Download to read offline
![THE ASTROPHYSICAL JOURNAL, 562 : 446È455, 2001 November 20
( 2001. The American Astronomical Society. All rights reserved. Printed in U.S.A.
X-RAYS AND PROTOSTARS IN THE TRIFID NEBULA
JEONGHEE RHO
Infrared Processing and Analysis Center, California Institute of Technology, MS 100-22, Pasadena, CA 91125 ; rho=ipac.caltech.edu
MICHAEL F. CORCORAN
NASA Goddard Space Flight Center, Code 662, Greenbelt, MD 20742
YOU-HUA CHU
Department of Astronomy, University of Illinois, 1002 West Green Street, Urbana, IL 61801
AND
WILLIAM T. REACH
Space Infrared Telescope Facility Center, California Institute of Technology, MS 100-22, Pasadena, CA 91125
Received 2001 February 19 ; accepted 2001 June 28
ABSTRACT
The TriÐd Nebula is a young H II region, recently rediscovered as a ““ pre-Orion ÏÏ star-forming region,
containing protostars undergoing violent mass ejections visible in optical jets as seen in images from the
Infrared Space Observatory and the Hubble Space T elescope. We report the Ðrst X-ray observations of
the TriÐd Nebula using ROSAT and ASCA. The ROSAT image shows a dozen X-ray sources, with the
brightest X-ray source being the O7 star, HD 164492, which provides most of the ionization in the
nebula. We also identify 85 T Tauri star and young, massive star candidates from near-infrared colors
using the JHK color-color diagram from the Two-Micron All-Sky Survey (2MASS). Ten X-ray sources
s
have counterpart near-infrared sources. The 2MASS stars and X-ray sources suggest there are potentially
numerous protostars in the young H II region of the TriÐd. ASCA moderate-resolution spectroscopy of
the brightest source shows hard emission up to 10 keV with a clearly detected Fe K line. The best model
Ðt is a two-temperature (T 1.2 ] 106 K and 39 ] 106 K) thermal model with additional warm absorb-
ing media. The hotter component has an unusually high temperature for either an O star or an H II
region ; a typical Galactic H II region could not be the primary source for such hot temperature plasma
and the Fe XXV line emission. We suggest that the hotter component originates in either the interaction
of the wind with another object (a companion star or a dense region of the nebula) or from Ñares from
deeply embedded young stars.
Subject headings : H II regions È infrared : stars È stars : formation È X-rays : individual (TriÐd Nebula)
1. INTRODUCTION
1995 ; Feigelson et al. 1993), Lupus (Krautter et al. 1997),
H II regions contain various types of X-rayÈemitting Ž
and Taurus-Auriga (Neuhuser et al. 1995 ; Wichmann et al.
sources, such as high- and low-mass stars, binaries, and 1997). Flarelike X-ray events have been detected from T
protostars. Single massive O and B stars emit soft X-rays Tauri stars (Kamata et al. 1997 ; Koyama et al. 1992). The
from shocks in their radiatively unstable, outwardly moving X-rayÈemitting sources in the Monoceros and Rosette
outer atmospheres. X-ray emission from late-type stars is molecular clouds are also mostly T Tauri and Herbig Ae/Be
attributed to magnetically heated stellar coronae and mag- stars and they typically show luminosities of L D
netically driven stellar Ñares. Protostars produce X-rays via X
1030È1032 ergs s~1 (Gregorio-Hetem et al. 1998). Recent
magnetic heating and perhaps from accretion or excretion Chandra observation of the Orion Nebula resolved a thou-
disks and/or interaction of stellar jets with the circumnebu- sand X-rayÈemitting PMS stars with a mass range of 0.05 to
lar gas. Massive binaries can produce X-ray emission from 50 M , and a combined infrared and X-ray study suggested
colliding stellar winds ; close low-mass binaries can produce _
that the X-ray luminosity depends on stellar mass, rotation-
X-rays via mass exchange and enhanced dynamo action. al history, and magnetic Ðeld (Garmire et al. 2000).
X-ray emission provides a sensitive tracer of these di†erent At a distance of 1.67 kpc (Lynds & OÏNeil 1985), the
stellar populations and, along with infrared colors, is one of TriÐd Nebula, M20, is one of the best-known astrophysical
the main probes for identifying protostars and preÈmain- objects and one of the prettiest : it glows brightly in red light
sequence (PMS) stars. and is trisected by obscuring dust lanes. At an age of
Studies of X-ray emission from H II/star-forming regions D 3 ] 105 yr, the TriÐd is one of the youngest known H II
have made signiÐcant advances because of ROSAT regions. Observations with the Infrared Space Observatory
(Casanova et al. 1995), ASCA (e.g., Koyama et al. 1996), (ISO) and the Hubble Space T elescope (HST ) show the
and, most recently, Chandra observations (Garmire et al. TriÐd as a dynamic, ““ pre-Orion ÏÏ star-forming region con-
2000). For example, in the o Oph core region, 70% of the taining young stars undergoing episodes of violent mass
near-infrared (NIR) sources associated with protostars and ejections, and protostars (like HH 399) losing mass and
molecular cores have X-ray counterparts (Casanova et al. energy to the nebula in optically bright jets (Cernicharo et
1995), and a large number of X-ray sources have been found al. 1998, hereafter CLC98 ; LeÑoch & Cernicharo 2000 ;
to be low-mass PMS stars in other star-forming regions Hester et al. 1999). The ionization of the nebular gas is
such as Orion (Alcala et al. 1996), Chamaeleon (Alcala et al. dominated by the O7.5 star HD 164492. HD 164492 is a
446](https://image.slidesharecdn.com/x-raysandprotostarsinthetrifidnebula-130128062932-phpapp01/75/X-rays-and-protostars_in_the_trifid_nebula-1-2048.jpg)
![X-RAYS AND PROTOSTARS IN THE TRIFID NEBULA 447
luminosity class V (Levato 1975) or III (Conti & Alschuler in Figure 1. This image reveals for the Ðrst time that the
1971) star with a bolometric luminosity of L D 0.5È1.6 TriÐd Nebula contains numerous X-ray sources. We have
] 1039 ergs s~1 and an X-ray luminosity of 6 bol 1032 ergs
] identiÐed X-ray sources in the PSPC image using the
s~1 (Chlebowski et al. 1989). The mass-loss rate of this O FTOOLS task SRCDETECT and estimated the count rates
star is M 2 ] 10~6 M yr~1 (Howarth & Prinja 1989)
0 and uncertainties. The detected point sources are presented,
_
and the wind terminal velocity is V 1580 km s~1 (Prinja, in order of increasing right ascension, in Table 1 and
=
Barlow, & Howarth 1990). marked in Figure 1. Table 1 lists the position, count rate,
X-ray observations of the TriÐd o†er a unique and p of detection for 10 sources detected at greater than 3
opportunity to study the inÑuence of a massive star on star p and two possible sources (sources 11 and 12) detected with
formation in an exceptionally young star-forming region. somewhat lower conÐdence. We deÐne here new X-ray
Initially, X-ray emission from the TriÐd was reported only sources detected in the ROSAT PSPC image, as ROSAT
from the O star in the Einstein IPC catalog (Chlebowski & X-ray source in the TriÐd (RXT). Since the count rate is very
Harnden 1989). We serendipitously discovered a complex of small except for the O star HD 164492, we estimated the
X-ray emission from the TriÐd Nebula in a PSPC obser- luminosity by assuming an absorption column density
vation of the nearby supernova remnant W28 (Rho et al. N 3 ] 1021cm~2 (A D 1.5m, see ° 4 for details), and a
H V
1995). Subsequently, we started an extensive investigation thermal spectrum with kT 1 keV. The X-ray emission of
of the X-ray emission from the TriÐd Nebula. This paper PMS stars is understood to be thermal emission from gas
presents the Ðrst detection of a dozen X-ray sources in the rapidly heated to a temperature of D1 keV by violent mag-
TriÐd Nebula, and we correlate these with protostar candi- netohydrodynamic reconnection events (Feigelson &
dates identiÐed using the Two-Micron All-Sky Survey Montmerle 1999). The correspondence between the PSPC
(2MASS) data. The ROSAT images show multiple point count rate and the X-ray unabsorbed Ñux is 1 ] 10~3
sources, including HD 164492 and several T Tauri stars, PSPC counts s~1 D2.95 ] 10~14 ergs s~1 cm~2. Using this
and the ASCA spectra show hard X-ray emission, including conversion, the luminosities of the X-ray sources are com-
detection of an Fe K line. We discuss identiÐcations of the puted and given in Table 1.
X-rayÈemitting sources and the origin of the unusually hard We have examined the source list database using
X-ray emission from the TriÐd. SIMBAD, identiÐed counterparts to the X-ray sources at
other wavelengths, and marked them in Figure 1. To visual-
2. X-RAY SOURCES IN THE TRIFID ize the correspondence of X-ray sources with either optical
2.1. X-Ray Observations or radio sources, we have plotted the X-ray contours over
The TriÐd Nebula was observed using the X-ray tele- an Ha image (F. Winkler, 2001, private communication) in
Ž
scope on ROSAT (Trumper 1993) with the Position Sensi- Figure 2. Twenty-four sources from the Guide Star Catalog
tive Proportional Counter (PSPC) as the imaging detector. (GSC) are visible in the Ha image in Figure 2. The brightest
The PSPC on-axis angular resolution is 25A (FWHM, at 1 X-ray point source, RXT8 in Table 1, corresponds to the O
keV), and the PSPC covers a 2¡ Ðeld of view in the 0.1È2.4 star HD 164492. The possible X-ray source RXT11 coin-
keV energy band. Two PSPC observations were analyzed cides with the B8 star HD 313596 (R.A. 18h02m35s and decl.
for this paper : rp900375 centered on HD 164492 and [22¡59@54A), and the possible X-ray source RXT12 with the
observed on 1993 September 8 for an exposure of 9365 s
(PI : S. Snowden), and rp500236 centered on the supernova
remnant W28 and observed on 1993 April 1 for an exposure
of 10,476 s (PI : R. Pisarski ; Rho et al. 1995).
We also performed an ASCA observation (PI : J. Rho ;
sequence number 26051000) toward the center of the TriÐd
Nebula. The observation took place on 1998 September 30
to October 2. ASCA (TI & H94) has two detector pairs : Gas
Imaging Spectrometers (GIS2 and GIS3) and Solid-State
Imaging Spectrometers (SIS0 and SIS1). The SIS covers an
energy band of 0.5È10 keV and the GIS 0.6È10 keV. The
on-axis angular resolution of the GIS and SIS is about 1@È2@.
Each GIS counter has a circular Ðeld of view of 35@ diameter
while the Ðeld of view of each SIS CCD is an 11@ square ;
thus, both GIS and SIS detectors sufficiently cover the
entire TriÐd Nebula. We Ðltered the data using a few cri-
teria such as cuto† rigidity (COR) and earth elevation
(based on Revison 2 processing). After Ðltering, the expo-
sure time was 57 ks for the GIS and 53.5 ks for the SIS. The
entire TriÐd Nebula region after background subtraction
has count rates of 0.050 ^ 0.001 counts s~1 for SIS0,
0.040 ^ 0.001 counts s~1 for SIS1, 0.030 ^ 0.008 counts s~1
for GIS2, 0.039 ^ 0.009 counts s~1 for GIS3, and
0.086 ^ 0.004 counts s~1 for the ROSAT PSPC, in their
respective energy bands (integrated over all channels).
FIG. 1.ÈPSPC X-ray image and detected sources (RXT) are marked
2.2. X-Ray Source IdentiÐcation with numbers. Count rates for each source are listed in Table 1. Known
The ROSAT PSPC image of the TriÐd Nebula is shown sources are also marked with the labels.](https://image.slidesharecdn.com/x-raysandprotostarsinthetrifidnebula-130128062932-phpapp01/75/X-rays-and-protostars_in_the_trifid_nebula-2-2048.jpg)
![448 RHO ET AL. Vol. 562
TABLE 1
X-RAY SOURCES DETECTED IN THE PSPC IMAGE OF THE TRIFID NEBULA
Name R.A. (2000) Decl. (2000) Count Rate Detection (p) log (L )
X
RXT1 . . . . . . . 18 02 52.8 [23 02 18.1 0.0014 ^ 0.0005 5 31.12
RXT2 . . . . . . . 18 02 39.3 [22 58 34.3 0.0011 ^ 0.0004 3.5 31.02
RXT3 . . . . . . . 18 02 41. [23 03 51.8 0.0014 ^ 0.0006 3.5 31.12
RXT4 . . . . . . . 18 02 36.0 [23 01 36.3 0.0019 ^ 0.0006 4 31.05
RXT5 . . . . . . . 18 02 35.0 [23 01 29.4 0.0021 ^ 0.0006 4.5 31.29
RXT6 . . . . . . . 18 02 27.9 [22 59 47.9 0.0016 ^ 0.0006 4 31.17
RXT7 . . . . . . . 18 02 25.4 [22 59 51.4 0.0010 ^ 0.0005 3 30.97
RXT8 . . . . . . . 18 02 23.35 [23 01 47.0 0.0131 ^ 0.0013 8 (see Table 4)
RXT9 . . . . . . . 18 02 21.1 [23 03 21.5 0.0010 ^ 0.0004 4 30.97
RXT10 . . . . . . 18 02 12.38 [22 55 37.0 0.0012 ^ 0.0004 4.5 31.05
RXT11 . . . . . . 18 02 34.92 [22 59 55.6 0.0006 ^ 0.0003 2.5 30.75
RXT12 . . . . . . 18 02 31.67 [23 02 25.6 0.0007 ^ 0.0003 2.5 30.82
NOTE.ÈUnits of right ascension are hours, minutes, and seconds, and units of declination are degrees, arcmin-
utes, and arcseconds.
optical star GSC–06842–00001 (R.A. 18h02m34s 8 and decl.
. sources showing NIR color excesses and present candidate
[23¡03@06A3). None of the other GSC stars coincide with
. protostars.
X-ray peaks. A radio source, GPSR5 6.980[0.286 (R.A.
18h02m28s 1 and decl. [23¡03@46A3 ; Becker et al. 1994), is
. . 3. NEAR-INFRARED SOURCES FROM 2MASS : YOUNG
close to the X-rayÈemitting area but does not have a corre- STELLAR OBJECTS
sponding X-ray peak. Four protostars (TC0, TC1, TC3, and We have identiÐed young stellar objects (YSOs) using the
TC4 sources1 in CLC98) have been reported in the TriÐd 2MASS data (Skrutskie et al. 1997). Using identical tele-
Nebula (CLC98), which are marked in Figure 1, but only scopes in the northern and southern hemispheres, 2MASS is
one is close to the X-ray peak at HD 164492. In the mapping the entire sky in the J (1.11È1.36 km), H (1.5È1.8
next section, we correlate the PSPC X-ray sources with km), and K (2È2.32 km) bands to a limiting point-source
s
sensitivity of approximately 16.5, 16.0, and 15.5 mag,
1 The source designation TC0, TC1, etc. was assigned by CLC98. Since respectively (Cutri et al. 2000).2 The data toward the TriÐd
this designation was already in use, for indexing purposes these sources Nebula were taken on 1998 June 14 using the southern
should be referred to as [CLC98] 0, etc. telescope, and most of these data were included in the
2MASS Second Incremental Release, but a small portion of
area was in the 2MASS Working Database because of large
photometric uncertainties at the time of the Incremental
Release. The photometry is typically better than 5% (Cutri
et al. 2000Èsee fn. 2).
We used the 2MASS point-source catalog to extract
sources within an 8@ radius centered on R.A. 18h02m30s and
decl. [23¡02@00A. We have selected sources with the follow-
ing criteria. First, we selected the sources that were detected
in all three J, H, and K bands. We then selected sources
s
with a signal-to-noise ratio greater than 10 (i.e., the J, H,
and K magnitudes are brighter than 15.8, 15.1, and 14.3
s
mag, respectively). These selections produced D1100 such
sources. We then accepted sources with photometric uncer-
tainties p 0.25 mag, for which the Ðt to the point-spread
function produced reduced s2 2. This last criterion
l
excluded blended sources that caused higher uncertainties
in the photometry. This is important in the Galactic plane
where NIR sources are crowded and the 2MASS has a
limited spatial resolution (3A5). This criterion removed inac-
.
curate blue points that appeared in the JHK color-color
s
diagram (as described below). The criteria we have used are
conservative for the magnitude limit and photometric
uncertainties of the 2MASS survey.
We plotted the sources in the JHK color-color diagram,
s
as shown in Figure 3a, in order to identify YSOs with infra-
FIG. 2.ÈPSPC X-ray contours superimposed on an Ha image of the red color excess. Figure 3b shows H[K versus K magni-
s s
TriÐd Nebula (courtesy of Frank Winkler). The optical image was tude for the selected sample. The interstellar reddening
obtained on 1994 July 4 (UT) from the Burrell Schmidt telescope of Case
Ž
Western Reserve University through a 25 A bandpass Ha Ðlter. Total
exposure time is 1800 s, and the scale is 2A0 pixel~1. The strongest X-ray
. 2 See R. M. Cutri et al. 2000, at http ://www.ipac.caltech.edu/2mass/
peak is at the O star, HD 164492. releases/ Ðrst/doc/explsup.html.](https://image.slidesharecdn.com/x-raysandprotostarsinthetrifidnebula-130128062932-phpapp01/75/X-rays-and-protostars_in_the_trifid_nebula-3-2048.jpg)
![No. 1, 2001 X-RAYS AND PROTOSTARS IN THE TRIFID NEBULA 449
FIG. 3b
FIG. 3a
FIG. 3c
FIG. 3.È(a) Near-infrared JHK color-color diagram : location in color-color diagram is used to determine T Tauri stars and massive YSOs. Normal stars
(dots), T Tauri stars (Ðlled circles), s
and massive YSOs (diamonds) are shown with di†erent symbols. The 2MASS counterparts of X-ray sources are marked
with crosses ; the sources above the extinction curve have high photometric uncertainties. They are 2MASS-red stars in Table 3 (see text for details).
Extinction vector is shown for 5 mag as the thick line labeled A . Thick curves are intrinsic colors of giant and dwarf stars. (b) Diagram of H [ K vs. K . T
v s s
Tauri stars and young massive protostars fall below (younger age) the PMS stars in this diagram. Symbols are the same as in (a). (c) T Tauri stars (circles) and
massive YSOs (diamonds), X-ray sources (crosses), and known embedded YSOs (triangles : from Cernicharo et al. 1998) are marked on optical image. (d)
2MASS three-color image (blue, green, and red for J, H, and K , respectively). Symbols are the same as in (c). The di†use, blue emission is probably Pab (in
s
the J band) from the H II region.
vector from Rieke & Lebofsky (1985) is also plotted. Adopt- the extinction curve (dashed line in Fig. 3a) and above the T
ing the intrinsic colors of giant and dwarf stars from Bessell Tauri star lines (solid line in Fig. 3a) are massive YSOs
& Brett (1988), we Ðnd the visual extinction toward this (Lada & Adams 1992) ; these stars are plotted as diamonds
direction is as high as A 30 mag for stars on the back in Figure 3a. In total, we found 41 T Tauri star candidates
V
side of the TriÐd Nebula. The observed JHK colors and 44 massive YSO candidates from the 2MASS sources
of YSOs can be explained by circumstellar disks models that are listed in Table 2. The T Tauri stars and YSOs are
(Lada & Adams 1992). We identiÐed the T Tauri stars with generally located along the ionization front. Figure 3c
(J[H) 0.58 ^ 0.11 ] (H[K ) ] 0.52 ^ 0.06 shows optical images, and T Tauri stars, massive YSOs, and
s and
(Meyer, Calvet, & Hillenbrand 1997) CTTS H[K [ 0.6 mag,
CTTS
X-ray sources are marked. Figure 3d shows a color compos-
where CTTS is classical T Tauri stars. These Ts Tauri stars ite of the J, H, and K images, with the J image in blue, H
fall between the two solid lines in Figure 3a. The stars below image in green, and Ks image in red. T Tauri stars, massive
s](https://image.slidesharecdn.com/x-raysandprotostarsinthetrifidnebula-130128062932-phpapp01/75/X-rays-and-protostars_in_the_trifid_nebula-4-2048.jpg)
![452 RHO ET AL. Vol. 562
TABLE 3
INFRARED COUNTERPARTS OF THE X-RAY SOURCES DETECTED IN PSPC IMAGES
2MASSa db Jd Hd Kd
s
RXT Counterparts (arcsec) R.A. (2000)c Decl. (2000)c (mag) (mag) (mag) Commentse
RXT1 . . . . . . . . YSO 10 18 02 52.4 [23 02 27.2 15.17 13.93 13.18 ...
RXT2 . . . . . . . . red 8 18 02 39.7 [22 58 29.0 17.72 15.06 13.50 J ““ Ðll ÏÏ
RXT3 . . . . . . . . red 5 18 02 41.4 [23 03 48.2 17.83 14.24 12.42 J ““ Ðll ÏÏ
red 13 18 02 40.4 [23 03 58.9 15.58 12.37 10.90 High psf *s2
RXT4 . . . . . . . . red 12.8 18 02 36.8 [23 01 43.9 15.85 12.78 11.34 High psf *s2
RXT5 . . . . . . . . NONE ... ... ... ... ... ... ...
RXT6 . . . . . . . . TTS 13 18 02 30.5 [22 59 28.7 12.71 11.65 10.91 ...
RXT7 . . . . . . . . YSO 12 18 02 26.6 [23 00 36.0 15.21 12.90 11.34 ...
RXT8f . . . . . . . Multiple ... ... ... ... ... ... ...
RXT9 . . . . . . . . red 6 18 02 21.5 [23 03 19.0 15.04 13.58 11.87 J ““ Ðll ÏÏ
RXT10 . . . . . . NONE ... ... ... ... ... ... ...
RXT11f . . . . . . red 4 18 02 34.7 [22 59 52.1 16.52 12.42 10.58 J ““ Ðll ÏÏ
RXT12 . . . . . . red 9 18 02 31.3 [23 02 18.2 15.12 13.17 11.36 J ““ Ðll ÏÏ
NOTE.ÈUnits of right ascension are hours, minutes, and seconds, and units of declination are degrees, arcminutes, and
arcseconds.
a Notes for the 2MASS sources : TTS T Tauri star ; YSO massive young stellar object ; red red 2MASS star.
b Projected distance between the X-ray and 2MASS sources.
c Coordinates of the 2MASS counterparts.
d J, H and K magnitudes of the 2MASS sources.
s
e J ““ Ðll ÏÏ means the J-band photometry is measured in band-Ðlled within the aperture, indicating their photometric
uncertainties are large. ““ High psf *s2 ÏÏ means the goodness of the Ðt of a point-spread function is high, indicating the sources are
either extended or unresolved double sources.
f RXT8 is HD 164492 (O7 star) and/or YSO (TC1 in CLC98), and RXT11 also coincides with HD 313596 (B8 star).
X-ray source in the TriÐd, and the shape of the spectrum of showed that the observed o†-source emission is dominated
the entire region in the ASCA data is not signiÐcantly di†er- by scattered emission from sources within the TriÐd
ent from the spectrum of this star alone because of the Nebula. The detection of hard emission from the TriÐd is
broad ASCA point-spread function. We made hard ([3 unusual because most single O stars have very little emis-
keV) and soft (3 keV) maps using the ASCA data, but no sion at energies above 2 keV and rarely show Fe K emission
obvious di†erence was noticeable at the spatial resolution (Corcoran et al. 1993). The only massive stars to show such
(1@) of ASCA. We also extracted an o†-source ASCA spec- hard X-ray spectra are binaries, either high-mass X-ray
trum in this direction, suspecting a contribution from the binaries (HMXBs) with collapsed companions or colliding
Galactic ridge emission. However, the o†-source spectrum wind binaries with noncollapsed (O or W-R star) compan-
ions that have strong stellar winds and signiÐcant colliding
wind X-ray emission.
We simultaneously Ðt the set of Ðve spectraÈthe
ROSAT /PSPC, ASCA/SIS0, SIS1, GIS2, and GIS3
spectraÈusing single- or two-temperature thermal models
(Mewe-Kaastra plasma model ; Kaastra 1992) with a single
absorbing column density N . The Ðts were unacceptable
H
(reduced s2 of 3). We next attempted a two-temperature
model. Following Corcoran et al.Ïs (1994) models of d Ori
and j Ori, we also included an additional ionized (““ warm ÏÏ)
absorbing medium, as representative of the photoionized
stellar wind material (Waldron 1984 ; Corcoran et al. 1994),
and allowed di†erent amounts of absorption for the hot and
cold components. The line-of-sight extinction value is
known toward this direction ; A 1.3È1.5 mag, i.e.,
V
E(B[V ) D 0.3È0.4 mag (Kohoutek, Mayer, & Lorenz 1999 ;
Lynds & OÏNeil 1985). Using E(B[V ) of 0.4 mag, we expect
an ISM N of D3 ] 1021 cm~2, with which we have Ðxed
H
the N value in our Ðt (also note that when we allow N to
H ] 1021 cm~2 falls within the errors). The model H
vary, 3
yielded an acceptable Ðt with kT D 0.14 keV (1.2 ] 106 K)
1
and N 5.9 ] 1021 cm~2, and a hotter component with
H,1 keV (3.9 ] 107 K) and N 2.7 ] 1021 cm~2,
kT D 3.3
with a line-of-sight ISM N 3 ] H,2 cm~2. The abun-
2
1021
H
dances are Ðxed at solar abundances. The Ðt results are
summarized in Table 4, along with the Fe K line character-
FIG. 4.ÈASCA SIS image (gray scale ranges 0.7È4.5 counts (6A4
. istics. The cold component arises from the O star atmo-
pixel~1) superposed on ROSAT contours. sphere, but the hot component might arise from a number](https://image.slidesharecdn.com/x-raysandprotostarsinthetrifidnebula-130128062932-phpapp01/75/X-rays-and-protostars_in_the_trifid_nebula-7-2048.jpg)
![No. 1, 2001 X-RAYS AND PROTOSTARS IN THE TRIFID NEBULA 453
of di†erent sources, such as unresolved interacting binaries,
active low-mass stars, or PMS stars.
5. THE NATURE OF THE HARD X-RAY COMPONENT
Our simultaneous ASCA/PSPC Ðts yield a total X-ray
luminosity of 1.9È2.5 ] 1034 ergs s~1 (0.3È10 keV) using the
two-temperature model shown in Table 4. If attributed to
HD 164492, then the ratio of X-ray and bolometric lumi-
nosities ( log L /L ) is between [5.0 and [4.5, which is
X bol
much higher than the typical ratios of log L /L D [7 for
X bol
a single O-type star (Chlebowski & Harnden 1989 ;
Ž
Berghofer, Schmitt, & Cassinelli 1996). The 3 keV X-ray
component is somewhat of a mystery since winds from
single O-type stars are not known to produce such high-
temperature emission. Typically the highest temperature
emission observed in O star X-ray spectra has kT 1 keV
(e.g., Corcoran et al. 1994).
FIG. 5a We discuss a few possibilities to explain the hot com-
ponent in the TriÐd Nebula. It may be that the hot com-
ponent arises in the interaction of the wind from the O7.5
star with another object (either a companion star or a dense
region of the nebula) outside the O star atmosphere. Collid-
ing winds between an early-type star and an early-type
companion (another O star or a Wolf-Rayet star) can
produce shock-heated material in the wind interaction
regions (Stevens, Blondin, & Pollock 1992), reaching tem-
peratures of 107[108 K, and emit X-rays. However, recent
photometric and spectroscopic studies of this region found
no evidence of a companion for HD 164492 (Kohoutek,
Mayer, & Lorenz 1999). Unless the collision occurs far from
the star (d [ 25@@), it will be unresolved to the ROSAT
PSPC. However, recent radio and NIR observations
toward the central region of the TriÐd detected three
sources close to the O star (Yusef-Zadeh et al. 2000), which
may be either stars or nebular knots photoionized by the
UV Ðeld of HD 164492.
FIG. 5b Hard emission could also arise in single O stars from
nonthermal emission produced by Fermi acceleration by
shocks in the O star wind (Chen & White 1991). The hard
component is suggested as a nonthermal tail produced by
inverse Comptonization of the photospheric UV Ðeld by a
population of fast particles accelerated by a distribution of
shocks. Although it is not possible to determine if the hard
emission is from a nonthermal tail or the two thermal tem-
perature components from goodness of the Ðt to the spectra,
the presence of the Fe K line suggests that the second com-
ponent is thermal. The sources of the two spectral com-
ponents need to be resolved spatially in order to determine
their origins. Enhanced hard X-ray emission might also be
produced by an oblique magnetic rotator, as is suspected in
another O7 star, h1 Orionis C, the central star of the Orion
Nebula (Gagne et al. 1997).
The X-ray properties of the core of the H II region W3
share a number of similarities to the TriÐd emission. For
FIG. 5c W3, the luminosity is a few 1033 ergs s~1 with a similarly
high temperature (Hofner & Churchwell 1997). In W3 (and
FIG. 5.È(a) GIS1 and GIS2 spectra of the TriÐd Nebula with its best Ðt possibly in the TriÐd), the high-temperature component
of a two-temperature thermal model with additional warm absorbing
media. Hard emission and Fe XXV line appear in the spectra. (b) SIS0 and may be produced by a hot, wind-shocked cavity that results
SIS1 and PSPC spectra with the best Ðts. (c) Each of two temperature when strong stellar winds interact with a surrounding dense
components (D1.2 ] 106 K and 3.9 ] 107 K) is marked on the SIS0 molecular cloud (e.g., Churchwell 1990). The presence of a
spectrum. known YSO (TC1 in CLC98) is consistent with the presence
of a molecular cloud in the TriÐd. In addition, IRAS obser-
vations of HD 164492 (van Buren, Noriega-Crespo, &
Dgani 1995) show a bow-shock structure around the star,](https://image.slidesharecdn.com/x-raysandprotostarsinthetrifidnebula-130128062932-phpapp01/75/X-rays-and-protostars_in_the_trifid_nebula-8-2048.jpg)
![454 RHO ET AL. Vol. 562
TABLE 4
SPECTRAL FIT RESULTS FROM SIMULTANEOUS FIT USING ASCA/PSPC
Parameters TriÐd
ISM N (1021 cm~2) . . . . . . . . . . . . . . . . . . . . . . . . . . 43
H
T ................................................ 3.3 ^ 0.6 keV ( 3.9 ] 107 K)
2
log EM (cm~3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55.8
2
Wind N , hot component (1021 cm~2) . . . . . . 2.7`2.3
H ~2.0
T ................................................ 0.14`0.06 keV ( 1.2 ] 106 K)
1 ~0.04
log EM (cm~3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57.4
1
Wind N , 1 component (1021 cm~2) . . . . . . . . 5.9`1.1
H ~0.9
Flux . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 ^ 0.5 ] 10~12
Unabsorbed Ñux . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 ^ 2 ] 10~11
Luminosity (ergs s~1) . . . . . . . . . . . . . . . . . . . . . . . . . 2.5 ] 1034 [d(kpc)/(1.67 kpc)]2
Fe K Line central energy (keV) . . . . . . . . . . . . . . 6.65`0.1
~0.2
Fe K equivalent width (keV) . . . . . . . . . . . . . . . . . 1.7 ^ 0.3
Fe K Ñux (photons s~1 cm~2) . . . . . . . . . . . . . . . 6( ^ 0.4) ] 10~4
perhaps indicative of a wind-cloud collision that might from unresolved emission from active PMS stars. With our
produce the high-temperature X-ray emission. The stellar current data, we cannot determine if one is more favored
wind outÑow at a speed of 1600 km s~1 should produce a than the other. To conclusively identify the hot component,
postshock temperature of D30 million degrees, though the a high-resolution image is needed to locate the emitting
observed temperature may be lower since radiative cooling object to determine whether the observed emission is pro-
is rapid. CLC98 suggested that the HCO` molecular clouds duced near the O star or whether a distributed group of
(likely the dust lanes) are fragmented shell around the active PMS stars dominates the observed emission. Along
nebula. In other words, the clouds and dust lanes we see in with the images, time-resolved spectra could allow us to
the optical image are located at the surface of the ionized distinguish whether the hard emission is Ñarelike (time
sphere. If this is the case, the 1600 km s~1 wind will be variable).
interacting with the lower density, ionized medium, and the
6. IDENTIFICATION OF X-RAY AND INFRARED SOURCES
shocked stellar wind can emit at high temperatures while
the photoionized materials of the edge of clouds could WITHIN THE TRIFID
supply sufficient density to emit strong X-rays. The detected X-ray sources and their counterparts are
The other possibility is that the hot component arises listed in Table 3. Most of X-ray sources are likely protostars
from deeply embedded young stars, especially since at least or PMS stars ; one source is a T Tauri star, two are massive
one embedded T Tauri star (TC1 in CLC98, which is protostars, and the others are unclassiÐed protostars. The
marked in Fig. 1) exists near the O star. The TC1 source in JHK color-color diagram using the 2MASS data suggests
s
CLC98 shows a large shift in the spectral energy distribu- that there are D80 protostars present in this nebula. It has
tion and violent ejections of high-velocity material (CLC98). been shown already (CLC98 ; LeÑoch & Cernicharo 2000)
Other ASCA observations showed bright X-ray sources that massive protostars (17È60 M ) are forming in the
_
with temperatures of 2È5 keV caused by Ñares of protostars TriÐd ; they are associated with molecular gas conden-
in the o Ophiuchi dark cloud (Koyama et al. 1992), the sations at the edges of clouds and their dynamical ages are
Orion Nebula (Yamauchi & Koyama 1993), and the R less than 104 yr. Whether low-mass protostars and T Tauri
Coronae Australis molecular cloud (Koyama et al. 1996). stars can be formed in a young (3 ] 105 yr) region such as
The hard emission was attributed to Ñares from individual the TriÐd is still an open question. Low-mass PMS stars of
PMS stars with typical X-ray luminosities in the range similarly young age were found in the Orion Nebula using
1030È1032 ergs s~1, and the peak luminosity of Ñares is new Chandra observations (Hillenbrand et al. 1998 ;
shown to be as large as 1033È1035 ergs s~1 and a tem- Garmire et al. 2000). The Chandra observation also showed
perature as high as 108 K (Feigelson & Montmerle 1999 ; the presence of young, low-mass (0.1È3 M ) PMS stars as
Grosso et al. 1997). The hard emission from W3 may be of a X-ray sources (Garmire et al. 2000). The_populations of
similar origin. Signs of active star formation in the TriÐd low-mass and massive protostars are similar in the TriÐd,
have been reported recently (LeÑoch et al. 2001) : there is a while we expect higher populations of low-mass protostars
dust cocoon or circumstellar disk around several members based on the initial mass function. This is consistent with
in the center of the TriÐd, and one young stellar source the fact that the TriÐd is a very young H II region ; T Tauri
shows a silicate feature in the circumstellar disk. Neither the stars have yet to form there. The distribution of T Tauri
o Ophiuchi dark cloud nor the R Coronae Australis molec- stars and massive YSOs is not obviously correlated with the
ular cloud is as bright in hard X-rays as the TriÐd Nebula, molecular cloud distribution. It is possible that they are
which may imply there are a higher number of protostars highly embedded in the molecular clouds, and their NIR
present in the TriÐd. A large number of protostars unre- colors cannot be fully obtained to identify protostars
solved to ASCA would dilute any Ñux variability produced because either J and/or H Ñux is unavailable. This is consis-
by Ñares. tent with the fact that 2MASS images show a higher popu-
In summary, one of two scenarios is likely responsible for lation of red stars in the southern part of the TriÐd. The
the hard emission : The emission may arise from HD 169942 HST images covering the southern part suggested the pres-
by the interaction of the wind from the O star with another ence of embedded stars at the head of the evaporating
object (a companion star or a dense region of the nebula) or globules (Hester et al. 1999). Deep NIR images and spec-](https://image.slidesharecdn.com/x-raysandprotostarsinthetrifidnebula-130128062932-phpapp01/75/X-rays-and-protostars_in_the_trifid_nebula-9-2048.jpg)
![No. 1, 2001 X-RAYS AND PROTOSTARS IN THE TRIFID NEBULA 455
troscopy will likely reveal hundreds of young protostars in similarly to 30 Dor, we would expect D1035 ergs s~1 di†use
the TriÐd as suggested by 2MASS and HST data. X-ray emission from the TriÐd. This is higher than the total
Whether di†use X-ray emission exists within the TriÐd X-ray luminosity of the TriÐd. It is likely that supernova
Nebula is currently unknown because of the limited spatial heating contributes signiÐcantly to the bright X-ray emis-
resolution of the PSPC images. For the unidentiÐed X-ray sion from 30 Dor. The TriÐd Nebula is too young to have
sources 5 and 10, we cannot determine whether they are a hosted supernova explosions, so its di†use X-ray emission
part of di†use emission or real point sources. They are very should be much fainter than that of 30 Dor.
