY

SUMMER 2005

Figure 4. The images on these post cards show mining activities at Mt. Mica around the turn of the 20th century.
A large derrick (left; printed by The Hugh C. Leighton Co., Portland, Maine) was used to remove boulders from a
deep trench (right; printed by The Metropolitan News Co., Boston, Massachusetts). Remnants of the stacked boulders shown in the right image still can be seen today. From the private collection of Jane C. Perham.

the metasedimentary host rock. It is exposed for


about 135 m along strike and ranges in thickness
from about 1.5 m at the western exposure near the
surface to over 8 m thick in places further down dip.
The pegmatite is poorly zoned, with a thin (25 cm)
wall zone and a 1.5-m-thick intermediate zone in
the thicker portion of the dike. The intermediate
zone consists principally of quartz and K-feldspar,
with lesser amounts of schorl and muscovite. The
core consists mainly of quartz, microcline, and
schorl, with local pods of cleavelandite and rare
areas of lepidolite with spodumene, pollucite, cassiterite, columbite, and very rare beryl. Pockets are
relatively abundant in the central area of the dike

(again, see figure 5), where Coromoto Minerals has


averaged about one cavity every 3 m. In this area,
the miners have learned to recognize several indications of pocket mineralization, including wisps of
lepidolite, large muscovite books (particularly
near pods of massive quartz), masses of friable
cleavelandite, rust-colored fractures crosscutting the
inner zones, and large schorl crystals that point
downward toward the cavities.

MINING AND PRODUCTION


Coromoto Minerals began mining at Mt. Mica in
July 2003; they have posted detailed reports about

Figure 5. The locations


of both historic and
modern mining activities at Mt. Mica are
shown on this diagram.
From a drawing by
Gary Freeman.

MT. MICA TOURMALINE

GEMS & GEMOLOGY

SUMMER 2005

153

Figure 6. Mt. Mica is


located in southwestern
Maine near the small
town of South Paris.

Figure 7. An excavator and dump truck are used


to remove mined material from the open cut at
Mt. Mica, as shown in this October 2004 photo
by Gary Freeman.

154

MT. MICA TOURMALINE

Figure 8. The entrance to a large pocket (no. 7) containing mostly green tourmaline is marked by the
small tunnel just in front of the men in the pit. Note
the location of this pocket within a pronounced bulge
in the thickness of the predominantly light-colored
pegmatite. Photo by Alexander Falster.

GEMS & GEMOLOGY

SUMMER 2005

significant finds on the Internet at www.coromotominerals.com. In addition, a separate article


(Simmons et al., 2005) will review the history and
recent specimen mining in more detail.
The miners (Gary and Mary Freeman, together
with Richard Edwards) initially used an excavator
to expose the pegmatite from beneath the old dump
material. Subsequently, they used drilling and blasting to mine the pegmatite in an open cut, and they
removed material from the pit using an excavator
and dump truck (figure 7).
Several small pockets containing smoky quartz
and blue apatite were encountered in 2003, but the
first significant tourmaline discovery did not occur
until May 2004, when pocket no. 7 was entered.
This cavity was found in an area where the pegmatite showed a local bulge in thickness (figure 8).
Hundreds of gemmy crystals of predominantly
green tourmaline were produced. Most ranged up to
3 cm long and 0.5 cm in diameter, with pink cores
that were commonly altered to a pink clay. The
zoned crystals with unaltered pink cores resembled
the classic watermelon pieces pictured in Hamlin
(1895). A modest number of small gem-quality pink
crystals also were recovered from this pocket. In all,
about 1 kg of gem rough was produced, and most of
the faceted stones examined for this article came
from this pocket.

Figure 9. In June 2004, fine tourmaline crystals were


recovered from pocket no. 10. The crystal on the left
(approximately 10 cm long) forms the top portion of
the reassembled specimen shown in figure 10. The
crystal on the lower right is part of a separate specimen. A chisel is shown for scale. The dark material
consists of fragments of the overlying metasedimentary host rock. Photo by Gary Freeman.

MT. MICA TOURMALINE

Figure 10. Two of the most significant tourmaline


crystals recently recovered by Coromoto Minerals
are shown here. On the left is a 19-cm-tall tricolored elbaite crystal (reassembled from three
pieces) on a matrix of cleavelandite and lepidolite
that was recovered from pocket no. 10. On the
right is a multicolored crystal from pocket no. 28
that measures 22 cm tall, and forms the top portion of a crystal that was originally 54 cm long.
Chemical analyses showed the colored portion to
be elbaite, while the thin black flat termination is
foitite. Composite photo by Gary Freeman.

In June 2004, the miners opened a series of pockets that appeared to line up with pocket no. 7 and
the 1979 Dagenais pocket (again, see figure 5). It is
possible that this zone of pockets extends even further down dip. Most notable was pocket no. 10,
which contained more well-formed green and multicolored tourmalines (figure 9), including a broken
19-cm-long color-zoned crystal that is probably the
finest tourmaline specimen found at Mt. Mica and
one of the best ever mined in New England (see the
left crystal in figure 10).
From June through December 2004, an additional 18 pockets were encountered. The companys
most important discovery was made in late
December 2004, with the opening of pocket no. 28
(figure 11). This cavity, which exceeded 7 m long,
consisted of three connected chambers. Several
cathedral-type smoky quartz crystals weighing ~20
kg were found at the edges of the pocket. The miners eventually recovered several hundred bicolored
green/pink and green/colorless tourmaline crystals
ranging up to 58 cm long and <1 cm in diameter

GEMS & GEMOLOGY

SUMMER 2005

155

Figure 11. The largest and most important cavity


found by Coromoto Minerals at Mt. Mica was 2004
pocket no. 28. Here, Richard Edwards (left) and Frank
Perham are shown at the beginning of the excavation
in late December. Photo by Gary Freeman.

by screening the pocket mud. They also recovered


some much larger tourmalines, including a 22-cmtall elbaite crystal that grades from reddish pink to
orange with a thin black flat termination (see figure
10, right). When four additional pieces of this crystal
were subsequently recovered, it was determined
that the entire crystal was originally 54 cm long
the largest elbaite tourmaline known from Maine
and perhaps the largest from North America.
The 2005 mining season began in March, with
the miners clearing debris from the pit and driving a

Figure 12. A small loader is used to remove pegmatite


material from the underground workings at Mt. Mica.
June 2005 photo by Brendan Laurs.

156

MT. MICA TOURMALINE

decline (tunnel) into the pegmatite from the northeastern portion of the 2004 open cut, near pocket no.
28. A new loader for working underground was purchased (figure 12), and Jim Claninan experienced
gem pegmatite miner from southern California
joined the mining crew. As of mid-June 2005, the
decline reached approximately 20 m deep and they
had encountered nine small pocketssome with
fine green and greenish blue gem tourmaline.
The crystal specimens produced by Coromoto
Minerals are being marketed to mineral collectors
through Graeber & Himes, Fallbrook, California. As
a byproduct of the crystal mining, the mine owners
have accumulated a few kilograms of gem rough.
Most of this material shows various shades of green
(from pocket nos. 7 and 11). So far, a few dozen gemstones (described below) have been faceted from
broken crystals. As mining progresses in 2005, such
rough material will continue to be stockpiled for
future cutting and marketing.

MATERIALS AND METHODS


Standard gemological properties were obtained on
45 faceted Mt. Mica tourmalines that were cut from
material produced from May 2004 to early 2005.
The samples weighed 0.7816.48 ct and were
faceted in a variety of shapes, including rectangular
(some with a checkerboard style), round, oval, pear,
square, and freeform. The samples reportedly were
not treated in any way.
We used a GIA Instruments Duplex II refractometer with a near-sodium equivalent light source
for refractive index readings, and determined specific gravity by the hydrostatic method. The samples
were tested for fluorescence in a darkened room
with four-watt long- and short-wave UV lamps.
Internal features were observed with a standard
gemological microscope, and a polariscope was used
to view optic figures and check for strain. Inclusions
in four samples were investigated by Raman spectroscopy at GIA in Carlsbad, using a Renishaw 2000
Ramascope.
Quantitative chemical analyses were obtained
by electron microprobe at the University of New
Orleans, Louisiana, on 363 fragments of tourmaline
from pocket nos. 7 and 28. Some of these samples
were taken from the same pieces of rough used to
facet the stones described above. The fragments
were mounted on 1 inch (2.5 cm) glass disks with
epoxy, and were ground and polished with 0.05 m
alumina powder. The mounted samples were then

GEMS & GEMOLOGY

SUMMER 2005

ultrasonically cleaned and carbon coated. When


possible, the samples were mounted so that analytical traverses could be performed from core to rim.
Also analyzed were two Mt. Mica tourmaline crystals that were collected in 1890 by A. C. Hamlin
(from the collection of Peter Lyckberg) and 10 additional tourmaline crystals mined in 1964 by Frank
Perham (from the collections of Ray Sprague and
Jane Perham). These crystals were partially mounted in epoxy so that the smoothest, glass-like prism
surfaces could be used for analysis.
Analyses were conducted with an ARL SEMQ
electron microprobe operated at an acceleration
potential of 15 kV, a beam current of 15 mA, and a
spot size of 2 m. An acquisition time of 45 seconds
per spot was used. The number of cations in the formula of each analysis was calculated so that the
data could be plotted and the tourmaline species
identified. Since some elements in tourmaline (i.e.,
boron, lithium, and hydrogen) cannot be measured
by electron microprobe, we calculated the cations
(and wt.% oxides) according to standard assumptions and conventions (see Deer et al., 1992).

RESULTS AND DISCUSSION


The gemological properties are summarized in table
2, with details described below.
Visual Appearance. Overall, the faceted samples
could be separated by color into groups of yellowish
green, pink (including orangy pink and brownish

purplish red), greenish blue, and bicolored or tricolored with green, pink, and/or near-colorless zones.
Most of the stones showed weak-to-moderate saturation and light-to-medium tonewith dark tones
seen in some of the green stones that were cut with
their tables oriented perpendicular to the c-axis.
All of the samples were transparent. In general,
the pink, yellowish green, and greenish blue stones
were lightly included (commonly with no inclusions visible to the naked eye), whereas the multicolored samples contained obvious partially healed
fractures and feathers.
Physical Properties. There were only slight variations
in the refractive indices, which could not be correlated to color. The most typical values were no = 1.639
and ne = 1.620 (yielding a birefringence of 0.019). Although S.G. values ranged from 3.04 to 3.08, the
lower values (3.043.05) were obtained for the multicolored stones, probably due to the presence of
abundant fluid inclusions. Most of the samples that
were only lightly included had S.G. values ranging
from 3.05 to 3.07.
In the polariscope, typical uniaxial optic figures
could be resolved in all of the stones in which the
faceting did not obscure view of the optic axis.
Subtle patchy or sector-like patterns were seen
along the optic axis direction in a few samples; no
evidence of significant strain was noted in any of
the tourmalines.
The gemological properties of the Mt. Mica samples are consistent with those reported in the litera-

TABLE 2. Properties of the 45 faceted samples of tourmaline from Mt. Mica, Maine.
Property
Color
Pleochroism

Clarity
Optic character
Refractive indices
no
ne
Birefringence
Specific gravity
UV fluorescence
Short-wave
Long-wave
Internal features

MT. MICA TOURMALINE

Description
Most were yellowish green or multicolored in yellowish green, pink, and/or near colorless. Other stones
were pink, orangy pink, brownish purplish red, and greenish blue.
Yellowish green: Weak to strong (depending on color saturation), in yellowish green and slightly bluish green
Pink: Weak, in purplish pink and orangy pink
Greenish blue: Weak or very weak, distinguishable by a slight brown tint
Transparent; lightly to heavily included
Uniaxial negative
1.6381.640
1.6191.622
0.0180.020
3.043.08
Inert
Inert
Most common were partially healed fractures, feathers, and growth zoning. Other features included color
zoning, fine needle-like tubes, linear or planar trails of pinpoints, planar two-phase (liquid-gas) inclusions,
mineral inclusions (feldspar and low-relief grains that could not be identified), and cavities.

GEMS & GEMOLOGY

SUMMER 2005

157

Figure 13. Network stringers of filamentary fluid


inclusions known as trichites are characteristic of
tourmalines from many localities, including Mt. Mica.
Photomicrograph by John I. Koivula; magnified 10x .

Figure 14. Sharp-edged, elongated straight-to-angular


growth zoning, as revealed by shadowed illumination
in this Mt. Mica tourmaline, is often encountered in
tourmalines in general. Photomicrograph by John I.
Koivula; magnified 15x .

ture for gem tourmaline (see, e.g., Webster, 1994).


Although the lowest R.I. values we obtained were
slightly below the range reported by Webster (1994),
they fell within the values given by Dunn (1975) for
tourmaline from Newry, Oxford Co., Maine
(1.6121.644). The lack of any correlation between
R.I. and color in our Mt. Mica samples also was documented for Newry tourmaline by Dunn (1975).
However, the weak violet fluorescence to shortwave UV observed by Dunn (1975) in pink and red
Newry tourmaline was not found in the similar-colored samples from Mt. Mica that we studied.
Microscopic Features. The most conspicuous inclusions consisted of partially healed fractures and

feathers. A wide diversity of characteristics related to healed fractures were recorded: flat planar-tocurved aggregates of minute fluid inclusions, spindle- and irregular-shaped trichites (figure 13), linear or planar trains of pinpoints, and flat two-phase
inclusions with conspicuous gas bubbles suspended
within fluid. Growth zoning was also quite common, but relatively inconspicuous, forming linear,
planar, or angular patterns (figure 14). The growth
zoning was most prevalent in the greenish blue
samples and the multicolored stones; in the latter,
the growth lines were oriented parallel to the color
zoning.
Other microscopic features included fine needle-like tubes (often in parallel arrays) and small

Figure 15. This inclusion in a Mt. Mica tourmaline


was identified as feldspar by Raman analysis. Fluid
inclusions also are visible in the immediate vicinity.
Photomicrograph by John I. Koivula; magnified 20x .

Figure 16. Cross-polarized light has produced interference colors in this inclusion, which is probably a crystal of tourmaline within tourmaline from Mt. Mica.
Photomicrograph by John I. Koivula; magnified 15x .

158

MT. MICA TOURMALINE

GEMS & GEMOLOGY

SUMMER 2005

cavities. Mineral inclusions were seen in six of


the 45 samples, typically as isolated, high-relief,
rather equant colorless grains. In three of the samples, these inclusions were identified as feldspar
by Raman analysis (see, e.g., figure 15), probably
sodic plagioclase (i.e., cleavelandite), since this
variety of albite is commonly associated with the
tourmaline in the Mt. Mica gem pockets. Two
additional samples contained low-relief colorless
grains exhibiting strong birefringence (figure 16),
but these could not be identified by Raman analysis due to their small size and/or position within
the stones. Their appearance suggested that they
were inclusions of tourmaline within their tourmaline hosts.

shown in table 4. As shown in figure 17, the analyses mostly fell in the elbaite field, although a few
corresponded to rossmanite. The elbaite colors
ranged from dark green/black to lighter green to
pink. Most of the pink and red elbaites were distinctly enriched in Ca (liddicoatite component) relative to other colors. The green tourmalines were
mostly lower in Ca and richer in Na (elbaite component). Near-colorless samples tended to have
low Ca and greater X-site vacancies (rossmanite
component).
TABLE 3. General chemical formulas of tourmaline
species found at Mt. Mica.
Formulaa

Species

Chemical Composition. The tourmalines analyzed


for this study were found to consist of four species:
elbaite, schorl, foitite, and rossmanite (table 3).
Notably, all four species were present in one pocket (no. 28). Representative electron-microprobe
data for tourmalines from pocket nos. 7 and 28 are

Elbaite
Schorl
Rossmanite
Foitite
a

Na(Li1.5,Al1.5)Al6(BO3)3Si6O18(OH)4
NaFe2+
3 Al6(BO3)3Si6O18(OH)4
(LiAl2)Al6(BO3)3Si6O18(OH)4
(Fe2+
2 Al)Al6(BO3)3Si6O18(OH)4

= vacancy

Figure 17. This diagram


shows the X-site composition of all the tourmaline samples analyzed from Mt. Mica
pocket nos. 7 and 28.
The analyses predominantly fell within the
elbaite field, although a
few points fell on the
border or slightly into
the rossmanite field.
Black tourmaline samples from these pockets
consisted of schorl or
foitite, but these compositions are not shown
on this diagram.

MT. MICA TOURMALINE

GEMS & GEMOLOGY

SUMMER 2005

159

TABLE 4. Representative compositions by electron microprobe of various colors and species of Mt. Mica tourmaline.a
Pale
pink
Elbaite

Chemical
composition
Oxides (wt.%)
SiO2
TiO2
B2O3 (calc.)
Al2O3
FeOtot
MnO
MgO
CaO
Li2O (calc.)
Na2O
K2O
H2O (calc.)
F
Subtotal
O=F
Total

36.80
bdl
11.02
42.89
0.07
0.12
bdl
0.07
1.90
2.00
bdl
3.33
0.99
99.18
0.42
98.76

Pink
Elbaite
36.78
bdl
11.10
42.46
0.11
0.55
bdl
1.11
2.11
2.01
0.04
3.36
1.01
100.62
0.42
100.20

Ions on the basis of 31 (O,OH,F)


Si
5.802
5.755
Al
0.198
0.245
Tet. sum
6.000
6.000
B
3.000
3.000
Al (Z)
6.000
6.000
Al
1.771
1.585
Ti
bdl
bdl
0.009
0.015
Fe2+
Mn
0.015
0.073
Mg
bdl
bdl
Li
1.204
1.326
Y sum
3.000
3.000
Ca
0.011
0.186
Na
0.610
0.609
K
bdl
0.007
Vacancy
0.378
0.198
X sum
1.000
1.000
F
0.496
0.498
OH
3.503
3.502

Red
Elbaite
36.84
bdl
11.07
42.70
bdl
0.29
bdl
1.01
2.02
1.55
0.01
3.40
0.88
99.78
0.37
99.41
5.783
0.217
6.000
3.000
6.000
1.683
bdl
bdl
0.039
bdl
1.277
3.000
0.170
0.471
0.002
0.357
1.000
0.435
3.564

Light
green
Elbaite
36.79
bdl
10.99
41.78
1.33
0.44
0.03
0.32
1.82
2.07
0.02
3.33
0.97
99.91
0.41
99.50
5.815
0.185
6.000
3.000
6.000
1.599
bdl
0.176
0.059
0.006
1.160
3.000
0.054
0.634
0.004
0.308
1.000
0.486
3.514

Green
Elbaite
36.59
0.08
10.80
38.67
4.46
1.70
0.03
0.15
1.43
2.38
bdl
3.18
1.14
100.60
0.48
100.12
5.889
0.111
6.000
3.000
6.000
1.226
0.009
0.600
0.231
0.007
0.927
3.000
0.026
0.743
bdl
0.231
1.000
0.582
3.417

Yellowgreen
Elbaite

Olive
green
Elbaite

Bluegreen
Elbaite

36.63
bdl
10.97
41.93
0.96
0.41
bdl
0.22
1.87
2.22
0.02
3.27
1.09
99.61
0.46
99.15

36.76
0.18
11.00
41.57
1.79
0.82
0.02
0.14
1.71
2.12
0.02
3.35
0.95
100.42
0.40
100.02

36.55
bdl
10.79
39.55
4.35
0.43
bdl
0.06
1.48
2.10
0.02
3.31
0.88
99.52
0.37
99.15

5.801
0.199
6.000
3.000
6.000
1.628
bdl
0.126
0.055
bdl
1.190
3.000
0.038
0.683
0.004
0.275
1.000
0.548
3.451

5.805
0.195
6.000
3.000
6.000
1.541
0.021
0.236
0.109
0.005
1.087
3.000
0.024
0.648
0.005
0.324
1.000
0.472
3.528

5.886
0.114
6.000
3.000
6.000
1.394
bdl
0.586
0.059
bdl
0.961
3.000
0.011
0.654
0.005
0.330
1.000
0.446
3.553

Blue
Elbaite
36.78
bdl
10.83
38.49
5.22
1.20
bdl
0.75
1.49
2.00
0.03
3.12
1.31
101.22
0.55
100.67
5.900
0.100
6.000
3.000
6.000
1.177
bdl
0.700
0.163
bdl
0.960
3.000
0.129
0.620
0.005
0.245
1.000
0.664
3.335

a Selected from analyses of 363 samples from pocket nos. 7 and 28, as well as the two 1890 Hamlin and the ten 1964 Perham crystals. Backgrounds
were determined by the mean atomic number (MAN) method (Donovan and Tingle, 1996). Standards used include both natural and synthetic materials:
albite (Na), adularia (K), quartz and clinopyroxene (Mg, Ca, Fe, Ti), chromite (Cr), rhodonite (Mn), V2O5 (V), PbO (Pb), ZnO (Zn), Bi-germanate (Bi), sillimanite (Si and Al), and fluortopaz (F). MAN standards used in addition to those above were corundum, fayalite, hematite, rutile, MgO, SrSO4, and ZrO2 .
Detection limits (in wt.% oxide): Ti = 0.008, Fe = 0.005, Mn = 0.006, Mg = 0.012, Ca = 0.007, and K = 0.014. Detection limits of elements analyzed but
not detected (in wt.% oxide): V = 0.007, Cr = 0.013, Zn = 0.022, Pb = 0.009, and Bi = 0.016. Li2O, B2O3, and H2O were calculated based on an
assumed elbaite tourmaline stoichiometry. Abbreviation: bdl = below detection limit.

