A Spitzer Survey of Dusty Disks in Scorpius-Centaurus Christine H. Chen (STScI) M. Bitner (STScI), E. Mamajek (Rochester), Marc Pecaut (Rochester), K. Su (Steward Observatory), A. Weinberger (Carnegie DTM)

Planet Formation in Our Solar System Once gas has dissipated, km-sized bodies agglomerate into oligarchs that stir small bodies Infrared observations of dust can help constrain disk properties during the period of oligarchic growth to determine average properties and magnitude of variation time Terrestrial Planet Formation Giant Planet Formation 1-3 Myr Myr Core Formation Envelope Accretion ~few Myr Myr Giant Impacts including Moon Formation and Late Patina Formation of km-sized bodies Oligarchic Growth

Oligarchic Growth Simulations Coagulation N-body simulations (Kenyon & Bromley 2004, 2005, 2008) Gas dissipates on a timescale ~10 Myr Parent bodies with sizes 1m-1km at AU in the disk Pluto-sized (1000 km) objects grow via collisions until gas dissipates They stir leftover planetesimals, which generates debris

Stellar Properties Impact Disk Evolution LuminosityMassWind Oligarch Formation Timescale Disk Thermal Radiation Dust Removal Mechanisms Kenyon & Bromley 2008 Chen et al. 2005

Corpuscular Stellar Wind Small grains can be “blown out” via transfer of momentum from corpuscular wind protons to circumstellar grains. In our Solar System, Q wind = 1, v = 400 km/sec, suggesting that radiation pressure is the dominant small grain removal mechanism around all but the lowest mass stars. Larger grains spiral inward because stellar wind produces a drag on circumstellar grains causing them to orbit with slower speeds. The increase in “drag” in the inward drift velocity, over that produced by Poynting-Robertson is given by the factor:

Scorpius-Centaurus OB Association Prebisch & Mamajek 2008 Nearest OB association from the Sun with typical stellar distances pc Contains 3 subgroups: –Upper Scorpius (5 Myr) –Upper Centaurus Lupus (15 Myr) –Lower Centaurus Crux (17 Myr) F- and G-type members have been identified based on Hipparcos common proper motion (de Zeeuw et al. 1999)

Observations Magellan MIKE high resolution (R~60,000) visual spectroscopy –Gas phase studies: H , CaII H and K/Na I D absorption –Stellar Activity Measurements: R HK, vsini Spitzer MIPS 24 and 70  m photometry –Identify excess candidates Follow-up Spitzer IRS low resolution (R~60) mid-infrared spectroscopy –Characterize grain composition, size, distance, mass Follow-up Spitzer SED mode (R~10) far-infrared observations –Characterize grain composition, size, distance, mass

MIPS 24  m Color-Color Diagram Stars with older ages possess more late spectral type stars with 24  m excess Stars with J-H < 0.02 possess significantly larger 24  m excesses than those of later spectral type 5 Myr 15 Myr 17 Myr Chen et al. 2010

FEPS Carpenter 09a US Carpenter 09b ScoCen Chen10 Total US 5 Myr 6/186/18 (33±15%) UCL 15 Myr 2/108/3910/49 (20±6%) LCC 17 Myr 7/1318/4425/57 (44±9%) ScoCen M  Stars: 24  m Disk Fraction M  stars possess a spectral type mid- to late- K-type at the age of US and F- to G-type at the ages of UCL and LCC

MIPS 70  m Color-Color Diagram Our MIPS 70  m observations were very shallow (1 cycle of 10 sec integrations) and were not sensitive to stellar photospheres We detect several objects with very bright 70  m excess 5 Myr 15 Myr 17 Myr Chen et al. 2010

Both Primordial and Debris Disks Present HD in LCC (17 Myr) MS 1.3" (170 AU) F3/F5 V Binary L IR /L * = Lisse et al Keller et al AK Sco in UCL (15 Myr) PMS Spectroscopic F5 Binary dM/dt = 9  M  yr -1

Chen et al MIPS 24  m Excess Evolution Our MIPS 24  m observations of F0-F5 stars are broadly consistent with the Kenyon & Bromley (2008) models, but do not indicate a peak in the upper envelope of 24  m excess at Myr The Carpenter et al. (2009) observation of US show that models must be updated to include dust within 30 AU around the late- type stars

HD : An Accreting Binary with a Weak Infrared Excess Time-variable, broad H  emission, consistent with an accretion rate ~10 -9 M  yr -1 MIPS 24  m and IRS excess (L IR /L * = ): if grains are large, then T gr = K. MIPS 70  m upper limit Bitner et al. 2010

MIPS 70  m Excess Evolution Kenyon & Bromley (2008) models predict that disks should be bright at 70  m Our MIPS 70  m observations of Sco-Cen are not sufficiently deep to test the self-stirred disk models at far- infrared wavelengths Chen et al. 2010

Spectral Type Dependence The probability that the F0-F5 stars and the F6-G5 stars possess the same distribution of K-[24] excess is 7.5% Chen et al. 2010

Why Do the Early and Late-Type Stars in Our Sample Behave Differently? Differences Anticipated by KB08 Models: Stellar Luminosity –F ex (24  m)/F * (24  m) Stellar Mass –Collisional Grinding Timescale Stellar Wind Drag: Stars with Spectral Type F0- F5 are evolving onto MS at Myr; their envelopes are changing from convective to radiative Late-type stars with winds would be expected to possess less dust

Stellar Mass Loss Rates and X-ray Activity Wood et al. (2005)

Dust Removal by Stellar Wind Drag? For HD , HD , and HD , M wind c 2/ L * = 76, 27, 15, and 77, suggesting that stellar wind drag is the dominant grain removal mechanism in these systems. At an age of 20 Myr, the Sun possessed M wind c 2/ L * = 460. Assuming stellar wind drag is larger than Poynting-Robertson drag, the lifetime of grains around a dusty star is For HD , the stellar wind drag life-time for the grains is ~12,000 years, significantly shorter than the estimate 16 Myr age of the system.

Stellar Activity Diagnostics: L x /L * Chen et al argued that a possible anti-correlation between L IR /L * and L x /L * argued for stellar wind drag as a possible dust removal mechanism ROSAT All Sky Survey is insensitive at the distance of Sco-Cen for inferring the presence of stellar winds <1000x that found in our solar system The probability that the x-ray “active” (ROSAT detected) and “inactive” stars are drawn from the same population is ~3.9%

Stellar Activity Diagnostics: R HK R HK measured from Magellan/MIKE R~60,000 visual spectra for (almost) all of the stars in our sample The probability that R HK “active” (stars with Ca II H and K core emission) and “inactive” stars are drawn from the same population is 0.15% R HK is typically used as an activity indicator for stars with spectral type later than F7V; therefore, anti-correlation may be tracing the effect of a parameter other than activity

Conclusions Disks within an OB association sub-group may possess a variety of evolutionary states The magnitude of the infrared excess detected is consistent with models of oligarchic growth; however, there is no convincing evidence for a “peak” in the infrared excess at Myr and disks around solar- mass stars are dustier than predicted The initial conditions assumed in the coagulation N-body models may be overly simplistic (e.g. no primordial grains, no parent bodies interior to 30 AU) The dust mass appears to be correlated with stellar mass with higher mass stars possessing more massive disks This finding is consistent with the assumptions that more massive stars (1) initially possess more massive disks, (2) are more luminous, and (3) do not possess strong stellar winds