The Astrophysical Journal, 744:112 (13pp), 2012 January 10 C 2012.
doi:10.1088/0004-637X/744/2/112
The American Astronomical Society. All rights reserved. Printed in the U.S.A.
A MOLECULAR HYDROGEN NEBULA IN THE CENTRAL cD GALAXY OF THE PERSEUS CLUSTER Jeremy Lim1,3 , Youichi Ohyama2 , Yan Chi-Hung2 , Dinh-V-Trung2,4 , and Wang Shiang-Yu2
2
1 Department of Physics, University of Hong Kong, Pokfulam Road, Hong Kong;
[email protected] Institute of Astronomy and Astrophysics, Academia Sinica, Taipei 10617, Taiwan;
[email protected],
[email protected],
[email protected],
[email protected] Received 2010 June 16; accepted 2011 October 6; published 2011 December 20
ABSTRACT We report narrowband imaging of the 1–0 S(1) ro-vibrational transition of molecular hydrogen (H2 ) from NGC 1275, the central cD galaxy of the Perseus Cluster. We find that the H2 gas has a spatial morphology identical to the optical emission-line nebula associated with this galaxy, a total luminosity in H2 1–0 S(1) only an order of magnitude less than in Hα, and if the line-emitting gas is thermalized a mass (at ∼2000 K) that is over two orders of magnitude smaller than that of the optical emission-line nebula (at ∼10,000 K). The ratio in H2 1–0 S(1) to Hα + [N ii] line intensities spans a characteristic range of ∼0.02–0.08 throughout the nebula; the brighter inner nebula exhibits patches with (nearly) constant line ratios unrelated to individual filaments. Recent models proposed to explain the peculiar nebular spectrum from the optical to infrared invoke thermalized along with non-thermalized injection of energy from ionizing particles. The energy density of highly relativistic electrons inferred to cause inverse-Compton scattering of hard X-ray emission from the core of the Perseus Cluster decreases steeply beyond a central radius of ∼20 kpc, yet we do not find any changes in the average or range spanned by the H2 1–0 S(1) to Hα + [N ii] line ratio between the inner (20 kpc) and outer (∼20–50 kpc) nebulae. On the other hand, saturated conduction from the surrounding X-ray gas produces, in the absence of magnetic fields, a heat flux that is approximately constant throughout the nebula: the change in the line ratio with position would then reflect the ability of the X-ray gas to penetrate presumably magnetically threaded filaments at different locations. Key words: galaxies: active – galaxies: clusters: intracluster medium – galaxies: individual (NGC 1275) – galaxies: ISM – infrared: ISM – ISM: molecules Online-only material: color figures
1955) or an explosive event in the nucleus of NGC 1275 (Burbidge et al. 1963; Burbidge & Burbidge 1965), much about the nature—composition, excitation, origin, and fate—of this nebula remains to this day poorly understood. Following the suggestions by Fabian & Nulsen (1977) and Cowie & Binney (1977) that relaxed galaxy clusters such as the Perseus Cluster—where the X-ray emission from the intracluster gas peaks sharply at the cluster center—should harbor an X-ray cooling flow, Kent & Sargent (1979) suggested that the optical emission-line nebula of NGC 1275 may have condensed from an accretion flow of the intracluster medium. Subsequently, many (although not all) central cD galaxies of clusters predicted to have strong X-ray cooling flows were found to exhibit much more luminous optical emission-line nebulae than their counterparts with at best weak X-ray cooling flows (e.g., Heckman et al. 1989, and references therein). This distinction provides circumstantial evidence for a physical link between the optical emission-line nebulae and X-ray cooling flows; alternatively, Nipoti & Binney (2004) argue that the optical emission-line filaments cannot survive in the higher temperature X-ray core of clusters that do not harbor X-ray cooling flows. The nebular Hα luminosities, however, are typically at least an order of magnitude (in the case of NGC 1275, about three orders of magnitude) higher than expected for gas cooling passively from an X-ray cooling flow at the then predicted mass-deposition rates (Heckman et al. 1989). Over the last decade, failures to detect X-ray-emitting gas at temperatures below about one-third the bulk ambient temperature of the intracluster medium in clusters suspected to harbor X-ray cooling flows (starting with the studies by Tamura et al. 2001 and Peterson et al. 2001) have led to severe (by at least an
1. INTRODUCTION The central cD (giant elliptical) galaxy of the Perseus Cluster, NGC 1275 (the host galaxy of the double-lobed (FR I) radio source 3C 84, which is also known as Perseus A), is among the most enigmatic galaxies in the local universe. Its most visually spectacular peculiarity is an optical emission-line (most often imaged in Hα + [N ii]; see Figure 1, left panel) nebula comprising a multitude of filaments that project chiefly radially or nearly so with respect to the galactic nucleus, although there also are tangential filaments lying well away from the center (Conselice et al. 2001). This nebula spans nearly 400 (144 kpc at a distance, for H0 = 70 km s−1 Mpc−1 , of 74 Mpc; thus 1 = 360 pc, as will be assumed throughout this article) in the north–south direction, compared with a half-light diameter for NGC 1275 of ∼69 (∼24.8 kpc) (Smith et al. 1990, scaled to the assumed distance) and Holmberg diameter of ∼690 (∼250 kpc) (Schombert 1987). The inner region of this nebula was first discovered by Minkowski (1957), although Baade & Minkowski (1954) had earlier presented tentative evidence that the bright nuclear line emission of NGC 1275 is extended, and subsequently studied by Burbidge et al. (1963), Burbidge & Burbidge (1965), and Minkowski (1968). The full splendor of this nebula, however, was not revealed until the photograph taken by Lynds (1970). Attributed early on to either a collision between NGC 1275 (then thought to be a spiral galaxy) and a foreground infalling (probably a spiral) galaxy (Minkowski 3 Adjunct Research Fellow, Institute of Astronomy and Astrophysics, Academia Sinica, Taipei 10617, Taiwan. 4 Current address: Institute of Physics, Vietnam Academy of Science and Technology, 10 DaoTan, ThuLe, BaDinh, Hanoi, Vietnam.
1
The Astrophysical Journal, 744:112 (13pp), 2012 January 10
Lim et al.
Figure 1. Image of the nebula associated with NGC 1275 in Hα + [N ii] as reproduced from Conselice et al. (2001) (left panel) and H2 1–0 S(1) as reported in this paper (right panel). There is an excellent morphological correspondence between the optical ionized-gas and near-IR molecular-gas nebulae. The green circles in both panels indicate imperfectly subtracted bright foreground stars. (A color version of this figure is available in the online journal.)
order of magnitude) downward revisions in the mass-deposition rates from such flows (e.g., reviews by Peterson & Fabian 2006; McNamara & Nulsen 2007). This has reinforced doubts among many about the existence of X-ray cooling flows, and further complicates our understanding of the link between the optical emission-line nebulae and supposed X-ray cooling flows. Important insights into the composition and excitation of these optical emission-line nebulae came with the discovery by Elston & Maloney (1994) of strong 2.12 μm (rest wavelength) emission from the v = 1–0 S(1) ro-vibrational transition of molecular hydrogen (H2 ) gas toward a number of central cluster cD galaxies, including NGC 1275. Soon thereafter, Jaffe & Bremer (1997) and Falcke et al. (1998) showed that only those central cD galaxies in clusters predicted to have strong X-ray cooling flows, but not others, exhibit detectable near-IR H2 lines. Edge et al. (2002) greatly expanded on the number of detections and showed that over two-thirds of central cD galaxies in clusters suspected to harbor strong X-ray cooling flows exhibit near-IR H2 lines. Donahue et al. (2000) demonstrated from narrowband Hubble Space Telescope (HST) imaging that the 2.12 μm H2 line in two relatively distant clusters closely resembles their optical emission-line nebulae, suggesting that the optical line-emitting and molecular gas phases are closely cospatial. Hatch et al. (2005) detected near-IR H2 lines at several regions in the optical emissionline nebula of NGC 1275 observed using slit spectroscopy, and found that the near-IR H2 lines have the same (radial) velocities as the Pα emission. They argued that the optical lineemitting and molecular hydrogen gas phases must therefore share a common excitation mechanism. Jaffe et al. (2005) showed from slit spectroscopy that the optical atomic/ionized and near-IR H2 lines from a number of central cD galaxies not only have similar (radial) velocities but also similar radial brightness profiles, and likewise argued for a common excitation mechanism for both gas phases. Just like for the nebular optical emission lines, Jaffe & Bremer (1997) pointed out that the H2 1–0 S(1) line is too luminous to constitute gas cooling passively from an X-ray
cooling flow at the then predicted mass-deposition rates, let alone the recent downward revisions on these rates to upper limits. They also pointed out that the observed ratio in the H2 1–0 S(1) to Hα luminosity (roughly 0.1) is about an order of magnitude larger than the same ratio observed in Galactic H ii regions and starburst galaxies (about 0.01), reinforcing earlier findings that (in general) the optical emission-line nebulae do not exhibit H ii-region-like spectra but instead LINER-like spectra (Heckman 1987, and references therein). Both Jaffe et al. (2001) and Wilman et al. (2002) found that most of the v = 1–0 H2 lines have line ratios indicative of thermalized gas at temperatures of ∼1000–2000 K. Reproducing the observed H2 1–0 S(1) to Hα line ratios has driven many subsequent models for simultaneously explaining the presumed heating of the molecular hydrogen gas and presumed ionization of the optical emission-line nebula. The way in which the luminous optical emission-line nebulae of central cluster cD galaxies are powered remains, in most cases, an outstanding puzzle. Many of the ideas debated to this day were critically examined over 20 years ago by Heckman et al. (1989): here, we confine our summary of the proposed models and their viability to the emission-line nebula in NGC 1275 (the same general ideas, although not necessarily in equal degrees of viability, apply to the luminous emission-line nebulae in other central cluster cD galaxies). These models can be divided into three broad categories: (1) photoionization (purely radiative energy injection); (2) collisional excitation/ionization in thermalized gas heated nonradiatively (non-radiative, purely thermalized energy injection); and (3) in addition to (2), excitation/ionization by collisions with energetic particles (non-radiative, non-thermalized energy injection). In the last model, the same energetic particles also produce thermal heating of the line-emitting gas. Virtually all the models proposed that invoke photoionization for powering the optical emission-line nebula in NGC 1275 face severe difficulties on energetic and/or other grounds. Photoionization by the active galactic nucleus (AGN) in NGC 1275, except in a compact region immediately around the nucleus, is 2
The Astrophysical Journal, 744:112 (13pp), 2012 January 10
Lim et al.
ruled out by the lack of strong radial gradients in the nebular optical emission-line ratios (Johnstone & Fabian 1988; Heckman et al. 1989); moreover, the AGN is not sufficiently luminous to produce the observed nebular luminosity (Heckman et al. 1989; Conselice et al. 2001). Instead, the lack of strong radial gradients in the nebular optical emission-line ratios strongly implies an in situ excitation source. Apart from a single isolated H ii region so far found (Shields & Filippenko 1990; see, however, Section 3.3.1), the nebula has a LINER-like rather than an H iiregion-like spectrum (Heckman 1987, and references therein), thus ruling out photoionization from massive stars (formed, if any, within the nebula) as the primary ionization mechanism. Photoionization by X-ray-emitting gas from the surrounding intracluster medium (Voit & Donahue 1990; Donahue & Voit 1991) provides an alternative to photoionization by the AGN or stellar radiation. The surface brightness of the X-ray-emitting gas in the core of the Perseus Cluster is ∼8 × 10−15 erg cm−2 s−1 arcsec−2 within a central radius of ∼1 (21.6 kpc) and drops steeply beyond (W. Forman, private communication). By comparison, utilizing the common estimate (e.g., Fabian et al. 2011) that the total nebular line intensity is ∼20 times (dominated in intensity by optical lines, but also including infrared and millimeter/submillimeter lines) than measured in Hα+[N ii], the surface brightness of the filaments (beyond the bright and compact nuclear source) exceeds ∼1 × 10−14 erg cm−2 s−1 arcsec−2 by up to two orders of magnitude (although relatively bright regions likely contain overlapping filaments) within a radius of ∼1 and drops, if at all, by at most a factor of a few in the outermost regions (Conselice et al. 2001). Thus, photoionization by the surrounding X-ray gas is by itself not sufficiently powerful to excite the nebula, with the deficiency becoming especially apparent beyond a central radius of ∼20 kpc. To simultaneously explain the excitation of both the optical line-emitting and molecular hydrogen gas under a single mechanism, Wilman et al. (2002) argued for UV irradiation by principally early-O stars (effective temperature ∼50,000 K), whereas Jaffe et al. (2005) advocated even harder UV radiation (blackbody temperatures 105 K; e.g., white dwarf or Wolf–Rayet stars). In both models, the Hα emission constitutes recombination lines from hydrogen gas that is ionized by the absorption of Lyman continuum photons. In the model of Wilman et al. (2002), molecular hydrogen is excited solely by absorption of photons in the Lyman and Werner lines of molecular hydrogen. In the model of Jaffe et al. (2005), molecular hydrogen is also excited by collisions with energetic photoelectrons released when molecular hydrogen absorbs photons at wavelengths significantly shorter than the Lyman limit of atomic hydrogen (hence the requirement for harder UV radiation than in the model of Wilman et al. 2002). The presence of any such stellar populations invoked in the models of Wilman et al. (2002) and Jaffe et al. (2005), however, is entirely speculative. Models that rely solely only on collisional excitation/ ionization in thermalized gas for producing the optical emissionline nebula, and simultaneously explaining the observed near-IR H2 to Hα line ratios, face severe difficulties. Shock heating by blast waves driven by supernovae, given the current upper limit on the mass-deposition rate from any X-ray cooling flow in the Perseus Cluster and hence the corresponding star formation rate in NGC 1275, falls short energetically (assuming that stars form with a Galactic initial mass function, and an efficiency of ∼2% in converting supernovae kinetic energy to gas thermal energy) by many orders of magnitude (Heckman et al. 1989). Furthermore,
the relatively high H2 1–0 S(1) to Hα line ratio requires slow (50 km s−1 ) so as not to appreciably ionize hydrogen and otherwise produce relatively low H2 1–0 S(1) to Hα line ratios (Jaffe & Bremer 1997). The requirement of only lightly ionizing hydrogen, however, means that such slow shocks cannot produce bright hydrogen recombination lines and hence the luminous optical emission-line nebula observed; in any case the generating of such slow shocks is entirely speculative (see the discussion in Jaffe & Bremer 1997 and Wilman et al. 2002). Pope et al. (2008b) and Pope et al. (2008a) suggest that the emission-line filaments comprise gas dragged out of infalling cold gas clumps, which condensed from the X-ray cooling flow, when intracluster X-ray gas, pushed along by the radio jets from or radio bubbles created by the AGN in NGC 1275, streams past these clumps. Energy dissipation due to the drag heats the resulting filaments, thus giving rise to the observed optical and infrared (from both atomic and molecular) emission lines. Ferland et al. (2009) have argued persuasively, however, that models which invoke a purely thermally excited gas cannot produce all the observed nebular emissionline ratios in the optical and infrared, in particular those involving the relatively bright (optical) He i, (mid-IR) [Ne ii], and (optical/mid-IR) [Ne iii] lines. The most widely discussed models of late invoke energetic particles to heat, excite, and dissociate molecular or ionize atomic gas so as to produce the observed IR H2 to optical/near-IR H i line ratios and indeed the other observed nebular emission-line ratios in the optical and infrared (e.g., Ferland et al. 2008, 2009; Fabian et al. 2011). Central to the model examined in greatest detail, described in Ferland et al. (2009), is the idea that energetic (ionizing) particles impinging on molecular gas produce a mixture of molecular, atomic, and ionized gas in the same parcel of gas. Because no single parcel of gas at a given density and temperature can produce the full range of molecular, atomic, and ionic emission lines observed, the emitting gas must comprise multiple components at different densities and temperatures. In this model, the IR H2 lines are produced in part in predominantly (almost fully) molecular (the remainder mostly atomic) gas excited by collisions with energetic particles, and in part in predominantly (∼90%) atomic (the remainder mostly molecular) gas at higher temperatures where molecular hydrogen is excited by collisions with atomic, molecular, or ionized hydrogen in thermalized gas. The H i lines are produced in part in predominantly (∼90%) atomic (the remainder mostly molecular) gas where hydrogen atoms are excited by collisions with energetic particles, and in part in weakly ionized (
