Low-ionization Emission Lines Research Articles

Low Ionization Emission Lines in Quasars: Clues from OI 8446 and the CaII Triplet

The formation of low emission lines in quasars and active galactic nuclei is still an open issue. Aided by the organizing power of the 4D eigenvector 1 scheme, we review basic developments since th...

The nature of the H2-emitting gas in the Crab nebula★

Understanding how molecules and dust might have formed within a rapidly expanding young supernova remnant is important because of the obvious application to vigorous supernova activity at very high redshift. In previous papers, we found that the H2 emission is often quite strong, correlates with optical low-ionization emission lines, and has a surprisingly high excitation temperature. Here we study Knot 51, a representative, bright example, for which we have available long slit optical and NIR spectra covering emission lines from ionized, neutral, and molecular gas, as well as HST visible and SOAR Telescope NIR narrow-band images. We present a series of CLOUDY simulations to probe the excitation mechanisms, formation processes and dust content in environments that can produce the observed H2 emission. We do not try for an exact match between model and observations given Knot 51's ambiguous geometry. Rather, we aim to explain how the bright H2 emission lines can be formed from within the volume of Knot 51 that also produces the observed optical emission from ionized and neutral gas. Our models that are powered only by the Crab's synchrotron radiation are ruled out because they cannot reproduce the strong, thermal H2 emission. The simulations that come closest to fitting the observations have the core of Knot 51 almost entirely atomic with the H2 emission coming from just a trace molecular component, and in which there is extra heating. In this unusual environment, H2 forms primarily by associative detachment rather than grain catalysis. In this picture, the 55 H2-emitting cores that we have previously catalogued in the Crab have a total mass of about 0.1 M_sun, which is about 5% of the total mass of the system of filaments. We also explore the effect of varying the dust abundance. We discuss possible future observations that could further elucidate the nature of these H2 knots.

Low Ionization Emission Lines in Quasars: Clues from OI 8446 and the CaII Triplet

Low Ionization Emission Lines in Quasars: Clues from OI 8446 and the CaII Triplet

ACCRETION PROPERTIES OF HIGH- AND LOW-EXCITATION YOUNG RADIO GALAXIES

Young radio galaxies (YRGs) provide an ideal laboratory to explore the connection between accretion disk and radio jet thanks to their recent jet formation. We investigate the relationship between the emission-line properties, the black hole accretion rate, and the radio properties using a sample of 34 low-redshift (z < 0.4) YRGs. We classify YRGs as high-excitation galaxies (HEGs) and low-excitation galaxies (LEGs) based on the flux ratio of high-ionization to low-ionization emission lines. Using the H{\alpha} luminosities as a proxy of accretion rate, we find that HEGs in YRGs have \sim1 dex higher Eddington ratios than LEGs in YRGs, suggesting that HEGs have higher mass accretion rate or higher radiative efficiency than LEGs. In agreement with previous studies, we find that the luminosities of emission lines, in particular H{\alpha}, are correlated with radio core luminosity, suggesting that accretion and young radio activities are fundamentally connected.

Stellar populations in the centers of nearby galaxies

AbstractThe great amount of data observed in recent years coupled with modelling using evolutionary synthesis codes (BPASS, COELHO, GALAXEV, GALEV, MILES, PÉGASE, etc. . .) to compute Single Stellar Populations (SSPs) and the availability of fast and ingenious spectral synthesis codes such as starlight, ULySS and VESPA, have significantly shed light on our knowledge about the formation and evolution of galaxies. However, there are still open issues concerning the stellar populations in nearby galaxies, particularly those harbouring Active Galactic Nuclei (AGN): can stellar populations mimic nuclear activity, leading to a misclassification based on optical emission line ratios (Stasińska et al. 2008)? We have applied the starlight code (Cid Fernandes et al. 2005) to a well studied sample of nearby galaxies' nuclear spectra (r &lt; ~ 200 pc), observed with the Hale 5 m telescope at Palomar Observatory in two different regions: ~ 4230-5110 Å and ~ 6210-6860 Å (Ho et al. 1995), with spectral resolutions of approximately 4 Å, and 2.5 Å. The aim is to properly derive the star-formation history (SFH), mean stellar age and metallicity and total stellar mass. Our results show that the star-formation history of Seyfert galaxies are very heterogeneous, i.e. these are composed of young, intermediate and old stellar populations, while the SFH of Low-Ionization Nuclear Emission-Line Regions (LINERs) are basically composed of old stellar populations. The absence of young stars in LINERs indicates that these are not responsible for the observed low-ionization emission lines. Furthermore, although a significant fraction of AGN spectra require a featureless continuum in their Spectral Energy Distribution (SED) modelling, this is not an indicative of the presence of an AGN, instead the continuum may simulate the presence of young stellar populations. The main objective of this research is to complement the study of spectroscopic parameters from 486 galaxies analyzed by Ho et al. (1995) that are public available in the VizieR catalog (Ho et al. 1997, 2009) and provide information about their stellar population content by means of the starlight. The base of Simple Stellar Populations used here was taken from Bruzual &amp; Charlot (2003) and spans 25 ages (from 1 Myr to 18 Gyr) and 6 metallicities (Z = 0.005, 0.02, 0.2, 0.4, 1 &amp; 2.5 Z⊙).

High-Dispersion Spectroscopy of BF Cygni at the Beginning of the 2006 Outburst

ABSTRACTBF Cygni (hereafter BF Cyg) is a symbiotic star that in 2006 July started into an optical outburst that has lasted for several years. Between 2006 August and 2007 January we obtained high-dispersion spectra (r ∼ 35,000) over the wavelength range of 3800–10000 Å of BF Cyg with the 3.5 m echelle spectrograph at Apache Point Observatory. The emission lines in the optical spectrum of BF Cyg changed dramatically during the interval of observations. The species found were mostly low-ionization emission lines of H, He, C, N, O, Mg, Si, Ca, Sc, Ti, V, Cr, Mn, Fe, and Ni. Both permitted and forbidden lines were seen. In particular, Fe I was observed in emission, which is not a common phenomenon. We discuss the evolution of the BF Cyg spectrum during the 5 month interval of our observations. Some emission lines developed P-Cygni profiles indicating gas expanding at about 100–200 km s-1. Other lines changed from emission to absorption. A few absorption lines that had not been observed in 2006 August were evident on the spectra obtained in 2006 November and 2007 January. Some interesting (but unidentified) persistent broad features were observed in the continuum. Comparisons of our observations with previous spectroscopy of BF Cyg demonstrate that the emission spectrum is in a lower state of excitation during this outburst than during quiescence.

3C 390.3: more stable evidence that the double-peaked broad Balmer lines originate from an accretion disc near a central black hole

In this manuscript, the structure of broad emission line regions (BLRs) of well-mapping double-peaked emitter (AGN with broad double-peaked low-ionization emission lines) 3C390.3 is studied. Besides the best fitted results for double-peaked broad optical balmer lines of 3C390.3 by theoretical disk model, we try to find another way to further confirm the origination of double-peaked line from accretion disk. Based on the long-period observed spectra in optical band around 1995 collected from AGN WATCH project, the theoretical disk parameters of disk-like BLRs supposed by elliptical accretion disk model (Eracleous et al. 1995) have been well determined. Through the theoretical disk-like BLRs, characters of observed light-curves of broad double-peaked H$\alpha$ of 3C390.3 can be well reproduced based on the reverberation mapping technique. Thus the accretion disk model is preferred as one better model for BLRs of 3C390.3. Furthermore, we can find that different disk parameters should lead to some different results about size of BLRs of 3C390.3 from the one measured through observational data, which indicates the measured disk parameters are significantly valid for 3C390.3. After that, the precession of theoretical elliptical disk-like BLRs being considered, we can find that the expected line profile in 2000 by theoretical model is consistent with the observed line profile by HST around 2000. Based on the results, we can further believe that the origination of broad double-peaked balmer emission lines of 3C390.3 are from accretion disk around central black hole.

JET-POWERED MOLECULAR HYDROGEN EMISSION FROM RADIO GALAXIES

H2 pure-rotational emission lines are detected from warm (100-1500 K) molecular gas in 17/55 (31% of) radio galaxies at redshift z<0.22 observed with the Spitzer IR Spectrograph. The summed H2 0-0 S(0)-S(3) line luminosities are L(H2)=7E38-2E42 erg/s, yielding warm H2 masses up to 2E10 Msun. These radio galaxies, of both FR radio morphological types, help to firmly establish the new class of radio-selected molecular hydrogen emission galaxies (radio MOHEGs). MOHEGs have extremely large H2 to 7.7 micron PAH emission ratios: L(H2)/L(PAH7.7) = 0.04-4, up to a factor 300 greater than the median value for normal star-forming galaxies. In spite of large H2 masses, MOHEGs appear to be inefficient at forming stars, perhaps because the molecular gas is kinematically unsettled and turbulent. Low-luminosity mid-IR continuum emission together with low-ionization emission line spectra indicate low-luminosity AGNs in all but 3 radio MOHEGs. The AGN X-ray emission measured with Chandra is not luminous enough to power the H2 emission from MOHEGs. Nearly all radio MOHEGs belong to clusters or close pairs, including 4 cool core clusters (Perseus, Hydra, A 2052, and A 2199). We suggest that the H2 in radio MOHEGs is delivered in galaxy collisions or cooling flows, then heated by radio jet feedback in the form of kinetic energy dissipation by shocks or cosmic rays.

Nuclear and extended spectra of NGC 1068 - I. Hints from near-infrared spectroscopy

We report the first simultaneous zJHK spectroscopy on the archetypical Seyfert 2 Galaxy NGC 1068 covering the wavelength region 0.9 to 2.4 micron. The slit, aligned in the NS direction and centred in the optical nucleus, maps a region 300 pc in radius at sub-arcsec resolution, with a spectral resolving power of 360 km s^-1. This configuration allow us to study the physical properties of the nuclear gas including that of the north side of the ionization cone, map the strong excess of continuum emission in the K-band and attributed to dust and study the variations, both in flux and profile, in the emission lines. Our results show that (1) Mid- to low-ionization emission lines are splitted into two components, whose relative strengths vary with the position along the slit and seem to be correlated with the jet. (2) The coronal lines are single-peaked and are detected only in the central few hundred of parsecs from the nucleus. (3) The absorption lines indicate the presence of intermediate age stellar population, which might be a significant contributor to the continuum in the NIR spectra. (4) Through some simple photoionization models we find photoionization as the main mechanism powering the emitting gas. (5) Calculations using stellar features point to a mass concentration inside the 100 - 200 pc of about 10^10 solar masses.

Theoretical spectra of photoevaporating protoplanetary discs: an atlas of atomic and low-ionization emission lines

We present a calculation of the atomic and low-ionization emission line spectra of photoevaporating protoplanetary discs. Line luminosities and profiles are obtained from detailed photoionization calculations of the disc and wind structures surrounding young active solar-type stars. The disc and wind density and velocity fields were obtained from the recently developed radiation-hydrodynamic models of Owen et al. that include stellar X-ray and EUV irradiation of protoplanetary discs at various stages of clearing, from primordial sources to inner hole sources of various hole sizes. Our models compare favourably with currently available observations, lending support to an X-ray driven photoevaporation model for disc dispersal. In particular, we find that X-rays drive a warm, predominantly neutral flow where the O i 6300 Å line can be produced by neutral hydrogen collisional excitation. Our models can, for the first time, provide a very good match to both luminosities and profiles of the low-velocity component of the O i 6300 Å line and other forbidden lines observed by Hartigan et al., which covered a large sample of T-Tauri stars. We find that the O i 6300 Å and the Ne ii 12.8 μm lines are predominantly produced in the X-ray-driven wind and thus appear blueshifted by a few km s−1 for some of the systems when observed at non-edge-on inclinations. We note, however, that blueshifts are only produced under certain conditions: X-ray luminosity, spectral shape and inner hole size all affect the location of the emitting region and the physical conditions in the wind. We caution therefore that while a blueshifted line is a tell-tale sign of an outflow, the lack of a blueshift should not be necessarily interpreted as a lack of outflow. Comparison of spectrally resolved observations of multiple emission lines with detailed model sets, like the one presented here, should provide useful diagnostics of the clearing of gaseous discs.

X-ray and multiwavelength view of NGC 4278

(Abridged) Based on UV to X-ray and radio to UV flux ratios, some argue that low ionization emission line regions (LINERs) and low luminosity AGN (LLAGN) are a scaled-down version of their more luminous predecessors Seyfert galaxies. Others, based on the lack of X-ray short (hours) time-scale variability, the non-detection of an iron line at 6.4 keV, and the faint UV emission, suggest the truncation of the classical thin accretion disk in the inner regions of the AGN where a radiatively inefficient accretion flow (RIAF) structure forms. We investigate the LINER-Seyfert connection by studying the unabsorbed, AGN powered, LINER galaxy NGC 4278. We analyzed one XMM-Newton and seven Chandra X-ray observations of NGC 4278 spread over a three year period. We detected a flux increase by a factor of ~3 on a ~3 months time-scale and by a factor of 5 between the faintest and the brightest observation separated by ~3 years. During only the XMM-Newton observation, where the highest flux level is detected, we found a 10% flux increase on a ~1 hour time-scale. A combination of an absorbed power law (N(H)~10^20 cm^-2, Gamma~2.2) plus a thermal component (kT~0.6 keV) were able to fit the Chandra spectra. The XMM-Newton spectra, where the highest X-ray flux is detected, are well fitted with a single absorbed power-law. No Fe K(alpha) emission line is detected at 6.4 keV. We constructed SEDs based on simultaneous or quasi simultaneous observations and compared them to LINER, radio-loud, and radio-quiet quasar SEDs. We find that at a low X-ray flux the NGC 4278 SED resembles that of typical LINER sources where the radio to X-ray emission can be considered as originating from a jet and/or RIAF, whereas at a high X-ray flux, NGC 4278 SED is more like a low luminosity Seyfert SED. Consequently, NGC 4278 could exhibit both LINER and Seyfert nuclear activity depending on the strength of its X-ray emission.

Partly obscured accretion disc model to explain shifted broad Balmer emission lines of active galactic nuclei

We present a new model to explain the appearance of red/blue-shifted broad low-ionization emission lines, especially emission lines in optical band, which is commonly considered as an indicator of radial motion of the line emitting gas in broad emission line regions (BLRs) of Active Galactic Nuclei (AGN). We show that partly obscured disk-like BLRs of dbp emitters (AGN with double-peak broad low-ionization emission lines) can also successfully produce shifted standard Gaussian broad balmer emission lines. Then we calculate two kinds of BH masses for AGN with shifted broad balmer emission lines selected from SDSS. We find that the BH masses calculated from M-sigma relation are systematically larger than virial BH masses for the selected objects, even after the correction of internal reddening effects in BLRs. The smaller virial BH masses than BH masses from M-sigma relation for objects with shifted broad emission lines are coincident with what we expect from the partly obscured accretion disk model. Thus, we provide an optional better model to explain the appearance of shifted broad emission lines, especially for those objects with underestimated virial BH masses.

Quasars with Super–Metal‐rich Emission‐Line Regions

We study the degree of chemical enrichment in the Broad Emission Line Regions (BELRs) of two QSOs with unusually strong nitrogen emission lines. The N V 1240/ C IV 1549 intensity ratio is often used as a metallicity indicator for QSOs. The validity of this approach can be tested by studying objects in which the N IV] and N III] lines, in addition to N V, are unusually strong and easily measurable. If all of these ionization states of nitrogen point to the same metallicity, it implies that the large N V strengths observed in most QSOs are not due to some peculiarity of the N V 1240 line. This test had previously been applied to Q0353-383, a QSO long known to have extremely strong N III] and N IV] lines, with the result supporting high metallicity in that object. Here we make the same check in two other QSOs with very strong nitrogen lines, as a step towards using such QSOs to better probe the early chemical enrichment histories of their host galaxies. J1254+0241 has a metallicity of about 10x solar, with good agreement between the abundance results from different line ratios. J1546+5253 has a more moderate metallicity, about 5x solar, but the abundances determined from different line ratios show a much wider scatter than they do for J1254+0241 or Q0353-383. This QSO also has an unusual low-ionization emission line spectrum similar to some low-ionization BAL QSOs and to the unusual AGN I Zw 1. We attribute the peculiarities in its spectrum to some combination of unusual structure and/or unusual physical conditions in its BELR. Our results further affirm the validity of the N V/C IV ratio as an abundance indicator in QSOs.

SDSS J1130+0058 an X-shaped radio source with double-peaked low-ionization emission lines: a binary black hole system?

In this paper we study the object SDSS J1130+0058 which is the only AGN known to have both double-peaked low-ionization broad emission lines, and also X-shaped radio structures. Emission from an accretion disk can reproduce the double-peaked line profile of broad H$\alpha$, but not the radio structure. Under the accretion disk model, the period of the inner emission line region is about 230 years. Using a new method to subtract the stellar component from the data of the SDSS DR4, we obtain an internal reddening factor which is less than previously found. The implied smaller amount of dust disfavors the backflow model for the X-shaped radio structure. The presence of a Binary Black Hole (BBH) system is the most natural way to explain ${\it both}$ the optical and radio properties of this AGN. Under the assumption of the BBH model, we can estimate the BBH system has a separation of less than 0.04 pc with a period less than 59 years, this may pose some problem to the BLRs sizes, still we conclude that the BBH model is favored on the basis of the present limited information.

The sizes of broad emission line regions and black hole masses of double-peaked broad low-ionization emission line objects

In this paper, the sizes of the broad emission line regions (BLRs) and black hole (BH) masses of double-peaked broad low-ionization emission line emitters (DBP emitters) are compared using different methods: virial BH masses versus BH masses from stellar velocity dispersions, the size of BLRs from the continuum luminosity versus the size of BLRs from the accretion disc model. First, the virial BH masses of DBP emitters estimated by the continuum luminosity and linewidth of broad Hβ are about six times (a much larger value, if including another DBP emitters, of which the stellar velocity dispersions are traced by the linewidths of narrow emission lines) larger than the BH masses estimated from the relation MBH‐σ which is a more accurate relation to estimate BH masses. Second, the sizes of the BLRs of DBP emitters estimated by the empirical relation of RBLR‐L 5100 A are about three times (a much larger value, if including another DBP emitters, of which the stellar velocity dispersions are traced by the linewidths of narrow emission lines) larger than the mean flux-weighted sizes of BLRs of DBP emitters estimated by the accretion disc model. The higher electron density of BLRs of DBP emitters would be the main reason which leads to smaller size of BLRs than the predicted value from the continuum luminosity.

The Star Formation History of the Virgo Early‐Type Galaxy NGC 4435: TheSpitzerMid‐Infrared View

We present a population synthesis study of NGC4435, an early-type Virgo galaxy interacting with NGC4438. We combine new spectroscopic observations obtained with the Spitzer Space Telescope IRS instrument with IRAC archival data and broad band data from the literature. The IRS spectrum shows prominent PAH features, low ionization emission lines and H_2 rotational lines arising from the dusty circumnuclear disk characterizing this galaxy. The central SED, from X-ray to radio, is well fitted by a model of an exponential burst superimposed on an old simple stellar population. From the lack of high excitation nebular lines, the [NeIII]15.5/[NeII]12.8 ratio, the temperature of molecular hydrogen, and the fit to the full X-ray to radio SED we argue that the present activity of the galaxy is driven by star formation alone. The AGN contribution to the ionizing flux is constrained to be less than 2%. The age of the burst is found to be around 190 Myr and it is fully consistent with the notion that the star formation process has been triggered by the interaction with NGC4438. The mass involved in the rejuvenation episode turns out to be less than 1.5% of the stellar galaxy mass sampled in a 5 central aperture. This is enough to render NGC4435 closely similar to a typical interacting early-type galaxy with inverted CaII[H+K] lines that will later turn into a typical cluster E+A galaxy and enforces the notion that these objects are the result of a recent rejuvenation episode rather than a genuine delayed formation.

Multicolour Photometry and Spectroscopy of the Slow Nova V475 Sct (Nova Scuti 2003)

The UBVRI photometry and 460–900 nm spectroscopy of a classical nova V475 Sct obtained after its outburst in August 2003 is discussed. The nova was classified as a Fe II slow nova with the t2,V = 48 d, t3,V = 53 d. The absolute magnitudes of the nova at maximum MVmax = −7.16±0.15, MBmax = −6.96±0.39, the color excess E(B−V) = 0.69±0.05 and the distance to the nova d = 4.8±0.9 kpc was determined. Observed 13.4-day periodicity of flares can be explained either by pulsation of the nova envelope or by the mass transfer bursts from the red to the white dwarf probably caused by a periastron passage of the third body. The rapid fade of the brightness, which started 57 days after the maximum, is related to a dust formation in the ejecta of the nova. The early optical spectra display the forest of low ionization emission lines, primarily Fe II and Balmer H, accompanied by two P Cygni absorptions, arising in the inner and outer envelope of the expanding nova shell ejected at brightness maximum and accelerated by continuous stellar wind. The spectrum taken in the nebular stage of the nova, which started in March 2004, shows very strong emission [O III] 495.9 nm and 500.7 nm lines, responsible for discrepancy of the B and V magnitudes determined from observations taken by different instruments. The nebular emission line profiles suggest a nonspherical ejection of the shell.

SDSS J2125−0813: the evidence for the origination of optical Fe ii emission lines from an accretion disc near a central black hole

Abstract We report on the radio-quiet quasar SDSS J2125−0813 which obviously emits optical Fe ii emission lines and double-peaked broad Balmer emission lines. Using the accretion disc model for double-peaked, broad, low-ionization emission lines, we reproduce the composite line spectra at the optical band between 4100 and 5600 Å. Broad Fe ii emission lines can be fitted simultaneously with the broad Hβ and Mg i lines, such that all broad lines have an elliptical disc profile with the same disc parameters. This result indicates that the optical Fe ii emission lines originate from the accretion disc near the central black hole which produces the double-peaked broad Balmer emission lines. Furthermore, we find that the object has a dimensionless accretion rate , which is much larger than the accretion rate for the advection-dominated accretion flow (ADAF) mode, and that the energy budget of the accretion disc is enough to power the double-peaked broad Balmer emission lines.

ChandraReveals Galaxy Cluster with the Most Massive Nearby Cooling Core: RXC J1504.1−0248

A Chandra follow-up observation of an X-ray-luminous galaxy cluster with a compact appearance, RXC J1504.1-0248, discovered in our REFLEX Cluster Survey, reveals an object with one of the most prominent cluster cooling cores. A β-model fit to the X-ray surface brightness profile gives a core radius of ~30 h kpc, which is much smaller than the cooling radius with ~140 kpc. As a consequence, more than 70% of the high X-ray luminosity of Lbol = 4.3 × 1045 h ergs s-1 of this cluster is radiated inside the cooling radius. A simple modeling of the X-ray morphology of the cluster leads to a formal mass deposition rate within the classical cooling flow model of 1500-1900 M☉ yr-1 (for h100 = 0.7; 2300-3000 M☉ yr-1 for h100 = 0.5). The center of the cluster is marked by a giant elliptical galaxy, which is also a known radio source. Thus, it is very likely that we are observing one of the interaction systems where the central cluster active galactic nucleus (AGN) is heating the cooling core region in a self-regulated way to prevent a massive cooling of the gas, similar to several such cases studied in detail in more nearby clusters. The interest raised by this system is then due to the high power recycled in RXC J1504-0248 over cooling timescales, which is about 1 order of magnitude higher than what occurs in the studied, nearby cooling core clusters. The assumption that cooling is exactly balanced by the AGN heating implies a central black hole mass growth rate of the order of 0.5 M☉ yr-1. This cluster is therefore a prime target for the study of AGN-intracluster medium interaction at very extreme conditions. Further features common to cooling cores found in this cluster are a strong temperature drop toward the center and narrow, low-ionization emission lines in the central cluster galaxy. The cluster is also found to be very massive, with a global X-ray temperature of about 10.5 keV and a total mass of about 1.7 × 1015 h M☉ inside 3 h Mpc.

Nitrogen-deficient and iron-rich associated absorbers with oversolar metallicities towards the quasar HE0141-3932

HE0141-3932 ( $z_{\rm em} = 1.80$ ) is a bright blue radio-quiet quasar which reveals an emission line spectrum with an unusually weak Lyα line. In addition, large redshift differences ( $\Delta z = 0.05$ ) are observed between high ionization and low ionization emission lines. Absorption systems identified at $z_{\rm abs} = 1.78, 1.71$ , and 1.68 show mild oversolar metallicities ( $Z \approx 1-2Z_\odot$ ) and can be attributed to the associated gas clouds ejected from the circumnuclear region. The joint analysis of the emission and absorption lines leads to the conclusion that this quasar is seen almost pole-on. Its apparent luminosity may be Doppler boosted by ∼10 times. The absorbing gas shows high abundance of Fe, Mg, and Al ([Fe, Mg, Al/C] $\simeq 0.15\pm0.10$ ) along with underabundance of N ([N/C] $\leq -0.5$ ). This abundance pattern is at variance with current chemical evolution models of QSOs predicting [N/C] $Z \sim Z_\odot$ . Full details of this work are given in Reimers et al . (2005).