1: % conf2esa.tex -- sample pages for CUP conference proceedings document class

2: % (based on v1.1 jfm2esam.tex)

3: % v0.1 released 20 September 2000 by Alison Woollatt

4: % Copyright (2000) Cambridge University Press

5: 

6: \NeedsTeXFormat{LaTeX2e}

7: 

8: \documentclass{cupconf}

9: \usepackage{epsfig}

10: \usepackage{subfigure}

11: % See if the author has AMS Euler fonts installed: If they have, attempt

12: % to use the 'upmath' package to provide upright math.

13: 

14:   \checkfont{eurm10}

15:   \iffontfound

16:     \IfFileExists{upmath.sty}

17:       {\typeout{^^JFound AMS Euler Roman fonts on the system,

18:                    using the 'upmath' package.^^J}%

19:        \usepackage{upmath}}

20:       {\typeout{^^JFound AMS Euler Roman fonts on the system, but you

21:                    dont seem to have the}%

22:        \typeout{'upmath' package installed. cupconf.cls can take advantage

23:                  of these fonts,^^Jif you use 'upmath' package.^^J}%

24:        \providecommand\upi{\pi}%

25:       }

26:   \else

27:     \providecommand\upi{\pi}%

28:   \fi

29: 

30: % See if the author has AMS symbol fonts installed: If they have, attempt

31: % to use the 'amssymb' package to provide the AMS symbol characters.

32: 

33:   \checkfont{msam10}

34:   \iffontfound

35:     \IfFileExists{amssymb.sty}

36:       {\typeout{^^JFound AMS Symbol fonts on the system, using the

37:                 'amssymb' package.^^J}%

38:        \usepackage{amssymb}%

39:        \let\le=\leqslant  \let\leq=\leqslant

40:        \let\ge=\geqslant  \let\geq=\geqslant

41:       }{}

42:   \fi

43: 

44: % See if the author has the AMS 'amsbsy' package installed: If they have,

45: % use it to provide better bold math support (with \boldsymbol).

46: 

47:   \IfFileExists{amsbsy.sty}

48:     {\typeout{^^JFound the 'amsbsy' package on the system, using it.^^J}%

49:      \usepackage{amsbsy}}

50:     {\providecommand\boldsymbol[1]{\mbox{\boldmath $##1$}}}

51: 

52: %%% Example macros (some are not used in this sample file) %%%

53: 

54: % For units of measure

55: \newcommand\dynpercm{\nobreak\mbox{$\;$dynes\,cm$^{-1}$}}

56: \newcommand\cmpermin{\nobreak\mbox{$\;$cm\,min$^{-1}$}}

57: 

58: % Various bold symbols

59: \providecommand\bnabla{\boldsymbol{\nabla}}

60: \providecommand\bcdot{\boldsymbol{\cdot}}

61: \newcommand\biS{\boldsymbol{S}}

62: \newcommand\etb{\boldsymbol{\eta}}

63: 

64: % For multiletter symbols

65: \newcommand\Real{\mbox{Re}} % cf plain TeX's \Re and Reynolds number

66: \newcommand\Imag{\mbox{Im}} % cf plain TeX's \Im

67: \newcommand\Rey{\mbox{\textit{Re}}}  % Reynolds number

68: \newcommand\Pran{\mbox{\textit{Pr}}} % Prandtl number, cf TeX's \Pr product

69: \newcommand\Pen{\mbox{\textit{Pe}}}  % Peclet number

70: \newcommand\Ai{\mbox{Ai}}            % Airy function

71: \newcommand\Bi{\mbox{Bi}}            % Airy function

72: 

73: % For sans serif characters:

74: % The following macros are setup in cupconf.cls for sans-serif fonts in text

75: % and math.

76: %

77: % \textsfi, \mathsfi   : sans-serif slanted

78: % \textsfb, \mathsfb   : sans-serif bold

79: % \textsfbi, \mathsfbi : sans-serif bold slanted (doesnt exist in CM fonts)

80: %

81: % For san-serif roman use \textsf and \mathsf as normal.

82: %

83: \newcommand\ssC{\mathsf{C}}    % for sans serif C

84: \newcommand\sfsP{\mathsfi{P}}  % for sans serif sloping P

85: \newcommand\slsQ{\mathsfbi{Q}} % for sans serif bold-sloping Q

86: 

87: % Hat position

88: \newcommand\hatp{\skew3\hat{p}}      % p with hat

89: \newcommand\hatR{\skew3\hat{R}}      % R with hat

90: \newcommand\hatRR{\skew3\hat{\hatR}} % R with 2 hats

91: \newcommand\doubletildesigma{\skew2\tilde{\skew2\tilde{\Sigma}}}

92: %       italic Sigma with double tilde

93: 

94: % array strut to make delimiters come out right size both ends

95: \newsavebox{\astrutbox}

96: \sbox{\astrutbox}{\rule[-5pt]{0pt}{20pt}}

97: \newcommand{\astrut}{\usebox{\astrutbox}}

98: 

99: \newcommand\gtrsimQ{\ensuremath{G_a(P,Q)}}

100: \newcommand\GsPQ{\ensuremath{G_s(P,Q)}}

101: \newcommand\p{\ensuremath{\partial}}

102: \newcommand\tti{\ensuremath{\rightarrow\infty}}

103: \newcommand\kgd{\ensuremath{k\gamma d}}

104: \newcommand\shalf{\ensuremath{{\scriptstyle\frac{1}{2}}}}

105: \newcommand\sh{\ensuremath{^{\shalf}}}

106: \newcommand\smh{\ensuremath{^{-\shalf}}}

107: \newcommand\squart{\ensuremath{{\textstyle\frac{1}{4}}}}

108: \newcommand\thalf{\ensuremath{{\textstyle\frac{1}{2}}}}

109: \newcommand\Gat{\ensuremath{\widetilde{G_a}}}

110: \newcommand\ttz{\ensuremath{\rightarrow 0}}

111: \newcommand\ndq{\ensuremath{\frac{\mbox{$\partial$}}{\mbox{$\partial$} n_q}}}

112: \newcommand\sumjm{\ensuremath{\sum_{j=1}^{M}}}

113: \newcommand\pvi{\ensuremath{\int_0^{\infty}%

114:   \mskip -33mu-\quad}}

115: 

116: \newcommand\etal{\mbox{\textit{et al.}}}

117: \newcommand\etc{etc.\ }

118: \newcommand\eg{e.g.\ }

119: 

120: \newtheorem{theorem}{Theorem}[section]

121: \newtheorem{lemma}{Lemma}

122: \newtheorem{corollary}{Corollary}

123: 

124: \title[Dynamics around supermassive black holes]

125: {Dynamics around supermassive black holes}

126: 

127: \author[A. Gualandris \& D. Merritt]{%

128: A\ls L\ls E\ls S\ls S\ls I\ls A\ls \ns G\ls U\ls A\ls L\ls A\ls

129: N\ls D\ls R\ls I\ls S\ls \ns \and\ns

130: D\ls A\ls V\ls I\ls D\ls \ns M\ls E\ls R\ls R\ls I\ls T\ls T\ls}

131: 

132: \affiliation{Department of Physics and Center for Computational 

133: Relativity and Gravitation, Rochester Institute of Technology,

134: Rochester, USA}

135: 

136: \pubyear{2007}

137: \volume{000}

138: \pagerange{000--000}

139: \date{?? and in revised form ??}

140: \setcounter{page}{000}

141: 

142: \begin{document}

143: 

144: \maketitle

145: 

146: \begin{abstract}

147: The dynamics of galactic nuclei reflects the presence of supermassive 

148: black holes (SBHs) in many ways.  

149: Single SBHs act as sinks, destroying a mass 

150: in stars equal to their own mass in roughly one relaxation time and 

151: forcing nuclei to expand. 

152: Formation of binary SBHs displaces a mass in stars roughly equal to

153: the binary mass, creating low-density cores and 

154: ejecting hyper-velocity stars.  

155: Gravitational radiation recoil can eject coalescing binary SBHs 

156: from nuclei, resulting in offset SBHs and lopsided cores.

157: We review recent work on these mechanisms and discuss the

158: observable consequences.

159: \end{abstract}

160: 

161: \firstsection % if your document starts with a section,

162:               % remove some space above using this command.

163: \section{Characteristic scales}

164: 

165: Supermassive black holes (SBHs) are ubiquitous components of

166: bright galaxies and many have been present

167: with roughly their current masses

168: ($\sim 10^9 M_\odot$) since very early times, as soon as 

169: $\sim 1$ Gyr after the Big Bang

170: (\cite[Fan {\it et~al.} 2003]{Fan:03};

171: \cite[Marconi {\it et~al.} 2004]{Marconi:04}).

172: A SBH strongly influences the motion of stars within a

173: distance $r_h$, the gravitational influence radius, where

174: \begin{equation}

175: r_h = {GM_\bullet\over \sigma_c^2};

176: \end{equation}

177: $M_\bullet$ is the SBH mass and $\sigma_c$ is the stellar

178: (1d) velocity dispersion in the core.

179: Using the tight empirical

180: correlation between $M_\bullet$ and $\sigma_c$:

181: \begin{equation}

182: \left({M_\bullet\over 10^8M_\odot}\right) = (1.66\pm 0.24) 

183: \left({\sigma_c\over 200\ {\rm km\ s}^{-1}}\right)^\alpha, \ \ \ 

184: \alpha = 4.86\pm 0.4

185: \label{eq:ms}

186: \end{equation}

187: (\cite[Ferrarese \& Ford 2005]{FF:05}),

188: this can be written

189: \begin{equation}

190: r_h \approx 18\ {\rm pc} \left({\sigma_c\over 200\ 

191: {\rm km\ s}^{-1}}\right)^{2.86} 

192: \approx 13\ {\rm pc} \left({M_\bullet\over 10^8M_\odot}\right)^{0.59}.

193: \label{eq:rhnew} 

194: \end{equation}

195: While the velocities of stars must increase -- by definition --

196: inside $r_h$, this radius is not

197: necessarily associated with any other observational marker.

198: Such is the case at the Galactic center, for instance, where the stellar

199: density exhibits no obvious feature at $r_h\approx 3$ pc.

200: However the most luminous elliptical galaxies always

201: have cores, regions near the center where the stellar density

202: is relatively low.

203: Core radii are of order $r_h$ in these galaxies,

204: and the stellar mass that was (apparently) removed in

205: creating the core is of order $M_\bullet$.

206: These facts suggest a connection between the cores and

207: the SBHs, and this idea has motivated much recent work,

208: reviewed here, on binary SBHs and on the consequences of displacing SBHs

209: temporarily or permanently from their central locations in galaxies.

210: 

211: \begin{figure}

212:   \vspace{0.5cm}

213:   \includegraphics[height=0.35\textheight]{fig_tr.ps}

214:   \caption{Relaxation time, measured at the SBH influence radius,

215:    in a sample of early-type galaxies (C\^ot\'e et al. 2004), {\it vs}.

216:    the central stellar velocity dispersion. Filled symbols are

217:    galaxies in which the SBH's influence radius is resolved; the

218:    star is the Milky Way bulge.

219:    (From Merritt, Mikkola \& Szell 2007).}\label{fig:trrh}

220: \end{figure}

221: 

222: An important time scale associated with galactic nuclei

223: (not just those containing SBHs) is the relaxation time, defined as 

224: the time for 

225: gravitational encounters between stars to establish a locally 

226: Maxwellian velocity distribution.

227: The nuclear relaxation time is

228: (\cite[Spitzer 1987]{Spitzer:87})

229: \begin{subequations}

230: \begin{eqnarray}

231: T_R &\approx& {0.34\sigma_c^3\over G^2 \rho_c m_\star \ln\Lambda} \\

232: &\approx& 1.2\times 10^{10}\ {\rm yr}

233: \left({\sigma_c\over 100\ {\rm km\ s}^{-1}}\right)^3

234: \left({\rho_c\over 10^5M_\odot {\rm pc}^{-3}}\right)^{-1} 

235: \left({m_\star\over M_\odot}\right)^{-1} 

236: \left({\ln\Lambda\over 15}\right)^{-1}

237: \label{eq:tr}

238: \end{eqnarray}

239: \end{subequations}

240: with $\rho_c$ the nuclear density 

241: and $\ln\Lambda$ the Coulomb logarithm.

242: Figure~\ref{fig:trrh} shows estimates of $T_R$

243: measured at $r_h$ in a sample of early-type galaxies,

244: assuming $m_\star = 1M_\odot$.

245: A least-squares fit to the points (shown as the dashed line

246: in the figure) gives

247: \begin{equation}

248: T_R(r_h) \approx 2.5\times 10^{13}\ {\rm yr}

249: \left({\sigma\over 200\ {\rm km\ s}^{-1}}\right)^{7.47}

250: \approx 9.6\times 10^{12}\ {\rm yr}

251: \left({M_\bullet\over 10^8 M_\odot}\right)^{1.54}.

252: \label{eq:combine}

253: \end{equation}

254: ``Collisional'' nuclei can be defined as those with

255: $T_R(r_h)\lesssim 10$ Gyr; figure~\ref{fig:trrh} shows

256: that such nuclei are uniquely associated with galaxies that 

257: are relatively faint,

258: as faint as or fainter than the Milky Way bulge,

259: which has $T_R(r_h)\approx 4\times 10^{10}$ yr.

260: Furthermore, relaxation-driven changes in the stellar distribution

261: around a SBH are generally confined to radii $\lesssim 10^{-1} r_h$,

262: making them all but unobservable in galaxies beyond the 

263: Local Group (T. Alexander, these proceedings).

264: But the relaxation time also fixes the rate of

265: gravitational scattering of stars into the central ``sink'' -- 

266: either a single or a binary SBH -- and this fact has important

267: consequences for nuclear evolution in low-luminosity galaxies, 

268: as discussed below.

269: 

270: \section{Core structure}\label{sec:corestructure}

271: 

272: The `core' of a galaxy can loosely be defined as the region

273: near the center where the density of starlight drops

274: significantly below what is expected based on an inward extrapolation

275: of the overall luminosity profile.

276: At large radii, the surface brightness profiles of early-type 

277: galaxies are well fit by the \cite{Sersic:68} model,

278: \begin{equation}

279: \ln I(R) = \ln I_e - b(n)\left[\left(R/R_e\right)^{1/n} - 1\right].

280: \label{eq:sersic}

281: \end{equation}

282: The quantity $b$ is normally chosen such that $R_e$ is the projected

283: radius containing one-half of the total light.

284: The shape of the profile is then determined by $n$;

285: $n= 4$ is the \cite{DeVauc:48} model, which is a good

286: representation of bright elliptical (E) galaxies

287: (\cite[Kormendy \& Djorgovski 1989]{DK:89}),

288: while $n=1$ is the exponential model, which approximates

289: the luminosity profiles of dwarf elliptical (dE) galaxies 

290: (\cite[Binggeli, Sandage \& Tarenghi 1984]{Binggeli:84}).

291: An alternative way to write (\ref{eq:sersic}) is

292: \begin{equation}

293: {d\ln I\over d\ln R} = -{b\over n}\left({R\over R_e}\right)^{1/n},

294: \label{eq:serslopes}

295: \end{equation}

296: i.e. the logarithmic slope varies as a power of the projected radius.

297: While there is no consensus on why the

298: S\'ersic model is such a good representation

299: of stellar spheroids, a possible hint comes from the

300: dark-matter halos produced in 

301: $N$-body simulations of hierarchical structure

302: formation, which are also well described by (\ref{eq:serslopes})

303: (\cite[Navarro {\it et~al.} 2004]{Navarro:04}),

304: suggesting that S\'ersic's model applies generally to systems that form

305: via dissipationless clustering

306: (\cite[Merritt {\it et~al.} 2005]{Universal:05}).

307: 

308: S\'ersic's model is known to accurately reproduce the luminosity

309: profiles in some galaxies over at least three decades

310: in radius (e.g. \cite[Graham {\it et~al.} 2003]{Graham:03}), 

311: but deviations often appear near the center.

312: Galaxies fainter than absolute magnitude 

313: $M_B\approx -19$ tend to have {\it higher} central surface brightness

314: than predicted by S\'ersic's model;

315: the structure of the central excess is typically unresolved but its

316: properties are consistent with those of a compact, 

317: intermediate-age star cluster

318: (\cite[Carollo, Stiavelli \& Mack 1998]{Carollo:98};

319: \cite[C\^ot\'e {\it et~al.} 2006]{ACS8};

320: \cite[Balcells {\it et~al.} 2007]{Balcells:07}).

321: Galaxies brighter than $M_B\approx -20$ have long been known

322: to exhibit central {\it deficits}

323: (e.g. \cite[Kormendy 1985a]{Kormendy:85a});

324: these have traditionally been called simply ``cores,'' 

325: perhaps because a flat central density profile was considered 

326: a priori most natural (\cite[Tremaine 1997]{Tremaine:97}).

327: 

328: For about two decades, it was widely  believed that

329: dE galaxies were distinct objects

330: from the more luminous E galaxies.

331: The dividing line between the two classes was put 

332: at absolute magnitude $M_B\approx -18$,

333: based partly on the presence of cores in bright galaxies,

334: and also on the relation between total luminosity and mean

335: surface brightness

336: (\cite[Kormendy 1985b]{Kormendy:85b}).

337: This view was challenged by \cite{Jerjen:97},

338: and in a compelling series of papers, A. Graham and

339: collaborators showed that -- aside from the cores --

340: early-type galaxies display a remarkable continuity of 

341: structural properties,

342: from $M_B\approx -13$ to $M_B\approx -22$

343: (\cite[Graham \& Guzman 2003]{GG:03}; 

344: \cite[Graham {\it et~al.} 2003]{Graham:03}; 

345: \cite[Trujillo {\it et~al.} 2004]{Trujillo:04}). 

346: 

347: \begin{figure}[t]

348: \begin{center}

349: \includegraphics[height=0.30\textheight]{fig_Graham.eps}

350: \includegraphics[height=0.30\textheight]{fig_mdefhist.ps}

351: \end{center}

352: \caption[]{{\it Left:} Surface brightness profile in the $R$ band

353: of NGC 3348, a ``core'' galaxy.

354: The dashed line is the best-fitting S\'ersic model;

355: the observed profile (points, and solid line) falls

356: below this inside of a break radius $r_b\approx 0''.35$.

357: (From Graham 2004.)

358: {\it Right:} Histogram of observed mass deficits for the

359: sample of core galaxies in Graham (2004) 

360: and Ferrarese {\it et al.} (2006).

361: (Adapted from Merritt 2006a.)}

362: \label{fig:mdef}

363: \end{figure}

364: 

365: The connection between nuclear star clusters and SBHs,

366: if any, is unclear; in fact it has been suggested that the two

367: are mutually exclusive 

368: (\cite[Ferrarese {\it et~al.} 2006]{Ferrarese:06};

369: \cite[Wehner \& Harris 2006]{Wehner:06}),

370: although counter-examples to this rule probably exist,

371: e.g. NGC 3384 which contains a nuclear cluster

372: (\cite[Ravindranath {\it et~al.} 2001]{Ravin:01})

373: and may contain a SBH

374: (\cite[Gebhardt {\it et~al.} 2003]{Gebhardt:03}).

375: 

376: Here we focus on the cores.

377: The cores extend outward to a break radius $r_b$ 

378: that is roughly a few times $r_h$,

379: or from $\sim 0.01$ to $\sim 0.05$ times $R_e$.

380: A more robust way of quantifying the cores is in

381: terms of their mass (i.e. light): the ``mass deficit''

382: (\cite[Milosavljevi{\'c} {\it et~al.} 2002]{Milos:02})

383: is defined as the difference in integrated mass 

384: between the observed density profile $\rho(r)$ and an

385: inward extrapolation of the outer profile,

386: $\rho_{out}(r)$, typically modelled as a S\'ersic profile

387: (figure~\ref{fig:mdef}):

388: \begin{equation}

389: M_{def}\equiv 4\pi \int_0^{r_b} 

390: \left[\rho_{out}(r) -\rho(r)\right] r^2 dr .

391: \label{eq:defmdef}

392: \end{equation}

393: Figure~\ref{fig:mdef} shows mass deficits for a sample of

394: ``core'' galaxies, expressed in units of the SBH mass.

395: There is a clear peak at $M_{\rm def}\approx 1 M_\bullet$,

396: although some galaxies have much larger cores.

397: 

398: The fact that core and SBH masses are often so similar 

399: suggests a connection between the two.

400: Ejection of stars by binary SBHs during galaxy mergers

401: is a natural model (\cite[Begelman, Blandford \& Rees 1980]{BBR:80});

402: the non-existence of cores in fainter galaxies could

403: then be due to regeneration of a steeper density profile

404: by star formation (e.g. \cite[McLaughlin {\it et.~al} 2006]{King:06})

405: or by dynamical evolution associated with the (relatively) short

406: relaxation times in faint galaxies 

407: (e.g. \cite[Merritt \& Szell 2006]{MS:06}).

408: However the largest cores are difficult to explain

409: via the binary model (\cite[Milosavljevi{\'c} \& Merritt 2001]{MM:01}).

410: 

411: \section{Massive binaries}\label{sec:massivebinaries}

412: 

413: A typical mass ratio for galaxy mergers in the local Universe

414: is $\sim 10:1$ (e.g. \cite[Sesana {\it et~al.} 2004]{Sesana:04}).

415: To a good approximation, the initial approach of the two

416: SBHs can therefore be modelled by assuming that the galaxy hosting

417: the smaller BH spirals inward under the influence 

418: of dynamical friction from the fixed distribution of

419: stars in the larger galaxy.

420: Modelling both galaxies as singular isothermal

421: spheres ($\rho\sim r^{-2}$) and assuming that the smaller

422: galaxy spirals in on a circular orbit, its tidally-truncated mass

423: is $m_2\approx \sigma_2^3r/2G\sigma$

424: where $\sigma_2$ and $\sigma$ are the velocity dispersion of the

425: small and large galaxy respectively

426: (\cite[Merritt 1984]{Merritt:84}).

427: Chandrasekhar's (1943) formula then gives for the orbital

428: decay rate and infall time 

429: \begin{equation}

430: {dr\over dt} = -0.30 {Gm_2\over\sigma r}\ln\Lambda,\ \ \ \ 

431: t_{infall} \approx 3.3 {r(0)\sigma^2\over \sigma_2^3}

432: \label{eq:df}

433: \end{equation}

434: where $\ln\Lambda$ has been set to 2.

435: Using (\ref{eq:ms})

436: to relate $\sigma$ and $\sigma_2$ to the respective

437: SBH masses $M_1$ and $M_2$, this becomes

438: \begin{equation}

439: t_{infall}\approx 3.3{r(0)\over\sigma} 

440: \left({M_2\over M_1}\right)^{-0.62},

441: \end{equation}

442: i.e. $t_{infall}$ exceeds the crossing time of the larger

443: galaxy by a factor $\sim q^{-0.6}$, $q\equiv M_2/M_1\le 1$.

444: Thus for mass ratios $q\gtrsim 10^{-3}$, 

445: infall requires less than $\sim 10^2 T_{cr}\approx 10^{10}$ yr.

446: This mass ratio is roughly the ratio between the masses of the

447: largest ($\sim 10^{9.5}M_\odot$) and smallest 

448: ($\sim 10^{6.5}M_\odot$) known SBHs and so it is reasonable

449: to assume that galaxy mergers will almost always lead to formation of

450: a binary SBH in a time less than $10$ Gyr.

451: This conclusion is strengthened if the effects of gas are 

452: taken into account (e.g. \cite[Mayer {\it et~al.} 2007]{Mayer:07}).

453: 

454: Equation (\ref{eq:df}) begins to break down when the

455: two SBHs approach more closely than $\sim r_h$,

456: the influence radius of the larger hole,

457: since the orbital energy of $M_2$ is absorbed by the stars, 

458: lowering their density and reducing the frictional force.

459: In spite of this slowdown, $N$-body integrations 

460: (\cite[Merritt \& Cruz 2001]{MC:01};

461: \cite[Merritt \& Milosavljevi{\'c} 2001]{MM:01};

462: \cite[Makino \& Funato 2004]{Makino:04};

463: \cite[Berczik {\it et~al.} 2005]{Berczik:05})

464: show that

465: the separation between the two SBHs continues to

466: drop rapidly until the binary semi-major axis

467: is $a\approx a_h$, where

468: \begin{subequations}

469: \begin{eqnarray}

470: a_h &\equiv& {G\mu\over 4\sigma^2} \approx {1\over 4}{q\over (1+q)^2}r_h \\

471: &\approx& 3.3 {\rm pc} {q\over (1+q)^2} \left({M_1+M_2\over 10^8 M_\odot}\right)^{0.59}

472: \end{eqnarray}

473: \label{eq:ah}

474: \end{subequations}

475: and $\mu\equiv M_1M_2/(M_1+M_2)$ is the binary reduced mass.

476: At this separation -- the ``hard binary'' separation --

477: the binary's binding energy per unit mass

478: is $\sim\sigma^2$ and it

479: ejects stars that pass within a distance $\sim a$

480: with velocities large enough to remove them from the nucleus 

481: (\cite[Mikkola \& Valtonen 1992]{MV:92};

482: \cite[Quinlan 1996]{Quinlan:96}).

483: 

484: What happens next depends on the density and geometry 

485: of the nucleus.

486: In a spherical or axisymmetric galaxy, 

487: the mass in stars on orbits that intersect the binary

488: is small, $\lesssim M_1+M_2$, and the 

489: binary rapidly ejects these stars;

490: no stars then remain to interact with the binary and its evolution

491: stalls (figure~\ref{fig:rbinoft}).

492: In non-axisymmetric (e.g. triaxial) nuclei,

493: the mass in stars on

494: centrophilic orbits can be much larger, allowing the binary

495: to continue shrinking past $a_h$.

496: And in collisional nuclei of any geometry, gravitational scattering

497: of stars can repopulate depleted orbits.

498: These different possbilities are discussed in more detail below.

499: 

500: \begin{figure}[t]

501: \vspace{-0.5cm}

502: \begin{center}

503: \includegraphics[height=0.60\textheight,angle=-90.]{fig_rbinoft.ps}

504: \end{center}

505: \caption[]{Early evolution of binary SBHs in $N$-body galaxies, 

506: for different values of the binary mass ratio,

507: $M_2/M_2=0.5,0.25,0.1,0.05,0.025$ (left to right).

508: The upper horizontal line indicates $r_h$,

509: the influence radius of the more massive hole.

510: Lower horizontal lines show $a_h$ (\ref{eq:ah}),

511: the ``hard-binary'' separation.

512: The evolution of the binary slows drastically when

513: $a\approx a_h$; in a real (spherical) galaxy with

514: much larger $N$, evolution would stall at this separation.

515: The smaller the infalling BH, the farther it spirals in before

516: stalling.

517: (Adapted from Merritt 2006a.)}

518: \label{fig:rbinoft}

519: \end{figure}

520: 

521: If the binary does stall at $a\approx a_h$,

522: it will have given up an energy

523: \begin{subequations}

524: \begin{eqnarray}

525: \Delta E &\approx& -{GM_1M_2\over 2r_h} + {GM_1M_2\over 2a_h} \\

526: &\approx& -{1\over 2}M_2\sigma^2 + 2(M_1+M_2)\sigma^2 \\

527: &\approx& 2 (M_1+M_2)\sigma^2

528: \end{eqnarray}

529: \end{subequations}

530: to the stars in the nucleus, i.e.,

531: the energy transferred from the binary to 

532: the stars is roughly proportional to the 

533: {\it combined} mass of the two SBHs.

534: The reason for this counter-intuitive 

535: result is the $a_h\propto M_2$

536: dependence of the stalling radius

537: (\ref{eq:ah}): smaller infalling BHs

538: form tighter binaries.

539: Detailed $N$-body simulations (\cite[Merritt 2006a]{Merritt:06}) 

540: verify that the mass deficit generated by the binary in evolving from

541: $\sim r_h$ to $\sim a_h$ is a weak function of the mass ratio,

542: \begin{equation}

543: M_{\rm def,h} \approx 0.4(M_1+M_2) \left({q\over 0.1}\right)^{0.2}

544: \label{eq:mdefh}

545: \end{equation}

546: for $0.025\lesssim q\lesssim 0.5$.

547: A mass deficit of $\sim 0.5 M_\bullet$ is still a factor 

548: $\sim 2$ too small to explain the observed peak

549: in the $M_{\rm def}/M_\bullet$ histogram

550: (figure~\ref{fig:mdef}).

551: On the other hand, bright elliptical galaxies 

552: have probably undergone numerous mergers,

553: and the proportionality between $M_{\rm def}$ and

554: $M_1+M_2$ (rather than, say, $M_2$) implies that 

555: the mass deficit following ${\cal N}$ mergers 

556: is $\sim 0.5{\cal N}$ times the {\it accumulated} BH mass.

557: Mass deficits in the range

558: $0.5\lesssim M_{\rm def}/M_\bullet\lesssim 1.5$

559: therefore imply $1\lesssim {\cal N}\lesssim 3$ mergers,

560: consistent with the number of major

561: mergers expected for bright galaxies since the epooch at which

562: most of the gas was depleted

563: (e.g. \cite[Haehnelt \& Kauffmann 2002]{HK:02}).

564: Hierarchical growth of cores tends to saturate after a few

565: mergers however 

566: making it difficult to explain mass deficits greater than

567: $\sim 2 M_\bullet$ in this way.

568: An effective way to enlarge cores still more is to kick

569: the SBH out, at least temporarily, as discussed in \S 4.

570: 

571: The first convincing evidence for a true, binary SBH 

572: was recently presented by \cite{Rodriguez:06},

573: who discovered two compact, flat-spectrum AGN 

574: at the center of a single elliptical galaxy,

575: with a projected separation of $\sim 7$ pc.

576: This is consistent with the radius at formation of a

577: $\sim 10^{7.5} M_\odot$ binary (\ref{eq:rhnew}), or the stalling

578: radius for a binary of at least $\sim 10^{9.3}M_\odot$ 

579: (\ref{eq:ah}).

580: Rodriguez {\it et~al.} estimate a binary mass of 

581: $\sim 10^8 M_\odot$ but with considerable uncertainty.

582: All other examples of ``binary'' SBHs in single galaxies

583: have separations $\gg r_h$ (\cite[Komossa 2006]{Komossa:06}).

584: 

585: \begin{figure}

586: \vspace{-1.0cm}

587: \centering

588: \includegraphics[height=0.65\textheight,angle=-90.]{fig_MMS6.eps}

589: \caption{Short-term {\it (a)} and long-term {\it (b)}

590: evolution of a massive binary in a series of $N$-body integrations.

591: Vertical axis is the inverse semi-major axis (i.e. energy) of the binary,

592: computed by averaging several independent $N$-body runs;

593: different curves correspond to different values of $N$, 

594: the number of ``star'' particles. 

595: The evolution of the binary is independent of $N$ until 

596: $a\approx a_h$ (horizontal line); thereafter

597: the evolution rate is limited by how quickly stars are scattered

598: onto orbits that intersect the binary, and decreases with

599: increasing $N$.

600: (From Merritt, Mikkola \& Szell 2007.)

601: \label{fig:MMS6}

602: }

603: \end{figure}

604: 

605: Even in a spherical galaxy, the stalling that occurs at 

606: $a\approx a_h$ can be avoided if stars continue to be scattered 

607: onto orbits that intersect the binary 

608: (\cite[Valtonen 1996]{Valtonen:96}; \cite[Yu 2002]{Yu:02};

609: \cite[Milosavljevi{\'c} \& Merritt 2003]{MM:03}).

610: Such ``collisional loss-cone repopulation'' requires that 

611: the two-body (star-star) relaxation time at $r\approx r_h$ 

612: be less then $\sim 10^{10}$ yr; 

613: according to (\ref{eq:combine}), this is the case in

614: galaxies with $M_\bullet\lesssim 10^6M_\odot$,

615: i.e. at the extreme low end of the SBH mass distribution.

616: %Even the Milky Way nucleus ($T_R(r_h)\approx 10^{10.5}$ yr) 

617: %is barely in this regime.

618: Collisional loss cone repopulation is therefore irrelevant to 

619: the luminous galaxies that are observed to have cores

620: but may be important in the mass range 

621: ($M_\bullet \lesssim 10^7 M_\odot$) of most interest

622: to space-based gravitational wave interferometers like LISA

623: (\cite[Hughes 2006]{Hughes:06}).

624: 

625: \begin{figure}

626:   \vspace{0.5cm}

627:   \includegraphics[height=0.36\textheight]{fig_MMS20.eps}

628:   \includegraphics[height=0.36\textheight]{fig_MMS19.eps}

629:   \caption{Evolution of a binary SBH in a collisional nucleus,

630:    based on a Fokker-Planck model that allows for evolution

631:    of the stellar distribution (Merritt, Mikkola \& Szell 2007).

632:    {\it Left:} Probability of finding the binary in a unit

633:    interval of $\ln a$.

634:    From left to right, curves are for 

635:    $M_1+M_2=(0.1,1,10,100)\times 10^6M_\odot$.

636:    Solid (dashed) curves are for $M_2/M_1 =1 (0.1)$.

637:    Open circles indicate when the rate of energy loss to

638:    stars equals the loss rate to gravitational waves; 

639:    filled circles

640:    correspond to an elapsed time since $a=a_h$ of $10^{10}$ yr.

641:    For the two smallest values of $M_\bullet$, the latter time

642:    occurs off the graph to the right.

643:    {\it Right:} Total time for a binary to evolve from $a=a_h$

644:    to gravitational wave coalescence as 

645:    a function of binary mass.

646:    The thick (black) curve is for $M_2/M_1=1$ and the 

647:    thin (blue) curve is for $M_2/M_1=0.1$.

648:    Dotted curves show the time spent in the gravitational

649:    radiation regime only.

650:    \label{fig:MMS1920}}

651: \end{figure}

652: 

653: $N$-body simulation would seem well suited to this problem

654: (e.g. \cite[Governato, Colpi \& Maraschi 1994]{Governator:94};

655: \cite[Makino 1997]{Makino:97};

656: \cite[Milosavljevi{\'c} \& Merritt 2001]{MM:01};

657: \cite[Makino \& Funato 2004]{Makino:04}).

658: The difficulty is noise -- or more precisely, getting the level

659: of noise just right.

660: In a real galaxy there is a clear separation of time scales

661: between an orbital period, and the time for stars to be scattered

662: onto depleted orbits: the first is typically much shorter than the second

663: which means that orbits intersecting the binary will remain empty 

664: for many periods before a new star is scattered in.

665: This is called the ``empty loss cone'' regime and it implies that

666: supply of stars to the binary will take place diffusively.

667: In an $N$-body simulation, however, $N$ is much smaller than its value

668: in real galaxies and orbits are repopulated too quickly.

669: This is the reason that binary evolution rates in $N$-body simulations

670: typically scale as $N^{-\alpha}, \alpha<1$ rather than the

671: $\sim T_R^{-1}\sim N^{-1}$ dependence expected if stars diffused gradually into

672: an empty loss cone (\cite[Merritt \& Milosavljevi{\'c} 2003]{MM:03}).

673: Figure~\ref{fig:MMS6} provides an illustration:

674: early evolution of the binary, until $a\approx a_h$, is $N$-independent;

675: formation of a hard binary then depletes the loss cone

676: and continued hardening occurs at a rate that

677: is a decreasing function of $N$, though less steep than $N^{-1}$.

678: 

679: An alternative approach is based on the Fokker-Planck equation.

680: Both single and binary SBHs can be modelled as 

681: ``sinks'' located at the centers of galaxies (\cite[Yu 2002]{Yu:02}).

682: The main differences are the larger physical extent

683: of the binary ($\sim G(M_1+M_2)/\sigma^2$ vs. $GM_\bullet/c^2$)

684: and the fact that the binary gives stars a finite kick

685: rather than disrupting or consuming them completely.

686: However the diffusion rate of stars into a central sink

687: varies only logarithmically with the size of the sink

688: (\cite[Lightman \& Shapiro 1978]{LS:78}),

689: and a hard binary ejects most stars well out of

690: the core with $V\gg\sigma$, so the analogy is fairly good.

691: The Fokker-Planck equation describing nuclei with sinks

692: is (\cite[Bahcall \& Wolf 1977]{BW:77})

693: \begin{equation}

694: {\partial N\over\partial t} = 4\pi^2 p(E) {\partial f\over \partial t}

695: = -{\partial F_E\over\partial E} - {\cal F}(E,t).

696: \label{eq:dndt}

697: \end{equation}

698: Here $N(E,t)=4\pi p(E)f(E,t)$ is the distribution of stellar energies,

699: $f(E,t)$ is the phase space density and

700: $p(E)$ is a phase-space volume element.

701: The first term on the RHS of (\ref{eq:dndt}) describes the response

702: of $f$ to the flux $F_E$ of stars in energy space due 

703: to encounters.

704: The second term, $-{\cal F}$, is the flux of stars into the

705: sink, which is dominated by scattering in angular momentum

706: (\cite[Frank \& Rees 1976]{FR:76}).

707: A proper treatment of the latter term requires a 2d (energy, angular

708: momentum) analysis, but a good approximation to ${\cal F}$

709: can be derived by assuming that the distribution of stars

710: has reached a quasi steady-state

711: near the loss cone boundary in phase space

712: (\cite[Cohn \& Kulsrud 1979]{CK:79}).

713: If the sink is a binary SBH, a second equation is needed

714: that relates the flux of stars into the loss cone to

715:  the rate of change of the binary's semi-major axis:

716: \begin{equation}

717: {d\over dt}\left({1\over a}\right) = {2\langle C\rangle\over a(M_1+M_2)}

718: \int {\cal F}(E,t) dE

719: \label{eq:hardb}

720: \end{equation}

721: with $\langle C\rangle\approx 1.25$ a dimensionless mean energy change for

722: stars that interact with the binary.

723: 

724: Both terms on the RHS of (\ref{eq:dndt}) imply changes in a time

725: $\sim T_R$.

726: The first term on its own implies evolution toward the Bahcall-Wolf (1976)

727: ``zero flux'' solution, $\rho\sim r^{-7/4}$.

728: The second term implies that a mass of order $\sim M_\bullet$

729: will be scattered into the sink

730: in a time of $T_R(r_h)$.

731: When the sink is a binary SBH, the binary responds by 

732: ejecting the incoming stars and shrinking,

733: according to (\ref{eq:hardb}).

734: As a result, changes in the structure of the nucleus on a relaxation

735: time scale

736: (e.g. growth of a core) are 

737: directly connected to changes in the binary semi-major axis.

738: 

739: Numerical solutions to (\ref{eq:dndt}), (\ref{eq:hardb}) 

740: (including also the effects of a changing gravitational potential)

741: have been presented by \cite{MMS:07}.

742: The solutions are well fit by

743: \begin{equation}

744: \ln\left({a_h\over a}\right) = -{B\over A} + \sqrt{{B^2\over A^2}

745: + {2\over A}{t\over T_R(r_h)}}

746: \label{eq:tdepend}

747: \end{equation}

748: where $t$ is defined as the time since the binary first became

749: hard ($a=a_h$), and the coefficients

750: $A\approx 0.016$, $B\approx 0.08$ depend weakly 

751: on the binary mass ratio.

752: Including the effect of energy lost to gravitational radiation:

753: \begin{equation}

754: {d\over dt}\left({1\over a}\right) = 

755: {d\over dt}\left({1\over a}\right)_{\rm stars} +  

756: {d\over dt}\left({1\over a}\right)_{\rm GR}

757: \end{equation}

758: allows one to compute the time to full coalescence,

759: $T_{\rm coal}$.

760: Figure~\ref{fig:MMS1920} shows $T_{\rm coal}$ (right panel),

761: and the time spent by the binary

762: in unit intervals of $\ln a$ prior to coalescence

763: (left panel), as functions of binary mass.

764: The time to coalescence is well fit by

765: \begin{subequations}

766: \begin{eqnarray}

767: Y &=& C_1 + C_2X+C_3X^2, \\

768: Y&\equiv&\log_{10}\left(T_{\rm coal}\over 10^{10}{\rm yr}\right), \\

769: X&\equiv&\log_{10}\left({M_1+M_2\over 10^6M_\odot}\right).

770: \end{eqnarray}

771: \label{eq:YvsX}

772: \end{subequations}

773: with

774: \begin{subequations}

775: \begin{eqnarray}

776: M_2/M_1=1:&&   C_1=-0.372,\ C_2=1.384,\ C_3=-0.025 \\

777: M_2/M_1=0.1:&& C_1=-0.478,\ C_2=1.357,\ C_3=-0.041.

778: \end{eqnarray}

779: \end{subequations}

780: Based on the figure, binary SBHs would be able to 

781: coalesce via interaction with stars alone in galaxies

782: with $M_\bullet \lesssim 2\times 10^6 M_\odot$.

783: For $M_1+M_2\gtrsim 10^7M_\odot$, evolution for 10 Gyr only

784: brings the binary separation slightly below $a_h$; in such

785: galaxies the most likely separation to find a massive binary

786: (in the absence of other sources of energy loss)

787: would  be near $a_h$.

788: 

789: The core continues to grow as the binary shrinks,

790: but the mass deficit is not related in a simple

791: way to  the mass in stars ``ejected'' by the binary

792: (e.g. \cite[Quinlan 1996]{Quinlan:96}).

793: Rather it results from a competition between

794: loss of stars to the binary, represented by $-{\cal F}(E,t)$, and 

795: the change in $N(E,t)$ due to diffusion of 

796: stars in energy, represented by

797: $-\partial F_E/\partial E$.

798: As the mass deficit increases, so do gradients

799: in $f$, which increases the flux of stars

800: toward the center and counteracts the drop in density.

801: In principle the two terms could balance,

802: but at some distance from the center the relaxation time

803: is so long that local $F_E(E)$ must drop below the

804: integrated loss term $\int_E^\infty {\cal F}(E) dE$ --

805: stars can not diffuse in fast enough to replace those

806: being lost to the binary and the density drops.

807: The Fokker-Planck solutions show that the mass deficit 

808: increases with binary binding energy as

809: \begin{equation}

810: M_{\rm def,c} \approx 1.7\left(M_1+M_2\right) \log_{10}\left(a_h/a\right)

811: \label{eq:mdefc}

812: \end{equation}

813: again with a weak dependence on $M_2/M_1$.

814: The mass deficit at the onset

815: of the gravitational radiation regime  is found to be

816: \begin{subequations}

817: \begin{eqnarray}

818: M_{\rm def,c} &\approx& (4.5,3.5,2.6,1.6)(M_1+M_2)\ \ (M_2/M_1=1) \\

819:             &\approx& (3.4,2.6,1.7,0.9)(M_1+M_2)\ \ (M_2/M_1=0.1)

820: \end{eqnarray}

821: \label{eq:mdefb}

822: \end{subequations}

823: for $M_1+M_2=(10^5,10^6,10^7,10^8)M_\odot$.

824: These values should be added to the mass deficits (\ref{eq:mdefh})

825: generated during formation of the binary when predicting

826: core sizes in real galaxies.

827: 

828: Are such mass deficits observed?

829: Only a handful of galaxies in the relevant mass range

830: ($M_{\rm gal} \lesssim 10^{10} M_\odot$) are near enough

831: that their cores could be resolved even if present;

832: of these, neither the Milky Way nor M32 exhibit cores.

833: Also, as noted above, many low-luminosity spheroids have

834: compact central excesses rather than cores.

835: These facts do not rule out the past existence 

836: of massive binaries in these galaxies however.

837: (1) Binary evolution might have been driven more by

838: gas dynamical torques than by ejection of stars;

839: gas content during the most recent major merger is believed 

840: to be a steep inverse function of galaxy luminosity

841:  (\cite[Kauffmann \& Haehnelt 2000]{Kauffmann:00}).

842: (2) Star formation can create a dense core after the two 

843: SBHs have coalesced 

844: (\cite[Mihos \& Hernquist 1994]{Mihos:94}).

845: (3) A two-body relaxation time short enough to bring

846: the two SBHs together would also allow a Bahcall-Wolf

847: cusp to be regenerated in a comparable time after the two SBHs combine,

848: tending to erase the core (\cite[Merritt \& Szell 2006]{MS:06}).

849: 

850: From the point of view of physicists hoping to detect

851: gravitational waves,

852: it is disappointing that this model only guarantees coalescence

853: at the extreme low end of the SBH mass distribution.

854: (Astronomers hoping to detect binary SBHs may take

855: the opposite point of view.)

856: Fortunately, there is no dearth of ideas for

857:  overcoming the ``final parsec problem'' and allowing

858: binary SBHs to merge efficiently, even in massive galaxies:

859: 

860: \begin{figure}

861:   \vspace{0.2cm}

862:   \includegraphics[height=0.35\textheight]{fig_rotate.ps}

863:   \caption{Efficient merger of binary SBHs in barred galaxies.

864:   Plots are based on $N$-body simulations (no gas) of

865:   equal-mass binaries at the centers of galaxy models,

866:   with and without rotation.

867:   {\it (a)} Spherical models.  The binary hardening rate

868:   declines with increasing $N$, as in figure~\ref{fig:MMS6},

869:   implying that the evolution

870:   would stall in the large-$N$ limit.

871:   {\it (b)} Binary evolution in a flattened, rotating version 

872:   of the same galaxy model.

873:   At $t\approx 10$, the rotating model forms a triaxial bar.

874:   Binary hardening rates in this model are essentially independent

875:   of $N$, indicating that the supply of stars to the binary

876:   is not limited by collisional loss-cone refilling as in 

877:   the spherical models.

878:   This is currently the only simulation that follows two SBHs 

879:   from kiloparsec to sub-parsec separations and that can be robustly

880:   scaled to real galaxies.

881:  (From Berczik {\it et al.} 2006.)}\label{fig:rotate}

882: \end{figure}

883: 

884: {\bf Non-axisymmetric geometries.} 

885: Real galaxies are not spherical nor even

886: axisymmetric; parsec-scale bars are relatively common

887: and departures from axisymmetry are often invoked 

888: to enhance fueling of AGN

889: (e.g. \cite[Shlosman {\it et~al.} 1990]{Shlosman:90}).

890: Orbits in a triaxial nucleus can be ``centrophilic,''

891: passing arbitrarily close to the center after a sufficiently long time

892: (Poon \& Merritt 2001, 2004).

893: This implies feeding rates for a central binary

894: that can approach  the ``full loss cone'' rate in 

895: spherical geometries, or 

896: \begin{equation}

897: {d\over dt} \left({1\over a}\right) 

898: \approx 2.5 F_{\rm c}{\sigma\over r_h^2}

899: \end{equation}

900: if the  fraction  $F_{\rm c}$ of centrophilic orbits is large

901: (\cite[Merritt \& Poon 2004]{Poon:04b}).

902: While $F_{\rm c}$ is impossible to know in any particular

903: galaxy, even small values imply much larger feeding rates

904: than in a diffusively-repopulated loss cone.

905: Figure~\ref{fig:rotate} shows results from $N$-body simulations

906: that support this idea.

907: 

908: {\bf Secondary slingshot.}

909: Stars ejected by a massive binary 

910: can interact with it again if they return to the nucleus

911: on nearly-radial orbits.

912: The total energy extracted from the binary via this 

913: ``secondary slingshot'' will be the sum

914: of the discret energy changes during the interactions.

915: \cite{MM:03} showed that a mass ${\cal M}_\star$ of stars

916: initially in the binary's loss cone causes the binary to evolve as

917: \begin{equation}

918: {1\over a} \approx {1\over a_h} + {4\over r_h}\ln\left(1+{t\over t_0}\right),

919: \ \ \ \ t_0 = {2\mu\sigma^2\over {\cal M}_\star \langle\Delta E\rangle} P(E)

920: \end{equation}

921: in the absence of diffusive loss cone repopulation,

922: where $\langle\Delta E\rangle$ is the specific energy change after one 

923: interaction with the binary, $E$ is the initial energy

924: and $P(E)$ is the orbital period.

925: The secondary slingshot runs its course after a few orbital periods.

926: \cite{Sesana:07} sharpened this analysis by carrying out

927: detailed three-body scattering experiments and recording the

928: precise changes in energy of stars as they underwent repeated

929: interactions with the binary.

930: They inferred modest ($\sim \times 2$) changes in $1/a$ due

931: to the secondary slingshot,

932: but their assumption of a $\rho\sim r^{-2}$ density profile

933: around the binary was probably over-optimistic; such steep

934: density profiles are never observed and even if present 

935: initially would be rapidly destroyed when the binary first formed.

936: 

937: {\bf Bound subsystems.}

938: As noted above, recent observational studies have greatly

939: increased the number of galaxies believed to harbor compact 

940: nuclear star clusters; inferred masses for the clusters are 

941: comparable with the mass

942: that would normally be associated with a SBH.

943: It is not yet clear whether these subsystems co-exist with

944: SBHs, but if they do, they could provide an extra

945: source of stars to interact with a massive binary.

946: \cite[Zier (2006, 2007)]{Zier:06,Zier:07} explored this idea, assuming

947: a steeply rising density profile around the binary,

948: $\rho \propto r^{-\gamma}$, at the time that its

949: separation first reached $\sim a_h$. Zier concluded that

950: a cluster having total mass $\sim M_1+M_2$, distributed

951: as a steep power law, $\gamma\gtrsim 2.5$, could

952: extract enough energy from the binary to allow gravitational

953: wave coalescence in less than $10$ Gyr.

954: $N$-body tests of this hypothesis are sorely needed;

955: as in the \cite{Sesana:07} study, Zier's approach did

956: not allow him to self-consistently follow the effect of

957: formation of the binary on the surrounding mass distribution.

958: 

959: {\bf Masive perturbers.}

960: In a nucleus containing a spectrum of masses,

961: the gravitational scattering rate is proportional to

962: \begin{equation}

963: \tilde{m} = {\int n(m) m^2 dm \over \int n(m) m dm}

964: \end{equation}

965: (e.g. \cite[Merritt 2004]{Merritt:04}).

966: \cite{Perets:07a} argued that ``massive perturbers''

967: near the center of the Milky Way -- massive stars,

968: star clusters, giant molecular clouds -- are

969: sufficiently numerous to dominate $\tilde m$, implying

970: potentially much higher rates of gravitational scattering into a 

971: central sink than in the case of solar-mass perturbers.

972: \cite{Perets:07b} extended this argument to

973: galaxies in general, emphasizing in particular the early

974: stages following a galactic merger, 

975: and concluded that collisional loss cone repopulation

976: would be sufficient to guarantee

977: coalescence of binary SBHs in less than 10 Gyr for

978: all but the most massive binaries.

979: As in the studies of \cite{Sesana:07} and \cite{Zier:07},

980: \cite{Perets:07b} optimistically assumed a steep ($\rho\propto r^{-2}$)

981: density profile around the binary, in spite of $N$-body studies

982: showing rapid destruction of the cusps.

983: Their arguments for massive perturbers in giant E galaxies

984: are also rather speculative.

985: 

986: {\bf Multiple SBHs.}

987: An extreme case of a ``massive perturber'' is a third SBH,

988: which might scatter stars into a central binary 

989: (\cite[Zhao {\it et~al.} 2002]{Zhao:02}),

990: or perturb the binary directly, driving the two SBHs into

991: an eccentric orbit and shortening the time scale for gravitational 

992: wave losses (\cite[Valtonen {\it et~al.} 1994]{Valtonen:94};

993: \cite[Makino \& Ebisuzaki 1994]{Makino:94};

994: \cite[Blaes {\it et~al.} 2002]{Blaes:02}; 

995: \cite[Volonteri {\it et~al.} 2003]{Volonteri:03};  

996: \cite[Iwazawa {\it et~al.} 2006]{Iwasawa:06};

997: \cite[Hoffman \& Loeb 2007]{Hoffman:07}).

998: The likelihood of multiple-SBH systems forming is probably

999: highest in the brightest E galaxies since massive binaries

1000: are most likely to stall (low stellar density, little gas) 

1001: and since large galaxies experience the most frequent mergers.

1002: Here again, more $N$-body simulations, including post-Newtonian

1003: terms, are needed; among other

1004: dynamical effects that could then be self-consistently included

1005: are changes in core structure, and BH-core oscillations like those

1006: described in the next section.

1007: 

1008: {\bf Gas.}

1009:  The same galaxy mergers that create binary SBHs can 

1010: also drive gas into the nucleus, and there is abundant

1011: observational evidence for cold

1012: (e.g. \cite[Jackson {\it et~al.} 1993]{Jackson:93};

1013: \cite[Gallimore {\it et~al.} 2001]{Gallimore:01};

1014: \cite[Greenhill {\it et~al.} 2003]{Greenhill:03})

1015: and hot (e.g. \cite[Baganoff {\it et~al.} 2003]{Baganoff:03}) 

1016: gas near the centers of at least some galaxies.

1017: Dense concentrations of gas can substantially accelerate 

1018: the evolution of a massive binary by increasing the drag on 

1019: the individual BHs

1020: (Escala {\it et~al.} 2004, 2005; \cite[Dotti {\it et~al.} 2007]{Dotti:07}).

1021: The plausibility of such dense accumulations of gas,

1022: with mass comparable to the mass of the SBHs, is unclear however

1023: (e.g. \cite[Sakamoto {\it et~al.} 1999]{Sakamoto:99};

1024: \cite[Christopher {\it et~al.} 2005]{Christopher:05}).

