Eur. J. Mineral.
2008, 20, 387–393
Published online April 2008
Special issue on Diamonds
Infrared and Raman spectroscopic observations of Central African
carbonado and implications for its origin
H KAGI* and S FUKURA†
Geochemical Laboratory, Graduate School of Science, The University of Tokyo, Hongo, Tokyo 113-0033, Japan
*Corresponding author, e-mail: [email protected]
Abstract: Carbonado diamonds from the Central African Republic (CAR) were investigated using spectroscopic observations.
Raman spectra for a polished section of carbonado showed the average Raman frequency as 1333.0 cm−1 ; measurements 10 µm
below the sample surface showed a bimodal distribution of the Raman frequency with the average at 1333.5 cm−1 . These contrasting results indicate that the carbonado interior retains considerable residual pressure. The maximum residual pressure detected
in this study from the 10 µm subsurface was 0.49 GPa. From infrared (IR) absorption spectra for crushed carbonado samples,
no absorption attributable to diamond was observed because absorption bands of mineral and fluid inclusions were too strongly
observed. Acid leaching treatment of the crushed grains elicited IR bands assignable to intrinsic diamond vibration and nitrogen
impurity in the diamonds. Nitrogen atoms in the CAR carbonado were not much aggregated and the degree of aggregation was
intermediate between type Ib and type IaA. In the chemically treated samples, IR absorption bands from liquid water and carbonate were ubiquitous suggesting that the CAR carbonado samples contain fluid inclusions and are similar to diamonds containing
mantle-derived fluids. The experimental results of this study suggest that the CAR carbonado originated from rapid heating event
with a presence of fluid in the mantle and subsequent rapid cooling before aggregation of nitrogen impurity in the diamond lattice.
Key-words: Raman spectra, infrared absorption spectra, diamond, residual pressure, fluid inclusions, carbonado.
Introduction
Polycrystalline aggregates of diamond, so-called carbonado, are characterized by a lack of mantle-derived mineral inclusions, a 13 C-depleted isotopic composition, trace
element abundance patterns that are distinct from kimberlite, and radiation effects from radioactive nuclides (Trueb
& Butterman, 1969; Trueb & de Wys, 1969; Trueb &
de Wys, 1971; Ozima et al., 1991; Shibata et al., 1993;
Kamioka et al., 1996; De et al., 1998). Based on these
observations, several assumptions on the genesis of carbonado have been proposed: metamorphism caused by a
large impact on the Earth’s crust during the Precambrian
era (e.g. Smith & Dawson, 1985); transformation of organic sedimentary carbon into diamond in a cold subducted slab (e.g. Robinson, 1978); and radiation-induced
diamond formation from organic carbon (Kaminskii, 1987;
Ozima et al., 1991; Daulton & Ozima, 1996). The origin of carbonado remains controversial. Heaney et al.
(2005) summarized the current understanding of their origin. Recently, a close relationship between photoluminescence (PL) spectra and the carbon isotope compositions of
carbonado has been reported (Kagi et al., 2007). Similar
correlation was also previously reported between cathodo† Passed away on October 10, 2006.
DOI: 10.1127/0935-1221/2008/0020-1817
luminescence (CL) and carbon isotopic distributions, suggesting a multi-step growth process, which results in the
heterogeneity of isotopic and optical signatures (De et al.,
2001).
For considering the carbonado origin, the most important issue is presumably the determination of whether carbonado is formed under static high-pressure and hightemperature conditions in the mantle or not. One means to
trace the depth origin of diamond is to detect the remnant
pressure (residual stress) in mineral inclusions of diamond
and the stress surrounding diamond (Izraeli et al., 1999;
Sobolev et al., 2000). Fukura et al. (2005) attempted to detect the surface stress distribution of carbonado from precise PL mapping with spatial resolution of less than 300 nm
using scanning near-field optical microscopy (SNOM).
They reported no stress on the carbonado surface. However, they detected a considerable upshift of Raman spectra
of carbonado from 5 µm beneath the sample surface, suggesting that the residual stress exists locally inside of carbonado. In their study, the detection of the residual stress
was limited to one sample point. Consequently, a systematic comparison of residual stress between the surface and
subsurface has not been performed. Clarification of the
residual pressure in carbonado will improve our understanding of the origin of carbonado, especially the highpressure aspect of its origin. Furthermore, it is important to
0935-1221/08/0020-1817 $ 3.15
c 2008 E. Schweizerbart’sche Verlagsbuchhandlung, D-70176 Stuttgart
"
388
H. Kagi, S. Fukura
observe intrinsic optical properties of carbonado diamond
using infrared (IR) spectroscopy. IR absorption spectra of
diamond involve information related to the aggregation
state of nitrogen which is the major impurity in diamonds
and which can be a useful thermal indicator. It is also
worthwhile to note that IR spectroscopy is sensitive to
volatile inclusions in diamond. For this study, Raman and
IR spectra of carbonado from the Central African Republic
(CAR) were measured in detail and the origin of carbonado
was reconsidered.
Samples and experimental methods
Sample and chemical preparations
Carbonado samples studied here were mined from alluvial
deposits in the Central African Republic (CAR) and provided by a commercial dealer; the average grain size was
5–10 mm. We studied a suite of samples using spectroscopic methods, chemical analyses, etc., as described in
the Introduction (Kagi et al., 1991, 1994a, 2007; Fukura
et al., 2005). For measuring Raman spectra, one face of
a carbonado grain of approximately 3 × 5 × 1.5 mm was
polished to optical grade. This sample specimen was identical to the sample we used to measure PL spectra with the
SNOM system (Fukura et al., 2005). The sample surface
is smooth, with grain boundaries of diamonds and brown
inclusions indicating the presence of hematite. The inclusions are visible using optical microscopy (see Fig. 1a).
For measuring infrared (IR) absorption spectra, specimens
were crushed to the size of several tens micrometers and
leached in a mixture of HF and HNO3 at 180 ◦ C in a sealed
Teflon! vessel for at least two days. By repeating the chemical treatments, most mineral inclusions were decomposed
and intrinsic absorption bands assignable to diamond were
obtained. Photomicrographs of representative grains for infrared spectra are shown in Fig. 1b.
Raman spectra of carbonado
Raman spectra of the polished carbonado specimen were
obtained with a Raman microprobe consisting of a 30 cm
single polychromator (250is; Chromex, Inc.), equipped
with an optical microscope (BX60; Olympus Optical Co.
Ltd.), an Ar+ laser (5500 A; Ion laser Technology, Inc.), a
CCD camera with 1024 × 128 pixels (DU420-OE; Andor
Technology) and a super notch filter (HIPF-514.5-1.0;
Kaiser Optical Systems, Inc.). Excitation light of the Ar+
gas laser was a 514.5 nm line; the laser power was approximately 3 mW at the sample surface. The Raman shift was
calibrated carefully using a naphthalene standard with energy resolution of approximately 0.1 cm−1 after applying
peak fitting procedures (Kawakami et al., 2003; Fukura
et al., 2006; Yamamoto et al., 2007). The beam size of
the incident light was approximately 1 µm using a 100×
objective lens. All measurements were conducted at room
temperature. The exposure time for obtaining one spectrum
Fig. 1. Photomicrograph of carbonado samples studied here. (a) Polished sample used for measuring Raman spectra. (b) Crushed samples used for measuring IR absorption spectra.
was 20 s. The optical configuration of the Raman microprobe is the confocal alignment, Raman spectra from a
certain depth can be observed selectively. For this study,
Raman spectra of carbonado were obtained from the surface and 10 µm below the surface. In general, carbonado
diamond particles are much smaller than this depth value.
In this study, subtle changes in Raman shift were detected
to determine the residual stress in carbonado. Recently,
Fukura et al. (2006) pointed out that a deviation in room
temperature can affect the Raman shift strongly. For that
reason, we carefully prepared the temperature conditions
of our laboratory, details of which are described later.
Infrared (IR) absorption spectra of carbonado
Infrared (IR) absorption spectra were measured using a
Fourier-transform infrared (FTIR) spectrometer (Spectrum
2000; Perkin-Elmer Inc.) equipped with an IR microprobe.
