1: \documentclass[useAMS,usenatbib]{mn2e}
2:
3: \usepackage{latexsym,graphicx,natbib}
4:
5:
6:
7: %%%% symbols definitions
8: \newcommand\simless{\mathbin{\lower 3pt\hbox
9: {$\rlap{\raise 5pt\hbox{$\char'074$}}\mathchar"7218$}}} %< or of order
10: \newcommand\simgreat{\mathbin{\lower 3pt\hbox
11: {$\rlap{\raise 5pt\hbox{$\char'076$}}\mathchar"7218$}}} %> or of order
12:
13: \def\apgt{\ {\raise-.5ex\hbox{$\buildrel>\over\sim$}}\ }
14: \def\aplt{\ {\raise-.5ex\hbox{$\buildrel<\over\sim$}}\ }
15:
16: %%%% units & names definitions
17: \newcommand\kms{{\rm\,km\,s^{-1}}}
18: \newcommand\kpc{{\rm\,kpc}}
19: \newcommand\masyr{\rm\,mas\,{yr^{-1}}}
20: \newcommand\msun{\rm\,M_\odot}
21: \newcommand\rsun{\rm\,R_\odot}
22: \newcommand\mbh{M_{\rm bh}}
23: \newcommand\mtot{M_{\rm tot}}
24: \newcommand\he{HE\,0437--5439}
25: \newcommand\rmin{R_{\rm min}}
26: \newcommand\rpmin{R^{'}_{\rm min}}
27: \newcommand\rcore{r_{\rm c}}
28: \newcommand\rhm{r_{\rm h}}
29: \newcommand\trlx{T_{\rm rlx}}
30: \newcommand\pcc{\rm\,pc^{-3}}
31:
32: %%%% journals definitions
33: \def\aj{AJ} %% Astronomical Journal
34: \def\apj{ApJ} %% Astrophysical Journal
35: \def\apjl{ApJL} %% Astrophysical Journal, Letters
36: \def\apjs{ApJS} %% Astrophysical Journal, Supplement
37: \def\aap{A\&A} %% Astronomy and Astrophysics
38: \def\aapr{A\&A~Rev.} %% Astronomy and Astrophysics Reviews
39: \def\aaps{A\&AS} %% Astronomy and Astrophysics, Supplement
40: \def\mnras{MNRAS} %% Monthly Notices of the RAS
41: \def\pasp{PASP} %% Publications of the ASP
42: \def\pasj{PASJ} %% Publications of the ASJ
43: \def\nat{Nature} %% Nature
44: \def\iaucirc{IAU~Circ.} %% IAU Cirulars
45: \def\bain{Bull.~Astron.~Inst.~Netherlands} %% Bulletin Astronomical Institute of the Netherlands
46:
47:
48:
49: \title[A hypervelocity star from the LMC]
50: {A hypervelocity star from the Large Magellanic Cloud}
51:
52: \author[Gualandris and Portegies Zwart]
53: {
54: Alessia Gualandris $^{1}$\thanks{E-mail: alessiag@science.uva.nl}
55: and Simon Portegies Zwart$^{1}$\\
56: $^1$ Astronomical Institute 'Anton Pannekoek'
57: and Section Computational Science,
58: University of Amsterdam,
59: the Netherlands \\
60: }
61:
62:
63: \begin{document}
64:
65:
66:
67: \maketitle
68:
69:
70: \begin{abstract}
71: We study the acceleration of the star \he\, to hypervelocity and
72: discuss its possible origin in the Large Magellanic Cloud (LMC). The
73: star has a radial velocity of 723$\kms$ and is located at a distance
74: of 61$\kpc$ from the Sun. With a mass of about 8$\msun$, the travel
75: time from the Galactic centre is of about 100\,Myr, much longer than
76: its main sequence lifetime. Given the relatively small distance to
77: the LMC (18$\kpc$), we consider it likely that \he\, originated in the
78: cloud rather than in the Galactic centre, like the other hypervelocity
79: stars. The minimum ejection velocity required to travel from the LMC
80: to its current location within its lifetime is about 500$\kms$.
