1: %%

2: %% Beginning of file 'xteapj13.tex'

3: %%

4: %% First modified 2007 May 10

5: %% Last modified 2008 January 10

6: %%

7: %% This is a sample manuscript marked up using the

8: %% AASTeX v5.x LaTeX 2e macros.

9: 

10: %% The first piece of markup in an AASTeX v5.x document

11: %% is the \documentclass command. LaTeX will ignore

12: %% any data that comes before this command.

13: 

14: %% The command below calls the preprint style

15: %% which will produce a one-column, single-spaced document.

16: %% Examples of commands for other substyles follow. Use

17: %% whichever is most appropriate for your purposes.

18: %%

19: %%\documentclass[12pt,preprint]{aastex}

20: 

21: %% manuscript produces a one-column, double-spaced document:

22: 

23: %%\documentclass[manuscript]{aastex}

24: 

25: %% preprint2 produces a double-column, single-spaced document:

26: 

27: \documentclass[11pt,preprint2]{aastex}

28: 

29: %% Sometimes a paper's abstract is too long to fit on the

30: %% title page in preprint2 mode. When that is the case,

31: %% use the longabstract style option.

32: 

33: %%\documentclass[preprint2,longabstract]{aastex}

34: 

35: %% If you want to create your own macros, you can do so

36: %% using \newcommand. Your macros should appear before

37: %% the \begin{document} command.

38: %%

39: %% If you are submitting to a journal that translates manuscripts

40: %% into SGML, you need to follow certain guidelines when preparing

41: %% your macros. See the AASTeX v5.x Author Guide

42: %% for information.

43: 

44: \newcommand{\Msun}{\hbox{$\hbox{M}_\odot\;$}}

45: \newcommand{\Rsun}{\hbox{$\hbox{R}_\odot\;$}}

46: \newcommand{\kms}{\hbox{${\rm km}\:{\rm s}^{-1}\;$}}

47: \newcommand{\Msuno}{\hbox{$\hbox{M}_\odot$}}

48: \newcommand{\Rsuno}{\hbox{$\hbox{R}_\odot$}}

49: \newcommand{\kmso}{\hbox{${\rm km}\:{\rm s}^{-1}$}}

50: \newcommand{\teff}{$T_{\rm eff}\;$}  

51: \newcommand{\teffo}{$T_{\rm eff}$}  

52: \newcommand{\logg}{$\log\;g\;$}  

53: \newcommand{\loggo}{$\log\;g$}  

54: \newcommand{\mic}{$\xi_{\rm t}$\;}

55: \newcommand{\mico}{$\xi_{\rm t}$}

56: \newcommand{\myemail}{Jonay.Gonzalez-Hernandez@obspm.fr}

57: 

58: %% You can insert a short comment on the title page using the command below.

59: 

60: %%\slugcomment{Gonz\'alez Hern\'andez et al.}

61: 

62: %% If you wish, you may supply running head information, although

63: %% this information may be modified by the editorial offices.

64: %% The left head contains a list of authors,

65: %% usually a maximum of three (otherwise use et al.).  The right

66: %% head is a modified title of up to roughly 44 characters.

67: %% Running heads will not print in the manuscript style.

68: 

69: \shorttitle{Stellar Abundances in XTE J1118+480}

70: \shortauthors{J. I. Gonz\'alez Hern\'andez et al.}

71: 

72: %% This is the end of the preamble.  Indicate the beginning of the

73: %% paper itself with \begin{document}.

74: 

75: \begin{document}

76: 

77: %% LaTeX will automatically break titles if they run longer than

78: %% one line. However, you may use \\ to force a line break if

79: %% you desire.

80: 

81: \title{Chemical Abundances of the Secondary Star \\ in the Black Hole

82: X-ray Binary XTE J1118+480}

83: 

84: %% Use \author, \affil, and the \and command to format

85: %% author and affiliation information.

86: %% Note that \email has replaced the old \authoremail command

87: %% from AASTeX v4.0. You can use \email to mark an email address

88: %% anywhere in the paper, not just in the front matter.

89: %% As in the title, use \\ to force line breaks.

