1
Supplementary Figure 1
Supplementary Figure 1: Oxygen isotope compositions of solar system materials.
Terrestrial and lunar oxygen lies on the terrestrial mass fractionation line (TFL) that
represents kinetic fractionation according to mass of the isotopes (slope = 0.5).
Cosmochemical variations in oxygen are dominated by 16O variations with up to 6 %
enrichments in refractory inclusions (Calcium, Aluminium-rich Inclusions, CAI), and lie
predominantly along the Carbonaceous Chondrite Anhydrous Mineral mixing line
(CCAM; slope = 0.95). The line of slope 0.95 was originally interpreted as requiring
nucleosynthetic addition of a mixture of oxygen isotopes. However, it has been found
that a line of slope 1.00 passes through CAI and the chondrite group of meteorites
suggesting that 16O chemical fractionation is the dominant process occurring (16O
fractionation line, 16OFL). The largest deficit previously recorded in 16O (ca. 1 %) is for
the R-chondrite family. Recently, an interpretation based on photochemical
predissociation of CO by UV photons has been proposed. By this process the nebula
becomes enriched in 17O and 18O, which stick to clays and react with protons to form
water ice. The accretion disk must therefore become progressively heavier in this
scenario, and the isotopic composition of the Sun, representing the composition of the
nebula, must be the most 16O-enriched composition. In light of this prediction, we
measured metal grains from the Moon that have been exposed to solar wind. Rather
than a composition enriched in 16O, we find a composition with a 5% 16O deficit (red
symbols). This composition is not consistent with the models of photochemical
predissociation in the solar nebula. If the oxygen systematics are inherited from the
molecular cloud preceding the solar nebula, the solar composition could reflect gross
disequilibrium in the oxygen isotope compositions between solid and gas phases.
2
Supplementary Methods
Samples
Lunar soil 10084 was sieved to remove fines less than c. 30 µm in diameter. The
coarse-grained fraction was separated with a hand magnet and the magnetic fraction was
searched for metal grains. The distinction between opaque glass spherules and the
metallic iron spherules was not always obvious and so grains were loaded onto tape and
mounted in epoxy. The mounts were lightly polished so that metallic grains could be
readily recognised with reflected light microscopy. The metallic grains were recovered
from the epoxy mounts and loaded onto gold foil. The grains were picked up with a
steel needle that has sufficient magnetization to hold the steel particles. Unfortunately
this also meant that the iron grains would not detach from the needle. It was found that
brushing the Fe particles with a single filament under the meniscus of ethanol detached
the particles from the needle.
Ilmenite grains which had been polished to expose mid sections were also recovered
and loaded on to the foil. The ilmenite particles were placed such that the polished side
was “up” allowing flat polished ilmenite grains to be exposed after pressing.
The particles on the gold foil were pressed under 15,000 kPa with a polished tungsten
carbide cube. Mounts were then imaged in reflected light microscopy. Images of grains
were also made from secondary electron emission.
Analytical Techniques
Oxygen isotope compositions were determined on the ANU SHRIMP II operating
with a Cs+ primary ion beam (15 keV) and negative secondary ion beam (10 keV). As
the samples were directly loaded into gold foil, no energy compensation was necessary
and so the electron gun was not used.
3
Data were collected in two sessions using three different instrumental
configurations. These configurations relate to the collector setup and we did this to
verify the data under different setups. We initially used multiple collection with a
faraday cup and two ion counters because there was sufficient 16O- intensity in the alpha
metal grain to do so. However, the faraday cup must be accurately calibrated to the ion
counters and so to reinforce our results from the first session, we carried out single
collector analyses of grain alpha on the ETP™ multiplier. We then analyzed grain beta
for a detailed depth profile although the statistics on the single counter 17O
measurement do not show resolution between the TFL and FL. However, the
succeeding analysis with three ion counters clearly shows elevated 17O and 18O.
Grain beta was not analyzed in the Faraday Cup – two ion counter mode so as to
preserve it for further experiments (Hydrogen, carbon, nitrogen etc.)
In session 1, a primary ion beam current of 0.6-2.0 nA was used with multiple
collection of the oxygen isotopes with 16O– in a Faraday cup and 17O and 18O collected
on Sjuts™ continuous dynode electron multipliers (CDEM; tuned count rate on
standards to yield 2  105 c/s for 18O). In session 2, both multiple collection and single
collection (magnet stepped through mass positions) were used. For this session, the
primary beam was reduced to 0.1-0.3 nA to allow the 16O– signal to be measured on the
ETP discrete dynode electron multiplier (tuned to yield approximately 5  105 c/s)
resulting in lower count rates and poorer counting statistics on the minor isotopes
compared to session 1. In multiple collection mode, three Sjuts CDEM were used. The
dead time correction was 20 ns. Differential gain between the multipliers was calibrated
on standards.
Mass resolution was individually set for the three mass positions in multiple
collection through different collector slit widths: 3,000 for mass 16, 5,500 for mass 17,
and 4,000 for mass 18. In single collector mode, the mass resolution was 5,500. The
4
tailing contribution for 16OH– under 17O– was less than 1 ‰ for session 1, but as high as
10‰ for session 2 because of the lower primary beam current and larger proportional
contribution of surface OH–.
Oxygen isotope ratios were normalized to an in-house southeast Australian
magnetite standard. Lunar ilmenites were also analyzed interspersed with the lunar
metal grains to . Data are reported in standard delta notation 17O, and 18O (‰).
Deviation from the Earth-Moon mass-dependent fractionation line is reported as 17O =
17O – 0.52*18O.
The first spot analyzed of grain alpha was used as an initial test of the analytical
protocol in terms of tuning and optimization of count rates, and selection of collectors.
Following the measurements of lunar ilmenite grains, the measured 16O count rates from
the metal grains were found to be suitable for Faraday cup measurement of 16O–. This
spot (1a) was analyzed to near exhaustion; we attempted to re-optimize the secondary
ion beam (spot 1b), but this led to a secondary ion signal characterized by a massfractionated O-isotopic composition and is probably extracting secondary ions from a
different region of the grain compared to the other analyses.
Sputter Rate
Direct measurement of the depth of the analytical pits produced in the Fe metal
grains was not feasible owing to the small size of the grains compared to the primary
spot size (ca. 25 µm), and the distorted shape of the initially spherical grains pressed in
to gold. The SHRIMP primary beam angle of incidence is 45° results in elliptical spots
on flat surfaces. Sputtered pits were measured by atomic force microscopy on polished
magnetite grains analyzed during session 1 (Figure 1). Based on the average depth of
pits (430 nm sputtered in 1120 s) and the primary beam intensity, the sputter rate is
5
0.15 nm/s/nA. A small correction was applied for density differences between magnetite
and ilmenite and Fe metal (c.f. R. G. Wilson and G.R. Brewer; Ion Beams, Krieger
Publishing, Malabar Florida, 1973).
Supplementary Figure 2. Ion probe spot profiles. Atomic Force Microscope image of a
spot (upper) and cross section (lower). Analyses performed by Dr Tim Senden, RSPSE,
ANU.
Instrumental mass fractionation
For multiple collection, the absolute instrumental mass fractionation cannot be
determined with any degree of accuracy because of the normalization to the individual
gains of the detectors. For the case of the single collector measurements, the mean
6
instrumental fractionation is c. +7‰. Variability in instrumental fractionation is much
greater for the grain mounts compared to polished surfaces. For the pooled ilmenite
analyses in session 1, the standard deviation for both 18O and 17O is ≈6 ‰. This
dispersion is much higher than normally produced in stable isotope measurements, but
is a result of variable sample orientation and its effect on secondary ion tuning and
hence instrumental fractionation.
7
Supplementary Table 1. Oxygen Isotope Data
Sample
18O
±
17O
±
17O
±
Session 1: MC - FC/CDEM/CDEM
Lunar Ilmenite
1.1
1.2
2.1
5.1
5.2
10.6
0.5
9.1
9.0
0.9
1.1
1.1
1.1
4.2
1.1
6.5
-2.5
2.0
-0.5
4.8
1.4
1.4
2.1
3.1
1.4
1.0
-2.8
-2.7
-5.2
4.3
1.5
1.5
2.2
3.8
1.5
53.0
-48.7
47.2
71.7
82.1
96.1
46.7
30.5
51.3
4.9
2.0
4.4
1.6
2.4
5.4
2.7
1.8
2.3
48.5
-4.8
51.4
60.8
70.1
73.3
58.8
44.7
45.6
4.9
3.8
5.1
5.8
4.5
7.0
4.1
3.7
5.6
20.9
20.5
26.8
23.5
27.4
23.4
34.5
28.8
18.9
5.5
3.9
5.6
5.8
4.7
7.6
4.4
3.8
5.7
-37.4
-40.5
1.1
1.1
-12.6
-20.9
2.7
3.0
6.8
0.2
2.8
3.1
Alpha
1.1a
1.1b
2.1
2.2
2.3
2.4
3.1
4.1
4.3
Nu
1.1
1.2
Session 2: SC – DDEM
Lunar Ilmenite
1.1
1.2
1.3
2.1
2.2
2.3
3.1
3.2
3.3
4.1
4.2
4.3
-12.3
2.1
2.3
30.5
20.1
20.6
4.6
-0.5
-9.2
-18.5
-4.4
-19.8
6.9
5.0
3.6
3.9
3.6
6.3
3.6
6.0
5.8
4.0
5.2
3.9
-7.2
-2.4
-3.8
21.8
7.6
-1.4
8.9
0.3
4.7
1.1
3.0
-3.6
5.6
5.8
5.8
5.6
5.8
5.7
5.8
6.0
6.1
6.2
8.4
6.3
-0.8
-3.5
-5.0
5.9
-2.9
-12.1
6.5
0.6
9.5
10.7
5.3
6.7
6.7
6.4
6.1
6.0
6.1
6.6
6.1
6.8
6.8
6.5
8.8
6.6
100.8
127.7
14.2
5.2
82.1
87.2
15.5
14.0
29.7
20.8
17.2
14.3
34.3
67.4
96.6
5.3
7.3
10.2
7.2
29.1
70.9
8.1
11.4
16.1
-10.6
-5.9
20.7
8.6
12.0
17.0
Alpha
5.1
5.2
Beta
1.1
1.2
1.3
8
Sample
18O
±
17O
±
17O
±
Session 2: MC – CDEM/CDEM/CDEM
Lunar ilmenite
3.1
3.2
3.3
4.1
6.1
6.2
0.6
-4.4
6.6
-5.7
0.1
3.1
1.1
1.2
1.1
1.3
1.3
1.2
0.5
-4.3
-0.5
6.6
1.8
-7.3
2.8
2.8
3.0
3.1
3.2
3.2
0.2
-2.0
-3.9
9.6
1.7
-8.9
2.9
2.9
3.1
3.2
3.3
3.3
65.0
79.0
92.2
6.3
4.7
4.9
75.1
82.4
86.8
7.9
12.8
11.1
41.3
41.3
38.9
8.6
13.0
11.4
Beta
2.1
2.2
2.3
Legend
18O = ([(18O/16O)meas/ (18O/16O)terr] – 1) × 1000
17O = ([(17O/16O)meas/ (17O/16O)terr] – 1) × 1000
meas = measured; terr = terrestrial
17O = 17O – 0.52*18O
Errors are 1.

SC = Single Collector Analysis; (10 ratios, each 2s 16O, 10s 17O, 5s 18O)
MC = Multiple Collector (10 sets of 20s collection of each isotope)
FC = Faraday Cup, DDEM = Discrete Dynode Electron Multiplier
CDEM = Continuous Dynode Electron Multiplier