The Infrared Emission and Vigorous Star Formation of Low-redshift Quasars
aa r X i v : . [ a s t r o - ph . GA ] F e b Draft version February 5, 2021
Typeset using L A TEX default style in AASTeX63
The Infrared Emission and Vigorous Star Formation of Low-redshift Quasars
Yanxia Xie, Luis C. Ho,
Ming-Yang Zhuang, and Jinyi Shangguan Kavli Institute for Astronomy and Astrophysics, Peking University, Beijing 100871, China Department of Astronomy, School of Physics, Peking University, Beijing 100871, China Max Planck Institute for Extraterrestrial Physics, Garching, Germany
ABSTRACTThe star formation activity of the host galaxies of active galactic nuclei (AGNs) provides valuableinsights into the complex interconnections between black hole growth and galaxy evolution. A majorobstacle arises from the difficulty of estimating accurate star formation rates in the presence of a strongAGN. Analyzing the 1 − µ m spectral energy distributions and high-resolution mid-infrared spectraof low-redshift ( z < .
5) Palomar-Green quasars with bolometric luminosity ∼ . − . erg s − ,we find, from comparison with an independent star formation rate indicator based on [Ne II] 12.81 µ mand [Ne III] 15.56 µ m, that the torus-subtracted, total infrared (8 − µ m) emission yields robuststar formation rates in the range ∼ − M ⊙ yr − . Combined with available stellar mass estimates,the vast majority ( ∼ − ∼ − Keywords: galaxies: active — galaxies: ISM — galaxies: nuclei — galaxies: Seyfert — (galaxies:)quasars: general — infrared: ISM INTRODUCTIONSupermassive black holes (BHs) are widely regarded as being closely connected with the evolution of galaxies (Rich-stone et al. 1998; Kormendy & Ho 2013; Heckman & Best 2014), but the exact manner in which active galactic nuclei(AGNs) truly impact their host galaxies remains a topic of lively debate. Does the BH grow in concert with the starsof the host, or does one component lag behind the other? Much attention has been devoted to the subject of AGNfeedback (Fabian 2012), but does AGN feedback inhibit or stimulate star formation? These issues can be clarified if wehave access to the ongoing star formation rate (SFR) and stellar mass ( M ∗ ) of the host galaxies of AGNs of differenttypes and in different stages of their evolution. In the context of the general galaxy population, star-forming galaxiesoccupy a well-defined main sequence, a relation between SFR and M ∗ that encodes vital information on the mannerand timescale in which galaxies acquire their stellar mass (e.g., Brinchmann et al. 2004; Elbaz et al. 2007; Noeske etal. 2007; Peng et al. 2010; Speagle et al. 2014; Barro et al. 2017). This same framework serves as a useful guide forprobing AGN host galaxies and their relation to the overall galaxy population.Previous attempts to investigate the star-forming main sequence of AGN host galaxies have yielded mixed results.The global SFRs of low to moderate-luminosity AGNs generally lie on or below the star-forming main sequence forredshifts 0 . < z < Xie et al. regime of extreme starbursts (e.g., Kirkpatrick et al. 2020). At the same time, seemingly contradictory conclusionsare reached by works that report that star formation activity is suppressed in luminous AGNs (e.g., Scholtz et al.2018; Stemo et al. 2020). The root causes of these diverse and potentially conflicting results stem from several factors,including systematics of sample selection, the accuracy of SFR and M ∗ measurements, and even the very definitionof the “starburst” phenomenon or the “main sequence” now commonly used as a reference to discuss the level of starformation in galaxies.The cold gas content of AGN host galaxies can shed additional light on these issues. The increasing availability ofcold gas measurements from direct observations of CO and H I (e.g., Evans et al. 2001, 2006; Scoville et al. 2003; Hoet al. 2008a; Wang et al. 2013, 2016; Walter et al. 2014; Xia et al. 2014; Brusa et al. 2015; Husemann et al. 2017;Kakkad et al. 2017; Shangguan et al. 2020a) or from indirect estimates based on dust emission and absorption (e.g.,Shangguan et al. 2018; Shangguan & Ho 2019; Yesuf & Ho 2019, 2020a, 2020b; Zhuang & Ho 2020) offers the possibilityof probing the star formation efficiency (SFE ≡ SFR /M gas ), which provides a complementary and independent probeof how effectively the cold gas converts to stars in AGN host galaxies (Husemann et al. 2017; S´anchez et al. 2018;Jarvis et al. 2020; Shangguan et al. 2020b; Zhuang et al. 2021).As the SFR plays a pivotal role in these considerations, the reliability with which it can be ascertained in AGN hostgalaxies becomes a central concern. Measuring the SFR in active galaxies is challenging because essentially all theconventional tracers of young stars in inactive galaxies (Kennicutt 1998a) are subject to contamination, to one extentor another, by radiation from the AGN. The problem is most severe for type 1 (unobscured, broad-line) sources, buttype 2 (obscured, narrow-line) systems also suffer. SFRs derived from conventional tracers such as H α are problematicbecause the AGN narrow-line region itself is a prodigious source of hydrogen recombination emission. Moreover, themid-infrared (IR) PAH features may be destroyed in the harsh environment of AGNs (Voit 1992). Motivated by Ho &Keto (2007), Zhuang et al. (2019) proposed an effective SFR estimator for AGNs based on the fine-structure lines of[Ne II] 12.81 µ m and [Ne III] 15.56 µ m. The principal limitation of this method is the current scarcity of the requisitemid-IR spectroscopy. In practice, the optical [O II] λ m = 0 . Λ = 0 . H = 67 . − Mpc − (Ade et al. 2016). TheSFRs and stellar masses in this paper are all scaled to the stellar initial mass function of Salpeter (1955). DATA AND MEASUREMENTSWe focus on the sample of 86 z < . from the Palomar-Green (PG) survey (Schmidt & Green 1983),as summarized in Boroson & Green (1992). Originally selected based on ultraviolet/optical colors, PG quasars forma representative sample of luminous, broad-line (type 1) AGNs unbiased with respect to dust or gas content. The The Boroson & Green (1992) sample officially contains 87 PG quasars. We exclude PG 1226+023 (3C 273) from this study because itsIR emission is dominated by synchrotron radiation from a prominent jet, which renders the modeling of the spectral energy distribution(SED) highly uncertain (Shangguan et al. 2018). low-redshift subset of PG quasars, in particular, has been extensively and systematically studied, providing a richlegacy of multi-wavelength data and physical parameters for the active nuclei and their host galaxies. For instance,BH masses can be estimated (Vestergaard & Peterson 2006; Ho & Kim 2015) from available optical spectra (Boroson& Green 1992; Ho & Kim 2009), which, in combination with optical luminosities or more complete broad-band SEDs(Neugebauer et al. 1987; Sanders et al. 1989; Shang et al. 2011) furnish bolometric luminosities and hence Eddingtonratios.An important parameter used in our study is the total stellar mass ( M ∗ ) of the host galaxy. Zhang et al. (2016)provide estimates of M ∗ for 55 of the PG quasars based on analysis of high-resolution Hubble Space Telescope (HST)optical and near-IR images. For the remaining 31 sources, we infer their M ∗ from the BH mass ( M BH ) using theempirical correlation (Greene et al. 2020) log (cid:18) M BH M ⊙ (cid:19) = (7 . ± .
09) + (1 . ± .
12) log (cid:18) M ∗ × M ⊙ (cid:19) , (1)which has an intrinsic scatter of 0.65 dex. Optical or near-IR morphologies have been published for 37 sources.Supplemented with new HST observations (Y. Zhao et al. 2021), 71 of the 86 sources now have HST images usefulfor morphological classification. The host galaxies of PG quasars exhibit diverse morphologies (Table 1). Apart fromelliptical galaxies and obvious major mergers, a number of sources display a normal or only a mildly perturbed disk.For the purposes of the present study, we consider all systems with a disturbed morphology or that possess a nearbycompanion as mergers.As our primary aim is to utilize the mid-IR emission lines of [Ne II] 12.81 µ m, [Ne III] 15.56 µ m, and [Ne V] 14.32 µ mto establish a reliable, independent estimate of the SFR (Zhuang et al. 2019), we rely on high-resolution spectra takenwith the Infrared Spectrograph (IRS; Houck et al, 2004) on the Spitzer Space Telescope (Werner et al. 2004), whichare available for 37 out of the 86 PG quasars. The high-resolution spectra comprise data for 34 objects taken with theshort-high mode, which samples 9 . − . µ m with a 4 . ′′ × . ′′ . − . µ m with a 11 . ′′ × . ′′ λ/ ∆ λ ≈ S Ne , defined as the ratioof the integrated line flux to the FWHM of the line. We consider a line detected if S Ne ≥ σ , with σ the standarddeviation of the flux density of the local continuum (Xie & Ho 2019). We calculate 3 σ upper limits by fixing the linewidth to the observed median value of the detected sources (FWHM = 459 km s − for [Ne II] and FWHM = 582km s − for [Ne III] and [Ne V]). A total of 26 sources are detected in [Ne II], [Ne III], and [Ne V]; 11 sources haveupper limits on [Ne V], among them seven with upper limits on [Ne II]; and one has none of the three lines detected.From Zhuang et al. (2019),SFR Ne ( M ⊙ yr − ) = 4 . × − ( Z ⊙ /Z ) (cid:20) L [Ne II]+[Ne III] − . L [Ne V] f + + 1 . f +2 (cid:21) , (2)where L [Ne II]+[Ne III] is the luminosity of the sum of [Ne II] 12.81 µ m and [Ne III] 15.56 µ m, L [Ne V] is the luminosityof [Ne V] 14.32 µ m (both in units of erg s − ), Z is the gas-phase metallicity of the galaxy, and f + and f +2 are the Equation 1 assumes a “diet” Salpeter initial mass function (Bell et al. 2003), which is very similar to Kroupa’s (2001) initial mass function.We adjust the zero point of Equation 1 to be consistent with our assumed Salpeter initial mass function.
Xie et al.
Wavelength ( µ m)12.71 12.81 12.91 F l ux D e n s it y ( J y ) Wavelength ( µ m)14.22 14.32 14.420.050.10 Wavelength ( µ m)15.46 15.56 15.660.050.10 [Ne II] 12.81 µ m [Ne V] 14.32 µ m [Ne III] 15.56 µ m (a) PG 1224+026 (b) (c) Figure 1.
Illustration of the fitting of the high-resolution IRS spectrum of PG 1224+026, for the lines (a) [Ne II] 12.81 µ m,(b) [Ne V] 14.32 µ m, and (c) [Ne III] 15.56 µ m. In each panel, the crosses with error bars denote the observed flux density andits corresponding 1 σ uncertainty. The red solid curve gives the best-fit single Gaussian profile to the line and the red dashedline marks the underlying continuum. fractional abundances of singly and doubly ionized neon, which can be determined from Equation 8 in Zhuang et al.(2019). We infer the metallicity of the host galaxy indirectly from the stellar mass-metallicity relation (see Section 4.1of Xie & Ho 2019). Since most quasar hosts have M ∗ & M ⊙ (Section 3.2), above which the stellar mass-metallicityrelation flattens (e.g., Kewley & Ellison 2008), we simply adopt the metallicity inferred at M ∗ = 10 M ⊙ for quasarsmore massive than this value. We use bootstrapping resampling to estimate the total error budget for SFR Ne , takinginto account uncertainties from the metallicity, line measurement, and other parameters involved in Equation 2. Theionization state of an AGN reflects its Eddington ratio, which, because of observational selection effects in mostsamples, is also related to its absolute luminosity (Ho 2008, 2009). The neon-based SFR formalism of Zhuang etal. (2019) only applies to high-ionization quasars and classical Seyfert nuclei, AGNs sufficiently powerful to emitdetectable [Ne V ] 14.32 µ m. Without the presence of [Ne V ] 14.32 µ m as a guide, it would be impossible to ascertainthe extent to which the lower ionization transitions of [Ne II ] and [Ne III ] might arise from the narrow-line region. Inlow-accretion rate, low-ionization AGNs (e.g., LINERs), most of the neon emission emerges as [Ne II ] and [Ne III ].Since the quasars in our sample are incontrovertibly luminous AGNs, the non-detection of [Ne V ] in 12 of the sourcesmust be due to observational sensitivity, and we treat the [Ne V ] non-detections as upper limits. The upper limits in[Ne V ] and in the lower-ionization lines translate to a range of allowed SFR Ne , depending on the limiting values of f + and f +2 that appear in the denominator of Equation 2.To evaluate whether reliable SFRs can be derived using the total IR dust continuum emission, we utilize the 70, 100,160, 250, 350, and 500 µ m photometry acquired by Petric et al. (2015) using the Herschel Space Observatory (Pilbrattet al. 2010). The scan-mode observations of Herschel ensure that all of the far-IR flux is captured. Shangguan et al.(2018) analyzed these far-IR observations, in combination with the IRS mid-IR spectra and sky-survey photometry atshorter wavelengths, to construct complete SEDs covering ∼ − µ m. To derive dust masses for the host galaxies,Shangguan et al. (2018) used several physical components to decompose the SEDs, including a population synthesismodel for the stars (Bruzual & Charlot 2003), hot dust emission for the AGN torus (Nenkova et al. 2008), and colddust emission for the global interstellar medium of the host galaxy (Draine & Li 2007). The total IR luminosity ( L IR )of the host is calculated by integrating the luminosity of the Draine & Li component of the best-fit model from 8 to1000 µ m, which converts to (Kennicutt 1998a)SFR IR ( M ⊙ yr − ) = 4 . × − L IR (erg s − ) . (3)The Draine & Li model contains four main parameters: (1) U min , the intensity of the radiation field of the diffuseinterstellar medium, which scales with the dust temperature; (2) γ , the mass fraction of the dust in the photodissoci-ation region; (3) q PAH , the mass fraction of the dust in the form of polycyclic aromatic hydrocarbons (PAHs); and (4) M d , the total dust mass, which sets the overall normalization of the model. The SEDs of 11 objects are significantly l og ( L bo l / e r g s - ) Ne / M O • yr -1 )0123 l og ( SF R I R / M O • y r - ) : δ (y-x) = 0.02 -0.44+0.34 dex Neon detection Neon upper limit
Figure 2.
Comparison between SFRs derived using the total IR (8 − µ m) luminosity with those derived using the luminosityof the mid-IR neon lines. The data points are color-coded according to L bol . Circles denote objects with all neon lines detected;triangles mark objects for which one or more of the neon lines is an upper limit, with the dotted line indicating the range ofSFR bracketed by Equation 2. The dashed line is the 1:1 relation. The median (16%, 84%) of the difference between the twoSFRs (log SFR IR − log SFR Ne ) is 0 . +0 . − . dex. influenced by non-detections in the Herschel bands. Following Shangguan et al. (2018), we adjust the parameters in thefit that would produce the most conservative upper limit on L IR consistent with the far-IR upper limits. After muchexperimentation, we choose to fix U min = 1 and q PAH = 0 .
47, as they barely affect the final results. The parameter γ directly correlates with L IR but only impacts the global fit negligibly because of the dominance of the AGN torusin the mid-IR. We fix γ in the range 0 . − .
3, adjusting only M d to best match the far-IR upper limits. A value of γ = 0 . RESULTS3.1.
Far-IR Luminosity as a SFR Estimator for Quasars
Using the mid-IR neon lines as an independent indicator of SFR in AGNs (Zhuang et al. 2019), we are now in aposition to assess the extent to which the far-IR luminosity tracks star formation, for the 37 PG quasars for which bothtracers can be measured. Figure 2 shows that SFR IR and SFR Ne follow a 1:1 relation, with no noticeable systematicdependence on AGN bolometric luminosity. Using the Weibull distribution as implemented in the Python package lifelines (Davidson-Pilon et al. 2020) to account for the interval of SFR Ne allowed by the upper limits in theneon lines, we find that the median (16%, 84%) of the difference between the two SFRs (log SFR IR − log SFR Ne ) is0 . +0 . − . dex. The consistency between these two measures of SFR suggests that the global dust content of the hostgalaxies is heated predominantly by young stars, and that the torus-subtracted, total IR (8 − µ m) luminosity https://lifelines.readthedocs.io/en/latest/ Xie et al. M ∗ / M O • )-1 l og ( SF R / M O • y r - ) z = 0, Saintonge et al. (2016)z = 0.2, Speagle et al. (2014) star-forming galaxies starburst LIRGs uncertain no data disks ellipticals mergers Figure 3.
PG quasars lie above the star-forming galaxy main sequence at z ≈ ± . σ scatter) and at z ≈ . ± . σ scatter). Filled red symbols indicate objects having direct stellarmass ( M ∗ ) measurements from high-resolution optical or near-IR images, while open red symbols represent objects with M ∗ estimated from the M BH − M ∗ relation. The typical uncertainties for SFR are 0.4 dex, and for M ∗ they are 0.2 dex for directand 0.65 dex for indirect measurements, as indicated by the red solid and dashed lines, respectively, in the lower-right corner.Different symbols represent different host galaxy morphologies, as given in the legend. Shown for reference are star-forminggalaxies from the xCOLD GASS sample (Saintonge et al. 2017; filled grey points) and starburst LIRGs (Shangguan et al. 2019;filled blue points). provides an unbiased estimator of the SFR, even for relatively powerful, unobscured quasars, such as those containedin the low-redshift PG sample. Reliable uncertainties are difficult to estimate, however. The total uncertainty ofSFR Ne is at least ∼ . IR or L IR , mostly . .
05 dex, likely are severely underestimated. As discussed in Shangguan et al. (2018), the SED fits used to derive L IR sample a limited grid of parameters of the Draine & Li (2007) models. From comparison of SED fits using differentassumptions for the torus component and analysis of mock data (Appendices C and D in Shangguan et al. 2018), thetrue uncertainties on L IR are closer to ∼ . − . L bol ≈ . − . erg s − ) and 2.5 orders of magnitude in star formationactivity (SFR ≈ − M ⊙ yr − ) spanned by our sample, the observed scatter of the correlation in Figure 2 is only ∼ IR to measure star formation activity, whose final errorbudget is assumed conservatively to be 0.4 dex.3.2. Quasar Hosts and the Galaxy Star-forming Main Sequence
Having established that we can trust the SFRs derived from the torus-subtracted, total IR luminosity, we now makeuse of SFR IR and M ∗ to examine our sample of quasar hosts in the context of the star-forming galaxy main sequence(Figure 3). For the present purposes, we adjust all quantities to a common scale normalized to the Salpeter (1955)stellar initial mass function . For comparison, we show the star-forming galaxies from xCOLD GASS (Saintonge etal. 2017), which comprises low-redshift (0 . < z < .
05) systems with M ∗ > M ⊙ whose SFRs were derivedfrom a combination of ultraviolet and mid-IR photometry (Saintonge et al. 2016). The black curve delineates theparametric relation of the z ≈ σ scatterof ± . /M ∗ ,we use the Kaplan-Meier estimator, as implemented in lifelines , to calculate the median and 16% −
84% intervalof log(sSFR / yr − ) = − . +0 . − . for quasar hosts, much higher than log(sSFR / yr − ) = − . ± .
73 for the star-forming galaxies from xCOLD GASS over the mass range of M ∗ ≥ . M ⊙ where they mostly overlap. Consideringthat the slope, zero point, and possibly even the shape of the star-forming main sequence evolve with cosmic time,perhaps a more appropriate reference should be made to the main sequence of Speagle et al. (2014), which is definedfor star-forming galaxies at z ≈ .
2, similar to the average redshift of the PG quasars. Relative to this reference frame,nearly half (41/86 or 48%) of the quasars still formally lie above the 1 σ scatter of ± − µ m SEDs wereanalyzed by Shangguan et al. (2019) following exactly the same methodology as Shangguan et al. (2018) employedfor the PG quasars. If we simply use as reference the GOALS sample, which has a minimum SFR ≈ M ⊙ yr − , then25% (22/86) of the PG quasars overlap with the LIRGs. Thus, by any reasonable measure, a significant fraction ofthe PG quasars live in starburst galaxies.It is interesting to note that the host galaxies of the quasars are highly heterogeneous. Among those that have fairlysecure classifications, the morphologies of the host galaxies range from ellipticals and mergers to seemingly undisturbeddisk systems, with no obvious correlation with location on the SFR − M ∗ plane. Disk galaxies can be found on or offthe main sequence, including the domain securely occupied by starbursts. As discussed in Section 4, this finding posesa challenge to the notion that major mergers are a necessary ingredient for triggering quasar activity.3.3. The Global Gas Content and Star Formation Efficiency of Quasar Hosts
The preceding subsection reveals that a significant fraction ( ∼ &
25% that are situated securely in theregime of starburst galaxies. Since we have access to total gas mass estimates for the sample (Shangguan et al. 2018),we can study the relation between cold gas content and global star formation activity (Schmidt 1959; Kennicutt1998b). Starburst galaxies are characterized by higher SFEs or, equivalently, shorter gas depletion timescales ( τ dep = SFE − ) than normal star-forming galaxies (e.g., Daddi et al. 2010; Genzel et al. 2010; Shangguan et al. 2019).The majority ( ∼ M gas /M ∗ ≈ .
1) compared to star-forming galaxies. The same holdsfor low-redshift ( z . .
5) type 2 quasars (Shangguan & Ho 2019). As with star-forming galaxies, quasars exhibit astrong inverse correlation between τ dep and sSFR (Figure 4). The relation for quasar hosts may be slightly steeperthan, but it is not inconsistent with, that for massive ( M ∗ & . M ⊙ ) star-forming galaxies defined in terms ofmolecular gas: τ dep (H ) ∝ sSFR − . (Saintonge et al. 2011). The majority of our sample (64/86 or 74%) have gasdepletion timescales distinctively offset from those of normal galaxies and comparable to those of starburst LIRGs( τ dep < . yr). The subset of starburst-like quasar hosts covers the entire range of stellar masses and morphologicaltypes of the parent sample, including, as mentioned, objects with no obvious evidence of significant ongoing or recentmerger activity.It is notable that a handful (10) of the elliptical host galaxies have exceptionally short gas depletion timescales(median τ dep = 10 . yr. All are massive (median M ∗ = 10 . M ⊙ ), presumably advanced or recent merger remnants.While their gas fraction is rather modest ( M gas /M ∗ ≈ . Stellar masses and SFRs for the stellar initial mass function of Salpeter (1955), Kroupa (2001), and Chabrier (2003) scale in the ratio1:1.49:1.58 (Kennicutt et al. 2009; Madau & Dickinson 2014).
Xie et al. -12 -11 -10 -9 log (sSFR / yr -1 )8.08.59.09.510.010.5 l og ( τ d e p / y r) star-forming galaxies starburst LIRGs uncertain no data disks ellipticals mergers τ d e p ( H ) ∝ s SF R - . Figure 4.
The variation of gas depletion timescale ( τ dep ) with specific star formation rate (sSFR) for the PG quasars arecompared with star-forming galaxies from xCOLD GASS (Saintonge et al. 2017; filled grey points for M ∗ < . M ⊙ ; filledblack points for M ∗ ≥ . M ⊙ ) and starburst LIRGs (Shangguan et al. 2019; filled blue circles). Typical uncertaintiesfor the PG quasars are given in the top-right corner. Filled red symbols indicate objects having direct M ∗ measurementsfrom high-resolution optical or near-IR images, while open red symbols represent objects with indirect M ∗ estimated from the M BH − M ∗ relation. Different symbol types represent different host galaxy morphologies, as given in the legend. Objects thathave both gas mass and SFR upper limits are highlighted with a large circle. The solid black line gives the slope of the relation τ dep (H ) ∝ sSFR − . for star-forming galaxies with M ∗ ≥ . M ⊙ (Saintonge et al. 2011). (SFR ≈ M ⊙ yr − ) that would qualify most of them as starbursts in terms of SFE. The starburst activity in thesesystems may be related to central gas concentration from recent merger activity (see, e.g., Husemann et al. 2017 forsimilar arguments for their sample), but high-resolution, high-sensitivity observations are needed to confirm or refutethis hypothesis. DISCUSSIONS4.1.
The Reliability of Infrared-based SFRs in AGN Host Galaxies
The heating source of the IR emission of AGNs continues to be a topic of intense debate. In light of the abundantcold interstellar medium detected in the host galaxies of luminous AGNs (e.g., Scoville et al. 2003; Ho et al. 2008b;Wang et al. 2013; Walter et al. 2014; Xia et al. 2014; Husemann et al. 2017; Kakkad et al. 2017; Shangguan et al. 2018;Shangguan & Ho 2019; Li et al. 2020; Zhuang & Ho 2020), partitioning between the contribution of BH accretionand young stars to the IR luminosity is inherently ambiguous and fraught with uncertainty. Depending on the spatialdistribution of the interstellar medium, dust grains can be exposed to the AGN even on large scales, large enough toreprocess the primary radiation in the far-IR (Sanders et al. 1989). The existence of extended narrow-line regions(e.g., Greene et al. 2011; Chen et al. 2019) attests to the potential reach of the AGN radiation field. Thus, no placein the host galaxy can be assumed to be immune from AGN heating. There are several strategies to confront thisproblem: (1) remove the likely strength of the AGN component in the IR by scaling to an empirical estimate of thestrength of the AGN anchored in another band (e.g., hard X-rays: Dai et al. 2018) or by modeling the contributionof the AGN-heated torus (e.g., Hatziminaoglou et al. 2010; Zhuang et al. 2018); (2) subtracting an empirical AGNtemplate that extends into the far-IR (e.g., Kirkpatrick et al. 2020; Li et al. 2020); or (3) choose a sufficiently longwavelength where the AGN contribution is assumed to be negligible (e.g., Rosario et al. 2012; Santini et al. 2012).However, while plausible, none of these approaches is truly foolproof. Indeed, based on their analysis of the intrinsicfar-IR SED of type 1 AGNs, Symeonidis et al. (2016) argue that no region shortward of ∼ µ m escapes AGNcontamination. While this assertion has been challenged subsequently (Lani et al. 2017; Lyu & Rieke 2017; Xu et al.2020), it remains the case that is extremely difficult to establish quantitatively the true extent to which AGN heatingaffects the far-IR regime. The AGN component for sure matters in the most powerful, distant quasars (e.g., Lutz etal. 2008; Li et al. 2020). Concern also has been raised frequently for low-redshift quasars, indeed for the PG sampleitself (Sanders et al. 1989; Haas et al. 2003; Symeonidis et al. 2016; Zhuang et al. 2018). In their detailed analysisof the 1 − µ m SED of the PG sample, Shangguan et al. (2018) find that the dust temperature (or the minimumradiation field intensity) of the quasar hosts mildly correlates with the quasar luminosity, suggesting that the AGNmay heat the interstellar medium on galactic scales. The same trend holds in a matched sample of z < . ∼ − µ m) Spitzer IRS spectra (Shi et al. 2014). From detailed SEDmodeling (Shangguan et al. 2018), we have available accurate total dust masses and hence cold gas masses, whoserobustness have been verified with ALMA CO observations (Shangguan et al. 2020a, 2020b). Among the 86 sourcesin the original z < . L bol = 10 λL λ (5100 ˚A) ≈ . − . erg s − ],BH mass ( M BH ≈ . − . M ⊙ ), and Eddington ratio ( λ E = L bol /L Edd ≈ . − ≈ − M ⊙ yr − , which agree surprisingly well (0 . +0 . − . dex; Figure 2) with the SFRs inferred fromthe torus-subtracted total IR (8 − µ m) luminosity. The ∼ The Mode of Star Formation in Quasars
The evolution of most star-forming galaxies is regulated by secular processes that situate them on a relatively tightmain sequence with a scatter of ∼ . − . − M ∗ diagram (e.g., Brinchmann et al. 2004; Daddiet al. 2007; Speagle et al. 2014). The exact slope and shape of the main sequence are controversial for galaxieswith M ∗ > M ⊙ (e.g., Renzini & Peng 2015; Schreiber et al. 2015; Whitaker et al. 2015; Tomczak et al. 2016),with differences attributable to variations in galaxy internal structure, systematics of sample selection, choice of SFRindicator, and methods adopted for statistical analysis of the SFR − M ∗ distribution (e.g., Abramson et al. 2014; Manciniet al. 2019; Popesso et al. 2019). Gas-rich major mergers can trigger starbursts (e.g., Sanders & Mirabel 1996), which0 Xie et al. manifest themselves as large excursions above the main sequence, although here, too, there is no universal agreementas to the exact criteria that define a starburst (e.g., Elbaz et al. 2011; Rodighiero et al. 2011; Bergvall et al. 2016).What is the dominant mode of star formation for AGNs in general, and for quasars in particular? The main sequencegives a useful framework for discussing the evolutionary status of AGN host galaxies and their relation to the galaxypopulation at large. The existing literature in this field, however, is complicated enormously by the diverse strategies ofAGN sample selection, the accuracy of the SFR and M ∗ tracers, and the myriad choices of main sequence prescription.While there is almost unanimous agreement that AGNs of low to moderate luminosity ( L bol . erg s − ) at z ≈ − L bol > erg s − . The situation is particularlycontentious at redshifts higher than ∼ .
5, where luminous AGNs have been reported to be above (e.g., Rovilos et al.2012; Florez et al. 2020; Kirkpatrick et al. 2020), on (e.g., Harrison et al. 2012; Xu et al. 2015; Stanley et al. 2017;Schulze et al. 2019), and below (e.g., Scholtz et al. 2018; Stemo et al. 2020) the main sequence. Fortunately, betterconvergence of opinion can be found for luminous ( L bol & erg s − ) AGNs at z . .
5. Most agree that low-redshiftquasars are located largely on and above the main sequence (Husemann et al. 2014; Xu et al. 2015; Zhang et al. 2016;Stanley et al. 2017; Jarvis et al. 2020). Regardless of redshift, it appears that the magnitude of an AGN’s offset fromthe main sequence correlates positively with its luminosity (Bernhard et al. 2019; Grimmett et al. 2020). For theirlarge sample of z ≈ . λ L bol at fixed M ∗ .Our current work takes advantage of the increasingly comprehensive set of physical parameters available for the PGquasars, spanning L bol ≈ . − . erg s − , for which we now have homogeneously measured, newly calibratedSFRs based on the total IR (8 − µ m) luminosity (Section 3.1), stellar masses (Section 2), and cold gas content(Shangguan et al. 2018, 2020a). Originally selected in the optical/ultraviolet, the PG quasars are not biased towardgas-rich systems, and our analysis covers the entire sample of 86 z < . ∼ −
30% of the quasars are within the 1 σ scatter of the main sequence, while 50% −
70% scatter aboveand 10% −
25% fall below it. If, instead, we use GOALS (Armus et al. 2009; Shangguan et al. 2019) as a guide,then roughly 25% (22/86) the quasars intermix with the LIRGs. In short, at least ∼
25% and as many as ∼
70% ofthe PG quasars can be deemed to be starbursts according to their sSFRs. Despite some of the ambiguities inherentin the interpretation of the SFR − M ∗ diagram, fortunately we can circumvent them using the available gas massmeasurements, which, when combined with the SFRs, yield SFEs (Section 3.3). We find that more than half of thequasar hosts will consume their gas reservoir in τ dep ≈ . z < . z ≈ . The Origin of High SFEs in Quasar Host Galaxies
What triggers starbursts? According to conventional wisdom: gas-rich, major mergers (e.g., Kim et al. 2002; Veilleuxet al. 2002; Larson et al. 2016). In this context, the present results for the PG quasars are somewhat puzzling. Asdiscussed in Section 4.2, we have established that the host galaxies of more than half of the quasars have specific SFRsand gas depletion time scales akin to those of starburst systems. Yet, careful scrutiny of Figures 3 and 4 reveals thatnot all the starbursting hosts display signatures of ongoing mergers or interactions. For example, for the 37 sourceswith τ dep . . ∼ ∼ ∼ L bol / erg s -1 )-12-11-10-9 l og ( s SF R / y r - ) τ = 0.02 p = 0.05 -1.5 -1.0 -0.5 0.0 log λ E τ = 0.23 p = 0.03 l og ( τ d e p / y r) τ = -0.26 p = 0.001 uncertain no data disks ellipticals mergers τ = -0.1 p = 0.19 (a) (b) (c) (d) Figure 5.
Dependence of τ dep on (a) AGN bolometric luminosity and (b) Eddington ratio. Dependence of sSFR on (c) AGNbolometric luminosity and (d) Eddington ratio. Filled red symbols indicate objects having direct M ∗ measurements from high-resolution optical or near-IR images, while open red symbols represent objects with indirect M ∗ estimated from the M BH − M ∗ relation. In panels (a) and (b), objects that have both gas mass and SFR upper limits are highlighted with a large circle. TheKendall’s correlation coefficient τ and p -value derived from the cenken function in R are given. In the correlation analysis weexclude objects with SFR and gas mass upper limits in panels (a) and (b), and we exclude objects with indirect M ∗ in panels(c) and (d). external dynamical perturbations are not detected universally among starbursts (see, e.g., Knapen & Cisternas 2015;Cibinel et al. 2019; Shangguan et al. 2019).Lacking clear evidence for an external agent, we turn to the AGN itself as a potential internal trigger of starformation. We anticipate, at the outset, that such an exercise may be frustrated by systematic uncertainties (Harrisonet al. 2017) and artificially strong correlations driven by the mutual dependence of SFR and L bol on redshift, stellarmass, and gas content (Zhuang & Ho 2020; Zhuang et al. 2021). The wide array of often conflicting results in theliterature range from reports of a positive correlation between AGN and star formation activity (e.g., Xu et al. 2015;Husemann et al. 2017; Lanzuisi et al. 2017; Schulze et al. 2019; Shangguan et al. 2020b; Stemo et al. 2020; Zhuang &Ho 2020; Zhuang et al. 2021), to weak or no correlation (e.g., Silvermann et al. 2009; Harrison et al. 2012; Mullaneyet al. 2012; Rosario et al. 2013a, 2013b; Shimizu et al. 2015; Stanley et al. 2015, 2017), or, indeed, to a negativecorrelation (e.g., Page et al. 2012). Figure 5 examines the possible connection between gas depletion time scale andspecific SFR on the one hand, and AGN luminosity and Eddington ratio on the other. No statistically significantcorrelation between τ dep and λ E is discerned, but we detect a moderate inverse correlation between τ dep and L bol ( τ = − . p = 0 . z ≈ . L bol & . erg s − . There seems to be a mildlysignificant positive correlation between sSFR and L bol and λ E ; the Kendall’s test yields a correlation coefficient of τ = 0 .
20 and p = 0 .
05 for Figure 5c and τ = 0 .
23 and p = 0 .
03 for Figure 5d.2
Xie et al.
In summary, we regard the evidence for a causal connection between star formation and AGN activity to be tenuous,at least insofar as the PG quasar sample itself is concerned. What seems abundantly clear, from this and many otherrecent studies, is that in the nearby universe galaxies hosting actively accreting BHs neither lack the fuel nor theability to form stars. CONCLUSIONSThe SFR and SFE are essential parameters to study how supermassive BHs coevolve with their host galaxies. We usethe sample of 86 z < . µ m) emission,through detailed comparison with independent SFRs derive from the [Ne II] 12.81 µ m and [Ne III] 15.56 µ m linesfollowing the recent calibration of Zhuang et al. (2019). We compare the SFRs with the stellar masses and gas massesof the hosts to investigate the nature of their star formation activity in relation to the general galaxy population. Wesummarize our main conclusions as follows: • The torus-subtracted total IR (8–1000 µ m) luminosity yields reliable SFRs for the host galaxies of quasars, atleast those with L bol ≈ . − . erg s − . • PG quasars form stars at a rate of ∼ − M ⊙ yr − . The majority (75% − ∼
25% having specific SFRs overlapping with thoseof low-redshift IR-luminous galaxies. • In conjunction with their short gas depletion time scales, we estimate that ∼
75% of the quasar hosts can beconsidered starburst systems. • The stellar morphologies of a significant fraction of the starburst hosts lack evidence of disturbances from recentmergers or interactions. • The star formation properties of the present sample of quasars are not strongly coupled to the properties of theAGN. ACKNOWLEDGMENTSWe thank our referee for helpful comments. This work was supported by the National Science Foundation of China(11721303, 11991052), the National Key R&D Program of China (2016YFA0400702), and the National Natural ScienceFoundation of China for Youth Scientist Project (11803001). Y.X. thanks Y. Zhao for providing HST morphologicaltypes and Tiago Costa, Yingjie Peng, Hassen Yesuf, Nadia Zakamska, and Chengpeng Zhang for helpful discussions.The Combined Atlas of Sources with Spitzer IRS Spectra (CASSIS) is a product of the IRS instrument team, supportedby NASA and JPL. CASSIS is supported by the “Programme National de Physique Stellaire” (PNPS) of CNRS/INSUco-funded by CEA and CNES and through the “Programme National Physique et Chimie du Milieu Interstellaire”(PCMI) of CNRS/INSU with INC/INP co-funded by CEA and CNES.REFERENCES
Abramson, L. E., Kelson, D. D., Dressler, A., et al. 2014,ApJL, 785, L36Ade, P. A. R., Aghanim, N., Arnaud, M., et al. 2016, A&A,594, A13Armus, L., Mazzarella, J. M., Evans, A. S., et al. 2009,PASP, 121, 559Baron, D., Netzer, H., Prochaska, J. X., et al. 2018,MNRAS, 480, 3993Barthel, P. D. 2006, A&A, 458, 107 Bell, E. F., McIntosh, D. H., Katz, N., & Weinberg, M. D.2003, ApJS, 149, 289Bergvall, N., Marquart, T., Way, M. J., et al. 2016, A&A,587, A72Bernhard, E., Grimmett, L. P., Mullaney, J. R., et al. 2019,MNRAS, 483, L52Boroson, T. A., & Green, R. F. 1992, ApJS, 80, 109Brinchmann, J., Charlot, S., White, S. D. M., et al. 2004,MNRAS, 351, 1151 Brusa, M., Feruglio, C., Cresci, G., et al. 2015, A&A, 578,A11Bruzual, G., & Charlot, S. 2003, MNRAS, 344, 1000Cano-D´ıaz, M., S´anchez, S. F., Zibetti, S., et al. 2016,ApJL, 821, L26Carilli, C. L., Walter, F., Bertoldi, F., et al. 2004, AJ, 128,997Carniani, S. 2017, Frontiers in Astronomy and SpaceSciences, 4, 24Catal´an-Torrecilla, C., Gil de Paz, A., Castillo-Morales, A.,et al. 2017, ApJ, 848, 87Chabrier, G. 2003, PASP, 115, 763Chen, J., Shi, Y., Dempsey, R., et al. 2019, MNRAS, 489,855Cibinel, A., Daddi, E., Sargent, M. T., et al. 2019,MNRAS, 485, 5631Daddi, E., Dickinson, M., Morrison, G., et al. 2007, ApJ,670, 156Daddi, E., Elbaz, D., Walter, F., et al. 2010, ApJL, 714,L118Dai, Y. S., Wilkes, B. J., Bergeron, J., et al. 2018, MNRAS,478, 4238Davidson-Pilon, C., Kalderstam, J., Jacobson, N., et al.2020, ZenodoDraine, B. T., & Li, A. 2007, ApJ, 657, 810Elbaz, D., Daddi, E., Le Borgne, D., et al. 2007, A&A, 468,33Elbaz, D., Dickinson, M., Hwang, H. S., et al. 2011, A&A,533, A119Ellison, S. L., Teimoorinia, H., Rosario, D. J., et al. 2016,MNRAS, 458, L34Evans, A. S., Frayer, D. T., Surace, J. A., et al. 2001, AJ,121, 1893Evans, A. S., Solomon, P. M., Tacconi, L. J., et al. 2006,AJ, 132, 2398Fabian, A. C. 2012, ARA&A, 50, 455Florez, J., Jogee, S., Sherman, S., et al. 2020, MNRAS, 497,3273Genzel, R., Tacconi, L. J., Gracia-Carpio, J., et al. 2010,MNRAS, 407, 2091Greene, J. E., Strader, J., & Ho, L. C. 2020, ARA&A, 58,257Greene, J. E., Zakamska, N. L., Ho, L. C., et al. 2011, ApJ,732, 9Grimmett, L. P., Mullaney, J. R., Bernhard, E. P., et al.2020, MNRAS, 495, 1392Groves, B., Dopita, M., & Sutherland, R. 2006, A&A, 458,405Haas, M., Klaas, U., M¨uller, S. A. H., et al. 2003, A&A,402, 87 Harrison, C. M. 2017, Nature Astronomy, 1, 0165Harrison, C. M., Alexander, D. M., Mullaney, J. R., et al.2012, ApJL, 760, L15Hatziminaoglou, E., Omont, A., Stevens, J. A., et al. 2010,A&A, 518, L33Heckman, T. M., & Best, P. N. 2014, ARA&A, 52, 589Ho, L. C. 2005, ApJ, 629, 680Ho, L. C. 2008, ARA&A, 46, 475Ho, L. C. 2009, ApJ, 699, 626Ho, L. C., Darling, J., & Greene, J. E. 2008a, ApJS, 177,103Ho, L. C., Darling, J., & Greene, J. E. 2008b, ApJ, 681, 128Ho, L. C., Filippenko, A. V., & Sargent, W. L. W. 2003,ApJ, 583, 159Ho, L. C., & Keto, E. 2007, ApJ, 658, 314Ho, L. C., & Kim, M. 2009, ApJS, 184, 398Ho, L. C., & Kim, M. 2015, ApJ, 809, 123Houck, J. R., Roellig, T. L., Van Cleve, J., et al. 2004,Proc. SPIE, 5487, 62Husemann, B., Davis, T. A., Jahnke, K., et al. 2017,MNRAS, 470, 1570Husemann, B., Jahnke, K., S´anchez, S. F., et al. 2014,MNRAS, 443, 755Jackson, T. M., Rosario, D. J., Alexander, D. M., et al.2020, MNRAS, 498, 2323Jarvis, M. E., Harrison, C. M., Mainieri, V., et al. 2020,MNRAS, 498, 1560Kakkad, D., Mainieri, V., Brusa, M., et al. 2017, MNRAS,468, 4205Kennicutt Jr., R. C. 1998a, ARA&A, 36, 189Kennicutt, R. C. 1998b, ApJ, 498, 541Kennicutt Jr., R. C., Hao, C.-N., Calzetti, D., et al. 2009,ApJ, 703, 1672Kewley, L. J., & Ellison, S. L. 2008, ApJ, 681, 1183Kim, D.-C., Veilleux, S., & Sanders, D. B. 2002, ApJS, 143,277Kim, M., Ho, L. C., & Im, M. 2006, ApJ, 642, 702Kim, M., Ho, L. C., Peng, C. Y., Barth, A. J., & Im, M.2017, ApJS, 232, 21Kirkpatrick, A., Urry, C. M., Brewster, J., et al. 2020, ApJ,900, 5Knapen, J. H., & Cisternas, M. 2015, ApJL, 807, L16Kormendy, J., & Ho, L. C. 2013, ARA&A, 51, 511Kroupa, P. 2001, MNRAS, 322, 231Lani, C., Netzer, H., & Lutz, D. 2017, MNRAS, 471, 59Lanzuisi, G., Delvecchio, I., Berta, S., et al. 2017, A&A,602, A123Larson, K. L., Sanders, D. B., Barnes, J. E., et al. 2016,ApJ, 825, 128 Xie et al.
Lebouteiller, V., Barry, D. J., Goes, C., et al. 2015, ApJS,218, 21Lebouteiller, V., Barry, D. J., Spoon, H. W. W., et al. 2011,ApJS, 196, 8Leslie, S. K., Kewley, L. J., Sanders, D. B., et al. 2016,MNRAS, 455, L82Li, Q., Wang, R., Fan, X., et al. 2020, ApJ, 900, 12Lutz, D., Sturm, E., Tacconi, L. J., et al. 2008, ApJ, 684,853Lyu, J., & Rieke, G. H. 2017, ApJ, 841, 76Madau, P., & Dickinson, M. 2014, ARA&A, 52, 415Mahoro, A., Povi´c, M., Nkundabakura, P., et al. 2019,MNRAS, 485, 452Mancini, C., Daddi, E., Juneau, S., et al. 2019, MNRAS,489, 1265Mullaney, J. R., Alexander, D. M., Aird, J., et al. 2015,MNRAS, 453, L83Mullaney, J. R., Alexander, D. M., Goulding, A. D., et al.2011, MNRAS, 414, 1082Mullaney, J. R., Pannella, M., Daddi, E., et al. 2012,MNRAS, 419, 95Nenkova, M., Sirocky, M. M., Ivezi´c, ˇZ., & Elitzur, M.2008, ApJ, 685, 147Netzer, H., Lutz, D., Schweitzer, M., et al. 2007, ApJ, 666,806Neugebauer, G., Green, R. F., Matthews, K., et al. 1987,ApJS, 63, 615Noeske, K. G., Weiner, B. J., Faber, S. M., et al. 2007,ApJL, 660, L43Page, M. J., Symeonidis, M., Vieira, J. D., et al. 2012,Nature, 485, 213Peng, Y.-J., Lilly, S. J., Kovaˇc, K., et al. 2010, ApJ, 721,193Petric, A. O., Ho, L. C., Flagey, N. J. M., & Scoville, N. Z.2015, ApJS, 219, 22Pilbratt, G. L., Riedinger, J. R., Passvogel, T., et al. 2010,˚a, 518, L1Popesso, P., Concas, A., Morselli, L., et al. 2019, MNRAS,483, 3213Renzini, A., & Peng, Y.-J. 2015, ApJL, 801, L29Richstone, D., Ajhar, E. A., Bender, R., et al. 1998,Nature, 395, A14Rodighiero, G., Daddi, E., Baronchelli, I., et al. 2011,ApJL, 739, L40Rosario, D. J., Santini, P., Lutz, D., et al. 2012, A&A, 545,A45Rosario, D. J., Santini, P., Lutz, D., et al. 2013a, ApJ, 771,63Rosario, D. J., Trakhtenbrot, B., Lutz, D., et al. 2013b,A&A, 560, A72 Rovilos, E., Comastri, A., Gilli, R., et al. 2012, A&A, 546,A58Saintonge, A., Catinella, B., Cortese, L., et al. 2016,MNRAS, 462, 1749Saintonge, A., Catinella, B., Tacconi, L. J., et al. 2017,ApJS, 233, 22Saintonge, A., Kauffmann, G., Wang, J., et al. 2011,MNRAS, 415, 61Salpeter, E. E. 1955, ApJ, 121, 161S´anchez, S. F., Avila-Reese, V., Hernandez-Toledo, H., etal. 2018, RMxAA, 54, 217Sanders, D. B. & Mirabel, I. F. 1996, ARA&A, 34, 749Sanders, D. B., Phinney, E. S., Neugebauer, G., et al. 1989,ApJ, 347, 29Santini, P., Rosario, D. J., Shao, L., et al. 2012, A&A, 540,A109Schmidt, M. 1959, ApJ, 129, 243Schmidt, M., & Green, R. F. 1983, ApJ, 269, 352Scholtz, J., Alexander, D. M., Harrison, C. M., et al. 2018,MNRAS, 475, 1288Schreiber, C., Pannella, M., Elbaz, D., et al. 2015, A&A,575, A74Schweitzer, M., Lutz, D., Sturm, E., et al. 2006, ApJ, 649,79Scoville, N. Z., Frayer, D. T., Schinnerer, E., et al. 2003,ApJL, 585, L105Shang, Z., Brotherton, M. S., Wills, B. J., et al. 2011,ApJS, 196, 2Shangguan, J., & Ho, L. C. 2019, ApJ, 873, 90Shangguan, J., Ho, L. C., Bauer, F. E., et al. 2020a, ApJS,247, 15Shangguan, J., Ho, L. C., Bauer, F. E., et al. 2020b, ApJ,899, 112Shangguan, J., Ho, L. C., Li, R., et al. 2019, ApJ, 870, 104Shangguan, J., Ho, L. C., & Xie, Y. 2018, ApJ, 854, 158Shao, L., Lutz, D., Nordon, R., et al. 2010, A&A, 518, L26Shi, Y., Rieke, G. H., Ogle, P. M., Su, K. Y. L., & Balog, Z.2014, ApJS, 214, 23Shimizu, T. T., Mushotzky, R. F., Mel´endez, M., et al.2015, MNRAS, 452, 1841Shimizu, T. T., Mushotzky, R. F., Mel´endez, M., et al.2017, MNRAS, 466, 3161Schulze, A., Silverman, J. D., Daddi, E., et al. 2019,MNRAS, 488, 1180Siebenmorgen, R., Heymann, F., & Efstathiou, A. 2015,A&A, 583, A120Silverman, J. D., Kovaˇc, K., Knobel, C., et al. 2009, ApJ,695, 171Speagle, J. S., Steinhardt, C. L., Capak, P. L., et al. 2014,ApJS, 214, 15 Stanley, F., Alexander, D. M., Harrison, C. M., et al. 2017,MNRAS, 472, 2221Stanley, F., Harrison, C. M., Alexander, D. M., et al. 2015,MNRAS, 453, 591Stemo, A., Comerford, J. M., Barrows, R. S., et al. 2020,ApJ, 888, 78Suh, H., Civano, F., Hasinger, G., et al. 2017, ApJ, 841, 102Symeonidis, M., Giblin, B. M., Page, M. J., et al. 2016,MNRAS, 459, 257Tomczak, A. R., Quadri, R. F., Tran, K.-V. H., et al. 2016,ApJ, 817, 118Veilleux, S., Kim, D.-C., & Sanders, D. B. 2002, ApJS, 143,315Vestergaard, M., & Peterson, B. M. 2006, ApJ, 641, 689Voit, G. M. 1992, MNRAS, 258, 841Walter, F., Bertoldi, F., Carilli, C., et al. 2003, Nature, 424,406Walter, F., Decarli, R., Sargent, M., et al. 2014, ApJ, 782,79Wang, R., Wagg, J., Carilli, C. L., et al. 2013, ApJ, 773, 44Wang, R., Wu, X.-B., Neri, R., et al. 2016, ApJ, 830, 53Werner, M. W., Roellig, T. L., Low, F. J., et al. 2004,ApJS, 154, 1 Whitaker, K. E., Franx, M., Bezanson, R., et al. 2015,ApJL, 811, L12Xia, X. Y., Gao, Y., Hao, C.-N., et al. 2012, ApJ, 750, 92Xie, Y., & Ho, L. C. 2019, ApJ, 884, 136Xu, J., Sun, M., & Xue, Y. 2020, ApJ, 894, 21Xu, L., Rieke, G. H., Egami, E., et al. 2015, ApJS, 219, 18Xue, Y. Q., Brandt, W. N., Luo, B., et al. 2010, ApJ, 720,368Yang, G., Brandt, W. N., Vito, F., et al. 2018, MNRAS,475, 1887Yesuf, H. M., & Ho, L. C. 2019, ApJ, 884, 177Yesuf, H. M., & Ho, L. C. 2020a, ApJ, 900, 107Yesuf, H. M., & Ho, L. C. 2020b, ApJ, 901, 42Zhang, Z., Shi, Y., Rieke, G. H., et al. 2016, ApJL, 819, L27Zhao, D., Ho, L. C., & Zhuang, M.-Y. 2021a, ApJ, inpreparationZhao, Y., Ho, L. C., Shangguan, J., et al. 2021b, ApJ,submittedZhuang, M.-Y., & Ho, L. C. 2019, ApJ, 882, 89Zhuang, M.-Y., & Ho, L. C. 2020, ApJ, 896, 108Zhuang, M.-Y., Ho, L. C., & Shangguan, J. 2018, ApJ, 862,118Zhuang, M.-Y., Ho, L. C., & Shangguan, J. 2019, ApJ, 873,103Zhuang, M.-Y., Ho, L. C., & Shangguan, J. 2021, ApJ, 906,38 Xie et al.
Table 1.
Physical Properties of PG Quasars
Object z D L log λL λ (5100 ˚A) log M BH log M ∗ M gas Morphology Reference(Mpc) (erg s − ) ( M ⊙ ) ( M ⊙ ) ( M ⊙ )(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)PG 0003+158 0.450 2572 45.99 9.45 11.83 ∗ . +0 . − . < · · · · · · PG 0003+199 0.025 113 44.17 7.52 10.38 ∗ . +0 . − . . +0 . − . D 3PG 0007+106 0.089 420 44.79 8.87 11.03 8 . +0 . − . . +0 . − . M 1PG 0026+129 0.142 693 45.07 8.12 11.07 8 . +0 . − . . +0 . − . E 3PG 0043+039 0.384 2133 45.51 9.28 11.13 8 . +0 . − . < ∗ . +0 . − . . +0 . − . · · · · · · PG 0050+124 0.061 282 44.76 7.57 11.31 8 . +0 . − . . +0 . − . D 5PG 0052+251 0.155 763 45.00 8.99 11.24 8 . +0 . − . . +0 . − . D 5PG 0157+001 0.164 811 44.95 8.31 11.72 8 . +0 . − . . +0 . − . M 4PG 0804+761 0.100 475 45.03 8.55 10.83 8 . +0 . − . . +0 . − . D 5PG 0838+770 0.131 635 44.70 8.29 11.33 8 . +0 . − . . +0 . − . D 6PG 0844+349 0.064 297 44.46 8.03 10.88 8 . +0 . − . . +0 . − . M 3PG 0921+525 0.035 159 43.60 7.45 10.32 ∗ . +0 . − . . +0 . − . D 3PG 0923+201 0.190 955 45.01 9.33 11.28 8 . +0 . − . . +0 . − . E,c 3PG 0923+129 0.029 131 43.83 7.52 10.38 ∗ . +0 . − . . +0 . − . D 5PG 0934+013 0.050 229 43.85 7.15 10.10 ∗ . +0 . − . . +0 . − . D 5PG 0947+396 0.206 1045 44.78 8.81 10.92 8 . +0 . − . . +0 . − . U 6PG 0953+414 0.239 1235 45.35 8.74 11.35 8 . +0 . − . . +0 . − . U 5
Note — Col. (1): object name. Col. (2): redshift. Col. (3): luminosity distance. Col. (4): monochromatic luminosity of the AGN at5100 ˚A, from Shangguan et al. (2018). Col. (5): mass of the BH, from Shangguan et al. (2018). Col. (6): stellar mass of the quasarhost galaxy from Zhang et al. (2016) or derived from the M ∗ − M BH relation of Greene et al. (2020; marked with an asterisk). Col.(7): metallicity of the quasar host galaxy estimated from the stellar mass-metallicity relation; see Section 2 for details. Col. (8):total gas mass, from Shangguan et al. (2018). Col. (9): morphological type of the host galaxy: “E” = elliptical, “D” = disk, “M”= merger, “U” = uncertain, “c” = companion, and “t” = tidal disturbance signatures. Col. (10): reference for morphological types:(1) Bentz & Manne-Nicholas 2018; (2) Crenshaw et al. 2003; (3) Kim et al. 2017; (4) Surace et al. 1998; (5) Y. Zhao et al. 2021;and (6) Zhang et al. 2016. (Table 1 is published in its entirety in machine-readable format. A portion is shown here for guidanceregarding its form and content.) Table 2.
SFRs of Quasar Host Galaxies
Object log L [Ne II] log L [Ne III] log L [Ne V] f + f +2 log SFR Ne log L IR log SFR IR (erg s − ) (erg s − ) (erg s − ) ( M ⊙ yr − ) (erg s − ) ( M ⊙ yr − )(1) (2) (3) (4) (5) (6) (7) (8) (9)PG 0003+158 · · · · · · · · · · · · · · · · · · < . < . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . − . +0 . − . PG 0007+106 41 . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . PG 0026+129 41 . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . PG 0043+039 · · · · · · · · · · · · · · · · · · < . < . · · · · · · · · · · · · · · · · · · . +0 . − . − . +0 . − . PG 0050+124 < .
91 41 . +0 . − . < . · · · · · · [0.19, 0.92] 44 . +0 . − . . +0 . − . PG 0052+251 · · · · · · · · · · · · · · · · · · . +0 . − . . +0 . − . PG 0157+001 42 . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . PG 0804+761 < .
23 41 . +0 . − . < . · · · · · · [0.76, 1.25] 43 . +0 . − . . +0 . − . PG 0838+770 41 . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . PG 0844+349 40 . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . PG 0921+525 40 . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . − . +0 . − . PG 0923+201 < .
23 41 . +0 . − . < . · · · · · · [0.62, 1.25] 43 . +0 . − . . +0 . − . PG 0923+129 40 . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . PG 0934+013 · · · · · · · · · · · · · · · · · · . +0 . − . . +0 . − . PG 0947+396 · · · · · · · · · · · · · · · · · · . +0 . − . . +0 . − . PG 0953+414 41 . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . Note — Col. (1): object name. Col. (2): luminosity of [Ne II] 12.81 µ m. Col. (3): luminosity of [Ne III] 15.56 µ m. Col. (4):luminosity of [Ne V] 14.32 µ m. Col. (5): fractional abundance of neon in the Ne + state. Col. (6): fractional abundanceof neon in the Ne +2 state. Col. (7): SFR calculated from neon luminosity. For SFR involving neon line upper limits, thevalues included in brackets present lower and upper limits of the true SFR. Col. (8): torus-subtracted IR luminosity in the8 − µ m band. Col. (9): SFR calculated from the torus-subtracted IR luminosity in the 8 − µµ