Foitite is a rare tourmaline species that is iron


rich in the Y-site and has an X-site that is more than
50% vacant (MacDonald et al., 1993); it formed the
thin black flat terminations on some of the pink
elbaite crystals (e.g., figure 10, right). The schorl
analyses corresponded to the black basal portions of
some of the elbaite crystals. The analyzed crystals
that were mined in 1890 and 1964 were all elbaite,
with compositions that overlapped the more recently mined tourmaline.
As expected, the chromophoric elements Fe, Ti,
and Mn showed strong correlation with color (see
table 4 and figure 18). Fe was highest in the black

160

MT. MICA TOURMALINE

tourmaline, and virtually absent in the near-colorless and pink tourmaline. The next-highest Fe values were found in the olive green, green, and
blue tourmaline. Ti correlated with elevated Fe in
the black and green to olive green samples. All
other colors had very low Ti contents. On average,
Mn was highest in the blue and green tourmaline,
and was lowest in all light-colored tourmaline.
Fluorine contents were relatively constant, averaging 1.08 wt.% and ranging from 0.8 to 1.31 wt.%
in all analyses. Mg was present mainly in the black
tourmaline, where it averaged 0.54 wt.% MgO and
ranged from below the detection limit to 1.38 wt.%

GEMS & GEMOLOGY

SUMMER 2005

Light
blue
Elbaite

Colorless
Elbaite

36.60
bdl
10.87
39.50
3.45
1.11
bdl
0.88
1.69
2.04
0.02
3.20
1.17
100.53
0.49
100.03

36.83
bdl
11.02
42.90
bdl
bdl
bdl
bdl
1.91
2.03
bdl
3.27
1.11
99.09
0.47
98.62

37.26
bdl
11.10
43.43
0.05
0.20
bdl
bdl
1.79
1.48
bdl
3.38
0.94
99.63
0.40
99.24

36.51
0.54
10.48
32.65
13.20
0.26
1.30
0.09
0.54
1.82
0.03
3.08
0.88
101.38
0.47
100.91

5.853
0.147
6.000
3.000
6.000
1.299
bdl
0.461
0.151
bdl
1.089
3.000
0.151
0.633
0.004
0.212
1.000
0.590
3.409

5.809
0.191
6.000
3.000
6.000
1.785
bdl
bdl
bdl
bdl
1.214
3.000
bdl
0.622
bdl
0.378
1.000
0.555
3.444

5.832
0.168
6.000
3.000
6.000
1.843
bdl
0.007
0.026
bdl
1.124
3.000
bdl
0.449
bdl
0.551
1.000
0.467
3.533

6.055
0.000
6.055
3.000
6.000
0.381
0.068
1.831
0.036
0.322
0.361
2.999
0.015
0.586
0.007
0.393
1.000
0.589
3.411

Colorless
Rossmanite

Black
Schorl

Black
Foitite
36.36
0.07
10.45
34.77
12.54
0.63
bdl
bdl
0.52
0.92
bdl
3.19
0.88
100.32
0.37
99.95
6.046
0.000
6.046
3.000
6.000
0.814
0.008
1.743
0.088
bdl
0.346
3.000
bdl
0.298
bdl
0.702
1.000
0.463
3.537

Elba, Italy (Pezzotta, 2001). Elba is the only locality


besides Mt. Mica where four tourmaline species
have been documented from a single gem pocket
(Pezzotta et al., 1998). Although it is not uncommon for multiple tourmaline species to be present
in a particular pegmatite deposit or district (see,
e.g., Selway, 1999), the occurrence of such a diverse
composition of tourmaline in a single pocket
appears to be quite unusual.

CONCLUSIONS AND FUTURE POTENTIAL


Gem- and specimen-quality tourmalines were produced in a wide variety of colors at Mt. Mica, mainly in the late 19th century and sporadically from the
1960s to 1990s. In 20042005, renewed mining by
Figure 18. These graphs show the average compositions of chromophoric elements Fe, Ti, and Mn (as
wt.% oxides) in various colors of tourmaline from Mt.
Mica. Fe and Ti are found mainly in the black and
green samples, whereas Mn is present in tourmaline of
all colors to varying degrees.

MgO. In all other colors, Mg was very low to below


the limit of detection.
Schorl is found as black tourmaline along the
pocket margins, and therefore formed early in crystallization of the miarolitic cavities. As the crystals
grew into the pocket, they were progressively
enriched in Al relative to Fe. As Fe diminished,
elbaite became the dominant tourmaline. Foitite,
and possibly rossmanite, represent the final stages
of tourmaline compositional evolution in the pockets. Interestingly, the black foitite caps noted on
some Mt. Mica tourmalines are similar to the
Mohrenkpfe found on elbaite crystals from

MT. MICA TOURMALINE

GEMS & GEMOLOGY

SUMMER 2005

161

Coromoto Minerals again produced attractive crystals and cut stones (figures 19 and 20), including
some of the finest tourmaline ever recovered from
the deposit.
Based on the deposit geology and on the projected
mineralization of the pegmatite down dip, Mt. Mica
shows considerable potential for additional tourmaline production. Future mining will focus on the
deeper extension of the 20042005 mineralized
zone, as well as on areas of the pegmatite to the
northeast that are closer to the surface. As the pegmatite is explored by both open-cut and underground methods, the miners anticipate that more
tourmaline will be found at this historic deposit.

Figure 20. Continued mining of the Mt. Mica deposit


is expected to yield additional gem tourmaline,
such as the bright pink and bluish green stones
shown here (9.86 and 4.77 ct). These tourmalines
were faceted in early 2005 by Dennis Creaser and
studied for this report. Courtesy of Coromoto
Minerals Inc.; photo by C. D. Mengason.

Figure 19. Well-formed gem-quality crystals of green


tourmaline such as this one (5.4 cm long) from Mt.
Mica pocket no. 11 are prized by collectors and connoisseurs. Courtesy of Graeber & Himes, Fallbrook,
California; photo Jeff Scovil.

ABOUT THE AUTHORS


Dr. Simmons is professor of mineralogy, and Mr. Falster is analytical instrumentation manager, in the Department of Geology and
Geophysics at the University of New Orleans, Louisiana. Mr. Laurs
is editor of Gems & Gemology. Mr. Koivula is chief gemologist at
the AGTA Gemological Testing Center in Carlsbad. Dr. Webber is
a research associate and adjunct professor at the University of
New Orleans and a graduate gemologist.
ACKNOWLEDGMENTS: The authors gratefully acknowledge
Gary and Mary Freeman, owners of Coromoto Minerals,
Oxford County, Maine, whose assistance and support made
this study possible. They generously provided descriptions of

162

MT. MICA TOURMALINE

the mining, photographs, and access to the property, and they


donated tourmaline samples for electron-microprobe analysis.
Their dedication to scientific investigation and particularly to
pegmatite research is greatly appreciated.
Additional tourmaline samples for electron-microprobe analysis
were donated by Peter Lyckberg (Luxembourg), Raymond
Sprague (Andover, Massachusetts), and Jane Perham (West
Paris, Maine). Raman analysis of selected inclusions was performed by senior research associate Sam Muhlmeister of the
GIA Gem Laboratory in Carlsbad. Gary Freeman, Raymond
Sprague, Roland Naftule (Nafco Gems, Scottsdale, Arizona),
and three additional reviewers made helpful suggestions for
improving the manuscript.

GEMS & GEMOLOGY

SUMMER 2005

REFERENCES
Bastin E.S. (1911) Geology of the pegmatites and associated
rocks of Maine. U.S. Geological Survey Bulletin 445, 152 pp.
Crowningshield R. (1966a) Developments and highlights at the
Gem Trade Lab in New York: Maine tourmaline. Gems &
Gemology, Vol. 12, No. 2, pp. 4344.
Crowningshield R. (1966b) Developments and highlights at the
Gem Trade Lab in New York: Maine tourmaline. Gems &
Gemology, Vol. 12, No. 3, pp. 7071.
Deer W.A., Howie R.A., Zussman J. (1992) An Introduction to
the Rock-forming Minerals, 2nd ed. Longman Scientific &
Technical, Essex, England.
Donovan J.J., Tingle T.N. (1996) An improved mean atomic number correction for quantitative microanalysis. Journal of
Microscopy, Vol. 2, No. 1, pp. 17.
Dunn P.J. (1975) On gem elbaite from Newry, Maine, U.S.A.
Journal of Gemmology, Vol. 14, No. 8, pp. 357368.
Fales M.G. (1995) Jewelry in America. Martha Gandy Fales and
Antique Collectors Club Ltd., Suffolk, England.
Foord E.E., Snee L.W., Aleinikoff J.N., King V.T. (1995) Thermal
histories of granitic pegmatites, western Maine, USA.
Abstracts with Programs, Geological Society of America
Annual Meeting, New Orleans, LA, Vol. 27, No. 6, p. A468.
Francis C.A. (1985) Maine tourmaline. Mineralogical Record,
Vol. 16, pp. 365388.
Hamlin A.C. (1873) The Tourmaline. James R. Osgood & Co.,
Boston, MA, 107 pp.; republished in 2004 by Rubellite Press,
New Orleans, LA.
Hamlin A.C. (1895) The History of Mount Mica of Maine, U.S.A.
and Its Wonderful Deposits of Matchless Tourmalines. Publ.

by A.C. Hamlin, 72 pp. plus color plates; republished in 2004


by Rubellite Press, New Orleans, LA.
King V.T., Ed. (2000) Mineralogy of Maine, Vol. 2: Mining
History, Gems, and Geology. Maine Geological Survey,
Augusta, ME, 524 pp.
MacDonald D.J., Hawthorne F.C., Grice J.D. (1993) Foitite,
3+
[Fe2+
2 (Al,Fe )]Al6Si6O18 (BO3) 3(OH)4, a new alkali-deficient
tourmaline: Description and crystal structure. American
Mineralogist, Vol. 78, pp. 12991303.
Perham J.C. (1987) Maines Treasure ChestGems and Minerals
of Oxford Co., 2nd ed. Quicksilver Publications, West Paris,
ME, 267 pp.
Pezzotta F. (2001) Heiss Begehrt: die Mohrenkpfe. Elba,
extraLapis No. 20, pp. 4849.
Pezzotta F., Guastoni A., Aurisicchio C. (1998) La rossmanite di
Rozn e dell Elba. Revista Mineralogica Italiana, Vol. 22, No.
1, pp. 4650.
Selway J.B. (1999) Compositional Evolution of Tourmaline in Granitic Pegmatites. Ph.D. dissertation, University of Manitoba,
Canada, 363 pp.
Simmons W.B., Freeman G., Falster A., Laurs B., Webber K. (2005)
New tourmaline production from Mount Mica: America's first
gem pegmatite. Rocks & Minerals, Vol. 80, No. 6 (accepted for
publication in November-December issue).
Webster R. (1994) Gems: Their Sources, Descriptions and
Identification, 5th ed. Revised by P.G. Read, ButterworthHeinemann, London.
Wise M.A., Francis C.A. (1992) Distribution, classification and
geological setting of granitic pegmatites in Maine.
Northeastern Geology, Vol. 14, No. 23, pp. 8293.

Mark your calendar for the

2006

GIA Gemological Research Conference


August 2627, 2006
To explore the most recent technical developments in gemology, GIA will host a Gemological Research Conference
in conjunction with the 4th International Gemological Symposium in San Diego, California.

nvited lectures, submitted oral presentations, and a poster session will


explore a diverse range of contemporary topics including the geology
of gem deposits, new gem occurrences, characterization techniques,
treatments, synthetics, and general gemology. Also scheduled is a oneday pre-conference field trip to the world-famous Pala pegmatite district
in San Diego County.
Abstracts should be submitted to gemconference@gia.edu (for oral
presentations) or ddirlam@gia.edu (for poster presentations). The abstract
deadline for all submissions is March 1, 2006. Abstracts of conference
presentations will be published in a special proceedings volume.

MT. MICA TOURMALINE

For further information,


contact the organizing
committee at
gemconference@gia.edu
or visit
www.gia.edu/gemsandgemology
Mark your calendar today!

GEMS & GEMOLOGY

SUMMER 2005

163

EDITORS
Thomas M. Moses and Shane F. McClure
GIA Gem Laboratory

CONTRIBUTING EDITORS
G. Robert Crowningshield
GIA Gem Laboratory, East Coast
Cheryl Y. Wentzell
GIA Gem Laboratory, West Coast

DIAMOND
Fracture Filled, with
Varying Results
Fracture filling is usually effective at
making a diamonds fractures less
obvious both to the unaided eye and
when viewed with magnification,
although a trained gemologist often
can identify the treatment relatively
easily. When the filling is not complete or is poorly executed, however,
identification can be much more challenging. Two diamonds recently submitted to the East Coast laboratory
highlighted these challenges.
The 9.01 ct rectangular modified
brilliant in figure 1 was quickly
identified as an artificially irradiated
diamond, but as noted in the last

Figure 1. This artificially irradiated 9.01 ct yellow diamond also


showed unusual clarity enhancement features.

164

LAB NOTES

issue of this journal (Spring 2005 Lab


Notes, p. 46) identification of one
treatment may not be the end of a
gemologists work. While the fractures in this diamond were generally
quite visible, some contained an
unusual texture and flow structure
(figure 2) that led us to examine the
stone more closely. A careful investigation at high magnification and
with the light at varying angles
revealed the telltale flash-effect colors associated with fracture filling.
We do not know exactly why the
treatment was ineffective in this
stone; however, a close look at figure
2 shows that the surface of the facet
has a mottled, hazy appearance,
which suggests that it may have been
burned. Excessive heat can cause a
high-lead glass filling to flow out of a

fracture, and the dendritic patterns


we observed (see also figure 3) support this possibility. One might further speculate that the excessive heat
could have been generated during the
color treatment process, since creating yellow color through artificial
irradiation requires subsequent
annealing. The heat necessary
(~600C) would certainly be sufficient to cause the filling material to
break down. Thus, the fractures likely became much more visible than
they were immediately after the fracture filling.
Another diamond, a 1.52 ct marquise, also had relatively obvious
fractures that would not be expected
in a clarity-enhanced stone. However, thorough examination with
magnification revealed a very unusu-

Figure 2. Heat treatment of the


diamond in figure 1 may have
created the unusual texture and
flow structure visible within this
fracture. Magnified 45x.

Figure 3. This dendritic pattern in


another fracture in the 9.01 ct
diamond indicates both the presence of a glass filler and exposure
to high heat. Magnified 60x.

GEMS & GEMOLOGY

SUMMER 2005

Figure 4. The unfamiliar pattern


within this fracture led to closer
examination of the 1.52 ct diamond for evidence of clarity
enhancement. Magnified 75x.

al pattern within the fractures (figure


4), which led us to look closely for
more classic indications of clarity
treatment. Figure 5 shows the very
weak flash-effect colors that we
found at higher magnification with
the aid of fiber-optic illumination.
Again, we do not know the exact reason for this poor result. However,
there may have been a problem with
the contact between the host diamond and the filler. These situations
usually arise when the glass does not
fill the entire fracture or it has been
altered in some way due to either
mishandling or possibly a flaw in
the treatment process.
Diamonds such as these can create great challenges for the practicing
gemologist. Anything unusual bears
careful scrutiny, but with diligence
the proper identification can often be
made using the standard instrumentation available to most gemologists
and jewelers.
Thomas Gelb and Matthew Hall

Editors note: The initials at the end of each item


identify the editor(s) or contributing editor(s)
who provided that item. Full names are given for
other GIA Gem Laboratory contributors.
GEMS & GEMOLOGY, Vol. 41, No. 2, pp. 164175

2005 Gemological Institute of America

LAB NOTES

Figure 5. These subtle flash-effect


colors in the 1.52 ct marquise are
further indications of clarity
treatment. Magnified 50x.

Large Diamond with Microinclusions of Carbonates


and Solid CO2
Natural diamonds occasionally contain microscopic inclusions of carbonates, water, apatite, and silicates (see,
e.g., O. Navon et al., Mantle-derived
fluids in diamond micro-inclusions,
Nature, Vol. 335, 1988, pp. 784789;
Winter 2004 Lab Notes, pp. 325326).
These inclusions are encapsulated
samples of mantle fluid, and thus are
important for geochemical study of
the mantle as well as for a better
understanding of diamond formation.
Although diamonds typically form at
a depth of around 140200 km, microinclusions of solid CO2 were reported
in a small rough type IaA diamond
that is thought to have formed at a
deeper origin, around 220270 km (M.
Schrauder and O. Navon, Solid carbon dioxide in a natural diamond,
Nature, Vol. 365, No. 6441, 1993, pp.
4244). The discovery of solid CO2 in
diamond has very important implications in the study of the earths mantle. However, that small crystal was
the only specimen reported in which
solid CO2 micro-inclusions had been
confirmed, except for a brief report of
CO2-bearing diamonds in a recent
publication (T. Hainschwang et al.,
HPHT treatment of different classes
of type I brown diamonds, Journal of
Gemmology, Vol. 29, 2005, pp.
261273). Thus, the East Coast laboratory was very interested to examine a

large faceted diamond that proved to


contain micro-inclusions of solid CO2
and carbonates, possibly from an even
greater depth in the mantle.
The 5.04 ct pear-shaped diamond
(14.94 x 9.26 x 6.27 mm; figure 6) of
unknown geographic origin was graded
D color. Although only a few very tiny
pinpoint inclusions were visible with
magnification, this stone displayed
extremely strong colorless graining
that created an undulating translucency throughout the entire diamond (figure 7). This severely affected its clarity,
resulting in a grade of SI2. Linear or planar graining occurred in a limited
region close to the point of the pear
shape. To the best of our recollection,
the undulatory graining in this stone is
the most prominent we have seen in
the GIA Gem Laboratory.
The infrared spectrum (figure 8)
showed no nitrogen- or boron-related
absorption, which suggests that it is a
type IIa diamond. This was supported
by the ultraviolet (UV)-visible absorption spectrum, which did not show any
absorption peaks. Unlike typical type
IIa diamonds, this stone displayed a
moderately strong yellow fluorescence
Figure 6. Spectroscopic analysis of
this 5.04 ct D-color diamond
revealed that it contains very
unusual micro-inclusions of solid
CO2 as well as carbonates.

GEMS & GEMOLOGY

SUMMER 2005

165

Figure 7. Hazy, undulatory graining is evident throughout the diamond in figure 6. This is the most
intense graining seen so far in the
laboratory, and it severely affected the diamonds clarity grade
(SI2). Image width is 8.0 mm.
to short-wave UV radiation. In the
Diamond Trading Company (DTC)
DiamondView, a moderately strong
blue-gray fluorescence and weak bluegray phosphorescence were observed;
the undulatory appearance was also
apparent (figure 9). Natural type IIa diamonds rarely phosphoresce.
The IR spectrum exhibited strong
absorptions of fundamental modes of
solid CO2 at ~2376 (3) and ~651 (2)
cm1. These two absorptions were so
strong that their positions could not be
precisely determined. Weak combination modes at 5141.3, 5019.5, 3753.8,
and 3625.3 cm1 also were observed.
Close examination of these absorption
positions and comparison with highpressure spectral data of solid CO2
revealed that the positions of these
combination modes closely fit those
seen in solid CO2 under a pressure of
6.3 0.4 GPa at room temperature (see
R. C. Hanson and L. H. Jones, Infrared
and Raman studies of pressure effects
on the vibrational modes of solid CO2,
Journal of Chemical Physics, Vol. 75,
1981, pp. 11021112). This pressure is
higher than the 5 0.5 GPa in the
stone reported by Schrauder and Navon
(1993). Extrapolation to a mantle temperature of 1,200C results in a pressure of 9.29.7 GPa, about 270290 km
in mantle depth. These very likely represent the conditions under which the

166

LAB NOTES

Figure 8. In the IR spectrum of the diamond in figure 6, strong absorptions of solid


CO2 occur at approximately 2376 and 651 cm1. Weak combination modes at
5141.3, 5019.5, 3753.8, and 3625.3 cm1 also were observed; the weak absorptions
at 1474.1 and 886.2 cm1 are most likely from micro-carbonate inclusions.
diamond originally crystallized. These
pressure regimes are much higher than
those in which most natural diamonds
have grown (56 GPa), which further
indicates that this diamond formed
deeper in the mantle.
The weak absorptions observed at
1474.1 and 886.2 cm1 (again, see figure
8) are most likely from micro-carbonate inclusions. These peak positions do
not entirely fit with those of published
calcite or magnesite positions at ambient conditions, and a pressure effect

Figure 9. The undulatory graining


seen in figure 7 is also apparent
in the DTC DiamondView fluorescence image.

like that of solid CO2 is highly possible.