81: Such a high velocity can only be obtained in a dynamical encounter
82: with a massive black hole. We perform 3-body scattering simulations in
83: which a stellar binary encounters a massive black hole and find that a
84: black hole more massive than $10^3\msun$ is necessary to explain the
85: high velocity of \he. We look for possible parent clusters for \he\, and
86: find that NGC\,2100 and NGC\,2004 are young enough to host stars
87: coeval to \he\, and dense enough to produce an intermediate mass black
88: hole able to eject an 8$\msun$ star with hypervelocity.
89: %We argue that the most likely birth place for
90: %\he\, in the LMC is the star cluster NGC\,330, which is young enough
91: %to host stars coeval to \he\, and dense enough to produce an
92: %intermediate mass black hole able to eject an 8$\msun$ star with
93: %hypervelocity.
94: \end{abstract}
95:
96:
97: \begin{keywords}
98: Stars: black-holes -- Stars: individual (HE\,0437-5439) --
99: Methods: N-body simulations -- Binaries: dynamics.
100: \end{keywords}
101:
102:
103: \section{Introduction}
104: \label{sec:intro}
105: Recent radial velocity measurements of young stars in the Galactic
106: halo has led to the discovery of a new class of stars, the so-called
107: {\it hypervelocity stars} (HVS), which travel with velocities
108: exceeding the escape speed of the Galaxy \citep{b05,b06}. A total of
109: 7 HVSs has been discovered so far, of which six are consistent with
110: an ejection from the centre of the Milky Way.
111:
112: The existence of a population of HVSs was first proposed by Hills
113: (1988), who considered hypervelocity ejections as a natural
114: consequence of galaxies hosting supermassive black holes (SMBHs). The
115: interaction between a SMBH and a stellar binary can result in the
116: dynamical capture of one of the binary components at the expense of
117: the high velocity ejection of its companion star
118: \citep{h88,yt03,gps05}. Later an alternative was proposed by
119: \citet{bgp06}, who argue that an intermediate-mass black hole (IMBH)
120: can eject stars with extremely high velocities as it spirals in
121: towards the Galactic centre. The latter mechanism tends to produce
122: HVSs in bursts which last a few Myr \citep{bgp06}. Since the
123: reconstructed ejection times of the observed HVSs are roughly evenly
124: distributed over about 100\,Myr \citep{b06}, it seems likely that they
125: were ejected in a continuous process. For 6 of the observed HVSs, in
126: fact, the travel time from the Galactic centre and the estimated age
127: are consistent and, by assuming a suitable proper motion, it is
128: possible to calculate a realistic trajectory.
129:
130: For \he, however, an ejection from the Galactic centre appears
131: problematic as the time required to travel from the centre to its
132: observed location exceeds the maximum lifetime of the star. \he\, is
133: a main-sequence star with spectral type B and a mass of about
134: 8.4$\msun$ for a solar metallicity or of about 8$\msun$ for a $Z =
135: 0.008$ metallicity, which corresponds to main-sequence lifetimes of 25
136: and 35\,Myr, respectively. Adopting a distance of 61$\kpc$ from the
137: Sun \citep{e05} and considering a total space velocity equal to the
138: radial velocity only, the travel time for a straight orbit from the
139: Galactic centre is of about 80\,Myr, much longer than the
140: main-sequence lifetime of the star. The estimated proper motion for
141: which the orbit of the star comes within 10\,pc from the centre is of
142: about 0.55$\masyr$ \citep{e05}. This results in a tangential velocity
143: of about 160$\kms$ and hence in a total space velocity of 740$\kms$,
144: which corresponds to a travel time of 100\,Myr for a realistic orbit.
