the origin and formation of metamorphic microdiamonds …cartigny/2001-chemgeol-cartignyetal.pdfthe...

Ž .Chemical Geology 176 2001 265–281www.elsevier.comrlocaterchemgeo

The origin and formation of metamorphic microdiamonds fromthe Kokchetav massif, Kazakhstan: a nitrogen and

carbon isotopic study

Pierre Cartigny a,b,), Katrien De Corte c,d, Vladislav S. Shatsky e, Magali Ader a,Paul De Paepe c, Nikolay V. Sobolev e, Marc Javoy a

a Laboratoire de Geochimie des Isotopes Stables, UniÕersite Paris 7 et Institut de Physique du Globe, Place Jussieu 2,´ ´F-75251 Paris, Cedex 05, France

b Geochemisches Institut der UniÕersitat Gottingen, Goldschmidtstraße 1, D-37077 Gottingen, Germany¨ ¨ ¨c Department of Geology and Soil Science, UniÕersity of Ghent, Krijgslaan 281, S8, B-9000 Ghent, Belgium

d Department of Geology and Mineralogy, Royal Museum for Central Africa, LeuÕensesteenweg 13, B-3080 TerÕuren, Belgiume Institute of Mineralogy and Petrography, NoÕosibirsk 630090, Russia

Received 1 October 1999; accepted 25 September 2000

Abstract

This study reports d13C, d

15N and N-content values for microdiamonds from ultrahigh-pressure metamorphic rocks ofthe Kokchetav massif in Kazakhstan. Both alluvial diamonds and in-situ diamonds from a garnet–clinopyroxene rock and a

Ž .marble i.e. a garnet–pyroxene dolomitic rock were investigated. In-situ diamonds were analysed in batches, because ofŽ .their small size average 40 mm , whereas the larger alluvial diamonds were analysed individually. The latter group has

d13C-values ranging from y15.92‰ to y10.57‰, d15N from y1.8‰ to q1.1‰ and N-contents from 2300 to 3650 ppm.

Diamonds from the garnet–clinopyroxene rock yield mean values of y10.50‰ for d13C, q5.9‰ for d

15N and a highaverage nitrogen content of 11,150 ppm. Values for diamonds in marble are y10.19‰, q8.5‰ and 2650 ppm,respectively.

For diamonds from garnet–clinopyroxene rock and marble, there is more nitrogen released by bulk combustion thanŽ .estimated by infrared IR spectroscopy, the differences being of about 7000 and 1500 ppm, respectively. These differences

suggest that a significant quantity of nitrogen is IR-inactive and may be present as fluid inclusions. Their carbon andnitrogen isotopic compositions are compatible with an in-situ crystallisation of diamond from dominantly metasedimentarysources, suggesting that sedimentary nitrogen can be subducted to very high pressures. Carbon isotopic fractionationbetween coexisting carbonate and diamond suggests crystallisation temperatures before the peak of metamorphism attemperatures probably below 7008C and deduced pressures of 3 GPa. Relative to the isotopic data reported for sediments,

) Corresponding author. Laboratoire de Geochimie des Isotopes, Universite Paris 7, Institut de Physique du Globe, Place Jussieu 2, 75251´ ´Paris, Cedex 05, France. Fax: q33-1-44-27-28-30.

Ž .E-mail address: [email protected] P. Cartigny .

0009-2541r01r$ - see front matter q 2001 Elsevier Science B.V. All rights reserved.Ž .PII: S0009-2541 00 00407-1

( )P. Cartigny et al.rChemical Geology 176 2001 265–281266

metasediments and in-situ diamonds, the slightly 15N-depleted compositions of alluvial diamonds is striking. These valuessuggest that the contribution of any metasedimentary source is unlikely and may point toward a maficrultramafic protolith.q 2001 Elsevier Science B.V. All rights reserved.

Keywords: Metamorphic microdiamond; Nitrogen; Carbon

1. Introduction

ŽGeological environments for microdiamond -.1mm occurrences are increasingly recognized on

Earth. They include kimberlites, lamproites andŽ .placer deposits e.g. Harris, 1992 for review , in

Žmeteorites and impact craters e.g. Hough et al.,. Ž .1995 , in lamprophyre dike MacRae et al., 1995 ,

Ž .volcaniclastic komatiites Capdevilla et al., 1999 ,Ž .K–T boundary sites e.g. Carlisle and Braman, 1991

Žand ultrahigh-pressure metamorphic rocks e.g..Sobolev and Schatsky, 1990 . Microdiamonds in the

last context occur in metamorphic rocks from theŽKokchetav massif in Kazakhstan Sobolev and

. ŽSchatsky, 1990 , in Dabie Shan, China Xu et al.,.1992 , in the Western Gneiss region of Norway

Ž .Dobrzhinetskaya et al., 1995 and in the SaxonianŽ .Erzgebirge, Germany Massonne, 1999 . However,

abundant crystals are found only in the Kokchetavmassif. In Kazakhstan, previous studies strongly sug-gest that diamonds were not incorporated in thesediments before subduction, deriving, for example,from an eroded kimberlite but have crystallised in-situwithin their stability field and therefore have formedin a paleo ultrahigh-pressure metamorphic eventŽ .Shatsky et al., 1989; De Corte et al., 1998 .

Despite the small size of in-situ Kokchetav dia-Ž .monds average size is 15 mm; Shatsky et al., 1989 ,

their morphology and Fourier Transform InfraRedŽ .FTIR spectroscopy data compared to larger alluvialdiamonds demonstrate that diamond characteristicsvary as a function of the host-rock, supporting the

Žpresence of distinct diamond populations De Corte.et al., 1998, 1999 . For example, diamonds with

octahedral habit are observed only in zoisite gneissesdevoid of symplectitic zoisite. Cubo-octahedral crys-tals are predominant in biotite gneisses and are char-acteristic also of diamonds found in the alluvialsderived from the massif. Garnet–clinopyroxene–

Ž .dolomitic rocks i.e. impure marbles , zoisite gneissesrich in symplectitic zoisite and garnet–clinopyroxenerock typically contain cuboidal diamonds. Similari-ties of morphology and real structure for microdia-monds from kimberlites and metamorphic rocks have

Ž .recently been discussed by Shatsky et al. 1998 .ŽPrevious work Finnie et al., 1994; Taylor et al.,

.1996; De Corte et al., 1998, 1999 has also shownthat the degree of substitution of carbon by nitrogenwithin the crystal structure as detected by FTIRmicrospectroscopy is the most pronounced for allu-

Ž .vial diamonds up to 5200 ppm N; ns20 . Dia-monds from garnet–clinopyroxene rocks form an

Ž .intermediate group 1050–2800 ppm N; ns18 ,Žfollowed by diamonds from marble less than 580

.ppm N; ns6 . The idea of distinct diamond popula-tions is also supported by the evidence that diamondsfrom garnet–clinopyroxene, zoisite gneiss with sym-plectitic zoisite and marbles typically contain waterand carbonate inclusions. In contrast, alluvial dia-monds do not contain carbonate inclusions or waterŽ .or only a minor amount .

The finding of carbonate and water inclusions inmetamorphic microdiamonds from garnet–clinopy-roxenite and marble implies diamond growth from a

Ž .C–O–H rich fluid De Corte et al., 1998, 1999 . TheŽ .source s of these fluids however remains unclear. In

Ž .an attempt to better constrain the source s of carbonand nitrogen of these microdiamonds, carbon andnitrogen isotopic compositions together with nitro-gen contents have been determined on a series ofwell-characterised samples. The study of microdia-monds is also simplified and significant relative to

Ž .other e.g. silicate minerals since they record onlyultra-high pressure metamorphic conditions and didnot reequilibrate during retrograde metamorphismoutside their stability field; otherwise, they wouldhave been graphitised. The study of microdiamondsmay thus provide an insight into the conditions

( )P. Cartigny et al.rChemical Geology 176 2001 265–281 267

prevailing under ultra high-pressure metamorphismand probably at the time of their crystallisation.

2. Sample description

In the present study, three well characterised sam-Ž .ple types were considered: 1 alluvial diamonds

from Tertiary sands located between 100 and 200 kmŽ .north of the Kokchetav massif; 2 diamonds ex-

Žtracted from a garnet–clinopyroxene rock sample. Ž .2-4i ; and 3 diamonds isolated from a garnet–

clinopyroxene dolomitic rock, also referred to asŽ . Ž .marble sample K92-99i . Diamonds from groups 2

Ž .and 3 are described in the text as Ain-situ dia-mondsB.

Alluvial diamonds range in size between 100 and300 mm. In contrast, in-situ diamonds are rarelylarger than 135 mm. They were isolated by a thermo-chemical extraction process described by De CorteŽ .2000 . Both in-situ diamonds larger than 60 mm insize and all alluvial diamonds have been previously

Žanalysed by FTIR spectroscopy ns9 for sample2-4i; ns3 for sample K92-99i; De Corte et al.,

.1998, 1999 .In addition, we determined carbonate contents and

carbon isotopic compositions of the garnet–clino-pyroxene rock 2-4 and the marble K92-99.

3. Analytical techniques

Alluvial and in-situ microdiamonds were wrappedin platinum foil and cleaned at 6008C in air toremove any organic contamination. Due to the smalldiamond size and thus a high surface to volume ratioŽ .to be degassed , special care was taken to removeany gas absorbed on the sample surface. Samples

Ž y7 .were therefore kept under vacuum f10 Torr for12 h and afterwards were degassed for 1 h by

Ž y8 .pyrolysis under higher vacuum f10 Torr at11008C before combustion. To demonstrate the ab-

40 Ž .sence of any air-contribution, Ar i.e. mrzs40was also monitored in the mass spectrometer. Theamount of 40Ar recovered from sample and blankanalyses being identical within experimental error, itshows that atmospheric nitrogen contamination wasindeed avoided.

Samples were analysed using the procedure de-Ž .scribed by Boyd et al. 1995 for conventional

macrodiamonds with a combustion apparatus slightlymodified for these samples. Compared to the on-line

Ž .combustion of Boyd et al. 1995 , the reaction vesselŽin these analyses was made of a quartz tube OD

.6rID 4 mm crimped in the middle to receive theŽ .sample see Fig. 1 .

Diamonds were combusted online in an oxygenatmosphere and carbon converted into CO . Nitro-2

gen was separated from carbon dioxide and anynitrogen oxide reduced to N using a CaO–Cu fur-2

nace. Nitrogen concentrations were measured withan accuracy better than 5% using a capacitancemanometer. Nitrogen isotopic composition was anal-ysed with a triple collector static vacuum mass spec-trometer connected directly to the extraction line. Anaccuracy of 0.5‰ for d

15N was established on thebasis of the analyses of international standards. CO2

was quantified with a precision better than 1% usinga piezoresistive gauge. Values for d

13C were deter-mined with an accuracy of 0.1‰ using a conven-

Žtional gas isotope ratio mass spectrometer for moredetails concerning this paragraph, see Boyd et al.,

.1995 and references therein .The relatively large size of alluvial diamonds

allowed single stones to be analysed, whereas in-situ

Fig. 1. Schematic outline of the carbon and nitrogen extractionŽ .line. See text and also Boyd et al. 1995 for details. PGs

piezoresistive gauge.


Table 1d

13C–d15 N and N contents of microdiamonds found in alluvialsand deriving from ultrahigh-pressure rocks of Kokchetav

13 15Name Weight d C d N N NŽ . Ž . Ž . Ž . Ž .mg ‰ ‰ ppm ppm

aFTIRy0 .7T86-4 0.0896 y11.95 y0.3 2352 2360–2713q0 .7y0.8T86-5 0.0517 y12.60 y0.3 2923 2696–3550q0 .8y0.9T86-8 0.0417 y12.29 q1.1 2735 2477–5076q1 .0y0.9T86-10 0.0535 y15.76 y1.1 2681 2498–2758q0 .8y1.0T86-11 0.0407 y12.80 q0.1 2361 2709–3109q1 .0y0.8T86-13 0.0438 y15.92 q0.0 3643 2128–2588q0 .8y0.8T86-14 0.0661 y10.57 y1.4 3164 2695–2840q0 .7y0.9T86-15 0.0574 y13.36 y1.8 2324 3814–5236q0 .8y1.1T86-16 0.0315 y12.29 y0.8 2860 2519–3367q1 .0

a Nitrogen contents obtained using infrared microspectroscopyŽ .DeCorte et al., 1998, 1999 are shown for comparison.

diamonds needed to be grouped prior to analysis.ŽWith a mean diamond size of ca. 30 mm computed

.from the 2-4 sample , a 1-mg sample is composed ofmore than 25,000 microdiamonds. Analyses of allu-vial diamonds allow the variability between singlesamples to be investigated whereas analyses of in-situdiamonds yield average d

13C–d15N–N-values.A further technical problem was the presence of

remaining zircons in the in-situ diamond mixture.Even after copious and very careful handpicking,about 10"5% of this mineral remained. However, a

Ž .complete combustion better than 99.9% of dia-monds is readily obtained under our experimentalconditions whereas zircons are absolutely undis-turbed under these conditions. Hence, nitrogen con-tents of in-situ microdiamonds were calculated usingthe amount of carbon released by diamond combus-tions, as quantified by the piezoresistive gauge. Us-ing this method, nitrogen contents of in-situ microdi-

Ž .amonds, see sample 2-4, Table 2 are reproducible.No correlation between carbon yields and calculatednitrogen contents was observed which confirms thatzircons did not contribute to the nitrogen recoveredfrom the combustion.

Carbonate and total reduced carbon contents andassociated carbon isotopic compositions were deter-

Žmined for the garnet–clinopyroxene rock sample. Ž .2-4-i and for the marble sample K92-99-i using a

step-heating procedure. The apparatus used is exactlythe same as the one described by Pineau and JavoyŽ .1994 ; the temperature and the oxygen fugacity

conditions of the steps changed and were givenbelow. The gases were sequentially extracted fromthe sample in three temperature steps of 1 h each.The first step at 5008C with 12 mm Hg oxygenpressure removed any organic carbon contamination.The second step at 8008C under vacuum decarbon-ated the sample, and the CO extracted allows the2

determination of the carbonate content and C-iso-topic composition. The third step at 13508C with 12mm Hg oxygen pressure combusted the total reduced

Ž .carbon graphiteqdiamond and the resulting CO2

allows the determination of the reduced carbon con-tent and isotopic composition. In the case of themarble, a good separation of the CO from the total2

reduced carbon was impossible by this techniquesince carbonate minerals were too abundant andwere present as inclusions in more refractory miner-als. In this case, the CO recovered from the last2

step contains a mixture of CO coming from the2

decarbonation of residual carbonate and from theoxidation of reduced carbon. Consequently, this stepgives a maximum value for the reduced carbon

Ž .content sample 92-99-i .

4. Results

The results for the alluvial diamonds are reportedin Table 1. They have d

13C-values ranging fromŽ .y15.92‰ to y 10.57‰ mean s y13.06‰ ,

Table 2d

13C– d15 N and N contents of microdiamonds from garnet–Ž . Žclinopyroxene rock sample 2-4-i and marble i.e. garnet–pyrox-

.ene dolomitic, sample 92-99-i from the ultrahigh-pressure massifof Kokchetav

13 15Name Weight d C d N N NŽ . Ž . Ž . Ž . Ž .mg ‰ ‰ ppm ppm

aFTIRy0 .52-4-i 1.5324 y10.71 q5.6 11,340 1050–2800q0 .6y0.52-4-i 0.7893 y10.65 q6.6 11,385 1050–2800q0 .6y0.52-4-i 0.6206 y10.34 q6.1 10,872 1050–2800q0 .6y0.52-4-i 1.6789 y10.50 q5.8 11,237 1050–2800q0 .5y0.52-4-i 0.5990 y10.32 q5.3 10,879 1050–2800q0 .6y0.592-99-i 0.1181 y10.22 q8.0 2620 -580q0 .5y0.592-99-i 0.2197 y10.21 q8.9 2762 -580q1 .4

a Nitrogen contents obtained using infrared microspectroscopyŽ . ŽDeCorte et al., 1998, 1999 are shown for comparison ns9 and

.ns3 for sample 2-4-i and 92-99-i, respectively .


Table 3Carbon isotopic composition and content of carbonate and totalreduced carbon of diamond-bearing garnet–clinopyroxene rockand marble

Name Carbonate Reduced carbon13 13Ž . Ž . Ž . Ž .C ppm d C ‰ C ppm d C ‰

2-4-i 480 y5.9"0.3 490 y10.4"0.292-99-i 88,200 y4.3"0.1 -2680 n.d.

d15N-values close to the atmospheric value, fromy

Ž .1.8‰ toq1.1‰ meansy0.5‰ and high N-con-tents, from f2300 to 3650 ppm. Results for in-situdiamonds are reported in Table 2. Five batches of

Ždiamonds from a garnet–clinopyroxene rock sample.2-4 were analysed. Results are consistent, yielding a

mean d13C-value of y10.50‰, a mean d

15N-valueof q5.9‰ and high average N contents of 11,150

Ž .ppm. Two batches of diamonds 92-99i, Table 2from the marble give a mean d

13C-value ofy10.22‰, with a mean d

15N-value of q8.5‰ andan average N content of 2650 ppm. Measured carbon

Ž .isotopic compositions Tables 1 and 2 are similar toŽ .the values reported by Pechnikov et al. 1993 forŽdiamonds from a pyroxene–carbonate rock y12.4‰

.to y10.6‰ from the Kokchetav massif. Carbonatecontents and carbon isotopic compositions from boththe garnet–clinopyroxene rock and the marble arereported in Table 3; carbonate carbon contents beingabout 480 ppm and 8.82 wt.%C with d

13C-values ofy5.9‰ and y4.3‰, respectively. The total reduced

Ž .carbon graphiteqdiamond content of the marble isbelow 2680 ppm. The total reduced carbon in gar-net–clinopyroxene rock is 490 ppm, very similar tocarbonate contents, with a d

13C-value of y10.4"0.2‰, very similar to the mean diamond d

13C-valueŽ .see Table 2 .

5. Discussion

5.1. Nitrogen as fluid inclusions: new eÕidence fordiamond crystallisation from a fluid phase

In contrast to alluvial diamonds for which there isa reasonable agreement between nitrogen contentsdetermined by combustion and by FTIR spec-

Ž .troscopy Table 1 , the two methods give very signif-Ž .icant differences for in-situ diamonds Table 2 . For

Ždiamonds from the garnet–clinopyroxene rock sam-. Ž .ple 2-4i and the marble 92-99i , there is more

nitrogen released by bulk combustion than estimatedŽ .by infrared IR spectroscopy. For diamonds from

Ž .the garnet–clinopyroxene rock sample 2-4i , theamount of nitrogen can reach more than 7000 ppm.Diamonds from this rock may therefore contain atleast 7000 ppm of an infrared inactive form. FTIRspectroscopy of diamonds from marble is not pre-cisely constrained as a consequence of poor qualityinfrared spectra, but the reliable data obtained on twocrystals suggest a maximum nitrogen content of 580

Ž .ppm De Corte et al., 1999; Table 2 . If this value isincreased to 1100 ppm, in order to account for apossible higher mean N-content over this diamondpopulation, an excess of f1500 ppm of infrared-inactive nitrogen must be present also in thesediamonds. Such a difference between FTIR spec-troscopy and bulk combustions has, to our knowl-edge, not been previously reported among mantle-de-

Ž .rived diamonds and may be symptomatic of somemetamorphic diamonds.

Within the diamond crystal lattice, nitrogen oc-curs in a variety of defects, most giving rise tocharacteristic IR absorptions. It is well establishedthat there exists four main nitrogen-bearing defects;

Ž . ŽA IaA diamond, N-pairs , B IaB, probably four N. Ž .atoms plus a vacancy , C Ib, isolated N and D

Ž . Žunknown structure see Davies, 1976; Woods, 1986;.Sobolev, 1991; Jones et al., 1992 . These four types

of defects are linked by a diffusion process, follow-ing a second order kinetic, resulting in conversion ofC to A centres and, when C is no longer present, to

ŽB and D defect centres e.g. Chrenko et al., 1977;.Evans and Qi, 1982 . The aggregation sequence can

Ž .be extended if platelet i.e. bearing D-defects degra-Ždation leads to the formation of voidites e.g. Evans

. Žet al., 1995 which are not detectable by FTIR e.g..Woods, 1994 . In diamonds characterised by an ad-

vanced nitrogen aggregation state, it has been shownthat IR inactive nitrogen may be present within

Ž .voidites e.g. Woods, 1994 and references therein .Metamorphic microdiamonds have low nitrogen

aggregation states, yet very high nitrogen contents.All the diamonds studied so far only contain nitrogen

ŽC- and A-defects Type Ib-IaA diamonds; Finnie et


al., 1994; Taylor et al., 1996; De Corte et al., 1998,.1999 and thus neither B- or D-defects. The missing

nitrogen therefore is unlikely to be within the crystalstructure initially in some form of substitution forcarbon, but rather outside.

The most likely explanation is that excess nitro-gen is present as fluid inclusions. This suggestion issupported by two observations. First, a C–O–H richfluid trapped within in-situ diamonds from both themarble and garnet–pyroxene rocks has been identi-fied by FTIR spectroscopy. Secondly, in the meantime, alluvial diamonds do not show any excess

Ž .nitrogen together with no or very minor amount ofŽ .water or carbonate De Corte et al., 1998, 1999 .

This observation links the presencerabsence of ex-cess nitrogen to the presencerabsence of trapped

Ž .fluid s . The likely presence of fluid inclusions withinthe microdiamonds may also explain the relativelyhigh concentration of noble gases observed by Ver-

Ž .chovsky and Begeman 1993 during step-heatingcombustion.

Because part of the nitrogen from in-situ dia-Žmonds is present within two phases diamond and

. 15probably fluid inclusions , the measured d N-valuewill represent a mixture of substitutional and non-substitutional nitrogen. This is clear that nitrogen

Žpresent under a substitutional form i.e. trapped.within the diamond lattice was initially present in

the C–O–H rich fluid from which the diamondcrystallised. In the same way, nitrogen under a non-substitutional form was trapped during growth, within

Žthe fibrous texture of the diamond Shatsky et al.,.1998; De Corte et al., 1998 and is originating from

the same fluid from which diamond crystallised. Thediamond d

15N-value is therefore likely to be repre-sentative of the d

15N of the growth medium fromwhich metamorphic diamond crystallised. In fact, themeasured d

15N-value could be difficult to interpretonly if diamond growth was associated with a strongfractionation of N-isotopes. There is however norelationship between the measured d

15N-value andthe relative amount of nitrogen under substitutional

Žand nonsubstitutional form i.e. from about 0% up to80% nonsubstitutional-N for diamonds from alluvials

.and garnet–clinopyroxene rock, respectively . Thisdoes not allow to suggest a strong kinetic fractiona-tion of N-isotopes between N and N in diamond2

during crystal growth. The d15N-value, therefore, is

considered as significantly representative of thegrowth medium.

5.2. Source of carbon and nitrogen within in-situdiamonds and growth scenario

The origin and formation of metamorphic micro-diamonds from the Kokchetav massif has been ad-dressed by several groups. It has been either arguedthat metamorphic diamonds grew within their host-

Ž .rocks Sobolev and Schatsky, 1990 or were detritalminerals incorported in the sediments which were

Ž .later metamorphosed Marakushev et al., 1998 . Fur-ther work demonstrated a strong dependence be-

Žtween the physical characteristics e.g. shape, size,.color, substitutional nitrogen content and abundance

of in-situ diamonds between the different host rocksŽ .De Corte et al., 1999 . Such evidence is seen asincompatible with a detrital origin of the diamondsand supports an in-situ crystallisation of the dia-monds. The present results show that there existsalso a strong difference between d

15N-values, bulkN-contents and the proportion of nonsubstitutionalnitrogen for the diamonds from the different host-rocks analysed. Although covering a quite narrowrange of d

13C-values, it could also be that mean13 Ždiamond d C-values i.e. y13.06, y10.50 and

y10.22 for diamonds from the alluvials, garnet–.clinopyroxene rock and marble rock, respectively

would show slight but significant differences and becharacteristics of their host rock. Taken together, thepresent results support the idea of a formation ofdiamonds within their host rock, a conclusion inagreement with previous conclusions summarisedabove. Moreover, in-situ diamonds have unusuallyhigh nitrogen content relative to kimberlitic dia-

Ž .monds Fig. 2 . This observation is not compatiblewith a model suggesting that the diamonds wouldderive from a kimberlite and were incorporated in

Žthe sediments prior to subduction Marakushev et al.,.1998 . Accordingly, the most plausible explanation

is that sources for C and N derived from the meta-morphic rocks.

Phanerozoic organic matter is characterised mostlyby positive d

15N-values, typically from 0‰ to q5‰Ž .e.g. Rau et al., 1987; Peters et al., 1978; Fig. 3 .Through diagenesis and metamorphism, organic mat-ter undergoes a strong loss of nitrogen but no major


13 Ž .Fig. 2. d C–N diagram for alluvial and in-situ metamorphic diamonds this study and worldwide diamonds from kimberlites andŽ .lamproites adapted from Cartigny et al., 1999 .


15 Ž . Ž .Fig. 3. d N-values for alluvial and in-situ metamorphic diamonds this study and for recent marine sediments Peters et al., 1978 ,Ž . Ž .ammoniacal nitrogen in metasediments Haendel et al., 1986 , ammoniacal nitrogen in subduction zones Bebout and Fogel, 1992 ,

Ž .worldwide fibrous diamonds and peridotitic diamonds from China Cartigny et al., 1997 and references therein .

15 Ž .variations of its d N-value Ader et al., 1998 . Withincreasing metamorphism, nitrogen can be fixed as

Ž .fluid inclusions Andersen et al., 1989 or as ammo-nium substituting for potassium within potassic min-

Žerals such as clay minerals e.g. Williams and Fer-. Žrell, 1991 , micas and feldspars e.g. Honma and

.Itihara, 1981 . It is well documented that devolatili-sation processes occur during prograde metamor-phism and lead to further loss of nitrogen and to anincrease in d

15N-values of the residual ammoniacalŽnitrogen; from q5‰ up to q16‰ Haendel et al.,.1986; Bebout and Fogel, 1992; Fig. 3 . As symbol-


ised by arrows in Fig. 4, previous data for metasedi-Ž .ments up to 18 kbar demonstrate that increasing

metamorphism is not only associated with an enrich-ment in 15N of ammoniacal nitrogen but also with anenrichment in 13C of the reduced carbon, the latter

12 Ž .due to the loss of C depleted-material e.g. CH 4Žandror partial reequilibration with carbonates e.g.

.Bebout, 1995 .Microdiamonds from garnet–clinopyroxene rock

and marble from Kokchetav clearly display positive15 Ž .d N-values Fig. 3; Table 2 . Nitrogen isotopic data

are therefore in agreement with a metasedimentaryorigin for the nitrogen. On Fig. 4, microdiamonds

13 15 Ž .d C- and d N-values Tables 1 and 2 are plottedŽ .together with those of Bebout 1995 . Fig. 4 includes

data obtained on microdiamonds formed from a fluidwhereas other data represent residual ammoniacal

nitrogen. In-situ diamonds indeed precipitated from aC–O–H fluid-phase, and therefore represent nitrogentrapped from a devolatilised phase whereas lowermetamorphic grade sediments represents residual

Ž .ammoniacal nitrogen Bebout, 1995 . Consideringthe nitrogen isotopic fractionation, D -valuesrock-fluid

for nitrogen bearing-species, in particular betweenNHq and N , isotopic fractionation values are likely4 2

Žto be below 2‰ at 7008C Hanschmann, 1981; Richet.et al., 1977; Bebout and Fogel, 1992 . From Fig. 4,

in-situ metamorphic microdiamonds fall on the ex-trapolated metamorphic trend and one can estimatethat ammoniacal nitrogen within potassic phases hadd

15N-values about 8‰ and 10‰.In the present study, it is most probable that

in-situ diamond d13C-values reflect the strong influ-

Žence of partial reequilibration with carbonates see

Fig. 4. Carbon and nitrogen isotopic compositions of microdiamonds from the Kokchetav massif and data for ammoniacal nitrogen inŽ .metamorphic rocks from Catalina Bebout, 1995 .


.below since carbonates are obviously present inŽ .high abundances Table 3 . In sediments, carbon

Ž .resides either as organic matter i.e. reduced or asŽ . 13carbonates oxidised , having initial mean d C-val-

Žues around y25 e.g. Craig, 1953; Schidlowski,. Ž1983 and 0‰ e.g. Craig, 1953; Veizer and Hoefs,.1976 , respectively. It has also been shown that with

increasing metamorphism, graphite–carbonate iso-topic equilibrium can be reached during graphitecrystallisation at temperatures as low as 5008C and isalmost always reached for temperatures above 6008CŽsee particularly Figs. 3 and 4 in Dunn and Valley,1992; and also Hahn-Weinheimer, 1965; Sheppardand Schwarz, 1970; Pineau et al., 1976; Valley andO’Neill, 1981; Wada and Suzuki, 1983; Scheele andHoefs, 1992; Hoffbauer and Spiering, 1994 and ref-

.erences therein . Despite there being no experimentaldata involving diamond, theoretical work suggeststhat the fractionation for carbon isotopic equilibrium

between graphite and diamond is likely to be smallerŽthan 0.5‰ at temperatures above 7008C Bottinga,

.1969; Polyakov and Kharlashina, 1995 . Therefore,under equilibrium conditions, fractionations in thediamond–carbonate system can be estimated. Weneglect the carbon isotopic fractionation betweendiamond and graphite. Taking it into account wouldyield even lower carbon isotopic temperatures. Thedifferent fractionation curves between graphite andcarbonate, as a function of temperature, are repre-sented on Fig. 5. For samples 2-4i and 92-99i,carbon isotopic fractionations between graphite andcarbonate are calculated from the data in Tables 2and 3 and represented on Fig. 5. Considering thewhole set of fractionation curves yields a range ofequilibrium isotopic temperatures between f4708Cand 6808C. Considering the most consistent fraction-ation curves yields temperatures between f5408C

Ž .and 6408C Fig. 5 . These temperatures are clearly

Fig. 5. Carbon isotopic thermometry of carbonate–diamond-bearing rocks according to theoretical, experimental and empirical determinedcarbon isotopic fractionation between calcite and graphite. Empirical or experimental carbon isotopic fractionation between graphite and

Ždiamond are lacking but theoretical work suggests fractionations below f0.5‰ for the deduced crystallisation temperatures adapted from.Scheele and Hoefs, 1992 .


below the 900–10008C envisaged for the peak ofŽmetamorphism in the Kokchetav massif Chopin and

.Sobolev, 1995; Shatsky et al., 1999 .From Fig. 6 where the graphiterdiamond transi-

tion is represented and the estimated temperatures, apressure of crystallisation between 3.0 and 3.5 GPacan be inferred. On Fig. 6 is also represented theP–T evolution of the massif as reviewed by Zhang et

Ž .al. 1997 . From this figure, one can first notice thatthe deduced temperature–pressure window is com-patible with the P–T evolution recorded by meta-morphic rocks of the massif. Second, diamond growthhas to take place during prograde metamorphism.

We conclude that these isotopic temperatures arecrystallisation temperatures. This implies that dia-mond records a crystallisation temperature, a situa-tion contrasting with graphite, which usually records

Žpeak metamorphism temperatures Dunn and Valley,.1992; Satish-Kumar and Wada, 1998 .

The isotopic temperature proposed above arebased on the assumption that isotopic equilibriumbetween carbonate and diamond was reached andthat no other process affect the carbon isotopic com-position of diamond or carbonate. There are howeverseveral factors that could affect slightly, or evenstrongly, the accuracy of isotopic temperatures. The

Fig. 6. The proposed P–T conditions for diamond crystallisation are compared with the P–T evolution of the Kokchetav massif as reviewedŽ .by Zhang et al. 1997 .


first would consist in a carbon isotopic disequilib-rium between carbonates and diamond during dia-mond formation, diamonds having crystallised athigher temperatures than those recorded by the car-bonate–diamond isotope thermometer. However,since isotopic equilibrium is almost always reached

Žat temperatures of about 7008C Sheppard andSchwarz, 1970; Pineau et al., 1976; Wada and Suzuki,1983; Dunn and Valley, 1992; Scheele and Hoefs,

.1992; Hoffbauer and Spiering, 1994 , it would bevery difficult to argue for the preservation of anyC-isotope disequilibrium at temperatures in excess of10008C. Thus, if a slight isotopic disequilibriumcannot be ruled out from the present data, it wouldcertainly not change the fact that diamond crystalli-sation did not occur at the peak of metamorphism.The second factor that could affect the accuracy ofisotopic temperatures can be the crystallisation of

Žgraphite from graphitised diamonds or retrograde.C-bearing fluids in isotopic equilibrium with car-

bonates during retrograde metamorphism. In that13 Žcase, carbonate d C-values may change and in-

.crease in the closed system hypothesis and accord-ingly affect the deduced isotopic temperature. Suchhypothesis is however incompatible with the ob-served data. For marble, from a mass balance pointof view, the high abundance of carbonate relative to

Ž .total reduced carbon Table 3 precludes any majorchanges of the carbonate d

13C. The garnet–clino-pyroxene rock displays similar carbon abundances in

Ž .carbonate and reduced carbon Table 3 . In that case,the d

13C-values measured for diamond and totalŽreduced carbon are within experimental error Tables

.2 and 3 . Therefore, the crystallisation of graphitefrom diamond is unlikely to affect the carbonated

13C-values since both graphite and diamond havenearly identical d

13C-values. We conclude that theprecipitation of graphite at equilibrium with carbon-ate is not indicated by the present observation. Infact, the present data rather suggest that graphiterepresent graphitised diamonds and that the amountof carbon brought in the rock during retrogrademetamorphism is relatively small.

Another difficulty linked with low crystallisationtemperatures exists in explaining how diamond nu-cleation can be reached at temperatures f7008Csince in experiment, diamond cannot be synthesisedat such low temperatures. In this particular case, we

Ž .question whether i.e. synthetic diamond tempera-ture constraints can be simply applied to natural

Ž .diamonds. For example, graphite as diamond canhardly be synthesised at temperatures below 8008C,whereas in natural systems graphite crystals are al-

Žready commonly formed at 5008C e.g. Tagiri and.Oba, 1986 . Thus, although several factors could

have affected the isotopic temperatures of diamondcrystallisation, the hypothesis that diamonds crys-tallised at temperatures about 7008C during progradepath, as opposed to peak of metamorphism, is fa-vored.

In summary, it is likely that, prior to diamondcrystallisation, reduced carbon was present asgraphite and that nitrogen was trapped within a

Žpotassium bearing phase, such as phengite since ithas been identified as intergrown with diamond,

.Shatsky et al., 1995 or one of its precursors. Dia-monds may have crystallised at temperatures lowerthan 7008C during the prograde path, the differentdiamond morphologies reflecting variable drivingforce conditions. Metamorphic microdiamond crys-tallisation may show similarities with the experimen-

Ž .tal work of Tanigushi et al. 1996 , Pal’yanov et al.Ž . Ž . Ž .1998, 1999 , Litvin 1998 and Litvin et al. 1998in which synthetic diamonds are grown from alkaline

Žcarbonate–carbon melted mixtures although in theseexperiments, synthetic diamonds are also grown un-

.der higher P–T conditions . The production of acarbonate-bearing fluid may catalyse the conversionof reduced carbon from graphite to diamond androrthe reduction of carbonate to diamond. In addition,the presence of carbonate in the fluid induces adecrease in the activity of NHq in silicate minerals,4

resulting in an increase of the fugacity of N in the2Ž .fluids e.g. Moine et al., 1994 . This can account for

both the likely presence of N in fluid inclusions and2

for the relatively high nitrogen contents of diamonds.

5.3. AlluÕial diamonds from the North of theKokchetaÕ massif

On one hand, previous investigations by De CorteŽ .et al. 1998, 1999 outlined the strong similarity in

the state of nitrogen aggregation between in-situ andalluvial diamonds, the latter deriving from the UHPmassif and found in alluvial placers located ca. 200


km to the north of the Kokchetav massif. Theseresults implied that the diamonds are likely to haveundergone similar T–t evolution, in other words that

Žthey are all metamorphic again see Fig. 6 of De.Corte et al., 1998 . On the other hand, this study

revealed strong differences between the two diamondsets and alluvial diamonds. This observation hasseveral implications. Firstly, alluvial diamonds maynot be representative of the whole diamond deposit,and secondly, these alluvial diamonds may derive

Ž .from a protolith or protoliths different from thediamond-bearing rocks under current investigation.Part of the present work was to further consider thispossibility.

Indeed the present set of data further supports theexistence of significant differences between alluvialand in-situ diamonds. For alluvial diamonds, there isa reasonable agreement between nitrogen contentsdetermined by combustion and FTIR spectroscopyŽ .Table 1 . Those diamonds contain no water or

Ž .carbonate inclusions or only minor amountswhereas this was not the case for in-situ diamondsŽthis study, see also Finnie et al., 1994; De Corte et

.al., 1998, 1999 . In addition, alluvial and in-situdiamonds show strikingly different nitrogen isotopic

15 Žcompositions, the low d N-values i.e. y1.8‰ to.q1.1‰; y0.5‰ mean of alluvial diamonds, con-

trasting with positive d15N-values displayed by in-

Ž .situ metamorphic diamonds fq5‰ to q8‰ .The present results further support the idea thatalluvial and in-situ diamonds do not derive from thesame protolith.

