Current Research and Future Plans

[Numbers in brackets refer to the bibliography at the end of this document]

Extragalactic Compact Objects

The idea of an object so massive that not even light can escape its gravitational pull remains one of the most fascinating concepts in astrophysics. Black holes come in at least two flavors: stellar-mass (3-20 M$_{\odot}$) black holes formed from the collapse of massive stars, and supermassive (10⁵-10⁹ M$_{\odot}$) black holes located at the centers of galaxies. Tantalizing (but debatable) evidence also exists for black holes with intermediate mass (100-10⁴ M$_{\odot}$). Black holes provide unique constraints on a variety of astrophysical problems, ranging from binary star evolution to accretion disk physics to galaxy formation to gravitational wave radiation. Furthermore, the existence of intermediate-mass black holes would seriously challenge our notions of how and where compact objects form.

The X-ray emission from the accretion disks of black holes not only provides a powerful diagnostic of accretion disk physics, it also provides the most efficient means of detecting black holes, especially in external galaxies. With the launch of Chandra and XMM-Newton, it is now possible to detect significant numbers of black hole X-ray binary systems in nearby galaxies (Figure 1), as well as study the X-ray spectra of the most luminous systems in greater detail. Much of my research has centered around studying black holes in nearby galaxies.

Stellar-Mass Black Holes

Given the meager number of black hole X-ray binary (XRB) systems in the Milky Way (less than 20!), population studies of BH XRBs are hardly feasible using just the Galactic population.

On the other hand, Chandra allows us to detect orders of magnitude more XRBs by summing XRB populations in nearby galaxies. Fed by Roche-lobe overflow (or strong stellar winds) from the companion star, stellar-mass black hole binary systems can shine at X-ray luminosities approaching 10³⁹ ergs s^-1 if accreting near their Eddington limit. However, most Milky Way XRBs (particularly low-mass XRBs) are highly variable, often accreting well below their Eddington limit. The paucity of Milky Way black hole candidates have made the determination of the duty cycles and burst durations of these sources quite difficult, two quantities that are crucial for understanding the accretion mechanism of black hole binaries. The ability to greatly increase the number of black hole candidates in external galaxies with Chandra provides an untapped resource for investigating the variable/transient nature of XRBs. I have begun a project (as part of my LTSA program "The X-ray Binary and Hot Gas Content of Early-type Galaxies") to investigate the transient nature of a large number of X-ray binaries using multiple epoch Chandra observations of nearby galaxies. The initial results are intriguing; the very luminous black hole X-ray binaries are not variable at all over month and year timescales[37], in stark contrast to what is observed in the (albeit few) Galactic counterparts, which are quite variable.

Do Intermediate-Mass Black Holes Exist?

The existence of black holes with masses of 100-10,000 M$_{\odot}$ has been hotly debated in recent years, but as of yet there is no incontrovertible evidence that black holes in this mass range exist. It has been suggested that a very luminous sub-class of X-ray binary in galaxies known as ultraluminous X-ray sources (ULXs) holds the key to demonstrating that intermediate mass black holes exist. Defined as non-nuclear X-ray point sources more luminous than 10³⁹ ergs s^-1, ULXs have luminosities that can exceed the Eddington limit of a 20 M$_{\odot}$ black hole by more than an order of magnitude, naturally suggesting a >100 M$_{\odot}$ accretor. However, theoretical studies have found that it is exceedingly difficult to create such a massive black hole through normal stellar evolution, prompting some researchers to propose alternate explanations involving 10-20 M$_{\odot}$ black holes for which the X-ray emission is beamed toward us or that these sources are emitting at a super-Eddington rate. My work in this field has included showing that ULXs are not found in old stellar populations such as elliptical galaxies or globular clusters, reinforcing the idea that ULXs are intimately associated with very recent star formation[26]. I am also working on assessing the feasibility of deriving the mass function of a ULX binary system for which an optical (secondary) counterpart has been identified. Such a determination will once and for all settle the issue of whether ULXs harbor intermediate mass black holes.

Globular Clusters

Testing the Dynamical History of Globular Clusters with X-ray Binaries

Globular clusters contain far more X-ray binaries per unit light than the rest of the galaxy. This is believed to result from the very high stellar density in globular clusters, which greatly facilitates the capture of a potential donor star by a solitary neutron star or black hole. With Joel Bregman, we have demonstrated that the probability that a Milky Way globular cluster hosts an XRB is proportional to the dynamical interaction rate of the cluster, $\propto L^{1.5}/r_{half}^{2.5}$, and that only clusters with half-mass relaxation times less than 10⁹ years harbor an XRB[33]. We are investigating the feasibility of applying this method to extragalactic globular cluster systems. I have also solved an old question regarding the origin of XRBs presently in the field of galaxies. While it had been suggested that all XRBs formed within globular clusters (with those XRBs presently in the field having formed in globular clusters but escaped to the field at a later time), by comparing the number of globular clusters to the number of XRBs in a sample of early-type galaxies observed with Chandra I have demonstrated that the slope of the relation between these two quantities is shallower than would be expected if there were a one-to-one correspondence between the number of globular clusters and the number of XRBs[28]. This implies that there is a significant field-born population of XRBs in galaxies.

Black Holes Within Extragalactic Globular Clusters

In the Milky Way, there is not a single documented case of a stellar-mass black hole contained within a globular cluster. It has been argued that multiple stellar interactions with a $\sim$10 M$_{\odot}$ black hole would tend to eject the black hole from the cluster on short timescales. Furthermore, there is only tenuous evidence that intermediate-mass black holes reside at the centers of globular clusters. However, the luminous X-ray sources in some extragalactic globular clusters strongly argue that these globular clusters do indeed harbor black holes (of unknown mass). Are these X-ray sources stellar-mass black holes accreting near their Eddington limit or intermediate-mass black accreting at 1% of their Eddington limit?

One means of distinguishing between stellar-mass and intermediate-mass black holes is by how the wind from the black hole interacts with material surrounding the black hole (primarily stellar mass loss from stars in the globular cluster). Such interaction would produce nebular optical emission lines; such emission lines have actually been observed in two extragalactic globular clusters that harbor luminous X-ray sources. These emission lines could result from either collisional or photoionization depending on the mass of the black hole. Thus, distinguishing between collisional and photoionization can provide a powerful discriminant between a stellar mass and intermediate mass black hole as the source of the X-ray emission. I have scheduled Magellan observations of these two objects to attempt to determine whether the lines are created by collisional or photoionization. The finding of an intermediate-mass black hole in a globular cluster would have important ramifications on our understanding of black hole formation on a galactic scale, since it is currently unknown whether the M_BH-$\sigma$ relation found for galactic bulges extends down to globular cluster size objects.

Early-type Galaxies

The Hot Interstellar Medium of Elliptical Galaxies

Since stellar mass loss is a major source of the hot X-ray-emitting interstellar medium of early-type galaxies, it should be expected that there is a strong correlation between the X-ray and optical luminosities of early-type galaxies. Studies have shown this relation to be $L_X \propto L_{opt}^{1.7-3.0}$, broadly consistent with prediction. However, there is a large dispersion (factor of 100) in the X-ray luminosities, L_X, of early-type galaxies at a given optical luminosity, L_opt. Although the X-ray emission in the X-ray "bright" (high L_X/L_opt) galaxies is almost entirely from hot gas, X-ray faint (low L_X/L_opt) galaxies must have had much of their gas removed from the system via galactic winds, or by ram pressure stripping by the ambient intracluster medium. What remains is primarily X-ray emission from a collection of stellar sources such as X-ray binaries, although a small amount of gas is generally present.

Also part of my LTSA project, I am using archival data from Chandra and XMM-Newton to assemble a list of temperatures and luminosities for a complete sample of the 34 optically-brightest elliptical galaxies that lie above a Galactic latitude of $\vert b\vert = 20^{\circ}$. With this sample, I am investigating the relation between the optical luminosity and X-ray luminosity of early-type systems, as well as the relation between the stellar velocity dispersion and the temperature of the X-ray gas. The slope of the L_X vs. L_opt relation is dependent on the hydrodynamical phase of the gas: a supernovae-driven supersonic wind, a subsonic outflow, or a cooling inflow. The X-ray luminosity of the gas changes with its hydrodynamical phase, and thus is dependent on the time since the formation of the galaxy. A comparison of the observed L_X to theoretical calculations will determine the time since the original heating of the gas.

The Role of Accretion onto Early-type Galaxies

The notion that early-type galaxies are devoid of any significant amount of cold gas is slowly being put to rest as a growing body of evidence points to the existence of large scale, massive H I structures surrounding the galaxies. These structures might originate either from galaxy merging or cold accretion of the intergalactic medium (IGM) by the galaxy. But is this external gas actually accreted by early-type galaxies as simulations suggest? If so, then what impact does this have on the evolution of early-type galaxies?

I have recently discovered evidence that the gas is indeed being accreted to the center regions of the galaxy. Since the ISM of early-type galaxies emanates from stellar winds and supernovae, one would expect the ISM to have the same metallicity as the stars, i.e. roughly solar. However, if significant amounts of the (pristine) H I gas is mixed with the solar metallicity gas lost from stars, then the mixed gas would have a substantially lower metallicity. This is exactly what I have found for X-ray faint (gas-poor) early-type galaxies I have analyzed via X-ray spectroscopy of the hot gas with Chandra and XMM-Newton; rather than solar metallicity as would be expected from stellar mass loss, the derived metal abundance of the hot gas is only 20% solar. This indicates that roughly 80% of the gas in the galaxy originated in the pristine H I halo surrounding the galaxy. Such a scenario requires a significant re-writing of our understanding of how early-type galaxies form.

Galaxy Clusters and Galaxy Evolution

Cluster-Induced Galaxy Morphological Evolution

It is now established observationally that the fraction of spiral galaxies in clusters is much higher at large redshifts (z > 0.4) than in the local Universe. This strongly argues that many spiral galaxies are transformed into elliptical and/or S0 galaxies over time. The spiral galaxy fraction is also lower in the inner regions of a cluster than in the outskirts, regardless of redshift. A potential explanation for both effects is that hot, X-ray-emitting intracluster gas (which is also centrally condensed) has systematically removed gas from galaxies via ram pressure stripping as the orbits of the galaxies bring them in toward the center of the cluster over time. The removal of gas from the galaxies would transform spirals into passive, slowly evolving elliptical or S0 galaxies. If this is indeed what is happening, there should be a close correspondence between the local spiral fraction in clusters and the local intracluster medium density. Such a correlation has not been searched for at z 00.4 (where the spiral transformation seems to be taking place), and I am beginning a project to do just this with my former postdoc Chris Mullis. Figure 2 illustrates the technique for Chris's recently discovered z = 1.4 cluster XMMU J2235.3-2557. For a sample of z > 0.4 clusters, we will use Chandra and XMM-Newton data to determine the intracluster gas density as a function of position, and archival HST data to determine the spiral/elliptical fraction as a function of the intracluster medium density. With a large sample, we will have enough statistics to test whether there is a correlation between these two quantities. Such a correlation will imply that stripping plays an important role in transforming spirals to ellipticals and S0s.

Temperature and Metallicity Profiles of the Intracluster Medium

I have also derived radial profiles of the temperature and iron abundance of the hot gas with ROSAT and BeppoSAX[8,11,14], and found that the iron abundance peaks in the center and declines smoothly with increasing radius. This indicates the galaxies that have contributed the metals to the intracluster medium are more spatially condensed than the primordial hot gas. The metallicity of the gas also provides clues to the enrichment mechanism of the intracluster medium. The ratio of iron to oxygen in the gas indicates the type of supernovae that enriched the primordial gas (Type Ia vs. Type II).

The Warm-Hot Intergalactic Medium (WHIM)

The number of baryons currently detected in the local Universe is considerably less than that predicted from Big Bang nucleosynthesis. It is believed that the majority of baryons in the local Universe lie outside of galaxies, in modest overdensity regions of the Universe, the so-called warm-hot intergalactic medium (WHIM), or "Cosmic Web". The baryons are predicted to be in the form of hot 10⁵-10⁷ K diffuse gas. Attempts to detect the WHIM in emission or absorption in the X-ray regime has produced tantalizing results, but have not conclusively demonstrated the existence of the WHIM. Indeed, our tentative detection of X-ray shadowing of the WHIM[18], where we used the neutral gas in a nearby spiral galaxy as an absorbing screen for X-ray photons of the WHIM emanating behind it, was not confirmed with a deeper Chandra observation.

Another approach for detecting the WHIM developed by Joel Bregman and myself involves using the hyperfine line of N VII, which is the equivalent of the 21 cm hyperfine transition for hydrogen. The hyperfine line of N VII occurs at a frequency of 53.2 GHz, but the atmosphere is opaque at this frequency, so it cannot be used to probe locally, but it can be used at modest redshift (z > 0.1) where the atmosphere is transparent. In this line, the emission from the WHIM will be too small to be detectable, but the detection of this line in absorption holds promise. For H I, the hyperfine absorption feature is greatly reduced by stimulated emission, since the two levels are close to their Boltzmann ratio, with the collision time being much shorter than the lifetime of the excited level. However, for N VII, the transition time is less than the time between collisions for WHIM conditions, so stimulated emission is unimportant and it is a much better absorber.

The idea behind this method is to observe a high redshift AGN as background "wall paper" in the 40-50 GHz range with the Green Bank Telescope (GBT) and search for N VII absorption lines from the WHIM along the line of sight to the AGN. We currently have a tentative 4$\sigma$ detection of an absorption line toward an AGN that could be evidence for the WHIM along the line of sight[38]. We have another upcoming observing run at the GBT to confirm this result and search for evidence of the WHIM in N VII absorption toward other lines of sight. This method holds the promise of providing a powerful constraint on the existence, density and temperature of the so-called "missing baryons" in the Universe.

PUBLICATIONS


(16 Refereed First Author Papers with 494 citations)

[38] "The Search for Million Degree Gas Through the N VII Hyperfine Line",
J. N. Bregman, & J. A. Irwin, ApJ, 666, 139, 2007

[37] "The Remarkable Stability of Probable Black Hole Low-Mass X-ray Binaries in
Nearby Galaxies",
J. A. Irwin, MNRAS, 371, 1903, 2006

[36] "The Size of the Cooling Region of Hot Gas in Two Elliptical Galaxies",
J. N. Bregman, B. Otte, E. D. Miller, & J. A. Irwin, ApJ, 642, 759, 2006

[35] "OVI Observations of Galaxy Clusters: Evidence for Modest Cooling Flows",
J. N. Bregman, A. C. Fabian, E. D. Miller, & J. A. Irwin, ApJ, 642, 746, 2006

[34] "Ultra-luminous X-ray Sources in Nearby Galaxies from ROSAT HRI Observations
II. Statistical Properties,
J.-F. Liu, J. N. Bregman, & J. A. Irwin, ApJ, 642, 171, 2006

[33] "Galactic Globular Clusters with Luminous X-Ray Binaries",
J. N. Bregman, J. A. Irwin, P. Seitzer, & M. Flores, ApJ, 640, 282, 2006

[32] "XMM-Newton Observation of Diffuse Gas and LMXBs in the Elliptical Galaxy
NGC 4649 (M60)",
S. W. Randall, C. L. Sarazin, & J. A. Irwin, ApJ, 636, 200, 2006

[31] "Optical Studies of Ultra-luminous X-ray Sources in Nearby Galaxies",
J.-F. Liu, J. N. Bregman, P. Seitzer, & J. Irwin, in press, astro-ph/0501310, 2005

[30] "OVI in Elliptical Galaxies: Indicators of Cooling Flows",
J. N. Bregman, E. D. Miller, A. E. Athey, & J. A. Irwin, ApJ, 635, 1031, 2005

[29] "The Cosmological Unimportance of Low Surface Brightness Galaxies",
C. Hayward, J. A. Irwin, & J. N. Bregman, ApJ, 635, 827, 2005

[28] "The Birthplace of Low-Mass X-Ray Binaries: Field Versus Globular Cluster
Populations",
J. A. Irwin, ApJ, 631, 511, 2005

[27] "A 2 Hour Quasi Period in an Ultraluminous X-Ray Source in NGC 628",
J.-F. Liu, J. N. Bregman, E. Lloyd-Davies, J. Irwin, C. Espaillat, & P. Seitzer,
ApJ, 621, L17, 2005

[26] "The Lack of Very Ultraluminous X-ray Sources (ULXs) in Early-type Galaxies",
J. A. Irwin, J. N. Bregman, & A. E. Athey, ApJ, 601 143, 2004

[25] "Chandra Observation of Diffuse Gas and LMXBs in the Elliptical Galaxy
NGC 4649 (M60)",
S. W. Randall, C. L. Sarazin, & J. A. Irwin, ApJ, 600, 729, 2004

[24] "Chandra Observations of Low Mass X-Ray Binaries and Diffuse Gas in the Early- Type Galaxies NGC 4365 and NGC 4382 (M85)",
G. R. Sivakoff, C. L. Sarazin, & J. A. Irwin, ApJ, 599, 218, 2004

[23] "Identifying X-ray Sources in Bulge and Disk Galaxies from X-ray Colors",
A. H. Prestwich, J. A. Irwin, R. E. Kilgard, M. I. Krauss, A. Zezas, F. Primini, P. Kaaret, & B. Boroson, ApJ, 595, 719 2003

[22] "Low Mass X-ray Binaries and Globular Clusters in Early-Type Galaxies",
C. L. Sarazin, A. Kundu, J. A. Irwin, G. R. Sivakoff, E. L. Blanton, & S. W. Randall, ApJ, 595, 743, 2003

[21] "X-ray Spectral Properties of Low-Mass X-ray Binaries in Nearby Galaxies",
J. A. Irwin, A. E. Athey, & J. N. Bregman, ApJ, 587, 356, 2003

[20] "An Ultra-Luminous X-Ray Object with a 2 Hour Period in M51",
J.-F. Liu, J. N. Bregman, J. A. Irwin, & P. Seitzer, ApJ, 581, L93, 2002

[19] "Untangling the X-ray Emission From the Sa Galaxy NGC 1291 With Chandra",
J. A. Irwin, C. L. Sarazin, & J. N. Bregman, ApJ, 570, 152, 2002

[18] "A Shadow of the Extragalactic X-ray Background",
J. N. Bregman, & J. A. Irwin, ApJ, 565, L13, 2002

[17] "Chandra X-ray Observations of the X-ray Faint Elliptical Galaxy NGC 4697",
C. L. Sarazin, J. A. Irwin, & J. N. Bregman, ApJ, 556, 533, 2001

[16] "The Detection of a Cooling Flow Elliptical Galaxy from O VI Emission",
J. N. Bregman, E. D. Miller, & J. A. Irwin, ApJ, 553, L125, 2001

[15] "Chandra X-ray Observations of the S0 Galaxy NGC 1553",
E. L. Blanton, C. L. Sarazin, & J. A. Irwin, ApJ, 552, 106, 2001

[14] "Iron Abundance Profiles of 12 Clusters of Galaxies Observed With BeppoSAX",
J. A. Irwin, & J. N. Bregman, ApJ, 546, 293, 2001

[13] "Resolving the Mystery of X-ray Faint Elliptical Galaxies: Chandra X-ray
Observations of NGC 4697",
C. L. Sarazin, J. A. Irwin, & J. N. Bregman, ApJ, 544, L101, 2000

[12] "The X-ray Faint Early-Type Galaxy NGC 4697",
J. A. Irwin, C. L. Sarazin, J. N Bregman, ApJ, 544, 293, 2000

[11] "Radial Temperature Profiles of 11 Clusters of Galaxies Observed with
Beppo-SAX",
J. A. Irwin, & J. N. Bregman, ApJ, 538, 543, 2000

[10] "Using the Bulge of M31 as a Template for the Integrated X-Ray Emission from Low-Mass X-Ray Binaries",
J. A. Irwin, & J. N. Bregman, ApJ, 527, 125, 1999

[9] "ROSAT HRI X-Ray Observations of the Open Globular Cluster NGC 288",
C. L. Sarazin, J. A. Irwin, R. T. Rood, F. R. Ferraro, & B. Paltrinieri, ApJ, 524, 220, 1999

[8] "Radial Temperature Profiles of X-Ray-Emitting Gas Within Clusters of Galaxies," J. A. Irwin, J. N. Bregman, & A. E. Evrard, ApJ, 519, 518, 1999

[7] "The Dependence of the Soft X-ray Properties of LMXBs on the Metallicity of Their Environment",
J. A. Irwin & J. N. Bregman, ApJ, 510, L21, 1999

[6] "ROSAT X-Ray Colors and Emission Mechanisms in Early-Type Galaxies,"
J. A. Irwin & C. L. Sarazin, ApJ, 499, 650, 1998

[5] "Low Mass X-ray Binaries As the Source of the Very Soft X-ray Emission in the X-ray Faintest Early-Type Galaxies,"
J. A. Irwin, & C. L. Sarazin, ApJ, 494, L33, 1998

[4] "Is There Molecular Gas in the H I Cloud Between NGC 4472 and UGC 7636?",
J. A. Irwin, D. T. Frayer, & C. L. Sarazin, AJ, 113, 1580, 1997

[3] "Heating of the Intracluster Gas in the Triangulum Australis Cluster,"
M. L. Markevitch, C. L. Sarazin, & J. A. Irwin, ApJ, 472, L17, 1996

[2] "X-Ray Evidence for the Interaction of the Giant Elliptical Galaxy NGC 4472 With Its Virgo Cluster Environment,"
J. A. Irwin, & C. L. Sarazin, ApJ, 471, 683, 1996

[1] "ROSAT X-ray Observations of the 2A 0335+096 Cluster of Galaxies,"
J. A. Irwin, & C. L. Sarazin, ApJ, 455, 497, 1995