The weak absorptions at 1274.3,
1113.5, 990.8, and 827.7 cm1 may be
from other components that also occur
as micro-inclusions. However, their
assignment is not clear at this
moment. Raman spectroscopy showed
no Raman peaks for solid CO2 or carbonates, but it did reveal many unusual
photoluminescence emissions at liquid-nitrogen temperature using 488 nm
laser excitation (figure 10). Except for
the 3H emission at 503.5 nm, little is
known about the assignment of other
observed lines, which rarely occur in
standard natural diamonds.
We believe the micro-inclusions
and associated graining are the main
cause of the undulatory hazy appearance. The micro-inclusions of solid
CO2 and carbonates in this gem diamond indicate a localized pocket of carbonate-rich material within the deep
mantle ( > 200 km). Furthermore, this
environment must have been stable
for an extended period, allowing for
the growth of this fairly large gem-quality diamond. This diamond represents
the largest of those so far reported
from the deep mantle. It is an unusual

GEMS & GEMOLOGY

SUMMER 2005

Figure 10. Except for the 3H emission at 503.5 nm, little is known about the assignment of the many photoluminescence emissions observed in the 5.04 ct diamond.
stone not only because of its special
gemological properties but also for its
importance in studying the earths
mantle and diamond formation.
Wuyi Wang, TM, Kyaw Soe Moe,
and Andy Hsi-Tien Shen

Light Blue Diamond,


with Type IIb and IIa Zones
Like some forms of nitrogen, boron
occurs as individual atoms substituting for carbon in the diamond structure. However, boron and nitrogen
typically do not coexist in natural diamonds. Unlike nitrogen-bearing diamonds and synthetic blue diamonds,
which frequently exhibit distinctive
growth zoning (best seen with the
DTC DiamondView), most natural
type IIb diamonds appear to have no
or only very subtlezoning.
An interesting 0.67 ct light blue
pear shape was recently submitted to
the West Coast laboratory for grading
(figure 11, left). Initial infrared absorption spectroscopy indicated that this
diamond was type IIa, which suggested
that the color was due to radiationrelated defects. Low-temperature visible absorption spectroscopy, however,
revealed the spectrum of a colorless IIa

LAB NOTES

diamond with no evidence of radiation


damage. The discrepancy between the
spectroscopy and the light blue bodycolor led us to undertake a careful
gemological examination.
The diamond was inert to longwave UV radiation and fluoresced a
very weak yellow to short-wave UV,
with brief, very weak blue phosphorescence. When tested with a gemological
conductometer, the stone showed
slight electrical conductivity in certain
directions. No color zoning was apparent when the pear shape was viewed

with the microscope and diffused


white light. Because in general these
properties are consistent with type IIb
diamonds, we decided to perform
detailed UV imaging and collect spectroscopic data from different parts of
the stone.
Phosphorescence in type IIb diamonds directly reflects the presence of
boron, so we examined the diamond
using the phosphorescence setup of
the DiamondView to try to map the
boron distribution. We were able to see
small blue phosphorescent areas near
the head and point of the pear shape
that were not visible in the belly (figure 11, right). The degree to which the
phosphorescent areas near the point
represented boron enrichment is
unclear, due to the tendency of pearshaped brilliant cuts to collect color at
the point. Our initial spectroscopic
data had been collected by passing the
beam through the girdle of the diamond at the belly; however, when the
same measurements were taken near
the head (where we saw the phosphorescent areas), the new absorption
spectra revealed the presence of low
concentrations of boron (figure 12).
Although it is possible that hydrogenboron complexes within the diamond
may contribute to some of the
observed features, the distinct differences in phosphorescence and the
spectroscopic data collected from various parts of the stone strongly suggest
that the diamond consisted of a mix-

Figure 11. This 0.67 ct Light blue pear-shaped diamond (left) was submitted for a colored diamond grading report. A DiamondView phosphorescence image (right) shows weak blue phosphorescence concentrated near
the head and point of the pear shape.

GEMS & GEMOLOGY

SUMMER 2005

167

Natural Type Ib Diamond with


Unusually High Nitrogen Content
Nitrogen is by far the most common
impurity in diamonds. If most of the
nitrogen in a diamond is present as
isolated nonaggregated atoms within
the crystal lattice (i.e., single substitutional N, or C-centers), the diamond is
considered to be type Ib. Even very
small amounts of this type of nitrogen
(<10 ppm) can produce vivid yellow or
orange-yellow colors, sometimes
resulting in the highly sought-after
canary diamonds (see, e.g., the article by J. M. King et al. on pp. 88115
of this issue). Most natural type Ib diamonds contain less than 100 ppm of
single substitutional N, though synthetic diamonds commonly contain
much more. The concentration of Ccenters in synthetic yellow diamonds
typically ranges up to ~200 ppm, but
may be even higher in deeply saturated yellow-orange or brown colors.
The West Coast laboratory recently had the opportunity to examine a
natural diamond with unusually high
Ib nitrogen content. The 0.24 ct Fancy
Dark pink-brown oval cut shown in
figure 13 contained so much nitrogen
that the signal intensity in the region
of the mid-infrared absorption spectrum where nitrogen is measured (the
one-phonon region, ~1400800 cm1)
exceeded the limits of our detector.
Higher-resolution data revealed that
the diamond was mostly type Ib with

1
2

Figure 13. This 0.24 ct Fancy


Dark pink-brown oval-cut diamond proved to have an unusually high type Ib nitrogen content.

Figure 12. Mid-infrared (top) and visible (bottom) absorption spectra were collected from two different portions of the Light blue diamond, as indicated by
the two lines in the phosphorescence image (top, inset). As would be expected
from that image, the presence of boron is apparent only in the no. 2 spectra, as
evident from the 2803 cm1 IR feature and the increase in absorption at the
red end of the visible spectrum (~600700 nm).
ture of type IIa and IIb zones.
The Lab Notes section has published reports on a few other mixed
type IIa and IIb diamonds (see Fall
1963, p. 85; Winter 19661967, p. 116;
Fall 1993, p. 199; Summer 2000, pp.
156157). Although spectroscopic data

168

LAB NOTES

were not available from these reports,


the nature of the type IIa/IIb zoning
appears very similar to that of the diamond described here. This type of
zoning in blue diamonds may be more
common than we have believed.
Christopher M. Breeding

GEMS & GEMOLOGY

SUMMER 2005

Figure 14. The mid-infrared absorption spectrum of the diamond in


figure 13 displays peaks at 1344 and 1130 cm1, indicating a very high
type Ib nitrogen content, as well as some type IaA nitrogen and small
amounts of hydrogen.

Figure 15. The visible spectrum of the diamond in figure 13 shows


complete absorption below ~515 nm and a broad band centered at
~610 nm; the latter probably contributes the pink component to
the diamonds color.

LAB NOTES

some type IaA (aggregated) N and also


contained minor hydrogen (figure 14).
Using FTIR software specially
designed to measure nitrogen concentrations in diamond, we calculated levels of 364 ppm Ib nitrogen and 132
ppm IaA. This diamond contained
more Ib nitrogen than any previous
natural or synthetic diamond that we
can recall examining in the laboratory.
It is unusual that a diamond with such
a high N content escaped aggregation
and remained in its isolated form.
When examined with magnification, the diamond showed irregular
color zoning and only a few small
feathers and needle-like inclusions,
similar to those we have previously
seen in type Ib yellow diamonds (see,
e.g., Spring 1994 Lab Notes, pp.
4142; King et al., 2005). Strain was
very weak and was localized around
the inclusions. The diamond showed
no reaction to either long- or shortwave UV radiation from a handheld
lamp. In the desk-model spectroscope,
general absorption was present below
~510 nm. More detailed data, collected at cryogenic temperatures using a
high-resolution visible absorption
spectrometer, revealed very strong
absorption through most of the visible spectrum, with complete absorption below ~515 nm due to the abundance of Ib nitrogen (figure 15). The
steady increase in absorption toward
the blue end of the spectrum was
accompanied by a broad band centered at ~610 nm that likely contributed the pink component of the
face-up color. The origin of this band
is unclear, but it may be an extension
of the deformation-related 550 nm
band that produces pink-red color in
other natural diamonds.
The diamond was also examined
with the DTC DiamondView. Several
irregularly zoned areas of very weak
green fluorescence were separated by
more intense green boundaries. The
zoning pattern of one region on the
table was particularly interesting
because it resembled the skeleton of a
fish (figure 16). The reason for this
unusual zoning pattern is not known
at this time.

GEMS & GEMOLOGY

SUMMER 2005

169

With Unusual Laser Drill Holes

Figure 16. When exposed to


strong short-wave UV radiation
in the DTC DiamondView, the
0.24 ct pink-brown diamond
shows zones that resemble the
skeleton of a fish.

This stone is an excellent example of a rare and unique scientific


treasure that belies its unremarkable
appearance.
Christopher M. Breeding

Figure 17. This large cavity in an


8.38 ct diamond does not resemble any natural etch channels or
laser drill holes seen previously in
the laboratory. Magnified 35x.

Over the past few years, the challenges


posed by innovative diamond laser
drilling techniques have increased
markedly (see S. F. McClure et al., A
new lasering technique for diamond,
Summer 2000 Gems & Gemology, pp.
138146; Summer 2002 Lab Notes, pp.
164165). Recently, a sharp-eyed diamond grader in the East Coast laboratory noticed something very strange in
an 8.38 ct round brilliant and brought
it to the attention of the Identification
Department.
Our experience has been that
laser drill holes are circular in appearance, while etch channels are angular, as they follow the diamonds
crystal structure. The large cavities
in this diamond (see, e.g., figure 17)
were unlike any natural etching we
have previously encountered. Yet,
where they reached the surface, these
holes were neither round nor angular; rather, their outlines consisted of
numerous semicircles (figure 18).
Long grooves, some of them with a
black residue, ran down the length of
the cavities (figure 19). These unusual striations mimicked what is characteristically seen in naturally occurring etch channels; if not scrutinized
carefully, they could have been dismissed as natural inclusions.

Figure 18. Where the drill holes


reach the surface of the diamond, they show an unusual
outline, again unlike any previously encountered in the lab.
Magnified 105x.

170

LAB NOTES

In this case, however, all of the


questionable cavities tapered to solid
inclusions (again, see figure 17). In
addition, the grooves were rounded
along their edges, and the dark color
and sugary texture were reminiscent
of laser drilling, which is known to
leave a black residue. All of these
clues led us to conclude that the cavities in this diamond were caused by
laser drilling, perhaps the result of several holes drilled into the same area.
One could speculate that the
drillers of this diamond were either
trying to mimic natural etch channels or perhaps had outlined some
near-surface crystals and drilled them
completely out.
Joseph Astuto and Thomas Gelb

Yellowish Orange
MAGNESIOAXINITE
The West Coast laboratory recently
examined a 4.90 ct transparent brownish yellowish orange piece of rough
(figure 20) that was submitted for
identification by JOEB Enterprises of
San Diego, California. This piece of
rough, reportedly from Tanzania, was
tabular in shape (17.75 x 8.86 x 3.64
mm) with parallel striations on some
of its crystal faces.
Observation with a polariscope
revealed a distinct biaxial optic figure.

Figure 19. Dark, rounded grooves


along the length of this cavity in
the 8.38 ct diamond are further
indications of laser drilling.
Magnified 45x.

GEMS & GEMOLOGY

SUMMER 2005

Figure 20. This 4.90 ct piece of


rough, reportedly from Tanzania,
is the first known sample of yellowish orange magnesioaxinite
seen in the GIA Gem Laboratory.

The material also displayed moderate


yellow and orange pleochroism. A
small flat that had been polished on
the surface allowed us to obtain refractive indices of = 1.657, = 1.660, and
= 1.668, resulting in a birefringence
of 0.011 and indicating that the optic
sign was positive. The specific gravity,
measured hydrostatically, was 3.18.
The sample fluoresced moderate
orange to long-wave UV radiation and
weak orange to short-wave UV. Using
a desk-model spectroscope, we observed only a broad absorption band
from 450 to 520 nm. Examination
with a gemological microscope
revealed fingerprints, fractures, needles, negative crystals, and distinct
orange-yellow and near-colorless zoning. The identification of this sample
did not appear to be straightforward
on the basis of physical and optical
properties alone, so we proceeded with
more advanced testing.
Both Raman and X-ray powder
diffraction analysis identified the
material as a member of the axinite
group. Axinite, Ca2(Fe, Mn, Mg)Al2B
Si4O15(OH), is a borosilicate mineral
group made up of a solid-solution
series with three end members: ferroaxinite, manganaxinite, and magnesioaxinite. The species designation
depends on the relative amounts of
iron, manganese, and magnesium,
respectively. The R.I. and S.G. of the
sample in question were almost identical to those reported for blue magnesioaxinite from Tanzania (E. A.

LAB NOTES

Jobbins et al., Magnesioaxinite, a


new mineral found as a blue gemstone from Tanzania, Journal of
Gemmology, Vol. 14, 1975, pp.
368375). Puzzled by the fact that we
were unable to locate any reference to
magnesioaxinite with a yellowish
orange color, we asked one of our
research scientists to take the sample
to Rutgers University in New Jersey
for electron-microprobe analysis.
Analyses from three different spots
indicated that the sample was chemically homogenous. Average results
(reported in weight percent) were as
follows: 46.74 SiO2, below the detection limit for TiO2, 18.39 Al2O3, 0.01
FeO, 0.80 MnO, 6.73 MgO, 21.53
CaO, 0.03 Na2O, and 0.01 K2O. This
composition was consistent with that
of the blue magnesioaxinite from
Tanzania (see Jobbins et al., 1975; G.
B. Andreozzi et al., Crystal chemistry
of the axinite-group minerals: A
multi-analytical approach, American
Mineralogist, Vol. 85, 2000, pp.
698706). Vanadium was present as a
minor constituent (0.13 wt.% V2O3) in
the blue magnesioaxinite, and EDXRF
analysis detected trace amounts of
vanadium in the yellowish orange
sample; unfortunately, quantitative
microprobe data for vanadium were
not collected.
In addition to the blue samples
from Tanzania discussed above,
brown-to-pink magnesioaxinite has
been reported from Luning, Nevada.
The material from Nevada has considerably less magnesium and more
iron and manganese than the
Tanzanian material (P. J. Dunn et al.,
Magnesioaxinite from Luning,
Nevada, and some nomenclature designations for the axinite group,
Mineralogical Record, Vol. 11, 1980,
pp. 1315; Andreozzi et al., 2000).
The Nevada samples contained just
enough magnesium to fall within the
broad magnesioaxinite range of the
series, whereas the composition of
the Tanzanian material was much
closer to that of the pure magnesioaxinite end member. The difference in
major-element chemistry between
these two sources most likely

explains their differences in color.


The brownish yellowish orange
magnesioaxinite described here was
similar in properties and chemistry to
the blue Tanzanian material, which
suggests that the yellowish orange
color is probably not due simply to
variations in iron, magnesium, and
manganese. It is possible that variations in trace elements such as vanadium may contribute to the color, but
as yet we do not know its true cause.
Elizabeth P. Quinn and
Christopher M. Breeding

CULTURED PEARL
With Cultured-Pearl Nucleus
It is always exciting to discover the
unexpected on a pearl radiograph.
When the pendant in figure 21 arrived
at the West Coast laboratory for a routine identification, we anticipated
nothing unusual. However, the Xradiographs used in pearl analysis
sometimes reveal surprises, and this
proved to be one such specimen.
The baroque pearl measured 16.00
x 13.50 mm in width and depth. (We
could not measure the length because

Figure 21. This 16.00 x 13.50 mm


baroque cultured pearl contained
an unusual surprise.

GEMS & GEMOLOGY

SUMMER 2005

171

Figure 23. The ~18 ct cultured


pearl in this X-radiograph appears
to have a separate tissue-nucleated cultured pearl at its core.

Figure 22. The X-radiograph of


the cultured pearl in figure 21
revealed another cultured pearl
as its nucleus, with bead and
nacre clearly defined.

part of the pearl was covered by the


mounting.) The X-ray luminescence
was very weak, indicating that it
might be a cultured pearl with thick
nacre. However, X-radiography
revealed that the nucleus did not
merely consist of the expected mother-of-pearl bead, but it was in fact an
entire cultured pearl with its own
bead nucleus and nacre clearly
defined on the radiograph (figure 22).
We have seen other cultured
pearls with cultured-pearl nuclei; for
example, we recently examined a
semibaroque cultured pearl (approximately 18 ct) that appeared to contain a separate tissue-nucleated cultured pearl in its core (figure 23).
The use of poor-quality cultured
pearls as bead nuclei is a practice
that has been known for a number
of years, though it is unclear how
common this practice is and we see
very few such cultured-pearl nuclei
in the laboratory. Nevertheless,
gemologists relish having their routines broken by these small but
delightful discoveries.
CYW

172

LAB NOTES

Dyed Golden Freshwater


Cultured Pearls
A graduated strand of what appeared to
be golden South Sea cultured pearls
arrived at the West Coast laboratory
for identification (figure 24). We routinely examine these types of cultured
pearls to determine whether or not
they are dyed. This strand, however,
provided another unexpected result.
Magnification revealed that the
string knots between the cultured
pearls were yellow, while the rest of

the string was white. There also were


concentrations of darker yellow
around some of the drill holes (figure
25). These preliminary observations
suggested that the color was artificial
in origin. The reaction to long-wave
UV radiation also was characteristic
of treated-color pearls: Some fluoresced medium orangy yellow (occasionally mottled with blue), while
others were medium pink with
uneven orange mottling (figure 26).
Naturally and uniformly colored yellow pearls evenly fluoresce yellow to
greenish yellow or greenish brown to
brown to long-wave UV; pink and
orange components, along with blue
patches and other mottling, are
inconsistent with an evenly distributed natural color.
UV-Vis reflectance spectra were
collected by research gemologist
Shane Elen. The spectra consistently
revealed a strong absorption trough in
the blue region at approximately
415440 nm rather than a deeper
trough in the UV region at approximately 330385 nm. These results
are consistent with dyed yellow saltwater cultured pearls (see S. Elen,
Spectral reflectance and fluorescence
characteristics of natural-color and
heat-treated golden South Sea cultured pearls, Summer 2001 Gems &

Figure 24. This strand of 34 dyed freshwater cultured pearls (approximately 10.6011.70 mm) at first appeared to be typical golden South Sea cultured pearls.

GEMS & GEMOLOGY

SUMMER 2005

Figure 25. Color concentrations


around the drill holes of these
freshwater cultured pearls, along
with the yellow coloring on the
knots of the string, indicate the
presence of dye.
Gemology, pp. 114123).
Treated-color yellow South Sea
cultured pearls have become commonplace. The surprise with this
necklace came while examining the
X-radiographs. Rather than the
expected bead nuclei, the radiographs
revealed the characteristic structure
of tissue-nucleated cultured pearls,
immediately raising suspicion of
freshwater origin.
EDXRF analysis (also performed
by Shane Elen) confirmed a level of
manganese that was consistent with
a Texas freshwater pearl in the reference file. Mn can be found as a trace
element in some saltwater pearls, but
the amount present in these cultured
pearls excluded the possibility of saltwater origin, thus proving they were
freshwater cultured pearls that had
been dyed to resemble golden
South Sea cultured pearls. It is interesting to note that while the varied

Figure 26. The mottled pink and


orange long-wave UV fluorescence of these cultured pearls is
consistent with treated color.
intensity (weak to strong) of the Xray luminescence had supported the
possibility of bead nucleation, it also
was consistent with colored freshwater tissue-nucleated cultured pearls.
Although we do not have comparative UV-Vis spectra for dyed or natural-color yellow freshwater cultured
pearls, the concentrations of color
around the drill holes and unusual
UV fluorescence behavior proved that
these samples were dyed.
CYW

More on Copper-Bearing
Color-Change TOURMALINE
from Mozambique
The Fall 2004 Lab Notes section (pp.
250251) reported on three unusual
samples of copper-bearing tourmaline that displayed a strong and distinct reverse color change from
grayish purple or purple in fluorescent light to graybluish green or

gray in incandescent light (figure 27).


We have now obtained quantitative
chemical data on all three of these
samples; after the table of the pearshaped preform was polished (with
the clients permission), the stones
were analyzed by electron microprobe by Dr. William B. (Skip)
Simmons and Alexander U. Falster
at the University of New Orleans in
Louisiana.
The analyses (table 1) show that
the tourmalines are elbaite, with
slight variations in composition. Their
major-element compositions are comparable to those reported for cuprian
elbaite from the Paraba and Rio
Grande do Norte states in Brazil (see,
e.g., J. E. Shigley et al., An update on
Paraba tourmaline from Brazil,
Winter 2001 Gems & Gemology, pp.
260276). However, their Cu contents
were significantly lower than those
reported for Paraba tourmaline:
They ranged from below the detection
limit (< 0.008) to 0.09 wt.% CuO.
These values also are significantly
lower than those reported for most of
the cuprian tourmalines from Nigeria
that have been analyzed (see Gem
News International, Fall 2001, pp.
239240, and Spring 2002, pp. 99100).
Only minute amounts of Fe and
Ti were detected in some analyses of
each stone. Although EDXRF analysis had detected Ga and Bi (again, see
the Fall 2004 Lab Note), these elements were below the detection limits of the electron microprobe. The
traces of Pb measured in some analyses are consistent with those found
in cuprian elbaite from other localities, and the client confirmed that no

Figure 27. These three color-change tourmalines, reported to be from Mozambique (viewed in fluorescent
light at left and incandescent light at right), contain Mn and Cu as their primary chromophores. They weigh
5.37, 5.47, and 5.68 ct.