likely normal stars, but the possibility that they are knots of Future high-resolution X-ray observations by new tele-
di†use emission cannot be ruled out. Di†use X-ray emission scopes such as Chandra and XMM should be able to
from H II regions has been detected, although it is rare. A resolve the O star from its immediate environment and dis-
few examples have been found, such as in the Carina crete X-ray sources, such as T Tauri stars, numerous pro-
Nebula (Seward & Chlebowski 1982), RCW 49 (Goldwurm, tostars, young and old normal stars, and to resolve the
Caraveo, & Bignami 1987 ; Belloni & Mereghetti 1994), and source of the high-temperature emission. A deep NIR image
the Cygnus Superbubble (Bochkarev & Sitnik 1985), and with other wavelength observations can identify the PMS
recently Wang (1999) reported di†use X-ray emission from stars and protostars and their mass populations. The TriÐd
the giant H II region 30 Dor in the Large Magellanic Cloud. Nebula is an exciting laboratory to understand the early
ROSAT and Broadband X-Ray T elescope observations of stage of star-forming activities in H II region.
the Carina Nebula show large-scale di†use emission over at
least 40@, as well as discrete X-ray sources and hot gas sur-
rounding g Car (Corcoran et al. 1995). The nature of di†use We thank Lynne Hillenbrand for helpful discussion on
emission is unclear in H II regions. Seward & Chlebowski near-infrared colors of protostars and protostar disk
(1982) suggested that stellar winds from the OB association models and for useful comments on the manuscript and
adequately heat the plasma. Wang (1999) suggested that the John Carpenter for helpful discussion on 2MASS data. We
X-ray thermal di†use emission arises in blister-shaped thank Frank Winkler for allowing us to reproduce his
region by loops of ionized gas and the structure is explained optical image. For J. R., this work is partially supported by
by the mass loading of the hot gas produced by the central NASA/ADP grant, NASA-1407. This publication makes
OB association. use of data products from the Two-Micron All-Sky Survey,
We compare the TriÐd Nebula with 30 Dor to determine which is a joint project of the University of Massachusetts
whether stellar winds in the TriÐd may produce observable and the Infrared Processing and Analysis Center, funded by
di†use X-rays. The stellar wind luminosity in 30 Dor is a the National Aeronautics and Space Administration and
few ] 1039 ergs s~1, and the X-ray luminosity of 30 Dor is the National Science Foundation. J. R. and W. T. R.
D1038 ergs s~1 (Wang 1999). The stellar wind luminosity of acknowledge the support of the Jet Propulsion Laboratory,
HD 146692 is L 1.7 ] 1036(M/2 ] 10~6 M yr~1)(V /
0 California Institute of Technology, which is operated under
w _
1580 km s~1)2 ergs s~1. If the TriÐd emits di†use X-rays contract with NASA.
REFERENCES
Alcala, J. M., et al. 1995, A&AS, 114, 109 Kaastra, J. S. 1992, An X-Ray Spectral Code for Optically Thin Plasmas,
ÈÈÈ. 1996, A&AS, 119, 7 SRON-Leiden Rept. (Leiden, Netherlands : SRON)
Becker, R. H., White, R. L., Helfand, D. J., & Zoonematkermani, S. 1994, Kamata, Y., Koyama, K., Tsuboi, Y., & Yamauchi, S. 1997, PASJ, 49, 461
ApJS, 91, 347 Kohoutek, L., Mayer, P., & Lorenz, R. 1999, A&AS, 134, 129
Belloni, R., & Mereghetti, S. 1994, A&A, 286, 935 Koyama, K., Asaoka, I., Kuriyama, T., & Tawara, Y. 1992, PASJ, 44, L255
Ž
Berghofer, T. W., Schmitt, J. H. M. M., & Cassinelli, J. P. 1996, A&AS, 118, Koyama, K., Ueno, S., Kobayashi, N., & Feigelson, E. D. 1996, PASJ, 48,
481 L87
Bessell, M. S., & Brett, J. M. 1988, PASP, 100, 1134 Krautter, J., et al. 1997, A&AS, 123, 329
Bochkarev, N. G., & Sitnik, T. G. 1985, Ap&SS, 108, 237 Lada, C. J., & Adams F. C. 1992, ApJ, 393, 278
Casanova, S., Montmerle, T., Feigelson, E. D., & Andre, P. 1995, ApJ, 439, LeÑoch, B., & Cernicharo, J. 2000, ApJ, 545, 340
752 LeÑoch, B., et al. 2001, A&A, 368, L13 (erratum 372, L65)
Cernicharo, J., et al. 1998, Science, 282, 462 (CLC98) Levato, A. 1975, A&AS, 19, 91
Chen, W., & White, R. L. 1991, ApJ, 366, 512 Lynds, B. T., & OÏNeil, E. J., Jr. 1985, ApJ, 294, 578
Chlebowski, T., Harnden, F. R., Jr., & Sciortino, S. 1989, ApJ, 341, 427 Meyer, M. R., Calvet, N., & Hillenbrand, L. 1997, AJ, 114, 288
Churchwell, E. 1990, A&A Rev., 2, 79 Ž
Neuhauser, R., Sterzik, M. F., Schmitt, J. H. M. M, Wichmann, R., &
Conti, P. S., & Alschuler, W. R. 1971, ApJ, 170, 325 Krautter, J. 1995, A&A, 297, 391
Corcoran, M. F., Swank, J., Rawley, G., Petre, R., Schmitt, J., & Day, C. Prinja, R. K., Barlow, M. J., & Howarth, I. D. 1990, ApJ, 361, 607
1995, Rev. Mexicana Astron. AstroÐs. Ser. de Conf., 2, 97 Ž
Rho, J.-H., Petre, R., Pisarski, R., & Jones, L. R. 1995, in Rontgenstrahlung
Corcoran, M., et al. 1993, ApJ, 412, 792 from the Universe, ed. H.-U. Zimmermann et al. (Garching : MPE), 273
ÈÈÈ. 1994, ApJ, 436, L95 Rieke, G. H., & Lebofsky, M. J. 1985, ApJ, 288, 618
Feigelson, E. D., Casanova, S., Montmerle, T., & Guibert, J. 1993, ApJ, Seward, F. D., & Chlebowski, T. 1982, ApJ, 256, 530
416, 623 Skrutskie, M. F., et al. 1997, in The Impact of Large-Scale Near-IR Sky
Feigelson, E. D., & Montmerle, T. 1999, ARA&A, 37, 363 Surveys, ed. F. Garzon, N. Epchtein, A. Omont, B. Burton, & P. Persi
Gagne, M., Caillault, J.-P., Stau†er, J. R., & Linsky, J. L. 1997, ApJ, 478, (Dordrecht : Kluwer), 25
L87 Stevens, I. R., Blondin, J. M., & Pollock, A. M. 1992, ApJ, 386, 265
Garmire, G., Feigelson, E. D., Broos, P., Hillenbrand, L. A., Pravdo, S. H., Tanaka, Y., Inoue, H., & Holt, S. S. 1994, PASJ, 46, L37 (TI of H94)
Townsley, L., & Tsuboi, Y. 2000, AJ, 120, 1426 Ž
Trumper, J. 1993, Science, 260, 1769
Goldwurm, A., Caraveo, P. A., & Bignami, G. F. 1987, ApJ, 322, 349 van Buren, D., Noriega-Crespo, A., & Dgani, R. 1995, AJ, 110, 2914
Gregorio-Hetem, J., Montmerle, T., Casanova, S., & Feigelson, E. D. 1998, Waldron, W. L. 1984, ApJ, 282, 256
A&A, 331, 193 Wang, Q. D. 1999, ApJ, 510, L139
Grosso, N., et al. 1997, Nature, 387, 56 Wichmann, R., Sterzik, M., Krautter, J., Metanomski, A., & Voges, W.
Hester, J. J., et al. 1999, BAAS, 194.681 1997, A&A, 326, 211
Hillenbrand, L. A., et al. 1998, AJ, 116, 1816 Yamauchi, S., & Koyama, K. 1993, ApJ, 405, 268
Hofner, P., & Churchwell, E. 1997, ApJ, 486, L39 Yusef-Zadeh, F., Shure, M., Wardle, M., & Kassim, N. 2000, ApJ, 540,
Howarth, I. D., & Prinja, R. K. 1989, ApJS, 69, 527 842](https://image.slidesharecdn.com/x-raysandprotostarsinthetrifidnebula-130128062932-phpapp01/75/X-rays-and-protostars_in_the_trifid_nebula-10-2048.jpg)
![THE ASTROPHYSICAL JOURNAL, 562 : 446È455, 2001 November 20
( 2001. The American Astronomical Society. All rights reserved. Printed in U.S.A.
X-RAYS AND PROTOSTARS IN THE TRIFID NEBULA
JEONGHEE RHO
Infrared Processing and Analysis Center, California Institute of Technology, MS 100-22, Pasadena, CA 91125 ; rho=ipac.caltech.edu
MICHAEL F. CORCORAN
NASA Goddard Space Flight Center, Code 662, Greenbelt, MD 20742
YOU-HUA CHU
Department of Astronomy, University of Illinois, 1002 West Green Street, Urbana, IL 61801
AND
WILLIAM T. REACH
Space Infrared Telescope Facility Center, California Institute of Technology, MS 100-22, Pasadena, CA 91125
Received 2001 February 19 ; accepted 2001 June 28
ABSTRACT
The TriÐd Nebula is a young H II region, recently rediscovered as a ““ pre-Orion ÏÏ star-forming region,
containing protostars undergoing violent mass ejections visible in optical jets as seen in images from the
Infrared Space Observatory and the Hubble Space T elescope. We report the Ðrst X-ray observations of
the TriÐd Nebula using ROSAT and ASCA. The ROSAT image shows a dozen X-ray sources, with the
brightest X-ray source being the O7 star, HD 164492, which provides most of the ionization in the
nebula. We also identify 85 T Tauri star and young, massive star candidates from near-infrared colors
using the JHK color-color diagram from the Two-Micron All-Sky Survey (2MASS). Ten X-ray sources
s
have counterpart near-infrared sources. The 2MASS stars and X-ray sources suggest there are potentially
numerous protostars in the young H II region of the TriÐd. ASCA moderate-resolution spectroscopy of
the brightest source shows hard emission up to 10 keV with a clearly detected Fe K line. The best model
Ðt is a two-temperature (T 1.2 ] 106 K and 39 ] 106 K) thermal model with additional warm absorb-
ing media. The hotter component has an unusually high temperature for either an O star or an H II
region ; a typical Galactic H II region could not be the primary source for such hot temperature plasma
and the Fe XXV line emission. We suggest that the hotter component originates in either the interaction
of the wind with another object (a companion star or a dense region of the nebula) or from Ñares from
deeply embedded young stars.
Subject headings : H II regions È infrared : stars È stars : formation È X-rays : individual (TriÐd Nebula)
1. INTRODUCTION
1995 ; Feigelson et al. 1993), Lupus (Krautter et al. 1997),
H II regions contain various types of X-rayÈemitting Ž
and Taurus-Auriga (Neuhuser et al. 1995 ; Wichmann et al.
sources, such as high- and low-mass stars, binaries, and 1997). Flarelike X-ray events have been detected from T
protostars. Single massive O and B stars emit soft X-rays Tauri stars (Kamata et al. 1997 ; Koyama et al. 1992). The
from shocks in their radiatively unstable, outwardly moving X-rayÈemitting sources in the Monoceros and Rosette
outer atmospheres. X-ray emission from late-type stars is molecular clouds are also mostly T Tauri and Herbig Ae/Be
attributed to magnetically heated stellar coronae and mag- stars and they typically show luminosities of L D
netically driven stellar Ñares. Protostars produce X-rays via X
1030È1032 ergs s~1 (Gregorio-Hetem et al. 1998). Recent
magnetic heating and perhaps from accretion or excretion Chandra observation of the Orion Nebula resolved a thou-
disks and/or interaction of stellar jets with the circumnebu- sand X-rayÈemitting PMS stars with a mass range of 0.05 to
lar gas. Massive binaries can produce X-ray emission from 50 M , and a combined infrared and X-ray study suggested
colliding stellar winds ; close low-mass binaries can produce _
that the X-ray luminosity depends on stellar mass, rotation-
X-rays via mass exchange and enhanced dynamo action. al history, and magnetic Ðeld (Garmire et al. 2000).
X-ray emission provides a sensitive tracer of these di†erent At a distance of 1.67 kpc (Lynds & OÏNeil 1985), the
stellar populations and, along with infrared colors, is one of TriÐd Nebula, M20, is one of the best-known astrophysical
the main probes for identifying protostars and preÈmain- objects and one of the prettiest : it glows brightly in red light
sequence (PMS) stars. and is trisected by obscuring dust lanes. At an age of
Studies of X-ray emission from H II/star-forming regions D 3 ] 105 yr, the TriÐd is one of the youngest known H II
have made signiÐcant advances because of ROSAT regions. Observations with the Infrared Space Observatory
(Casanova et al. 1995), ASCA (e.g., Koyama et al. 1996), (ISO) and the Hubble Space T elescope (HST ) show the
and, most recently, Chandra observations (Garmire et al. TriÐd as a dynamic, ““ pre-Orion ÏÏ star-forming region con-
2000). For example, in the o Oph core region, 70% of the taining young stars undergoing episodes of violent mass
near-infrared (NIR) sources associated with protostars and ejections, and protostars (like HH 399) losing mass and
molecular cores have X-ray counterparts (Casanova et al. energy to the nebula in optically bright jets (Cernicharo et
1995), and a large number of X-ray sources have been found al. 1998, hereafter CLC98 ; LeÑoch & Cernicharo 2000 ;
to be low-mass PMS stars in other star-forming regions Hester et al. 1999). The ionization of the nebular gas is
such as Orion (Alcala et al. 1996), Chamaeleon (Alcala et al. dominated by the O7.5 star HD 164492. HD 164492 is a
446](https://crownmelresort.com/image.slidesharecdn.com/x-raysandprotostarsinthetrifidnebula-130128062932-phpapp01/75/X-rays-and-protostars_in_the_trifid_nebula-1-2048.jpg)
![X-RAYS AND PROTOSTARS IN THE TRIFID NEBULA 447
luminosity class V (Levato 1975) or III (Conti & Alschuler in Figure 1. This image reveals for the Ðrst time that the
1971) star with a bolometric luminosity of L D 0.5È1.6 TriÐd Nebula contains numerous X-ray sources. We have
] 1039 ergs s~1 and an X-ray luminosity of 6 bol 1032 ergs
] identiÐed X-ray sources in the PSPC image using the
s~1 (Chlebowski et al. 1989). The mass-loss rate of this O FTOOLS task SRCDETECT and estimated the count rates
star is M 2 ] 10~6 M yr~1 (Howarth & Prinja 1989)
0 and uncertainties. The detected point sources are presented,
_
and the wind terminal velocity is V 1580 km s~1 (Prinja, in order of increasing right ascension, in Table 1 and
=
Barlow, & Howarth 1990). marked in Figure 1. Table 1 lists the position, count rate,
X-ray observations of the TriÐd o†er a unique and p of detection for 10 sources detected at greater than 3
opportunity to study the inÑuence of a massive star on star p and two possible sources (sources 11 and 12) detected with
formation in an exceptionally young star-forming region. somewhat lower conÐdence. We deÐne here new X-ray
Initially, X-ray emission from the TriÐd was reported only sources detected in the ROSAT PSPC image, as ROSAT
from the O star in the Einstein IPC catalog (Chlebowski & X-ray source in the TriÐd (RXT). Since the count rate is very
Harnden 1989). We serendipitously discovered a complex of small except for the O star HD 164492, we estimated the
X-ray emission from the TriÐd Nebula in a PSPC obser- luminosity by assuming an absorption column density
vation of the nearby supernova remnant W28 (Rho et al. N 3 ] 1021cm~2 (A D 1.5m, see ° 4 for details), and a
H V
1995). Subsequently, we started an extensive investigation thermal spectrum with kT 1 keV. The X-ray emission of
of the X-ray emission from the TriÐd Nebula. This paper PMS stars is understood to be thermal emission from gas
presents the Ðrst detection of a dozen X-ray sources in the rapidly heated to a temperature of D1 keV by violent mag-
TriÐd Nebula, and we correlate these with protostar candi- netohydrodynamic reconnection events (Feigelson &
dates identiÐed using the Two-Micron All-Sky Survey Montmerle 1999). The correspondence between the PSPC
(2MASS) data. The ROSAT images show multiple point count rate and the X-ray unabsorbed Ñux is 1 ] 10~3
sources, including HD 164492 and several T Tauri stars, PSPC counts s~1 D2.95 ] 10~14 ergs s~1 cm~2. Using this
and the ASCA spectra show hard X-ray emission, including conversion, the luminosities of the X-ray sources are com-
detection of an Fe K line. We discuss identiÐcations of the puted and given in Table 1.
X-rayÈemitting sources and the origin of the unusually hard We have examined the source list database using
X-ray emission from the TriÐd. SIMBAD, identiÐed counterparts to the X-ray sources at
other wavelengths, and marked them in Figure 1. To visual-
2. X-RAY SOURCES IN THE TRIFID ize the correspondence of X-ray sources with either optical
2.1. X-Ray Observations or radio sources, we have plotted the X-ray contours over
The TriÐd Nebula was observed using the X-ray tele- an Ha image (F. Winkler, 2001, private communication) in
Ž
scope on ROSAT (Trumper 1993) with the Position Sensi- Figure 2. Twenty-four sources from the Guide Star Catalog
tive Proportional Counter (PSPC) as the imaging detector. (GSC) are visible in the Ha image in Figure 2. The brightest
The PSPC on-axis angular resolution is 25A (FWHM, at 1 X-ray point source, RXT8 in Table 1, corresponds to the O
keV), and the PSPC covers a 2¡ Ðeld of view in the 0.1È2.4 star HD 164492. The possible X-ray source RXT11 coin-
keV energy band. Two PSPC observations were analyzed cides with the B8 star HD 313596 (R.A. 18h02m35s and decl.
for this paper : rp900375 centered on HD 164492 and [22¡59@54A), and the possible X-ray source RXT12 with the
observed on 1993 September 8 for an exposure of 9365 s
(PI : S. Snowden), and rp500236 centered on the supernova
remnant W28 and observed on 1993 April 1 for an exposure
of 10,476 s (PI : R. Pisarski ; Rho et al. 1995).
We also performed an ASCA observation (PI : J. Rho ;
sequence number 26051000) toward the center of the TriÐd
Nebula. The observation took place on 1998 September 30
to October 2. ASCA (TI & H94) has two detector pairs : Gas
Imaging Spectrometers (GIS2 and GIS3) and Solid-State
Imaging Spectrometers (SIS0 and SIS1). The SIS covers an
energy band of 0.5È10 keV and the GIS 0.6È10 keV. The
on-axis angular resolution of the GIS and SIS is about 1@È2@.
Each GIS counter has a circular Ðeld of view of 35@ diameter
while the Ðeld of view of each SIS CCD is an 11@ square ;
thus, both GIS and SIS detectors sufficiently cover the
entire TriÐd Nebula. We Ðltered the data using a few cri-
teria such as cuto† rigidity (COR) and earth elevation
(based on Revison 2 processing). After Ðltering, the expo-
sure time was 57 ks for the GIS and 53.5 ks for the SIS. The
entire TriÐd Nebula region after background subtraction
has count rates of 0.050 ^ 0.001 counts s~1 for SIS0,
0.040 ^ 0.001 counts s~1 for SIS1, 0.030 ^ 0.008 counts s~1
for GIS2, 0.039 ^ 0.009 counts s~1 for GIS3, and
0.086 ^ 0.004 counts s~1 for the ROSAT PSPC, in their
respective energy bands (integrated over all channels).
FIG. 1.ÈPSPC X-ray image and detected sources (RXT) are marked
2.2. X-Ray Source IdentiÐcation with numbers. Count rates for each source are listed in Table 1. Known
The ROSAT PSPC image of the TriÐd Nebula is shown sources are also marked with the labels.](https://crownmelresort.com/image.slidesharecdn.com/x-raysandprotostarsinthetrifidnebula-130128062932-phpapp01/75/X-rays-and-protostars_in_the_trifid_nebula-2-2048.jpg)
![448 RHO ET AL. Vol. 562
TABLE 1
X-RAY SOURCES DETECTED IN THE PSPC IMAGE OF THE TRIFID NEBULA
Name R.A. (2000) Decl. (2000) Count Rate Detection (p) log (L )
X
RXT1 . . . . . . . 18 02 52.8 [23 02 18.1 0.0014 ^ 0.0005 5 31.12
RXT2 . . . . . . . 18 02 39.3 [22 58 34.3 0.0011 ^ 0.0004 3.5 31.02
RXT3 . . . . . . . 18 02 41. [23 03 51.8 0.0014 ^ 0.0006 3.5 31.12
RXT4 . . . . . . . 18 02 36.0 [23 01 36.3 0.0019 ^ 0.0006 4 31.05
RXT5 . . . . . . . 18 02 35.0 [23 01 29.4 0.0021 ^ 0.0006 4.5 31.29
RXT6 . . . . . . . 18 02 27.9 [22 59 47.9 0.0016 ^ 0.0006 4 31.17
RXT7 . . . . . . . 18 02 25.4 [22 59 51.4 0.0010 ^ 0.0005 3 30.97
RXT8 . . . . . . . 18 02 23.35 [23 01 47.0 0.0131 ^ 0.0013 8 (see Table 4)
RXT9 . . . . . . . 18 02 21.1 [23 03 21.5 0.0010 ^ 0.0004 4 30.97
RXT10 . . . . . . 18 02 12.38 [22 55 37.0 0.0012 ^ 0.0004 4.5 31.05
RXT11 . . . . . . 18 02 34.92 [22 59 55.6 0.0006 ^ 0.0003 2.5 30.75
RXT12 . . . . . . 18 02 31.67 [23 02 25.6 0.0007 ^ 0.0003 2.5 30.82
NOTE.ÈUnits of right ascension are hours, minutes, and seconds, and units of declination are degrees, arcmin-
utes, and arcseconds.
optical star GSC–06842–00001 (R.A. 18h02m34s 8 and decl.
. sources showing NIR color excesses and present candidate
[23¡03@06A3). None of the other GSC stars coincide with
. protostars.
X-ray peaks. A radio source, GPSR5 6.980[0.286 (R.A.
18h02m28s 1 and decl. [23¡03@46A3 ; Becker et al. 1994), is
. . 3. NEAR-INFRARED SOURCES FROM 2MASS : YOUNG
close to the X-rayÈemitting area but does not have a corre- STELLAR OBJECTS
sponding X-ray peak. Four protostars (TC0, TC1, TC3, and We have identiÐed young stellar objects (YSOs) using the
TC4 sources1 in CLC98) have been reported in the TriÐd 2MASS data (Skrutskie et al. 1997). Using identical tele-
Nebula (CLC98), which are marked in Figure 1, but only scopes in the northern and southern hemispheres, 2MASS is
one is close to the X-ray peak at HD 164492. In the mapping the entire sky in the J (1.11È1.36 km), H (1.5È1.8
next section, we correlate the PSPC X-ray sources with km), and K (2È2.32 km) bands to a limiting point-source
s
sensitivity of approximately 16.5, 16.0, and 15.5 mag,
1 The source designation TC0, TC1, etc. was assigned by CLC98. Since respectively (Cutri et al. 2000).2 The data toward the TriÐd
this designation was already in use, for indexing purposes these sources Nebula were taken on 1998 June 14 using the southern
should be referred to as [CLC98] 0, etc. telescope, and most of these data were included in the
2MASS Second Incremental Release, but a small portion of
area was in the 2MASS Working Database because of large
photometric uncertainties at the time of the Incremental
Release. The photometry is typically better than 5% (Cutri
et al. 2000Èsee fn. 2).
We used the 2MASS point-source catalog to extract
sources within an 8@ radius centered on R.A. 18h02m30s and
decl. [23¡02@00A. We have selected sources with the follow-
ing criteria. First, we selected the sources that were detected
in all three J, H, and K bands. We then selected sources
s
with a signal-to-noise ratio greater than 10 (i.e., the J, H,
and K magnitudes are brighter than 15.8, 15.1, and 14.3
s
mag, respectively). These selections produced D1100 such
sources. We then accepted sources with photometric uncer-
tainties p 0.25 mag, for which the Ðt to the point-spread
function produced reduced s2 2. This last criterion
l
excluded blended sources that caused higher uncertainties
in the photometry. This is important in the Galactic plane
where NIR sources are crowded and the 2MASS has a
limited spatial resolution (3A5). This criterion removed inac-
.
curate blue points that appeared in the JHK color-color
s
diagram (as described below). The criteria we have used are
conservative for the magnitude limit and photometric
uncertainties of the 2MASS survey.
We plotted the sources in the JHK color-color diagram,
s
as shown in Figure 3a, in order to identify YSOs with infra-
FIG. 2.ÈPSPC X-ray contours superimposed on an Ha image of the red color excess. Figure 3b shows H[K versus K magni-
s s
TriÐd Nebula (courtesy of Frank Winkler). The optical image was tude for the selected sample. The interstellar reddening
obtained on 1994 July 4 (UT) from the Burrell Schmidt telescope of Case
Ž
Western Reserve University through a 25 A bandpass Ha Ðlter. Total
exposure time is 1800 s, and the scale is 2A0 pixel~1. The strongest X-ray
. 2 See R. M. Cutri et al. 2000, at http ://www.ipac.caltech.edu/2mass/
peak is at the O star, HD 164492. releases/ Ðrst/doc/explsup.html.](https://crownmelresort.com/image.slidesharecdn.com/x-raysandprotostarsinthetrifidnebula-130128062932-phpapp01/75/X-rays-and-protostars_in_the_trifid_nebula-3-2048.jpg)
![No. 1, 2001 X-RAYS AND PROTOSTARS IN THE TRIFID NEBULA 449
FIG. 3b
FIG. 3a
FIG. 3c
FIG. 3.È(a) Near-infrared JHK color-color diagram : location in color-color diagram is used to determine T Tauri stars and massive YSOs. Normal stars
(dots), T Tauri stars (Ðlled circles), s
and massive YSOs (diamonds) are shown with di†erent symbols. The 2MASS counterparts of X-ray sources are marked
with crosses ; the sources above the extinction curve have high photometric uncertainties. They are 2MASS-red stars in Table 3 (see text for details).
Extinction vector is shown for 5 mag as the thick line labeled A . Thick curves are intrinsic colors of giant and dwarf stars. (b) Diagram of H [ K vs. K . T
v s s
Tauri stars and young massive protostars fall below (younger age) the PMS stars in this diagram. Symbols are the same as in (a). (c) T Tauri stars (circles) and
massive YSOs (diamonds), X-ray sources (crosses), and known embedded YSOs (triangles : from Cernicharo et al. 1998) are marked on optical image. (d)
2MASS three-color image (blue, green, and red for J, H, and K , respectively). Symbols are the same as in (c). The di†use, blue emission is probably Pab (in
s
the J band) from the H II region.
vector from Rieke & Lebofsky (1985) is also plotted. Adopt- the extinction curve (dashed line in Fig. 3a) and above the T
ing the intrinsic colors of giant and dwarf stars from Bessell Tauri star lines (solid line in Fig. 3a) are massive YSOs
& Brett (1988), we Ðnd the visual extinction toward this (Lada & Adams 1992) ; these stars are plotted as diamonds
direction is as high as A 30 mag for stars on the back in Figure 3a. In total, we found 41 T Tauri star candidates
V
side of the TriÐd Nebula. The observed JHK colors and 44 massive YSO candidates from the 2MASS sources
of YSOs can be explained by circumstellar disks models that are listed in Table 2. The T Tauri stars and YSOs are
(Lada & Adams 1992). We identiÐed the T Tauri stars with generally located along the ionization front. Figure 3c
(J[H) 0.58 ^ 0.11 ] (H[K ) ] 0.52 ^ 0.06 shows optical images, and T Tauri stars, massive YSOs, and
s and
(Meyer, Calvet, & Hillenbrand 1997) CTTS H[K [ 0.6 mag,
CTTS
X-ray sources are marked. Figure 3d shows a color compos-
where CTTS is classical T Tauri stars. These Ts Tauri stars ite of the J, H, and K images, with the J image in blue, H
fall between the two solid lines in Figure 3a. The stars below image in green, and Ks image in red. T Tauri stars, massive
s](https://crownmelresort.com/image.slidesharecdn.com/x-raysandprotostarsinthetrifidnebula-130128062932-phpapp01/75/X-rays-and-protostars_in_the_trifid_nebula-4-2048.jpg)
![452 RHO ET AL. Vol. 562
TABLE 3
INFRARED COUNTERPARTS OF THE X-RAY SOURCES DETECTED IN PSPC IMAGES
2MASSa db Jd Hd Kd
s
RXT Counterparts (arcsec) R.A. (2000)c Decl. (2000)c (mag) (mag) (mag) Commentse
RXT1 . . . . . . . . YSO 10 18 02 52.4 [23 02 27.2 15.17 13.93 13.18 ...
RXT2 . . . . . . . . red 8 18 02 39.7 [22 58 29.0 17.72 15.06 13.50 J ““ Ðll ÏÏ
RXT3 . . . . . . . . red 5 18 02 41.4 [23 03 48.2 17.83 14.24 12.42 J ““ Ðll ÏÏ
red 13 18 02 40.4 [23 03 58.9 15.58 12.37 10.90 High psf *s2
RXT4 . . . . . . . . red 12.8 18 02 36.8 [23 01 43.9 15.85 12.78 11.34 High psf *s2
RXT5 . . . . . . . . NONE ... ... ... ... ... ... ...
RXT6 . . . . . . . . TTS 13 18 02 30.5 [22 59 28.7 12.71 11.65 10.91 ...
RXT7 . . . . . . . . YSO 12 18 02 26.6 [23 00 36.0 15.21 12.90 11.34 ...
RXT8f . . . . . . . Multiple ... ... ... ... ... ... ...
RXT9 . . . . . . . . red 6 18 02 21.5 [23 03 19.0 15.04 13.58 11.87 J ““ Ðll ÏÏ
RXT10 . . . . . . NONE ... ... ... ... ... ... ...
RXT11f . . . . . . red 4 18 02 34.7 [22 59 52.1 16.52 12.42 10.58 J ““ Ðll ÏÏ
RXT12 . . . . . . red 9 18 02 31.3 [23 02 18.2 15.12 13.17 11.36 J ““ Ðll ÏÏ
NOTE.ÈUnits of right ascension are hours, minutes, and seconds, and units of declination are degrees, arcminutes, and
arcseconds.
a Notes for the 2MASS sources : TTS T Tauri star ; YSO massive young stellar object ; red red 2MASS star.
b Projected distance between the X-ray and 2MASS sources.
c Coordinates of the 2MASS counterparts.
d J, H and K magnitudes of the 2MASS sources.
s
e J ““ Ðll ÏÏ means the J-band photometry is measured in band-Ðlled within the aperture, indicating their photometric
uncertainties are large. ““ High psf *s2 ÏÏ means the goodness of the Ðt of a point-spread function is high, indicating the sources are
either extended or unresolved double sources.
f RXT8 is HD 164492 (O7 star) and/or YSO (TC1 in CLC98), and RXT11 also coincides with HD 313596 (B8 star).
X-ray source in the TriÐd, and the shape of the spectrum of showed that the observed o†-source emission is dominated
the entire region in the ASCA data is not signiÐcantly di†er- by scattered emission from sources within the TriÐd
ent from the spectrum of this star alone because of the Nebula. The detection of hard emission from the TriÐd is
broad ASCA point-spread function. We made hard ([3 unusual because most single O stars have very little emis-
keV) and soft (3 keV) maps using the ASCA data, but no sion at energies above 2 keV and rarely show Fe K emission
obvious di†erence was noticeable at the spatial resolution (Corcoran et al. 1993). The only massive stars to show such
(1@) of ASCA. We also extracted an o†-source ASCA spec- hard X-ray spectra are binaries, either high-mass X-ray
trum in this direction, suspecting a contribution from the binaries (HMXBs) with collapsed companions or colliding
Galactic ridge emission. However, the o†-source spectrum wind binaries with noncollapsed (O or W-R star) compan-
ions that have strong stellar winds and signiÐcant colliding
wind X-ray emission.
We simultaneously Ðt the set of Ðve spectraÈthe
ROSAT /PSPC, ASCA/SIS0, SIS1, GIS2, and GIS3
spectraÈusing single- or two-temperature thermal models
(Mewe-Kaastra plasma model ; Kaastra 1992) with a single
absorbing column density N . The Ðts were unacceptable
H
(reduced s2 of 3). We next attempted a two-temperature
model. Following Corcoran et al.Ïs (1994) models of d Ori
and j Ori, we also included an additional ionized (““ warm ÏÏ)
absorbing medium, as representative of the photoionized
stellar wind material (Waldron 1984 ; Corcoran et al. 1994),
and allowed di†erent amounts of absorption for the hot and
cold components. The line-of-sight extinction value is
known toward this direction ; A 1.3È1.5 mag, i.e.,
V
E(B[V ) D 0.3È0.4 mag (Kohoutek, Mayer, & Lorenz 1999 ;
Lynds & OÏNeil 1985). Using E(B[V ) of 0.4 mag, we expect
an ISM N of D3 ] 1021 cm~2, with which we have Ðxed
H
the N value in our Ðt (also note that when we allow N to
H ] 1021 cm~2 falls within the errors). The model H
vary, 3
yielded an acceptable Ðt with kT D 0.14 keV (1.2 ] 106 K)
1
and N 5.9 ] 1021 cm~2, and a hotter component with
H,1 keV (3.9 ] 107 K) and N 2.7 ] 1021 cm~2,
kT D 3.3
with a line-of-sight ISM N 3 ] H,2 cm~2. The abun-
2
1021
H
dances are Ðxed at solar abundances. The Ðt results are
summarized in Table 4, along with the Fe K line character-
FIG. 4.ÈASCA SIS image (gray scale ranges 0.7È4.5 counts (6A4
. istics. The cold component arises from the O star atmo-
pixel~1) superposed on ROSAT contours. sphere, but the hot component might arise from a number](https://crownmelresort.com/image.slidesharecdn.com/x-raysandprotostarsinthetrifidnebula-130128062932-phpapp01/75/X-rays-and-protostars_in_the_trifid_nebula-7-2048.jpg)
![No. 1, 2001 X-RAYS AND PROTOSTARS IN THE TRIFID NEBULA 453
of di†erent sources, such as unresolved interacting binaries,
active low-mass stars, or PMS stars.
5. THE NATURE OF THE HARD X-RAY COMPONENT
Our simultaneous ASCA/PSPC Ðts yield a total X-ray
luminosity of 1.9È2.5 ] 1034 ergs s~1 (0.3È10 keV) using the
two-temperature model shown in Table 4. If attributed to
HD 164492, then the ratio of X-ray and bolometric lumi-
nosities ( log L /L ) is between [5.0 and [4.5, which is
X bol
much higher than the typical ratios of log L /L D [7 for
X bol
a single O-type star (Chlebowski & Harnden 1989 ;
Ž
Berghofer, Schmitt, & Cassinelli 1996). The 3 keV X-ray
component is somewhat of a mystery since winds from
single O-type stars are not known to produce such high-
temperature emission. Typically the highest temperature
emission observed in O star X-ray spectra has kT 1 keV
(e.g., Corcoran et al. 1994).
FIG. 5a We discuss a few possibilities to explain the hot com-
ponent in the TriÐd Nebula. It may be that the hot com-
ponent arises in the interaction of the wind from the O7.5
star with another object (either a companion star or a dense
region of the nebula) outside the O star atmosphere. Collid-
ing winds between an early-type star and an early-type
companion (another O star or a Wolf-Rayet star) can
produce shock-heated material in the wind interaction
regions (Stevens, Blondin, & Pollock 1992), reaching tem-
peratures of 107[108 K, and emit X-rays. However, recent
photometric and spectroscopic studies of this region found
no evidence of a companion for HD 164492 (Kohoutek,
Mayer, & Lorenz 1999). Unless the collision occurs far from
the star (d [ 25@@), it will be unresolved to the ROSAT
PSPC. However, recent radio and NIR observations
toward the central region of the TriÐd detected three
sources close to the O star (Yusef-Zadeh et al. 2000), which
may be either stars or nebular knots photoionized by the
UV Ðeld of HD 164492.
FIG. 5b Hard emission could also arise in single O stars from
nonthermal emission produced by Fermi acceleration by
shocks in the O star wind (Chen & White 1991). The hard
component is suggested as a nonthermal tail produced by
inverse Comptonization of the photospheric UV Ðeld by a
population of fast particles accelerated by a distribution of
shocks. Although it is not possible to determine if the hard
emission is from a nonthermal tail or the two thermal tem-
perature components from goodness of the Ðt to the spectra,
the presence of the Fe K line suggests that the second com-
ponent is thermal. The sources of the two spectral com-
ponents need to be resolved spatially in order to determine
their origins. Enhanced hard X-ray emission might also be
produced by an oblique magnetic rotator, as is suspected in
another O7 star, h1 Orionis C, the central star of the Orion
Nebula (Gagne et al. 1997).
The X-ray properties of the core of the H II region W3
share a number of similarities to the TriÐd emission. For
FIG. 5c W3, the luminosity is a few 1033 ergs s~1 with a similarly
high temperature (Hofner & Churchwell 1997). In W3 (and
FIG. 5.È(a) GIS1 and GIS2 spectra of the TriÐd Nebula with its best Ðt possibly in the TriÐd), the high-temperature component
of a two-temperature thermal model with additional warm absorbing
media. Hard emission and Fe XXV line appear in the spectra. (b) SIS0 and may be produced by a hot, wind-shocked cavity that results
SIS1 and PSPC spectra with the best Ðts. (c) Each of two temperature when strong stellar winds interact with a surrounding dense
components (D1.2 ] 106 K and 3.9 ] 107 K) is marked on the SIS0 molecular cloud (e.g., Churchwell 1990). The presence of a
spectrum. known YSO (TC1 in CLC98) is consistent with the presence
of a molecular cloud in the TriÐd. In addition, IRAS obser-
vations of HD 164492 (van Buren, Noriega-Crespo, &
Dgani 1995) show a bow-shock structure around the star,](https://crownmelresort.com/image.slidesharecdn.com/x-raysandprotostarsinthetrifidnebula-130128062932-phpapp01/75/X-rays-and-protostars_in_the_trifid_nebula-8-2048.jpg)
![454 RHO ET AL. Vol. 562
TABLE 4
SPECTRAL FIT RESULTS FROM SIMULTANEOUS FIT USING ASCA/PSPC
Parameters TriÐd
ISM N (1021 cm~2) . . . . . . . . . . . . . . . . . . . . . . . . . . 43
H
T ................................................ 3.3 ^ 0.6 keV ( 3.9 ] 107 K)
2
log EM (cm~3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55.8
2
Wind N , hot component (1021 cm~2) . . . . . . 2.7`2.3
H ~2.0
T ................................................ 0.14`0.06 keV ( 1.2 ] 106 K)
1 ~0.04
log EM (cm~3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57.4
1
Wind N , 1 component (1021 cm~2) . . . . . . . . 5.9`1.1
H ~0.9
Flux . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 ^ 0.5 ] 10~12
Unabsorbed Ñux . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 ^ 2 ] 10~11
Luminosity (ergs s~1) . . . . . . . . . . . . . . . . . . . . . . . . . 2.5 ] 1034 [d(kpc)/(1.67 kpc)]2
Fe K Line central energy (keV) . . . . . . . . . . . . . . 6.65`0.1
~0.2
Fe K equivalent width (keV) . . . . . . . . . . . . . . . . . 1.7 ^ 0.3
Fe K Ñux (photons s~1 cm~2) . . . . . . . . . . . . . . . 6( ^ 0.4) ] 10~4
perhaps indicative of a wind-cloud collision that might from unresolved emission from active PMS stars. With our
produce the high-temperature X-ray emission. The stellar current data, we cannot determine if one is more favored
wind outÑow at a speed of 1600 km s~1 should produce a than the other. To conclusively identify the hot component,
postshock temperature of D30 million degrees, though the a high-resolution image is needed to locate the emitting
observed temperature may be lower since radiative cooling object to determine whether the observed emission is pro-
is rapid. CLC98 suggested that the HCO` molecular clouds duced near the O star or whether a distributed group of
(likely the dust lanes) are fragmented shell around the active PMS stars dominates the observed emission. Along
nebula. In other words, the clouds and dust lanes we see in with the images, time-resolved spectra could allow us to
the optical image are located at the surface of the ionized distinguish whether the hard emission is Ñarelike (time
sphere. If this is the case, the 1600 km s~1 wind will be variable).
interacting with the lower density, ionized medium, and the
6. IDENTIFICATION OF X-RAY AND INFRARED SOURCES
shocked stellar wind can emit at high temperatures while
the photoionized materials of the edge of clouds could WITHIN THE TRIFID
supply sufficient density to emit strong X-rays. The detected X-ray sources and their counterparts are
The other possibility is that the hot component arises listed in Table 3. Most of X-ray sources are likely protostars
from deeply embedded young stars, especially since at least or PMS stars ; one source is a T Tauri star, two are massive
one embedded T Tauri star (TC1 in CLC98, which is protostars, and the others are unclassiÐed protostars. The
marked in Fig. 1) exists near the O star. The TC1 source in JHK color-color diagram using the 2MASS data suggests
s
CLC98 shows a large shift in the spectral energy distribu- that there are D80 protostars present in this nebula. It has
tion and violent ejections of high-velocity material (CLC98). been shown already (CLC98 ; LeÑoch & Cernicharo 2000)
Other ASCA observations showed bright X-ray sources that massive protostars (17È60 M ) are forming in the
_
with temperatures of 2È5 keV caused by Ñares of protostars TriÐd ; they are associated with molecular gas conden-
in the o Ophiuchi dark cloud (Koyama et al. 1992), the sations at the edges of clouds and their dynamical ages are
Orion Nebula (Yamauchi & Koyama 1993), and the R less than 104 yr. Whether low-mass protostars and T Tauri
Coronae Australis molecular cloud (Koyama et al. 1996). stars can be formed in a young (3 ] 105 yr) region such as
The hard emission was attributed to Ñares from individual the TriÐd is still an open question. Low-mass PMS stars of
PMS stars with typical X-ray luminosities in the range similarly young age were found in the Orion Nebula using
1030È1032 ergs s~1, and the peak luminosity of Ñares is new Chandra observations (Hillenbrand et al. 1998 ;
shown to be as large as 1033È1035 ergs s~1 and a tem- Garmire et al. 2000). The Chandra observation also showed
perature as high as 108 K (Feigelson & Montmerle 1999 ; the presence of young, low-mass (0.1È3 M ) PMS stars as
Grosso et al. 1997). The hard emission from W3 may be of a X-ray sources (Garmire et al. 2000). The_populations of
similar origin. Signs of active star formation in the TriÐd low-mass and massive protostars are similar in the TriÐd,
have been reported recently (LeÑoch et al. 2001) : there is a while we expect higher populations of low-mass protostars
dust cocoon or circumstellar disk around several members based on the initial mass function. This is consistent with
in the center of the TriÐd, and one young stellar source the fact that the TriÐd is a very young H II region ; T Tauri
shows a silicate feature in the circumstellar disk. Neither the stars have yet to form there. The distribution of T Tauri
o Ophiuchi dark cloud nor the R Coronae Australis molec- stars and massive YSOs is not obviously correlated with the
ular cloud is as bright in hard X-rays as the TriÐd Nebula, molecular cloud distribution. It is possible that they are
which may imply there are a higher number of protostars highly embedded in the molecular clouds, and their NIR
present in the TriÐd. A large number of protostars unre- colors cannot be fully obtained to identify protostars
solved to ASCA would dilute any Ñux variability produced because either J and/or H Ñux is unavailable. This is consis-
by Ñares. tent with the fact that 2MASS images show a higher popu-
In summary, one of two scenarios is likely responsible for lation of red stars in the southern part of the TriÐd. The
the hard emission : The emission may arise from HD 169942 HST images covering the southern part suggested the pres-
by the interaction of the wind from the O star with another ence of embedded stars at the head of the evaporating
object (a companion star or a dense region of the nebula) or globules (Hester et al. 1999). Deep NIR images and spec-](https://crownmelresort.com/image.slidesharecdn.com/x-raysandprotostarsinthetrifidnebula-130128062932-phpapp01/75/X-rays-and-protostars_in_the_trifid_nebula-9-2048.jpg)
![No. 1, 2001 X-RAYS AND PROTOSTARS IN THE TRIFID NEBULA 455
troscopy will likely reveal hundreds of young protostars in similarly to 30 Dor, we would expect D1035 ergs s~1 di†use
the TriÐd as suggested by 2MASS and HST data. X-ray emission from the TriÐd. This is higher than the total
Whether di†use X-ray emission exists within the TriÐd X-ray luminosity of the TriÐd. It is likely that supernova
Nebula is currently unknown because of the limited spatial heating contributes signiÐcantly to the bright X-ray emis-
resolution of the PSPC images. For the unidentiÐed X-ray sion from 30 Dor. The TriÐd Nebula is too young to have
sources 5 and 10, we cannot determine whether they are a hosted supernova explosions, so its di†use X-ray emission
part of di†use emission or real point sources. They are very should be much fainter than that of 30 Dor.
likely normal stars, but the possibility that they are knots of Future high-resolution X-ray observations by new tele-
di†use emission cannot be ruled out. Di†use X-ray emission scopes such as Chandra and XMM should be able to
from H II regions has been detected, although it is rare. A resolve the O star from its immediate environment and dis-
few examples have been found, such as in the Carina crete X-ray sources, such as T Tauri stars, numerous pro-
Nebula (Seward & Chlebowski 1982), RCW 49 (Goldwurm, tostars, young and old normal stars, and to resolve the
Caraveo, & Bignami 1987 ; Belloni & Mereghetti 1994), and source of the high-temperature emission. A deep NIR image
the Cygnus Superbubble (Bochkarev & Sitnik 1985), and with other wavelength observations can identify the PMS
recently Wang (1999) reported di†use X-ray emission from stars and protostars and their mass populations. The TriÐd
the giant H II region 30 Dor in the Large Magellanic Cloud. Nebula is an exciting laboratory to understand the early
ROSAT and Broadband X-Ray T elescope observations of stage of star-forming activities in H II region.
the Carina Nebula show large-scale di†use emission over at
least 40@, as well as discrete X-ray sources and hot gas sur-
rounding g Car (Corcoran et al. 1995). The nature of di†use We thank Lynne Hillenbrand for helpful discussion on
emission is unclear in H II regions. Seward & Chlebowski near-infrared colors of protostars and protostar disk
(1982) suggested that stellar winds from the OB association models and for useful comments on the manuscript and
adequately heat the plasma. Wang (1999) suggested that the John Carpenter for helpful discussion on 2MASS data. We
X-ray thermal di†use emission arises in blister-shaped thank Frank Winkler for allowing us to reproduce his
region by loops of ionized gas and the structure is explained optical image. For J. R., this work is partially supported by
by the mass loading of the hot gas produced by the central NASA/ADP grant, NASA-1407. This publication makes
OB association. use of data products from the Two-Micron All-Sky Survey,
We compare the TriÐd Nebula with 30 Dor to determine which is a joint project of the University of Massachusetts
whether stellar winds in the TriÐd may produce observable and the Infrared Processing and Analysis Center, funded by
di†use X-rays. The stellar wind luminosity in 30 Dor is a the National Aeronautics and Space Administration and
few ] 1039 ergs s~1, and the X-ray luminosity of 30 Dor is the National Science Foundation. J. R. and W. T. R.
D1038 ergs s~1 (Wang 1999). The stellar wind luminosity of acknowledge the support of the Jet Propulsion Laboratory,
HD 146692 is L 1.7 ] 1036(M/2 ] 10~6 M yr~1)(V /
0 California Institute of Technology, which is operated under
w _
1580 km s~1)2 ergs s~1. If the TriÐd emits di†use X-rays contract with NASA.
REFERENCES
Alcala, J. M., et al. 1995, A&AS, 114, 109 Kaastra, J. S. 1992, An X-Ray Spectral Code for Optically Thin Plasmas,
ÈÈÈ. 1996, A&AS, 119, 7 SRON-Leiden Rept. (Leiden, Netherlands : SRON)
Becker, R. H., White, R. L., Helfand, D. J., & Zoonematkermani, S. 1994, Kamata, Y., Koyama, K., Tsuboi, Y., & Yamauchi, S. 1997, PASJ, 49, 461
ApJS, 91, 347 Kohoutek, L., Mayer, P., & Lorenz, R. 1999, A&AS, 134, 129
Belloni, R., & Mereghetti, S. 1994, A&A, 286, 935 Koyama, K., Asaoka, I., Kuriyama, T., & Tawara, Y. 1992, PASJ, 44, L255
Ž
Berghofer, T. W., Schmitt, J. H. M. M., & Cassinelli, J. P. 1996, A&AS, 118, Koyama, K., Ueno, S., Kobayashi, N., & Feigelson, E. D. 1996, PASJ, 48,
481 L87
Bessell, M. S., & Brett, J. M. 1988, PASP, 100, 1134 Krautter, J., et al. 1997, A&AS, 123, 329
Bochkarev, N. G., & Sitnik, T. G. 1985, Ap&SS, 108, 237 Lada, C. J., & Adams F. C. 1992, ApJ, 393, 278
Casanova, S., Montmerle, T., Feigelson, E. D., & Andre, P. 1995, ApJ, 439, LeÑoch, B., & Cernicharo, J. 2000, ApJ, 545, 340
752 LeÑoch, B., et al. 2001, A&A, 368, L13 (erratum 372, L65)
Cernicharo, J., et al. 1998, Science, 282, 462 (CLC98) Levato, A. 1975, A&AS, 19, 91
Chen, W., & White, R. L. 1991, ApJ, 366, 512 Lynds, B. T., & OÏNeil, E. J., Jr. 1985, ApJ, 294, 578
Chlebowski, T., Harnden, F. R., Jr., & Sciortino, S. 1989, ApJ, 341, 427 Meyer, M. R., Calvet, N., & Hillenbrand, L. 1997, AJ, 114, 288
Churchwell, E. 1990, A&A Rev., 2, 79 Ž
Neuhauser, R., Sterzik, M. F., Schmitt, J. H. M. M, Wichmann, R., &
Conti, P. S., & Alschuler, W. R. 1971, ApJ, 170, 325 Krautter, J. 1995, A&A, 297, 391
Corcoran, M. F., Swank, J., Rawley, G., Petre, R., Schmitt, J., & Day, C. Prinja, R. K., Barlow, M. J., & Howarth, I. D. 1990, ApJ, 361, 607
1995, Rev. Mexicana Astron. AstroÐs. Ser. de Conf., 2, 97 Ž
Rho, J.-H., Petre, R., Pisarski, R., & Jones, L. R. 1995, in Rontgenstrahlung
Corcoran, M., et al. 1993, ApJ, 412, 792 from the Universe, ed. H.-U. Zimmermann et al. (Garching : MPE), 273
ÈÈÈ. 1994, ApJ, 436, L95 Rieke, G. H., & Lebofsky, M. J. 1985, ApJ, 288, 618
Feigelson, E. D., Casanova, S., Montmerle, T., & Guibert, J. 1993, ApJ, Seward, F. D., & Chlebowski, T. 1982, ApJ, 256, 530
416, 623 Skrutskie, M. F., et al. 1997, in The Impact of Large-Scale Near-IR Sky
Feigelson, E. D., & Montmerle, T. 1999, ARA&A, 37, 363 Surveys, ed. F. Garzon, N. Epchtein, A. Omont, B. Burton, & P. Persi
Gagne, M., Caillault, J.-P., Stau†er, J. R., & Linsky, J. L. 1997, ApJ, 478, (Dordrecht : Kluwer), 25
L87 Stevens, I. R., Blondin, J. M., & Pollock, A. M. 1992, ApJ, 386, 265
Garmire, G., Feigelson, E. D., Broos, P., Hillenbrand, L. A., Pravdo, S. H., Tanaka, Y., Inoue, H., & Holt, S. S. 1994, PASJ, 46, L37 (TI of H94)
Townsley, L., & Tsuboi, Y. 2000, AJ, 120, 1426 Ž
Trumper, J. 1993, Science, 260, 1769
Goldwurm, A., Caraveo, P. A., & Bignami, G. F. 1987, ApJ, 322, 349 van Buren, D., Noriega-Crespo, A., & Dgani, R. 1995, AJ, 110, 2914
Gregorio-Hetem, J., Montmerle, T., Casanova, S., & Feigelson, E. D. 1998, Waldron, W. L. 1984, ApJ, 282, 256
A&A, 331, 193 Wang, Q. D. 1999, ApJ, 510, L139
Grosso, N., et al. 1997, Nature, 387, 56 Wichmann, R., Sterzik, M., Krautter, J., Metanomski, A., & Voges, W.
Hester, J. J., et al. 1999, BAAS, 194.681 1997, A&A, 326, 211
Hillenbrand, L. A., et al. 1998, AJ, 116, 1816 Yamauchi, S., & Koyama, K. 1993, ApJ, 405, 268
Hofner, P., & Churchwell, E. 1997, ApJ, 486, L39 Yusef-Zadeh, F., Shure, M., Wardle, M., & Kassim, N. 2000, ApJ, 540,
Howarth, I. D., & Prinja, R. K. 1989, ApJS, 69, 527 842](https://crownmelresort.com/image.slidesharecdn.com/x-raysandprotostarsinthetrifidnebula-130128062932-phpapp01/75/X-rays-and-protostars_in_the_trifid_nebula-10-2048.jpg)
Uploaded bySérgio Sacani
X rays and-protostars_in_the_trifid_nebula
This document summarizes an astronomical study of the Trifid Nebula using X-ray and infrared observations. The ROSAT X-ray image revealed over a dozen X-ray sources in the nebula, including the bright O7 star that provides most of the ionization. Infrared observations identified 85 young star candidates. Ten X-ray sources had infrared counterparts. Spectroscopy of the brightest X-ray source with ASCA showed unusually hot plasma up to 10 keV, suggesting interaction between winds or flares from embedded young stars.
More Related Content
Similar to X rays and-protostars_in_the_trifid_nebula
X rays and-protostars_in_the_trifid_nebula
- 1. THE ASTROPHYSICAL JOURNAL,562 : 446È455, 2001 November 20 ( 2001. The American Astronomical Society. All rights reserved. Printed in U.S.A. X-RAYS AND PROTOSTARS IN THE TRIFID NEBULA JEONGHEE RHO Infrared Processing and Analysis Center, California Institute of Technology, MS 100-22, Pasadena, CA 91125 ; rho=ipac.caltech.edu MICHAEL F. CORCORAN NASA Goddard Space Flight Center, Code 662, Greenbelt, MD 20742 YOU-HUA CHU Department of Astronomy, University of Illinois, 1002 West Green Street, Urbana, IL 61801 AND WILLIAM T. REACH Space Infrared Telescope Facility Center, California Institute of Technology, MS 100-22, Pasadena, CA 91125 Received 2001 February 19 ; accepted 2001 June 28 ABSTRACT The TriÐd Nebula is a young H II region, recently rediscovered as a ““ pre-Orion ÏÏ star-forming region, containing protostars undergoing violent mass ejections visible in optical jets as seen in images from the Infrared Space Observatory and the Hubble Space T elescope. We report the Ðrst X-ray observations of the TriÐd Nebula using ROSAT and ASCA. The ROSAT image shows a dozen X-ray sources, with the brightest X-ray source being the O7 star, HD 164492, which provides most of the ionization in the nebula. We also identify 85 T Tauri star and young, massive star candidates from near-infrared colors using the JHK color-color diagram from the Two-Micron All-Sky Survey (2MASS). Ten X-ray sources s have counterpart near-infrared sources. The 2MASS stars and X-ray sources suggest there are potentially numerous protostars in the young H II region of the TriÐd. ASCA moderate-resolution spectroscopy of the brightest source shows hard emission up to 10 keV with a clearly detected Fe K line. The best model Ðt is a two-temperature (T 1.2 ] 106 K and 39 ] 106 K) thermal model with additional warm absorb- ing media. The hotter component has an unusually high temperature for either an O star or an H II region ; a typical Galactic H II region could not be the primary source for such hot temperature plasma and the Fe XXV line emission. We suggest that the hotter component originates in either the interaction of the wind with another object (a companion star or a dense region of the nebula) or from Ñares from deeply embedded young stars. Subject headings : H II regions È infrared : stars È stars : formation È X-rays : individual (TriÐd Nebula) 1. INTRODUCTION 1995 ; Feigelson et al. 1993), Lupus (Krautter et al. 1997), H II regions contain various types of X-rayÈemitting Ž and Taurus-Auriga (Neuhuser et al. 1995 ; Wichmann et al. sources, such as high- and low-mass stars, binaries, and 1997). Flarelike X-ray events have been detected from T protostars. Single massive O and B stars emit soft X-rays Tauri stars (Kamata et al. 1997 ; Koyama et al. 1992). The from shocks in their radiatively unstable, outwardly moving X-rayÈemitting sources in the Monoceros and Rosette outer atmospheres. X-ray emission from late-type stars is molecular clouds are also mostly T Tauri and Herbig Ae/Be attributed to magnetically heated stellar coronae and mag- stars and they typically show luminosities of L D netically driven stellar Ñares. Protostars produce X-rays via X 1030È1032 ergs s~1 (Gregorio-Hetem et al. 1998). Recent magnetic heating and perhaps from accretion or excretion Chandra observation of the Orion Nebula resolved a thou- disks and/or interaction of stellar jets with the circumnebu- sand X-rayÈemitting PMS stars with a mass range of 0.05 to lar gas. Massive binaries can produce X-ray emission from 50 M , and a combined infrared and X-ray study suggested colliding stellar winds ; close low-mass binaries can produce _ that the X-ray luminosity depends on stellar mass, rotation- X-rays via mass exchange and enhanced dynamo action. al history, and magnetic Ðeld (Garmire et al. 2000). X-ray emission provides a sensitive tracer of these di†erent At a distance of 1.67 kpc (Lynds & OÏNeil 1985), the stellar populations and, along with infrared colors, is one of TriÐd Nebula, M20, is one of the best-known astrophysical the main probes for identifying protostars and preÈmain- objects and one of the prettiest : it glows brightly in red light sequence (PMS) stars. and is trisected by obscuring dust lanes. At an age of Studies of X-ray emission from H II/star-forming regions D 3 ] 105 yr, the TriÐd is one of the youngest known H II have made signiÐcant advances because of ROSAT regions. Observations with the Infrared Space Observatory (Casanova et al. 1995), ASCA (e.g., Koyama et al. 1996), (ISO) and the Hubble Space T elescope (HST ) show the and, most recently, Chandra observations (Garmire et al. TriÐd as a dynamic, ““ pre-Orion ÏÏ star-forming region con- 2000). For example, in the o Oph core region, 70% of the taining young stars undergoing episodes of violent mass near-infrared (NIR) sources associated with protostars and ejections, and protostars (like HH 399) losing mass and molecular cores have X-ray counterparts (Casanova et al. energy to the nebula in optically bright jets (Cernicharo et 1995), and a large number of X-ray sources have been found al. 1998, hereafter CLC98 ; LeÑoch & Cernicharo 2000 ; to be low-mass PMS stars in other star-forming regions Hester et al. 1999). The ionization of the nebular gas is such as Orion (Alcala et al. 1996), Chamaeleon (Alcala et al. dominated by the O7.5 star HD 164492. HD 164492 is a 446
- 2. X-RAYS AND PROTOSTARSIN THE TRIFID NEBULA 447 luminosity class V (Levato 1975) or III (Conti & Alschuler in Figure 1. This image reveals for the Ðrst time that the 1971) star with a bolometric luminosity of L D 0.5È1.6 TriÐd Nebula contains numerous X-ray sources. We have ] 1039 ergs s~1 and an X-ray luminosity of 6 bol 1032 ergs ] identiÐed X-ray sources in the PSPC image using the s~1 (Chlebowski et al. 1989). The mass-loss rate of this O FTOOLS task SRCDETECT and estimated the count rates star is M 2 ] 10~6 M yr~1 (Howarth & Prinja 1989) 0 and uncertainties. The detected point sources are presented, _ and the wind terminal velocity is V 1580 km s~1 (Prinja, in order of increasing right ascension, in Table 1 and = Barlow, & Howarth 1990). marked in Figure 1. Table 1 lists the position, count rate, X-ray observations of the TriÐd o†er a unique and p of detection for 10 sources detected at greater than 3 opportunity to study the inÑuence of a massive star on star p and two possible sources (sources 11 and 12) detected with formation in an exceptionally young star-forming region. somewhat lower conÐdence. We deÐne here new X-ray Initially, X-ray emission from the TriÐd was reported only sources detected in the ROSAT PSPC image, as ROSAT from the O star in the Einstein IPC catalog (Chlebowski & X-ray source in the TriÐd (RXT). Since the count rate is very Harnden 1989). We serendipitously discovered a complex of small except for the O star HD 164492, we estimated the X-ray emission from the TriÐd Nebula in a PSPC obser- luminosity by assuming an absorption column density vation of the nearby supernova remnant W28 (Rho et al. N 3 ] 1021cm~2 (A D 1.5m, see ° 4 for details), and a H V 1995). Subsequently, we started an extensive investigation thermal spectrum with kT 1 keV. The X-ray emission of of the X-ray emission from the TriÐd Nebula. This paper PMS stars is understood to be thermal emission from gas presents the Ðrst detection of a dozen X-ray sources in the rapidly heated to a temperature of D1 keV by violent mag- TriÐd Nebula, and we correlate these with protostar candi- netohydrodynamic reconnection events (Feigelson & dates identiÐed using the Two-Micron All-Sky Survey Montmerle 1999). The correspondence between the PSPC (2MASS) data. The ROSAT images show multiple point count rate and the X-ray unabsorbed Ñux is 1 ] 10~3 sources, including HD 164492 and several T Tauri stars, PSPC counts s~1 D2.95 ] 10~14 ergs s~1 cm~2. Using this and the ASCA spectra show hard X-ray emission, including conversion, the luminosities of the X-ray sources are com- detection of an Fe K line. We discuss identiÐcations of the puted and given in Table 1. X-rayÈemitting sources and the origin of the unusually hard We have examined the source list database using X-ray emission from the TriÐd. SIMBAD, identiÐed counterparts to the X-ray sources at other wavelengths, and marked them in Figure 1. To visual- 2. X-RAY SOURCES IN THE TRIFID ize the correspondence of X-ray sources with either optical 2.1. X-Ray Observations or radio sources, we have plotted the X-ray contours over The TriÐd Nebula was observed using the X-ray tele- an Ha image (F. Winkler, 2001, private communication) in Ž scope on ROSAT (Trumper 1993) with the Position Sensi- Figure 2. Twenty-four sources from the Guide Star Catalog tive Proportional Counter (PSPC) as the imaging detector. (GSC) are visible in the Ha image in Figure 2. The brightest The PSPC on-axis angular resolution is 25A (FWHM, at 1 X-ray point source, RXT8 in Table 1, corresponds to the O keV), and the PSPC covers a 2¡ Ðeld of view in the 0.1È2.4 star HD 164492. The possible X-ray source RXT11 coin- keV energy band. Two PSPC observations were analyzed cides with the B8 star HD 313596 (R.A. 18h02m35s and decl. for this paper : rp900375 centered on HD 164492 and [22¡59@54A), and the possible X-ray source RXT12 with the observed on 1993 September 8 for an exposure of 9365 s (PI : S. Snowden), and rp500236 centered on the supernova remnant W28 and observed on 1993 April 1 for an exposure of 10,476 s (PI : R. Pisarski ; Rho et al. 1995). We also performed an ASCA observation (PI : J. Rho ; sequence number 26051000) toward the center of the TriÐd Nebula. The observation took place on 1998 September 30 to October 2. ASCA (TI & H94) has two detector pairs : Gas Imaging Spectrometers (GIS2 and GIS3) and Solid-State Imaging Spectrometers (SIS0 and SIS1). The SIS covers an energy band of 0.5È10 keV and the GIS 0.6È10 keV. The on-axis angular resolution of the GIS and SIS is about 1@È2@. Each GIS counter has a circular Ðeld of view of 35@ diameter while the Ðeld of view of each SIS CCD is an 11@ square ; thus, both GIS and SIS detectors sufficiently cover the entire TriÐd Nebula. We Ðltered the data using a few cri- teria such as cuto† rigidity (COR) and earth elevation (based on Revison 2 processing). After Ðltering, the expo- sure time was 57 ks for the GIS and 53.5 ks for the SIS. The entire TriÐd Nebula region after background subtraction has count rates of 0.050 ^ 0.001 counts s~1 for SIS0, 0.040 ^ 0.001 counts s~1 for SIS1, 0.030 ^ 0.008 counts s~1 for GIS2, 0.039 ^ 0.009 counts s~1 for GIS3, and 0.086 ^ 0.004 counts s~1 for the ROSAT PSPC, in their respective energy bands (integrated over all channels). FIG. 1.ÈPSPC X-ray image and detected sources (RXT) are marked 2.2. X-Ray Source IdentiÐcation with numbers. Count rates for each source are listed in Table 1. Known The ROSAT PSPC image of the TriÐd Nebula is shown sources are also marked with the labels.
- 3. 448 RHO ET AL. Vol. 562 TABLE 1 X-RAY SOURCES DETECTED IN THE PSPC IMAGE OF THE TRIFID NEBULA Name R.A. (2000) Decl. (2000) Count Rate Detection (p) log (L ) X RXT1 . . . . . . . 18 02 52.8 [23 02 18.1 0.0014 ^ 0.0005 5 31.12 RXT2 . . . . . . . 18 02 39.3 [22 58 34.3 0.0011 ^ 0.0004 3.5 31.02 RXT3 . . . . . . . 18 02 41. [23 03 51.8 0.0014 ^ 0.0006 3.5 31.12 RXT4 . . . . . . . 18 02 36.0 [23 01 36.3 0.0019 ^ 0.0006 4 31.05 RXT5 . . . . . . . 18 02 35.0 [23 01 29.4 0.0021 ^ 0.0006 4.5 31.29 RXT6 . . . . . . . 18 02 27.9 [22 59 47.9 0.0016 ^ 0.0006 4 31.17 RXT7 . . . . . . . 18 02 25.4 [22 59 51.4 0.0010 ^ 0.0005 3 30.97 RXT8 . . . . . . . 18 02 23.35 [23 01 47.0 0.0131 ^ 0.0013 8 (see Table 4) RXT9 . . . . . . . 18 02 21.1 [23 03 21.5 0.0010 ^ 0.0004 4 30.97 RXT10 . . . . . . 18 02 12.38 [22 55 37.0 0.0012 ^ 0.0004 4.5 31.05 RXT11 . . . . . . 18 02 34.92 [22 59 55.6 0.0006 ^ 0.0003 2.5 30.75 RXT12 . . . . . . 18 02 31.67 [23 02 25.6 0.0007 ^ 0.0003 2.5 30.82 NOTE.ÈUnits of right ascension are hours, minutes, and seconds, and units of declination are degrees, arcmin- utes, and arcseconds. optical star GSC–06842–00001 (R.A. 18h02m34s 8 and decl. . sources showing NIR color excesses and present candidate [23¡03@06A3). None of the other GSC stars coincide with . protostars. X-ray peaks. A radio source, GPSR5 6.980[0.286 (R.A. 18h02m28s 1 and decl. [23¡03@46A3 ; Becker et al. 1994), is . . 3. NEAR-INFRARED SOURCES FROM 2MASS : YOUNG close to the X-rayÈemitting area but does not have a corre- STELLAR OBJECTS sponding X-ray peak. Four protostars (TC0, TC1, TC3, and We have identiÐed young stellar objects (YSOs) using the TC4 sources1 in CLC98) have been reported in the TriÐd 2MASS data (Skrutskie et al. 1997). Using identical tele- Nebula (CLC98), which are marked in Figure 1, but only scopes in the northern and southern hemispheres, 2MASS is one is close to the X-ray peak at HD 164492. In the mapping the entire sky in the J (1.11È1.36 km), H (1.5È1.8 next section, we correlate the PSPC X-ray sources with km), and K (2È2.32 km) bands to a limiting point-source s sensitivity of approximately 16.5, 16.0, and 15.5 mag, 1 The source designation TC0, TC1, etc. was assigned by CLC98. Since respectively (Cutri et al. 2000).2 The data toward the TriÐd this designation was already in use, for indexing purposes these sources Nebula were taken on 1998 June 14 using the southern should be referred to as [CLC98] 0, etc. telescope, and most of these data were included in the 2MASS Second Incremental Release, but a small portion of area was in the 2MASS Working Database because of large photometric uncertainties at the time of the Incremental Release. The photometry is typically better than 5% (Cutri et al. 2000Èsee fn. 2). We used the 2MASS point-source catalog to extract sources within an 8@ radius centered on R.A. 18h02m30s and decl. [23¡02@00A. We have selected sources with the follow- ing criteria. First, we selected the sources that were detected in all three J, H, and K bands. We then selected sources s with a signal-to-noise ratio greater than 10 (i.e., the J, H, and K magnitudes are brighter than 15.8, 15.1, and 14.3 s mag, respectively). These selections produced D1100 such sources. We then accepted sources with photometric uncer- tainties p 0.25 mag, for which the Ðt to the point-spread function produced reduced s2 2. This last criterion l excluded blended sources that caused higher uncertainties in the photometry. This is important in the Galactic plane where NIR sources are crowded and the 2MASS has a limited spatial resolution (3A5). This criterion removed inac- . curate blue points that appeared in the JHK color-color s diagram (as described below). The criteria we have used are conservative for the magnitude limit and photometric uncertainties of the 2MASS survey. We plotted the sources in the JHK color-color diagram, s as shown in Figure 3a, in order to identify YSOs with infra- FIG. 2.ÈPSPC X-ray contours superimposed on an Ha image of the red color excess. Figure 3b shows H[K versus K magni- s s TriÐd Nebula (courtesy of Frank Winkler). The optical image was tude for the selected sample. The interstellar reddening obtained on 1994 July 4 (UT) from the Burrell Schmidt telescope of Case Ž Western Reserve University through a 25 A bandpass Ha Ðlter. Total exposure time is 1800 s, and the scale is 2A0 pixel~1. The strongest X-ray . 2 See R. M. Cutri et al. 2000, at http ://www.ipac.caltech.edu/2mass/ peak is at the O star, HD 164492. releases/ Ðrst/doc/explsup.html.
- 4. No. 1, 2001 X-RAYS AND PROTOSTARS IN THE TRIFID NEBULA 449 FIG. 3b FIG. 3a FIG. 3c FIG. 3.È(a) Near-infrared JHK color-color diagram : location in color-color diagram is used to determine T Tauri stars and massive YSOs. Normal stars (dots), T Tauri stars (Ðlled circles), s and massive YSOs (diamonds) are shown with di†erent symbols. The 2MASS counterparts of X-ray sources are marked with crosses ; the sources above the extinction curve have high photometric uncertainties. They are 2MASS-red stars in Table 3 (see text for details). Extinction vector is shown for 5 mag as the thick line labeled A . Thick curves are intrinsic colors of giant and dwarf stars. (b) Diagram of H [ K vs. K . T v s s Tauri stars and young massive protostars fall below (younger age) the PMS stars in this diagram. Symbols are the same as in (a). (c) T Tauri stars (circles) and massive YSOs (diamonds), X-ray sources (crosses), and known embedded YSOs (triangles : from Cernicharo et al. 1998) are marked on optical image. (d) 2MASS three-color image (blue, green, and red for J, H, and K , respectively). Symbols are the same as in (c). The di†use, blue emission is probably Pab (in s the J band) from the H II region. vector from Rieke & Lebofsky (1985) is also plotted. Adopt- the extinction curve (dashed line in Fig. 3a) and above the T ing the intrinsic colors of giant and dwarf stars from Bessell Tauri star lines (solid line in Fig. 3a) are massive YSOs & Brett (1988), we Ðnd the visual extinction toward this (Lada & Adams 1992) ; these stars are plotted as diamonds direction is as high as A 30 mag for stars on the back in Figure 3a. In total, we found 41 T Tauri star candidates V side of the TriÐd Nebula. The observed JHK colors and 44 massive YSO candidates from the 2MASS sources of YSOs can be explained by circumstellar disks models that are listed in Table 2. The T Tauri stars and YSOs are (Lada & Adams 1992). We identiÐed the T Tauri stars with generally located along the ionization front. Figure 3c (J[H) 0.58 ^ 0.11 ] (H[K ) ] 0.52 ^ 0.06 shows optical images, and T Tauri stars, massive YSOs, and s and (Meyer, Calvet, & Hillenbrand 1997) CTTS H[K [ 0.6 mag, CTTS X-ray sources are marked. Figure 3d shows a color compos- where CTTS is classical T Tauri stars. These Ts Tauri stars ite of the J, H, and K images, with the J image in blue, H fall between the two solid lines in Figure 3a. The stars below image in green, and Ks image in red. T Tauri stars, massive s
- 5. 450 RHO ET AL. Vol. 562 FIG. 3d YSOs, and X-ray sources are marked. The image also shows tive criteria as they are not detected in one or both of the J a number of protostars at the ionization front along the and H bands. Therefore, we have relaxed the selection cri- dust lane. teria and extracted 2MASS sources from the same area, These protostar candidates were cross-correlated with excluding only sources with an artifact Ñag. We made a list the X-ray sources. Taking into account the resolution of the of the 2MASS sources that are within 15A from X-ray PSPC, we identify a possible coincidence if the separation sources. Each X-ray source has D10 2MASS counterpart between a 2MASS source and an X-ray source is less than candidates ; only the red stars are identiÐed as possible 15A. Three X-ray sources (RXT1, RXT6, and RXT7) are counterparts to the X-ray sources. Seven red 2MASS star coincident with either T Tauri stars or YSOs ; they are counterparts of X-ray sources are identiÐed and listed in noted in Table 3. Figure 3d shows that many red 2MASS Table 3, with J, H, and K magnitudes and reasons that s stars within the X-ray source error boxes could be con- they were not selected in the earlier list of T Tauri star or sidered coincidences, but they are not included in the list of massive YSO candidates (either J ““ Ðll ÏÏ or high reduced sources selected according to the aforementioned conserva- s2, indicating multiple or extended sources with large
- 6. No. 1, 2001 X-RAYS AND PROTOSTARS IN THE TRIFID NEBULA 451 TABLE 2 2MASS YOUNG STELLAR OBJECT AND T TAURI STAR CANDIDATES IN THE TRIFID NEBULAa YSO Candidates TTS Candidates Number Designation Number Designation Number Designation Number Designation 1 2MASSI 23 2MASSW 1 2MASSI 23 2MASSW J1802121[230439 J1802379[230211 J1802158[230555 J1802331[230135 2 2MASSI 24 2MASSW 2 2MASSI 24 2MASSW J1802211[230437 J1802211[230234 J1802102[230403 J1802197[230136 3 2MASSI 25 2MASSI 3 2MASSI 25 2MASSW J1801592[230406 J1802441[230245 J1802129[230348 J1802281[230142 4 2MASSW 26 2MASSI 4 2MASSW 26 2MASSW J1802139[230213 J1802367[230302 J1802197[230136 J1802336[230154 5 2MASSW 27 2MASSI 5 2MASSI 27 2MASSI J1802142[230144 J1802450[230332 J1801575[230121 J1802199[230305 6 2MASSI 28 2MASSI 6 2MASSI 28 2MASSI J1802168[230110 J1802500[230334 J1802124[225856 J1802395[230343 7 2MASSI 29 2MASSI 7 2MASSI 29 2MASSI J1801561[230009 J1802361[230428 J1802173[225614 J1802211[230437 8 2MASSI 30 2MASSW 8 2MASSI 30 2MASSI J1802044[225829 J1802319[230440 J1802108[225532 J1802409[230459 9 2MASSI 31 2MASSI 9 2MASSI 31 2MASSI J1802122[225825 J1802372[230503 J1802205[225458 J1802349[230505 10 2MASSW 32 2MASSI 10 2MASSI 32 2MASSI J1802202[225807 J1802369[230656 J1802205[225458 J1802263[230730 11 2MASSI 33 2MASSI 11 2MASSI 33 2MASSW J1802401[225721 J1802470[230702 J1802162[225530 J1802242[230910 12 2MASSI 34 2MASSI 12 2MASSI 34 2MASSI J1802377[225850 J1802290[230704 J1802464[225546 J1802403[230913 13 2MASSI 35 2MASSW 13 2MASSI 35 2MASSI J1802463[225912 J1802247[230729 J1802473[225729 J1802384[230934 14 2MASSI 36 2MASSI 14 2MASSI 36 2MASSI J1802433[225931 J1802344[230806 J1802367[225804 J1803003[230134 15 2MASSI 37 2MASSI 15 2MASSW 37 2MASSI J1802491[225945 J1802339[230924 J1802312[225911 J1802467[230130 16 2MASSI 38 2MASSI 16 2MASSW 38 2MASSI J1802503[225949 J1802518[230810 J1802305[225929 J1802538[230033 17 2MASSI 39 2MASSI 17 2MASSW 39 2MASSI J1802438[225959 J1802499[230740 J1802228[225935 J1802566[230026 18 2MASSI 40 2MASSI 18 2MASSI 40 2MASSI J1802441[230020 J1802541[230736 J1802175[225938 J1803023[230022 19 2MASSI 41 2MASSI 19 2MASSI 41 2MASSI J1802266[230036 J1802473[230428 J1802409[225941 J1802490[225656 20 2MASSW 42 2MASSI 20 2MASSI J1802226[230101 J1802482[230309 J1802344[230102 21 2MASSI 43 2MASSI 21 2MASSI J1802467[230130 J1802524[230227 J1802284[230115 22 2MASSW 44 2MASSI 22 2MASSW J1802271[230201 J1802548[225901 J1802408[230131 a 2MASSI and 2MASSW indicate the Incremental Release and Working Database catalogs, respectively. photometric uncertainties). These seven sources are likely expect there are large numbers of such embedded protostars protostars because their infrared colors are red and they in the TriÐd Nebula not detected in the NIR observations. also emit X-rays. By itself, an infrared color excess indicates This population is likely correlated with highly extincted either the presence of a YSO or a heavily extincted main- dust lanes (see Fig. 3d) and molecular clouds. sequence star. The 2MASS counterparts of the X-ray sources are marked in the JHK color-color diagram (Fig. 4. X-RAY SPECTRAL ANALYSIS s The ASCA SIS image is shown in Figure 4 with the PSPC 3a) and on the 2MASS composite image (Fig. 3d). A few X-ray sources do not have obvious optical or 2MASS contours superposed. The X-ray emission is dominated by counterparts. the emission from the O star. We extracted ASCA and There are Ðve known massive YSOs in the TriÐd with ROSAT spectra from the entire region of the TriÐd Nebula. mid-infrared emission detected using ISO (CLC98). These The spectrum (Fig. 5a) shows clear Fe K line emission, protostars were not identiÐed as protostellar candidates along with weak Si and S lines, and a hard continuum tail from the 2MASS data, probably because they are too up to 10 keV, the highest energy observable by ASCA. We deeply embedded to be detected in the NIR. A 2MASS red extracted a background-corrected spectrum from a smaller star is the counterpart of the protostar TC2 in CLC98 ; we region centered on HD 164492. HD 164492 is the brightest
- 7. 452 RHO ET AL. Vol. 562 TABLE 3 INFRARED COUNTERPARTS OF THE X-RAY SOURCES DETECTED IN PSPC IMAGES 2MASSa db Jd Hd Kd s RXT Counterparts (arcsec) R.A. (2000)c Decl. (2000)c (mag) (mag) (mag) Commentse RXT1 . . . . . . . . YSO 10 18 02 52.4 [23 02 27.2 15.17 13.93 13.18 ... RXT2 . . . . . . . . red 8 18 02 39.7 [22 58 29.0 17.72 15.06 13.50 J ““ Ðll ÏÏ RXT3 . . . . . . . . red 5 18 02 41.4 [23 03 48.2 17.83 14.24 12.42 J ““ Ðll ÏÏ red 13 18 02 40.4 [23 03 58.9 15.58 12.37 10.90 High psf *s2 RXT4 . . . . . . . . red 12.8 18 02 36.8 [23 01 43.9 15.85 12.78 11.34 High psf *s2 RXT5 . . . . . . . . NONE ... ... ... ... ... ... ... RXT6 . . . . . . . . TTS 13 18 02 30.5 [22 59 28.7 12.71 11.65 10.91 ... RXT7 . . . . . . . . YSO 12 18 02 26.6 [23 00 36.0 15.21 12.90 11.34 ... RXT8f . . . . . . . Multiple ... ... ... ... ... ... ... RXT9 . . . . . . . . red 6 18 02 21.5 [23 03 19.0 15.04 13.58 11.87 J ““ Ðll ÏÏ RXT10 . . . . . . NONE ... ... ... ... ... ... ... RXT11f . . . . . . red 4 18 02 34.7 [22 59 52.1 16.52 12.42 10.58 J ““ Ðll ÏÏ RXT12 . . . . . . red 9 18 02 31.3 [23 02 18.2 15.12 13.17 11.36 J ““ Ðll ÏÏ NOTE.ÈUnits of right ascension are hours, minutes, and seconds, and units of declination are degrees, arcminutes, and arcseconds. a Notes for the 2MASS sources : TTS T Tauri star ; YSO massive young stellar object ; red red 2MASS star. b Projected distance between the X-ray and 2MASS sources. c Coordinates of the 2MASS counterparts. d J, H and K magnitudes of the 2MASS sources. s e J ““ Ðll ÏÏ means the J-band photometry is measured in band-Ðlled within the aperture, indicating their photometric uncertainties are large. ““ High psf *s2 ÏÏ means the goodness of the Ðt of a point-spread function is high, indicating the sources are either extended or unresolved double sources. f RXT8 is HD 164492 (O7 star) and/or YSO (TC1 in CLC98), and RXT11 also coincides with HD 313596 (B8 star). X-ray source in the TriÐd, and the shape of the spectrum of showed that the observed o†-source emission is dominated the entire region in the ASCA data is not signiÐcantly di†er- by scattered emission from sources within the TriÐd ent from the spectrum of this star alone because of the Nebula. The detection of hard emission from the TriÐd is broad ASCA point-spread function. We made hard ([3 unusual because most single O stars have very little emis- keV) and soft (3 keV) maps using the ASCA data, but no sion at energies above 2 keV and rarely show Fe K emission obvious di†erence was noticeable at the spatial resolution (Corcoran et al. 1993). The only massive stars to show such (1@) of ASCA. We also extracted an o†-source ASCA spec- hard X-ray spectra are binaries, either high-mass X-ray trum in this direction, suspecting a contribution from the binaries (HMXBs) with collapsed companions or colliding Galactic ridge emission. However, the o†-source spectrum wind binaries with noncollapsed (O or W-R star) compan- ions that have strong stellar winds and signiÐcant colliding wind X-ray emission. We simultaneously Ðt the set of Ðve spectraÈthe ROSAT /PSPC, ASCA/SIS0, SIS1, GIS2, and GIS3 spectraÈusing single- or two-temperature thermal models (Mewe-Kaastra plasma model ; Kaastra 1992) with a single absorbing column density N . The Ðts were unacceptable H (reduced s2 of 3). We next attempted a two-temperature model. Following Corcoran et al.Ïs (1994) models of d Ori and j Ori, we also included an additional ionized (““ warm ÏÏ) absorbing medium, as representative of the photoionized stellar wind material (Waldron 1984 ; Corcoran et al. 1994), and allowed di†erent amounts of absorption for the hot and cold components. The line-of-sight extinction value is known toward this direction ; A 1.3È1.5 mag, i.e., V E(B[V ) D 0.3È0.4 mag (Kohoutek, Mayer, & Lorenz 1999 ; Lynds & OÏNeil 1985). Using E(B[V ) of 0.4 mag, we expect an ISM N of D3 ] 1021 cm~2, with which we have Ðxed H the N value in our Ðt (also note that when we allow N to H ] 1021 cm~2 falls within the errors). The model H vary, 3 yielded an acceptable Ðt with kT D 0.14 keV (1.2 ] 106 K) 1 and N 5.9 ] 1021 cm~2, and a hotter component with H,1 keV (3.9 ] 107 K) and N 2.7 ] 1021 cm~2, kT D 3.3 with a line-of-sight ISM N 3 ] H,2 cm~2. The abun- 2 1021 H dances are Ðxed at solar abundances. The Ðt results are summarized in Table 4, along with the Fe K line character- FIG. 4.ÈASCA SIS image (gray scale ranges 0.7È4.5 counts (6A4 . istics. The cold component arises from the O star atmo- pixel~1) superposed on ROSAT contours. sphere, but the hot component might arise from a number
- 8. No. 1, 2001 X-RAYS AND PROTOSTARS IN THE TRIFID NEBULA 453 of di†erent sources, such as unresolved interacting binaries, active low-mass stars, or PMS stars. 5. THE NATURE OF THE HARD X-RAY COMPONENT Our simultaneous ASCA/PSPC Ðts yield a total X-ray luminosity of 1.9È2.5 ] 1034 ergs s~1 (0.3È10 keV) using the two-temperature model shown in Table 4. If attributed to HD 164492, then the ratio of X-ray and bolometric lumi- nosities ( log L /L ) is between [5.0 and [4.5, which is X bol much higher than the typical ratios of log L /L D [7 for X bol a single O-type star (Chlebowski & Harnden 1989 ; Ž Berghofer, Schmitt, & Cassinelli 1996). The 3 keV X-ray component is somewhat of a mystery since winds from single O-type stars are not known to produce such high- temperature emission. Typically the highest temperature emission observed in O star X-ray spectra has kT 1 keV (e.g., Corcoran et al. 1994). FIG. 5a We discuss a few possibilities to explain the hot com- ponent in the TriÐd Nebula. It may be that the hot com- ponent arises in the interaction of the wind from the O7.5 star with another object (either a companion star or a dense region of the nebula) outside the O star atmosphere. Collid- ing winds between an early-type star and an early-type companion (another O star or a Wolf-Rayet star) can produce shock-heated material in the wind interaction regions (Stevens, Blondin, & Pollock 1992), reaching tem- peratures of 107[108 K, and emit X-rays. However, recent photometric and spectroscopic studies of this region found no evidence of a companion for HD 164492 (Kohoutek, Mayer, & Lorenz 1999). Unless the collision occurs far from the star (d [ 25@@), it will be unresolved to the ROSAT PSPC. However, recent radio and NIR observations toward the central region of the TriÐd detected three sources close to the O star (Yusef-Zadeh et al. 2000), which may be either stars or nebular knots photoionized by the UV Ðeld of HD 164492. FIG. 5b Hard emission could also arise in single O stars from nonthermal emission produced by Fermi acceleration by shocks in the O star wind (Chen & White 1991). The hard component is suggested as a nonthermal tail produced by inverse Comptonization of the photospheric UV Ðeld by a population of fast particles accelerated by a distribution of shocks. Although it is not possible to determine if the hard emission is from a nonthermal tail or the two thermal tem- perature components from goodness of the Ðt to the spectra, the presence of the Fe K line suggests that the second com- ponent is thermal. The sources of the two spectral com- ponents need to be resolved spatially in order to determine their origins. Enhanced hard X-ray emission might also be produced by an oblique magnetic rotator, as is suspected in another O7 star, h1 Orionis C, the central star of the Orion Nebula (Gagne et al. 1997). The X-ray properties of the core of the H II region W3 share a number of similarities to the TriÐd emission. For FIG. 5c W3, the luminosity is a few 1033 ergs s~1 with a similarly high temperature (Hofner & Churchwell 1997). In W3 (and FIG. 5.È(a) GIS1 and GIS2 spectra of the TriÐd Nebula with its best Ðt possibly in the TriÐd), the high-temperature component of a two-temperature thermal model with additional warm absorbing media. Hard emission and Fe XXV line appear in the spectra. (b) SIS0 and may be produced by a hot, wind-shocked cavity that results SIS1 and PSPC spectra with the best Ðts. (c) Each of two temperature when strong stellar winds interact with a surrounding dense components (D1.2 ] 106 K and 3.9 ] 107 K) is marked on the SIS0 molecular cloud (e.g., Churchwell 1990). The presence of a spectrum. known YSO (TC1 in CLC98) is consistent with the presence of a molecular cloud in the TriÐd. In addition, IRAS obser- vations of HD 164492 (van Buren, Noriega-Crespo, & Dgani 1995) show a bow-shock structure around the star,
- 9. 454 RHO ET AL. Vol. 562 TABLE 4 SPECTRAL FIT RESULTS FROM SIMULTANEOUS FIT USING ASCA/PSPC Parameters TriÐd ISM N (1021 cm~2) . . . . . . . . . . . . . . . . . . . . . . . . . . 43 H T ................................................ 3.3 ^ 0.6 keV ( 3.9 ] 107 K) 2 log EM (cm~3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55.8 2 Wind N , hot component (1021 cm~2) . . . . . . 2.7`2.3 H ~2.0 T ................................................ 0.14`0.06 keV ( 1.2 ] 106 K) 1 ~0.04 log EM (cm~3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57.4 1 Wind N , 1 component (1021 cm~2) . . . . . . . . 5.9`1.1 H ~0.9 Flux . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 ^ 0.5 ] 10~12 Unabsorbed Ñux . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 ^ 2 ] 10~11 Luminosity (ergs s~1) . . . . . . . . . . . . . . . . . . . . . . . . . 2.5 ] 1034 [d(kpc)/(1.67 kpc)]2 Fe K Line central energy (keV) . . . . . . . . . . . . . . 6.65`0.1 ~0.2 Fe K equivalent width (keV) . . . . . . . . . . . . . . . . . 1.7 ^ 0.3 Fe K Ñux (photons s~1 cm~2) . . . . . . . . . . . . . . . 6( ^ 0.4) ] 10~4 perhaps indicative of a wind-cloud collision that might from unresolved emission from active PMS stars. With our produce the high-temperature X-ray emission. The stellar current data, we cannot determine if one is more favored wind outÑow at a speed of 1600 km s~1 should produce a than the other. To conclusively identify the hot component, postshock temperature of D30 million degrees, though the a high-resolution image is needed to locate the emitting observed temperature may be lower since radiative cooling object to determine whether the observed emission is pro- is rapid. CLC98 suggested that the HCO` molecular clouds duced near the O star or whether a distributed group of (likely the dust lanes) are fragmented shell around the active PMS stars dominates the observed emission. Along nebula. In other words, the clouds and dust lanes we see in with the images, time-resolved spectra could allow us to the optical image are located at the surface of the ionized distinguish whether the hard emission is Ñarelike (time sphere. If this is the case, the 1600 km s~1 wind will be variable). interacting with the lower density, ionized medium, and the 6. IDENTIFICATION OF X-RAY AND INFRARED SOURCES shocked stellar wind can emit at high temperatures while the photoionized materials of the edge of clouds could WITHIN THE TRIFID supply sufficient density to emit strong X-rays. The detected X-ray sources and their counterparts are The other possibility is that the hot component arises listed in Table 3. Most of X-ray sources are likely protostars from deeply embedded young stars, especially since at least or PMS stars ; one source is a T Tauri star, two are massive one embedded T Tauri star (TC1 in CLC98, which is protostars, and the others are unclassiÐed protostars. The marked in Fig. 1) exists near the O star. The TC1 source in JHK color-color diagram using the 2MASS data suggests s CLC98 shows a large shift in the spectral energy distribu- that there are D80 protostars present in this nebula. It has tion and violent ejections of high-velocity material (CLC98). been shown already (CLC98 ; LeÑoch & Cernicharo 2000) Other ASCA observations showed bright X-ray sources that massive protostars (17È60 M ) are forming in the _ with temperatures of 2È5 keV caused by Ñares of protostars TriÐd ; they are associated with molecular gas conden- in the o Ophiuchi dark cloud (Koyama et al. 1992), the sations at the edges of clouds and their dynamical ages are Orion Nebula (Yamauchi & Koyama 1993), and the R less than 104 yr. Whether low-mass protostars and T Tauri Coronae Australis molecular cloud (Koyama et al. 1996). stars can be formed in a young (3 ] 105 yr) region such as The hard emission was attributed to Ñares from individual the TriÐd is still an open question. Low-mass PMS stars of PMS stars with typical X-ray luminosities in the range similarly young age were found in the Orion Nebula using 1030È1032 ergs s~1, and the peak luminosity of Ñares is new Chandra observations (Hillenbrand et al. 1998 ; shown to be as large as 1033È1035 ergs s~1 and a tem- Garmire et al. 2000). The Chandra observation also showed perature as high as 108 K (Feigelson & Montmerle 1999 ; the presence of young, low-mass (0.1È3 M ) PMS stars as Grosso et al. 1997). The hard emission from W3 may be of a X-ray sources (Garmire et al. 2000). The_populations of similar origin. Signs of active star formation in the TriÐd low-mass and massive protostars are similar in the TriÐd, have been reported recently (LeÑoch et al. 2001) : there is a while we expect higher populations of low-mass protostars dust cocoon or circumstellar disk around several members based on the initial mass function. This is consistent with in the center of the TriÐd, and one young stellar source the fact that the TriÐd is a very young H II region ; T Tauri shows a silicate feature in the circumstellar disk. Neither the stars have yet to form there. The distribution of T Tauri o Ophiuchi dark cloud nor the R Coronae Australis molec- stars and massive YSOs is not obviously correlated with the ular cloud is as bright in hard X-rays as the TriÐd Nebula, molecular cloud distribution. It is possible that they are which may imply there are a higher number of protostars highly embedded in the molecular clouds, and their NIR present in the TriÐd. A large number of protostars unre- colors cannot be fully obtained to identify protostars solved to ASCA would dilute any Ñux variability produced because either J and/or H Ñux is unavailable. This is consis- by Ñares. tent with the fact that 2MASS images show a higher popu- In summary, one of two scenarios is likely responsible for lation of red stars in the southern part of the TriÐd. The the hard emission : The emission may arise from HD 169942 HST images covering the southern part suggested the pres- by the interaction of the wind from the O star with another ence of embedded stars at the head of the evaporating object (a companion star or a dense region of the nebula) or globules (Hester et al. 1999). Deep NIR images and spec-
- 10. No. 1, 2001 X-RAYS AND PROTOSTARS IN THE TRIFID NEBULA 455 troscopy will likely reveal hundreds of young protostars in similarly to 30 Dor, we would expect D1035 ergs s~1 di†use the TriÐd as suggested by 2MASS and HST data. X-ray emission from the TriÐd. This is higher than the total Whether di†use X-ray emission exists within the TriÐd X-ray luminosity of the TriÐd. It is likely that supernova Nebula is currently unknown because of the limited spatial heating contributes signiÐcantly to the bright X-ray emis- resolution of the PSPC images. For the unidentiÐed X-ray sion from 30 Dor. The TriÐd Nebula is too young to have sources 5 and 10, we cannot determine whether they are a hosted supernova explosions, so its di†use X-ray emission part of di†use emission or real point sources. They are very should be much fainter than that of 30 Dor. likely normal stars, but the possibility that they are knots of Future high-resolution X-ray observations by new tele- di†use emission cannot be ruled out. Di†use X-ray emission scopes such as Chandra and XMM should be able to from H II regions has been detected, although it is rare. A resolve the O star from its immediate environment and dis- few examples have been found, such as in the Carina crete X-ray sources, such as T Tauri stars, numerous pro- Nebula (Seward & Chlebowski 1982), RCW 49 (Goldwurm, tostars, young and old normal stars, and to resolve the Caraveo, & Bignami 1987 ; Belloni & Mereghetti 1994), and source of the high-temperature emission. A deep NIR image the Cygnus Superbubble (Bochkarev & Sitnik 1985), and with other wavelength observations can identify the PMS recently Wang (1999) reported di†use X-ray emission from stars and protostars and their mass populations. The TriÐd the giant H II region 30 Dor in the Large Magellanic Cloud. Nebula is an exciting laboratory to understand the early ROSAT and Broadband X-Ray T elescope observations of stage of star-forming activities in H II region. the Carina Nebula show large-scale di†use emission over at least 40@, as well as discrete X-ray sources and hot gas sur- rounding g Car (Corcoran et al. 1995). The nature of di†use We thank Lynne Hillenbrand for helpful discussion on emission is unclear in H II regions. Seward & Chlebowski near-infrared colors of protostars and protostar disk (1982) suggested that stellar winds from the OB association models and for useful comments on the manuscript and adequately heat the plasma. Wang (1999) suggested that the John Carpenter for helpful discussion on 2MASS data. We X-ray thermal di†use emission arises in blister-shaped thank Frank Winkler for allowing us to reproduce his region by loops of ionized gas and the structure is explained optical image. For J. R., this work is partially supported by by the mass loading of the hot gas produced by the central NASA/ADP grant, NASA-1407. This publication makes OB association. use of data products from the Two-Micron All-Sky Survey, We compare the TriÐd Nebula with 30 Dor to determine which is a joint project of the University of Massachusetts whether stellar winds in the TriÐd may produce observable and the Infrared Processing and Analysis Center, funded by di†use X-rays. The stellar wind luminosity in 30 Dor is a the National Aeronautics and Space Administration and few ] 1039 ergs s~1, and the X-ray luminosity of 30 Dor is the National Science Foundation. J. R. and W. T. R. D1038 ergs s~1 (Wang 1999). The stellar wind luminosity of acknowledge the support of the Jet Propulsion Laboratory, HD 146692 is L 1.7 ] 1036(M/2 ] 10~6 M yr~1)(V / 0 California Institute of Technology, which is operated under w _ 1580 km s~1)2 ergs s~1. If the TriÐd emits di†use X-rays contract with NASA. REFERENCES Alcala, J. M., et al. 1995, A&AS, 114, 109 Kaastra, J. S. 1992, An X-Ray Spectral Code for Optically Thin Plasmas, ÈÈÈ. 1996, A&AS, 119, 7 SRON-Leiden Rept. (Leiden, Netherlands : SRON) Becker, R. H., White, R. L., Helfand, D. J., & Zoonematkermani, S. 1994, Kamata, Y., Koyama, K., Tsuboi, Y., & Yamauchi, S. 1997, PASJ, 49, 461 ApJS, 91, 347 Kohoutek, L., Mayer, P., & Lorenz, R. 1999, A&AS, 134, 129 Belloni, R., & Mereghetti, S. 1994, A&A, 286, 935 Koyama, K., Asaoka, I., Kuriyama, T., & Tawara, Y. 1992, PASJ, 44, L255 Ž Berghofer, T. W., Schmitt, J. H. M. M., & Cassinelli, J. P. 1996, A&AS, 118, Koyama, K., Ueno, S., Kobayashi, N., & Feigelson, E. D. 1996, PASJ, 48, 481 L87 Bessell, M. S., & Brett, J. M. 1988, PASP, 100, 1134 Krautter, J., et al. 1997, A&AS, 123, 329 Bochkarev, N. G., & Sitnik, T. G. 1985, Ap&SS, 108, 237 Lada, C. J., & Adams F. C. 1992, ApJ, 393, 278 Casanova, S., Montmerle, T., Feigelson, E. D., & Andre, P. 1995, ApJ, 439, LeÑoch, B., & Cernicharo, J. 2000, ApJ, 545, 340 752 LeÑoch, B., et al. 2001, A&A, 368, L13 (erratum 372, L65) Cernicharo, J., et al. 1998, Science, 282, 462 (CLC98) Levato, A. 1975, A&AS, 19, 91 Chen, W., & White, R. L. 1991, ApJ, 366, 512 Lynds, B. T., & OÏNeil, E. J., Jr. 1985, ApJ, 294, 578 Chlebowski, T., Harnden, F. R., Jr., & Sciortino, S. 1989, ApJ, 341, 427 Meyer, M. R., Calvet, N., & Hillenbrand, L. 1997, AJ, 114, 288 Churchwell, E. 1990, A&A Rev., 2, 79 Ž Neuhauser, R., Sterzik, M. F., Schmitt, J. H. M. M, Wichmann, R., & Conti, P. S., & Alschuler, W. R. 1971, ApJ, 170, 325 Krautter, J. 1995, A&A, 297, 391 Corcoran, M. F., Swank, J., Rawley, G., Petre, R., Schmitt, J., & Day, C. Prinja, R. K., Barlow, M. J., & Howarth, I. D. 1990, ApJ, 361, 607 1995, Rev. Mexicana Astron. AstroÐs. Ser. de Conf., 2, 97 Ž Rho, J.-H., Petre, R., Pisarski, R., & Jones, L. R. 1995, in Rontgenstrahlung Corcoran, M., et al. 1993, ApJ, 412, 792 from the Universe, ed. H.-U. Zimmermann et al. (Garching : MPE), 273 ÈÈÈ. 1994, ApJ, 436, L95 Rieke, G. H., & Lebofsky, M. J. 1985, ApJ, 288, 618 Feigelson, E. D., Casanova, S., Montmerle, T., & Guibert, J. 1993, ApJ, Seward, F. D., & Chlebowski, T. 1982, ApJ, 256, 530 416, 623 Skrutskie, M. F., et al. 1997, in The Impact of Large-Scale Near-IR Sky Feigelson, E. D., & Montmerle, T. 1999, ARA&A, 37, 363 Surveys, ed. F. Garzon, N. Epchtein, A. Omont, B. Burton, & P. Persi Gagne, M., Caillault, J.-P., Stau†er, J. R., & Linsky, J. L. 1997, ApJ, 478, (Dordrecht : Kluwer), 25 L87 Stevens, I. R., Blondin, J. M., & Pollock, A. M. 1992, ApJ, 386, 265 Garmire, G., Feigelson, E. D., Broos, P., Hillenbrand, L. A., Pravdo, S. H., Tanaka, Y., Inoue, H., & Holt, S. S. 1994, PASJ, 46, L37 (TI of H94) Townsley, L., & Tsuboi, Y. 2000, AJ, 120, 1426 Ž Trumper, J. 1993, Science, 260, 1769 Goldwurm, A., Caraveo, P. A., & Bignami, G. F. 1987, ApJ, 322, 349 van Buren, D., Noriega-Crespo, A., & Dgani, R. 1995, AJ, 110, 2914 Gregorio-Hetem, J., Montmerle, T., Casanova, S., & Feigelson, E. D. 1998, Waldron, W. L. 1984, ApJ, 282, 256 A&A, 331, 193 Wang, Q. D. 1999, ApJ, 510, L139 Grosso, N., et al. 1997, Nature, 387, 56 Wichmann, R., Sterzik, M., Krautter, J., Metanomski, A., & Voges, W. Hester, J. J., et al. 1999, BAAS, 194.681 1997, A&A, 326, 211 Hillenbrand, L. A., et al. 1998, AJ, 116, 1816 Yamauchi, S., & Koyama, K. 1993, ApJ, 405, 268 Hofner, P., & Churchwell, E. 1997, ApJ, 486, L39 Yusef-Zadeh, F., Shure, M., Wardle, M., & Kassim, N. 2000, ApJ, 540, Howarth, I. D., & Prinja, R. K. 1989, ApJS, 69, 527 842