Submitted to ApJ Chandra Observation of the Trifid Nebula: X-ray emission from the O star complex and actively forming Pre-main sequence stars Jeonghee Rho and Solange V. Ram´ırez 4 SIRTF Science Center, Mail Stop 220-06, California Institute of Technology. 0 0 2 Michael F. Corcoran1 and Kenji Hamaguchi n a Code 662, NASA/Goddard Space Flight Center, Greenbelt, MD 20771 J 9 Bertrand Lefloch 1 1 Laboratoire d’Astrophysique, Observatoire de Grenoble, BP 53, F-38041, Grenoble CEDEX9, v France 7 7 3 ABSTRACT 1 0 4 0 The Trifid Nebula, a young star-forming HII region, was observed for 16 hours / h by the ACIS-I detector on board of the Chandra X-ray Observatory. We detected p 304 X-ray sources, thirty percent of which are hard sources and seventy percent of of - o whichhavenear-infraredcounterparts. Chandra resolvedtheHD164492multiplesystem r t into a number of discrete X-ray sources. X-ray emission is detected from components s a HD164492A (an O7.5III star which ionizes the nebula), B and C (a B6V star), and : v possibly D (a Be star). Component C is blended with an unidentified source to the i X NW. HD164492A has a soft spectrum (kT 0.5 keV) while the component C blend ≈ r shows much harder emission (kT 6 keV). This blend and other hard sources are a ≈ responsible for the hard emission and Fe K line seen by ASCA, which was previously attributedentirelytoHD164492A.ThesoftspectrumoftheOstarissimilartoemission seen from other single O stars and is probably produced by shocks within its massive stellar wind. Lack of hard emission suggests that neither a magnetically confined wind shock nor colliding wind emission is important in HD164492A. A dozen stars are found to have flares in the field and most of them are pre-main sequence stars (PMS). Six sources with flares have both optical and 2MASS counterparts. These counterparts are not embedded and thus it is likely that these sources are in a later stage of PMS evolution, possiblyClassIIorIII.Twoflaresourcesdidnothaveanynear-IR,optical, or radiocounterparts. WesuggesttheseX-ray flarestars areinanearly pre-mainsequence 1Universities Space Research Association, 7501 Forbes Blvd, Ste206, Seabrook, MD 20706 – 2 – stage (Class I or earlier). We also detected X-ray sources apparently associated with two massive star forming cores, TC1 and TC4. The spectra of these sources show high extinction and X-ray luminosities of 2 5 1031 erg s−1. If these sources are Class 0 − × objects, it is unclear if their X-ray emission is due to solar-type magnetic activities as in Class I objects, or some other mechanism. Subject headings: stars: activity — stars: pre-main sequence — X-rays: stars 1. Introduction Despite of the fact that X-ray observations of star-forming regions have been a powerful tool for discovering young stellar objects (YSO) and T Tauri stars (TTS) since Einstein (references in Feigelson & Montmerle 1999), long wavelength (IR and submillimeter) emission is used to define evolutionary classes for young stars. Classes I-III are solely based on the excess seen in the IR spectral energy distribution (SED) with respect to stellar blackbody photospheric emission, as measured by the spectral index α = d log (λ F ) / d log λ between λ = 2.2 and 10-25µm: IR λ Class I , II and III correspond to α > 0, -2 <α < 0, and α < -2 (Lada 1987). Class 0 IR IR IR stars represent the earliest phase of star formation and are identified by the ratio of submillimeter to bolometric luminosity (Andr´e et al. 1993) and are seen as condensations in submillimeter far- infrared dust continuum maps; these condensations often show collimated CO outflows or internal heating sources. X-ray bright T Tauri stars usually belong to either Class II or III. High resolution images obtained by the Chandra X-ray Observatory open a new era in the study of star formation because the 1′′ resolution of Chandra is necessary to resolve individual sources in nearby star forming regions, and because confusion due to foreground and background stars is much less important at X-ray energies than in the optical and near-infrared regime. Recent Chandra observations of Orion (Garmire et al. 2000; Feigelson et al. 2002) detected hundreds of X-ray sources, e.g. pre-main sequence stars (PMS) with masses in the range 0.05 M⊙ 50 M⊙ and − a combined infrared and X-ray study suggests that the X-ray luminosity of PMS dependson