Earth and Planetary Science Letters 357–358 (2012) 179–193

Contents lists available at SciVerse ScienceDirect

Earth and Planetary Science Letters journal homepage: www.elsevier.com/locate/epsl

Evidence for deep mantle convection and primordial heterogeneity from nitrogen and carbon stable isotopes in diamond M. Palot a,n, P. Cartigny a, J.W. Harris b, F.V. Kaminsky c, T. Stachel d a

Laboratoire de Ge´ochimie des Isotopes Stables, Institut de Physique du Globe de Paris, Universite´ Paris Diderot, CNRS UMR 7154, Sorbonne Paris-Cite´, 1 rue de Jussieu, 75238 Paris Cedex 05, France b School of Geographical and Earth Sciences, Gregory Building, University of Glasgow, Glasgow G12 8QQ, United Kingdom c KM Diamond Exploration Ltd., 2446 Shadbolt Lane, West Vancouver, British Columbia, Canada V7S 3J1 d Department of Earth and Atmospheric Sciences, University of Alberta, Edmonton, Canada T6G 2E3

a r t i c l e i n f o

abstract

Article history: Received 1 June 2012 Received in revised form 7 September 2012 Accepted 13 September 2012 Communicated by: B. Marty Available online 23 October 2012

Diamond, as the deepest sample available for study, provides a unique opportunity to sample and examine parts of the Earth’s mantle not directly accessible. In order to provide further constraints on mantle convection and deep volatile cycles, we analysed nitrogen and carbon isotopes and nitrogen abundances in 133 diamonds from Juina (Brazil) and Kankan (Guinea). Host syngenetic inclusions within these diamonds indicate origins from the lithosphere, the asthenosphere-transition zone and the lower mantle. Juina and Kankan diamonds both display overall carbon isotopic compositions within the current upper mantle range but the d13C signatures of diamonds from the asthenosphere-transition zone extend toward very negative and positive values, respectively. Two Kankan diamonds with both lower mantle and asthenosphere-transition zone inclusions (KK-45 and KK-83) are zoned in d13C, and have signatures consistent with multiple growth steps likely within both the lower mantle and the asthenosphere-transition zone illustrating the transfer of material through the 670 km seismic discontinuity. At a given locality, diamonds from the upper and the lower mantle show similar d15N distributions with coinciding modes within the range defined by typical upper mantle samples, as one might expect for a well stirred reservoir resulting from whole mantle convection. Kankan diamonds KK-11 (lower mantle), KK-21 and KK-92 (both lithospheric) display the lowest d15N values (-24.9%, -39.4% and -30.4%) ever measured in terrestrial samples, which we interpret as reflecting primordial heterogeneity preserved in an imperfectly mixed convective mantle. Our diamond data thus provide support for deeply rooted convection cells, together with the preservation of primordial volatiles in an imperfectly mixed convecting mantle, thereby reconciling the conflicting interpretations regarding mantle homogeneity derived from geochemical and geophysical studies. & 2012 Elsevier B.V. All rights reserved.

Keywords: mantle volatiles stable isotopes mantle convection diamonds primordial heterogeneity

1. Introduction The dynamic and compositional structure of the mantle is critical to understand the evolution of our planet. Mantle convection provides the central framework linking geochemistry, geophysics, mineral physics and geology. Noble gas and radiogenic lithophile element geochemistry of mid-ocean ridge basalts (MORB) and oceanic island basalts (OIB) lead to the suggestion that distinct reservoirs exist (e.g. Alle gre et al., 1986; Hilton and

n Corresponding author. Present address: Department of Earth and Atmospheric Sciences, University of Alberta, Edmonton, Canada T6G 2E3. E-mail address: [email protected] (M. Palot).

0012-821X/$ - see front matter & 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.epsl.2012.09.015

Porcelli, 2003; Hofmann, 2003; Ballentine et al., 2005; Kurz et al., 2009). The difference between chondrites and most terrestrial rocks in 142Nd/144Nd has been interpreted as resulting from early global differentiation of the Earth’s mantle composed of an incompatible element-depleted source of MORB and a complementary enriched reservoir located deeply within the mantle (e.g. Boyet and Carlson, 2005; O’Neil et al., 2008). This has been taken as evidence for layered mantle convection (often assumed to be separated at 670 km depth) but the size and depth of these reservoirs remain mostly unconstrained. Seismic tomography studies have shown that some slabs subduct through the 670 km seismic discontinuity down to the core–mantle boundary (e.g. Fukao et al., 1992; Grand, 1994; Van der Hilst et al., 1997; Masters et al., 2000; Lay, 2007), providing

Paragenesis

d13C

1-6 1-22 1-24 1-37 1-102 4-27 5-16 5-35 5-58 5-59 5-60 5-78 5 ¼80 5-105 5-109-2 5-109-3 BZ-208 BZ-211 BZ-212b BZ-212d BZ-214 1-35 4-4 5-55 BZ-213 BZ-124 BZ-127 BZ-129 BZ-209 BZ-215 BZ-217 1-1 1-2 1-5 1-8 1-30-1 1-30-4 1-32 1-34 4-7-2 4-7-3 4-101 4-102 5-107 BZ-237 1-4-3 1-4-4 1-7 1-31 1-33 1-36 1-38

Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Eclogitic? Lithospheric Lithospheric Lithospheric Lithospheric Asthe. B Asthe. B Asthe. B Asthe. B Asthe. B Asthe. B LM/TZ P LM/TZ P LM/TZ P LM/TZ P LM/TZ P LM/TZ P LM/TZ P LM/TZ P LM/TZ P LM/TZ P LM/TZ P LM/TZ B? LM/TZ B? LM/TZ B? LM P? LM P? LM P? LM P LM P LM P LM P

 5.0  4.4  2.8  5.1  4.4  4.2  5.6  7.8  4.9  3.9  4.8  4.6  6.2  4.4

P P P P

 5.2  13.0  8.5  8.5  11.6  5.5  3.5  4.3  3.4  12.3  8.4  6.3  12.1  8.8  8.6  4.7  4.5  4.3  4.2  4.9  10.3  4.2  4.2  9.7  4.7  4.6  4.6  4.3  5.1  4.0  5.3  4.4  4.7  4.3  5.4  4.2

d15N

 4.6

Error 7(2s)

1.0

 3.4

0.5

 3.3 þ1 þ 0.2 þ 1.2 þ 3.8 þ 1.3 þ 1.2  2.1

0.5 2.0 0.5 0.5 0.5 0.5 0.5 0.5

 5.8  8.8

0.5 0.5

 1.5

0.5

 4.4  5.6  4.0  0.8

0.5 0.5 0.7 0.5

 3.2

0.5

 1.0

0.5

[N]FTIR (ppm)

%B

87 0 0 75 17 0 20 70 0 0 0 0 61 0 56 30 14 586 21 507 0 68 0 n.d. 78 47 112 0 559 289 101 361 0 112 30 n.d. 25 0 0 0 0 0 0 40 148 124 410 0 41 0 20 0

98

97 90

[N]BULK (ppm)

14

100 69

16 100 38 100 92 17 92 100

48

27 64 79 91 71 54 100

30 510 77 66 750 166 62 331

95 97

61 125

86

34

95 97 100 98

42 166 26 477

97

74

52

20

Inclusions

Unknown Unknown Unknown Unknown Calcite, Fe-altered Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Ruby Spinel Spinel Spinel cpx Majorite, Cpx Majorite Majorite cpx Majorite Majorite Ilmenite Ilmenite Ilmenite Ilmenite Ilmenite Ilmenite Ilmenite Ilmenite Ilmenite Ilmenite CaSiO3, olivine SiO2 SiO2 MgSiO3 (Al), majorite fPer, ilmenite fPer, ilmenite fPer, ilmenite fPer fPer fPer fPer

Main characteristics of inclusion chemistry

Hi-Cr, Lo–Al Hi-Ti,Fe, Lo–Cr Hi-Ti,Fe, Lo–Cr Hi-Ti,Fe, Lo–Cr Hi-Cr Hi-Si at.pfu, Lo-Cr Hi-Si at.pfu, Lo-Cr Hi-Si at.pfu, Lo-Cr Lo-Cr Hi-Si at.pfu, Lo-Cr Hi-Si at.pfu, Lo-Cr Hi-Mn, Lo-Mg Hi-Mn, Lo-Mg Hi-Mn, Lo-Mg Hi-Mn, Lo-Mg Hi-Mn, Lo-Mg Hi-Mn, Lo-Mg Hi-Mn, Lo-Mg Hi-Mn, Lo-Mg Hi-Mn, Lo-Mg Hi-Mn, Lo-Mg Hi-traces (Ca–Si), Lo-Fo (ol.) almost pure SiO2 almost pure SiO2 Hi-Al, Lo-Ni (Mg–Si), Hi-Si at.pfu Hi-Fe (fPer), Hi-Mn, Lo-Mg (ilm.) Hi-Fe (fPer), Hi-Mn, Lo-Mg (ilm.) Hi-Mn, Lo-Mg (ilm.) Lo-Fe Lo-Fe Lo-Fe Lo-Fe

M. Palot et al. / Earth and Planetary Science Letters 357–358 (2012) 179–193

Sample

180

Table 1 Chemical and mineralogical characteristics of Juina diamonds. This includes measurements of d13C by mass spectrometry in dual inlet mode, d15N and nitrogen content (Nbulk) by mass spectrometry in static mode, [N]FTIR and % of B defect by infrared spectroscopy for Juina diamonds. Errors on d13C are better than 70.1% (2s), and 7 0.5% (2s) for d15N except when blank contributed to more than 20%, in which case a Monte Carlo correction was applied.

0.5 2.6

0.5

1.5

0.5

 2.5  2.8

 1.5

 8.2

 3.1

Abbreviations: P ¼‘‘peridotitic’’, B¼ ‘‘basaltic composition’’, ol. ¼olivine, fPer¼ ferropericlase, cpx¼ clinopyroxene, grt ¼garnet, TAPP ¼Tetragonal Almandine-Pyrope Phase, MgSiO3 (Al) represents aluminium rich magnesian perovskites, Hi¼high, Lo¼ low, Mg-Si¼ MgSiO3-silicate, Ca–Si ¼CaSiO3-silicate, Fo ¼forsterite and LM/TZ boundary¼ limit between the Lower Mantle and the Transition Zone. Inclusions compositions derived from Hutchison et al. (1999) and Kaminsky et al. (2001).

36 100

8 99

13 66

63 13 86 100

0.5

3-1 3-4 3-101 4-18 4-33 5-17 5-103 5-104 BZ-88 max BZ-88-2 BZ-88-3 BZ-201 BZ-205 BZ-206 BZ-207 BZ-210 BZ-241 BZ-242

LM LM LM LM LM LM LM LM LM LM LM LM LM LM LM LM LM LM

P P P P P P P P? P P P or Asthe. B? P P P P P P P

 4.9  2.8  4.4  3.0  3.9  4.9  5.1  4.2  7.0  9.6  12.8  4.8  4.6  4.7  4.5  4.6  5.3  5.2

 3.7

325 0 0 0 0 83 40 0 0 52 0 15 0 0 13 0 0 0

100

806

fPer fPer fPer fPer fPer fPer, spinel 2 fPer 3 CaSiO3, Ni native fPer fPer fPer fPer 2 fPer, TAPP fPer, TAPP fPer, MgSiO3, TAPP fPer, MgSiO3 (Al) fPer, ruby, 2 MgSiO3 (Al) fPer, MgSiO3 (Al)

Hi-Fe Lo-Fe Lo-Fe Lo-Fe Lo-Fe Lo-Fe Hi-Fe Hi traces Lo-Fe Lo-Fe Lo-Fe Hi-Fe Lo-Fe Lo-Fe Hi-Fe (fPer), Lo-Ni (Mg-Si) Lo-Fe (fPer), Hi-Al, Lo-Ni (Mg-Si) Lo-Fe (fPer), Lo-Cr (ruby), Lo-Ni (Mg-Si) Lo-Fe (fPer), Hi-Al, Lo-Ni (Mg-Si)

M. Palot et al. / Earth and Planetary Science Letters 357–358 (2012) 179–193

181

evidence for deeply rooted convection cells, and casting doubt on the preservation of deep primordial reservoirs. The motivation still persists to reconcile the geochemical inference of distinct reservoirs in the mantle and geophysical evidence for whole mantle convection. This leads to alternative models of convection (for review see Tackley, 2007). Samples from the deep mantle ( 4300 km) are extremely scarce (e.g. Sautter et al., 1991). Diamonds derived from the base of the lithosphere (Stachel and Harris, 2008 for review) down to the lower mantle (Wilding, 1990; Moore et al., 1991; Harte and Harris, 1994; Harte et al., 1999; Hutchison et al., 1999, 2001; Stachel et al., 2000a, b, 2002, 2005; Kaminsky et al., 2001; Hayman et al., 2005; Tappert et al., 2005, 2009; Kaminsky, 2012) are resistant to high temperatures and can survive exhumation to Earth’s surface. They contain both inclusions that are similar to typical upper mantle rocks (peridotite and eclogite) (Stachel and Harris, 2008 for review), and that are expected for the transition zone and the lower mantle (Stachel et al., 2002 and Kaminsky, 2012 for review). Diamonds are therefore the deepest samples available providing a unique opportunity to study parts of the Earth’s mantle which normally are inaccessible. The present paper focusses on nitrogen and carbon isotopes and nitrogen abundances of 133 diamonds from Juina (Brazil) and Kankan (Guinea) derived from the lithosphere, the asthenospheretransition zone and the lower mantle and discusses these new data in the framework of deep mantle convection.

2. Mineral inclusions in diamonds Most diamonds ( 499%) are formed in thick continental lithosphere at  150-250 km depth, but diamonds recovered especially from Juina (Brazil) (Hutchison et al., 1999, 2001; Harte, 1999, 2010; Kaminsky et al., 2001; Hayman et al., 2005; Kaminsky, 2012) and Kankan (Guinea) (Stachel et al., 2000a, b, 2002) show, through their inclusion content, formation not only in the lithosphere, but also in the asthenosphere-transition zone (250–670 km) and in the upper regions of the lower mantle ( 4670 km depth). Syngenetic inclusions recovered in our diamonds (Tables 1 and 2) have been extensively discussed by Hutchison et al. (1999, 2001), Harte et al. (1999), Kaminsky et al. (2001) and Harte (2010) for Juina samples, and by Stachel et al. (2000a, b) for Kankan samples. The following paragraph is a summary of the main assemblages recovered in our diamonds (Tables 1 and 2). The suite of mineral inclusions comprises majoritic garnets with basaltic-like compositions formed over most of the asthenosphere-transition zone (Wilding, 1990; Harte, 1992, 2010). Ferropericlase (fPer) inclusions [(Mg, Fe)O], associated with tetragonal pyrope-almandine phase (TAPP), MgSiperovskite (with low or high Al2O3 values), or CaSi-perovskites (pvk), are symptomatic of a formation in a ‘‘peridotitic’’ lower mantle (Harte et al., 1999; Stachel et al., 2000b; Kaminsky et al., 2001; Harte, 2010). Mg-perovskites with low-Al contents (Kankan) are formed in the topmost  20 km of the lower mantle, whilst Al-rich inclusions at Juina indicate the uppermost 100 km of the lower mantle (Stachel et al., 2005; Harte, 2010). Harte (2010) argues, that based on present evidence, diamonds do not form below 800 km depth. Single ferropericlase represents the most common lower mantle inclusion in Juina and Kankan diamonds. Their occasional occurrence together with olivine inclusions indicates that in some rare instances ferropericlase could derive from dunites within the upper mantle (Brey et al., 2004). However, the restriction of ferropericlase inclusions in diamonds to localities where diamonds with lower mantle assemblages occur, and the similarities

Paragenesis

d13C

KK-6 KK-10 KK-19 KK-22 KK-46 KK-47 KK-50 KK-51 KK-52 KK-53 KK-59 KK-76 KK-3 KK-4 KK-40 KK-67 KK-68 KK-75 KK-77 KK-80 KK-86 KK-21 KK-26 KK-28 KK-49 KK-65 KK-70 KK-73 KK-78 KK-79 KK-90 KK-92-1 KK-92-2 KK-93 KK-7 KK-9 KK-11 KK-13 KK-16 KK-23 KK-29 KK-30 KK-31 KK-32 KK-35 KK-36 KK-37 KK-39-5 KK-39-1 KK-43 KK-45-1 KK-45-2 KK-48 KK-54

Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Lithospheric Lithospheric Lithospheric Lithospheric Lithospheric Lithospheric Lithospheric Lithospheric Lithospheric Lithospheric Lithospheric Lithospheric Lithospheric Lithospheric Lithospheric Lithospheric Lithospheric Lithospheric Lithospheric Lithospheric Lithospheric Lithospheric LM P LM P or B LM P LM P LM P LM P LM P LM P LM P LM P or B LM P LM P LM P LM P LM P LM P LM B LM B LM P LM P

 2.7  3.1 þ1.2

 4.1

d15N

 2.9

Error 7 (2s)

0.5

 0.2

 0.2 E E E E E E E E E P P P P P P P P P P P P P

 4.8  0.8  5.1  9.2  10.4  4.8  2.0  4.7  4.8  3.4  4.1  4.8  4.8  2.4  3.8  2.9  2.8  4.8  4.6  3.2  3.2  4.4  3.7  3.9  3.5  3.5  2.8  4.4  3.3  3.5  3.9  3.5  4.5  4.0  3.5  1.1  1.4  3.7  2.8 þ1.4  3.7  3.3

 4.7 þ8.5  0.3 þ7.8  4.1 -11.2 þ0.6  7.0  4.3 –39.4 þ4.9  0.2  4.9 þ9.6  5.1 þ5.8 þ1.0 þ2.4 þ4.5

0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 1.5 0.5 0.5 0.5 0.5

–30.4  1.2  4.8

0.5 0.5 0.5

–24.9

0.5

 1.3

1.0

 1.0  6.5  2.7 þ4.6  3.0

0.7 1.6 2.3 3.0 1.0

 1.3

0.9

[N]FTIR (ppm) 0 17 0 397 356 1196 125 0 866 31 0 683 922 166 450 1047 1651 403 131 464 956 123 69 34 262 180 32 36 197 0 255 115 418 906 0 0 20 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5

%B

[N]BULK (ppm)

987

4 0 0 19 90 9 4 0 9 24 14 26 18 25 7 16 13 9 19 18 12 28 71 26

100

Main characteristics of inclusion chemistry

Unknown Unknown Unknown

100 37 99 0 95

Inclusions

873 279 481 1193 1292 499 134 907 1151 296 81 107 426 163 10 21 80 38 295 251 794 47 115 0 0 0

8 0 7 2 2 4 4 0

3

Altered Sulphide unknown 3 altered Unknown Unknown Altered Unknown cpx grt cpx Coesite 2 rutiles cpx 2 cpx, altered 2 cpx grt 2 olivines grt, 2olivines grt, 2 olivines, cpx olivine grt, 2 ol., cpx-opx 2 grt, 2 chromites 2 olivines grt, 2 olivines cpx Chromite, altered 2 olivines 2 olivines Chromite fPer CaSiO3 fPer 2 fPer 2 fPer, MgSiO3, Siderite fPer fPer fPer fPer, SiO2 CaSiO3 fPer 2 fPer fPer fPer fPer fPer MgSiO3, K-fsp MgSiO3, K-fsp fPer fPer

Hi-Ca, Lo-Cr Hi-Ca, Lo-Cr almost pure SiO2 almost pure TiO2 Hi-Ca, Lo-Cr Hi-Ca, Lo-Cr Hi-Ca, Lo-Cr Hi-Ca, Lo-Cr Lo-Fo Hi-Cr (grt), Lo-Fo (ol) Hi-Cr (grt), Lo-Fo (ol), Lo-Na (cpx) Lo-Fo Hi-Cr (grt), Lo-Fo (ol.), Lo-Ca (cpx) Hi-Cr (grt), Hi-Ti (chromite) Lo-Fo Hi-Cr (grt), Lo-Fo (ol) Lo-Na (cpx) Hi-Fe,Mn, Lo-Al Lo-Fo Lo-Fo Hi-Fe,Mn, Lo-Al Lo-Fe Hi-traces Lo-Fe Lo-Fe Lo-Fe (fPer), Lo-Ni (Mg–Si) Lo-Fe Lo-Fe Lo-Fe Lo-Fe Hi-traces Lo-Fe Lo-Fe Lo-Fe Lo-Fe Lo-Fe Lo-Fe Lo-Ni (Mg–Si) Lo-Ni (Mg–Si) Lo-Fe Lo-Fe

M. Palot et al. / Earth and Planetary Science Letters 357–358 (2012) 179–193

Sample

182

Table 2 Chemical and mineralogical characteristics of Kankan diamonds. Measurements of d13C by mass spectrometry in dual inlet mode, d15N and nitrogen content (Nbulk) by mass spectrometry in static mode, [N]FTIR and % of B defect by infrared spectroscopy for Kankan diamonds. Errors on d13C are better than 7 0.1% (2s), and 7 0.5% (2s) for d15N except when blank contributed to more than 20%, in which case a Monte Carlo was applied.

1.0 1.5

1.0

2.0

0.6

0.7

 3.4  0.1

 2.8

 8.2

þ2.6

 3.9

Abbrevations: P ¼ ‘‘peridotitic’’, E ¼ ‘‘eclogitic’’, B ¼‘‘basaltic composition’’, webste. ¼ websteritic, ol.¼ olivine, fPer¼ ferropericlase, cpx¼ clinopyroxene, grt ¼ garnet, K-fsp ¼ potassic feldspar (primary hollandite?), amph¼ amphibole, Hi¼ high, Lo¼ low, Mg–Si¼ MgSiO3-silicate, Ca–Si¼ CaSiO3-silicate, Fo ¼forsterite. Inclusions compositions derived from Stachel et al. (2000a, b).

10 100

0

6 100

6

1.0 1.2

KK-57 KK-63 KK-66 KK-71 KK-74 KK-82 KK-83 KK-83-1 KK-83-2 KK-87 KK-88 KK-89 KK-94 KK-95 KK-103 KK-104 KK-108 KK-84 KK-1 KK-5 KK-61 KK-61-1 KK-61-2

LM P LM P LM P LM P LM P LM P LM P LM P LM P LM P LM P LM P LM P LM P LM P LM P LM P dunitic Asthe. B Asthe. webste. Asthe. webste. Asthe. webste. Asthe. webste.

