Skip to main content

Full text of "X-ray Emission from an Expanding Supergiant Shell in IC 2574"

See other formats

accepted for publication in the ApJ Letters 

X— ray Emission from an Expanding Supergiant Shell in IC 2574 

Fabian Walter 

Radioastronomisches Institut der Universitat, Auf dem Hiigel 71, D-53121 Bonn, Germany 

Electronic mail: 

Jiirgen Kerp 

Radioastronomisches Institut der Universitat, Auf dem Hiigel 71, D-53121 Bonn, Germany 

Electronic mail: 

Neb Duric 

Department of Physics and Astronomy, University of New Mexico, Albuquerque, NM 87131 

Electronic mail: 

Elias Brinks 

Departamento de Astronomia, Apartado Postal 144, Guanajuato, Gto. 36000, Mexico 
National Radio Astronomy Observatory, P.O. Box O, Socorro, NM 87801 
Electronic mail: 


Uli Klein 

Radioastronomisches Institut der Universitat, Auf dem Hiigel 71, D-53121 Bonn, Germany 

Electronic mail: 


We present a multi-wavelength study of a supergiant shell within the violent 
interstellar medium of the nearby dwarf galaxy IC2574, a member of the 
M81 group of galaxies. Neutral hydrogen (H I) observations obtained with 
the Very Large Array (VLA) reveal a prominent expanding supergiant H I 
shell in the northeast quadrant of IC 2574 which is thought to be produced by 
the combined effects of stellar winds and supernova explosions. It measures 
roughly 1000 x 500 pc in size and is expanding at about 25 kms" 1 . The H I 
data suggest an age of about 1.4 x 10 6 yrs; the energy input must have been 
of order (2.6 ± 1) x 10 53 ergs. Massive star forming regions, as traced by 
Ha emission, are situated predominantly on the rim of this H I shell. This 

- 2 - 

supports the view that the accumulated H I on the rim has reached densities 
which are high enough for secondary star formation to commence. VLA radio 
continuum observations at A 6 cm show that these star-forming regions are 
the main sources of radio continuum emission in this galaxy. This emission 
is mainly thermal in origin. Soft X-ray emission from within the H I hole is 
detected by a pointed ROSAT PSPC observation. The emission is resolved, 
coinciding in size and orientation with the H I shell. These spatial properties 
suggest that the emission is generated by an X-ray emitting plasma located 
within the H I shell although a contribution from X-ray binaries cannot be 
completely ruled out. The X-ray luminosity within the 0. 11-2.4 keV energy 
range is Lx = (1.6 ± 0.5) x 10 38 ergss _1 . The X-ray data are compatible with 
emission coming from a Raymond & Smith (1977) plasma at a temperature of 
about log(T[K]) = 6.8 and a density of 0.03 cm -3 . The energy content of the 
coronal gas corresponds to (4 ± 2) x 10 53 ergs, or broadly in agreement with the 
energy input derived on the basis of the H I observations. 

Subject headings: galaxies: individual (IC2574) — galaxies: ISM — ISM: 
bubbles — ISM: structure — X-rays: ISM 

1. Introduction 

The interstellar medium (ISM) of galaxies is shaped by the aftermath of recent, massive 
star formation (SF). Huge cavities filled with coronal gas are believed to be blown in the 
ambient medium through strong stellar winds of the most massive stars as well as through 
subsequent Type II supernova (SN) explosions of stars more massive than ^8 M (see 
Tenorio-Tagle & Bodenheimer 1988 for a review). The structures that form during such 
processes are often referred to as bubbles (diameter ~10 pc), superbubbles (~200 pc) or 
supergiant shells (~1 kpc). A recent development is the finding that supergiant shells reach 
much larger sizes in dwarf galaxies than in more massive spiral galaxies (Puche et al. 1992, 
Walter & Brinks 1998 - hereafter referred to as WB98). This is attributed to dwarfs 
having a lower gravitational potential, which allows shells to grow more easily, and to the 
absence of differential rotation and spiral density waves, which prevents them from being 
destroyed prematurely. Dwarf galaxies are therefore ideally suited to study the origin and 
evolution of supergiant shells. 

Cox & Smith (1974) not only realized that the energy input of SNe creates cavities 
filled with hot gas, but that the cooling time is relatively long (10 6 to 10 9 years, depending 

- 3- 

on density and temperature of the matter within), allowing the creation of a tunneling 
network (the swiss cheese model). Weaver et al. (1977) discussed a model in which fast 
stellar winds are adiabatically shocked to temperatures of order 10 6 -10 7 K. These winds 
then act like a piston, driving the expansion of the outer shell of swept-up ambient material. 
The basic picture still stands today, although several refinements have been proposed, e.g., 
by Chu & Mac Low (1990) who calculate the expected X-ray emission from superbubbles. 

The ISM that is swept up accumulates in the form of an expanding shell and forms the 
rim of a superbubble or supergiant shell. The resulting shocks on their rims may induce 
new generations of SF, or SF regions may be created simply via gravitational fragmentation 
(Elmegreen 1994, for an early discussion see Mueller & Arnett 1976). 

Direct observational evidence for the theoretical scenarios mentioned above is still 
scarce. Very few, if any, supergiant shells have been detected to date through the emission 
of the coronal gas (see Bomans et al. 1996 for a review). In fact, it is the neutral and 
ionized gas which has proven to be the best tool to trace the outlines of superbubbles. The 
best direct evidence has come from ROSAT PSPC observations, giving information about 
the plasma properties of the interior of some superbubbles. Examples are the supergiant 
shell LMC4 (Bomans, Dennerl & Kiirster 1994), the superbubbles N44 (Kim et al. 1998) 
and Nil (Mac Low et al. 1998), all three situated in the Large Magellanic Cloud (LMC), 
the supergiant shell SGS2 in NGC4449 (Bomans, Chu & Hopp 1997) and the possible 
supershell near HolmberglX (Miller 1995). In the following we will present the probable 
detection of coronal gas filling a supergiant shell within the nearby (3.2 Mpc) dwarf galaxy 

2. Observational Evidence for a Hot— Gas filled expanding Supergiant Shell in 

IC 2574 

2.1. HI observations 

The neutral interstellar medium of IC 2574 has been extensively studied by WB98 using 
H I observations obtained with the NRAOQ Very Large Array. An H I surface brightness 
map of IC 2574 is shown in Fig. 1 a. WB98 find a total of 48 H I holes in IC 2574 most 
of which are expanding. The most prominent of these holes (No. 35 in their study) is the 
subject of this letter. An H I map of the region around this hole is shown as a blowup 

1 The National Radio Astronomy Observatory (NRAO) is operated by Associated Universities, Inc., under 
cooperative agreement with the National Science Foundation. 


in Fig. 2 a. The size of the hole is RdOOO pc x 500 pc (which corresponds to an effective 
radius r c g « 350 pc) and its radial expansion velocity is ~25 kms -1 . The dynamical age 
is therefore about 1.4 x 10 7 years. WB98 show that the lcr-scale height of the H I disk is 
approximately 350 pc, or of the same order as the radius of the hole. Hence, the hole is still 
contained within the disk (no material lost due to blow-out of the gas into the halo). Given 
the derived scaleheight and ambient average H I column density of Ahi ~ 10 21 cm -2 , the 
ambient H I volume density before the creation of the hole is estimated at n H i = 0.5 cm -3 . 
The H I mass that was present before the evacuation of the H I hole which now must be 
largely accumulated on the rim is therefore about 2 x 10 6 M . Using these numbers and 
correcting the H I mass for the contribution of primordial helium, we estimate the kinetic 
energy of the expanding shell to be « 1.7 x 10 52 ergs. 

No signature of an infalling high-velocity cloud (HVC) is found in the H I data and 
WB98 propose an internal origin for the creation of the hole. Using the numerical models 
of Chevalier (1974), who calulates the late phase of supernova remnants in which the shell 
accretes matter from both the interstellar medium and the hot interior, an energy input 
of (2.6 ± 1) x 10 53 ergs by a star forming region is needed to create this hole. Note that 
the kinetic energy of the hole is of order 10% of this value, which is about half the value 
originally predicted by Weaver et al. (1977). This energy is equivalent to about 100 Type 
II SNe and the strong stellar winds of their progenitors; a major SF event. 

2.2. Ha observations 

In order to trace regions of current SF activity, Ha maps were obtained by WB98. 
A greyscale representation of their Ha map is given in Fig. 1 b. Although the global SF 
activity in IC2574 is rather low, a ring of H II regions is prominent in the northeast 
quadrant of the galaxy. A blowup of the Ha emission around the same region as in Fig. 2 a 
is given in Fig. 2 b. Note that virtually all Ha emission in that region is situated on 
the rim of the H I shell (as first noticed by Martimbeau, Carignan & Roy 1994). Three 
Wolf-Rayet stars have been detected in the most luminous H II region on the northern 
rim by Drissen, Roy & Moffat (1993). IUE spectra through the same region show the 
presence of P Cygni profiles indicating the presence of strong stellar winds (Rosa, Joubert 
& Benvenuti 1984). Using the calibration by Miller & Hodge (1994), the total Ha flux on 
the rim is 2.3 x 10~ 12 erg cm" 2 s _1 which corresponds to a luminosity of L(Ha) = 7 x 10 5 
L (L = 3.85 x 10 33 erg s" 1 ) or an energy equivalent of about 50 stars of spectral type 05 
(WB98). This is almost one third of the Ha luminosity of 30 Dor, which is believed to be 
the most active SF region in the Local Group (Chu & Kennicutt 1994). The average volume 

- 5 - 

densities derived on the rim of the shell are 3 cm" 3 for the neutral H I (southern part) 
and 2 cm" 3 for the ionized hydrogen in the northern part. This suggests that most of the 
neutral gas component in the northern part of the hole has been ionized by the prominent 
SF regions. 

2.3. Radio Continuum observations 

Radio continuum maps of IC 2574 were obtained at A 6 cm using the VLA D-array. 
A map of the A 6-cm emission is presented in Fig. 2 d. We detect four radio sources in 
the region of the supergiant shell; their positions and flux densitities (in parentheses) are 
B102459.7+684320 (S = 0.9mJy), B102503. 1+684348 (S = l.lmJy), B102507.5+684325 
(S = 0.8mJy) and B102507. 7+684346 (S = 0.8mJy). The flux densities are accurate to 
within 20%. Three of the four sources coincide with regions of intense Ha emission on the 
rim of the shell. These sources have Radio/Ha flux ratios that are consistent with the radio 
emission being purely thermal (Caplan & Deharveng 1986). This result confirms that the 
Ha emission traces photo-ionized gas and that these sources are indeed H II regions and 
not shock-ionized regions due to the expansion of the hole. 

The source furthest to the southwest, B102459. 7+684320, is clearly nonthermal (see 
Fig. 2 d) and its radio emission is well in excess of what one would expect from an H II 
region on the basis of the rather weak Ha emission. It most likely is a supernova remnant 
given its nonthermal nature, its appearance as a point source and a flux density which is 
consistent with the brighter SNRs found in nearby galaxies (e.g., Duric et al. 1995; Lacey 
et al. 1997). 

2.4. X— ray observations 

In order to examine the X-ray properties of IC 2574, we have analysed a ROSAT PSPC 
observation towards IC2574 (US600145P, total integration time: 7.3 ks). The PSPC point 
spread function at E — 1.23 keV is about 28" (FWHM) across the area of interest. The 
pointed ROSAT PSPC observation reveals a significant X-ray feature which is coincident 
with the very center of the H I hole. A contour map of the X-ray emission is overlaid 
on the Ha-emission in Figs, lb (whole galaxy) and 2b (blowup). An overlay with the 
H I is presented in Fig. 2 c. This X-ray feature has an angular extent of about 45" x 30" 
(FWHM), with its major axis aligned with the H I hole. Note that the emission is extended 
even if we account for the residual pointing uncertainties of the ROSAT telescope of about 

- 6- 

10" (Hasinger 1997, priv. comm.). In addition, ROSAT's wobble direction during the 
observations was perpendicular to the orientation of the major axis of the H I hole. Based 
on the ROSAT PSPC observations we therefore find evidence for extended X-ray emission 
originating from the interior of the H I hole. However higher resolution observations will be 
needed to confirm this. 

IC2574 is located at a high galactic latitude (b ~ +43°), which means that the 
photoelectric absorption due to the Galactic interstellar medium is moderate. The 
Leiden/Dwingeloo H I A21-cm line survey (Hartmann & Burton 1997) gives an H I column 
density of Am = 2 x 10 20 cm^ 2 after separating the Galactic diffuse H I emission from that 
of IC2574. Because of the apparently clear line of sight through the hole (see Fig. la) we 
considered it justified to neglect any contribution to the H I column density within IC 2574. 

To separate the source counts from that of the diffuse X-ray background, we performed 
an on/off analysis. This method also removes the contribution of the so-called non-cosmic 
X-ray background radiation (Snowden et al. 1994). In order to obtain quantitative 
information about the spectrum of the X-ray feature, we evaluated the standard ROSAT 
hardness ratios (HR1 and HR2), which range from -1 (soft) to 1 (hard). We took this 
approach because the number of photons is too low for a standard analysis of the X-ray 
spectrum. We divided the ROSAT energy band into the three main energy bands, denoted 
as ROSAT C-band (0.25 keV, PI channels 11-41) the M-band (0.75 keV, 52-90) and 
J-band (1.5 keV, 91-201). The evalution of HR2 which is nearly unaffected by galactic 
absorption yields a value of HR2 = = 0.2 ± 0.2. This leads to a first estimate of the 

temperature of log(T[K]) pa6.8. Adopting this estimate we can correct the soft bands for 
photoelectric absorption by the Galactic ISM (Snowden et al. 1994). This analysis yields 
HR1 = (j+m+c) — 0-2 ± 0.4 and the following photon counts for energies below and above 
1 keV: I(E < lkeV,ll - 90) = 106 ± 39cts and J(E > 1 keV, 91 - 201) = 38 ± 9cts. 
This simple approach indicates that the X-ray feature emits most of its photons below 
E — 1 keV. Combined with the fact that the source is elongated, this suggests that a high 
mass, or even low mass X-ray binary is unlikely to be the source of the detected emission. 
It should be noted though that a possible contribution from X-ray binaries to the observed 
X-ray emission cannot be ruled out completely at this point. 

Using the hardness ratios and assuming a reduced metallicity of about 15% solar (as 
derived in a few H II regions in IC2574 by Miller & Hodge 1994), we calculate a plasma 
temperature range, according to Raymond & Smith (1977), of log(T 15 %[K]) = 6.9-7.1. If 
the X-ray plasma within the H I hole contains more metals the temperature will be lower 
(log(T solar [K]) = 6.4-6.8). Note that a higher abundance is more likely since the interior of 
the shell has recently been enriched by the heavy elements ejected by the SNe that created 


the hole. As a result the true temperature will lie somewhere in between; we adopt a 
temperature of log(T[K]) = 6.8 ± 0.3 for the remainder of this paper. 

For a plasma of log(T[K]) = 6.8 ±0.3 and a reduced metallicity we evaluate an emission 
measure of EM = (0.65 ± 0.15) cm _6 pc. Assuming an extent of the line of sight through 
the X-ray cavity of 700 pc, we derive an electron density of n e = (0.03 ± 0.01) cm -3 . This 
value is similar to the densities found in the supergiant shell SGS2 in NGC4449 (0.03 cm -3 ) 
or in the LMC4 (ps 0.01 cm" 3 ) (Bomans et al. 1997). The total X-ray luminosity of the 
interior of the hole is Lx(0.28 — 2.4 keV) = (1.6 ± 0.5) x 10 38 ergss _1 which lies in between 
the values for SGS2 in NGC4449 («7x lO^ergss -1 ) and LMC4 (« 0.15 x 10 38 ergss" 1 ). 
This luminosity is about one order of magnitude higher than predicted by theoretical 
models (see, e.g., the calculations by Chu et al. 1995), as is the case with most X-ray 
luminosities determined thus far (Nil being the only exception to date - see the discussion 
by Mac Low et al. 1998). 

3. Discussion 

If we assume that the supergiant shell is the result of a recent SF event, the picture 
which emerges from the observations discussed above is the following: about 1.4 x 10 7 years 
ago (the dynamical age of the H I hole), a major SF event took place at the center of what 
today shows up as a prominent H I hole. Since its creation some 2.6 x 10 53 ergs of energy 
have been deposited by the most massive stars, into the ambient ISM. About 10% of this 
energy is now still present in the form of the kinetic energy of the expanding shell. Note 
that the least massive stars that go off as SN are most probably still present in the cavity 
since their lifetime (^5x 10 7 years) is somewhat longer then the dynamical age of the hole. 
In other words, there is still some energy input going on today which might explain why 
the derived temperature for the interior of the H I hole is at the upper end for supergiant 
shells. On the other hand, a larger than expected contribution from X-ray binaries 
may also explain the high temperature estimate. Adopting a sound speed of 100 kms -1 
for the coronal gas filling the hole, we derive a time of only « 3 x 10 6 years before a 
soundwave actually reaches the rim of the shell. The gas has therefore had enough time to 
establish a relatively uniform distribution within the cavity. For a plasma temperature of 
log(T[K]) = 6.8 ± 0.3 and an internal density of (0.03 ± 0.01) cm -3 we derive an internal 
pressure of P = 2 n c T « (4 ± 2) x 10 5 Kcm" 3 (LMC4: P = 2 x 10 4 Kcm" 3 ). This value 
has to be compared to the much lower pressure of the ambient warm ionized medium 
(P « 10 3 — 10 4 Kcm~ 3 ). This means that it is probably this hot gas which is still driving 
the expansion of the shell (see, e.g., Weaver et al. 1977). The shell, as it is sweeping up 

- 8- 

material, has already reached the point at which secondary star formation can take place, 
as evidenced by the H II regions located along its perifery. The energy of the hot-gas 
interior can be approximated by | (2n c ) kTV and yields a value of (4 ± 2) x 10 53 ergs. 
The model by Weaver et al. (1977) predicts that about 50 % of the total energy deposited 
by SNe goes into the thermal heating of the gas interior a superbubble. Our values indicate 
that, within the errors, indeed most of the energy deposited by the SN explosions has gone 
into the heating of the gas. Note that a small decrease in temperature and increase in the 
ambient density would bring these numbers in even closer agreement. In fact, it is the 
broad agreement of the observed numbers with those derived from theory which gives us 
added confidence that it is indeed hot diffuse gas that fills the cavity rather than a cluster 
of X-ray binaries. A definitive answer will have to await the better spectral and spatial 
resolution of future X-ray missions, such as AXAF and XMM, which are expected to shed 
further light on this unique object. 

FW gratefully acknowledges the 'Deutsche Forschungsgemeinschaft (DFG)' for the 
award of a stipendium in the Graduate School "The Magellanic Clouds and other Dwarf 
Galaxies" . EB acknowledges a grant from CONACyT (0460P-E) and thanks the Graduate 
School for financially supporting a visit to Bonn during which the foundations for this paper 
were laid. We thank Dominik Bomans, You-Hua Chu, and Mordecai-Mark Mac Low for 
fruitful discussions during the Dwarf Galaxy workshop organised by the Graduate School 
and an anonymous referee for valuable comments which helped us to improve this paper. 


Bomans, D.J., Dennerl, K., & Kurster, M. 1994, A&A, 283, L21 

Bomans, D.J., Chu, Y.-H., Magnier, E.A., & Points, S. 1996, in Rontgenstrahlung from the 
Universe, ed. H.U. Zimmermann, J.E. Triimper & H. Yorke (Garching: Max-Planck 
Institut fiir Extraterrestrische Physik), p. 237 

Bomans, D.J., Chu, Y.-H., & Hopp, U. 1997, AJ, 113, 1678 

Caplan, J., & Deharveng, L. 1986, A&A, 155, 297 

Chevalier, R.A. 1974, ApJ, 188, 501 

Chu, Y.-H., & Mac Low, M.-M. 1990, ApJ, 365, 510 

Chu, Y.-H., & Kennicutt, R.C. 1994, ApJ, 425, 720 

Chu, Y.-H., Chang, H.-W., Su, Y.-L., & Mac Low, M.-M. 1995, ApJ, 450, 156 

- 9- 

Cox, D.P., k Smith, B.W. 1974, ApJ, 189, L105 

Drissen, L., Roy, J.-R., & Moffat, A. F. J. 1993, AJ, 106, 1460 

Duric, N., Gordon, M.S., Goss, W.M., Viallefond, F., & Lacey, C. 1995, ApJ, 445, 173 

Elmegreen, B.G. 1994, ApJ, 427, 384 

Fabbiano, G., & Trinchieri, G. 1987, ApJ, 315, 46 

Hartmann D., & Burton W.B. 1997, in An Atlas of Galactic Neutral Hydrogen, Cambridge 
University Press 

Kim, S., Chu, Y.-H., Staveley-Smith, L. & Smith, R.C. 1998, AJ, submitted 
Lacey, C, Duric, N., & Goss, W.M. 1997, ApJS, 109, 417 

Mac Low, M.-M., Chang, T.H., Chu, Y.-H., & Points, S.D. 1998, ApJ, 493, 260 

Martimbeau, N., Carignan, C, & Roy, J.-R. 1994, A J, 427, 656 

Miller, B.W. 1995, ApJ, 446, L75 

Miller, B. W., & Hodge, P. 1994, ApJ, 427, 656 

Mueller, M. W., & Arnett, W.D. 1976, ApJ, 210, 670 

Puche, D., Westpfahl, D., Brinks, E., & Roy, J.-R. 1992, AJ, 103, 1841 

Raymond, J.C., & Smith, B.W. 1977, ApJS, 35, 419 

Rosa, M., Joubert, M., & Benvenuti, P. 1984, A&AS, 57, 361 

Snowden S.L., McCammon D., Burrows D.N., & Mendenhall J. A. 1994, ApJ, 424, 714 
Tenorio-Tagle, C, & Bodenheimer, P. 1988, ARA&A, 26, 145 
Weaver, R., McCray, R., Castor, J., Shapiro, P., & Moore, R. 1977, ApJ, 218, 377 
Walter, F., & Brinks, E., 1998, AJ, submitted (WB98) 

This preprint was prepared with the AAS IATeX macros v4.0. 


Fig. 1. — Fig. 1 a: H I surface brightness map of IC 2574 as presented in WB98. Note that the 
ISM of IC 2574 is dominated by many hole-like structures. The dashed square indicates the 
H I hole which is the focus of this letter. The dotted square shows the extent of the optical 
Ha-map. Fig. lb: Ha-map of IC2574 as presented in WB98 (greyscale). The contours 
(plotted at 15, 20, 30, 50, 70, 90% of the peak emission) represent the X-ray emission of 
IC 2574 as observed with the ROSAT PSPC camera. The dashed square again indicates the 
supergiant shell which is the focus of this letter. 

Fig. 2. — Fig. 2 a: Blowup of the H I surface brightness map around the prominent H I hole 
(the size of the beam is indicated in the lower left). Fig. 2b: Blowup of the Ha-image with 
X-ray contours superimposed. The area shown is identical to that in Fig. 2 a; the X-ray 
contour levels are the same as in Fig. 1 b. Fig. 2 c: X-ray contours overlaid on the H I hole. 
Fig. 2d: 5A 6 cm radio continuum emission of IC 2574 (contours) overlaid on the Ha map 
(greyscale). The beamsize is indicated in the lower left. Contours are plotted at 0.14, 0.22, 
0.31, 0.39 and 0.48 mJy beam" 1 . 

This figure "walterfig.jpg" is available in "jpg" format from: