Lunar and Planetary Science XLVIII (2017)
2309.pdf
THE CARBON ISOTOPE COMPOSITION OF THE SUN. J. R. Lyons1, E. Gharib-Nezhad2, T. R. Ayres3,
1
School of Earth and Space Exploration, Arizona State University, 781 S. Terrace Rd, Tempe, AZ, 85287, USA,
[email protected], 2School of Molecular Sciences, Arizona State University, Tempe, AZ 85287, USA, 3Center for
Astrophysics and Space Astronomy, University of Colorado, Boulder, CO 80309, USA.
.
Introduction: Light stable isotope compositions
of meteorites, planets, and the Sun constrain how our
solar system formed and the nature of the formation
environment, with the Sun considered most representative of the bulk starting material. The NASA Genesis
mission succeeded in measuring O and N isotope ratios
in returned solar wind samples [1], [2]. After accounting for isotope fractionation due to ion collisions in the
solar atmosphere [3], extrapolation of the solar wind
results to the solar photosphere demonstrate that
Earth’s O and N reservoirs experienced a very different
isotopic history than that of the bulk Sun. For oxygen
isotopes this result was anticipated based on isotope
data from inclusions in primitive meteorite [4], [5], and
it was suggested that photochemical processes in the
solar nebula were responsible for the isotopic difference between the bulk Sun and planetary materials [5],
[6], [7]. Carbon isotopes were not expected to be similarly affected by nebular photochemistry due to the
vigor of exchange reactions [5]. We find here that
Earth is, in fact, highly enriched in 13C relative to the
Sun.
Spectroscopy of the solar photosphere: CO
rovibrational transitions dominate the 2-5 micron spectral window of the photosphere. Shuttle-borne ATMOS
FTS data contains thousands of CO fundamental
(∆v=1) and first-overtone (∆v = 2) lines, recorded at
high signal-to-noise ratio and at high spectral resolution (ω/Δω~150,000) [8]. Following Ayres et al. [9],
we constructed ‘hybrid’ line profiles by co-adding absorption lines of CO isotopologues with similar excitation energy (Elow), wavenumber range, and absorption
depth (Fig. 1). To compute accurate O and C isotope
ratios from CO photospheric lines, accurate line
strengths are necessary for each rovibrational transition. Using the latest dipole moment functions [10] we
have found differences of about 3-6% compared to
some previously published values [11], and nearagreement with others [12] (Fig. 2). A radiationhydrodynamic model of the solar photosphere was used
to capture convection-related temperature variations
associated with solar granulation [13]. The radiative
transfer scheme used is described in [9].
Isotope ratios of O and C in the photosphere:
Our 18O abundance for the temperature-enhanced pho-
tosphere is 18OSMOW = -50 ± 11‰, which is the same
within errors as the inferred ratio from Genesis (Fig.
3). The 1- error includes contributions from sample
random errors, the variability of the 3-D hydrodynamic
snapshots, and uncertainty in the f-values. Our 17O value is 17OSMOW = -65 ± 33‰, which cannot distinguish
between the Genesis photosphere value and a terrestrial
value at the 2- level due to the low SNR of the 12C17O
lines. Our results provide the first accurate, directly
determined 18O/16O isotope ratio for the solar photosphere, and provide support for the dominant role of
inefficient Coulomb drag in fractionation of O isotopes
from the corona to the solar wind [3].
From our analysis of the ATMOS CO data we also
obtain the carbon isotope composition of the photosphere. We find the photosphere is depleted in 13C relative to terrestrial carbonates (VPDB standard) by -48 ±
7‰ (Fig. 4). Mantle carbon is believed to have a mean
13C ~ -5 ‰ [14], implying that bulk terrestrial C is
enriched in 13C by nearly as much as bulk terrestrial O
is enriched in 18O. Our results confirm measurements
of solar wind in lunar regolith grains [15], and disagree
with TiC data from CAIs in the Isheyevo meteorite
which have 13C ~ 0 ‰ [16]. Our photosphere ratio is
near the -40‰ mean determined for Jupiter [17], but
with large uncertainties on the latter.
Implications: Our results demonstrate that bulk
Earth, Mars, and asteroids, are enriched in 13C relative
to the starting material that formed the solar system.
This enrichment may have been a result of photochemical processing of CO in the nebula, freeze-out of CO
in the outer nebula, or could be inherited from the parent cloud. These scenarios will be discussed.
References: [1] McKeegan K. D. et al. (2011) Science, 332, 1528–1532. [2] Marty B. et al., (2011) Science, 332, 1533–1536 (2011). [3] Bodmer R. and
Bochsler P. (2000) J. Geophys. Res., 105, 47-60.
[4] Clayton R. N., Grossman L., Mayeda T. K. (1973)
Science, 182, 485–4S88. [5] Clayton R. N. (2002) Nature, 415, 860–861. [6] Yurimoto H. and Kuramoto K.
(2004) Science, 305, 1763–1766. [7] Lyons J. R. and
Young E. D. (2005) Nature, 435, 317–320. [8] Abrams
M. C. et al. (1996) Appl. Opt., 35, 2747–2751. [9]
Ayres T. R. et al. (2013) Astrophys. J., 765, 46–71.
[10] Li G. et al. (2015) Astrophys. J. Suppl. Ser., 216,
Lunar and Planetary Science XLVIII (2017)
15–32. [11] Goorvitch D. (1994) Astrophys. J. Suppl.
Ser., 95, 535-552. [12] Hure J. M. and Roueff E.
(1996) Astron. Astrophys. Suppl.,117, 561–568. [13]
Freytag B. et al. (2012) J. Comp. Phys., 231, 919-959.
[14] Deines P. (2002) Earth Sci. Rev., 58, 247-278.
[15] Hashizume K. et al. (2004) Astrophys. J., 600,
480-484. [16] Meibom A. et al. (2007) Astrophys.,
656, L33-L36. [17] Niemann H. B. et al. (1998) J.
Geophys. Res., 103, 22,831-22,845. [18] Scott P. C. et
al. (2006) Astron. Astrophys., 456, 675-688. [19] Lyons J. R., Gharib-Nezhad E., Ayres T. R. (submitted).
Figure 1. Lower energy levels versus line center frequency for hybridized lines constructed from ATMOS
FTS data for all CO isotopologues analyzed [9]. Most
hydridized lines consist of 3-6 individual lines of similar lower energy level and line center frequency. Overtone points (v = 2) are used only for 12C16O. Analyses
of rare isotopologues use only the fundamental transitions (v = 1).
Figure 2. New 12C16O gf-values for groups of co-added
lines normalized to previously published (‘old’) gfvalues [11], [12]. g = 2J″ +1 is the multiplicity for
lower state rotational quantum number J″, and the f-
2309.pdf
value is the oscillator strength. Blue points are normalized by HR96 [12]; red points are normalized by G94
[11]. Circles are for v=1 transitions, and squares are
for v=2. This figure shows that the new gf-values are
considerably closer to the gf-values of HR96 [12].
Figure 3. Directly measured oxygen isotope ratios of
the photosphere from ATMOS observations of CO.
Genesis solar wind and inferred photosphere values [1]
also shown, together with theoretical mass-dependently
fractionated values for inefficient Coulomb drag for
O6+ and O7+ ions in the solar corona. The value for
18O of the photosphere determined here is consistent
with the inferred value from Genesis and is distinct
from a terrestrial value. Error bars are 1-.
Figure 4. Carbon isotope ratios in the solar system.
The value for the Sun (in red) reported here is among
the lightest in the solar system for bulk materials, and
lighter than previously reported values (AL13 [9],
SA06 [18]). The computed values for C5+ and C6+ ions
in the solar wind (in red) are similar to solar wind in
lunar regolith [15], but differ from the ratio in TiC
from Isheyevo [16]. References for other data are in
[19].