Short course “Diamond as a messenger from the
Earth’s interior: natural samples and experiment”
Part 2:
What do we learn from diamonds?
Yana Fedortchouk
Department of Earth Sciences, Dalhousie University
Halifax, Canada (yana@dal.ca)
Outline
• What do we know about diamond formation?
• Phenocrysts or xenocrysts?
• Complex growth – resorption patterns in natural diamonds
• Role of oxidation – reduction processes
• Role of carbon saturation in mantle fluids/melts
• Applications
• Carbon isotopes – carbon cycle
• Craton-formation
• “window” into the mantle
• How studies of diamond inclusions help in kimberlite
prospecting and exploration
2
FIG. 1. HRXCT three-dimensional image of the eclogite
xenolith U51, created by stacking the 80 two-dimensional
images. The appearances of various phases are the same as
those in Figure 2.
entations to look for any associations or alignments.
Volume visualization software makes it possible to
view any aspect of the 3-D model from any perspective.
It is possible to render some of the model as
transparent and display only one or two mineral
phases at a time. Then by rotating the model, it is
possible to look for spatial relationships between
different crystals of the same mineral or between
different minerals. These visualizations are difficult
to display here as 2-D figures (Fig. 1), but an animation
of the diamonds, garnets, and clinopyroxenes
of this U51 eclogite xenolith rotating in
space can be viewed at the following address: http://
www.ctlab.geo.utexas.edu/imfoframes/imfoani.html.
A complete 3-D model is more representative of
the sample than are thin sections. Modal analyses of
five thin sections taken from this U51 xenolith
ranged from 25 to 40% garnet. The entire xenolith is
actually 25.9% garnet by volume, so most of the thin
sections are poor representations of the complete
xenolith. This results partly from the coarse grain
size of the xenolith compared to the size of a thin
section. Also, the location of a thin section is commonly
chosen to include some interesting feature,
rather than to accurately represent the entire volume
of the rock.
Chemical and isotopic analytical techniques
Major- and minor-element compositions of minerals
were determined with a Cameca SX-50 electron
microprobe at the University of Tennessee.
Minerals and metals were used as standards. Analytical
conditions employed an accelerating voltage
of 15 keV, a beam current of 20 nA, beam size of 5
µm, and 20 second counting times for all elements,
except K in clinopyroxene and Na in garnet (60 seconds
each). All analyses underwent a full ZAF correction.
Cathodoluminescence (CL) images of
diamonds were collected on the EMP by beam rastering
and translation of the sample stage. Each
frame of the CL images is 384 µm x 384 µm in size.
Concentrations of REE and other trace elements
were obtained using a Secondary Ion Mass Spectrometer
(SIMS, Cameca IMS 3f) at the Tokyo Institute
of Technology. A well-calibrated augite
megacryst from an alkali basalt in Japan and a
quenched glass of JB-1 rock were used as standards.
An energy-filtering technique with an offset voltage
of -40V for REE and -100V for other trace elements
was applied to eliminate possible molecular
interference. The primary ion beam 16
Owas
about
20 µm in diameter. Analytical uncertainties are 10-
20% for REE and 5-10% for other trace elements.
Details of the SIMS technique were presented in
Yurimoto et al. (1989).
Isotopic compositions of carbon and nitrogen,
and the concentrations of nitrogen in micro-areas on
the surfaces of polished diamonds were analyzed
using the SIMS (Cameca IMS 6f) at the Carnegie
Institution of Washington. Carbon isotopes were
measured using an Cs+
ion beam (0.5-2.0 nA) and
collection of negatively charged C-ions at low mass
resolution and high energy offset (+250 ± 100 eV).
Nitrogen isotopes and concentrations were determined
using the 15
N12
C/14
N12
C ratio to measure
δ15
N in nitrogen-bearing diamonds, because nitrogen
itself does not ionize appreciably by sputtering.
The analysis used a Cs+
ion beam (5-40 nA,
depending on N concentration) and collection of
negatively charged CN molecules at high mass resolution
(MRP = 7000-9000). The analytical uncertainties
are ±0.6‰ for δ13
C, ±3‰ for δ15
N, and ±10 %
in nitrogen concentrations. Details about the analytical
techniques were presented by Hauri et al.
(1999).
Downloadedby[DalhousieUniversity]at06:0914June2012
3
Phenocrysts or xenocrysts?
Diamond nucleation and growth by reduction of carbonate melts
under high-pressure and high-temperature conditions
Makoto Arima
Yusuke Kozai
Geological Institute, Yokohama National University 79-7, Tokiwadai, Hodogaya-ku, Yokohama 240-8501, Japan
Minoru Akaishi
Advanced Materials Laboratory, National Institute for Materials Science 1-1, Namiki, Tsukuba, Ibaraki 305-0044, Japan
ABSTRACT
We report for the first time experimental evidence for the nucleation and growth of
diamonds from carbonatitic melts by reduction in reactions with silicon metal or silicon
carbide. Experiments were carried out in the CaMg(CO3)2-Si and CaMg(CO3)2-SiC systems
at 7.7 GPa and temperatures of 1500–1800 8C. No graphite was added to the run
powder as a carbon source; the carbonate-bearing melts supply the carbon for diamond
formation. Diamond grows spontaneously from the carbonatitic melt by reducing reactions:
CaMg(CO3)2 1 2Si 5 CaMgSi2O6 1 2C in the CaMg(CO3)2-Si system, and
CaMg(CO3)2 1 2SiC 5 CaMgSi2O6 1 4C in the CaMg(CO3)2-SiC system. Our results
provide strong experimental support for the view that some natural diamonds crystallized
from carbonatitic melts by metasomatic reducing reactions with mantle solid phases.
Keywords: diamond formation, moissanite, metasomatic reducing reaction, carbonatitic melts.