90: 

91: \author{Jonay I. Gonz\'alez Hern\'andez\altaffilmark{1,2,3}, Rafael

92: Rebolo\altaffilmark{3,4}, Garik Israelian\altaffilmark{3},

93: Alexei V. Filippenko\altaffilmark{5}, Ryan Chornock\altaffilmark{5},

94: Nozomu Tominaga\altaffilmark{6}, Hideyuki Umeda\altaffilmark{6}, and 

95: Ken'ichi Nomoto\altaffilmark{6,7,8}}

96: 

97: %% Notice that each of these authors has alternate affiliations, which

98: %% are identified by the \altaffilmark after each name.  Specify alternate

99: %% affiliation information with \altaffiltext, with one command per each

100: %% affiliation.

101: 

102: \altaffiltext{1}{Observatoire de Paris-Meudon, GEPI, 5 place Jules

103: Janssen, 92195 Meudon Cedex, France; Jonay.Gonzalez-Hernandez@obspm.fr}

104: \altaffiltext{2}{CIFIST Marie Curie Excellence Team}

105: \altaffiltext{3}{Instituto de Astrof{\'\i }sica de Canarias, E-38205 La Laguna,

106: Tenerife, Spain; rrl@iac.es, gil@iac.es}

107: \altaffiltext{4}{Consejo Superior de Investigaciones 

108: Cient{\'\i}ficas, Spain.}

109: \altaffiltext{5}{Department of Astronomy, Uni\-ver\-si\-ty of

110: California, Berkeley, CA 94720-3411, USA; \\ alex@astro.berkeley.edu,

111: rchornock@astro.berkeley.edu} 

112: \altaffiltext{6}{Department of Astronomy, School of Science, University of

113: Tokyo, Bunkyo-ku, Tokyo 113-0033, Japan;

114: tominaga@astron.s.u-tokyo.ac.jp, umeda@astron.s.u-tokyo.ac.jp, 

115: nomoto@astron.s.u-tokyo.ac.jp}

116: \altaffiltext{7}{Research Center for the Early Universe, School of

117: Science, University of Tokyo, Bunkyo-ku, Tokyo 113-0033, Japan.} 

118: \altaffiltext{8}{Institute for the Physics and Mathematics of the

119: Universe, University of Tokyo, Kashiwa, Chiba 277-8582, Japan.} 

120: 

121: %% Mark off your abstract in the ``abstract'' environment. In the manuscript

122: %% style, abstract will output a Received/Accepted line after the

123: %% title and affiliation information. No date will appear since the author

124: %% does not have this information. The dates will be filled in by the

125: %% editorial office after submission.

126: 

127: \begin{abstract}

128: 

129: Following the recent abundance measurements of Mg, Al, Ca, Fe, and Ni

130: in the black hole X-ray binary \mbox{XTE J1118+480} using

131: medium-resolution Keck~II/ESI spectra of the secondary star

132: (Gonz\'alez Hern\'andez et al. 2006), we perform a detailed abundance

133: analysis including the abundances of Si and Ti. These element

134: abundances, higher than solar, indicate that the black hole in this

135: system formed in a supernova event, whose nucleosynthetic products

136: could pollute the atmosphere of the secondary star, providing clues on

137: the possible formation region of the system, either Galactic halo,

138: thick disk, or thin disk. We explore a grid of explosion models

139: with different He core masses, metallicities, and geometries.

140: Metal-poor models associated with a formation scenario in the Galactic

141: halo provide unacceptable fits to the observed abundances, allowing

142: us to reject a halo origin for this X-ray binary. The thick-disk

143: scenario produces better fits, although they require substantial

144: fallback and very efficient mixing processes between the inner layers

145: of the explosion and the ejecta, making quite unlikely an origin

146: in the thick disk. The best agreement between the model predictions

147: and the observed abundances is obtained for metal-rich progenitor

148: models. In particular, non-spherically symmetric models are able to

149: explain, without strong assumptions of extensive fallback and mixing,

150: the observed abundances. Moreover, asymmetric mass ejection in

151: a supernova explosion could account for the required impulse necessary

152: to launch the system from its formation region in the Galactic thin

153: disk to its current halo orbit.

154:  

155: \end{abstract}

156: 

157: \keywords{black holes: physics --- stars: abundances --- stars: 

158: evolution --- stars: individual (\mbox{XTE J1118+480}) --- supernovae:

159: general --- X-rays: binaries}  

160: 

161: \section{Introduction}

162:        

163: The low-mass X-ray binary \mbox{XTE J1118+480} is the first identified

164: black hole moving in Galactic halo regions (Wagner et al. 2001;

165: Mirabel et al. 2001). Since it was discovered during a faint outburst

166: on UT 2000 March 29 (Remillard et al.\ 2000), it has been intensively

167: studied in both the X-ray and optical spectral regions. During the

168: decay of the outburst, McClintock et al. (2001) and Wagner et al.