1025: Large-scale galaxy merger simulations 

1026: (\cite[Kazantzidis {\it et~al.} 2005]{Kazant:05};

1027: \cite[Mayer {\it et~al.} 2007]{Mayer:07})

1028: show that the presence of gas leads to more rapid formation

1029: of the massive binary, but these simulations still lack the

1030: resolution to follow the binary past $a\approx  r_h$

1031: and so have nothing relevant to say (yet) concerning the 

1032: final parsec problem.

1033: 

1034: \bigskip

1035: As this summary indicates, many possible solutions

1036: to the ``final parsec problem'' exist, but none is guaranteed to

1037: be effective in all or even most galaxies.

1038: The safest bet is that both coalesced and uncoalesced binary

1039: SBHs exist, but with what relative frequency is still anyone's guess.

1040: 

1041: \section{SBH/IBH binaries}\label{IBH}

1042: 

1043: Secure dynamical evidence exists for SBHs in the mass range

1044: $10^{6.5}\lesssim M_\bullet/M_\odot\lesssim 10^{9.5}$

1045: (\cite[Ferrarese \& Ford 2005]{FF:05})

1046: and compelling arguments have been made for BHs with masses

1047: $10^5\lesssim M_\bullet/M_\odot\lesssim 10^7$

1048: in active nuclei (\cite[Greene \& Ho 2004]{Greene:04}).

1049: Binary mass ratios as extreme as $1000:1$, and possibly

1050: greater, are therefore to be expected.

1051: This possibility has received most attention in the context

1052: of intermediate-mass black holes

1053: (IBHs) in the Milky Way, where they could form 

1054: in dense star clusters like the Arches or Quintuplet

1055: before spiralling into the center and forming a tight

1056: binary with the $\sim 3.5\times 10^6 M_\odot$ 

1057: SBH (\cite[Portegies Zwart \& McMillan 2002]{Zwart:02};

1058: \cite[Hansen \& Milosavljevi{\'c} 2003]{Hansen:03}).

1059: 

1060: \begin{figure}

1061:   \vspace{0.5cm}

1062:   \includegraphics[height=0.305\textheight]{fig_Baumgardtetal1.eps}

1063:   \includegraphics[height=0.31\textheight]{fig_RB1.ps}

1064:   \caption{{\it Left:} $N$-body  simulations of the inspiral of IBHs into

1065:   the center of the Milky Way. Solid lines show the separation

1066:   between the two BHs and dashed lines are theoretical predictions

1067:   that ignore loss-cone depletion or changes in the structure of the

1068:   core.

1069:   Smaller IBHs spiral in farther before ``stalling'';

1070:   the $M_{\rm IBH}=10^3M_\odot$ simulation ends before the stalling

1071:   radius is reached.

1072:    (From Baumgardt, Gualandris \& Portegies Zwart 2006.)

1073:    {\it Right:} Evolution beyond $a=a_h$, based on the Fokker-Planck

1074:    model of Merritt, Mikkola \& Szell (2007). 

1075:    Dashed lines indicate when the evolution time due to gravitational

1076:    radiation losses is less than $10$ Gyr.}\label{fig:IBH}

1077: \end{figure}

1078: 

1079: The predicted hard-binary separation for a SBH/IBH pair is

1080: (\ref{eq:ah})

1081: \begin{equation}

1082: a_h \approx 0.5 {\rm mpc} \left({q\over 10^{-3}}\right) 

1083: \left({M_\bullet\over 3.5\times 10^6M_\odot}\right)^{0.59},

1084: \ \ \ \ q\equiv {M_{\rm IBH}\over M_\bullet}.

1085: \end{equation}

1086: This separation -- $\sim 10^2$ AU -- is comparable to

1087: the periastron distances of the famous ``S'' stars

1088: (\cite[Eckart {\it et~al.} 2002]{Eckart:02};

1089: \cite[Ghez {\it et~al.} 2005]{Ghez:05}).

1090: Dynamical constraints on the existence of an IBH at this

1091: distance from the SBH are currently weak

1092: (\cite[Yu \& Tremaine 2003]{Yu:03};

1093: \cite[Hansen \& Milosavljevi{\'c} 2003]{Hansen:03};

1094: \cite[Reid \& Brunthaler 2004]{Reid:04}).

1095: Figure~\ref{fig:IBH}a shows $N$-body simulations designed to 

1096: mimic inspiral of IBHs into the Galactic center.

1097: The figure confirms the expected slowdown in the inspiral rate

1098: at a separation $\sim a_h$.

1099: Figure~\ref{fig:IBH}b plots evolutionary tracks for the same

1100: three IBH masses as in the 

1101: left panel, based on the Fokker-Planck model of

1102: \cite{MMS:07}.

1103: For $M_{\rm IBH}\lesssim 10^3 M_\odot$, evolution of the

1104: binary is dominated by gravitational wave losses already

1105: at $a=a_h$.

1106: 

1107: \begin{figure}

1108:   \vspace{0.2cm}

1109:   \includegraphics[height=0.315\textheight]{fig_Baumgardtetal4.eps}

1110:   \includegraphics[height=0.335\textheight]{fig_baum_FP.ps}

1111:   \caption{{\it Left:} Creation of a core by inspiral of IBHs

1112:   into the Galactic center.  The initial density profile is shown

1113:   by the solid line.  Green (dotted) line shows the core $10$ Myr 

1114:   after the $10^4M_\odot$ IBH has merged with the SBH; almost

1115:   no change occurs during this time.

1116:    (From Baumgardt, Gualandris \& Portegies Zwart 2006.)

1117:    {\it Right:} Fokker-Planck model showing how the cusp

1118:    regenerates due to two-body scattering, on Gyr timescales.

1119:    (Adapted from Merritt, Mikkola \& Szell 2007.)}\label{fig:IBH2}

1120: \end{figure}

1121: 

1122: The same inward flux of stars that allows the binary to shrink

1123: also implies an outward flux of stars ejected by the binary.

1124: The latter are a potential source of ``hyper-velocity stars''

1125: (HVSs),

1126: stars moving in the halo with greater than Galactic escape velocity 

1127: (\cite[Hills 1988]{Hills:98}).

1128: The relation between the stellar ejection rate and the binary

1129: hardening rate, when $a\le a_h$, 

1130: is given by (\ref{eq:hardb}) after rewriting it as

1131: \begin{equation}

1132: T_{\rm hard}\equiv a{d\over dt}\left({1\over a}\right) \approx 

1133: {2.5 \over M_\bullet} \times {\rm flux};

1134: \label{eq:hardc}

1135: \end{equation}

1136: here ``flux'' is the total mass in stars per unit time,

1137: from all energies, that are scattered

1138: into (and ejected by) the binary.

1139: Combining (\ref{eq:hardc}) with (\ref{eq:tdepend}), the flux is

1140: \begin{subequations}

1141: \begin{eqnarray}

1142: &\sim& 5.0 {M_\bullet\over T_R(r_h)} 

1143: \left[1+{5t\over T_R(r_h)}\right]^{-1/2} \\

1144: &\lesssim& 350 M_\odot {\rm yr}^{-1} 

1145: \end{eqnarray}

1146: \end{subequations}

1147: where the second line uses values appropriate to the Galactic center.

1148: %This estimate is consistent with $N$-body results

1149: %(\cite[Baumgardt, Gualandris \& Portegies Zwart 2006]{Baumgardt:06};

1150: %\cite[Matsubayashi {\it et~al.} 2007]{Matsu:07}).

1151: Relating the total ejected flux to the number of HVSs

1152: that would be observed is not straighforward;

1153: for instance, only a fraction ($\lesssim 10$\%) 

1154: would be ejected with high enough velocity

1155: to still be moving faster than $\sim 500$ km s$^{-1}$

1156: after climbing through the Galactic potential

1157: (\cite[Gualandris {\it et~al.} 2005]{Gualandris:05};

1158: \cite[Baumgardt {\it et~al.} 2006]{Baumgardt:06}),

1159: and targeted searches for HVSs only detect certain stellar types

1160: so that knowledge of the stellar mass function is also required

1161: (\cite[Brown {\it et~al.} 2006]{Brown:06}).

1162: 

1163: Inspiral of the IBH creates a core of radius 

1164: $\sim 0.05\ {\rm pc} \approx 1''$ 

1165: (figure~\ref{fig:IBH2}a).

1166: Such a core might barely be detectable at the center of the 

1167: Milky Way from star counts.

1168: There is no clear indication of a core 

1169: (\cite[Schoedel {\it et.~al} 2007]{Schoedel:07}),

1170: but if the inspiral occurred more than a few Gyr

1171: ago, star-star gravitational scattering would have

1172: gone some way toward ``refilling'' the region depleted

1173: by the binary 

1174: (\cite[Merritt \& Wang 2005]{Wang:05};

1175: \cite[Merritt \& Szell 2006]{MS:06}; figure~\ref{fig:IBH}b).

1176: In this case, however, the ejected stars

1177: would almost all have moved beyond the range of HVS surveys

1178: by now.

1179: 

1180: The angular distribution of the ejected stars has been proposed

1181: as a test for their origin; unlike other possible sources of

1182: HVSs, a SBH/IBH binary tends to eject

1183: stars parallel to the orbital plane or, if the orbit is eccentric,

1184: in a particular direction 

1185: (\cite[Levin 2006]{Levin:06};

1186: \cite[Sesana {\it et~al.} 2006]{Sesana:06}).

1187: In two $N$-body simulations of IBH inspiral however

1188: (\cite[Baumgardt, Gualandris \& Portegies Zwart 2006]{Baumgardt:06};

1189: \cite[Matsubayashi {\it et~al.} 2007]{Matsu:07}),

1190: the orientation of the binary began to change appreciably,

1191: in the manner of a random walk, after it became hard.

1192: This was due to ``rotational Brownian motion''

1193: (\cite[Merritt 2002]{Merritt:02}):

1194: torques from passing stars -- the same stars that extract

1195: energy and angular momentum from the binary -- also change

1196: the direction of the binary's orbital angular momentum vector.

1197: In one hardening time $|a/\dot{a}|$ of the binary,

1198: its orientation changes by

1199: \begin{subequations}

1200: \begin{eqnarray}

1201: \Delta\theta &\approx& q^{-1/2} \left({m_\star\over M_\bullet}\right)^{1/2} 

1202: \left(1-e^2\right)^{-1/2} \\

1203: &\approx& 9.0^\circ \left({q\over 10^{-3}}\right)^{-1/2} 

1204: \left({M_\bullet\over 10^6 m_\star}\right)^{-1/2}

1205: \left[{\left(1-e^2\right)^{-1/2}\over 5}\right].

1206: \label{eq:RBM}

1207: \end{eqnarray}

1208: \end{subequations}

1209: (The eccentricity dependence in (\ref{eq:RBM}) is approximate;

1210: the numerical coefficient in this equation has only been confirmed by detailed

1211: scattering experiments for $e=0$.)

1212: In both of the cited $N$-body studies, the binary eccentricity

1213: evolved appreciably away from zero before the orientation changes

1214: became signficant.

1215: Rotational Brownian motion  might not act quickly enough to randomize

1216: the orienation of a SBH/IBH binary in a time of $\sim 10^8$ yr,

1217: the flight time from the Galactic center to the halo,

1218: unless perturbers more massive than Solar-mass stars

1219: are present near the binary however

1220: (\cite[Merritt 2002]{Merritt:02};

1221: \cite[Perets \& Alexander 2007]{Perets:07}).

1222: 

1223: \section{Kicks and cores}\label{sec:kicksandcores}

1224: 

1225: After seeming to languish for several decades, the field

1226: of numerical relativity has recently experienced exciting progress.

1227: Following the breakthrough papers of \cite{Pretorius:05},

1228: \cite{Campanelli:06} and \cite{Baker:06a}, several groups

1229: have now successfully simulated the evolution of binary BHs 

1230: all the way to coalescence.

1231: The final inspiral is driven by emission of gravitational waves,

1232: and in typical (asymmetric) inspirals, a net impulse -- a ``kick'' --

1233: is imparted to the system due to anisotropic emission of the waves

1234: (\cite[Bekenstein 1973]{Bekenstein:73}; \cite[Fitchett 1984]{Fitchett:84};

1235: \cite[Favata et al. 2004]{Favata:04}).  Early arguments that the

1236: magnitude of the recoil velocity would be modest for non-spinning BHs

1237: %(\cite[Redmount \& Rees 1989]{Redmount:89}) 

1238: were confirmed by the simulations,  which found

1239: $V_{\rm kick}\lesssim 200$ km s$^{-1}$ in the absence of spins

1240: (\cite[Baker et al. 2006]{Baker:06b}; 

1241: \cite[Gonzalez et al. 2007a]{Gonzalez:07a};

1242: \cite[Herrmann et al. 2007]{Herrmann:07}).  

1243: The situation changed

1244: dramatically following the first 

1245: (\cite[Campanelli {\it et~al.} 2007a]{Campanelli:07a})

1246: simulations of ``generic'' binaries,

1247: in which the individual BHs were spinning and tilted with respect

1248: to the orbital angular momentum vector.

1249: Kicks as large as $\sim 2000$ km s$^{-1}$ have now

1250: been confirmed (\cite[Campanelli {\it et~al.} 2007b]{Campanelli:07b}; 

1251: \cite[Gonzalez {\it et~al.} 2007b]{Gonzalez:07b};

1252: \cite[Tichy \& Marronetti 2007]{Tichy:07}), 

1253: and scaling arguments based on the post-Newtonian approximation

1254: suggest that the maximum kick velocity would

1255: probably increase to $\sim 4000$ km s$^{-1}$ in the case of maximally-spinning

1256: holes (\cite[Campanelli {\it et~al.} 2007b]{Campanelli:07b}).  

1257: The most propitious configuration for the kicks appears to be

1258: an equal-mass binary in which the individual

1259: spin vectors are oppositely aligned and oriented parallel to the

1260: orbital plane.  

1261: For unequal-mass binaries, the maximum kick is

1262: \begin{equation}

1263: V_{\rm max} \approx 6\times 10^4 {\rm km\ s}^{-1} {q^2\over (1+q)^4}

1264: \end{equation}

1265: where $q\equiv M_2/M_1\le 1$ is the binary mass ratio

1266: and maximal spins have been assumed

1267: (\cite[Campanelli {\it et~al.} 2007c]{Campanelli:07c}).

1268: Orienting the BHs with their spins perpendicular to the orbital

1269: angular momentum may seem odd

1270: (\cite[Bogdanovi\'c, Reynolds \& Miller 2007]{BRM:07}),

1271: but there is considerable evidence that SBH spins bear no relation

1272: to the orientations of the gas disks that surround them

1273: (e.g. \cite[Kinney {\it et~al.} 2000]{Kinney:00}; \cite[Gallimore {\it et.~al.} 2006]{Gallimore:06}) and this is presumably even more true with

1274: respect to the directions of infalling BHs.

1275: Galaxy escape velocities are $\lesssim 3000$ km s$^{-1}$ 

1276: (\cite[Merritt {\it et~al.} 2004]{MMFHH:04}),

1277: so gravitational wave recoil can in principle eject 

1278: coalescing SBHs completely from galaxies.

1279: 

1280: \begin{figure}

1281:  \vspace{-0.2cm}

1282:  \includegraphics[height=14cm]{fig_cont_1.ps}

1283:  \includegraphics[height=15.cm]{fig_lauer.ps}

1284:  \caption{{\it Left:} Core oscillations in an $N$-body simulation

1285:   of ejection of a SBH from the center of a galaxy; 

1286:   the kick velocity was 60\% of the escape velocity.  

1287:   Contour plots show the stellar density at equally spaced times,

1288:   spanning $\sim 1/2$ of the SBH's orbital period.

1289:   Filled circles

1290:   are the SBH and crosses indicate the location of the (projected)

1291:   density maxima. (From Gualandris \& Merritt 2007.)

1292:   {\it Right:} Surface brightness contours of three ``core'' galaxies 

1293:   with double or offset nuclei, from Lauer {\it et~al.} (2005).

1294:   {\it Top:} NGC 4382; {\it middle:} NGC 507; {\it bottom:} NGC 1374.

1295:   \label{fig:cont}}

1296: \end{figure}

1297: 

1298: Detailed $N$-body simulations show that the motion of a SBH that 

1299: has been kicked with enough velocity to eject it out of the 

1300: core, but not fast enough to escape the galaxy entirely, 

1301: exhibits three distinct phases (\cite[Gualandris \& Merritt 2007]{GM:07}):

1302: 

1303: \begin{itemize}

1304: \item {\it Phase I:}

1305: The SBH oscillates with decreasing amplitude, losing

1306:   energy via dynamical friction each time it passes through the core.

1307:   Chandrasekhar's theory accurately reproduces the motion of the SBH

1308:   in this regime for values $2\lesssim\ln\Lambda\lesssim 3$ of the

1309:   Coulomb logarithm, if the gradually-decreasing core density 

1310:   is taken into account.  

1311: \item {\it Phase II:} When the amplitude of the motion has decayed

1312:    to roughly the core radius, the SBH

1313:   and core begin to exhibit oscillations about their common center of

1314:   mass (figure~\ref{fig:cont}). These oscillations decay exponentially

1315:   (figure~\ref{fig:tests}), but with a time constant that is $10-20$ times 

1316:   longer than would be predicted by a naive application of the 

1317:   dynamical friction formula.

1318: \item {\it Phase III:} Eventually the SBH's kinetic energy drops

1319:   to an average value

1320: \begin{equation}

1321: {1\over 2}M_\bullet V_\bullet^2 \approx {1\over 2}m_\star {\rm v}_\star^2

1322: \label{eq:vbrown}

1323: \end{equation}

1324: i.e. to the kinetic energy of a single star.

1325: This is the regime of gravitational Brownian motion

1326: (\cite[Bahcall \& Wolf 1976]{BW:76}; \cite[Young 1977]{Young:77}; 

1327: \cite[Merritt {\it et~al.} 2007]{MBL:07}).

1328: \end{itemize}

1329: 

1330: \begin{figure}

1331:   \vspace{0.5cm}

1332:   \includegraphics[height=0.62\textheight,angle=-90.]{fig_tests.ps}

1333:   \caption{Evolution of the SBH kinetic energy following a kick of

1334:     60\% the central escape velocity, in two $N$-body simulations of

1335:     a galaxy represented by $N$ stars, with

1336:     $N=(2.5\times 10^5, 2\times 10^6)$.

1337:     The mass of the SBH, and the total mass of the galaxy, 

1338:     are the same in the two simulations; all that varies 

1339:     is the mass of the ``star'' particles.

1340:     The right-hand panel shows binned values of $V^2$.

1341:     Most of the elapsed time is spent in SBH/core oscillations

1342:     like those illustrated in figure~\ref{fig:cont}.

1343:     Eventually, the SBH's kinetic energy decays to the Brownian

1344:     value, shown as the horizontal dashed lines in the right panel.

1345:     The Brownian velocity scales as $m_\star^{1/2}$ and so is

1346:     smaller for larger $N$.

1347:     Scaled to a $3\times 10^{10}M_\odot$ galaxy, the

1348:     time to reach the Brownian regime would be $\sim 10^8$ yr.

1349:     (Adapted from Gualandris \& Merritt 2007.)

1350:    }\label{fig:tests}

1351: \end{figure}

1352: 

1353: A natural definition of the ``return time'' of a kicked SBH is the 

1354: time to reach the Brownian regime.  

1355: Unless the kick is very close to the escape velocity,

1356: the return time is dominated by the time spent in ``Phase II;''

1357: during this time, the SBH's energy decays roughly as 

1358: \begin{equation}

1359: E \approx \Phi(0) + \Phi(r_c) e^{-(t-t_c)/\tau}

1360: \end{equation}

1361: (\cite[Gualandris \& Merritt 2007]{GM:07});

1362: $t_c$ is the time when the SBH re-enters the core whose

1363: radius is $r_c$.

1364: The damping time in the $N$-body simulations, $\tau$, is 

1365: \begin{subequations}

1366: \begin{eqnarray}

1367: \tau &\approx & 15{\sigma_c^3\over G^2\rho_cM_\bullet}\\

1368: &\approx &1.2\times 10^7{\rm yr} 

1369: \left({\sigma_c\over 250{\rm km\ s}^{-1}}\right)^3

1370: \left({\rho_c\over 10^3M_\odot {\rm pc}^{-3}}\right)^{-1}

1371: \left({M_\bullet\over 10^9M_\odot}\right)^{-1}, 

1372: \end{eqnarray}

1373: \end{subequations}

1374: with $\rho_c$ and $\sigma_c$ the core density and velocity

1375: dispersion respectively.

1376: The number of decay times required for the SBH's energy

1377: to reach the Brownian level is 

1378: $\sim\ln\left({M_\bullet/m_\star}\right)

1379: \approx 20$, implying that a kicked SBH will remain significantly

1380: off-center for a long time, as long as $\sim 1$ Gyr in a bright galaxy

1381: with a low-density core.

1382: 

1383: In fact, asymmetric cores are rather common.

1384: These include off-center nuclei 

1385: (\cite[Bingelli {\it et~al.} 2000]{Bingelli:00};

1386: \cite[Lauer {\it et~al.} 2005]{Lauer:05});

1387: double nuclei (\cite[Lauer {\it et~al.} 1996]{Lauer:96});

1388: and cores with a central minimum in the surface brightness

1389: (\cite[Lauer {\it et~al.} 2002]{Lauer:02}).

1390: Three examples, from \cite{Lauer:05}, are reproduced here

1391: on the right side of figure~\ref{fig:cont}; 

1392: all are luminous ``core'' galaxies, and each

1393: strikingly resembles at least one frame from the $N$-body montage

1394: on the left.

1395: The longevity of the ``Phase II'' oscillations makes the kicks

1396: a plausible model for the observed asymmetries.

1397: This explanation is probably not appropriate for the famous

1398: double nucleus of M31, since M31 is not a ``core'' galaxy,

1399: and since one of the brightness peaks in M31 (the one associated

1400: with the SBH) lies essentially at the galaxy photocenter;

1401: Figure~\ref{fig:cont} suggests that an oscillating SBH

1402: would typically (though not always) be found on the opposite side 

1403: of the galaxy from the point of peak brightness.

1404: The M31 double nucleus has been successfully modelled as a clump

1405: of stars on eccentric orbits which maintain their lopsidedness

1406: by virtue of moving deep within the Keplerian potential of the SBH

1407: (\cite[Tremaine 1995]{Tremaine:95}).

1408: 

1409: \begin{figure}

1410:  \vspace{0.2cm}

1411:  \includegraphics[height=0.30\textheight]{fig_GM_12.ps}

1412:  \includegraphics[height=0.30\textheight]{fig_GM_13.eps}

1413:  \caption{{\it Left:} 

1414:   Points show projected  density profiles computed from $N$-body

1415:   models after a kicked SBH has returned to the center,

1416:   for three different values of the kick velocity 

1417:   ($V_{\rm kick}=(0.2,0.4,0.8)\times V_{\rm esc}$). 

1418:   Each set of points is compared with the best-fitting core-Sersic

1419:   model (lines). 

1420:   The insert shows a  zoom into the central region.

1421:   {\it Right:}  Mass deficits generated by kicked SBHs.

1422:   The differents sets of points correspond to different

1423:   galaxy models.

1424:   (Adapted from Gualandris \& Merritt 2007.)

1425:   \label{fig:GMmdef}}

1426: \end{figure}

1427: 

1428: The kicks are quite effective at inflating cores

1429: (\cite[Merritt {\it et~al.} 2004]{MMFHH:04};

1430: \cite[Boylan-Kinchin {\it et~al.} 2004]{Boylan:04}).

1431: Figure~\ref{fig:GMmdef}, from \cite{GM:07}, illustrates this:

1432: the mass deficit generated by the kick is approximately

1433: \begin{equation}

1434: M_{\rm def,k} \approx 5M_\bullet \left(V_{\rm kick}/V_{\rm esc}\right)^{1.75}.

1435: \label{eq:mdefk}

1436: \end{equation}

1437: Even modest kicks can generate cores substantially larger 

1438: than those produced during the formation of a massive binary

1439: ($\sim 1 M_\bullet$; eq.~\ref{eq:mdefh}).

1440: Furthermore, this mechanism is potentially effective in even the

1441: most luminous galaxies, unlike the relaxation-driven model

1442: for core growth discussed above (eq.~\ref{eq:mdefc})

1443: which only applies to nuclei with short relaxation times.

1444: Gravitational radiation recoil is therefore a tenable explanation 

1445: for the subset

1446: of luminous E galaxies with large mass deficits (figure~\ref{fig:mdef}). 

1447: An alternative explanation for the over-sized cores

1448: (\cite[Lauer {\it et~al.} 2007]{Lauer:07})

1449: postulates that the SBHs in these galaxies are ``hypermassive,'' 

1450: $M_\bullet\gtrsim 10^{10}M_\odot$ and that the cores

1451: are a consequence of slingshot ejection by a massive binary.

1452: 

1453: The kicks have a number of other potentially observable 

1454: consequences, including spatially and/or kinematically offset AGN

1455: (\cite[Madau \& Quataert 2004]{MQ:04}; 

1456: \cite[Haehnelt {\it et~al.} 2006]{Haehnelt:06};

1457: \cite[Merritt {\it et~al.} 2006]{Naked:06};

1458: \cite[Bonning {\it et~al.} 2007]{Bonning:07})

1459: and distorted or wiggling radio jets

1460: (\cite[Gualandris \& Merritt 2007]{GM:07}).

1461: Many of these manifestations were first discussed by R. Kapoor in

1462: a remarkably prescient series of papers

1463: (\cite[Kapoor 1976,1983a,b,1985]{Kapoor:76,Kapoor:1983a,Kapoor:1983b,

1464: Kapoor:85}).

1465: 

1466: \section{Black-hole-driven expansion}

1467: 

1468: The growth of a core around a shrinking, binary SBH

1469: was discussed above: beyond a certain radius, the

1470: relaxation time becomes so long that the encounter-driven 

1471: flux of stars toward the center cannot compensate for 

1472: losses to the binary, forcing the density to drop.

1473: A similar process takes place around a single SBH

1474: (\cite[Shapiro 1977]{Shapiro:77}; \cite[Dokuchaev 1989]{Dokuchaev:89}): 

1475: stars coming too near are consumed, or disrupted, and the

1476: density drops.

1477: This effect is absent from the classical equilibrium

1478: models for stars around a BH

1479: (e.g. \cite[Bahcall \& Wolf 1976]{BW:76}; \cite[Cohn \& Kulsrud 1978]{CK:78})

1480: since these solutions fix the phase-space density far from the BH,

1481: enforcing an inward flux of stars 

1482: precisely large enough to replace the stars 

1483: being consumed by the sink.

1484: In reality, the BH acts as a heat source, in much the same way that 

1485: hard binary stars inject energy into a post-core-collapse globular 

1486: cluster and cause it to re-expand.

1487: 

1488: A simple model that produces self-similar

1489: expansion of a nucleus containing a SBH can be constructed

1490: by simply changing the outer boundary condition in the

1491: \cite{BW:76} problem

1492: from $f(0)=f_0$ to $f(0)=0$.

1493: One finds that the evolution after $\sim$one relaxation

1494: time can be described as 

1495: $\rho(r,t) = \rho_c(t)\rho^*(r)$, with

1496: $\rho^*(r)$ slightly steeper than the $\rho\sim r^{-7/4}$

1497: Bahcall-Wolf form; the normalization drops

1498: off as $\rho_c\propto t^{-1}$ at late times.

1499: Figure~\ref{fig:m32expand} shows the results of

1500: a slightly more realistic calculation in a model

1501: designed to mimic the nearby dE galaxy M32.

1502: After reaching approximately the Bahcall-Wolf form, 

1503: the density drops in amplitude

1504: with roughly fixed slope for $r\lesssim r_h$.

1505: This example suggests that the nuclei of galaxies

1506: like M32 or the Milky Way might have been $\sim$ a few

1507: times denser in the past than they are now,

1508: with correspondingly higher rates of stellar

1509: tidal disruption and stellar collisions.

1510: 

1511: Expansion due to a central BH has been observed in a 

1512: handful of studies based on fluid 

1513: (\cite[Amaro-Seoane {\it et~al.} 2004]{Amaro:04}),

1514: Monte-Carlo (\cite[Shapiro \& Marchant 1978]{SM:78};

1515: \cite[Marchant \& Shapiro 1980]{MS:80};

1516: \cite[Freitag {\it et~al.} 2006]{Freitag:06}),

1517: Fokker-Planck (\cite[Murphy {\it et~al.} 1991]{MCD:91}),

1518: and $N$-body (\cite[Baumgardt {\it et~al.} 2000]{Baumgardt:04}) algorithms.

1519: All of these studies allowed stars to be lost into or destroyed

1520: by the BH; however most adopted parameters more

1521: suited to globular clusters than to nuclei,

1522: e.g. a constant-density core.

1523: \cite{MCD:91} applied the isotropic, multi-mass

1524: Fokker-Planck equation to the evolution of nuclei

1525: containing SBHs, including an approximate loss term

1526: in the form of (\ref{eq:dndt}) to model the

1527: scattering of low-angular-momentum stars into the SBH.

1528: Most of their models had what would now be considered

1529: unphysically high densities and the evolution was dominated

1530: by physical collisions between stars.

1531: However in two models with lower densities, they reported

1532: observing significant expansion over $10^{10}$ yr; 

1533: these models had initial central relaxation times of 

1534: $T_r\lesssim 10^9$ yr when scaled to real galaxies,

1535: similar to the relaxation times near the centers of

1536: M32 and the Milky Way.

1537: The $\rho\sim r^{-7/4}$ form of the density profile

1538: near the SBH was observed to be approximately conserved

1539: during the expansion.

1540: \cite{Freitag:06} carried out Monte-Carlo

1541: evolutionary calculations of a suite of models

1542: containing a mass spectrum, some of which were

1543: designed to mimic the Galactic center star cluster.

1544: After the stellar-mass BHs in their models

1545: had segregated to the center,

1546: they observed a roughly self-similar expansion.

1547: \cite{Baumgardt:04} followed core collapse

1548: in $N$-body models with and without a 

1549: massive central particle;

1550: ``tidal destruction'' was modelled by simply

1551: removing stars that came within a certain distance

1552: of the massive particle.

1553: When the ``black hole'' was present, the cluster 

1554: expanded almost from the start

1555: and in an approximately self-similar way.

1556: These important studies notwithstanding, there is a crucial

1557: need for more work on this problem in order to

1558: understand how the rates of processes like stellar tidal

1559: disruption vary over cosmological times

1560: (e.g. \cite[Milosavljevi{\'c} {\it et~al.} 2006]{milos:06}).

1561: 

1562: \begin{figure}

1563: \vspace{-2.cm}

1564: \begin{center}

1565: \includegraphics[width=0.65\textwidth,angle=-90.]{fig_m32expand.ps}

1566: \caption

1567: {Black-hole-driven expansion of a nucleus; this Fokker-Planck

1568: model was given parameters such that the density at the end

1569: is similar to what is currently observed in the nucleus

1570: of M32, with a final influence radius $r_h\approx 3$ pc.

1571: The left panel shows density profiles at constant time intervals

1572: after a Bahcall-Wolf cusp has been established;

1573: the right panel shows the evolution of the density at

1574: $0.1$ pc as a function of $M_{acc}$, 

1575: the accumulated mass in tidally-disrupted stars.

1576: As scaled to M32, the final time is roughly $2\times 10^{10}$ yr.

1577: This plot suggests that the densities of collisional

1578: nuclei like those of M32 and the Milky Way were once higher, 

1579: by factors of $\sim $ a few, than at present.

1580: (From Merritt 2006b.)

1581: \label{fig:m32expand}

1582: }

1583: \end{center}

1584: \end{figure}

1585: 

1586: \begin{acknowledgments}

1587: We thank H. Baumgardt, A. Graham and T. Lauer for permission to

1588: reproduce figures from their published work, and A. Graham

1589: for comments on the manuscript.

1590: We acknowledge support from the

1591: National Science Foundation under grants 

1592: AST-0420920 and AST-0437519 and from the National

1593: Aeronautics and Space Administration under grants

1594: NNG04GJ48G and NNX07AH15G.

1595: \end{acknowledgments}

1596: 

1597: 

1598: \begin{thebibliography}{}

1599: 

1600: \bibitem[Amaro-Seoane {\it et~al.} (2004)]{Amaro:04} 

1601: \textsc{Amaro-Seoane, P., Freitag, M., \& Spurzem, R.} 2004

1602: Accretion of stars on to a massive black hole: 

1603: a realistic diffusion model and numerical studies.

1604: \textit{Mon.~Not.~R.~Astron.~Soc.} \textbf{352}, 655--672.

1605: 

1606: \bibitem[Armitage \& Natarajan(2002)]{Armitage:02} 

1607: \textsc{Armitage, P.~J., \& Natarajan, P.} 2002

1608: Accretion during the Merger of Supermassive Black Holes.

1609: \textit{Astrophys.~J.~Letts.} \textbf{567}, L9--L12.

1610: 

1611: \bibitem[Baganoff {\it et~al.} 2003]{Baganoff:03} 

1612: \textsc{Baganoff, F.~K., et al.} 2003 

1613: Chandra X-ray spectroscopic imaging of Sagittarius A* and the central 

1614: parsec of the Galaxy.

1615: \textit{Astrophys.~J.} \textbf{591}, 891--899.

1616: 

1617: \bibitem[Bahcall \& Wolf (1976)]{BW:76} 

1618: \textsc{Bahcall, J.~N., \& Wolf, R.~A.} 1976

1619: Star distribution around a massive black hole in a globular cluster.

1620: \textit{Astrophys.~J.} \textbf{209}, 214--232.

1621: 

1622: \bibitem[Bahcall \& Wolf (1977)]{BW:77} 

1623: \textsc{Bahcall, J.~N., \& Wolf, R.~A.} 1977

1624: The star distribution around a massive black hole in a globular cluster. 

1625: II Unequal star masses.

1626: \textit{Astrophys.~J.} \textbf{216}, 883--907.

1627: 

1628: \bibitem[Baker {\it et~al.} (2006a)]{Baker:06a} 

1629: \textsc{Baker, J.~G., Centrella, J., Choi, D.-I., Koppitz, M., \& 

1630: van Meter, J.} 2006a

1631: Gravitational-wave extraction from an inspiraling configuration of 

1632: merging black holes. 

1633: \textit{Phys.~Rev.~Letts.} \textbf{96}, 111102.

1634: 

1635: \bibitem[Baker {\it et~al.} (2006b)]{Baker:06b} 

1636: \textsc{Baker, J.~G., Centrella, J., Choi, D.-I., Koppitz, M., 

1637: van Meter, J.~R., \& Miller, M.~C.} 2006b

1638: Getting a kick out of numerical relativity.

1639: \textit{Astrophys.~J.} \textbf{653}, L93--L96.

1640: 

1641: \bibitem[Balcells {\it et~al.} (2007)]{Balcells:07} 

1642: \textsc{Balcells, M., Graham, A.~W., \& Peletier, R.~F.} 2007

1643: Galactic Bulges from Hubble Space Telescope NICMOS Observations: 

1644: Central Galaxian Objects, and Nuclear Profile Slopes.

1645: \textit{Astrophys.~J.} \textbf{665}, 1084--1103.

1646: 

1647: \bibitem[Batcheldor {\it et~al.} (2007)]{Batcheldor:07} 

1648: \textsc{Batcheldor, D., Marconi, A., Merritt, D., \& Axon, D.~J.} 2007

1649: How Special Are Brightest Cluster Galaxies? 

1650: The Impact of Near-Infrared Luminosities on Scaling Relations for BCGs.

1651: \textit{Astrophys.~J.} \textbf{663}, L85-L88.

1652: 

1653: \bibitem[Baumgardt et al. (2006)]{Baumgardt:06} 

1654: \textsc{Baumgardt, H., Gualandris, A., \& Portegies Zwart, S.} 2006

1655: Ejection of hypervelocity stars from the Galactic Centre 

1656: by intermediate-mass black holes.

1657: \textit{Mon.~Not.~R.~Astron.~Soc.} \textbf{372}, 174--182.

1658: 

1659: \bibitem[Baumgardt {\it et~al.} (2004)]{Baumgardt:04} 

1660: \textsc{Baumgardt, H., Makino, J., \& Ebisuzaki, T.} 2004

1661: Massive black holes in star clusters. I. Equal-mass clusters.

1662: \textit{Astrophys.~J.} \textbf{613}, 1133--1142.

1663: 

1664: \bibitem[Begelman {\it et~al.} (1980)]{BBR:80} 

1665: \textsc{Begelman, M.~C., Blandford, R.~D., \& Rees, M.~J.} 1980

1666: Massive black hole binaries in active galactic nuclei.

1667: \textit{Nature} \textbf{287}, 307--309.

1668: 

1669: \bibitem[Bekenstein(1973)]{Bekenstein:73} 

1670: \textsc{Bekenstein, J.~D.} 1973

1671: Gravitational-Radiation Recoil and Runaway Black Holes.

1672: \textit{Astrophys.~J.~Letts.} \textbf{183}, 657--664. 

1673: 

1674: \bibitem[Berczik et al.(2005)]{Bercik:05} 

1675: \textsc{Berczik, P., Merritt, D., \& Spurzem, R.} 2005

1676: Long-Term Evolution of Massive Black Hole Binaries. II. 

1677: Binary Evolution in Low-Density Galaxies.

1678: \textit{Astrophys.~J.} \textbf{633}, 680--687. 

1679: 

1680: \bibitem[Berczik {\it et~al.} (2006)]{Berczik:06} 

1681: \textsc{Berczik, P., Merritt, D., Spurzem, R., \& Bischof, H.-P.} 2006

1682: Efficient Merger of Binary Supermassive Black Holes in Nonaxisymmetric 

1683: Galaxies.

1684: \textit{Astrophys.~J.~Letts.} \textbf{642}, L21--L24.

1685: 

1686: \bibitem[Binggeli {\it et~al.} (2000)]{Binggeli:00} 

1687: \textsc{Binggeli, B., Barazza, F., \& Jerjen, H.} 2000

1688: Off-center nuclei in dwarf elliptical galaxies.

1689: \textit{Astron.~Ap.} \textbf{359}, 447--456.

1690: 

1691: \bibitem[Binggeli {\it et~al.} (1984)]{Binggeli:84} 

1692: \textsc{Binggeli, B., Sandage, A., \& Tarenghi, M.} 1984

1693: Studies of the Virgo Cluster. I - Photometry of 109 galaxies 

1694: near the cluster center to serve as standards.

1695: \textit{Astron.~J.} \textbf{89}, 64--82.

1696: 

1697: \bibitem[Blaes {\it et~al.} (2002)]{Blaes:02} 

1698: \textsc{Blaes, O., Lee, M.~H., \& Socrates, A.} 2002

1699: The Kozai Mechanism and the Evolution of Binary Supermassive Black Holes.

1700: \textit{Astrophys.~J.} \textbf{578}, 775--786.

1701: 

1702: \bibitem[B{\"o}ker {\it et~al.} (2004)]{Boker:04} 

1703: \textsc{B{\"o}ker, T., Sarzi, M., McLaughlin, D.~E., van der Marel, 

1704: R.~P., Rix, H.-W., Ho, L.~C., \& Shields, J.~C.} 2004

1705: A Hubble Space Telescope Census of Nuclear Star Clusters 

1706: in Late-Type Spiral Galaxies. II. 

1707: Cluster Sizes and Structural Parameter Correlations.

1708: \textit{Astron.~J.} \textbf{127}, 105--118.

1709: 

1710: \bibitem[Bogdanovi{\'c} {\it et~al.} (2007)]{BRM:07} 

1711: \textsc{Bogdanovi{\'c}, T., Reynolds, C.~S., \& Miller, M.~C.} 2007

1712: Alignment of the Spins of Supermassive Black Holes Prior to Coalescence.

1713: \textit{Astrophys.~J.~Letts.} \textbf{661}, L147--L151.

1714: 

1715: \bibitem[Bonning {\it et~al.} (2007)]{Bonning:07} 

1716: \textsc{Bonning, E.~W., Shields, G.~A., \& Salviander, S.} 2007 

1717: Recoiling black holes in quasars.

1718: \textit{ArXiv e-prints, 705, arXiv:0705.4263}.

1719: 

1720: \bibitem[Boylan-Kolchin et al.(2004)]{Boylan:04} 

1721: \textsc{Boylan-Kolchin, M., Ma, C.-P., \& Quataert, E.} 2004

1722: Core Formation in Galactic Nuclei due to Recoiling Black Holes.

1723: \textit{Astrophys.~J.} \textbf{613}, L37--L40.

1724: 

1725: \bibitem[Brown {\it et~al.} (2006)]{Brown:06} 

1726: \textsc{Brown, W.~R., Geller, M.~J., Kenyon, S.~J., \& Kurtz, M.~J.} 2006

1727: A successful targeted search for hypervelocity stars.

1728: \textit{Astrophys.~J.~Letts.} \textbf{640}, L35--L38.

1729: 

1730: \bibitem[Campanelli {\it et~al.} (2006)]{Campanelli:06} 

1731: \textsc{Campanelli, M., Lousto, C.~O., Marronetti, P., \& 

1732: Zlochower, Y.} 2006

1733: {Accurate Evolutions of Orbiting Black-Hole Binaries without Excision.}

1734: \textit{Phys.~Rev.~Letts.} \textbf{96}, 111101.

1735: 

1736: \bibitem[Campanelli {\it et~al.} (2007a)]{Campanelli:07a} 

1737: \textsc{Campanelli, M., Lousto, C.~O., Zlochower, Y., Krishnan, B., 

1738: \& Merritt, D.} 2007a

1739: Spin flips and precession in black-hole-binary mergers.

1740: \textit{Phys.~Rev.~D} \textbf{75}, 064030.

1741: 

1742: \bibitem[Campanelli {\it et~al.} (2007b)]{Campanelli:07b} 

1743: \textsc{Campanelli, M., Lousto, C., Zlochower, Y., \& Merritt, D.} 2007b

1744: Large Merger Recoils and Spin Flips from Generic Black Hole Binaries.

1745: \textit{Astrophys.~J.} \textbf{659}, L5--L8.

1746: 

1747: \bibitem[Campanelli {\it et~al.} (2007c)]{Campanelli:07c} 

1748: \textsc{Campanelli, M., Lousto, C.~O., Zlochower, Y., \& Merritt, D.} 2007c

1749: Maximum Gravitational Recoil.

1750: \textit{Phys.~Rev.~Letts.} \textbf{98}, 231102.

1751: 

1752: \bibitem[Carollo {\it et~al.} (1998)]{Carollo:98} 

1753: \textsc{Carollo, C.~M., Stiavelli, M., \& Mack, J.} 1998

1754: Spiral Galaxies with WFPC2. II. The Nuclear Properties of 40 Objects.

1755: \textit{Astron.~J.} \textbf{116}, 68--84.

1756: 

1757: \bibitem[Chandrasekhar (1943)]{Chandra:43} 

1758: \textsc{Chandrasekhar, S.} 1943

1759: Dynamical friction. I. General considerations: 

1760: the coefficient of dynamical friction.

1761: \textit{Astrophys.~J.} \textbf{97}, 255--273.

1762: 

1763: \bibitem[Christopher {\it et~al.} (2005)]{Christopher:05} 

1764: \textsc{Christopher, M.~H., Scoville, N.~Z., Stolovy, S.~R., 

1765: \& Yun, M.~S.} 2005

1766: \textit{Astrophys.~J.} \textbf{622}, 346--365.

1767: 

1768: \bibitem[Cohn \& Kulsrud (1978)]{CK:78} 

1769: \textsc{Cohn, H., \& Kulsrud, R.~M.} 1978

1770: The stellar distribution around a black hole - Numerical 

1771: integration of the Fokker-Planck equation.

1772: \textit{Astrophys.~J.} \textbf{226}, 1087--1108.

1773: 

1774: \bibitem[C{\^o}t{\'e} {\it et~al.} (2004)]{ACS1} 

1775: \textsc{C{\^o}t{\'e}, P., et al.} 2004

1776: The ACS Virgo Cluster Survey. I. Introduction to the Survey.

1777: \textit{Astrophys.~J.~Suppl.} \textbf{153}, 223--242.

1778: 

1779: \bibitem[C{\^o}t{\'e} {\it et~al.} (2006)]{ACS8} 

1780: \textsc{C{\^o}t{\'e}, P., et al.} 2006

1781: The ACS Virgo Cluster Survey. VIII. The Nuclei of Early-Type Galaxies.

1782: \textit{Astrophys.~J.~Suppl.} \textbf{165}, 57--94.

1783: 

1784: \bibitem[de Vaucouleurs (1948)]{DeVauc:48} 

1785: \textsc{de Vaucouleurs, G.} 1948

1786: Recherches sur les Nebuleuses Extragalactiques.

1787: \textit{Annales d'Astrophysique} \textbf{11}, 247--287.

1788: 

1789: \bibitem[Dokuchaev (1989)]{Dokuchaev:89} 

1790: \textsc{Dokuchaev, V.~I.} 1989

1791: The evolution of a massive black-hole in the nucleus of a normal galaxy.

1792: \textit{Sov.~Astron.~Letts.} \textbf{15}, 167--170.

1793: 

1794: \bibitem[Dotti {\it et~al.} (2007)]{Dotti:07} 

1795: \textsc{Dotti, M., Colpi, M., Haardt, F., \& Mayer, L.} 2007

1796: \textit{Mon.~Not.~R.~Astron.~Soc.} \textbf{582}, 956--962.

1797: 

1798: \bibitem[Eckart {\it et~al.} (2002)]{Eckart:02} 

1799: \textsc{Eckart, A., Genzel, R., Ott, T., \& Sch{\"o}del, R.} 2002

1800: Stellar orbits near Sagittarius A*.

1801: \textit{Mon.~Not.~R.~Astron.~Soc.} \textbf{331}, 917--934.

1802: 

1803: \bibitem[Escala {\it et~al.} (2004)]{Escala:04} 

1804: \textsc{Escala, A., Larson, R.~B., Coppi, P.~S., \& Mardones, D.} 2004

1805: The role of gas in the merging of massive black holes in 

1806: galactic nuclei. I. Black hole merging in a spherical gas cloud.

1807: \textit{Astrophys.~J.} \textbf{607}, 765--777.

1808: 

1809: \bibitem[Escala {\it et~al.} (2005)]{Escala:05} 

1810: \textsc{Escala, A., Larson, R.~B., Coppi, P.~S., \& Mardones, D.} 2005

1811: The role of gas in the merging of massive black holes in 

1812: galactic nuclei. II. Black hole merging in a nuclear gas disk.

1813: \textit{Astrophys.~J.} \textbf{630}, 152--166.

1814: 

1815: \bibitem[Fan {\it et~al.} (2003)]{Fan:03} 

1816: \textsc{Fan, X., et al.} 2003

1817: A survey of z>5.7 quasars in the Sloan digital sky survey. II. 

1818: Discovery of three additional quasars at z>6.

1819: \textit{Astron.~J.} \textbf{125}, 1649--1659.

1820: 

1821: \bibitem[Favata {\it et~al.} (2004)]{Favata:04} 

1822: \textsc{Favata, M., Hughes, S.~A., \& Holz, D.~E.} 2004

1823: How Black Holes Get Their Kicks: Gravitational Radiation 

1824: Recoil Revisited.

1825: \textit{Astrophys.~J.~Letts} \textbf{607}, L5--L8. 

1826: 

1827: \bibitem[Ferrarese {\it et~al.} (2006)]{Ferrarese:06} 

1828: \textsc{Ferrarese, L., et al.\ 2006}

1829: A Fundamental Relation between Compact Stellar Nuclei, 

1830: Supermassive Black Holes, and Their Host Galaxies.

1831: \textit{Astrophys.~J.~Letts} \textbf{644}, L21--L24. 

1832: 

1833: \bibitem[Ferrarese \& Ford (2005)]{FF:05} 

1834: \textsc{Ferrarese, L., \& Ford, H.} 2005

1835: Supermassive Black Holes in Galactic Nuclei: 

1836: Past, Present and Future Research.

1837: \textit{Sp.~Sci.~Rev.} \textbf{116}, 523--624. 

1838: 

1839: \bibitem[Fitchett (1983)]{Fitchett:83} 

1840: \textsc{Fitchett, M.~J.}1983

1841: The influence of gravitational wave momentum losses on the 

1842: centre of mass motion of a Newtonian binary system.

1843: \textit{Mon.~Not.~R.~Astron.~Soc.} \textbf{203}, 1049--1062.

1844: 

1845: \bibitem[Frank \& Rees (1976)]{FR:76} 

1846: \textsc{Frank, J., \& Rees, M.~J.} 1976

1847: Effects of massive central black holes on dense stellar systems.

1848: \textit{Mon.~Not.~R.~Astron.~Soc.} \textbf{176}, 633--647.

1849: 

1850: \bibitem[Freitag {\it et~al.} (2006)]{Freitag:06} 

1851: \textsc{Freitag, M., Amaro-Seoane, P., \& Kalogera, V.} 2006

1852: Stellar remnants in galactic nuclei: mass segregation.

1853: \textit{Astrophys.~J.} \textbf{649}, 91--117.

1854: 

1855: \bibitem[Gallimore {\it et~al.} 2001]{Gallimore:01}

1856: \textsc{Gallimore, J.~F. et al.} 2001

1857: The nature of the nuclear H2O masers of NGC 1068: Reverberation and 

1858: evidence for a rotating disk.

1859: \textit{Astrophys.~J.} \textbf{556}, 694--703.

1860: 

1861: \bibitem[Gallimore {\it et~al.} (2006)]{Gallimore:06} 

1862: \textsc{Gallimore, J.~F., Axon, D.~J., O'Dea, C.~P., Baum, S.~A., \& 

1863: Pedlar, A.} 2006

1864: A Survey of Kiloparsec-Scale Radio Outflows in Radio-Quiet Active 

1865: Galactic Nuclei.

1866: \textit{Astron.~J.} \textbf{132}, 546--569. 

1867: 

1868: \bibitem[Gebhardt {\it et~al.} (2003)]{Gebhardt:03} 

1869: \textsc{Gebhardt, K., et al.} 2003

1870: Axisymmetric Dynamical Models of the Central Regions of Galaxies.

1871: \textit{Astrophys.~J.} \textbf{583}, 92--115.

1872: 

1873: \bibitem[Gerhard (2001)]{Gerhard:01} 

1874: \textsc{Gerhard, O.} 2001

1875: The Galactic Center HE I Stars: Remains of a Dissolved Young Cluster?

1876: \textit{Astrophys.~J.} \textbf{546}, L39--L42.

1877: 

1878: \bibitem[Gonz{\'a}lez {\it et~al.} (2007a)]{Gonzalez:07a} 

1879: \textsc{Gonz{\'a}lez, J.~A., Sperhake, U., Br{\"u}gmann, B., Hannam, M., 

1880: \& Husa, S.} 2007a

1881: Maximum Kick from Nonspinning Black-Hole Binary Inspiral.

1882: \textit{Phys.~Rev.~Letts.} \textbf{98}, 091101.

1883: 

1884: \bibitem[Gonz{\'a}lez {\it et~al.} (2007b)]{Gonzalez:07b} 

1885: \textsc{Gonz{\'a}lez, J.~A., Hannam, M., Sperhake, U., Br{\"u}gmann, B., 

1886: \& Husa, S.} 2007b

1887: Supermassive Recoil Velocities for Binary Black-Hole Mergers with 

1888: Antialigned Spins.

1889: \textit{Phys.~Rev.~Letts.} \textbf{98}, 231101.

1890: 

1891: \bibitem[Governato et al. (1994)]{Governato:94} 

1892: \textsc{Governato, F., Colpi, M., \& Maraschi, L.} 1994

1893: The fate of central black holes in merging galaxies.

1894: \textit{Mon.~Not.~R.~Astron.~Soc.} \textbf{271}, 317--322.

1895: 

1896: \bibitem[Graham {\it et~al.} (2003)]{Graham:03} 

1897: \textsc{Graham, A.~W., Erwin, P., Trujillo, I., \& Asensio Ramos, A.} 2003

1898: A New Empirical Model for the Structural Analysis of Early-Type Galaxies, 

1899: and A Critical Review of the Nuker Model.

1900: \textit{Astron.~J.} \textbf{125}, 2951--2963. 

1901: 

1902: \bibitem[Graham \& Guzm{\'a}n (2003)]{GG:03} 

1903: \textsc{Graham, A.~W., \& Guzm{\'a}n, R.} 2003

1904: HST Photometry of Dwarf Elliptical Galaxies in Coma, 

1905: and an Explanation for the Alleged Structural Dichotomy 

1906: between Dwarf and Bright Elliptical Galaxies.

1907: \textit{Astron.~J.} \textbf{125}, 2936--2950.

1908: 

1909: \bibitem[Greene \& Ho (2004)]{Greene:04} 

1910: \textsc{Greene, J.~E., \& Ho, L.~C.} 2004

1911: Active galactic nuclei with candidate intermediate-mass black holes.

1912: \textit{Astrophys.~J.} \textbf{610}, 722--736.

1913: 

1914: \bibitem[Greenhill {\it et~al.} (2003)]{Greenhill:03} 

1915: \textsc{Greenhill, L.~J., et al.} 2003

1916: A warped accretion disk and wide-angle outflow 

1917: in the inner parsec of the Circinus galaxy.

1918: \textit{Astrophys.~J.} \textbf{590}, 162--173.

