Content uploaded by Peng Liang
Author content
All content in this area was uploaded by Peng Liang on Jan 07, 2021
Content may be subject to copyright.
Deciphering the enigmatic origin of Guyana’s diamonds
1,111,†2
1Department of Geosciences, Baylor University, One Bear Place, no. 97354, Waco, Texas 76798, U.S.A.
2Department of Geosciences, University of Arkansas, 340 N. Campus Drive, Fayetteville, Arkansas 72701, U.S.A.
Diamonds have long been mined from alluvial terrace deposits within the rainforest of Guyana,
South America. No primary kimberlite deposits have been discovered in Guyana, nor have there been
previous studies on the mineralogy and origin of the diamonds. Paleoproterozoic terranes in Guyana are
prospective to diamond occurrences because the most productive deposits are associated spatially with
the eastern escarpment of the Paleoproterozoic Roraima Supergroup. Geographic proximity suggests
that the diamonds are detrital grains eroding from the <1.98 Ga conglomerates, metamorphosed to
zeolite and greenschist facies. The provenance and paragenesis of the alluvial diamonds are described
are typically small, and those in our collection range from 0.3 to 2.7 mm in diameter; octahedral and
dodecahedral, with lesser cubic and minor macle forms. The diamonds are further subdivided into those
with abraded and non-abraded surfaces. Abraded diamonds show various colors in cathodolumines-
cence, whereas most non-abraded diamonds appear blue. In all populations, diamonds are predominantly
colorless, with lesser brown to yellow and very rare white. Diamonds are predominantly Type IaAB
and preserve moderate nitrogen aggregation and total nitrogen concentrations ranging from trace to
~1971 ppm. The kinetics of nitrogen aggregation indicate mantle-derived residence temperatures of
1124 ± 100 °C, assuming residence times of 1.3 and 2.6 Ga for abraded and non-abraded diamonds,
respectively. The diamonds are largely sourced from the peridotitic to eclogitic lithospheric upper
13C values of –5.82 ± 2.45‰ (VPDB-LSVEC) and inclusion suites predominantly
comprised of forsterite, enstatite, Cr-pyrope, chromite, rutile, clinopyroxene, coesite, and almandine
garnet. Detrital, accessory minerals are non-kimberlitic. Detrital zircon geochronology indicates
diamondiferous deposits are predominantly sourced from Paleoproterozoic rocks of 2079 ± 88 Ma.
Keywords: Diamond, Guyana, Guiana shield, Roraima supergroup
Diamonds have been mined for the past century from alluvial
gravels along the rivers and creeks deep within Guyana’s Amazon
rainforest. The diamonds are found as placers in paleo-to-modern
channels, terraces, and streambeds. Guyana’s diamonds are found
in headless placers, with the most productive gravels associated
spatially with the eastern escarpment of the Roraima Supergroup,
which suggests that the diamonds originate from this detrital
source (Fig. 1). The humid tropical climate and ancient weath-
ering profile of the Guiana Shield makes addressing diamond
provenance a difficult task. Exploration has been driven by
artisanal miners who prospect using detrital indicator minerals.
No primary kimberlites are known in the region, and there is a
lack of prior research on the nature and origin of diamonds in
Guyana (e.g., Gibbs and Barron 1993; Shields and Letendre
1999; Persaud 2010).
Several hypotheses exist on the origin of Guyana’s diamonds.
Diamonds may be sourced from yet undiscovered primary
kimberlites or igneous intrusions (Shields and Letendre 1999;
Persaud 2010). The most likely location would be associated
spatially with the highly magnesian ultramafic intrusives of the
1.7 ± 0.2 Ga, Badidku suite (Olszewski et al. 1977). These rocks
intruded during a tectonically quiet period of the late Trans-
Amazonian orogeny (Gibbs and Barron 1999) and might be
associated with other ultramafic intrusions such as kimberlites.
The only confirmed examples of primary diamond deposits in the
Guiana Shield are in Guaniamo, Venezuela, and Dachine, French
Guiana (Fig. 1), where diamonds are found in Neoproterozoic
kimberlite sills and metamorphosed Paleoproterozoic ultra-
mafic and pyroclastic shoshonites or lamprophyres, respectively
(Capdevila et al. 1999; Magee and Taylor 1999; Kaminsky et
al. 2000, 2004; Channer et al. 2001; Wyman et al. 2008; Smith
et al. 2016). Kimberlites exist in Roraima, Brazil, but these are
not diamondiferous (Svisero et al. 2017; Cabral et al. 2017).
Known diamondiferous kimberlites from West Central Africa
are an additional potential source for Guyana’s alluvial deposits
(Reid 1974; Briceno 1984). These regions were immediately
adjacent prior to the Jurassic rifting and opening of the Atlantic
Ocean; separated by 200 to 400 km (Fig. 1). If diamonds were
derived from Africa then they would be more prevalent in the
northeastern Guiana Shield where diamonds are not found. Fur-
thermore, West African diamonds have been traced to the host
American Mineralogist, Volume 106, pages 54–68, 2021
0003-004X/21/0001–054$05.00/DOI: https://doi.org/10.2138/am-2020-7486 54
* E-mail: Roy_Bassoo1@Baylor.edu
† ORCID 0000-0002-2080-7915
BASSOO ET AL.: DECIPHERING THE ENIGMATIC ORIGIN OF GUYANA’S DIAMONDS 55
American Mineralogist, vol. 106, 2021
diamondiferous kimberlites, which range in age from 92–846
Ma (Bardet and Vachette 1966; Andrews-Jones 1968; Hall 1972;
Janse 1996; Skinner et al. 2004). More locally, the Avanavero
mafic dikes and sills (<1.79 Ga) have been suggested to be dia-
mondiferous, but these dikes are tholeiitic norites and gabbros
and thus unlikely to be diamond-bearing (Da Silva Rodrigues
1991; Reis et al. 2000; Heesterman et al. 2005). Finally, the dia-
monds may be derived as detrital grains weathering directly from
diamondiferous conglomerates of the Paleoproterozoic Roraima
Supergroup (e.g., Gibbs and Barron 1993; Meyer and McCal-
lum 1993). This may be plausible because alluvial diamonds are
sometimes found in gravels with abundant, rounded quartz, red
jasper, and quartzite clasts; also observed in Roraima Supergroup
conglomerates. Placer diamonds in Mutum, Brazil, are mined
from alluvial terraces shed possibly from conglomerates of the
Tepequém Formation, which is equivalent to the Arai Formation
pebbly sandstones and volcaniclastics of the Roraima Supergroup
(Santos et al. 2003; Reis et al. 2017). Similarly, in Suriname,
Geological Map of the Guiana Shield and inferred adjacent
West African craton during the Jurassic, showing primary and placer
diamond deposits. (AM = Amatuk, AR = Arenápolis, BA = Baoule, BV
= Boa Vista, CA = Canastra, CO = Los Coquitos, CR = Coromandel,
DC = Dachine, EK = Ekereku, GU = Guaniamo, JW = Jawalla, KM =
Kamarang, KU = Kurupung, KW = Konawaruk, MG = Mano Gadua,
MK = Maikwak, MM = Monkey Mt., NA = Nassau, RO = Rosebel,
from Santos et al. (2003), Nadeau and Heesterman (2010), Daoust et
al. (2011), Kroonenberg et al. (2016), and Bassoo and Murphy (2018)].
(Color online.)
BASSOO ET AL.: DECIPHERING THE ENIGMATIC ORIGIN OF GUYANA’S DIAMONDS56
American Mineralogist, vol. 106, 2021
diamonds are found in alluvium shed from metamorphosed
Proterozoic conglomerates or ultramafic volcaniclastics of the
Rosebel group, but their provenance remains unknown (van
Kooten 1954; Schönberger and de Roever 1974; Schönberger
1975; Bosma et al. 1983; Ramlal 2018; Naipal et al. 2019). The
Roraima Supergroup and similar sedimentary sequences may
have been a sink for diamonds erupted by unknown Paleopro-
terozoic kimberlites, and may be detrital evidence for some of
Earth’s earliest kimberlites.
Guyana’s diamonds are a known commodity in the global
diamond trade, referred to as “British diamonds,” stemming from
a colonial past. Economically valuable diamonds in Guyana are
small, most commonly ranging from 0.1 to 0.4 carats, although
larger stones are found. Guyanese diamonds are predominantly
colorless (G–J), yellow (K–M), and brown. Pink and green body
colors are rare. Exploration activities conducted by Golden Star
resources recovered alluvial diamonds and accessory minerals
from Amatuk (Fig. 1) and are described as octahedral with moder-
ate resorption textures. Green and brown spotting and skins are
attributed to radiation damage from accessory zircon, monazite,
uraninite, micas, and K-feldspar (Breeding et al. 2018), when
residing in a placer environment. Surface abrasion and breakage
post-dating resorption along crystal edges were interpreted to be
derived from either a high energy fluvial environment or agitation
during recovery (Shields and Letendre 1999).
We present the morphology, abrasion and dissolution tex-
tures, nitrogen concentration, carbon isotopic composition, and
inclusion mineralogy of Guyana’s alluvial diamonds to learn
about their provenance, host magma characteristics, residence
temperatures, and mantle source beneath the Guiana Shield. We
characterize Guyana’s diamonds and their accessory minerals,
thereby facilitating meaningful comparisons with other diamond
deposits in the Guiana Shield and elsewhere. Diamonds were
obtained from the Ekereku, Jawalla, Konawaruk, Kamarang,
Kurupung, Maikwak, and Monkey Mountain alluvial deposits
in Guyana (Fig. 1). Most diamonds crystallized from peridotitic
lithosphere, although there is a lesser population of eclogitic
diamonds. These diamonds can be divided into two basic groups;
those with and without sedimentological abrasive textures. Im-
portantly, non-abraded and abraded diamonds have statistically
different cathodoluminescence, ultraviolet blue light lumines-
cence, and Fourier transform infrared (FTIR) responses. Abraded
diamonds are evidence for a detrital, polycyclic source weather-
ing from the Roraima Supergroup. Non-abraded diamonds may
be derived from the Roraima Supergroup but have experienced
less transport and recycling.
Diamonds occur across the Guiana Shield and lesser por-
tions of the Guaporé Shield within the Amazonian craton of
South America (Fig. 1). The Guiana Shield is exposed for
over 900 000 km2 along the northern margin of the Amazonian
Craton. Chemical weathering prevails throughout much of the
Guiana Shield, resulting in thick sequences of saprolite capped
by “tor” formations of large boulders of resistive and residual
host rock, with the saprolite regolith stripped away by weathering
(Kroonenberg and Gersie 2019). Although exposures are limited
by accessibility and extreme tropical weathering, there is broad
consensus that the Paleoproterozoic evolution of this craton was
dominated by episodes of accretionary orogenic events around
an Archean core complex during the Main Trans-Amazonian
(2.26–2.08 Ga) to Late Trans-Amazonian (2.07–1.93 Ga) orog-
enies (Tassinari 1997; Vanderhaeghe et al. 1998; Reis et al. 2000;
Delor et al. 2003; Cordani and Teixeira 2007; Fraga et al. 2009;
Daoust et al. 2011; Kroonenberg et al. 2016).
