Insights from the Galactic center Elisabeth A.C. Mills (San Jose State University Boston University) Take advantage of the opportunity to give the last talk to make this a bit more responsive to everything I have heard this week, and to share an assortment of thoughts on how the center of our own galaxy fits in to things.
d=8.0 kpc The Galactic Center M(H 2 ): ~4x10 7 M. SFR: 0.1 M /yr. Sgr B2 100 pc Sgr A Central Molecular Zone NOTE: This is not the Circumnuclear Disk. Start with an orders of magnitude trip into the center of our galaxy, or the accretion journey of gas. These are the properties. Beware of nomenclature confusion.
d=8.0 kpc The Galactic Center Sgr B2 100 pc Sgr A 7 pc Central Molecular Zone Sgr A East First thing we encounter is one of the most recent supernovae, estimated to be ~10,000 years old and interacting with the surrounding gas. Remember this because it will come up again.
The Galactic d=8.0 kpc Center Sgr B2 100 pc Sgr A 7 pc Torus Central Molecular Zone Sgr A East 3 pc Circumnuclear Disk THIS is the circumnuclear disk, or what we have been calling a Torus, as seen on similar scales in other galaxies. Estimated mass of a few 10^4 solar masses (Requena Torres et al. 2012). Last reservoir of molecular gas we will encounter on our way to the black hole.
The Galactic d=8.0 kpc Center Sgr B2 0.5 pc 100 pc Minispiral Sgr A Central Molecular Zone 7 pc 3 pc Sgr A East Circumnuclear Disk Inside of the CND there is estimated to be a few hundred solar masses of atomic gas (Jackson et al. 1993), and a few tens of solar masses of ionized gas in the minispiral. This is a radio image, so now can see synchrotron emission directly from the black hole.
The Galactic d=8.0 kpc Center 0.15 pc Central lightyear Sgr B2 0.5 pc 100 pc Minispiral Sgr A Central Molecular Zone 7 pc 3 pc Sgr A East Circumnuclear Disk The central light year- now primarily seeing stars of the nuclear cluster. Well inside of the sphere of influence of the black hole right now (the radius inside of which the mass contribution from the nuclear cluster and black hole are comparable), which is around the radius of the CND.
The Galactic d=8.0 kpc Center S stars 0.15 pc 0.03 pc Central lightyear Sgr B2 0.5 pc 100 pc Minispiral Sgr A Central Molecular Zone 7 pc 3 pc Sgr A East Circumnuclear Disk At last we reach the S stars. These are the young (B-type) stars famously used to determine the parameters of the central supermassive black hole, and are suggested to have formed in-situ in one of the most extreme environments imaginable: in an extremely dense gas disk less than a tenth of a parsec from that black hole (e.g., Hobbs & Nayakshin 2009). This means a few million years ago, there was *another* torus, and even gas within a tenth of a parsec may be making stars rather than feeding the black hole.
The Galactic d=8.0 kpc Center S stars 0.15 pc 0.03 pc Central lightyear Sgr B2 Event Horizon (simulation) 0.5 pc 100 pc 0.000001 pc Sgr A Central Molecular Zone Minispiral 7 pc 3 pc Sgr A East Circumnuclear Disk And of course there is yet further to go scales of current accretion onto our black hole (the stuff that absolutely ends up in the black hole + not stopped by star formation), which hopefully will be reached by the Event Horizon telescope. This is really the full range you need to look at in our Galaxy when you ask, how does the large scale connect to the small scale?
The Galactic d=8.0 kpc Center S stars 0.15 pc 0.03 pc Central lightyear Sgr B2 Event Horizon (simulation) Quintuplet 3-4 Myrs (Liermann+ 2012) Arches 2-3 Myrs 100 pc Fermi Bubbles: 4-9 Myrs (Bordoloi+ 2017, Miller+ 2016) 0.000001 pc Sgr A 0.5 pc Minispiral 7 pc Young Nuclear Cluster 2-6 Myrs (Lu+ 2013) Central Molecular Zone 3 pc Sgr A East Circumnuclear Disk That s the current mass budget. What is the history? We know of three massive clusters in the GC, of similar age. Note that this includes both stars directly surrounding the black hole (~1 pc) and a more distributed population in and out of the other two clusters (within 100 pc, e.g., Mauerhan et al. 2009,2010). Might suggest prior starburst activity. Supporting this idea, we know there was either starburst or AGN activity within the past 10 Myrs from the Fermi bubbles. Either a <10 Myr duration constant AGN activity, or momentum injection events 4/6 Myrs ago. So: we have an inactive black hole, a starburst that happened on sub-parsec scales < 10 Myrs ago apparently coincident with a starburst on hundreds of pc scales (and Fermi bubbles) Can we make this fit e.g., the picture of a duty cycle given in K. Alatalo s talk?
Central Molecular Zone The Galactic Center Synchrotron and ionized gas, Cold dust, Stars and hot dust S stars Stars 2.1 μm Central lightyear Stars & hot dust 3.8 μm Keck L-band Ghez et al. 2005 ApJ 635,1087 20 cm (Radio), 1.1 mm, 8 μm VLA, CSO-Bolocam, Spitzer-IRAC Yusef-Zadeh et al. 2004 ApJS 155,421 Bally et al. 2010 ApJ 721,137 Stolovy et al. 2006 JPhCS 54,176 Image credit: A. Ginsburg, NRAO Keck K-band Ghez et al. 2008 ApJ 689,1044 Sgr A East Synchrotron (Sgr A East) and ionized gas 6 cm (Radio) VLA Zhao et al. 2016 ApJ 817,171 Minispiral Ionized gas & synchrotron (Sgr A*) 1.3 cm (Radio) VLA Zhao et al. 2009 ApJ 699, 186 Circumnuclear Disk Hot dust 19.7, 31.5, 37.1 μm SOFIA- FORCAST Lau et al. 2013 ApJ 775, 37 (Image credits)
TAKE-HOME POINT #1: A (molecular) torus does not an AGN make Milky Way: > 10 4 MSUN in 3 pc. (Requena Torres+ 2012) NGC 1068: 10 5 MSUN in 10 pc (Garcia-Burillo+ 2016) Need to understand prevalence of tori in all types of galaxies While it is exciting to begin to characterize compact tori around AGN, really need to explain what is so different about them and our own Galaxy. What properties of the available mass reservoir predict whether a black hole is active. How ubiquitous are tori?
A quick aside: The nitrogen isotope ratio in the Galactic center might not be so confusing 14 N/ 15 N ~ 150 Adande + Ziurys 2012 H 13 CN 1-0 HC 15 N 1-0 Mills+ In prep Let s talk about what we know about our own torus. Have current ALMA data covering the circumnuclear disk in bands 3,6,7 for excitation, covering a wide range of species and isotopologues. Thanks to Christian s talk, I took a look at this and have a rough number for 14N/15N from our HCN data (making big assumptions, like that H13CN 1-0 is optically thin, and that 12C/13C~25). This is roughly consistent with recent measurements of the gradient.
Mills+ In prep HCN J=3-2 Feldmeier+ 2014 H2 1-0 Q(1) 2.4 μm But perhaps most surprising to me has been, not the ALMA data, but infrared data covering H2. The CND is supposed to be behind a curtain of dust (and maybe even providing part of the curtain), requiring mm/submm wavelengths to look at the (presumably) cool, dense gas in that structure, and yet here we are, directly seeing the H2 in the near infrared, and it is behaving a lot like the stuff we typically use as a proxy for this molecule.
HCN J=3-2 H2 1-0 Q(1) If the prior image didn t convince you that there is a nice coincidence here, check out the velocities.
Detect 13 purerotational transitions of H2 l,b 1 H2 1-0 Q(1) 2 3 3 ISO spectra Mills+ (ApJ, in referee) So now we want to explore the potential of H2 in nuclear environments as a really interesting direct probe of conditions that can be compared to measurements made with proxies. Have archival ISO spectra from 2.4 to 45 microns. Although spectra toward Sgr A* were published, the H2 lines were not previously analyzed for any of these positions.
Can fit these with 3 discrete temperature components: 500 K 1300 K > 2600 K 3 Mills+ (ApJ, in referee) See a range of quite hot temperatres. Hottest gas that is present is at least 2600 K.
Or- fit a power-law distribution of temperatures dn T n dt n = 2.83 A small power law index = flatter slope 3 = more gas at high temperatures Mills+ (ApJ, in referee) Since 3 components is arbitrary, we also constrain a continuous, power-law distribution of temperatures. Find a good fit for both Northern and Southern parts of the CND. What is interesting is how flat it is. We can also try to extrapolate this to measure a mass, but that is a story for another day, though I will say it does not do as bad a job as I thought it might do! And presumably the hotter the environment (hint hint, extreme extragalactic sources) the better you can do.
CND: n = 2.8 Other Galactic center clouds: n = 4.7-5.0 LIRGS, ULIRGS, LINERS, star forming galaxies: n = 3.8-6.4 (Togi et al. ApJ 2016) Stephan s Quintet: n = 4.5 (Appleton et al., ApJ Accepted) TAKE-HOME POINT #2: It doesn t have to be active to be extreme. CND has a larger fraction of hot gas than any other extragalactic source yet probed with H2. Mills+ (ApJ, in referee) But a main take-away for right now: This source really is unusually hot. Can compare it to other clouds in the central 300 pc, and the kiloparsec-scale centers of other galaxies, and there is much more hot gas here. Now, it is very likely that once we can start to look at other galaxies/nuclei on similarly small scales (JWST), we will see other comparable hot spots, but it is notable how different it is. Truly is an extreme environment.
What traces molecular mass in the Galactic center? Although H2 shows some interesting possibilities for tracing mass as well, in general we want something a little better-excited and less likely to be obscured. So what best traces the molecular mass in our own CMZ?
N H2 NH2 Sgr B2 Sgr A 150 pc Battersby et al., in prep. [CI] 2P1-3P0 13CO 1-0 CO 4-3 CO 7-6 Martin+ 2006 Compare H2 column derived from Herschel dust measurements to various molecules, look at how peaks compare. One prominent feature near Sgr A. Even in this less extreme environment, places where you may need CI. Here is that supernova remnant we saw (just 1, as opposed to the dozens in arp 299) and already it is wreaking havoc in the gas. Not clear if it is due to mechanical or CR destruction of CO, but the cloud interacting with the SNR is booming bright in CI.
N H2 NH2 Sgr B2 Sgr A 150 pc Battersby et al., in prep. [CI] 2P1-3P0 Tanaka+ 2011 13CO 1-0 CO 4-3 CO 7-6 Martin+ 2006 Compare H2 column derived from Herschel dust measurements to various molecules, look at how peaks compare. One prominent feature near Sgr A. Even in this less extreme environment, places where you may need CI. Here is that supernova remnant we saw (just 1, as opposed to the dozens in arp 299) and already it is wreaking havoc in the gas. Not clear if it is due to mechanical or CR destruction of CO, but the cloud interacting with the SNR is booming bright in CI.
How good are different column density tracers? N H2 150 pc Battersby et al., in prep. HCN 1-0 Jones et al. HCN HCN 1-0:H2 Sgr D M0.50+0.00 Sgr B2- core Dust Ridge M0.25+0.01 M0.07+0.04 M1.6-0.03 M0.83-0.18 M0.68-0.12 Sgr B2 - halo M-0.13-0.08 M0.11-0.08 M-0.02-0.07 Sgr C 25 20 15 10 5 Mills & Battersby 2017 So the supernova is kind of messing things up, but what is the best way to trace the rest of the gas that is present? Continue to ask what best matches the dust. Ask if any band 3 tracer has a brightness that well-matches the column. And Gao and Solomon aside that tracer is not HCN. In addition to enhancement in the supernova cloud, Sgr B2 is too faint, due to self-absorption. And H13CN is even worse- it s nearly a scatter plot. Intrinsic abundance of HCN seems to be changing significantly.