1919: 

1920: \bibitem[Gualandris \& Merritt (2007)]{GM:07}

1921: \textsc{Gualandris, A. \& Merritt, D.} 2007

1922: {Ejection of Supermassive Black Holes from Galaxy Cores.}

1923: \textit{ArXiv e-prints, 708, arxiv:0708.0771}.

1924: 

1925: \bibitem[Gualandris {\it et~al.} (2005)]{Gualandris:05} 

1926: \textsc{Gualandris, A., Portegies Zwart, S., \& Sipior, M.~S.} 2005

1927: Three-body encounters in the Galactic Centre: the origin of the hypervelocity star SDSS J090745.0+024507

1928: \textit{Mon.~Not.~R.~Astron.~Soc.} \textbf{363}, 223--228.

1929: 

1930: \bibitem[Haehnelt \& Kauffmann (2002)]{HK:02} 

1931: \textsc{Haehnelt, M.~G., \& Kauffmann, G.} 2002

1932: Multiple supermassive black holes in galactic bulges.

1933: \textit{Mon.~Not.~R.~Astron.~Soc.} \textbf{336}, L61-L64.

1934: 

1935: \bibitem[Haehnelt {\it et~al.} (2006)]{Haehnelt:06} 

1936: \textsc{Haehnelt, M.~G., Davies, M.~B., \& Rees, M.~J.} 2006

1937: Possible evidence for the ejection of a supermassive black hole 

1938: from an ongoing merger of galaxies.

1939: \textit{Mon.~Not.~R.~Astron.~Soc.} \textbf{366}, L22-L25.

1940: 

1941: \bibitem[Hansen \& Milosavljevi{\'c} (2003)]{Hansen:03} 

1942: \textsc{Hansen, B.~M.~S., \& Milosavljevi{\'c}, M.} 2003

1943: The need for a second black hole at the Galactic center.

1944: \textit{Astrophys.~J.} \textbf{593}, L77--L80.

1945: 

1946: \bibitem[Herrmann {\it et~al.} (2007)]{Herrmann:07} 

1947: \textsc{Herrmann, F., Hinder, I., Shoemaker, D., \& Laguna, P.} 2007

1948: Unequal mass binary black hole plunges and gravitational recoil.

1949: \textit{Class.~Quantum Gravity} \textbf{24}, 33--42. 

1950: 

1951: \bibitem[Hills (1988)]{Hills:88} 

1952: \textsc{Hills, J.~G.} 1988

1953: Hyper-velocity and tidal stars from binaries disrupted by a massive 

1954: Galactic black hole.

1955: \textit{Nature} \textbf{331}, 687--689.

1956: 

1957: \bibitem[Hoffman \& Loeb (2007)]{Hoffman:07} 

1958: \textsc{Hoffman, L., \& Loeb, A.} 2007

1959: Dynamics of triple black hole systems in hierarchically merging 

1960: massive galaxies.

1961: \textit{Mon.~Not.~R.~Astron.~Soc.} \textbf{377}, 957--976.

1962: 

1963: \bibitem[Holley-Bockelmann et al. (2002)]{HB:02} 

1964: \textsc{Holley-Bockelmann, K., Mihos, J.~C., Sigurdsson, S., Hernquist, L., \& 

1965: Norman, C.} 2002

1966: The Evolution of Cuspy Triaxial Galaxies Harboring Central Black Holes.

1967: \textit{Astrophys.~J.} \textbf{567}, 817--827.

1968: 

1969: \bibitem[Holley-Bockelmann \& Sigurdsson (2006)]{HB:06} 

1970: \textsc{Holley-Bockelmann, K., \& Sigurdsson, S.} 2006

1971: A Full Loss Cone For Triaxial Galaxies.

1972: ArXiv Astrophysics e-prints, arXiv:astro-ph/0601520.

1973: 

1974: \bibitem[Hughes (2006)]{Hughes:06} 

1975: \textsc{Hughes, S.~A.} 2006

1976: A brief survey of LISA sources and science.

1977: In \textit{Laser Interferometer Space Antenna: 6th International 

1978: LISA Symposium},

1979: AIP Conference Proceedings, vol.\ 873, pp. 13--20.

1980: 

1981: \bibitem[Ivanov {\it et al.} (1999)]{Ivanov:99} 

1982: \textsc{Ivanov, P.~B., Papaloizou, J.~C.~B., \& Polnarev, A.~G.} 1999

1983: The evolution of a supermassive binary caused by an accretion disc.

1984: \textit{Mon.~Not.~R.~Astron.~Soc.} \textbf{307}, 79--90.

1985: 

1986: \bibitem[Iwasawa {\it et~al.} (2006)]{Iwasawa:06} 

1987: \textsc{Iwasawa, M., Funato, Y., \& Makino, J.} 2006

1988: Evolution of massive black hole triples. I. Equal-mass binary-single systems.

1989: \textit{Astrophys.~J.} \textbf{651}, 1059--1067.

1990: 

1991: \bibitem[Jackson {\it et~al.} 1993]{Jackson:93}

1992: \textsc{Jackson, J.~M. et al.} 1993

1993: Neutral gas in the central 2 parsecs of the Galaxy.

1994: \textit{Astrophys.~J.} \textbf{402}, 173--182.

1995: 

1996: \bibitem[Jerjen \& Binggeli (1997)]{Jerjen:97} 

1997: \textsc{Jerjen, H., \& Binggeli, B.} 1997

1998: Are ``Dwarf'' Ellipticals Genuine Ellipticals? 

1999: In \textit{The Nature of Elliptical Galaxies; 2nd Stromlo Symposium}

2000: (ed.\ M. Arnaboldi, G. S. Da Costa, \& P. Saha)

2001: ASP Conference Series, vol.\ 116, pp.\ 239--250.

2002: 

2003: \bibitem[Kapoor (1976)]{Kapoor:76} 

2004: \textsc{Kapoor, R.~C.} 1976

2005: Ejection of massive black holes from galaxies.

2006: \textit{Pramana} \textbf{7}, 334--343. 

2007: 

2008: \bibitem[Kapoor (1983a)]{Kapoor:83a} 

2009: \textsc{Kapoor, R.~C.} 1983a

2010: On the hypothesis of ejection of supermassive black holes from 

2011: centers of galaxies and its application to quasar-galaxy associations.

2012: \textit{Astrophys.~Sp.~Sci.} \textbf{93}, 79--85.

2013: 

2014: \bibitem[Kapoor (1983b)]{Kapoor:83b} 

2015: \textsc{Kapoor, R.~C.} 1983b

2016: Some implications of escape of supermassive black holes from 

2017: galactic nuclei due to plasmoid ejection.

2018: \textit{Astrophys.~Sp.~Sci.} \textbf{95}, 425--429.

2019: 

2020: \bibitem[Kapoor (1985)]{Kapoor:85} 

2021: \textsc{Kapoor, R.~C.} 1985

2022: Effect of dynamical friction on the escape of a supermassive 

2023: black hole from a galaxy.

2024: \textit{Astrophys.~Sp.~Sci.} \textbf{112}, 347--359.

2025: 

2026: \bibitem[Kauffmann \& Haehnelt (2000)]{Kauffmann:00} 

2027: \textsc{Kauffmann, G., \& Haehnelt, M.} 2000

2028: A unified model for the evolution of galaxies and quasars.

2029: \textit{Mon.~Not.~R.~Astron.~Soc.} \textbf{311}, 576--588. 

2030: 

2031: \bibitem[Kazantzidis {\it et al.} (2005)]{Kazant:05} 

2032: \textsc{Kazantzidis, S., et al.} 2005

2033: The fate of supermassive black holes and the evolution of the 

2034: MBH-$\sigma$ relation in merging galaxies: The effect of gaseous dissipation.

2035: \textit{Astrophys.~J.~Letts.} \textbf{623}, L67--L70.

2036: 

2037: \bibitem[Kinney {\it et~al.} (2000)]{Kinney:00} 

2038: \textsc{Kinney, A.~L., Schmitt, H.~R., Clarke, C.~J., Pringle, J.~E., 

2039: Ulvestad, J.~S., \& Antonucci, R.~R.~J.} 2000

2040: Jet Directions in Seyfert Galaxies.

2041: \textit{Astrophys.~J.} \textbf{537}, 152--177.

2042: 

2043: \bibitem[Komossa (2006)]{Komossa:06} 

2044: \textsc{Komossa, S.} 2006

2045: Observational evidence for binary black holes and active double nuclei.

2046: \textit{Memorie della Societa Astronomica Italiana} \textbf{77}, 733--771. 

2047: 

2048: \bibitem[Kormendy (1985a)]{Kormendy:85a} 

2049: \textsc{Kormendy, J.} 1985a

2050: Brightness profiles of the cores of bulges and elliptical galaxies.

2051: \textit{Astrophys.~J.~Letts.} \textbf{292}, L9--L12.

2052: 

2053: \bibitem[Kormendy (1985b)]{Kormendy:85b} 

2054: \textsc{Kormendy, J.} 1985b

2055: Families of ellipsoidal stellar systems and the formation 

2056: of dwarf elliptical galaxies.

2057: \textit{Astrophys.~J.} \textbf{295}, 73--79.

2058: 

2059: \bibitem[Kormendy \& Djorgovski (1989)]{KD:89} 

2060: \textsc{Kormendy, J., \& Djorgovski, S.} 1989

2061: Surface photometry and the structure of elliptical galaxies.

2062: \textit{Ann.~Rev.~Astron.~Ap.} \textbf{27}, 235--277.

2063: 

2064: \bibitem[Lauer {\it et~al.} (1993)]{Lauer:93} 

2065: \textsc{Lauer, T.~R., et al.} 1993

2066: Planetary camera observations of the double nucleus of M31.

2067: \textit{Astron.~J.} \textbf{106}, 1436--1447. 

2068: 

2069: \bibitem[Lauer {\it et~al.} (1996)]{Lauer:96} 

2070: \textsc{Lauer, T.~R., et al.} 1996

2071: Hubble Space Telescope Observations of the Double Nucleus of NGC 4486B.

2072: \textit{Astrophys.~J.~Letts.} \textbf{471}, L79--L82.

2073: 

2074: \bibitem[Lauer {\it et~al.} (2002)]{Lauer:02} 

2075: \textsc{Lauer, T.~R., et al.} 2002

2076: Galaxies with a central minimum in stellar luminosity density.

2077: \textit{Astron.~J.} \textbf{124}, 1975--1987.

2078: 

2079: \bibitem[Lauer {\it et~al.} (2005)]{Lauer:05} 

2080: \textsc{Lauer, T.~R., et al.} 2005

2081: The Centers of Early-Type Galaxies with Hubble Space Telescope. 

2082: V. New WFPC2 Photometry.

2083: \textit{Astron.~J.} \textbf{129}, 2138--2185.

2084: 

2085: \bibitem[Lauer {\it et~al.} (2007)]{Lauer:07} 

2086: \textsc{Lauer, T.~R., et al.} 2007

2087: The Masses of Nuclear Black Holes in Luminous Elliptical Galaxies and 

2088: Implications for the Space Density of the Most Massive Black Holes.

2089: \textit{Astron.~J.} \textbf{662}, 808--834. 

2090: 

2091: \bibitem[Levin (2006)]{Levin:06} 

2092: \textsc{Levin, Y.} 2006

2093: Ejection of high-velocity stars from the Galactic center 

2094: by an inspiraling intermediate-mass black hole.

2095: \textit{Astrophys.~J.} \textbf{653}, 1203--1209.

2096: 

2097: \bibitem[Macfadyen \& Milosavljevi{\'c} (2006)]{Milos:06} 

2098: \textsc{Macfadyen, A.~I., \& Milosavljevi{\'c}, M.} 2006 

2099: ArXiv Astrophysics e-prints, arXiv:astro-ph/0607467 

2100: 

2101: \bibitem[Madau \& Quataert (2004)]{MQ:04} 

2102: \textsc{Madau, P., \& Quataert, E.} 2004

2103: The Effect of Gravitational-Wave Recoil on the Demography of Massive 

2104: Black Holes.

2105: \textit{Astrophys.~J.~Letts.} \textbf{606}, L17--L20.

2106: 

2107: \bibitem[Makino (1997)]{Makino:97} 

2108: \textsc{Makino, J.} 1997

2109: Merging of Galaxies with Central Black Holes. II. 

2110: Evolution of the Black Hole Binary and the Structure of the Core.

2111: \textit{Astrophys.~J.} \textbf{478}, 58--65. 

2112: 

2113: \bibitem[Makino \& Ebisuzaki (1994)]{Makino:94} 

2114: \textsc{Makino, J., \& Ebisuzaki, T.} 1994

2115: Triple black holes in the cores of galaxies.

2116: \textit{Astrophys.~J.} \textbf{446}, 607--610.

2117: 

2118: \bibitem[Makino \& Funato (2004)]{Makino:04} 

2119: \textsc{Makino, J., \& Funato, Y.} 2004

2120: Evolution of Massive Black Hole Binaries.

2121: \textit{Astrophys.~J.} \textbf{602}, 93--102.

2122: 

2123: \bibitem[Marchant \& Shapiro (1980)]{MS:80} 

2124: \textsc{Marchant, A.~B., \& Shapiro, S.~L.} 1980

2125: Star clusters containing massive, central black holes. 

2126: III - Evolution calculations.

2127: \textit{Astrophys.~J.} \textbf{239}, 685--704.

2128: 

2129: \bibitem[Marconi {\it et~al.} (2004)]{Marconi:04} 

2130: \textsc{Marconi, A., Risaliti, G., Gilli, R., Hunt, L.~K., Maiolino, R., 

2131: \& Salvati, M.} 2004

2132: Local supermassive black holes, relics of active galactic nuclei 

2133: and the X-ray background.

2134: \textit{Mon.~Not.~R.~Astron.~Soc.} \textbf{351}, 169--185.

2135: 

2136: \bibitem[Matsubayashi {\it et~al.} (2007)]{Matsu:07} 

2137: \textsc{Matsubayashi, T., Makino, J., \& Ebisuzaki, T.} 2007

2138: Orbital evolution of an IMBH in the Galactic nucleus with a 

2139: massive central black hole.

2140: \textit{Astrophys.~J.} \textbf{656}, 879--896.

2141: 

2142: \bibitem[Mayer {\it et~al.} (2007)]{Mayer:07} 

2143: \textsc{Mayer, L., Kazantzidis, S., Madau, P., Colpi, M., Quinn, T., \& 

2144: Wadsley, J.} 2007

2145: Rapid Formation of Supermassive Black Hole Binaries in Galaxy Mergers with Gas.

2146: \textit{Science} \textbf{316}, 1874--1876. 

2147: 

2148: \bibitem[McLaughlin {\it et~al.} (2006)]{King:06} 

2149: \textsc{McLaughlin, D.~E., King, A.~R., \& Nayakshin, S.} 2006

2150: The $M-\sigma$ relation for nucleated galaxies.

2151: \textit{Astrophys.~J.~Letts.} \textbf{650}, L37--L40.

2152: 

2153: \bibitem[Merritt (1984)]{Merritt:84} 

2154: \textsc{Merritt, D.} 1984

2155: Relaxation and tidal stripping in rich clusters of galaxies. 

2156: II. Evolution of the luminosity distribution.

2157: \textit{Astrophys.~J.} \textbf{276}, 26--37.

2158: 

2159: \bibitem[Merritt (2000)]{Merritt:00} 

2160: \textsc{Merritt, D.} 2000

2161: Black Holes and Galaxy Evolution.

2162: In \textit{Dynamics of Galaxies: from the Early Universe to the Present}

2163: (ed.\ F. Combes, G. A. Mamon \& V. Charmandaris)

2164: ASP Conference Series, vol.\ 197, pp. 221--230.

2165: 

2166: \bibitem[Merritt (2002)]{Merritt:02} 

2167: \textsc{Merritt, D.} 2002

2168: Rotational Brownian motion of a massive binary.

2169: \textit{Astrophys.~J.} \textbf{568}, 998--1003.

2170: 

2171: \bibitem[Merritt (2004)]{Merritt:04} 

2172: \textsc{Merritt, D.} 2004

2173: Evolution of the dark matter distribution at the galactic center.

2174: \textit{Phys.~Rev.~Letts.} \textbf{92}, 201304.

2175: 

2176: \bibitem[Merritt(2006a)]{Merritt:06} 

2177: \textsc{Merritt, D.} 2006a

2178: Mass Deficits, Stalling Radii, and the Merger Histories of Elliptical Galaxies.

2179: \textit{Astrophys.~J.} \textbf{648}, 976--986.

2180: 

2181: \bibitem[Merritt (2006b)]{ROP:06} 

2182: \textsc{Merritt, D.} 2006b

2183: Dynamics of galaxy cores and supermassive black holes.

2184: \textit{Rept. Prog. Phys.} \textbf{69}, 2513--2579.

2185: 

2186: \bibitem[Merritt et al.(2007)]{MBL:07} 

2187: \textsc{Merritt, D., Berczik, P., \& Laun, F.} 2007

2188: Brownian Motion of Black Holes in Dense Nuclei.

2189: \textit{Astron.~J.} \textbf{133}, 553--563.

2190: 

2191: \bibitem[Merritt \& Cruz (2001)]{MC:01} 

2192: \textsc{Merritt, D., \& Cruz, F.} 2001

2193: Cusp Disruption in Minor Mergers.

2194: \textit{Astrophys.~J.~Letts.} \textbf{551}, L41-L44.

2195: 

2196: \bibitem[Merritt et al. (2006b)]{Haloes:06} 

2197: \textsc{Merritt, D., Graham, A.~W., Moore, B., Diemand, J., 

2198: \& Terzi{\'c}, B.} 2006

2199: Empirical Models for Dark Matter Halos. I. Nonparametric 

2200: Construction of Density Profiles and Comparison with Parametric Models.

2201: \textit{Astron.~J.} \textbf{132}, 2685--2700.

2202: 

2203: \bibitem[Merritt {\it et~al.} (2007)]{MMS:07} 

2204: \textsc{Merritt, D., Mikkola, S., \& Szell, A.} 2007 

2205: Long-Term Evolution of Massive Black Hole Binaries. III. 

