Pōniuā'ena: A Luminous z=7.5 Quasar Hosting a 1.5 Billion Solar Mass Black Hole
Jinyi Yang, Feige Wang, Xiaohui Fan, Joseph F. Hennawi, Frederick B. Davies, Minghao Yue, Eduardo Banados, Xue-Bing Wu, Bram Venemans, Aaron J. Barth, Fuyan Bian, Konstantina Boutsia, Roberto Decarli, Emanuele Paolo Farina, Richard Green, Linhua Jiang, Jiang-Tao Li, Chiara Mazzucchelli, Fabian Walter
DDraft version June 25, 2020
Typeset using L A TEX twocolumn style in AASTeX63
P¯oniu¯a‘ena: A Luminous z = 7 . Quasar Hosting a 1.5 Billion Solar Mass Black Hole
Jinyi Yang, Feige Wang, ∗ Xiaohui Fan, Joseph F. Hennawi,
2, 3
Frederick B. Davies,
2, 4
Minghao Yue, Eduardo Banados, Xue-Bing Wu,
5, 6
Bram Venemans, Aaron J. Barth, Fuyan Bian, Konstantina Boutsia, Roberto Decarli, Emanuele Paolo Farina, Richard Green, Linhua Jiang, Jiang-Tao Li, Chiara Mazzucchelli, and Fabian Walter Steward Observatory, University of Arizona, 933 N Cherry Ave, Tucson, AZ 85721, USA Department of Physics, University of California, Santa Barbara, CA 93106-9530, USA Max Planck Institut f¨ur Astronomie, K¨onigstuhl 17, D-69117, Heidelberg, Germany Lawrence Berkeley National Laboratory, 1 Cyclotron Rd, Berkeley, CA 94720, USA Kavli Institute for Astronomy and Astrophysics, Peking University, Beijing 100871, China Department of Astronomy, School of Physics, Peking University, Beijing 100871, China Department of Physics and Astronomy, University of California, Irvine, CA 92697, USA European Southern Observatory, Alonso de C´ordova 3107, Casilla 19001, Vitacura, Santiago 19, Chile Las Campanas Observatory, Carnegie Observatories, Colina El Pino, Casilla 601, La Serena, Chile INAF – Osservatorio di Astrofisica e Scienza dello Spazio di Bologna, via Gobetti 93/3, I-40129 Bologna, Italy Max Planck Institut f¨ur Astrophysik, Karl–Schwarzschild–Straße 1, D-85748, Garching bei M¨unchen, Germany Department of Astronomy, University of Michigan, 311 West Hall, 1085 S. University Ave, Ann Arbor, MI, 48109-1107, USA (Received; Revised; Accepted)
ABSTRACTWe report the discovery of a luminous quasar, J1007+2115 at z = 7 .
515 (“P¯oniu¯a‘ena”), from ourwide-field reionization-era quasar survey. J1007+2115 is the second quasar now known at z > .
5, deepinto the reionization epoch. The quasar is powered by a (1 . ± . × M (cid:12) supermassive black hole(SMBH), based on its broad Mg II emission-line profile from Gemini and Keck near-IR spectroscopy.The SMBH in J1007+2115 is twice as massive as that in quasar J1342+0928 at z = 7 .
54, the currentquasar redshift record holder. The existence of such a massive SMBH just 700 million years afterthe Big Bang significantly challenges models of the earliest SMBH growth. Model assumptions ofEddington-limited accretion and a radiative efficiency of 0.1 require a seed black hole of (cid:38) M (cid:12) at z = 30. This requirement suggests either a massive black hole seed as a result of direct collapseor earlier periods of rapid black hole growth with hyper-Eddington accretion and/or a low radiativeefficiency. We measure the damping wing signature imprinted by neutral hydrogen absorption in theintergalactic medium (IGM) on J1007+2115’s Ly α line profile, and find that it is weaker than thatof J1342+0928 and two other z (cid:38) (cid:104) x HI (cid:105) = 0 . +0 . − . . This range of values suggests a patchy reionization history toward different IGMsightlines. We detect the 158 µ m [C II ] emission line in J1007+2115 with ALMA; this line centroidyields a systemic redshift of z = 7 . ± . ∼ M (cid:12) yr − in its host galaxy. Keywords: galaxies: active — galaxies: high-redshift — quasars: individual (UHSJ100758.264+211529.207) — cosmology: observations — early universe INTRODUCTION
Corresponding author: Jinyi [email protected] ∗ NHFP Hubble Fellow
Luminous reionization-era quasars ( z > .
5) pro-vide unique probes of supermassive black hole (SMBH)growth, massive galaxy formation, and intergalacticmedium (IGM) evolution in the first billion years of theUniverse’s history. However, efforts to find such objectshave proven to be difficult because of a combination of a r X i v : . [ a s t r o - ph . GA ] J un Yang et al. the declining spatial density of quasars at high redshift,the limited sky coverage of near-infrared (NIR) photom-etry, and the low efficiency of spectroscopic follow-upobservations.During the past few years, high-redshift quasarsearches using newly available wide-area optical and IRsurveys have resulted in a sixfold increase in the num-ber of known quasars at z > .
5: 47 luminous quasars at z > . z ≥ z > . z = 7 . z (cid:38) z > z ∼ − z = 7 . z ∼ .
5, closeto the mid-point redshift of reionization (Planck Col-laboration et al. 2018). Its discovery enables new mea-surements of a quasar Ly α damping wing and providesnew constraints on the earliest SMBH growth. In thispaper, we adopt a ΛCDM cosmology with parametersΩ Λ = 0 .
7, Ω m = 0 .
3, and h = 0 . CANDIDATE SELECTION ANDOBSERVATIONSIn this section we describe the selection method thatled to the discovery of J1007+2115 and the spectro-scopic observations. This quasar was selected basedon the same photometric dataset used for our previous z ∼ . − Selection Method
We have constructed an imaging dataset by combin-ing all available optical and infrared photometric sur- veys that covers ∼ of high Galactic lati-tude sky area with z/y, J , and WISE photometry to thedepth of J ∼
21 (5 σ ), and have used this dataset tocarry out a wide-field systematic survey for quasars at z > . Wide-field Infrared Survey Explorer survey (
WISE,
Wright et al. 2010). For the
WISE pho-tometry, when we applied the selection cuts, we usedthe photometric data from the ALLWISE catalog . Toidentify quasars at z (cid:38) .
5, we required the object tobe undetected in all optical bands. We used a simpleIR color cut J − W > − . J >
5, S/N W >
5. Forced aperture photometry in all PS1 and DECaLSbands was used to reject contaminants further. Afterthe selection cuts, we visually inspected images of eachcandidate in all bands. In this step, both the ALLWISEand unWISE (Lang 2014; Meisner et al. 2018) imageswere included. All photometric data are summarized inTable 1. 2.2.
Near-infrared Spectroscopy
J1007+2115 was confirmed as a quasar during ourGemini/GNIRS run in 2019 May. The discovery spec-trum was of low quality because of the high airmasswhen it was observed. A one-hour (on-source) obser-vation with Magellan/FIRE was used further to con-firm this new quasar shortly after the GNIRS observa-tions. To obtain higher quality spectra, we observedthe quasar for 5.5 hours (on-source) with GNIRS andfor 2.2 hours (on-source) with Keck/NIRES in Mayand June of 2019. The redshift measured from theMg II line is z MgII = 7 . ± . µ m. A 1 . (cid:48)(cid:48) R ∼ . (cid:48)(cid:48)
675 slit ( R ∼ . (cid:48)(cid:48)
75 slit( R ∼ µ m http://wise2.ipac.caltech.edu/docs/release/allwise/ Rest-frame wavelength (˚A) . . . . f λ ( − e r g s − c m − ˚A − ) . . . f λ MgII . . . . f λ CIV
Observed wavelength (˚A) . . . . Observed Frequency (GHz) F l u x D e n s i t y ( m J y / b e a m ) [CII] Figure 1. Upper-left : The combined spectrum of J1007+2115 from GNIRS and NIRES data, compared with photometricdata in the
Y, J, H , and K bands (orange points with error bars). The J -band data point is from the UHS and data in other threebands are from our photometry with UKIRT obtained after the discovery of this quasar. The photometric data are consistentwith the spectrum. The purple dashed line represents the best-fit pseudo-continuum. The two inner plots show the fits to theC IV and Mg II lines. The red solid lines represent the best-fit spectra. The orange lines are the Fe II components and the bluelines denote the best-fit emission lines. Upper-right : The spectrum of the [C II ] emission line with the uncertainty (grey) andbest fit Gaussian profile (red). The [C II ] line peaks at the observed frequency 223.2 ± ± Bottem : Images (15 (cid:48)(cid:48) × (cid:48)(cid:48) , north is up and east is to the left) of J1007+2115 in PS1 z , PS1 y , DECaLS z ,UKIRT Y , UHS J , UKIRT H , and UKIRT K bands. This quasar is not detected in PS1 z , PS1 y , and DECaLS z . The 3 σ fluxlimits in these three bands are measured from our forced photometry and reported in Table 1. with a fixed 0 . (cid:48)(cid:48)
55 narrow slit resulting in a resolvingpower of R ∼ PypeIt (Prochaska etal. 2020). We corrected the telluric absorption by fittingan absorption model directly to the quasar spectra us-ing the telluric model grids produced from the Line-By-Line Radiative Transfer Model ( LBLRTM ; Clough et al.2005). We stacked the spectra from GNIRS and NIRES,weighted by inverse-variance, and scaled the result withthe K -band magnitude. The final stacked spectrum isshown in Figure 1.2.3. [C II ]-based Redshift and Dust Continuum fromALMA We observed J1007+2115 with ALMA (configurationC43-4, Cycle 7) to detect the [C II ] emission line and un-derlying dust continuum emission from the quasar host https://github.com/pypeit/PypeIt http://rtweb.aer.com/lblrtm.html galaxy. The observations were taken in 2019 Octoberwith 15 min on-source integration time. The synthe-sized beam size is 0 . (cid:48)(cid:48) × . (cid:48)(cid:48)
34 and the final data cubereaches an rms noise level of 0.4 mJy beam − per 10 kms − channel. The ALMA data were reduced with theCASA 5.4 pipeline (McMullin et al. 2007). J1007+2115is strongly detected in both the [C II ] emission line andthe continuum. The source is not spatially resolved.The [C II ] emission line provides the most accu-rate measurement of the quasar systemic redshift. AGaussian fit to the [C II ] line yields a redshift of7.5149 ± II ]-based redshift as thesystemic redshift of the quasar. We obtain a line fluxof F [CII] = 1 . ± . − , and an FWHM [CII] =331 . ± . − , corresponding to a line luminosityof L [CII] = (1 . ± . × L (cid:12) . Applying the rela-tion between star formation rate (SFR) and L [CII] forhigh-redshift ( z > .
5) galaxies from De Looze et al.(2014) which has a systematic uncertainty of a factor of ∼ [CII] ∼ − M (cid:12) yr − . This issimilar to the SFR of quasar J1342+0928 at z = 7 . Yang et al. uum is also detected, and we measure 1.2 ± µ m) and total infrared luminosities (TIR:8–1000 µ m) of L FIR = (3.3 ± × L (cid:12) and L TIR = (4.7 ± × L (cid:12) , assuming an optically thin greybody with dust temperature T d = 47 K and emissivityindex β = 1.6 (Beelen et al. 2006) and taking the effectof the cosmic microwave background (CMB) on the dustemission into account (e.g., da Cunha et al. 2013). TheSFR TIR is estimated as ∼ M (cid:12) /yr by applying thelocal scaling relation from Murphy et al. (2011). A 1.5 BILLION SOLAR MASS BLACK HOLEThe central black hole mass of the quasar can be es-timated based on its luminosity and the FWHM of theMg II line. We fit the near-IR spectrum with a pseudo-continuum, including a power-law continuum, Fe II tem-plate (Vestergaard & Wilkes 2001; Tsuzuki et al. 2006),and Balmer continuum (De Rosa et al. 2014). Gaus-sian fits of the C IV and Mg II emission lines are per-formed on the continuum-subtracted spectrum. A two-component Gaussian profile is used. The uncertainty isestimated using 100 mock spectra created by randomlyadding Gaussian noise at each pixel with its scale equalto the spectral error at that pixel (e.g. Shen et al. 2019;Wang et al. 2020). All uncertainties are then estimatedbased on the 16th and 84th percentile of the distribu-tion. The best-fit pseudo continuum and the line fittingof C IV and Mg II are shown in Figure 1.From the spectral fit, we find that the power-law con-tinuum has a slope α = − . ± .
01 ( f λ ∝ λ α ). Therest-frame 3000 ˚A luminosity is measured to be λL =(3 . ± . × erg s − , corresponding to a bolometricluminosity of L bol = (1 . ± . × erg s − assum-ing a bolometric correction factor of 5.15 (Richards etal. 2006). The apparent and absolute rest-frame 1450 ˚Amagnitudes are derived to be m , AB = 20 . ± . M , AB = − . ± .
07 from the best-fit powerlaw continuum. The line fitting of Mg II yields anFWHM = 3247 ±
188 km s − and a Mg II -based redshiftof z MgII = 7 . ± . ±
35 km s − blueshift relative to the [C II ] line, similar to other z (cid:38) IV fitting results inan FWHM of 6821 ± − . The C IV line has a3220 ±
362 km s − blueshift compared to the Mg II line.These measurements are summarized in Table 1.We estimate the black mass based on the bolomet-ric luminosity and the FWHM of the Mg II line byadopting the local empirical relation from Vestergaard& Osmer (2009). The black hole mass is derived to be M BH = (1 . ± . × M (cid:12) , resulting in an Eddington M B H [ M (cid:12) ] J1342+0928 at z = 7 . z = 6 . z = 7 . z = 7 . Universe Age (Gyr)
Figure 2.
Black hole growth of J1007+2115, comparedwith those of quasars J1342+0928 at z = 7 .
54 (Ba˜nadoset al. 2018), J1120+0641 at z = 7 .
09 (Mortlock et al. 2011),J0252–0503 at z = 7 .
00 (Wang et al. 2020), and J0100+2802at z = 6 .
33 (Wu et al. 2015). The black hole growth ismodeled as M BH = M seed exp[t/(0.05 Gyr)], assuming thatthe black holes accrete at the Eddington limit with a radia-tive efficiency of 0.1 since seed formation. The curves arenormalized to the observed black hole mass and redshift ofthese quasars. J1007+2115 requires the most massive seedblack hole under the same assumptions of black hole growth. ratio of L bol /L Edd = 1 . ± .
2. Note that the black holemass uncertainty estimated here does not include thesystematic uncertainties of the scaling relation, whichcould be up to ∼ . z (cid:38) . z = 7 . z = 7 .
5, a seed black hole with a mass of ∼ (or 3 × ) M (cid:12) would have to accrete continuouslyat the Eddington limit starting at z = 30 (or 15), assum-ing a radiative efficiency of 0.1 (see Figure 2). Under thissame set of fixed assumptions about black hole growth,J1007+2115 requires the most massive seed black holecompared to any other known quasar. This is consistentwith the direct collapse black hole seed model ratherthan the Pop III stellar remnant seed model. Even with Figure 3. a).
The intrinsic quasar spectrum from our PCA fit (red-side) and prediction (blue-side), compared with the observedspectrum in Figure 1. b).
The zoom-in Ly α region with 100 draws (thinner blue lines) from the covariant prediction errorcalibrated from the 1% of most similar quasars in the PCA training sample. c). The mock quasar transmission spectra with thevolume-averaged neutral fraction (cid:104) x HI (cid:105) = 0 . t Q = 10 . yr which are from the maximum pseudo-likelihoodmodel. The solid blue line represents the median of mock spectra and the shaded region is the the 16th–84th percentile range. a massive seed black hole, Eddington accretion witha high duty cycle and low radiative efficiency ( ∼ . CONSTRAINT ON THE IGM NEUTRALFRACTION FROM A WEAK DAMPING WINGAT Z = 7 . z >
7, the damping wing profile, detectable as ab-sorption redward of the Ly α emission line caused by thehighly neutral IGM, is one of the most promising trac-ers of the IGM neutral fraction. J1007+2115 provides uswith a new sightline to estimate the IGM neutral frac-tion through damping wing analysis at a time deep intothe reionization epoch.To estimate the IGM neutral fraction through damp-ing wing analysis, we follow the procedures describedin Davies et al. (2018a,b), which has also been used to analyze the spectra of three other luminous z (cid:38) α region using the principal component analy-sis (PCA) approach in Davies et al. (2018b). Thisapproach predicts the intrinsic blue-side quasar spec-trum (rest-frame 1175–1280 ˚A) from the red-side spec-trum (1280–2850 ˚A) using a training sample of ∼ (cid:104) x HI (cid:105) . Thismethod models the quasar transmission spectrum witha multi-scale hybrid model, which is a combination ofthe density, velocity, and temperature fields, large-scalesemi-numerical reionization simulations around massivequasar-hosting halos (Davies & Furlanetto in prep),and one-dimensional radiative transfer of ionizing pho-tons emitted by the quasar (Davies et al. 2016). Weconstruct realistic forward-modeled representations ofquasar transmission spectra, accounting for the covari-ant intrinsic quasar continuum uncertainty. We thenperform Bayesian parameter inference on the mock spec- Yang et al. . . . . . . h x HI i . . . . . P o s t e r i o r P r o b a b ili t y D e n s i t y J0252 − z = 7 . z = 7 . z = 7 . z = 7 . . . . . . . . . . . . . N e u tr a l F r a c t i o n h x H I i McGreer+2015Davies+2018Wang+2020This Work . . . . . . . Redshift − − − Fan+2006
Universe Age (Gyr)
Figure 4. Left : The posterior PDF of the volume-averaged neutral fraction (cid:104) x HI (cid:105) for J1007+2115, compared to (cid:104) x HI (cid:105) estimatedfrom the other z > Right : Constraints on theIGM neutral fraction derived from high redshift quasars through measurements of Ly α optical depth (Fan et al. 2006, blacksquares), dark gaps (McGreer et al. 2015, blue squares), and damping wings (Davies et al. 2018a; Wang et al. 2020, blue andorange pentagons). The new measurement for J1007+2115 is shown as the red filled pentagon. The dark and light grey shadedregions represent the 68% and 95% credible intervals from Planck observations (Planck Collaboration et al. 2018). These quasarmeasurements indicate a rapidly changing phase from z = 7 . z = 6 with large scatter in the neutral fraction. tra to recover the joint posterior probability distributionfunctions (PDF) of (cid:104) x HI (cid:105) and log t Q from the observedspectrum. In the Bayesian inference, the likelihood iscomputed from maximum pseudo-likelihood model pa-rameters and the pseudo-likelihood is defined as theproduct of individual flux PDFs of 500 km/s binned pix-els, equivalent to the likelihood function of the binnedtransmission spectrum in the absence of correlations be-tween pixels (see more details in Davies et al. 2018a).To measure (cid:104) x HI (cid:105) , we set a broad t Q range of 10 yr
48 at z =7 .
09 and (cid:104) x HI (cid:105) ∼ .
60 at z = 7 .
54 have been reported(Davies et al. 2018a). Recent analysis of the dampingwing feature of the quasar J0252–0503 at z = 7 . (cid:104) x HI (cid:105) = 0 .
7. All of these measurements are based onthe same methodology used in this work. We compareour result with these estimates, as shown in Figure 4.It is evident that the damping wing absorption is muchweaker in J1007+2115 compared to that in the otherthree z > α wavelength,the observed spectrum of J1007+2115 does not deviatefrom the blue-side prediction based on red-side PCA re-construction. This result is to be compared with J0252–0503 (Wang et al. 2020) where the observed spectrum is ∼
40% lower than the prediction without damping wingabsorption. The (cid:104) x HI (cid:105) estimated from J1007+2115 at z = 7 .
54 is lower than the measurements from all ofthe other three sightlines. Studies of the Ly α emissionfrom z > (cid:104) x HI (cid:105) = 0 . +0 . − . at z ∼ (cid:104) x HI (cid:105) > .
76 at z ∼ σ outlier, compared to the previous results.Although it is difficult to draw solid conclusions becauseof the large uncertainties (and the broad PDF) on thevalue of (cid:104) x HI (cid:105) , the much weaker damping wing seen inJ1007+2115’s spectrum indicates a significant scatter ofthe IGM neutral fraction in the redshift range z = 7 . z = 7 .
0, which can be interpreted as observationalevidence of patchy reionization. SUMMARYWe report the discovery of a new quasar J1007+2115with a [C II ]-based redshift of z = 7 . ± . II ] and dust continuum emis-sion from the quasar host galaxy are well detected, andimply a SFR [CII] ∼ − M (cid:12) yr − . It is only the sec-ond quasar known at such high redshift and thus pro-vides a valuable new data point for early SMBH andreionization history studies.By fitting the NIR spectrum, we derive M BH = (1 . ± . × M (cid:12) and an Eddington ratio of L bol /L Edd =1 . ± . II emission line. Theblack hole in J1007+2115 is twice as massive as thatof J1342+0928 at a very similar redshift of z = 7 . ∼ (3 × ) M (cid:12) at z = 30 (15). Through damp-ing wing modeling of the quasar spectrum, we esti-mate the volume-averaged neutral fraction to be (cid:104) x HI (cid:105) =0 . +0 . − . at z = 7 .
5. Together with three previous mea-surements from quasar damping wing analyses, our newresult indicates a large scatter of the IGM neutral frac-tion from z = 7 . z = 7 .
0, indicative of a patchyreionization process.ACKNOWLEDGMENTSThanks to Dave Osip for approving the request ofFIRE spectropsopy which is important to the confirma-tion of this quasar. J. Yang, X. Fan and M. Yue ac-knowledge the supports from the NASA ADAP GrantNNX17AF28G. F. Wang thanks the support providedby NASA through the NASA Hubble Fellowship grant
Table 1.
Photometric Properties and Derived Pa-rameters of J1007+2115.
R.A. (J2000) 10:07:58.26Decl. (J2000) +21:15:29.20 z [CII] ± m ± M –26.66 ± f λ,z PS1 a × − erg s − cm − ˚A − f λ,y PS1 a × − erg s − cm − ˚A − f λ,z DECaLS a × − erg s − cm − ˚A − Y ± J ± H ± K ± W ± W ± z MgII ± z CIV ± α λ –1.14 ± MgII − [CII] -736 ±
35 km s − ∆v CIV − MgII -3220 ±
362 km s − FWHM
MgII ±
188 km s − FWHM
CIV ± − λL (3.8 ± × erg s − L bol (1.9 ± × erg s − M BH (1.5 ± × M (cid:12) L bol /L Edd ± F [CII] ± − FWHM [CII] ± − L [CII] (1.5 ± × L (cid:12) S . ± [CII]
80 – 520 M (cid:12) yr − SFR
TIR M (cid:12) yr − a They are 3- σ flux limits in PS1 z , PS1 y , and DECaLS z bands, from our forced photometry with 3 arcsec aperturediameter. Observatory was made possible by the generous finan-cial support of the W. M. Keck Foundation. The au-thors wish to recognize and acknowledge the very sig-nificant cultural role and reverence that the summit ofMaunakea has always had within the indigenous Hawai-ian community. We are most fortunate to have theopportunity to conduct observations from this moun-tain. This research is based in part on observationsobtained at the Gemini Observatory (GN-2019B-Q-135,GN-2019A-DD-109), which is operated by the Associ-ation of Universities for Research in Astronomy, Inc.,under a cooperative agreement with the NSF on be-half of the Gemini partnership: the National ScienceFoundation (United States), National Research Coun-cil (Canada), CONICYT (Chile), Ministerio de Ciencia,Tecnolog´ıa e Innovaci´on Productiva (Argentina), Min-ist´erio da Ciˆencia, Tecnologia e Inova¸c˜ao (Brazil), andKorea Astronomy and Space Science Institute (Republicof Korea). This paper makes use of the following ALMA
Yang et al. data: ADS/JAO.ALMA
Facilities:
Gemini(GMOS), Keck(NIRES), Magel-lan(FIRE), UKIRT(WFCam), ALMA
Software:
PypeIt (Prochaska et al. 2020)REFERENCES