Suppressed star formation in circumnuclear regions in Seyfert galaxies
aa r X i v : . [ a s t r o - ph ] A p r T HE A STROPHYSICAL J OURNAL L ETTERS : IN P RESS
Preprint typeset using L A TEX style emulateapj v. 11/12/01
SUPPRESSED STAR FORMATION IN CIRCUMNUCLEAR REGIONS IN SEYFERT GALAXIES J IAN -M IN W ANG , Y AN -M EI C HEN , C
HANG -S HUO Y AN , C HEN H U AND W EI -H AO B IAN
Key Laboratory for Particle Astrophysics, Institute of High Energy Physics, CAS, 19B Yuquan Road, Beijing 100049, China
Received 2007 January 17; accepted 2007 March 30
ABSTRACTFeedback from black hole activity is widely believed to play a key role in regulating star formation and blackhole growth. A long-standing issue is the relation between the star formation and fueling the supermassive blackholes in active galactic nuclei (AGNs). We compile a sample of 57 Seyfert galaxies to tackle this issue. Weestimate the surface densities of gas and star formation rates in circumnuclear regions (CNRs). Comparing withthe well-known Kennicutt-Schmidt (K-S) law, we find that the star formation rates in CNRs of most Seyfertgalaxies are suppressed in this sample. Feedback is suggested to explain the suppressed star formation rates.
Subject headings: galaxies: active — galaxies: Seyfert — galaxies: feedback INTRODUCTION
The implications of the well-known relations between blackhole masses and bulge magnitudes (Magorrian et al. 1998), andthe velocity dispersions (Gebhardt et al. 2000; Frarreasse &Merrit 2000) show a coevolution of the black holes and theirhost galaxies. However, how do black holes know the evolu-tion stage of the galaxies and how to control the growth of theblack holes are currently understood via the the feedback fromthe black hole (Silk & Rees 1998; Croton et al. 2006; Schaw-inski et al. 2006). Numerical simulations show two roles offeedback from the black hole activity: (1) modulating the starformation rates; (2) heating the medium and finally quenchingthe black hole activity (Di Matteo et al. 2005). The direct evi-dence for the presence of the feedback from active black holeshas to be shown from observations, yet.The main goal of the present paper is to show one piece ofevidence for the feedback role in active galaxies. We showthe AGN feedback domain, where starburst should be sup-pressed. We find the star formation rates in Seyfert galaxiesis significantly lower than the rates predicted by the Kennicut-Schmidt’s law. We use the cosmological parameters H =75km s - Mpc - , Ω M = 0 . Ω Λ = 0 . AGN FEEDBACK DOMAIN
When the CNR medium is optically thick, namely, the opti-cal depth τ = κ abs Σ gas ≥
1, where κ abs is opacity and Σ gas thegas surface density, the radiation from the black hole activitywill continuously heat the medium and blow the gas away so asto lower the star formation rates. The condition of τ = 1 yieldsa critical density Σ c gas = 9 . × (cid:0) κ abs / (cid:1) - M ⊙ pc - , (1) κ abs has a mean value of 5 for the CNR medium (Semenov etal. 2003). This is feedback driven by AGN radiation. We noteoutflows from Seyfert active nucleus have much low kineticluminosities, typically ∼ - (3 - L Bol based on X-ray warmerabsorbers (Blustin et al. 2005), where L Bol is the bolometricluminosity. Feedback from outflows could be thus neglected.When Σ gas > Σ c gas , the AGN radiation-driven feedback will sup-press the star formation. On the other hand, AGN feedbackreaches its maximum when an AGN radiates at the Eddingtonlimit L AGN = L Edd = 1 . × ( M • / M ⊙ )erg/s. In the case of L Edd ≤ L IRSFR , AGN have inefficient feedback to star formation. With the help of SFR = 4 . (cid:0) L IR / erg s - (cid:1) M ⊙ yr - , Σ c gas isgiven by using the K-S law ˙ Σ SFR = A Σ γ gas (Kennicutt 1998a), Σ c gas = 8 . × M . R - . M ⊙ pc - , (2)where ˙ Σ SFR = SFR /π R is the surface density of the star forma-tion rate, A = 2 . × - , γ = 1 . M = M • / M ⊙ is the blackhole mass and R = R / Σ gas ≥ Σ c gas , the gas is so dense that theluminosity from star formation dominates over the AGN. Wecall Σ c gas ≤ Σ gas ≤ Σ c gas , (3)the AGN feedback domain as shown in Fig. 1, in which the K-Slaw is broken.The strong radiation pressure from the black hole accretiondisk at Eddington limit is P AGN ≈ . × - M R - dyn cm - ,where M = M • / M ⊙ . The pressure from supernovae explo-sion is P SN ≈ ǫ ˙ Σ SFR c = 2 . × - ˙ Σ SFR , ǫ - dyn cm - , where ˙ Σ SFR , = ˙ Σ SFR / M ⊙ yr - kpc - and ǫ - = ǫ/ - is the ef-ficiency converting the mass into radiation (Thompson et al2005). We find P AGN ≥ P SN within CNRs of radius ∼ ǫ , M • and ˙ Σ SFR . Thisindicates that the radiation from AGN dominates over the localfeedback from supernovae explosion. After an AGN switcheson, the star formation is suppressed and then feedback fromsupernovae is further weakened. The timescale of the AGNfeedback to the starburst regions can be estimated by t FB ∼ E gas / f FB C L AGN , where L AGN is AGN luminosity, C = ∆Ω / π is the covering factor, the thermal energy E gas ≈ kT M gas / m p , k is the Boltzmann constant, m p is the proton mass, T is thegas temperature, M gas = π R Σ gas is the gas mass and f FB is thefeedback efficiency. We have t FB ∼ . × f - , - R T Σ gas , C - . L - yr , (4)where Σ gas , = Σ gas / M ⊙ pc - , T = T / × K, f FB , - = f FB / - , L = L AGN / erg s - and C . = C / . ABLE HE S EYFERT G ALAXY S AMPLESeyfert 1Object Redshift FWHM log λ L λ Ref. log M • ˙ M • log Σ gas S PAH R log ˙ Σ SFR (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)3C120 0.033 ... 44.17 2 7.74 a a - Table 1 is published in its entirety in the electronic edition. A portion is shown here for guidance regarding its formand content. a the blackhole mass are directly taken from Peterson et al. (2004). b refers to [O III ] FWHM. c based on M • - M bulge relation, F01475-0740: M bulge = - .
80; NGC 3660: M bulge = - . β for Seyfert 1s or stellar velocity dispersion σ for Seyfert 2s (in km s - );(4) luminosity of 5100Å deduced from extrapolation of F ν ∝ ν - . or [O III ] λ - ); (5) references for columns(3) and (4) are given below, respectively; (6) black hole mass (in M ⊙ ); (7) accretion rate (in M ⊙ yr - ); (8) gas surface density(in M ⊙ pc - ); (9) surface brightness of the 3.3 µ m PAH emission feature (in unit of 10 ergs s - kpc - ); (10) the scale of thestarburst regions (in kpc); (11) and (12) are the lower ( ˙ Σ LSFR ) and upper ( ˙ Σ USFR ) limits of surface density of star formationrates, respectively (in M ⊙ yr - kpc - ).Reference.-(1) NED; (2) Peterson et al. (2004); (3) Spinogilio et al. (1995); (4) Doroshenko & Terebezh (1979); (5) Kinneyet al. (1993); (6) Nelson & Whittle (1995); (7) Dahari & Robertis (1988); (8) Lipari et al. (1991); (9) Corral et al. (2005);(10) Kirhakos & Steiner (1990); (11) Visvanathan & Griersmith (1977); (12) Cid Fernandes et al. (2004); (13) Gu & Huang(2002); (14) Kailey & Lebofsky (1988); (15) Heraudeau & Simien (1998); (16) Bassani (1999); (17) Whittle et al. (1988); (18)Whittle (1992); (19) Garcia-Rissmann et al. (2005); (20) Crenshaw et al. (2003); (21) Marzini et al. (2003); (22) Veron-Cettyet al. (2001); (23) Postman & Lauer (1995). a universal rule of cosmic star formation, we may get the un-dergoing physics in the CNRs.We have to point out here that the short feedback time doesNOT mean the same timescale of the starburst. The present t FB means the starburst rates will be suppressed once AGN istriggered and make it possible for AGN and starburst coexist. APPEARANCE OF FEEDBACK IN SEYFERT GALAXIES
For the goal to test the above scenario, we compile 57 Seyfertgalaxies (Imanishi 2002; Imanishi 2003; Imanishi & Wada2004). The star formation rates in CNRs of Seyfert galaxiescan be traced by several indicators, particularly, PAH featuresat 3.3, 6.2, 7.7, 8.6 and 11.2 µ m, which radiate from vibration ofPAH grains containing about 50 carbon atoms. Among the fea-tures, 3.3 µ m emission is intrinsically strong and less affectedby broad silicate dust absorption (Imanishi 2002). We choose3.3 µ m emission as an indicator of the star formation rate. Weconvert the PAH emission into IR luminosity via L IR = 10 L PAH relation with a scatter by a factor of 2-3 for pure star formation(Imanishi 2002). Since some PAH grains would be destroyedby EUV and X-ray photons from the central engine, we havethe lower limit of the surface density of the star formation rates ˙ Σ LSFR = 35 . L PAH , R - ( M ⊙ yr - kpc - ) , (5)by using the relation of the star formation rate and the in-frared luminosity (eq. 7) (Kennicutt 1998a), where L PAH , = L PAH / erg s - . On the other hand, the infrared emissionfrom Seyfert galaxies covers the contribution from starburst andreprocessing radiation from AGNs, we have the upper limit ofthe surface density of the star formation rates ˙ Σ USFR = 35 . L FIR , R - ( M ⊙ yr - kpc - ) , (6)where L FIR , = L FIR / erg s - is the observed far-IR luminos-ity. We take the geometric average ˙ Σ SFR = (cid:16) ˙ Σ LSFR ˙ Σ USFR (cid:17) / andthe error bars correspond to ˙ Σ LSFR and ˙ Σ USFR . We have to stress this average only represents the central value of logarithm of ˙ Σ USFR and ˙ Σ LSFR and the upper and lower limits of ˙ Σ SFR are themost important. Table 1 gives the sample of Seyfert galaxies,which have been observed through IRTF SpeX or Subaru IRCSwith spatial resolution of 0 . ′′ - . ′′ .For Seyfert 1 galaxies, we estimate L Bol = 9 L , where L is the luminosity at 5100Å and then the accretion rate ˙ M • = L Bol /η c , where η = 0 . Σ tot = 2 . × α - / . ˙ M / • , f - / • M / R - / M ⊙ pc - , (7)given by the disk model (King et al. 2002; Yi & Black-man 1994; Tan 2005), where the opacity κ abs = 5 in this re-gion, f • is the ratio of the black hole mass to the total, ˙ M • , = ˙ M • / . M ⊙ yr - and α . = α/ . Σ gas = f g Σ tot Σ gas = 1 . × f g , . α - / . ˙ M / • , f - / • M / R - / M ⊙ pc - , (8)where f g , . = f g / .
05 is the gas fraction to the to-tal. Considering the disk is located inside thebulge, we have f g = M gas / M disk > M gas / M Bulge = (cid:0) M gas / M dust (cid:1) (cid:0) M dust / M • (cid:1) (cid:0) M • / M Bulge (cid:1) , where M disk is the to-tal mass of the disk, M gas / M dust is the gas-to-dust mass ra-tio and M Bulge ≈ M • is the bulge mass (Kormendy &Gebhardt 2001; McLure & Dunlop 2002). It has beenfound that the dust mass in PG quasars is comparablewith in Seyfert galaxies (Spinoglio et al. 2002; Haas etuppressed Starburst in Seyfert Galaxies 3 F IG . 1.— The plot of gas and star formation rate surface densities. The yellow region is the AGN feedback domain given by Σ c gas ≤ Σ gas ≤ Σ c gas . The Comptonthick region has Σ gas ≥ . × M ⊙ pc - (i.e. N H ≥ cm - ). The red squares are starburst galaxies taken from Kennicutt (1998b). The cyan and blue-magentastars are Seyfert 1 and 2 galaxies, respectively. The blue star is NGC 3227, in which the star formation rate is 0.05 M ⊙ yr - and the gas mass M gas = (2 - × M ⊙ within 65pc taken from Davies et al. (2006). al. 2003). The mean value of gas-to-dust mass ratio is h M gas / M dust i ∼
250 (Haas et al. 2003). We estimated dustmass from M dust ∼ . f µ D (cid:2) exp(143 . / T dust ) - (cid:3) M ⊙ ,and the dust temperature is estimated by T dust = (1 + z ) (cid:2) . - / ln(0 . f µ / f µ ) (cid:3) K, where f µ and f µ are thefluxes at 100 µ m and 60 µ m in unit of Jy, respectively, D L is theluminosity distance in unit of Mpc (Evans et al. 2005). We findthe mean value of h M dust / M • i = 0 . ± . f g ≥ .
05 as a lower limit in this paper. Thompsonet al. (2005) used f g = 0 .
1. We note Σ gas ∝ f - / • , resulting inuncertainties of Σ gas by a factor of 4 for f • = 10 - - α = 0 . L Bol = 3500 L [O III] with a mean uncertainty of 0.38 dex (Heckman et al. 2004),where L [O III] is the [O III ] λ M • - σ relation (Tremaine et al. 2002), where thedispersion velocity σ = FWHM([O III ]) / .
35 if the dispersionvelocity is not available.Fig. 1 shows the Σ gas - ˙ Σ SFR plot of Seyfert CNRs. Wefind that CNR gas surface densities of Seyfert galaxies are lo-cated within the AGN feedback domain. There are clearly threebranches in the figure, separating the Seyfert galaxies, when Σ c gas > Σ gas > Σ c gas . Seyfert galaxies marked in Zone I still sat-isfy the K-S law. Those (Mrk 273, Mrk 938, NGC 5135 andNGC 1068) marked in Zone II are located between the K-Slaw and Zone III. These are ultra-luminous infrared galaxies,or mixed with strong starbursts. The main energy sources inthe CNRs are in a transition state from a starburst to an AGNin these galaxies. The fraction of the transiting galaxies is only4 / ∼ /
10. Though the completeness of the present sample isuncertain, this fraction implies that the transition is quite short and indicated by the feedback timescale from equation (4). TheSeyfert galaxies in Zone III are undergoing suppressed star for-mation strongly, being 1-2 orders lower than that predicted bythe K-S law. The suppressed ˙ Σ SFR is obviously caused by thefeedback. Galaxies obeying the K-S law are powered by nu-clear energy from stars, however gravitational energy releasedfrom accretion onto the black holes is powering AGNs if a tran-sition from starburts to active galaxies happens. With the dis-sipation of CNR gas due to star formation and accretion ontothe black holes, Σ gas is decreasing and the galaxies may returnto the K-S law once AGNs switch off. Such a behavior likesevolution of stellar energy sources in the Hertzprung-Russelldiagram.It has been found that black hole duty cycles follow the his-tory of star formation rate density (Wang et al. 2006). Theabove scenario then implies that both the black hole activitiesand starbursts are episodic (Davies et al. 2006). The multiplecycles of the black holes and starbursts make it impossible tomeasure the time delay between the two episodes. Howeverthe stellar synthesis may tell the star formation history and thengive the black hole activity history. CONCLUSIONS AND DISCUSSIONS
We find direct evidence for the feedback from active blackholes in Seyfert galaxies. Once a black hole is triggered, thefeedback will significantly suppress the starbursts within a quiteshort timescale of a few 10 years. The duty cycles of Seyfertgalaxies strongly indicate there is an efficient way to frequentlytrigger black holes and quench starbursts.The data presented in this paper are only lower limits of thegas densities. Future VLT (Very Large Telescope) and ALMA(Atacama Large Millimiter Array) measurements of the starformation rates and gas densities will finally identify roles ofthe feedback from the black hole activities. Wang et al.The helpful comments from the referee are acknowledged.The authors are grateful to R. C. Kennicutt, L. C. Ho, S. N.Zhang and X.-Y. Xia for useful discussions. We appreciatethe stimulating discussions among the members of IHEP (Insti- tute of High Energy Physics) AGN group. J.-M.W. thanks theNatural Science Foundation of China for support via NSFC-10325313 and 10521001, CAS key project via KJCX2-YW-T03. REFERENCESBassani, L., et al. 1999 ApJS, 121, 473Blustin, A. J., Page, M. J., Fuerst, S. V., Branduardi-Raymont, G., Ashton, C.E. 2005, A&A, 431, 111Cid Fernandes, R., et al. Rodrigues Lacerda, R.; Joguet, B. 2004 MNRAS, 355,273Crenshaw, D. M., Kraemer, S. B. & Gabel, J. R. 2003 ApJ, 126, 1690Croton, D. J., Springel, V., White, S. D. M., De Lucia, G., Frenk, C. S., Gao,L., Jenkins, A., Kauffmann, G., Navarro, J. F., Yoshida, N. 2006, MNRAS,365, 11Corral, A., Barcons, X., Carrera, F. J., Ceballos, M. T., Mateos, S. 2005 A&A,431, 97Dahari, O. & Robertis, M. M. D. 1988 ApJS, 67, 249Davies, R. I., et al. 2006, ApJ, 646, 754Doroshenko, V. T. & Terebezh, V. Yu. 1979 SvAL, 5, 305Di Matteo, T., Springel, V., Hernquist, L. 2005, Nature, 433, 604Evans, A. S., Mazzarella, J. M., Surace, J. A., Frayer, D. T., Iwasawa, K. &Sanders, D. B. 2005, ApJS, 159, 197Ferrarese, L. & Merritt, D., 2000, ApJ, 539, L9Garcia-Rissmann, A., Vega, L. R.; Asari, N. V.; Cid Fernandes, R.; Schmitt, H.;Gonzoulez Delgado, R. M.; Storchi-Bergmann, T. 2005 MNRAS, 359, 765Gebhardt, K. et al., 2000, ApJ, 543, L5Gu, Q. & Huang, J. 2002 ApJ, 579, 205Haas, M., et al. 2003, A&A, 402, 87Heckman, T. M. et al. 2004, ApJ, 613, 109Heraudeau, P. & Simien, F. 1998 A&AS, 133, 317Imanishi, M., 2002, ApJ, 569, 44Imanishi, M. 2003, ApJ, 599, 918Imanishi, M. & Wada, K. 2004, ApJ, 617, 214Kailey, W. F. & Lebofsky, M. J. 1988 ApJ, 326, 653Kaspi, S., Smith, P. S., Netzer, H., Maoz, D., Jannuzi, B. T., Giveon, U., 2000,ApJ, 533, 631Kauffmann, G., et al. 2003, MNRAS, 346, 1055Kennicutt, R. C., Jr. 1998a, ARA&A, 36, 189Kennicutt, R. C. Jr. 1998b, ApJ, 498, 541 King, A., Frank, J. & Raine, D. J., 2002, Accretion Power in Astrophysics,Cambridge University Press, p.90Kinney, A. L., Bohlin, R. C., Calzetti, D., Panagia, N., & Wyse, R. F. G. 1993ApJS, 86, 5Kirhakos, S. D. & Steiner, J. E. 1990 AJ, 99, 1722Kormendy, J. & Gebhardt, K., 2001, in The 20th Texas Symposiumon Relativistic Astrophysics, ed. H. Martel & J.C. Wheeler, AIP,(astro-ph/0105230)Krumholz, M. R. & McKee, C. F. 2005, ApJ, 630, 250Lipari, S., Bonatto, C. & Pastoriza, M. 1991 MNRAS, 253, 19Magorrian, J. et al., 1998, AJ, 115, 2285Marzini, P., Sulentic, J. W., Zamanov, R., Calvani, M. & Dultzin-Hacyan, D.,2003 ApJS, 145, 199McLure, R. J. & Dunlop, J. S. 2002, MNRAS, 331, 795Nelson, C. H. & Whittle, M. 1995 ApJS, 99, 67Peterson, B. M. et al. 2004, ApJ, 613, 682Postman, M., Lauer, T. R. 1995 ApJ, 440, 28Silk, J. & Rees, M. J. 1998, A&A, 331, L1Schawinski, K., et al. 2006, Nature, 442, 888Semenov, D., Henning, Th., Helling, Ch., Ilgner, M., Sedlmayr, E. 2003, A&A,410, 611Shakura, N. I. & Sunyaev, R. A. 1973, A&A, 24, 337Spinoglio, L., Andreani, P., Malkan, M. A. 2002, ApJ, 572, 105Spinoglio, L., Malkan, M. A., Rush, B., Carrasco, L. & Recillas-cruz, E. 1995ApJ, 453, 616Tan, J. C. & Blackman, E. G. 2005, MNRAS, 362, 983Thompson, T. A., Quataert, E., Murray, N. 2005, ApJ, 630, 167Tremaine, S. et al. 2002, ApJ, 574, 740Veron-Cetty, M.-P., Veron, P. & Gongalves, A.C. 2001 A&A, 372, 730Visvanathan, N., Griersmith, D. 1977 A&A, 59, 317Wang, J.-M., Chen, Y.-M. & Zhang, F. 2006, ApJ, 647, L17Whittle, M. 1992 ApJS, 79, 49Whittle, M., Pedlar, A., Meurs, E. J. A., Unger, S. W., Axon, D. J. & Ward, M.J. 1988 ApJ, 326, 125Yi, I., Field, G. B. & Blackman, E. G. 1994, ApJ, 432, L31 uppressed Starburst in Seyfert Galaxies 5
Table 1 The Seyfert Galaxy Sample
Seyfert 1Object Redshift FWHM log λ L λ Ref. log M • ˙ M • log Σ gas S PAH R log ˙ Σ SFR (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)3C120 0.033 ... 44.17 2 7.74 a a a a a a a a a σ log L [O III] Ref. log M • ˙ M • log Σ gas S PAH R log ˙ Σ SFR (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)F01475-0740 0.017 ... 41.69 13 7.55 c b b b b b b b b b c b Wang et al. a the blackhole mass are directly taken from Peterson et al. (2004). b refers to [O III ] FWHM. c based on M • - M bulge relation, F01475-0740: M bulge = - .
80; NGC 3660: M bulge = - . β for Seyfert 1s or stellar velocity dispersion σ for Seyfert 2s (in km s - );(4) luminosity of 5100Å deduced from extrapolation of F ν ∝ ν - . or [O III ] λ - ); (5) references for columns(3) and (4) are given below, respectively; (6) black hole mass (in M ⊙ ); (7) accretion rate (in M ⊙ yr - ); (8) gas surface density(in M ⊙ pc - ); (9) surface brightness of the 3.3 µ m PAH emission feature (in × ergs s - kpc - ); (10) the scale of the starburstregions (in kpc); (11) and (12) are the lower ( ˙ Σ LSFR ) and upper ( ˙ Σ USFR ) limits of surface density of star formation rates, respectively(in M ⊙ yr - kpc -2