2206: Binary Evolution in Collisional Nuclei.

2207: \textit{ArXiv e-prints, 705, arXiv:0705.2745}.

2208: 

2209: \bibitem[Merritt {\it et~al.} (2004)]{MMFHH:04} 

2210: \textsc{Merritt, D., Milosavljevi{\'c}, M., Favata, M., Hughes, S.~A., 

2211: \& Holz, D.~E.} 2004

2212: Consequences of Gravitational Radiation Recoil.

2213: \textit{Astrophys.~J.~Letts.} \textbf{607}, L9--L12.

2214: 

2215: \bibitem[Merritt {\it et~al.} (2005)]{Universal:05} 

2216: \textsc{Merritt, D., Navarro, J.~F., Ludlow, A., \& Jenkins, A.} 2005

2217: A Universal Density Profile for Dark and Luminous Matter?

2218: \textit{Astrophys.~J.~Letts.} \textbf{624}, L85--L88.

2219: 

2220: \bibitem[Merritt \& Poon (2004)]{Poon:04b} 

2221: \textsc{Merritt, D., \& Poon, M.~Y.} 2004

2222: Chaotic loss cones and black hole fueling.

2223: \textit{Astrophys.~J.} \textbf{606}, 788--798.

2224: 

2225: \bibitem[Merritt {\it et~al.} (2006a)]{Naked:06} 

2226: \textsc{Merritt, D., Storchi-Bergmann, T., Robinson, A., Batcheldor, D., 

2227: Axon, D., \& Cid Fernandes, R.} 2006a

2228: The nature of the HE0450-2958 system.

2229: \textit{Mon.~Not.~R.~Astron.~Soc.} \textbf{367}, 1746--1750.

2230: 

2231: \bibitem[Merritt \& Szell (2006)]{MS:06} 

2232: \textsc{Merritt, D., \& Szell, A.} 2006

2233: Dynamical cusp regeneration.

2234: \textit{Astrophys.~J.} \textbf{648}, 890--899. 

2235: 

2236: \bibitem[Merritt \& Wang (2005)]{Wang:05} 

2237: \textsc{Merritt, D., \& Wang, J.} 2005

2238: Loss cone refilling rates in galactic nuclei.

2239: \textit{Astrophys.~J.~Letts.} \textbf{621}, L101--L104.

2240: 

2241: \bibitem[Mihos \& Hernquist (1994)]{Mihos:94} 

2242: \textsc{Mihos, J.~C., \& Hernquist, L.} 1994

2243: Dense stellar cores in merger remnants.

2244: \textit{Astrophys.~J.~Letts.} \textbf{437}, L47--L50. 

2245: 

2246: \bibitem[Mikkola \& Valtonen (1992)]{MV:92} 

2247: \textsc{Mikkola, S., \& Valtonen, M.~J.} 1992

2248: Evolution of binaries in the field of light particles and the 

2249: problem of two black holes.

2250: \textit{Mon.~Not.~R.~Astron.~Soc.} \textbf{259}, 115--120. 

2251: 

2252: \bibitem[Milosavljevi{\'c} \& Merritt (2001)]{MM:01} 

2253: \textsc{Milosavljevi{\'c}, M., \& Merritt, D.} 2001

2254: Formation of galactic nuclei.

2255: \textit{Astrophys.~J.} \textbf{563}, 34--62.

2256: 

2257: \bibitem[Milosavljevi{\'c} \& Merritt (2003)]{MM:03} 

2258: \textsc{Milosavljevi{\'c}, M., \& Merritt, D.} 2003

2259: Long-term evolution of massive black hole binaries.

2260: \textit{Astrophys.~J.} \textbf{596}, 860--878.

2261: 

2262: \bibitem[Milosavljevi{\'c} {\it et~al.} (2006)]{milos:06} 

2263: \textsc{Milosavljevi{\'c}, M., Merritt, D., \& Ho, L.~C.} 2006

2264: Contribution of stellar tidal disruptions to the X-ray 

2265: luminosity function of active galaxies.

2266: \textit{Astrophys.~J.} \textbf{652}, 120--125.

2267: 

2268: \bibitem[Milosavljevi{\'c} {\it et~al.} (2002)]{Milos:02} 

2269: \textsc{Milosavljevi{\'c}, M., Merritt, D., Rest, A., \& van den Bosch, F.~C.} 

2270: 2002

2271: Galaxy cores as relics of black hole mergers.

2272: \textit{Mon.~Not.~R.~Astron.~Soc.} \textbf{331}, L51--L54.

2273: 

2274: \bibitem[Murphy {\it et~al.} (1991)]{MCD:91} 

2275: \textsc{Murphy, B.~W., Cohn, H.~N., \& Durisen, R.~H.} 1991

2276: Dynamical and luminosity evolution of active galactic nuclei - 

2277: Models with a mass spectrum.

2278: \textit{Astrophys.~J.} \textbf{370}, 60--77.

2279: 

2280: \bibitem[Navarro {\it et~al.} (2004)]{Navarro:04} 

2281: \textsc{Navarro, J.~F., et al.} 2004, 

2282: The inner structure of LCDM haloes - III. Universality and asymptotic slopes.

2283: \textit{Mon.~Not.~R.~Astron.~Soc.} \textbf{349}, 1039--1051. 

2284: 

2285: \bibitem[Perets {\it et al.} (2007)]{Perets:07a} 

2286: \textsc{Perets, H.~B., Hopman, C., \& Alexander, T.} 2007

2287: Massive perturber-driven interactions between stars and a massive black hole.

2288: \textit{Astrophys.~J.} \textbf{656}, 709--720.

2289: 

2290: \bibitem[Perets \& Alexander (2007)]{Perets:07b} 

2291: \textsc{Perets, H.~B., \& Alexander, T.} 2007

2292: Massive perturbers and the efficient merger of binary massive black holes.

2293: ArXiv e-prints, 705, arXiv:0705.2123 

2294: 

2295: \bibitem[Poon \& Merritt (2001)]{Poon:01} 

2296: \textsc{Poon, M.~Y., \& Merritt, D.} 2001

2297: Orbital Structure of Triaxial Black Hole Nuclei.

2298: \textit{Astrophys.~J.} \textbf{549}, 192--204.

2299: 

2300: \bibitem[Poon \& Merritt (2004)]{Poon:04a} 

2301: \textsc{Poon, M.~Y., \& Merritt, D.} 2004

2302: A self-consistent study of triaxial black hole nuclei.

2303: \textit{Astrophys.~J.} \textbf{606}, 774--787.

2304: 

2305: \bibitem[Portegies Zwart \& McMillan (2002)]{Zwart:02} 

2306: \textsc{Portegies Zwart, S.~F., \& McMillan, S.~L.~W.} 2002

2307: The runaway growth of intermediate-mass black holes in dense star clusters.

2308: \textit{Astrophys.~J.} \textbf{576}, 899--907.

2309: 

2310: \bibitem[Pretorius (2005)]{Pretorius:05} 

2311: \textsc{Pretorius, F.} 2005 

2312: Evolution of Binary Black-Hole Spacetimes.

2313: \textit{Phys.~Rev.~Letts.} \textbf{95}, 121101.

2314: 

2315: \bibitem[Pringle (1991)]{Pringle:91} 

2316: \textsc{Pringle, J.~E.} 1991

2317: The properties of external accretion discs.

2318: \textit{Mon.~Not.~R.~Astron.~Soc.} \textbf{248}, 754--759.

2319: 

2320: \bibitem[Quinlan (1996)]{Quinaln:96} 

2321: \textsc{Quinlan, G.~D.} 1996

2322: The dynamical evolution of massive black hole binaries I. 

2323: Hardening in a fixed stellar background.

2324: \textit{New Astron.} \textbf{1}, 35--56.

2325: 

2326: \bibitem[Ravindranath et al. (2001)]{Ravin:01} 

2327: \textsc{Ravindranath, S., Ho, L.~C., Peng, C.~Y., Filippenko, A.~V., 

2328: \& Sargent, W.~L.~W.} 2001

2329: Central Structural Parameters of Early-Type Galaxies as Viewed with 

2330: Nicmos on the Hubble Space Telescope.

2331: \textit{Astron.~J.} \textbf{122}, 653--678.

2332: 

2333: \bibitem[Ravindranath {\it et~al.} (2002)]{Ravin:02} 

2334: \textsc{Ravindranath, S., Ho, L.~C., \& Filippenko, A.~V.} 2002

2335: Nuclear cusps and cores in early-type galaxies as relics of 

2336: binary black hole mergers.

2337: \textit{Astrophys.~J.} \textbf{566}, 801--808.

2338: 

2339: \bibitem[Redmount \& Rees(1989)]{Redmount:89} 

2340: \textsc{Redmount, I.~H., \& Rees, M.~J.} 1989 

2341: Gravitational-radiation rocket effects and galactic structure.

2342: \textit{Comm.~Astrophys.} \textbf{14}, 165--178.

2343: 

2344: \bibitem[Reid \& Brunthaler (2004)]{Reid:04} 

2345: \textsc{Reid, M.~J., \& Brunthaler, A.} 2004

2346: The proper motion of Sagittarius A*. II. The mass of Sagittarius A*.

2347: \textit{Astrophys.~J.} \textbf{616}, 872--884.

2348: 

2349: \bibitem[Rodriguez {\it et~al.} (2006)]{Rodriguez:06} 

2350: \textsc{Rodriguez, C., Taylor, G.~B., Zavala, R.~T., Peck, A.~B., 

2351: Pollack, L.~K., \& Romani, R.~W.} 2006

2352: A Compact Supermassive Binary Black Hole System.

2353: \textit{Astrophys.~J.} \textbf{646}, 623--632.

2354: 

2355: \bibitem[Sakamoto {\it et~al.} (1999)]{Sakamoto:99} 

2356: \textsc{Sakamoto, K., Okumura, S.~K., Ishizuki, S., \& Scoville, N.~Z.} 1999

2357: CO images of the central regions of 20 nearby spiral galaxies.

2358: \textit{Astrophys.~J.~Suppl.} \textbf{124}, 403--437.

2359: 

2360: \bibitem[Sch{\"o}del {\it et~al.} (2007)]{Schoedel:07} 

2361: \textsc{Sch{\"o}del, R., et al.} 2007

2362: The structure of the nuclear stellar cluster of the Milky Way.

2363: \textit{Astron.~Ap.} \textbf{469}, 125--146.

2364: 

2365: \bibitem[S\'ersic (1968)]{Sersic:68} 

2366: \textsc{S\'ersic, J.~L.} 1968

2367: Atlas de Galaxies Australes.

2368: Cordoba: Observatorio Astronomico.

2369: 

2370: \bibitem[Sesana {\it et~al.} (2006)]{Sesana:06} 

2371: \textsc{Sesana, A., Haardt, F., \& Madau, P.} 2006

2372: Interaction of massive black hole binaries with their stellar environment. 

2373: I. Ejection of hypervelocity stars

2374: \textit{Astrophys.~J.} \textbf{651}, 392--400.

2375: 

2376: \bibitem[Sesana {\it et~al.} (2007)]{Sesana:07} 

2377: \textsc{Sesana, A., Haardt, F., \& Madau, P.} 2007

2378: Interaction of massive black hole binaries with their stellar environment. 

2379: II. Loss cone depletion and binary orbital decay.

2380: \textit{Astrophys.~J.} \textbf{660}, 546--555.

2381: 

2382: \bibitem[Sesana {\it et~al.} (2004)]{Sesana:04} 

2383: \textsc{Sesana, A., Haardt, F., Madau, P., \& Volonteri, M.} 2004

2384: Low-frequency gravitational radiation from coalescing massive 

2385: black hole binaries in hierarchical cosmologies.

2386: \textit{Astrophys.~J.} \textbf{611}, 623--632.

2387: 

2388: \bibitem[Shapiro \& Marchant (1978)]{SM:78} 

2389: \textsc{Shapiro, S.~L., \& Marchant, A.~B.} 1978

2390: Star clusters containing massive, central black holes - 

2391: Monte Carlo simulations in two-dimensional phase space

2392: \textit{Astrophys.~J.} \textbf{225}, 603--624.

2393: 

2394: \bibitem[Shlosman et al. (1990)]{Shlosman:90} 

2395: \textsc{Shlosman, I., Begelman, M.~C., \& Frank, J.} 1990

2396: The fuelling of active galactic nuclei.

2397: \textit{Nature} \textbf{345}, 679--686.

2398: 

2399: \bibitem[Spitzer (1987)]{Spitzer:87} 

2400: \textsc{Spitzer, L.} 1987

2401: Dynamical Evolution of Globular  Clusters.

2402: Princeton University Press.

2403: 

2404: \bibitem[Tichy \& Marronetti (2007)]{Tichy:07} 

2405: \textsc{Tichy, W., \& Marronetti, P.} 2007 

2406: Binary black hole mergers: large kicks for generic spin orientations.

2407: ArXiv e-prints, arXiv:gr-qc/0703075 

2408: 

2409: \bibitem[Tremaine (1995)]{Tremaine:95} 

2410: \textsc{Tremaine, S.} 1995

2411: An Eccentric-Disk Model for the Nucleus of M31.

2412: \textit{Astron.~J.} \textbf{110}, 628--633. 

2413: 

2414: \bibitem[Tremaine (1997)]{Tremaine:97}

2415: \textsc{Tremaine, S.} 1997

2416: The centers of elliptical galaxies.

2417: In. \textit{Unsolved Problems in Astrophysics}

2418: (ed. J.N.~Bahcall \& J.P.~Ostriker),

2419: pp.~137--158. Princeton University Press.

2420: 

2421: \bibitem[Tremaine et al. (2002)]{Tremaine:02} 

2422: \textsc{Tremaine, S., et al.} 2002

2423: The Slope of the Black Hole Mass versus Velocity Dispersion Correlation.

2424: \textit{Astrophys.~J.} \textbf{574}, 740--753.

2425: 

2426: \bibitem[Trujillo {\it et~al.} (2004)]{Trujillo:04} 

2427: \textsc{Trujillo, I., Erwin, P., Asensio Ramos, A., \& Graham, A.~W.} 2004

2428: Evidence for a New Elliptical-Galaxy Paradigm: Sersic and Core Galaxies.

2429: \textit{Astron.~J.} \textbf{127}, 1917--1942.

2430: 

2431: \bibitem[Valtonen {\it et~al.} (1994)]{Valtonen:94} 

2432: \textsc{Valtonen, M.~J., Mikkola, S., Heinamaki, P., \& Valtonen, H.} 1994

2433: Slingshot ejections from clusters of three and four black holes.

2434: \textit{Astrophys.~J.~Suppl.} \textbf{95}, 69--86.

2435: 

2436: \bibitem[Valtonen (1996)]{Valtonen:96} 

2437: \textsc{Valtonen, M.~J.} 1996, 

2438: Are supermassive black holes confined to galactic nuclei?

2439: \textit{Comm.~Astrophys.} \textbf{18}, 191--206.

2440: 

2441: \bibitem[Vicari {\it et~al.} (2007)]{Vicari:07} 

2442: \textsc{Vicari, A., Capuzzo-Dolcetta, R., \& Merritt, D.} 2007

2443: Consequences of Triaxiality for Gravitational Wave Recoil of Black Holes.

2444: \textit{Astrophys.~J.~Letts.} \textbf{662}, 797--807.

2445: 

2446: \bibitem[Volonteri {\it et~al.} (2003)]{Volonteri:03} 

2447: \textsc{Volonteri, M., Haardt, F., \& Madau, P.} 2003

2448: The assembly and merging history of supermassive black holes in 

2449: hierarchical models of galaxy formation.

2450: \textit{Astrophys.~J.} \textbf{582}, 559--573.

2451: 

2452: \bibitem[Wehner \& Harris (2006)]{Wehner:06} 

2453: \textsc{Wehner, E.~H., \& Harris, W.~E.} 2006

2454: From Supermassive Black Holes to Dwarf Elliptical Nuclei: A Mass Continuum.

2455: \textit{Astrophys.~J.~Letts.} \textbf{644}, L17--L20.

2456: 

2457: \bibitem[Young(1977)]{Young:77} 

2458: \textsc{Young, P.~J.} 1977

2459: The black tide model of QSOs. II - Destruction in an isothermal sphere.

2460: \textit{Astrophys.~J.} \textbf{215}, 36--52.

2461: 

2462: \bibitem[Yu (2002)]{Yu:02} 

2463: \textsc{Yu, Q.} 2002

2464: Evolution of massive binary black holes.

2465: \textit{Mon.~Not.~R.~Astron.~Soc.} \textbf{331}, 935--958. 

2466: 

2467: \bibitem[Yu \& Tremaine (2003)]{Yu:03} 

2468: \textsc{Yu, Q., \& Tremaine, S.} 2003

2469: Ejection of hypervelocity stars by the (binary) black hole in the 

2470: Galactic center.

2471: \textit{Astrophys.~J.} \textbf{599}, 1129--1138.

2472: 

2473: \bibitem[Zhao {\it et~al.} (2002)]{Zhao:02} 

2474: \textsc{Zhao, H., Haehnelt, M.~G., \& Rees, M.~J.} 2002

2475: Feeding black holes at galactic centres by capture from isothermal cusps.

2476: \textit{New Astron.} \textbf{7}, 385--394.

2477: 

2478: \bibitem[Zier (2006)]{Zier:06} 

2479: \textsc{Zier, C.} 2006

2480: Merging of a massive binary due to ejection of bound stars.

2481: \textit{Mon.~Not.~R.~Astron.~Soc.} \textbf{371}, L36--L40.

2482: 

2483: \bibitem[Zier (2007)]{Zier:07} 

2484: \textsc{Zier, C.} 2007

2485: Merging of a massive binary due to ejection of bound stars - II.

2486: \textit{Mon.~Not.~R.~Astron.~Soc.} \textbf{378}, 1309--1327.

2487: 

2488: \end{thebibliography}

2489: 

2490: \end{document}

2491: 

2492: