New Astronomy Reviews 47 (2003) 689–691www.elsevier.com/ locate/newastrev

O n the connection between radio jet morphology and accretiondisk state

*Thomas J. Maccarone , Annalisa CelottiAstrophysics Sector, Scuola Internazionale Superiore di Studi Avanzati, via Beirut, n. 2–4, Trieste 34014,Italy

Abstract

We present a unified model for the formation of jets in X-ray binaries and active galactic nuclei. It has been previouslyshown that the dividing line between FR I and FR II galaxies in luminosity occurs at the same luminosity as the dividing linebetween gas-pressure and radiation-pressure dominated disks. We combine past results showing that the scale heights for thecorona of ‘low state’ flows tend to be larger than those for radiation pressure dominated disks and that pair production shouldoccur when coronal photons enter a relativistic proto-jet to show that, for reasonable assumptions, the jets of FR I galaxieswill be dynamically dominated by pairs while the jets of FR II systems will be dynamically dominated by protons. 2003 Elsevier B.V. All rights reserved.

Keywords: Accretion discs; Galaxies: jets; X-rays: binaries

Relativistic jets have been seen from both stellar dominated while the jets of the high luminositymass and supermassive black holes. The jets from systems will be proton-dominated.the supermassive black holes show a dichotomy in Recently, it has been suggested that jets should betheir morphological properties which seems to corre- ejected only when two criteria are met: (1) thelate well with the accretion flow’s luminosity in accreting black hole is rapidly rotating; and (2) theEddington units (Ghisellini and Celotti, 2001). It accretion flow onto the black hole has a substantialseems likely that the jets from X-ray binaries show a thickness (Meier, 2001). The first criterion comessimilar morphological dependence on luminosity from the hypothesis that jets are powered by the[compare, e.g. the low luminosity jets observed in extraction of the considerable energy stored in theStirling et al. (2001)with the high luminosity jets rotation of the black hole (Blandford and Znajek,observed inCorbel et al. (2002)].In this proceed- 1977). The second comes from the fact that theings, we suggest a scheme for explaining how the Blandford–Znajek process is efficient only when theluminosity of the accretion flow can determine the poloidal magnetic fields are strong. The poloidalmorphology of the radio jets. We show that the hot field strengths are likely to vary strongly with theparts of the accretion disk are likely to have larger disk scale height (Livio et al., 1999). The low/hardscale heights in low luminosity systems than high states of X-ray binaries are thought to occur whenluminosity systems, and that as a consequence, the the accretion flow includes a hot and geometricallyjets of these low luminosity systems will be pair thick component (e.g.Shapiro et al., 1976; Haardt

and Maraschi, 1993). The very high states have beensuggested to be puffed up by radiation pressure,*Corresponding author. Fax:139-040-378-7528.

E-mail address: maccarone@ap.sissa.it(T.J. Maccarone). albeit perhaps only for some fraction of the time in a

1387-6473/03/$ – see front matter 2003 Elsevier B.V. All rights reserved.doi:10.1016/S1387-6473(03)00125-8

690 T.J. Maccarone, A. Celotti / New Astronomy Reviews 47 (2003) 689–691

limit cycle (e.g. Shakura and Sunyaev, 1976; Ab- 21n LH Gpairs jetS D]] ]] ] ]]]ramowicz et al., 1988; Chen et al., 1995; Szus- 510 V S DS D in R 10 0.1Lprotons SCH EDDzkiewicz and Miller, 2001). These disks will be‘slim’ rather than thick, with the distinction being if the jet becomes optically thick to pair productionthat a thick disk has a scale height approximately (Sikora and Madejski, 2000). This equation is de-equal to the radius, while a slim disk has a scale rived by assuming a pair-loading scheme where softheight less than, but a substantial fraction of the (| 200 keV)g-rays from the ‘corona’ are upscatteredradius. The transition luminosity between the low/ to | few MeV by the bulk motion of the electrons inhard state and the high/soft state is thought to be| the proto-jet. The | few MeV g-rays then pair5–10% of the Eddington luminosity for all black produce by interacting with additional| 200 keVhole masses, while the critical luminosity for radia-

g-rays from the corona. The pairs produced by these21 / 8tion pressure to dominate is about 0.33M LBH EDD interactions then allow for more coronal photons to(Shakura and Sunyaev, 1976; Chen et al., 1995; be upscattered, driving a pair cascade which reaches9Meier, 2001). For a typical AGN with a 10 M( completion when the proto-jet becomes opticallyblack hole, the transition luminosity above which the thick to the coronal photons. The result will be avery high state of the system sets in is about 0.02 weak pair cascade. Because the maximum energiesL , very close to the value found byGhisellini andEDD reached by the photons will be of order a few MeVCelotti (2001),so it is natural to suggest that FR I’s and there will be a quasi-thermal cutoff to the photoncome from the low state accretors and FR II’s come distribution, the pairs produced by this processfrom the very high state accretors. should be only mildly energetic, with typical ener-

Previous work has attempted to explain the mor- gies of about 1–2 MeV.phology differences between FR I and FR II galaxies The low/hard states of X-ray binaries are general-in terms of their environment—as FR I systems are ly modelled with geometrically thick accretionpredominantly seen in clusters and dense groups andflows—either due to magnetic coronae with largeFR II in loose groups or isolated galaxies (Bicknell, scale heights (e.g.di Matteo et al., 1999) or due to1995), in terms of differences in the central engines adiabatic accretion flows (e.g.Rees et al., 1982;properties (Baum et al., 1995), or in terms of Narayan and Yi, 1994). In such systems, we have adifferences in pair content (Reynolds et al., 1996). thick disk (i.e. H /R | 1, with R | 30R ) at ahot SCHThe work ofGhisellini and Celotti (2001),combined high temperature and a low luminosity. Above, wewith the finding that the critical accretion rate for the have assumed that the radiation field from the hotFR I / II transition is equal to the accretion rate at thick disk is strong enough that pair cascades go towhich accretion discs become dominated by radia- completion over the full scale height of the hot disk.tion pressure strengthens the argument that the This will not be true for very low luminosities, as theproperties of the central engines cause the FR I / II seed photon density far from the black hole will bedichotomy in some way. It is plausible, then, that the too low to drive pair cascades. By taking thecorrelation between environment and morphology is expression fromSikora and Madejski (2000)for theseen because the environment controls the accretionoptical depth of the proto-jet as seen by the corona,rate by regulating the gas supply to the accreting we find that the criterion for pair production cascadesblack hole. to occur is

Here, we show that because the very high statedisks are ‘slim’ and the low state disks are ‘thick’, Lcor

]]]H 515 R ,pairs SCHpair loading processes should be more effective in 0.01LEDDlow state disks than in very high state disks. Given amaximal scale height of the hot part of the accretion whereH is the scale height at which pairpairs

disk, H, a Schwarzschild radius of the central black production ceases andL is the luminosity of thecor

hole,R , an opening angle of the jet near its base, ‘corona’. Typical values for the coronal radius andSCH

V , a jet bulk Lorentz factor,G, and a jet luminosity, hence its maximal height are larger than the corre-I

L , the ratio of pairs to protons in a jet will be sponding values ofH , so the cutoff of the pairjet pairs

T.J. Maccarone, A. Celotti / New Astronomy Reviews 47 (2003) 689–691 691

cascade will always occur atH . This criterion morphology can be explained by FR I’s having apairs

yields higher fraction of their initial mass-energy in radia-tively efficient pairs rather than protons, because pairn LGpairs cor loading should be more efficient in FR I / low stateS D]] ] ]51500V .S Din 10 Lprotons jet objects.

We note that this mechanism fails to produce pairAt the boundary between the FR I and FR II24dominated jets below luminosities of about 10 ofstates,L /L |0.4 (Ghisellini and Celotti, 2001),cor jet

the Eddington rate, so further investigation is re-so n /n |600, giving a jet which is mildlypairs protonsquired to determine whether the model is correct.dynamically dominated by the leptons, since there

will be two leptons per ‘pair’ and the pairs will becreated by the cascade with typically about 1–2 MeV

A cknowledgementseach, while the protons are assumed to be cold. SinceL /L should remain roughly constant through thecor jet

We gratefully acknowledge useful conversationslow state,n /n should also remain roughlypairs protonswith John Miller regarding slim disks in very highconstant. In the very high state, the luminosity of thestate systems and Dave Meier for discussions regard-coronal X-rays/g-rays is always high enough for theing jet formation scenarios.pair cascade to go to completion. The disks, how-

ever, tend to be ‘slim’ rather than ‘thick’, withH /Ralways less than| 0.3. Theoretical modelling of the

R eferencesvery high state is still in its infancy, but the bestcurrent models suggest thatH should always be less

A bramowicz, M.A. et al., 1988. ApJ 332, 646.than about 5R and typically about 3R (e.g.SCH SCHB aum, S.A., Zirbel, E.L., O’Dea, C.P., 1995. ApJ 451, 88.Szuszkiewicz and Miller, 2001). Thus n /npairs protons B icknell, V G., 1995. ApJS 101, 29.

will be less than| 1000 for all luminosities, with B landford, R.D., Znajek, R.L., 1977. MNRAS 179, 433.values as low as|5–10 for L | L . We thus find C hen, X. et al., 1995. ApJL 443, 61L.EDD

that the jets produced in low/hard state black holes C orbel, S. et al., 2002. Science 298, 196.d i Matteo, T., Celotti, A., Fabian, A.C., 1999. MNRAS 304, 809.should be dynamically dominated by pairs while theH aardt, F., Maraschi, L., 1993. ApJ 413, 507.jets produced in very high state systems should beG hisellini, G., Celotti, A., 2001. A&A 379, L1.

dynamically dominated by protons. L ivio, M., Ogilvie, G.I., Pringle, J.E., 1999. ApJ 512, 100.We have shown that the observed dichotomies M eier, D.L., 2001. ApJL 548, 9L.

between FR I and FR II galaxies are consistent with N arayan, R., Yi, I., 1994. ApJL 428, 13L.R ees, M.J. et al., 1982. Nature 295, 17.FR I jets being produced by black holes accreting asR eynolds, C.S. et al., 1996. MNRAS 283, 873.‘low state’ objects and FR II jets being produced byS hakura, N.I., Sunyaev, R.A., 1976. MNRAS 175, 613.

black holes accreting as ‘very high state’ objects. S hapiro, S.L., Lightman, A.P., Eardley, D.M., 1976. ApJ 204, 187.The differences in total jet power are a straight- S ikora, M., Madejski, G., 2000. ApJ 534, 109.forward consequence of the higher accretion rates in S tirling, A.M. et al., 2001. MNRAS 327, 1273.

S zuszkiewicz, E., Miller, J.C., 2001. MNRAS 328, 36.the very high state than in the low state. The