Chemical abundances along the quasar main sequence

A&A, 689, A321 (2024) h ps://doi.o g/10.1051/0004-6361/202450458 c The Au ho s 2024 As onomy & As ophysics Chemical abundances along he quasa main sequence A. Flo is1,2,?, P. Ma ziani1,?, S. Panda3,?,?? , M. Sniegowska4, M. D’Ono io2, A. Decon o-Machado5, A. del Olmo5, and B. Cze ny6 1Na ional Ins i u e o As ophysics (INAF), As onomical Obse a o y o Pado a, 35122 Pado a, I aly 2Dipa imen o di Fisica e As onomia, Uni e si à di Pado a, Vicolo dell’ Osse a o io 3, 35122 Pado a, I aly 3Labo a ó io Nacional de As o ísica, MCTI, R. dos Es ados Unidos 154, Nações 37504-364, I ajubá, B azil 4School o Physics and As onomy, Tel A i Uni e si y, Tel A i , 69978, Is ael 5Ins i u o de As o ísica de Andalucía (IAA-CSIC), Glo ie a de As onomía s/n, 38038 G anada, Spain 6Cen e o Theo e ical Physics, Polish Academy o Sciences, Al. Lo ników 32/46, 02-668 Wa saw, Poland Recei ed 20 Ap il 2024 /Accep ed 27 May 2024 ABSTRACT Con ex . The main sequence o quasa s has eme ged as a powe ul ool o o ganizing he obse a ional and physical cha ac e is ics o ype-1 ac i e galac ic nuclei (AGNs). Aims. In his s udy, we p esen a comp ehensi e analysis o he me allici y o he gas in he b oad-line egion, inco po a ing bo h new da a and p e iously published indings, o assess he p esence o any end along he main sequence. Me hods. We pe o med a mul icomponen analysis on he s onges ul a iole (UV) and op ical emission lines o a sample o 13 adio quie quasa s in he 0.009 ≤z≤0.472 edshi ange, selec ed based on he a ailabili y o mul iwa eleng h da a. We employed UV and op ical da a ob ained om he Hubble Space Telescope (mainly om he Cosmic O igins Spec og aph and Fain Objec Spec og aph) and se e al g ound-based obse a o ies, espec i ely. We hen measu ed en diagnos ic a ios and compa ed hem wi h he p edic ion o CLOUDY pho oioniza ion simula ions, iden i ying he closes pho oioniza ion solu ion o he da a. Resul s. Ou in es iga ion e eals a consis en pa e n along he main sequence. We obse e a sys ema ic p og ession in me allici y, anging om subsola alues o me allici y le els se e al imes highe han sola alues. Conclusions. These indings unde sco e he undamen al ole o me allici y in co ela ing wi h he main sequence o quasa s. Ex eme me allici y alues, a leas se e al dozen imes he sola me allici y, a e con i med in low-zAGNs adia ing a a high Edding on a io, al hough he o igin o he ex eme en ichmen emains open o deba e. Key wo ds. me hods: obse a ional – galaxies: ac i e – quasa s: emission lines – quasa s: gene al – quasa s: supe massi e black holes – galaxies: Sey e 1. In oduc ion In ecen yea s, he main sequence (MS) o quasa s (Sulen ic e al. 2000b;Ma ziani e al. 2001,2010;Shen & Ho 2014) has p o ed o be a c ucial ool o un a eling he spec- al p ope ies o ype 1 ac i e galac ic nuclei (AGNs, e.g., Panda e al. 2019b;Panda & ´ Sniegowska 2024;Fu e al. 2022; Wang e al. 2022;Nagoshi & Iwamu o 2022, among he mos ecen wo ks), e en a he highes edshi s (e.g., Yang e al. 2023;Loiacono e al. 2024). The MS is he pa ame e space de ined by RFeII ( he a io be ween he in ensi y o he Feii emis- sion a λ4570 o Hβ) and he Full Wid h a Hal Maximum o Hβ: FWHM(Hβ) (Gaskell 1985;Bo oson & G een 1992). The MS allows o he de ini ion o wo main popula ions o quasa s occupying di e en egions o his pa ame e space (Sulen ic e al. 2000a,2011;Ma ziani e al. 2018b;Panda e al. 2024): – Popula ion A (Pop. A), ound below 4000kms−1o FWHM(Hβ), and – Popula ion B (Pop. B), ound abo e 4000kms−1o FWHM(Hβ). ?Co esponding au ho s; [email p o ec ed], [email p o ec ed],[email p o ec ed], ?? CNPq Fellow. Asubpopula iono Pop. A AGNs,cha ac e izedbyRFeII &1, is he ex eme Pop. A (xA), which is associa ed wi h ex eme- acc e ing AGNs (Ma ziani & Sulen ic 2014;Du e al. 2016; Panda & Ma ziani 2023). These popula ions exhibi di e - en spec al ea u es associa ed wi h an a ay o physical condi ions in he b oad-line egion (BLR) o hese objec s (F aix-Bu ne e al. 2017;Ma ziani e al. 2018b;Panda e al. 2019b). Se e al pa ame e s a y along he MS (see e.g., Sulen ic e al. 2011;F aix-Bu ne e al. 2017;Panda 2021a, o summa ies). T ends in ol e pa ame e s associa ed wi h line o ma ion, such as he ioniza ion pa ame e , U, which has been shown o g adually inc ease mo ing om xA o Pop. B objec s (Sulen ic e al. 2000b;Ma ziani e al. 2010). The gas densi y, nH, seems o ollow an in e se end o U, wi h xA objec s associa ed wi h much highe densi ies han Pop. B objec s (Ma ziani e al. 2001;Ma suoka e al. 2008;Neg e e e al. 2012; Ma ínez-Aldama e al. 2015;Panda e al. 2018). No ably, RFeII is belie ed o be p ima ily d i en by he Edding on a io, L/LEdd (Bo oson & G een 1992;Ma ziani e al. 2001;Sun & Shen 2015;D’Ono io e al. 2021). The posi ion o objec s along he MS is hus connec ed o he adia ion emi ed by he AGN con inuum sou ce, which is i sel connec ed o he a e o acc e ion o ma e on o he supe massi e black hole (SMBH). Open Access a icle, published by EDP Sciences, unde he e ms o he C ea i e Commons A ibu ion License (h ps://c ea i ecommons.o g/licenses/by/4.0), which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly ci ed. This a icle is published in open access unde he Subsc ibe o Open model.Subsc ibe o A&A o suppo open access publica ion. A321, page 1 o 31 Flo is, A., e al.: A&A, 689, A321 (2024) The FWHM(Hβ) is s ongly a ec ed by he iewing angle, e en i he objec s conside ed in he MS a e all ype 1 AGNs. The e is inc easing e idence ha he low-ioniza ion lines (LILs) a e emi ed in a la ened con igu a ion co esponding o he acc e ion disk (e.g., Deca li e al. 2011;Mejía-Res epo e al. 2018;Neg e e e al. 2018;Panda 2021b), as was sugges ed a long ime ago (Collin-Sou in e al. 1988). The e ec o he iewing angle on he line wid h ( oughly p opo ional o sinθ, whe e θis he angle be ween he line o sigh and he acc e ion disk axis) is compounded by he e ec o he SMBH mass ha is weak bu ha becomes signi ican i AGNs in luminosi y anges o se e al o de s o magni udes a e conside ed (Ma ziani e al. 2018a). I no co ec ion o he iewing angle is applied, he Pop. B objec s, by de ini ion wi h la ge FWHM, a e cha ac e ized by sys ema ically mo e massi e SMBHs han hei Pop. A coun e - pa s (Ma ziani e al. 2003b). The ole o he chemical con en o he BLR is, howe e , no ye comple ely unde s ood in he p ocess o line o ma- ion. P e ious s udies like hose o ´ Sniegowska e al. (2021) and Ga nica e al. (2022) ha e ocused on he me al con en o xA quasa s and de e mined high me allici y, Z, in he BLR o hese objec s, eaching ex eme alues. High Zis a po en ial explana ion o he obse ed ends o densi y and ioniza ion pa ame e , as esonan sca e ing becomes mo e e icien when me als a e mo e abundan (e.g., Mu ay e al. 1995). Addi ion- ally, a ious s udies ha e indica ed he sys ema ic en ichmen o he BLR o AGNs compa ed o hei hos galaxies, nea ly always eaching supe -sola alues (e.g., Hamann & Fe land 1999;Nagao e al. 2006;Xu e al. 2018;Maiolino & Mannucci 2019, and e e ences he ein). High Z alues ha e been con- i med h ough independen echniques, such as i on emis- sion and abso p ion ea u es de ec ed in he nuclea egion (Jiang e al. 2018;Maiolino & Mannucci 2019). Some p oposed explana ions include selec i e en ichmen o he ci cumnu- clea egion o he AGN (´ Sniegowska e al. 2021). Me allici ies a ound he sola me allici y a e possible o xA sou ces only unde he mos ex eme condi ions (Temple e al. 2021). In addi- ion, he e is s ill he possibili y ha he adop ed empi ical and pho oioniza ion models a e inadequa e. Recen wo ks sugges ha en ichmen may no occu o he BLR gas o all AGNs: owa d he opposi e ex eme o he sequence, he es ima ion o me allici y indica es a sola me allic- i y ( o he adio-quie Pop. B) o a subsola one ( o he adio- loud Pop. B; Punsly e al. 2018;Ma ziani e al. 2022). P e ious s udies in ol ed a ew objec s and samples in di e en edshi and luminosi y domains (S21,G22,M23), ocusing on speci ic spec al bins o popula ions o he MS. I is necessa y o unde - s and wha happens be ween he ex emes o he MS, and a emp o unco e any end in me allici y along he MS ha could un eil possible mechanisms behind he e olu ion o quasa s and hei hos galaxies: e idence o blueshi s (Zamano e al. 2002;Komossa e al. 2008;Bo oson 2005;Aoki e al. 2005; Zhang e al. 2011;Neg e e e al. 2018) in he lines o he spec- a o mos AGNs hin s a he possibili y ha ou lows could chemically en ich he gas ound in he in e s ella medium (ISM) o e en he in e galac ic medium (IGM) (Choi e al. 2024; Sabhlok e al. 2024). Indeed, Ha ison e al. (2014) poin ed ou he a - eaching impac o such ou lows beyond he BLR o hese objec s, sugges ing an in e play be ween he en iched ma e ial p oduced in he egions close o he ac i e nucleus and expelled o e ime in he di ec ion o he hos galaxy. In his wo k, we p esen he sys ema ic s udy o he me allic- i y o a sample o objec s co e ing mos spec al ypes along he MS a low edshi (Sec ion 2). Sec ion 3desc ibes he me hod- ology and echniques used. Resul s o he objec s o he sample a e p o ided in Sec ion 4. In Sec ion 5, we add ess he po en- ial issues o ou me hodology. The discussion o he esul s in Sec ion 6 ocuses on he co ela ion o he me allici y o objec s wi h acc e ion pa ame e s and he possible en ichmen mecha- nisms ha may explain ou esul s. The conclusions a e summa- ized in Sec ion 7. 2. Sample The sample o objec s used in his wo k was selec ed based on a ew c i e ia: 1. The a ailabili y o g ound obse a ions in he op ical wa e- leng h ange o objec s below a edshi zo 0.8, whe e Hβ can be obse ed. 2. The a ailabili y o a chi al ul a iole (UV) da a om HST co e ing he λ1300–2100Å es - ame ange o he objec . 3. The equi emen s o adio quie (RQ) objec s, o s udy he p ope ies o he BLR wi hou he in luence o a collima ed je (Bake e al. 1994;Punsly e al. 2018;Gau e al. 2019). Table 1 epo s he main obse a ional pa ame e s, in he ol- lowing o de : common name, IAU coo dina e name, adop ed edshi , HβFWHM, he FeII- o-Hβin ensi y a io, RFeII (wi h Feii emission in eg a ed o e he λ4434−4684Å ange, and Hβmeasu ed on he ull b oad p o ile), he spec al ype (ST) based on HβFWHM and RFeII, and Kelle man’s a io (Kelle mann e al. 1989), which is necessa y o dis inguish RQ objec s om RL ones, de ined as RK= Radio B ,(1) using da a om he NED da abase. He e, Radio is he speci ic lux a a wa eleng h o λ=6 cm (5GHz) and Bis he speci ic lux a λ=4400Å (680 THz) in he Bband. When he NED da a was no a ailable, he speci ic lux used in he calcula ion was ob ained om o he wo ks. The las wo columns summa ize he UV and op ical da a sou ces o each objec . Mos objec s ha sa is ied hese selec ion c i e ia we e included in he sample o objec s in Ma ziani e al. (2003a). The spec a o PHL 1092, which is he mos ex eme objec in he sample, we e ob ained om Ma inello e al. (2020). Figu e 1desc ibes he dis ibu ion o he objec s o he sam- ple in bolome ic luminosi y (and compu ed in Sec ion 6.2) and z, con on ing hem wi h a e age alues om he samples used in he wo ks o ´ Sniegowska e al. (2021), Ga nica e al. (2022), and Ma ziani e al. (2023) (he ea e e e ed o as S21,G22, and M23, espec i ely). 3. Me hodology 3.1. Subdi ision o spec al ypes An ad an age o he MS is he possibili y o de ining spec- al ypes along he op ical plane o he sequence wi h sim- ple selec ion c i e ia (Sulen ic e al. 2002;Shen & Ho 2014; Panda & ´ Sniegowska 2024). The expec a ion is ha he e ec o sys ema ic ends along he sequence is minimized in each spec al ype, so ha an a e age spec um o he sou ces in each bin is ep esen a i e o all o mos o he indi idual sou ces, no only phenomenologically, bu also in e ms o physical p ope - ies. This has been shown o be he case (Sulen ic e al. 2002; Bache e al. 2004;Ma ziani e al. 2010), e en wi h some lim- i a ions (Ma ziani e al. 2001,2018a,b;Panda e al. 2019b), as A321, page 2 o 31 Flo is, A., e al.: A&A, 689, A321 (2024) Table 1. Basic sample p ope ies. Common name IAU Code zFWHM(Hβ)RFeII Spec al ype RKUV Op ical (1) (2) (3) (4) (5) (6) (7) (8) (9) M k 335 J00063+2012 0.0256 2209 ±220 0.21 ±0.04 A1 0.42 COS M03 I Zw 1 J00535+1241 0.061 1506 ±151 1.83 ±0.28 A4 0.40 FOS, COS M03 Fai all 9 J01237−5848 0.046 5958 ±596 0.43 ±0.08 B1 0.92 FOS, COS M03 PHL 1092 J01399+0619 0.3965 2494 ±249 1.76 ±0.39 A4 0.42 STIS, M20 M20 A k 120 J05161−0008 0.033 5863 ±586 0.49 ±0.04 B1 0.30 FOS M03 M k 110 J09252+5217 0.036 2048 ±205 0.19 ±0.04 A1 2.18 COS M03 Ton 28 J10040+2855 0.329 2398 ±240 1.04 ±0.18 A3 <0.40 FOS M03 NGC 3783 J11390−3744 0.009 3514 ±351 0.20 ±0.04 B1 0.30 FOS, COS M03 LB 2522 J13012+5902 0.472 3999 ±400 1.30 ±0.32 A3 <0.25 FOS M03 LEDA 51016 J14170+4456 0.114 2720 ±272 1.01 ±0.16 A3 0.17 FOS, COS M03 M k 478 J14421+3526 0.077 1322 ±132 1.04 ±0.10 A3 0.20 FOS M03 M k 509 J20441−1043 0.035 3309 ±331 0.02 ±0.02 A1 0.35 COS M03 A k 564 J22426+2943 0.025 1011 ±101 0.67 ±0.22 A2 0.88 FOS, COS M03 No es. (1) Objec ’s common name. (2) IAU coo dina e name. (3) Redshi . (4) FWHM o Hβ, in uni s o kms−1. (5) RFeII =Feiiλ4570/Hβmeasu ed in his wo k. (6) Spec al ype. (7) Kelle man a io, om NED, o om he ollowing sou ces: Wang e al. (2023a) o M k 110, Jä elä e al. (2022) o A k 564; and Cape i e al. (2021), Baldi e al. (2022), and Kelle mann e al. (1994) o Ton 28, LB 2522, and LEDA 51016, espec i ely. (8) HST came a used o collec ing UV spec a o he objec , using he abb e ia ion FOS o he Fain Objec Spec og aph, and COS o he Cosmic O igins Spec og aph. (9) Op ical da a sou ce, using he abb e ia ions M03 and M20 o he pape s Ma ziani e al. (2003a) and Ma inello e al. (2020), espec i ely. Fig. 1. S acked his og ams depic ing he dis ibu ion o sample objec s in log(Lbol) and z, ca ego ized by he espec i e popula ion. Assigned colo s co espond o he legend, while e ical lines deno e a e age al- ues om he objec s used in he e e enced wo ks S21,G22, and M23. he 2D op ical MS is no a simple 1D end (Wildy e al. 2019; Ma ínez-Aldama e al. 2021), bu a he he p ojec ion o a end in a 4D space ( he so-called 4DE1 space, Sulen ic e al. 2000b). In his pape , we ollow he naming con en ion and he sub- di ision o Sulen ic e al. (2002) in egula s eps o δRFeII = 0.5 (inc easing nume als wi h inc easing RFeII), and δFWHM Hβ=4000km s−1(A, B, and a e wa d B+, B++, e c.). 3.2. Emission line in e p e a ion and empi ical modeling In his wo k, we adop an in e p e a ion o he emission line p o- iles o Type-I AGNs based on h ee main componen s: – B oad componen (BC): A i ialized componen ha cons i- u es he bulk o he line p o iles o he main emission lines. I ollows a Lo en zian p o ile in Pop. A objec s and a Gaus- sian p o ile in Pop. B objec s (Ma ziani e al. 2010). I is a dis inc i e componen o Type-I AGN line p o iles, iden i y- ing he BLR o hese objec s. – Blueshi ed componen (BLUE): A blueshi ed ou low com- ponen p esen in he lines o almos all AGNs, a ec ing b oad and na ow lines (Bache e al. 2004;Zamano e al. 2002). I is ypically modeled by a skewed Gaussian ha exhibi s an asymme y owa d he blue (Zhang e al. 2011). This componen is p edominan in xA objec s ha a e cha - ac e ized by s ong winds, and is mainly p esen in high ion- iza ion lines (HILs) like Ci . – Ve y b oad componen (VBC): A i ialized componen wi h a la ge FWHM han he BC, as i models egions o he BLR ha a e close o he SMBH (a e y b oad-line egion (VBLR), Pe e son & Fe land 1986;Sulen ic e al. 2000c; Snedden & Gaskell 2007). I is a componen exclusi e o Pop. B objec s ha a e modeled wi h a Gaussian ha is ed- shi ed by se e alkms−1 o he es - ame wa eleng h o he emission line. I is a simpli ica ion in oduced o accoun o he g adien s mo e p ope ly modeled in he “locally op imally emi ing clouds” (LOCs) scheme (Baldwin e al. 1995) ha accoun s o he s a i ica ion o p ope ies in he BLR, assuming ha he e is a g adual g adien o a ia ion in p ope ies in he BLR ha ollows a powe law. An addi ional componen ha can be encoun e ed is he na - ow componen (NC), which is, howe e , no associa ed wi h he BLR, and hus modeled wi h na ow lines and dis ega ded in he es ima ion o he me allici y o he gas in he BLR. 3.3. Mul icomponen i ing echnique Ou sample was analyzed using he spec i (K iss 1994) ask om IRAF, inspec ing he egions o ou spec a cen- e ed on he Sii +Oi ]λ1400, Ci λ1549, Ciii]λ1909 (he ea e , Sii +Oi ], Ci , Ciii], espec i ely), and Hβlines. The labo- a o y acuum wa eleng hs o hese lines, as well as all o he lines measu ed du ing he analysis, a e om Vanden Be k e al. (2001). The spec i ask uns on he Space Telescope Science Da a Analysis Sys em (STSDAS) con ib package o IRAF, which was de eloped by he Na ional Op ical As onomy Obse - a o y (NOAO) in Tucson, A izona (Tody 1986,1993). Spec i allows use s o in e ac i ely i a wide ange o emission and abso p ion lines on he con inuum model o he A321, page 3 o 31 Flo is, A., e al.: A&A, 689, A321 (2024) Table 2. Flux and unce ain y measu emen s. Objec name λ1300–1450Å λ1450–2100 Å Op ical ange A(V) λ1350 RMSSiIV λ1700 RMSCIV RMSCIII] λ5100 RMSop (1) (2) (3) (4) (5) (6) (7) (8) (9) M k 335 22.11 0.65 14.83 0.70 1.30 5.25 0.12 0.096 I Zw 1 10.15 0.38 16.43 0.70 0.37 6.65 0.12 0.176 Fai all 9 35.19 0.91 26.36 1.63 1.19 2.72 0.09 0.071 PHL 1092 3.40 0.08 2.77 0.06 0.07 0.28 0.01 0.110 A k 120 33.13 2.32 13.81 1.00 0.88 9.67 0.15 0.349 M k 110 24.01 1.57 16.02 1.74 ... 0.76 0.03 0.034 Ton 28 10.36 3.11 7.98 1.38 1.25 1.92 0.04 0.060 NGC 3783 79.16 2.05 63.26 2.84 2.09 10.36 0.20 0.332 LB 2522 6.25 2.08 5.89 1.67 1.27 0.54 0.02 0.022 LEDA 51016 8.90 0.24 7.22 0.45 0.26 0.79 0.02 0.024 M k 478 36.23 2.06 26.93 0.93 0.52 2.00 0.02 0.039 M k 509 84.18 1.71 73.99 2.10 1.91 8.87 0.31 0.157 A k 564 6.86 0.52 6.06 0.84 0.27 0.75 0.04 0.166 No es. (1) Objec ’s common name. (2), (4), (7) Speci ic lux o he con inuum emission measu ed a 1350Å, 1700 Å, and 5100 Å, espec i ely, and measu ed in uni s o 10−15 e gs−1cm−2Å−1. (3), (5), (6), (8) RMS o he lux associa ed wi h he con inuum emission, measu ed in uni s o 10−15 e gs−1cm−2Å−1, in he anges o he a ious lines inspec ed in his wo k. (9) Abso p ion pa ame e in he V band ob ained om he NED da abase. inpu spec um. To achie e his, i employs inpu ASCII iles c e- a ed using he dbc ea e ask, con aining he necessa y unc- ions o model a ious componen s o he spec um. Gaussian and Lo en zian unc ions, widely used in mul icomponen i ing, a e each desc ibed by he alues o he line in ensi y, cen al wa e- leng h,FWHM,andasymme ypa ame e .Inaddi ion o hes an- da d unc ionsp o ided,spec i allows heuseo use -p o ided unc ions. In pa icula , we u ilized he use con and use line unc ions o model he complex Feiii ( ound in heUV ange) and Feii (also ound in he op ical ange) emissions, espec i ely. The spec i mul icomponen i ing p ocedu e is in e ac- i e, based on he inpu ile desc ibed in he p e ious pa a- g aphs. The ini ial pa ame e alues in he inpu ile se e as s a ing guesses o modeling he componen s. Each pa ame e can be e ined using a minimiza ion i p o ided by he ou ine. When he algo i hm con e ges success ully, i p o ides he bes - i pa ame e s ha minimize he χ2be ween he model and he obse ed da a, esul ing in an op imal solu ion. This i ing is ca - ied ou wi h wo di e en algo i hms (ou o he i e a ailable in spec i ): – Ma qua d is based on he Le enbe g-Ma qua d (LM) algo- i hm, which i e a i ely minimizes he χ2using he Jaco- bian ma ix. This ma ix con ains he pa ial de i a i es o he model unc ion i ing he da a o each i pa ame e . The algo i hm combines g adien descen and Gauss-New on me hods, adjus ing he s ep size a each i e a ion. I akes la ge s eps when he i ing imp o es and smalle s eps when i does no . – The downhill Simplex algo i hm u ilizes an i e a i e me hod o i ing, and o adjus ing alues o use -p o ided com- ponen s, s a ing om ini ial guesses and cons ain s. I i e - a i ely imp o es he solu ion owa d he minimum χ2wi h ope a ions on an N-dimensional ( he numbe o ee pa am- e e s) simplex (P ess e al. 1992). Unlike he LM me hod, he simplex algo i hm equi es only unc ion e alua ions and no de i a i es. E en i i is no e icien in e ms o he i e a ion numbe , i is e y obus and ne e causes he spec i ask o c ash. We s a ed by i ing he con inuum emission in he wa eleng h ange wi h a powe law unc ion, ca e ully selec ing egions o he spec um una ec ed by he p esence o emission o abso p- ion lines. The lux alues o he con inuum a a ious e e ence wa eleng hs a e p o ided in Table 2. In he op ical wa eleng h ange, we inco po a ed a Feii empla e o co ec o Feii con amina ion, which can ac as a pseudo-con inuum on he modeled spec um and which is pa - icula ly p oblema ic in objec s wi h s ong Feii emission. Ou Feii empla e was de i ed om a Eu opean Sou he n Obse - a o y (ESO) spec um o I Zw 1 (Ma ziani e al. 2009,2010, 2021). To add ess Feiii con amina ion, which is p e alen in he λ1800–2100Å wa eleng h ange, we u ilized a empla e om Ves e gaa d & Wilkes (2001) based on he Feiii emission o I Zw 1. In cases in which he Feiii emission is s ong, and he e is e idence o a s onge ea u e a 1914Å o he empla e, we modeled he 1914Å ea u e wi h a Gaussian p o ile, e en i i was se e ely blended wi h Ciii]. This s ep is ypically necessa y o xA objec s. When i ing emission lines, we allowed hem o a y, while imposing a ule ha he NC o lines emain ixed a he es ame o he objec . Addi ionally, we ensu ed ha he FWHM o Sii +Oi ] was simila o hose measu ed o Ci and Hβ (Ma ziani e al. 2010). This is because lines associa ed wi h he i ialized componen o Hβ, Ciii], Ci , and Sii +Oi ] a e p o- duced in app oxima ely he same egions. By he same oken, BLUE and VBC wid hs and asymme ies a e also made consis- en in he i . All o he measu emen s epo ed in he ables h oughou he pape and hose employed in he me allici y compu a ions ha e been co ec ed o galac ic ex inc ion o each objec using he equa ions o Ca delli e al. (1989). 3.4. Es ima ion o unce ain ies The placemen o he con inuum is one o he main sou ces o unce ain y in measu ing emission-line in ensi ies. In pa icula , he con inuum placemen s ongly in luences ex ended ea u es A321, page 4 o 31 Flo is, A., e al.: A&A, 689, A321 (2024) such as he Feiii and Heiiλ1640 emissions in he UV ange, and he Feii emission in he Hβ ange. The unce ain y associa ed wi h con inuum placemen was es ima ed as ollows (c . ´ Sniegowska e al. 2021): ini ially, he in ensi y o he con inuum was measu ed using IRAF in each ange in which he bes i was conduc ed, yielding he oo mean squa e (RMS) o he lux (measu ed in uni s o 10−15 e gs−1cm−2Å−1), as is shown in Table 2. Subsequen ly, he in ensi y o each line was measu ed by ixing he con- inuum’s in ensi y a bo h maximum and minimum alues, by adding o sub ac ing he RMS, o ob ain he line’s in ensi y a i s minimum and maximum, espec i ely. When he con inuum is highe (lowe ), he line in ensi y dec eases (inc eases), compen- sa ing o he lux change, esul ing in lowe (uppe ) limi s ha a e asymme ic and ep esen he ange o emission line in ensi- ies.To handle hese unce ain ies, we ha e assumed ha he e o dis ibu ion ollows a iangula dis ibu ion (D’Agos ini 2003). This simpli ies e o compu a ion and is he p e e ed me hod o dealing wi h asymme ic unce ain ies. Fo each line mea- su emen , he a iance σ2(X) o any pa ame e Xwas calcula ed using he o mula o he iangula dis ibu ion: σ2(X)=∆2x++ ∆2x−+ ∆x++ ∆x− 18 ,(2) whe e ∆x+and ∆x− ep esen he di e ences be ween measu e- men s wi h he maximum and bes con inuum, and wi h he bes and minimum con inuum, espec i ely. Finally, e o s in spec al diagnos ic a ios we e calcula ed using s anda d e o p opaga- ion o mulas. 3.5. CLOUDY pho oioniza ion simula ions To cons ain he p ope ies o he gas in he BLR, we conduc ed pho oioniza ion simula ions using he CLOUDY spec al syn he- sis code (Fe land e al. 2017). CLOUDY is a C++ code ha models he ioniza ion, chemical, and he mal s a e o he ma e ial ha may be exposed o an ex e nal adia ion ield o o he sou ces o hea ing. I p edic s obse ables, including emission and abso p ion spec a, by sol ing he s a is ical and he mal equi- lib ium equa ions o many chemical species, including ions. The code accoun s o collisional exci a ions and adia i e p o- cesses, and ea s adia ion ans e h ough an escape p obabili y o malism. Pho oioniza ion simula ions equi e inpu pa ame e s ha de ine ioniza ion and le el popula ions, along wi h elec on em- pe a u e and op ical hickness as a unc ion o he geome ical dep h wi hin he emi ing gas cloud. The assumed geome y is open and plane-pa allel, ep esen ing a cloud o emi ing gas as a slab exposed o a adia ion ield on only one side. The inpu pa ame e s o a comple e compu a ion o he gas s a e, used o p edic emission line in ensi ies, include: 1. The ioniza ion pa ame e , U, de ined by he equa ion U=R∞ ν0 Lν hνdν 4π 2 BLRcnH =Q(H) 4π 2 BLRcnH ,(3) whe e Q(H) is he numbe o ionizing pho ons and BLR is he dis ance be ween he con inuum sou ce and he line emi - ing gas. I p o ides he a io be ween he hyd ogen-ionizing pho ons and he hyd ogen numbe densi y. 2. Hyd ogen gas numbe densi y, nH[cm−3]. 3. Me allici y, Z, o sola me allici y, Z, de ined as log Z= loghZ Hi−loghZ Hi, whe e hZ Hiand hZ Hia e he a io o he measu ed numbe densi y o me als o hyd ogen, and he same quan i y o he Sun, espec i ely. 4. Spec al ene gy dis ibu ion (SED) o he quasa . 5. Column densi y, Nc. 6. Mic o u bulence pa ame e . This esul ing 6D pa ame e space is no ypically ully explo ed, and in ou case we examined ends using a ays o simula ions ha we e o ganized as ollows. We used h ee di e en SEDs, one o each o he ollowing cases: Pop. B RQ (Lao e al. 1997a), an SED o Pop. A sou ces (Ma hews & Fe land 1987), and one o xA quasa s (Fe land e al. 2020). Zwas assumed o scale as sola (Z), wi h 12 alues anging be ween 0.01 and 1000Z o Pop. A and xA sou ces, and 14 alues be ween 0.001 and 20Z o Pop. B. The mic o u bulence pa ame e was se o 0km s−1. This is negligible o esonance UV lines, as is ou lined in ´ Sniegowska e al. (2021), bu is expec ed o lead o an unde p edic ion o Feii emission (Sigu & P adhan 2003;B uhweile & Ve ne 2008;Panda e al. 2018,2019a; ´ Sniegowska e al. 2021;Panda 2021b). Fo each me allici y alue, we conside ed an a ay o simula ions co e ing he nHand Upa ame e plane in he ange o 7 ≤log nH≤ 14cm−3,−4.5≤log U≤1 o Pop. A (667 simula ions) and 7≤lognH≤13 cm−3,−3≤log U≤1 o Pop. B (425 simula ions). The sepa a ion o he h ee componen s (BC, VBC, and BLUE) can be heu is ically associa ed wi h h ee di e - en egions expec ed o be in di e en physical condi ions, BLR, VBLR, and BLUE (see also he scheme in Fig. 2 Decon o-Machado e al. 2023). A sou ce o unce ain y comes om he ac ha pho oioniza ion compu a ions we e ca ied ou unde he assump ion o single-zone emission, albei sepa- a ely o each o he h ee egions. While his assump ion seems a good one o he BLR o ex eme Pop. A, in which densi y and ioniza ion end owa d limi ing alues (Ma ziani e al. 2001; Panda e al. 2019b;Panda 2021b), his migh no be he case o Pop. B sou ces (Ko is a & Goad 2000), o o he less-ex eme pa o Pop. A. 3.6. Diagnos ic a ios Diagnos ic a ios o lines p o ide cons ain s on he ioniza ion, me allici y, and densi y o he emi ing gas in he BLR. The ol- lowing diagnos ic a ios we e used o compu ing he me allici y o ou low-zquasa sample. 1. (SiIV+OIV])/CIV: widely used as a me allici y indica o (Nagao e al. 2006;Shin e al. 2013). An explana ion o why his a io is a eliable me allici y indica o is p o- ided by Huang e al. (2023). In sho , he ionic ac ion o Si+3inc eases wi h me allici y ( he con inuum igh abo e he ioniza ion po en ial o Si+22.5 Ryd emains signi i- can ), while he ionic ac ion o C+3, con e sely, s ongly dec eases wi h me allici y, as he con inuum abo e 4 Ryd is almos comple ely abso bed in he inne mos laye s o he emi ing gas. E en i bo h lines a e diminished because he elec on empe a u e goes down as me al abundance goes up, he a io o Sii +Oi ]/Ci inc eases because o he dec eas- ing C+3ionic ac ion wi h me allici y. 2. (SiIV+OIV])/HeII: sensi i e o me allici y, assuming ha he He abundance ela i e o hyd ogen is cons an . The OIV] line has a c i ical densi y o nc∼1011 cm−3(Ma ziani e al. 2020), making his diagnos ic a io a ec ed by densi y. A321, page 5 o 31 Flo is, A., e al.: A&A, 689, A321 (2024) 3. CIV/HeII: diagnos ic a io sensi i e o me allici y, again assuming a cons an He abundance ela i e o hyd ogen. The easons why he Sii +Oi ]/Heiiλ1640 and Ci /Heiiλ1640 a e e y eliable me allici y indica o s a e simila . Fi s , he e is he linea inc ease in he abundance o Si and C o He wi h inc easing Z. In he Ci /Heiiλ1640 case, he ioniza- ion po en ials o C2+and He+a e almos coinciden ; how- e e , he main di e ence is ha Heiiλ1640 is a ecom- bina ion line and he Ci is due o collisional exci a ion (Ma ziani e al. 2020;´ Sniegowska e al. 2021). As he me al- lici y Zinc eases, he elec on empe a u e, Te, dec eases (e.g., Pagel e al. 1979), and hence me al line exci a ion is dis a o ed. Con e sely, he lowe Teinc eases he a e o He++ ecombina ion, ein o cing he inc ease in he a ios wi h Z. The a io Ci /Heiiλ1640 has been used (Shin e al. 2013;Sulen ic e al. 2014) despi e he Heiiλ1640 weakness. Howe e , wi h high S/N and esolu ion measu emen s om he Fain Objec Spec og aph (FOS), and especially he Cos- mic O igins Spec og aph (COS), we can accu a ely es ima e he line in ensi y o Heiiλ1640, which was p e iously e y di icul o assess, and sel -consis en ly deblend i om he much s onge Ci emission. 4. CIV/Hβ: unlike all o he o he a ios employed in his wo k, he Ci /Hβ a io is subjec o majo obse a ional unce - ain y in he pho ome ic p ecision o he op ical da a, and a lack o concu en op ical and UV obse a ions. The dep h o abso p ion ea u es close o he es ame in he Ci p o ile is di icul o assess and may signi ican ly lowe he Ci /Hβ (one o he mos se ious cases could be PHL 1092, Table 10). The diagnos ic alue o he Ci /Hβ a io is p ima ily due o i s s ong dependence on he ioniza ion pa ame e . Fig. 2 shows he beha io o he Ci /Hβin ensi y a io in he plane logZ, log U: he isopho al con ou s a e almos pa allel o he Zaxis. In he p esence o a e y low Ci /Hβ(∼1) a io, he Umus be e y low, wi h log U.−2.5. A highe ioniza ion pa am- e e s o −1.logU.0, and mode a e densi y, he Ci emis- sion o he Balme line is op imized a ela i ely low alues o Z. A highe Z(&10Z), o he same U, he elec on em- pe a u e is signi ican ly lowe ed, dec easing he collisional exci a ion a e, and he e o e lowe ing he line emission. The e a e wo implica ions ele an o his wo k: a e y weak Ci equi es e y low U, wi h he lowes Ci p ominence ob ained o low Z(Fig. 2). Con e sely, o log Ube ween 1 and 0, whe e Ci is e icien ly p oduced, Ci /Hβ&10 can be achie ed o logZ∼ −2, wi h a densi y lognHbe ween 11 and 13. 5. AlIII/CIV: a diagnos ic a io sensi i e o he ioniza ion pa ame e . This is because i ela es lines o elemen s wi h di e en ioniza ion po en ials. I is po en ially use ul o con- s ain he e ec o eedback om supe no æ since Aliii is o e p oduced in he explosion o ype Ibc supe no ae o he Ci sola abundances (Chie i& Limongi 2013). This diag- nos ic a io may end o bias he es ima ed Z owa d highe alues. 6. AlIII/SiIII]: a diagnos ic a io sensi i e o densi y, since i in ol es an in e combina ion line wi h a well-de ined c i i- cal densi y o nc∼1011 cm−3 o Siiii] (Neg e e e al. 2013). I he a io Aliii/Siiii] is 1, he esul ing densi y will be much g ea e han he c i ical densi y. His o ically, diagnos- ic a ios employing Aliii ha e a ely been used, since he measu emen o he in ensi y o Aliii was qui e unce ain, bu wi h high-quali y COS and FOS spec a, we ha e been able o measu e i accu a ely. Fig. 2. Beha io o he diagnos ic a io, Ci /Hβ, (shown on a loga i h- mic scale) as a unc ion o he ioniza ion pa ame e , U, and me allici y, Z, o ou alues o hyd ogen densi y, log nH=7,9,11,13 [cm−3], using he Ma hews & Fe land (1987) Pop. A SED. 7. SiIII]/CIII]: a diagnos ic a io sensi i e o densi y, simila o Aliii/Siiii] bu a ound lowe alues, nc∼1010 cm−3 o Ciii] (Hamann e al. 2002). Like he a io desc ibed abo e, i Siiii]/Ciii] is 1, he esul ing densi y will be much g ea e A321, page 6 o 31 Flo is, A., e al.: A&A, 689, A321 (2024) Fig. 3. Le : plo o he χ2/χ2 min dis ibu ion as a unc ion o he ela i e in ensi y o he FeIIBC o PHL 1092 wi h espec o he spec i e ie ed in ensi y, s. he ela i e in ensi y o he FeIIBLUE wi h espec o he o iginal in ensi y o FeIIBC. Middle: same, bu o he AlIIIVBC emission o Fai all 9. Righ : same, bu o he FeIIVBC emission o Fai all 9. han he c i ical densi y o Ciii] as well, and is a e y pow- e ul densi y es ima o , cons aining densi y e y well. The inclusion o s ong Ciii] may, howe e , bias he (U,nH,Z) solu ion owa d lowe densi ies. This issue is analyzed in Sec ion 5. 8. CIII]/CIV: a diagnos ic a io sensi i e o he ioniza ion pa ame e as i compa es he in ensi y o di e en ioniza ion deg ees o he same a omic species. Howe e , since Ciii] is an in e combina ion line, i is also sensi i e o densi y. Fo xA sou ces, he a io Ciii]/Ci is subjec o highe unce ain- ies due o Feiii con amina ion, making i di icul o measu e he in ensi y o Ciii] accu a ely. Howe e , since xA Ciii] is weakes along he MS (Bache e al. 2004;Aoki & Yoshida 1999), a ai ly high-densi y solu ion is sugges ed anyway despi e he unce ain ies. 9. FeII/Hβ: a diagnos ic a io sensi i e o me allici y, as i in ol es a me al and hyd ogen. Howe e , his diagnos ic a io is also sensi i e o he densi y, ioniza ion pa ame e , and column densi y o he line-emi ing gas (Ve ne e al. 1999, 2004;B uhweile & Ve ne 2008;Panda e al. 2018,2019b; Panda 2021b). High Feii/Hβis only possible i condi ions o low ioniza ion, high densi y (nH&1011 cm−3), and a la ge column densi y (Nc&1023 cm−2) a e sa is ied (Panda e al. 2018,2019b;Panda 2021b). The a io Feii/Hβis ela- i ely easy o measu e wi h good p ecision (∼0.1a a1σ con idence le el). Howe e , he mean escape p obabili y o - malism is known o pe o m poo ly in he p edic ion o he emission line s eng h o gas i adia ed by s ong X- ay sou ces (e.g., Dumon e al. 2003, and e e ences he ein). In he case o his pa icula a io, he e is he e o e conside - able unce ain y associa ed wi h he pho oioniza ion compu- a ions. 10. HeIIop /Hβ: a diagnos ic a io sensi i e o he ioniza ion pa ame e . Assuming he abundance o He ela i e o hyd o- gen is cons an , i is he a io be ween wo species wi h e y di e en ioniza ion po en ials: Heiiλ4686 equi es ∼50eV o be ionized, while hyd ogen equi es ∼10eV. The Heiiλ4686/Hβ a io is ela i ely insigni ican in he Zes i- ma es. These diagnos ic a ios a e used di e en ly depending on he quasa popula ion. No ably, no all componen s a e p esen in all objec s. Al hough mos objec s can be modeled assuming a BLUE componen o HILs, i is no always p esen o LILs. By he same oken, he VBC appea s only in Pop. B sou ces. Fo Pop. A objec s, eliable es ima ion is only possible o physical pa ame e s associa ed wi h he BC. Fo Pop. B quasa s, i is pos- sible o es ima e he physical pa ame e s o gas loca ed in he VBLR and BLR, while o xA quasa s i is possible o es ima e he p ope ies o he BLUE and BC. In his s udy, we exclusi ely examined he BLUE o PHL 1092, since i was he only objec wi h a su icien numbe o diagnos ic a ios enabling an accu a e cons ain o i s physical condi ions. The solu ion o he single-zone model, ep esen ed by a sin- gle poin in he 3D space (nH,U,Z), was de e mined by minimiz- ing he χ2compu ed om he di e ence be ween he obse ed line a ios and he p edic ed line a ios ac oss he en i e 3D space (Ma ziani e al. 2024), applying he ollowing ela ion: χ2 kc(nH,U,Z)=X i Rkci −Rkci,mod(nH,U,Z) δRkci !2 ,(4) wi h Rkc iden i ying he diagnos ic a io conside ed o he objec and componen , and δRkc iden i ying he e o associa ed wi h each componen . The “mod” subsc ip iden i ies he alue o he speci ic diagnos ic a io p edic ed by he CLOUDY simula ions o each alue in he 3D g id o he (nH,U,Z) pa ame e space. 3.7. FeII and AlIII uppe limi s CLOUDY pho oioniza ion simula ions ha e p e iously hin ed a he possibili y o a con ibu ion o he in ensi y o Aliii and Feii emission om he VBC o Pop. B Quasa s and om he BLUE o xA Quasa s (Zhou e al. 2023). This possibili y was hus in es iga ed by damping he in ensi y o Aliii emission ob ained om he mul icomponen i ing o ou spec a, adding a Gaussian componen o he i wi h he same blueshi ( edshi ) o xA quasa s ( o Pop. B quasa s) co esponding o he shi be ween he HβBC and he HβBLUE (HβVBC), and he same FWHM as he BLUE (VBC). A e including his addi ional componen , he χ2associa ed wi h he i was calcula ed o a y- ing in ensi y alues o he Gaussian line. As a esul , egions associa ed wi h 1σ, 2σ, and 3σcompa ibili y we e d awn in he plo s shown in Figu e 3. The same p ocedu e was applied o he Feii emission in he op ical wa eleng h ange, wi h he di - e ence ha ins ead o adding a Gaussian componen , he addi- ional Feii emission was modeled om he ESO Feii empla e (Ma ziani e al. 2009,2010). The esul s o his compu a ion we e hen added o he known diagnos ic a ios associa ed wi h he VBC o he BLUE o ou objec s, by using he 1σuppe limi as he diagnos ic a io. 4. Resul s 4.1. Es ima ion o me allici y and physical pa ame e s Tables 3,4,5, and 6 epo he in ensi y o each di e en line componen o he objec s o he sample measu ed om he mul- A321, page 7 o 31 Flo is, A., e al.: A&A, 689, A321 (2024) Table 3. Sii +Oi ] blend line measu emen s. Objec name Sii +Oi ] BLUE BC VBC (1) (2) (3) (4) M k 335 85.2±25.4 634.9±74.6 ... M k 110 ... 706.3±104.9 ... M k 509 251.7±98.9 1859 ±300 ... A k 564 11.5±10.5 115.5±24.5 ... M k 478 6.9±22.8 541 ±122 ... Ton 28 63.5±40.7 200 ±139 ... LB 2522 146 ±67 76.9±113.5 ... LEDA 51016 34.3±12.2 305 ±35 ... I Zw 1 104.9±22.8 481 ±56 ... PHL 1092 27.1±5.6 7.31 ±3.55 ... Fai all 9 108 ±26 363 ±32 390 ±70 A k 120 220 ±114 881 ±155 501 ±188 NGC 3783 404 ±141 1013 ±159 539 ±145 No es. (1) Objec ’s common name. (2) In ensi y o he line associ- a ed wi h he BLUE, wi h associa ed e o s, measu ed in uni s o 10−15 e gs−1cm−2. (3) In ensi y o he line associa ed wi h he BC, wi h associa ed e o s, measu ed in uni s o 10−15 e gs−1cm−2. (4) In ensi y o he line associa ed wi h he VBC, wi h associa ed e o s, measu ed in uni s o 10−15 e gs−1cm−2. icomponen i ing, wi h associa ed e o s, o he ou spec- al anges ha ha e been i sepa a ely; namely, Sii +Oi ], Ci +Heiiλ1640, Aliii+Siiii]+Ciii], and Hβ+Feii. The i s a e shown in he igu es o Appendix A, and he p ojec ions o he (U,nH,Z) pa ame e space in Appendix B. Table 7summa izes he diagnos ic in ensi y a ios used o he es ima ion o he phys- ical pa ame e s. The es ima es o he me allici y and physical condi ions in he BLR o ou sample we e ob ained using he me hodology desc ibed in Sec ion 3, and he esul s a e epo ed in Table 8. Resul s o indi idual sou ces a e discussed in mo e de ail in Appendix C. 4.2. The end along he main sequence Figu e 4depic s he MS ep esen a ion o ou sample, along- side ep esen a i e alues om S21,G22, and M23. The main esul is ha he highes -me allici y objec s a e clus e ed in he xA egion o he MS. Con e sely, all o he objec s wi h sola o subsola me allici ies a e si ua ed in he le po ion o he MS, demons a ing a g adien o me allici y ha peaks in he xA segmen o he MS and g adually dec eases owa d RL Pop. B sou ces – an ou s anding esul ha con i ms he heo e ical p edic ions o Panda e al. (2019b). Fo he sou ces o ST A1 and B1 wi h s onge Feii, we encoun e Za ound sola , which becomes supe sola o RFeII &0.5. An impo an esul is ha hese sou ces a e a low z, and each Zcompa able o he sam- ples o S21 and G22 ha a e high zand ha a e associa ed wi h quasa s o e y high MBH and ex eme luminosi y. Con e sely, xA sou ces om he p esen sample a e AGNs o mode a e lumi- nosi y, na ow line Sey e 1s (NLSy1s) wi h MBH .108M. Since he high-zobjec s a e ex eme adia o s a a high Edding- on a io o Lbol/LEdd ∼1, as he xA o ou sample a e, he en ichmen phenomenon appea s o be independen o cosmic epoch and o black hole mass, a leas up o cosmic ages o ∼3Gy a e he Big Bang. The mechanisms ha could explain he high-Z alues and he p esence o a g adien o me allici y a e b ie ly discussed in Sec ion 6.5. 5. Valida ion In he ollowing sec ions, we discuss he limi s o ou echnique and he es s ha we used o add ess hese limi s. 5.1. S a is ical e o s The minimum χ2c i e ion and he F- es a e no sui ed o es i- ma e he in luence o s a is ical e o s. In his espec , boo - s ap epe i ion o he Z,nH, and Ues ima e allows o de ia- ion om a pe ec solu ion (i.e., om he diagnos ic a io al- ues p edic ed by a de ini e model Z,nH, and U) and enables one o assess how la ge he ela i e unce ain ies in he in en- si y a ios can be un il he co ec solu ion in e ms o me allici y is los . Boo s ap simula ions we e ca ied ou , assuming an- dom a ia ions and ollowing a Gaussian dis ibu ion wi h σco - esponding o se e al ac ional unce ain ies. In p ac ice, each eplica ion assigned new alues o he a ios, ollowing Gaus- sian de ia es wi h a dispe sion co esponding o he ac ional unce ain y. Fig. 5shows ha he Zin o ma ion is p ese ed o s a is ical e o s up o ≈0.45 o he somewha idealized case o en diagnos ic a ios ha ing a cons an ac ional unce ain y, assuming diagnos ic a ios om simula ed cases wi h me allic- i y anging om logZ=0.3 o 1.7, which could be ai ly ypi- cal o Pop. A and xA. The s a is ical unce ain ies do no a ec he Zes ima es up o 0.3–0.35 ac ional e o s. Inc easing he s a is ical unce ain ies by 50% o a ios Ci /Heiiλ1640 and Sii +Oi ]/Heiiλ1640, because o he ain ness o Heiiλ1640, lowe s he highes unce ain y p ese ing he Z alue by 0.05 (i.e., om 35% o 30%). The s a is ical e o s end o lowe Z (as well as Uand nH, which a e no shown). The e o e, s a is i- cal luc ua ions a e unlikely o induce sys ema ically highe al- ues o he me allici y o he ones de i ed om he model and o accoun o he e y high Zdeduced o he ex eme Pop. A sou ces. Mo e ele an o he analysis o he esul is he possi- bili y o sys ema ic disc epancies in emission line a ios o he expec a ion o a pho oioniza ion solu ion. 5.2. Al e na i e χ2e alua ions The ques ion is whe he he echnique employed o he χ2es i- ma e could be a he o igin o he ou -o -scale Zo PHL 1092, and o he e y high alues (Z.100) encoun e ed among xA sou ces. The χ2exp ession o Eq. (4) is weigh ing he con i- bu ion o each measu emen on he in e se o he unce ain y squa ed. This is expec ed o be e ec i e in inc easing he likeli- hood o he solu ion, bu a he same ime gi es minimum weigh o “bad” da a ha ha e la ge unce ain y. Se e al a ios in Table 7a e be ween wo de ec ed lines, bu ha e an unce ain y la ge han he epo ed alue. We he e o e conside ed wo al e na i es o he χ2 echnique epo ed in Table 8, and hose a e shown in Table 9. Ins ead o di iding each addendum by he e o squa ed, he addenda could be di ided by he alue expec ed om he sim- ula ion o each (U,nH,Z) g id elemen (di e en o each U, nH,Z), o by he obse ed a ios ha a e cons an . This las op ion especially o e s a alid al e na i e o he con en ional χ2 because i e ec i ely quan i ies he ela i e di e ence be ween any model and he obse ed se o a ios. Each a io will ha e a A321, page 8 o 31 Flo is, A., e al.: A&A, 689, A321 (2024) Table 4. Ci and Heiiλ1640 line measu emen s. Objec name Ci Heiiλ1640 BLUE BC VBC BLUE BC VBC (1) (2) (3) (4) (5) (6) (7) M k 335 24.9±16.9 3943 ±274 ... 153 ±33 454 ±77 ... M k 110 68.2±40.2 5526 ±427 ... ... 413 ±99 ... M k 509 266 ±130 11100 ±994 ... 394 ±154 1096 ±300 ... A k 564 29.5±21.5 409 ±54 ... 42.1±24.9 140 ±30 ... M k 478 213 ±30 820 ±100 ... 124 ±25 102 ±35 ... Ton 28 203 ±65 221 ±88 ... 63.9±46.5 36.4±62.7 ... LB 2522 201 ±83 87.6±99.2 ... 53.5±59.8 31.8±95.1 ... LEDA 51016 151 ±25 446 ±67 ... 58.4±17.1 42.5±25.0 ... I Zw 1 987 ±97 678 ±96 ... 197 ±47 143 ±45 ... PHL 1092 26.2±4.4 5.23 ±2.84 ... 3.17 ±2.06 0.56 ±2.23 ... Fai all 9 448 ±84 2354 ±185 1772 ±295 14.7±33.2 172 ±59 127 ±86 A k 120 1500 ±155 2500 ±169 1002 ±154 ... 150 ±48 401 ±107 NGC 3783 2318 ±282 8720 ±534 6119 ±867 85.6±86.0 794 ±175 2180 ±570 No es. (1) Objec ’s common name. (2),(5) In ensi y o he line associa ed wi h he BLUE, wi h associa ed e o s, measu ed in uni s o 10−15 e gs−1cm−2. (3),(6) In ensi y o he line associa ed wi h he BC, wi h associa ed e o s, measu ed in uni s o 10−15 e gs−1cm−2. (4),(7) In ensi y o he line associa ed wi h he VBC, wi h associa ed e o s, measu ed in uni s o 10−15 e gs−1cm−2. Table 5. λ1900 blend line measu emen s. Objec name AlIII SiIII] CIII] BLUE BC BLUE BC BLUE BC VBC (1) (2) (3) (4) (5) (6) (7) (8) M k 335 ... 42.5±45.9 ... 186 ±68 ... 705 ±125 ... M k110 ... ... ... ... ... ... ... M k 509 ... 441 ±210 ... 916 ±277 ... 2034 ±401 ... A k 564 ... 14.6±7.4 ... 36.5±10.5 ... 231 ±25 ... M k 478 ... 101 ±26 ... 217 ±38 ... 298 ±46 ... Ton 28 ... 135 ±81 ... 84.8±74.3 ... 51.2±51.8 ... LB 2522 ... 60.2±87.4 ... 13.1±89.4 ... 85.7±92.0 ... LEDA 51016 ... 30.9±14.6 ... 93.9±23.0 ... 94.0±23.1 ... I Zw 1 ... 153 ±34 ... 270 ±46 ... 162 ±35 ... PHL 1092 3.9±1.6 9.6±4.1 4.1±1.7 9.7±4.2 3.3±1.5 1.2±3.9 ... Fai all 9 ... 137 ±43 ... 440 ±70 ... 294 ±59 189 ±51 A k 120 ... 116 ±51 ... 396 ±82 ... 457 ±77 301 ±93 NGC 3783 ... 499 ±151 ... 661 ±170 ... 1160 ±201 1024 ±344 No es. (1) Objec ’s common name. (2),(4),(6) In ensi y o he line associa ed wi h he BLUE, wi h associa ed e o s, measu ed in uni s o 10−15 e gs−1cm−2. (3),(5),(7) In ensi y o he line associa ed wi h he BC, wi h associa ed e o s, measu ed in uni s o 10−15 e gs−1cm−2. (8) In ensi y o he line associa ed wi h he VBC, wi h associa ed e o s, measu ed in uni s o 10−15 e gs−1cm−2. weigh independen om i s associa ed unce ain y. I his exe - cise is ca ied ou , he minimum χ2solu ions con i m he high Z alues o bo h LB 2522 and PHL 1092. The Zde i ed o PHL 1092 BLUE e en inc eases o 20–50Z. The e o e, he me al- lici y es ima es a high Zdo no appea s ongly sensi i e o he di e en choices o χ2 ha we ha e conside ed. 5.3. In luence o op ical-UV misma ch Se e al sou ces a e a ec ed by an op ical-UV lux misma ch, likely o igina ed by obse a ions conduc ed a di e en momen s using bo h g ound-based and space-based obse a o ies. Op ical obse a ions a e suscep ible o uncon ollable ligh loss, signi i- can ly impac ing he eliabili y o diagnos ic a ios based on bo h op ical and UV lines. Howe e , diagnos ic a ios dependen on lines ound a simila wa eleng hs a e less suscep ible o his issue, as he in insic a iabili y o he quasa should simila ly a ec hem. The mos likely candida e o his p oblem is he Ci /Hβ a io, as i is he sole diagnos ic a io in ol ing a UV line and an op ical one. Indeed, when es ima ing he physical pa ame e s associa ed wi h he BLR o ou sample o quasa s, he Ci /Hβ a io o en dis up ed he compu a ion, e u ning combina ions o U,Z, and nH ha we e no compa ible wi h ou p e ious assump ions, o solu ions s ongly incompa ible wi h any o he CLOUDY simu- la ions used in his wo k: e y low Ci /Hβ alues a e possible o low U, o mos Zs (Fig. 2). The alues o Ci /Hβalong wi h he logUand log Z esul s ob ained om he calcula ion a e shown in Table 10. I is impo an o no e ha he Ci /Hβ a io was excluded in mos calcula ions (Table 7). The p ima y A321, page 9 o 31 Flo is, A., e al.: A&A, 689, A321 (2024) Fig. 7. Beha io o he diagnos ic a io Ci /Heiiλ1640 (magen a), Ci /Hβ(blue), RFeII (g een) on a loga i hmic scale s. mic o u bulence eloci y in kms−1. Le : using he solu ion o I Zw 1 (squa es) and PHL 1092 (ci cles). Righ : Using a low-ioniza ion, high-densi y solu- ion app op ia e o xA sou ces (see ex o de ails), wi h h ee di e en Z alues: 200Z(ci cles), 100Z(squa es), and 50 Z( iangles). Fig. 8. Diag am illus a ing he end o Zac oss di e en spec al ypes, wi h he e e ence alues o each spec al ype ob ained as he geome - ic mean o all he measu emen s assumed o all he objec s belonging o a speci ic spec al ype, including ou me hod, he use o he Ci /Hβ a io, and he p esence o mic o u bulence o all sou ces belonging o he A2, A3, and A4 spec al ypes. Con idence in e als a e ob ained as he s anda d de ia ion o ou esul s. The numbe o objec s con- ibu ing o each spec al ype is shown in g een, along wi h he a e age RFeII o he sou ces conside ed. Two e e ence poin s (in ed) a e shown desc ibing he RL and RQ Pop. B sou ces s udied in M23. and Ga nica e al. (2022). The p esence o such s ong co ela- ions is no i ial, as hese quan i ies we e de i ed independen ly and using di e en p ocedu es. Ou esul s con i m he ea lie inding conce ning low-zPG quasa s ha he BLR gas Zco ela es wi h Edding on a io, while Zshows a much weake co ela ion wi h MBH (Shin e al. 2013). Typical alues o Sii +Oi ]/Ci and N /Ci a low edshi s sugges ed alues a ound he sola one o a ew imes highe han i . The ange co e ed by he wo diagnos ic a ios was ound o be consis en wi h hose measu ed o a sample o in e media e- edshi quasa s o a luminosi y compa able o he one o luminous AGNs a low z(Sulen ic e al. 2014). The end wi h Lbol/LEdd is due o he inclusion o Pop. B, and spec- al ype A1s ha a e adia ing a a lowe Edding on a io wi h espec o xA sou ces. The absence o a clea end a in e - media e zco esponding o he cosmic noon wi h a maximum o nuclea ac i i y and s a o ma ion (e.g., Madau & Dickinson 2014;Flo ez e al. 2020;Fö s e Sch eibe & Wuy s 2020), and Table 12. De i ed p ope ies o he sample. Objec name log(Lbol) log(MBH)L/LEdd [e g s−1] [M] (1) (2) (3) (4) M k 335 44.95 ±0.03 7.78 ±0.05 0.10 ±0.01 M k 110 44.46 ±0.03 7.48 ±0.05 0.06 ±0.01 M k 509 45.34 ±0.03 8.21 ±0.16 0.09 ±0.03 A k 564 44.25 ±0.03 6.94 ±0.06 0.14 ±0.02 M k 478 45.32 ±0.03 7.51 ±0.04 0.43 ±0.05 Ton 28 46.26 ±0.03 8.26 ±0.05 0.67 ±0.10 I Zw 1 45.62 ±0.03 7.45 ±0.05 0.98 ±0.12 LB 2522 46.03 ±0.03 8.22 ±0.06 0.43 ±0.07 LEDA 51016 45.26 ±0.03 7.75 ±0.05 0.22 ±0.02 PHL 1092 45.72 ±0.03 7.71 ±0.05 0.67 ±0.12 Fai all 9 45.08 ±0.03 8.14 ±0.05 0.06 ±0.01 A k 120 45.40 ±0.03 8.29 ±0.04 0.09 ±0.01 NGC 3783 44.52 ±0.03 7.71 ±0.05 0.04 ±0.01 No es. (1) Objec ’s common name. (2) Bolome ic luminosi y. (3) Mass o he SMBH a he cen e o he galaxy. (4) Edding on a io associa ed wi h acc e ion on o he SMBH. he la ge Zes ima ed o quasa s a ha cosmic epoch, migh jus be he esul o a selec ion e ec dis a o ing he lowes Lbol/LEdd o any gi en MBH (Sulen ic e al. 2014). The h ee emaining co ela ions depic ed in Figu e 9we e no ound o be s a is ically signi ican . The absence o a sig- ni ican co ela ion be ween logZand log MBH was pa icula ly puzzling, as hese quan i ies ha e been ound o co ela e in o he wo ks (Xu e al. 2018;Maiolino & Mannucci 2019), and la ge SMBHs we e ound o be he ones wi h he highes me allici y, since hei s onge g a i a ional po en ial is able o ap he mo e me allic ma e in hei i ialized egion. 6.4. Compa ison be ween he pho oioniza ion and e e be a ion solu ions o BLR The compu a ion o physical pa ame e s in he BLR, cou- pled wi h Equa ion (3) o he ioniza ion pa ame e , yields an independen measu emen o BLR, which could be com- pa ed wi h he esul s ob ained om Hβ e e be a ion map- ping o ou sou ces (c . Wandel 2000;Neg e e e al. 2013). The e e be a ion, BLR, o ime delays a e epo ed when- e e a ailable in he p esen a ion o he indi idual objec s o he sample (Appendix C). The BLR es ima ions a e compa ed in Figu e 10. The wo independen es ima ions yield simi- la esul s o mos sou ces o ou sample, wi hin a ac o o ∼3 (Ves e gaa d & Pe e son 2006;Neg e e e al. 2013;Panda 2022). Th ee es ima ions o he BLR adius using he physi- cal pa ame e s ob ained om ou spec al measu emen s yield esul s incompa ible wi h he e e be a ion mapping es ima ion: M k 335, M k 110m, and A k 564 (Fig. 10). This disc epancy a ises om he es ima ions e u ning a low-densi y solu ion, and hus likely o e es ima ing he ex en o he Hβ-emi ing egion. The Ciii] in ensi y o A k 564 migh be o e es ima ed because o he co ec ion needed o accoun o hea y abso p ion close o he line es - ame (Fig. A.1). Remo ing he Ciii] diagnos ic a ios mo es he da a poin o his sou ce, in ag eemen wi h he Hβ e e be a ion mapping es ima e. Howe e , in he o he wo cases, he in ensi y o Ciii] and he p esence o o he spec al ea u es sugges ed he p esence o low-densi y gas in he BLR, a A321, page 16 o 31 Flo is, A., e al.: A&A, 689, A321 (2024) Fig. 9. Figu es showing a co ela ion be ween a ious acc e ion pa ame e s wi h logZand RFeII. Poin s a e displayed using he same colo -coding as in Figu e 4, wi h hei e o ba s. When he co ela ion is conside ed signi ican ( P&0.7, co esponding o a con idence le el o ≈0.01), he co ela ion line is shown, wi h i s de ining pa ame e s in e cep (A) and slope (B) shown wi h hei espec i e e o s. Fo each co ela ion a e also shown he Pea son co ela ion coe icien , P, om he SLOPES p og am (Feigelson & Babu 1992) and he p obabili y, P, o a spu ious co ela ion. s a i ica ion o p ope ies ha may be consis en wi h he p es- ence o an ex ended BLR. This seems o be especially ue o M k 110, in which he edwa d asymme y o a ain VBC migh equi e an inne mos adius o ∼100 g a i a ional adii. Ne e heless, he bulk o he measu emen s con i m he physical condi ions ob ained om he es ima ion associa ed wi h CLOUDY pho oioniza ion simula ions, p o iding an independen measu emen o he BLR. 6.5. En ichmen mechanisms along he quasa main sequence The esul s ob ained om he es ima ion o me allici y in he BLR o ou sou ces sugges he p esence o di e en en ichmen mechanisms ac ing in he di e en objec s o ou sample, gi ing ise o he sola o subsola me allici y ha we obse e in Pop. A and B sou ces, and he highes me allici y ha we obse e in ex eme acc e o s. Ma eucci & Pado ani (1993) discussed he possibili y o me al en ichmen in quasa s and p edic ed ha he gas can each sola abundances in less han 108yea s a e s a o ma ion s a s o a Salpe e -like ini ial mass unc ion (IMF). The e a e hus many possible mechanisms ha explain he me al abundances obse ed in Pop. A and Pop. B quasa s. We ha e, howe e , obse ed a ema kable numbe o objec s displaying a supe sola me allici y, wi h he o ali y o xA sou ces exhibi ing Z&10Z. One o he p ima y p oposals o explain his phe- nomenon is he possibili y o s a o ma ion in he icini y o he ac i e nucleus, s a ing om he models de eloped by A ymowicz e al. (1993) and Collin & Zahn (1999). In his con ex , Wang e al. (2009) p edic ed he p esence o highly en iched BLRs in quasa s, wi h Zup o 100Z, sugges ing he possibili y o s a o ma ion in he icini y o he acc e- ion disk o he cen al nucleus as a means o chemical en ich- men . The SMBH could hen ap hese me als in he egion due o i s g a i a ional po en ial. A simila mechanism in ol es acc e ion-modi ied s a s (AMSs): a s ella popula ion cha ac e - ized by s a s acc e ing ma e ial om he disk and apidly a ain- ing high masses, subsequen ly exploding as supe no ae and u he en iching he su ounding ma e ial (Wang e al. 2021; Can iello e al. 2021;Wang e al. 2023b). In ou sample, I Zw 1 and PHL 1092 display blueshi s ha domina e he emis- sion o hei HILs, pa icula ly Ci . The p esence o s ong winds o igina ing om he high-acc e ion phase expe ienced by hese objec s p oduce ou lows capable o escaping he BLR and en iching he su ounding egions (Vie i e al. 2016,2018, M23). Fu he mo e, he e is a lack o edshi e olu ion in he me allici y o quasa s (Jua ez e al. 2009;Fan & Wu 2023, and e e ences he ein). Me allici y in he BLRs is highe han in he na ow-line egions, sugges ing ha he chemical en ich- men occu s in si u, nea he molecula o us (Collin & Zahn 1999). A op-hea y IMF can explain he obse ed me al abun- dances, which could p oduce a s ong me al en ichmen in a sho imescale (τ < 109y , e en hough acc e ion-modi ied s a s could li e longe han isola ed s a s; see Chen & Lin 2024). An addi ional line o e idence suppo ing an in si u en ich- men is p o ided by he s ong simila i y obse ed be ween he objec s in ou sample and hose ound in high- edshi samples (Bañados e al. 2016;D’Odo ico e al. 2023), since a high ed- shi he e is no ime o p oduce he ex eme amoun s o me als obse ed (Xu e al. 2018). A321, page 17 o 31 Flo is, A., e al.: A&A, 689, A321 (2024) Fig. 10. Top: sca e plo showing he compa ison o BLR es ima ed wi h e e be a ion mapping ( BLR,RM), and using CLOUDY pho oioniza- ion simula ions ( BLR,Φ), on a loga i hmic plane. Poin s a e displayed using he same colo -coding as in Figu e 4, wi h hei espec i e e o ba s. The dash-do ed line desc ibes he posi ion o objec s o which he wo independen de e mina ions coincide. The alues o BLR,Φ o objec s whose de e mina ions do no coincide wi h BLR,RM by a ma gin a e shown in he uppe igh quad an o he igu e (exp essed in pc), connec ed o hei ep esen a i e poin s h ough an a ow. An al e na i e esul o he de e mina ion o physical pa ame e s is shown o A k 564 wi h a squa e, ob ained by emo ing he diagnos ic a ios employing Ciii]. Bo om: plo showing he di e ence be ween he wo es ima- ions. The ho izon al dash-do ed line desc ibes he posi ion o objec s o which he wo independen de e mina ions coincide. The do ed line connec s he wo di e en de e mina ions o A k 564 desc ibed abo e. The simila i y be ween low-zquasa s acc e ing a a high a e and mos in e media e- o-high-zquasa s (Sulen ic e al. 2000a) has been in e p e ed as a selec ion e ec (Sulen ic e al. 2014). In lux-limi ed su eys, o a gi en mass, he only obse able objec s a e hose o a highe luminosi y o , equi alen ly, a highe Edding on a io. The e o e, he p esence o a high pe cen age o high acc e o s a high edshi should no be in e p e ed as hem being he only quasa s p esen a high edshi . Low-luminosi y quasa s obse ed a low edshi s, such as NGC 3783 in ou sample, a e as o ye unobse able a in e media e- o-high ed- shi s. I , as ou esul s sugges , Zis co ela ed wi h Lbol/LEdd, a popula ion o quasa s wi h me allici ies a ound he sola me al- lici y migh be p esen up o z≈4, as has indeed been ound om a small sample o deep obse a ions o e y ain quasa s (Sulen ic e al. 2014). The mo phological ypes o he objec s o he sam- ple we e e ie ed om he NED ex agalac ic da abase and o he wo ks (McKe nan e al. 2010;Kho unzhe e al. 2012; Kim e al. 2017,2021;Zhao e al. 2021). The sample is howe e oo small o each e en a en a i e conclusion. A eliable analysis o he nonionizing pa o he SED (especially adio, a -in a ed (FIR), and mid-in a ed (MIR) emission) ha may yield in o - ma ion on he ci cumnuclea o hos s a o ma ion p ope ies is de e ed o an e en ual wo k. 7. Conclusions We analyzed a sample o 14 Type-I AGNs dis ibu ed along he MS o quasa s. We ca ied ou a mul icomponen i ing, as was desc ibed in Sec ion 3, o measu e he in ensi y o majo emis- sion lines in he λ1300 – λ2100 and op ical wa eleng h anges, and con on ed he esul s om app oxima ely en diagnos ic a ios wi h hose ob ained h ough CLOUDY pho oioniza ion sim- ula ions, cons aining he physical pa ame e s ha desc ibe he BLRs o hese sou ces. The ou comes o ou analysis can be summa ized as ollows: – We iden i ied a s ong end in me allici y along he MS, exhibi ing an inc easing g adien o me allici y om Pop. B quasa s (wi h subsola o sola me allici y) o xA sou ces (wi h supe sola me allici y Z>10Z). The hypo hesis ega ding Zas one o he p incipal co ela es o he MS (Panda e al. 2019b;Ma ziani e al. 2024) is con i med. – We p oposed a “ ecipe” o es ima e he me al con en in he BLR, u ilizing app oxima ely en diagnos ic a ios sensi i e o Z,U, and nH. We ecommend employing he Ci /Hβ a io only in he case o UV and op ical spec a acqui ed simul a- neously, as i was iden i ied as a majo sou ce o e o in his wo k. – We con i med me allici y alues se e al dozen imes he sola me allici y, among sou ces classi ied as xA ha a e consis en on a e age wi h he alues epo ed o he highly acc e ing quasa s o ´ Sniegowska e al. (2021), Ga nica e al. (2022). – The highes -Zobjec s migh be signi ican ly a ec ed by he in oduc ion o mic o u bulence, a he le el expec ed o he line he mal b oadening (∼10km s−1). The main ca ea s conce ning he Zes ima es a e ela ed o he p ecision o he measu emen s (ideally, ela i e unce ain- ies should be less han ∼30%), he lack o con empo anei y o op ical and UV obse a ions, and he simpli ica ion in o- duced by he single zone app oxima ion in he pho oioniza ion modeling, e en i he sepa a ion be ween he BLR, VBLR, and ou lowing componen akes in o accoun majo physical di e - ences. In his espec , we plan o implemen a locally op imized emi ing cloud scheme (Baldwin e al. 1995). The applica ion o he scheme o highe zwould equi e pho oioniza ion compu a- ions employing SEDs app op ia e o sou ces a high luminosi y (Du as e al. 2020;K awczyk e al. 2013). Howe e , es ima ions o me allici y a e y high edshi s (e.g., Rojas-Ruiz e al. 2024, a z≈7.24) a e now becoming possible hanks o he combina- ion o nea -in a ed (NIR) and MIR obse a ions and may o e an impo an cons ain on he ea ly e olu ion o he SMBHs and hei hos galaxies (e.g., Ca aneo e al. 2009). Acknowledgemen s. We hank Pu Du, Jian-Min Wang and Chen Hu o p o- iding us wi h he en a i e alue o he ime-delay o Hβ o PG 1259+593. SP acknowledges he inancial suppo o he Conselho Nacional de Desen- ol imen o Cien í ico e Tecnológico (CNPq) Fellowships 300936/2023-0 and 301628/2024-6. The au ho s would like o hank D . Jaime Pe ea o his use ul commen s and o his help and gene ous alloca ion o compu ing ime on hype ca and idilico a IAA (CSIC). A.D.M. and A.d.O. acknowledge inan- cial suppo om he Spanish MCIU h ough p ojec PID2022-140871NB-C21 by “ERDF A way o making Eu ope” and he Se e o Ochoa g an CEX2021- 515001131-S unded by MCIN/AEI/10.13039/501100011033. This esea ch is A321, page 18 o 31 Flo is, A., e al.: A&A, 689, A321 (2024) based in pa on obse a ions made wi h he NASA/ESA Hubble Space Tele- scope ob ained om he Space Telescope Science Ins i u e, which is ope a ed by he Associa ion o Uni e si ies o Resea ch in As onomy, Inc., unde NASA con ac NAS 5-26555. These obse a ions a e associa ed wi h p og ams 2424, 3418, 4045, 4210, 5182, 5486, 6766, 6781, 9894, 12212, 12604, 12916, 13811, 13814, 14481, 15669 and 16301. Some o he da a p esen ed in his pape we e ob ained om he Mul imission A chi e a he Space Telescope Science Ins i u e (MAST). STScI is ope a ed by he Associa ion o Uni e si ies o Resea ch in As onomy, Inc., unde NASA con ac NAS5-26555. Re e ences Aoki, K., & Yoshida, M. 1999, ASP Con . Se ., 162, 385 Aoki, K., Kawaguchi, T., & Oh a, K. 2005, ApJ, 618, 601 A ymowicz, P., Lin, D. N. C., & Wample , E. J. 1993, ApJ, 409, 592 Bache , R., Ma ziani, P., Sulen ic, J. W., e al. 2004, ApJ, 617, 171 Bake , J. C., Huns ead, R. W., Kapahi, V. K., & Sub ahmanya, C. R. 1994, ASP Con . Se ., 54, 195 Baldi, R. D., Lao , A., Beha , E., e al. 2022, MNRAS, 510, 1043 Baldwin, J., Fe land, G., Ko is a, K., & Ve ne , D. 1995, ApJ, 455, L119 Bañados, E., Venemans, B. P., Deca li, R., e al. 2016, ApJS, 227, 11 Ben z, M. C., Denney, K. D., G ie , C. J., e al. 2013, ApJ, 767, 149 Ben z, M. C., Williams, P. R., S ee , R., e al. 2021, ApJ, 920, 112 Bo oson, T. 2005, AJ, 130, 381 Bo oson, T. A., & G een, R. F. 1992, ApJS, 80, 109 B uhweile , F., & Ve ne , E. 2008, ApJ, 675, 83 Buendia-Rios, T. M., Neg e e, C. A., Ma ziani, P., & Dul zin, D. 2023, A&A, 669, A135 Can iello, M., Je myn, A. S., & Lin, D. N. C. 2021, ApJ, 910, 94 Cape i, A., Lao , A., Baldi, R. D., Robinson, A., & Ma coni, A. 2021, MNRAS, 502, 5086 Ca delli, J. A., Clay on, G. C., & Ma his, J. S. 1989, ApJ, 345, 245 Ca aneo, A., Fabe , S. M., Binney, J., e al. 2009, Na u e, 460, 213 Chen, Y.-J., Liu, J.-R., Zhai, S., e al. 2023, MNRAS, 522, 3439 Chen, Y.-X., & Lin, D. N. C. 2024, ApJ, 967, 88 Chie i, A., & Limongi, M. 2013, ApJ, 764, 21 Choi, E., Some ille, R. S., Os ike , J. P., Hi schmann, M., & Naab, T. 2024, ApJ, 964, 54 Collin, S., & Zahn, J.-P. 1999, Ap&SS, 265, 501 Collin-Sou in, S., Dyson, J. E., McDowell, J. C., & Pe y, J. J. 1988, MNRAS, 232, 539 D’Agos ini, G. 2003, Bayesian Reasoning in Da a Analysis: A C i ical In oduc ion (Singapo e: Wo ld Scien i ic) Deca li, R., Do i, M., & T e es, A. 2011, MNRAS, 413, 39 Decon o-Machado, A., del Olmo O ozco, A., Ma ziani, P., Pe ea, J., & S i pe, G. M. 2023, A&A, 669, A83 Diamond-S anic, A. M., Fan, X., B and , W. N., e al. 2009, ApJ, 699, 782 D’Odo ico, V., Bañados, E., Becke , G. D., e al. 2023, MNRAS, 523, 1399 D’Ono io, M., Ma ziani, P., & Chiosi, C. 2021, F on . As on. Space Sci., 8, 157 Du, P., & Wang, J.-M. 2019, ApJ, 886, 42 Du, P., Hu, C., Lu, K.-X., e al. 2015, ApJ, 806, 22 Du, P., Wang, J.-M., Hu, C., e al. 2016, ApJ, 818, L14 Du, P., Zhang, Z.-X., Wang, K., e al. 2018, ApJ, 856, 6 Dumon , A. M., Collin, S., Pale ou, F., e al. 2003, A&A, 407, 13 Du as, F., Bongio no, A., Ricci, F., e al. 2020, A&A, 636, A73 Fan, X., & Wu, Q. 2023, ApJ, 944, 159 Feigelson, E. D., & Babu, G. J. 1992, ApJ, 397, 55 Fe land, G. J., Cha zikos, M., Guzmán, F., e al. 2017, Re . Mex. As on. As o is., 53, 385 Fe land, G. J., Done, C., Jin, C., Land , H., & Wa d, M. J. 2020, MNRAS, 494, 5917 Flo ez, J., Jogee, S., She man, S., e al. 2020, MNRAS, 497, 3273 Fö s e Sch eibe , N. M., & Wuy s, S. 2020, ARA&A, 58, 661 F aix-Bu ne , D., Ma ziani, P., D’Ono io, M., & Dul zin, D. 2017, F on . As on. Space Sci., 4, 1 Fu, S., B and , W. N., Zou, F., e al. 2022, ApJ, 934, 97 Ganci, V., Ma ziani, P., D’Ono io, M., e al. 2019, A&A, 630, A110 Ga cia-Rissmann, A., Rod íguez-A dila, A., Sigu , T. A. A., & P adhan, A. K. 2012, ApJ, 751, 7 Ga nica, K., Neg e e, C. A., Ma ziani, P., e al. 2022, A&A, 667, A105 Gaskell, C. M. 1985, ApJ, 291, 112 Gau , H., Gu, M., Ramya, S., & Guo, H. 2019, A&A, 631, A46 Ga ilo i´ c, N., Popo i´ c, L. ˇ C., & Kolla schny, W. 2007, in Black Holes om S a s o Galaxies - Ac oss he Range o Masses, eds. V. Ka as, & G. Ma , 238, 369 GRAVITY Collabo a ion (Dex e , J., e al.) 2020, A&A, 635, A92 G ie , C. J., Pe e son, B. M., Pogge, R. W., e al. 2012, ApJ, 744, L4 G upe, D., Komossa, S., & Gallo, L. 2018, ATel, 11815 Gup a, A., Ma hu , S., K ongold, Y., & Nicas o, F. 2013, ApJ, 772, 66 Hamann, F., & Fe land, G. 1999, ARA&A, 37, 487 Hamann, F., Ko is a, K. T., Fe land, G. J., Wa ne , C., & Baldwin, J. 2002, ApJ, 564, 592 Ha ison, C. M., Alexande , D. M., Mullaney, J. R., & Swinbank, A. M. 2014, MNRAS, 441, 3306 Hu, C., Li, S.-S., Yang, S., e al. 2021, ApJS, 253, 20 Huang, Y.-K., Hu, C., Zhao, Y.-L., e al. 2019, ApJ, 876, 102 Huang, J., Lin, D. N. C., & Shields, G. 2023, MNRAS, 525, 5702 Jä elä, E., Dahale, R., C epaldi, L., e al. 2022, A&A, 658, A12 Jiang, J., Pa ke , M. L., Fabian, A. C., e al. 2018, MNRAS, 477, 3711 Jiang, B.-W., Ma ziani, P., Sa i´ c, D., e al. 2021, MNRAS, 508, 79 Jua ez, Y., Maiolino, R., Mujica, R., e al. 2009, A&A, 494, L25 Ju ᡠno á, A., Cos an ini, E., Di Gesu, L., e al. 2024, A&A, 681, A101 Kelle mann, K. I., S amek, R., Schmid , M., Sha e , D. B., & G een, R. 1989, AJ, 98, 1195 Kelle mann, K. I., S amek, R. A., Schmid , M., G een, R. F., & Sha e , D. B. 1994, AJ, 108, 1163 Kho unzhe , G. A., Sazono , S. Y., Bu enin, R. A., & Tkachenko, A. Y. 2012, As on. Le ., 38, 475 Kim, M., Ho, L. C., Peng, C. Y., Ba h, A. J., & Im, M. 2017, ApJS, 232, 21 Kim, M., Ba h, A. J., Ho, L. C., & Son, S. 2021, ApJS, 256, 40 Kolla schny, W. 2004, in The In e play Among Black Holes, S a s and ISM in Galac ic Nuclei, eds. T. S o chi-Be gmann, L. C. Ho, & H. R. Schmi , 222, 105 Kolla schny, W., Bischo , K., Robinson, E. L., Welsh, W. F., & Hill, G. J. 2001, A&A, 379, 125 Komossa, S., Xu, D., Zhou, H., S o chi-Be gmann, T., & Bine e, L. 2008, ApJ, 680, 926 Komossa, S., G upe, D., Gallo, L. C., e al. 2020, A&A, 643, L7 Ko is a, K. T. 1992, ApJS, 79, 285 Ko is a, K. T., & Goad, M. R. 2000, ApJ, 536, 284 Ko is a, K., Fe land, G., & Baldwin, J. 1997, ApJ, 487, 555 K awczyk, C. M., Richa ds, G. T., Meh a, S. S., e al. 2013, ApJS, 206, 4 K iss, G. 1994, ASP Con . Se ., 61, 437 K iss, G. A., A a , N., Edmonds, D., e al. 2019, A&A, 623, A82 Kuehn, C. A., Baldwin, J. A., Pe e son, B. M., & Ko is a, K. T. 2008, ApJ, 673, 69 Lao , A., Fio e, F., El is, M., Wilkes, B. J., & McDowell, J. C. 1997a, ApJ, 477, 93 Lao , A., Jannuzi, B. T., G een, R. F., & Bo oson, T. A. 1997b, ApJ, 489, 656 Li, Y.-R., Wang, J.-M., Ho, L. C., Du, P., & Bai, J.-M. 2013, ApJ, 779, 110 Li, Y.-R., Wang, J.-M., Zhang, Z.-X., e al. 2019, ApJS, 241, 33 Li, S.-S., Yang, S., Yang, Z.-X., e al. 2021, ApJ, 920, 9 Li a, P., Kishimo o, M., Goosmann, R. W., e al. 2021, MNRAS, 507, 579 Loiacono, F., Deca li, R., Mignoli, M., e al. 2024, A&A, 685, A121 Lyu, J., Rieke, G. H., & Smi h, P. S. 2019, ApJ, 886, 33 Madau, P., & Dickinson, M. 2014, ARA&A, 52, 415 Maiolino, R., & Mannucci, F. 2019, A&A Re ., 27, 3 Ma inello, M., Rod íguez-A dila, A., Ma ziani, P., Sigu , A., & P adhan, A. 2020, MNRAS, 494, 4187 Ma ínez-Aldama, M. L., Dul zin, D., Ma ziani, P., e al. 2015, ApJS, 217, 3 Ma ínez-Aldama, M. L., del Olmo, A., Ma ziani, P., e al. 2018, A&A, 618, A179 Ma ínez-Aldama, M. L., Panda, S., Cze ny, B., e al. 2021, ApJ, 918, 29 Ma ziani, P. 2023, Symme y, 15, 1859 Ma ziani, P., & Sulen ic, J. W. 2014, MNRAS, 442, 1211 Ma ziani, P., Cal ani, M., & Sulen ic, J. W. 1992, ApJ, 393, 658 Ma ziani, P., Sulen ic, J. W., Zwi e , T., Dul zin-Hacyan, D., & Cal ani, M. 2001, ApJ, 558, 553 Ma ziani, P., Sulen ic, J. W., Zamano , R., e al. 2003a, ApJS, 145, 199 Ma ziani, P., Zamano , R. K., Sulen ic, J. W., & Cal ani, M. 2003b, MNRAS, 345, 1133 Ma ziani, P., Sulen ic, J. W., S i pe, G. M., Zam i , S., & Cal ani, M. 2009, A&A, 495, 83 Ma ziani, P., Sulen ic, J. W., Neg e e, C. A., e al. 2010, MNRAS, 409, 1033 Ma ziani, P., Sulen ic, J. W., Plauchu-F ayn, I., & del Olmo, A. 2013, ApJ, 764, 150 Ma ziani, P., Ma ínez Ca ballo, M. A., Sulen ic, J. W., e al. 2016, Ap&SS, 361, 29 Ma ziani, P., del Olmo, A., D’Ono io, M., e al. 2018a, P oceedings o Science, Vol. PoS(NLS1-2018), Re isi ing Na ow-lineSey e 1 Galaxies and Thei Place in he Uni e se. 9-13 Ap il 2018.Pado a Bo anical Ga den, I aly Online a h ps://pos.sissa.i /cgi-bin/ eade /con .cgi?con id=328 Ma ziani, P., Dul zin, D., Sulen ic, J. W., e al. 2018b, F on . As on. Space Sci., 5, 6 A321, page 19 o 31 Flo is, A., e al.: A&A, 689, A321 (2024) Ma ziani, P., del Olmo, A., Pe ea, J., D’Ono io, M., & Panda, S. 2020, A oms, 8, 94 Ma ziani, P., Be on, M., Panda, S., & Bon, E. 2021, Uni e se, 7, 484 Ma ziani, P., Bon, E., Bon, N., e al. 2022, As on. Nach ., 343 Ma ziani, P., Panda, S., Decon o Machado, A., & Del Olmo, A. 2023, Galaxies, 11, 52 Ma ziani, P., Flo is, A., Decon o-Machado, A., e al. 2024, Physics, 6, 216 Mas ose io, G., Ing am, A., & an de Klis, M. 2020, MNRAS, 498, 4971 Ma hews, W. G., & Fe land, G. J. 1987, ApJ, 323, 456 Ma suoka, Y., Kawa a, K., & Oyabu, S. 2008, ApJ, 673, 62 Ma eucci, F., & Pado ani, P. 1993, ApJ, 419, 485 McKe nan, B., Fo d, K. E. S., & Reynolds, C. S. 2010, MNRAS, 407, 2399 Mejía-Res epo, J. E., Li a, P., Ne ze , H., T akh enb o , B., & Capellupo, D. M. 2018, Na . As on., 2, 63 Mi chell, J. A. J., Wa d, M. J., Kynoch, D., e al. 2024, MNRAS, 529, 4824 Mülle , A., & Wold, M. 2006, A&A, 457, 485 Mu ay, N., Chiang, J., G ossman, S. A., & Voi , G. M. 1995, ApJ, 451, 498 Nagao, T., Maiolino, R., & Ma coni, A. 2006, A&A, 459, 85 Nagoshi, S., & Iwamu o, F. 2022, PASJ, 74, 1198 Neg e e, C. A., Dul zin, D., Ma ziani, P., & Sulen ic, J. W. 2012, ApJ, 757, 62 Neg e e, C. A., Dul zin, D., Ma ziani, P., & Sulen ic, J. W. 2013, ApJ, 771, 31 Neg e e, C. A., Dul zin, D., Ma ziani, P., e al. 2018, A&A, 620, A118 Ne ze , H. 2019, MNRAS, 488, 5185 Pagel, B. E. J., Edmunds, M. G., Blackwell, D. E., Chun, M. S., & Smi h, G. 1979, MNRAS, 189, 95 Panda, S. 2021a, PhD Thesis, Polish Academy o Sciences, Ins i u e o Physics, Poland Panda, S. 2021b, A&A, 650, A154 Panda, S. 2022, F on . As on. Space Sci., 9, 850409 Panda, S., & Ma ziani, P. 2023, F on . As on. Space Sci., 10, 1130103 Panda, S., & ´ Sniegowska, M. 2024, ApJS, 272, 13 Panda, S., Cze ny, B., Adhika i, T. P., e al. 2018, ApJ, 866, 115 Panda, S., Cze ny, B., Done, C., & Kubo a, A. 2019a, ApJ, 875, 133 Panda, S., Ma ziani, P., & Cze ny, B. 2019b, ApJ, 882, 79 Panda, S., Ma ínez-Aldama, M. L., Ma inello, M., e al. 2020, ApJ, 902, 76 Panda, S., Kozłowski, S., G omadzki, M., e al. 2024, ApJS, 272, 11 Pe e son, B. M., & Fe land, G. J. 1986, Na u e, 324, 345 Pe e son, B. M., Wande s, I., Be am, R., e al. 1998, ApJ, 501, 82 Pe e son, B. M., Fe a ese, L., Gilbe , K. M., e al. 2004, ApJ, 613, 682 Phillips, M. M. 1978, ApJS, 38, 187 P ess, W. H., Flanne y, B. P., & Teukolsky, S. A. 1986, Nume ical Recipes. The A o Scien i ic Compu ing (Camb idge: Camb idge Uni e si y P ess) P ess, W. H., Teukolsky, S. A., Ve e ling, W. T., & Flanne y, B. P. 1992, Nume ical Recipes in FORTRAN. The A o Scien i ic Compu ing (Camb idge: Camb idge Uni e si y P ess) Punsly, B., Ma ziani, P., Benne , V. N., Nagai, H., & Gu well, M. A. 2018, ApJ, 869, 143 Rojas-Ruiz, S., Mazzucchelli, C., Finkels ein, S. L., e al. 2024, ApJ, 967, 27 Sabhlok, S., W igh , S. A., Vayne , A., e al. 2024, ApJ, 964, 84 Shapo alo a, A. I., Popo i´ c, L. ˇ C., Bu enko , A. N., e al. 2012, ApJS, 202, 10 Shemme , O., & Liebe , S. 2015, ApJ, 805, 124 Shemme , O., T akh enb o , B., Ande son, S. F., e al. 2010, ApJ, 722, L152 Shen, Y., & Ho, L. C. 2014, Na u e, 513, 210 Shin, J., Woo, J.-H., Nagao, T., & Kim, S. C. 2013, ApJ, 763, 58 Sigu , T. A. A., & P adhan, A. K. 2003, ApJS, 145, 15 Snedden, S. A., & Gaskell, C. M. 2007, ApJ, 669, 126 ´ Sniegowska, M., Ma ziani, P., Cze ny, B., e al. 2021, ApJ, 910, 115 ´ Sniegowska, M., Panda, S., Cze ny, B., e al. 2023, A&A, 678, A63 Sulen ic, J. W., Ma ziani, P., & Dul zin-Hacyan, D. 2000a, ARA&A, 38, 521 Sulen ic, J. W., Ma ziani, P., Zwi e , T., Dul zin-Hacyan, D., & Cal ani, M. 2000b, ApJ, 545, L15 Sulen ic, J. W., Zwi e , T., Ma ziani, P., & Dul zin-Hacyan, D. 2000c, ApJ, 536, L5 Sulen ic, J. W., Ma ziani, P., Zamano , R., e al. 2002, ApJ, 566, L71 Sulen ic, J., Ma ziani, P., & Zam i , S. 2011, Bal . As on., 20, 427 Sulen ic, J. W., Ma ziani, P., del Olmo, A., e al. 2014, A&A, 570, A96 Sulen ic, J. W., del Olmo, A., Ma ziani, P., e al. 2017, A&A, 608, A122 Sun, J., & Shen, Y. 2015, ApJ, 804, L15 Temple, M. J., Fe land, G. J., Rankine, A. L., Cha zikos, M., & Hewe , P. C. 2021, MNRAS, 505, 3247 Tody, D. 1986, SPIE Con . Se ., 627, 733 Tody, D. 1993, ASP Con . Se ., 52, 173 U, V., Ba h, A. J., Vogle , H. A., e al. 2022, ApJ, 925, 52 Vanden Be k, D. E., Richa ds, G. T., Baue , A., e al. 2001, AJ, 122, 549 an G oningen, E. 1987, A&A, 186, 103 Ve ne , E. M., Ve ne , D. A., Ko is a, K. T., e al. 1999, ApJS, 120, 101 Ve ne , E., B uhweile , F., Ve ne , D., e al. 2004, ApJ, 611, 780 Ves e gaa d, M., & Pe e son, B. M. 2006, ApJ, 641, 689 Ves e gaa d, M., & Wilkes, B. J. 2001, ApJS, 134, 1 Vie i, G., Piconcelli, E., Fio e, F., e al. 2016, in Ac i e Galac ic Nuclei 12: A Mul i-Messenge . Pe spec i e, AGN12, 22 Vie i, G., Piconcelli, E., Fio e, F., e al. 2018, AGN13: Beau y and he Beas , 13, 42 Wandel, A. 2000, New As on. Re ., 44, 427 Wang, J.-M., Yan, C.-S., Li, Y.-R., e al. 2009, ApJ, 701, L7 Wang, J.-M., Liu, J.-R., Ho, L. C., & Du, P. 2021, ApJ, 911, L14 Wang, Y., Shang, Z., & B o he on, M. S. 2022, MNRAS, 514, 1595 Wang, A., An, T., Cheng, X., e al. 2023a, MNRAS, 518, 39 Wang, J.-M., Zhai, S., Li, Y.-R., e al. 2023b, ApJ, 954, 84 Wildy, C., Cze ny, B., & Panda, S. 2019, A&A, 632, A41 Woo, J.-H., Wang, S., Rakshi , S., e al. 2024, ApJ, 962, 67 Xu, F., Bian, F., Shen, Y., e al. 2018, MNRAS, 480, 345 Yang, J., Wang, F., Fan, X., e al. 2023, ApJ, 951, L5 Yu, L.-M., Bian, W.-H., Zhang, X.-G., e al. 2020, ApJ, 901, 133 Zamano , R., Ma ziani, P., Sulen ic, J. W., e al. 2002, ApJ, 576, L9 Zam i , S., Sulen ic, J. W., Ma ziani, P., & Dul zin, D. 2010, MNRAS, 403, 1759 Zhang, K., Dong, X.-B., Wang, T.-G., & Gaskell, C. M. 2011, ApJ, 737, 71 Zhao, Y., Ho, L. C., Shangguan, J., e al. 2021, ApJ, 911, 94 Zhou, Y., Jiang, P., Shi, X., & Hao, L. 2023, ApJ, 959, 47 Zu, Y., Kochanek, C. S., & Pe e son, B. M. 2011, ApJ, 735, 80 A321, page 20 o 31 Flo is, A., e al.: A&A, 689, A321 (2024) Appendix A: Fi spec a o he sample The spec al analysis o he 14 objec s o ou sample is shown in Figu es A.1-A.4. The spec al i o he Ciii] spec al ange o M k 110 is absen because he e a e no Ciii] da a o his objec . Fig. A.1. Spec al i s o ( om op o bo om) M k 335, M k 110, M k 509, and A k 564 in he wa eleng h anges conside ed, wi h speci ic luxes measu ed in uni s o 10−15 e g s−1cm−2Å−1and he wa eleng h measu ed in Å. A lowe box desc ibing he esiduals is associa ed wi h each spec al i , wi h he wa eleng h exp essed in uni s o adial eloci ies (km s−1). The o iginal spec um is plo ed wi h a black line and he modeled spec um is plo ed in magen a. Each addi ional line componen is plo ed wi h a di e en colou , acco ding o he legend. Dashed e ical lines ep esen he es - ame wa eleng hs o he main lines o he selec ed spec al ange. Le :λ1320-1450 Å ange. Middle le :λ1450-1700 Å ange. Middle igh :λ1800-2100 Å ange. Righ :λ4430-5300 Å ange. A321, page 21 o 31 Flo is, A., e al.: A&A, 689, A321 (2024) Fig. A.2. Same as in Figu e A.1 bu o M k 478, Ton 28, LB 2522, and LEDA 51016, espec i ely. A321, page 22 o 31 Flo is, A., e al.: A&A, 689, A321 (2024) Fig. A.3. Same as in Figu e A.1 bu o I Zw 1, PHL 1092, Fai all 9, and A k 120, espec i ely. Fig. A.4. Same as in Figu e A.1 bu o NGC 3783 A321, page 23 o 31 Flo is, A., e al.: A&A, 689, A321 (2024) Appendix B: Dis ibu ion o U,nH, and Z o each objec The dis ibu ion o he physical pa ame e s o U,nH, and Z o he 14 objec s in ou sample is shown in Figu es B.1-B.5. Fig. B.1. Two-dimensional pa ame e spaces illus a ing he physical condi ions o he gas in he BLR o he objec s. Each plo ep esen s he BC, wi h a g een do indica ing he minimum χ2compu ed be ween measu ed diagnos ic a ios and CLOUDY simula ions. The o ange egion co esponds o a 1σaccu acy. Blue and ed do s and shades ep esen he minimum χ2and associa ed egions o he BLUE and VBC, espec i ely. Le : Loga i hmic plane o nHand Z.Middle: Loga i hmic plane o nHand U.Righ : Loga i hmic plane o Uand Z. A321, page 24 o 31 Flo is, A., e al.: A&A, 689, A321 (2024) Fig. B.2. Con inued. A321, page 25 o 31