All samples were heated to 100 ◦ C for at least 8 h before measurements to remove adsorbed water. Chemically
treated carbonado grains were set on a KBr plate and IR incident light was applied through a square optical aperture
of 50 µm × 50 µm at minimum. Dry air was pumped continuously into the IR microscope to prevent interference from
atmospheric water vapor. The IR absorption spectra were
Central African carbonado: spectroscopic study and origin
389
obtained using a globar light source, a KBr beam splitter
and an MCT detector. Spectral measurements were carried
out at 4 cm−1 and interferograms from at least 300 scans
were averaged to obtain one spectrum.
Results and discussion
Figure 2 shows representative Raman spectra of carbonado.
Figure 2a shows a Raman spectrum obtained from 10 µm
below the surface of the polished carbonado sample. Figure 2b shows a Raman spectrum of a single-crystal diamond with diameter of approximately 4 mm. This diamond
is a natural type Ia diamond that is free from mineral inclusions. Raman scattering attributable to the F2g mode
of lattice vibration of diamond was observed at approximately 1333 cm−1 for both spectra. Several remarkable
differences are apparent between the two spectra. First, the
baseline of the carbonado spectrum is much higher than the
standard diamond. As described in previous reports, carbonado samples emit very intense photoluminescence (PL)
because of the presence of radiation-induced lattice defects
(Kagi et al., 1991, 1994a, 1994b, 2007). Excitation with the
514.5 nm Ar+ laser induces PL lines at 504 nm (2.46 eV),
575 nm (2.16 eV) and 638 nm (1.95 eV) for the CAR carbonado samples. The uplifted baseline for carbonado results from the contribution of intense PL of carbonado.
Second, the half width of Raman spectra of carbonado
is much greater than that of the standard diamond sample. The half width of Raman spectra of diamond correlates respectively with the degree of crystallization and
the internal stress. With decreasing diamond crystallinity,
the peak width of Raman spectra increases (Miyamoto,
1998). The difference in the half width is attributable to
the lower crystallinity of carbonado diamond than that of a
millimetre-sized single crystal of diamond. Furthermore, a
considerable up shift in the diamond Raman band was also
observed for some spectra of carbonado as described later.
With increasing internal stress, the peak position of Raman
spectra shifts toward a higher wavenumber (Boppart et al.,
1985). Fukura et al. (2005) reported an upshifted Raman
spectrum of carbonado. According to their results, the upshift (2.0 cm−1 ) in the Raman spectrum corresponding to
0.72 GPa was detected from 5 µm below the surface of the
carbonado sample. They inferred a high-pressure origin of
carbonado from the remnant internal stress inside the specimen. However, no systematic comparison between the surface of carbonado sample and the inside was conducted. In
the present study, we acquired Raman spectra on the surface of the polished specimen and 10 µm below the sample surface. As explained in the experimental procedure,
using the confocal optical alignment with of a ×100 objective lens, the Raman signal was collected locally from a
spheroidal region with 1 µm diameter and 5 µm depth.
Raman spectra of the polished carbonado sample were
obtained at arbitrary positions for 25 points at the sample
surface and 24 points 10 µm below the surface. To determine the peak position of Raman spectra precisely, curve
I
Raman spectra of carbonado
Fig. 2. Raman spectra of a carbonado sample (a) and a single crystal
of natural type Ia diamond (b). Insets show the expanded peak to
clarify differences between peak shapes.
fitting to a Lorentzian function was applied after subtracting a baseline in the region of 1250–1400 cm−1 . Raman
spectra of a standard diamond crystal shown in Fig. 2b were
also measured at arbitrary positions. The average Raman
shift of the standard sample for 8 points was 1333.0 cm−1 ,
with the standard deviation of 0.02 cm−1 . This Raman shift
is a typical value for a single crystal diamond without
strain. Figure 3 shows histograms of the Raman shift for
the polished carbonado sample. Average values of Raman
shift for the carbonado surface and 10 µm below the surface
were, respectively, 1333.0 cm−1 and 1333.5 cm−1 . Raman
spectra collected inside of carbonado sample were slightly
higher in frequency than those collected on the sample surface. The energy resolution of our spectrometer was approximately 0.1 cm−1 by applying the numerical peak fitting procedure and is much smaller than this difference.
As described already, an upshift of Raman spectra toward
higher frequency corresponds to an increase in pressure (internal stress). This property is useful to estimate the pressure of diamond under a stress field. Figure 3a shows that
390
H. Kagi, S. Fukura
Fig. 4. Representative IR absorption spectra of carbonado before
acid leaching. Absorption bands are dominated by fluid, silicates
and carbonate. No absorption bands assignable to diamond were observed.
Fig. 3. Histogram of the peak position of diamond Raman band of
carbonado. (a) Surface of polished carbonado sample. (b) 10 µm
below the surface of the sample.
the average Raman shift of the surface of carbonado was
equal to that of the standard diamond sample. Some of the
data are very low in frequency. These downshifted data presumably resulted from the increased surface temperature
induced by an incident laser (Kagi et al., 1994b). Although
the laser power applied in the present study was sufficiently
low (3 mW) to prevent laser-induced heating, the rise in
sample surface temperature could cause the low Raman frequency because Raman spectra obtained from 10 µm below
the surface did not show such downshifted data. This artefact affects the uppermost surface of the sample with small
grain size whose thermal conductivity is relatively low.
Moreover, it is noteworthy that the histogram of Raman
data from the subsurface is bimodal (see Fig. 3b): one distribution apparently exists at around 1333.8 cm−1 . In the
present study, the maximum Raman frequency was observed at 1334.4 cm−1 , corresponding to the residual pressure of 0.49 GPa by applying the pressure-dependence of
2.87 cm−1 /GPa (Boppart et al., 1985) and resembling the
result of Fukura et al. (2005). The results of the present
study confirm that the residual stress was retained inside
of the carbonado. Furthermore, the bimodal distribution
of Raman shift, as shown in Fig. 3b, suggests that some
regions were stress-free even under the same depth from
the surface, which implies that the residual stress was retained heterogeneously in carbonado and that the residual stress was distributed locally in the grain boundary
or the interface between diamond and mineral inclusions.
The presence of residual stress itself does not necessarily demonstrate that carbonado is of high-pressure origin,
because residual stress is present even in polycrystalline
ceramic materials synthesized at ambient pressure by thermal expansion anisotropy and misfitting intragranular dispersion (e.g. Kovalev et al., 2000; Nakagawa et al., 2006).
However, trapping of mineral inclusions in the mantle can
also induce residual stress because of the difference in the
bulk modulus and thermal expansivity between mineral
inclusions and diamond. The presence of residual stress
clarified in this study cannot be the direct evidence of the
high-pressure origin of carbonado. However, the spatial relationship between residual stress and grain distribution in
future will give a clue to the origin of residual stress in carbonado.
Infrared spectra of carbonado
Figure 4 depicts two representative infrared (IR) spectra of carbonado before acid leaching treatment. These
spectra involve the absorption bands assignable to H2 O
fluid, carbonates clay minerals and silicate. It is noteworthy that these spectra do not exhibit the intrinsic absorption
Central African carbonado: spectroscopic study and origin
Fig. 5. IR absorption spectrum of carbonado after acid leaching.
Intrinsic absorption bands derived from diamond lattice vibration
and nitrogen-related impurity absorption are observed respectively
at around 1900–2300 cm−1 and 1000–1300 cm−1 .
bands derived from diamond lattice in the region between
1800 cm−1 and 2300 cm−1 , which means that all IR absorption bands are attributable to the abundant inclusions
in carbonado. Kagi et al. (1994a) first reported IR absorption spectra of CAR carbonado. The presence of nitrogen
platelet was suggested from the characteristic absorption
band at around 1384 cm−1 . However, Kagi et al. (1994a)
showed no intrinsic absorption band of diamond; it remained unclear whether the band at 1384 cm−1 can truly be
assigned to nitrogen platelets, as pointed out by Garai et al.
(2006). They suggested BN contamination in the stainless
container that was used to crush the carbonado as one possibility.
Figure 5 shows a representative IR spectrum after acid
leaching is shown; the intrinsic diamond absorption and
the absorption bands originating from nitrogen impurity are
readily apparent. Mineral-originated absorption bands disappeared after the treatment. In contrast, fluid-originated
absorption bands at 3400 cm−1 and 1640 cm−1 were observed even after the acid treatments. Before the acid leaching procedure, the IR spectra differed from each other
as shown in Fig. 4. However, after acid leaching, the IR
spectra were mutually similar and actually no difference
is visible grain-by-grain (data not shown). This similarity
strongly suggests that some fluid inclusions are primary
inclusions, the carbonado was secondarily altered; that alteration was heterogeneous within a carbonado sample.
That characteristic is consistent with a report of De et al.
(2001) of heterogeneous textures of CAR carbonado by
cathodoluminescence and carbon isotope distributions.
Furthermore, it is noteworthy that the OH stretching band
at 3400 cm−1 and HOH bending vibration at 1640 cm−1 derived from H2 O liquid in addition to the carbonate band at
1420 cm−1 were ubiquitously observed for the acid treated
grains. We carefully dried the sample before and after
the acid leaching procedure. For that reason, contamination of water during laboratory processes is unlikely. Garai
et al. (2006) also observed absorption bands assignable
391
to liquid water. Fluid inclusions and carbonate inclusions
were highly durable during the acid leaching procedures
in contrast to the hydrous silicate inclusions, which decomposed eventually after the treatment. These observations strongly suggest that the fluid inclusions are primarily
trapped within the carbonado crystals and carbonate minerals precipitate in the inclusions. If that is the case, it can be
suggested that C-O-H fluid or volatile-rich melt had been
involved during the carbonado crystallization. The presence of C-O-H mantle fluid in natural diamonds has been
reported and the close relationship between the mantle fluid
and diamond crystallization has been clarified (Chrenko
et al., 1967; Navon et al., 1988; Guthrie et al., 1991; Navon
1991; Izraeli et al., 2001; Zedgenizov et al., 2004). Considering the residual pressure observed using Raman spectroscopy and the similarity to the mantle-originated diamonds in the involvement of C-O-H fluid in carbonado,
it is reasonable to infer that carbonado was crystallized in
the mantle under high-pressure and high-temperature conditions.
Nitrogen aggregation state of carbonado
The IR absorption bands attributable to nitrogen impurity
involved in the diamond lattice are visible in the region of
1000–1500 cm−1 . Impurities comprising nitrogen atoms in
diamond undergo aggregation in the diamond lattice by kinetic diffusion. The aggregation state depends strongly on
the mantle residence time of diamond, its nitrogen content
and the temperature history (Chrenko et al., 1977; Walker,
1979). The shape of IR absorption bands changes according to the aggregation state of nitrogen. The nitrogen aggregation sequence progresses from isolated substituted nitrogen atoms in type Ib diamonds, nitrogen pairs in type IaA
diamonds and finally type IaB diamond containing nitrogen atoms in tetragonal arrangement and platelets (Evans
& Qi, 1982; Taylor et al., 1996 and references therein).
Figure 6a presents the infrared spectrum of carbonado
corresponding to the region covering absorption bands derived from nitrogen impurities in diamond lattice. The
spectrum is identical to that shown in Fig. 5. The absorption
spectrum is overlapped with a couple of absorption bands
attributable to silicate inclusions. However, the remaining
absorption bands are assignable to absorption bands of nitrogen impurities of type Ib and type IaA. For comparison,
IR absorption spectra of type Ib and type IaA are displayed
respectively in Fig. 6b and 6c. These observations imply
that the aggregation state of carbonado is intermediate between type Ib and type IaA and that the aggregation state
of nitrogen in carbonado is not high.
In addition to the representative carbonado sample whose
IR spectrum is shown in Fig. 6, carbonado diamonds having various nitrogen concentrations were studied. Table 1
lists the nitrogen concentration and the extent of nitrogen
aggregation (IaA (%) = IaA / (Ib + IaA) × 100) for four
grains. The extent of nitrogen aggregation was calculated
using fitting of a linear combination between the type IaA
spectrum (see Fig. 6b) and type Ib spectrum (see Fig. 6c)
into observed IR spectra. The estimated values contain
392
H. Kagi, S. Fukura
of carbonado suggest that the maximum temperature carbonado exposed after the formation of radiation center was
not higher than 400 ◦ C (Kagi et al., 1994a; Kagi et al.,
2007). Recently, attempts were made to duplicate the carbonado microstructure by sintering diamond or graphite
powders at high pressure and high temperature without a
catalyst (Irifune et al., 2003, 2004; De et al., 2004). The
grains of the synthetic polycrystalline diamond aggregates
are one or two orders of magnitude smaller than those of
natural carbonado. In particular, the direct conversion of
graphite to diamond reported by Irifune et al. (2003, 2004)
suggests that diamond aggregates with small grains crystallized at a high super saturation state over a short period.
Taken together, the experimental results obtained in the
present study suggest that some rapid heating event such
as an upwelling hot plume, with the presence of mantle
fluid and subsequent cooling, could be a possible origin of
diamond. The origin of carbonado remains difficult to conclude definitively. Further studies of isotopic compositions
and trace element patterns, including platinum group elements, will elucidate the formation history of carbonado in
the mantle.
Fig. 6. Nitrogen-related band region of IR absorption spectrum of diamonds. (a) Carbonado sample after acid leaching. (b) Natural type
Ia diamond. (c) Synthetic type Ib diamond. An IR spectrum of carbonado is intermediate between type Ia and type Ib diamonds.
Table 1. Concentration and aggregation state of nitrogen in the carbonado samples.
Sample number
ML30_2
ML33_11
ML37_11
ML37_12
N (ppm)
30
240
440
390
IaA (%)
0
50
60
55
5% errors overall. Nitrogen contents were estimated from
the relative absorbance ratio of nitrogen-induced bands to
intrinsic diamond bands around 2000 cm−1 (Clark et al.,
1992). In four samples, except for one sample (ML30_2)
having very low nitrogen concentration, the respective aggregation state was intermediate between those of type Ib
and type IaA and mutually similar.
For examining the precise temperature history and residence time at high temperature activating the aggregation
of nitrogen impurity, we used the relationship between the
extent of aggregation (type IaA %) and nitrogen concentration reported by Taylor et al. (1996). Given that the surrounding temperatures were 2000 ◦ C, 1400 ◦ C, 1000 ◦ C
and 900 ◦ C, the mantle residence time can be estimated
respectively as 300 s, 1 yr, 1 Myr, and 1 Gyr. Ozima &
Tatsumoto (1997) reported that the formation age of carbonado was 2.6–3.8 Gyr ago. It is expected that carbonado
crystallized in a short period and was quenched after crystallization of diamond to reconcile the less-aggregated nitrogen with the geologically greater age of carbonado.
The radiation from radioactive elements was known to be
present after the crystallization event, because the PL bands
Acknowledgements: With great sadness, we announce
that the second author, Satoshi Fukura, passed away on October 10, 2006; it was a tragic loss for us and for this field
of a young researcher full of promise for the future. We are
grateful to Professor Takehiko Yagi for technical advice related to polishing diamonds, Dr. Hiroaki Ohfuji, Dr. Tetsuo
Nakagawa and Mrs. Shoko Odake for valuable discussion.
This study was supported by a Grant-in-Aid for Scientific
Research (16654088, 18340177, 18654098; 19GS0205)
from the Japan Society for Promotion of Science (JSPS),
Iketani Foundation, Sumitomo Fundamental Science Foundation and by a Grant-in-aid for The 21st Century COE
Program for Frontiers in Fundamental Chemistry from the
Ministry of Education, Culture, Sports, Science and Technology.
References
Boppart, H., van Straaten, J., Silvera, I.F. (1985): Raman-spectra of
diamond at high-pressures. Phys. Rev., B 32, 1423-1425.
Chrenko, R.M., McDonald, R.S., Darrow, K.A. (1967): Infra-red
spectra of diamond coat. Nature, 213, 474-476.
Chrenko, R.M., Tuft, R.E., Strong, H.M. (1977): Transformation of
the state of nitrogen in diamond. Nature, 270, 141-144.
Clark, C.D., Collins, A.T., Woods, G.S. (1992): Absorption and
Luminescence Spectroscopy. In “The properties of natural and
synthetic diamond”, J.E. Field, eds., Academic Press, London,
35-79.
Daulton, T.L. & Ozima, M. (1996): Radiation-induced diamond formation in uranium-rich carbonaceous materials. Science, 271,
1260-1263.
De, S., Heaney, P.J., Hargraves, R.B., Vicenzi, E.P., Taylor, P.T.
(1998): Microstructural observations of polycrystalline diamond: a contribution to the carbonado conundrum. Earth
Planet. Sci. Lett., 164, 421-433.
Central African carbonado: spectroscopic study and origin
De, S., Heaney, P.J., Vicenzi, E.P., Wang, J. (2001): Chemical heterogeneity in carbonado, an enigmatic polycrystalline diamond.
Earth Planet. Sci. Lett., 185, 315-330.
De, S., Heaney, P.J., Fei, Y., Vicenzi, E.P. (2004): Microstructural
study of synthetic diamond and comparison with carbonado, a
natural polycrystalline diamond. Am. Mineral., 89, 438-445.
Evans, T. & Qi, Z. (1982): The kinetics of the aggregation of nitrogen atoms in diamond. Proc. R. Soc. London, A381, 159-178.
Fukura, S., Nakagawa, T., Kagi, H. (2005): High spatial resolution photoluminescence and Raman spectroscopic measurements of a natural polycrystalline diamond, carbonado. Diam.
Relat. Mat., 14, 1950-1954.
Fukura, S., Mizukami, T., Odake, S., Kagi, H. (2006): Factors determining the stability, resolution, precision of a conventional
Raman spectrometer. Appl. Spectrosc., 60, 946-950.
Garai, J., Haggerty, S.E., Pekhi, S., Chance, M. (2006): Infrared absorption investigations confirm the extraterrestrial origin of carbonado diamonds. Astrophys. J., 653, L153–L156.
Guthrie Jr., G.D., Veblen, D.R., Navon, O., Rossman, G.R. (1991):
Submicrometer fluid inclusions in turbid-diamond coats. Earth
Planet. Sci. Lett. 105, 1-12.
Heaney, P.J., Vicenzi, E.P., De, S. (2005): Strange diamonds: The
mysterious origins of carbonado and framesite. Elements, 1, 8589.
Irifune, T., Kurio, A., Sakamoto, S., Inoue, T., Sumiya, H. (2003):
Ultrahard polycrystalline diamond from graphite. Nature, 421,
599-600.
Irifune, T., Kurio, A., Sakamoto, S., Inoue, T., Sumiya, H.,
Funakoshi, K. (2004): Formation of pure polycrystalline diamond by direct conversion of graphite at high pressure and high
temperature. Phys. Earth Planet. Inter., 143-144, 593-600.
Izraeli, E.S., Harris, J.W., Navon, O. (1999): Raman barometry of
diamond formation. Earth Planet. Sci. Lett., 173, 351-360.
—,—,— (2001): Brine inclusions in diamonds: a new upper mantle
fluid. Earth Planet. Sci. Lett., 187, 323-332.
Kagi, H., Masuda, A., Takahashi, K. (1991): Laser-induced luminescence from natural polycrystal diamond “Carbonado”.
Naturwissenschaften, 78, 355-358.
Kagi, H., Takahashi, K., Hidaka, H., Masuda, A. (1994a): Chemical
properties of Central African carbonado and its genetic implications. Geochim. Cosmochim. Acta, 58, 2629-2638.
Kagi, H., Tsuchida, I., Wakatsuki, M., Takahashi, K., Kamimura,
N., Iuchi, K. Wada, H. (1994b): Proper understanding of
down-shifted Raman spectra of natural graphite: Direct estimation of laser-induced rise in sample temperature. Geochim.
Cosmochim. Acta, 58, 3527-3530.
Kagi, H., Sato, S., Akagi, T., Kanda, H. (2007): Generation history of carbonado inferred from photoluminescence spectra,
cathodoluminescence image and carbon isotopic composition.
Am. Mineral., 92, 217-224.
Kaminskii, F.V. (1987): Genesis of carbonado; polycrystalline aggregate of diamond. Dokl. Akad. Nauk SSSR, 291, 439-440 (in
Russian).
Kamioka, H., Shibata, K., Kajizuka, I., Ohta, T. (1996): Rare-earth
element patterns and carbon isotopic composition of carbonados: implications for their crustal origin. Geochem. J., 30, 189194.
Kawakami, Y., Yamamoto, J., Kagi, H. (2003): Micro-Raman densimeter for CO2 inclusions in mantle-derived minerals. Appl.
Spectrosc., 57, 1333-1339.
393
Kovalev, S., Ohji, T., Yamauchi, Y., Sakai, M. (2000): Grain boundary strength in non-cubic ceramic polycrystals with misfitting
intragranular inclusions (nanocomposites). J. Mater. Sci., 35,
1405-1412
Miyamoto, M. (1998): Micro Raman spectroscopy of diamonds in
the Canyon Diablo iron meteorite: Implication for the shock origin. Antarct. Meteorite Res., 11, 173-179.
Nakagawa, T., Fukura, S., Nakai, M., Sugiyama, K., Kagi, H.
(2006): Development of a “Built-in” SNOM head for an AFM
system and stress mapping of an Al2 O3 /ZrO2 eutectic composite. Opt. Rev., 13, 269-275.
Navon, O. (1991): High internal pressures in diamond fluid inclusions determined by infrared absorption. Nature, 353, 746-748.
Navon, O., Hucheon, I.D., Rossman, G.R., Wasserburg, G.J. (1988):
Mantle-derived fluids in diamond micro-inclusions. Nature,
335, 784-789.
Ozima, M. & Tatsumoto, M. (1997): Radiation-induced diamond
crystallization: Origin of carbonados and its implication on meteorite nano-diamonds. Geochim. Cosmochim. Acta, 61, 369376.
Ozima, M., Zashu, S., Tomura, K., Matsuhisa, Y. (1991): Constraints
from noble-gas contents on the origin of carbonado diamonds.
Nature, 351, 472-474.
Robinson, D.N. (1978): The characteristics of natural diamond and
their interpretation. Miner. Sci. Eng., 10, 55-72.
Shibata, K., Kamioka, H., Kaminsky, F.V., Koptil, V.I., Svisero, D.P.
(1993): Rare earth element patterns of carbonado and yakutite:
evidence for their crustal origin. Mineral. Mag., 57, 607-611.
Smith, V.J. & Dawson, J.B. (1985): Carbonados; Diamond aggregates from early impacts of crustal rocks? Geology, 13, 342343.
Sobolev, N.V., Fursenko, B.A., Goryainov, S.V., Shu, J., Hemley,
R.J., Mao, H.-K., Boyd, F.R. (2000): Fossilized high pressure
from the Earth’s deep interior: The coesite-in-diamond barometer. Proc. Natl. Acad. Sci. U.S.A., 917, 11875-11879.
Taylor, W.R., Canil, D., Milledge, H.J. (1996): Kinetics of Ib to IaA
nitrogen aggregation in diamond. Geochim. Cosmochim. Acta,
60, 4725-4733.
Trueb, L.F. & Butterman, W.C. (1969): Carbonado: A microstructural study. Am. Mineral., 54, 412-425.
Trueb, L.F. & de Wys, E.C. (1969): Carbonado: Natural polycrystalline diamond. Science, 165, 799-802.
—, — (1971): Carbon from Ubangi – a microstructural study. Am.
Mineral., 56, 1252-1256.
Walker, J. (1979): Optical absorption and luminescence in diamond.
Rep. Prog. Phys., 42, 1605-1659.
Yamamoto, J., Kagi, H., Kawakami, Y., Hirano, H., Nakamura, M.
(2007): Paleo-Moho depth determined from the pressure of CO2
fluid inclusions: Raman spectroscopic barometry of mantle- and
crust-derived rocks. Earth Planet. Sci. Lett., 253, 369-377.
Zedgenizov, D.A., Kagi, H., Shatsky, V.S., Sobolev, N.V. (2004):
Carbonatitic melts in cuboid diamonds from Udachnaya kimberlite pipe (Yakutia): evidence from vibrational spectroscopy.
Mineral. Mag., 68, 61-73.
Received 20 April 2007
Modified version received 20 December 2007
Accepted 10 March 2008