145:
146: We discard the possibility that \he\, is a blue straggler originated
147: in the Galactic centre since, even in the case of a merger between two
148: less massive stars, the merger product would not survive as an
149: 8$\msun$ star for much longer than the main-sequence lifetime of such
150: a star.
151:
152: Given the fact that \he\, is much closer to the LMC galaxy than to the
153: Milky Way, \citet{e05} suggest that \he\, might have been ejected from
154: the LMC. Considering a total distance of 18$\kpc$ from the LMC
155: centre, the minimum ejection velocity required to travel to the
156: current position is about 500$\kms$. (Here we adopted a travel time
157: of 35\,Myr, which is equal to the main-sequence lifetime in a low
158: metallicity environment.) The radial velocity of \he\, relative to the
159: LMC is 461$\kms$, and the required tangential velocity is then
160: 160$\kms$, which in turns implies a proper motion of 1.9$\masyr$. A
161: velocity of the order of 500$\kms$ can only be obtained in a dynamical
162: interaction with a massive black hole (see \citet{gps05}).
163:
164: If \he\, was ejected by an IMBH and if it came from the LMC, the
165: natural consequence is that there must then be an IMBH in the LMC. We
166: investigate this hypothesis by studying the dynamical ejection of an
167: 8$\msun$ main-sequence star in an interaction with a massive
168: black hole. We simulate the encounter between a stellar binary and a
169: black hole to study the minimum mass of the black hole required to
170: accelerate one of the interacting stars to a velocity of at least
171: 500$\kms$. We then investigate where in the LMC such an interaction
172: could have taken place.
173:
174: We find that a black hole of $\apgt 10^3\msun$ is required to explain
175: the velocity of \he. Such an IMBH could form in a young and dense
176: star cluster \citep{p04} and would help driving the frequent strong
177: encounters with massive stars needed to make high-velocity ejections
178: likely \citep{2006MNRAS.372..467B}. The IMBH must still be present in
179: a star cluster which contains stars as massive as \he. In addition,
180: the parent star cluster must have been massive enough and with a short
181: enough relaxation time to produce an IMBH \citep{p04}. The most likely
182: clusters are NGC\,2100 and NGC\,2004.
183:
184:
185:
186: \section{Three-body scatterings with a massive black hole in the LMC}
187: \label{sec:scatter}
188:
189: In this section we explore the hypothesis of a dynamical ejection from
190: the LMC by means of numerical simulations of three-body scatterings
191: with a massive black hole. In \S\,\ref{sec:nb} we focus on
192: interactions in which a binary containing a young 8$\msun$
193: main-sequence star (representing \he) encounters a single black hole
194: of $10^2\msun$ to $10^4\msun$, whereas in \S\,\ref{sec:bbh} a single
195: 8$\msun$ main-sequence star encounters a binary black hole. In the
196: first case, an exchange interaction can lead to the high velocity
197: ejection of one of the binary components while in the latter case the
198: main-sequence star can be accelerated in a fly-by.
199:
200: The simulations are carried out using the {\tt sigma3} package, which
201: is part of the STARLAB\footnote{\tt
202: http://www.manybody.org/manybody/starlab.html} software environment
203: \citep{mh96,p01}.
204:
205: For each simulation we select the masses of the three stars, the
206: semi-major axis and eccentricity of the binary and the relative
207: velocity at infinity between the binary's centre of mass and the
208: single star. The phase of the encountering binary is randomly drawn
209: from a uniform distribution \citep{hb83} and the orbital eccentricity
210: is taken randomly between circular and hyperbolic from the thermal
211: distribution. The impact parameter $b$ is randomised according to
212: $b^2$ in the range $[0-b_{\rm max}]$ to guarantee that the probability
213: distribution is homogeneously sampled. The maximum value $b_{\rm max}$
214: is determined automatically for each set of experiments (see
215: \citet{gpe04} for a description). Energy conservation is typically
216: better than one part in $10^6$ and, in case the error exceeds
217: $10^{-5}$, the encounter is rejected. The accuracy in the integrator
218: is chosen in such a way that at most 5\% of the encounters are
219: rejected. During the simulations we allow for physical collisions when
220: the distance between any two stars is smaller than the sum of their
221: radii. The black hole is assumed to be a point mass, while for the
222: stars we adopt zero-age main-sequence radii, taken from
223: \citet{1989ApJ...347..998E}.
224:
225:
226:
227: \subsection{Encounters between a stellar binary and a massive black hole}
228: \label{sec:nb}
229:
230: We consider encounters between a binary consisting of two
231: main-sequence stars with masses $m_1$ and $m_2$ and a single black
232: hole with mass $\mbh$. The mass of one binary component is fixed to
233: 8$\msun$ while the mass of the companion star has values of 2$\msun$,
234: 4$\msun$, 8$\msun$ and 16$\msun$. For each choice of stellar masses,
235: the semi-major axis $a$ is varied between a minimum value, which is
236: set by the physical radii of the stars so that the two components do
237: not touch at the first pericentre passage, and a maximum of 1\,AU. We
238: consider black holes of masses $10^2\msun$, $10^3$ and $10^4\msun$.
239: The relative velocity at infinity between the black hole and the
240: binary's centre of mass is set equal to 20$\kms$.
241: %, consistent with the velocity dispersion in the LMC \citep{vdm02}.
242:
243:
244: \begin{figure}
245: \begin{center}
246: \includegraphics[width=8cm]{fig1.eps}
247: \end{center}
248: \caption{Branching ratio for the outcome of encounters between a
249: stellar binary and a single $10^3\msun$ black hole as a function of
250: the initial binary semi-major axis. The two binary components are
251: assumed to be 8$\msun$ main-sequence stars. The different outcomes
252: are: merger (stars), preservation (triangles) and exchange
253: (squares). The subset of exchanges with a high velocity escaper
254: ($V_{\infty} \ge 500\kms$) is indicated with bullets. The error
255: bars represent the formal ($1\sigma$) Poissonian uncertainty of the
256: measurement.}
257: \label{fig:branch}
258: \end{figure}
259: In Fig.\,\ref{fig:branch} we present the probability of different
260: outcomes (branching ratios) in the simulations of an encounter between
261: an equal mass binary ($m_1 = m_2 = 8\msun$) and a $10^3\msun$ black
262: hole. For each of the initial semi-major axes we perform a total of
263: 1500 scattering experiments, which result either in a fly-by, an
264: exchange or a merger. Mergers occur preferentially in the case of
265: tight binaries, leaving place to preservations and exchanges for wider
266: binaries. In an exchange interaction one of the main-sequence stars
267: is captured by the black hole while the other star is ejected,
268: possibly with high velocity. If the velocity exceeds 500$\kms$, we
269: regard the star as a possible candidate for \he. This occurs in about
270: 10\% or less of all encounters for semi-major axes in the range
271: $0.1-1.0\,\rm AU$.
272:
273: \begin{figure}
274: \begin{center}
275: \includegraphics[width=8cm]{fig2.eps}
276: \end{center}
277: \caption{Fraction of exchange encounters between a stellar binary and
278: a $10^3\msun$ black hole that results in the ejection of a
279: main-sequence stars with a velocity $\apgt 500\kms$. The
280: bullets indicate the relative probability for an incoming
281: binary with two 8$\msun$ stars (see
282: Fig.\,\ref{fig:branch}). The other curves are computed with
283: different companion masses to an 8$\msun$ star: 2$\msun$
284: (crosses), 4$\msun$ (squares), 16$\msun$ (triangles). In the
285: latter case, the ejected star is the secondary.}
286: \label{fig:branchHVS}
287: \end{figure}
288: In Fig.\,\ref{fig:branchHVS} we show the branching ratios for high
289: velocity ejections for binaries with a 8$\msun$ star and a companion
290: with mass in the range $2.0-16.0\msun$. While the 8$\msun$
291: main-sequence star is ejected, the companion star is captured by the
292: IMBH. As in the previous plot, the black hole mass is of $10^3\msun$.
293: Exchange encounters with fast escapers are more likely in the case of
294: binaries with larger companion masses and/or shorter orbital periods.
295:
296:
297: Analytical estimates by \citet{yt03} for the case of tidal breakup
298: predict an ejection velocity at infinity
299: \begin{equation}
300: \label{eq:tidal}
301: V_{\infty} = v^{'}\left(\frac{\mbh}{m}\right)^{1/6}\left(\frac{0.1\,\rm AU}{a}\frac{m}{1\msun}\right)^{1/2}\,,
302: \end{equation}
303: where $v^{'}$ is a function of the distance of closest approach
304: $\rmin$.
305: If we define a dimensionless closest approach parameter
306: $\rpmin = \frac{\rmin}{a} \left( \frac{\mbh}{m}\right)^{-1/3}\,,$
307: $v^{'}$ varies in the range 130--160$\kms$ for $\rpmin = 0-1$.
308: For $m=8\msun$ and $a=0.1\,\rm AU$ Eq.\,\ref{eq:tidal} yields
309: $v_{\infty} \approx 560\kms$ for $\mbh = 10^2\msun$; a 100$\msun$
310: IMBH would in principle be sufficient to explain the origin of \he.
311:
312:
313: \begin{figure*}
314: \begin{center}
315: \includegraphics[width=5.5cm]{fig3a.eps}
316: \includegraphics[width=5.5cm]{fig3b.eps}
317: \includegraphics[width=5.5cm]{fig3c.eps}
318: \end{center}
319: \caption{Velocity distributions for different values of the black hole
320: mass $\mbh=10^2\msun$ (left), $10^3\msun$ (middle),
321: $10^4\msun$ (right), and of the initial semi-major
322: axis $a$ = 0.1, 0.3, 1.0\,AU.}
323: \label{fig:veld}
324: \end{figure*}
325: In Fig.\,\ref{fig:veld} we show the velocity distributions of the
326: escaping single star in the case of equal mass binaries ($m_1 = m_2 =
327: 8\msun$). The different panels refer to different values of the black
328: hole mass, from $10^2\msun$ (left) to $10^3\msun$ (middle) and
329: $10^4\msun$ (right). In each panel, the different distributions refer
330: to three different values of the initial semi-major axis. While it
331: appears possible to reach the required recoil velocity in the case of
332: $\mbh=10^2\msun$, the smallest possible semi-major axis is needed
333: ($a=0.1\,\rm AU$) for this to happen, and only in about 10\% of
334: exchange encounters, which corresponds to about 3\% of all simulated
335: scatterings. In the case of the $10^4\msun$ black hole it is possible
336: to achieve such high velocities even for rather wide binaries. Based
337: on these data we conclude that a 100$\msun$ black hole is unlikely to
338: have resulted in the ejection of \he\, and a black hole mass $\apgt
339: 10^3\msun$ is favoured for typical values of $a$.
340:
341: We note that the velocity distributions are not sensitive to the
342: initial velocity between the binary and the single star as the total
343: energy of the system is dominated by the binding energy of the binary
344: rather than by the kinetic energy of the incoming star. Additional
345: experiments performed with initial velocities of $5\kms$ and $10\kms$
346: showed no appreciable difference in the ejection velocities.
347:
348: We also note that for detached binaries the binary tidal radius is
349: always larger than the tidal radius of the single stars and therefore
350: tidal breakup and ejection occur before the stars can be disrupted by
351: the black hole. For (near) contact binaries, instead, it is possible
352: to disrupt a star before an ejection can take place. This process
353: does not depend on the black hole mass but on the size of the binary
354: compared to the size of the stars. Our simulations only include
355: detached binaries as the proper treatment of contact binaries would
356: require hydrodynamical simulations.
357:
358:
359: The analytical estimates obtained with Eq.\,\ref{eq:tidal} are
360: considerably larger than the results of our simulations (see
361: Fig.\,\ref{fig:veld}). This discrepancy is probably caused by the
362: assumption of \citet{yt03} that the binary is disrupted well within
363: the tidal radius, i.e. $\rpmin \aplt 1$, whereas in 10\% to 20\% of
364: our simulated encounters $\rpmin > 1$. For example, for $\rpmin
365: \simeq 3-4$, a capture can occur which eventually leads to the
366: ejection of one star. Since the ejection velocity depends largely on
367: the Keplerian velocity of the binary at the position of tidal breakup,
368: we expect a dependence of the final velocity of escapers on the
369: parameter $\rpmin$.
370: \begin{figure}
371: \begin{center}
372: \includegraphics[width=8cm]{fig4.eps}
373: \end{center}
374: \caption{Velocity of the ejected star after an exchange interaction
375: with a $10^3\msun$ black hole as a function of the distance of closest
376: approach to the black hole. The different symbols indicate different
377: initial values of the binary semi-major axis.}
378: \label{fig:rmvel}
379: \end{figure}
380: Fig.\,\ref{fig:rmvel} shows the velocity at infinity of the ejected
381: star versus the dimensionless distance of closest approach. In this
382: example we consider equal mass binaries with $m = 8\msun$ and a black
383: hole of mass $10^3\msun$.
384: The figure shows that the highest ejection velocities are preferentially
385: achieved in close encounters.
386: The neglect of relatively wide encounters by \citet{yt03} might
387: therefore be the origin of the apparent discrepancy in the velocities.
388: The agreement between our numerical simulations and the analytical
389: estimates of \citet{yt03} improves for higher black hole masses, and
390: is very good for encounters with a SMBH \citep[see][]{gps05}.
391:
392: In our systematic study of the effect of the initial semi-major axis
393: of the interacting binary we adopted a homogeneous sampling in $\log
394: a$. Furthermore, the number of scattering experiments performed per
395: initial selection of $a$ are weighted with equal cross section. If the
396: distribution of orbital separations in a star cluster is flat in $\log
397: a$, like in the case of young star clusters
398: \citep{2005A&A...430..137K}, we can superpose the results of these
399: experiments in order to acquire a total velocity distribution of the
400: ejected star. In Fig.\,\ref{fig:all3} we present this superposed
401: velocity distribution for a binary consisting of two 8$\msun$
402: stars. The three histograms in this figure give the velocity
403: distribution for a 100$\msun$ (left), $10^3\msun$ (middle) and
404: $10^4\msun$ (right) black hole.
405: \begin{figure}
406: \begin{center}
407: \includegraphics[width=8cm]{fig5.eps}
408: \end{center}
409: \caption{Velocity distribution for the ejected star after an
410: interaction between an equal mass ($m=8\msun$) binary and a
411: black hole of $10^2\msun$ (black), $10^3\msun$ (blue)and
412: $10^4\msun$ (red). These velocity distributions are
413: integrated over the entire range of orbital separations for
414: the initial binary.}
415: \label{fig:all3}
416: \end{figure}
417:
418:
419: \subsection{Encounters between a single star and a binary black hole}
420: \label{sec:bbh}
421:
422: We now consider encounters between a single star and a binary black
423: hole. In a recent study \citet{2006ApJ...640L..39G} argue that it is
424: possible to form a black hole binary as a result of two collision
425: runaways in a star cluster. The binary eventually merges due to the
426: emission of gravitational wave radiation \citep{peters64} but, before
427: coalescing, the two black holes experience frequent interactions with
428: other cluster stars, possibly accelerating them to larger velocities.
429:
430: We simulate these encounters and study the probability of high
431: velocity ejections as a function of the binary parameters: the masses
432: of the two black holes $M_{\rm Bh1}$ and $M_{\rm Bh2}$ and the initial
433: semi-major axis $a$. The mass of the single star is fixed to
434: 8$\msun$, in accordance with the observed mass of \he. We consider
435: black hole masses of $10^2\msun$, $10^3\msun$ and $10^4\msun$ in both
436: equal mass and unequal mass binaries. For each choice of black hole
437: masses, the semi-major axis is varied between a minimum value of 1\,AU
438: and a maximum value of 1000\,AU. The minimum value is chosen in order
439: to guarantee that the binary remains unaffected by the emission of
440: gravitational wave radiation during the period over which the
441: encounter takes place\footnote{Two $10^3\msun$\, black holes in a
442: 1\,AU orbit will merge due to the emission of gravitational waves in
443: about 24\,Myr.}.
444:
445: Except for a few mergers (about 1-2\% of all cases), the vast majority
446: of the encounters result in a fly-by. The velocity at infinity of the
447: escaping star depends sensitively on the impact parameter of the
448: encounter. Only in encounters for which the impact parameter is
449: comparable to the binary separation are high velocity ejections
450: realized. Most of these encounters are resonant, allowing the incoming
451: star to come as close to one the black holes as a fraction of the
452: orbital separation. Only a few \% of the encounters result in ejection
453: velocities as high as 500$\kms$ and we argue that this mechanism is
454: unlikely to have produced the observed velocity of \he.
455:
456:
457: \section{An IMBH in the LMC?}
458: The analysis carried out in the previous sections suggest the
459: intriguing idea that \he\, was ejected by an encounter with an IMBH of
460: $\apgt 10^3\msun$ in the LMC. Such an IMBH would most likely be found
461: in a young dense cluster containing stars coeval to \he. Recently
462: \citet{mg2003} measured the structural parameters for 53 LMC clusters.
463: A total of nine clusters in this database are younger than 35\,Myr.
464: Three of these clusters, NGC\,1818, NGC\,1847 and NGC\,1850, have the
465: appropriate age to be the host of \he\, but their current densities
466: are much smaller than what is required to produce an IMBH
467: \citep{p04,2002MNRAS.330..232M}. Of the six remaining clusters,
468: three (NGC\,1805, NGC\,1984, NGC\,2011) have a mass below 5000$\msun$
469: and are therefore unlikely to have produced an IMBH. Also, the
470: presence of an IMBH in these clusters would have been noticed easily
471: \citep{2005ApJ...620..238B}.
472:
473: \begin{table}
474: \caption{ List of young (age $\aplt 35$\,Myr) LMC clusters
475: with an estimated initial relaxation time smaller than 100\,Myr. The
476: columns give the name of the cluster, the mass, the core radius, the
477: ratio of the core radius to the half-mass radius, the age, the current
478: two-body relaxation time, the initial two-body relaxation time
479: (computed using Eq.\,1 of \citet{2006astro.ph.10659P} with $\kappa =
480: 0.1$) and the black hole mass based on \citet{hh06}. For estimating
481: the relaxation time we adopted a mean mass $\langle m \rangle = 1$ and
482: a Coulomb logarithm parameter $\gamma = 0.4$ for the lower limit, and
483: $\langle m \rangle = 0.65$ and $\gamma = 0.01$ for the upper limit.
