Transcript Meteoritic Constraints on Protoplanetary Disks

Meteoritic Constraints on our
Protoplanetary Disk
Steve Desch
Arizona State University
School of Earth and Space
Exploration NEW!
Meteorites severely constrain models
of how our protoplanetary disk
formed and evolved.
Chondrule
Formation
Short-Lived Radionuclides
in CAIs (Ca,Al-rich inclusions)
Meteorites: under most astronomers’ radar?
Daylight fireball, Grand Tetons National Park, August 10, 1972
In response to the fireball / meteorite
of Luce, France, 1772...
“There are no rocks in the sky.
Therefore rocks do not fall from
the sky.”
Antoine Lavoisier (1743-1797)
Thomas Jefferson (1743-1826)
“...it is easier for me to believe that
two Yankee professors would lie than
that stones would fall from heaven.”
Chondrites are rocks from asteroids (2-3 AU).
They are mostly unaltered since the birth of the Solar System.
rocks
sky
Chondrites: Leftover crumbs
from solar system formation
Cross section of
Carraweena (L3.9)
MATRIX GRAINS
CHONDRULES
CAIs
Chondrule Formation: Key Constraints
•Peak temperatures > 1800 K
•Cooled from peak temperatures to liquidus (1800 K) in minutes
•Cooled from liquidus to solidus (1400 K) in hours
•During formation, chondrule densities > 10 m-3
•Scale of chondrule-forming events were > 1000 km
•Chondrule formation took place in presence of gas and of
matrix dust, probably in present-day asteroid-belt region
•Chondrule formation took place repeatedly over many Myr.
Chondrule Thermal Histories
Chondrules crystallized from a nearly complete melt: peak
temperatures exceeded liquidus temperature, ~ 1575C
Chondrules did not lose significant sulfur: chondrules were
completely molten for only minutes (Yu & Hewins 1998)
Chondrule textures require cooling rates ~ 100 K/hr (porphyritic) to ~
1000 K/hr (radial, barred): chondrules took hours to crystallize.
radial pyroxene texture
porphyritic texture
Chondrule Densities
About 5% of all chondrules are compound chondrules, stuck
together while molten (Ciesla et al. 2004)
Given likely relative velocities, chondrule densities were ~ 10 m-3
in zone where chondrules melted (Gooding & Keil 1981)
Chondrules lost K, Fe, Mg, Si
through evaporation, but did
not experience measureable
isotopic fractionation.
K, Fe, Mg and Si vapor didn’t
leave chondrule formation
region: demands chondrule
density ~ 10 m-3
(Cuzzi & Alexander 2006)
Chondrule Densities
Chondrule densities nc ~ 10 m-3 are high!
nc = (c / gas) gas
4/3 s ac3
= C (0.01) (10-9 g cm-3) (gas / 10-9 g cm-3)
4/3 (2.5 g cm-3) (300 m)3
nc = 0.03 C (gas /
10-9
g
cm-3)
C (gas / 10-9 g cm-3) ~ 300
m-3
Gas was very dense:
•formation at midplane
•massive disk
•gas compressed?
Chondrules concentrated
At midplane at 2 AU
locally
in
MMSN,
Hard to concentrate
gas = 2 x 10-10 g cm-3
solids to C > 100
Chondrule Formation Region
In order for chondrules to collide, region must be >> 10 km
In order for K, Fe, Mg, Si vapor to not diffuse away from
chondrules, size of chondrule formation region must exceed
1000 km (Cuzzi & Alexander 2006)
Chemical kinetic models show evaporation took place in presence
of gas, P > 10-3 atm (Alexander 2004)
Chondrules and matrix grains have complementary chemical
compositions (Wood 1985; Palme et al. 1993): chondrules formed in the
presence of dust, in the asteroid belt region
Chondrule Formation Epoch
Relict chondrules are found within other
chondrules: chondrule formation occurred
repeatedly
Chondrule formation was
still taking place 2 Myr
after CAIs formed (Amelin
et al. 2002)
Some chondrules formed at
same as CAI formation
(Bizarro et al. 2004)
Chondrules formed over a
2 Myr span
Chondrule Formation: Models
•Asteroid impacts (Urey & Craig 1953; Urey 1967; Sanders 1996)
•Asteroid magmatic processes (Merill 1920; Chen et al. 1998; Lugmair &
Shukolyukov 2001) No gas, matrix dust. Thermal histories? Scale?
No gas, matrix dust. Thermal histories?
•Formation by ablation in bipolar outflows (Skinner 1990; Liffman 1995)
•Flares near early Sun / X-wind model (Shu et al. 1996,1997, 2001)
Scales too small. Thermal histories?
•Magnetic flares within the solar nebula (Levy & Araki 1989)
•Lightning (Morfill et al. 1993; Pilipp et al. 1998; Desch & Cuzzi 2000)
•Shock waves in solar nebula gas (Wood 1963; Iida et al. 2001; Desch &
Connolly 2002; Ciesla & Hood 2002)
Thermal histories just right!
Chondrule Formation: Shock Models
gas ~ 0.3 - 3 x 10-9 g cm-3
C < 10 (porphyritic) to
C ~ 100 (radial, barred)
Vs ~ 6 - 9 km s-1
Reproduce chondrule thermal histories
Desch & Connolly (2002)
Iida et al. (2001)
Ciesla & Hood (2002)
Desch et al. (2005)
Chondrule Formation: Shock Models
•X-ray flares impinging on top of disk (Nakamoto et al. 2005)
•Clumpy accretion (Boss and Graham 1993) Densities too low
•Accretion shock on top of disk (Ruzmaikina & Ip 1994)
•Accretion shocks in Jovian subnebula (Ruffert & Nelson 2005)
•Bow shocks of eccentric planetesimals (Hood 1998; Weidenschilling
et al. 1998; Ciesla et al. 2004)
Scales too small?
Timing?
•Large-scale shocks from gravitational instabilities (Wood 1984;
Wood 1996; Desch & Connolly 2002; Boss & Durisen 2005)
Satisfies all constraints!
Chondrule Formation: Shock Models
spiral-density waves
shock front at 2-3 AU,
propagating at about
5-10 km s-1 relative to
surrounding gas
Boss & Durisen (2005)
Short-Lived Radionuclides
CAIs & chondrules held radionuclides like 26Al, (t1/2 = 0.7 Myr)
26Mg
24Mg
26Al/27Al
27Al
24Mg
26Mg
24Mg
=
26Mg
24Mg
+
0
26Al
27Al
x
0
27Al
24Mg
~ 5 x 10-5
Short-Lived Radionuclides
WARNING! 26Al/27Al = 5 x 10-5 is NOT the value at the beginning
of the Solar System, nor when the first CAIs formed. It is the value
when most CAIs stopped being isotopically disturbed (melted).
Bulk isotopic analyses show that 26Al/27Al was as high as
6 x 10-5 (Young et al. 2005) at some point. Most CAIs experienced
isotopic closure (stopped melting) 0.4 Myr after that.
Solar System existed for an unknown period of time before the time
when 26Al/27Al = 6 x 10-5. Was it many Myr? Was it effectively
zero (i.e., the Solar System formed already containing 26Al)?
Short-Lived Radionuclides
FUN inclusions
Some CAIs are FUN inclusions (Fractionation and Unknown
Nuclear effects), and are thermally processed more than other
CAIs, have very odd stable isotope anomalies, and show no
evidence at all for 26Al or other radionuclides (26Al/27Al < 10-8)
Most likely explanation: these CAIs formed first, before Solar
System acquired 26Al, etc. Other CAIs formed later.
“Late injection” (Sahijpal & Goswami 1998)
t ~ - 1 Myr?
t=0
t=+0.4 Myr
FUN
CAIs
Solar CAIs Solar System
CAIs stopped
System
acquired
melting,
26Al/27Al =
26Al/27Al =
formed
t=+2.4 Myr
Chondrules
Short-Lived Radionuclides
McKeegan et al. (2000)
10Be/9Be
10-3
Tachibana et al. (2006)
60Fe/56Fe
~ 10-6
~
And also 10Be, t 1/2 = 1.5 Myr (McKeegan et al. 2000; Sugiura et al. 2001;
MacPherson & Huss 2001; Chaussidon et al. 2001; Srinivasan 2002)
and 60Fe, t 1/2 = 1.5 Myr (Tachibana & Huss 2003; Huss & Tachibana
2004; Mostefaoui et al. 2005; Quitte et al. 2005; Tachibana et al. 2006)
Short-Lived Radionuclides
In fact, 9 radionuclides with t1/2 < 16 Myr
have been confirmed from meteorites...
Where did they come from?!
Especially 60Fe!!
41Ca
(t1/2 = 0.1 Myr) (Srinivasan et al. 1994, 1996)
36Cl (t
1/2 = 0.3 Myr) (Murty et al. 1997; Lin et al. 2004)
26Al (t
1/2 = 0.7 Myr) (Lee et al. 1976; MacPherson et al. 1995)
60Fe (t
1/2 = 1.5 Myr) (Tachibana & Huss 2003; Mostefaoui et al. 2004)
10Be (t
1/2 = 1.5 Myr) (McKeegan et al. 2000; Sugiura et al. 2001)
53Mn (t
1/2 = 3.7 Myr) (Birck & Allegre 1985)
107Pd (t
1/2 = 6.5 Myr) (Kelly & Wasserburg 1978)
182Hf (t
Myr) (Harper & Jacobsen 1994)
1/2 = 9
129I (t
1/2 = 15.7 Myr) (Jeffery & Reynolds 1961)
Short-Lived Radionuclides
Irradiation / Spallation within
the Solar System (a la X-wind)?
Irradiation of rocky material at
0.1 AU in principle can produce
radionuclides like 60Fe
Only 64Ni(p,p)60Fe reaction
can happen, but 64Ni is rare and
cross section is < 0.1 mbarn
Predicted yields:
60Fe/56Fe
~ 10-11
(Lee et al 1998; Leya et al 2003)
Shu et al. (1996)
Short-Lived Radionuclides
Inherited from the Interstellar Medium?
Steady-state “average” abundance of 60Fe in Galaxy is
60Fe/56Fe ~ 3 x 10-8 (Wasserburg et al 1996) to
60Fe/56Fe
~ 3 x 10-7 (Harper 1996), even with prompt delivery.
The ratio
decreases the longer molecular clouds take to form stars.
Injected by AGB star just before Solar System formation?
AGB star would have to be < 1 pc away, < 1 Myr before Solar
System formed. Observed probability of this event is < 3 x 10-6
(Kastner & Myers 1994)
Same arguments would hold for Type Ia supernovae, novae, etc.
Short-Lived Radionuclides
Injected by supernova during Solar System formation?
Type II supernovae are the only source of 60Fe
naturally associated with star-forming regions.
70 - 90% of all low-mass stars within 2 kpc form in rich
embedded clusters, with > 100 members (Lada & Lada 2003).
Probability of a such clusters containing a massive (> 25 M)
star are ~ 70% (Adams & Laughlin 2001). Confirmed by
census of such clusters.
Therefore > 50% of all low-mass stars are associated with
massive stars that will go supernova within a few Myr.
Short-Lived Radionuclides
Injected by supernova during Solar System formation?
Type II supernovae are the only source of 60Fe
naturally associated with star-forming regions.
protoplanetary
disks
HST image,
Orion Nebula
~ 0.2 pc
1 Ori C: 40 M star will
supernova in 1-4 Myr
Short-Lived Radionuclides
G353.2+0.9 H II
region in NGC 6357
(Healy et al. 2004;
Hester & Desch 2005)
Pismis 24-1 (O3 If*),
Pismis 24-17 (O3 IIIf*)
and Wolf-Rayet stars
(Massey et al. 2001)
~ 0.4 pc
These stars will
supernova in < 1 Myr
Short-Lived Radionuclides
Injection of radioactive grains directly into protoplanetary
disk supplies just enough 60Fe!
Iron likely in form of dust grains: gas-phase Fe disappeared
from SN 1987A ejecta at same time (2 years post-explosion)
that 10-3 M of dust formed (Colgan et al 1994)
Mass of 60Fe ejected by 25 M supernova ~ 8 x 10-6 M
(Woosley & Weaver 1995)
Fraction intercepted by 30 AU radius disk at 0.3 pc away ~
(30 AU)2 / 4 (0.3 pc)2 ~ 6 x 10-8
Mixed with 0.01 M of solar composition material,
60Fe
/ 56Fe ~ 1 x 10-6
One supernova can also inject the other short-lived radionuclides
with observed abundances (Ellinger et al. 2007, in preparation)
Short-Lived Radionuclides
Protoplanetary disks ~ 0.3 pc from a supernova (1051 erg) are not
destroyed! (Chevalier 2000; Ouellette et al. 2005; Ouellette et al. 2006 in prep.)
Radioactive
grains aren’t
deflected
around disk
Ouellette et al. 2006 (in preparation)
Conclusions
CHONDRULE FORMATION:
Disk was likely massive (10 x minimum-mass solar nebula), and
experienced gravitational instabilities and shocks.
Other solids would have been heated by shocks (crystalline silicates)
SHORT-LIVED RADIONUCLIDES:
Disk was too hot (at least transiently) for first 1-2 Myr for most solids
to survive.
60Fe
shows Solar System formed near a supernova in an H II region,
definitely not in a quiescent molecular cloud like Taurus.
High UV fluxes would have photoevaporated our disk down to 30
AU, and stellar encounters would disrupt orbits, consistent with
observations of Kuiper Belt.