- Email: [email protected]
Unusual noble gas compositions in polycrystalline diamonds: preliminary results from the Jwaneng kimberlite, Botswana
Chemical Geology 203 (2004) 347 – 358 www.elsevier.com/locate/chemgeo
Unusual noble gas compositions in polycrystalline diamonds: preliminary results from the Jwaneng kimberlite, Botswana Masahiko Honda a,*, David Phillips b, Jeffrey W. Harris c, Igor Yatsevich a a
Research School of Earth Sciences, The Australian National University, Canberra, ACT 0200, Australia b School of Earth Sciences, The University of Melbourne, Melbourne, Victoria 3010, Australia c Division of Earth Sciences, University of Glasgow, Glasgow G12 8QQ, UK Received 4 September 2002; accepted 23 October 2003
Abstract We have undertaken noble gas analyses of four polycrystalline framesite diamonds from the Jwaneng kimberlite pipe, Botswana. These samples yielded complex, multiple noble gas components (crustal, atmospheric and in situ radiogenic/ fissiogenic), which were successfully deconvoluted by combining vacuum crushing and step-heating experiments and examining a full suite of noble gas isotope and elemental abundances. The most striking observation is the presence of crustal nucleogenic neon, released on graphitisation of the framesites. Neon of this composition can only have been produced in the crust and subsequently incorporated during formation of the framesites in the mantle. This may indicate that noble gases produced in the crust, such as nucleogenic neon and possibly xenon, and some radiogenic 4He and 40Ar and fissiogenic xenon, were introduced into the subcontinental mantle source during ancient subduction-related processes. If correct, then some parts of the mantle may contain significant quantities of crustal noble gases. D 2003 Elsevier B.V. All rights reserved. Keywords: Diamond; Noble gas; Mantle evolution; Mantle lithosphere
1. Introduction Research goals in noble gas geochemistry include understanding the structure of the Earth’s mantle and the creation of a coherent model for its evolution. In this regard, studies of noble gas compositions of midocean-ridge basalts (MORBs) and ocean island basalts (OIBs) such as found in Loihi Seamount, Hawaii, Iceland and Re´union (e.g., Alle`gre et al., 1983; Dixon et al., 2000; Honda et al., 1993; Kurz et al., 1983;
* Corresponding author. Fax: +61-2-6125-0738. E-mail address: [email protected] (M. Honda). 0009-2541/$ - see front matter D 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.chemgeo.2003.10.012
Moreira et al., 1998; Staudacher and Alle`gre, 1982; Trieloff et al., 2002), have provided a wealth of information. However, virtually all these data are from samples that are effectively of zero-age and only provide information on the present-day composition of mantle noble gases. It is critically important to determine if there is any systematic variation in mantle noble gas composition with time. This evidence is particularly relevant to current debates on mantle geodynamics. For example, noble gas data for mantle-derived samples of different ages could be used to evaluate the degree to which the upper mantle has interacted with the lower mantle over time and allow further refinement of models concerning two
348
M. Honda et al. / Chemical Geology 203 (2004) 347–358
layered convection vs. whole mantle convection (e.g., Alle`gre, 1997; Kellogg et al., 1999) and mass transport (including volatiles) in the mantle (e.g., Davies, 1999; Kamijo et al., 1998; Porcelli and Wasserburg, 1995). Previous attempts to determine the noble gas geochemistry of the ancient mantle by analysis of older geological samples have been largely unsuccessful, in part owing to the lack of suitable sample material. A possible exception is the determination of a high 3He/4He ratio in olivines separated from Archaean komatiites from Canada (Matsumoto et al., 2002; Richard et al., 1996). Diamonds have several unique characteristics that make them potentially very useful as robust containers of mantle noble gases: (1) diamonds are chemically inert and typically have experienced little interaction with the surrounding mantle, crust or atmosphere, despite their depth of origin and rapid transportation to shallow crustal levels in kimberlite and lamproite volcanic magmas, (2) many diamonds, particularly those of industrial quality, contain significant quantities of mantle-derived noble gases (e.g., Burgess et al., 1998; Ozima and Zashu, 1983), (3) diamonds are highly retentive of all noble gases, including helium (Honda et al., 1987), (4) most diamonds appear to be derived from 150- to 200-km depth in the Earth’s subcontinental mantle (e.g., Wyllie, 1995), although some may be derived from as deep as the 670-km discontinuity (e.g., Harte et al., 1998) and (5) diamonds cover a wide range of crystallization ages mostly between 1.0 and 3.2 billion years (e.g., Richardson et al., 1984, 1993). Thus, diamonds provide a direct window into the ancient mantle, and they are unique repositories of mantle-derived noble gases. In this paper, we report preliminary noble gas compositions for four polycrystalline framesite diamonds from the Jwaneng kimberlite, Botswana. Previous noble gas studies on polycrystalline diamonds (framesites) showed large quantities of 36Ar and radiogenic 4He and 40Ar (Burgess et al., 1998, Phillips, unpublished data). Thus, the expectation was that we could extract information on mantlederived noble gas compositions (helium, neon, argon, krypton and xenon) from single, small fragments of framesite. The present study represents part of a wider investigation into the evolution of the noble gases through time.
2. Jwaneng diamonds The Jwaneng kimberlite cluster is situated in southern Botswana, approximately 120 km west of Gaborone. Kinny et al. (1989) reported a U – Pb zircon intrusion age of 235 F 4 Ma for the Jwaneng DK2 body and a Rb –Sr age of 250 F 17 Ma, from the nearby DK7 kimberlite. Although arguably the richest diamond mine in the world, some 60% of the diamonds are classed as aggregates (Deines et al., 1997). Framesites are a polycrystalline form of diamond and exhibit variable, sub-millimetre grain sizes. They are usually gray to black, comprising randomly oriented fine crystallites that are sometimes intergrown with silicates and oxides of eclogitic and lherzolitic parageneses (Kirkley et al., 1994). Framesites are also characterized by an unusual bimodal carbon isotopic distribution, with average values of 1.5x d13C and 19x d13C, which has lead to suggestions of an origin from subduction of oceanic sediments associated with ophiolite sequences (Burgess et al., 1998; Kirkley et al., 1994). Therefore, eclogitic diamonds may provide the means for evaluating the recycling of noble gases into the mantle. Inclusions from gem-quality Jwaneng diamonds comprise an even mix of peridotitic and eclogitic compositional affinities (Deines et al., 1997). Using composites of garnet and clinopyroxene inclusions, Richardson et al. (1999) obtained a two-point isochron Re – Os genesis age of 1540 F 20 Ma for eclogitic diamonds from Jwaneng. More recent Re – Os results for eclogitic sulphide inclusions from Jwaneng diamonds indicate two age groups, one at f 1.5 Ga and another of f 2.9 Ga (Richardson et al., 2001b; Shirey et al., 2001). No age information is available for diamonds of peridotitic association.
3. Experimental procedures Noble gas analyses were conducted on four polycrystalline (framesite) diamonds with each aggregate measuring approximately 2 2 1 cm in size. The diamonds in each sample exhibit variable, sub-millimetre grain sizes. In an effort to remove any silicate contaminants, the diamond samples were initially coarsely crushed into millimetre-size fragments, leached with 7% HF, and ultrasonically cleaned in
M. Honda et al. / Chemical Geology 203 (2004) 347–358
distilled water and analytical grade acetone. After bake-out at 200 jC for at least 12 h, gases were extracted by step-heating or by vacuum crushing at room temperature. The procedures for noble gas extraction, purification and mass spectrometric analysis are essentially the same as those described by Honda et al. (1993) and Matsumoto et al. (1998). Typical blank levels both for step-heating and vacuum crushing experiments were 4He = 2 –6 10 11, 20Ne = 1– 4 10 12, 40Ar = 4 –9 10 9, 84Kr = 1 – 5 10 13, and 132Xe = 1– 4 10 13 cm3 STP. There were no differences in blank amounts, regardless of gas extraction methods (heating or vacuum crushing) or extraction temperatures. This probably indicates that blanks were dominated by background in the sample handling system. Within uncertainties, noble gas isotopic ratios in blank runs were atmospheric.
4. Results Table 1 lists the noble gas results obtained from the four Jwaneng framesites both by step-heating and vacuum crushing (the character ‘H’ or ‘C’, respectively, is added to the sample name to identify the method of gas extraction). The total noble gas concentrations of the Jwaneng framesites are very high compared with those in other mantle-derived materials. For example, total amounts of 22Ne in the Jwaneng framesites released by step-heating range from 1.0 10 10 to 1.1 10 9 ccSTP/g, whereas average 22 Ne concentrations in MORB, OIB glasses and ultramafic xenoliths are 7 10 11, 8 10 11 and 3 10 11 ccSTP/g, respectively (Ozima and Podosek, 2002). The majority of noble gases were released in the low temperature fractions of step-heating experiments or in the first fraction of the vacuum crushing experiments (e.g., sample Jwaneng 2C). In order to examine the characteristics of noble gas release from Jwaneng framesites in more detail, sample Jwaneng 4H was heated incrementally at 600, 1000, 1800 and 2000 jC (Fig. 1). In this case, more than half of the 22Ne, 36Ar, 84Kr and 130Xe were released at the lowest temperature fraction of 600 jC and much of the remainder was released at 1000 jC. Noble gas isotopic compositions at these lower temperatures (600 and 1000 jC) are close to atmospheric values (Table 1). In contrast, the highest temperature
349
steps (1800 and 2000 jC) yielded virtually all the radiogenic 4He and 40Ar*, fissiogenic 136Xe* and excess 129Xe relative to atmospheric. 4.1. Helium The total helium concentrations, released from the Jwaneng framesite samples by step-heating, range from 55 to 173 10 6 ccSTP/g, with the majority of helium being released on graphitisation of the diamonds (f 2000 jC; Fig. 1). Helium concentrations from vacuum crushing experiments are two orders of magnitude lower than the total helium concentrations released from the step-heating experiments. 3 He/4He ratios observed in the framesites, from both step-heating and vacuum crushing experiments, are less than the atmospheric value of 1.4 10 6 and range from 0.81 RA to 0.024 RA (where RA is the atmospheric 3He/4He ratio). Thus, these samples do not contain significant amounts of primordial mantlederived 3He as observed in some diamonds (Ozima and Zashu, 1983, 1988, 1991; Wada and Matsuda, 1998). The enrichment of radiogenic helium in the current samples is consistent with previous observations from Jwaneng framesites by Burgess et al. (1998). These authors reported uranium contents of 114 and 33 ppb for two Jwaneng framesites and suggested that uranium is likely to be in microinclusions within framesites. Using an emplacement age of 235 Ma for the Jwaneng kimberlite (Kinny et al., 1989) and a U content of 100 ppb (with Th/U = 3), the amount of in situ produced radiogenic 4He in the current framesite samples is calculated to be 5 10 6 ccSTP/g. This amount corresponds to less than 10% of the total 4He released from the samples during stepheating. On the other hand, if the formation age of the Jwaneng framesites is 2.9 Ga, as postulated from the Re – Os ages of eclogite inclusions (Shirey et al., 2001), then virtually all radiogenic 4He could have been produced by the decay of U and Th in the samples. This subject will be discussed further in later sections by deconvolving radiogenic, nucleogenic and fissiogenic components in the samples. 4.2. Neon Most neon was released in the lower temperature fractions of the step-heating experiments (Fig. 1) or
350
Table 1 Noble Gas Isotopic Compositions of Jwaneng framesite diamondsa Sampleb
Jwaneng 2C (0.525 g) Jwaneng 2H (0.805 g) Jwaneng 3C (0.765 g) Jwaneng 3H (0.507 g) Jwaneng 4C (0.735 g) Jwaneng 4H (0.464 g)
Air
3 He/4He ( 10 6)
22
Ne ( 10 12)
20
1700 2000 Total 1 2 Total 1800 2000 Total
1.85 (4) 53.1 (11) 54.9 (11) 2.5 (1) 0.87 (2) 3.4 (1) 7.0 (1) 166.1 (33) 173.1 (33) 0.36 (1)
0.33 (2) ? – 1.13 (4) 0.47 (5) 0.96 (3) 0.40 (1) 0.034 (1) 0.050 (1) 0.35 (4)
1167 (14) 1.8 (1) 1169 (14) 61.4 (11) 4.7 (8) 65.9 (14) 142 (7) b.b. 142 (7) 14.0 (3)
9.78 8.70 9.78 9.74 9.81 9.74 9.62 b.b. 9.62 9.75
(5) (3) (5) (9) (28) (9) (13)
1800 2000 Total
4.9 (1) 77.1 (15) 82.0 (16) 4.0 (1)
0.84 (4) 0.109 (4) 0.16 (1) 0.26 (1)
93.3 (12) 8.1 (13) 101.5 (19) 105.5 (18)
600 1000 1800 2000 Total
0.226 (3) 0.273 (3) 11.8 (1) 154.5 (2) 166.0 (2)
b.b. b.b. 0.15 (1) 0.011 (2) – 1.40
132 (2) 106 (2) 6.9 (4) 6.9 (5) 251 (3)
Ne/22Ne
21
Ne/22Ne
36
Ar ( 10 9)
40
Ar/36Ar
84
Kr ( 10 12)
130 Xe ( 10 12)
128
Xe/130Xe
864 (3) 24 325 (566) 2063 (49) 3218 (31) 22 926 (4460) 3458 (78) 1583 (12) 33 241 (1348) 5275 (224) 1184 (12)
217 (14) 3.8 (3) 221 (14) 99 (6) 2.4 (2) 100 (6) 179 (11) 11.1 (7) 190 (11) 11.3 (7)
4.1 (2) 0.15 (1) 4.3 (2) 1.5 (1) 0.07 (1) 1.6 (1) 2.6 (1) 0.63 (3) 3.3 (1) 0.24 (2)
0.472 0.499 0.473 0.482 0.453 0.480 0.477 0.493 0.480 0.497
(8) (16) (8) (9) (34) (8) (8) (8) (6) (22)
129
Xe/130Xe
6.55 8.33 6.61 6.93 8.14 6.98 6.77 8.26 7.06 6.86
(8) (21) (8) (8) (42) (8) (13) (9) (10) (28)
131
Xe/130Xe
5.26 5.56 5.27 5.12 6.15 5.17 5.27 5.28 5.27 5.41
(6) (13) (6) (5) (32) (5) (6) (5) (5) (24)
132
Xe/130Xe
6.68 7.42 6.71 6.64 7.40 6.67 6.67 7.07 6.75 6.66
(5) (14) (5) (4) (25) (4) (5) (4) (4) (26)
134
Xe/130Xe (3) (9) (3) (3) (19) (3) (3) (4) (2) (13)
Xe/130Xe
(3) (54) (3) (5) (33) (6) (8)
(13) (8)
0.0294 0.0839 0.0295 0.0284 0.0277 0.0284 0.0316 b.b. 0.0316 0.0284
(8) (5)
6.4 (1) 0.34 (1) 6.7 (1) 3.8 (1) 0.05 (1) 3.9 (1) 3.53 (7) 0.47 (2) 4.01 (8) 0.55 (1)
9.85 7.19 9.64 9.78
(6) (57) (8) (6)
0.0301 0.1058 0.0362 0.0309
(3) (118) (13) (3)
2.78 (5) 0.11 (1) 2.89 (5) 2.84 (5)
1710 (6) 53 070 (2871) 3638 (155) 1572 (6)
105 (7) 2.3 (2) 107 (7) 99 (6)
1.61 (9) b.b. 1.61 (9) 1.25 (7)
0.449 (14) b.b. 0.449 (14) 0.461 (15)
6.47 (16) b.b. 6.47 (16) 6.47 (14)
5.08 (14) b.b. 5.08 (14) 5.27 (11)
6.41 (15) b.b. 6.41 (15) 6.74 (12)
2.51 (7) b.b. 2.51 (7) 2.61 (5)
2.08 (7) b.b. 2.08 (7) 2.23 (5)
9.79 9.64 8.14 7.75 9.63 9.80
(6) (6) (22) (30) (4)
0.0282 (3) 0.0281 (3) 0.0865 (21) 0.105 (5) 0.0318 (3) 0.0290
3.82 (3) 3.07 (2) 0.155 (5) 0.31 (1) 7.35 (4)
301 (2) 364 (2) 24 800 (730) 47 200 (1830) 2830 (109) 295.5
110 (5) 86.4 (4) 7.1 (3) 3.8 (2) 207 (6)
2.3 (1) 0.95 (4) 0.19 (9) 0.14 (2) 3.6 (1)
0.470 (7) 0.467 (8) 0.485 (14) 0.22 (14) 0.460 (7) 0.472
6.50 (5) 6.48 (6) 6.67 (8) 7.01 (17) 6.52 (6) 6.496
5.21 (6) 5.24 (7) 5.10 (17) 5.73 (15) 5.23 (5) 5.213
6.60 (7) 6.56 (6) 6.98 (7) 9.89 (23) 6.74 (4) 6.607
2.57 (2) 2.52 (3) 2.93 (5) 4.95 (24) 2.67 (2) 2.563
2.19 (2) 2.18 (3) 2.58 (6) 4.29 (23) 2.29 (2) 2.176
b.b.: below procedural blank. ?: 3He peak was not found. a 38 36 Noble gas amounts are in ccSTP/g. Quoted errors are one standard deviation, shown in parentheses as the last digits. Ar/ Ar and Kr isotopic ratios in the samples are atmospheric within uncertainties. b H: heated for 30 min, C: crushed for 300 strokes. Jwaneng 2C was crushed twice successively.
2.59 3.37 2.62 2.74 3.30 2.77 2.65 3.38 2.80 2.73
136
2.20 2.95 2.22 2.39 2.92 2.41 2.26 3.08 2.42 2.12
(3) (8) (3) (3) (18) (3) (3) (4) (3) (11)
M. Honda et al. / Chemical Geology 203 (2004) 347–358
Jwaneng 1H (0.523 g)
4 He ( 10 6)
M. Honda et al. / Chemical Geology 203 (2004) 347–358
351
Fig. 1. Noble gas release by step-heating from Jwaneng framesite 4H. More than half of the total amounts of 22Ne, 36Ar and 130Xe were released from the lowest temperature fraction (600 jC). Gas release was virtually complete at 1000 jC. In contrast, radiogenic 4He and 40Ar* and fissiogenic 136Xe* were mostly released during graphitisation of diamond at 2000 jC.
during the first fraction of vacuum crushing (sample Jwaneng 2C). The isotopic compositions in these fractions are atmospheric within the uncertainties. In contrast, the neon isotopic compositions of higher temperature step-heating fractions are non-atmospheric and have close affinities with crustal nucleogenic neon. In Fig. 2, the neon isotopic results from the Jwa-
neng framesites are plotted in 20 Ne/22Ne – 21Ne/ 22 Ne space. Neon data from the high temperature fractions of samples Jwaneng 1H, 3H and 4H plot close to a mixing line between atmospheric and crustal nucleogenic neon (e.g., Kennedy et al., 1990). Crustal nucleogenic neon, which has been found in natural gases and brines, comprises a mixture of nucleogenic
Fig. 2. Neon isotope diagram, showing the results from four Jwaneng framesites. The MORB correlation line of Moreira et al. (1998) is also plotted. Neon data from the high temperature fractions of samples 1H, 3H and 4H lie close to a mixing line between atmospheric neon and crustal nucleogenic neon (e.g., Kennedy et al., 1990). The nucleogenic neon observed in the Jwaneng framesites is likely to have been produced in the crust and subsequently incorporated into the samples.
352
M. Honda et al. / Chemical Geology 203 (2004) 347–358
neon isotopes produced from nuclear reactions 18 O(a,n)21Ne, 19F(a, n)22Na(h+/2.6a) ! 22Ne and 19 F(a,p)22Ne in crustal material with a constant O/ F (atomic) ratio of about 100 (Kennedy et al., 1990). Crustal nucleogenic neon has also been found in Archean anorthosites in West Greenland (Azuma et al., 1993) and in the continental metamorphic rocks (gneiss and amphibolites) collected from the KTB deep bore hole in NE Bavaria, Germany (Drescher et al., 1998). It is highly unlikely that the crustal nucleogenic neon observed in the Jwaneng framesites is a product of in situ nuclear reactions, for the following reasons. First, it would be fortuitous that the framesites have the specific O/F abundance ratio of f 100 to produce crustal-like nucleogenic neon within the samples. Second, given a crustal composition (e.g., oxygen f 47 w.t%) and U content of 100 ppb, the amount of nucleogenic 21Ne produced over 235 Ma (the
emplacement age for the Jwaneng kimberlite) and 2.9 Ga (Re – Os model age) is calculated to be 0.2– 4 10 12 ccSTP/g (Yatsevich and Honda, 1997), which is comparable to the integrated contents of nucleogenic 21Ne extracted by step-heating experiments for the samples (0.4 – 1.0 10 12 ccSTP/g; Table 2). However, as the bulk oxygen abundance in the framesites must be several orders of magnitude lower than that used in the above calculation, the quantity of nucleogenic neon generated within the framesites must be negligible compared to the amounts actually observed. It could be argued that the crustal-like nucleogenic neon observed in the Jwaneng framesites originated from the surrounding lithospheric mantle, which had undergone earlier enrichment in lithophile elements, resulting in elevated 21Ne and 22Ne abundances. However, we believe that this scenario cannot account for the observations of fluorine produced-nucleogenic
Table 2 Radiogenic, nucleogenic and fissiogenic components in Jwaneng framesite diamonds (in ccSTP/g)a 4 He ( 10 6)
21 Ne* ( 10 12)
21
Ne*/4He ( 10 8)
40 Ar* ( 10 6)
40
1700C 2000C Total 1 2 Total 1800C 2000C Total
1.85 (4) 53.1 (11) 54.9 (11) 2.5 (1) 0.87 (2) 3.4 (1) 7.0 (1) 166.1 (33) 173.1 (33) 0.36 (1)
0.53 (35) 0.102 (14) 0.65 (35) – – – 0.44 (13) – 0.44 –
29 (19) 0.19 (3) 1.2 (6) – – – 6.3 (19) – 0.26 –
3.62 (7) 8.23 (31) 11.8 (4) 11.3 (3) 1.1 (3) 12.4 (4) 4.5 (1) 15.4 (9) 20.0(10) 0.48 (1)
1800C 2000C Total
4.9 (1) 77.1 (15) 82.0 (16) 4.0 (1)
0.088 (31) 0.69 (17) 0.77 (14) 0.21 (4)
1.8 (6) 0.89 (22) 0.94 (17) 5.2 (11)
600C 1000C 1800C 2000C Total
0.226 (3) 0.273 (3) 11.8 (1) 154.5 (2) 166.0 (2)
– – 0.43 (4) 0.57 (6) 1.0
– – 3.6 (3) 0.37 (4) 0.6
Sample Jwaneng 1H (0.523 g) Jwaneng 2C (0.525 g) Jwaneng 2H (0.805 g) Jwaneng 3C (0.765 g) Jwaneng 3H (0.507 g) Jwaneng 4C (0.735 g) Jwaneng 4H (0.464 g)
Productionb (1) Instantaneous (2) Integrated for 4.5 Ga
4.5c 4.5c
Ar*/4He
136 Xe* ( 10 15)
136
1.96 (5) 0.16 (1) 0.22 (1) 4.5 (1) 1.2 (3) 3.7 (1) 0.65 (2) 0.09 (1) 0.12(1) 1.35 (4)
99 (125) 115 (14) 188 (13) 327 (50) 52 (15) 374 (53) 218 (81) 566 (39) 795 (95) –
53 (68) 2.2 (3) 3.4 (2) 129 (20) 61 (18) 110 (16) 31 (12) 3.4 (2) 4.6 (6) –
3.93 (7) 5.7 (5) 9.7 (5) 3.62 (7)
0.80 0.07 0.12 0.91
– – – 68 (59)
– – – 17 (15)
0.020 (8) 0.21 (1) 3.8 (2) 14.6 (8) 18.6 (8)
0.09 (4) 0.77 (3) 0.32 (1) 0.095 (5) 0.112 (5)
27 (57) 4 (29) 78 (11) 290 (47) 399 (81)
121 (254) 15 (105) 6.6 (10) 1.9 (3) 2.4 (5)
0.23 0.59
(2) (1) (1) (1)
Xe*/4He ( 10 9)
3.1 2.6
a The radiogenic, nucleogenic and fissiogenic amounts in the samples are calculated by subtracting atmospheric components. Quoted errors are one standard deviation, shown in parentheses as the last digits. b Production ratios are calculated by assuming a Th/U ratio of 3 and a K/U ratio of 12 000. c From Yatsevich and Honda (1997).
M. Honda et al. / Chemical Geology 203 (2004) 347–358 22
Ne in the samples. Average fluorine concentrations in the mantle are estimated to be 25 times lower than in the crust (Mason and Moore, 1982; McDonough and Sun, 1995), whereas oxygen abundances are approximately the same in the mantle and crust. Thus, in order to produce the crustal-like nucleogenic 21 Ne*/22Ne* ratio observed in the samples, the surrounding lithospheric mantle must have been enriched in fluorine concentration by 25 times the average mantle abundance. The consistency in noble gas and halogen (Cl, Br and I) abundances between African (including Jwaneng) diamonds and MORB samples (Johnson et al., 2000) suggests a MORB-like mantle source for lithospheric volatiles, and there is no evidence of an enrichment in halogen concentrations. Therefore, we suggest that the nucleogenic neon observed in the Jwaneng framesites is likely to have been produced in the crust and subsequently incorporated into the framesites during their crystallisation in the subcontinental mantle. Mantle neon is thought to have an end-member 20 Ne/22Ne ratio of 13.7 (Honda et al., 1991; Moreira et al., 1998) or 12.5 (Trieloff et al., 2002, 2000), and is the only known neon component with a 20Ne/22Ne ratio that is higher than the atmospheric value. However, all 20Ne/22Ne ratios observed in the Jwaneng framesites were close to or lower than the atmospheric value of 9.8 (Table 1), indicating that the Jwaneng framesites contain no detectable mantle neon. In contrast, some cubic diamonds contain elevated 20 Ne/22Ne ratios attributed to the presence of a mantle neon component (Ozima and Zashu, 1988, 1991; Wada and Matsuda, 1998). Thus, it seems likely that any mantle neon in the Jwaneng framesites was overwhelmed by the addition of crustal nucleogenic neon (higher temperature fractions) or by atmospheric neon (lower temperature fractions). This observation is consistent with the paucity of primordial 3He in the Jwaneng framesites. 4.3. Argon 38
Ar/36Ar ratios of the Jwaneng framesites are atmospheric within the uncertainties. Integrated 40 Ar/36Ar ratios vary between 1572 and 5275, significantly higher than the atmospheric value of 295.5. The 2000 jC graphitisation steps produced the highest 40 Ar/36Ar ratios, with values up to 53 000 (sample
353
Jwaneng 3H). These high 40Ar/36Ar ratios could result from the addition of in situ radiogenic argon and/or the presence of excess 40Ar. As with neon, the majority of 36Ar was released in the first step-heating and vacuum crushing experiments, resulting in relatively low 40Ar/36Ar ratios. This indicates that significant amounts of atmospheric argon was present within low-retention sites in the diamond samples. 4.4. Krypton The krypton isotopic ratios measured in the Jwaneng framesites are atmospheric at the one sigma level, and are not listed in Table 1. 4.5. Xenon As with neon and argon, most xenon was released in the lower temperature step-heating or first vacuum crushing fractions and is characterised by atmospheric isotopic compositions (Table 1). The xenon isotopic compositions of the higher temperature stepheating (samples Jwaneng 1H, 2H and 4H) and the second vacuum crushing (sample Jwaneng 2C) fractions, exhibit xenon isotopic anomalies relative to atmospheric compositions. On a 129Xe/130Xe vs. 136 Xe/130Xe diagram, the data for Jwaneng samples 1H, 2H and 2C plot adjacent to the MORB correlation line (slope = 3.13 F 0.08, Kunz et al., 1998, Fig. 3). The 129Xe isotope anomalies observed in samples Jwaneng 1H, 2H and 2C are comparable to those in mantle-derived samples, including MORBs (Kunz et al., 1998; Marty, 1989; Moreira et al., 1998; Staudacher and Alle`gre, 1982), Loihi dunites and Icelandic picritic glasses (Trieloff et al., 2000). Excess 129Xe, relative to the atmospheric value, observed in mantle-derived samples is generally attributed to radioactive decay of an extinct nuclide 129I (half life of 16 Ma), once present in the Earth (e.g., Ozima and Podosek, 2002). However, a similar excess in 129Xe, accompanied by excess 131 – 136Xe (as observed in the Jwaneng framesites), has recently been reported for continental metamorphic rocks (gneiss and amphibolites) collected from the KTB deep bore hole (Drescher et al., 1998), and in Archean cherts in the metasedimentary sequences of Pilbara Craton, Western Australia (Pinti et al., 2001). Drescher et al. (1998) and Pinti et al. (2001) postulated that the excess
354
Fig. 3. Plot of 129Xe/130Xe versus Kunz et al. (1998) is also plotted. 129
M. Honda et al. / Chemical Geology 203 (2004) 347–358
136
Xe/130Xe observed in framesite diamonds from the Jwaneng kimberlite. The MORB correlation line of
Xe in these rocks was in fact due to nucleogenic production from tellurium 128 Te(n,g) 129 Te(h/ 69m) ! 129I(h/69Ma) ! 129Xe in the crust. The observation of crustal-like noble gases (e.g., radiogenic helium and argon, nucleogenic neon) and the apparent absence of mantle noble gases (e.g., primordial 3He and solar-like neon) in the Jwaneng framesites are somewhat analogous to the compositions reported for the KTB metamorphic rocks. Therefore, one interpretation is that the 129Xe excesses observed in the Jwaneng framesites represent crustal nucleogenic xenon, rather than mantle xenon. It is also important to point out that the 131 Xe excesses in the Jwaneng framesites are systematically higher than those observed in MORBs (Kunz et al., 1998), possibly also reflecting the nucleogenic production of 131Xe from tellurium 130 Te(n,g)131Te(h/25m) ! 131I(h/8d) ! 131Xe. In addition, the (129Xe/131Xe)excess ratios in the Jwaneng framesites are consistent with the range of nucleogenic 129Xe/131Xe ratios produced from tellurium by neutron capture (0.7 – 4.0; Browne and Berman, 1973), thus supporting a nucleogenic origin for the excess 129Xe. It must be stressed, however, that it would require reasonably high neutron fluxes and tellurium concentrations in the crust to provide the level of 129Xe and 131Xe excesses observed in the Jwaneng framesites.
Thus, although a crustal nucleogenic origin for excess 129Xe in the Jwaneng framesites requires verification, particularly in relation to the absolute production of nucleogenic xenon in the crust (Drescher et al., 1998; Pinti et al., 2001), we tentatively suggest that the Jwaneng framesites do not contain detectable amounts of mantle xenon. This argument is consistent with the observation of crustal neon in the framesites. As discussed earlier, it is likely that some fraction of the radiogenic 4He in the samples results from in situ production—therefore, a corresponding amount of in situ fissiogenic 136Xe also exists. Consequently, the 136Xe in the Jwaneng framesites can be interpreted as a mixture of crustal (dominated by fissiogenic xenon produced from 238 U), in situ produced fissiogenic xenon and atmospheric components.
5. Discussion As shown in the previous section, the noble gas compositions of the Jwaneng framesites have unusual characteristics: (1) large quantities of atmospheric noble gases associated with the first step-heating and vacuum crushing fractions; (2) no detectable MORBlike (mantle) neon and xenon; and (3) large quantities of radiogenic 4He and 40Ar, nucleogenic neon and
M. Honda et al. / Chemical Geology 203 (2004) 347–358
possibly xenon and fissiogenic xenon, released in the higher temperature step-heating or second vacuum crushing experiments. In this section, we further examine these unusual noble gas compositions. 5.1. Atmospheric noble gases in Jwaneng framesites The step-heating experiments demonstrated that large quantities of atmospheric noble gases appear to be loosely trapped within the framesites (Fig. 1). In contrast, radiogenic 4He and 40Ar, nucleogenic neon and xenon and fissiogenic xenon were released from the framesites during graphitisation at f 2000 jC. Thus, it is unlikely that the atmospheric noble gas components reflect an intrinsic noble gas component associated with framesite crystallisation in the subcontinental mantle. We suggest that the atmospheric noble gases in the framesites are likely to be secondary contaminants. As the framesites contain traces of mineral intergrowths, a possible explanation is that the atmospheric noble gas components were released from altered mineral phases that were not removed by acid treatment. Small amounts of atmospheric noble gases were, however, also noted at 2000 jC (Fig. 2), possibly due to incomplete outgassing of altered mineral phases during the lower temperature step-heating experiments. This possibility will be addressed in future studies by crushing the framesites into smaller fragments, more rigorous acid cleaning and SEM examination of framesite surfaces. 5.2. Radiogenic, nucleogenic and fissiogenic noble gases in Jwaneng framesites The combined results for radiogenic, nucleogenic and fissiogenic noble gas components observed in the Jwaneng framesites are presented in Table 2. In this tabulation, we assume that the total amounts of 4 He released from the samples are radiogenic. In order to calculate the amounts of nucleogenic 21Ne (21Ne*), radiogenic 40Ar (40Ar*) and fissiogenic 136 Xe (136Xe*), corrections for atmospheric noble gas compositions are made according to the following equation: X * ¼ Xref fðX =Xref Þobserved ðX =Xref Þatmospheric g ð1Þ
355
where X = 21Ne, 40Ar or 136Xe and Xref = 20Ne, 36Ar or 130Xe, respectively. Abundance ratios of 21Ne*, 40 Ar* and 136Xe*, relative to 4He, as well as the theoretical 21Ne*/4He (Yatsevich and Honda, 1997) and 40Ar*/4He ratios, calculated for both instantaneous and integrated production for 4.5 Ga in the crust, are also listed in Table 2. A precise value of [Esf 136Ysf] ( = 6.8 10 18/a) has been determined from the study of U-bearing accessory minerals (Ragettli et al., 1994), where Esf and 136Ysf are the decay constant for spontaneous fission of 238U and the fission yield of 136Xe, respectively. These values were used to calculate the 136Xe* production in the current samples, assuming a Th/U ratio of 3 and a K/U ratio of 12,000. It is noted that the production ratios of 40Ar*/4He and 21Ne*/4He are similar for both the crust and the mantle, because the relative abundances of the parent elements (U, Th and K, and oxygen—the main target element to produce nucleogenic 21Ne) are believed to be analogous in the two regions (e.g., Mason and Moore, 1982; Sun and McDonough, 1989). The 21Ne*/4He ratios for Jwaneng samples 4C (crushing) and 4H (1800 jC step) are 5.2 10 8 and 3.6 10 8, respectively. These abundance ratios are consistent with the theoretically calculated 21 Ne*/4He production ratio ( = 4.7 10 8) for the crust. Noble gases, released from the vacuum crushing and 1800 jC experiments, probably represent the noble gases trapped in fluid inclusions and defects within the framesites, and are dominated by crustal noble gas compositions. Thus, the consistency between 21Ne*/4He ratios obtained by vacuum crushing (sample 4C) and step-heating (1800 jC step; sample 4H), and the theoretical value suggests that the degree of elemental fractionation between 4He and 21Ne* is insignificant during transportation of crustal noble gases into the mantle and encapsulation in the Jwaneng framesites. On the other hand, the integrated 21Ne*/4He ratio ( = 0.6 10 8) for sample 4H (step-heating) is systematically lower than the theoretical crustal value. Integrated 21Ne*/4He ratios for Jwaneng diamonds 1H, 2H and 3H are also lower than the theoretical value (Table 2). As previously discussed in Section 4.2, it is unlikely that appreciable amounts of nucleogenic neon can be produced in situ even if uranium and thorium are unusually abundant, and we have
356
M. Honda et al. / Chemical Geology 203 (2004) 347–358
suggested that nucleogenic neon was produced in the continental crust and subsequently incorporated into the framesites. As radiogenic 4He and nucleogenic 21 Ne do not appear to have been fractionated significantly when incorporated into the Jwaneng framesites, the difference in integrated 21Ne*/4He ratios between the step-heating experiments and the theoretical value can be used to estimate the fraction of in situ produced radiogenic 4He in the samples. On this basis, 74%, 94%, 80% and 88% of the total radiogenic 4 He observed in samples 1H, 2H, 3H and 4H, respectively, are estimated to be in situ products. The remaining radiogenic 4He in the samples can be regarded as inherited, probably crustal. This implies that the formation age of the Jwaneng framesites is older than the emplacement age of the Jwaneng kimberlite (see Section 4.1). As the integrated 136Xe*/4He ratios of the Jwaneng framesites obtained from step-heating experiments are consistent with the theoretical production ratio (Table 2), the fractions of in situ produced fissiogenic 136Xe from 238U are the same as those estimated for radiogenic 4He. The remaining 136Xe* in the samples can be interpreted as fissiogenic 136Xe produced in the crust. The integrated 40Ar*/4He ratios for the Jwaneng framesites are lower than 40 Ar*/4He* production ratios calculated for the crust and mantle (K/U = 12,000, Th/U = 3; Table 2). If the 40 Ar* in the Jwaneng framesites is dominated by in situ produced radiogenic 40Ar (i.e. analogous to radiogenic 4He), this would imply that K/U ratios in the Jwaneng framesites are lower than the average K/U abundance ratio in the crust and mantle.
sites suggests that these volatiles were transported into the diamond-forming region of the subcontinental mantle, possibly by subduction-related processes. This may imply that the framesites formed in a different environment to the majority of macro-crystalline diamonds and that framesites represent a distinct diamond population. Comparative studies of noble gases in macro-crystalline diamonds from the same kimberlite pipe are thus important for exploring this aspect further. The unusual carbon isotopic and mineral intergrowth compositions of Jwaneng framesites led Kirkley et al. (1994) to conclude that these diamonds formed under carbon supersaturation conditions during subduction of oceanic sediments and intercalated slab components. The incorporation of subducted crustal material into the mantle is an attractive explanation for the presence of crustal noble gases in diamonds. Although the parageneses of the framesites in this study were not constrained, they are likely to be eclogitic, as eclogitic diamonds are generally believed to have been formed from recycled carbon (e.g., Kirkley et al., 1991). Recent Re – Os analyses of eclogite xenoliths and eclogitic inclusions in diamonds from several southern African kimberlites suggest that subduction-related crustal recycling may have been a viable process during continent formation in the Archean (f 2.9 Ga) and resulted in widespread formation of eclogitic diamonds at that time (Richardson et al., 2001a; Shirey et al., 2001). If this is true, then crustal noble gases could have been introduced into the mantle at a very early stage in Earth history.
5.3. Crustal noble gases in the Jwaneng framesites 6. Conclusions The Jwaneng framesites likely formed in the subcontinental lithospheric mantle and acquired their volatiles, including noble gases, during crystallisation. As diffusion of noble gases in diamond is expected to be very slow (e.g., Honda et al., 1987; Ozima, 1989), late incorporation of noble gases during crustal residence of diamonds is highly unlikely. We are not aware of any experimental or other evidence that questions this premise. Hence, the observation of crustal noble gases (nucleogenic neon and possibly xenon, some of radiogenic 4He and 40 Ar and fissiogenic xenon) in the Jwaneng frame-
We have undertaken a full suite of noble gas analyses of four framesite diamonds using vacuum crushing and step-heating experimental techniques. We found that the Jwaneng framesites have unique characteristics in terms of noble gas elemental and isotope abundances: (1) large quantities of atmospheric noble gases, (2) no detectable MORB-like (mantle) neon, and (3) large quantities of radiogenic 4 He and 40Ar, nucleogenic neon and fissiogenic xenon and excess 129Xe, released in the higher temperature step-heating or second vacuum crushing
M. Honda et al. / Chemical Geology 203 (2004) 347–358
experiments; these compositions are analogous to those found in some continental metamorphic rocks (Drescher et al., 1998) and in Archean cherts (Pinti et al., 2001). Large quantities of atmospheric noble gases were released from lower temperature and initial vacuum crushing experiments. We interpret the atmospheric noble gases in the Jwaneng framesites as secondary, possibly derived from silicate and other mineral phases intergrown with the framesites. Although further authentication is necessary, we tentatively conclude that excess 129Xe in the Jwaneng framesites resulted from the addition of crustal nucleogenic xenon produced from neutron capture of tellurium in the crust (Drescher et al., 1998; Pinti et al., 2001) and that, as is the case for neon, there is no detectable mantle xenon in the samples. The observations of 21Ne/22Ne ratios consistent with the crustal nucleogenic production ratio strongly suggest that this component was produced in the crust and then incorporated into the Jwaneng framesites during their formation in the mantle. Crustal noble gases (nucleogenic neon and possibly xenon, some radiogenic 4He and 40Ar and fissiogenic xenon) could have been introduced into the diamond stability field of the subcontinental mantle by subduction-related processes. It is critically important to test this hypothesis through further investigations of eclogitic diamonds from various localities. Such studies will clarify whether crustal noble gas inventories can be retained during subduction of crustal material into the mantle lithosphere. The present work demonstrates the versatility of undertaking a full suite of noble gas isotope and elemental studies, using both vacuum crushing and step-heating methods. This approach allows deconvolution of the complex, multiple noble gas components (e.g., crustal, atmospheric and in situ) found in framesite diamonds.
Acknowledgements De Beers Consolidated Mines are thanked for donating the diamond specimens for this project and for allowing publication of this manuscript. We thank Mario Trieloff and an anonymous reviewer for their constructive comments on the manuscript. [PD]
357
References Alle`gre, C.J., 1997. Limitation on the mass exchange between the upper and lower mantle: the evolving convection regime of the Earth. Earth Planet. Sci. Lett. 150, 1 – 6. Alle`gre, C.J., Staudacher, T., Sarda, P., Kurz, M., 1983. Constraints on evolution of Earth’s mantle from rare gas systematics. Nature 303, 762 – 766. Azuma, S., Ozima, M., Hiyagon, H., 1993. Anomalous neon and xenon in an Archaean anorthosite from West Greenland. Earth Planet. Sci. Lett. 114, 341 – 352. Browne, J.C., Berman, B.L., 1973. Neutron-capture cross sections for 128Te and 130Te and the xenon anomaly in old tellurium ores. Phys. Rev. C 8, 2405 – 2411. Burgess, R., Johnson, L.H., Mattey, D.P., Harris, J.W., Turner, G., 1998. He, Ar and C isotopes in coated and polycrystalline diamonds. Chem. Geol. 146, 205 – 217. Davies, G.F., 1999. Geophysically constrained mass flows and the 40 Ar budget: a degassed lower mantle? Earth Planet. Sci. Lett. 166, 149 – 162. Deines, P., Harris, J.W., Gurney, J.J., 1997. Carbon isotope ratios, nitrogen content and aggregation state, and inclusion chemistry of diamonds from Jwaneng, Botswana. Geochim. Cosmochim. Acta 61, 3993 – 4005. Dixon, E.T., Honda, M., McDougall, I., Campbell, I.H., Sigurdsson, I., 2000. Preservation of near-solar neon isotopic ratios in Icelandic basalts. Earth Planet. Sci. Lett. 180, 309 – 324. Drescher, J., Kirsten, T., Scha¨fer, K., 1998. The rare gas inventory of the continental crust, recovered by the KTB Continental Deep Drilling Project. Earth Planet. Sci. Lett. 154, 247 – 263. Harte, B., Harris, J.W., Hutchison, M.T., Watt, G.R., Wilding, M.C., 1998. Mineral facies and source materials for lower mantle inclusions in diamonds from Sao Luiz, Brazil. Min. Mag. 62A, 575 – 576. Honda, M., Reynolds, J.H., Roedder, E., Epstein, S., 1987. Noble gases in diamonds: occurrences of solarlike helium and neon. J. Geophys. Res. 92, 12507 – 12521. Honda, M., McDougall, I., Patterson, D.B., Doulgeris, A., Clague, D.A., 1991. Possible solar noble gas component in Hawaiian basalts. Nature 349, 149 – 151. Honda, M., McDougall, I., Patterson, D.B., Doulgeris, A., Clague, D.A., 1993. Noble gases in submarine pillow basalt glasses from Loihi and Kilauea, Hawaii: a solar component in the Earth. Geochim. Cosmochim. Acta 57, 859 – 874. Johnson, L.H., Burgess, R., Turner, G., Milledge, H.J., Harris, J.W., 2000. Noble gas and halogen geochemistry of mantle fluids: comparison of African and Canadian diamonds. Geochim. Cosmochim. Acta 64, 717 – 732. Kamijo, K., Hashizume, K., Matsuda, J.-I., 1998. Noble gas constraints on the evolution of the atmosphere – mantle system. Geochim. Cosmochim. Acta 62, 2311 – 2321. Kellogg, L.H., Hager, B.H., Hilst, R.D.v.D., 1999. Compositional stratification in the deep mantle. Science 283, 1881 – 1884. Kennedy, B.M., Hiyagon, H., Reynolds, J.H., 1990. Crustal neon: a striking uniformity. Earth Planet. Sci. Lett. 98, 277 – 286. Kinny, P.D., Compston, W., Bristow, J.W., Williams, I.S., 1989.
358
M. Honda et al. / Chemical Geology 203 (2004) 347–358
Archean mantle xenocrysts in a Premian kimberlite: two generations of kimberlitic zircon in Jwaneng DK2, southern Botswana. In: Ross, N.Kimberlites and Related Rocks. Geol. Soc. Australia Special Pub., Sydney, Australia, pp. 833 – 842. Kirkley, M.B., Gurney, J.J., Otter, M.L., Hill, S.J., Daniels, L.R., 1991. The application of C isotope measurement to the identification of the sources of C in diamonds: a review. Appl. Geochem. 6, 447 – 494. Kirkley, M.B., Gurney, J.J., Rickard, R.S., 1994. Jwaneng framesites: carbon isotopes and intergrowth compositions. In: Meyer, H.O.A., Leonardos, O.H. (Eds.), Diamonds: characterisation, Genesis and Exploration. CPRM Spec. Publ. Companhia de Presquisa de Recursos Minerais, Brasilia, Brazil, pp. 127 – 135. Kunz, J., Staudacher, T., Alle`gre, C.J., 1998. Plutonium-fission xenon found in Earth’s mantle. Science 280, 877 – 880. Kurz, M.D., Jenkins, W.J., Hart, S.R., Clague, D., 1983. Helium isotopic variations in volcanic rocks from Loihi Seamount and the Island of Hawaii. Earth Planet. Sci. Lett. 66, 388 – 406. Marty, B., 1989. Neon and xenon isotopes in MORB: implications for the earth – atmosphere evolution. Earth Planet. Sci. Lett. 94, 45 – 56. Mason, B., Moore, C.B., 1982. Principles of Geochemistry. Wiley, New York. 344 pp. Matsumoto, T., Honda, M., McDougall, I., O’Reilly, S.Y., 1998. Noble gases in anhydrous lherzolites from the Newer Volcanics, south Australia: a MORB-like reservoir in the subcontinental mantle. Geochim. Cosmochim. Acta 62, 2521 – 2533. Matsumoto, T., Seta, A., Matsuda, J., Chen, Y., Arai, S., 2002. Helium in the Archaean komatiites revisited: significantly high 3 He/4He ratios revealed by fractional crushing gas extraction. Earth Planet. Sci. Lett. 196, 213 – 225. McDonough, W.F., Sun, S.-S., 1995. The composition of the Earth. Chem. Geol. 120, 223 – 253. Moreira, M., Kunz, J., Alle`gre, C., 1998. Rare gas systematics on popping rock: isotopic and elemental compositions in the upper mantle. Science 279, 1178 – 1181. Ozima, M., 1989. Gases in diamonds. Annu. Rev. Earth Planet. Sci. 17, 361 – 384. Ozima, M., Podosek, F.A., 2002. Noble Gas Geochemistry, 2nd ed. Cambridge Univ. Press, Cambridge, UK, 286 pp. Ozima, M., Zashu, S., 1983. Primitive helium in diamonds. Science 219, 1067 – 1068. Ozima, M., Zashu, S., 1988. Solar-type Ne in Zaire cubic diamonds. Geochim. Cosmochim. Acta 52, 19 – 25. Ozima, M., Zashu, S., 1991. Noble gas state of the ancient mantle as deduced from noble gases in coated diamonds. Earth Planet. Sci. Lett. 105, 13 – 27. Pinti, D., Matsuda, J., Maruyama, S., 2001. Anomalous xenon in Archean cherts from Pilbara Craton, Western Australia. Chem. Geol. 175, 387 – 395. Porcelli, D., Wasserburg, G.J., 1995. Mass transfer of He, Ne, Ar,
and Xe through a steady state upper mantle. Geochim. Cosmochim. Acta 59, 4921 – 4937. Ragettli, R.A., Hebeda, E.H., Signer, P., Wieler, R., 1994. Uranium – xenon chronology: precise determination of Esf*136Ysf for spontaneous fission of 238U. Earth Planet. Sci. Lett. 128, 653 – 670. Richard, D., Marty, B., Chaussidon, M., Arndt, N., 1996. Helium isotopic evidence for a lower mantle component in depleted Archean komatiite. Science 273, 93 – 95. Richardson, S., Gurney, J.J., Erlank, A.J., Harris, J.W., 1984. Origin of diamonds in old enriched mantle. Nature 310, 198 – 202. Richardson, S.H., Harris, J.W., Gurney, J.J., 1993. Three generations of diamonds from old continental mantle. Nature 366, 256 – 258. Richardson, S.H., Chinn, I.L., Harris, J.W., 1999. Age and origin of eclogitic diamonds from the Jwaneng kimberlite, Botswana. In: Gurney, J.J., Gurney, J.L., Pascoe, M.D., Richardson, S.H.The P.H. Nixon Volume. Red Roof Design, Cape Town, pp. 709 – 713. Richardson, S.H., Shirey, S.B., Harris, J.W., Carlson, R.W., 2001a. Archean subduction recorded by Re – Os isotopes in eclogitic sulfide inclusions in Kimberley diamonds. Earth Planet. Sci. Lett. 191, 257 – 266. Richardson, S.H., Shirey, S.B., Harris, J.W., Carlson, R.W., 2001b. The Life and Times of Multiple Generations of Diamonds From the Kaapvaal Craton Keel Slave-Kaapvaal Workshop. Merrickville, Canada. Shirey, S.B., et al., 2001. Archean emplacement of eclogitic components into the lithospheric mantle during formation of the Kaapvaal Craton. Geophys. Res. Lett. 28, 2509 – 2512. Staudacher, T., Alle`gre, C.J., 1982. Terrestrial xenology. Earth Planet. Sci. Lett. 60, 389 – 406. Sun, S.-S., McDonough, W.F., 1989. Chemical and isotopic systematics of oceanic basalts: implications for mantle composition and processes. In: Saunders, A.D., Norry, M.J.Magmatism in Ocean Basins. Geological Society Special Publication No. 42, Boulder, USA, pp. 313 – 345. Trieloff, M., Kunz, J., Clague, D.A., Harrison, D., Alle`gre, C.J., 2000. The nature of pristine noble gases in mantle plumes. Science 288, 1036 – 1038. Trieloff, M., Kunz, J., Alle`gre, C.J., 2002. Noble gas systematics of the Re´union mantle plume source and the origin of primordial noble gases in Earth’s mantle. Earth Planet. Sci. Lett. 200, 297 – 313. Wada, N., Matsuda, J., 1998. A noble gas study of cubic diamonds from Zaire: constraints on their mantle source. Geochim. Cosmochim. Acta 62, 2335 – 2346. Wyllie, P.J., 1995. Experimental petrology of upper mantle materials, processes and products. J. Geodyn. 20, 429 – 468. Yatsevich, I., Honda, M., 1997. Production of nucleogenic neon in the Earth from natural radioactive decay. J. Geophys. Res. 102, 10291 – 10298.
Unusual noble gas compositions in polycrystalline diamonds: preliminary results from the Jwaneng kimberlite, Botswana Masahiko Honda a,*, David Phillips b, Jeffrey W. Harris c, Igor Yatsevich a a
Research School of Earth Sciences, The Australian National University, Canberra, ACT 0200, Australia b School of Earth Sciences, The University of Melbourne, Melbourne, Victoria 3010, Australia c Division of Earth Sciences, University of Glasgow, Glasgow G12 8QQ, UK Received 4 September 2002; accepted 23 October 2003
Abstract We have undertaken noble gas analyses of four polycrystalline framesite diamonds from the Jwaneng kimberlite pipe, Botswana. These samples yielded complex, multiple noble gas components (crustal, atmospheric and in situ radiogenic/ fissiogenic), which were successfully deconvoluted by combining vacuum crushing and step-heating experiments and examining a full suite of noble gas isotope and elemental abundances. The most striking observation is the presence of crustal nucleogenic neon, released on graphitisation of the framesites. Neon of this composition can only have been produced in the crust and subsequently incorporated during formation of the framesites in the mantle. This may indicate that noble gases produced in the crust, such as nucleogenic neon and possibly xenon, and some radiogenic 4He and 40Ar and fissiogenic xenon, were introduced into the subcontinental mantle source during ancient subduction-related processes. If correct, then some parts of the mantle may contain significant quantities of crustal noble gases. D 2003 Elsevier B.V. All rights reserved. Keywords: Diamond; Noble gas; Mantle evolution; Mantle lithosphere
1. Introduction Research goals in noble gas geochemistry include understanding the structure of the Earth’s mantle and the creation of a coherent model for its evolution. In this regard, studies of noble gas compositions of midocean-ridge basalts (MORBs) and ocean island basalts (OIBs) such as found in Loihi Seamount, Hawaii, Iceland and Re´union (e.g., Alle`gre et al., 1983; Dixon et al., 2000; Honda et al., 1993; Kurz et al., 1983;
* Corresponding author. Fax: +61-2-6125-0738. E-mail address: [email protected] (M. Honda). 0009-2541/$ - see front matter D 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.chemgeo.2003.10.012
Moreira et al., 1998; Staudacher and Alle`gre, 1982; Trieloff et al., 2002), have provided a wealth of information. However, virtually all these data are from samples that are effectively of zero-age and only provide information on the present-day composition of mantle noble gases. It is critically important to determine if there is any systematic variation in mantle noble gas composition with time. This evidence is particularly relevant to current debates on mantle geodynamics. For example, noble gas data for mantle-derived samples of different ages could be used to evaluate the degree to which the upper mantle has interacted with the lower mantle over time and allow further refinement of models concerning two
348
M. Honda et al. / Chemical Geology 203 (2004) 347–358
layered convection vs. whole mantle convection (e.g., Alle`gre, 1997; Kellogg et al., 1999) and mass transport (including volatiles) in the mantle (e.g., Davies, 1999; Kamijo et al., 1998; Porcelli and Wasserburg, 1995). Previous attempts to determine the noble gas geochemistry of the ancient mantle by analysis of older geological samples have been largely unsuccessful, in part owing to the lack of suitable sample material. A possible exception is the determination of a high 3He/4He ratio in olivines separated from Archaean komatiites from Canada (Matsumoto et al., 2002; Richard et al., 1996). Diamonds have several unique characteristics that make them potentially very useful as robust containers of mantle noble gases: (1) diamonds are chemically inert and typically have experienced little interaction with the surrounding mantle, crust or atmosphere, despite their depth of origin and rapid transportation to shallow crustal levels in kimberlite and lamproite volcanic magmas, (2) many diamonds, particularly those of industrial quality, contain significant quantities of mantle-derived noble gases (e.g., Burgess et al., 1998; Ozima and Zashu, 1983), (3) diamonds are highly retentive of all noble gases, including helium (Honda et al., 1987), (4) most diamonds appear to be derived from 150- to 200-km depth in the Earth’s subcontinental mantle (e.g., Wyllie, 1995), although some may be derived from as deep as the 670-km discontinuity (e.g., Harte et al., 1998) and (5) diamonds cover a wide range of crystallization ages mostly between 1.0 and 3.2 billion years (e.g., Richardson et al., 1984, 1993). Thus, diamonds provide a direct window into the ancient mantle, and they are unique repositories of mantle-derived noble gases. In this paper, we report preliminary noble gas compositions for four polycrystalline framesite diamonds from the Jwaneng kimberlite, Botswana. Previous noble gas studies on polycrystalline diamonds (framesites) showed large quantities of 36Ar and radiogenic 4He and 40Ar (Burgess et al., 1998, Phillips, unpublished data). Thus, the expectation was that we could extract information on mantlederived noble gas compositions (helium, neon, argon, krypton and xenon) from single, small fragments of framesite. The present study represents part of a wider investigation into the evolution of the noble gases through time.
2. Jwaneng diamonds The Jwaneng kimberlite cluster is situated in southern Botswana, approximately 120 km west of Gaborone. Kinny et al. (1989) reported a U – Pb zircon intrusion age of 235 F 4 Ma for the Jwaneng DK2 body and a Rb –Sr age of 250 F 17 Ma, from the nearby DK7 kimberlite. Although arguably the richest diamond mine in the world, some 60% of the diamonds are classed as aggregates (Deines et al., 1997). Framesites are a polycrystalline form of diamond and exhibit variable, sub-millimetre grain sizes. They are usually gray to black, comprising randomly oriented fine crystallites that are sometimes intergrown with silicates and oxides of eclogitic and lherzolitic parageneses (Kirkley et al., 1994). Framesites are also characterized by an unusual bimodal carbon isotopic distribution, with average values of 1.5x d13C and 19x d13C, which has lead to suggestions of an origin from subduction of oceanic sediments associated with ophiolite sequences (Burgess et al., 1998; Kirkley et al., 1994). Therefore, eclogitic diamonds may provide the means for evaluating the recycling of noble gases into the mantle. Inclusions from gem-quality Jwaneng diamonds comprise an even mix of peridotitic and eclogitic compositional affinities (Deines et al., 1997). Using composites of garnet and clinopyroxene inclusions, Richardson et al. (1999) obtained a two-point isochron Re – Os genesis age of 1540 F 20 Ma for eclogitic diamonds from Jwaneng. More recent Re – Os results for eclogitic sulphide inclusions from Jwaneng diamonds indicate two age groups, one at f 1.5 Ga and another of f 2.9 Ga (Richardson et al., 2001b; Shirey et al., 2001). No age information is available for diamonds of peridotitic association.
3. Experimental procedures Noble gas analyses were conducted on four polycrystalline (framesite) diamonds with each aggregate measuring approximately 2 2 1 cm in size. The diamonds in each sample exhibit variable, sub-millimetre grain sizes. In an effort to remove any silicate contaminants, the diamond samples were initially coarsely crushed into millimetre-size fragments, leached with 7% HF, and ultrasonically cleaned in
M. Honda et al. / Chemical Geology 203 (2004) 347–358
distilled water and analytical grade acetone. After bake-out at 200 jC for at least 12 h, gases were extracted by step-heating or by vacuum crushing at room temperature. The procedures for noble gas extraction, purification and mass spectrometric analysis are essentially the same as those described by Honda et al. (1993) and Matsumoto et al. (1998). Typical blank levels both for step-heating and vacuum crushing experiments were 4He = 2 –6 10 11, 20Ne = 1– 4 10 12, 40Ar = 4 –9 10 9, 84Kr = 1 – 5 10 13, and 132Xe = 1– 4 10 13 cm3 STP. There were no differences in blank amounts, regardless of gas extraction methods (heating or vacuum crushing) or extraction temperatures. This probably indicates that blanks were dominated by background in the sample handling system. Within uncertainties, noble gas isotopic ratios in blank runs were atmospheric.
4. Results Table 1 lists the noble gas results obtained from the four Jwaneng framesites both by step-heating and vacuum crushing (the character ‘H’ or ‘C’, respectively, is added to the sample name to identify the method of gas extraction). The total noble gas concentrations of the Jwaneng framesites are very high compared with those in other mantle-derived materials. For example, total amounts of 22Ne in the Jwaneng framesites released by step-heating range from 1.0 10 10 to 1.1 10 9 ccSTP/g, whereas average 22 Ne concentrations in MORB, OIB glasses and ultramafic xenoliths are 7 10 11, 8 10 11 and 3 10 11 ccSTP/g, respectively (Ozima and Podosek, 2002). The majority of noble gases were released in the low temperature fractions of step-heating experiments or in the first fraction of the vacuum crushing experiments (e.g., sample Jwaneng 2C). In order to examine the characteristics of noble gas release from Jwaneng framesites in more detail, sample Jwaneng 4H was heated incrementally at 600, 1000, 1800 and 2000 jC (Fig. 1). In this case, more than half of the 22Ne, 36Ar, 84Kr and 130Xe were released at the lowest temperature fraction of 600 jC and much of the remainder was released at 1000 jC. Noble gas isotopic compositions at these lower temperatures (600 and 1000 jC) are close to atmospheric values (Table 1). In contrast, the highest temperature
349
steps (1800 and 2000 jC) yielded virtually all the radiogenic 4He and 40Ar*, fissiogenic 136Xe* and excess 129Xe relative to atmospheric. 4.1. Helium The total helium concentrations, released from the Jwaneng framesite samples by step-heating, range from 55 to 173 10 6 ccSTP/g, with the majority of helium being released on graphitisation of the diamonds (f 2000 jC; Fig. 1). Helium concentrations from vacuum crushing experiments are two orders of magnitude lower than the total helium concentrations released from the step-heating experiments. 3 He/4He ratios observed in the framesites, from both step-heating and vacuum crushing experiments, are less than the atmospheric value of 1.4 10 6 and range from 0.81 RA to 0.024 RA (where RA is the atmospheric 3He/4He ratio). Thus, these samples do not contain significant amounts of primordial mantlederived 3He as observed in some diamonds (Ozima and Zashu, 1983, 1988, 1991; Wada and Matsuda, 1998). The enrichment of radiogenic helium in the current samples is consistent with previous observations from Jwaneng framesites by Burgess et al. (1998). These authors reported uranium contents of 114 and 33 ppb for two Jwaneng framesites and suggested that uranium is likely to be in microinclusions within framesites. Using an emplacement age of 235 Ma for the Jwaneng kimberlite (Kinny et al., 1989) and a U content of 100 ppb (with Th/U = 3), the amount of in situ produced radiogenic 4He in the current framesite samples is calculated to be 5 10 6 ccSTP/g. This amount corresponds to less than 10% of the total 4He released from the samples during stepheating. On the other hand, if the formation age of the Jwaneng framesites is 2.9 Ga, as postulated from the Re – Os ages of eclogite inclusions (Shirey et al., 2001), then virtually all radiogenic 4He could have been produced by the decay of U and Th in the samples. This subject will be discussed further in later sections by deconvolving radiogenic, nucleogenic and fissiogenic components in the samples. 4.2. Neon Most neon was released in the lower temperature fractions of the step-heating experiments (Fig. 1) or
350
Table 1 Noble Gas Isotopic Compositions of Jwaneng framesite diamondsa Sampleb
Jwaneng 2C (0.525 g) Jwaneng 2H (0.805 g) Jwaneng 3C (0.765 g) Jwaneng 3H (0.507 g) Jwaneng 4C (0.735 g) Jwaneng 4H (0.464 g)
Air
3 He/4He ( 10 6)
22
Ne ( 10 12)
20
1700 2000 Total 1 2 Total 1800 2000 Total
1.85 (4) 53.1 (11) 54.9 (11) 2.5 (1) 0.87 (2) 3.4 (1) 7.0 (1) 166.1 (33) 173.1 (33) 0.36 (1)
0.33 (2) ? – 1.13 (4) 0.47 (5) 0.96 (3) 0.40 (1) 0.034 (1) 0.050 (1) 0.35 (4)
1167 (14) 1.8 (1) 1169 (14) 61.4 (11) 4.7 (8) 65.9 (14) 142 (7) b.b. 142 (7) 14.0 (3)
9.78 8.70 9.78 9.74 9.81 9.74 9.62 b.b. 9.62 9.75
(5) (3) (5) (9) (28) (9) (13)
1800 2000 Total
4.9 (1) 77.1 (15) 82.0 (16) 4.0 (1)
0.84 (4) 0.109 (4) 0.16 (1) 0.26 (1)
93.3 (12) 8.1 (13) 101.5 (19) 105.5 (18)
600 1000 1800 2000 Total
0.226 (3) 0.273 (3) 11.8 (1) 154.5 (2) 166.0 (2)
b.b. b.b. 0.15 (1) 0.011 (2) – 1.40
132 (2) 106 (2) 6.9 (4) 6.9 (5) 251 (3)
Ne/22Ne
21
Ne/22Ne
36
Ar ( 10 9)
40
Ar/36Ar
84
Kr ( 10 12)
130 Xe ( 10 12)
128
Xe/130Xe
864 (3) 24 325 (566) 2063 (49) 3218 (31) 22 926 (4460) 3458 (78) 1583 (12) 33 241 (1348) 5275 (224) 1184 (12)
217 (14) 3.8 (3) 221 (14) 99 (6) 2.4 (2) 100 (6) 179 (11) 11.1 (7) 190 (11) 11.3 (7)
4.1 (2) 0.15 (1) 4.3 (2) 1.5 (1) 0.07 (1) 1.6 (1) 2.6 (1) 0.63 (3) 3.3 (1) 0.24 (2)
0.472 0.499 0.473 0.482 0.453 0.480 0.477 0.493 0.480 0.497
(8) (16) (8) (9) (34) (8) (8) (8) (6) (22)
129
Xe/130Xe
6.55 8.33 6.61 6.93 8.14 6.98 6.77 8.26 7.06 6.86
(8) (21) (8) (8) (42) (8) (13) (9) (10) (28)
131
Xe/130Xe
5.26 5.56 5.27 5.12 6.15 5.17 5.27 5.28 5.27 5.41
(6) (13) (6) (5) (32) (5) (6) (5) (5) (24)
132
Xe/130Xe
6.68 7.42 6.71 6.64 7.40 6.67 6.67 7.07 6.75 6.66
(5) (14) (5) (4) (25) (4) (5) (4) (4) (26)
134
Xe/130Xe (3) (9) (3) (3) (19) (3) (3) (4) (2) (13)
Xe/130Xe
(3) (54) (3) (5) (33) (6) (8)
(13) (8)
0.0294 0.0839 0.0295 0.0284 0.0277 0.0284 0.0316 b.b. 0.0316 0.0284
(8) (5)
6.4 (1) 0.34 (1) 6.7 (1) 3.8 (1) 0.05 (1) 3.9 (1) 3.53 (7) 0.47 (2) 4.01 (8) 0.55 (1)
9.85 7.19 9.64 9.78
(6) (57) (8) (6)
0.0301 0.1058 0.0362 0.0309
(3) (118) (13) (3)
2.78 (5) 0.11 (1) 2.89 (5) 2.84 (5)
1710 (6) 53 070 (2871) 3638 (155) 1572 (6)
105 (7) 2.3 (2) 107 (7) 99 (6)
1.61 (9) b.b. 1.61 (9) 1.25 (7)
0.449 (14) b.b. 0.449 (14) 0.461 (15)
6.47 (16) b.b. 6.47 (16) 6.47 (14)
5.08 (14) b.b. 5.08 (14) 5.27 (11)
6.41 (15) b.b. 6.41 (15) 6.74 (12)
2.51 (7) b.b. 2.51 (7) 2.61 (5)
2.08 (7) b.b. 2.08 (7) 2.23 (5)
9.79 9.64 8.14 7.75 9.63 9.80
(6) (6) (22) (30) (4)
0.0282 (3) 0.0281 (3) 0.0865 (21) 0.105 (5) 0.0318 (3) 0.0290
3.82 (3) 3.07 (2) 0.155 (5) 0.31 (1) 7.35 (4)
301 (2) 364 (2) 24 800 (730) 47 200 (1830) 2830 (109) 295.5
110 (5) 86.4 (4) 7.1 (3) 3.8 (2) 207 (6)
2.3 (1) 0.95 (4) 0.19 (9) 0.14 (2) 3.6 (1)
0.470 (7) 0.467 (8) 0.485 (14) 0.22 (14) 0.460 (7) 0.472
6.50 (5) 6.48 (6) 6.67 (8) 7.01 (17) 6.52 (6) 6.496
5.21 (6) 5.24 (7) 5.10 (17) 5.73 (15) 5.23 (5) 5.213
6.60 (7) 6.56 (6) 6.98 (7) 9.89 (23) 6.74 (4) 6.607
2.57 (2) 2.52 (3) 2.93 (5) 4.95 (24) 2.67 (2) 2.563
2.19 (2) 2.18 (3) 2.58 (6) 4.29 (23) 2.29 (2) 2.176
b.b.: below procedural blank. ?: 3He peak was not found. a 38 36 Noble gas amounts are in ccSTP/g. Quoted errors are one standard deviation, shown in parentheses as the last digits. Ar/ Ar and Kr isotopic ratios in the samples are atmospheric within uncertainties. b H: heated for 30 min, C: crushed for 300 strokes. Jwaneng 2C was crushed twice successively.
2.59 3.37 2.62 2.74 3.30 2.77 2.65 3.38 2.80 2.73
136
2.20 2.95 2.22 2.39 2.92 2.41 2.26 3.08 2.42 2.12
(3) (8) (3) (3) (18) (3) (3) (4) (3) (11)
M. Honda et al. / Chemical Geology 203 (2004) 347–358
Jwaneng 1H (0.523 g)
4 He ( 10 6)
M. Honda et al. / Chemical Geology 203 (2004) 347–358
351
Fig. 1. Noble gas release by step-heating from Jwaneng framesite 4H. More than half of the total amounts of 22Ne, 36Ar and 130Xe were released from the lowest temperature fraction (600 jC). Gas release was virtually complete at 1000 jC. In contrast, radiogenic 4He and 40Ar* and fissiogenic 136Xe* were mostly released during graphitisation of diamond at 2000 jC.
during the first fraction of vacuum crushing (sample Jwaneng 2C). The isotopic compositions in these fractions are atmospheric within the uncertainties. In contrast, the neon isotopic compositions of higher temperature step-heating fractions are non-atmospheric and have close affinities with crustal nucleogenic neon. In Fig. 2, the neon isotopic results from the Jwa-
neng framesites are plotted in 20 Ne/22Ne – 21Ne/ 22 Ne space. Neon data from the high temperature fractions of samples Jwaneng 1H, 3H and 4H plot close to a mixing line between atmospheric and crustal nucleogenic neon (e.g., Kennedy et al., 1990). Crustal nucleogenic neon, which has been found in natural gases and brines, comprises a mixture of nucleogenic
Fig. 2. Neon isotope diagram, showing the results from four Jwaneng framesites. The MORB correlation line of Moreira et al. (1998) is also plotted. Neon data from the high temperature fractions of samples 1H, 3H and 4H lie close to a mixing line between atmospheric neon and crustal nucleogenic neon (e.g., Kennedy et al., 1990). The nucleogenic neon observed in the Jwaneng framesites is likely to have been produced in the crust and subsequently incorporated into the samples.
352
M. Honda et al. / Chemical Geology 203 (2004) 347–358
neon isotopes produced from nuclear reactions 18 O(a,n)21Ne, 19F(a, n)22Na(h+/2.6a) ! 22Ne and 19 F(a,p)22Ne in crustal material with a constant O/ F (atomic) ratio of about 100 (Kennedy et al., 1990). Crustal nucleogenic neon has also been found in Archean anorthosites in West Greenland (Azuma et al., 1993) and in the continental metamorphic rocks (gneiss and amphibolites) collected from the KTB deep bore hole in NE Bavaria, Germany (Drescher et al., 1998). It is highly unlikely that the crustal nucleogenic neon observed in the Jwaneng framesites is a product of in situ nuclear reactions, for the following reasons. First, it would be fortuitous that the framesites have the specific O/F abundance ratio of f 100 to produce crustal-like nucleogenic neon within the samples. Second, given a crustal composition (e.g., oxygen f 47 w.t%) and U content of 100 ppb, the amount of nucleogenic 21Ne produced over 235 Ma (the
emplacement age for the Jwaneng kimberlite) and 2.9 Ga (Re – Os model age) is calculated to be 0.2– 4 10 12 ccSTP/g (Yatsevich and Honda, 1997), which is comparable to the integrated contents of nucleogenic 21Ne extracted by step-heating experiments for the samples (0.4 – 1.0 10 12 ccSTP/g; Table 2). However, as the bulk oxygen abundance in the framesites must be several orders of magnitude lower than that used in the above calculation, the quantity of nucleogenic neon generated within the framesites must be negligible compared to the amounts actually observed. It could be argued that the crustal-like nucleogenic neon observed in the Jwaneng framesites originated from the surrounding lithospheric mantle, which had undergone earlier enrichment in lithophile elements, resulting in elevated 21Ne and 22Ne abundances. However, we believe that this scenario cannot account for the observations of fluorine produced-nucleogenic
Table 2 Radiogenic, nucleogenic and fissiogenic components in Jwaneng framesite diamonds (in ccSTP/g)a 4 He ( 10 6)
21 Ne* ( 10 12)
21
Ne*/4He ( 10 8)
40 Ar* ( 10 6)
40
1700C 2000C Total 1 2 Total 1800C 2000C Total
1.85 (4) 53.1 (11) 54.9 (11) 2.5 (1) 0.87 (2) 3.4 (1) 7.0 (1) 166.1 (33) 173.1 (33) 0.36 (1)
0.53 (35) 0.102 (14) 0.65 (35) – – – 0.44 (13) – 0.44 –
29 (19) 0.19 (3) 1.2 (6) – – – 6.3 (19) – 0.26 –
3.62 (7) 8.23 (31) 11.8 (4) 11.3 (3) 1.1 (3) 12.4 (4) 4.5 (1) 15.4 (9) 20.0(10) 0.48 (1)
1800C 2000C Total
4.9 (1) 77.1 (15) 82.0 (16) 4.0 (1)
0.088 (31) 0.69 (17) 0.77 (14) 0.21 (4)
1.8 (6) 0.89 (22) 0.94 (17) 5.2 (11)
600C 1000C 1800C 2000C Total
0.226 (3) 0.273 (3) 11.8 (1) 154.5 (2) 166.0 (2)
– – 0.43 (4) 0.57 (6) 1.0
– – 3.6 (3) 0.37 (4) 0.6
Sample Jwaneng 1H (0.523 g) Jwaneng 2C (0.525 g) Jwaneng 2H (0.805 g) Jwaneng 3C (0.765 g) Jwaneng 3H (0.507 g) Jwaneng 4C (0.735 g) Jwaneng 4H (0.464 g)
Productionb (1) Instantaneous (2) Integrated for 4.5 Ga
4.5c 4.5c
Ar*/4He
136 Xe* ( 10 15)
136
1.96 (5) 0.16 (1) 0.22 (1) 4.5 (1) 1.2 (3) 3.7 (1) 0.65 (2) 0.09 (1) 0.12(1) 1.35 (4)
99 (125) 115 (14) 188 (13) 327 (50) 52 (15) 374 (53) 218 (81) 566 (39) 795 (95) –
53 (68) 2.2 (3) 3.4 (2) 129 (20) 61 (18) 110 (16) 31 (12) 3.4 (2) 4.6 (6) –
3.93 (7) 5.7 (5) 9.7 (5) 3.62 (7)
0.80 0.07 0.12 0.91
– – – 68 (59)
– – – 17 (15)
0.020 (8) 0.21 (1) 3.8 (2) 14.6 (8) 18.6 (8)
0.09 (4) 0.77 (3) 0.32 (1) 0.095 (5) 0.112 (5)
27 (57) 4 (29) 78 (11) 290 (47) 399 (81)
121 (254) 15 (105) 6.6 (10) 1.9 (3) 2.4 (5)
0.23 0.59
(2) (1) (1) (1)
Xe*/4He ( 10 9)
3.1 2.6
a The radiogenic, nucleogenic and fissiogenic amounts in the samples are calculated by subtracting atmospheric components. Quoted errors are one standard deviation, shown in parentheses as the last digits. b Production ratios are calculated by assuming a Th/U ratio of 3 and a K/U ratio of 12 000. c From Yatsevich and Honda (1997).
M. Honda et al. / Chemical Geology 203 (2004) 347–358 22
Ne in the samples. Average fluorine concentrations in the mantle are estimated to be 25 times lower than in the crust (Mason and Moore, 1982; McDonough and Sun, 1995), whereas oxygen abundances are approximately the same in the mantle and crust. Thus, in order to produce the crustal-like nucleogenic 21 Ne*/22Ne* ratio observed in the samples, the surrounding lithospheric mantle must have been enriched in fluorine concentration by 25 times the average mantle abundance. The consistency in noble gas and halogen (Cl, Br and I) abundances between African (including Jwaneng) diamonds and MORB samples (Johnson et al., 2000) suggests a MORB-like mantle source for lithospheric volatiles, and there is no evidence of an enrichment in halogen concentrations. Therefore, we suggest that the nucleogenic neon observed in the Jwaneng framesites is likely to have been produced in the crust and subsequently incorporated into the framesites during their crystallisation in the subcontinental mantle. Mantle neon is thought to have an end-member 20 Ne/22Ne ratio of 13.7 (Honda et al., 1991; Moreira et al., 1998) or 12.5 (Trieloff et al., 2002, 2000), and is the only known neon component with a 20Ne/22Ne ratio that is higher than the atmospheric value. However, all 20Ne/22Ne ratios observed in the Jwaneng framesites were close to or lower than the atmospheric value of 9.8 (Table 1), indicating that the Jwaneng framesites contain no detectable mantle neon. In contrast, some cubic diamonds contain elevated 20 Ne/22Ne ratios attributed to the presence of a mantle neon component (Ozima and Zashu, 1988, 1991; Wada and Matsuda, 1998). Thus, it seems likely that any mantle neon in the Jwaneng framesites was overwhelmed by the addition of crustal nucleogenic neon (higher temperature fractions) or by atmospheric neon (lower temperature fractions). This observation is consistent with the paucity of primordial 3He in the Jwaneng framesites. 4.3. Argon 38
Ar/36Ar ratios of the Jwaneng framesites are atmospheric within the uncertainties. Integrated 40 Ar/36Ar ratios vary between 1572 and 5275, significantly higher than the atmospheric value of 295.5. The 2000 jC graphitisation steps produced the highest 40 Ar/36Ar ratios, with values up to 53 000 (sample
353
Jwaneng 3H). These high 40Ar/36Ar ratios could result from the addition of in situ radiogenic argon and/or the presence of excess 40Ar. As with neon, the majority of 36Ar was released in the first step-heating and vacuum crushing experiments, resulting in relatively low 40Ar/36Ar ratios. This indicates that significant amounts of atmospheric argon was present within low-retention sites in the diamond samples. 4.4. Krypton The krypton isotopic ratios measured in the Jwaneng framesites are atmospheric at the one sigma level, and are not listed in Table 1. 4.5. Xenon As with neon and argon, most xenon was released in the lower temperature step-heating or first vacuum crushing fractions and is characterised by atmospheric isotopic compositions (Table 1). The xenon isotopic compositions of the higher temperature stepheating (samples Jwaneng 1H, 2H and 4H) and the second vacuum crushing (sample Jwaneng 2C) fractions, exhibit xenon isotopic anomalies relative to atmospheric compositions. On a 129Xe/130Xe vs. 136 Xe/130Xe diagram, the data for Jwaneng samples 1H, 2H and 2C plot adjacent to the MORB correlation line (slope = 3.13 F 0.08, Kunz et al., 1998, Fig. 3). The 129Xe isotope anomalies observed in samples Jwaneng 1H, 2H and 2C are comparable to those in mantle-derived samples, including MORBs (Kunz et al., 1998; Marty, 1989; Moreira et al., 1998; Staudacher and Alle`gre, 1982), Loihi dunites and Icelandic picritic glasses (Trieloff et al., 2000). Excess 129Xe, relative to the atmospheric value, observed in mantle-derived samples is generally attributed to radioactive decay of an extinct nuclide 129I (half life of 16 Ma), once present in the Earth (e.g., Ozima and Podosek, 2002). However, a similar excess in 129Xe, accompanied by excess 131 – 136Xe (as observed in the Jwaneng framesites), has recently been reported for continental metamorphic rocks (gneiss and amphibolites) collected from the KTB deep bore hole (Drescher et al., 1998), and in Archean cherts in the metasedimentary sequences of Pilbara Craton, Western Australia (Pinti et al., 2001). Drescher et al. (1998) and Pinti et al. (2001) postulated that the excess
354
Fig. 3. Plot of 129Xe/130Xe versus Kunz et al. (1998) is also plotted. 129
M. Honda et al. / Chemical Geology 203 (2004) 347–358
136
Xe/130Xe observed in framesite diamonds from the Jwaneng kimberlite. The MORB correlation line of
Xe in these rocks was in fact due to nucleogenic production from tellurium 128 Te(n,g) 129 Te(h/ 69m) ! 129I(h/69Ma) ! 129Xe in the crust. The observation of crustal-like noble gases (e.g., radiogenic helium and argon, nucleogenic neon) and the apparent absence of mantle noble gases (e.g., primordial 3He and solar-like neon) in the Jwaneng framesites are somewhat analogous to the compositions reported for the KTB metamorphic rocks. Therefore, one interpretation is that the 129Xe excesses observed in the Jwaneng framesites represent crustal nucleogenic xenon, rather than mantle xenon. It is also important to point out that the 131 Xe excesses in the Jwaneng framesites are systematically higher than those observed in MORBs (Kunz et al., 1998), possibly also reflecting the nucleogenic production of 131Xe from tellurium 130 Te(n,g)131Te(h/25m) ! 131I(h/8d) ! 131Xe. In addition, the (129Xe/131Xe)excess ratios in the Jwaneng framesites are consistent with the range of nucleogenic 129Xe/131Xe ratios produced from tellurium by neutron capture (0.7 – 4.0; Browne and Berman, 1973), thus supporting a nucleogenic origin for the excess 129Xe. It must be stressed, however, that it would require reasonably high neutron fluxes and tellurium concentrations in the crust to provide the level of 129Xe and 131Xe excesses observed in the Jwaneng framesites.
Thus, although a crustal nucleogenic origin for excess 129Xe in the Jwaneng framesites requires verification, particularly in relation to the absolute production of nucleogenic xenon in the crust (Drescher et al., 1998; Pinti et al., 2001), we tentatively suggest that the Jwaneng framesites do not contain detectable amounts of mantle xenon. This argument is consistent with the observation of crustal neon in the framesites. As discussed earlier, it is likely that some fraction of the radiogenic 4He in the samples results from in situ production—therefore, a corresponding amount of in situ fissiogenic 136Xe also exists. Consequently, the 136Xe in the Jwaneng framesites can be interpreted as a mixture of crustal (dominated by fissiogenic xenon produced from 238 U), in situ produced fissiogenic xenon and atmospheric components.
5. Discussion As shown in the previous section, the noble gas compositions of the Jwaneng framesites have unusual characteristics: (1) large quantities of atmospheric noble gases associated with the first step-heating and vacuum crushing fractions; (2) no detectable MORBlike (mantle) neon and xenon; and (3) large quantities of radiogenic 4He and 40Ar, nucleogenic neon and
M. Honda et al. / Chemical Geology 203 (2004) 347–358
possibly xenon and fissiogenic xenon, released in the higher temperature step-heating or second vacuum crushing experiments. In this section, we further examine these unusual noble gas compositions. 5.1. Atmospheric noble gases in Jwaneng framesites The step-heating experiments demonstrated that large quantities of atmospheric noble gases appear to be loosely trapped within the framesites (Fig. 1). In contrast, radiogenic 4He and 40Ar, nucleogenic neon and xenon and fissiogenic xenon were released from the framesites during graphitisation at f 2000 jC. Thus, it is unlikely that the atmospheric noble gas components reflect an intrinsic noble gas component associated with framesite crystallisation in the subcontinental mantle. We suggest that the atmospheric noble gases in the framesites are likely to be secondary contaminants. As the framesites contain traces of mineral intergrowths, a possible explanation is that the atmospheric noble gas components were released from altered mineral phases that were not removed by acid treatment. Small amounts of atmospheric noble gases were, however, also noted at 2000 jC (Fig. 2), possibly due to incomplete outgassing of altered mineral phases during the lower temperature step-heating experiments. This possibility will be addressed in future studies by crushing the framesites into smaller fragments, more rigorous acid cleaning and SEM examination of framesite surfaces. 5.2. Radiogenic, nucleogenic and fissiogenic noble gases in Jwaneng framesites The combined results for radiogenic, nucleogenic and fissiogenic noble gas components observed in the Jwaneng framesites are presented in Table 2. In this tabulation, we assume that the total amounts of 4 He released from the samples are radiogenic. In order to calculate the amounts of nucleogenic 21Ne (21Ne*), radiogenic 40Ar (40Ar*) and fissiogenic 136 Xe (136Xe*), corrections for atmospheric noble gas compositions are made according to the following equation: X * ¼ Xref fðX =Xref Þobserved ðX =Xref Þatmospheric g ð1Þ
355
where X = 21Ne, 40Ar or 136Xe and Xref = 20Ne, 36Ar or 130Xe, respectively. Abundance ratios of 21Ne*, 40 Ar* and 136Xe*, relative to 4He, as well as the theoretical 21Ne*/4He (Yatsevich and Honda, 1997) and 40Ar*/4He ratios, calculated for both instantaneous and integrated production for 4.5 Ga in the crust, are also listed in Table 2. A precise value of [Esf 136Ysf] ( = 6.8 10 18/a) has been determined from the study of U-bearing accessory minerals (Ragettli et al., 1994), where Esf and 136Ysf are the decay constant for spontaneous fission of 238U and the fission yield of 136Xe, respectively. These values were used to calculate the 136Xe* production in the current samples, assuming a Th/U ratio of 3 and a K/U ratio of 12,000. It is noted that the production ratios of 40Ar*/4He and 21Ne*/4He are similar for both the crust and the mantle, because the relative abundances of the parent elements (U, Th and K, and oxygen—the main target element to produce nucleogenic 21Ne) are believed to be analogous in the two regions (e.g., Mason and Moore, 1982; Sun and McDonough, 1989). The 21Ne*/4He ratios for Jwaneng samples 4C (crushing) and 4H (1800 jC step) are 5.2 10 8 and 3.6 10 8, respectively. These abundance ratios are consistent with the theoretically calculated 21 Ne*/4He production ratio ( = 4.7 10 8) for the crust. Noble gases, released from the vacuum crushing and 1800 jC experiments, probably represent the noble gases trapped in fluid inclusions and defects within the framesites, and are dominated by crustal noble gas compositions. Thus, the consistency between 21Ne*/4He ratios obtained by vacuum crushing (sample 4C) and step-heating (1800 jC step; sample 4H), and the theoretical value suggests that the degree of elemental fractionation between 4He and 21Ne* is insignificant during transportation of crustal noble gases into the mantle and encapsulation in the Jwaneng framesites. On the other hand, the integrated 21Ne*/4He ratio ( = 0.6 10 8) for sample 4H (step-heating) is systematically lower than the theoretical crustal value. Integrated 21Ne*/4He ratios for Jwaneng diamonds 1H, 2H and 3H are also lower than the theoretical value (Table 2). As previously discussed in Section 4.2, it is unlikely that appreciable amounts of nucleogenic neon can be produced in situ even if uranium and thorium are unusually abundant, and we have
356
M. Honda et al. / Chemical Geology 203 (2004) 347–358
suggested that nucleogenic neon was produced in the continental crust and subsequently incorporated into the framesites. As radiogenic 4He and nucleogenic 21 Ne do not appear to have been fractionated significantly when incorporated into the Jwaneng framesites, the difference in integrated 21Ne*/4He ratios between the step-heating experiments and the theoretical value can be used to estimate the fraction of in situ produced radiogenic 4He in the samples. On this basis, 74%, 94%, 80% and 88% of the total radiogenic 4 He observed in samples 1H, 2H, 3H and 4H, respectively, are estimated to be in situ products. The remaining radiogenic 4He in the samples can be regarded as inherited, probably crustal. This implies that the formation age of the Jwaneng framesites is older than the emplacement age of the Jwaneng kimberlite (see Section 4.1). As the integrated 136Xe*/4He ratios of the Jwaneng framesites obtained from step-heating experiments are consistent with the theoretical production ratio (Table 2), the fractions of in situ produced fissiogenic 136Xe from 238U are the same as those estimated for radiogenic 4He. The remaining 136Xe* in the samples can be interpreted as fissiogenic 136Xe produced in the crust. The integrated 40Ar*/4He ratios for the Jwaneng framesites are lower than 40 Ar*/4He* production ratios calculated for the crust and mantle (K/U = 12,000, Th/U = 3; Table 2). If the 40 Ar* in the Jwaneng framesites is dominated by in situ produced radiogenic 40Ar (i.e. analogous to radiogenic 4He), this would imply that K/U ratios in the Jwaneng framesites are lower than the average K/U abundance ratio in the crust and mantle.
sites suggests that these volatiles were transported into the diamond-forming region of the subcontinental mantle, possibly by subduction-related processes. This may imply that the framesites formed in a different environment to the majority of macro-crystalline diamonds and that framesites represent a distinct diamond population. Comparative studies of noble gases in macro-crystalline diamonds from the same kimberlite pipe are thus important for exploring this aspect further. The unusual carbon isotopic and mineral intergrowth compositions of Jwaneng framesites led Kirkley et al. (1994) to conclude that these diamonds formed under carbon supersaturation conditions during subduction of oceanic sediments and intercalated slab components. The incorporation of subducted crustal material into the mantle is an attractive explanation for the presence of crustal noble gases in diamonds. Although the parageneses of the framesites in this study were not constrained, they are likely to be eclogitic, as eclogitic diamonds are generally believed to have been formed from recycled carbon (e.g., Kirkley et al., 1991). Recent Re – Os analyses of eclogite xenoliths and eclogitic inclusions in diamonds from several southern African kimberlites suggest that subduction-related crustal recycling may have been a viable process during continent formation in the Archean (f 2.9 Ga) and resulted in widespread formation of eclogitic diamonds at that time (Richardson et al., 2001a; Shirey et al., 2001). If this is true, then crustal noble gases could have been introduced into the mantle at a very early stage in Earth history.
5.3. Crustal noble gases in the Jwaneng framesites 6. Conclusions The Jwaneng framesites likely formed in the subcontinental lithospheric mantle and acquired their volatiles, including noble gases, during crystallisation. As diffusion of noble gases in diamond is expected to be very slow (e.g., Honda et al., 1987; Ozima, 1989), late incorporation of noble gases during crustal residence of diamonds is highly unlikely. We are not aware of any experimental or other evidence that questions this premise. Hence, the observation of crustal noble gases (nucleogenic neon and possibly xenon, some of radiogenic 4He and 40 Ar and fissiogenic xenon) in the Jwaneng frame-
We have undertaken a full suite of noble gas analyses of four framesite diamonds using vacuum crushing and step-heating experimental techniques. We found that the Jwaneng framesites have unique characteristics in terms of noble gas elemental and isotope abundances: (1) large quantities of atmospheric noble gases, (2) no detectable MORB-like (mantle) neon, and (3) large quantities of radiogenic 4 He and 40Ar, nucleogenic neon and fissiogenic xenon and excess 129Xe, released in the higher temperature step-heating or second vacuum crushing
M. Honda et al. / Chemical Geology 203 (2004) 347–358
experiments; these compositions are analogous to those found in some continental metamorphic rocks (Drescher et al., 1998) and in Archean cherts (Pinti et al., 2001). Large quantities of atmospheric noble gases were released from lower temperature and initial vacuum crushing experiments. We interpret the atmospheric noble gases in the Jwaneng framesites as secondary, possibly derived from silicate and other mineral phases intergrown with the framesites. Although further authentication is necessary, we tentatively conclude that excess 129Xe in the Jwaneng framesites resulted from the addition of crustal nucleogenic xenon produced from neutron capture of tellurium in the crust (Drescher et al., 1998; Pinti et al., 2001) and that, as is the case for neon, there is no detectable mantle xenon in the samples. The observations of 21Ne/22Ne ratios consistent with the crustal nucleogenic production ratio strongly suggest that this component was produced in the crust and then incorporated into the Jwaneng framesites during their formation in the mantle. Crustal noble gases (nucleogenic neon and possibly xenon, some radiogenic 4He and 40Ar and fissiogenic xenon) could have been introduced into the diamond stability field of the subcontinental mantle by subduction-related processes. It is critically important to test this hypothesis through further investigations of eclogitic diamonds from various localities. Such studies will clarify whether crustal noble gas inventories can be retained during subduction of crustal material into the mantle lithosphere. The present work demonstrates the versatility of undertaking a full suite of noble gas isotope and elemental studies, using both vacuum crushing and step-heating methods. This approach allows deconvolution of the complex, multiple noble gas components (e.g., crustal, atmospheric and in situ) found in framesite diamonds.
Acknowledgements De Beers Consolidated Mines are thanked for donating the diamond specimens for this project and for allowing publication of this manuscript. We thank Mario Trieloff and an anonymous reviewer for their constructive comments on the manuscript. [PD]
357
References Alle`gre, C.J., 1997. Limitation on the mass exchange between the upper and lower mantle: the evolving convection regime of the Earth. Earth Planet. Sci. Lett. 150, 1 – 6. Alle`gre, C.J., Staudacher, T., Sarda, P., Kurz, M., 1983. Constraints on evolution of Earth’s mantle from rare gas systematics. Nature 303, 762 – 766. Azuma, S., Ozima, M., Hiyagon, H., 1993. Anomalous neon and xenon in an Archaean anorthosite from West Greenland. Earth Planet. Sci. Lett. 114, 341 – 352. Browne, J.C., Berman, B.L., 1973. Neutron-capture cross sections for 128Te and 130Te and the xenon anomaly in old tellurium ores. Phys. Rev. C 8, 2405 – 2411. Burgess, R., Johnson, L.H., Mattey, D.P., Harris, J.W., Turner, G., 1998. He, Ar and C isotopes in coated and polycrystalline diamonds. Chem. Geol. 146, 205 – 217. Davies, G.F., 1999. Geophysically constrained mass flows and the 40 Ar budget: a degassed lower mantle? Earth Planet. Sci. Lett. 166, 149 – 162. Deines, P., Harris, J.W., Gurney, J.J., 1997. Carbon isotope ratios, nitrogen content and aggregation state, and inclusion chemistry of diamonds from Jwaneng, Botswana. Geochim. Cosmochim. Acta 61, 3993 – 4005. Dixon, E.T., Honda, M., McDougall, I., Campbell, I.H., Sigurdsson, I., 2000. Preservation of near-solar neon isotopic ratios in Icelandic basalts. Earth Planet. Sci. Lett. 180, 309 – 324. Drescher, J., Kirsten, T., Scha¨fer, K., 1998. The rare gas inventory of the continental crust, recovered by the KTB Continental Deep Drilling Project. Earth Planet. Sci. Lett. 154, 247 – 263. Harte, B., Harris, J.W., Hutchison, M.T., Watt, G.R., Wilding, M.C., 1998. Mineral facies and source materials for lower mantle inclusions in diamonds from Sao Luiz, Brazil. Min. Mag. 62A, 575 – 576. Honda, M., Reynolds, J.H., Roedder, E., Epstein, S., 1987. Noble gases in diamonds: occurrences of solarlike helium and neon. J. Geophys. Res. 92, 12507 – 12521. Honda, M., McDougall, I., Patterson, D.B., Doulgeris, A., Clague, D.A., 1991. Possible solar noble gas component in Hawaiian basalts. Nature 349, 149 – 151. Honda, M., McDougall, I., Patterson, D.B., Doulgeris, A., Clague, D.A., 1993. Noble gases in submarine pillow basalt glasses from Loihi and Kilauea, Hawaii: a solar component in the Earth. Geochim. Cosmochim. Acta 57, 859 – 874. Johnson, L.H., Burgess, R., Turner, G., Milledge, H.J., Harris, J.W., 2000. Noble gas and halogen geochemistry of mantle fluids: comparison of African and Canadian diamonds. Geochim. Cosmochim. Acta 64, 717 – 732. Kamijo, K., Hashizume, K., Matsuda, J.-I., 1998. Noble gas constraints on the evolution of the atmosphere – mantle system. Geochim. Cosmochim. Acta 62, 2311 – 2321. Kellogg, L.H., Hager, B.H., Hilst, R.D.v.D., 1999. Compositional stratification in the deep mantle. Science 283, 1881 – 1884. Kennedy, B.M., Hiyagon, H., Reynolds, J.H., 1990. Crustal neon: a striking uniformity. Earth Planet. Sci. Lett. 98, 277 – 286. Kinny, P.D., Compston, W., Bristow, J.W., Williams, I.S., 1989.
358
M. Honda et al. / Chemical Geology 203 (2004) 347–358
Archean mantle xenocrysts in a Premian kimberlite: two generations of kimberlitic zircon in Jwaneng DK2, southern Botswana. In: Ross, N.Kimberlites and Related Rocks. Geol. Soc. Australia Special Pub., Sydney, Australia, pp. 833 – 842. Kirkley, M.B., Gurney, J.J., Otter, M.L., Hill, S.J., Daniels, L.R., 1991. The application of C isotope measurement to the identification of the sources of C in diamonds: a review. Appl. Geochem. 6, 447 – 494. Kirkley, M.B., Gurney, J.J., Rickard, R.S., 1994. Jwaneng framesites: carbon isotopes and intergrowth compositions. In: Meyer, H.O.A., Leonardos, O.H. (Eds.), Diamonds: characterisation, Genesis and Exploration. CPRM Spec. Publ. Companhia de Presquisa de Recursos Minerais, Brasilia, Brazil, pp. 127 – 135. Kunz, J., Staudacher, T., Alle`gre, C.J., 1998. Plutonium-fission xenon found in Earth’s mantle. Science 280, 877 – 880. Kurz, M.D., Jenkins, W.J., Hart, S.R., Clague, D., 1983. Helium isotopic variations in volcanic rocks from Loihi Seamount and the Island of Hawaii. Earth Planet. Sci. Lett. 66, 388 – 406. Marty, B., 1989. Neon and xenon isotopes in MORB: implications for the earth – atmosphere evolution. Earth Planet. Sci. Lett. 94, 45 – 56. Mason, B., Moore, C.B., 1982. Principles of Geochemistry. Wiley, New York. 344 pp. Matsumoto, T., Honda, M., McDougall, I., O’Reilly, S.Y., 1998. Noble gases in anhydrous lherzolites from the Newer Volcanics, south Australia: a MORB-like reservoir in the subcontinental mantle. Geochim. Cosmochim. Acta 62, 2521 – 2533. Matsumoto, T., Seta, A., Matsuda, J., Chen, Y., Arai, S., 2002. Helium in the Archaean komatiites revisited: significantly high 3 He/4He ratios revealed by fractional crushing gas extraction. Earth Planet. Sci. Lett. 196, 213 – 225. McDonough, W.F., Sun, S.-S., 1995. The composition of the Earth. Chem. Geol. 120, 223 – 253. Moreira, M., Kunz, J., Alle`gre, C., 1998. Rare gas systematics on popping rock: isotopic and elemental compositions in the upper mantle. Science 279, 1178 – 1181. Ozima, M., 1989. Gases in diamonds. Annu. Rev. Earth Planet. Sci. 17, 361 – 384. Ozima, M., Podosek, F.A., 2002. Noble Gas Geochemistry, 2nd ed. Cambridge Univ. Press, Cambridge, UK, 286 pp. Ozima, M., Zashu, S., 1983. Primitive helium in diamonds. Science 219, 1067 – 1068. Ozima, M., Zashu, S., 1988. Solar-type Ne in Zaire cubic diamonds. Geochim. Cosmochim. Acta 52, 19 – 25. Ozima, M., Zashu, S., 1991. Noble gas state of the ancient mantle as deduced from noble gases in coated diamonds. Earth Planet. Sci. Lett. 105, 13 – 27. Pinti, D., Matsuda, J., Maruyama, S., 2001. Anomalous xenon in Archean cherts from Pilbara Craton, Western Australia. Chem. Geol. 175, 387 – 395. Porcelli, D., Wasserburg, G.J., 1995. Mass transfer of He, Ne, Ar,
and Xe through a steady state upper mantle. Geochim. Cosmochim. Acta 59, 4921 – 4937. Ragettli, R.A., Hebeda, E.H., Signer, P., Wieler, R., 1994. Uranium – xenon chronology: precise determination of Esf*136Ysf for spontaneous fission of 238U. Earth Planet. Sci. Lett. 128, 653 – 670. Richard, D., Marty, B., Chaussidon, M., Arndt, N., 1996. Helium isotopic evidence for a lower mantle component in depleted Archean komatiite. Science 273, 93 – 95. Richardson, S., Gurney, J.J., Erlank, A.J., Harris, J.W., 1984. Origin of diamonds in old enriched mantle. Nature 310, 198 – 202. Richardson, S.H., Harris, J.W., Gurney, J.J., 1993. Three generations of diamonds from old continental mantle. Nature 366, 256 – 258. Richardson, S.H., Chinn, I.L., Harris, J.W., 1999. Age and origin of eclogitic diamonds from the Jwaneng kimberlite, Botswana. In: Gurney, J.J., Gurney, J.L., Pascoe, M.D., Richardson, S.H.The P.H. Nixon Volume. Red Roof Design, Cape Town, pp. 709 – 713. Richardson, S.H., Shirey, S.B., Harris, J.W., Carlson, R.W., 2001a. Archean subduction recorded by Re – Os isotopes in eclogitic sulfide inclusions in Kimberley diamonds. Earth Planet. Sci. Lett. 191, 257 – 266. Richardson, S.H., Shirey, S.B., Harris, J.W., Carlson, R.W., 2001b. The Life and Times of Multiple Generations of Diamonds From the Kaapvaal Craton Keel Slave-Kaapvaal Workshop. Merrickville, Canada. Shirey, S.B., et al., 2001. Archean emplacement of eclogitic components into the lithospheric mantle during formation of the Kaapvaal Craton. Geophys. Res. Lett. 28, 2509 – 2512. Staudacher, T., Alle`gre, C.J., 1982. Terrestrial xenology. Earth Planet. Sci. Lett. 60, 389 – 406. Sun, S.-S., McDonough, W.F., 1989. Chemical and isotopic systematics of oceanic basalts: implications for mantle composition and processes. In: Saunders, A.D., Norry, M.J.Magmatism in Ocean Basins. Geological Society Special Publication No. 42, Boulder, USA, pp. 313 – 345. Trieloff, M., Kunz, J., Clague, D.A., Harrison, D., Alle`gre, C.J., 2000. The nature of pristine noble gases in mantle plumes. Science 288, 1036 – 1038. Trieloff, M., Kunz, J., Alle`gre, C.J., 2002. Noble gas systematics of the Re´union mantle plume source and the origin of primordial noble gases in Earth’s mantle. Earth Planet. Sci. Lett. 200, 297 – 313. Wada, N., Matsuda, J., 1998. A noble gas study of cubic diamonds from Zaire: constraints on their mantle source. Geochim. Cosmochim. Acta 62, 2335 – 2346. Wyllie, P.J., 1995. Experimental petrology of upper mantle materials, processes and products. J. Geodyn. 20, 429 – 468. Yatsevich, I., Honda, M., 1997. Production of nucleogenic neon in the Earth from natural radioactive decay. J. Geophys. Res. 102, 10291 – 10298.