INTRODUCTION
It has been suggested that some diamonds
in Earth’s mantle crystallized from carbonatitic
melts by reducing reactions and/or
formed from volatile (C-H-O) rich fluids
(Haggerty, 1986; Luth, 1993; Blundy et al.,
1991; Navon, 1999). Studies of inclusions
METHODS
Series of experiments were carried out in
the CaMg(CO3)2-Si and CaMg(CO3)2-SiC
systems at 7.7 GPa and temperatures of 1500–
1800 8C for 60–1440 min (Table 1) using a
modified belt-type high-pressure apparatus
with a 32-mm-diameter bore (Yamaoka et al.,
1992). Methods of pressure calibration and
temperature measurement were described previously
(Akaishi et al., 1990). No direct control
or estimation of oxygen fugacity was
made in this study. The starting materials with
various Si/C ratios (3.0–15.6) were carefully
prepared from highly pure synthetic calcite
on July 8, 2013geology.gsapubs.orgDownloaded from
iamond nucleation and growth by reduction of carbonate melts
under high-pressure and high-temperature conditions
Makoto Arima
Yusuke Kozai
Geological Institute, Yokohama National University 79-7, Tokiwadai, Hodogaya-ku, Yokohama 240-8501, Japan
Minoru Akaishi
Advanced Materials Laboratory, National Institute for Materials Science 1-1, Namiki, Tsukuba, Ibaraki 305-0044, Japan
ABSTRACT
We report for the first time experimental evidence for the nucleation and growth of
diamonds from carbonatitic melts by reduction in reactions with silicon metal or silicon
carbide. Experiments were carried out in the CaMg(CO3)2-Si and CaMg(CO3)2-SiC systems
at 7.7 GPa and temperatures of 1500–1800 8C. No graphite was added to the run
powder as a carbon source; the carbonate-bearing melts supply the carbon for diamond
formation. Diamond grows spontaneously from the carbonatitic melt by reducing reactions:
CaMg(CO3)2 1 2Si 5 CaMgSi2O6 1 2C in the CaMg(CO3)2-Si system, and
CaMg(CO3)2 1 2SiC 5 CaMgSi2O6 1 4C in the CaMg(CO3)2-SiC system. Our results
provide strong experimental support for the view that some natural diamonds crystallized
from carbonatitic melts by metasomatic reducing reactions with mantle solid phases.
Keywords: diamond formation, moissanite, metasomatic reducing reaction, carbonatitic melts.
ODUCTION
as been suggested that some diamonds
rth’s mantle crystallized from carbonmelts
by reducing reactions and/or
METHODS
Series of experiments were carried out in
the CaMg(CO3)2-Si and CaMg(CO3)2-SiC
systems at 7.7 GPa and temperatures of 1500–
1992). Methods of pressure calibration and
temperature measurement were described previously
(Akaishi et al., 1990). No direct control
or estimation of oxygen fugacity was
on July 8, 2013geology.gsapubs.orgDownloaded from
Figure 1. Sample assembly of high-pressure
experiments. 1—pyrophyllite, 2—NaCl, 3—
starting powder in Pt capsule, 4—graphite
heater, 5—steel ring, 6—NaCl with 10 wt%
ZrO2, 7—NaCl with 20 wt% ZrO2.
Figure 2. Scanning electron microscope
photographs of diamond (run MK-17). A:
Newly crystallized octahedral diamonds in
matrix of quench phases. B: Octahedral diamonds
with spinel twin.
and natural magnesite from Victoria, Australia
(Brey et al., 1983), with 9-grade silicon metal
for the CaMg(CO3)2-Si system and with reagent-grade
a-SiC (moissanite) for the
CaMg(CO3)2-SiC system (Table 1).
No graphite was added to the run powder
as a carbon source. The carbonate-Si or carbonate-SiC
mixture was sealed into a Pt capsule
made of a platinum tube 6 mm in diameter,
6 mm long, and 0.2 mm thick. The
capsule was packed in a dry NaCl pressure
medium and placed in the center of a graphite
heater in a high-pressure sample assembly
(Fig. 1). In some runs, a natural octahedral
diamond with an edge length of 0.8–1.0 mm
was placed in the center of the starting powder
in the run capsule. The run products were examined
with optical microscope, scanning
electron microscope, X-ray diffraction, Raman
spectroscopy, and electron microprobe.
Weighing the run capsule before and after
under the pressure and temperature conditions
employed in this study.
RESULTS
The addition of higher amounts of reduction
agents (Si or SiC) to the carbonate mixture, as
well as higher run temperature and longer run
duration, greatly enhanced the formation of
diamond (Table 1). No seed growth or nucleation
was observed in the run with the minimum
Si (run MK-7), but all runs with higher
amounts of Si contain newly crystallized diamonds.
In the CaMg(CO3)2-SiC systems, all
runs at 1600 and 1800 8C contain newly crystallized
diamonds, except for two short time
runs at 1600 8C: 240 min for run MK-15 and
120 min for run MK-1. In run MK-1, we observed
growth on the seed, but no new nucleation.
Longer run duration (1440 min) in the
CaMg(CO3)2-SiC system led to spontaneous
nucleation of diamond, even at 1500 8C (run
MK-18).
To test the effect of the Pt capsule, dummy
experiments were carried out in which carbonate
powder was encapsulated in a Pt capsule
without the reduction agents (Table 1). After
the experiment (run MK-14 with seed diamond
at 1800 8C and 7.7 GPa for 60 min and
run MK-11 without seed at 1800 8C and 7.7
GPa for 50 min), no diamond nucleation or
seed growth was detected in run MK-14 and
there was no diamond nucleation in run MK-
11. The results confirm that Pt metal does not
act as a solvent catalyst in the experiments.
and growth were detected (runs MK-1 and
MK-15).
No significant difference in morphology
was noticed between newly crystallized diamonds
in both the CaMg(CO3)2-Si and
on July 8, 2013geology.gsapubs.orgDownloaded from
Figure 3. Differential interference micrograph
of {111} face of seed diamond. A: Resorbed
dull edge of seed crystal was sharpened
after run by newly grown fibrous
diamonds (run MK-8). B: Triangular growth
hillocks having both positive and negative
orientation to {111} faces of seed (run MK-
5).
CaMg(CO3)2 1 2Si 5 CaMgSi2O6 1 2C in
the CaMg(CO3)2-Si system and CaMg(CO3)2
1 2SiC 5 CaMgSi O 1 4C in the
served that flaky graphite crystals, similar to
those of our study, first crystallized from and
coexisted with supercritical fluids, then transformed
into diamond after a substantial incubation
period. Akaishi et al. (2000) suggested
that the incubation period is probably the induction
time, which determines the diamond
formation in the fluid systems. In this study,
relatively larger concentrations of graphite
crystals occurred in the runs with shorter duration,
in which no diamond nucleation and
growth were detected. The spontaneous diamond
growth occurred after a substantial incubation
period. The metastable graphite
would be a quench phase that crystallized
from excess carbon dissolved in the carbonate
melt at higher temperatures. Alternatively, the
graphite could be a precursor that first crystallized
from the carbonatitic melt, and was
the substrate on which diamond crystallized
after a substantial incubation period via a
mechanism similar to that suggested by Akaishi
et al. (2000).