stellar mass, rotational history, and magnetic field (Garmire et al. 2000). A high percentage of Class I PMS were also found to be X-ray emitters; in ρ Oph, 70% of identified Class I stars are X-ray bright. Strong X-ray flares from PMS in ρ Oph (Imanishi et al. 2001), Monoceros R2 (Kohno et al. 2002), and Orion (Feigelson et al. 2002) were detected from Class I, II and III objects, possibly because of magnetic activity. Moreover, X-ray emission from Class 0 candidates was detected in OMC-3 (Tsuboi et al. 2001). The Trifid Nebula (M20) is one of the best-known astrophysical objects. It is a classical nebula ionized by an O7.5 star, HD164492A, and the ionized nebula glows in red light. The nebula is trisected by obscuring dust lanes giving the Trifid its name. A blue reflection nebula appears to the north of the red nebula. At an age of 3 105 years, the Trifid is a young H II region. ∼ × Recent studies using the Infrared Space Observatory (ISO) and the Hubble Space Telescope (HST) – 3 – (Cernicharo et al. 1998; Lefloch & Cernicharo 2000; Hester et al. 1999) show the Trifid to be a dynamic, “pre-Orion” star forming region containing young stars undergoing episodes of violent massejection, withprotostarslikeHH399 (Leflochetal.2002) losingmassandenergytothenebula in optically bright jets. Four massive (17–60 M⊙) protostellar cores were discovered in the Trifid from millimeter-wave observations. These cores are associated with molecular gas condensations at the edges of clouds (Lefloch & Cernicharo 2000). T Tauri stars and young stellar object candidates were also identified using near-infrared color-color diagrams from 2MASS data (Rho et al. 2001). Unlike better-studied nearby star-forming regions such as Orion and ρ Oph, star forming activities in the Trifid have only recently been recognized (Cernicharo et al. 1998), and as a result of this (and also due to contamination by foreground and background stars in the optical and IR), the population of PMS in the Trifid has not been fully investigated. The distance to the Trifid Nebula is between 1.68 and 2.84 kpc (Lynds et al. 1985; Kohoutek et al. 1999 and references therein) and a distance of 1.68 kpc is adopted in this paper. Inthispaper,wereportresultsofChandra observationsoftheTrifidNebula. Chandra resolved the HD 164492 multiple system into a number of discrete X-ray sources and we present their X-ray properties which include components A (an O star), C (a B6V star) and B (A2Ia). Our Chandra observations revealed 304 sources and we found that 30% of the sources have hard emission similar to that from PMS. Among these candidate PMS, we report properties of about a dozen flare sources, which include unusual variability from the O star and unusual emission from an A-type supergiant. We also discuss the X-ray properties of the HD 164492 complex, the properties of X-ray sources which are apparently associated with two protostellar cores, and also the properties of some apparently strongly variable objects. 2. Observations TheTrifid Nebula was observed with the Advanced CCDImaging Spectrometer (ACIS) detec- tor on board the Chandra X-ray Observatory (Weisskopf et al. 2002) on 2002 June 13. The results presented here arise from the imaging array (ACIS-I), which consists of four 1024 1024 front-side × illuminated CCDs. The array was centered at R.A. 18h02m28s and Dec. 22◦56′50′′ (J2000) and − covered an area in the sky of about 17′ 17′. The total exposure time of the ACIS observations × was 58 ksec. This observation is sensitive to X-ray luminosities of 5 1029 erg s−1, assuming an × appropriate PMS X-ray spectrum(temperature of 1 keV and an absorption 1.6 1021 cm−1) at the × distance (1.68 kpc) of the Trifid. We started data analysis with the Level 1 processed event list provided by the pipeline pro- cessing at the Chandra X-ray Center. The energy and grade of each data event were corrected for charge transfer inefficiency (CTI), applying the algorithm described by Townsley et al. (2000). The event file was filtered to include event grades of 0, 2, 3, 4, and 6, and filtered by time intervals to exclude background flaring intervals or other bad times. The filtering process was done using – 4 – the Chandra Interactive Analysis of Observations (CIAO) package1 provided by the Chandra X-ray Center. Figure 1 shows the Chandra ACIS-I threecolor image of the