The Guiana Shield is divided into four major Paleoproterozoic
terranes (Daoust et al. 2011), one Mesoproterozoic terrane, and
two isolated Archean terranes (Norcross 1997; Kroonenberg et
al. 2016). Crustal development in the Guiana Shield created a
series of greenstone belts, associated gneisses, and amphibolites
of the Maroni-Itacaiunas Belt (Tassinari 1997). The greenstone
belts underwent several episodes of deformation, intrusion, and
metamorphism between 2.26 and 2.08 Ga, followed by cooling
between 2.08 and 1.93 Ga. These events are interpreted to reflect
the expression of the Trans-Amazonian Orogeny in the Guiana
Shield (Cordani and de Brito Neves 1982; Gibbs and Barron
1993; Norcross 1997; Daoust et al. 2011). Much of the region
was then unconformably overlain by the Roraima Supergroup,
which is a stratigraphic succession dominated by interbedded
1.98–1.78 Ga sandstones and conglomerates deposited from
rocks eroding from earlier greenstone terranes (Priem et al. 1973;
Santos et al. 2003). Finally, coeval mafic dikes of the Avanavero
Suite intruded the entire sequence at ~1.79 Ga (Reis et al. 2000).
The region has been a stable craton, only modified by ero-
sional processes, throughout much of the late Phanerozoic. The
upper Proterozoic was marked by a period of prolonged uplift
with no evidence of sedimentation (Gibbs and Barron 1993).
Changing climate and repeated uplift cycles since the late Triassic
have resulted in the continuing surface evolution of the Guiana
Shield. Rifting of South America from Africa beginning in the
early Jurassic was associated with accelerated weathering and
fluvial activity (McConnell 1968). The regional drainage patterns
and depositional systems have evolved in response to millions
of years of river capture, rejuvenation, degradation, and aggra-
dation initiated by faulting and rifting since the opening of the
Atlantic Ocean. Today the region preserves a complex network
of high alluvial, terrace, alluvial flat, river bed, buried channel,
and plateau deposits. Repeated cycles of erosion and deposition
have led to complex diamond placer deposits with detrital assem-
blages reflecting variable provenance and timing. The diamonds
are found alongside accessory phases that include quartz, topaz,
jasper, rutile, anatase, zircon, ilmenite, gold, corundum, and tour-
maline, with minor garnet and chromite. Olivine and perovskite
are notably absent. Diamond is the only residual mineral that
certainly eroded from kimberlite rock. Despite the possibility for
primary kimberlite sources and the complex geologic history of
the Guiana Shield, placer deposits are the only diamond sources
that have been discovered in Guyana.
We acquired a collection of diamonds from miners and a local diamond
merchant (Kay’s Diamond Enterprise Ltd.). We first cataloged each stone by
measuring its dimensions and mass. The overall morphology of the diamonds
was then assessed using a Zeiss Scope AX IO petrographic microscope. A focus
was delineating crystal shape, dissolution textures, and abrasion. Some features
were documented using high-resolution scanning electron microscopy energy-
dispersive X-ray (SEM-EDS) spectroscopy at the Baylor University Center for
Microscopy and Imaging.
BASSOO ET AL.: DECIPHERING THE ENIGMATIC ORIGIN OF GUYANA’S DIAMONDS 57
American Mineralogist, vol. 106, 2021
The nitrogen content of 415 diamonds was measured, and the aggregation
state calculated for the majority of samples using a Thermoscientific Nicolet iN10
FTIR spectrometer. Analyses were performed across 675–4000 cm–1 in cooled
resolution of 4 cm–1. Nitrogen concentration was calculated from individual
spectra by applying the Beer-Lambert law and absorption values of nitrogen
bands at 1365, 1284, and 1175 cm–1, using the least-squares fitting approach (e.g.,
Howell et al. 2012).
Effective thickness (x) was calculated by normalizing the absorption coeffi-
–12443/5;
21582026/12 and the average effective thickness was then calculated
13C, in
‰VPBD-LSVEC) were performed at the Baylor University Stable Isotope Laboratory us-
ing a Costech Elemental Combustion System 4010 connected to a Thermo-Electron
Delta V Advantage continuous flow Isotope Ratio Mass Spectrometer (CF-IRMS)
through a Thermo-Conflo IV interface. Fifty-seven crushed and powdered diamond
samples (<0.5 mm) which range from 0.05–0.90 mg (mean = 0.26 ± 0.15 mg) were
loaded into silver capsules, flash combusted at 1000 °C to convert the diamond to
CO2, which was carried to the CF-IRMS by a constant helium gas flow.
Optical cathodoluminescence (CL) of 556 diamonds was observed using a
Nikon-Japan LV UEPI microscope equipped with a low vacuum Reliotron III
that operated at 7.5–9 kV and 0.3–0.5 amps. The CL data was supplemented with
ultraviolet blue light (UV) luminescence (425–495 nm) observed in 558 diamonds
using an Olympus BX51 petrographic microscope equipped with an EXFO X-cite
120 fluorescence illumination system, with an exposure time of approximately
10 s for most samples. Each crystal’s color, or lack thereof, was documented for
both CL and UV observations.
Inclusions within 91 diamonds were identified using Raman spectra collected
with a Thermoscientific DXR Raman microscope equipped with a 532 nm laser
–1 grating.
Inclusion species were identified by matching diagnostic peak positions and
heights of unknown spectra with those in the RRUFF spectral database (Lafuente
et al. 2016).
Accessory minerals were extracted from heavy mineral separates that we col-
lected from buried alluvial and colluvial deposits in the Kurupung region (UTM
WGS 84 Zone 20N, 804678E, 675615N), ~4 km from the base of the Roraima
Supergroup escarpment. Rare diamonds were also recovered from these heavy
mineral separates. Samples were sieved, separated by mass in a shaker table, and
cleaned in a sonic bath. Hand-picked accessory minerals of garnet, chromite, and
ilmenite were mounted in epoxy, polished at Baylor University and analyzed at
the University of Texas at Austin, using a JEOL JXA-8200 Electron Microprobe
(EPMA). Zircon crystals were also hand-picked and analyzed for U-Pb crystal-
lization ages using laser ablation-inductively coupled plasma-mass spectrometry
(LA-ICP-MS) at the University of Arkansas Trace Element and Radiogenic Isotope
Laboratory. For each sample, 150 detrital zircon grains were mounted on a glass
slide using double-sided tape. Grain surfaces were ablated using an ESI 193 nm
Excimer laser ablation system and a Thermo-iCap quadropole ICP-MS. Data
helium flow rate of 0.8 L min–1, and a fluence of ~4.5 J cm–2. Zircon from Pleso-
vice (337.13 Ma; Slama et al. 2008) was used as a primary standard, with R33
(419 Ma; Black et al. 2004) as a secondary standard. The weighted mean average
of R33 analyses was within 1% of the accepted age. Data were reduced in Iolite v.
3.71 (Paton et al. 2011), and individual analyses were manually trimmed to avoid
zones of high discordance and 204Pb-bearing inclusions. Twenty-four analyses
were excluded because they were not zircon or were contaminated by inclusions.
An additional 17 zircon grains (5.6% of the total) were excluded on the basis of
discordance (207Pb/206Pb vs. 206Pb/238U) exceeded 10% or 5% reverse discordance.
In total, 259 concordant zircon U-Pb analyses were retained (Sharman et al. 2018;
Online Material1 Fig. OM1 and Table OM5).
Morphology
Diamonds in our collection range in size from 0.3 to 2.7 mm
in diameter; with a mean of 1.1 ± 0.2 mm. Masses range from
0.01 to 0.95 g. Diamonds are predominantly colorless (93%)
with lesser amounts of brown (4%) to yellow (3%) and very rare
white (1%) variants (Fig. 2). Approximately half of the diamonds
show no radiation spotting (55%). The remainder has green to
green-blue skins or green spotting that covers up to ~100% of
the surface area. A small population (5%) show combined over-
printing of brown and green spotting (Table 1).
Diamonds across all regions are octahedral (30%) and
dodecahedral (30%), with lesser combination (16%), flattened
or elongate (15%), and cubic (3%) forms (Fig. 3) (Table 1).
Those from Ekereku present more dodecahedrons (46%) than
octahedrons (31%). Kamarang has the most flattened or elon-
gated diamonds (26%), with the remainder being octahedral,
Body and skin color. (a) Colorless, (b) white, (c) yellow, (d) brown, (e) brown skins (f) green skins. (Color online.)
BASSOO ET AL.: DECIPHERING THE ENIGMATIC ORIGIN OF GUYANA’S DIAMONDS58
American Mineralogist, vol. 106, 2021
dodecahedral, and various combinations. Twinned diamonds and
aggregates are rare from any location. Fragments account for
9% of diamonds analyzed and resorbed fracture or fragmented
diamond surfaces were not observed.
Fine stepwise, lamellar trigonal faces are common, whereas
flat faces and sharp edges are uncommon (Fig. 3). Resorption
textures, including terraces, teardrop hillocks, and dissolution
pits are common in 95% of diamonds. Most octahedral diamonds
have resorbed edges, and in the case of Ekereku, more diamonds
are fully resorbed to dodecahedrons. Tear drop hillocks are the
most common dissolution texture (Fig. 3f). The most common
pit dissolution textures are flat-bottom (37%) and point-bottom
Table 1. Summary morphology
Color Habit Spotting Late stage etching Sedimentological abrasion
Location n B Br Y Gr P W X O D Co C Fl Unk Gr Br Co X Etching No etching Abraded Non-abraded
Ekereku 226 0 9 6 0 0 0 211 69 103 29 9 11 5 97 9 18 205 90 208 159 53
Jawalla 19 0 1 2 0 0 0 16 6 7 4 1 1 0 8 0 0 11 6 13 19 0
Kamarang 154 0 6 2 0 0 1 145 44 16 35 0 40 19 95 4 9 46 53 101 140 13
Konawaruk 6 0 0 0 0 0 0 6 3 1 0 0 2 0 3 0 1 2 0 5 6 0
Kurupung 81 0 1 4 0 0 1 75 25 17 11 3 18 8 30 2 6 90 28 55 78 5
Maikwak 3 0 0 0 0 0 1 2 1 2 0 0 0 0 3 0 0 0 0 3 3 0
Monkey Mt. 3 0 0 0 0 0 0 3 2 1 0 0 0 0 2 0 0 1 0 3 3 0
% 0.0 3.5 2.8 0.0 0.0 0.6 92.7 30.4 29.8 16.0 2.6 14.6 6.5 37.1 2.3 5.3 55.3 31.3 68.7 85.2 14.8
Notes: B = blue; Br = brown; Y = yellow; P = pink; W = white; X = no color/spotting. O = octahedral; D = dodecahedral; Co = combination; C = cubic; Fl = flattened/
elongate; Unk = unknown.
Morphological features. (a) octahedral, (b) dodecahedral, (cd) trigons, (e) hexagons, (f) hillocks, (g) microdiscs,
(h) fully abraded, (i) edge abrasion, (j) percussive marks. (Color online.)
(12%) trigons and tetragons (19%) (Table 2). Very fine (8%) to
fine (3%) point-bottom trigons are more abundant than coarse to
very coarse point-bottom trigons. Very fine (5%) to fine (14%)
flat-bottom trigons are slightly more abundant than coarse (12%)
to very coarse (1%) flat-bottom trigons. Point bottom hexagons,
and trapezoids are notably rare. Late-stage etching features such
as corrosion sculptures, shallow depressions, ruts, and glossy
surfaces are observed in two-thirds of the diamonds. Less than
5% of diamonds show fine to medium microdisk textures, but
when present, occur in swarms.
Edge abrasion is found in 44% of abraded diamonds and
is the most common sedimentologic-induced surface texture
BASSOO ET AL.: DECIPHERING THE ENIGMATIC ORIGIN OF GUYANA’S DIAMONDS 59
American Mineralogist, vol. 106, 2021
(Fig. 3i). Percussive marks along crystal edges, scratches,
and crescentiform fissures are less common (Fig. 3j). Most
diamonds have some degree of surface abrasion from minor
scratches and edge abrasion, and some have no apparent surface
abrasion. This distinction in surface texture is the basis to sub-
divide Guyana’s diamonds into abraded (85%) and non-abraded
(15%) populations. Abraded diamonds may reflect an older
population with extensive fluvial communition, whereas the
non-abraded diamond population may reveal a shorter distance
and/or time of fluvial transport. Ekereku has the highest relative
proportion of non-abraded diamonds followed by Kamarang
and Kurupung, whereas Jawalla, Konawaruk, Monkey Mt.,
and Maikwak, have no abraded diamonds. Abraded diamonds
average mass (9.3 ± 21.5 mg) is larger than that of non-abraded
diamonds (5.5 ± 4.2 mg). Non-abraded diamonds are more
dodecahedral in form (45% compared to 25%) (Online Mate-
rial1 Table OM1). Resorbed abraded diamonds present more
flat-bottomed dissolution textures (85.3%) than point-bottom
dissolution textures (18.7%). Non-abraded diamonds also
present greater flat-bottomed dissolution textures (56.5%)
than point-bottom dissolution textures (43.5%), but to a lesser
degree (Online Material1 Table OM1).
Cathodoluminescence and UV luminescence
Guyana’s diamonds cathodoluminesce green (35%), blue
(25%), and turquoise (18%) with lesser orange, red, and yellow
(Table 3). Abraded and non-abraded diamonds demonstrate
different responses to CL. Guyana’s abraded diamonds
dominantly show green CL (39%), turquoise (19%), and blue
(19%), with minor orange, red, and yellow (~10%), with 11%
non-luminescent (Fig. 4a). In contrast, non-abraded diamonds
cathodoluminesce blue (59%), minor green (13%), turquoise
(10%), and orange (4%), with 13% showing no CL. No yellow
2 tests)
demonstrate with >99% certainty that the frequencies of CL
response between abraded and non-abraded diamonds are
independent (Table 4).
Ultraviolet luminescence also yields a clear distinction be-
tween abraded and non-abraded diamonds. Abraded diamonds
UV luminesce green (84%) with very minor orange (1%), red
(2%), and yellow (3%) variants, and 10% with no apparent lu-
minescence. In contrast, non-abraded diamonds show a smaller
population of green-luminescent grains (52%), and minor popula-
tions of red- (1%) and yellow-luminescent (1%) grains. A large
proportion of non-abraded diamonds yielded no luminescence
response (45%) (Fig. 4b). We note that similar CL and UV
response distributions are observed for all diamonds regardless
of geographic location.
Nitrogen and carbon isotope composition
The nitrogen concentration of diamonds range between trace
(<5 ppm ~ Type IIa) and 1971 ppm, with a mean of 315 ± 24
ppm and rare outliers from 1175 to 1971 ppm. Diamonds are
predominantly Type IaAB, accounting for 53%, with Type 1aA
at 22%, and Type IaB at 9%. Some of the diamonds are Type
IIa (17%) and thus show no nitrogen aggregation (Fig. 5). No
Type IIb diamonds are observed.
Marked differences in nitrogen concentration exist between
abraded and non-abraded diamonds (Table 4). Non-abraded
diamonds have higher average nitrogen concentrations compared
to abraded diamonds with means of 621 ± 47 and 271 ± 21 ppm,
respectively (Fig. 5; Table 4). The nitrogen concentrations of
non-abraded diamonds preserve a more uniform distribution
with lower skewness and kurtosis coefficients of 0.26 and 1.91,
whereas the abraded diamonds display a pronounced positive
skewness (1.82) and a higher kurtosis (6.62) (Table 4; Online
Material1 Fig. OM1). Multivariate density Finite Mixture
Modeling of nitrogen concentrations was applied to determine
if subpopulations could be readily identified within the greater
abraded and non-abraded diamond populations, based on the
Bayesian Information Criteria which demonstrates that a 21-com-
ponent mixture is the most appropriate choice (Online Material1
Fig. OM2) (e.g., Schwartz 1978; Galbraith and Green 1990;
Liang and Forman 2019; McLachlan et al. 2019). The Finite Mix-
ture Modeling identified 10 distinct subpopulations within the
abraded diamonds and 7 subpopulations within the non-abraded
diamonds (Table 4). The abraded and non-abraded diamonds may
share 6 subpopulations, with nitrogen concentrations of 604 ±
48, 390 ± 48, 180 ± 16, 47 ± 3, 23 ± 1, and 4 ± 0.3 ppm. The
non-abraded population may have one additional group centered
at 968 ± 79 ppm, whereas the abraded diamonds may have ad-
Table 3. Optical spectroscopy
Cathodoluminescence UV fluorescence n
Location B Gr T O R Y X Z Gr O R Y X Z
Ekereku 80 53 32 9 3 1 28 2 154 1 3 1 50 0 209
Jawalla 3 4 7 4 0 0 0 1 18 1 0 0 0 0 19
Kamarang 20 66 26 12 5 1 19 3 127 0 7 5 13 1 152
Konawaruk 1 3 1 0 0 0 1 0 6 0 0 0 0 0 6
Kurupung 10 36 16 4 6 1 3 4 65 2 3 3 7 0 80
Maikwak 1 2 0 0 0 0 0 0 3 0 0 0 0 0 3
Monkey Mt. 2 0 1 0 0 0 0 0 3 0 0 0 0 0 3
% 24.8 34.8 17.6 6.2 3.0 0.6 10.8 2.1 79.5 0.8 2.7 1.9 14.8 0.2
Notes: Cathodoluminescence, n = 471; UV flourescence, n = 473. B = blue; Gr = green; T = turquoise; Y = yellow; O = orange; R = red; X = no response; Z = combination.
Table 2. Pit dissolution textures of Guyana’s diamonds
Type Trigons Tetragons Hexagons Trapezoids
Point Bottom (%)
very fine (<25 µm) 7.7 1.3 0.0 0.0
fine (25–50 µm) 2.6 3.8 0.0 0.0
medium (50–75 µm) 1.3 5.1 1.3 0.0
coarse (75–100 µm) 0.0 6.4 0.0 0.0
very coarse (>100 µm) 0.0 1.3 0.0 0.0
Total % 11.5 17.9 1.3 0.0
Flat Bottom (%)
very fine (<25 µm) 5.1 0.0 5.1 0.0
fine (25–50 µm) 14.1 0.0 9.0 5.1
medium (50–75 µm) 5.1 0.0 1.3 1.3
coarse (75–100 µm) 11.5 3.8 3.8 0.0
very coarse (>100 µm) 1.3 0.0 1.3 1.3
Total % 37.2 3.8 20.5 7.7
Notes: Error ±5 µm; # of diamonds = 370; # of diamonds with dissolution pits = 78.
BASSOO ET AL.: DECIPHERING THE ENIGMATIC ORIGIN OF GUYANA’S DIAMONDS60
American Mineralogist, vol. 106, 2021
ditional populations at 13 ± 1, 84 ± 8, 1234 ± 99, and 1911 ± 150
ppm. Nitrogen concentrations reported here are average and bulk
compositions intended to characterize the diamond populations
as a whole. Heterogeneity within each individual diamond is
expected but not elaborated further.
13C value of –5.82 ±
13C values
are slightly more depleted than diamonds from Arenápolis and
Boa Vista, Brazil, and are significantly more enriched in 13C than
those of Guaniamo, Venezuela (–15 ± 2.78‰), and Dachine,
French Guiana (–24.65 ± 4.12‰) (Fig. 6). Diamonds from West
and Central Africa preserve a mode of –3.5‰, making them
more enriched than Guyana’s diamonds (Cartigny et al. 2014).
13C value of –6.11 ± 2.71‰
13C value of –5.06
13C,
morphology, and inclusion paragenesis of Guyana’s diamonds
please see Online Material1 Tables OM1–OM4.
Inclusions
Diamond-hosted inclusions in order of decreasing abundance
include forsteritic olivine, enstatitic orthopyroxene, rutile, chro-
Cathodoluminescence and UV luminescence optical responses. (Color online.)
BASSOO ET AL.: DECIPHERING THE ENIGMATIC ORIGIN OF GUYANA’S DIAMONDS 61
American Mineralogist, vol. 106, 2021
mite, garnet, clinopyroxene, and coesite (Table 5). Epigenetic
inclusions consist of graphite, which is restricted to internal
fractures within the diamond or along relaxed inclusion margins.
Sulfide inclusions are rare. Most primary inclusions of olivine
and orthopyroxene are cubo-octahedral, whereas clinopyroxene
is prismatic to elongate cubo-octahedral. Forsteritic olivine is
the most common inclusion and found in 6.3% of all diamonds
examined. When found, olivine occurs as single inclusions or in
clusters. Olivine inclusions are colorless and range in size from
inclusions occur in 0.6% of diamonds. Garnets range in size from
Accessory mineral geochemistry
Guyana’s alluvial diamond deposits are frequently found with
non-kimberlite, water-worn quartz, topaz, jasper, rutile, anatase,
zircon, ilmenite, gold, corundum, and tourmaline, with minor
garnet and chromite. Traditional kimberlite indicator minerals
are uncommon, but chromite, garnet, and ilmenite occur in our
samples (Online Material1 Table OM4). Two populations of
Table 4. Spectroscopic comparisons between abraded and non-
abraded diamonds
Spectroscopy Abraded Non-abraded
FTIRa N (ppm) n N (ppm) n
mean 271 ± 21 349 621 ± 47 44
median 109 590
σ 353 473
skewness 1.82 0.26
kurtosis 6.62 1.91
FMM populations
Pop. 1 1911 ± 150 2 – –
Pop. 2 1234 ± 99 16 – –
Pop. 3 – – 968 ± 79 18
Pop. 4 610 ± 52 54 604 ± 48 7
Pop. 5 351 ± 29 60 390 ± 26 4
Pop. 6 193 ± 16 30 180 ± 16 6
Pop. 7 84 ± 8 34 – –
Pop. 8 39 ± 3 46 47 ± 3 1
Pop. 9 21 ± 2 26 23 ± 1 4
Pop. 10 13 ± 1 16 – –
Pop. 11 6 ± 0.4 5 4 ± 0.3 2
Cathodoluminescenceb % n % n
Blue 19.0 76 58.6 41
Green 38.7 155 12.9 9
Turquoise 19.0 76 10.0 7
Orange 7.0 28 1.4 1
Red 2.7 11 4.3 3
Yellow 0.7 3 – –
No response 10.5 42 12.9 9
Combination 2.5 10 – –
UV luminescence % n % n
Blue – – – –
Green 84.3 339 52.1 37
Turquoise – – – –
Orange 1.0 4 – –
Red 3.0 12 1.4 1
Yellow 2.0 8 1.4 1
No Response 9.5 38 45.1 32
Combination 0.2 1 – –
a Abraded vs. non-abraded log N (ppm), |t|-test statistic: 12.7 > |t|-threshold value:
1.3, with 95% compatibility interval and p < 0.00001.
b Abraded vs. non-abraded CL (%), χ2-test statistic: 287 > χ2-threshold value:
14.1, with 7 degrees of freedom, a 90% compatibility interval, and p < 0.00001.
10
100
1000
10000
0 10 20 30 40 50 60 70 80 90 100
Abraded 271 ± 21ppm
Total N (ppm)
NB Aggregation (%)
Non Abraded 621 ± 47ppm
52.7%
16.8% 8.7%
21.9%
Relative
absorbance
Wavenumber (cm-1)
40019003400
Type IaB
Type IaA
Type IIa
Type IaAB
1
3
0
0
°
C
1
2
5
0
°
C
1
1
7
5
°
C
1
1
0
0
°
C
Figure 5
2443
2157 2026 1284
1175
1365
Nitrogen
Diamond
a) N concentration vs. degree of nitrogen aggregation b) N aggregation type proportions
c) Representative FTIR spectrum
0.2
1.2
Nitrogen aggregation of Guyana’s diamonds. (a) Total N (ppm) vs. NB (%) plot of abraded and non-abraded diamonds (Total N ±
24 ppm, %NB ± 1.7%), (b c) representative FTIR spectrum showing carbon and nitrogen in diamond
regions. (Color online.)
Table 5. Guyana diamond inclusion types
Inclusion No. of diamonds Paragenesis
Chromite 11 Peridotitic
Enstatitea 15 Peridotitic
Forsteriteb 42 Peridotitic
Cr-pyrope garnet 3 Peridotitic
Almandine garnet 1 Eclogitic
Coesite 1 Eclogitic
Clinopyroxene 1 Eclogitic
Rutile 15 Eclogitic
a Raman spectra doublet peaks indicate a more Mg-rich and enstatite
composition.
b Raman spectra doublet peak and lower intensity of ν3 mode peak suggest a
more forsteritic composition (Mg# ~ 90).
BASSOO ET AL.: DECIPHERING THE ENIGMATIC ORIGIN OF GUYANA’S DIAMONDS62
American Mineralogist, vol. 106, 2021
mantle-derived and crustal-derived garnets are observed based
on discrimination diagrams of Schulze (2003; Fig. 7). Crustal
garnets are Mn-rich spessartines (Alm27±19Py2±5Sp70±23). The
mantle-derived garnets (Alm3±3Py12±2Sp84±3) are also low Cr and
have moderately low TiO2 (~0.19 wt%). SEM-EDS profiles of
both garnet populations show no obvious compositional zoning
or evidence of metamorphic overprint. Accessory ilmenite grains
are MgO-poor (0.17 ± 0.4 wt%) and TiO2-rich ulvospinels (72.2
± 7.6 wt%) (Wu72±16Ru28±16). We analyzed chromite in our samples
and compared them to previous analyses of accessory chromite
conducted by Goldenstar resources from diamondiferous alluvial
deposits in Amatuk (Shields and Letendre 1999). Most chromites
have moderate to low TiO2 (0.5 ± 0.8 wt%) and Al2O3 (14.2 ± 5.6
wt%) demonstrating magmatic arc (ARC) and mid ocean ridge
basalt (MORB) affinities.
Detrital zircon geochronology
Two representative diamond-bearing heavy mineral separates
from the Kurupung region were targeted for reconnaissance
detrital zircon geochronology. The first was from colluvium in a
raised terrace, ~3 m above the modern river (KU002). The second
was collected from an alluvial gravel horizon within a buried
fluvial channel. Most zircons are well rounded by abrasion. A
minor proportion (~8%) of zircons occur as slightly abraded to
non-abraded euhedral, doubly terminated prisms. Detrital zircon
age distributions of both buried alluvium and terrace colluvium
samples are dominated by zircon between 2.0 and 2.2 Ga in age;
recording similar peak ages of 2.06 and 2.04 Ga, respectively
(Online Material1 Fig. OM2). Only five grains of earliest Pa-
leoproterozoic or Archean in age (i.e., >2.3 Ga) were identified
(~2% of the total), and these grains are generally late Archean
(2.6–2.8 Ga) (Fig. 8). All zircon U-Pb analyses are older than
the depositional age range of the 1.78 ± 0.03 to 1.98 ± 0.08 Ga
Roraima Supergroup (Santos et al. 2003).
Guyana’s alluvial diamond deposits represent an economic
resource that also preserve valuable insights into the evolu-
tion of surface and subsurface processes of the Guiana Shield.
Hypotheses regarding the mantle paragenesis and surface prov-
enance of Guyana’s diamonds exist, but each are based largely
on speculation. Our morphology, composition, and inclusion
data sets provide empirically based reasoning to characterize
the mantle conditions of diamond formation and identify the
possible provenance of placer alluvial deposits.
The Guiana Shield has existed as a thick craton with litho-
13C values (in %VPDB-LSVEC) of Guyana’s diamonds
compared to other South American diamonds [1Tappert et al. (2006);
2Kaminsky et al. (2000); 3Smith et al. (2016)]. (Color online.)
Ca# vs. Mg# plot of mantle and crustal derived accessory
garnets (Schulze 2003). (Color online.)
Comparative detrital zircon U-Pb distributions from alluvial
and colluvial samples at Kurupung. [1Xie et al. 2010; 2De Waele et al.
2015, 3Santos et al. 2003; 4Smith et al. 2016; 5Kroonenberg et al. 2015;
6Skinner et al. 2004; 7Kaminsky et al. 2004; 8Bardet and Vachette 1966,
Andrews-Jones 1968, Hall 1972, and Skinner et al. 2004.] (Color online.)
BASSOO ET AL.: DECIPHERING THE ENIGMATIC ORIGIN OF GUYANA’S DIAMONDS 63
American Mineralogist, vol. 106, 2021
spheric roots extending >130 km depth since at least 1.9 Ga
(Gibbs and Barron 1993; Schulze et al. 2006). Such depths are
enough to intersect the diamond stability field beneath the craton
(Stachel and Harris 2009), depending on the thermal regime
during that time. However, kimberlite deposits in Guyana have
not been discovered, unlike elsewhere in the Guiana Shield
(Kaminsky et al. 2000, 2004; Channer et al. 2001; Svisero et
al. 2017; Cabral et al. 2017). Instead, diamonds are recovered
from Quaternary to modern alluvial deposits. Tectonic and sedi-
mentologic processes may have mixed diamonds from different
temporal and geographic sources into the same alluvial basins.
Thus, Guyana’s diamonds reflect complex surface and mantle
processes across the Guiana Shield, since the Paleoproterozoic.
Paragenesis
Diamonds preserve both compositional and physical evidence
of their parental mantle source. During crystal growth, diamonds
may entrap particular suites of minerals as inclusions. Guyanese
diamonds primarily contain inclusions of forsteritic olivine, en-
statitic orthopyroxene, and chromite. This assemblage indicates
most of the diamonds crystallized in, and were extracted from,
the harzburgite peridotite dominated upper mantle. Less common
diamonds preserve an eclogitic assemblage containing rutile, al-
mandine garnet, clinopyroxene, and coesite inclusions. Cratonic
lithospheric mantle assemblages are typically stable between 1
and 7 GPa and <1300 °C (Mather et al. 2011). Both peridotite
and eclogite assemblages spatially coexist beneath stable cratons,
consistent with model geothermal gradients of 38–42 mWm–2
(Pollack and Chapman 1977; Stachel and Harris 2009).
The carbon isotopic composition of diamonds can be used
to further resolve mantle paragenesis. Guyana’s diamonds yield
13C value of –5.82 ± 2.45‰ (Fig. 6), falling within
the range of –8 to –2‰ found in most P-type diamonds sampled
from the peridotitic upper mantle (Cartigny 2005). Few diamonds
13C values are E-type,
having crystallized from eclogite and contain a possible organic
carbon source derived from metasedimentary rocks (Cartigny
13C values (< –16‰) might be indica-
tive of websteritic paragenesis (Deines and Harris 2004) but
are rare in Guyana’s diamonds. Together, Guyana’s diamond
inclusion suites and carbon isotopic compositions confirm that
the diamonds are largely sourced from the lithospheric mantle
and derived from predominantly peridotite and eclogite rocks.
Accessory mineral compositions can also be used to infer the
paragenesis of Guyana’s diamond deposits. Accessory garnet is
spessartine (Alm27±19Py2±5Sp70±23) in composition and most are
likely derived from granite and granodiorite rocks, which are
common throughout the Guiana Shield. A subpopulation of the
spessartine preserves Mg/Ca ratios indicative of a mantle-derived
affinity (Fig. 7; Mg#/Ca# ~3.80). These mantle spessartines
(Alm3±3Py12±2Sp84±3) have low-Cr and low-TiO2 concentrations
and are likely derived from eclogitic source rocks (Schulze 2003).
Although sourced from the mantle, these garnets are not likely
of kimberlitic origin. Instead, these low-Cr and low-TiO2 spes-
sartine were more likely derived from subducted oceanic crust
rich in Mn. In Guyana there are early Proterozoic manganiferous
metasedimentary rocks (Gibbs and Barron 1993); the largest of
which is at Mathews Ridge and outcrops north of the Roraima
Supergroup (Westerman 1969). These could be remnants of sub-
ducted oceanic crust from which the mantle-affiliated garnets are
derived. Accessory ulvospinels (Wu72±16Ru28±16) are Mg-poor and
are similarly derived from non-kimberlite mafic intrusions. Most
chromites have moderate to low TiO2 and Al2O3 concentrations
indicative of ARC and MORB tectonic environments, which is
typical of the Guiana shield; consisting of volcanic arc segments
accreted over an Archean core (Tassinari 1997; Cordani and
Teixeira 2007; Daoust et al. 2011; Reis et al. 2000; Fraga et al.
2009; Kroonenberg et al. 2016).
Paleo-thermal conditions of the mantle
The concentration and aggregation state of lattice bound
nitrogen in diamond provides additional constraints on the
crystallization temperature and mantle residence conditions of
Guyanese diamonds. Nitrogen is the most abundant impurity
in diamond, with concentrations sometimes reaching >0.3 wt%
(Woods et al. 1990; Cartigny et al. 2001; Stachel and Harris
2009; Smart et al. 2011). Quantitative and FTIR derived nitrogen
concentrations range from trace to 1971 ppm, with population
means from different localities ranging between 120 and 620 ppm
(Fig. 5a). Guyanese diamonds are predominantly Type I, includ-
ing Type IaAB, Type IaA, and Type IaB. When compared to other
South American diamonds (Kamarang, Kurupung, Konawaruk
and Maikwak), nitrogen concentrations are within similar ranges
to diamonds from Boa Vista, Brazil (Table 6). The aggregation
state of nitrogen within diamond is largely a function of residence
temperature (Chrenko et al. 1977; Evans and Qi 1982). Kinetic
modeling of the change in nitrogen aggregation state can be used
to quantify the thermal evolution of the continental lithosphere
and paleotectonic processes (Taylor et al. 1990). Individual total
N ppm vs. %NB (degree of nitrogen aggregation) ratios were
used to calculate mean residence temperatures of 1124 ± 100 °C
(Fig. 5) using the kinetics of the nitrogen A-B center aggregation
reaction (e.g., Taylor et al. 1990, 1996). Residence temperature
distribution is slightly negatively skewed and slightly leptokurtic
(Table 7). We considered a minimum age of eruption of 2.0 Ga
and an assumed residence time of 1.3 Ga, if Guyana’s diamonds
Table 6. Nitrogen concentrations of Guiana Shield diamonds
Location Total N ± σ (ppm) %NB n
Guyana
Ekereku 407 ± 472 38 ± 34 134
Jawalla 123 ± 201 27 ± 34 19
Kamarang 275 ± 310 38 ± 34 150
Kurupung 242 ± 352 36 ± 41 79
Konawaruk 249 ± 265 18 ± 19 6
Maikwak 250 ± 78 34 ± 5 3
Monkey Mt. 610 ± 138 65 ± 16 3
Venezuela
Los Coquitosa 641 ± 324 62 ± 19 77
Guaniamob 620 69 192
French Guiana
Dachinec 20 ± 29 NA 18
Brazil
Boa Vistad 345 ± 412 39 ± 30 34
a Kaminsky et al. (2006).
b Kaminsky et al. (2000).
c Smith et al. (2016).
d Tappert et al. (2006).
BASSOO ET AL.: DECIPHERING THE ENIGMATIC ORIGIN OF GUYANA’S DIAMONDS64
American Mineralogist, vol. 106, 2021
are derived from the Roraima Supergroup (maximum age of peri-
dotitic diamonds ~3.3 Ga; maximum age of Roraima Supergroup
~2.0 Ga). We also considered the possibility that Guyana’s dia-
monds might be derived from more recent kimberlites, of which
the youngest of the Guiana Shield have an eruption age of ~0.7
Ga (Channer et al. 2001) and an assumed residence time of 2.6
Ga (maximum age of peridotitic diamonds 3.3 Ga; maximum
age of Guaniamo kimberlites ~0.7 Ga). Assumptions regarding
residence times ranging from the Paleoproterozoic of 1.3 Ga and
Neoproterozoic of 2.6 Ga vary residence temperature by <3%.
These formation temperatures indicate that both diamond popu-
lations can be derived from similar thermal conditions beneath
the Guiana Shield, despite a wide range in possible emplace-
ment ages. Projecting the average residence temperature along
geothermal gradients based upon heat flows of 38–42 mW/m2
gives minimum crystallization depths between 165 and 185 km,
which is evidence for crystallization in the lithospheric mantle
(Stachel and Harris 2009).
Magma composition
Diamond habit, size, and form reflect crystal development
during growth in the mantle, whereas dissolution and abra-
sion textures reveal resorption during ascent and transport,
respectively. Guyana’s diamonds are predominantly octahedral
(Fig. 3a), indicating slow growth in near-equilibrium conditions
found beneath a thick and stable subcratonic lithosphere in the
diamond stability field (Harrison and Tolansky 1964; Seal 1965;
Sunagawa 1984; Stachel and Harris 2009). Dodecahedrons
(Fig. 3b) represent the second most common crystal habit for
diamonds across Guyana; this form is particularly observed
at Ekereku. Dodecahedrons may form in the mantle or during
magmatic ascent via dissolution of octahedron apices and edges
in response to disequilibria produced by high-temperature oxida-
tion in the presence of CO2 and H2O fluid phases (Fedortchouk
et al. 2007). Diamond populations from other regions in Guyana
do not preserve a distribution so enriched in dodecahedral form,
providing evidence for different conditions of mantle departure
or host magma volatile contents.
Additional aspects of the diamond-entraining magmatic
fluid can be inferred from dissolution features on crystal faces.
Chemical oxidation caused by the interaction of fluid volatile
phases in a kimberlite or related rock exploits surface defects on
Table 7. Formation temperatures of Guiana Shield diamonds
Location T (°C) n Derivation
Guyana
Abraded 1128 ± 104 208 N aggregation (1.3 Ga residence)
Non-abraded 1101 ± 72 34 N aggregation (2.6 Ga residence)
mean 1124 ± 100 242
median 1148 242
skewness –1.49 242
kurtosis 1.72 242
Brazila
Boa Vista 1208 ± 74 27 chromite-olivine inclusion
equilibration
Venezuelab
Guaniamo 1116 ± 99 6 garnet-diopside inclusion
equilibration
a Tappert et al. (2006).
b Kaminsky et al. (2000).
a diamond, causing it to dissolve. Dissolution produces pits and
hillocks, whose geometries are controlled by internal disloca-
tions, crystal defects, and intensive variables, including pressure,
temperature, and fO2 (Fedortchouk 2015). Dissolution features
(Fig. 3f) occur on 95% of the diamonds. Teardrop hillocks are the
most common resorption texture and indicate moderate resorp-
tion (Tappert and Tappert 2011). Many different pit dissolution
textures are recognized but are indistinguishable among different
geographic locations. Flat- and point-bottomed trigons, tetragons,
and hexagon pits are observed. Coarse flat-bottom trigons are
evidence for pre-eruptive mantle/melt dissolution at temperatures
ranging from 1250 to 1300 °C (Fedortchouk 2015). Average
diamond residence temperatures of ~1120 °C and mantle/melt
dissolution temperatures of 1250 to 1300 °C suggest diamonds
encounter variable mantle thermal conditions during residence,
such as higher temperatures of kimberlitic melt during ascent.
The greater abundance of flat-bottom trigons relative to point-
bottom trigons in abraded diamond populations indicates that
the erupting host magma was enriched in H2O relative to CO2
(Fedortchouk 2015). Experimental data confirm that dissolution
pit textures begin as point-bottom features. In a fluid enriched in
H2O relative to CO2, dissolution of the diamond proceeds more
rapidly in the lateral direction rather than the vertical or {111}
direction, creating coarse and flat-bottomed pits (Fedortchouk
2015). Late-stage etching microdisk textures (Fig. 3g) found
on ~5% of the diamonds also provide evidence for an H2O-rich
magma. These features form when the surface of the diamond
is etched by a H2O fluid phase in the ascending host magma
(Fedortchouk et al. 2007). Kimberlite magmas are inferred to
be often CO2-rich and have a high CO2/H2O ratio (Sparks 2013).
The dissolution textures observed in Guyana’s diamonds instead
suggest that both abraded and non-abraded diamond populations
were hosted by a magma with more H2O, and therefore a lower
CO2/H2O ratio. This characteristic is more commonly inferred for
lamproitic melts (Bergman 1987; Mitchell and Bergman 1991),
although H2O-rich kimberlite melts are possible [e.g., Ekati mine
kimberlites, see Fedortchouk et al. (2010)]. Because the non-
abraded diamonds present a greater proportion of point-bottom
dissolution textures than abraded diamonds, these diamonds
may have been entrained in a different host melt with a different
ratio of H2O/CO2.
Diamond is strongly resistant in the natural environment to
abrasion. It is brittle, which permits only microcleavage abra-
sion in the faceting process, and it is extremely hard (Wilks and
Wilks 1972; Oganov et al. 2013). Accordingly, percussive marks,
scratches, and edge abrasion textures require repeated cycles of
erosion, transport, and deposition and/or a lengthy residence
time within surface environments. Using abrasion textures, we
recognized two populations of alluvial diamonds. Pristine, non-
abraded diamonds account for 15% of all diamonds. Ekereku
has the highest relative proportion of non-abraded diamonds.
Jawalla, Konawaruk, Maikwak, and Monkey Mt. have no
abraded diamonds, but we acknowledge that the sample sizes
from those regions are small.
The non-abraded stones are found in the same deposits as
the abraded diamonds. Nitrogen concentrations indicate that
BASSOO ET AL.: DECIPHERING THE ENIGMATIC ORIGIN OF GUYANA’S DIAMONDS 65
American Mineralogist, vol. 106, 2021
non-abraded and abraded groups may have statistically distinct
subpopulations, six of which are common to both groups. These
shared subpopulations suggest that some diamonds are sourced
from similar mantle residence and thermal conditions. The
distinct subpopulations identified in only the abraded or non-
abraded groups may reflect isolated mantle and emplacement
conditions. Statistically distinct spectroscopic responses between
abraded and non-abraded populations also present evidence for
different provenance and source (Table 4). Ultraviolet lumi-
nescence responses in abraded diamonds are overwhelmingly
green and non-responsive, whereas non-abraded diamonds are
split between non-responsive or green (Fig. 4). Luminescence
is caused by the presence of impurities, such as nitrogen, and
related defects which create vacancy centers in diamonds.
Combined substitutional nitrogen and vacancy centers (H3 and
H4 defects) within the diamond produced by irradiation fol-
lowed by high-temperature annealing are commonly ascribed
to green luminescence (Shigley and Breeding 2013). This sug-
gests that Guyana’s abraded diamond population, which has a
greater proportion of luminescent green stones, were subjected
to some form of metamorphic overprint. Abraded diamonds
cathodoluminesce predominantly green, with moderate to minor
blue and other colors. The majority of non-abraded diamonds
however, cathodoluminesce blue. Worldwide, most diamonds
from unmetamorphosed deposits cathodoluminesce blue with
zero phono line emissions at 415–440 nm and 480–490 nm
(Bulanova et al. 1995; Lindblom et al. 2005). This is attributed
to the N3 defect where three nitrogen atoms surround a vacancy.
With increasing grades of metamorphism and mild annealing
at higher temperatures, nitrogen related defect centers become
increasingly more complex and blue CL shifts to other colors,
with emissions between 490–670 nm (Bruce et al. 2011; Kopy-
lova et al. 2011). In diamond, vacancies and interstitials become
more mobile at 500–800 and ~300 °C, respectively (Clark et al.
1992; Collins et al. 2005; Collins and Kiflawi 2009). Enhanced
diffusion and entrapment of vacancies and interstitials at higher
temperatures at the site of various nitrogen forms to produce
new optical centers is likely how blue CL shifts to other colors
(Iakoubovskii and Adriaenssens 1999; Collins and Ly 2002). In
the crust, these temperatures are typically inferred for zeolite
to greenschist facies metamorphic environments. The abraded
population likely experienced some metamorphic overprint.
Comparatively, non-abraded diamonds may have been less
metamorphosed. This can be also influenced by pre-existing
nitrogen and vacancy related defects (N3, NV0, and NVN) in
Type Ia diamonds and manifest as photoluminescence emis-
sions more commonly at 575 nm and less so at 430–450 nm but
rarely correlates with cathodoluminescence (Bruce et al. 2011).
Alternatively, irradiation damage during long placer residence
would create charge vacancies and interstitials in the diamond
lattice, which would also shift CL emittance from blue to other
colors, but this damage is usually more common within the
-
ing et al. 2018). Abraded diamonds are also larger in mass (9.3
± 2.2 mg) than non-abraded diamonds (5.5 ± 4.2 mg), present
fewer dodecahedral forms, and a greater ratio of flat-bottom to
point-bottom dissolution textures (5:1 compared to 1:1). This
is evidence that abraded and non-abraded diamond populations
may have been entrained in different host melts. At present,
both abraded and non-abraded populations are mixed into the
same alluvial deposits, reflecting a complex interplay of mantle
storage, eruptive history, and fluvial transport throughout the
>2.5 Ga evolution of the Guiana Shield.
Provenance
Abrasion, nitrogen concentration, CL, and UV observations
indicate that alluvial deposits from each productive region
across Guyana host both abraded, metamorphosed, nitrogen-
poor diamonds as well as non-abraded, less metamorphosed,
nitrogen-rich diamonds. Guyana’s abraded diamonds are most
likely derived from recycled paleo-placers with a metamorphic
overprint. The probable source is the greenschist facies pebbly
conglomerates or volcaniclastics of the 1.98 to 1.78 Ga Roraima
Supergroup. Diamondiferous alluvial deposits occur in close
proximity to the margin of the Roraima Supergroup. In addition,
diamond-bearing alluvial deposits contain abundant quartzite and
jasper clasts derived from the nearby Roraima Supergroup. The
Roraima Supergroup has been exposed at the surface since at least
the early Mesozoic (McConnell 1968). Increased weathering and
erosion associated with rifting and opening of the Atlantic since
the Jurassic produced initial alluvial deposits derived from the
Roraima Supergroup, which have been continually exhumed and
redeposited since that time by repeated cycles of river capture,
rejuvenation, degradation, and aggradation (McConnell 1968;
Gibbs and Barron 1993). The majority of diamond populations
in Guyana were likely abraded within this complex polycyclic
alluvial environment. Non-abraded diamonds were also likely
sourced from the Roraima Supergroup basal conglomerates and
volcaniclastic horizons but were trapped in sediment packages
that experienced less energetic deposition, reworking, and expe-
rienced less metamorphic overprint and longer primary storage.
The Roraima Supergroup source is supported by comparison
to alluvial diamond deposits in Mutum, Brazil. Mutum diamonds
are mined from alluvial terraces shed from conglomerates of the
Tepequém Formation in Brazil (Santos et al. 2003; Reis et al.
2017). The diamonds are similar to Guyana’s diamonds in size,
13C, and nitrogen concentrations. These diamonds,
however, also present more flattened forms (60%), gray dia-
monds (40%), and fewer proportions of diamonds with trigons
(Araújo et al. 2011). This suggests that these diamonds may be
derived from a different magma source than Guyana’s diamonds.
The Arai Formation of the Roraima Supergroup is the local age
and lithologic equivalent of the Tepequém Formation. Accord-
ingly, conglomerates and volcaniclastics of the Arai Formation
may be the detrital source of Guyana’s diamonds, representing a
regional Paleoproterozoic high energy, siliciclastic depositional
basin that was a sink for diamonds emplaced by several episodes
of kimberlite volcanism within the Guiana Shield, during the
middle Paleoproterozoic or older (>2.0 Ga).
Other hypotheses regarding the source of the diamonds have
been put forward over the past decades, including sources in
West Africa, Boa Vista (Brazil), Dachine (French Guiana), and
Guaniamo (Venezuela). West Africa is home to kimberlite and
alluvial diamond deposits. Most diamonds from West Africa
are alluvial in origin, but diamonds have also been mined from
kimberlites (Deines and Harris 1995; Skinner et al. 2004). The
BASSOO ET AL.: DECIPHERING THE ENIGMATIC ORIGIN OF GUYANA’S DIAMONDS66
American Mineralogist, vol. 106, 2021
alluvial diamonds are largely weathered from Mesozoic kimber-
lite pipes and dikes. Colorless to brown body color and green
skins are common, similar to those in Guyana (Sutherland 1982;
Janse 1996; Stachel et al. 2002). Size and quality are inversely
related to distance from the kimberlite source (Knopf 1970;
Norman et al. 1996; Grantham and Allen 1960; Janse 1996).
The smallest carat sizes and best quality gems are found in beach
placers along the West African coastline. The oldest Mesozoic
sources may have produced microdiamonds that could have
reached Guyana’s eastern coast prior to rifting associated with
the Atlantic Ocean, but diamonds are not found on the coast.
The more recent Mesozoic kimberlites are too young and shed
detrital diamonds into the juvenile Atlantic Ocean instead. We
conclude that Guyana’s diamonds are unlikely to be related to
sources in West Africa.
The closest primary diamond-bearing igneous sources in the
modern tectonic setting are the Guaniamo kimberlite dikes in
Venezuela and metamorphosed ultramafic and pyroclastic sho-
shonites or lamprophyres in Dachine, French Guiana (Capdevila
et al. 1999; Channer et al. 2001; Kaminsky et al. 2000, 2004;
Smith et al. 2016). Although some similarities exist, including
upper mantle temperatures, sizes, and colors between Guyana’s
and Guaniamo’s diamonds, we contend that those sources are
unlikely to be genetically associated with Guyanese diamonds
based on geographic distance, distinct 13C signatures, nitro-
gen aggregation, predominantly eclogitic inclusions, and CL
responses (Channer et al. 2001; Kaminsky et al. 2000, 2004).
Furthermore, Dachine diamonds are smaller (<1 mm), have a
lower total nitrogen range from trace to 110 ppm, are strictly Ib
to IaA, and have mostly sulfide inclusions (Smith et al. 2016).
Detrital zircon geochronology provides additional compel-
ling evidence supporting provenance. Detrital zircon U-Pb ages
from both colluvial and alluvial sediment demonstrate that these
diamondiferous deposits are likely derived from rocks of late
Trans-Amazonian age, with lesser contributions of Archean
grains (Fig. 8). These deposits are inferred to be Pleistocene
to recent in age (McConnell 1968; Gibbs and Barron 1993).
The similar ages between the diamond-bearing colluvium and
alluvium is consistent with the interpretation that both types of
placer diamond deposit are derived from similarly aged bedrock
sources of the Barama-Mazaruni group such as Surumu group
volcaniclastics (Basei and Teixeira 1975; Schobbenhaus et al.
1994; Reis et al. 2017), or alternatively, as inherited minerals
from the once extensive Roraima Supergroup in a foreland basin
fill (Santos et al. 2003). The dominantly unimodal characteristic
of these detrital zircon age spectra is consistent with derivation
from a relatively small geographic region, as a more multi-modal
age distribution would be expected if the diamond-bearing sedi-
ment was sourced from a large fluvial catchment. Furthermore,
the narrow range of ages suggesting a local geographic origin,
preponderance of coarse quartz and jasper lithic fragments, and
presence of non-kimberlitic or related rock type indicator miner-
als, lends evidence to the Roraima Supergroup being the prov-
enance area for most of Guyana’s detrital zircons and diamonds.
In an environment prone to extreme tropical weathering,
Guyana’s diamonds are likely the only remaining mantle xeno-
crysts, which can offer insight into volcanic and tectonic controls
that existed during the early evolution of the Guiana Shield.
Detrital zircon U-Pb ages suggest that the kimberlite or related
rock sources that erupted Guyana’s diamonds represent episodes
of kimberlite volcanism during at least the middle Paleopro-
terozoic or during the Archean. Also, the surfaces of diamonds
demonstrate they were emplaced by various mantle-derived
melts enriched or balanced in H2O relative to CO2. Guyana’s
diamonds are likely xenocrysts from some of the Earth’s oldest
kimberlites or lamproites. These source rocks have long since
eroded or are still undiscovered.
Guyana’s diamonds are divided into two populations of
recycled and abraded diamonds and primary and non-abraded
diamonds. Both abraded and non-abraded diamonds can be
distinguished according to unique spectroscopic, luminescence,
and morphological characteristics. This classification can aid in
characterizing the prospectivity of underexplored diamond fields,
especially when other mantle xenocrysts are absent. The abraded
diamonds are likely recycled detrital grains eroding from rocks
of the Roraima Supergroup, which could therefore represent a
regional placer diamond source terrane with a lateral extent of
at least 450 000 km2. The provenance of non-abraded diamonds
still remains enigmatic. It may be that there are undiscovered
primary kimberlites, but it is more likely that these non-abraded
diamonds may also be sourced from Roraima Supergroup basal
conglomerates and volcaniclastic horizons that have experienced
less metamorphic overprint. Both diamond populations, inclusion
13C, and nitrogen concentrations suggest diamond forma-
tion at residence temperatures of approximately 1120 °C in the
peridotitic and eclogitic lithospheric mantle of the Guiana Shield.
We thank M. Harden of Alicanto Minerals and G. Nestor of the Guyana Geology
and Mines Commission for their logistical support. A. Jagnandan of the Guyana
Gold and Diamond Miners Association. R. Melville and N. Blackman provided
invaluable on field support, for which we are grateful. We also thank J. Krakowsky
(General Manager for Kays Diamond Enterprise Ltd.), for technical advice and
diamond samples, which were crucial to this study. A special thanks is given to
M. Kopylova of the University of British Columbia for her technical input. We
thank B. Schaulis at the University of Arkansas for assistance with LA-ICP-MS
data collection.
Andrews-Jones, D.A. (1968) Petrogenesis and geochemistry of the rocks of the Ken-
ema district, Sierra Leone. Ph.D. thesis, abstract, Leeds University, England.
Araújo, D.P., Santos, R.V., Souza, V., Chemale, F., and Dantas, E. (2011) Diamantes
Serra do Tepequém: resultados preliminares. 12th Simpósio da Amazônia,
2 a 5 de outubro de 2011-Boa Vista-Roraima. Documento da Conferência.
Bardet, M.G., and Vachette, M. (1966) Determination d’ages de kimberlites de
l’ouest African et assai d’interpretation de dataions des diverses venus dia-
mantiferes dans le monde. Bureau de Recherches Géologiques et Miniéres
Report D866A59.
Basei, M.A.S., and Teixeira, W. (1975) Geocronologia do Território de Roraima,
in Anais, Conferência Geológica Intergüianas, 10th, Belém, Brazil: Belém,
Brazil. Departamento Nacional de Pesquisa Mineral, 453–473.
Bassoo, R., and Murphy, B. (2018) The 9 Mile deposit of the Barama-Mazaruni
Greenstone belt of the Guiana Shield: Geochemistry, geochronology and
regional significance. Brazilian Journal of Geology, 48, 671–683.
Bergman, S.C. (1987) Lamproites and other potassium-rich igneous rocks, a re-
view of their occurrence, mineralogy and geochemistry. Geological Society
of London, Special Publications, 30, 103–190.
Black, L.P., Kamo, S.L., Allen, C.M., Davis, D.W., Aleinikoff, J.N., Valley, J.W.,
Mundil, R., Campbell, I.H., Korsch, R.J., Williams, I.S., and Foudoulis, C.
(2004) Improved 206Pb/238U microprobe geochronology by the monitoring of
a trace-element-related matrix effect: SHRIMP, ID–TIMS, ELA–ICP–MS
and oxygen isotope documentation for a series of zircon standards. Chemical
Geology, 205, 115–140.
BASSOO ET AL.: DECIPHERING THE ENIGMATIC ORIGIN OF GUYANA’S DIAMONDS 67
American Mineralogist, vol. 106, 2021
Bosma, W., Kroonenberg, S.B., van Lissa, R.V., Maas, K., and de Roever, E.W.F.
(1983) Igneous and metamorphic complexes of the Guiana Shield in Suriname.
Geologie en Mijnbouw, 62, 241–254.
Breeding, C.M., Eaton-Magaña, S., and Shigley, J.E. (2018) Natural-color green
diamonds: A beautiful conundrum. Gems and Gemology, 54, 2–27.
Briceno, H.O. (1984) Genesis of Venezuelan mineral deposits II. Acta Scientifica
Venezolana 36.
Bruce, L.F., Kopylova, M.G., Longo, M., Ryder, J., and Dobrzhinetskaya, L.F.
(2011) Luminescence of diamonds from metamorphic rocks. American
Mineralogist, 96, 14–22.
Bulanova, G.P. (1995) The formation of diamond. Journal of Geochemical Ex-
ploration, 53, 1–23.
Cabral, N.I., Nannini, F., Silveira, F.V., Cunha, L.M., Bezerra, A.K., and Souza,
W.S. (2017) Mapa das áreas kimberlíticas e diamantíferas dos estados de
Roraima, Pará, Piauí, Rio Grande do Norte, Santa Catarina e Rio Grande do Sul.
Programa Geologia do Brasil (PGB), Ação Avaliação dos Recursos Minerais do
Brasil. Natal, CPRM 2017, 1 mapa colorido, 215 × 90 cm. Escala 1:5 000 000.
Capdevila, R., Arndt, N., Letendre, J., and Sauvage, J.F. (1999) Diamonds in
volcaniclastic komatiite from French Guiana. Nature, 399, 456–458.
Cartigny, P. (2005) Stables isotopes and the origin of diamond. Elements, 1, 79–84.
Cartigny, P., De Corte, K., Vladislav, S., Ader, M., De Parpe, P., Sobolev, N.K., and
Javoy, M. (2001) The origin and formation of metamorphic microdiamonds
from the Kokchetav massif, Kazakhstan: a nitrogen and carbon isotopic study.
Chemical Geology, 176, 265–281.
Cartigny, P., Palot, M., Thomassot, E., and Harris, J.W. (2014) Diamond formation:
A stable isotope perspective. Annual review of Earth and Planetary Sciences,
42, 699–732.
Channer, D.M.D., Egorov, A., and Kaminsky, F. (2001) Geology and structure of
the Guaniamo diamondiferous kimberlite sheets, south-west Venezuela. Revista
Brasileira de Geociências, 31, 615–630.
Chrenko, R.M., Tuft, R.E., and Strong, H.M. (1977) Transformation of the state
of nitrogen in diamond. Nature, 20, 141–144.
Clark, C.D., Collins, A.T., and Woods, G.S. (1992) Absorption and luminescence
spectroscopy. In J.E. Field, Ed., The Properties of Natural and Synthetic
Diamond, 35–69. Academic.
Collins, A., and Kiflawi, I. (2009) The annealing of radiation damage in Type Ia
diamond. Journal of Physics: Condensed Matter, 21, 364209.
Collins, A., and Ly, C-H. (2002) Misidentification of nitrogen-vacancy absorption
in diamond. Journal of Physics: Condensed Matter, 14, L467–L471.
Collins, A., Connor, A., Ly, C-H., and Shareef, A. (2005) High-temperature an-
nealing of optical centers in type-I diamond. Journal of Applied Physics, 97,
083517.
Cordani, U.G., and de Brito Neves, B.B. (1982) The geologic evolution of South
America during the Archaean and early Proterozoic. Revista Brasileira de
Geociências, 12, 78–88.
Cordani, U.G., and Teixeira, W. (2007) Proterozoic accretionary belts in the
Amazonian Craton. Geological Society of America Memoirs, 200, 297–320.
Da Silva Rodrigues, A.F. (1991) Depositos diamantiferos de Roraima. In C. Schob-
benhaus, E.T. Queiroz, and C.E.S. Coelho, Eds., Principais depositos minerais
do Brasil. Rochas e minerais industriais. Gemas e rochas ornamentais: Brasilia,
Departmento Nacional de Producao Mineral (DNPM)/Companhia de Pesquisa
de Recursos Minerais (CPRM), 177–198.
De Waele, B., Lacorde, M., Vergara, and Chan, G. (2015) New insights on
proterozoic tectonics and sedimentation along the peri-Gondwanan West
African margin based on zircon U-Pb SHRIMP geochronology. Precambrian
Research, 259, 156–175.
Daoust, C., Voicu, G., Brisson, H., and Gauthier, M. (2011) Geological setting of
the Paleoproterozoic Rosebel gold district, Guiana Shield, Suriname. Journal
of South American Earth Sciences, 32, 222–245.
Deines, P., and Harris, J.W. (2004) New insights into the occurrence of 13C-depleted
carbon in the mantle from two closely associated kimberlites: Letlhakane and
Orapa, Botswana. Lithos, 77, 125–142.
——— (1995) Sulfide inclusion chemistry and carbon isotopes of African dia-
monds. Geochimica Cosmochimica Atca, 59, 3173–3188.
Delor, C., Lahondère, D., Egal, E., Lafon, J.M., Cocherie, A., Guerrot, C., Rossi,
P., Truffert, C., Théveniaut, H., Phillips, D., and de Avelar, V.G. (2003)
Trans-Amazonian crustal growth and reworking as revealed by the 1:500 000
scale geological map of French Guyana (2nd ed.). Geology of France and
surrounding areas. Special Guyana Shield, BRGM–SGF, 2-3-4, 5–57.
Evans, T., and Qi, Z. (1982) The kinetics of the aggregation of nitrogen atoms in
diamond. Proceedings of the Royal Society A, 381, 169–178.
Fedortchouk. Y (2015) Diamond resorption features as a new method for examin-
ing conditions of kimberlite emplacement. Contributions to Mineralogy and
Petrology, 170, 36.
Fedortchouk, Y., Canil, D., and Semenets, E. (2007) Mechanisms of diamond
oxidation and their bearing on the fluid composition in kimberlite magmas.
American Mineralogist, 92, 1200–1212.
Fedortchouk, Y., Matveev, S., and Carlson, J.A. (2010) H2O and CO2 in kimberlitic
fluid as recorded by diamonds and olivines in several Ekati Diamond Mine
kimberlites, Northwest Territories, Canada. Earth and Planetary Science Let-
ters, 289, 549–559.
Fraga, L.M.B., Macambira, M.J.B., Dall’Agnol R., and Costa, J.B.S. (2009)
1.94–1.93 Ga charnockitic magmatism from the central part of the Guyana
Shield, Roraima, Brazil: single-zircon evaporation data and tectonic implica-
tions. Journal of South American Earth Sciences, 27, 247–257.
Galbraith, R., and Green, P. (1990) Estimating the component ages in a finite
mixture. International Journal of Radiation Applications and Instrumentation,
Part D, Nuclear Tracks and Radiation Measurements, 17, 197–206.
Gibbs, A.K., and Barron, C.N. (1993) The Geology of the Guiana Shield. Oxford
University Press, 246 p.
Grantham, D.R., and Allen, J.B. (1960) Kimberlite in Sierra Leone. Overseas
Geology and Mineral Resources, 8(1), 5–25.
Hall, P.K. (1972) The diamond fields of Sierra Leone. Geological Survey of Sierra
Leone Bulliten, 5(1), 133.
Harrison, E.R., and Tolansky, S. (1964) Growth history of a natural octahedral
diamond. Proceedings of the Royal Society of London, Series A, 279, 490–496.
Heesterman, L., Kemp A.W., Arjune B.K., and Cole, E. (2005) Pashanamu project
geology, structure and geochemistry. Guyana Geology and Mines Commission.
Howell, D., O’Neill, C.J., Grant, K.J., Griffin, W.L., Pearson, N.J., and O’Reilly,
diamond growth. Diamond and Related Materials, 29, 29–36.
Iakoubovskii, K., and Adriaenssens, G.J. (1999) Photoluminescence in CVD dia-
mond films. Physica Status Solidi Applied Research, 172, 123–129.
Janse, A.J.A. (1996) A history of diamond sources in Africa: Part II. Gems and
Gemology.
Kaminsky, F.V., and Khachatryan, G. (2012) Nitrogen in diamond. KM Diamond
Exploration, Ltd. West Vancouver, Canada. Indian Institute of Technology,
Mumbai.
Kaminsky, F.V., Zakharchenko, O.D., Griffin, W.L., Channer, D.M. DeR., and
Khacchatryan-Blinova, G.K. (2000) Diamond from the Guaniamo area,
Venezuela. Canadian Mineralogist, 38, 1347–1370.
Kaminsky, F.V., Sablukov, S.M., Sablukova, L.I. and Channer, D.M.D. (2004)
Neoproterozoic “anomalous” kimberlites of Guaniamo, Venezuela: Mica
kimberlites of “isotopic transitional” type. Lithos, 76, 565–590.
Kaminsky, F.V., Zakharchenko O.D., Khachatryan G.K., Griffin W.L., and Channer,
D.M.D. (2006) Diamond from the Los Coquitos area, Bolivar State Venezuela.
Canadian Mineralogist, 44, 323–340.
Knopf, D. (1970) Les kimberlites et les roches apparentés de Côte d’Ivoire (Kim-
berlite and related rocks of Ivory Coast). Sodemi, Abidjan, pp 202.
Kopylova, M., Afanasiev, V.P., Bruce, L.F., Thurston, P.C., and Ryder, J. (2011)
Metaconglomerate preserves evidence for kimberlite, diamondiferous root and
medium grade terrane of a pre-2.7 Ga Southern Superior protocraton. Earth
and Planetary Science Letters, 312, 231–225.
Kroonenberg, S.B., and Gersie, K. (2019) Quaternary climate change and its impor-
tance for gold exploration in the Guiana Shield. Conference paper. 11th Inter
Guiana Geological Conference, Paramaribo, Suriname, 19–20 February, 2019.
Kroonenberg, S.B., De Roever, E.W.F., Fraga, L.M., Reis, N.J., Faraco, M.T.,
Cordani, U.G., Lafon, J.M., and Wong, Th.E. (2016) Paleoproterozoic evolu-
tion of the Guiana Shield in Suriname—A revised model. Netherlands Journal
of Geosciences, 95, 491–522.
Lafuente, B., Downs, R.T., Yang, H., and Stone, N. (2016) The power of databases:
The RRUFF project. In T. Armbruster and R.M. Danisi, Eds., Highlights in
Mineralogical Crystallography, p. 1–29. W. de Gruyter GmbH.
Liang, P., and Forman, S. (2019) LDAC: Luminescence dose and age calculator (v
1.0) (computer software). Luminescence dating research lab. Baylor University.
https://www.baylor.edu/geosciences/index.php?id=962356
Lindblom, J., Holsa, J., Papunen, H., and Hakkanen, H. (2005) Luminescence study
of defects in synthetic as-grown and HPHT diamonds compared to natural
diamonds. American Mineralogist, 90, 428–440.
Magee, C.W., and Taylor, W.R. (1999) Diamond and chromite geochemical con-
straints on the nature of the Dachine complex, French Guiana, Annual Report.
Australian National University, Canberra, Australia, 85 p.
Mather, K.A., Pearson, D.G., McKenzie, D., Kjarsgaard, B.A., and Priestly, K.
(2011) Constraints on the depth and thermal history of the cratonic lithosphere
from peridotite xenoliths, xenocrysts and seismology. Lithos, 125, 729–742.
McConnell, R.B. (1968) Planation surfaces in Guyana. The Geographical Journal,
134, 506–520.
McLachlan, G.J., Lee, S.X., and Rathnayake, S.I. (2019). Finite mixture models.
Annual Review of Statistics and its Application, 6(1), 355–378.
Meyer, H.O.A., and McCallum, M.E. (1993) Diamonds and their sources in the
Venezuelan portion of the Guyana Shield. Economic Geology, 88, 989–998.
Mitchell, R.H., and Bergman, S.C. (1991) Petrology of Lamproites. Plenum Press.
Nadeau, S., and Heesterman, L. (2010) Guyana Geology and Mines Commission,
Geological Map of Guyana—scale 1:1 million. Guyana Geology and Mines
Commission, Georgetown, Guyana.
Naipal, R., Kroonenberg, S., and Mason, P. (2019) Ultramafic rocks of the
Paleoproterozoic greenstone belt in the Guiana Shield of Suriname, and
their mineral potential. SAXI- XI Inter Guiana Geological Conference 2019:
BASSOO ET AL.: DECIPHERING THE ENIGMATIC ORIGIN OF GUYANA’S DIAMONDS68
American Mineralogist, vol. 106, 2021
Paramaribo, Suriname.
Norcross, C.E. (1997) U-Pb geochronology of the Omai intrusion-hosted Au-quartz
vein deposit and host rocks, Guiana Shield, South America. Master of Science
thesis, Department of Geology, University of Toronto.
Norman, D.I., Ward, J., and McKittrick, S. (1996) Hosts and sources of Ghana
diamonds. SME/AIME Annual Meeting, Phoenix, Arizona, U.S.A.
Oganov, A., Hemley, R.J., Hazen, R.M., and Jones, A.P. (2013) Structure, bonding,
and mineralogy of carbon at extreme conditions. Reviews in Mineralogy and
Geochemistry, 75, 47–77.
Olszewski, W.J. Jr., Gaudette, H., and Mendonza, V. (1977) Rb-Sr geochronology
of the basement rocks, Amazonas Territory, Venezuela: A progress report.
Proceedings, V Congreso Geologico Venezolano, 2, 519–525.
Paton, C., Hellstrom, J., Paul, B., Woodhead, J., and Hergt, J. (2011) Iolite: Free-
ware for the visualisation and processing of mass spectrometric data. Journal
of Analytical Atomic Spectrometry, 26(12), 2508.
Persaud, K. (2010) The diamond industry and exploration for diamonds in Guyana,
Guyana. Geology and Miners Commission.
Pollack, H.N., and Chapman, D.S. (1977) On the regional variation of heat flow,
geotherms and lithospheric thickness. Tectonophysics, 38, 179–196.
Priem H.N.A., Boelrijk N.A.I.M, Hebed E.H., Verdurmen E.A., and Verschure
R.H. (1973) Age of the Precambrian Roraima Formation in northeastern South
America: Evidence from isotopic dating of Roraima pyroclastic volcanic rocks
in Suriname. GSA Bulletin, 84(5), 1677–1684.
Ramlal, S. (2018) An investigation of the Brincks intrusion and its relationship to
the surrounding gold deposits, Brokopondo, Suriname, South America. M.S.
thesis. Anton de Kom University of Suriname, 146–147.
Reid, A.R. (1974) Proposed origin for Guianian diamonds. Geology, 2, 67–68.
Reis, N.J., de Faria, M.S.G., Fraga, L.M., and Haddad, R.C. (2000) Orosirian
calc-alkaline volcanism and the Orocaima event in the northern Amazonian
Craton, eastern Roraima State. Revista Brasileira de Geociências, 30, 380–383.
Reis, N.J., Nadeau, S., Fraga, L.M., Betiollo, L.M., Faraco, M.T.L., Reece, J.
Lachhman, D., and Ault, R. (2017) Stratigraphy of the Roraima Supergroup
along the Brazil-Guyana border in the Guiana Shield, Northern Amazonian
Craton—Results of the Brazil-Guyana Geology and Geodiversity mapping
project. Brazillian Journal of Geology, 47, 43–57.
Santos, J.O.S., Potter, P.E., Reis, N.J., Hartman, L.A., Fletcher, I.R., and
McNaughton, N.J. (2003) Age, source and regional stratigraphy of the Roraima
Supergroup and Roraima-like outliers in northern South America based on
U-Pb geochronology. Geological Society of America Bulletin, 115, 331–348.
Schobbenhaus, C., Hoppe, A., Lork, A., and Baumann, A. (1994) Idade U/Pb do
magmatismo Uatumã no norte do cráton Amazônico, Escudo das Guianas
(Brasil): Primeiros resultados. In Anais, Congresso Brasileiro de Geologia,
37th, Camboriú, Brazil: Porto Alegre, Brazil, Sociedade Brasileira de Geo-
logia, 2, 395–397.
Schönberger, J.M.H. (1975) Diamond exploration between the Suriname and
Saramacca Rivers (NE Suriname). Mededeling Geologisch Mijnbouwkundige
Dienst, 23, 228–238.
Schönberger, H., and de Roever, E.W.F. (1974) Possible origin of diamonds in the
Guiana Shield: Comment. Geology, 2, 475–475.
Schulze, D.J. (2003) A classification scheme for mantle-derived garnets in
kimberlite: A tool for investigating the mantle and exploring for diamonds.
Lithos, 71, 195–213.
Schulze, D.J., Canil, D.M., Channer, D.M.D., and Kaminsky, F.V. (2006) Layered
mantle structure beneath the western Guyana Shield, Venezuela: Evidence
from diamonds and xenocrysts in Guaniamo kimberlites. Geochimica et
cosmochimica Acta, 70, 192–205.
Schwartz, G. (1978) Estimating the dimensions of a model. Annals of Statistics,
6, 461–464.
Seal, M. (1965) Structure in diamonds as revealed by etching. American Miner-
alogist, 50, 105–123.
Sharman, G.R., Sharman, J.P., and Sylvester, Z. (2018) DetritalPy: A Python-based
toolset for visualizing and analysing detrital geo-thermochronologic data. The
Depositional Record, 4(2), 202–215.
Shields, H.N., and Letendre, J. (1999) Final exploration report on Amatuk Pros-
pecting License for the period January 2, 1996 to February 9, 1998. Golden
Star Resources Ltd.
Shigley, J.E., and Breeding, M. (2013) Optical defects in diamond: A quick refer-
ence chart. Gems and Gemology, 49.
Skinner, E.M.W., Apter, D.B., Morelli, C., and Smithson, N.K. (2004) Kimberlites
of the Man Craton, West Africa. Lithos, 76, 233–259.
Slama, J., Košler, J., Condon, D.J., Crowley, J.L., Gerdes, A.E., Hanchar, J.M.,
Horstwood, M.S., Morris, G.A., Nasdala, L., Norberg, N.S., and others. (2008)
Plešovice zircon: A new natural reference material for U-Pb and Hf isotopic
microanalysis. Chemical Geology, 249(1-2), 1–35.
Smart, K., Chacko, T., Stachel, T., Muehlenbachs, K., Stern, R., and Heaman, L.
(2011) Diamond growth from oxidized carbon sources beneath the Northern
13C-N study of eclogite-hosted diamonds from
the Jericho kimberlite. Geochimica et Cosmochimica Acta, 75, 6027–6047.
Smith, B.M., Walter, J.W., Galina, P.B., Mikhail, S., Burnham, A.D., Gobbo, L.,
and Kohn, S.C. (2016) Diamonds from Dachine, French Guiana: A unique
record of early Proterozoic subduction. Lithos, 265, 82–95.
Sparks, R.S.J. (2013) Kimberlite volcanism. Annual Review of Earth and Planetary
Sciences, 41, 497–582.
Stachel, T., and Harris, J.W. (2009) Formation of diamond in the Earth’s mantle.
Journal of Physics: Condensed Matter, 21, 364206 (10pp).
Stachel, T., Harris, J.W., Aulbach, S., and Deines, P. (2002) Kankan diamonds
13C and N characteristics of deep diamonds. Contributions to
Mineralogy and Petrology, 142, 465–475.
Sunagawa, I. (1984) Morphology of natural and synthetic diamond crystals. In
I. Sunagawa, Ed., Materials Science of the Earth’s Interior. Terra Scientific,
Tokyo, 303–330.
Sutherland, D.G. (1982) The transport and sorting of diamonds by fluvial and
marine processes. Economic Geology, 77, 1613–1620.
Svisero, D.P., Shigley, J.E., and Weldon, R. (2017) Brazilian diamonds: A historical
and recent perspective. Gems and Gemology, 53(1).
Tappert, R., and Tappert, M. (2011) Diamonds in Nature: A Guide to Rough
Diamonds. Springer.
Tappert, R., Stachel, T., Harris, J.W., Muehlenbachs, K., and Brey, G.P. (2006)
Placer diamonds from Brazil: Indicators of the composition of the Earth’s
mantle and the distance to their kimberlitic sources. Economic Geology,
101, 453–470.
Tassinari, C.C.G. (1997) The Amazonian Craton. In M.J. De Wit and M.J. Ashwal,
Eds., Greenstone Belts. Clarendon Press, 558–566.
Taylor, W.R., Jaques, A.L., and Ridd, M. (1990) Nitrogen-defect aggregation
characteristics of some Australasian diamonds: Time-temperature constrains
on the source regions of pipe and alluvial diamonds. American Mineralogist,
75, 1290–1310.
Taylor, W.R., Canil, D., and Milledge, H.J. (1996) Kinetics of Ib to IaA nitrogen
aggregation in diamond. Geochimica et Cosmochimica Acta, 60, 4725–4733.
van Kooten, C. (1954) Eerste onderzoek op diamant: Rosebel-Sabanpassie.
Mededeling Geologisch Mijnbouwkundige Dienst, Suriname 11, 63–64 pp.
in diamonds from kimberlite and alluvial sources. Mineralogical Magazine,
39, 349–360.
Vanderhaeghe, O., Ledru, P., Thiéblemont, D., Egal, E., Cocherie, A., Tegyey, M.,
and Milési, J.P. (1998) Contrasting mechanism of crustal growth Geodynamic
evolution of the Paleoproterozoic granite greenstone belts of French Guiana.
Precambrian Research, 92, 165–193.
Westerman, A.R. (1969) The structural analysis of Matthews Ridge manganese
mine, North West District, Guyana—with an appendix on supergene enrich-
ment. Paper presented at the 8th Guiana Geological Conference, August 1969.
Wilks, E.M., and Wilks, J. (1972) The resistance of diamond to abrasion. Journal
of Physics D: Applied Physics, 5, 1902–1919.
Woods, G.S., Van Wyk, J.A., and Collins, A.T. (1990) The nitrogen content of
type Ib synthetic diamond. Philosophical Magazine B, 62, 589–595.
Wyman, D.A., O’Neill, C., and Ayer, J.A. (2008) Evidence for modern-style
subduction to 3.1 Ga; a plateau-adakite-gold (diamond) association. In K.C.
Condie and V. Pease, Eds., When Did Plate Tectonics Begin on Planet Earth?
Geological Society of America, 129–148.
Xie, X., Mann, P., and Escalona, A. (2010) Regional provenance study of Eocene
clastic sedimentary rocks within the South America-Caribbean plate boundary
zone using detrital zircon geochronology. Earth and Planetary Science Letters,
291(1-4), 159–171.
Endnote:
1Deposit item AM-21-17486, Online Material. Deposit items are free to all readers
and found on the MSA website, via the specific issue’s Table of Contents (go to http://
www.minsocam.org/MSA/AmMin/TOC/2021/Jan2021_data/Jan2021_data.html).