LAB NOTES

GEMS & GEMOLOGY

SUMMER 2005

173

TABLE 1. Electron-microprobe analyses of three samples of color-change


elbaite tourmaline from Mozambique.a
Chemical
compostion

5.37 ct grayish
purple oval

5.68 ct purple
emerald cut

5.47 ct purple
pear preform

Oxides (wt.%)
SiO2
TiO2
B2O3b
Al2O3
V2O3
FeO
MnO
CuO
SrO
CaO
PbO
ZnO
Li2Ob
Na2O
K2O
H2Ob
F
Sum
-O=F
Total

36.6136.77
0.010.04
10.9811.00
41.5341.76
bdl0.02
0.020.05
1.942.14
bdl0.09
0.020.08
0.450.51
bdl0.02
bdl
1.771.84
1.952.14
bdl
3.353.39
0.840.95
99.98100.31
0.350.40
99.6099.91

36.6136.86
bdl0.01
11.0111.06
42.5142.77
0.080.15
bdl0.01
0.070.11
bdl0.03
0.020.04
0.720.89
bdl0.04
bdl0.03
1.942.06
1.481.78
bdl0.04
3.333.37
0.940.99
99.1399.78
0.400.42
98.7399.36

36.6936.87
bdl0.01
11.0311.12
42.5342.79
0.080.11
bdl0.01
0.110.15
bdl0.03
0.020.03
0.600.68
bdl
bdl0.04
1.952.16
1.671.80
bdl
3.323.36
0.971.03
99.3299.69
0.410.43
98.8999.25

Ions on the basis of 31 (O,OH,F)


Si
5.7965.817
Al
0.1830.204
T sum
6.000
B
3.000
Al (Z)
6.000
Al
1.5451.575
bdl0.002
V3+
Ti
0.0020.004
Fe2+
0.0030.006
Mn
0.2590.287
Cu
bdl0.011
Sr
0.0010.007
Pb
bdl0.001
Zn
bdl
Li
1.1261.173
Y sum
3.000
Ca
0.0760.086
Na
0.5980.658
K
bdl
Vacancy
0.2560.326
X sum
1.000
F
0.4220.475
OH
3.5253.578

5.7695.789
0.2110.231
6.000
3.000
6.000
1.6611.741
0.0110.018
bdl0.001
bdl0.001
0.0090.014
bdl0.004
0.0020.003
bdl0.002
bdl0.004
1.2331.302
3.000
0.1210.150
0.4520.543
bdl0.008
0.3090.426
1.000
0.4680.491
3.5093.532

5.7645.791
0.2090.236
6.000
3.000
6.000
1.6011.729
0.0090.014
bdl0.001
bdl0.001
0.0140.019
bdl0.003
0.0010.003
bdl
bdl0.004
1.2391.361
3.000
0.1010.115
0.5090.548
bdl
0.3400.391
1.000
0.4800.511
3.4893.520

Analyzed with an ARL SEMQ electron microprobe, using an accelerating voltage of 15 kV, a beam current of 15 nA, a spot size of 2 m, and 100-second count times for each spot. Ranges are shown for five
points analyzed on each sample. Mg, Cr, Sc, Ga, Bi, and Cl were analyzed but were at or below the detection limit (except for 0.01 wt.% MgO in one analysis). Abbreviation: bdl = below detection limit.
b Values for B O , Li O, and H O were calculated based on an assumed elbaite tourmaline stoichiometry.
2 3
2
2

174

LAB NOTES

lead-containing compounds were


used in the cutting and polishing process. Thus, the speculation in the previous entry about Pb contamination
from the polishing compound residue
was probably erroneous.
The transition metals Mn, Fe, Ti,
and Cu are the primary color-causing
agents in tourmaline, in addition to V
and Cr (W. B. Simmons et al., Gem
tourmaline chemistry and paragenesis, Australian Gemmologist, Vol. 21,
2001, pp. 2429). The oval stone contained much more Mn than the other
two samples. Its Mn content was comparable to that found in Paraba
tourmaline and similar to or slightly
higher than that reported in cuprian
elbaite from Nigeria (again, see Gem
News International, Fall 2001, pp.
239240, and Spring 2002, pp. 99100).
The emerald cut and pear-shaped preform had appreciably less Mn than has
been reported in Paraba tourmaline
and in three violet-blue cuprian
elbaites from Nigeria, but it fell within
the lower range reported for four greenish blue Nigerian cuprian elbaites.
By correlating the chemical composition with the UV-Vis-NIR spectra
of all three stones (figure 28), the particular chromophores and the colorchange phenomenon can be better
understood. Each stone had a broad
absorption peak centered around 515
nm (attributed to Mn3+). In addition,
Cu2+ peaks were centered at approximately 690 and 900 nm (E | c), or 720
and 910 nm (E||c). The latest data confirm our preliminary conclusion that
manganesespecifically, Mn3+ residing in the Y crystallographic siteis
one of the primary coloring agents of
this material. Although the Cu concentrations are significantly lower
than those of cuprian elbaites from
Brazil and Nigeria, the presence of copper is also contributing to the color;
however, the absorption caused by
Mn3+ in these specimens most likely
explains their less saturated and more
purplish color in fluorescent light,
along with the presence of some pinkish zones, as compared to the more
intensely colored copper-induced blue
hues of Paraba tourmaline.

GEMS & GEMOLOGY

SUMMER 2005

Figure 28. These polarized UV-Vis-NIR spectra for the three Mozambique
tourmalines (E| c, top; E||c, bottom) show a broad absorption peak centered
around 515 nm (attributed to Mn3+), with transmission windows on either
side centered at approximately 420450 nm and 570580 nm. These separate transmission maxima could account for the color-change phenomenon
observed in these stones. The peaks from approximately 1400 to 1500 nm are
due to OH groups. The inferred optical path lengths are as follows: emerald
cut ~7 mm, pear-shaped preform ~8 mm, and oval ~68 mm. Spectra have
been shifted vertically for clarity.

Although the concentrations of V


were nearly equal to those of Mn in
both the emerald cut and pear-shaped
preform, and they exceeded the contents of Cu in the same stones, the
absorption spectra do not show any
indication of the V3+ peaks (599610
nm and 418440 nm) that have been
reported for tourmaline (K. Schmetzer,

LAB NOTES

The color-change phenomenon


may be explained by the transmission
windows on either side of the Mn3+
absorption peak. These separate transmission maxima are centered at
approximately 420450 nm (the violet
to purple spectral region) and
570580 nm (the greenish yellow to
yellow-orange region; again, see figure
28). However, often it is not possible
to explain the observed color of a sample on the basis of the absorption spectra alone. Although the colors of our
samples correspond to the approximate regions of maximum transmission in the absorption spectra, they do
not correlate exactly; additional factors such as variables in the composition of the stones, lighting and viewing conditions, the manner in which
color perception varies between individuals, and other considerations combine to determine the actual hues
observed by each viewer (Y. Liu et al.,
A colorimetric study of the alexandrite effect in gemstones, Journal of
Gemmology, 1999, Vol. 26, No. 6, pp.
371385). Thus, the exact cause of the
reverse nature of the color change (relative to the alexandrite effect)
remains unknown. We are confident,
however, that the unusual phenomenon could be further explained
with a detailed colorimetric study in
conjunction with the absorption
spectra.
Although these Mozambique
tourmalines were found to have
much lower copper contents than
their Brazilian and Nigerian counterparts, they do expand the geographic
locales where copper-bearing elbaite
has been found.
CYW, Eric Fritz, and
Sam Muhlmeister

Absorption spectroscopy and colour


of V 3+-bearing natural oxides and
silicatesA contribution to the crystal
chemistry of vanadium, Neues Jahrbuch
fr Mineralogie, Abhandlungen, Vol.
144, No. 1, 1982, pp. 73 106 [in
German]); therefore, there is no evidence that V3+ is contributing to the
color of these stones.

GEMS & GEMOLOGY

PHOTO CREDITS
Thomas Gelb15, 17, 18, and 19; Jessica
Arditi6; Wuyi Wang7 and 9; Maha
Calderon11 (left), 13, 20, 21, and 26;
Christopher M. Breeding11 (right) and 16;
Cheryl Y. Wentzell22 and 23; C. D.
Mengason24, 25, and 27.

SUMMER 2005

175

EDITOR
Brendan M. Laurs (blaurs@gia.edu)

CONTRIBUTING EDITORS
Emmanuel Fritsch, IMN, University of
Nantes, France (fritsch@cnrs-imn.fr)
Henry A. Hnni, SSEF, Basel, Switzerland
(gemlab@ssef.ch)
Kenneth V. G. Scarratt, GIA, Bangkok, Thailand
(kscarratt@aol.com)
James E. Shigley, GIA Research,
Carlsbad, California (jshigley@gia.edu)
Christopher P. Smith, GIA Gem Laboratory,
New York (chris.smith@gia.edu)

COLORED STONES AND


ORGANIC MATERIALS
Demantoid from northern Pakistan. At the February
2004 Tucson gem shows, one of us (BML) was shown
some attractive demantoid that was reportedly from
northern Pakistan. The dealer who had the material,
Syed Iftikhar Hussain (Syed Trading Co., Peshawar,
Pakistan), indicated that it came from the Kaghan Valley
area (Hazara district), which is the same region that hosts
Pakistans famous peridot deposits. A visit to the mining
area by Mr. Hussain revealed shallow open cuts within
ultramafic rocks. The garnets occur as well-formed crystals in a soft white host rock that appears to consist
mostly of talc. A small amount of rough was produced in
2003, yielding a few dozen carats of faceted stones up to

Figure 1. A small demantoid deposit, reportedly in


northern Pakistan, is the source of these round brilliants (0.300.60 ct) and matrix specimen. Note the
strong color zoning in the two larger cut stones. GIA
Collection nos. 3193031932; photo by Maha Calderon.

3 ct (most ranged from melee to 1 ct). Since that time,


there has been little additional production.
One matrix specimen and three faceted demantoids
(0.300.60 ct; figure 1) were donated to GIA by Mr.
Hussain, and the cut stones were characterized by one of
us (EPQ). Their color ranged from yellow-green to green,
with two of the stones having fairly strong color zoning
in green and lighter brownish greenish yellow. The faceup combination of these colors within the faceted stones
created an almost pleochroic appearance. The three samples showed the following additional characteristics:
diaphaneitytransparent; R.I.over the limits of a standard refractometer; S.G.3.853.95; anomalous birefringencemoderate to strong; fluorescenceinert to both
long- and short-wave UV radiation; Chelsea filter reactionweak pink; and a strong broad absorption band at
440 nm and lines in the red end of the spectrum were
observed with the desk-model spectroscope. Microscopic
examination revealed fine wavy and/or curved needles in
all three stones, as well as a few fractures. In two samples, these needles occurred in clusters, creating partial
horsetail inclusions. All three stones also contained

Editors note: The initials at the end of each item identify the
editor or contributing editor who provided it. Full names and
affiliations are given for other contributors. Shane F. McClure
and Dr. Mary L. Johnson of the GIA Gem Laboratory in
Carlsbad are thanked for their internal review of the Gem
News International section.
Interested contributors should send information and
illustrations to Brendan Laurs at blaurs@gia.edu (e-mail),
760-603-4595 (fax), or GIA, 5345 Armada Drive, Carlsbad,
CA 92008. Original photos will be returned after consideration or publication.
GEMS & GEMOLOGY, Vol. 41, No. 2, pp. 176187
2005 Gemological Institute of America

176

GEM NEWS INTERNATIONAL

GEMS & GEMOLOGY

SUMMER 2005

angular growth lines, and in two samples these lines


formed an hourglass-like outline.
These properties are similar to those reported for
demantoid from Iran (see Spring 2002 Gem News International, p. 96). However, while the Iranian material typically appears too dark in stones weighing more than 0.70
ct, this was not the case for the 13 ct samples from
Pakistan that Mr. Hussain showed us.
Elizabeth P. Quinn (equinn@gia.edu)
GIA Gem Laboratory, Carlsbad

Figure 2. These samples (1.35, 1.31, and 0.89 ct) were


selected to show the range of color of pyrope-almandine
being produced from southern Ethiopia. GIA Collection
nos. 3175531757; photo by Maha Calderon.

BML
Pyrope-almandine from Ethiopia. According to Scott
Davies of American-Thai Trading, Bangkok, small quantities of purple to orange-red garnet from Ethiopia have been
entering the gem trade in Thailand since June 2004. The
slightly rounded rough pieces range up to 12 grams or larger. While most of the material is dark and contains abundant inclusions, a small percentage shows attractive color
and good clarity. According to one supplier, Teferi Gobezie
of Addis Ababa, Ethiopia, the garnet is mined at Baya
Gundi, in the Hagare Mariam region of southern Ethiopia
about 200 km from Kenya. Garnets are known from several localities in southern Ethiopia (see Gem resources in
Ethiopia, Jewellery News Asia, No. 209, 2002, p. 52)
Mr. Davies kindly loaned and donated several rough
and cut samples of this garnet to GIA. Examination of
three faceted stones (1.35, 1.31, and 0.89 ct; figure 2) by one
of us (EPQ) showed the following properties (listed in
respective order, as appropriate): colorred, dark orangered, and purplish pink; R.I.1.741, 1.744, and 1.757; S.G.
3.85, 3.84, and 4.00; and fluorescenceall three samples
were inert to long- and short-wave UV radiation. With the
desk-model spectroscope, all three samples showed weak
absorption lines at 430, 460, and 620 nm, and stronger
bands at 505, 520, and 570 nm. Microscopic examination
revealed clusters of small, rounded, transparent, doubly
refractive crystals in the red stone (which senior research
associate Sam Muhlmeister identified as zircon by Raman
analysis); small crystals, needles, and clouds of pinpoints in
the dark orange-red sample; and pinpoint inclusions as well
as large, transparent, flattened, doubly refractive crystals in
the purplish pink stone (which could not be identified due
to their position within the sample). The properties of these
garnets are consistent with those reported for pyropealmandine (see C. M. Stockton and D. V. Manson, A proposed new classification for gem-quality garnets, Winter
1985 Gems & Gemology, pp. 205218).
Elizabeth P. Quinn and BML

able, along with both rough and cut samples of this new
corundum that reportedly have not been treated in any way.
The deposit is located in the Kenya Rift Valley in the
west-central part of the country, approximately 70 km east
of Eldoret near Lake Baringo. In October 2001, a local gem
prospector was shown samples of pink and red corundum
from this area by members of the Tugen tribe who lived in
Barsemoi. The prospector later visited the locality and collected samples of gem-quality corundum from a dry river
bed. He applied for a license with the Kenya Commissioner
of Mines, and after a two-year process received a mining
permit for a 5 km2 area. In June 2004, he resumed collecting surface material from the site and has obtained more
than 2 kg of corundum, ranging from red to pink to pinkish
violet. So far, approximately 350 stones have been faceted
at his office in Eldoret, and the largest weigh 2.80 and 3.71
ct (both oval cuts). Most of the stones range from 0.5 to 1.6
ct, and are bright purplish pink to purplish red (figure 3).
As of May 2005, approximately 400 local villagers were
active in the region surrounding the original mining claim,
expanding the corundum-bearing area to approximately 20
km2. Preliminary indications suggest that this area has the
potential to become a significant source of untreated ruby

Figure 3. A new corundum deposit in west-central


Kenya is the source of these rubies and pink sapphires
(0.341.64 ct). Courtesy of Dudley Blauwet Gems;
photo by Maha Calderon.

New ruby and pink sapphire deposit in the Lake Baringo


area, Kenya. At the February 2005 Tucson gem shows, one of
us (DB) was shown about 180 grams of rough ruby and pink
sapphire that were reportedly mined from a new locality in
Kenya. Subsequently more information has become avail-

GEM NEWS INTERNATIONAL

GEMS & GEMOLOGY

SUMMER 2005

177

Figure 4. Faceted sunstone from India was available


in large quantities at this years Tucson gem show.
The loose stones shown here weigh 5.138.06 ct,
and the pendant contains a 20 ct sunstone set in
18K white gold. Courtesy of Anil B. Dholakia Inc.;
photo by Maha Calderon.
and pink sapphire. The gemological properties of this
corundum are currently being studied at GIA, and will be
reported in a future issue of G&G.
Dudley Blauwet (mtnmin@attglobal.net)
Dudley Blauwet Gems, Louisville, Colorado

was notable for the huge quantity of faceted stones available. Mr. Dholakia had been accumulating the material for
the past two years, and estimated that 75,000 carats were
available, in calibrated rectangle and oval shapes from 6 x
4 to 12 x 10 mm. Some of the stones were set in bracelets
and pendants (see, e.g., figure 4).
Mr. Dholakia loaned several samples to GIA for examination, and gemological properties were collected on three
emerald cuts (5.138.06 ct; again see figure 4) by one of us
(EPQ): colorbrownish orange and brown (color created by
inclusions); diaphaneitysemitransparent to translucent
(due to numerous inclusions); R.I.1.530 to 1.541; birefringence0.009 or 0.010; S.G.2.632.64; Chelsea filter reactionweak orangy pink; and fluorescenceinert to longwave and weak red to short-wave UV radiation. Weak
absorption in the blue and green areas of the spectrum was
visible with the desk-model spectroscope. These properties
are comparable to those reported for sunstone (aventurescent oligoclase feldspar) by R. Webster (Gems, 5th ed.,
revised by P. Read, Butterworth-Heinemann, Oxford,
England, 1994, pp. 215216). Microscopic examination of
the three samples revealed numerous angular transparentto-semitransparent thin platelets that were reddish orange
in transmitted light and iridescent in reflected light. These
inclusions had the distinctive appearance of hematite, and
this was confirmed with Raman analysis by senior research
associate Sam Muhlmeister.
Elizabeth P. Quinn and BML

BML
Faceted orange and brown sunstone from India. At the
2005 Tucson gem shows, Anil Dholakia of Anil B.
Dholakia Inc., Franklin, North Carolina, had a large
amount of faceted sunstone from Orissa, India. Although
similar material has been produced from India in the past
(see Summer 1995 Gem News, pp. 130131), this sunstone

INCLUSIONS IN GEMS

Figure 5. Each of these three rectangular step-cut


kyanites (2.317.39 ct) from Minas Gerais, Brazil,
contains a well-formed inclusion of staurolite. Photo
by Maha Calderon for microWorld of Gems.

Figure 6. This transparent dark yellowish brown crystal of staurolite formed a conspicuous inclusion in one
of the Brazilian kyanites. Photomicrograph by John I.
Koivula for microWorld of Gems; magnified 10x.

178

GEM NEWS INTERNATIONAL

Staurolite in kyanite from Brazil. Faceted kyanite is generally considered to be a gem for collectors of the unusual. While the intense blue to greenish blue color that it
sometimes displays can be quite attractive, its perfect
cleavage and somewhat low directional hardness (57 on

GEMS & GEMOLOGY

SUMMER 2005

the Mohs scale) will always limit its use as a mainstream


gemstone. One of the appealing features of kyanite for
the gemologist, however, is the variety of interesting and
colorful mineral inclusions it sometimes hosts. Among
these are orange garnets, bright red rutile, and prisms of
green tourmaline.
At the 2005 Tucson gem shows, a small selection of
faceted light bluish green kyanites containing another
interesting mineral inclusion were offered by Luciana
Barbosa of the Gemological Center, Belo Horizonte, Minas
Gerais, Brazil. They were cut from material recovered near
Barra do Salinas in Minas Gerais.
Three rectangular step-cut kyanites (2.31, 3.44, and
7.39 ct; figure 5) were examined for this report. Two of
the samples were obtained from Ms. Barbosa, and the
third was provided by William Pinch of Pittsford, New
York. Included in each stone was a single, transparent,
dark yellowish brown crystal (figure 6), which was wellformed with chisel-point terminations and a diamondshaped cross-section. Ms. Barbosa stated that these inclusions were pseudo-orthorhombic crystals of staurolite, a
monoclinic iron-aluminum silicate. The smallest stone
also contained a partial crystal of the same mineral that
was exposed on the pavilion surface and polished flat during faceting. This made it ideal for Raman analysis, by
which we verified the identification as staurolite.
Staurolite is virtually unknown as an inclusion. It has
apparently only been reported as an unusual single inclusion in an eclogitic diamond (L. R. M. Daniels and J. J.
Gurney, A crustal mineral in a mantle diamond,
Nature, Vol. 379, No. 6561, 1996, pp. 153156) and may
also be found in metamorphic corundum. The identification of staurolite inclusions in kyanite is not completely
surprising, however, since both minerals are primarily
products of medium-grade regional metamorphism and
are known to form in association with each other.
John I. Koivula (johnkoivula@hotmail.com)
West Coast AGTA Gemological Testing Center
Carlsbad, California
Maha Calderon
Carlsbad, California

Obsidian with spessartine inclusions. Although generally


uninspiring as gems from a superficial macroscopic standpoint, transparent-to-translucent natural volcanic glasses
frequently contain interesting inclusions. For example,
very rare euhedral indialite crystals and transparent rods of
sillimanite have been found in obsidian (see J. Hyrsl and V.
Zcek, Obsidian from Chile with unusual inclusions,
Journal of Gemmology, Vol. 26, No. 5, 1999, pp. 321323).
At the 2005 Tucson gem shows, Dr. Jaroslav Hyrsl,
from Kolin, Czech Republic, had a few semitransparent
brown obsidians that were faceted from rough he had
obtained in Bolivia. One of these gems, pictured in figure
7, had been cut into a 10.25 ct octagonal step cut. In strong
incandescent light or bright direct sunlight, tiny sparkling

GEM NEWS INTERNATIONAL

Figure 7. Even without magnification, small orange


spessartine inclusions are clearly visible in this 10.25
ct Bolivian obsidian. Photo by Maha Calderon for
microWorld of Gems.

orange inclusions could be seen even without magnification (again, see figure 7). Dr. Hyrsl stated that the inclusions had been identified as spessartine by X-ray diffraction and chemical analysis.
Gemological testing confirmed that the brown host
gem was obsidian. With magnification, as seen in figure 8,
the orange inclusions showed abundant crystal faces, and
their isometric crystal form was consistent with the general appearance of garnet. Examination between crossed
polarizers confirmed they were isotropic, while their high
relief and Becke line reaction showed that they had a much
higher refractive index than the surrounding obsidian. As a
further gemological exercise, microspectroscopy of these
inclusions (performed through the optical system of a
Leica gemological microscope) showed weak but visible

Figure 8. Shown here are two of the more than a


dozen well-formed spessartine inclusions observed in
the Bolivian obsidian. Photomicrograph by John I.
Koivula for microWorld of Gems; magnified 10x.

GEMS & GEMOLOGY

SUMMER 2005

179

absorption features that would be expected from spessartine.


A literature search found only one other report of garnet inclusions in obsidian, from a volcano in northwestern
Argentina (P. J. Gauthier et al., Grenats des rhyolites de la
caldra de La Pava-Ramadas (NW Argentine) et de leurs
xnolites granitiques, Comptes Rendus de lAcadmie
des Sciences, Serie II. Sciences de la Terre et des Planetes,
Vol. 318, 1994, pp. 16291635). Obsidian blocks within
rhyolite lava contained euhedral microcrystals of almandine-spessartine that ranged from 100 to 400 m in diameter. By comparison, the garnet inclusions in the Bolivian
obsidian reported in the present entry were much larger
(i.e., up to about 1.5 mm in diameter).
John I. Koivula and Maha Calderon
Fluorite inclusions in quartz from Madagascar. One of the
more interesting recent inclusion discoveries being offered
to collectors at the 2005 Tucson gem shows were rough
and cut samples of rock crystal quartz containing euhedral
crystals of violet-to-blue fluorite. Most of this material
was being sold at the Clarion Hotel by Frederic Gautier of
Little Big Stone Co., Antananarivo, Madagascar. Although
the locality was represented at the Tucson show as
Miandrivazo, Mr. Gautier recently reported that the mine
is actually situated about 17 km northeast of the village of
Amborompotsy, Fianarantsoa Province, central Madagascar. The deposit can be reached from Amborompotsy
only by motorcycle or on foot, due to the poor condition of
the road. Large quantities of quartz crystals have been
mined from this region for several years, but material containing fluorite inclusions was first recovered in late
November 2004. The crystals were mined from a quartz
vein approximately 80 cm wide that cross-cuts quartzite
host rocks. The deposit was worked with hand tools by
local miners until the end of February 2005, when a large
cave-in buried the productive area.

Figure 9. These intergrown crystals of quartz (up to


12.5 cm long) from central Madagascar contain directionally deposited inclusions of fluorite. Photo by
Maha Calderon for microWorld of Gems.

180

GEM NEWS INTERNATIONAL

Figure 10. Three large blue fluorite inclusions make


this faceted 21.68 ct rock crystal gem from
Madagascar much more interesting. Photo by Maha
Calderon for microWorld of Gems.

Mr. Gautier estimated the total production of fluoriteincluded quartz at about 300 kg, with an approximately
equal amount of quartz that did not contain any fluorite
inclusions. Most of the fluorite-included material consists
of single crystals, some doubly terminated, ranging up to
23 cm long. Only a few crystal clusters were produced. Mr.
Gautier reported that Little Big Stone Co. acquired most of
the production, which is being sold in three categories: (1)
mineral specimens and rough for cutting, (2) polished
points, and (3) polished materialconsisting of emerald
cuts (ranging from 2 to 30 ct), square slabs (from 1 cm to a
few centimeters across), and cabochons. He believed the
largest stone cut to date weighed approximately 100 ct.
One of us (JIK) has examined several hundred of these
quartz crystals (see, e.g., figure 9). In all samples, the fluorite inclusions were confined to the near-surface outer

Figure 11. Displaying sharp octahedral faces and an


intense violet color, this 2.85 mm fluorite crystal was
trapped during the final stage of growth near the surface
of a quartz crystal from Madagascar. Photomicrograph
by John I. Koivula for microWorld of Gems.

GEMS & GEMOLOGY

SUMMER 2005

zones of their rock crystal hosts, which indicates late-stage


crystallization of the fluorite in relation to the growth
cycle of the quartz. The inclusions were situated only
under certain crystal faces and on only one side of the
quartz (again, see figure 9), suggesting a directional deposition of the fluorite.
The fluorite inclusions are particularly well displayed
in faceted stones such as the 21.68 ct example in figure 10,
which was cut by Leon M. Agee of Agee Lapidary, Deer
Park, Washington. The fluorite inclusions ranged from
blue to violet in light to dark tones, and many were strongly color zoned (figure 11). Some of the more prominent
individual fluorite inclusions measured up to 7 mm in
diameter, with clusters being somewhat larger. In a small
percentage of the crystals, the fluorite formed together
with clouds of green particles that had the appearance of
chlorite. The fluorite inclusions were identified as such by
means of Raman analysis. The refractive index (1.43) and
isotropic nature of one partial fluorite inclusion extracted
from the surface of a broken crystal was determined
through Becke line testing using Cargille liquids.
A microscopic examination of several quartz crystals
showed that some of the fluorite inclusions had been
exposed to the surface through fine surface-reaching
cracks. These inclusions were often either partially or
completely dissolved away. The former case resulted in a
loose crystallite of fluorite that was trapped, but free to
move about in a fluorite-shaped cavity. In the latter case,
the resulting cavity appeared to be colorless or, in some
instances, was partially filled with light yellowish brown
epigenetic debris.
The potential for large faceted stones, together with
the relatively large size and the attractive violet-to-blue
color of the fluorite inclusions, should make this new
material from Madagascar quite popular among gemologists and others who appreciate inclusions in gemstones.
John I. Koivula, Maha Calderon, and BML
Bismuthinite inclusions in rose quartz from Madagascar.
Eye-visible inclusions are common in almost all quartz
varieties except rose quartz. Although massive rose
quartz is colored by submicroscopic inclusions (see J. S.
Goreva et al., Fibrous nanoinclusions in massive rose
quartz: The origin of rose coloration, American
Mineralogist, Vol. 86, No. 4, 2001, pp. 466472), only very
rarely does this material contain macroscopic inclusions
(see, e.g., Summer 2003 Gem News International, pp.
159162). At the 2005 Tucson gem shows, one of these
contributors (FD) had some unusual rose quartz from
Madagascar that displayed abundant, conspicuous inclusions with a metallic luster (figure 12).
The material was recovered from the Itongafeno pegmatite (also known as Tsaramanga), located 23 km west of
Antsirabe near the village of Mahaiza. This deposit is part
of the Analalava pegmatite district and has been worked
for asteriated rose quartz and dark blue aquamarine for

GEM NEWS INTERNATIONAL

Figure 12. These faceted samples of rose quartz (5.74


and 12.55 ct) from the Itongafeno pegmatite in central Madagascar contain conspicuous inclusions of
bismuthinite. Photo by J. Hyrsl.

almost 100 years. In addition to common pegmatite minerals, it has produced rare examples of emerald and native
bismuth. Several vertically disposed zones of mineralization have been identified in the pegmatite, including rose
quartz, smoky quartz adjacent to a thin Nb-Ta oxide and
biotite layer, and rose quartz with metallic inclusions.
About 60 kg of the included rose quartz was extracted
between September 2003 and June 2004, although only a
small percentage was suitable for cutting.
The inclusions in 10 faceted stones were examined,
and with magnification they appeared metallic gray to
black and had peculiar shapes (figure 13). They occurred in
parallel rows of strongly elongated, sometimes striated
crystals about 0.5 mm thick. Powder X-ray diffraction
(XRD) analysis of an extracted crystal and chemical analyses of several surface-reaching inclusions in one sample
with a Jeol 5800LV scanning electron microscope

Figure 13. The bismuthinite inclusions in the rose


quartz from Madagascar appeared metallic gray to
black, and showed a variety of irregular forms. A network of needle-like inclusions, possibly rutile, was
responsible for the asterism in this rose quartz.
Photomicrograph by J. Hyrsl; the field of view is
approximately 3 mm.

GEMS & GEMOLOGY

SUMMER 2005

181

equipped with a Princeton Gamma-Tech energy-dispersive


IMIX detector identified the inclusions as bismuthinite
(Bi2S3). These were sometimes accompanied by smaller,
more equant, yellow-to-black chalcopyrite grains. XRD
also suggested the possible presence of a Cu-Bi-sulfosalt
close to gladite in composition, which probably was present as thin metallic needles seen in some samples.
In addition to the inclusions described above, the rose
quartz also contained abundant tiny colorless needles
(possibly rutile; again, see figure 13) oriented in three
directions at angles of 120. These needles are apparently
the cause of asterism in samples that are cut as spheres
or cabochons.
To our knowledge, this is the first confirmed report of
bismuthinite in faceted quartz from a pegmatite.
Previously, bismuthinite was identified in quartz from tintungsten deposits and from Alpine-type fissures (see J.
Hyrsl and G. Niedermayr, Magic World: Inclusions in
Quartz, Bode Verlag, Haltern, Germany, 2003, p. 48).
Jaroslav Hyrsl (hyrsl@kuryr.cz)
Kolin, Czech Republic
Fabrice Danet
Style Gems, Antsirabe, Madagascar
EF

SYNTHETICS AND SIMULANTS


Yellow hydrothermal synthetic sapphires seen in India.
Since about mid-2004, the Gem Testing Laboratory in
Jaipur, India, has encountered approximately 20 synthetic yellow sapphires (3.505.30 ct) grown by a hydrothermal technique (see, e.g., figure 14). Previously, the only
yellow synthetic sapphires we had encountered were
grown by a melt technique (e.g., flame-fusion). The physical properties of these hydrothermal samples were comparable to both natural and hydrothermal synthetic yellow sapphires (see, e.g., V. G. Thomas et al., Tairus
hydrothermal synthetic sapphires doped with nickel and
chromium, Fall 1997 Gems & Gemology, pp. 188202).
Refractive indices were 1.7601.770, with a birefringence
of 0.0070.008. Specific gravity varied between 3.98 and
4.01. All the specimens fluoresced weak orange to both
long- and short-wave UV radiation, with a stronger reaction to short-wave.
With magnification, most of these synthetics exhibited some scattered flake-like inclusions (similar to bread
crumbs seen in synthetic quartz). In approximately
eight samples, these flake-like inclusions were concentrated in a single plane along one side of the seed plate
(figure 15). Inclusions such as these have not been seen in
natural sapphires (Thomas et al., 1997). A weak
chevron growth pattern that is characteristic of synthetic hydrothermal growth (see, e.g., Thomas, et al.,
1997; K. Schmetzer and A. Peretti, Some diagnostic features of Russian hydrothermal synthetic rubies and sap-

182

GEM NEWS INTERNATIONAL

Figure 14. The slightly brownish yellow body color of


this 4.30 ct hydrothermal synthetic sapphire is characteristic of those encountered at the Gem Testing
Laboratory in Jaipur, India. Photo by G. Choudhary.

phires, Spring 1999 Gems & Gemology, pp. 1728) was


visible in only about five of the specimens examined.
Infrared spectroscopy of all the synthetic sapphires
showed absorption peaks in the region between
3600 3300 cm 1, at 3563, 3483, 3383, and 3303 cm1.
These peaks are related to OH- groups, and have been documented in both natural and synthetic corundum
(Thomas et al., 1997). However, the samples were readily
identified as synthetic by their characteristic bread
crumb inclusions (present as scattered flakes and/or concentrated in a flat plane), as well as their occasional
chevron growth patterns.
Gagan Choudhary (gtljpr_jp1@sancharnet.in)
Gem Testing Laboratory, Jaipur, India

Figure 15. Some of the yellow hydrothermal synthetic


sapphires showed bread crumb inclusions that
were concentrated in a plane along the seed plate.
These flake-like inclusions have not been seen in natural sapphires. Photomicrograph by G. Choudhary;
magnified 30x.

GEMS & GEMOLOGY

SUMMER 2005

TREATMENTS
Interesting heated citrine. It is relatively well known that
the vast majority of citrine quartz is the product of heat
treating amethyst. Material from Brazil and Uruguay has
often been used for this purpose. Both large and small
amethyst-lined geodes are converted to citrine using simple low-temperature heat treatment in air. Because so
much of this heat-treated material is available, and it is
cost prohibitive to attempt to determine the cause of color
(natural or heated) for most of these stones, the trade typically assumes that all citrine is heated and this yellow
quartz is priced accordingly.
At the 2005 Tucson gem shows, well-known quartz specialists Si and Ann Frazier (El Cerrito, California) noticed
clusters of brownish orange heat-treated citrine crystals
from Uruguay that showed an unusual near-surface phenomenon. Many of the faces on these crystals had a pearly,
reflective, almost mirror-like luster that was reminiscent of
the silvery white adularescence characteristic of Sri Lankan
moonstones. According to the dealer, Marcelo Valadares of
General Brazilian Gems, Governador Valadares, Brazil, only
a small proportion of the citrines showed this unusual surface feature after they were heated to approximately 400C.
The Fraziers provided one of these contributors (JIK)
with a specimen of the heated citrine that measured 23.9 x
21.4 x 11.1 mm (figure 16). With magnification, the reflective surfaces appeared to show true adularescence (figure
17), though the effect was apparent only on the minor
rhombohedral crystal faces.

Figure 16. The interesting adularescent areas shown


by the minor rhombohedral faces on this 23.9-mmwide citrine crystal group formed during the heat
treatment of an amethyst cluster from Uruguay.
Photo by Maha Calderon for microWorld of Gems.

Figure 17. The adularescence shown by certain crystal faces of the heated citrine specimen provides an
example of an interesting treatment-induced optical
phenomenon. Photomicrograph by John I. Koivula for
microWorld of Gems; magnified 10x.

The cause of this heat-induced adularescence is


unknown. A curious rumor concerning this phenomenon
was that the pearly-looking surfaces were actually individually cut and glued onto the surfaces of the crystals, but we
found no evidence of this procedure (which, from an economic standpoint alone, seems quite unviable). The adularescence could be related to the delamination of twin
planes or of ultra-fine growth structures during heating.
Prior to this delamination, these submicroscopic planar features were perhaps similar to those that produce an unusual play-of-color in the minor rhombohedral faces of some
amethyst and rock crystal specimens (referred to as the
Lowell Effect; see Winter 1987 Gem News, p. 240).
Other theories are possible, and we welcome any observations from Gems & Gemology readers.
John I. Koivula and Maha Calderon
Chocolate Tahitian cultured pearls. At the 2005 Tucson
gem shows, Emiko Pearls International Inc. (Bellevue,
Washington) had some Tahitian cultured pearls, marketed
as chocolate pearls, that were processed to lighten their
color (figure 18). Although this attractive product has been
available for 45 years, many in the trade are still unaware
of it (see Trade raises questions about chocolate pearls,
Jewellery News Asia, No. 241, September 2004, pp. 160,
162). The colors range from a pleasing dark brown to light
yellow-brown, and have been described in the trade as
copper, bronze, or honey colored; these hues are not
typical of natural-color Tahitian cultured pearls. The luster
of the chocolate cultured pearls varies from satin-like to
metallic. The available sizes range from 9 to 15 mm,
although larger ones have been produced. Production is
limited to a few hundred per month because only certain
Tahitian cultured pearls can be successfully treated.
The coloration of these cultured pearls results from a
proprietary two-stage process. Ron Greenidge of Emiko

GEM NEWS INTERNATIONAL

GEMS & GEMOLOGY

SUMMER 2005

183

were first introduced to the market, they have not encountered any problems with color stability.
The exact nature of the process remains undisclosed,
and future studies are needed to establish identification
criteria. Preliminary research suggests that careful observation of the long-wave UV fluorescence will likely be helpful in identifying these cultured pearls, when compared to
known natural colors as a reference (see US gem labs
seek to uncover process behind brown pearls, Jewellery
News Asia, No. 241, September 2004, p. 162).
Marisa Zachovay (marisa.zachovay@gia.edu)
GIA Education, Carlsbad

Pearls International kindly provided the following general


information from the company that performs the procedure. The first step is to remove color from the selected
Tahitian cultured pearls, which could be compared to the
bleaching process used on Akoya cultured pearls. In the
second step, the bleached color is stabilized. The company
reported that the process does not involve the use of dye or
heat, and the brown color is stable under conditions associated with routine wear. Since Emiko Pearls International
began selling the chocolate cultured pearls when they

Lead glassfilled rubies appear in the Middle East. The


Middle Eastespecially the countries along the Arabian
Gulfhas long been an important gold market, but recently the region has gained importance in trading gems and
jewelry. This can clearly be observed in Dubai, where the
Dubai Metal and Commodities Centre has been established, incorporating the Dubai Diamond Exchange and a
future GIA campus. Moreover, since the Gem & Pearl
Testing Laboratory of Bahrain issued its first report in
1991, the region has also seen an increase in the number of
gemological laboratories.
When a new treatment such as the lead-glass filling of
rubies appears, it usually takes some time to be recognized
in the smaller markets around the world. Most new treatments are initially reported in the U.S., the Far East, and
Europe. In recent months, these major gem markets have
encountered lead glassfilled rubies (see Fall 2004 Lab
Notes, pp. 247249, and references therein). Now such
rubies are being seen in the Middle East as well.

Figure 19. This 10.35 ct faceted ruby with lead


glassfilled fractures was submitted to the Gem & Pearl
Testing Laboratory of Bahrain. Photo by N. Sturman.

Figure 20. With magnification, the presence of a leadglass filler was evident from the vivid flash-effect colors and gas bubbles seen in some of the fractures.
Photomicrograph by N. Sturman; magnified 20x.

Figure 18. The coloration of the chocolate


Tahitian cultured pearls in this strand (13.616.7
mm in diameter) is the result of a proprietary twostage process. Courtesy of Emiko Pearls
International Inc.; photo by Maha Calderon.

184

GEM NEWS INTERNATIONAL

GEMS & GEMOLOGY

SUMMER 2005

In February 2005, the 10.35 ct orangy red natural ruby


shown in figure 19 was submitted via a Saudi
Arabiabased client to the Gem & Pearl Testing Laboratory of Bahrain, and two more lead glassfilled rubies have
subsequently appeared from a separate source. The 10.35
ct stone exhibited all the characteristic features associated
with this treatment: loupe-visible flash effects and flattened gas bubbles within surface-reaching fractures (figure
20); differences in surface luster between the corundum
and the glass filling in reflected light; and a filling material
that was opaque to X-rays (again, see the Lab Note referenced above).
The appearance of these lead glassfilled rubies in the
Middle East confirms that this treatment is reaching gem
markets worldwide.
Nick Sturman (metalgem@commerce.gov.bh)
Gem & Pearl Testing Laboratory
Manama, Bahrain

CONFERENCE REPORTS
GemmoBasel 2005. This International Colloquium of
Gemmology was held April 29May 2 in honor of the
60th birthday of renowned Swiss gemologist (and GNI
contributing editor) Dr. Henry A. Hnni. Approximately
170 participants from 29 countries gathered at the
University of Basel to hear two days of presentations. Dr.
W. Stern of the University of Basel opened the session
with an entertaining biography of Dr. Hnni, noting that
this celebration coincided with the anniversary in August
of Dr. Hnnis 25 years with the SSEF Swiss Gemmological Institute.
The first technical session, on diamonds, began with a
presentation by Dr. Paul Spear of the Diamond Trading
Company (DTC) Research Centre in Maidenhead, U.K.
Dr. Spear noted that the DTC Research Centre has developed methodology to rapidly screen for and identify both
HPHT- and CVD-grown synthetic diamonds, down to 0.03
ct, as well as synthetic moissanite, using the DiamondSure and DiamondView verification instruments. Dr.
Emmanuel Fritsch of the University of Nantes, France,
and collaborators reported on advances in identifying and
classifying the origin of color in brown and black diamonds. Significantly, most type Ia (nitrogen containing)
diamonds with brown graining have a series of absorptions
in the near-infrared region called amber centers, the
most common of which (AC1) shows a main absorption
around 4168 cm1. The more intense the brown color is,
the more intense this AC1 absorption is, and the more A
aggregates are detected (figure 21). This and other results
indicate that AC1 may be the cause of the visible-range
absorption inducing the brown color. Jean-Pierre Chalain
of SSEF stressed the importance of a database of naturalcolor diamonds for treatment research. Microscopic observation and extensive spectroscopic analyses of an historic
(1907) 1.18 ct greenish yellow diamond crystal showed

GEM NEWS INTERNATIONAL

Figure 21. The intensity of the infrared absorption of the


most common amber center in diamond (AC1) at
about 4168 cm-1 correlates to that of the A aggregates,
as well as with brown color. This indicates that the
brown color of certain gem diamonds may be caused
by this amber center. The spectra are offset vertically
for clarity, and have all been normalized to the intrinsic absorption of diamond at about 2450 cm-1. From top
to bottom, the samples consist of a very dark brown
(nearly black) 1.22 ct briolette, a dark reddish brown
0.75 ct round brilliant, a dark reddish brown 0.45 ct
round brilliant, and a pinkish brown 0.31 ct single-cut
diamond. All spectra were taken with the beam in the
girdle plane, except for the briolette, in which the shortest dimension was chosen; the inferred optical path
lengths are 5.2, 5.2, 4.9, and 4.1 mm, respectively.
strong overlap with the properties of HPHT-treated diamonds of similar color, indicating the need for further testing of such material.
Leading the Gemstone session was John I. Koivula of
the West Coast AGTA Gemological Testing Center (AGTA
GTC) in Carlsbad, California. Reiterating his conviction
that the microscope is the gemologists most important
tool, he showed photomicrographs of distinctive inclusions,
such as pink covellite in quartz, reportedly from Bahia,
Brazil; intense blue fluorite in quartz from Madagascar; and
turquoise in selenite from New Mexico. Tay Thye Sun of
the Far East Gemmological Institute, Singapore, stressed the
importance of classifying jade correctly, not only jade vs.
nephrite and their simulants, but also the many treated
jades. Today, most of the jadeite in the market is B-jade,
which is bleached and polymer impregnated; it can be identified by distinctive features in the infrared spectrum.
Edward Boehm of Joeb Enterprises, Solana Beach,
California, presented an update on mining at Mogok,
Myanmar, for Richard Hughes, also of the West Coast
AGTA GTC. He reported that the mining methods used at
Mogok have been virtually unchanged for decades, with the
area pockmarked by tunnels and as many as 250,000 people
working the area at the height of its activity.

