www.sciencemag.org/cgi/content/full/310/5749/839/DC1
Supporting Online Material for
Meteorite Phosphates Show Constant 176Lu Decay Rate Since 4557 Million
Years Ago
Yuri Amelin
E-mail: [email protected]
Published 4 November 2005, Science 310, 839 (2005)
DOI: 10.1126/science.117919
This PDF file includes:
Materials and Methods
SOM Text
Fig. S1
Tables S1 to S3
References and Notes
Analytical procedures
The surface of the Acapulco sample was cleaned by ultrasonic agitation in ethanol. The
sample was gently crushed in an alumina mortar, and washed with ethanol to remove a very fine
fraction. Metal was removed with a hand magnet. The remaining material was separated by
magnetic susceptibility using a Frantz barrier magnetic separator. Phosphate grains are found in
the two least magnetic fractions. These fractions also contain feldspar, therefore splits of the
least magnetic fractions were additionally separated using a heavy liquid (methylene iodide
diluted to the density of 3.15 g/cm3). All fractions were checked under a binocular microscope,
and some of the fractions were additionally purified by hand picking.
Phosphate fractions were spiked with mixed 176Lu-180Hf and 230Th-235U-205Pb mixed tracers
before dissolution. The 176Lu-180Hf mixed tracer, made of highly enriched 176Lu (98.5%) and
180
Hf (98.2%), has been calibrated against the mixed Lu-Hf reference solution prepared by F.
Corfu (1) of high purity Ames metals dissolved together. Spike calibration was additionally
tested against a mixed Lu-Hf reference solution (2) provided by M. Bizzarro. Calibration of
176
Lu/180Hf ratio in the spike against these two reference solutions agreed within 0.3%.
Phosphates were dissolved by overnight treatment with 0.5M HNO3-0.2M HBr at 90°C.
Insoluble residues of feldspar and minor pyroxene, present in some fractions, were separated by
centrifugation. In order to assure isotopic equilibration of Hf between sample and spike,
dissolved samples were evaporated with 0.05 ml of HClO4, following (3). Samples were redissolved in 3M HCl, and passed through columns containing 0.05 ml of the Bio-Rad1Wx8
resin. After removing the bulk sample with additional amount of 3M HCl, U and Pb were
extracted with 0.5M HNO3-0.2M HBr and 0.5M HNO3-0.03M HBr, respectively (procedure
modified from (4)). The washes from the U-Pb column, containing Lu and Hf (along with Ca
and most other elements) in 3M HCl, were passed through a column with 0.05 ml of Eichrom
Ln-Spec resin http://www.eichrom.com/products/tech/lnresin.cfm (procedure modified from (3,
5)). After removal of the major elements and light rare earth elements with additional 3M HCl,
Lu (together with Yb and other heavy rare earth elements) was extracted with 6M HCl.
Separation of Yb from Lu was not necessary, because the mass spectrometric procedure for Lu
isotopic analysis used here effectively eliminates Yb. Hf, together with Zr, was extracted with
0.5M HNO3-0.2M HF. The Hf-Zr fractions in 0.5M HNO3-0.2M HF were passed through
columns with 0.05 ml of the Bio-Rad50Wx8 resin, to assure complete removal of REE. The
column washes containing Hf were diluted with additional 0.5M HNO3, to produce ca. 1 ml of
sample solution in 0.5M HNO3-0.07M HF, ready for isotopic analysis.
Isotopic composition of Hf was measured on a Nu Plasma MC-ICPMS at the Geological
Survey of Canada (GSC), using a desolvating nebulizer DSN-100 for sample introduction. The
typical sample uptake rate was 0.1 ml/min, and the duration of analysis was about 10 min.
Because the samples studied here are very small, obtaining the lowest possible instrument
background was a necessary condition for analysis. At the beginning of each of the two
analytical sessions, the DSN was thoroughly cleaned by filling with hot 2% HNO3 (this
procedure, recommended by the Nu Instruments, the manufacturer of the mass spectrometer and
the DSN, is similar to the procedure described in (6)). In addition to reduction of the
background, this washing procedure greatly accelerates washout between sample analysis. Final
cleaning of the DSN, aimed at removal of any Hf possibly insoluble in HNO3 was achieved by
overnight aspiration of 0.5M HNO3-0.2M HF.
Ion beams were measured using the array of three ion counting secondary electron
multipliers in three mass steps as shown below:
Mass step
IC0
IC1
IC2
1
180
178
176
2
179
177
175
3
178
176
174
A correction factor including mass discrimination (assuming power law) and the ratio of
multiplier gains was calculated from the 179Hf/177Hf ratio measured in the JMC-475 standard and
unspiked samples. This correction factor was applied to the isotopic ratios 180Hf/178Hf and
178
Hf/176Hf, measured in the mass steps 1 and 3 using the multipliers IC0 and IC1. The corrected
value of the 178Hf/176Hf ratio was used to calculate the correction factor for the multipliers IC1
and IC2, using this ratio measured in the mass step 1. This procedure is similar to the multidynamic method of U isotope analysis using an array of two multipliers (7). The values of the
correction factor measured during one day were combined, and the median value was used to
correct the measurements of spiked samples. The variations of the correction factor during one
day were between 0.03-0.21% (2 MAD (8)). The variations between the median values for the
five days of analyses in May-June 2005 are 0.28% (2 MAD).
After correction for mass discrimination and multiplier gains, the 178Hf/176Hf isotopic ratio
was corrected for spike contribution, and converted to 176Hf/177Hf assuming 178Hf/177Hf= 1.46743
(the average value from static multicollector analyses of larger loads of unspiked Hf on the GSC
Nu Plasma). The 180Hf/178Hf ratio was converted to 180Hf/177Hf using the same value of
178
Hf/177Hf for unspiked samples, and the matching 180Hf/177Hf and 180Hf/178Hf for model sample
– spike mixtures.
The typical intensity of ion beams of all Hf isotopes during sample and standard (20 pg/g and
50 pg/g solutions of the JMC-475 Hf in 0.5M HNO3) analyses was below 105 cps (counts per
second), and in some cases below 104 cps. For this intensity range, accurate background
subtraction is a crucial condition for obtaining accurate isotopic data. The background ion
beams were measured using the same procedure as the samples. The median values of 259-511
cps 180Hf, 41-90 cps 176Hf (as well as the values for all other isotopes) measured in 0.5M HNO30.07M HF, and 123-146 cps 180Hf, 28-35 cps 176Hf, measured in 0.5M HNO3, were subtracted
from the sample and standard ion beams, respectively, during off-line data reduction. In all
sample and standard analyses, no detectable ion beam after background correction was observed
on the mass 175, and the signal on the mass 174 corresponds to the abundance of 174Hf.
Therefore no correction for Yb and Lu interferences were applied.
Analyses of the Hf isotopic standard JMC-475 are summarized in the table below:
Standard solution
JMC-475 0.2ppb
JMC-475 0.2ppb
JMC-475 0.2ppb
JMC-475 0.2ppb
180
Hf_IC0
mV
176
Hf/177Hf
2Mad
mean%
Hf/177Hf
Average
2 sigma
2 sigma %
1.1179
0.0016
0.14
1.8860
0.0022
0.11
0.28237
0.00072
0.25
Median
2MAD
2MAD%
2MADmean%
1.1179
0.0014
0.12
0.060
1.8860
0.0019
0.098
0.049
0.28245
0.00031
0.11
0.056
1.10091
1.10362
1.10527
1.10044
1.10314
1.10802
1.10446
1.8902
1.8878
1.8910
1.8793
1.8764
1.8849
1.8822
3.5207
6.8360
17.7536
Average
2 sigma
2 sigma %
1.1037
0.0052
0.47
1.8845
0.0110
0.59
0.2839
0.0026
0.92
Median
2MAD
2MAD%
2MADmean%
1.1036
0.0033
0.30
0.11
1.8849
0.0105
0.56
0.21
0.2844
0.0014
0.49
0.15
0.00012
0.00012
0.00012
0.00012
0.00027
0.00029
0.00029
0.00014
0.00014
0.00015
0.031
0.031
0.029
0.028
176
1.8868
1.8849
1.8870
1.8852
0.00078
0.00078
0.00078
0.00078
0.00174
0.00185
0.00183
0.00162
0.00327
0.00918
0.00120
0.00120
0.00122
0.00123
180
1.11843
1.11734
1.11869
1.11706
JMC-475 0.02ppb
JMC-475 0.02ppb
JMC-475 0.02ppb
JMC-475 0.02ppb
JMC-475 0.05ppb
JMC-475 0.05ppb
JMC-475 0.05ppb
JMC-475 0.02 ppb spiked 2x
JMC-475 0.02 ppb spiked 4x
JMC-475 0.02 ppb spiked 11x
0.00734
0.00723
0.00733
0.00735
Hf_IC2 F_01 *)
mV
IC0/IC1
0.087
0.134
0.112
0.167
0.079
0.073
0.058
0.046
0.056
0.040
0.28240
0.28271
0.28187
0.28251
0.28445
0.28382
0.28454
0.28521
0.28284
0.28184
0.28170
0.28468
0.28445
0.28520
2Mad
mean%
0.036
0.037
0.048
0.040
0.188
0.126
0.115
0.165
0.080
0.106
0.099
0.121
0.103
0.126
*) F_01 IC0/IC1 is the correction factor including mass discrimination (assuming power law) and the ratio of
multiplier gains for multipliers IC0 and IC1.
The median values of 1.885±0.011 (2 MAD) and 0.2844±0.0014, were measured in 20 pg/g
and 50 pg/g solutions of the JMC-475 Hf (20-50 pg of Hf per analysis, respectively). The Hf
isotopic ratios in Table S1 in the “Results” section below are reported relative to the accepted
176
Hf/177Hf=0.28216 in the JMC-475 standard, and their uncertainties are quadratic sums of the
within-run statistics and the reproducibility of standard analysis. The accuracy of the procedure
was additionally tested by analysis of spiked JMC-475 standard, with 180Hf/177Hf ratios between
3.5-17.7 (matching the range of these ratios in most sample analyses). The corrected 176Hf/177Hf
ratios in the spiked JMC-475 standard are within the range of unspiked analyses, thus
conforming the accuracy of spike subtraction.
The isotopic ratios of Lu were measured on a Triton TI mass spectrometer, from single Re
filament loads with colloidal graphite, using a secondary electron multiplier. This procedure is
similar to the used previously in Lu-Hf analyses of single zircon grains (12). The presence of Yb
was monitored on the mass 174, and the correction to the signal measured on mass 176 was
applied, but the ion beam of Yb was always negligible at the temperature optimal for Lu
ionization (1370-1450°C).
Analytical blanks between 0.9-1.5 pg Hf, 0.03-0.07 pg Lu (n=4) were measured in the first
analytical session, and the blanks between 0.09-0.14 pg Hf, 0.02-0.04 pg Lu (n=5), were
measured in the second session. Blank measurements were processed like sample
measurements, including calculation of the 176Hf/177Hf ratio, which was similar to 0.28, within
error, in all cases. Blank correction was applied to each isotope using the average blank values
for each session, and uncertainties in the blank values were propagated into the final errors of the
176
Lu/177Hf and 176Hf/177Hf ratios. Despite the low blank values, the uncertainty in the blank
correction is the main source of uncertainty in 176Lu/177Hf in the fractions with the most
radiogenic Hf ratios but the lowest content of unradiogenic Hf (e.g., fractions 2, 6, 22, 23).
Isochron regressions, MAD values, and weighted average calculations were performed using
Isoplot-Ex 3.00 (9).
The quality of the entire procedure is additionally verified by Lu-Hf dating of milligramsized apatite fractions from carbonatites and phoscorites of the Kovdor massif (13), acquired as a
part of an ongoing search for a standard mineral for Lu-Hf dating. The Lu-Hf data for the
Kovdor apatites are summarized in the table below:
Sample
ng Hf
ng Lu
176
Lu/
(Hf)
(Lu)
mg
ppm
ppm
KV-2
6.67
0.0306
0.3784
0.204
2.52
1.7522
0.85
0.2952
0.28
KV-5
5.65
0.0703
0.4197
0.397
2.37
0.8471
0.48
0.2884
0.28
KV-6
5.17
0.0787
0.4169
0.407
2.16
0.7520
0.49
0.2879
0.28
KV-8
3.75
0.0938
0.4267
0.352
1.60
0.6454
0.52
0.2866
0.27
KV-9
6.91
0.0065
0.2003
0.045
1.38
4.3726
3.69
0.3129
0.45
KV-18
6.56
0.1384
0.3638
0.908
2.39
0.3730
0.39
0.2856
0.27
KV-21
3.44
0.0255
0.2336
0.088
0.80
1.2981
1.92
0.2916
0.29
KV-22
6.10
0.0088
0.5783
0.054
3.53
9.3224
3.12
0.3503
0.58
KV-25
3.32
0.0710
0.3879
0.236
1.29
0.7749
0.74
0.2883
0.29
177
2σ%
176
Weight
Hf/
2σ%
177
Hf
Hf
Lu-Hf isochron regression of all these analyses, shown in the figure below, yielded the age of
382±14 Ma using the 176Lu decay constant of 1.867*10-11 year-1. This age is in excellent
agreement with previous U-Pb and Th-Pb age determinations between 377-382 Ma for
baddeleyite, zircon and apatite fractions from the same samples, and with the best estimate for
the age of Kovdor massif of 378.54±0.23 Ma, based on the 238U-206Pb systematics of baddeleyite
free from Pb loss (13).
data-point error crosses are 2σ
0.37
176
Hf/177 Hf
0.35
0.33
0.31
Age = 382 ± 14 Ma
0.29
Initial
176
Hf/177Hf =0.28247 ± 0.00041
MSWD = 0.75
0.27
0
2
4
6
176
8
10
12
Lu/177Hf
References to Analytical Procedures
1. F. Corfu, S. R. Noble, Genesis of southern Abitibi greenstone belt, Superior Province,
Canada: Evidence from zircon Hf isotope analyses using a single filament technique.
Geochim. Cosmochim. Acta 56, 2081-2097 (1992).
2. M. Bizzarro, J. A. Baker, D. Ulfbeck, A new digestion and chemical separation technique
for rapid and highly reproducible determination of Lu/Hf and Hf isotope ratios in
geological materials by MC-ICP-MS. Geostandards Newsletter 27, 133-145 (2003).
3. G. H. Barfod, O. Otero, F. Albarède, Phosphate Lu-Hf geochronology. Chem. Geol. 200,
241-253 (2003).
4. G. W. Lugmair, S. J. C. Galer, Age and isotopic relationships among the angrites Lewis
Cliff 86010 and Angra dos Reis, Geochim. Cosmochim. Acta 56, 1673-1694 (1992).
5. C. Münker, S. Weyer, E. Scherer, K. Mezger, Separation of high field strength elements
(Nb, Ta, Zr, Hf ) and Lu from rock samples for MC-ICPMS measurements.
Geochemistry, Geophysics, Geosystems (G-Cubed) 2, Paper # 2001GC000183.
6. K. D. Collerson, B. S. Kamber, R. Schoenberg Applications of accurate, high-precision
Pb isotope ratio measurement by multi-collector ICP-MS. Chem. Geol. 188, 65– 83
(2002).
7. J. E. Snow, J. M. Friedrich, Multiple ion counting ICPMS double spike method for
precise U isotopic analysis at ultra-trace levels. International Journal of Mass
Spectrometry 242, 211–215 (2005).
8. MAD = Median Absolute Deviation from the Median. The MAD is a robust estimator of
variability, which is used with medians in the same way that standard deviations are used
with averages (the definition is adapted from the manual to the Isoplot 3.0 (9)).
Following (10-11), I used a robust (nonparametric) data treatment (median and MAD,
calculated with a user-defined function provided with the Isoplot/Ex), which makes no
assumptions about the nature or distribution of the scatter of the data points, for
processing ion beams and isotopic ratios throughout this study. This was necessary
because normal distribution cannot be assumed in the presence of small but significant
irregularities in the washout, which add a skewed population of spikes to the dataset.
9. K. R. Ludwig, Isoplot/Ex version 3.00, A Geochronological Toolkit for Microsoft Excel,
Berkeley Geochronology Center Special Publ. 4, 71pp (2003),
http://www.bgc.org/Isoplot3betaManual.pdf.
10. N. M. S. Rock, J. A. Webb, N. J. McNaughton, et al., Nonparametric-estimation of
averages and errors for small data-sets in isotope geoscience - a proposal. Chem. Geol.
66, 163-177 (1987).
11. C. Reimann, P. Filzmoser, Normal and lognormal data distribution in geochemistry:
death of a myth. Consequences for the statistical treatment of geochemical and
environmental data. Environmental Geology 39, 1001-1014 (2000).
12. Y. Amelin, D.-C. Lee, A. N. Halliday, R. T. Pidgeon, Nature of the Earth’s earliest crust
from hafnium isotopes in single detrital zircons. Nature 399, 252-255 (1999).
13. Y. Amelin, A.N. Zaitsev Precise geochronology of phoscorites and carbonatites: The
critical role of U-series disequilibrium in age interpretations. Geochim. Cosmochim. Acta
66, 2399–2419 (2002).
Fraction description:
Acapulco:
1. One large slightly yellowish apatite grain.
2. One large colorless clear merrillite grain.
3. One large colorless clear apatite grain.
4. 14 mid-sized clear colorless grains.
5. 16 mid-sized grains, some contain small dark inclusions and minor yellow staining.
6. 20 mid-sized grains with extensive yellow staining.
7. Multiple grains with yellow staining from “non-magnetic @0.8A” fraction with density
ρ<3.15 g/cm3.
8. Multiple grains with dark inclusions from “non-magnetic @0.8A” fraction with density
ρ<3.15 g/cm3.
9. Multiple colorless clear grains from “non-magnetic @0.8A” fraction with density ρ<3.15
g/cm3.
10. Multiple grains (assorted) from “non-magnetic @0.8A” fraction with density ρ<3.15
g/cm3.
11. Multiple grains with minor staining and inclusions from “non-magnetic @0.8A” fraction
with density ρ>3.15 g/cm3.
12. An aliquot of “non-magnetic @0.8A” fraction without density separation.
13. An aliquot of “non-magnetic @0.8A” fraction without density separation (similar to the
fraction 12).
14. Multiple grains (assorted) from “non-magnetic @0.8A” fraction with density ρ<3.15
g/cm3 (similar to the fraction 10).
15. Multiple grains with minor staining and inclusions from “non-magnetic @0.8A” fraction
with density ρ>3.15 g/cm3 (similar to the fraction 11).
Richardton:
16. Multiple grains with yellow surface staining from “non-magnetic @1.0A” fraction.
17. Multiple grains with dark inclusions or turbidity from “non-magnetic @1.0A” fraction.
18. Multiple colourless clear grains from “non-magnetic @1.0A” fraction.
19. Multiple assorted grains from “non-magnetic @1.0A” fraction.
20. Multiple yellow phosphate grains (previously hand-picked for U-Pb work).
21. One clear colorless grain (merrillite?).
22. 15 clear colorless grains.
23. Multiple second best colorless grains (rough surface and rare inclusions).
24. –100 mesh fraction “magnetic @1.7A” – a mixture of phosphates and feldspar with
minor pyroxene.
25. Bulk “non-magnetic @1.0 A” fraction with density ρ>3.15 g/cm3.
Results
Table S1. Lu-Hf data
Frac.#
Weight
(Hf)
(Lu)
mg
ppm
ppm
ng Hf
ng Lu
176
Lu/
2σ%
177
176
Hf/
2σ%
177
Hf
Radiogenic
176
Hf
Hf, %
Acapulco
1
0.0399 0.3239
0.945
0.0129
0.038
0.4143
4.17
0.3155
0.64
10.4
2
0.0232 0.0370
3.510
0.0009
0.081
13.4662
28.65
1.4566
6.56
80.6
3
0.0366 0.4058
0.765
0.0149
0.028
0.2677
3.70
0.3052
0.55
7.4
4
0.0948 0.1633
0.659
0.0155
0.063
0.5733
3.54
0.3327
0.63
15.0
5
0.1372 0.0688
0.399
0.0094
0.055
0.8232
5.53
0.3544
0.96
20.2
6
0.0873 0.0078
0.670
0.0007
0.059
12.1919
31.36
1.3271
7.53
78.7
7
0.31 0.0077
0.007
0.0024
0.002
0.1256
1.84
0.2908
0.70
2.8
8
0.35 0.0083
0.034
0.0029
0.012
0.5870
1.43
0.3296
0.66
14.2
9
0.38 0.0088
0.039
0.0033
0.015
0.6278
1.19
0.3330
0.68
15.1
10
0.61 0.0098
0.052
0.0060
0.032
0.7511
0.72
0.3442
0.56
17.9
11
0.33 0.0325
0.420
0.0107
0.138
1.8310
0.51
0.4390
0.61
35.6
12
0.0523 0.0211
0.349
0.0011
0.018
2.3504
3.33
0.4850
1.05
41.7
13
0.0784 0.0432
0.466
0.0034
0.037
1.5331
1.19
0.4136
0.65
31.6
14
0.5551 0.0072
0.057
0.0040
0.032
1.1251
1.00
0.3788
0.60
25.4
15
0.2765 0.1099
0.907
0.0304
0.251
1.1716
0.30
0.3835
0.51
26.3
Richardton
16
0.17 0.0055
2.37
0.0009
0.402
61.1493
3.92
5.7520
1.65
95.1
17
0.24 0.0158
2.21
0.0038
0.531
19.8575
2.63
2.1451
1.07
86.8
18
0.28 0.0046
2.24
0.0013
0.626
69.7214
4.28
6.3600
3.80
95.6
19
0.43 0.0060
1.73
0.0026
0.742
40.5509
2.16
3.9355
1.23
92.8
20
0.1141 0.0040
2.87
0.0005
0.328
103.1703
7.25
9.6756
1.92
97.1
21
0.0251 0.0185
3.71
0.0005
0.093
28.4022
6.91
2.7509
2.14
89.7
22
0.0502 0.0033
3.31
0.0002
0.166
142.5724
14.38
13.1792
2.00
97.9
23
0.1252 0.0018
1.66
0.0002
0.208
131.6541
11.80
11.8603
1.41
97.6
24
1.0000 0.0410
0.24
0.0410
0.237
0.8206
0.24
0.3519
0.50
19.6
25
0.6468 0.0089
1.89
0.0058
1.225
30.0838
0.71
2.9357
0.54
90.4
Fractions 1-6 are analyzed during the first analytical session in May 2005, the other fractions –
during the second analytical session in June 2005. The sample weights shown with two decimal
places are measured using a Sartorius analytical balance with precision of 10-5 g; the sample
weights shown with four decimal places are measured using a Mettler microbalance with
precision of 10-7 g. Lu and Hf concentrations and amounts, and the isotopic ratios are corrected
for analytical blank. The percentage of radiogenic 176Hf is calculated from the difference
between the measured 176Hf/177Hf ratios, and the median chondritic176Hf/177Hf = 0.282737,
calculated from the bulk chondrite analyses of Bizzarro et al. (2003) and Patchett et al. (2004).
Table S2. Pb isotopic data for Acapulco phosphates.
Fraction #
206
Pb/204Pb
206
Pb/204Pb
204
Pb/206Pb
204
Pb/206Pb
207
%err
Pb/206Pb
(tot.)
207
Pb/206Pb
%err
Rho 207*/206* 207*/206*
(raw)
(tot)
(tot.)
4/6-7/6 Age (Ma)
age err
1
187.7
189.7
5.27E-03
0.70
0.64427
0.083
-0.100
4556.5
1.3
2
25.70
25.60
3.91E-02
1.11
0.79730
0.218
0.478
4557.8
6.6
3
147.4
149.0
6.71E-03
0.58
0.65065
0.083
-0.077
4556.3
1.3
4
102.8
103.4
9.67E-03
0.33
0.66436
0.081
-0.419
4557.1
1.4
5
120.0
121.0
8.26E-03
0.44
0.65783
0.083
-0.160
4556.7
1.4
6
76.96
78.76
1.27E-02
1.35
0.67745
0.127
0.647
4555.6
1.6
Fraction numbers are keyed to Table S1. The errors are 2σ of the mean. The 206Pb/204Pb ratio
labeled “raw” is the measured ratio. Isotopic ratios labeled “total” (tot.) are corrected for
fractionation, analytical blank of 1.0 pg Pb, and spike, but not corrected for initial Pb. “Rho 4/67/6” is the error correlation between 204Pb/206Pb and 207Pb/206Pb. The 207Pb*/206Pb* age is
calculated assuming primordial Pb as the initial Pb.
Table S3. U-Pb data for Acapulco phosphates.
Fraction # ppm U ppm Pb 206Pb/204Pb 238U/206Pb 238U/206Pb 207Pb/206Pb 207Pb/206Pb 204Pb/206Pb 204Pb/206Pb Rho
(tot)
(tot.)
%err
(tot.)
%err
(tot.)
%err
4/6-7/6
1
10.69
19.43
189.3
0.92418
0.66
0.64427
0.083
5.27E-03
0.70
-0.100
2
0.23
1.19
25.55
0.63695
6.49
0.79730
0.218
3.91E-02
1.11
0.478
3
8.66
16.24
148.7
0.91354
0.43
0.65065
0.083
6.71E-03
0.58
-0.077
4
6.03
12.50
103.2
0.87099
0.43
0.66436
0.081
9.67E-03
0.33
-0.419
5
2.23
4.37
120.8
0.89738
0.36
0.65783
0.083
8.26E-03
0.44
-0.160
6
0.57
1.28
78.60
0.85831
0.73
0.67745
0.127
1.27E-02
1.35
0.647
Fraction numbers are keyed to Table S1. The errors are 2σ of the mean. Isotopic ratios are
shown in the format requited for calculation of a three-dimensional total Pb/U isochron. Isotopic
ratios labeled “total” (tot.) are corrected for fractionation, analytical blank of 1.0 pg Pb and 0.3pg
U, and spike, but not corrected for initial Pb. U and Pb concentrations are corrected for
analytical blank. “Rho 4/6-7/6” is the error correlation between 204Pb/206Pb and 207Pb/206Pb.
Error correlations between 207Pb/206Pb-238U/206Pb and
zero for all samples.
204
Pb/206Pb-238U/206Pb are assumed to be
1. Real initial heterogeneity in 176Hf/177Hf could have been caused by incomplete mixing of
different nucleosynthetic components, or by radiogenic effects from very early Lu/Hf
fractionation in precursor material of the studied meteorites. Apparent initial
heterogeneity could be a result of later migration of Lu and/or Hf.
2. Meteorites contain two calcium phosphate minerals: apatite Ca5(PO4)3(OH,F,Cl), and
merrillite (also called whitlockite) –(Ca,Mg)3(PO4)2, http://www.mindat.org/min6577.html. Merrillite can contain up to a few percent of Na, Y and heavy rare earth
elements. The difference in density, magnetic susceptibility, and optical properties
between apatite and merrillite is small, and these minerals are not separated by
conventional mineral separation procedure as used here, and cannot be easily
distinguished by appearance during hand-picking under a binocular microscope. Apatite
and merrillite are easily recognized by the difference in chemical composition (the
presence of Cl in apatite, and Mg and Na in merrillite) using electron microprobe or
electron microscope.
3. MSWD = Mean Square of Weighted Deviates. It is, roughly, a measure of the ratio of the
observed scatter of the points (from the best-fit line) to the expected scatter (from the
assigned errors and error correlations). If the assigned errors are the only cause of scatter,
the MSWD will tend to be near unity. MSWD values much greater than unity generally
indicate either underestimated analytical errors, or the presence of non-analytical scatter.
MSWD values much less than unity generally indicate either overestimated analytical
errors or unrecognized error-correlations. The Probability of Fit is the probability that, if
the only reason for scatter from a straight line is the analytical errors assigned to the data
points, the scatter of the data points will exceed the amount observed for the data (the
definitions are adapted from K. R. Ludwig, Berkeley Geochronology Center Special
Publ. 4, 71 pp (2003), http://www.bgc.org/Isoplot3betaManual.pdf).
4. The Lu-Hf data for the Richardton phosphates reported here can be used to estimate the
contribution of phosphates to the balance of Lu in chondrites. Ordinary chondrites
contain about 0.6% of Ca phosphates, mostly merrillite (A. J. Brearley, R. H. Jones, In:
Planetary materials (J. J. Papike, ed.), Reviews in Mineralogy 36, Mineralogical Society
of America, Washington, D.C., USA., 398 pp. (1998)). By combining this abundance
estimate with the median Lu concentration of 2.2 ppm in Richardton phosphate fractions,
and the median Lu concentration of 0.036 ppm in chondrite whole rocks (3, 10), we find
that ca. 38% of total chondritic Lu is contained in phosphates. Considering the relatively
low 176Lu/177Hf=0.03102, reported for Richardton whole rock (3), this is a minimum
estimate. Variations of phosphate abundance between 0 and 1-1.5% can explain the
observed range of variations of Lu/Hf ratios in ordinary chondrites.