How good are different column density tracers? N H2 Battersby et al., in prep. HCN 1-0 5.5 5.0 4.5 4.0 HCN 150 pc 50 km/s Sgr B2 Jones et al. HCN HCN 1-0:H2 M1.6-0.03 Sgr D M0.83-0.18 M0.50+0.00 Sgr B2- core Dust Ridge M0.25+0.01 M0.07+0.04 M0.68-0.12 Sgr B2 - halo 25 20 15 10 5 Log ) 3.5 4.0 4.5 5.0 5.5 6.0 M-0.13-0.08 M0.11-0.08 M-0.02-0.07 Sgr C Mills & Battersby 2017 So the supernova is kind of messing things up, but what is the best way to trace the rest of the gas that is present? Continue to ask what best matches the dust. 2.4 Ask if any band 3 tracer has a brightness that well-matches the column. And Gao and Solomon aside that tracer is not HCN. In addition to enhancement in the supernova 2.2 cloud, Sgr B2 is too faint, due to self-absorption. And H13CN is even worse- it s nearly a scatter plot. Intrinsic abundance of HCN seems to be changing significantly. 4.0 4.5 5 4.5 4.0 3.5 3.0 2.5 Log C 2 H 2.0 4.0 4.5 5.0 5.5 6.0 2.9 2.8 2.7 2.6 2.5 2.4 2.3 2.2 2.1 13 CS 2.0 4.0 4.5 5.0 5.5 6.0 4.0 3.8 3.6 3.4 3.2 3.0 2.8 2.6 3.2 3.1 3.0 2.9 2.8 2.7 CH 3 CN H 13 CO 2.6 4.0 4.5 5 Log [
5.5 How good are different HCN column density tracers? N H2 Battersby et al., in prep. 4.0 HCN 1-0 Jones et al. HCN HCN 1-0:H2 Total Integrated Line Intensity (K km s 1 ) 5.0 4.5 3.5 4.0 4.5 5.0 5.5 6.0 5.0 Log ) 5.5 3.8 5.0 3.4 4.5 3.0 4.0 2.6 HCNH 13 CN M0.50+0.00 Sgr B2- core Dust Ridge M0.25+0.01 M0.07+0.04 150 pc 3.5 2.2 4.0 4.5 5.0 5.5 6.0 HNCO 3.8 3.8 4.5 Sgr D 3.0 4.0 3.6 3.6 M-0.13-0.08 Sgr C M1.6-0.03 3.4 3.4 2.8 M0.11-0.08 4.0 M0.83-0.18 M-0.02-0.07 M0.68-0.12 Sgr B2 - halo 3.2 3.2 3.5 2.6 3.0 3.0 3.5 Mills & Battersby 2017 2.8 25 20 15 10 5 2.8 2.4 3.0 3.0 2.6 2.6 2.2 So the supernova is kind of messing things up, but what is the best way to trace the rest of the gas that is present? Continue to ask 2.4 what best matches the dust. 2.4 Ask if any band 3 tracer has a brightness that well-matches the column. 2.5 And Gao and Solomon aside 2.5 that tracer is not HCN. In addition 2.2 to enhancement in the supernova 2.2 2.0 cloud, Sgr B2 is too faint, due to self-absorption. And H13CN is even 4.0worse- 4.5 it s 5.0nearly 5.5 a 6.0 scatter plot. 4.0Intrinsic 4.5 5.0 abundance 5.5 6.0 of HCN seems 4.0 4.5 to be 5.0 changing 5.5 6.0 significantly. 4.0 4.5 5 4.5 4.0 3.5 3.0 2.5 C 2 H 2.0 4.0 4.5 5.0 5.5 6.0 3.6 3.2 2.8 2.4 4.5 4.5 2.9 2.8 4.0 2.7 2.6 3.5 2.5 3.0 2.4 2.3 50 km/s Sgr B2 Log HC 3 N C 2 H 13 CS 2.5 2.2 2.1 2.0 4.0 4.5 5.0 5.5 6.0 4.5 4.0 3.5 3.0 2.5 4.0 4.5 5.0 5.5 6.0 4.0 2.9 3.2 2.8 3.1 2.7 2.6 3.0 2.5 2.9 2.4 2.3 2.8 2.2 2.7 2.1 SiO CH 3 CN 13 CS H 13 CO + 2.0 2.6 4.0 4.5 5.0 5.5 6.0 Log [cloud mass] 5.0 4.5 4.0 3.5 3.0 4.0 4.5 4.0 3.2 3.2 3.4 3.1 3.2 3.0 3.0 2.8 2.9 2.6 2.8 2.4 2.7 2.2 HNC CH 3 CN C HHN 13 CO 13 C 2.6 2.0 4.0 4.5 5 Log [
