New insight into the Solar System's transition disk phase provided by the unusual meteorite Isheyevo
NNew insight into the Solar System’s transition disk phaseprovided by the unusual meteorite Isheyevo
Short Title: The Solar Transition Disk Article Type: ApJL
Melissa A. MorrisState University of New York, Cortland, NY 13045-0900School of Earth and Space Exploration, Arizona State University, Tempe, AZ 85287-6004 [email protected] andLaurence A. J. GarvieCenter for Meteorite Studies, Arizona State University, Tempe, AZ 85287-6004andL. Paul KnauthSchool of Earth and Space Exploration, Arizona State University, Tempe, AZ 85287-6004Received ; accepted a r X i v : . [ a s t r o - ph . E P ] D ec ABSTRACT
Many aspects of planet formation are controlled by the amount of gas re-maining in the natal protoplanetary disk (PPDs). Infrared observations showthat PPDs undergo a transition stage at several Myr, during which gas densitiesare reduced. Our Solar System would have experienced such a stage. However,there is currently no data that provides insight into this crucial time in our PPD’sevolution. We show that the Isheyevo meteorite contains the first definitive ev-idence for a transition disk stage in our Solar System. Isheyevo belongs to aclass of metal-rich meteorites whose components have been dated at almost 5Myr after the first solids in the Solar System, and exhibits unique sedimentarylayers that imply formation through gentle sedimentation. We show that suchlayering can occur via gentle sweep-up of material found in the impact plumeresulting from the collision of two planetesimals. Such sweep-up requires gasdensities consistent with observed transition disks (10 − - 10 − g cm − ). Assuch, Isheyevo presents the first evidence of our own transition disk and providesnew constraints on the evolution of our solar nebula. Subject headings: planets and satellites: formation — protoplanetary disks —planet-disk interactions — meteorites, meteors, meteoroids
1. Introduction
The main body of evidence for processes occurring during the early Solar Systemis found within meteorites. The combination of this evidence and theoretical modelinghas led to a greater understanding of the formation and evolution of the Solar System.However, several aspects of the evolution of our protoplanetary disk remain unresolved;in particular, the accretion of planetesimals and the formation of planets. It is generallythought that planet formation depends on the presence of gas within the disk (Kokubo& Ida 2002; Ikoma & Genda 2006), but little is known about later densities, during theso-called transition phase. However, the unusual metal-rich meteorite, Isheyevo, providesinsight into this important phase of the early solar nebula.The metal-rich carbonaceous chondrites (CH: ALH85085-like, CB: Bencubbin-like, andIsheyevo) are intriguing, being mixtures of chondrules, chemically zoned metals, unzonedmetal, hydrated lithic clasts, and refractory inclusions, while matrix is largely absent(e.g., Weisberg et al. 2001; Krot et al. 2002; Rubin 2003; Campbell et al. 2005). Thezoned metals are hypothesized to have formed from a vapor-melt plume produced duringan impact between planetesimals (e.g., Krot et al. 2005; Olsen et al. 2013). An impactorigin is consistent with the majority of components found in these chondrites. Mostchondrules in metal-rich chondrites are cryptocrystalline or skeletal in texture (Weisberget al.2001; Rubin et al. 2003; Krot et al. 2005; Scott1988; Krot et al. 2001; Hezel etal. 2003), and seem to have formed differently from other chondrules, based upon theirinferred thermal histories and their young age. Whereas the formation of the majority ofchondrules has been dated to a range of disk ages 2-3 Myr after the formation of calcium-,aluminum-rich inclusions (CAIs) (Kurahashi et al 2008; Villeneuve et al. 2009), chondrulesfrom CH/CB/Isheyevo chondrites formed almost 3 Myr later (Krot et al. 2005; Krot etal. 2008; Bollard et al. 2013). Isheyevo contains lithologies characteristic of both CH 4 –and CB chondrites (Ivanova & Lorenz 2006; Ivanova et al. 2008) and phyllosilicate-bearingclasts with extreme N-enrichments (Krot et al. 2005; Krot et al. 2001; Greshake 2001 ;Briani et al. 2009; Bonal et al. 2010), likely of primitive origin (Olsen et al. 2013).Most primitive meteorites show evidence of having been compacted, fragmented, andmixed beneath their surfaces, wherein signs of primary accretion are obliterated. Evidencefor accretionary growth of planetesimals is absent in our collection of extraterrestrialmaterials because such structures would not survive the processes associated with planetarydifferentiation that gave rise to the iron meteorites and achondrites. However, Isheyevopreserves primary accretionary structures delineated by a well-sorted mixture of smallmetal grains, chondrules, CAIs, and clay-bearing matrix lumps (Ivanova et al. 2008). Thejuxtaposition of fine-grained clasts that experienced extensive aqueous alteration withmaterials that formed at high temperature shows that Isheyevo is a mechanical mixture ofdisparate materials.
2. Observations
As described in detail in Garvie et al. 2015, Isheyevo preserves primary accretionarystructures exemplified by prominent layering and lobe-like structures delineated by themetallic and nonmetallic components. The laminations are consistent with settling ofparticles that have not coagulated into aggregates. Layers richer in iron grains protrudedownward into layers richer in silicate grains indicating that these are sedimentary loadstructures akin to those in terrestrial aqueous deposits where dense sediment is depositedover, and protrudes into, a less dense layer (Allen 1984). The stratigraphic up position isthus established. Also evident are faults that disrupt the planarity of the laminations andcan be traced at high angles to the layering. These faults show necking and attenuation ofthe layers (Figures 1a, 1b), suggesting that the aggregate was weakly cohesive and behaved 5 –Fig. 1.—
Left : Photograph of a 10 x 14 cm slice from Isheyevo. Metal is white and the non-metalcomponents are dark. Evident is the fine laminations near the top of the specimen.
Center : False colorimage of the slice. Whites and yellows represent metal, and the darker colors correspond to the non-metalliccomponents. Layer A is fine-grained with weakly defined laminations and B shows several alternating metal-rich and silicate-rich layers. The base of layer C is silicate rich with lobe-like structures protruding intothe coarser-grained layer D. Layer E exhibits several silicate-rich layers and has a largely metal-rich basewith well-developed finger-like lobes protruding into layer F. The dashed white lines delineate the faultedsediment.
Right : Reflected-light image showing evidence of cold deformation of a metal particle. The imageshows silicates as a uniform dark color - the round structure on the left is a chondrule. The metal has beenstained with sodium bisulfite, which stains the metal producing colors that reveal the structure of the metal.The metal particle at the center of the image shows contorted bands of kamacite (stained light brown) andtetrataenite (bright), this contortion is evidence of low temperature deformation. macroscopically like soft sediment. Microscopic examination shows plastically deformedmetal grains (Figure 1c), which together with the clay clasts and chemically zoned metalspheres is evidence of post sedimentary, low-temperature deformation. These stratigraphicfeatures provide clues to the accretion and formation of the Isheyevo parent body.Terrestrial processes that result in sedimentation, such as declining velocity in a fluidflow do not apply in this case. Therefore, layering in Isheyevo reflects accumulation ofmaterial in an accreting environment. Such layering could occur as a result of high-velocityimpacts, but would result in disruption of the layers around larger impacting grains anddestruction of clay clasts upon impact, which is not seen. As such, we investigate other 6 –processes that result in gentle sedimentation and layering.The mixture of chondrules, having textures indicating rapid cooling (Campbell et al. 2005),and metal spheres, which require cooling over days to weeks, (e.g., Goldstein et al. 2007),is consistent with formation in an impact plume (Krot et al. 2005; Olsen et al. 2013). Theimplication is that chondrules and metal spheres were mixed with remnants of solid materialfrom the original impactor and then subsequently reaccreted by the surviving planetesimalon relatively short timescales. Reaccretion via gravitational settling is suggested (e.g.,Asphaug et al. 2011). However, we show that gravitational settling is unlikely, and proposeinstead that a fan-like sheet of ejecta from the impact was slowed by gas drag and overtakenby the surviving planetesimal at speeds that allowed gentle sedimentation.
3. Astrophysical Setting
Quantitative modeling of glancing blows between molten planetesimals shows that animpact plume, originating primarily from the impactor, can be produced downrange of thecollision (Asphaug et al. 2011). The model predicts the flow of the material and the size ofdroplets produced (Asphaug et al. 2011). Chondrule sizes in Isheyevo constrain the sizes ofthe impacting bodies to be in the tens of km range (Asphaug et al. 2011). While it hasbeen argued that some of the components in Isyehevo, such as the refractory inclusionsand porphyritic chondrules, may have been incorporated from material in the nebula(Krot et al. 2008; Krot et al. 2009), such a scenario is inconsistent with our understandingof Solary System evolution, unless they originated as accretionary breccias of impact debris(Krot & Bizzarro 2014). As such, we propose these particular components originate from asolid carapace ( ∼ ∼ ∼ V esc ∼
72 m s − (Case 5, Asphaug et al. 2011). At this rate, theleading edge of the sheet of material will reach a distance that is comparable to the Hillradius ( r H = ( M p / M (cid:12) ) / a ≈ . × km at 3 AU) in ∼ ∼ Size sorting of materials in the plume of ejecta occurs because components will travelvarying distances before stopping. The stopping time before spherical droplets of different 8 –Fig. 2.—
Left : Distance traveled by metal and silicate spheres of varying radii, ejected from a commonpoint at speed V = 30 m s − , before being stopped by gas of density ρ g = 10 − g cm − . Right : Same as onthe left, except the gas density is ρ g = 10 − g cm − . radii and density recouple to the gas is given by (Cuzzi et al. 2001): t s = ρ s a s c s ρ g , (1)where ρ s is the particle density, a s is the particle radius, c s is the sound speed at 150 K,and ρ g is the gas density. For silicates, we use ρ s = 3.3 g cm − (representative of forsterite)and ρ s = 7.2 g cm − for Fe-rich metals. We consider a range of droplet sizes of 10-300 µ m,typical of the sizes of chondrules and metal grains in chondrites. The distance traveled froma common point of origin by droplets of different radii and material density is given by l s ≈ V t s , where V is the initial velocity of the droplet. In gas densities typical of the solarnebula at 2-3 AU at ∼ ρ g = 10 − g cm − ; Desch & Connolly 2002, Morris & Desch2010; Desch et al. 2012), silicate spheres will travel ∼ ∼ l s on t s results in sortingbased on size and composition. Following the breakup of the impact plume, metal sphereswill travel farther than silicates of similar size before their motions are arrested. In thecase of lower gas density ( ρ g = 10 − g cm − ), typical of transition disks (Figure 2, Right),metal spherules will travel thousands of kilometers farther than similarly sized silicates.The difference in stopping times results in the aerodynamical sorting of the particles as 9 –shown in Figure 3.Fig. 3.— A snapshot showing the distribution of silicate and metal spheres of different radii, originatingfrom a common point, after becoming recoupled to the gas. Silicate spheres are shown in green and metalspheres are shown in blue. The Isheyevo parent body (indicated by the thin orange area) will sweep upparticles as it travels from right to left.
After silicate and metal particles are stopped and recouple to the nebular gas, they mustthen reaccrete onto the Isheyevo parent body; otherwise the components would disperseinto the nebula. Previously proposed scenarios for reaccretion have invoked gravitationalsettling (Krot et al. 2008; Asphaug et al. 2011), but we argue that this process is too slowto present a reasonable method for reaccretion.
The gravitational settling time is given by z/v t , where z is the distance from thetarget, and the terminal velocity v t = gt s , where g is the local gravity. In a nebula with ρ g = 10 − g cm − , the time for silicate particles of 10-300 µ m to settle from the Hill sphereof a planetesimal 70 km in diameter is 83 - 2500 yr. For similarly sized metal particles,the settling time is 38 - 1150 yr. Over such long time periods, particles would disperse 10 –through the nebula due to turbulence in the gas long before they are able to gravitationallysettle to the body. Even were it possible for the particles to settle before dispersion,layering would not occur. As a consequence of the long timescales involved for gravitationalsettling, particles would reach terminal velocity long before nearing the body, reaching thesurface simultaneously and erasing the effects of the size sorting. For higher gas densities,gravitational settling is even more improbable. For example, at ρ g = 10 − g cm − , the timefor particles to settle from the Hill sphere are two orders of magnitude higher than for ρ g =10 − g cm − . Therefore, reaccretion by gravitational settling is implausible. Our calculations and data from Isheyevo are consistent with the rotating impactedbody sweeping up material downrange of the collision, as it continues on with a slightvelocity relative to the gas and particles. This sweep-up scenario is compatible with thesize sorting of silicates and metals in Isheyevo. Our determination of the distribution ofparticles (Figure 3) predicts that chondrules will be swept up by the parent body withsmaller-sized metals. In general, we observe that this is the case within Isheyevo. Sizemeasurements from a representative piece of Isheyevo show that metals have a radius of 33 µ m (n=161) and silicates, including clay clasts, have a radius of 60 µ m (n=56). In addition,sweep-up at low velocity is necessary to preserve the clay-rich clasts.Our calculations suggest that sufficient mass can be swept up to produce meters-thicklayers of particles resembling those found in Isheyevo. The fraction of particles swept upby a parent body with diameter D ≈
70 km will be f = ( πD ) / (4 A ), where A is the areaof the fan-like sheet of material at the time it is reaccreted. If we assume the sheet ofmaterial stops moving as a unit after spreading to a distance r ∼ Hill radius, and assuminghomologous expansion of a 500 km x 500 km square at 3.3 hours, the sheet has area 11 – ≈ × km , so that f ∼ × − . Gravitational focusing can increase this by a factor ∼ ( v esc / V ) , where V ∼ − may reflect the random velocities of particles, yielding f ∼ × − . In gas of density ρ g = 10 − g cm − , and for an ejected mass 3 × g(Asphaug et al. 2011), the mass of solids reaccreted is ∼ × g. This mass of solids issufficient to coat the entire asteroid surface to a depth of about 1 meter, or cover a fractionof the asteroid surface to greater depth, depending on its rotation rate. These preliminarycalculations demonstrate that for the stopping lengths we consider typical, it is possible forthe parent body to sweep up sufficient mass to produce the CH/CB/Isheyevo chondrites,provided the asteroid moves in the same direction as the ejecta, the initial velocity of thesheet of material, V , is low ( ∼ V esc ), and the gas density is high enough to arrest themotions of the particles before they travel much farther than ∼ × km.
4. Astrophysical Implications
Reaccreation by sweep-up at low velocity, while retaining evidence of the aerodynamicsorting, places constraints on the the density of nebular gas and the degree of turbulence inthe nebula.
Zoned metal particles, such as those found in Isheyevo, are interpreted to haveformed and cooled in a matter of days to weeks, based on their chemical zoning profiles(Goldstein et al. 2007; Meibom et al. 2000; Petaev et al. 2001; Petaev & Jacobsen 2003;Campbell & Humayun 2004). This time constraint imposes bounds on the gas densityof the protoplanetary disk during their formation. Metal spheres formed in impacts anddispersed in gas at densities typical of the formation of most chondrules at around 2 Myr 12 –post CAIs ( ρ g = 10 − g cm − ) would be stopped and reaccreated on timescales too shortto allow for their condensation. In order to meet the constraints on cooling times formetal spheres, sweep-up must occur within days to weeks. According to our parameterstudies (Table 1), gas densities in the range of ρ g = 10 − to 10 − g cm − are indicated foraccreting bodies moving with a sweep-up velocity of v su = 25 - 500 m s − . Gas densitieslower than ρ g = 10 − g cm − would result in longer stopping times, causing dispersal ofthe impact products into the nebula, without reaccretion.In order to preserve the aerodynamic sorting effects of the particles, they must beswept up before they are mixed by turbulence. We have employed the methods describedby (Cuzzi & Zahnle 2004) to determine the mixing timescale for individual components,once they have recoupled to the gas. The effective viscosity in a weakly turbulent nebulais given by ν t = αc s H , where α is a dimensionless parameter determined by the massaccretion rate of the nebula, c s is the sound speed, and H is the scale height of the disk.Both models and observations suggest that typically α ∼ − to 10 − and H ∼ R/ R is the distance from the central star (Cuzzi & Zahnle 2004). The diffusivity due toturbulence is D = ν t /P r t , where P r t is the Prandtl number, typically assumed to be P r t =1, giving D = ν t (Cuzzi & Zahnle 2004). The timescale for mixing of particles separated bya distance L is then t mix = L /D .Given α = 10 − , and a maximum particle separation distance L max ∼ t mix ∼
100 hours. The largest components must be swept up within thistime frame in order to preserve the sorting depicted in Figure 3, which requires that theIsheyevo parent body must move at a velocity ≥
24 m s − relative to the particles. This isa lower limit to the relative velocity, V rel , we expect as a result of the collision. At minimumparticle separation of ∼
13 km, applicable to smaller particles, mixing can occur within onesecond. The average size of particles in Isheyevo indicate separation distances of L ∼ t mix ∼
70 hours, which requires that the body move at V rel ∼
28 m s − . An upperlimit to the relative velocity is provided by the preservation of the integrity, both thermallyand mechanically, of the clay-like clasts. It is therefore likely that the sweep-up velocity fallsat the lower limit of the range indicated by particle size. This relative velocity, combinedwith the constraints on cooling rates of metal particles, is consistent with gas densities of ρ g = 10 − to 10 − g cm − (Table 1). It is also important to note that our calculations of themixing timescale (combined with the size sorting observed in Isheyevo) place constraints onthe degree of turbulence in the disk, since for larger α , t mix is correspondingly shorter. Wefind that for α > − , preservation of sorting, such as that observed in Isheyevo, is unlikely.
5. Conclusion
Our calculations show that the components found in the Isheyevo meteoriteare consistent with sweep-up at low velocity within gas of density ρ g = 10 − to10 − g cm − . These densities are consistent with those of observed transition disks(Salyk et al. 2009; Williams & Cieza 2011). Through Isheyevo’s association with meteoritesthat have components dated at around 5 Myr (Krot et al. 2005; Ivanova et al. 2006;Bollard et al. 2013), we infer that this important stage in the evolution of the Solar Systemoccurred at ∼ α < − . We thereforeconclude that Isheyevo, the oldest known sedimentary rock, accreted in the Solar System’smildly turbulent transition disk, a heretofore purely theoretical phase in the primordialsolar nebula. 14 –
6. Acknowledgements
We wish to thank Bill Bottke and Jeff Cuzzi for their thoughtful comments andsuggestions. M.A.M. was supported by NASA Cosmochemistry grant NNX14AN58G.L.A.J.G was supported by NASA Origins of Solar System grant NNX11AK58G. 15 –Table 1. Results of Parameter Study on Sweep-up Time v su −
09 a − − − −
25 m s − < b − < − < − < − < a Gas density in units of g cm − . b Sweep-up in units of days. 16 –
REFERENCES
Allen, J. R. L. 1984, Sedimentary Structures, 663.Asphaug, E., Jutzi, M., & Movshovitz, N. 2011, Earth and Planetary Science Letters, 308,369.Boley, A. C., Morris, M. A., & Desch, S. J. 2013, ApJ, 776, 101.Bollard, J., Connelly, J. N., & Bizzarro, M 2013.