GEMS & GEMOLOGY

SUMMER 2005

185

Figure 22. Increasingly sophisticated gem treatments


and synthetics have made laboratory reports indispensable to auction houses. This 8.01 ct untreated
Burmese ruby recently sold for $2.2 milliona record
auction price of $275,000 per caratat Christies
New York. Courtesy of Christies Images.

Dr. Karl Schmetzer of Petershausen, Germany, reported on taaffeite and musgravite, and compared the optical
and chemical properties of typical samples of both gem
minerals with iron-rich (taaffeite and musgravite) and
zinc-rich (taaffeite) samples, all from Sri Lanka, to expand
the compositional range of this material. During the question-and-answer session, gem dealer Werner Spaltenstein
of Chanthaburi, Thailand, noted that gem-quality taaffeites have been found in Tunduru (Tanzania) as well. Dr.
Ulrich Henn of the German Gemmological Laboratory,
Idar-Oberstein, reported on feldspars, the most common
minerals in the earths crust, which provide an interesting
link between mineralogy and gemology. The main gem
feldspars are orthoclase, amazonite (microcline), moonstone (perthite or antiperthite), albite, sunstone (oligoclase
or labradorite), and andesine.
Dr. Hnni began the Instrumentation session with a
review of instruments used in gem identification at SSEF
over the past 25 years. In addition to the adoption of analytical techniques such as EDXRF, FTIR, Raman spectroscopy, andmost recentlylaser-induced breakdown
spectroscopy (LIBS), the gem laboratory has seen significant advances in lighting, especially fiber-optic illumination. He recommended that if a laboratory could not
afford to purchase needed instrumentation, they should
make arrangements to access equipment at a local university. This contributor gave participants a glimpse behind
the pages of Gems & Gemology, emphasizing the
importance of a strict peer review, editing, and production
process. She also offered several tips for publishing in
G&G: Choose a topic with a strong application to gemology; organize the paper logically, including separate sections on the materials and methods, results, and discus-

186

GEM NEWS INTERNATIONAL

sion; describe every test and say how many (and which)
samples were used for each; and do a thorough literature
review. Manfred Eickhorst of Eickhorst & Co., Hamburg,
Germany, discussed various lighting needs for aging eyes:
less glare, brighter light, and stability (i.e., nonflickering
light). A high-frequency (40,000 Hz) lamp can eliminate
flicker, whereas the use of two dimmable bulbs (with a
color temperature of 5,500 K and 6,500 K) can help
refresh the eyes as well as provide more efficient color
evaluation. Dr. Gaston Giuliani of IRD and
CRPG/CNRS, Vandoeuvre-ls-Nancy, France, reported
on the use of oxygen isotopes to study emeralds from
nearly 70 deposits. Although there is overlap among
deposits, he found that emeralds could be classed into
three groups based on their 18O values. He also found
that oxygen isotopic analysis could be used to separate
Mogok from Mong Hsu rubies.
In the Trade Issues session, Helen Molesworth of
Christies London reported that auction houses today
require gem lab reports on virtually all stones (especially
the most important ones, as shown in figure 22).
However, gemology is advancing rapidly, and clients are
reminded that reports are only an opinion based on current knowledge and technology. Werner Spaltenstein
reported on his mining and buying activities in Tunduru
and Madagascar, two areas that have great future
promise because there was so little systematic gem
exploration in the past. Also, tastes and interests
change; for example, spinel was actually discovered at
Mahenge (Tanzania) 20 years ago, but not until recently
was there a market for it.
Eric Emms of the Precious Stone Laboratory,
London, led the Treatments and Synthetics session by
describing the three-fold impact on gemologists of the
proliferation of treatments: (1) They must have sufficient funds to acquire the needed equipment and technical support staff, (2) they must have knowledge
beyond the basic FGA or GG diploma, and (3) they
must disclose to buyers/clients if treatment is detected
or may be present. Dr. Roy Walters of Ocean Optics,
Dunedin, Florida, and Dr. Michael Krzemnicki of SSEF
recently determined that surrounding a sample with
argon during LIBS analysis greatly improves the resolution of the main Be feature at 313 nm, so Be can be
detected down to 2 ppm.
Dr. V. S. Balitsky of the Institute of Experimental Mineralogy, Moscow, reported that inclusions, growth zoning,
and IR spectra may be useful to separate some synthetic
from natural quartz, but not all types can be identified. Dr.
Lore Kiefert of the East Coast AGTA GTC in New York
explained the various factors involved in determining the
geographic origin of gemstones: the inclusion scene, traceelement chemistry, and spectroscopic characteristics.
Because of overlap in some features in stones from various
sources, it is critical that multiple lines of evidence are
used in the determination.

GEMS & GEMOLOGY

SUMMER 2005

The last session, on Pearls, was led by Dora Fourcade


of Pacific Perles, Tahiti. She reported that only 5% of the
cultured pearls produced in Tahiti are superior quality; by
law, the 10% classified as rejects must be destroyed.
Typically, only 2025% of the harvest are rounds. Also in
the pearl session, Andy Mller of Hinata Trading Co.,
Kyoto, Japan, reported that production of South Sea cultured pearls has increased 160% since 1999, with
Indonesia leading by weight in 2005 and Australia the
leader by value. He emphasized the need to balance supply
and demand, and warned of the problem posed by global
warming. Water temperatures exceeding 33C are deadly
to pearl oysters. Elisabeth Strack of the Hamburg Gemmological Institute, Germany, provided a comprehensive
review of European freshwater mussel pearls. The
European pearl mussel, Margaritifera margaritifera, can
live 130 years, and is mature at 15. Ubiquitous at one
time, the mussel population dropped by 8090% in the
20th century, so only a few hundred thousand mussels
remain today. Nicholas Sturman of the Gem & Pearl
Testing Laboratory, Manama, Bahrain, described the natural pearls from the Arabian Gulf, once considered the best
in the world, but now in short supply. Bahrain law forbids
the sale of cultured pearls, so the local commerce depends
on natural pearls.
Alice S. Keller (akeller@gia.edu)
GIA, Carlsbad

ANNOUNCEMENTS
Exhibits
Cameos at the Met. Cameo Appearances, a display of
more than 160 examples of the art of gem carving from
Greco-Roman antiquity to the 19th century, will be on
display until October 30, 2005, at the Metropolitan
Museum of Art in New York City. Among the pieces featured will be works by the neoclassical Italian masters
Benedetto Pistrucci, Giuseppe Girometti, and Luigi
Saulini. A variety of educational programs will be offered in
conjunction with the exhibition. Also on display at the Met
(through January 2006) is the exhibit The Bishop Jades, a
selection of fine Chinese and Mughal Indian jades from the
collection of Heber R. Bishop that was donated to the museum in 1902. Visit www.metmuseum.org/news/news_pr.asp
or call (212) 535-7710.
Diamonds in London. Diamonds, an exhibit of some of
the worlds largest and most valuable faceted diamonds,
will be on display at the Natural History Museum in
London through February 26, 2006. The exhibit includes
the De Beers Millennium Star, the Steinmetz Pink, the
Ocean Dream, and the Moussaieff Red (see J. M. King and J.
E. Shigley, An important exhibition of seven rare gem dia-

GEM NEWS INTERNATIONAL

monds, Summer 2003 Gems & Gemology, pp. 136143),


in addition to a 616 ct rough diamond and a number of
other notable gems. Visit www.nhm.ac.uk/diamonds.

Conferences
Raman spectroscopy. The 3rd International Conference
on the Application of Raman Spectroscopy in Art and
Archaeology will be held August 31September 3, 2005,
in Paris, France. The Raman analysis of gem materials will
be among the topics covered. Visit www.ladir.cnrs.fr/
ArtRaman2005/or e-mail artraman2005@glvt-cnrs.fr.
Agate symposium. On September 1011, 2005, a symposium on agate and cryptocrystalline quartz will take place
at the Colorado School of Mines in Golden, Colorado.
Field trips to Colorado mineral localities are planned for
September 1213. These events will take place in conjunction with the 38th Annual Denver Gem and Mineral
Show (September 1618). Call 303-202-4766 or e-mail
pmodreski@usgs.gov.
Diamond 2005. A review of the latest scientific and technological aspects of natural and synthetic diamond (as well
as related materials) will take place at the 16th European
Conference on Diamond, Diamond-like Materials, Carbon
Nanotubes and Nitrides, September 1116, 2005, in
Toulouse, France. Visit www.diamond-conference.com.
Rapaport International Diamond Conference 2005. This
conference will be held in New York on September 19, 2005,
and feature an insiders look at the international diamond
and jewelry industry. Visit www.diamonds.net/conference.
Hong Kong Jewellery & Watch Fair. Held at the Hong
Kong Convention and Exhibition Centre on September
2125, this show will also host GIA GemFest Asia 2005
on September 23. The free educational event, Update on
Diamond Cut, will take place from 9 am to 12 noon in
Room 301B; RSVP to events@gia.edu, or call 760-603-4205
in the U.S. or +852-2303-0075 in Hong Kong. An educational seminar given by the Gemmological Association of
Hong Kong will also take place on September 23. Visit
www.jewellerynetasia.com/exhibitions.

ERRATUM
The Spring 2005 article Coated pink diamond: A cautionary tale contained an error in figure 2 on p. 37. The labels
for the room temperature and cryogenic spectra in this figure were inadvertently transposed. We thank Nick
Sturman of the Gem & Pearl Testing Laboratory of
Bahrain for bringing this to our attention.

GEMS & GEMOLOGY

SUMMER 2005

187

Book REVIEWS

2005
101 Bench Tips for Jewelers
By Alan Revere, 120 pp., illus., publ.
by MJSA/AJM Press, Providence, RI,
2004. US$42.95 ($36.50 for MJSA
members)
Alan Reveres latest literary contribution to the goldsmith trade ranges
from how to maximize the efficiency
and functionality of a jewelers bench
to troubleshooting a wide array of situations that make up a typical day in
the life of a bench jeweler.
There is one tip to each page, with
the illustrations taking up a large portion of that page. As a seasoned educator, Mr. Revere has made the book easy
to navigate via its index. The writing is
concise, yet understandable and often
quite witty in spite of the technical
content that many might otherwise
consider esoteric. Bear in mind that
such a consistent and user-friendly format will occasionally sacrifice completeness. After all, how much detail
can you give when discussing the intricacies of essential techniques such as
soldering, riveting, engraving, and the
like, in only a few hundred words?
Evidently, Mr. Revere has chosen
hand-renderings by Sean Kane over
photographs to avoid distracting glare,
misleading reflections, and similar
problems associated with shooting at
the bench. Some illustrations have
been cleverly augmented by the selective use of transparency, exploded
views, cross-sections, and sequences.
In an era where conventional
trade skills are no longer passed on
from old masters to their apprentices
through long and intensive relationships (that typically averaged more
like 1001 bench tips per year), this
book fills a need. Self-taught tradespeople, students, and graduates from
trade schools and vocational schools

BOOK REVIEWS

alike, as well as bench jewelers learning on the job, will find 101 Bench
Tips a valuable reference, as will business owners and gemologists-cumdesigners when planning out procedures for the shop floor.
Mr. Reveres oeuvre makes every
attempt to transmit that professional
trade mentality that we goldsmiths
have cultivated for millennia: a tradition of ingenuity.
ROBERT ACKERMANN
Gemological Institute of America
Carlsbad, California

Identification of Gemstones
By Michael ODonoghue and Louise
Joyner, 313 pp., illus., publ. by
Butterworth-Heinemann, Oxford,
UK, 2003. US$59.95
The authors have boldly taken on the
subject of gemstone identification
properties, appearances, sources, imitations, synthetics, treatments, and
separation proceduresand the result
is a book of amazing scope.
To understand the depth of information covered, one only has to look
at the first chapteron diamonds
which begins with diamond properties, types, appearances, and typical
inclusions. The more common diamond imitations are detailed, along
with separation techniques such as
the use of a 10 loupe, reflectivity
meter, thermal conductivity probe,
and immersion in various liquids.
Gem-quality synthetic diamonds from
General Electric, Sumitomo, De Beers,
and Russia are covered next.
Treatments such as the use of ink,
radium salts, various irradiation procedures (sometimes combined with heat
treatment), HPHT, fracture filling, and

EDITORS
Susan B. Johnson
Jana E. Miyahira-Smith
Stuart Overlin

diamond thin films are covered. Their


effect on a diamonds clarity and/or
color and the various methods used to
detect them are discussed. This chapter (like most of the other gem chapters) ends with a section titled
Reports of Interesting and Unusual
Examples from the Literature.
The next two chapters cover
corundum and beryl, giving varieties,
properties, descriptions (including by
source), treatments (heat and diffusion for corundum, fracture filling for
emerald), synthetics by method and
manufacturer, imitations, and composites (assembled stones). As in all
gem chapters, the authors describe
procedures and properties that can be
used to separate the gem covered
from other gem materials with which
they are commonly confused.
The remaining gem chapters cover
opal, quartz, the jade minerals, chrysoberyl (alexandrite, cats-eye, and other
varieties), topaz, tourmaline, the garnet group, spinel, peridot, zircon,
moonstone and other feldspars,
rhodochrosite, tanzanite (zoisite),
rhodonite, turquoise, lapis lazuli,
pearl, amber, ivory, jet, coral, shell,
and natural and artificial glass.
Other chapters highlight information on basic gem testing equipment, metals, ceramics, plastics,
assembled stones, less common gemstones, rarely fashioned species, gem
materials in their rough state, how
crystals are grown, synthetics for the
collector, treatments, and locality

GEMS & GEMOLOGY

*This book is available for purchase


through the GIA Bookstore, 5345
Armada Drive, Carlsbad, CA 92008.
Telephone: (800) 421-8161; outside
the U.S. (760) 603-4200. Fax: (760)
603-4266. E-mail: myorder@gia.edu

SUMMER 2005

189

information. The book ends with a


six-page glossary, recommended reference materials, and an index.
As stated earlier, the book covers
an amazing scope and depth of information, but unfortunately the title is a
bit misleading. Because of its organization mostly by gem species or group,
one would have to be fairly far along
in the identification of an unknown
gemstone for this book to be of benefit. By not choosing to use a color or
property (e.g., refractive index) format,
identifying a gemstone with this book
is a little like trying to look up a number in a telephone directory without
knowing the persons name.
Although the book includes a collection of 60 full-color photomicrographs of various gemstone features,
there is unfortunately no reference to
any of the individual images in the
text. As a result, the reader must try
to understand difficult concepts
through written descriptions alone.
For example, the doubling of facet
junctions seen in some directions of
some diamond simulants is described
as looking like tramlines (p. 15).
This description could have been
made much clearer by directing the
reader to color plate 3, which shows
the appearance of doubling as seen in
synthetic moissanite.
One of the authors stated goals
is to keep updating gemstone information by use of a companion website. At the time of this writing,
however, the only information that
appeared on the website were biographies of the authors and some additional instrumentation procedures.
For example, there was no mention
of CVD synthetic diamonds or
beryllium diffusion treatment of
corundum.
Despite these drawbacks, the
information in this book is considerable and valuable. The authors personalities come through in the writing, which adds a human touch to the
subject without compromising the
science of gemology.
DOUGLAS KENNEDY
Gemological Institute of America
London, England

190

BOOK REVIEWS

Minerals and Their Localities


By Jan H. Bernard and Jaroslav
Hyrsl, edited by Vandall T. King, 807
pp., illus., publ. by Granit, Praha,
Czech Republic, 2004. US$145.00
This superb mineralogical compendium provides basic descriptive information for more than 4,200 mineral
species and lists more than 8,500 localities from which they come. Each
mineral species entry (arranged alphabetically) begins with its name in boldface and its chemical formula, followed by a classification number
(Strunz system), descriptive characteristics, physical properties, and crystallographic data, and concludes with a
fairly comprehensive listing of its
localities. There are also entries (in
bold italics) for certain other materials
that the authors deem worthy of inclusion (such as inadequately described
minerals, intermediate compositional
members, well-known mixtures,
organic materials, etc.). A select number of varietal names (in italics) are
cross-referenced in this alphabetical
listing and are further described under
the corresponding species entries.
Several minerals that are pending formal approval as species are so marked,
and species that may have utility as
gemstones are also identified.
The 1,000 color photographs,
though small, are well chosen to
demonstrate a given species, highlight
unusual occurrences, and add visual
appeal to the overall presentation. The
thick, glossy paper stock is durable and
enhances the reproduction of the
images. A colored border on the pages
serves as an alphabetical index that is
useful as well as attractive.
The book concludes with a table
of The Richest Type Localities of the
World, a list of references, and an
alphabetical list of mineral localities.
This last section, which includes all
8,500 localities with associated
species, is particularly valuable. While
the locality list is extensive and well
chosen to include virtually all important sources, it is by necessity not allinclusive.
A unique feature of this compendi-

um is the organization of locality


information for each mineral according to geologic environment: magmatic, pegmatitic, contact metasomatic,
hydrothermal, alpine fissure, sedimentary, regional metamorphic, and products of weathering. The authors call
these environments genetic classes.
The discussion of this format in the
introduction is required reading if one
is to use the book to best advantage.
This is particularly true when referring to entries for minerals that occur
in several geologic environments. For
example, all of the localities where
quartz has formed by magmatic processes are listed first, followed by its
pegmatitic occurrences, and so forth.
While this organization is quite
instructive, it can make scanning for
specific information somewhat protracted, if not tedious, for extended
entries. Both current and historical
localities (including those no longer
producing) are included.
The descriptive information is
well written, even lively, and is effective in enabling readers to visualize
the described species and varieties.
Considering the enormous amount of
information presented, there are relatively few spelling errors, although
they may occasionally be disconcerting, such as when the same locality or
mineral feature is spelled several different ways in the same entry.
Perhaps because of the length, the
authors have employed many spacesaving tricks, which create some
minor readability problems. For example, there is no space between mineral
entries, and while the boldface species
names serve as clear demarcation, the
non-boldface variety entries get lost. If
space had not been an issue, it would
have been useful to provide a glossary
of geologic terms. As it is, the authors
occasional use of technical jargon
makes it a good idea for the nontechnical user to have a geologic dictionary close at hand.
This book was a monumental
undertaking, the product of a lifetime
of work and a hands-on understanding of the sort of information that is
helpful to anyone interested in miner-

GEMS & GEMOLOGY

SUMMER 2005

als and their localities (and geologic


environments). An international flavor comes through, assuring the reader that the book is truly global in perspective. The authors also have paid
special attention to localities that produce gemstones. The descriptions
associated with each genetic class
and locality make this book unique.
Each species entry can become a
learning experience and enhance the
readers general knowledge of mineral
occurrences.
Minerals and Their Localities is
an excellent and extremely useful reference. For anyone interested in minerals and gems with particular reference to their localities, there is no
better resource available.
DOROTHY ETTENSOHN
Natural History Museum of
Los Angeles County
Los Angeles, California

Masterpieces of the Mineral


World: Treasures from
The Houston Museum of
Natural Science
By Wendell E. Wilson and Joel A.
Bartsch with Mark Mauthner, 264
pp., illus., publ. by The Houston
Museum of Natural Science and The
Mineralogical Record in association
with Harry N. Abrams Inc., New
York, 2004. US$75.00
Mineral collecting has been around for
centuries and continually becomes
more sophisticated in its outlook. The
finest mineral specimens are objets
dart, and there is a growing overlap
between the mineral and gem markets. Therefore, this is an excellent
time to expose gem dealers to a book
explaining the finer points of mineral
specimen appreciation. Unfortunately,
the text of this handsome book is a little too biased to be entirely useful.
Masterpieces of the Mineral World
uses a selection of gorgeous mineral
specimens from the Houston Museum
of Natural Science to explain connoisseurship in mineral collecting. There
are three main parts: The Discerning

BOOK REVIEWS

Eye, which explains the factors that


make a mineral specimen desirable
from an aesthetic point of view; a
series of plates and a section called
The Paths of Discovery, which detail
the individual histories of 80 mineral
specimens; and A Royal Passion,
which reviews the history of mineral
collecting by aristocrats in Europe.
In the first part, Wilson lists several factors that might qualify a mineral
specimen for masterpiece status:
crystal form, color, transparency, luster, size, perfection, crystal orientation/grouping and related aesthetic
visual composition, and provenance
(both natural origin and history in
human hands). Other factors include
associated species, durability, preparation, authenticity (persistence of the
specimens unmodified state), matrix,
and rarity.
Yet some important considerations
to a gemological audience are only
touched on or are missing from this
text. Every mineral specimen is a collaboration between nature and human
beings, from the original collector to
the final owner; at every stage, a
human has had to recognize, preserve,
and sometimes improve the specimen. Many aesthetic factors are subject to modification by human activity,
and standards of preparation and
restoration have changed over time.
Current practices include trimming
matrix with hand tools or saws; cleaning specimens with chemical solutions; improving color, transparency,
and luster by oiling with mineral oil;
and gluing broken crystals back together, or reattaching broken crystal groups
to their matrix. All these practices can
improve the aesthetic factors mentioned above, and some are specifically
mentioned for individual specimens in
the section that follows.
As with gems, though, the definition of what is considered an acceptable modification changes over time,
and some previously accepted mineral
enhancement practices may now be
considered fakery. In addition, a minerals locality is often an important
factor in its value, yet often that information only loosely accompanies the

piece. There is no overall disclosure


practice in the mineral field; practices
vary from dealer to dealer. Thus the
transition from the gem market to the
mineral market is as challenging as
the opposite transition would be.
Photos of the 80 specimens by Jeff
Scovil and Harold and Erica Van Pelt
are excellent. The accompanying text,
however, reads as though written from
a sellers perspective, and sometimes
verges on hyperbole. For instance,
well-formed transparent grossular crystals are not extremely rare. Another
entry claims for size, luster, transparency, sharpness, and form, the
Piuva [Russian] ferro-axinites have
no rivalexcept for, in this reviewers opinion, the best from the New
Melones Dam, California. Yet the
mineral specimens are beautiful
enough to speak for themselves, and
the authors should be congratulated
for including rare and humble species
as well as conventional beauties.
As the final essay on princely collections explains, mineral collecting is
an odd passion, involving scientific
curiosity, aspirations to completeness,
and the competitive urge for bragging
rights. Early mineral collections were
intended for medical research, but they
later became hubs of social activity and
tourism for the learned leisure class.
The mineral collections of counts and
royal physicians were purchased by
museum patrons and eventually
became national treasures. Today,
mineral collectors come from all socioeconomic backgrounds, all cultures,
and all age groups. One can never accumulate the best collection, only the
best available. All the standard caveat
emptor principles apply: Learn from
many sources, not just one; be suspicious of anything too good to be true;
and take everything with a grain of salt.
Nevertheless, as shown in this
beautiful volume, we can only marvel
at the treasures our planet provides
and at the talents their discoverers
and preparers demonstrate. Together,
what masterpieces they create!
MARY L. JOHNSON
Gemological Institute of America
Carlsbad, California