Moreover, it is of importance to notice that allu-vial diamonds cover a narrow range of d

15N-values.This suggests that alluvial diamonds are more likelyto derive from a single protolith as opposed to a

15 Žvariety of protoliths. Their d N-values y1.8‰ to.q1.1‰; meany0.5‰ contrast with moderate to

15 Ž .high d N-values of sediments f0‰ to q5‰ ,Ž 15metamorphosed crustal rocks strictly positive d N-

values; Haendel et al., 1986; Bebout and Fogel,.1992; Fig. 3 and in-situ metamorphic diamonds

specifically from this study. Accordingly, the ab-sence of clearly positive d

15N-values as those dis-played by in-situ diamonds seems to exclude anymetasedimentary nitrogen source. The absence of

15 Ždiamond with strictly positive d N-values e.g..q5‰ to q8‰ like in-situ diamonds is at odds the

fact that several metasedimentary diamond-bearingŽ 15 .rocks again with strictly positive d N-values have

likely contributed to alluvial diamond deposit. Wesuggest that microdiamonds from, e.g. garnet–pyrox-ene or marble rocks eventually contributed to theso-called alluvial diamond deposit but were not sam-pled because of their generally smaller size or weredestroyed during transport, their fibrous structuregiving them a more fragile character. An investiga-

Ž .tion of small size i.e. -40 mm alluvial diamondswould provide additional constraints.

Ž .Abdulkabirova and Zayachkovsky 1997 high-lighted the presence of diamondiferous lamproites inand around the Kokchetav massif. There is no dataavailable on diamond from these lamproites yet.Therefore, the relationship between the magmatic

Ž .and the metamorphic unit s remains unknown.Namely, it is unknown whether the diamonds in thelamproites could be for example mantle-derived orinherited from digested metamorphic rocks and in-corporated during emplacement. Nevertheless, con-sidering published data on diamonds from other lam-proites, several points can be drawn. In the presentcase, d13C- and d

15N-values do not allow to discrim-inate lamproitic from alluvial diamonds since the

Ž .field they define overlap not shown . However,diamonds found within lamproites show rather high

Žnitrogen aggregation state i.e. IaA-IaB diamonds;.Harris and Collins, 1985 relative to alluvial dia-

monds from the Kokchetav massif. These alluvialdiamonds also show rather high nitrogen contents

Ž .relative to kimberliticrlamproitic diamonds Fig. 2 .From these two evidences, we conclude that alluvialdiamonds are unlikely related to lamproites intruding

Ž .the massif Abdulkabirova and Zayachkovsky, 1997 .Considering that alluvial diamonds are metamor-

phic and derive from UHP rocks of the KokchetavŽmassif as suggested on the basis of nitrogen aggre-

.gation state; De Corte et al., 1998 , several hypothe-ses can be proposed to account for the rather lowd

15N-values. If one accepts that metamorphic de-volatilisation identified for sediments leads to a 15Nenrichment of q1‰ to q10‰ in the residueŽHaendel et al., 1986; Bebout and Fogel, 1992; Boyd

.and Philippot, 1998 , then the pre-metamorphic pro-tolith for alluvial diamonds requires a protolith withnegative or very close to zero d

15N. Such a lowd

15N-value would preclude any metasedimentary


Žsource for the nitrogen Haendel et al., 1986; Bebout.and Fogel, 1992; Boyd and Philippot, 1998 and

Ž .rather points towards to either 1 a mantle-derivedsource for nitrogen, as such a source is the only

15 Žreservoir known to provide low d N signatures see. Ž .Javoy et al., 1986; Cartigny et al., 1997 or 2 some

atmospheric nitrogen. In this specific case, if atmo-Ž .spheric nitrogen was trapped as fluid inclusions ?

and subducted before alluvial diamond growth, theŽ .noble gase in particular Ne signatures should also

be kept. The recognition of the protolith of alluvialdiamonds is however difficult and, given the presentstate of knowledge, the interpretation of associatedd

13C-values would be hazardous. At least, the con-tribution of any metasedimentary source seemsunlikely. The recognition of an olivine inclusion,typical of kimberlitic and lamproitic diamonds

Žworldwide Sobolev, Shatsky and Yefimova, unpub-.lished data within one alluvial diamond also points

towards a maficrultramafic protolith.In summary, the present result definitely supports

Ž .the idea that 1 Kokchetav alluvial diamonds are notderived from the most representative diamond-

Žbearing rocks found in the UHP massif i.e. garnet–. Ž .clinopyroxene rock and marble and 2 that these

alluvial diamonds are not representative of the dia-monds from the massif as a whole. This supports

Ž . Ž .previous conclusions of De Corte et al. 1999 . 3The rather low d

15N-values seem to reject the contri-bution of any metasedimentary nitrogen source.

6. Conclusions

The aim of this study has been to provide a firstnitrogen and carbon isotopic systematic of microdia-monds as a function of their host-rock in order tobetter constrain their origin and formation. In conclu-sion, the following seven points are emphasised.

Ž .1 A major proportion of the nitrogen presentwithin diamonds recovered from a garnet–clino-pyroxene rock and a marble, is likely present as

Ž .fluid-inclusions. 2 For these diamonds, carbon andnitrogen stable isotopes strongly suggest a metasedi-

Ž .mentrary origin. 3 Metamorphic microdiamondsmay not have crystallised at the peak of metamor-

Ž .phism i.e. 10008C but at temperatures f7008C.

Ž .4 Diamonds of AalluvialB type appear to containno significant amount of nitrogen as fluid inclusions,

Ž .and 5 their nitrogen isotopic composition suggestthat they may not derive from a metasedimentaryprotolith. An ultra-high pressure metamorphic rocksof mafic or ultramafic composition, i.e. differentfrom the other investigated diamond-bearing rocks,is possible.

Ž .6 When compared with diamonds from kimber-lites and lamproites, metamorphic diamonds haveunusually high concentrations of nitrogen.

Ž .7 The present set of data shows that it is possi-ble to subduct sedimentary nitrogen and carbon togreat depths.

Acknowledgements

We express special thanks to J. Harris and W.Taylor for comments and strong improvements to themanuscript. Comments by J. Klerkx, J. Hoefs and N.Jendrzejewski are also greatly acknowledged. PCbenefited from a grant by Naturalia and Biologiaduring acquisition of the results. KDC acknowledgesreceipt of a grant from the Flemish Institute forSupport of Scientific and Technological Research in

Ž .the Industry IWT . The authors wish to thank W.Griffin, R. Burgess and the anonymous for veryvaluable comments and reviews. IPGP contribution1720 and CNRS 249.

References

Abdulkabirova, M.A., Zayachkovsky, A.A., 1997. Diamond ofKazakhstan. In: Kazhegeldin, A.M., Abdulin, A.A., Bespaev,H.A., Votsalevsky, E.S., Daukeev, S.Z., Miroshnichenko, L.A.Ž .Eds. , Almaty, Ministry of Geology and Preservation ofUnderground Resources. p. 83.

Ader, M., Boudou, J.P., Javoy, M., Goffe, B., Daniels, E., Vieth-´Redemann, A., 1998. Isotope study of organic nitrogen of

Ž .Westphalian anthracites from eastern Pennsylvania USA andŽ .from Bramsche Massif Germany . Org. Geochem. 29, 315–

323.Andersen, T., Burke, E.A.J., Austraheim, H., 1989. Nitrogen

bearing-aqueous fluid inclusions in some eclogites from West-ern Gneiss region of the Norwegian caledonides. Contrib.Mineral. Petrol. 103, 153–165.

Bebout, G.E., 1995. The impact of subduction-zone metamor-


phism on mantle-ocean chemical cycling. Chem. Geol. 126,191–218.

Bebout, G.E., Fogel, M.L., 1992. Nitrogen-isotope compositionsof metasedimentary rocks in the Catalina Schist, California:implications for metamorphic devolatilization history.Geochim. Cosmochim. Acta 56, 2839–2849.

Bottinga, Y., 1969. Carbon isotope fractionation between graphite,diamond and carbon dioxide. Earth Planet. Sci. Lett. 5, 301–307.

Boyd, S.R., Philippot, P., 1998. Precambrian ammonium biogeo-chemistry; a study of the Moine metasediments. ScotlandChem. Geol. 144, 257–268.

Boyd, S.R., Rejou-Michel, A., Javoy, M., 1995. Improved tech-´niques for the extraction, purification and quantification ofnanomole quantities of nitrogen gas: the nitrogen content of adiamond. Meas. Sci. Technol. 6, 297–305.

Capdevilla, R., Arndt, N., Letendre, J., Sanvage, J.F., 1999.Diamonds in volcaniclastic komatite from French Guiana.Nature 399, 456–458.

Carlisle, D.B., Braman, D.R., 1991. Nanometre-size diamonds inthe CretaceousrTertiary boundary clay of Alberta. Nature352, 708–709.

Cartigny, P., Boyd, S.R., Harris, J.W., Javoy, M., 1997. Nitrogenisotopes in peridotitic diamonds from Fuxian, China; the man-tle signature. Terra Nova 4, 175–179.

Cartigny, P., Harris, J.W., Javoy, M., 1999. Diamond genesis,mantle fractionations and mantle nitrogen content: a study ofd

13C–N concentrations in diamonds. Earth Planet. Sci. Lett.Ž .submitted for publication Oct. 1999 .

Chopin, C., Sobolev, N.V., 1995. Principal indicators of UHP inŽ .crustal rocks. In: Coleman, R.G., Wang, X. Eds. , Ultra-High

Pressure Metamorphism. Cambridge Univ. Press, Cambridge,UK, pp. 96–131.

Chrenko, R.M., Tuft, R.E., Strong, H.M., 1977. Transformation ofthe state of nitrogen in diamond. Nature 270, 141–144.

Craig, H., 1953. The geochemistry of the stable carbon isotopes.Geochim. Cosmochim. Acta 3, 53–92.

Davies, G., 1976. The A nitrogen aggregate in diamond—itssymetry and possible structure. J. Phys. C: Solid State Phys. 9,L537–542.

De Corte, K., 2000. Study of microdiamonds from UHP metamor-Ž .phic rocks of the Kokchetav massif Northern Kazakhstan :

characterization and genesis, PhD thesis, University of Ghent,Belgium, 173 pp.

De Corte, K., Cartigny, P., Shatsky, V.S., Sobolev, N.V., Javoy,M., 1998. First evidence of fluid inclusions in metamorphicmicrodiamonds from the Kokchetav massif, Northern Kaza-khstan. Geochim. Cosmochim. Acta 62, 3765–3773.

De Corte, K., Cartigny, P., Shatsky, V.S., De Paepe, P., Sobolev,N.V., Javoy, M., 1999. Characteristics of microdiamonds fromUHPM rocks of the Kokchetav massif. In: Gurney, J.J., Gur-

Ž .ney, L.G., Pascoe, M.D., Richardson, S.H. Eds. , Proceedingsof the 7th International Kimberlite Conference, April 13–17,Cape Town, Red Roof Design cc, Goodwood, South-Africa.pp. 174–182.

Dobrzhinetskaya, L.F., Eide, E.A., Larsen, R.B., Sturt, B.A.,

Tronnes, R.G., Smith, D.C., Taylor, W.R., 1995. Microdia-mond in high-grade metamorphic rocks of the Western Gneissregion, Norway. Geology 23, 597–600.

Dunn, S.R., Valley, J.W., 1992. Calcite–graphite isotope ther-mometry: a test for polymetamorphim in marble, Tudor gab-bro aureole, Ontario, Canada. J. Metamorph. Geol. 10, 487–501.

Evans, T., Qi, Z., 1982. The kinetics of the aggregation ofnitrogen atoms in diamond. Philos. Trans. R. Soc. LondonA381, 159–178.

Evans, T., Kiflawi, I., Luyten, W., Van Tendeloo, G., Woods,G.S., 1995. Conversion of platelets into dislocation loops andvoidite formation in type IaB diamonds. Philos. Trans. R. Soc.London A449, 295–313.

Finnie, K.S., Fisher, D., Griffin, W.L., Harris, J.W., Sobolev,N.V., 1994. Nitrogen aggregation in metamorphic diamondsfrom Kazakhstan. Geochim. Cosmochim. Acta 58, 5173–5177.

Javoy, M., Pineau, F., Delorme, H., 1986. Carbon and nitrogenisotopes in the mantle. Chem. Geol. 68, 399–412.

Hahn-Weinheimer, V.P., 1965. Die isotopische Verteilung vonKohlenstoff und Schwefel in Marmor und anderen Metamor-phiten. Sonderdruck aus der Geologischen Rundschau Band55, 197–209.

Haendel, D., Muhle, K., Nitzsche, H., Stiehl, G., Wand, U., 1986.¨Isotopic variations of the fixed nitrogen in metamorphic rocks.Geochim. Cosmochim. Acta 50, 749–758.

Hanschmann, G., 1981. Berechnung von Isotopeeffeckten aufquantenchemischer grundlage am Beispiel stickstoffhaltigerMolekule. ZfI-Mitt. 41, 19–39.¨

Ž .Harris, J.W., 1992. Diamond geology. In: Field, J.E Ed. , TheProperties of Natural and Synthetic Diamonds. pp. 345–393.

Harris, J.W., Collins, A.T., 1985. Studies of Argyle diamonds.Ind. Diamond Rev. 3, 128–130.

Hoffbauer, R., Spiering, B., 1994. Petrologic phase equilibria andstable isotope fractionation of carbonate–silicate paragenesesfrom granulite-grade rocks of Sri Lanka. Precambrian Res. 66,325–349.

Honma, H., Itihara, Y., 1981. Distribution of ammonium inminerals of metamorphic and granitic rocks. Geochim. Cos-mochim. Acta 45, 983–988.

Hough, R.M., Gilmour, I., Pillinger, C.T., Arden, J.W., Gilkes,K.W.R., Yuan, J., Milledge, H.J., 1995. Diamond and siliconcarbide in impact melt rock from the Ries impact crater.Nature 378, 41–44.

¨Jones, R., Briddon, P.R., Oberg, S., 1992. First-principles theoryof nitrogen aggregates in diamond. Philos. Mag. Lett. 66,67–74.

Litvin, Y.A., 1998. Hot spots of the mantle and experiment to 10GPa: alkaline reactions, lithosphere carbonatization and newdiamond-generating systems. Geol. Geofiz. 39, 1772–1779,Ž .in Russian , English Translation: Russ. Geol. Geophys. 39,1761–1767.

Litvin, Y.A., Chudinovskikh, L.T., Zharikov, V.A., 1998. Crys-Ž . Ž .tallisation of diamond in the NaMg CO –K CO –C sys-3 2 2 3 2

tem at 8–10 GPa. Dokl. Earth Sci. 359A, 433–435.MacRae, N.D., Armitage, A.E., Jones, A.L., Miller, A.R., 1995. A


diamondiferous lamprophyre dike, Gibson Lake area, North-west Territories. Intern. Geol. Rev. 37, 212–229.

Marakushev, A.A., Paneyakh, N.A., Zotov, I.A., 1998. The prob-Žlem of mantle petrology. Geol. Geofiz. 39, 1750–1755 in

.Russian , English Translation: Russ. Geol. Geophys. 39,1740–1745.

Massonne, H.-J., 1999. A new occurrence of microdiamonds inquartzofeldspathic rocks of the Saxonian Erzgebirge, Ger-many, and their metamorphic evolution. In: Gurney, J.J.,

Ž .Gurney, J.L., Pascoe, M.D., Richardson, S.H. Eds. , Proceed-ings of the 7th International Kimberlite Conference, vol. 2.Red Roof Design, Cape Town, pp. 533–539.

Moine, B., Guillot, C., Gibert, F., 1994. Controls of the composi-tion of nitrogen-rich fluids originating from reaction withgraphite and ammonium-bearing biotite. Geochim. Cos-mochim. Acta 58, 5503–5523.

Pal’yanov, Y.N., Sokol, A.G., Borzdov, Y.M., Khokhryakov,A.F., Sobolev, N.V., 1998. Diamond crystallisation in the

Ž .systems CaCO –C, MgCO –C, and CaMg CO –C. Dokl.3 3 3 2

Earth Sci. 363, 1156–1159.Pal’yanov, Yu.N., Sokol, A.G., Borzdov, Yu.M., Khokhryakov,

A.F., Sobolev, N.V., 1999. Diamond formation from mantlecarbonate fluids. Nature 400, 417–418.

Pechnikov, V.A., Bobrov, V.A., Podkuyko, Y.A., 1993. Isotopiccompositions of diamond and accompanying graphite in NorthKazakhstan metamorphic rocks. Geol. Int. 30, 153–157.

Peters, K.E., Sweeney, R.E., Kaplan, I.R., 1978. Correlation ofcarbon and nitrogen stable isotope ratios in sedimentary or-ganic matter. Limnol. Oceanogr. 23, 598–604.

Pineau, F., Javoy, M., 1994. Strong degassing at ridge crests: thebehavior of dissolved carbon and water in basalt glasses at148N, Mid-Atlantic Ridge. Earth Planet. Sci. Lett. 123, 179–198.

Pineau, F., Latouche, L., Javoy, M., 1976. L’origine du graphite etles fractionnements isotopiques du carbone dans les marbres

Ž .metamorphiques des Gour Oumeladen Ahaggar, Algerie , des´ ´Ž . ŽAdirondacks New Jersey, USA et du Damara Namibie,

.Sud-Ouest africain . Bull. Soc. Geol. Fr. 18, 1713–1723.Polyakov, V.B., Kharlashina, N.N., 1995. The use of heat capacity

data to calculate carbon isotope fractionation between graphite,diamond and carbon dioxide: a new approach. Geochim. Cos-mochim. Acta 59, 2561–2572.

Rau, G.H., Arthur, M.A., Dean, W.E., 1987. 15 Nr14 N variationsin Cretaceous Atlantic sedimentary sequences: implication forpast changes in marine nitrogen biogeochemistry. Earth Planet.Sci. Lett. 82, 269–279.

Richet, P., Bottinga, J., Javoy, M., 1977. A review of hydrogen,carbon, oxygen, sulphur and chlorine stable isotope fractiona-tion among gaseous molecules. Annu. Rev. Earth Planet. Sci.5, 65–110.

Satish-Kumar, M., Wada, H., 1998. Carbon isotope fractionationbetween calcite and graphite during high temperature meta-morphism. Min. Mag. 62A, 1318–1319.

Scheele, N., Hoefs, J., 1992. Carbon isotope fractionation betweencalcite, graphite and CO : an experimental study. Contrib.2

Mineral. Petrol. 112, 35–45.

Shatsky, V.S., Sobolev, N.V., Yefimova, E.S., 1989. Morphologi-cal features of accessory microdiamonds from metamorphicrocks of the Earth’s crust. In: Boyd, F.R., Meyer, H.O.A.,

Ž .Sobolev, N.V. Eds. , Workshop on Diamonds. Extended Ab-stracts, Washington, DC, pp. 94–95.

Shatsky, V.S., Sobolev, N., Vavilov, V., 1995. Diamond-bearingmetamorphic rocks of the Kokchetav massif, northern Kaza-

Ž .khstan. In: Coleman, R.G., Wang, X. Eds. , Ultra-High Pres-sure Metamorphism. Cambridge Univ. Press, Cambridge, UK,pp. 427–455.

Shatsky, V.S., Rylov, G.M., Yefimova, E.S., De Corte, K.,Sobolev, N.V., 1998. Morphology and real structure of micro-

Ž .diamonds from metamorphic rocks Kokchetav Massif , kim-Žberlites and alluvial placers. Geol. Geofiz. 39, 942–955 in

.Russian , English Translation: Russ. Geol. Geophys. 39, 949–961.

Shatsky, V.S., Jagoutz, E., Sobolev, N.V., Kozmenko, O.A.,Parkhomenko, V.S., Troesch, M., 1999. Geochemistry and ageof ultrahigh pressure metamorphic rocks from Kokchetav mas-

Ž .sif Northern Kazakhstan . Contrib. Mineral. Petrol. 137, 185–205.

Sheppard, S.M.F., Schwarz, H.P., 1970. Fractionation of carbonand oxygen isotopes and magnesium between between coexist-ing metamorphic calcite and graphite. Contrib. Mineral. Petrol.26, 161–198.

Schidlowski, M., 1983. A 3800-million-year isotopic record of lifefrom carbon in sedimentary rocks. Nature 333, 313–318.

Sobolev, E.V., 1991. The impurity centres and some problems ofdiamond genesis. Ext. Abstr. Vol., Fifth Intern. Kimb. Conf.,Araxa, Brazil, 388–390.

Sobolev, N.V., Schatsky, V.S., 1990. Diamond inclusions ingarnet from metamorphic rocks: a new environment for dia-mond formation. Nature 343, 742–745.

Tagiri, M., Oba, T., 1986. Hydrothermal syntheses of graphitefrom bituminous coal at 0.5–5 kbar water vapor pressure and300–6008C. Jpn. Assoc. Mineral., J. Pet. Econ. Geol. 81,260–271.

Tanigushi, T., Dobson, D., Jones, A.P., Rabe, R., Milledge, H.J.,1996. Synthesis of cubic diamond in the graphite–magnesium

Ž .carbonate and graphite–K Mg CO systems at high pres-2 3 2

sure of 9–10 GPa region. J. Mater. Res. 11, 2622–2633.Taylor, W.R., Canil, D., Milledge, H.J., 1996. Kinetics of Ib to

IaA nitrogen aggregation in diamond. Geochim. Cosmochim.Acta 60, 4725–4733.

Valley, J.W., O’Neill, J.R., 1981. 13Cr12 C exchange betweencalcite and graphite: a possible thermometer in Grenvillemarbles. Geochim. Cosmochim. Acta 45, 411–419.

Veizer, J., Hoefs, J., 1976. The nature of 18 Or16O and 13Cr12 Csecular trends in sedimentary carbonate rocks. Geochim. Cos-mochim. Acta 40, 1387–1395.

Verchovsky, A.B., Begemann, U.O., 1993. Implanted radiogenicand other noble gases in crustal diamonds from NorthernKazakhstan. Earth Planet. Sci. Lett. 120, 87–102.

Wada, H., Suzuki, K., 1983. Carbon isotopic thermometry cali-brated by dolomite–calcite solvus temperatures. Geochim.Cosmochim. Acta 47, 697–706.


Williams, L.B., Ferrell, R.E., 1991. Ammonium substitution inillite during maturation of organic matter. Clays Clay Miner.39, 400–408.

Woods, G.S., 1986. Platelets and the infrared absorption of type Iadiamonds. Philos. Trans. R. Soc. London A407, 219–238.

Ž .Woods, G.S., 1994. Voidites in diamonds. In: Davies, G. Ed. ,Properties and Growth of Diamond. Institution of ElectricalEngineers, London, pp. 97–98.

Xu, S.T., Okay, A.I., Ji, S., Sengor, A.M.C., Su, W., Liu, Y.,Jiang, L., 1992. Diamond from the Dabie Shan metamorphicrocks and its implication for tectonic setting. Science 256,80–82.

Zhang, R.Y., Liou, J.G., Ernst, W.G., Coleman, R.G., Sobolev,N.V., Shatsky, V.S., 1997. Metamorphic evolution of dia-mond-bearing and associated rocks from the Kokchetav Mas-sif, northern Kazakhstan. J. Metamorph. Geol. 13, 479–496.

the origin and formation of metamorphic microdiamonds …cartigny/2001-chemgeol-cartignyetal.pdfthe...

Documents