The chronology of planetesimal processing

The cosmochronology of Rb-Sr and the initial 87Sr/86Sr ratios

As briefly discussed in Section 10.3, the most common isochron dating methods use measurements of the parameters of Eqn (10.7a) for several different samples that originated (according to a priori data) simultaneously and from the same material. If the samples have remained closed systems, Eqn (10.7a) is valid for each sample,

Table 12.3 Average compositions of iron-meteorite chemical groups. From Mittlefehldt et al. (1998). Reproduced by permission of the Mineralogical Society of America

Table 12.3 Average compositions of iron-meteorite chemical groups. From Mittlefehldt et al. (1998). Reproduced by permission of the Mineralogical Society of America

Group

Ni

S

P

Co

Ga

Ge

Cr

Cu

Mo

Ir

W

Pd

Au

As

Sb

wt

%

ppm

ppb

IAB

16

n.d.a

0.21

0.46

96

400

16

130

8.2

2.7

1.6

3.5

1.5

11

270

IC

7.1

n.d.

0.43

0.46

52

230

70

160

7.7

0.38

1.3

3.5

1.1

11

98

IIAB

6.1

0.2

0.6

0.53

58

170

38

130

6.9

10

2.1

2.6

1.1

9.9

200

IIC

11

n.d.

0.53

0.65

37

95

87

260

8.4

6.4

n.d.

6

1.1

8.2

150

IID

11

n.d.

0.98

0.47

76

87

31

280

9.4

9.9

2.4

5.3

1.1

10

220

IIIAB

8.5

1.4

0.56

0.5

20

39

40

160

7.2

4.1

1

3.5

1.2

10

265

IIIF

8

n.d.

0.22

0.36

6.8

0.91

210

170

7.2

3.2

1.2

4.4

0.91

11

86

IVA

8

0.8

0.09

0.4

2.2

0.12

140

150

5.9

1.9

0.6

4.6

1.5

7.6

9

IVB

17

0.03

0.1

0.74

0.22

0.05

88

12

27

22

3

12

0.15

1.1

a Not determined.

0.01

12 Highly processed meteorites Au, ppm

IIIAB iron meteorites

Au, ppm

Fig. 12.5 (a), (b) Fractionation of Au, As and Ir in magmatic IIIAB irons. The clear trends reflect the partitioning of the elements between solid and liquid metal in the course of solidification of the metallic cores of meteorite parent bodies. From Wasson (1999), © Elsevier Science 1999, reproduced by permission.

and therefore a set of these equations is available. The factors (r D/s D)INI (= a) and [exp(AT) — 1] (= b) are expected to be constants for this given set. Therefore the evolution of (rD/sD)t (= y) as a function of (rR/sD)t (= x) is expressed by the straight line y = a + bx (an isochron) on a y versus x plot (see the inset in Fig. 12.7), yielding both the age and the initial ratio of the suite of samples.

The method further provides a test for the integrity of the systematics in a sample suite. If the measured (r D/s D)t and (r R/s D)t ratios in several different samples constitute a straight line then it can be concluded that (1) the samples originated simultaneously and the regression slope b gives their age; (2) the parent material

10"

■1x10-5=Solar ratio of 109Ag/Ni

'l0'1x solar ratio

A IIIAB

A PAL

Fig. 12.6 Silver depletion in iron, pallasite and mesosiderite meteorites: Ag shows a strong depletion in a number of iron meteorites, down to 10-5 relative to the solar abundance. After Chen and Wasserburg (1996), © 1996 American Geophysical Union, modified by permission.

was homogeneous with the (initial) ratio (rD/sD)INI fixed by the intercept a; and (3) the samples have not experienced loss or gain of the species relevant to Eqn (10.7a).

Figure 12.7 illustrates the results of the Rb-Sr isochron dating of a large number of chondritic and achondritic meteorites: radioactive 87Rb j-decays to 87Sr, and 86Sr is the reference stable isotope. Rubidium is a highly volatile element whereas Sr is refractory (Table 3.1). The proximity of all the data points to a straight line indicates that all these meteorites were formed from the same well-mixed initial material. The age fits (to within the uncertainty limits) with the Pb/Pb ages. The far lower precision of the Rb-Sr isochron compared to the Pb/Pb isochron is a result of the much longer half-life of 87Rb, and, further, it must be noted that there is still a ~ 1% uncertainty associated with the 87Rb half-life.

Non-chondritic meteorites show extreme Rb depletion, and the range of 87Rb/86Sr ratios in eucrites (Fig. 12.8) corresponds to the extreme left of Fig. 12.7. The age derived from the isochron plot in Fig. 12.8 agrees with other results, indicating that the rocks have remained closed systems for Rb and Sr. Although this age is not highly significant owing to its large error, a precise initial ratio can be obtained.

Fig. 12.7 Isochron dating: Rb-Sr systematics in meteorites. The inset at lower right shows the principles of the isochron-dating technique (Section 10.3). The Rb-Sr meteoritic isochron indicates the simultaneous formation of a number of meteorites, the inferred age is similar to the Pb/Pb age but much less precise. After Minster et al. (1982).

Fig. 12.7 Isochron dating: Rb-Sr systematics in meteorites. The inset at lower right shows the principles of the isochron-dating technique (Section 10.3). The Rb-Sr meteoritic isochron indicates the simultaneous formation of a number of meteorites, the inferred age is similar to the Pb/Pb age but much less precise. After Minster et al. (1982).

Further such results for the Lewis Cliff (LEW), Angra dos Reis (ADOR) and Moore County angrites are indistinguishable from each other: 87Sr/86SrINI = 0.698 970 ± 0.000015 (Lugmair and Galer, 1992). Calcium-aluminium-rich intrusions (CAIs), which also have extremely low Rb/Sr ratios, have lower values of 87Sr/86SrINI = 0.698 80 ± 0.000 03 in accord with their greater age. Assuming that the achondrite precursors evolved with a Rb/SrSOS ratio ~ 0.3 since the time of CAI formation would give a reasonable (model-dependent) 10-15 Myr time difference between CAI formation and the achondrites (Papanastassiou and Wasserburg, 1969; Gray et al., 1973; Birck and Allegre, 1978).

The Pb/Pb chronometry of processed meteorites

Lead is a volatile element and is in addition chalcophile and slightly siderophile, while U is a lithophile refractory element. Therefore U-Pb and Pb/Pb systematics provide a suitable chronometer for the events that generated the parent bodies of non-chondritic meteorites generally indicating metal (and probably sulphide) fractionation as well as the loss of volatile elements.

Again, as in the case of CAI dating (Section 10.3), high-^ minerals are most appropriate for Pb/Pb dating. However, because of the extreme Pb loss from achondrite parent bodies, not only phosphates such as whitlockite but also common

Fig. 12.8 The Rb-Sr systematics of eucrites. Ordinary eucrites show rather low present-day 87Sr/86Sr ratios, corresponding to time-integrated Rb/Sr ratios a factor almost 100 lower than the chondritic values 0.85) on average. Even lower ratios are typical for cumulate eucrites. These meteorites exhibit a great loss of volatile Rb early in the history of their parent body. From Smoliar (1993, see references on original contributions there), © Meteoritical Society 1993, reproduced by permission.

Fig. 12.8 The Rb-Sr systematics of eucrites. Ordinary eucrites show rather low present-day 87Sr/86Sr ratios, corresponding to time-integrated Rb/Sr ratios a factor almost 100 lower than the chondritic values 0.85) on average. Even lower ratios are typical for cumulate eucrites. These meteorites exhibit a great loss of volatile Rb early in the history of their parent body. From Smoliar (1993, see references on original contributions there), © Meteoritical Society 1993, reproduced by permission.

silicate minerals such as pyroxenes have high p values. There are three high-quality Pb/Pb ages for non-chondritic meteorites.

Whole-rock samples of the angrites SAH99555 and NWA1296 show extremely high p values up to ~ 3000; these samples along with mineral separates and leachates from them yield a Pb/Pb isochron in the 207pb/206pb versus 204pb/206pb diagram giving an age of 4566.2 ± 0.1 Myr, which is the most precise absolute date of any early solar system object (Baker et al., 2005). This age is very similar to the chondrite-formation age and both are only ~ 3 Myr younger than the earliest objects of the solar system, CAIs (Fig. 11.6).

Pyroxenes from LEW 86010 and Angra dos Reis also show p values up to ~ 5500; in this case Pb/Pb dating yields identical ages of 4557.8 ± 0.5 and 4557.8 ± 0.4 Myr respectively for the two meteorites (Lugmair and Galer,

1992; see Fig. 13.1). The dated minerals are not secondary but are direct products of melt crystallization. They do not show signs of later disturbance and therefore yield an age for the magmatic differentiation. The Acapulco meteorite also contains primary magmatic phosphate, which has yielded a Pb/Pb age of 4557 ± 2 Myr, indistinguishable from those obtained for LEW 86010 and Angra dos Reis (Gopel etal, 1994).

The difference of almost 10 Myr between two very well-defined and robust dates on similar achondrite parent bodies is surprising. While the older set could record a 26Al heating effect, this is not possible for the younger set, which therefore more probably reflects collision heating. Independently of their precise interpretation, however, these Pb/Pb dates provide reliable absolute-age markers for early solar system history, and have been used to construct combined time scales (Chapter 13; Lugmair and Shukolyukov, 2001; Baker et al., 2005).

The Pd-Ag and Hf-W isotopic systematics in irons and achondrites

The extreme depletion of iron meteorites in moderately volatile Ag (Fig. 12.6) and the correspondingly high Pd/Ag ratios allow the measurement of radiogenic 107 Ag* and thus the use of the Pd-Ag chronometer for iron meteorites, even though 107 Ag is the most abundant isotope of silver. The 107Pd-107Ag systematics (half-life 6.5 Myr; Table 3.3) present an important constraint on the timing of the devolatilization of iron meteorites and associated processes. The best (reference) Pd-Ag evolution diagram was obtained for the Gibeon metal meteorite (Fig. 12.9). The diagram implies closed-system evolution, therefore the slope of the regression line gives the 107Pd/108Pd ratio at the time of the loss of moderately volatile Ag (Eqn 10.10): 107Ag*/108Pd = 107Pd/108PdINI = (2.40 ± 0.05) x 10-5. The Canyon Diablo IA iron shows a similar value (Carlson and Hauri, 2001).

High-quality Gibeon-like isochrons are not available for other iron meteorites. However, reasonable assumptions allow the relative chronology of volatile loss events. These assumptions are: (1) the solar nebula was homogeneous with respect to the initial 107Ag/109Ag ratio; (2) a meteorite exhibited a simple two-stage evolution, with solar values of the ratios 109Ag/108PdSoS and 107Ag/109AgSOS before the loss event. Then the measured ratios on a single sample suffice to determine the 107Pd/108PdINI ratio at the time of the event:

107Pd/108PdINI = (107Ag/109AgMET - 107Ag/109AgSOS) x 109Ag/108PdMET.

Substitution of the 107Pd/108PdINI ratios for the meteorite and for Gibeon in Eqn (10.11) gives the age of the loss event relative to the Gibeon age: most meteorites were formed within ~ 10 Myr after Gibeon (Fig. 12.10).

0

Fig. 12.9 The Pd-Ag evolution diagram for Gibeon metal. The great depletion of Ag makes it possible to measure the excess of 107 Ag*, the product of extinct radioactive107Pd. From Chen and Wasserburg (1996), © 1996 American Geophysical Union, modified by permission.

The 182Hf-182W systematics are useful for providing time constraints on metal-silicate melt segregation (Section 11.6). Iron meteorites have essentially zero Hf/W ratios and thus record the relative time of metal formation from the nebula, or the last equilibration of metallic melt with silicate melt. The range of s182 in iron meteorites, which is ±0.5s182 relative to the assumed SOS value, along with the respective solar system ratios (Fig. 11.8, Table 3.3) yield model ages within an interval of ~ 8 Myr, again similar to the age ranges discussed above.

It must, however, be noted that a large population of iron meteorite samples plot close to the least radiogenic value. This means either that these highly processed iron meteorites were formed at the same time as the earliest metal particles in chondrites or that they inherited unradiogenic W from their precursors, these same metal particles. The first possibility is unlikely in view of the Pd/Ag chronological evidence and the age relationships between chondritic and non-chondritic meteorites discussed above. The second possibility implies that the assumption of complete isotopic resetting does not hold for the W systematics in iron meteorites; the W-isotopic ratios are more a tracer than a chronometer for these meteorites.

As well as in metal-silicate segregation, Hf/W ratios are fractionated in silicate melting and crystallization processes. Quitte et al. (2000) found a large range of Hf/W ratios in eucrites and inferred a time of formation of the eucrite parent body

1000

Fig. 12.10 Relative ages for iron, pallasite and mesosiderite meteorites inferred from the 107Pd-107Ag systematics, plotted against Ge content as a measure of volatile depletion. Groups IVA and IVB irons are highly depleted in volatile elements and therefore these meteorites could have originated as metal condensates; correspondingly, the 107Ar*/108Pd = 107Pd/108Pd ratio inferred from the Gibeon isochron is expected to be close to the SOS initial value. Note the short formation interval, within 10 Myr (grey band), for all but two meteorites. This is in agreement with the Sr initial ratio systematics and with the Pb/Pb ages of angrites (Fig. 13.1). After Chen and Wasserburg (1996), © 1996 American Geophysical Union, modified by permission.

1000

Fig. 12.10 Relative ages for iron, pallasite and mesosiderite meteorites inferred from the 107Pd-107Ag systematics, plotted against Ge content as a measure of volatile depletion. Groups IVA and IVB irons are highly depleted in volatile elements and therefore these meteorites could have originated as metal condensates; correspondingly, the 107Ar*/108Pd = 107Pd/108Pd ratio inferred from the Gibeon isochron is expected to be close to the SOS initial value. Note the short formation interval, within 10 Myr (grey band), for all but two meteorites. This is in agreement with the Sr initial ratio systematics and with the Pb/Pb ages of angrites (Fig. 13.1). After Chen and Wasserburg (1996), © 1996 American Geophysical Union, modified by permission.

shortly (less than 1 Myr) after the Ste Marguerite chondrite; this follows from substitution of the respective initial 182Hf/180HfINI ratios (Fig. 12.11 and the inset in Fig. 11.8) into Eqn (10.11). A similar time scale follows from a simple two-stage model for the metal-silicate differentiation in Vesta: assuming the SOS abundances of the species of interest during the first stage (Table 3.3), a post-differentiation 180Hf/184WVEsTA,MANTLE ratio ~ 21 and the present-day s182 ~ 17 (Jacobsen, 2005) and substituting these values into Eqn (18.2) gives 2.5 Myr (after SOS formation). Thus core segregation on Vesta occurred very early, at about the same time as the SAH 99555 and NWA 1296 angrite parent body was being formed. The early differentiation of Vesta is further confirmed by the Mn-Cr chronometer.

The 53Mn-53Cr isotope systematics in achondrites

We have encountered the 53Mn-53Cr chronometer in the context of secondary carbonate in a chondrite (Section 11.6). In the course of the processing of asteroid-like

0.866

0.864

0.866

180Hf/1 84 W

Fig. 12.11 The 182Hf-182W systematics of eucrite meteorites. The best-fit line was calculated without Millbillillie (MIL) and the duplicate for Bereba (BER). See Eqn (19.1b) for the e notation. Substituting the initial 182Hf/180Hfsos = 1.01 x 10-4 and the inferred 182Hf/180HfINI = 7.96 x 10-5 into Eqn (10.11) gives a ~ 3.1 Myr time interval between the formation of chondrites and silicate fractionation of the eucrite parent body. For meteorite names and classes, see the List of meteorites. After Quitte et al. (2000) and Jacobsen (2005).

180Hf/1 84 W

Fig. 12.11 The 182Hf-182W systematics of eucrite meteorites. The best-fit line was calculated without Millbillillie (MIL) and the duplicate for Bereba (BER). See Eqn (19.1b) for the e notation. Substituting the initial 182Hf/180Hfsos = 1.01 x 10-4 and the inferred 182Hf/180HfINI = 7.96 x 10-5 into Eqn (10.11) gives a ~ 3.1 Myr time interval between the formation of chondrites and silicate fractionation of the eucrite parent body. For meteorite names and classes, see the List of meteorites. After Quitte et al. (2000) and Jacobsen (2005).

bodies, both Mn and Cr behave as lithophiles. They are fractionated in silicate melting and crystallization, Cr and Mn entering spinel and Ca-bearing minerals respectively. Therefore the silicate magmatic activity on achondrite parent bodies has led to differences in the Mn/Cr ratio both between batches of magma, which are now individual stony meteorites, and between minerals that crystallized in equilibrium with each other within a single meteorite. The 53Mn-53Cr systematics among minerals of the LEW 86010 and Angra dos Reis meteorites indicate a 53Mn/55Mn ratio equal to (1.25 ± 0.07) x 10-6 at the time of their crystallization at 4557.8 ± 0.4 Myr (Lugmair and Shukolyukov, 1998). Thereby the relative 53Mn-53Cr time scale is now also calibrated to an absolute age.

The steepest Mn-Cr isochron found is that yielded by whole-rock howardites, eucrites and diogenites (Fig. 12.12), which gives a 53Mn/55Mn ratio equal to (4.7 ± 0.5) x 10-6. The HED parent body therefore was likely to have been differentiated 7.1 ± 0.8 Myr before the 4557.8 ± 0.5 Myr Pb/Pb age obtained for LEW and ADOR, i.e. at 4564.9 ± 1 Myr, which is the same time as that when, according to Pb/Pb-isochron age measurements, chondrules (Fig. 11.6) and some highly

HED parent body 53Mn / 5 5Mn = (4.7 ± 0.5) x10-6 F5 " £53.INI=+0.25 ± 0.07

0

HED parent body 53Mn / 5 5Mn = (4.7 ± 0.5) x10-6 F5 " £53.INI=+0.25 ± 0.07

Fig. 12.12 TheMn-Cr isochron for whole-rock samples of thehowardites, eucrites and diogenites (HED). The average ordinary-chondrite data point is shown for comparison. The slope of the isochron gives a common 53Mn/55Mn ratio for all meteorites separated presumably from a common 4565-Myr-old parent body (see the main text). At that time the precursor material had already evolved to 0.25 e units (Eqn 19.1b) above the present day terrestrial value. From Lugmair and Shukolyukov (1998), © Elsevier Science 1998, reproduced by permission.

processed meteorites were formed. This shows that the differentiation of asteroids and the formation of parent bodies of non-chondritic meteorites started almost simultaneously with chondrite formation.

In addition to its Pb/Pb age of4557 ± 2 Myr, phosphate from the Acapulco meteorite has yielded a highly precise 129I-129Xe relative age, 9.26 ± 0.25 Myr younger than that of the Bjurbole chondrite (Brazzle et al., 1999), and this now allows us to assign an age of 4566.3 ± 2 Myr to Bjurbole, whereby the I-Xe relative time scale of Fig. 11.9 is calibrated with much greater precision and confidence than if only chondritic ages were used.

The two calibrations obtained (1) via Acapulco and 129I-129Xe systematics and (2) via angrite 53Mn-53Cr systematics can be compared, as they have independently yielded ages for the feldspar of the Ste Marguerite H4 chondrite. Using I-Xe

Relationships between extinct and live chronometers systematics and the Acapulco calibration, the feldspar has an apparent absolute age of 4567 ± 2 Myr (Brazzle etal., 1999). Using the angrite calibration, a Mn-Cr age of 4565 ± 0.7 Myr results (Lugmair and Shukolyukov, 2001). The second, more precise, result is within experimental error of the first, and the confidence in both approaches is very high. In Chapter 13 these results are considered further and combined with 26Al-26Mg systematics to provide a more complete overview of early solar system chronology.

0 0

Post a comment