484: The core and half-mass radius for R\,136 are from
485: \citet{1996ApJ...466..254B}.}
486: \label{Table:LMC_clusters}
487: \begin{tabular}{lccccccc}
488: \hline
489: name & $M$ & $\rcore$ & $\rcore/\rhm$ & age & $\trlx$ & $\trlx^0$ & $\mbh$\\
490: &$\msun$ & pc & & Myr & Myr & Myr & $\msun$\\
491: \hline
492: R\,136 & 35000 & 0.32 &0.29& 3 & 7--17 & 6--14 & 1000 \\
493: NGC\,2004 & 27000 & 1.57 &0.50& 20 & 71--170 & 41--98 & 1600 \\
494: NGC\,2100 & 30200 & 1.22 &0.62& 16 & 51--120 & 31--73 & 2200 \\
495: \hline
496: \end{tabular}
497: \end{table}
498: The remaining three clusters are listed in
499: Tab.\,\ref{Table:LMC_clusters}. For these clusters we reconstruct the
500: initial relaxation time via the method described by
501: \citet{2006astro.ph.10659P}. In this relatively simple model, the
502: current relaxation time has a one-parameter relation with the
503: cluster's initial relaxation time. Of the three clusters listed in
504: Tab.\,\ref{Table:LMC_clusters}, one (R\,136) is too young to produce
505: an IMBH and experience a strong encounter with a binary. The best
506: candidate clusters to host an IMBH are then NGC\,2100 and
507: NGC\,2004. We can estimate the mass of a possible black hole in these
508: clusters by adopting the relation between the cluster structural
509: parameters and the mass of a central black hole presented by
510: \citet{hh06}. The authors performed direct N-body simulations to
511: quantify the relation between the mass of a central IMBH and the ratio
512: of the core radius to the half-mass radius. The last column in
513: Tab.\,\ref{Table:LMC_clusters} provides an estimate for the mass of
514: the IMBH using this relation. We find that, given the structure of the
515: observed clusters, a 1600$\msun$ to 2200$\msun$ IMBH could be present.
516:
517:
518: We argue that NGC\,2100 and NGC\,2004 form the most promising birth
519: places for \he. If this were the case, the travel time from the parent
520: cluster to the current location would be less than about 20\,Myr.
521: Such a short travel time would require an ejection velocity $\apgt
522: 800\kms$ for the adopted total distance to the LMC of 18$\kpc$. This
523: large velocity could only be obtained with a black hole of several
524: $10^3\msun$, somewhat higher than the estimated mass of a possible
525: black hole in these clusters.
526:
527:
528: We now compute the rate for ejection of hypervelocity stars from
529: NGC\,2100 and NGC\,2004 using the parameters listed in
530: Tab.\,\ref{Table:LMC_clusters} and a dimensionless cross-section for
531: exchange encounters $\tilde{\sigma} \simeq 5000$, as derived from the
532: scattering experiments. Adopting a central density of
533: $2\times10^4\pcc$ (consistent with a King $W_0 = 9$ initial profile),
534: a velocity of $10\kms$ and a semi-major axis of 0.2\,AU, we obtain an
535: encounter rate of about one per 2\,Myr for both clusters. Since only
536: one in about ten exchange encounters produces a fast escaper (see
537: Fig.\,\ref{fig:branchHVS}), we derive a production rate of one per
538: 20\,Myr. This value represents a lower limit as we don't take into
539: account the effects of a mass function in the core and the semi-major
540: distribution. It is therefore conceivable that NGC\,2100 or NGC\,2004
541: has produced one high velocity escaper, in which case an IMBH must be
542: present in one of these clusters.
543:
544: %This exciting possibility, though
545: %controversial, has interesting consequences regarding the presence of
546: %an IMBH in the LMC.
547:
548: \section{Acknowledgments}
549: We are grateful to Jes\'us Ma\'iz Apell\'aniz for interesting comments on
550: the manuscript. This work was supported by the Netherlands
551: Organization for Scientific Research (NWO under grant No. 635.000.001
552: and 643.200.503), the Royal Netherlands Academy of Arts and Sciences
553: (KNAW) and the Netherlands Research School for Astronomy (NOVA).
554:
555: \bibliographystyle{mn2e}
556: \bibliography{biblio}
557:
558: \end{document}
559: