Rapidly accreting black hole of the Lyα-luminous quasar PSO J006.1240+39.2219
Ekaterina Koptelova, Chorng-Yuan Hwang, Matthew A. Malkan, Po-Chieh Yu
aa r X i v : . [ a s t r o - ph . GA ] J u l Draft version July 18, 2019
Typeset using L A TEX twocolumn style in AASTeX63
Rapidly accreting black hole of the Ly α -luminous quasar PSO J006.1240+39.2219 Ekaterina Koptelova, Chorng-Yuan Hwang, Matthew A. Malkan, and Po-Chieh Yu Graduate Institute of AstronomyNational Central UniversityTaoyuan City 32001, Taiwan Physics and Astronomy DepartmentUniversity of CaliforniaLos Angeles, CA 90095-1547 (Received July 23, 2018)
Submitted to ApJABSTRACTWe present near-infrared 1.1–1.3 and 1.3–1.6 µ m spectra of the Ly α -luminous quasar PSOJ006.1240+39.2219 at z = 6 . ± .
003 obtained with the NIRSPEC spectrograph at the Keck-IItelescope. The spectra cover the CIV λ λ α λ = − . ± .
26 and measure an absolute magnitude of M = − .
60. Using the scaling relation between black hole mass, width of the CIV line and ultra-violet continuum luminosity, we derive a black hole mass of (2 . ± . × M J , which is consistentbut somewhat smaller than the typical black hole masses of z & L Bol /L Edd & / CIV > Keywords: quasars: emission lines — quasars: supermassive black holes — quasars: individual (PSOJ006.1240+39.2219) INTRODUCTIONThe high-redshift quasars discovered at z & M ≈ −
26 (Ba˜nados et al.2016; Jiang et al. 2016). There is also a growing num-ber of less luminous high-redshift quasars with M & −
25 (e.g., Willott et al. 2010; Matsuoka et al. 2018;Koptelova et al. 2019). The generally high luminosi-ties of high-redshift quasars are powered by accre-tion onto massive supermassive black holes (SMBHs).The masses of the SMBHs of these quasars are usu-ally estimated from their single-epoch spectra usingthe widths of the CIV λ λ M J (e.g., Jiang et al.2006, 2007; Kurk et al. 2007, 2009; De Rosa et al. 2011,2014; Willott et al. 2010; Wu et al. 2015; Onoue et al.2019). Recently, Ba˜nados et al. (2018) reported the dis-covery of the most distant luminous quasar at z = 7 . × M J . The existence of these SMBHs provides evi-dence of their extremely rapid and efficient mass growthat early epochs. The discussed scenarios of the massgrowth of SMBHs are multiple mergers of smaller ini-tial black holes and gas accretion (e.g., Volonteri & Rees Koptelova et al. z & ∼ M J ) but ac-creting at the Eddington and super-Eddington lumi-nosities have also been discovered (Willott et al. 2010;Mazzucchelli et al. 2017). They might be at the earlystages of the growth of their SMBHs. The observationsof these less massive, less evolved SMBHs can help usto test different scenarios of the formation of SMBHs athigh redshift.The high-redshift quasars, including the most distantof them at z > z ≈ . α line of PSO J006+39is 3 to 4 times narrower (FWHM ∼ − ) than atypical broad emission line (see Peterson 1997) and im-plies a less massive SMBH than typically in the quasarsknown at z > .
5. At the same time, the luminos-ity of PSO J006+39 of M . −
26, inferred fromits discovery spectrum, is comparable to the luminosi-ties of other high-redshift luminous quasars at z > . α λ α -luminous quasar as the luminosity of its Ly α line( ∼ × L J ) constitutes almost 3% of the total lu-minosity of PSO J006+39 and is larger than the typ-ical Ly α line luminosities of quasars by 2 to 3 times.We also found evidence of a significant quasar contri-bution to the Ly α emission by observing fast variabilityof the Ly α line of PSO J006+39 on timescales of daysand weeks in the quasar rest frame (Koptelova et al.2017). The narrow and variable Ly α emission line ofPSO J006+39 makes it similar to local Narrow-lineSeyfert 1 galaxies (NLSy1; Osterbrock & Pogge 1985).NLSy1s exhibit narrow broad emission lines as theresult of their smaller SMBHs and often show vari-ability of UV lines and continuum (e.g., Collier et al.2001; Romano et al. 2002). Based on the higher ac-cretion rates and metallicities of NLSy1s in comparisonwith broad line quasars (e.g., Kuraszkiewicz et al. 2000;Shemmer & Netzer 2002; Nagao et al. 2002), Mathur(2000) concluded that they might be at early evolution-ary stages similar to high-redshift quasars. The signs ofa recent star-formation activity in the host galaxies of in-dividual NLSy1s support the idea that these objects arerelatively young (e.g., Wang et al. 2004; Wang & Wei2006). The properties of PSO J006+39 in comparisonwith local NLSy1s and known high-redshift quasars canconstrain the evolutionary stage of its SMBH.Here, we present new near-infrared observations ofPSO J006+39 obtained at the Keck Observatory whichcover the CIV λ λ . We alsopresent the analysis of the metallicity of the circumnu-clear gas of PSO J006+39 using the flux ratios of theobserved emission lines. The results presented in ourpaper provide new evidence that PSO J006+39 is in anearly phase of the black hole growth. In Section 2, we de-scribe the observations and data reduction. In Section 3,we present analysis of the continuum and emission linesof PSO J006+39, and derive the mass of its SMBH. InSection 4, we compare the black hole mass and accretionrate of PSO J006+39 with those of other high-redshiftquasars and NLSy1s, and discuss its metal abundance.In Section 5 we present our main conclusions. In the During review of our paper, Tang et al. (2019) published theresults of their analysis of the near-infrared spectrum of PSOJ006+39 obtained using Gemini North Near Infra-Red Spectro-graph (GNIRS). As suggested by our referee, we added a compar-ison between our results derived from the Subaru/FOCAS andKeck/NIRSPEC data, and from the Gemini/GNIRS data. Thecontinuum and line properties of PSO J006+39 estimated by usfrom the Gemini/GNIRS spectrum are presented in Appendix A. apidly accreting black hole of PSO J006+39 H = 67 . − Mpc − , Ω M = 0 .
31 and Ω Λ = 0 . −10 −5 0 5 10arcsec−50510 a r cs e c PSO J006+39S G NE Figure 1.
NIRSPEC-6 21 . ′′ × . ′′ OBSERVATIONS AND DATA REDUCTIONThe observations of PSO J006+39 were conducted us-ing NIRSPEC, a cross-dispersed, cryogenic, echelle spec-trograph on the Keck-II telescope (McLean et al. 1998).The data were acquired on the second half night of2017 July 28 (UT) under ∼ . ′′ . ′′
77 seeing and cloudysky with decreasing cloud coverage during the obser-vations. We used Low Resolution mode of NIRSPECwith a 0 . ′′ ′′ -long slit which provided aresolving power of R ≈ . The observed wave-length intervals corresponding to our spectral setupwere 1.1–1.3 and 1.3–1.6 µ m and included the CIV andCIII] emission lines in the NIRSPEC-2 and NIRSPEC-4 bands, respectively. During these observations, wetook 8 ×
300 s and 6 ×
300 s exposures in NIRSPEC-2 andNIRSPEC-4, respectively. The spectra of the quasarwere taken simultaneously with the spectra of a nearbystar within ∼ . ′′ The data are available at https://koa.ipac.caltech.edu/ the slit by an angle of 21 . ◦
6. This star is denoted asstar S in Figure 1. The PS1 mean y -band magnitude ofstar S is y PS1 = 19 . ± .
07 AB mag. The y PS1 -bandmulti-epoch mean magnitude of PSO J006+39 whichrepresents the typical magnitude of the quasar measuredover several PS1 epochs is y PS1 = 20 . ± .
08 AB mag(see Koptelova et al. 2017). Additionally, we took two60-s images of the quasar in each of the NIRSPEC-2,NIRSPEC-4 and NIRSPEC-6 filters using a slit-viewingcamera (SCAM). The SCAM images had a pixel scale of0 . ′′
18 pixel − and 21 . ′′ tasks andIRAF-based package WMKONSPEC . The wavelengthcalibration was performed using night sky emission linesfrom the quasar’s frames. For the identification of thesky lines we used the spectral atlas of Rousselot et al.(2000). The resulting dispersions of the NIRSPEC-2and NIRSPEC-4 spectra as measured from the sky lineswere 2.179 and 2.894 ˚A pixel − , respectively. The wave-length calibrated 2-dimensional frames of the quasarwere sky corrected, aligned and combined together. The1D NIRSPEC-2 and NIRSPEC-4 spectra were extractedfrom the combined 2-dimensional frames using IRAFtask apall . The continuum of PSO J006+39 was de-tected with low signal-to-noise ratios of SN R ≈ apall , we usedthe spectrum of brighter star S as a reference spec-trum to trace the emission of PSO J006+39. The re-sulting NIRSPEC-2 and NIRSPEC-4 spectra were cor-rected for telluric lines and absolute flux calibrated usingthe telluric standard stars, HIP8535 (type A1V, 2MASS J = 8 . ± .
030 Vega mag) and HIP114716 (type A0V,2MASS J = 6 . ± .
034 Vega mag), observed at sim-ilar airmass as the quasar. The accuracy of the fluxcalibration of the resulting NIRSPEC-2 and NIRSPEC-4 spectra is limited by the photometric accuracy ( ∼ IRAF is distributed by the National Optical Astronomy Observa-tories, which are operated by the Association of Universities forResearch in Astronomy, Inc., under cooperative agreement withthe National Science Foundation. Koptelova et al.
Figure 2.
Subaru/FOCAS and Keck/NIRSPEC spectrum of PSO J006+39 with the prominent emission lines marked. Thedisplayed spectrum is smoothed with a 10-pixel boxcar filter. Red points show the fluxes of PSO J006+39 based on the FOCAS Y -band and NIRSPEC N N N y PS1 -band multi-epoch flux of PSO J006+39 is shown by a bluepoint. The transmission curves of the NIRSPEC N N N − erg s − cm − ˚A − is presented in the lower panel. The FOCAS–NIRSPECspectrum of this figure is available as the Data behind the Figure. Table 1.
Photometry of PSO J006+39 in the Sub-aru/FOCAS Y band, and Keck/NIRSPEC N N N λ eff Magnitude(s) (˚A) (AB mag)FOCAS Y / Nov 2, 2015 270 10036 20.28 ± N ± N ± N ± night. The NIRSPEC N N N Y band on November2, 2015 simultaneously with the spectrum of the quasar.Table 1 presents the FOCAS Y band magnitude of PSOJ006+39 calibrated relative to the flux of BD+28d4211observed together with PSO J006+39 at similar airmass. RESULTS3.1.
CONTINUUM
In Koptelova et al. (2017) we derived a spectral slopeof the PSO J006+39 UV continuum of α λ = − . ± .
48 using wavelength intervals 9500–9900 and 10000–10150 ˚A (where α λ is defined such that F λ ∝ λ α λ ). In this previous analysis, the FOCAS spectrum ofPSO J006+39 was not corrected for telluric absorp-tion at wavelengths 9300–9800 ˚A. Moreover, the 9500–9900 ˚A wavelength interval includes the red wing ofthe NV line which likely affected the previous slopemeasurement. Here, we analyze the FOCAS spectrumcorrected for telluric absorption. The fraction of thequasar flux absorbed in the earth’s atmosphere was esti-mated by dividing the FOCAS spectrum of the standardstar BD+28d4211 by the flux-scaled black body modelof the star with an effective temperature of 82000 K(Latour et al. 2015). The derived atmospheric trans-mission is displayed in Figure 3. We note that the y PS1 -band magnitude of PSO J006+39 estimated fromthe corrected FOCAS spectrum is brighter by ∼ y -band light curve of PSO J006+39 shows bright-ness variations with a peak-to-peak amplitude of ∼ ∼ α line of PSO J006+39. To infer the bright- apidly accreting black hole of PSO J006+39 Figure 3.
Profiles of the Ly α , NV, OI+SiII and CII emis-sion lines of PSO J006+39. The Ly α and NV lines werefitted with Lorentzian profiles, and the OI+SiII and CIIlines – with Gaussians after subtracting the power-law con-tinuum. The fitted Ly α and NV profiles are shown withgrey lines, while the total fit is shown by a red line. The at-mospheric transmission between 9300–9800 ˚A is overplottedwith a solid thick line. ness state of PSO J006+39 at the epochs of its FO-CAS and NIRSPEC observations, we first calculatedthe spectral slope of the quasar continuum from theNIRSPEC spectrum with a wider wavelength cover-age than that of the FOCAS spectrum. Using wave-length intervals of 11100–11300, 11400–11600, 13085–13400 and 14700–15200 ˚A we measured a spectral slopeof α λ = − . ± .
26, where the quoted uncertainty isthe statistical error of the fit. The fitted power law isshown in Figure 2 with a solid line. The estimated con-tinuum slope is consistent but somewhat flatter thanthe typical slope of luminous quasars (Zheng & Malkan1993; Vanden Berk et al. 2001; Selsing et al. 2016). Wethen fitted the FOCAS data using the power law with afixed spectral slope of α λ = − .
35 and spectral windowsof 9700–9850 and 10050–10100 ˚A. The spectral windowsadopted for the analysis of the FOCAS and NIRSPECspectra were taken to be similar to the rest-frame wave-length intervals commonly used to fit the continua ofquasars (Vanden Berk et al. 2001; Decarli et al. 2010;Lusso et al. 2015) and less affected by the contributionfrom emission lines and the Big Blue Bump (BBB) (e.g.,Malkan 1983). The estimated continuum flux of PSO J006+39 at the epoch of its FOCAS observations isshown in Figure 2 with a dashed line. By comparing thecontinuum flux at the epochs of the FOCAS and NIR-SPEC observations, we find that the brightness state ofPSO J006+39 was different at these two epochs. PSOJ006+39 was brighter by about 0.8 mag during the FO-CAS observations than during the NIRSPEC observa-tions. Thus, the continuum flux of PSO J006+39 mightbe different at different epochs depending on the bright-ness state of the quasar. Figure 2 also shows the fluxesof PSO J006+39 in the FOCAS Y , and NIRSPEC N N N EMISSION LINES
To estimate the fluxes and widths of the emission linesin the FOCAS and NIRSPEC spectra of PSO J006+39,we subtracted the power-law continuum ( F λ ∝ λ − . )and fitted the lines with analytical profiles. The Ly α ,NV, CIV and CIII] lines were fitted using Lorentzianprofiles. The intrinsic profile of Ly α is likely altered byneutral hydrogen absorption seen as the series of absorp-tion lines at its blue side. However, these absorbtion fea-tures probably do not significantly affect the total fluxof the Ly α line due to its intrinsically narrow width (seeKoptelova et al. 2017). The OI+SiII and CII emissionlines were modelled using Gaussian profiles. The fittedprofiles of the emission lines are shown in Figures 3 and4. The estimated properties of the lines are summarizedin Tables 2 and 3. From the line fit, we find that, simi-lar to the Ly α line, the CIV and CIII] emission lines ofPSO J006+39 are somewhat narrower compared withtheir usual widths in broad line quasars ( > − ;see Figure 4). The redshift of PSO J006+39 esti-mated as the mean of the Ly α +NV, OI+SiII, CIVand CIII] redshifts is z = 6 . ± . µ m emission by Mazzucchelli et al. (2017)( z [CII] = 6 . ± . − relative to the [CII] 158- µ m line.The absolute magnitude of the continuum of PSOJ006+39 at rest frame wavelength 1450 ˚A estimatedfrom the NIRSPEC data is M = − . ± .
07 mag-nitudes, where the error includes a flux calibration un-certainty of 0.03 mag and an uncertainty of 0.06 magintroduced by the slope error added in quadrature. Theuncertainty introduced by the slope error was estimatedfrom a sample of simulated NIRSPEC spectra with thecontinuum slopes normally distributed around a meanof –1.35 and a standard error of 0.26. We also note that
Koptelova et al.
Table 2.
Properties of the emission lines.
Line λ peak Redshift 10 − × F line EW FWHM(˚A) (erg s − cm − ) (˚A) (km s − )Ly α . .
617 67 . ± . ± ± . .
617 24 . ± . ± ± . .
620 3 . ± . ± ± . −− . ± . ± ± . .
613 13 . ± . ± ± . .
613 4 . ± . ± ± Table 3.
Fluxes of the Ly α , NV and CIII]lines relative to the flux of the CIV line(see Table 2) measured from the FOCASand NIRSPEC spectra of PSO J006+39. Line Ly α NV CIII] F line / F CIV . ± . a . ± . a . ± . a The uncertainties take into account possi-ble changes in the line fluxes by up to 10%between the epochs of the Subaru/FOCASand Keck/NIRSPEC observations. the absolute magnitude of PSO J006+39 at the epoch ofthe FOCAS observations was M = − .
43 (assum-ing a slope of α λ = − . Figure 4.
CIV and CIII] emission lines of PSO J006+39fitted with Lorentzian profiles (red). The scaled profiles ofthe Ly α and NV lines are overplotted in blue and green forcomparison. The power-law continuum is subtracted fromthe line profiles. BLACK HOLE MASS
To estimate the mass of the SMBH of PSO J006+39,we used the empirical relation between black hole mass,CIV line width and UV continuum luminosity found byVestergaard & Peterson (2006, hereafter VP06). This
Table 4. α λ , M , λL λ (1350˚A), λL λ (3000˚A), M BH and L Bol /L Edd derived from the Keck/NIRSPEC spectrum ofPSO J006+39. α λ –1.35 ± M –25.60 ± λ L λ (1350˚A) (10 erg s − ) 1.57 ± λ L λ (3000˚A) (10 erg s − ) 1.18 ± M BH (10 M J ), VP06 2.19 ± L Bol , /L Edd , VP06 2.17 ± L Bol , /L Edd , VP06 2.21 ± M BH (10 M J ), P17 1.20 ± L Bol , /L Edd , P17 3.95 ± L Bol , /L Edd , P17 4.02 ± relation has been used for the estimation of the blackhole masses of other high-redshift quasars which allowsfor the direct comparison of our results with the previ-ous works. The intrinsic scatter of this relation is 0.36dex. It is based on the virial equation in which blackhole mass is proportional to the size of the broad lineregion ( R ) and to the square of the emission line width(FWHM ). The size of the broad line region of high-redshift quasars is generally unknown. It is estimatedusing the empirical relation between R and the contin-uum luminosity of the form R ∝ L β found from rever-beration mapping of local AGNs ( β is equal to 0.5 in therelation of VP06).The method of the black hole mass measurement frombroad emission lines of quasars relies on the assumptionthat the broad line region is virialized. However, theprofile of the CIV line of quasars is often altered bynon-virial effects (outflows or absorption in the circum-nuclear region) which make the CIV line a less reliableblack hole mass estimator compared to low-ionizationemission lines such as MgII. As discussed in VP06 andCoatman et al. (2017), the blueshifted by > − CIV line generally leads to overestimation of the CIVblack hole masses by a few times.The CIV line of PSO006+39 is symmetric and is notsignificantly blueshifted with respect to the other UVemission lines. This suggests that any possible bias inthe black hole mass measurement due to the blueshiftof the CIV line is likely small. The estimated black holemass and Eddington ratio of PSO J006+39 are summa-rized in Table 4. The quoted errors include the uncer-tainties in the line width of the CIV line and monochro-matic luminosity λL λ (1350˚A). We estimated the bolo-metric luminosity of PSO J006+39 by applying bolo-metric correction factors of 3.81 and 5.15 to the con-tinuum luminosities at 1350 and 3000 ˚A, respectively(Richards et al. 2006; Shen et al. 2008). We note that apidly accreting black hole of PSO J006+39 L Edd = 1 . × ( M BH / M J ) erg s − (e.g.,Peterson 1997). The accretion rate of PSO J006+39presented in Table 4 was estimated for two values of thebolometric luminosity, L Bol , and L Bol , , derivedfrom the continuum luminosity at 1350 and 3000 ˚A, re-spectively. In previous works, the bolometric luminosi-ties of high-redshift quasars were estimated from theircontinuum luminosities at 3000 ˚A by applying a bolo-metric correction factor of 5.15 (see De Rosa et al. 2011,2014; Mazzucchelli et al. 2017). In Section 4, to com-pare the accretion rate of PSO J006+39 with those ofother high-redshift quasars, we use the bolometric lumi-nosity estimated in the same way. De Rosa et al. (2014)used both CIV and MgII emission lines and relations ofVP06 and Vestergaard & Osmer (2009) to measure theblack hole masses of a few quasars at z & .
5. The CIVand MgII black hole masses of these quasars are consis-tent within 0.10–0.34 dex. We note that unlike the CIVline of PSO J006+39, the CIV lines of these quasarsare significantly blueshifted by & − (see alsoMazzucchelli et al. 2017). Figure 5 presents the compar-ison of the CIV black hole masses and accretion rates ofthese quasars and those of PSO J006+39.We also derived the black hole mass of PSO J006+39using the relation of Park et al. (2017, hereafter P17)found for a sample of low–redshift AGNs with high-quality HST spectra (see Table 4). P17 considered twocases, when black hole mass depends on line width asFWHM and when it depends as FWHM γ , allowing γ to be different from a physically expected value of 2.For our comparison, we used the P17 relation found for γ fixed to 2 which has an intrinsic scatter of 0.43 dex. Theblack hole mass of PSO J006+39 derived from the rela-tions of VP06 and P17 differs by a factor of two whichreflects the typical systematic uncertainties in black holemass measurements using different relations and emis-sion lines. The uncertainties in the black hole mass andaccretion rate of PSO J006+39 related to continuumvariability are also within the typical systematic uncer-tainties. DISCUSSION4.1.
Accretion rate
The early analysis of the UV-to-infrared spectra ofquasars and Seyfert galaxies by Sun & Malkan (1989)showed that luminous quasars typically have higher ac-cretion rates than low-redshift Seyfert galaxies. Theincreasing number of quasars found in different sur- -2 -1 0 1Log(L
Bol, 3000 /L Edd )7.58.08.59.09.510.010.5
Log ( M B H / M O • ) -25 -26 -27 -28 M Figure 5.
Black hole mass as a function of the Eddingtonfraction and absolute magnitude. PSO J006+39 is shownwith an open star. The arrow indicates the high lumi-nosity brightness states of PSO J006+39. The z > . veys at z > − < log( L Bol / L Edd ) < & M J are also typically between − < log( L Bol /L Edd ) <
0. There are also less luminoushigh-redshift quasars with luminosities of M > − < M J which exhibit higher accretionrates (e.g., Willott et al. 2010). De Rosa et al. (2011)analyzed a sample of 19 quasars at 4 . < z < . L Bol / L Edd ) = − . ± .
20 and log( L Bol / L Edd ) = − . ± .
24, respec-tively. The recent analysis of 11 z & . z > . M ∼ − z > . Koptelova et al. quasars. The high accretion rate of the SMBH of PSOJ006+39 compared to other luminous quasars mightsuggest that it is in an unusual phase of the rapidgrowth (Pezzulli et al. 2016; Lupi et al. 2016). The ef-ficiency of the growth of SMBHs is expected to declineat masses of ∼ M J caused by gas depletion dueto SMBH accretion, quasar outflow and star formation(e.g., Hamann et al. 2002). The high accretion rate ofPSO J006+39 indicates that there is probably plenty ofgas surrounding its SMBH which provides material forits accretion. The presence of the strong outflow thatmoves gas away from a SMBH might be indicated by theblueshift of the CIV line. In the known luminous high-redshift quasars, the CIV line is usually significantlyblueshifted by & − (e.g., De Rosa et al. 2014;Mazzucchelli et al. 2017). Compared to these quasars,the CIV line of PSO J006+39 does not show any signifi-cant blueshift. Thus, the outflow of gas in PSO J006+39is probably not too strong to prevent its SMBH from theactive growth.The accretion rate of PSO J006+39 and the am-plitude of its intrinsic brightness variations are con-sistent with the accretion rate–variability distributionof low-luminosity quasars ( L Bol < erg s − ; e.g.,Lu et al. 2019). The amplitude of the intrinsic bright-ness variations of PSO J006+39 estimated from its y PS1 -band light curve covering ∼ ∼
24% (calculated as inEdelson et al. 2002). This amplitude is also consistent(although slightly higher) with brightness variations of ∼ Metal abundance and ionization ofcircumnuclear gas
It is usually assumed that the density and ioniza-tion state of the circumnuclear gas in low- and high-redshift quasars are similar (Hamann & Ferland 1999;Nagao et al. 2006). If this assumption is correct,the flux ratios between different UV lines of high-redshift quasars indicate that their gas metallicitiesare 3–10 times of solar metallicity even at redshift z > z = 7 . N V / C I V Z/Z O • C II/ C I V Z/Z O • C III]/ C I V Z/Z O • ( O I + S i II ) / C I V Figure 6.
NV/CIV, CII/CIV, CIII]/CIV, and(OI+SiII)/CIV line ratios as a function of redshift.The right axes show the approximate metallicities estimatedusing Table 10 of Nagao et al. (2006). The NV/CIV line ra-tio of PSO J006+39 estimated from the FOCAS–NIRSPECand GNIRS data is marked with open and filled stars,respectively. The sample of 3 . < z < . z ∼ z = 6 .
08 from Kurk et al. (2009) is marked with a diamond.The CIII]/CIV line ratio of ULAS J1120+0641 at z = 7 . richment of high-redshift quasars happens rapidly andmostly ends before their luminous phase (Hamann et al.2002).To infer the metal abundance of PSO J006+39, we ex-amined the following flux ratios: NV/CIV, CIII]/CIV,CII/CIV, and (OI+SiII)/CIV presented in Figure 6.The NV/CIV, CII/CIV, and (OI+SiII)/CIV line ratiosinvolve emission lines observed with FOCAS and NIR-SPEC at two different epochs. In general, the fluxesof the same emission lines observed at different epochscould be different. In particular, the high-ionizationlines such as NV and CIV might be more variable com-pared to the CII and OI+SiII lines. The typical bright-ness change of high-ionization lines in some stronglyvariable quasars is about 10% over a few years (e.g.,Lira et al. 2018). To account for the variations of theNV and CIV line fluxes of PSO J006+39, we added inquadrature 10% of the measured line fluxes to the fluxuncertainties of these lines quoted in Table 2. Thus, theerrorbars shown in Figure 6 include uncertainties in themeasured line fluxes and uncertainties due to the varia- apidly accreting black hole of PSO J006+39 Z J (see Table 10 of Nagao et al. 2006).To derive the metallicity using the observed line fluxratios, we extrapolated the model predicted line ratiosof Nagao et al. (2006). Figure 6 shows the approxi-mate metallicity of PSO J006+39 corresponding to mea-sured line ratios NV/CIV, CIII]/CIV, and CII/CIV.For comparison, Figure 6 also shows the approximatemetallicities of the known high-redshift quasars fromDietrich et al. (2003a); Jiang et al. (2007); Kurk et al.(2009); De Rosa et al. (2014) also estimated based onthe calculations of Nagao et al. (2006). The relativeabundance of nitrogen is proportional to metallicityand therefore the line ratios such as NV/CIV are oftenused as metallicity indicators of the broad line regionsof quasars (Hamann & Ferland 1993). The NV/CIVline ratio of PSO J006+39 of > > Z J . The CIII]/CIV and CII/CIV lineratios resulting in metallicities of ∼ Z J and ∼ Z J might not be reliable metallicity indicators since theCIII] and CII lines originate at different emitting re-gions with the CIV line (e.g., Hamann et al. 2002). Be-sides, the CII/CIV and also (OI+SiII)/CIV line ratiosmight be increasing with redshift as seen for the sampleof Jiang et al. (2007). Earlier, Nagao et al. (2006) notedthat the (OI+SiII)/CIV line ratio might marginally cor-relate with redshift (see Figure 24 of Nagao et al. 2006).The analysis of a large sample of quasars byMatsuoka et al. (2011) also showed that the NV/CIVline ratio tends to be larger for quasars withmore massive SMBHs or higher accretion rates(see also Warner et al. 2003; Shemmer et al. 2004; Matsuoka et al. 2017; Xu et al. 2018). The dependanceof the NV/CIV line ratio on black hole mass may possi-bly result from the mass-metallicity relation of galaxies(e.g., Tremonti et al. 2004). However, its dependence onaccretion rate is less clear. Matsuoka et al. (2011) sug-gested that mass accretion rates onto growing SMBHsmight be associated with the post-starburst phase whenthe mass loss of the post-starburst population of stars(AGB stars) triggers quasar activity. These stars canquickly enrich the central regions of quasars’ host galax-ies with nitrogen. Nitrogen might be then ingested intothe broad line regions of quasars with stellar winds fu-elling the black hole accretion (see also Davies et al.2007). Given this scenario, the high NV/CIV line ra-tio of PSO J006+39 might be due to the local over-abundance of nitrogen rather than due to the overallhigh metallicity of its broad line region and the cen-tral region of its host galaxy. The quasars showing highNV/CIV line ratios are found to be rare at any red-shift which might suggest that this evolutionary phaseis relatively short (Araki et al. 2012; Matsuoka et al.2017). NLSy1 galaxies also typically exhibit some-what high NV/CIV line ratios for their small blackhole masses (Shemmer & Netzer 2002; Shemmer et al.2004). In Figure 7, we show the NV/CIV line ratio ofPSO J006+39 in comparison with that of the quasars at2 . < z < . z ∼ Z J (Shemmer & Netzer2002; Nagao et al. 2002; Fields et al. 2005). Given therelatively small SMBH of PSO J006+39 compared toother z > . α , CIII] and CIV lines aresensitive to the density and ionization state of the cir-cumnuclear gas. In Figure 8, we plot the Ly α /CIV andCIII]/CIV line ratios of PSO J006+39 in comparisonwith the line ratios of z < z > . α /CIV0 Koptelova et al. BH /M O • )0.11.0 N V / C I V -1.0 -0.5 0.0 0.5 1.0Log(L Bol, 3000 /L Edd ) Figure 7.
NV/CIV line ratio vs. black hole mass and ac-cretion rate. The NV/CIV line ratio of PSO J006+39 es-timated from the FOCAS–NIRSPEC and GNIRS data ismarked with open and filled stars, respectively. The quasarsat z ∼ L Bol /L Edd ) = − . − . − .
5. The dashed, thick andthin lines in the right panel correspond to median black holemasses log( M BH /M J ) = 9 .
7, 9.3, 8.9. and CIII]/CIV line ratios of PSO J006+39 correspondto a lower ionization parameter of the circumnuclear gasin comparison with the known z > . α /CIV line ra-tio from the FOCAS–NIRSPEC, and GNIRS data in-dicates that the ionization parameter of PSO J006+39may vary significantly depending on the brightness stateof the quasar. The low level of gas ionization due to thehigh density of gas is more typical for the broad lineregions of NLSy1s. As also seen in Figure 8, the gasdensity in the broad line region of PSO J006+39 is con-sistent with the typical gas densities in the broad lineregions of quasars at z <
5. Therefore, the overall metalabundance of its circumnuclear gas should not be toodifferent from the typical metal abundances of quasarsexcept for the high nitrogen abundance. The densityof the circumnuclear gas of PSO J006+39 seems to beone order higher than typically in the known quasars at z > . z > . α /CIV line ratios as their Ly α lineflux is usually severely reduced by neutral hydrogen ab-sorption. The inferred density of gas and low level ofionization similar to that of NLSy1s might imply that PSO J006+39 is at the early evolutionary stage of lumi-nous quasars. L y α / C I V l o g U = − . − . − . l o g n H = . l og n H = . l og n H = . ULAS J1120 VISTA J0109VISTA J0305PSO J006
Figure 8.
CIII]/CIV vs. Ly α /CIV. The line ratios ofPSO J006+39 from the FOCAS and NIRSPEC, and GNIRSdata are shown with open and filled stars, respectively.The line ratios of NLSy1 galaxies from Kuraszkiewicz et al.(2000) (open diamonds), of quasars at 3 . < z < . < z < . z = 7 . z = 6 .
75 and VISTAJ0305-3150 at z = 6 .
61 (see De Rosa et al. 2014). (Astheir CIII]/CIV line ratios are not measured, we assumedthat they are close to those of ULAS J1120+0641 andPSO J006+39). The measured Ly α /CIV line ratios ofPSO J006+39, ULAS J1120+0641, VISTA J0109-3047 andVISTA J0305-3150 represent lower limits, since the Ly α flux of high-redshift quasars is typically reduced by neu-tral hydrogen absorption. The grid of the theoretical lineratios calculated by Kuraszkiewicz et al. (2000) for gas den-sities log n H = 9 . , . , . U = − . , − . , − . CONCLUSIONSWe presented the new Keck/NIRSPEC observationsand combined analysis of the Subaru/FOCAS andKeck/NIRSPEC data of the Ly α -luminous quasar PSOJ006.1240+39.2219 at z = 6 . ± . apidly accreting black hole of PSO J006+39 α λ = − . ± .
26, whichagrees but is slightly flatter than the typical contin-uum slope of quasars. The estimated rest-frame UVabsolute magnitude of PSO J006+39 at the epoch of itsKeck/NIRSPEC observations is M = − . ± . M BH =(2 . ± . × M J and is about 3–4 times smallerthan the typical masses of the SMBHs of high-redshiftquasars within a luminosity range of − . < M < − . & & Z J indicated by the NV/CIV line ratiois somewhat high for the mass of its SMBH. Such high metallicity is unlikely in the early phases of black holeformation. On the other hand, the NV/CIV line ratioof PSO J006+39 might not be due to its high metal-licity. It might reflect a particular evolutionary phasecharacterized by the high abundance of nitrogen. Thishigh abundance of nitrogen might have been producedby the post-starburst population of stars which couldtrigger the quasar activity of PSO J006+39 by provid-ing the fuel for black hole accretion. The Ly α /CIV andCIII]/CIV line ratios also indicate a lower level of ion-ization of the circumnuclear gas than usually in quasars,which is more typical for NLSy1 galaxies, – the low-redshift analogues of high-redshift quasars. Thus, theevolutionary phase of PSO J006+39 differs from that ofother known high-redshift quasars of similar luminosi-ties. Its intensively growing black hole might be in theearly phase of quasar activity.ACKNOWLEDGMENTSThe data presented herein were obtained at the W.M. Keck Observatory (Program U055, PI: M. Malkan),which is operated as a scientific partnership among theCalifornia Institute of Technology, the University of Cal-ifornia and the National Aeronautics and Space Admin-istration. The Observatory was made possible by thegenerous financial support of the W. M. Keck Founda-tion. Support for this work was provided by the Ministryof Science and Technology of Taiwan, grant Nos MOST105-2119-M-007-022-MY3, MOST 107-2119-M-008-009-MY3, and MOST 107-2811-M-008-2524. The authorswish to recognize and acknowledge the very significantcultural role and reverence that the summit of Mau-nakea has always had within the indigenous Hawaiiancommunity. We are most fortunate to have the oppor-tunity to conduct observations from this mountain.REFERENCES Araki, N., Nagao, T., Matsuoka, K., et al. 2012, A&A, 543,A143Baldwin, J., Ferland, G., Korista, K., & Verner, D. 1995,ApJS, 455, 119Ba˜nados, E., Venemans, B. P., Decarli, R., et al. 2016,ApJS, 227, 11Ba˜nados, E., Venemans, B. P., Mazzucchelli, C., et al. 2018,Nature, 553, 473Chambers, K.C., Magnier, E. A., Metcalfe, N., et al. 2016,arXiv:1612.05560Coatman, L., Hewett, P. C., Banerji, M., et al. 2017,MNRAS, 465, 2120Collier, S., Crenshaw, D. M., Peterson, B. M., et al. 2001,ApJ, 561, 146 Davies, R. I., M¨uller S´anchez, F., Genzel, R., et al. 2007,ApJ, 671, 1388De Rosa, G., Decarli, R., Walter, F., et al. 2011, ApJ, 739,56De Rosa, G., Venemans, B. P., Decarli, R., et al. 2014, ApJ,790, 145Decarli, R., Falomo, R., Treves, A., et al. 2010, MNRAS,402, 2441Dietrich, M., Appenzeller, I., Hamann, F., et al. 2003a,A&A, 398, 891Dietrich, M., Hamann, F., Shields, J. C., et al. 2003b, ApJ,589, 722Dye, S., Lawrence, A., Read, M. A., et al. 2018, MNRAS,473, 5113 Koptelova et al.
Edelson, R., Turner, T. J., Pounds, K., et al. 2002, ApJ,568, 610Fields, D. L., Mathur, S., Pogge, R. W., et al. 2005, ApJ,634, 928Hamann, F., & Ferland, G. 1993, ApJ, 418, 11Hamann, F., & Ferland, G. 1999, ARA&A, 37, 487Hamann, F., Korista, K. T., Ferland, G. J., Warner, C., &Baldwin, J. 2002, ApJ, 564, 592Hawarden, T. G., Leggett, S. K., Letawsky, M. B.,Ballantyne, D. R., & Casali, M. M. 2001, MNRAS, 325,563Hewett, P. C., Warren, S. J., Leggett, S. K., & Hodgkin,S. T. 2006, MNRAS, 367, 454Jiang, L., Fan, X., Hines, D. C., et al. 2006, AJ, 132, 2127Jiang, L., Fan, X., Vestergaard, M., et al. 2007, AJ, 134,1150Jiang, L., McGreer, I. D., Fan, X., et al. 2016, ApJ, 833, 222Juarez, Y., Maiolino, R., Mujica, R., et al. 2009, A&A, 494,L25Kaiser, N., Aussel, H., Burke, B. E., et al. 2002,Proc. SPIE, 4836, 154Kaiser, N., Burgett, W., Chambers, K., et al. 2010,Proc. SPIE, 7733, 77330EKoptelova, E., Hwang, C.-Y., Yu, P.-C., Chen, W.-P., &Guo, J.-K. 2017, Scientific Reports, 7, 41617Koptelova, E., & Hwang, C.-Y. 2019, arXiv:1907.02721,ApJL in pressKuraszkiewicz, J., Wilkes, B. J., Czerny, B., & Mathur, S.2000, ApJ, 542, 692Kurk, J. D., Walter, F., Fan, X., et al. 2007, ApJ, 669, 32Kurk, J. D., Walter, F., Fan, X., et al. 2009, ApJ, 702, 833Latour, M., Fontaine, G., Green, E. M., & Brassard, P.2015, A&A, 579, 39Lira, P., Kaspi, S., Netzer, H., et al. 2018, ApJ, 865, 56Lu, K.-X., Huang, Y.-K., Zhang, Z.-X., et al. 2019, ApJ,877, 23Lupi, A., Haardt, F., Dotti, M., et al. 2016, MNRAS, 456,2993Lusso, E., Worseck, G., Hennawi, J. F., et al. 2015,MNRAS, 449, 4204MacLeod, C. L., Ivezi´c, ˇZ., Kochanek, C. S., et al. 2010,ApJ, 721, 1014Madau, P., Haardt, F., & Dotti, M. 2014, ApJL, 784, L38Malkan, M. A. 1983, ApJ, 268, 582Mathur, S. 2000, MNRAS, 314, L17Matsuoka, K., Nagao, T., Maiolino, R., et al. 2017, A&A,608, A90Matsuoka, K., Nagao, T., Marconi, A., Maiolino, R., &Taniguchi, Y. 2011, A&A, 527, A100 Matsuoka, Y., Iwasawa, K., Onoue, M., et al. 2018, ApJS,237, 5Mazzucchelli, C., Ba˜nados, E., Venemans, B. P., et al. 2017,ApJ, 849, 91McLean, I. S., Becklin, E. E., Bendiksen, O., et al. 1998,Proc. SPIE, 3354, 566Mortlock, D. J., Warren, S. J., Venemans, B. P., et al. 2011,Nature, 474, 616Nagao, T., Marconi, A., & Maiolino, R. 2006, A&A, 447,157Nagao, T., Murayama, T., Shioya, Y., & Taniguchi, Y.2002, ApJ, 575, 721Onoue, M., Kashikawa, N., Matsuoka, Y., et al. 2019,arXiv:1904.07278, ApJ, in pressOsterbrock, D. E., & Pogge, R. W. 1985, ApJ, 297, 166Park, D., Barth, A. J., Woo, J.-H., et al. 2017, ApJ, 839, 93Pentericci, L., Fan, X., Rix, H.-W., et al. 2002, AJ, 123,2151Peterson, B. M. 1997, An Introduction to Active GalacticNuclei (Cambridge: Cambridge Univ. Press)Pezzulli, E., Valiante, R., & Schneider, R. 2016, MNRAS,458, 3047Planck Collaboration XIII, 2016, A&A, 594, A13Richards, G. T., Lacy, M., Storrie-Lombardi, L. J., et al.2006, ApJS, 166, 470Romano, P., Turner, T. J., Mathur, S., & George, I. M.2002, ApJ, 564, 162Rousselot, P., Lidman, C., Cuby, J.-G., Moreels, G., &Monnet, G. 2000, A&A, 354, 1134Selsing, J., Fynbo, J. P. U., Christensen, L., & Krogager,J.-K. 2016, A&A, 585, A87Shemmer, O., & Netzer, H. 2002, ApJL, 567, L19Shemmer, O., Netzer, H., Maiolino, R., et al. 2004, ApJ,614, 547Shen, Y., Greene, J. E., Strauss, M. A., Richards, G. T., &Schneider, D. P. 2008, ApJ, 680, 169Sun, W.-H., & Malkan, M. A. 1989, ApJ, 346, 68Tang, J.-J., Goto, T., Ohyama, Y., et al. 2017, MNRAS,466, 4568Tang, J.-J., Goto, T., Ohyama, Y., et al. 2019, MNRAS,484, 2575Tremonti, C. A., Heckman, T. M., Kauffmann, G., et al.2004, ApJ, 613, 898Vanden Berk, D. E., Richards, G. T., Bauer, A., et al. 2001,AJ, 122, 549Vestergaard, M., & Peterson, B. M. 2006, ApJ, 641, 689Vestergaard, M., & Osmer, P. S. 2009, ApJ, 699, 800Volonteri, M. 2010, A&A Rv, 18, 279Volonteri, M., & Rees, M. J. 2005, ApJ, 633, 624 apidly accreting black hole of PSO J006+3913