GEMS & GEMOLOGY

SUMMER 2005

191

Gemological ABSTRACTS

2005
EDITOR
A. A. Levinson
University of Calgary
Calgary, Alberta, Canada

REVIEW BOARD
Christopher M. Breeding
GIA Gem Laboratory, Carlsbad
Maha Calderon
Carlsbad, California
Jo Ellen Cole
Vista, California
Michelle Walden Fink
GIA Gem Laboratory, Carlsbad
Eric Fritz
GIA Gem Laboratory, Carlsbad
R. A. Howie
Royal Holloway, University of London
Alethea Inns
GIA Gem Laboratory, Carlsbad
David M. Kondo
GIA Gem Laboratory, New York
Taijin Lu
GIA Research, Carlsbad
Wendi M. Mayerson
GIA Gem Laboratory, New York
Kyaw Soe Moe
GIA Gem Laboratory, New York
Keith A. Mychaluk
Calgary, Alberta, Canada

COLORED STONES AND


ORGANIC MATERIALS
Gemological and mineralogical properties of black jadeite. M.
Zhang, P. Hou, and J. Wang, Jiangsu Geology, Vol. 28,
No. 2, 2004, pp. 100102 [in Chinese with English
abstract].
Mozui, which means black jadeite in Chinese, is mined in
Myanmar and southern China. Recently, it has become common in Asian jade markets. It is essentially monomineralic,
consisting of > 90% omphacite, which is a pyroxene (like
jadeite). This article reports on the gemological and mineralogical properties of this blackish green to black material and provides a comparison to jadeite.
The refractive indices (1.6671.670) and specific gravity
(3.343.44) of black jadeite are almost identical to those of
jadeite. It has a strong vitreous luster, and polished stones have
a smooth surface. Petrographic studies show that the omphacite
occurs as prisms or fibers of varying length; the grain size is usually < 0.3 mm. Albite, tremolite, and some minute opaque
grains with metallic luster are minor constituents. Only subtle
differences are found in the infrared absorption spectra and Xray powder diffraction patterns of the two materials. The chemical formula of the omphacite in black jadeite is
(Ca,Na)(Mg,Fe2+,Fe3+,Al)(Si2O6); the ratio Na/(Na+Ca) is 0.45,
which lies between that of jadeite (>0.8) and diopside (<0.2). The
main trace elements are Sr, Cr, and Mn.
Even though there is a clear difference in the chemical
compositions of the main pyroxene minerals in black
jadeite (omphacite) and jadeite, the authors suggest that the
black material should nevertheless be considered jadeite, since
their physical properties are almost same. Yet, they acknowledge resistance to this suggestion; the name omphacite jade

Joshua Sheby
GIA Gem Laboratory, New York
James E. Shigley
GIA Research, Carlsbad
Boris M. Shmakin
Russian Academy of Sciences, Irkutsk, Russia
Russell Shor
GIA, Carlsbad
Rolf Tatje
Duisburg University, Germany
Christina Taylor
Boulder, Colorado
Sharon Wakefield
Northwest Gem Lab, Boise, Idaho

192

GEMOLOGICAL ABSTRACTS

This section is designed to provide as complete a record as


practical of the recent literature on gems and gemology. Articles
are selected for abstracting solely at the discretion of the section
editor and his reviewers, and space limitations may require that
we include only those articles that we feel will be of greatest
interest to our readership.
Requests for reprints of articles abstracted must be addressed to
the author or publisher of the original material.
The reviewer of each article is identified by his or her initials at the
end of each abstract. Guest reviewers are identified by their full
names. Opinions expressed in an abstract belong to the abstracter and in no way reflect the position of Gems & Gemology or GIA.
2005 Gemological Institute of America

GEMS & GEMOLOGY

SUMMER 2005

was recommended by C. M. Ou Yang et al. (Recent studies on inky black omphacite jade, a new variety of pyroxene jade, Journal of Gemmology, Vol. 28, No. 6, 2003,
pp. 337344).
TL
Harvesting impacts and invasion by alien species
decrease estimates of black coral yield off Maui,
Hawaii. R. W. Grigg, Pacific Science, Vol. 58, No. 1,
2004, pp. 16.
After 40 years of successful management, three new developments threaten the sustainability of the black coral fishery in Hawaii. First, harvesting pressure on the colonies
has increased, while the biomass of black coral off Maui
has decreased 25% since 1976. Second, an alien species of
coral, Carijoa riisei (often called cutthroat coral), has
overgrown many of the black coral colonies. Initially
detected in Pearl Harbor, Oahu, in 1972, this invasive
species is believed to have been brought from the Atlantic
Ocean on the hulls and/or in ballast water of ships
between 1940 and 1970. During the 1990s, C. riisei was
reported at sites in the eight major Hawaiian Islands.
Third, increasing sales of black coral in recent years (about
$30 million retail in Hawaii alone) is placing more
demands on the resource.
A larger size (height) limit of the coral and a reduction
in the amount harvested are measures that can be implemented to help conserve the resource. Divers and the
coral industry must abide by stricter regulations to ensure
the future of the black coral fishery in Hawaii.
MC
Study on compositions and colouring mechanism of
freshwater cultured pearls. M. Yang, S. Guo, L. Shi,
and W. Wang, Journal of Gems and Gemmology,
Vol. 6, No. 2, 2004, pp. 1013 [in Chinese with
English abstract].
Chinese freshwater cultured pearls display a large range of
colors (e.g., white, pink, gray, purple, red), however, there
is no reliable explanation for these variations. In this article, the authors attempt to correlate the color variations
with trace-element contents and other parameters. Three
color varieties (white, orange-red, and dark purple; two
samples of each), representing the color range of Chinese
freshwater cultured pearls, were analyzed for their inorganic mineral components by X-ray diffraction, for selected trace metals by plasma emission spectroscopy, for conchiolin by FTIR spectroscopy, and for carotenoids by
Raman spectroscopy.
There were no significant differences in the inorganic mineral components of the three color varieties.
However, the concentrations of the trace metals Zn, Mg,
Ti, V, Ag, and Co varied significantly, being particularly
high in the dark purple cultured pearls. For example, the
Zn content (945 ppm) in the dark purple specimens was
nearly eight times higher than in the white ones (121
ppm). The concentrations of Ti and V were higher in the

GEMOLOGICAL ABSTRACTS

orange-red cultured pearls than in the white samples,


but less than in the dark purple specimens. Cu, Fe, Cr,
and Al were not detected (<0.1 ppm) in any of the samples. From these data, the authors conclude that selected
trace elements are responsible, at least in part, for the
coloration of Chinese freshwater cultured pearls.
Among the organic components, there were no significant differences in the conchiolin in the various samples. However, Raman spectra showed four carotenoidrelated peaks in both the orange-red and the dark purple
samples, but not in the white ones, suggesting that
carotenoids also play a role in the coloration of these
freshwater cultured pearls.
TL
Study on star spinel. L. Tian, W. Huang, H. Liu, and Y.
Chen, Journal of Gems and Gemmology, Vol. 6, No.
2, 2004, pp. 13 [in Chinese with English abstract].
Spinel occasionally displays asterism with either a sixrayed or a four-rayed star; however, multiple sets of stars
in the same sample are relatively rare. This article
describes three oval cabochons of bluish or grayish black
Sri Lankan spinel (2.88, 3.21, and 3.31 ct) that displayed
four sets of six-rayed stars and three sets of four-rayed
stars, all of which were bright and sharp. An abundance
of tiny needle-like inclusions parallel to several octahedral planes were observed with the microscope. Electronmicroprobe imaging showed these inclusions to be < 1
m in diameter and 520 m long. They were composed
of two different minerals based on their cross-sectional
shapes and chemical compositions: (1) a triangular Fecontaining mineral, and (2) an irregular-shaped Ca-containing mineral. The multiple sets of stars were attributed to these tiny inclusions distributed along the three
edges of each octahedral (111) face in the spinel crystals.
Eight sets of six-rayed stars together with 12 sets of fourrayed stars would be seen if the spinel crystals had been
polished into spheres.
TL

DIAMONDS
About signs of mechanical and chemical effects upon diamond crystals from the Urals deposits. V. N.
Anfilogov, Proceedings of the Russian Mineralogical Society, Vol. 133, No. 3, 2004, pp. 105108
[in Russian with English abstract].
The complex morphology of diamond crystals is reviewed
for the possible application of locating their primary
source. Dislocations on the surface of crystals cannot be
explained by mechanical distortion; rather, the dislocations formed during crystal growth due to internal defects.
The occurrence of rounded diamonds in the Urals placer
deposits cannot be explained by tumbling in water. It is
suggested that the primary sources of these diamonds are
nearby kimberlites.
RAH

GEMS & GEMOLOGY

SUMMER 2005

193

Carbonatitic melts in cuboid diamonds from Udachnaya


kimberlite pipe (Yakutia): Evidence from vibrational
spectroscopy. D. A. Zedgenizov [zed@uiggn.nsc.ru],
H. Kage, V. S. Shatsky, and N. V. Sobolev, Mineralogical Magazine, Vol. 68, No. 1, 2004, pp. 6173.
Micro-inclusions (110 m) in 55 diamonds of cubic habit
from the Udachnaya kimberlite pipe in Russia were
shown by IR and Raman spectroscopy to contain a multiphase assemblage that includes carbonates, olivine,
apatite, graphite, water, and silicate glasses. These microinclusions had an elevated internal pressure that indicated
the original materials were trapped during growth of the
host diamond. These internal pressuresextrapolated to
mantle temperatureslie within the stability field of diamond, but the temperatures are relatively low. Such temperatures are typical for the formation of cuboid diamonds.
In contrast to previous data for African diamonds, the
micro-inclusions in the Udachnaya cuboids were extremely carbonatitic in composition [H2O/(H2O+CO2) ~520%],
and the assemblage contained in their micro-inclusions
was similar to that found in some types of carbonatites.
The low water and silica contents testify that the material
in the micro-inclusions of the Udachnaya diamonds was a
near-solidus carbonatitic melt.
RAH
Do Arraial do Tejuco Diamantina: 290 anos de produo
de diamantes [From the Arraial do Tejuco to
Diamantina: 290 years of diamond production]. R.
H. Corra-Silva, Diamond News, Part I. Vol. 5, No.
17, 2004, pp. 4044; Part II. Vol. 5, No. 18, 2004, pp.
1319 [in Portuguese].
Part I of this article describes the early history of the
Diamantina mining district in Minas Gerais, Brazil, from
the discovery of gold in the 17th century and of diamonds
around 1714 to the end of the 19th century, by which time
Diamantina had lost its position as the worlds leading diamond producer to South Africa. It focuses on certain
aspects of the industry, such as the regulation of diamond
production and trade by the Portuguese Crown and the
role of some prominent personalities.
Part II gives a description of the geologic setting, mineralogy, and mining at Diamantina. The diamonds are
found in conglomerates, and in colluvial and alluvial
deposits (the primary source rocks have not yet been identified). They are generally small and of gem quality.
Diamonds from the conglomerates frequently have a
green or brown coating from natural irradiation. Mining is
done both by garimpeiros who also rework old diggings,
and by mechanical operations (dragas) mostly on the
Rio Jequitinhonha. Minerals that accompany the diamonds are anatase, kyanite, lazulite, magnetite, gold,
quartz, rutile, and zircon.
RT
Prospects of diamond-bearing ability in Ukraine and
trends of geological prospecting works. D. S.
Gursky, V. S. Metalidi, V. L. Prykhodko, and Yu. V.

194

GEMOLOGICAL ABSTRACTS

Geiko, Mineralogical Journal (Ukraine), Vol. 26,


No. 1, 2004, pp. 717.
This evaluation of possible diamond-bearing occurrences
in the six major structural units of the Ukrainian shield
takes into consideration potential secondary (placer)
deposits as well as primary (i.e., kimberlite and lamproite)
sources in the region. Although no deposits are currently
known, there are encouraging indications that the
Ukrainian shield has diamond potential. For example, (1)
detrital Cr-bearing pyropes often associated with diamonds are reported from several areas, and (2) diamonds
0.20.3 mm in size have been recovered from sands of the
Baltian Neogene region and in the beach deposits of the
northern coast of the Sea of Azov. The authors recommend detailed mapping of heavy minerals, as well as the
use of other prospecting methods. [Abstracters note: This
is one of six articles, all in English, in a thematic issue of
the Mineralogical Journal (Ukraine) on the potential for
diamond occurrences in the Ukraine. Abstracts of two
additional articles follow.]
RAH
Age correlation of endogenic processes of the Slave
(Canada) and Middle-Peri-Dnieper (Ukraine) cratons
in connection with the diamond-bearing ability
problems. N. P. Shcherbak, G. V. Artemenko, and
A. V. Grinchenko, Mineralogical Journal (Ukraine),
Vol. 26, No. 1, 2004, pp. 1823.
The Slave craton of the Canadian shield and the MiddlePeri-Dnieper (MPD) craton of the northwestern Ukrainian
shield both stabilized in the Archean and have not been
subjected to any considerable reworking in the Proterozoic
or Phanerozoic. This stability is an essential condition for
the diamond occurrences in the Slave craton. In the MPD
craton, diamonds have been found only in Tertiary and
Quaternary sedimentary formations, but, based on the
occurrence of 23002000 and 14001200 million-year-old
dike complexes in both cratons, the MPD craton is considered a promising prospect for diamond-bearing rocks.
RAH
Paleotectonic, petrological and mineralogical criteria of
diamond-bearing ability of the Ukrainian Shield. V.
M. Kvasnytsya, Ye. B. Glevassky, and S. G. Kryvdik,
Mineralogical Journal (Ukraine), Vol. 26, No. 1,
2004, pp. 2440.
Regional estimates are given for the possibility of diamond
discoveries in the various structural blocks of the
Ukrainian shield on the basis of Cliffords rule, which
states that economic diamond deposits occur only in
ancient stable cratons, or archons, with a crustal age of
not less than 2,800 million years. The Middle-Peri-Dnieper
craton in the Ukrainian shield has thickened crust, and the
most promising region for the discovery of mantle (i.e.,
kimberlite or lamproite) diamond deposits is where the
archon slopes adjacent to a sediment-filled linear trough
called the Dnieper-Donets aulacogen.
RAH

GEMS & GEMOLOGY

SUMMER 2005

GEM LOCALITIES
Archaeological geology of the worlds first emerald mine.
J. A. Harrell [james.harrell@utoledo.edu], Geoscience Canada, Vol. 31, No. 2, 2004, pp. 6976.
An archaeological and geologic survey of the mountain
valley region of Wadi Sikait in the Eastern Desert of Egypt
was conducted to map the distribution of ancient mine
workings, deduce ancient mining methods, and describe
the geologic occurrence of emerald at this, the worlds first
emerald mine. Several other sites in the area also have
indications of ancient emerald extraction. Mining probably began toward the end of the Ptolemaic period in the
1st century BC. Most of the activity, however, dates to the
Roman and early Byzantine periods, from the late 1st century BC through the 6th century AD. Renewed small-scale
mining during the first three decades of the 20th century
proved unsuccessful, probably due to the poor quality of
the emeralds.
The author provides a map of the mine workings,
which combines topographic and geologic information.
The Wadi Sikait area is divided into sections known as
the North Village, Middle Village and Road, and South
Village. Archaeological excavations of the South Village
area have uncovered ancient temples and administration
buildings, possibly from the first century AD.
The emerald deposits are characterized by the intrusion of Be-bearing quartz or pegmatite veins into Cr- or Vbearing mafic or ultramafic rocks (i.e., the source of the
chromophores). Emeralds formed in the contact zone
between phlogopite schist and the quartz/pegmatite
veins.
JEC
Marble-hosted ruby from Vietnam. P. V. Long [vggc@fpt.vn],
H. Q. Vinh, V. Garnier, G. Giuliani, and D. Ohnenstetter, Canadian Gemmologist, Vol. 25, No. 3,
2004, pp. 8395.
This article compares the gemological characteristics and
geologic setting of rubies from Luc Yen and Quy Chau in
northern Vietnam. At Luc Yen, rubies occur as disseminated crystals in small veinlets or fissures within marbles
of Upper Proterozoic to Lower Cambrian ages. Secondary
ruby deposits occur in associated alluvial gravels. At Quy
Chau, the economic ruby and sapphire deposits are found
only in placers.
The ruby crystals from Luc Yen examined were typically hexagonal bipyramids, whereas those from Quy
Chau commonly formed short hexagonal prisms.
Compared to the Luc Yen material, the color of the Quy
Chau rubies was usually more saturated and showed a
stronger red fluorescence to long- and short-wave UV
radiation. R.I. values of rubies from both areas were
1.7621.770 (birefringence ~0.008). S.G. values for samples from Luc Yen were 3.924.01, whereas those from
Quy Chau were 3.944.05. Rubies from both areas exhibited twinning, straight and angular growth zoning, and (in

GEMOLOGICAL ABSTRACTS

some instances) various forms of color zoning. The most


common mineral inclusions in material from both localities were apatite, calcite, dolomite, rutile, diaspore, phlogopite, and zircon. Three kinds of fluid inclusions (both
primary and secondary) also were found. The inclusion
characteristics and geologic settings indicate that rubies
from these deposits are the result of high-grade (amphibolite facies) regional metamorphism.
JES
NuristanSouth Pamir province of Precambrian gems. A.
K. Litvinenko [litvinenko78@yahoo.com], Geology
of Ore Deposits, Vol. 46, No. 4, 2004, pp. 263268.
A highly metamorphosed Precambrian tectonic block
measuring 660 165 km in the Afghanistan-PakistanTajikistan region has a remarkable abundance of gem
deposits. More than three dozen localities contain minerals suitable for use as gems or ornamental stones. A partial
list of these materials includes lazurite, spinel, clinohumite, ruby, sapphire, scapolite, cordierite, emerald, and
tourmaline. Many of these come from various types of
metamorphic deposits, the nature of which is determined
by the composition of the sedimentary protoliths (e.g., the
presence of highly aluminous clay) and by the intensity
and type of metamorphism.
EF

INSTRUMENTS AND TECHNIQUES


Ein notwendiger Test in der Perlenuntersuchung [A necessary test for the pearl identification]. H. A. Hnni
[gemlab@ssef.ch], Gemmologie: Zeitschrift der
Deutschen Gemmologischen Gesellschaft, Vol. 53,
No. 1, 2004, pp. 3942 [in German with English
abstract].
X-radiography is a standard method for distinguishing natural from cultured pearls by revealing their different internal characteristics. With recent progress in pearl culturing,
the structures of natural and cultured pearls are becoming
increasingly similar. Unlike their saltwater counterparts,
freshwater pearls (natural or cultured) luminesce to X-ray
excitation because of the Mn content of their nacre. As the
nuclei of most cultured pearls are made of freshwater
nacre, this luminescence (as viewed through the subsequent cultured overgrowth) can be used as an additional
test to identify various kinds of natural and cultured
pearls. The author gives a brief description of the practical
application of this technique.
RT
Impurity measurements in diamond using IR. W. Taylor
[pterodia@wxc.com.au], Rough Diamond Review,
No. 4, 2004, pp. 4042.
IR spectroscopy provides a quantitative method for identifying nitrogen and other light-element (e.g., H, O, B) impurities and their associated defects in diamonds; these elements may substitute for carbon atoms in the diamond
lattice in amounts up to several thousand parts per mil-

GEMS & GEMOLOGY

SUMMER 2005

195

lion. The abundance and configuration of such impurities


are of importance for determining the post-growth history
of diamonds in the earths mantle and for distinguishing
diamonds from different sources. Within individual crystals, the zonation of nitrogen content and aggregation state
can be mapped and correlated to growth history. Nitrogenaggregation plots show a correlation between nitrogen
population types and diamond properties, such as color
and morphology.
IR spectra that are characteristic of the atomic configuration and concentration of nitrogen and the other lightelement impurities, and molecular submicroscopic fluid
inclusions (e.g., H2O, CO2), are illustrated and/or discussed in this article. Also presented are practical aspects
of measuring and interpreting IR spectra of diamonds (e.g.,
using software that deconvolutes the spectra to isolate the
contributions from various nitrogen centers).
AI