TAKE-HOME POINT #3: HCN has problems N H2 150 pc Battersby et al., in prep. HCN 1-0 Jones et al. HCN 1-0:H2 HCN 1-0 60 50 40 30 20 10 Mills & Battersby 2017 See more evidence of abundance variations. For example, all of the diffuse HCN outside of the typically studied cloud cores (a region, which according to the column density, contributes around half of the mass here) has proportionately brighter HCN than the clouds do. This molecule is not a straightforward tracer.
N H2 150 pc Battersby et al., in prep. So what traces the dust column density best? HNCO 404-303 Jones et al. (see also: Henshaw et al. 2016) What is a good tracer? The best molecule we found in that has been surveyed in the GC in Band 3 is HNCO. And it is NICE. Everywhere you see strong dust emission/ large column you see nearly proportionally bright HNCO. And this is coming from a relatively complex, heavy, shock-tracing molecule, and not even its ground transition.
N H2 150 pc Battersby et al., in prep. So what traces the dust column density best? HNCO 404-303 Jones et al. (see also: Henshaw et al. 2016) Mills & Battersby 2017 This should concern you. What is a good tracer? The best molecule we found in that has been surveyed in the GC in Band 3 is HNCO. And it is NICE. Everywhere you see strong dust emission/ large column you see nearly proportionally bright HNCO. And this is coming from a relatively complex, heavy, shock-tracing molecule, and not even its ground transition.
TAKE-HOME POINT #4: 300 μm HNCO can be radiatively excited by far-ir emission. (Churchwell et al. 1986) 100 μm We should probably all be more careful with HNCO. B-type transitions connecting the different K ladders. Excited by FIR, and quickly decay back to the K=0 ladder. This is not a new idea: 30 years ago, radiative excitation of HNCO was measured to be important for Sgr B2. So if you are in an intense radiation environment, worth a thought as to what the spatial distribution of HNCO might really telling you, in addition to tracing shocks.
Our Galactic center N H2 150 pc All of this so far is trying to draw conclusions from a sample size of 1. What happens when we start being able to expand our sample size of resolved galaxy centers?
Which properties can be predicted? May understand what sets the size of a central molecular zone, (100s of parsecs in size) (Krumholz+ 17) Unaware of a theoretical framework that sets the expected size/ lifetime of gas structures on the size scales of tori Which of the global (~100s pc) properties are predictive? Does gas mass alone imply the magnitude of a starburst? What sets the presence of a molecular outflow? On what scales are gas properties linked with an AGN? I don t know of theoretical framework that sets the expected size of a torus, based on dynamical arguments, but please correct me if I am wrong!