X-rays and their implications for
star formation, young stars and protoplanetary
Stars form in cold molecular clouds (T~10K) which collapse into protostars
(surface T~103 K) surrounded by protoplanetary disks (T~102).
Eric Feigelson and his colleagues were leaders in the unexpected discoveriy
that yount stellar systems also produce gas with T~106-108
K and energetic particles with MeV energies. Today, the phenomenology
of X-ray and radio emission from pre-main sequence stars is very rich. But
the fraction of energy entering kev and MeV processes is still very small,
so it is unclear whether they are incidental or important to the astrophysics
of star formation and early stellar evolution. Our achievements during
the 1980-90s include:
The early years
We had a number of notable achievements during the 1980-90s. We were
the first to: detect their variable X-ray emission (1981),
report X-ray-discovered weak-lined T Tauri (WTT) stars (1981),
encounter a powerful WTT radio continuum flare (1985),
uncover VLBI-scale redio structure in active WTT stars (1991),
organize a multiwavelength campaign on a flaring WTT star (1994
, detect X-rays from protostars (1996),
detect circularly polarized radio emission from a protostar (1997),
discover the nearest young stellar cluster found in a century using X-ray
Several Penn State graduate and undergraduate students participated in these
efforts, including first-authored papers by Pete Stine, Doug O'Neal, Jim
Leous, Lee Carkner and Eric Mamajek. These findings, and those of many other
researchers in the field, are brought together in a 1999 Annual Reviews
of Astronomy & Astrophysics article entitled High energy processes in
young stellar objects by Feigelson & Montmerle.
The Chandra years
In mid-1999, NASA Great Observatory for X-ray astronomy
was launched with the Advanced CCD Imaging Spectrometer developed by Penn
State and MIT. Evan Pugh Professor Gordon Garmire is PI of the ACIS
instrument, and Eric Feigelson serves on the ACIS Team specializing in issues
relating to young stars. A Chandra image of an active star formation
region will often show hundreds of X-ray emitting stars ranging from OB stars
to pre-main sequence brown dwarfs, from Class I protostars (age ~105
yrs) to post-T Tauri stars (age ~ 107 yrs). The X-ray populations
are comparable to those in the latest infrared images, and can penetrate
extraordinarily deeply into the cloud (AV ~ 500 mags).
Eric has particularly emphasized study of the Orion Nebula where ~2000
X-ray stars appear in a single field, including both the full initial mass
function of ~1 Myr stars and deeply embedded populations associated with
the BN/KL and other molecular cores (2000, 2002a,
2003) He is leading the Chandra Orion Ultradeep Project (COUP) which is analyzing and interpreting a 10-day near-continuous
Chandra ACIS observation made in January 2003. Kosta Getman led the
huge data analysis effort which uncovered 1616 X-ray sources (2004).
A Special Issue of the Astrophysical Journal Supplement devoted to the initial
COUP studies is planned for early 2005. He also studies older, dispersed
pre-main sequence stellar populations in collaboration with Warrick Lawson
of UNSW. Here are some of the issues addressed by these efforts:
- Origins of pre-main sequence
magnetic activity The enhanced magnetic activity of
T Tauri stars is largely consistent with stellar activity seen in the Sun
and other late-type stars where magnetic fields are generated in a dynamo
driven by the differential rotation in the stellar interior. However,
our Chandra data show that the expected link between activity and surface
rotation, which is the principal correlate of X-ray emission in older main
sequence stars, is completely absent in pre-main sequence stars (2003).
Instead, correlations between X-ray emission and stellar mass, size/volume
and bolometric luminosity are present. The reasons are not clear: either
the solar-type alpha-Omega magnetic dynamo is fully saturated, or a new
distributed turbulent dynamo (widelly predicted to be present in fully-convective
stars) is operative.
- Evolution of stellar magnetic
activity From two complementary long Chandra observations,
the COUP study of the Orion Nebula Cluster and the CDF-N study of a high-Galactic
latitude field, we have improved our knowledge of the decay of magnetic activity
in late-type stars (2005).
A monotonic decay is now traced from 106 to 1010
yrs, although the simple link between activity and rotational decay may not
be the only cause.
- X-ray effects on the circumstellar
environment Theoretical calculations
show that young stellar X-rays are likely to be the dominant source of ionization
in the vicinity of pre main sequence stars. See our two reviews on
this fascinating topic (2000, 2005). Ionization
of the circumstellar disk may have critical effects on disk dynamics (promoting
MHD turbulence and viscosity) and on coupling between the disk and bipolar
outflows. Less clearly, the X-ray Dissociation Regions around
embedded stars may be sufficiently extended to affect ambipolar diffuse
and future star formation in molecular clouds. The violent magnetic
reconnection events producing X-ray flares will also accelerate particles
to MeV energies. These may impact the disk, produce radioactive nuclides
in disk solids via spallation reactions, and thereby account for the anomalous
abundances of short-lived radionuclides in CAI inclusions of carbonaceous
chondritic meteorities which give direct information about condition in the
protoplanetary disk that gave rise to our solar system (2002a).
- The fate of OB stellar winds
For 25 years, theorists have discussed whether the
massive high-velocity winds from the highest mass stars would slam against
surrounding molecular material producing an X-ray emitting shock. The
issue is critical for understanding HII regions around young OB clusters,
as the standard theory of photoionized nebulae omits the effects of wind
kinetic energy. Our Chandra studies of high-mass star forming regions,
led by Leisa Townsley, have clearly discovered the predicted shock plasma.
We find that most of the HII region volume is filled with 10 MK gas,
but most of the wind energy and mass flows unimpeded away from the cloud
into the Galactic interstellar medium (2003).
These X-ray flows are small-scale analogies of the Galactic winds of starburst
galaxies. A major Chandra survey of high-mass
star forming regions across the Galaxy is underway.
- X-rays from Herbig-Haro jets
Collimated bipolar outflows are a ubiquitous feature
of protostars, probably arising from magnetic interactions between the accreting
star and the inner disk. Material is quickly accelerated to 100-500
km/s velocities, producing the well-known low-excitation emission line Herbig-Haro
objects, and then entrains surrounding materials to produce the slower bipolar
outflows seen in mm molecular lines. We have found weak X-ray emission
from the terminal shock of one HH object (2001)
and from shocks at the base of another HH object (2003).
- Discovering 10 Myr old stars
and disks While the earliest stages of
stellar evolution are now readily studied in molecular clouds, it has proved
difficult to establish quality samples of somewhat older pre-main sequence
stars that have dispersed into the Galaxy. X-ray surveys are
the most effective tool for finding these post-T Tauri stars.
Using the ROSAT and Chandra satellites, we found two nearby, poor but
compact groups of pre-main sequence stars: the eta Cha cluster (1999)
and the quintet around HD 104237 (2003b).
Both of these clusters, like the TW Hya Association and beta Pic moving
group, are kinematically consistent with membership in the large Sco-Cen
OB association (2000,
The existence of young stars dispersed over such a wide area is explained
by a model in which stars inherit velocities from supersonic turbulent gas
motions seen in giant molecular clouds (1996).
The eta Cha cluster is now the most complete sample of coeval 9+/-1 Myr stars
ranging from spectral types B8 to M5 (2001).
Infrared and optical study shows that 2/3 of them still possess infrared
excess disks and 1/3 are still accreting at a low level (2003a)
. We are now in the process of identifying other ~10 Myr-old stars for
studies of planet formation.