 4.0  3.2  4.1  3.4  4.4  2.6  3.4 þ0.1  3.5  4.6  3.0  3.5  3.5  3.6  4.0  3.9  3.5  3.3 þ0.9 þ0.6 þ0.5 þ3.8  2.3

þ0.8 þ0.8

0 0 0 9 27 0 0 0 0 0 0 6 0 0 10 0 0 0 0 0 0 0 0

95 95

4 3 0 4 8 0 4

fPer fPer 2 fPer, 2 CaSiO3 fPer fPer fPer fPer, 3 cpx, ‘‘ol’’ Fe fPer, 3 cpx, ‘‘ol’’ Fe fPer, 3 cpx, ‘‘ol’’ Fe fPer, CaSiO3 fPer fPer 2 fPer 2 fPer fPer, MgSiO3 fPer 2 fPer, 2 MgSiO3, amph fPer-ol., fPer, ol. majorite, cpx majorite 2 majorites 2 majorites 2 majorites

Lo-Fe Lo-Fe Lo-Fe (fPer), Hi-traces (Ca–Si) Lo-Fe Lo-Fe Lo-Fe Lo-Fe (fPer), Hi-Mg,Sr (cpx), Lo-Mg (ol) Lo-Fe (fPer), Hi-Mg,Sr (cpx), Lo-Mg (ol) Lo-Fe (fPer), Hi-Mg,Sr (cpx), Lo-Mg (ol) Lo-Fe (fPer), Hi-traces (Ca-Si) Lo-Fe Lo-Fe Lo-Fe Lo-Fe Lo-Fe (fPer), Lo-Ni (Mg–Si) Lo-Fe Lo-Fe (fPer), Lo-Ni (Mg–Si) Lo-Fe (fPer) Hi-Si at.pfu, Hi-Ca, Lo–Cr Hi-Si at.pfu, Lo-Ca,Cr Hi-Si at.pfu, Lo-Ca,Cr Hi-Si at.pfu, Lo-Ca,Cr Hi-Si at.pfu, Lo-Ca,Cr

M. Palot et al. / Earth and Planetary Science Letters 357–358 (2012) 179–193

183

of d13C–d15N-[N] and aggregation states of nitrogen between diamonds containing ferropericlase only and diamonds of proven lower mantle origin (i.e. assemblages of ferropericlaseþCaSiO3pvk or MgSiO3-pvk) indicate that most if not all ferropericlase inclusions in diamonds derived from lower mantle (Harte et al., 1999; Kaminsky et al., 2001; Stachel et al., 2000b, 2005; Hayman et al., 2005; Kaminsky, 2012).

3. Analytical methods 3.1. Nitrogen content and aggregation state Samples were previously broken to release their mineral inclusion(s). Nitrogen contents (NFTIR) and aggregation states (%B) of Juina (Table 1) and Kankan (Table 2) diamonds were obtained by micro Fourier Transform Infrared (FTIR) Spectroscopy on 1.0–3.0 mm diamond fragments, many of which were cleavage slices. Detection limit is about 10 ppm of nitrogen. Spectra (650– 4000 cm  1) were accumulated over 300 scans with a resolution of 4 cm  1 and an aperture size of 100 mm using a Nicolet MagnaIR 550 Spectrometer (bench) fitted with a continuum infrared microscope. Using the absorption coefficients for the nitrogen A- and B-centres of Boyd et al. (1994, 1995a), N-abundance were calculated after spectral decomposition using a least-square method. The total uncertainty is better than 10% for N-abundance and accuracy is about 73% (2s) for aggregation state. Aggregation of nitrogen, expressed as percentage of B-defects in diamond (i.e. cluster of four nitrogen atoms surrounding a vacancy), is a second order kinetic diffusion process, depending upon nitrogen abundance, time-integrated mantle storage temperature (T) and mantle residence time (t) (Chrenko et al., 1977):   1=½Adefect1=½N ¼ A expEa=RT t N (total nitrogen¼[A-defect] þ[B defect]) and nitrogen in A-defects are in atomic ppm, t (time) is in s, T is in Kelvin, A¼294,000 ppm/s, R¼8.317 kJ/mol/K and Ea (activation energy)¼ 7 eV (Cooper, 1990). 3.2. Nitrogen isotope composition and bulk N-content Following FTIR analyses, diamond fragments (0.3–4.0 mg) were combusted in a quartz crucible externally heated at 1100 1C in a pure O2 atmosphere (Boyd et al., 1995b). The produced CO2 was then trapped using CaO (to produce calcium carbonate). Trace amounts of CO, hydrogen, CH4 or H2O in the nitrogen gas that may cause isobaric interferences were removed chemically after diamond combustion through a double purification with Cu–CuO and CaO held at 600 1C and 450 1C, respectively (Boyd et al., 1995b). After quantification of nitrogen amounts using a capacitance manometer with a precision better than 5% (2s) and a detection limit of 0.01 nmol N2, nitrogen was analysed using a purpose built triple collector static mass spectrometer with an accuracy better than 70.5% (2s) established by the analysis of the international IAEA-N1 and -N2 standards (Table 1A). Traces of carbonaceous and air components have been monitored measuring the intensity on the masses 12 (12C þ ) and 40 (40Ar þ ), and the half-life of m/z¼28 and 29. The occurrence of CO, CO2 and CH4 leading to increased 29 half-life. Typical nitrogen blanks were 0.3770.09 nmol N2 (1s). Blanks were measured before each diamond analysis and a Monte Carlo method was applied for error estimation when the blank contribution was higher than 20%. When it was possible, several fragments of a single diamond were analysed. Individual analyses are listed (Tables 1 and 2) for internal isotopic heterogeneity 42%.

184

M. Palot et al. / Earth and Planetary Science Letters 357–358 (2012) 179–193

3.3. Carbon isotope composition The previously trapped CO2 gas (see above) was recovered after heating the CaO to 850 1C (Boyd et al., 1995b) and carbon isotopes determined using a conventional dual-inlet mass spectrometer (Finnigan delta XP þ ) with an accuracy better than70.1%(2s). Diamonds with no nitrogen detected during FTIR spectroscopy were directly combusted for d13C analysis using a dedicated extraction line. Values of d13C for each diamond in Tables 1 and 2 provide average d13C values measured on several fragments. For samples with an internal isotopic heterogeneity41.3% (i.e. average variability within individual diamonds worldwide, Cartigny, 2005) individual analyses are listed (Tables 1 and 2). After recognising that some diamonds are isotopically heterogeneous, for a number of large diamonds (BZ-205 to BZ-207, BZ-210, 1-4, 5-17, KK-16, KK-61, KK-83, KK-84 and KK-108) multiple fragments were analysed using an elemental analyser (Thermo Flash 1112 series) coupled with a mass spectrometer Finnigan MAT 253 in continuous flow mode. This new technique allows fast analysis of small diamond chips (o0.5 mg) with an accuracy of 70.1% (2s). Combustion of diamonds occurred in an oxidation column (chromium oxide) at 1020 1C during 20 s. Gases released were purified in a copper section and separated in Porapak QS gas chromatograph and introduced into the mass spectrometer with helium as a carrier. Internal standards used for the calibration of d13C consisted of graphite and diamond powder whose carbon isotopic composition were determined using conventional techniques at d13C¼-23.3170.03% (1s) and d13C¼-8.4270.24% (1s), respectively.

4. Results 4.1. Nitrogen abundance and aggregation state FTIR analyses of 62 Juina diamonds indicate that 32 contain nitrogen above the detection limit ( 410 ppm, i.e. Type I) ranging from 13 to 586 ppm with a median of 69 ppm (Table 1). The proportion of diamonds with no detectable nitrogen in FTIR ( o10 ppm, i.e. Type II) at Juina is 47% among diamonds of unknown paragenesis (N ¼9 out of 19, Unknown), 33% among lithospheric diamonds (N ¼1 out of 3, Litho.), 39% among asthenospheric-transition zone diamonds (N ¼7 out of 18, Asthe.) and 59% among lower mantle samples (N ¼13 out of 22, L.M). The proportion of Type I diamonds is somewhat higher compared to previous studies (Hutchison et al., 1999; Kaminsky et al., 2001) and reflects the fact that we analysed almost every piece of material available. Our samples range from 17% to 100% of B defects, with a median of 96% (Table 1, Fig. 1). The proportion of fully aggregated nitrogen (B 495%) is 50% (N ¼5 out of 10, Unknown), 50% (N¼ 1 out of 2, Litho.), 45% (N ¼5 out of 11, Asthe.) and 67% (N¼ 6 out of 9, L.M), in good agreement with previous studies. Diamonds 1–36 and BZ-88 containing lower mantle inclusions display relatively low nitrogen aggregation states of 52% and 66%, respectively. Of 77 Kankan diamonds analysed by FTIR, 35 are Type I ranging from 5 to 1651 ppm with a median of 188 ppm (Table 2). The proportion of Type II diamonds at Kankan is 33% (N ¼4 out of 12, Unknown), 5% (N ¼1 out of 22, Litho.), 100% (N ¼6 out of 6, Asthe.) and 84% (N ¼31 out of 37, L.M), which again is slightly higher compared to previous study (Stachel et al., 2002). Aggregation states range from 0% to 100% with a median of 19% (Table 2, Fig. 1). The percentage of diamonds with fully aggregated nitrogen is 37% (N ¼3 out of 8, Unknown), 0% (N ¼0 out of 21, Litho.) and 100% (N ¼5 out of 5, L.M).

Among both the Juina and Kankan sample sets, diamonds having inclusion of ‘‘basaltic’’ and ‘‘peridotitic’’ composition do not display distinct [N] and %B signatures. 4.2. Bulk nitrogen abundance and isotopic composition The d15N of 62 diamonds from Juina (N¼ 24) and Kankan (N¼ 38) derived from the lithosphere, the asthenosphere-transition zone and the lower mantle, related to either ‘‘basaltic’’ or ‘‘peridotitic’’ compositions are presented in Tables 1 and 2, respectively. Juina diamonds define a narrow range in d15N from 8.8% to þ 3.8% (Table 1) with a median of  3.1% (Fig. 2). Diamonds from the lithosphere, the asthenosphere-transition zone and lower mantle-transition zone boundary (but still within the upper mantle) and the lower mantle at Juina display similar d15N median values close to  3% (Fig. 2). Diamonds with asthenospheric-transition zone inclusions of ‘‘basaltic’’ composition have all positive d15N values. Bulk N-abundances are 30–750 ppm (median 71 ppm), and 8–806 ppm (median 31 ppm) in diamonds from the upper and lower mantle, respectively (Table 1). At Kankan a much larger d15N range is observed, from -39.4% (the lowest value ever measured in a terrestrial sample) to þ9.6%, with a median of  1.3% (Table 2, Fig. 2). Diamonds from the lithosphere and lower mantle at Kankan (asthenospheric-transition zone samples were below the detection limit) display similar negative d15N median values (Fig. 2). In contrast to Juina, at Kankan there is no obvious difference in d15N between diamonds of ‘‘peridotitic’’ and ‘‘basaltic’’ paragenesis. Bulk nitrogen abundance are 10–1292 ppm (median 296 ppm), and 0  115 ppm (median 4 ppm) in diamonds from the lithosphere and the lower mantle, respectively (Table 2). Overall nitrogen abundances determined by FTIR are generally comparable to those obtained by bulk combustion for both Juina and Kankan diamonds (Tables 1 and 2), suggesting a fairly homogeneous distribution of nitrogen within most of our diamonds. At a given locality, diamonds from the upper and lower mantle show very similar d15N distributions with coinciding modes within the range defined by typical upper mantle samples worldwide (Fig. 2). 4.3. Carbon isotopic composition Further d13C analyses of 122 diamonds produced ranges from 12.3% to  2.8% (median 4.8%) at Juina (N ¼55, Table 1), and 10.4% to þ3.8% (median  3.5%) at Kankan (N ¼67, Table 2), and compare well with previous measurements on these diamonds (Hutchison et al., 1999; Kaminsky et al., 2001; Stachel et al., 2002). Diamonds derived from the lithosphere and the lower mantle display similar d13C distributions centred at  4.6% at Juina and  3.5% at Kankan (Fig. 3). The d13C distributions of asthenospheric-transition zone diamonds at Juina (N¼ 6) and Kankan (N ¼5) contrast through distinct tails extending toward highly negative (down to 12.3%, Juina), and positive values (up to þ3.8%, Kankan) (Fig. 3). Diamonds 1–30, 4–7 (Mn-rich ilmenite) and BZ-88 (fPer) are zoned in d13C by more than 5.0% (Table 1), whilst KK-45 (MgSiO3 þ ‘‘K-fsp’’) and KK-83 (fPerþ 3cpxþnative Fe-olivine) display variations 43.6% (Table 2). These variations cannot be placed into a core-rim stratigraphic context as we studied fragments from broken diamonds. 5. Discussion Diamonds from Juina and Kankan have been subject of a number of previous studies (see reviews by Kaminsky, 2012 and

M. Palot et al. / Earth and Planetary Science Letters 357–358 (2012) 179–193

185

Fig. 1. Nitrogen content (atomic.ppm) and nitrogen aggregation state (%B) of Juina (squares) and Kankan (triangles) diamonds from the lithosphere (  150–250 km depth), the asthenosphere-transition zone (250–670 km depth) and the lower mantle ( 4670 km depth). Isotherms are based on the second order kinetic law for nitrogen diffusion (Chrenko et al., 1977) assuming a mantle residence of 100 Ma. Total errors of N-abundance and aggregation state of nitrogen are indicated by error bars and are better than 10% and 7 3% (2s), respectively.

Fig. 2. Nitrogen isotopic compositions of Juina and Kankan diamonds. Data are expressed as per mil deviation relative to the composition of air (Mariotti, 1983) with a typical error of 70.5% (2s). Upper mantle samples refer to diamonds from the lithosphere and the asthenosphere-transition zone, while lower mantle samples refer to diamonds from below 670 km depth. Upper mantle and lower mantle diamonds from Juina (left) are within the range of the current upper mantle (see text for references) illustrated by the shaded area. At Kankan, upper and lower mantle also display similar ranges of d15N (right). The three lowest nitrogen isotope values ever recorded in terrestrial samples are circled.

Stachel et al., 2005). From these studies, a general consensus has emerged that deep diamond formation in the asthenosphere and lower mantle is related to subducted crustal material. This conclusion is based on evidence such as extreme LREE and Sr enrichment and presence of slight Eu-anomalies in former Ca-perovskite inclusions (Harte et al., 1999; Stachel et al., 2000b), the ‘‘basaltic’’ composition of majorite inclusions (Stachel et al., 2000a; Tappert et al., 2005), the non-mantle like C-isotope compositions of the host diamonds (Hutchison et al., 1999; Stachel et al., 2002; Tappert et al., 2005, 2009; Walter et al., 2011) and the discovery of unusual inclusions at Juina interpreted

to represent primary phases of basaltic bulk composition at lower mantle depths (Walter et al., 2011). Harte (2010) suggested that asthenospheric-transition zone and lower mantle diamonds, down to 800 km depth, are linked to dehydration of wadsleyiteringwoodite in subducting lithospheric mantle. However, d13C and d15N of two lower mantle diamonds at Juina (Hutchison et al., 1999) were within the range defined by typical mantle samples casting doubt that super deep diamonds are all related to crustal material. In the following section, we will discuss further d13C–d15N-N analyses in Juina and Kankan diamonds and their implications for

186

M. Palot et al. / Earth and Planetary Science Letters 357–358 (2012) 179–193

Fig. 3. Carbon isotopic compositions of Juina and Kankan diamonds. Data are expressed in % deviation relative to Vienna-Pee Dee Belemnite with a typical error better than 7 0.1% (2s). Data are grouped into three depths origin. Asthenospheric-transition zone diamonds from Juina and Kankan are characterised by general depletion and enrichment of 13C, respectively, compared to the mantle range (see text for references).

the formation of sublithospheric diamonds, mantle convection and deep mantle volatile cycle. 5.1. Origin of carbon and nitrogen involved in the formation of Juina and Kankan diamonds. Co-variations in d13C–d15N constrain the origin of Juina diamonds to a mixing process rather than a high temperature isotopic fractionation process (Fig. 4). The first end-member (1 in Fig. 4), defined by the mean of diamonds BZ-209 and BZ-124, displays d15N¼ þ2.3%, d13C¼  12.2% and N/C¼ 592 ppm. These characteristics are consistent with recycled material dominated by organic compounds (Haendel et al., 1986, Bebout and Fogel, 1992, Ader et al., 1998, Cartigny et al., 2001a, 2004). The second end-member (2 in Fig. 4), defined by the mean of diamonds 1-8 and BZ-201, displays values of d15N¼ 8.5%, d13C¼  4.4% and N/C¼67 ppm, consistent with mantle-derived samples which are close to -5% in average (Javoy et al., 1986, Boyd et al., 1987, Javoy and Pineau, 1991, Marty and Humbert, 1997, Marty and Zimmermann, 1999, Cartigny et al., 2001b, c, Cartigny, 2005). Note that the estimate N/C of diamond is linking to the N/C of the growth medium by the behaviour of nitrogen.

Carbon and nitrogen isotopic compositions of asthenospherictransition zone diamonds of ‘‘basaltic’’ compositions at Juina (Fig. 4) are consistent with a crustal origin and compare well with previous studies on similar samples (Harte et al., 1999, Harte, 2010, Stachel et al., 2005, Tappert et al., 2005, 2009, Walter et al., 2011). However, Juina diamonds of peridotitic composition from the lithosphere, transition zone/lower mantle boundary and lower mantle fall within the main range of mantle-derived samples, distinct from positive d15N signatures of subducted material (i.e. subducted metaperidotites, metagabbros, metabasalts, and metasediments, Busigny et al., 2003, 2011, Philippot et al., 2007). The majority of Juina diamonds therefore derived from mantle-material rather than from deeply recycled crustal material. Kankan diamonds display much larger variations in d15N from -39.4% to þ9.6%, with no distinctive isotopic signatures according to their paragenesis (‘‘peridotitic’’/’’basaltic’’) and depths of origin (e.g. upper/lower mantle). Unlike at Juina, positive d15N values are not restricted to asthenospheric-transition zone diamonds with inclusions of ‘‘basaltic’’ composition, and are associated with mantle-like signatures in d13C. Positive d13C values are restricted to asthenospheric-transition zone diamonds with ‘‘basaltic’’ inclusion and possibly relate to diamond formation

M. Palot et al. / Earth and Planetary Science Letters 357–358 (2012) 179–193

from recycled carbonates (Craig, 1953, Veizer and Hoefs, 1976, Stachel et al., 2002, McCammon et al., 2004). No detectable nitrogen was found in these diamonds and d15N analyses cannot be used to confirm their proposed subduction origin. If lithospheric and lower mantle Kankan diamonds reflected mixing between mantle-derived and recycled material, we would expect systematic correlations between 13C enriched or depleted compositions and d15N values which deviate to higher values ( 4   1%) compared to the mantle range. The lack of such a signature indicates that most of the studied Kankan diamonds (excluding the majoritic garnet-bearing samples) is unlikely to be derived from recycled material. Slight co-variations in d13C–d15N–N in Kankan diamonds may be consistent with isotopic fractionation in an open system (i.e. Rayleigh distillation) where the isotopic and elemental evolution of the fluid is related to diamond crystallisation (e.g. Cartigny et al., 2001; Thomassot et al., 2007; Palot et al., 2009, 2012; Stachel et al., 2009; Smart et al., 2011). Modelling a general decrease of d13C with decreasing N (Fig. 5a) and, increasing d15N (Fig. 5b) indicates that nitrogen behaved compatibly. Fractionation factors of DC -3.5% and DN   1.3% are in agreement with Bottinga’s (1969) calculations for equilibrium between diamond and an oxidised CO2-bearing fluid at 1200 1C, and with Deines et al.’s (1989) estimate of isotopic fractionation between diamond and N2, respectively. Overall, the modelled fluids display mantlelike d13C-d15N–N characteristics (Fig. 5). Because of scatter in the data, it cannot, however, be ruled out that some Kankan diamonds derived from recycled material. It is also possible that some diamonds precipitated from more reducing fluids, consistent with the picture of a generally reduced deep mantle with locally more oxidised regions (e.g. McCammon et al., 2004; Rohrbach and Schmidt, 2011). However, the proposed isotopic fractionation processes at Kankan fail to account for the extremely negative d15N values observed for three diamonds KK-11, KK-21 and KK-92. 5.2. Evidences for multiple episodes of diamond growth At Juina there are indications that sublithospheric diamonds had a short mantle residence time of o100 Ma (Bulanova et al., 2010;

187

Harte and Richardson, 2011). If these tentative genesis ages were to be confirmed through future work, then their mantle residence times had been very short compared to lithospheric diamonds worldwide (typically 42 Ga, e.g. Pearson and Shirey, 1999). Assuming that an old diamond crystallisation also holds true for lithospheric diamonds from Juina and Kankan no temporal relationship exists between the formation of lithospheric ( 150–250 km depth) and sublithospheric diamonds ( 250–800 km) at these locations. The proposed transport of sublithospheric diamonds to the surface by a rising plume soon after their subduction-related formation (Harte and Richardson, 2011) is therefore a coincidence and there would be no time available for further diamond growth in the sublithospheric mantle, i.e. single episode of diamond growth. From the present dataset two lower mantle diamonds (BZ-88 and 1–36, both from Juina) display low aggregation. From the aggregation state of nitrogen in diamond a time–temperature integrated history can be obtained (see Section 3.1), and used to put constraints on the rate of diamond ascent in the mantle. In the following calculation we assumed an adiabatic gradient from 1600 1C at the bottom of the transition zone to 1300 1C at the base of the lithosphere (e.g. Katsura et al., 2004), constant ascent rates, a 10% error on the activation energy (Cooper, 1990) and used quadratic error propagation. Calculated exhumation rates are 3187100 cm/yr (2s) for 1–36 and 3827140 cm/yr (2s) for BZ-88, which are significantly higher than upward convective flow and plume flow (e.g. Maclennan et al., 2001). There are two possible scenarios that would allow reconciling the presence of lower mantle inclusions in diamonds with low nitrogen aggregation states: (1) Exhumation through a much more rapidly ascending medium, e.g. a proto-kimberlite magma (Ringwood et al., 1992) or (2) Multiple episodes of diamond growth during ascent, with a major portion of poorly aggregated nitrogen being incorporated en route to lithosphere depths (Palot et al., in press). Rapid ascent from the lower mantle is inconsistent with transport rates modelled by Harte and Cayzer (2007), based on partially unmixed majoritic garnet inclusions in Juina diamonds, which are, within uncertainties, compatible with upper mantle convection or plume flow velocities. This ascending plume model is consistent with paleo-plate reconstructions which show that the Juina region was located at the margin of the African low

Fig. 4. Modelling of d13C–d15N co-variations in Juina diamonds as a mixing process (solid line). Asthenospheric-transition zone diamonds BZ-209 and BZ-124, derived from of ‘‘basaltic’’ sources, define the ‘‘recycled material’’ end-member 1 (top left), while diamonds from the lower mantle-transition zone boundary and lower mantle BZ-201 and 1–8 with inclusion of ‘‘peridotitic’’ composition define the mantle end-member 2 (bottom right).

188

M. Palot et al. / Earth and Planetary Science Letters 357–358 (2012) 179–193

Fig. 5. Modelling of d13C–N (a) and d13C–d15N (b) co-variations in Kankan diamonds (dashed lines) through open system isotopic fractionation of different fluids. The initial d13C0, d15N0 and nitrogen content [N]0 of the fluids responsible for diamond precipitation are given for each calculated fractionation trend. The behaviour of nitrogen during diamond growth (KN) and the carbon isotopic fractionation factors (DC and DN) are varied for calculations [1] and [2] to fit to the large spread of data points.

shear velocity zone during the Cretaceous, a potential source region for deep mantle plumes according to Torsvik et al. (2010). Multiple episodes of diamond growth is supported by the observation of zonations in C-isotopes in diamonds from Kankan (KK-45 and KK-83) and Juina (BZ-88, 1-30 and 4–7) consistent with distinct episodes of diamond growth in a changing environment (see Section 5.3.1). The mantle signatures in d13C-d15N for most of the studied lower mantle diamonds and the illustration of multiple episodes of diamond growth for two Juina diamonds demonstrate that the transport to surface of sublithospheric diamonds soon after their subduction-related formation (e.g. Harte and Richardson, 2011) is very unlikely. 5.3. Evidence for deep mantle convection from carbon and nitrogen isotopes. 5.3.1. Diamond transport through the 670 km discontinuity inferred from C-isotope zonation within individual samples Carbon derived from the convective mantle (i.e. ‘‘mantle carbon’’) is usually described as being isotopically homogeneous

with values around  473% as seen in MORB, OIB, carbonatites, carbonates from kimberlites and diamonds (e.g. Deines, 1980; Javoy et al., 1982, 1986; Marty and Zimmermann., 1999; Cartigny, 2005; Stachel et al., 2009). This signature is also found in most diamonds from the lower mantle (specifically, those having inclusions of ‘‘peridotitic’’ affinity), but does not generally hold for asthenospheric-transition zone diamonds worldwide (Hutchison et al., 1999; Stachel et al., 2002; Hayman et al., 2005; Tappert et al., 2005). As the chemistry of inclusions in available asthenospheric-transition zone diamonds generally has a ‘‘basaltic’’ affinity (e.g. Stachel et al., 2000a; Tappert et al., 2005), it remains unclear whether these C-isotopic values are symptomatic of the convective mantle or not. Such diamonds are generally either enriched or depleted in 13C and have often been attributed to the recycling of sedimentary carbon (e.g. Stachel et al., 2002; Tappert et al., 2005, see Section 5.1). In a recent study of diamonds from a komatiite (Cartigny, 2010), it was argued that the very negative d13C-values (mode at 27%) would reflect mantle values in the asthenosphere-transition zone, but no direct constraints about the depth of volatiles could be deduced. The

M. Palot et al. / Earth and Planetary Science Letters 357–358 (2012) 179–193

following paragraph little relies on the mechanism leading to anomalous C-isotope composition in basaltic diamonds from the asthenospheric-transition zone. The lower mantle diamonds (of ‘‘peridotitic’’ affinity) from Juina and Kankan display small range of C-isotope composition (d13C between  8 and 1%). However, the observation of five Juina and Kankan diamonds with lower mantle inclusions showing zonation in d13C (Tables 1 and 2) provides new insights into the origin of C-isotope variability in sublithospheric diamonds. Such a zonation within individual diamonds has been also observed in other Juina lower mantle diamonds by Hutchison et al. (1999). These zonations could either reflect C-isotope variability induced by a single diamond growth (Deines, 1980; Smart et al., 2011; Palot et al., 2012) or a change in the C-isotope composition of their growth environment (Hutchison et al., 1999; Palot et al., 2012). As the present study is based on bulk measurements of individual fragments, we cannot address the presence of continuous changes in C-isotopes from the core to the rim expected for Rayleigh fractionation processes. Yet, if the large internal C-isotope variability recorded in our five diamonds was related to Rayleigh fractionation during diamond crystallisation and if Juina and Kankan lower mantle diamond growth had occurred under similar conditions, excursion from the normal lower mantle range of d13C should occur in the same direction (either toward enriched or 13C depleted compositions). This is, however, not the case, with zoned Juina diamonds extending toward more negative values, whilst the Kankan diamonds extend toward more positive values. These distinct distributions could reflect diamond growth in the lower mantle from fluids/melts having different d13Ccomposition and/or different speciation. However, the observation that zoned Juina and Kankan diamonds overlap with the respective ranges defined by asthenospheric-transition zone diamonds from the same localities is not likely to be a coincidence. This is particularly obvious at Kankan where asthenospherictransition zone diamonds display some of the highest d13C-values (up to þ3.8%) ever measured (only 1% of diamonds worldwide having d13C40%). Therefore, we interpret the zonations to be symptomatic of a change in growth environment. Samples KK-45 and KK-83 both contain inclusions typical for the lower mantle (MgSiO3-pvk and fPer, respectively, Stachel et al., 2000b), in agreement with the d13C values of parts of the diamonds ( 2.8% and -3.4%, respectively, Fig. 6 and Table 2). Other fragments of the same diamonds, however, display inclusions of former microclinehollandite and ringwoodite, respectively (Stachel et al., 2000b), with positive d13C values (þ0.1% and þ 1.4% respectively, Table 2), consistent with a growth in the asthenosphere-transition zone below Kankan (Fig. 6). Since no information is available on the chronology of the observed zonations (i.e. diamond fragments), these could have been generated during either the ascent or the descent of the diamonds through the lower mantle at the asthenospheretransition zone. At Juina, asthenospheric-transition zone diamonds extend toward 13C depleted values (Hutchison et al., 1999 and Table 1). As for Kankan, three diamonds from Juina (see above) are zoned in d13C (Table 1) supporting multiple growth events in a changing environment. However, at Juina, we cannot firmly ascribe these zonations to material transfer through the 670 km discontinuity given that at this locality low d13C are not restricted to diamonds from the asthenosphere-transition zone but have been recorded in four lower mantle diamonds with inclusions of ‘‘basaltic’’ composition (Walter et al., 2011). In summary, the zonation in C-isotopes in some Kankan diamonds represents the first direct geochemical evidence for transfer of material through the 670 km seismic discontinuity, being therefore consistent with geophysical and geochemical evidence (e.g. Fukao et al., 1992; Grand, 1994; Van der Hilst et al., 1997; Masters et al., 2000; Lay, 2007; Walter et al., 2011).

189

5.3.2. Large scale material exchange between the upper and lower mantle inferred from similar N-isotopes in lithospheric to lower mantle diamonds at a given locality. Compared to mantle carbon which is at, or close to a steady state (e.g. Javoy et al., 1982, 1986), nitrogen is far from that condition. This is well illustrated by the gross isotopic imbalance between nitrogen subducted into and degassed from the mantle, characterised by positive and negative d15N, respectively. The origin of the nitrogen isotopic imbalance (see Boyd et al., 1992) between surface material (atmosphere, crust, sediments) and the upper mantle remains unclear and may relate either to hydrodynamic escape (Tolstikhin and Marty, 1998), core–mantle fractionation (Marty and Dauphas, 2003) or heterogeneous accretion of volatiles (Javoy, 1995, 1997; Marty, 2012). Each of these models predicts secular evolution of both major reservoirs as exchange of matter occurs, but such a record remains elusive (see discussion by Marty and Dauphas, 2003, Cartigny and Ader, 2003, Kerrich and Jia, 2004). It has been suggested that the Earth’s primitive mantle originally had d15N-values as low as  25% (e.g. Javoy, 1997; Jia and Kerrich, 2004; Mohapatra et al., 2009; Marty, 2012) which evolved to present day-values close to  5% recorded by fibrous diamonds (Boyd et al., 1987, 1992), MORB (e.g. Javoy and Pineau, 1991; Marty and Humbert, 1997; Nishio et al., 1999; Marty and Zimmermann, 1999; Cartigny et al., 2001c) and peridotitic diamonds (e.g. Cartigny, 2005). This value of 5% is largely accepted even if another suggestion exist (e.g. Mohapatra and Murty, 2004). This suggestion relies, however, on series of data that are obtained using stepwise heating technique which has been questioned subsequently (Yokochi and Marty, 2006). The early Earth atmosphere might have d15N signature either as high as þ30% (Javoy, 1997; Jia and Kerrich, 2004) or nearly constant over the last 3.5 Ga (Sano and Pillinger, 1990; Pinti, 2001). Positive d15N-values of recycled nitrogen since the last 3.5 Ga have been proposed by several authors (see review by Thomazo et al., 2009). This makes nitrogen isotopes a powerful tracer of crustal material in the mantle and, consequently, d15N analyses have been applied to constrain the origin of eclogitic diamonds (e.g. Cartigny, 2005) and lamproites (e.g. Jia et al., 2003). The N-isotopes provide a time-integrated image of crustal recycling and mantle dynamics. The predicted secular evolution of mantle d15N leads to either distinct or similar d15N signatures for the upper and lower mantle if the Earth’s mantle is characterised by either two separate convection cells or a single convection cell, respectively. Overall, upper mantle (lithospheric and asthenospherictransition zone) and lower mantle diamonds at each of the two studied localities display similar median d15N values, which are both within the range defined by mantle samples worldwide (Fig. 2). This is in clear support for a well stirred Earth’s mantle with at least the upper parts of the lower mantle being convectively coupled to the upper mantle. Five diamonds from Juina (1-4, 3-1, 5-103, BZ-201 and BZ-207, Table 1) containing iron-rich ferropericlase inclusions have previously been interpreted to derive from deepest lower mantle (Harte et al., 1999; Kaminsky et al., 2001; Hayman et al., 2005). If that interpretation was correct, then the observation of similar d15N values for these diamonds, those from the shallow lower mantle, and the upper mantle suggests that convective coupling include even the deep lower mantle. This interpretation is however not unique, as subducted oceanic crust (Stachel et al., 2005) or Mg–Fe partitioning between ringwoodite and Ca-perovskiteþferropericlase during the phase transition may also create Fe-enrichment (Brey et al., 2004). Therefore, from the present data we can only derive that the 670 km discontinuity does not represent a major barrier to the transfer of material, but the existence of an isolated reservoir deeper in the lower mantle cannot be precluded.

190

M. Palot et al. / Earth and Planetary Science Letters 357–358 (2012) 179–193

Fig. 6. Carbon isotopic zonation observed in super-deep diamonds from Kankan (KK-45 and KK-83). Multiple growth events both in the lower mantle and the asthenospheretransition zone are consistent with both presence of lower and upper mantle inclusions and distinct carbon isotopic compositions at different depth levels. Arrows illustrate the movement of diamonds. The d13C ranges of Kankan diamonds from the lithosphere, the asthenosphere and the lower mantle are reported for comparison.

5.4. Preservation of primordial heterogeneities inferred from very low d15N signatures. Diamonds KK-11 (lower mantle), KK-21 and KK-92 (lithosphere), all of ‘‘peridotitic’’ affinity, display the lowest d15N values ever measured in terrestrial samples (-24.9%, -39.4% and -30.4%, respectively). Such d15N values, in principle, reflect either an isotopic fractionation process or a source component. If diffusion-related kinetic isotopic fractionation was responsible for these values, then light isotopes would diffuse faster than heavier isotopes and both N- and C-isotopes would be affected. We would expect both very negative d15N and d13C signatures, but instead, the d13C values (-3.5% to -3.2%) are indistinguishable from other Kankan diamonds. Alternatively, an open system Rayleigh fractionation is potentially responsible for large isotopic fractionation. To investigate such a possibility, we can model the d15N evolution of the diamond forming medium using the following equation: 15

15

ðd NÞ ¼ ðd NÞ0 þ DAB  lnf N where fN is the remaining fraction of nitrogen in the diamond forming medium, d15N0 ¼  4% is the initial nitrogen isotopic composition (within the range of current mantle values), and D is the fractionation factor between diamond (phase A) and diamond growth medium (phase B). The decrease of d15N toward very negative signatures requires a positive D. Assuming a fractionation factor DN2NH3 ¼ þ1% at 1200 1C (Richet et al., 1977, see also Thomassot et al., 2007), calculation indicates that an extreme fractionation with almost complete removal of nitrogen from the growth medium (less than 1 ppm of the initial N-abundance) is required to reach d15N values as low as 20% (Fig. 7). This contrasts with the studied diamonds which displayed N-abundances higher (up to 10 times) than observed for Kankan diamonds of lower mantle paragenesis (Table 2). In addition, high degree of Rayleigh fractionation should be accompanied by d13Cvalues strongly deviating from the mantle value (to positive or highly negative values, depending on the oxygen fugacity), which again is not observed. Thus, isotopic fractionation processes cannot account for the observed highly negative d15N values. More likely, these signatures characterise a source component which strongly contrast with samples from the present day upper

Fig. 7. Modelling of the nitrogen isotopic composition of a diamond precipitating fluid following a Rayleigh distillation process. The remaining nitrogen in the fluid is expressed in fraction of 1 (logarithmic scale). See text for further details. The shaded area represents the nitrogen loss required to account for the very low values of d15N determined for three Kankan diamonds.

mantle (d15N   5%). We propose that these diamonds record primordial heterogeneity of volatile elements in the mantle. Interestingly, the studied samples have C- and N-isotopes compositions close to diamond CHL 12 from China previously measured (d13C ¼  3.5%, d15N¼-24.2%), and interpreted to represent primordial mantle material (Cartigny et al., 1997). This striking similarity between these is compatible with the existence of an additional well homogenised reservoir. Values of d15No-40% are only observed in meteorites (Grady et al., 1986; Rubin and Choi, 2009) and are predicted for the Earth’s primordial mantle by models invoking enstatite chondrite as the principal building material (e.g. Clayton and Mayeda, 1984; Meisel et al., 1996; Javoy, 1995, 1997; Dauphas et al., 2004). Note that these d15N signatures also fall within the estimated range for Mars parent body (e.g. Becker and Pepin, 1984, 1986, Mohapatra and Murty, 2003). The lack of continuity of d15N values toward the observed highly negative values and their large difference to the present-

M. Palot et al. / Earth and Planetary Science Letters 357–358 (2012) 179–193

191

Fig. 8. Histogram of nitrogen isotopic compositions of lower mantle diamonds from Juina and Kankan. Plume lavas and oceanic island basalts (OIB) are from Dauphas and Marty (1999), Marty and Dauphas (2003) and Mohapatra et al. (2009). Only subglacial basalts DICE 2 and 12 from Iceland from Mohapatra et al. (2009) are used for this compilation as the other samples represent phenocrysts and xenocrysts from ultramafic nodules. Air is the international reference for the d15N notation.

day mantle range support our interpretation of preserved primordial heterogeneities, in both the upper and the lower mantle, stored either as blobs or deep isolated layers (e.g. Becker et al., 1999; Deschamps et al., 2011). Such a long-term preservation of isolated mantle domains has been recognised in abyssal peridotites from the Gakkel ridge (Liu et al., 2008), Cenozoic lavas from Baffin Island and West Greenland (Jackson et al., 2010), and from noble gases in OIB (e.g. Alle gre et al., 1986; Kurz et al., 2009). High 3 He/4He ratios and solar Ne isotopic composition have been also observed in diamonds from South Africa and Zaire, respectively, and were interpreted as reflecting incorporation of primordial material (Ozima and Zashu, 1983, 1988). As discussed earlier, the formation of lithospheric and lower mantle diamonds is generally interpreted as being temporally independent (e.g. Walter et al., 2011; Harte, 2010). The present data rather illustrate striking similarities between shallow and deep diamonds at both Juina and Kankan, yet with distinct isotopic characteristics for each locality, which may point to a genetic link between deeper and shallow diamonds.

5.5. Implication for the origin of oceanic island basalts (OIB) and mantle structure Our observation of almost exclusively negative d15N values for lower mantle diamonds has important implications for source of mantle plumes. ‘‘Primitive’’ noble gas signatures from OIB samples and fluid inclusions in carbonatites are associated with d15N40% (Marty et al., 1998; Dauphas and Marty, 1999; Marty and Dauphas, 2003; Mohapatra et al., 2009), requiring the involvement of a reservoir distinct from the MORB source. This distinct reservoir has been suggested to reside within the lower mantle (e.g. Kaneoka, 1981; White and Hofmann, 1982; Alle gre et al., 1986; Ballentine et al., 2005; Kurz et al., 2009). The observation of variable 3He/4He, 20Ne/22Ne and 40Ar/36Ar ratios illustrates diversity among the geochemical sources for distinct oceanic island provinces (Marty and Dauphas, 2003; Mohapatra et al., 2009), their d15N values are invariably positive and display no clear correlation with the noble gas isotope signature. This distinct positive d15N signature of OIB contrasts with the almost exclusively negative d15N signature of lower mantle diamonds

(Fig. 8), suggesting that OIB and lower mantle diamonds have distinct nitrogen sources. OIB appears to be buffered by deeply subducted material which shares common isotopic signatures for Pb, Sr, Nd, O (e.g. Hofmann, 2003) including nitrogen. Given that the studied diamonds likely derived from the uppermost parts of the lower mantle (Harte et al., 1999; Stachel et al., 2000b; Kaminsky et al., 2001), the distinct nitrogen isotopic signature of OIB can be interpreted to reflect sources located deeply within the lower mantle. Alternatively, some OIB could derive from the transition zone (e.g. Ringwood, 1994), consistent with positive d15N signatures in asthenospheric-transition zone diamonds with inclusions from ‘‘basaltic’’ sources at Juina.

6. Conclusions Diamonds from Juina and Kankan illustrate that large scale transfer of material occurs through the 670 km seismic discontinuity. Kankan diamonds also record the contribution of primordial nitrogen—which we infer to be d15No 40%. It remains unclear however whether such primordial heterogeneities exist as blobs dispersed in the convective mantle or as a deeper layer isolated from the mantle convection cell. Our data support the view that subducted slabs penetrate into the deep lower mantle, yet also support the existence of mantle domains that are preserved from homogenisation during mantle convection, crustal extraction and slab recycling.

Acknowledgments We are very grateful to Editor Bernard Marty and two anonymous reviewers for their constructive and careful reviews that improved the strength of this manuscript. Analyses were performed at Laboratoire de Geochimie des Isotopes Stables, Institut de Physique du Globe, Universite´ Paris Diderot. Michel Girard, Guillaume Landais and Jean-Jacques Bourrand are acknowledged for their technical assistance. This work was supported by grants from the CNRS/INSU ‘‘Dyeti’’ programme. IPGP contribution 3327.

192

M. Palot et al. / Earth and Planetary Science Letters 357–358 (2012) 179–193

Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.epsl.2012.09.015.

References Ader, M., Boudou, J.P., Javoy, M., Goffe´, B., Daniels, E., Vieth-Redemann, A., 1998. Isotope study of organic nitrogen of Westphalian anthracites from eastern Pennsylvania (USA) and from Bramsche Massif Germany. Org. Geochem. 29, 315–323. Alle gre, C.J., Staudacher, T., Sarda, P., 1986. Rare gas systematics, formation of the atmosphere, evolution and structure of the Earth’s mantle. Earth Planet. Sci. Lett. 81, 127–150. Ballentine, C.J., Marty, B., Sherwood Lollar, B., Cassidy, M., 2005. Neon isotopes constrain convection and volatil origin in the Earth’s mantle. Nature 433, 33–38. Bebout, G.E., Fogel, M.L., 1992. Nitrogen-isotope compositions of metasedimentary rocks in the Catalina Schist, California: implications for metamorphic devolatilization history. Geochim. Cosmochim. Acta 56, 2839–2849. Becker, R.H., Pepin, R.O., 1984. The case for a Martian origin of the shergottites: nitrogen and noble gases in EETA 79001. Earth Planet. Sci. Lett. 69, 225–242. Becker, R.H., Pepin, R.O., 1986. Nitrogen and light noble gases in Shergotty. Geochim. Cosmochim. Acta 50, 993–1000. Becker, T., Kellogg, J.B., O’Connell, R.J., 1999. Thermal constraints on the survival of primitive blobs in the lower mantle. Earth Planet. Sci. Lett. 171, 351–365. Bottinga, Y., 1969. Carbon isotope fractionation between graphite, diamond and carbon dioxide. Earth Planet. Sci. Lett. 5, 301–307. Boyd, S.R., Kiflawi, I., Woods, G.S., 1994. The relationship between infrared absorption and the A defect concentration in diamond. Philos. Mag. 6B9, 1149–1153. Boyd, S.R., Kiflawi, I., Woods, G.S., 1995a. Infrared absorption by the B nitrogen aggregate in diamond. Philos. Mag. B72, 351–361. Boyd, S.R., Re´jou-Michel, A., Javoy, M., 1995b. Improved techniques for the extraction, purification and quantification of nanomole quantities of nitrogen gas: the nitrogen content of a diamond. Meas. Sci. Technol. 6, 297–305. Boyd, S.R., Mattey, D.P., Pillinger, C.T., Milledge, H.J., Mendelssohn, M., Seal, M., 1987. Multiple growth events during diamond genesis: an integrated study of carbon and nitrogen isotopes and nitrogen aggregation state in coated stones. Earth Planet. Sci. Lett. 86, 341–353. Boyd, S.R., Pillinger, C.T., Milledge, H.J., Mendelssohn, M., Seal, M., 1992. C and N isotopic composition and infrared absorption spectra of coated diamonds: evidence for the regional uniformity of CO2–H2O rich fluids in lithospheric mantle. Earth Planet. Sci. Lett. 109, 633–644. Boyet, M., Carlson, R.W., 2005. 142Nd evidence for early (44.53 Ga) global differentiation of the silicate Earth. Science 309, 576–581. Brey, G.P., Bulatov, V., Girnis, A., Harris, J.W., Stachel, T., 2004. Ferropericlase—a lower mantle phase in the upper mantle. Lithos 77, 655–663. Bulanova, G.P., Walter, M.J., Smith, C.B., Kohn, S.C., Armstrong, L.S., Blundy, J., Gobbo, L., 2010. Mineral inclusions in sublithospheric diamonds from Collier 4 kimberlite pipe, Juina, Brazil: subducted protoliths, carbonated melts and primary kimberlite magmatism. Contrib. Mineral. Petrol. 160, 489–510. Busigny, V., Cartigny, P., Philippot, P., 2011. Nitrogen isotopes in ophiolitic metagabbros: a re-evaluation of modern nitrogen fluxes in subduction zones and implication for the early Earth atmosphere. Geochim. Cosmochim. Acta 75, 7502–7521. Busigny, V., Cartigny, P., Philippot, P., Ader, M., Javoy, M., 2003. Massive recycling of nitrogen and other fluid-mobile elements (K, Rb, Cs, H) in a cold slab environment: evidence from HP to UHP oceanic metasediments of the Schistes Lustres nappe (western Alps, Europe). Earth Planet. Sci. Lett. 215, 27–42. Cartigny, P., 2005. Stables isotopes and the origin of diamond. Elements 1, 79–84. Cartigny, P., 2010. Mantle-related carbonados? Geochemical insights from diamonds from the Dachine komatiite (French Guiana). Earth Planet. Sci. Lett. 296, 329–339. Cartigny, P., Ader, M., 2003. A comment on ‘‘The nitrogen record of crust-mantle interaction and mantle convection from Archean to Present’’ by B. Marty and N. Dauphas. Earth Planet. Sci. Lett. 216, 425–432. Cartigny, P., Boyd, F.R., Harris, J.W., Javoy, M., 1997. Nitrogen isotopes in peridotitic diamonds from Fuxian, China: the mantle signature. Terra Nova 9, 175–179. Cartigny, P., Chinn, I., Viljoen, K.S., Robinson, D., 2004. Early Proterozoic ( 41.8 Ga) ultrahigh pressure metamorphism: evidence from Akluilˆak microdiamonds (NW Canada). Science 304, 853–855. Cartigny, P., De Corte, K., Shatsky, V.S., Ader, M., De Paepe, P., Sobolev, N.V., Javoy, M., 2001a. The origin and formation of metamorphic microdiamonds from the Kokchetav massif, Kazakhstan: a nitrogen and carbon isotopic study. Chem. Geol. 176, 267–283. Cartigny, P., Harris, J.W., Javoy, M., 2001b. Diamond genesis, mantle fractionations and mantle nitrogen content: a study of d13C–N concentrations in diamonds. Earth Planet. Sci. Lett. 185, 85–98. Cartigny, P., Jendrzejewski, N., Pineau, F., Javoy, M., 2001c. Volatile (C, N, Ar) variability in MORB and the respective role of mantle source heterogeneity

and degassing: the case of the southwest Indian ridge. Earth Planet. Sci. Lett. 194, 241–257. Chrenko, R.M., Tuft, R.E., Strong, H.M., 1977. Transformation of the state of nitrogen in diamond. Nature 270, 141–144. Clayton, R., Mayeda, T., 1984. Oxygen isotope composition of enstatite chondrites and aubrites. J. Geophys. Res. 89, C245–C249. Cooper, G.I., 1990. Infrared spectroscopy of diamond in relation to mantle processes. Unpubl. Ph.D. Thesis, University of London, p. 262. Craig, H., 1953. The geochemistry of the stable carbon isotopes. Geochim. Cosmochim. Acta 3, 53–92. Dauphas, N., Davis, A.M., Marty, B., Reisberg, L., 2004. The cosmic molybdenum– ruthenium isotope correlation. Earth Planet. Sci. Lett. 226, 465–475. Dauphas, N., Marty, B., 1999. Heavy nitrogen in carbonatites of the Kola Peninsula: a possible signature of deep mantle. Science 286, 2488–2490. Deines, P., 1980. The carbon isotopic composition of diamonds: relationship to diamond shape, color, occurrence and vapor composition. Geochim. Cosmochim. Acta 44, 943–961. Deines, P., Harris, J.W., Spear, P.M., Gurney, J.J., 1989. Nitrogen and 13C content of Finsch and Premier diamonds and their implications. Geochim. Cosmochim. Acta 53, 1367–1378. Deschamps, F., Kaminski, E., Tackley, P., 2011. A deep mantle origin for the primitive signature of ocean island basalt. Nat. Geosci. 4, 879–882. Fukao, Y., Obayashi, M., Inoue, H., Nembai, M., 1992. Subducting slabs stagnant in the mantle transition zone. J. Geophys. Res.-Sol. Earth 97, 4809–4822. Grady, M.M., Wright, I.P., Carr, L.P., Pillinger, C.T., 1986. Compositional differences in enstatite chondrites based on carbon and nitrogen stable isotope measurements. Geochim. Cosmochim. Acta 50, 2799–2813. Grand, S.P., 1994. Mantle shear structure beneath the America and surrounding oceans. J. Geophys. Res.-Sol. Ea. 99, 11591–11621. 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. Harte, B., 1992. Trace element characteristics of deep-seated eclogite parageneses—an ion microprobe study of inclusions in diamonds. Abstracts V.M. Goldschmidt Conference, Geochem. Soc. Reston, Virginia, A48. Harte, B., 2010. Diamond formation in the deep mantle: the record of mineral inclusions and their distribution in relation to mantle dehydration zones. Mineral. Mag. 74, 189–215. Harte, B., Cayzer, N., 2007. Decompression and unmixing of crystals included in diamonds from the mantle transition zone. Phys. Chem. Miner. 34, 647–656. Harte, B., Harris, J.W., 1994. Lower mantle mineral associations preserved in diamonds. Miner. Mag. A58, 384–385. Harte, B., Harris, J.W., Hutchison, M.T., Watt, G.R., Wilding, M., 1999. Lower mantle mineral associations in diamonds from Sa~ o Luiz, Brazil. Mantle Petrology: Field Observations and High Pressure Experimentation: A Tribute to Francis R. Boyd The Geochemical Society, 125–153. Harte, B. and Richardson, S.H., 2011. Mineral inclusions in diamonds track the evolution of a Mesozoic subducted slab beneath West Gondwanaland. Gondwana Res. 21, 236–245. Hayman, P.C., Kopylova, M.G., Kaminsky, F.V., 2005. Lower mantle diamonds from Rio Soriso (Juina area, Mato Grosso, Brazil). Contrib. Mineral. Petrol. 149, 430–445. Hilton, D.R., Porcelli, D., 2003. Noble gases as mantle tracers. In: Carlson, R.W., Holland, K.K., Turekian, K.K. (Eds.), Treatise on Geochemistry, vol. 2. Elsevier, Oxford, pp. 277–318. Hofmann, A.W., 2003. Sampling mantle heterogeneity through oceanic basalts: isotopes and trace elements. In: Carlson, R.W., Holland, H.D., Turekian, K.K. (Eds.), Treatise on Geochemistry, vol. 2. Elsevier, Amsterdam, pp. 61–101. Hutchison, M.T., Cartigny, P., and Harris, J.W., 1999. Carbon and nitrogen compositions and physical characteristics of tranzition zone and lower mantle diamonds from Sa~ o Luiz, Brazil. In: Gurney, J.J., Gurney, J.L., Pascoe, M.D., Richardson, S. H. (Eds.), Proceedings of the 7th International Kimberlite Conference, vol. 1). Red Rood Design, Cape Town, pp. 372–382. Hutchison, M.T., Hursthouse, M.B., Light, M.E., 2001. Mineral inclusions in diamonds: associations and chemical distinctions around the 670 km discontinuity. Contrib. Mineral. Petrol. 142, 119–126. Jackson, M.G., Carlson, R.W., Kurz, M.D., Kempton, P.D., Francis, D., Blusztajn, J., 2010. Evidence for the survival of the oldest terrestrial mantle reservoir. Nature 466, 853–856. Javoy, M., 1995. The integral enstatite chondrite model of the Earth. Geophys. Res. Lett. 22, 2219–2222. Javoy, M., 1997. The major volatile element of the Earth: their origin, behavior, and fate. Geophys. Res. Lett. 24, 177–180. Javoy, M., Pineau, F., 1991. The volatiles record of a ‘‘popping’’ rock from the MidAtlantic ridge 141N: chemical and isotopic composition of a gas trapped in the vesicules. Earth Planet. Sci. Lett. 107, 598–611. Javoy, M., Pineau, F., Alle gre, C.J., 1982. Carbon geodynamic cycle. Nature 300, 171–173. Javoy, M., Pineau, F., Delorme, H., 1986. Carbon and nitrogen isotopes in the mantle. Chem. Geol. 57, 41–62. Jia, Y., Kerrich, R., Gupta, A.K., Fyfe, W.S., 2003. 15N-enriched Gondwana lamproites, eastern India: crustal N in the mantle source. Earth Planet. Sci. Lett. 215, 43–56. Jia, Y., Kerrich, R., 2004. A reinterpretation of the crustal N-isotope record: evidence for a 15N-enriched Archean atmosphere? Terra Nova 16, 102–108. Kaminsky, F.V., 2012. Mineralogy of the lower mantle: a review of ‘super-deep’ mineral inclusions in diamond. Earth-Sci. Rev. 110, 127–147.

M. Palot et al. / Earth and Planetary Science Letters 357–358 (2012) 179–193

Kaminsky, F.V., Zakharchenko, O.D., Davies, R., Griffin, W.L., Khachatryan-Blinova, G.K., Shiryaev, A.A., 2001. Superdeep diamonds from Juina area, Mato Grosso State, Brazil. Contrib. Mineral. Petrol. 140, 734–753. Kaneoka, I., 1981. Noble gas constraints on the layered structure of the mantle. Rock Magn. Paleogeophys. 8, 94–99. Katsura, T., Yamada, H., Nishikawa, O., et al., 2004. Olivine–Wadsleyite transition in the system (Mg,Fe)SiO4. J. Geophys. Res. 109, B02209. Kerrich, R., Jia, Y.A., 2004. Comment on ‘‘The nitrogen record of crust–mantle interaction and mantle convection from Archean to Present’’ by B. Marty and N. Dauphas. Earth Planet. Sci. Lett. 225, 435–440. Kurz, M.D., Curtice, J., Fornari, D., Geist, D., Moreira, M., 2009. Primitive neon from the center of the Galapagos hotspot. Earth Planet. Sci. Lett. 286, 23–34. Lay, T., 2007. Deep earth structure—lower mantle and D’’. In: Romanowicz, B., Dziewonski, A. (Eds.), Treatise on Geophysics, vol. 1. Elsevier, Amsterdam, pp. 619–654. Liu, C.Z., Snow, J.E., Hellebrand, E., Brugmann, G., von der Handt, A., Buchl, A., Hofmann, A.W., 2008. Ancient, highly heterogeneous mantle beneath Gakkel ridge, Arctic Ocean. Nature 452, 311–316. Maclennan, J., McKenzie, D., Gronvold, K., 2001. Plume-driven upwelling under Iceland. Earth Planet. Sci. Lett. 194, 67–82. Mariotti, A., 1983. Atmospheric nitrogen is a reliable standard for natural 15N abundance measurements. Nature 303, 685–687. Marty, B., 2012. The origins and concentrations of water, carbon, nitrogen and noble gases on Earth. Earth Planet. Sci. Lett. 313–314, 56–66. Marty, B., Dauphas, N., 2003. The nitrogen record of crust-mantle interaction and mantle convection from Archean to present. Earth Planet. Sci. Lett. 206, 397–410. Marty, B., Humbert, F., 1997. Nitrogen and argon isotopes in oceanic basalts. Earth Planet. Sci. Lett. 152, 101–112. Marty, B., Tolstikhin, I.N., Kamensky, I.L., Nivin, V., Balaganskaya, E., Zimmermann, L., 1998. Plume-derived rare gases in 380 Ma carbonatites from the Kola region (Russia) and the argon isotopic composition in the deep mantle. Earth Planet. Sci. Lett. 164, 179–192. Marty, B., Zimmermann, L., 1999. Volatiles (He, C, N, Ar) in mid-ocean ridge basalts: assessment of shallow-level fractionation and characterization of source composition. Geochim. Cosmochim. Acta 63, 3619–3633. Masters, G., Laske, G., Bolton, H., Dziewonski, A., 2000. The relative behavior of shear velocity, bulk sound speed, and compressional velocity in the mantle: Implications for chemical and thermal structure. Geophys. Monogr. Am. 117, 63–87. McCammon, C.A., Stachel, T., Harris, J.W., 2004. Iron oxidation state in lower mantle mineral assemblages II. Inclusions in diamonds from Kankan, Guinea. Earth Planet. Sci. Lett. 222, 423–434. Meisel, R., Walker, R., Morgan, J., 1996. The osmium isotopic composition of the Earth’s primitive upper mantle. Nature 383, 517–520. Mohapatra, R.K., Harrison, D., Ott, U., Gilmour, J.D., Trieloff, M., 2009. Noble gas and nitrogen isotopic components in Oceanic Island Basalts. Chem. Geol. 266, 29–37. Mohapatra, R.K., Murty, S.V.S., 2003. Precursors of Mars: constraints from nitrogen and oxygen isotopic compositions of martian meterorites. Meteorit. Planet. Sci. 38, 225–241. Mohapatra, R.K., Murty, S.V.S., 2004. Nitrogen isotopic composition of the MORB mantle: a reevaluation. Geochem. Geophys. Geosyst. 5 , http://dx.doi.org/ 10.1029/2003GC000612. Moore, R.O., Gurney, J.J., Griffin, W.L., Shimizu, N., 1991. Ultra-high pressure garnet inclusions in Monastery diamonds: trace element abundance patterns and conditions of origin. Eur. J. Mineral. 3, 213–230. Nishio, Y., Ishii, T., Gamo, T., Sano, Y., 1999. Volatile element isotopic systematics of the Rodringues Triple Junction Indian Ocean MORB: implications for mantle heterogeneity. Earth Planet. Sci. Lett. 170, 241–253. O’Neil, J., Carlson, R.W., Francis, D., Stevenson, R.K., 2008. Neodynium-142 evidence for Hadean mafic crust. Science 321, 1828–1831. Ozima, M., Zashu, S., 1983. Primitive helium in diamonds. Nature 219, 1067–1068. Ozima, M., Zashu, S., 1988. Solar-type Ne in Zaire cubic diamonds. Geochim. Cosmochim. Acta 52, 19–25. Palot, M., Cartigny, P., Viljoen, K.S., 2009. Diamond origin and genesis: a C and N stable isotope study on diamonds from a single eclogitic xenolith (Kaalvallei, South Africa). Lithos 112S, 758–766. Palot, M., Pearson, D.G., Stachel T. and Harris, J.W. Multiple growth episodes or prolonged formation of diamonds? Inferences from infrared absorption data. J. Geol. Soc. India, in press. Palot, M., Pearson, D.G., Stern, R.A., Stachel, T., Harris, J.W., 2012. Multiple growth events, processes and fluid sources involved in the growth of diamonds from Finsch mine, RSA: a micro-analytical study. In: 10th International Kimberlite Conference, Extended Abstract No. 10IKC-68.

193

Pearson, D.G. and Shirey, S.B., 1999. Isotopic dating of diamonds. In : Lambert David, D., Ruiz, J. (Eds.), Application of Radiogenic Isotopes to Ore Deposit Research and Exploration, vol. 12: Rev. Econ. Geol., Boulder, CO, United States, Society of Economic Geologists, pp. 143–171. Philippot, P., Busigny, V., Scamberulli, M., Cartigny, P., 2007. Nitrogen and oxygen isotopes as tracers of fluid activities in serpentinites and metasediments during subduction. Miner. Petrol. 91, 11–24. Pinti, D.L., 2001. The isotopic record of Archean nitrogen and the evolution of the early Earth. Trends Geochem. 2, 1–17. Richet, P., Bottinga, Y., Javoy, M., 1977. A review of hydrogen, carbon, nitrogen, oxygen, sulphur and chlorine stable isotope fractionation among gaseous molecules. Annu. Rev. Earth Planet. Sci. 82, 269–279. Ringwood, A.E., 1994. Role of the transition zone and 660 km discontinuity in mantle dynamics. Phys. Earth Planet. 86, 5–24. Ringwood, A.E., Kesson, S.E., Hibberson, W., Ware, N., 1992. Origin of kimberlites and related magmas. Earth Planet. Sci. Lett. 113, 521–538. Rohrbach, A., Schmidt, M.W., 2011. Redox freezing and melting in the Earth’s deep mantle resulting from carbon-iron redox coupling. Nature 472, 209–212. Rubin, A.E., Choi, B.G., 2009. Origin of halogens and nitrogen in Enstatite Chondrites. Earth Moon Planets 105, 41–53. Sano, Y., Pillinger, C.T., 1990. Nitrogen isotopes and N2/Ar ratios in cherts: an attempt to measure time evolution of atmospheric d15N. Geochem. J. 24, 317–327. Sautter, V., Haggerty, S.E., Field, S., 1991. Ultradeep ( 4300 km) ultramafic xenoliths: petrological evidence from the transition zone. Science 252, 827–830. Smart, K.A., Chacko, T., Stachel, T., Muehlenbachs, K., Stern, R.A., Heaman, L.M., 2011. Diamond growth from oxidized carbon sources beneath the Northern Slave Craton, Canada: a d13C–N study of eclogite-hosted diamonds from the Jericho kimberlite. Geochim. Cosmochim. Acta 75, 6027–6047. Stachel, T., Brey, G.P., Harris, J.W., 2000a. Kankan diamonds (Guinea) I: from the lithosphere down to the transition zone. Contrib. Mineral. Petrol. 140, 1–15. Stachel, T., Harris, J.W., Brey, G.P., Joswig, W., 2000b. Kankan diamonds (Guinea) II: lower mantle inclusion parageneses. Contrib. Mineral. Petrol. 140, 16–27. Stachel, T., Brey, G.P., Harris, J.W., 2005. Inclusions in sublithospheric diamonds: glimpses of deep earth. Elements 1, 73–78. Stachel, T., Harris, J.W., 2008. The origin of cratonic diamonds—constraints from mineral inclusions. Ore Geol. Rev. 34, 5–32. Stachel, T., Harris, J.W., Aulbach, S., Deines, P., 2002. Kankan diamonds (Guinea) III: d13C and nitrogen characteristics of deep diamonds. Contrib. Mineral. Petrol. 142, 465–475. Stachel, T., Harris, J.W., Muehlenbachs, K., 2009. Sources of carbon in inclusion bearing diamonds. Lithos 112S, 625–637. Tackley, A.W., 2007. Mantle geochemical geodynamics. In: Bercovici, D., Schubert, G. (Eds.), Treatise on Geophysics, vol. 7. Elsevier, Amsterdam, pp. 437–505. Tappert, R., Foden, J., Stachel, T., Muehlenbachs, K., Tappert, M., Wills, K., 2009. Deep mantle diamonds from South Australia: a record of Pacific subduction at the Gondwanan margin. Geology 37, 43–46. Tappert, R., Stachel, T., Harris, J.W., Muehlenbachs, K., Ludwig, T., Brey, G.P., 2005. Diamonds from Jagersfontein (South Africa): messengers from the sublithospheric mantle. Contrib. Mineral. Petr. 150, 505–522. Thomassot, E., Cartigny, P., Harris, J.W., Viljoen, K.S., 2007. Methane-related diamond crystallization in Earth’s mantle: stable isotope evidences from a single diamond-bearing xenolith. Earth Planet. Sci. Lett. 257, 362–371. Thomazo, C., Pinti, D.L., Busigny, V., Ader, M., Hashizume, K., 2009. Biological activity and Earth’s surface evolutions: insights from carbon, sulfur, nitrogen and iron stable isotopes in the rock record. CR Palevol. 8, 665–678. Tolstikhin, I.N., Marty, B., 1998. The evolution of terrestrial volatiles: a view from helium, neon, argon and nitrogen isotope modelling. Chem. Geol. 147, 27–52. Torsvik, T.H., Burke, K., Steinberger, B., Webb, S.J., Ashwal, L.D., 2010. Diamonds sampled by plumes from the core-mantle boundary. Nature 466, 352–355. Van der Hilst, R.D., Widiyantoro, S., Engdahl, E.R., 1997. Evidence for deep mantle circulation from global tomography. Nature 386, 578–584. Veizer, J., Hoefs, J., 1976. The nature of 18O/16O and 13C/12C secular trends in sedimentary carbonate rocks. Geochim. Cosmochim. Acta 40, 1387–1395. Walter, M.J., Kohn, S.C., Araujo, D., Bulanova, G.P., Smith, C.B., Gaillou, E., Wang, J., Steel, A., Shirey, S.B., 2011. Deep mantle cycling of oceanic crust: evidence from diamonds and their mineral inclusions. Nature 334, 54–57. White, W.M., Hofmann, A.W., 1982. Sr and Nd isotope geochemistry of oceanic basalts and mantle evolution. Nature 296, 821–825. Wilding, M.C., 1990. A study of diamonds with syngenetic inclusions. Unpubl. PhD Thesis, University of Edinburgh, pp. 281. Yokochi, R., Marty, B., 2006. Fast chemical and isotopic exchange of nitrogen during reaction with hot molybdenum. Geochem. Geophys. Geosyst. 7 , http:// dx.doi.org/10.1029/2006GC001253.