JEWELRY RETAILING
Are natural diamond engagement rings forever? E. H.
Fram and R. Baron, International Journal of Retail
& Distribution Management, Vol. 32, No. 7, 2004,
pp. 340345.
The authors conducted a study of the attitudes toward natural diamond engagement rings among 459 students at the
Rochester Institute of Technology, New York; these
young people are the prime target market for such rings.
One-third of the respondents were in the 1822 age group,
while the rest were 23 and older. The male:female ratio
was 55:45, with 93% of the students coming from the U.S.
Of the full sample, 15% were already married or engaged,
54% said they had no expectation of becoming engaged
within the next three years, and 31% believed they might
become engaged in that time period.
While a majority (61%) of respondents believed that a
natural diamond engagement ring was very or somewhat
important, a significant number were either neutral (22%)
or believed such rings were somewhat/very unimportant
(16%). In addition, 45% of the women believed it was
appropriate to celebrate an engagement without a natural
diamond ring; only 28% of the men agreed with that
statement. About one-third of the respondents would consider alternatives to natural diamond engagement rings,
either jewelry or nonjewelry, while nearly one-fourth
were neutral on the subject. The report concluded that
the number of women predisposed to acquiring a diamond engagement ring might not be as high as believed,
and this might represent a downward trend that could
impact future diamond engagement ring sales.
RS
Diamond geezers. M. Prince [princem1@southernct.edu],
Business Strategy Review, Vol. 16, No. 1, 2005, pp.
2227.
The diamond engagement ring is a product that creates

196

GEMOLOGICAL ABSTRACTS

desire beyond reason in the eye of the beholder, especially


those rings that fulfill the requirements of a prestige
brandproducts for which consumers will pay a premium. The characteristics of these brands are that they set
themselves on a higher plane than the competition, represent high expectations of quality and value, and make a
personal statement about the user/wearer.
Diamond engagement rings offer a conspicuous display of romantic emotion and social position. Diamonds
from a prestige jeweler can also stimulate a feeling of
attachment toward that brand, kinship with other purchasers of that brand, disdain for competing products in
the same market, and an elitist mentality. Brand choices
can be affected by age, gender, ethnic background, education, and sophistication levels.
RS
Evolution of an advertising campaign. A. DeMarco, JCK,
Vol. 176, No. 5, 2005, pp. 149152.
This article details the development of a new advertising
campaign for Lockes Diamantaires, a boutique-type retail
diamond jewelry store established by a De Beers Diamond
Trading Company sightholder in Manhattan. The advertising agency, AgencySacks, began by proposing a campaign
that would combine credibility and good value to a predominantly female clientele in the age range 26 to 65 and
earning more than $75,000/year. The goal was to position
the store as a professional, quality-oriented alternative to
both wholesale-to-the-public operations and high-end
retailers. The creative staff of the agency developed three
potential advertisements that focused on image and emotions rather than product. However, some members of
Lockes staff wanted some products featured, so two different advertisements were produced. One was an image ad
for magazines, while the other, for newspapers, stressed
product and value.
RS

TREATMENTS
Doping by diffusion and implantation of V, Cr, Mn and Fe
ions in uncoloured beryl crystals. J. C. R. Mittani
[juan.mittani@dfn.if.usp.br], S. Watanbe, M. Matsuoka, D. L. Baptista, and F. C. Zawislak, Nuclear
Instruments and Methods in Physics Research B,
Vol. 218, 2004, pp. 255258.
The coloration of slabs from a natural crystal of colorless
beryl (from Salinas, Minas Gerais, Brazil) was carried out
through doping by diffusion and ion implantation with V,
Cr, Mn, and Fe. [Editors note: The authors do not describe
the colors produced by this treatment process.] The beryl
slabs were cut to test whether treatment was easier perpendicular or parallel to the c-axis. Specimens were placed
in an alumina crucible with chemical-grade compounds
(VCl3, Cr2Cl3, MnCl2, or FeCl3) and heated to 700, 725, or
750C for periods of 7, 14, or 21 days. After treatment,
they were analyzed using optical absorption, electron para-

GEMS & GEMOLOGY

SUMMER 2005

magnetic resonance, and inductively coupled plasmalaser


ablationmass spectroscopy. The results showed that the
ion species could be doped into the beryl by the diffusion
procedure. However, the low dosage used during the
implantation process failed to introduce any significant
level of the ions. One interesting feature documented during the ion implantation process was the changing of preexisting Mn2+ and Fe3+ into Mn3+ and Fe2+. Diffusion parallel to the c-axis was easier than perpendicular to this axis
due to the structural channels in the beryl.
EF
If you cant stand the heat, get out of the gem business:
Part IV. G. Roskin. JCK, Vol. 175, No. 9, 2004, pp.
109114.
While the vast majority of HPHT-treated diamonds can be
identified, some still confound gemological laboratories.
GIA has documented nearly 11,000 HPHT-annealed diamonds submitted for reports. Only a small percentage (primarily colored diamonds) were submitted without having
been declared as HPHT-treated. Other laboratories have
reported seeing limited numbers of HPHT-annealed diamonds. The range of diamonds that are treated has
expanded beyond the original type IIa, most commonly to
some varieties of type Ia.
The article questions the claim, made by some HPHT
diamond treaters, that the process finishes what Mother
Nature started. Quite often the process of treating diamonds involves temperatures far higher than those found
naturally.
RS
Quality enhancement of Vietnamese ruby by heat treatments. P. Winotai [scpwn@mahidol.ac.th], P. Limsuwan, I. M. Tang, and S. Limsuwan, Australian
Gemmologist, Vol. 22, No. 2, 2004, pp. 7277.
The heating of Vietnamese rubies in an oxygen atmosphere improves their color and clarity. Trace amounts of
Fe3+ in these rubies causes a pale yellow coloration, whereas the intervalence charge-transfer mechanism between
Fe2+ and Ti4+ gives them an undesirable bluish color. The
latter can be diminished by heating the ruby in oxygen so
that most of its Fe2+ is converted to Fe3+, resulting in a
more intense red. X-ray diffraction analysis showed that
the cation/anion ratio of the structure was smallest after
heat treatment at 1,300C. The number of Fe2+ ions converted to Fe3+ (as detected by electron spin resonance spectrometry) was found to increase with temperature.
Vietnamese ruby may acquire an intense red color (with
no yellow or blue modifier) after optimal heating at
1,500C for 12 hours.
RAH
Study on colorific mechanism of topaz after irradiated by
positron annihilation technique. Y. Deng and Y.
Zhang, Nuclear Electronics & Detection Technology, Vol. 24, No. 1, 2004, pp. 6266 [in Chinese
with English abstract].
Topaz can be color enhanced by irradiation with gamma

GEMOLOGICAL ABSTRACTS

rays, neutrons, and high-energy electrons; however, the


mechanism(s) by which the radiation changes the color is
still unclear. To gain insight into this problem, samples of
untreated and irradiated topaz were studied using the
positron annihilation technique. This technique yields
information on crystal lattice defects, such as vacancies
and micro-voids, by measuring the average lifetimes of
positrons after irradiation. From such data, causes and
mechanisms of color enhancement can be deduced.
Natural topaz samples from Taishan, Guangdong,
China, were divided into three groups for the experiments:
colorless, blue (from neutron irradiation), and light blue
(from high-energy electron irradiation). The average lifetimes of the positrons were about the same in samples
within each group, but there were significant differences
in the average lifetimes between the three groups. The
average lifetimes were proportional to color intensity (i.e.,
longer in darker samples). Irradiation by neutrons and/or
electrons induces many point defects in the topaz lattice,
and these defects are the main cause of color. Some of
these induced defects and associated colors disappear with
annealing. The authors suggest that all methods capable of
producing vacancies and micro-voids in the topaz lattice
could be used for color enhancement.
TL
Topical coatings on diamonds to improve their color: A
long history. T. Moses, C. P. Smith, W. Wang, and
M. Hall, Rapaport Diamond Report, Vol. 27, No.
20, 2004, pp. 9193.
The oldest form of treatment applied to diamonds to mask
or alter their color appearance involves coatings. Coatings
with a colored substance were applied to the backs of diamonds as early as the 16th century in Italy. At that time,
for example, greenish colors were commonly achieved by
adding a green or blue coating to the back of yellow stones.
Yet virtually any color can be imparted to a diamond with
an appropriate coating. Such coatings, which are still occasionally seen today, are relatively unsophisticated, requiring only a steady hand and the proper coating medium.
Through the years, the methods by which coatings have
been applied to diamonds have changed. Since the 1950s,
sputter or vapor deposition techniques have enabled
metallic and fluoride coatings (originally developed for
lenses during World War II) to be applied to diamonds. In
these instances, typically a blue or purple coating is
applied to the entire pavilion or just near the girdle to neutralize or mask a yellow bodycolor.
Coatings are best detected by microscopic examination. Two light sources, darkfield and overhead diffused
light, combined with a subtle diffuser placed between the
diamond and the light well of the microscope, provide the
most effective viewing conditions. The coatings may
appear as blotches or uneven areas of color and frequently
have a blue or purplish tint in those cases where metallic
or fluoride coatings have been applied; gas bubbles are
observed in rare instances. Indications of the presence of a

GEMS & GEMOLOGY

SUMMER 2005

197

coating may be encountered during color grading, from


the different colors seen when the diamond is viewed
face-up and table-down; in addition, the color may be concentrated along the girdle. In the future, coatings of synthetic diamond on natural diamond, applied by chemical
vapor deposition (CVD) techniques, have the potential
not only to modify the less desirable color of a diamond,
but also to add weight to the original stone. Such CVDcoated diamonds may be difficult to detect.
JEC
Yellow and brown coloration in beryllium-treated sapphires.
V. Pisutha-Arnond [pvisut@geo.sc.chula.ac.th], T.
Hger, P. Wathanakul, and W. Atichat, Journal of
Gemmology, Vol. 29, No. 2, 2004, pp. 77103.
Elevated levels of Be (up to ~11 ppm) have been discovered
in both natural and synthetic sapphires heat treated in
Thailand. This article investigates the causes of yellow
and brown coloration in such Be-treated sapphires. Using a
series of irradiation and Be-heating experiments on both
synthetic and natural colorless sapphires, the authors
show that the geochemical behavior of Be mimics that
previously proposed for Mg. Stable yellow or brown colors
can be produced or removed depending on the (Be+Mg)/Ti
ratio of the sapphires, the presence or absence of Fe, and
whether heating is done in an oxidizing or reducing environment.
WMM

MISCELLANEOUS
Diamonds and civil conflicts in AfricaThe conflicts in
central Africa and west Africa. M. K. Kachikwu,
Journal of Energy & Natural Resources Law, Vol.
22, No. 2, 2004, pp. 171193.
This article reiterates the belief that the most serious and
prolonged civil conflicts within African states are the
result of rebel groups and governments battling to control
resources, especially diamonds. Because of their easy
transport across borders and conversion to cash, diamonds
are particularly susceptible to illicit purposes.
The author recounts the histories of civil conflicts in
diamond-producing countries, specifically Angola, Sierra
Leone, and the Democratic Republic of the Congo.
Patterns are cited of colonial plunder and superpower
interference, followed by world neglect, as background for
the recent conflicts. In two of the cases (Sierra Leone and
Angola), revolutions borne out of political or social unrest
were eventually subverted into grabs for diamond wealth.
In the Congo, the author maintains, diamonds were the
prize for outside invasion forces seeking influence within
that vast nation. The case of Liberia is also cited. That
country has essentially no indigenous diamond production, but illicit diamond trade running through its borders
supported the corrupt regime of Charles Taylor. The article also describes the efforts of the Kimberley Process to
stop trade in conflict diamonds. It notes, however, that

198

GEMOLOGICAL ABSTRACTS

without massive intervention from the United Nations


and the superpowers, African nations may again fall into
the same patterns of commodity-driven civil strife.
RS
The diamond frontier. From bleak and barren to blingbling: How a growing industry is changing Canadas
northern communities. D. McDonald, Time
(Canadian edition), Vol. 163, No. 14, 2004, pp. 4855.
The 1991 discovery of diamonds in Canada, and the subsequent opening of the Ekati and Diavik diamond mines,
have transformed economic prospects for the 20,000 aboriginal people of the Northwest Territories (NWT). Native
communities are experiencing increased self-esteem and a
better quality of life than a decade ago, yet they have still
found ways to maintain their traditional values. Much of
the credit for these achievements can be attributed to the
mining companies acknowledgment that they are operating on lands traditionally used for hunting and trapping,
and their negotiating agreements favorable to the native
people. These agreements include such matters as hiring (a
target of 40% in the mines), training (to overcome a lack of
education leading to skilled jobs), and business opportunities (such as service contracts). Never before have the aboriginal people participated to such an extent in resource
development on their traditional lands even though, by
legal statute, royalties from mines in the NWT go to the
federal government.
The results have been spectacular. For example, unemployment among native people has been greatly reduced,
and when mines under construction (e.g., Snap Lake) come
into production, there may even be a shortage of aboriginal
workers to fill the target quotas. Nevertheless, diamonds
also present problems for the NWT aboriginal communities. For example, increased wealth has brought pressures
on communities that have a tradition of sharing. Native
people now distinguish between goods that have been
shared historically (e.g., harpoons and rifles) and those that
have not (e.g., snowmobiles). In addition, the increased
salaries have not solvedand in some cases, have contributed tofamiliar social problems such as alcohol and
substance abuse.
AAL
The Great Diamond Hoax of 1872. R. Wilson, Smithsonian, Vol. 35, No. 3, 2004, pp. 7079.
In 18711872, Philip Arnold and his partner, John Slack,
played on peoples greed and conned millionaires and
politicians into buying stock in a company that supposedly had found a diamond field in Colorado. Mr. Arnold had
acquired knowledge of diamonds from his previous
employer, Diamond Drill Co. With a bag of rough diamonds (presumably taken from the drill maker) mixed
with rubies and garnets, the two con men approached
businessmen and bankers in early 1871 telling of their diamond find and looking for investors. After getting an initial payment of $50,000, the two went to England in July
1871 to buy $20,000 of uncut diamonds and rubies for use

GEMS & GEMOLOGY

SUMMER 2005

in obtaining more investment capital. Seeing actual diamonds in hand (and, later, salted at the supposed
deposit), the speculators were easily duped. In fact, when
some of the investors sent 10% of the uncut stones to
Charles Lewis Tiffany in New York (in October 1871) for
appraisal, he unwittingly supported the scam by reporting
their value as $150,000. (After the hoax was exposed, it
was revealed that Mr. Tiffany did not have much experience with uncut stones.)
The hoax unraveled in October 1872, when Clarence
King, a geologist conducting a survey for the U.S. government, and his men overheard a few people talking about a
diamond find that happened to be in the surveys purview.
If diamonds actually occurred on land he was surveying,
King feared, he would have to explain to Congress why he
and his men had not found them. When they inspected the
so-called diamond field, King noticed that everywhere
there was a diamond there were rubies surrounding it,
which was too perfect a coincidence to be natural. Also,
anthills containing gems were always surrounded by footprints, which was quite suspicious. King concluded that
the diamond field was a hoax and published a letter exposing it in November 1872. Although a grand jury did indict
Arnold and Slack for fraud, it seems the investors, out of
embarrassment, let the case drop.
JS
In the wake of things: Speculating in and about sapphires
in northern Madagascar. A. Walsh, American
Anthropologist, Vol. 106, No. 2, 2004, pp. 225237.
Sapphires were discovered in the area around
Ambondromifehy in the northern part of Madagascar in
1996; before then, the locals held them in such low esteem
that children used them as slingshot pellets. The fact that
foreigners place so much value on these stones, which
have no local use, is a source of great speculation and suspicion among local inhabitantseven the Malagasy traders
who make their livings by buying and selling these stones.
This article describes the complex trading culture that
has grown up around the mining areas. Local miners,
traders, and demarchers (brokers) have developed a trading ethic divided between businessy for legitimate deals
and katramo for deceptive practices. Businessy encompasses many practicessome unquestionably ethical, and
others that require a great deal of awareness on the part of
both parties. Katramo is pure deception. Local traders,
however sharp, consider themselves at a disadvantage to
foreigners because of their ultimate secret, that sapphires have no use or value within their country. Thus,
because the Malagasy do not understand why foreigners
value sapphires, they believe that foreign traders are constantly getting the better of them. Nor do the locals
believe that sapphires are used primarily for jewelry.
Common suspicions hold that the stones are used in
armaments for the Iraq War, electronic gear, watches, and
even to construct impenetrable walls for the houses of
foreign billionaires.
RS

GEMOLOGICAL ABSTRACTS

Injuries and injury care among child labourers in gem


polishing industries in Jaipur, India. R. R. Tiwari
[rajtiwari2810@yahoo.co.in], A. Saha, J. R. Paikh,
and H. N. Saiyed, Journal of Occupational Health,
Vol. 46, No. 3, 2004, pp. 216219.
The gem polishing industry in Jaipur, India, employs about
200,000 people who process colored stones imported mostly from various African sources. About 20,000 of these
workers are children (< 18 years old). For this study,
researchers at Indias National Institute of Occupational
Health interviewed 588 child workers randomly selected
from various cutting operations. Most (78% in this study)
were in the 1014 year age range. They found that these
workers were frequently exposed to minor injuries associated with the cutting, faceting, and drilling of gem materials and were not covered by any type of health plan. The
prevalence of injuries within this age group could indicate
they are too immature to understand safety practices.
Their injuries were often complicated by problems with
hygiene and lack of knowledge regarding wound care and
first aid. Chemicals such as chromium oxide were often
absorbed into their wounds, contaminating those injuries.
They also found that children working longer shifts
(> 6 hours) were more likely to injure themselves than
those working a few hours a day, as fatigue negatively
affects concentration. Those employed less than two
years tended to sustain more injuries because of their lack
of experience.
MC
The next African revolution. S. Benson, New York
Diamonds, Vol. 86, January 2005, pp. 4046.
The South African government is considering legislation
that could radically transform the diamond industry in
that country. Its aim is to increase beneficiation (i.e.,
deriving greater added value from diamond resources in
the form of employment, majority ownership, and government revenues). Proposed amendments to the South
African Diamond Act of 1986 would allow the countrys
Diamond Board to require De Beerss Diamond Trading
Company (DTC) and other miners to supply local manufacturing operations directly. Currently, De Beers exports
all of its rough diamonds to its DTC office in London and
supplies South African manufacturers from sights with a
mix containing rough diamonds from mines in Africa and
Russia. In addition, the proposed new Mining Charter
would require that 26% of equity in the countrys existing
mines be held by historically disadvantaged people
within 10 years.
DTC managing director Gareth Penny said he agreed
with the basic goals of beneficiation but believed the proposed legislation ultimately would make the mining and
cutting industries smaller and less competitive.
Neighboring governments of Namibia and Botswana are
watching results of South Africas beneficiation laws as a
potential model for their countries.
RS

GEMS & GEMOLOGY

SUMMER 2005

199

The

LAST Page

The

Other Side

of the

Lens

Photo by Karen Myers

of Diamonds. Though they are quick to downplay the accolades they have received, the Van Pelts are honorary life
members of the American Gem Trade Association, as well as
honorary fellows of the Bowers Museum of Cultural Art in
Santa Ana, California.

ith 81 G&G cover shots to their creditand


countingits time we introduced you to Harold
and Erica Van Pelt, the husband-and-wife team
whose gem and jewelry photos have appeared in every issue
of Gems & Gemology since 1981.
Kansas native Harold, known to his friends as Van, studied
at the famed Brooks Institute of Photography in Santa
Barbara, California. Erica was born and raised in Germany,
where she trained as a photographer. The two met in 1962,
at a Los Angeles museum exhibit of Harolds photos, and
have worked together since then.
The Van Pelts originally specialized in the photography of
high-end furniture, but their focus began to shift in the late
1960s, when a private collector commissioned them to photograph a group of tourmalines (still their favorite gem to
shoot). They were entering largely uncharted waters. With the
exception of Lee Boltin, whose work appeared in 1965s
Gems in the Smithsonian Institution, there were no true artists
working in gem and mineral photography. Free to follow their
own vision, the Van Pelts developed innovative compositions
and lighting techniques that revolutionized the way we look
at these treasures.
In addition to G&G, their photographs have been featured in
Lapidary Journal, Mineralogical Record, JCK, and countless
other publications, as well as dozens of books, including
their own popular Birthday Book of Gems and Birthday Book

200

THE LAST PAGE

One vital ingredient to their success is teamwork. The most


interesting part of our relationship is that we both have our
departments, even though were always treading on each
others territory, Erica explains. The layout, the designs for
the photos, and the props are my department; Vans is the
technical lighting and operating the camera. A single shot,
like the yellow diamonds cover for this issue of G&G, can
require hours of meticulous work to set up the stones, the
background, the lights, and the best camera angle.
The Van Pelts have taken that same complementary
approach in becoming world-class lapidaries. Since the
1970s, they have carved objects such as candlesticks, vases,
containers, and eggs from rock crystal quartz and other
materials. They have even designed or modified much of
the machinery they use for rock carving. Their creations
have been exhibited at a number of international gem and
mineral shows, though none have ever been offered for sale
(see J. Sinkankas, Artistry in rock crystal: The Van Pelt collection, Winter 1982 G&G, pp. 214220, for several
notable examples).
In between photo shoots and their travels, Harold is putting
the finishing touches on a full-sized agate chair, eight years
in the making. Erica is involved in setting up a gem exhibit
for a collector that will open in Shanghai this October.
Another ongoing project at their Hollywood home/studio is
cataloguing the thousands of shots they have taken over the
years. Most of the information on these images resides only
in the Van Pelts collective memories. Once they have compiled all the data and had the photos digitally scanned, their
photo library will be a valuable resource for the gem and
mineral world.
In the meantime, each photo still offers a new challenge.
We always start from the beginning, Erica says. We
always strive to build on our last shot, so we have to find a
different approach every time.
Stuart Overlin
Associate Editor

GEMS & GEMOLOGY

SUMMER 2005

You might also like