Trifid. Hundredsof point sources are detected with little diffuseemission. X-ray sources were located usingthe wavdetect tool within the CIAO package. This tool performs a Mexican hat wavelet decomposition and reconstruction of the image after accounting for the spatially varying point spread function as described by Freeman et al. (2002). We used wavelet scales ranging from 1 to 16 pixels in steps of √2, and a default source threshold probability of 1 10−6. The wavdetect tool was run using an exposure map and × it produced a catalog of 353 sources from the entire ACIS-I FOV. Then we identified false sources producedbycosmic rays orcases in whichthesourcecounts arebelow thebackgroundcounts. This observation finally resulted in 304 X-ray sources detected from the total ACIS-I FOV (17′ 17′). × The full source list is given in Table 1 in order of R.A. Thirty percent of these sources are shown to be hard (shown in blue in Figure 1); the hard sources have a spectral hardness ratio (SHR) > 0.2, where SHR is the ratio of the net counts in the hard 2.0 8.0 keV band to those in the soft − − 0.5 2.0 keV band. Diffuse emission was not obvious in the Chandra images of the Trifid Nebula − with our current data processing. None-detection of diffuse emission in M20 is consistent with a claim that high mass star-forming regions without stars earlier than O6 may be unlikely to exhibit diffuse soft X-rays (Townsley et al. 2003; Abbott 1982) The brightest source in the field corresponds to the O star HD164492A (source 102) and has 884 counts. This is equivalent to 0.047 photons per frame, which is small enough so as not to be affected by pileup effects (the pileup fraction is << 0.05). In order to verify the hard emission and Fe K line detected in the ASCA spectra of HD164492A (Rho et al. 2001), we extracted a spectrum usingthesameextractionregionaswasusedintheASCAanalysis,andconfirmedthattheChandra and ASCA spectra are the same within the errors. However, we found that HD164492A is actually asoftsource(see section 6for details)whilethehardemission andFeKline(which wereattributed to the O star in the ASCA analysis) is actually produced by the hard sources resolved by Chandra as shown Figure 2. In addition, the emission previously attributed entirely to HD164492A in an analysis of ROSAT PSPC data (Rho et al. 2001) is now resolved into a dozen X-ray sources as shown in Figure 1 and 2. To confirm the X-ray astrometry, we used the 2MASS2 final release point source catalog to search for near-IR counterparts to the X-ray sources in the full source list. We cross-correlated the positions of the 2MASS sources with the positions of our X-ray sources. Within 5′ of the Chandra aimpoint, we found 72 2MASS sources that coincide with X-ray source positions. We computed the total R. A. and DEC offset of these sources and obtained offsets of 0.16” in R. A., and 0.09” − − in DEC, which are less than one third of a pixel. Therefore, the astrometry of Trifid ACIS image 1http://cxc.harvard.edu/ciao/index.html 2http://www.ipac.caltech.edu/2mass – 5 – was confirmed with the minor systematic errors in the Chandra aspect solution. 3. The Multiple System HD164492 The HD164492 complex is a multiple stellar system composed of 7 physically related compo- nents, A-G (Kohoutek et al. 1999). Component HD164492A is the O7.5 III((f)) star responsible for ionizing the nebula. This complex was observed as an unresolved bright X-ray source in PSPC images (Rho et. al. 2001). Our Chandra images (Figures 1 and 2) resolve this complex into a number of discrete X-ray sources. X-ray emission is clearly detected from components A, B, and C as shown in Table 2 (Fig. 2). HD164492A is the brightest X-ray source in this complex. Compo- nent HD164492C (a B6V star, Gahm et al. 1983) is a bright X-ray source (source 94) and is barely resolved from a nearby X-ray source to the NW of C (source 91) which has no obvious optical counterpart (Fig. 2). Source D (a Be star) is a strong H-α emitter (Herbig 1957) and has a disk (resolved by a radio image), similar to proplydsin the Orionnebula(O’Dell 2001). The component D is much fainter in X-rays than component C. 3.1. HD 164492A The X-ray light curve of the HD164492A shows small but significant variability ( 20% in the ∼ 10 hour observation, Figure 4). This is unsual since most OB stars are not known X-ray variables. X-ray variability is notwidely known from early typestars based on previous Einstein and ROSAT surveys (Chlebowski et al. 1989; Bergho¨fer et al. 1996). However, recent Chandra observations show significant variability from some early type stars (Feigelson et al. 2002). For example, θ2 OriA, an O9.5 star, shows a 50% drop in 10 hr with multiple 10-20% flares (Feigelson et al. 2002). Some level of time variability is also observed from other early type stars in Orion (Schulz et al. 2001). The spectrum of the O star shows thermal emission as shown in Fig. 3a. We fit the X-ray spectra of HD164492A using two collisional ionization equilibrium (CIE) thermal models (a Mewe- Kaastra plasma model (Kaastra 1992) and an updated Raymond-Smith model (Raymond & Smith 1977); MEKAL and APEC in XSPEC). Both models gave similar results. We assumed sub-solar metal abundance (0.3 solar); assuming solar abundances produced less than a 20% change in the derived spectral parameters. The best fit of the spectrum of O star yields N = 1(< 5) 1021 H × cm−2 and a temperature of kT = 0.5(< 0.6) keV, which implies an X-ray luminosity of 2.8 1031 s × erg s−1. The derived temperature is similar to the X-ray temperatures of other single massive stars (Corcoran et al. 1994; Moffat et al. 2002). Although some of the luminosity may be due to an unresolved low-mass companion such as a T-Tauri star which could be responsible for the variability, the low temperature obtained from the spectral fit is largely consistent with a single massive star. Assuming a bolometric luminosity of L 0.5–1.6 1039 ergs s−1, the ratio of X-ray bol ∼ × to bolometric luminosity is 7.76 < logL /L < 7.25, which is smaller than the canonical value x bol − − L /L 7 but it is within the scatter (Bergho¨fer et al. 1996). The lack of high temperature X- x bol ≈ − – 6 – ray emission in HD164492A suggests that a magnetically confined wind shock has not developed in theOstar of HD164492A as hasbeensuggested in other variableOstars showinghightemperature X-ray emission (like θ Ori A, C, and E Schulz et al. (2001)). Similarly the lack of hard emission in HD164492A also suggests that colliding wind emission (as considered in Rho et al. (2001)) is not important. X-rays from HD164492A are likely produced by shocks distributed through the star’s massive stellar wind. The variability may be from instabilities within a radiatively driven wind (Lucy & White 1980; Feldmeier et al. 1997), but if so the rapid timescale of the changes ( 2 ∼ hours) suggests that the X-ray emission is dominated by a small number of strong shocks rather than a large distribution of weak shocks. 3.2. HD164492C WealsomodelledtheextractedspectrumfromsourceC,whichisblendedwiththeunidentified X-ray source to the NW (source 91: Component C2) which is barely resolved from source C. The component D, a Be star, is not identified as a source by wavelet analysis; however, there is some faint X-ray emission coinciding with the Be star as shown in Figure 2. Component C2 (source 91) shows variability, and the light curve is shown in Fig. 4. The best two-temperature absorbed thermal fit in Fig. 3b has N (1) = 7.8+5.2 1021 cm−2 and a temperature of kT = 0.6 0.4 keV, H −4.8× 1 ± and N (2) = 1.6(+4.4) 1021 cm−2 and a temperature of kT = 5.9(+∞ ) keV. The luminosity of H −1.1 × 2 −3.4 soft and hard components are 1.4 1032 and 6 1031 erg s−1, respectively. The lower absorption × × N (1) = 1.6 1021 cm−2 is comparable to the optical extinction A = 1 mag (Kohoutek et al. H v × 1999), suggesting that the wind(s) and the strong ionizing radiation from the O and B stars strip the surrounding materials. The 6 keV component is partially responsible for the spectra seen by ASCA (Rho et al. 2001). The X-ray emission from B and Be stars is not well characterized at present since they are often faint. The L /L ratio is below 10−7 10−5 for stars of spectral type x Bol ∼ B1.5(III-V)andlater(Feigelson etal.2002), andthelatetypestarshavealuminosityrangeof1029- 1031 ergs−1. ThustheobservedX-rayemissionof2 1032ergs−1 fromthesourceCblendisdifficult × to reconcile with emission from a typical B6V type star or a late type main-sequence companion star. We extracted separate spectra from both components C and C2, but the spectra could not be distinguished from each other and additionally we lose the hard component emission dueto lack of photon statistics in the hard energy band, although separate light curves show stronger time variability in the component C2. The B6V star is an X-ray source as shown in Figure 2, and the component C2 is a companion candidate which is also strong X-ray emitter. The companions of later B-type stars with strong X-ray emission, are suggested to be often PMS (Stelzer et al. 2003), which show characteristics of high X-ray luminosities and hard X-rays. The presence of an unidentified X-ray source near the B6V star and the presence of hard X-ray emission in this system suggest that C2 may be a PMS. Other possibilities such as emission from corona of the Herbig Be star, or a mechanism related to the proplyds remain open at this point. – 7 – 3.3. HD 164492B Component B (source 106) is classified as A2Ia (Gahm et al. 1983) or possibly as an A2 III (Lindroos 1985). A-type stars are not often detected as X-ray sources (Guillout et al. 1999) since they lack subsurface convection zones to power strong magnetic fields, and they lack a strong UV flux to power massive stellar winds. This may indicate that this star has a low-mass, X-ray bright companion or the spectral type may be incorrect. Its X-ray luminosity is listed in Table 2; we assumed a temperature of 0.15 keV from other A stars (Simon & Drake, 1993) and the absorption of 1.6 1021 cm−2. × 4. Pre-main Sequence star candidates We created lightcurves for each of the detected X-ray sources, using a bin length of 2500 sec, which gives a total number of 24 bins. We calculated the ratio between max count in the highest bin and minimum count in the lowest bin and found 40 sources have theratio greater than 1 after ∼ accounting for the errors. If a source was only detected duringa flare, we used the mean count rate instead of minimum count rate. In Table 2 we list sources in which this ratio greater than 3 (sources 8, 23, 97, 166, 170, 194, 211, 237, 246, 256, 283and285). Thesesourcesshouldrepresentthemoststronglyvariablesources. Toconfirmthisvariability wealsoperformedasimpleχ2 test against anassumedtheconstant light curve. The probability of constancy for each source from this test is given in Table 2, for which we used additional binning as needed for each source. Nine sources (sources 23, 91, 97, 166, 194, 237, 256, 283 and 285) in Table 2 are definitely variable (null probability < 10%) based on the χ2 test and our ratio test. We show sample light curves of the variable sources in Fig. 4. We searched for near-infrared and optical counterparts for these variable sources. Unfortu- nately no previous optical identification of T Tauri stars exists in the Trifid. The only identified T Tauri stars or young stellar objects are 85 PMS candidates identified from infrared excesses using 2MASS second release data (Rho et al. 2001). Here we re-evaluate the identification of PMS candi- dates –TTS or YSO– from near-infrared color-color diagrams using the 2MASS final release data. The final release data have improved photometry by allowing multiple-components of point spread function for the blended sources with the improved zero point, and a complete source catalog by thoroughly identifying artifacts. Among variable sources in Table 2, only source 283 was previously identified as a YSO (Rho et al. 2001) and now sources 23, 97, 166, 194 and 237 are identified as TTS or YSO based on the 2MASS data. We also searched for optical counterparts to these X-ray sources using the Guide star catalog (GSC 2.2) and USNO-B catalog, but since these catalogs do not include any sources near bright optical diffuse emission, we also directly compared the X-ray sources in Table 1 to digital all sky images (DPOSS). Among the variable sources listed in Table 2, five (source 97, 166, 237, 256 and 285) have SHR > 0.2 (and they appear blue in Figure 1). − – 8 – Sources 8, 23, 194, 211, 246 and 283 have both optical and 2MASS counterparts. These sources are not embedded and thus are likely in later stage of PMS evolution, possibly Class II or III. Sources 8, 211 and 246 have 2MASS counterparts, but not identified as TTS or YSO from near-infrared colors. They are still likely PMS, because the near-IR excess is shown to be time- dependent (Carpenter et al. 2000). Sources 97, 166, 237, 256 and 285 have no optical counterparts and have SHR > 0.2 and the light curve of source 166 is shown in Fig. 4. In particular, source − 256 and 285 have neither near-IR nor optical counterparts but exhibit flares in their X-ray light curves as shown in Fig. 4. No radio counterparts are known for these sources. We suggest these X-ray flare stars are in an early pre-main sequence stage (Class I or earlier). Source 211 has sufficient counts for spectral fitting, and the best-fit using a one-temperature thermal model yields N =1(< 8) 1021 cm−2 and a temperature of kT = 2 (>1) keV. Theinferred X-ray luminosity of H × 1.1 1031 erg s−1, is comparable to those from TTS and YSO (Feigelson et al. 1999), and brighter × than those of typical low-mass main-sequence stars ( 1030 erg s−1). The sources with flares seem ∼ to be found preferentially concentrated along the dust lane and at the edge of the HII region, i.e. along the ionization fronts. This reinforces the conclusion that most of flaring sources are PMS. We estimated the X-ray luminosities of sources 8, 23, 97, 166, 170, 194, 246, 256 and 283 assuming the spectral parameters are the same as those of source 211. For sources 237 and 285, we assumed the spectral parameters are the same as those of TC1 source, because their hardness ratios are comparable to that of TC1. 5. Identification of X-ray Emission from Massive Protostellar Cores We compared the X-ray sources with the known massive star forming cores detected in the 1300µm dust continuum map (Lefloch et al. 2001). There are five such cores, four of which are known to have bipolar wings. We detected X-ray emission from TC1 (source 117) and possibly TC4 (source 38), two cores with bipolar wings and associated Class 0 candidates (Cernicharo et al. 1998; Lefloch & Cernicharo 2000). The sources associated with both TC1 and TC4 are hard X-ray sources (which appear as blue sources in Figure 1). Figure 5 shows the positions of the X-ray sources superposed on the 1300µm dust continuum map. The X-ray source associated with TC1 has no 2MASS counterpart. But the X-ray identification of the source associated with the TC4 core is less clear, because this source is located at the edge of the ACIS field of view where the width of PSF is 12′′. The TC4 source coincides with the 1300µm dust continuum source ∼ within 11 12′′, and also has two 2MASS counterparts within 12′′. The closest 2MASS source − is 2MASS180212.5-230549 for which J band emission is not detected and H(=14.8mag) and K s (=12.9mag) magnitudes are contaminated by a nearby star. It is not clear if this 2MASS source is related to the TC4 and the X-ray source. TC4 satisfies the Class 0 definition of Andre et al. (1993) in terms of its dust temperature ( 20 K), the ratio of its submillimeter to bolometric luminosity, ∼ and the presence of a bipolar outflow. An outflow from TC4 shows a kinematic age of 6.8 103 yr × and suggests it to be an intermediate- or high-mass object (Lefloch et al. 2002). For TC1, the SED is not available so it’s not clear if TC1 satisfies the Class 0 criteria or not; however, the object is – 9 – deeply embedded in a dust core so it is at least a Class I object, and possibly younger than a Class I source. TC1 and TC4 have hardness ratios of 0.24 and 0.67, respectively (see Table 1 and 2). The source associated with TC4 is one of the hardestsources in Table 1. Thesesources are either highly absorbedor have high temperatures. Ouranalysis of theX-ray spectraof TC1andTC4showsthat the best fit thermal model parameters are N =6 1022 cm−2, kT = 1.7 keV and L = 4.7 1031 H x × × erg s−1 for the TC 4 source, and N = 3.4 1022 cm−2, kT = 1.1 keV, and L = 1.9 1031 H x × × erg s−1 for the TC1 source. The large column densities we derive show that these sources are highly embedded. The only detection of X-ray emission from a Class 0 object until now is a source located in OMC-3 (CSO6) (Tsuboi et al. 2001). In comparison, the Class 0 source in OMC-3 (CSO6) has N = (1 3) 1023 cm−2 and a luminosity of 1030 erg s−1. X-ray emission from Class H − × 0 objects which are in the dynamical infall phase is poorly understood since so far only a few Class 0 candidates have been identified and their properties are not well constrained. It is unclear if the X-ray emission is due to solar-type magnetic activities as in Class I objects. The fact that these two counterparts have high hardness ratios does not supportthe X-ray emission is from a low-mass companion to the protostellar core. If these objects are accreting, then the X-ray emission may be related to the accumulation and release of angular momentum toward the growing central star by outflow processes. 6. Summary YoungHIIregionsliketheTrifidarerichsourcesofX-rayemitters. OurChandra imagesreveal a few hundred X-ray sources including variable and hard sources, along with pre-main sequence stars and more evolved OB stars. We summarize our findings here. 1. Chandra images show 304 X-ray sources; thirty percent of the sources are hard, and two- thirds have near-infrared counterparts. The full list of Chandra X-ray sources is given in Table 1. 2. The multiple star system HD164492 is resolved for the first time in X-rays into individual components. X-ray emission is detected from components A, B, C (a B6V star), which is blended with an unidentified source in the Chandra images. This blend has comparable X-ray brightness to the O star. The O star HD 164492A shows small but significant variability and has a soft spectrum with a temperature of 0.5 keV. The temperature is comparable to those of other single massive stars and theratio of X-ray and bolometric luminosities is smaller than the canonical value L /L 7 but it is within the scatter of distribution. The lack of any hard component suggests x bol ≈ − that neither a magnetically confined wind shock nor colliding wind shock is needed to describe the X-ray emission from the O star. The variability of the X-ray emission implies that the emission is produced by a small number of strong shocks in the wind of HD 164492A. 3. TheX-rayspectrumfromthecomponentCblendrequiresatwo-temperaturethermalmodel – 10 – with kT = 0.6 0.4 keV and kT = 5.9(+∞ ) keV. Theinferred X-ray luminosity is 2 1032 erg s−1. 1 ± 2 −3.4 × This blend is highly variable in X-rays which suggests that one of the stars dominates the emission. 4. We found a dozen stars which show evidence of flaring activity and there could be many as 40 variable stars in the full source list. Nine sources (sources 8, 23, 97, 166, 194, 237, 256, 283 ∼ and 285 in Table 2) have significant variability based on the χ2 statistics. We searched for their near-infrared and optical counterparts, and found that six stars have both optical and 2MASS counterparts. These sources are likely in later stages of PMS evolution. Four sources which have nooptical counterpartsandhaveSHRgreather than 0.2arelikely inearlystages ofPMS,possibly − Class I or earlier. There are a few stars with neither near-IR nor optical counterpart whose light curves show strong flares, suggesting that they are very early stage pre-main sequence stars. 5. We detected X-ray emission from TC1 and possibly TC4, two massive star forming cores with bipolar wings and associated Class 0 candidates. Both TC1 and TC4 show extremely hard X-rays and their spectra imply very high absorption (N =3.5 - 6 1022 cm−2) and high luminosity H × (2 5 1031 erg s−1). Only one other detection of X-ray emission from a Class 0 object has been − × previously reported. Thusourtwo detections imply thesecond and thirdX-ray detection of aClass 0 object. It is unclear if the X-ray emission of these objects is due to solar-type magnetic activities as in Class I objects. If the X-ray emission is from the accreting stage, the X-ray emission may be related to the competing processes of accumulation of angular momentum toward the growing central star and release of angular momentum by outflow processes. ThirtypercentoftheX-ray sources inthefullfieldareshowntobehard(SHR> 0.2) sources − (shown in blue in Figure 1), and 16 percent of sources are extremely hard sources (SHR > +0.2). A high proportion of these sources are probably PMS because they are either highly embedded or extemely high temperature. These hard sources along with the sources near HD 164492C are responsibleforthehardspectraseenbyASCA.Theaccurate positionsfromthecomplete listof the X-ray sources are provided in Table 1. These sources provide an opportunity to identify interesting PMS using the multi-wavelength follow-up observations, which will help us to further understand the population and evolution of protostellar objects. Partial support for this work was provided by NASA through Chandra grant G02-3095A. J. R. and S. V. R. acknowledge the support of California Institute of Technology, the Jet Propulsion Laboratory, which is operated under contract with NASA.