Graphite microinclusions, located in the genetic
center of Yakutian diamonds, were described
by Bulanova et al. (1998). The graphite
is closely associated with metallic iron,
wu¨stite, and carbonate phases. They suggested
nucleation of the Yakutian diamonds on the
matrix of graphite–metallic iron-wu¨stite in the
presence of carbonate-rich fluids or melts. Our
experimental results support the genetic model
proposed by Bulanova et al. (1998) for the
Yakutian diamonds.
Recent high-pressure melting experiments
for valu
roda for
to H. W
isotope
REFER
Akaishi
tio
hig
tio
p.
Akaishi
Sy
sy
su
p.
Akaishi
ma
mo
hig
tio
p.
Arima,
S.,
am
po
ex
Blundy,
Ca
of
p.
Brey, G
Ha
ro
tle
p.
Bulanov
19
fro
M
Dalton,
me
the
on July 8, 2013geology.gsapubs.orgDownloaded from
Taylor	et	al.	(2000)
4
as the ca. 88 Ma Buffalo Head Hills kimberlites in northern
Alberta, are situated on a Paleoproterozoic accreted terrane
without an apparent Archean basement (e.g., Eccles et al.,
2004).
Diamondiferous lamproites, which are generally in Proterozoic
terranes, span an age range from ca. 1400 Ma, at the
Bobi dyke, Ivory Coast (Bardet, 1974), to 22 to 20 Ma at Ellendale,
in the West-Kimberley province, on the southwestern
margin of the Kimberley Block of Western Australia
(Jaques et al., 1986). The most important lamproitic diamond
deposit, Argyle, on the southeastern margin of the Kimberley
Block, has been dated at 1150 Ma (Pidgeon et al., 1989).
Ultrahigh-pressure diamonds: The type locality for diamond-bearing
ultrahigh-pressure rocks is the Kokchetav massif,
Kazakhstan, which is located near the collisional suture
between a Proterozoic microcontinental nucleus and a Vendian
to Early Cambrian arc system along the southwestern
margin of the Siberian platform (e.g., Sengor et al., 1993).
The ultrahigh-pressure metamorphism took place between
ca. 540 and 530 Ma (Jagoutz et al., 1990), and exhumation of
the ultrahigh-pressure rocks to midcrustal levels was achieved
by ca. 517 to 515 Ma (Troesch and Jagoutz, 1993).
Although traced to metamorphic source rocks in the 1970s
(Rozen et al., 1972), the microdiamonds were not recognized
694 GURNEY ET AL.
0361-0128/98/000/000-00 $6.00 694
TABLE 1. Kimberlite Ages and Diamond Ages from Southern African Diamond Mines
Name of Emplacement P-type Archean P-type Proterozoic
kimberlite age (Ma) Harzburgitic (Ga) E-type (Ga) Iherzolitic (Ga) E-type (Ga) FD References
Premier 1180 ± 30 छ ~2.0 ~2.0 1,2,3
~1.2
Venetia 519 ࡗ ~2.0 ~2.0 3, 4
Jwaneng 235 ± 2 ࡗ ~2.9 ~1.5
✵ 5, 6
Klipspringer 155 छ ~2.6 7
Finsch 118 ± 3 ~3.3–3.2 1.58 ± 0.05 8, 9, 10
Orapa 93.1 छ ~2.9 0.99 ± 0.05
✵ 10, 11
Kimberley pool 95 ~3.3– 3.2 2.89 ± 0.06 8, 12
Koffiefontein 90.4 ࡗ ~2.9 ~1.1 13
Jagersfontein 86 छ ~1.7 14
~1.1
Notes: Filled diamonds = G-10 inclusions common, but not dated; Open diamond = G-10 inclusions present, but P-type diamonds form only minor part
of production; FD = fibrous diamond
References: For kimberlite ages see compilation of Field et al. (2008); Inclusion ages: 1 = Richardson et al. (1993), 2 = Richardson (1986), 3 = Richardson
and Shirey (2008), 4 = Richardson et al. (2009), 5 = Richardson et al. (1999), 6 = Richardson et al. (2004), 7 = Westerlund et al. (2004), 8 = Richardson
et al. (1984), 9 = Smith et al. (1991), 10 = Richardson et al. (1990), 11 = Shirey et al. (2001), 12 = Richardson et al. (2001), 13 Pearson et al. (1998), 14 =
Aulbach et al. (2009)
TABLE 2. Kimberlite Ages and Diamond Ages from Slave Province Kimberlites and Diamond Mines (*)
Name of Emplacement P-type P-type
kimberlite age (Ma) harzburgitic (Ga) Iherzolitic E-type (Ga) FD References
Anuri 613 1
Gahcho Kué 542 ࡗ 2
Snap Lake* 533–535 ࡗ छ ✵ 3
Victoria Island 256–286 2
Jericho 172.3 छ ࡗ 4
Diavik* 55 ~3.5–3.3 2.2–1.8
✵ 5,6
Panda* 53 3.5.± 0.17 छ ✵ 7,8
Inclusion ages: 6 = Aulbach et al. (2008), 8 = Westerlund et al. (2006)
References: Kimberlite ages: 1 = Masun et al. (2004), 2 = Heaman et al. (2003), 3 = Heaman et al. (2004), 4 = Heaman et al. (1997), 5 = Graham et al.
(1999), 7 = Creaser et al. (2004)
From	Gurney	et	al.	(2010)
as the ca. 88 Ma Buffalo Head Hills kimberlites in northern
Alberta, are situated on a Paleoproterozoic accreted terrane
without an apparent Archean basement (e.g., Eccles et al.,
2004).
Diamondiferous lamproites, which are generally in Proterozoic
terranes, span an age range from ca. 1400 Ma, at the
Bobi dyke, Ivory Coast (Bardet, 1974), to 22 to 20 Ma at Ellendale,
in the West-Kimberley province, on the southwestern
margin of the Kimberley Block of Western Australia
(Jaques et al., 1986). The most important lamproitic diamond
deposit, Argyle, on the southeastern margin of the Kimberley
Block, has been dated at 1150 Ma (Pidgeon et al., 1989).
Ultrahigh-pressure diamonds: The type locality for diamond-bearing
ultrahigh-pressure rocks is the Kokchetav massif,
Kazakhstan, which is located near the collisional suture
between a Proterozoic microcontinental nucleus and a Vendian
to Early Cambrian arc system along the southwestern
margin of the Siberian platform (e.g., Sengor et al., 1993).
The ultrahigh-pressure metamorphism took place between
ca. 540 and 530 Ma (Jagoutz et al., 1990), and exhumation of
the ultrahigh-pressure rocks to midcrustal levels was achieved
by ca. 517 to 515 Ma (Troesch and Jagoutz, 1993).
Although traced to metamorphic source rocks in the 1970s
(Rozen et al., 1972), the microdiamonds were not recognized
694 GURNEY ET AL.
0361-0128/98/000/000-00 $6.00 694
TABLE 1. Kimberlite Ages and Diamond Ages from Southern African Diamond Mines
Name of Emplacement P-type Archean P-type Proterozoic
kimberlite age (Ma) Harzburgitic (Ga) E-type (Ga) Iherzolitic (Ga) E-type (Ga) FD References
Premier 1180 ± 30 छ ~2.0 ~2.0 1,2,3
~1.2
Venetia 519 ࡗ ~2.0 ~2.0 3, 4
Jwaneng 235 ± 2 ࡗ ~2.9 ~1.5
✵ 5, 6
Klipspringer 155 छ ~2.6 7
Finsch 118 ± 3 ~3.3–3.2 1.58 ± 0.05 8, 9, 10
Orapa 93.1 छ ~2.9 0.99 ± 0.05
✵ 10, 11
Kimberley pool 95 ~3.3– 3.2 2.89 ± 0.06 8, 12
Koffiefontein 90.4 ࡗ ~2.9 ~1.1 13
Jagersfontein 86 छ ~1.7 14
~1.1
Notes: Filled diamonds = G-10 inclusions common, but not dated; Open diamond = G-10 inclusions present, but P-type diamonds form only minor part
of production; FD = fibrous diamond
References: For kimberlite ages see compilation of Field et al. (2008); Inclusion ages: 1 = Richardson et al. (1993), 2 = Richardson (1986), 3 = Richardson
and Shirey (2008), 4 = Richardson et al. (2009), 5 = Richardson et al. (1999), 6 = Richardson et al. (2004), 7 = Westerlund et al. (2004), 8 = Richardson
et al. (1984), 9 = Smith et al. (1991), 10 = Richardson et al. (1990), 11 = Shirey et al. (2001), 12 = Richardson et al. (2001), 13 Pearson et al. (1998), 14 =
Aulbach et al. (2009)
TABLE 2. Kimberlite Ages and Diamond Ages from Slave Province Kimberlites and Diamond Mines (*)
Name of Emplacement P-type P-type
kimberlite age (Ma) harzburgitic (Ga) Iherzolitic E-type (Ga) FD References
Anuri 613 1
Gahcho Kué 542 ࡗ 2
Snap Lake* 533–535 ࡗ छ ✵ 3
Victoria Island 256–286 2
Jericho 172.3 छ ࡗ 4
Diavik* 55 ~3.5–3.3 2.2–1.8
✵ 5,6
Panda* 53 3.5.± 0.17 छ ✵ 7,8
Inclusion ages: 6 = Aulbach et al. (2008), 8 = Westerlund et al. (2006)
References: Kimberlite ages: 1 = Masun et al. (2004), 2 = Heaman et al. (2003), 3 = Heaman et al. (2004), 4 = Heaman et al. (1997), 5 = Graham et al.
(1999), 7 = Creaser et al. (2004)
DIAMONDS THROUGH TIME 695
TABLE 3. Kimberlite Ages and Diamond Ages from Kimberlites of the Siberian Craton
Name of Emplacement P-type P-type
kimberlite age (Ma) harzburgitic (Ga) E-type (Ga) Iherzolitic (Ga) FD References
Chomur) 436-421
(Upper Olenek
Nakyn 364 ࡗ
Udachnaya 361 ± 6 ~3.5 – 3.1 2.9 ± 0.4 ~2.01 ± 0.06
✵ 1,2, 3
(Daldyn)
Yubileynaya (Alakit) 358 ࡗ ✵
Mir 360 ࡗ ✵(Malo-Botuoba)
23 Party Congress ࡗ(Malo-Botuoba)
Upper Muna 345
Kharamai 235
Kuoika 128–148
References: For kimberlite ages see compilation by Griffin et al. (1999); Inclusion ages: 1 = Pearson et al. (1999), 2 = Pearson et al. (1995), 3 = Richardson
and Harris (1997)
TABLE 4. Lamproite, Kimberlite and Diamond Ages from the Kimberley Block, NW Australia
Name of Emplacement P-type P-type
kimb./lampr age (Ma) harzburgitic Iherzolitic (Ga) E-type (Ga) FD References
Argyle 1178 ± 47 छ छ 1.58 ± 0.06 Ga 1,2
Lamproite
Seppelt 800 छ 3
Kimberlite
Aries Kimberlite 815 छ छ 4
Ellendale 20 छ 1.43 ± 0.13 ࡗ 5,6
Lamproite
Inclusion ages: 2 = Richardson (1986); 6 = Smit et al. (2008)
References: Kimberlite ages: 1 = Pidgeon et al. (1989); 3 = Wyatt et al. (1999); Downes et al. (2006); 5 = Allsopp et al. (1985)
Phenocrysts or xenocrysts?core by their textural characteristics above (again, see figure 9). Recognizing kimberlite in the field is important because diamonds are always so
scarce in kimberlite that they are never visible in outcrop. Starting with the right rock is the first step to finding diamonds.
Figure 10. This diagram shows the ages of continental keels and their relationships with tectonic processes, diamond-hosting magmas, and different diamond types. Note the antiquity
of mantle keels and lithospheric diamonds (parallel to the rhombohedral face (parallel to the rhombohedral face (greater than 1 billion years) and the youth of most kimberlite eruptions
(less than 550 million years). Solid vertical bars depict the duration of ongoing processes, magmas, or diamond-forming events, while solid dots indicate single known occurrences.
Dashed lines connect known ages or indicate when these events might have occurred. Adapted from Gurney et al. (2010), with modifications.
Typically, the host rocks that carry diamonds are younger than the diamonds and the ancient continental cratons they intrude, as shown in figure
10. With only a few exceptions (Argyle, Premier, and Wawa), all known diamond-bearing kimberlites are less than about 550 Ma (million years old)
and most of them less than 300 Ma, with abundant episodes of kimberlite eruption at less than 120 Ma in southern Africa and less than 80 Ma in
North America. Kimberlites are very quickly weathered and eroded rocks, so quickly—years, in fact—that this degradation cannot explain the
preponderance of young kimberlitic volcanism. Instead there are likely some unique changes in mantle volatiles and the relationship of plate
From	Shirey (2013)	and	Gurney	et	al.	(2010)
5
Complex growth – resorption patterns in natural
diamonds Smart	et	al.	(2011)
6
Role of carbon saturation in mantle fluids/melts
Diamond may form in Earth's mantle by a variety of processes
(Stachel and Luth, 2015):
• recrystallization of the low-pressure graphite polymorph,
• Precipitation from a fluid or melt saturated with carbon,
• by oxidation–reduction reactions involving carbonate or methane.
Graphite
Carbonate
Pressure(GPa)
Temperature,C
o
C-O fluid
Diam
ΔlogfO2
2
4
6
8
10
1300
1340
1380
1420
-2.5 -2.1 -1.7 -1.3 -0.9 -0.5
Frost & Wood, 1997
the Fe and Ni contained in the carbide to leave a
contains all iron as Fe21
and Fe31
in silicates and
ll carbon as diamond. Owing to its low viscosity
operties27,28
, any excess carbonatite not consumed
would percolate upwards along grain boundaries
(Fe,Ni)-metal and carbide until complete redox
mmobilization due to reduction of CO2 to C0
—is
umably very efficient process will eventually exmetal
and carbide through precipitation of dian
a metal-free mantle domain where diamond
earing garnet or perovskite (Fig. 3).
of such domains, where the supply of carbonatite
the redox capacity of Fe,Ni-metal, an iron carbide
orm. The redox capacity of Fe31
in such mantle
xactly equivalent or slightly superior (due to the
disproportionation of Fe21
) to that necessary to
nd to CO2. Similarly, the maximum increase in C
domains metasomatized by carbonatites derived
g lithosphere is restricted to ,1,000 p.p.m., equint
of carbonatite that may be immobilized by the
content of 1 wt% expected in the lower mantle5
.
pristine mantle itself does not contain sufficient C to form diamonds25
,
and thus diamonds are expected to form dominantly through the
above redox freezing process wherever carbonatites percolate. The
inverse process—that is, redox melting which destroys most of these
diamonds through oxidation—then leads to the generation of melts
that carry remnants of such deep mantle domains to the Earth’s surface.
Average mantle is expected to contain ,1 wt% Fe0
formed from
Fe21
disproportionation in the lower mantle5
; such metal fractions are
stable and would not segregate to Earth’s core30
. For reaction (1), the
redox capacity of 1 wt% Fe0
is equivalent to 0.8 wt% magnesite. As
argued above, the properties of carbonatite melt result in a selfregulating
mechanism, where infiltrating carbonatite melt oxidizes
all Fe0
, leaving behind a diamond-bearing mantle domain with exactly
the same redox-capacity (that is, that of 0.8 wt% magnesite). In the
reverse process, it can thus be expected that about 1 wt% of carbonatite
melt forms in such upwelling mantle domains. Although upwelling
occurs at speeds comparable to plate tectonic movements (that is,
1–10 cm yr21
), 1% low viscosity melt in the mantle matrix rises with
speeds of at least 10–100 m yr21
(ref. 31). Consequently, carbonatite
flow will tend to escape from the upwelling mantle matrix but will
suffer redox freezing as long as progressing carbonatite melts encounter
~250 km
~410 km
Depth
~660 km
Carbonatite escape
Metal saturation
Abyssal
peridotite
Fe3+/Fe2+
control
Fe0–FeO
control
Buffer capacity
IW
Redox
freezing
Redox
melting
Δlog fO2
–1.5 0 +3 1 0.5 0.10
Rel. to IW Metal (wt%)
c redox freezing and redox melting caused by redox
rth’s mantle. Main panel, cartoon illustrating a possible
zing and redox melting events driven by oxidation state
ducted lithosphere and ambient asthenospheric mantle.
e fO2
(red line) and redox buffer capacity (blue line) as
decreases in the subcratonic upper mantle13
to reach
ion at ,250km depth6,14
. Change in redox control from
0
plus ,3,200 p.p.m. Fe0
calculated from mass balance (mantle: 8 wt% FeOtotal
;
40 vol% majoritic garnet: 5.5wt% FeOtotal
, Fe31
/SFe 5 0.33 (ref. 6); 60 vol.%
ringwoodite: 9.7wt% FeOtotal
, Fe31
/SFe 5 0.02 (ref. 14)). Transition to the
lower mantle is accompanied by a gain in buffer capacity because MgSiperovskite
incorporates more Fe31
at metal saturation fO2
than majoritic
garnet, resulting in ,1.0wt% metal5
. Remixing of subducted carbonated
lithosphere into the mantle at these depths leads to redox freezing and
MgCO3 +	2(Fe,Ni)0	 =	3(Fe,Ni,Mg)O	 +	C
Oxidation – reduction processes
Rohrbach and	
Schmidt	 (2011)
without permission.
Luth (2013) Stagno et	al.	(2013)
En	+	Mgn =	Ol +	Dia +	O2
Dol +	Coes =	Di	+ Dia +	O2
8
Carbon saturation in mantle fluids/melts
Stachel and	Luth (2015) Stachel and	Luth (2015)
Harzburgite:
• Xenoliths P-T below solidus à Diamond growth
from fluid
• extremely limited redox buffering capacity of
cratonic peridotites à redox reactions cannot
produce notable diamond growth
What can we learn from diamonds?
• Carbon cycle
• Formation of cratons
• “Window” into the mantle
• Kimberlite exploration
10
Carbon isotopes – carbon cycle
Diamonds&theGeologyofMantleCarbon377
-20-25-30-35-40 -15 -10 -5 0
-20-25-30-35-40 -15 -10 -5 0
-15
Polycrystalline
from kimberlites
(n = 120)
Komatiitic
from Dachine
French Guyana
(n=181)
-20-25-30-35-40 -15 -10 -5 0
-20-25-30-35-40 -15 -10 -5 0
Carbonados
(n=54)
δ13C (‰)
F) Metamorphic diamonds
(n = 120)
Recycled carbon
-20-25-30-35-40 -10 -5 0
Eclogitic
from Jericho, Slave Craton,
Canada (n=42)
-20-25-30-35-40 -15 -10 -5 0
lowest value
Jagersfontain
(South Africa)
Sao Luiz
(Brazil)
Kankan
(Guinea)
highest
value
Transition Zone
(n = 31)
main mantle-range
(fibrous diamonds, mid-ocean ridge basalts
carbonatites and kimberlites)
Peridotitic diamonds
(n = 1357)
Eclogitic diamonds
(n = 997)
Fibrous diamonds
and diamond coats
(n = 127)
Lower mantle diamonds
(n = 78)
δ13C (‰)
of
of
ied
nds
and
ond
in
itic
wn
e or
ds,
ado
wn
ge”
ite-
and
see
From	Shirey et	al.		(2013)
• Worldwide carbon isotopic
composition of diamonds ranges from
−41 to +5‰, close to the range in
sedimentary rocks.
• ~72% of diamonds have carbon
isotopic composition within of −8 to
−2‰ (mean −5‰) = similar to mantlederived
rocks (mid-ocean ridge
basalts, ocean island basalts,
carbonatites, kimberlites).
11
Formation of cratons
Underthrusting of oceanic slabs or upwelling plume magmatism ?
From	Stachel and	Harris	(2008)	
Oceanic Crust
Continental Crust
Lithosphere
Oceanic
LithosphereLithosphere
Astenosphere
Continental Crust
Oceanic Crust
Oceanic
Lithosphere
Eclogites
Peridotites
E-type Dia
P-type Dia
0
P-type
E-type
δC
13
-10-20-30-40
From	Stachel and	Harris	(2008)	and	
Helmstaedt and	Schulze	(1989)
Origin of cratonic lithosphere during
Archean subduction events
Formation of cratons
From	Stachel and	Harris	(2008)
Mantle residence
temperatures for
peridotitic, websteritic
and eclogitic diamonds
based on the nitrogen
thermometer shows that
diamonds forming in very
distinct environments
(peridotite vs. eclogite)
show identical
equilibration
temperatures.
13
When plate tectonics has started?
Insights from diamond ages
From	Stachel and	Harris	(2008)	and	
Helmstaedt and	Schulze	(1989)
Diamonds as a “time-capsule”
of ancient processes
14
examples of the same phenomena. The idea that diamonds can form beneath the younger mobile belts
surrounding the ancient cratonic nuclei opens up new tectonic settings for exploration and ties some diamond
formation to deep continent-scale geologic processes.
Specific Mantle Geologic Settings. From the data gathered on diamond ages within some cratons, patterns of age and mineral inclusion
composition can be linked to broad-scale regional cratonic lithosphere evolution. The way diamonds form in the lithosphere is better understood if
diamond formation pulses can be correlated with thermo-tectonic events for which there is independent evidence. The best examples where this
correlation can be drawn are the Kaapvaal craton of southern Africa (box A) and the Slave craton of Canada. In both cases, suites of diamonds
that form with initial craton stabilization can be distinguished from diamonds produced by later fluids added to the base of the mantle keel by
underthrusting of oceanic slabs or upwelling plume magmatism.
Figure 27. This figure illustrates the use of inclusions in diamonds to understand the onset of a global process. The absence of diamonds with E-type (eclogitic) inclusions in the oldest
diamond populations suggests that something changed around 3,200 Ma to create and preserve diamonds with E-type inclusions. The letters refer to specific localities around the
world (Pa = Panda, M =Murowa, L = Letseng, U = Udachnaya, W = Wellington, Dv = Diavik, K = Kimberley, Kl = Klipspringer, Jw = Jwaneng, V = Venetia, Ja = Jagersfontein, E =
Ellendale, and Ko = Koffiefontein). Closed symbols are from many inclusion isochron ages, and open symbols are single inclusion model ages. Shirey and Richardson (2011)
hypothesized that the change recorded was the onset of the major cycle of plate tectonics known as the Wilson Cycle (WC). The Wilson Cycle is comprised of the opening of an ocean
basin (stages 1–2, thin gray band) and its closing (stages 5–6, thicker gray band), which culminates in continental collision. Global patterns in diamond composition, style of formation,
Shirey and	Shigley (2013)
• Significant difference in age between E- and
P-type diamonds: no eclogitic diamonds (Etype)
older than 3 billion years
• E-type diamonds (3 Ga) capture the first
record of basaltic rock (eclogite is a basalt
at high pressure metamorphism) in the
mantle keel of the continents.
• This indicates ocean basin closure and
continental collision (modern plate tectonics
or Wilson Cycle), because
• Basalt is derived from the ocean floor and is
incorporated into the mantle keel during
collision.
• Mark a transition from a planet dominated
by vertical geodynamic processes (plumes) to
lateral tectonics and subduction.
When plate tectonics has started?
Insights from diamond ages
15
Diamonds & the Geology of Mantle Carbon 381
the case for most of the deep diamonds
from the mantle transition zone such as
at Jagersfontein, South Africa and the
Sao Luis-Juina fields, Brazil (Tappert
et al. 2005a; Palot et al. 2012) and possibly
at Dachine, French Guyana (Cartigny
2010; Smith et al. 2012). But many
eclogitic diamonds worldwide show a
d15
N-distribution similar to most peridotitic
diamonds, pointing to a mantle
origin of their nitrogen and therefore
of their carbon (Fig.  9). About half of
eclogitic diamonds with low-d13
C values
also show negative d15
N-values. These
d13
C-d15
N-N co-variations have been
argued to be inconsistent with simple
mixing of subduction components (see
Cartigny 2005 for review) and the data
thus have been interpreted to reflect
decarbonation reactions occurring in
eclogites. However, the remarkably low-
d15
N compositions of the peridotitic diamonds
from Pipe 50, China (Cartigny et
al. 1997) suggests that variability in the
N isotopic composition of mantle exists,
and might mask the straightforward assignment
of a positive d15
N to recycled
components involved in diamond formation.
With a heterogeneous mantle, the
d13
C-d15
N mixing relationships become
more complex. Evidence for subduction
is clear from many studies: sulfur
isotope of sulfide inclusions, oxygen
isotopes of eclogitic silicate inclusions
and eclogite nodules, and geological
considerations. But subduction seems
to require some decoupling of carbon
from other elements, like N. This view
can be reconciled by metasomatic processes
during diamond formation. But it
remains unclear, if carbon is a massively
cycled element, why the presence of low
d13
C for recycled carbon in eclogitic
diamonds remains rarer than the normal
mantle-like d13
C. Perhaps the amount of
C recycled form the mantle portion of
the slab which should have d13
C near
−5‰ has been underestimated.
10
20
30
40
10
20
30
20
40
60
5
10
10
20
40
60
δ15N (‰)
-24 -18 +18-12 +12-6 +60
Nitrogen in
Metasediments
(n=146) Devolatilisation
Metamorphic
Diamonds
(n=20)
Fibrous/Coated
Diamonds
(n=124)
Peridotitic
Diamonds
(n=218)
Mid Ocean
Ridge Basalts
(n=38)
Eclogitic
Diamonds
(n=152)
Figure 9. The rationale in using the 15
N/14
N-values in
diamonds lies in the distinct distributions displayed by
surface and mantle reservoirs. Nitrogen in metasedimentary
and metamorphic diamonds, metagabbros and
metaophioloites (not shown) being enriched in 15
N
compared to the mantle shown here by fibrous/coated
diamond, mid-ocean ridge basalts (not shown) and peridotitic
diamonds. The similarities of the 15
N/14
N-distributions
of eclogitic and peridotitic diamonds must be
explained if the source of carbon and nitrogen in eclogitic
diamond is to be largely related to subduction (see
text for discussion). From Cartigny (2005).
From	Shirey et	al.		(2013)
From	Smart	et	al.		(2016)
• Witwatersrand diamonds have enriched nitrogen but mantle
carbon isotopic compositions.
• This nitrogen values suggest contamination of the mantle by
nitrogen-rich Archaean sediments.
• Modern-style plate tectonics operated as early as 3.5 billion
years ago
When plate tectonics has started?
Insights from nitrogen and carbon isotopes
Lower mantle inclusions
McCammon, 2001
150 µm
Occurences
Mineral Assemblage
(Mg,Fe)O
MgSiO3
Nitrogen-
free
1. Material exchange
2. Material uplift from the
base of the lower mantle
3. Upper and Lower mantle are
chemically different
“Window” into the mantle
17
Weiss	(2015)
Composition of fluids in diamonds
The infiltration of fluids into continental lithospheric mantle is a
key mechanism for controlling abrupt changes in the chemical and
physical properties of the lithospheric root1,2
, as well as diamond
formation3
, yet the origin and composition of the fluids involved
are still poorly constrained. Such fluids are trapped within diamonds
when they form4–7
and so diamonds provide a unique
means of directly characterizing the fluids that percolate through
the deep continental lithospheric mantle. Here we show a clear
chemical evolutionary trend, identifying saline fluids as parental
to silicic and carbonatitic deep mantle melts, in diamonds from the
Northwest Territories, Canada. Fluid–rock interaction along with
in situ melting cause compositional transitions, as the saline fluids
traverse mixed peridotite–eclogite lithosphere. Moreover, the
chemistry of the parental saline fluids—especially their strontium
isotopic compositions—and the timing of host diamond formation
suggest that a subducting Mesozoic plate under western North
America is the source of the fluids. Our results imply a strong
association between subduction, mantle metasomatism and
fluid-rich diamond formation, emphasizing the importance of
subduction-derived fluids in affecting the composition of the deep
lithospheric mantle.
Ancientsectionsof continental lithospheric mantle (CLM) are characterized
by multi-stage evolution, involvingstrong depletionand melt
removal followed by variable degrees of ephemeral refertilization1,2
.
Refertilization, or enrichment, occurs by mantle metasomatism,
whereby invading fluids or melts transport mobile components
between different mantle reservoirs. This process plays a major part
in shaping the mineralogical and geochemical variation in the CLM, as
well as in determining its long-term stability, rheology and oxidation
state1,8
. While many mantle samples reflect the action of metasomatism,
including mantle xenoliths and mineral inclusions in diamonds,
the nature of the fluids involved can normally only be constrained
Omphacite
OPX
CPX
Olivine
Chromite
(SiO2 + Al2O3)
(MgO + FeO + CaO)
Saline HDF
and peridotitic
microinclusions
Silicic HDF and
eclogitic
microinclusions
Saline
Silicic
Carbonatitic
0.001
0.01
0.1
1
10
100
Cs
Rb
Ba
Th
U
K
Ta
Nb
La
Ce
Pr
Sr
Nd
Sm
Hf
Zr
Eu
Ti
Gd
Tb
Dy
Y
Ho
Er
Tm
Yb
Lu
0.001
0.01
0.1
1
10
Cs
Rb
Ba
Th
U
K
Ta
Nb
La
Ce
Pr
Sr
Nd
Sm
Hf
Zr
Eu
Ti
Gd
Tb
Dy
Y
Ho
Er
Tm
Yb
Lu
PM-normalized
Saline HDFs
Silicic HDFs
(Na2O + K2O)
Range of HDF composition
in fibrous diamonds
0.01
0.1
1
Nd
Sm
Eu
Gd
Tb
Dy
0.001
0.1
10
Nd
Sm
Eu
Gd
Tb
Dy
a c
Diamond
E191
500 μm
Laser ablation pit
b
E111
E151
E153
E191
E231
E141
E152 E142
E11014E154
E217
HDF composition in
eclogitic fibrous diamond
(PAN4) from Panda
HDF composition in peridotitic fibrous
diamond (ON-DVK-294) from Diavik
Figure 1 | Microinclusion compositions in fibrous diamonds from the
Fox kimberlite, Ekati mine. a, Photomicrograph of diamond E191 with the
location of the microinclusions analysed by electron probe micro-analyser
(EPMA). Filled symbols indicate HDFs; open symbols indicate olivine and
orthopyroxene (OPX). b, Composition of HDFs and micro-mineral
inclusions associated with specific Fox diamonds coded by colour. The global
compositional range of HDFs (delineated by average compositions for
individual diamonds) and the wide range of compositions shown by individual
diamonds PAN4 (ref. 5) and ON-DVK-294 (ref. 6) from neighbouring central
Slave kimberlites are also shown. Shaded arrows define the compositional
evolution trajectories of HDFs due to fluid–rock interaction and melting in
carbonated peridotite (taupe arrow) and hydrous eclogite (pink arrow)
lithologies (see also Fig. 2). c, Primitive-mantle (PM) normalized trace element
and chondrite-normalized (CN) REE patterns of saline and silicic HDFs in
fibrous diamond from the Fox kimberlite. Full analyses and additional figures
are in Supplementary Tables 1–4 and Supplementary Fig. 1.
1
Lamont-Doherty Earth Observatory,ColumbiaUniversity,New York,New York10964, USA. 2
Departmentof Earth Sciences,Durham University, Durham,DH1 3LE, UK. 3
Departmentof Earth & Atmospheric
Sciences, University of Alberta, Edmonton, Alberta T6G 2E3, Canada.
2 0 A U G U S T 2 0 1 5 | V O L 5 2 4 | N A T U R E | 3 3 9
The infiltration of fluids into continental lithospheric mantle is a
key mechanism for controlling abrupt changes in the chemical and
physical properties of the lithospheric root1,2
, as well as diamond
formation3
, yet the origin and composition of the fluids involved
are still poorly constrained. Such fluids are trapped within diamonds
when they form4–7
and so diamonds provide a unique
means of directly characterizing the fluids that percolate through
the deep continental lithospheric mantle. Here we show a clear
chemical evolutionary trend, identifying saline fluids as parental
to silicic and carbonatitic deep mantle melts, in diamonds from the
Northwest Territories, Canada. Fluid–rock interaction along with
in situ melting cause compositional transitions, as the saline fluids
traverse mixed peridotite–eclogite lithosphere. Moreover, the
chemistry of the parental saline fluids—especially their strontium
isotopic compositions—and the timing of host diamond formation
suggest that a subducting Mesozoic plate under western North
America is the source of the fluids. Our results imply a strong
association between subduction, mantle metasomatism and
fluid-rich diamond formation, emphasizing the importance of
subduction-derived fluids in affecting the composition of the deep
lithospheric mantle.
Ancientsectionsof continental lithospheric mantle (CLM) are characterized
by multi-stage evolution, involvingstrong depletionand melt
removal followed by variable degrees of ephemeral refertilization1,2
.
Refertilization, or enrichment, occurs by mantle metasomatism,
whereby invading fluids or melts transport mobile components
between different mantle reservoirs. This process plays a major part
in shaping the mineralogical and geochemical variation in the CLM, as
well as in determining its long-term stability, rheology and oxidation
state1,8
. While many mantle samples reflect the action of metasomatism,
including mantle xenoliths and mineral inclusions in diamonds,
the nature of the fluids involved can normally only be constrained
Omphacite
OPX
CPX
Olivine
Chromite
(SiO2 + Al2O3)
(MgO + FeO + CaO)
Saline HDF
and peridotitic
microinclusions
Silicic HDF and
eclogitic
microinclusions
Saline
Silicic
Carbonatitic
0.001
0.01
0.1
1
10
100
Cs
Rb
Ba
Th
U
K
Ta
Nb
La
Ce
Pr
Sr
Nd
Sm
Hf
Zr
Eu
Ti
Gd
Tb
Dy
Y
Ho
Er
Tm
Yb
Lu
0.001
0.01
0.1
1
10
Cs
Rb
Ba
Th
U
K
Ta
Nb
La
Ce
Pr
Sr
Nd
Sm
Hf
Zr
Eu
Ti
Gd
Tb
Dy
Y
Ho
Er
Tm
Yb
Lu
PM-normalized
Saline HDFs
Silicic HDFs
(Na2O + K2O)
Range of HDF composition
in fibrous diamonds
0.01
0.1
1
Nd
Sm
Eu
Gd
Tb
Dy
0.001
0.1
10
Nd
Sm
Eu
Gd
Tb
Dy
a c
Diamond
E191
500 μm
Laser ablation pit
b
E111
E151
E153
E191
E231
E141
E152 E142
E11014E154
E217
HDF composition in
eclogitic fibrous diamond
(PAN4) from Panda
HDF composition in peridotitic fibrous
diamond (ON-DVK-294) from Diavik
Figure 1 | Microinclusion compositions in fibrous diamonds from the
Fox kimberlite, Ekati mine. a, Photomicrograph of diamond E191 with the
location of the microinclusions analysed by electron probe micro-analyser
(EPMA). Filled symbols indicate HDFs; open symbols indicate olivine and
orthopyroxene (OPX). b, Composition of HDFs and micro-mineral
inclusions associated with specific Fox diamonds coded by colour. The global
compositional range of HDFs (delineated by average compositions for
individual diamonds) and the wide range of compositions shown by individual
diamonds PAN4 (ref. 5) and ON-DVK-294 (ref. 6) from neighbouring central
Slave kimberlites are also shown. Shaded arrows define the compositional
evolution trajectories of HDFs due to fluid–rock interaction and melting in
carbonated peridotite (taupe arrow) and hydrous eclogite (pink arrow)
lithologies (see also Fig. 2). c, Primitive-mantle (PM) normalized trace element
and chondrite-normalized (CN) REE patterns of saline and silicic HDFs in
fibrous diamond from the Fox kimberlite. Full analyses and additional figures
are in Supplementary Tables 1–4 and Supplementary Fig. 1.
1
Lamont-Doherty Earth Observatory,Columbia University,New York,New York10964, USA. 2
Departmentof Earth Sciences,Durham University, Durham,DH1 3LE, UK. 3
Departmentof Earth & Atmospheric
Sciences, University of Alberta, Edmonton, Alberta T6G 2E3, Canada.
Weiss	(2015)
Fibrous diamonds from
Ekati Mine, Canada:
• Diamonds can trap
fluids percolating
through lithosphere
• Saline fluids as
parental to silicic and
carbonatitic deep
mantle melts
• The timing of diamond
formation suggest that
a subducting Mesozoic
plate under western
North
• America is the source
of the fluids
18
Kimberlite prospecting and exploration
• Pyrope Garnet
• Chromite
• Ilmenite
• Chrome-diopside
• (Olivine)
• Heavy minerals, resistant
to chemical and
mechanical weathering
Diamond grade of a
kimberlite depends on:
• How much diamond-bearing
peridotite and eclogite are
present
• What the diamond grade of the
source rocks were
• How well the diamonds were
preserved during transportation
to the surface
• Recognition of mineral composition
known to be associated with diamond
• Confirmation that these are derived
from the diamond stability field
(“diamond window”)
• Assessment of the quantity of highinterest
mantle minerals sampled and
preserved by kimberlite
Indicator minerals:
Mineral
composition
Mineral
abundance
Recognition of mineral composition known to be associated with
diamond
Recognition of mineral composition known to be associated with
diamond
Ilmenite
Diamond Inclusions
From	Stachel and	Harris	(2008)	
1. Peridotite association vs. Eclogite association
2. Peridotite xenoliths:
Grt harzburgite > Spl harzburgite > Grt lherzolite
1. About 32-fold preferential association of carbon
for depleted harzburgite over lherzolite
The mantle source :
- Rich in carbon?
- In diamond stability field (“diamond window”)?
Garnets
From Stachel and Harris (2008)
G10		 G9
Garnets: “diamond window”
Temperature
Pressure
Graphite
Diamond
Thermobarometery
- Garnet
- Chrome Diopside
- Enstatite
Ni in Garnet thermometer:
-well-calibrated, reliable
- single grain thermobarometer
Diamond preservation in kimberlite
After Mitchell (1986)
Thank	you!