Onset of a Planetesimal Dynamo and the Lifetime of the Solar

46th Lunar and Planetary Science Conference (2015)
MAGNETIC FIELD. H. Wang1, B. P. Weiss1, B. G. Downey1, J. Wang2, Y. K. Chen-Wiegart2, J. Wang2, C. R.
Suavet1, R. R. Fu1, E. A. Lima1, and M. E. Zucolotto3, 1Department of Earth, Atmospheric, and Planetary Sciences,
Massachusetts Institute of Technology, Cambridge, MA, USA ([email protected]), 2Photon Sciences Directorate,
Brookhaven National Laboratory, Upton, NY, USA, 3Museu Nacional, Rio de Janeiro, Brazil
Introduction: The paleomagnetism of achondritic
meteorites provide evidence for advecting metallic
core dynamos and large-scale differentiation on their
parent planetesimals. Their small sizes (~102 km) relative to planets enable new opportunities to understand
the physics of dynamo generation in size regimes with
distinct thermal evolution parameters. Furthermore,
their extremely old ages, up to just several million
years (My) younger than the age of the solar system,
offer the possibility of constraining the nebular magnetic field environment and its lifetime.
One key unknown about planetesimal dynamos is
their onset time. Some theoretical studies have suggested that it might occur instantaneously after largescale melting [1, 2] while others have argued that a
dynamo could be delayed by several to tens of My or
longer [3, 4]. Here we present the first paleomagnetic
constraint on the onset time of a planetesimal dynamo,
with implications for the physics of core formation,
planetary thermal evolution, and dynamo generation
Another key unknown is the temporal evolution of
the solar nebula and its magnetic fields. Nebular fields
have been proposed to play a key role in the mass and
momentum evolution of protoplanetary disks [5, 6] and
may have been associated with the formation of chondrules [7]. Because the oldest basaltic achondrites are
thought to have formed during and soon after the observed ~3-5 My lifetime of protoplanetary nebulae
around Sun-like stars, they offer the possibility of constraining the late evolution and lifetime of nebular
magnetic fields and the nebula itself.
Samples and Experiments: Our study focused on
angrites, a group of ancient basaltic achondrites from
an early differentiated planetesimal. With unshocked,
unbrecciated textures, they are among the oldest
known and pristine planetary igneous rocks.
We selected two of the oldest angrites (D’Orbigny
and Sahara 99555; ~4563.4 million years old (Ma) [8,
9]) and a younger angrite (Angra dos Reis; ~4556.6
Ma [8]), which are least likely to have been contaminated by strong magnets. The two older angrites are
just ~4 My younger than the oldest known calciumaluminium inclusions (CAI, 4567.2-4567.9 Ma [10]).
Rock magnetic measurements, including hysteresis
loops, back-field demagnetization, first-order reversal
curves and thermomagnetic Curie temperature measurements, along with synchrotron transmission X-ray
microscopy [11], show that the major magnetization
carriers for all three angrites are fine-grained pseudo-
single domain magnetite particles, which are among
the most reliable paleomagnetic field recorders.
We used alternating field (AF) demagnetization
method for anhysteretic remanent magnetization
(ARM) paleointensities [1] as well as a new CO2+H2
gas mixture system [12] for controlled oxygen fugacity
thermal paleointensities. We found that the natural
remanent magnetizations (NRM) in D’Orbigny and
Sahara 99555 demagnetize at much lower coercivities
(~30 to ~50 mT, Fig. 1A) and temperatures (~300ºC)
than laboratory-applied total thermoremanent magnetization (TRM) (which persists to > 145 mT and
~500ºC). This indicates that their NRMs are not acquired during primary cooling in a paleomagnetic
field, but instead are later overprints from collectors’
hand magnets (low coercivity component, LC, < ~10
mT), viscous remanence acquired in Earth’s field and
possible partial TRMs from metamorphic events on the
angrite parent body (APB) (middle coercivity component, MC, ~10-50 mT). Unlike the MC components,
the high coercivity (HC, > 75 mT) magnetization in
both meteorites are internally non-unidirectional (Fig.
1B, C), indicating no detectable magnetic field during
initial cooling from the 580°C Curie temperature.
Fig. 1. (A) Two-dimensional projection of the endpoints of the NRM vector during progressive AF demagnetization for D’Orbigny subsample F7a. Open
(closed) symbols represent projections on the up-east
(Z-E) and north-east (N-E) planes. The LC and MC
components are labeled with purple and green arrows,
respectively. (B) Equal area projection showing paleomagnetic directions of HC magnetization in mutually
oriented D’Orbigny subsamples. (C) Directions of MC
magnetization for the same subsamples.
46th Lunar and Planetary Science Conference (2015)
The AF demagnetization spectra of NRMs in
D’Orbigny and Sahara 99555 closely resemble ~200ºC
partial TRMs acquired in a laboratory-applied field of
~10 µT, suggesting an origin from low-temperature
metamorphic reheating events during active APB dynamo period. ARM acquisition tests [13] show that
their magnetic carriers are stable below ~300ºC. We
estimate the initial cooling magnetic field paleointensities from the HC magnetization for D’Orbigny and
Sahara 99555 to be less than ~1 µT (Fig. 2A)
In contrast, the NRM of Angra dos Reis behaves
similarly to total TRM (also see [1]) and has a unidirectional HC magnetization, confirming its initial
thermal origin. It has a ~15 µT magnetic paleointensity (Fig. 2B).
Fig. 2. ARM paleointensities estimated from NRM
loss by AF demagnetization versus ARM acquisition
using HC components for Sahara 99555 and Angra dos
Reis. (A) Sahara 99555 subsample 7 for laboratory
direct current (DC) bias field of 50 µT. (B) Angra dos
Reis subsample AMC16 for laboratory DC bias fields
of 50 µT (squares), 200 µT (circles), and 600 µT (diamonds) (after [1]).
Implications: Our paleointensity results showed
that D’Orbigny and Sahara 99555 initially cooled in no
detectable magnetic field (paleointensities < ~1 µT) on
the APB ~4 My after CAI formation. On the other
hand, Angra dos Reis cooled in a ~15 µT APB core
dynamo paleomagnetic field ~11 My after CAI formation. This indicates that the APB dynamo initiated
between ~4 and ~11 My after solar system formation
(Fig. 3). This is consistent with planetesimal evolution
models calling for dynamos delayed by at least several
million years after core formation. In particular, thermal blanketing effects from 26Al decay in the mantle
could initially suppress core convection [3, 4].
The D’Orbigny and Sahara 99555 paleointensities
also suggest that external solar nebula magnetic fields
in the vicinity of the APB declined from ~50 µT (as
recorded by Semarkona chondrules) [7] at ~1.2-3 My
after CAI formation [14] to < ~1 µT at ~3.8-4.5 My
after CAI formation. These age and magnetic field
constraints suggest that the solar nebula dispersed between ~1.2-3 My and ~3.8-4.5 My after solar system
formation (Fig. 3), consistent with observed lifetimes
of infrared excesses around Sun-like stars [15, 16].
It is estimated that if magnetocentrifugal winds
and/or the magnetorotational instability play a central
role in driving stellar accretion and momentum transfer, then the observed accretion rates of Sun-like stars
would require fields of ~10-100 µT [5, 6, 7]. Because
the inferred paleointensity limits from the two older
angrites are at least an order of magnitude below these
values, magnetic fields may have ceased to play a major role in the Sun’s accretion by ~4 My after the formation of CAIs (Fig. 3). In addition, chondrules that
formed after this time in the vicinity of the APB would
have required nonmagnetic formation mechanisms like
nebular shocks [17] and planetesimal collisions [18]
rather than x-winds [19], magnetic reconnection flares
and current sheets [20].
Fig. 3. Constraints on solar nebular field evolution and
the onset of the APB dynamo from our paleomagnetic
measurements of angrites ([1] and this study) and
Semarkona chondrules [7].
References: [1] Weiss, B. P. et al. (2008) Science,
322, 713-716. [2] Elkins-Tanton, L. T. (2011) EPSL,
305, 1-10. [3] Sterenborg, M. G. and Crowley, J. W.
(2013) PEPI, 214, 53-73. [4] Roberts, J. H. et al.
(2013) LPI Conf. Abs., #8033. [5] Wardle, M. (2007)
Astrophys. Space Sci. 311, 35–45. [6] Bai, X.-N. and
Goodman, J. (2009) Astrophys. J. 701, 737–755. [7]
Fu, R. R. et al. (2014) Science, 346, 1089-1092. [8]
Brennecka, G. A. and Wadhwa, M. (2011) PNAS, 109,
9299-9303. [9] Spivak-Birndorf, L. et al. (2009) GCA,
73, 5202-5211. [10] Wadhwa, M. et al. (2013) Met.
Soc. Abs., 76, 5253. [11] Wang, J. et al. (2012) APL,
100, 143107. [12] Suavet, C. et al. (2014) GGG, 15,
2722-2743. [13] de Groot, L. et al. (2012) PEPI, 194195, 71-84. [14] Ushikubo, T. et al. (2013) GCA, 109,
280-295. [15] Haisch Jr., K. E. et al. (2001) Astrophys.
J. 553, L153–L156. [16] Mamajek, E. E. (2009) AIP
Conf. Proc., 1158, 3-10. [17] Desch, S. J. and Connolly Jr., H. C. (2002) Meteorit. Planet. Sci., 37, 183–207.
[18] Desch, S. J. and Mouschovias, T. C. (2001) Astrophys. J., 550, 314–333. [19] Shu, F. H. et al. (1996)
Science, 271, 1545–1552. [20] Levy, E. H. and Araki
S. (1989) Icarus, 81, 74–91.