169: (2001) determined the radial-velocity curve of the companion star,

170: yielding a mass function $f(M) \approx 6$~\Msuno. The companion star

171: was classified as a late-type main-sequence star with a mass of

172: 0.1--0.5~{\Msuno} (Wagner et al.\ 2001). 

173: 

174: By modelling the light curve, McClintock et al. (2001) 

175: derived a lower limit to the

176: orbital inclination, $i \ga 55^\circ$, and consequently an upper limit

177: to the black hole mass of $M_{\rm BH} \la 10$~\Msuno. Additional

178: evidence for a high inclination ($i \ga 60^\circ$) comes from

179: measurements of tidal distortion (Frontera et al. 2001), whereas the

180: lack of dips or eclipses for a Roche-lobe filling secondary yields

181: upper limits of $i \ga 80^\circ$ and $M_{\rm BH} \ga 7.1$~\Msuno. 

182: Later, Gelino et al. (2006) derived an orbital inclination of

183: $68^\circ\;\pm\;2^\circ$, by modeling the optical and infrared

184: ellipsoidal light curves of the system in quiescence. This value of

185: the inclination allowed them to better constrain the black hole mass

186: at $M_{\rm BH} = 8.53 \pm 0.60$~\Msuno.   

187: 

188: The system is placed in the Galactic halo, with an extraordinarily

189: high Galactic latitude ($b \approx 62.3^{\circ}$), and a height of

190: $\sim$1.6 kpc above the Galactic plane, according to its distance of

191: $1.85\pm0.36$ kpc (Wagner et al.\ 2001). This appears surprising since

192: all other black hole binaries are located in the Galactic disk. 

193: An accurate measurement of

194: its proper motion coupled with its distance provides space-velocity

195: components ($U$, $V$) which seem consistent with those of some old halo

196: globular clusters (Mirabel et al.\ 2001). This opened the possibility

197: that the system originated in the Galactic halo, and therefore, that the

198: black hole could be either the remnant of a supernova (SN) in the very

199: early Galaxy or the result of direct collapse of an ancient massive star.

200: However, the galactocentric orbit crossed the Galactic plane many

201: times in the past, and the system could have formed in the

202: Galactic disk and been launched into its present orbit as a

203: consequence of the ``kick'' imparted during the SN explosion of

204: a massive star (Gualandris et al.\ 2005). Recent observations with the

205: 10-m Keck~II telescope revealed that the secondary star has a

206: supersolar surface metallicity ($[{\rm Fe/H}]=0.2\pm0.2$, Gonz\'alez

207: H\'ernandez et al. 2006), confirming the origin of the black hole

208: in a SN event. Thus, if the system originated in the Galactic

209: halo, the element abundances of the secondary star must have been

210: enriched by a factor of 5--25 depending on whether its initial 

211: metallicity resembled a thick-disk star or a halo star.

212: 

213: Element abundances of secondary stars of X-ray binaries have been

214: studied for the systems Nova Scorpii 1994 (Israelian et al. 1999;

215: Gonz\'alez Hern\'andez et al. 2007), A0620--00 (Gonz\'alez Hern\'andez

216: et al. 2004), Centaurus X-4 (Gonz\'alez Hern\'andez et al. 2005),

217: \mbox{XTE J1118+480} (Gonz\'alez Hern\'andez et al. 2006, hereafter

218: Paper~I), and V4641 Sagittarii (Orosz et al. 2001; Sadakane et

219: al. 2006). All of these X-ray binaries show metallicities close to

220: solar independent of their location with respect to the Galactic

221: plane, and possible scenarios of pollution from a SN or hypernova have

222: been discussed. In this paper, we compare in detail different

223: scenarios of the possible enrichment of the secondary star from SN

224: yields, providing conclusions on the formation region (Galactic halo,

225: thick disk, or thin disk) of this halo black hole X-ray binary.

226: 

227: \begin{figure}[ht!]

228: %\epsscale{.80}

229: %\plotone{f1.eps}

230: \centering

231: \includegraphics[width=6.2cm,angle=90]{f1.eps}

232: \caption{\footnotesize{Radial velocities of \mbox{XTE J1118+480}

233: folded on the orbital solution of the data, together with the 

234: best-fitting sinusoid. Individual velocity uncertainties are $\le 7$ 

235: ${\rm km}\ {\rm s}^{-1}$ and are not plotted because they are always 

236: smaller than the symbol size. The bottom panel shows the residuals

237: of the fit.}}   

238: \label{fig1}

239: \end{figure}

240: 

241: \section{Observations}

242: 

243: As already reported in Paper~I, we obtained 74 medium-resolution spectra 

244: of the black hole X-ray binary \mbox{XTE J1118+480}, in quiescence, with

245: the Echelle Spectrograph and Imager (ESI; Sheinis et al.\ 2002) at the

246: 10-m Keck~II telescope on UT 14 February 2004. The data covered the

247: spectral range 4000--9000~{\AA} at a resolving power

248: $\lambda/\Delta\lambda \approx 6,000$. We also observed ten template 

249: stars with spectral types in the range K0V--M2V with the same

250: instrument and spectral configuration. The exposure time

251: was fixed at 300~s to minimize the effects of orbital smearing which, 

252: for the orbital parameters of \mbox{XTE J1118+480}, is in the

253: range 0.6--26.6 \kmso, smaller than the instrumental resolution of 50 

254: \kmso. All of the spectra were reduced in a standard manner.  

255: 

256: \begin{figure*}[ht!]

257: %\epsscale{.80}

258: %\plotone{f2.eps}

259: \centering

260: \includegraphics[width=11cm,angle=90]{f2.eps}

261: \caption{\footnotesize{Observed spectrum of the secondary star of

262: \mbox{XTE J1118+480} (top panel) and of two properly broadened templates 

263: (BD$-$053763, middle panel; BD+52857, bottom panel).}}  

264: \label{fig2}

265: \end{figure*}

266: 

267: \section{Revised Orbital Parameters}

268: 

269: We extracted the radial velocities by cross-correlating each target

270: spectrum with the spectrum of a K5V template star, using the software

271: MOLLY developed by T. R. Marsh. We fitted these data with a sine wave

272: using a $\chi2$ method providing the following orbital solution (see

273: Fig.~\ref{fig1}): $\gamma= 2.7 \pm 1.1$ \kmso, $K_2 = 708.8 \pm 1.4$ \kmso,

274: $P=0.16995 \pm 0.00012$~d, and $T_0=2,453,049.93346 \pm 0.00007$~d,

275: where $T_0$ is defined as the corresponding time of the closest

276: inferior conjunction of the companion star, and the quoted

277: uncertainties are 1$\sigma$. This orbital period, $P$, and the 

278: velocity amplitude of the orbital motion of the secondary

279: star, $K_2$, lead to a mass function of $f(M) = 6.27 \pm 0.04$ \Msuno,

280: consistent with (but more precise than) previous results (McClintock 

281: et al. 2001; Wagner et al. 2001; Torres et al. 2004). 

282: 

283: The derived radial velocity of the center of mass of the

284: system agrees somewhat (at the 3$\sigma$ level) with previous

285: studies ($\gamma=+26\pm17$ \kmso, McClintock et al. 2001;

286: $\gamma=-15\pm10$ \kmso, Wagner et al. 2001; 

287: $\gamma=+16\pm6$ \kmso, Torres et al. 2004). Note that our

288: medium-resolution data have a factor of $\geq 4$ higher spectral

289: resolution than the data these authors used. We estimate the

290: uncertainty in our individual radial-velocity measurements of

291: typically 6~\kms (see Fig.~\ref{fig1}).  

292: 

293: \subsection{Secondary Spectrum}

294: 

295: The individual spectra were corrected for their radial velocity and 

296: combined in order to improve the signal-to-noise ratio (S/N). After

297: binning in wavelength in steps of 0.3~{\AA}, the final spectrum had 

298: S/N $\approx 80$ in the continuum in the red spectral

299: region. This spectrum, displayed in Fig.~\ref{fig2}, was compared with

300: ten template stars having spectral types K0V to M2V. The best fit shows

301: a K5V star rather than the later spectral types (K7/8V or even M)

302: suggested in previous studies (Wagner et al. 2001; Torres et al.

303: 2004). 

304: 

305: We should remark that the spectrum of the secondary star in this

306: system and the spectra of the template stars were normalized using the

307: same procedure. The continuum was fitted with a low-order spline, 

308: in order to avoid the smoothing of possible existing broad TiO bands. 

309: Following Marsh et al. (1994), we computed the optimal value of

310: $v~\sin~i$ by subtracting broadened versions of the K5V template

311: (in steps of 1 \kmso) and minimizing the residual. We used a

312: spherical rotational profile with linearized limb-darkening $\epsilon

313: = 0.8$ (Al-Naimiy 1978) due to the spectral type K of the secondary

314: star. The best fit corresponds to $v~\sin~i=100^{+3}_{-11}$ \kmso,

315: where the uncertainties have been derived by assuming extreme cases for

316: $\epsilon = 0-1$. Our derived rotational velocity, combined with our

317: value of the velocity amplitude, $K_2$, implies a binary mass ratio

318: $q=0.027\pm0.009$, in agreement with previous results (Torres et al.

319: 2004, and references therein).

320: 

321: \begin{deluxetable}{lccccc}

322: \tabletypesize{\scriptsize}

323: \tablecaption{Ranges and Steps of Model Parameters}   

324: \tablewidth{0pt}

325: \tablehead{\colhead{Parameter} & \colhead{Range} & \colhead{Step}}

326: \startdata

327: $T_{\mathrm{eff}}$  & $3500 \rightarrow 5000$ K & 100 K\\ 

328: $\log [g/({\rm cm~s}2)]$  & $4 \rightarrow 5$  & 0.1\\ 

329: $\mathrm{[Fe/H]}$ & $-1.5 \rightarrow 1$  & 0.1\\ 

330: $f_{4500}$ &  $0 \rightarrow 2$  & 0.1\\ 

331: $m_0$  & $0 \rightarrow -0.00091$ & $-$0.00010\\ 

332: \enddata

333: %\tablecomments{} 

334: \label{tbl1}

335: \end{deluxetable}

336: 

337: %***Jonay: In the figure below, the abscissa (x-axis)

338: % labels are still too small.

339: 

340: \begin{figure*}[ht!]

341: %\epsscale{.80}

342: %\plotone{f3.eps}

343: \centering

344: \includegraphics[width=11cm,angle=0]{f3.eps}

345: \caption{\footnotesize{Distributions obtained for each parameter using

346: Monte Carlo simulations. The labels at the top of each bin indicate

347: the number of simulations consistent with the bin value. The total

348: number of simulations was 1000.}}

349: \label{fig3}

350: \end{figure*}

351: 

352: \begin{figure*}[ht!]

353: %\epsscale{.80}

354: %\plotone{f4.eps}

355: \centering

356: \includegraphics[width=9cm,angle=90]{f4.eps}

357: \caption{\footnotesize{Best synthetic spectral fits to the Keck/ESI

358: spectrum of the 

359: secondary star in the \mbox{XTE J1118+480} system (bottom panel) and

360: the same for a template star (properly broadened) shown for comparison

361: (top panel). Synthetic spectra are computed for solar abundances

362: (dashed line) and best-fit abundances (solid line).}}

363: \label{fig4}

364: \end{figure*}

365: 

366: \begin{figure*}[ht!]

367: %\epsscale{.80}

368: %\plotone{f5.eps}

369: \centering

370: \includegraphics[width=9cm,angle=90]{f5.eps}

371: \caption{\footnotesize{The same as in Fig.\ 4, but for the spectral range

372: 5510--5550 {\AA}.}}  

373: \label{fig5}

374: \end{figure*}

375: 

376: \begin{figure*}[ht!]

377: %\epsscale{.80}

378: %\plotone{f6.eps}

379: \centering

380: \includegraphics[width=9cm,angle=90]{f6.eps}

381: \caption{\footnotesize{The same as in Fig.\ 4, but for the spectral range

382: 6420--6480 {\AA}.}}  

383: \label{fig6}

384: \end{figure*}

385: 

386: \section{Chemical Analysis}

387: 

388: \subsection{Stellar Parameters}

389: 

390: The normalized spectra of X-ray transients, although observed in

391: quiescence, show apparently weaker stellar lines of the secondary star

392: due to the veiling caused by the accretion disk. The veiling from the

393: accretion disk was estimated to be $\sim65\% \pm 8\%$ in the spectral 

394: range 5800--6400~{\AA} in December 2000 and January 2001, and

395: $\sim 40\% \pm 10\%$ in January 2003, by performing standard optimal

396: subtraction techniques with K5V--M0V template stars. 

397: 

398: As shown in Paper~I, 

399: we tried to infer the stellar parameters, $T_{\mathrm{eff}}$ and $\log

400: g$, and the metallicity [Fe/H], of the secondary star taking into account

401: the veiling from the accretion disk, defined as a linear function of

402: wavelength and thus described with two additional parameters, the

403: veiling at 4500~{\AA}, $f_{4500} = F^{4500}_{\rm disk}/F^{4500}_{\rm

404: sec}$, and the slope, $m_0$. Note that the total flux is defined as

405: $F_{\rm total} = F_{\rm disk} + F_{\rm  sec}$, where $F_{\rm disk}$ and

406: $F_{\rm sec}$ are the flux contributions of the disk and the continuum

407: of the secondary star, respectively. This procedure involves a

408: $\chi2$ minimization routine which compares several features of the 

409: stellar spectrum with a grid of synthetic spectra computed using the

410: LTE code MOOG (Sneden 1973). We used a grid of LTE model atmospheres

411: (Kurucz 1993) and atomic line data from the Vienna Atomic Line

412: Database (VALD, Piskunov 1995). The oscillator strengths of relevant

413: lines were adjusted until they reproduced the solar atlas (Kurucz et al.\

414: 1984) with solar abundances (Grevesse et al. 1996). The changes we

415: applied to the $\log gf$ values taken from the VALD database were

416: $\Delta\log gf \lesssim 0.2$ dex. 

417: 

418: \begin{deluxetable}{lccccccc}

419: \tabletypesize{\scriptsize}

420: \tablecaption{Abundance Uncertainties in \mbox{XTE J1118+480}}

421: \tablewidth{0pt}

422: \tablehead{\colhead{Element} & $\mathrm{[X/H]}_{\rm

423: LTE}$\tablenotemark{a} & $\Delta_{\sigma}$ &

424: $\Delta_{T_{\mathrm{eff}}}$ & $\Delta_{\log g}$

425: & $\Delta_{\rm veil}$ & $\Delta_{\rm

426: tot}$\tablenotemark{b} & $n_{\rm lines}$\tablenotemark{c}}

427: \startdata

428: Mg   &  0.35 & 0.12 & 0    & $-$0.10  & 0.20 & 0.25 & 1\\

429: Al   &  0.60 & 0.12 & 0.05 &  0     & 0.15 & 0.20 & 1\\

430: Si   &  0.37 & 0.03 & $-$0.07&  0.07  & 0.18 & 0.21 & 2\\

431: Ca   &  0.15 & 0.03 & 0.13 & $-$0.16  & 0.11 & 0.23 & 5 \\

432: Ti   &  0.32 & 0.18 & 0.12 & $-$0.03  & 0.14 & 0.26 & 3\\

433: Fe   &  0.18 & 0.08 & 0.06 &  0.04  & 0.13 & 0.17 & 5 \\

434: Ni   &  0.30 & 0    & 0.10 &  0.12  & 0.14 & 0.21 & 2 \\

435: Li\tablenotemark{d} & 1.78 & 0.12 & 0.15 & 0.05 & 0.15  & 0.25  & 1\\

436: \enddata

437: 

438: \tablenotetext{a}{Element abundances of the secondary star (calculated 

439: assuming LTE) are $\mathrm{[X/H]}= \log [N(\mathrm{X})/N(\mathrm{H})]_{\rm

440: star} - \log [N(\mathrm{X})/N(\mathrm{H})]_{\rm Sun}$, where

441: $N(\mathrm{X})$ is the number density of atoms. Uncertainties, $\Delta

442: \mathrm{[X/H]}$, are at the 1$\sigma$ level and take into account the

443: uncertainties in the stellar and veiling parameters.}

444: 

445: \tablenotetext{b}{The total error was estimated as $\Delta_{\rm tot} 

446: = \sqrt{\Delta^2_{\sigma} + \Delta^2_{T_{\mathrm{eff}}} + 

447: \Delta^2_{\log g} + \Delta^2_{\rm veil}}.$} 

448: 

449: \tablenotetext{c}{Number of features analyzed for each

450: element.} 

451: 

452: \tablenotetext{d}{\mbox{Li} abundance is expressed as $\log 

453: \epsilon(\mathrm{Li})_{\rm NLTE} = \log 

454: [N(\mathrm{Li})/N(\mathrm{H})]_{\rm NLTE} + 12$.} 

455: 

456: 

457: \tablecomments{The uncertainties from the dispersion of the best

458: fits to different features, $\Delta_{\sigma}$, are estimated using the

459: following formula: $\Delta_{\sigma} =\sigma/\sqrt{N}$, where $\sigma$

460: is the standard deviation of the measurements.}

461: 

462: \label{tbl2}

463: \end{deluxetable}

464: 

465: We selected nine spectral features containing in total 30 lines of

466: Fe~I and 8 lines of Ca~I with excitation potentials between 1 and 5~eV. 

467: The five free parameters were varied in the ranges given in Table~\ref{tbl1}.

468: For each given iron abundance in the range $[{\rm Fe}/{\rm H}] < 0$, the

469: Ca abundance was fixed according to the Galactic trend of Ca (Bensby

470: et al.\ 2005), while for $[{\rm Fe}/{\rm H}] > 0$, we assumed $[{\rm

471: Ca}/{\rm Fe}]=0$. A rotational broadening of 100 {\kms} and a

472: limb darkening $\epsilon = 0.8$ were adopted. The microturbulence,

473: $\xi$, was computed using an experimental expression as a function of

474: effective temperature and surface gravity (Allende Prieto et al.\

475: 2004).   

476:  

477: The result, already presented in Paper~I, provides as most likely

478: values $T_{\mathrm{eff}} = 4700 \pm 100$~K, $\log [g/{\rm cm~s}^2] =

479: 4.6 \pm 0.3$, $\mathrm{[Fe/H]} = 0.18 \pm 0.17$, $f_{4500} =

480: 0.85 \pm 0.20$, and $m_0 = -0.0002 \pm 0.0001$. The 1$\sigma$

481: uncertainties of the five free parameters were determined using 1000

482: realizations whose corresponding histograms are displayed in

483: Fig.~\ref{fig3}. 

484: 

485: \begin{figure*}[ht!]

486: %\epsscale{.80}

487: %\plotone{f6.eps}

488: \centering

489: \includegraphics[width=13.5cm,angle=0]{f7.eps}

490: \caption{\footnotesize{Abundance ratios of the secondary star in

491: \mbox{XTE J1118+480} (blue wide cross) in comparison with the

492: abundances of G and K metal-rich dwarf stars. Galactic trends were

493: taken from Gilli et al. (2006). The size of the cross 

494: indicates the uncertainty. Filled and empty circles correspond to

495: abundances for exoplanet host stars and stars without known exoplanet

496: companions, respectively. The dashed-dotted lines indicate solar

497: abundance values.}} 

498: \label{fig7}

499: \end{figure*}

500: 

501: The stellar parameters derived, especially the effective temperature,

502: could provide important implications on the Gelino et al. (2006)

503: determination of the orbital inclination of the system ($i \approx 

504: 68^\circ$). These authors modeled the optical and infrared (IR)

505: ellipsoidal light curves of \mbox{XTE J1118+480} in quiescence,

506: assuming a K7V spectral type (\teff $\approx 4250$~K) for the secondary

507: star. However, our spectroscopic value (\teff $\approx 4700$~K) would

508: require more contribution of the flux from the accretion disk in the $K$

509: band. Gelino et al. (2001) derived an inclination of $i \approx 41^\circ$,

510: yielding a black hole mass of $M_{\rm BH} \approx 11$~\Msun for 

511: the \mbox{A0620--00} system, by adopting \teff $\approx 4600$~K, 

512: which is 300~K lower than the spectroscopic value

513: reported by Gonz\'alez Hern\'andez et al. (2004). Hynes et al. (2005)

514: suggested that this different effective temperature of the 

515: secondary star in \mbox{A0620--00} would require a larger disk

516: contribution in the $K$ band, and therefore a higher inclination. In

517: fact, Gelino et al. (2001) commented that if the $K$-band disk veiling

518: as high as 50\% of the total flux, the derived inclination would increase to

519: 60$^\circ$. The black hole mass would then drop to $M_{\rm BH} \approx 

520: 5$ \Msuno. Although this is an extreme case, milder IR veiling could

521: still have a substantial impact on the derived black hole mass.

522: Similarly, the black hole mass in \mbox{XTE J1118+480} might be 

523: significantly affected when using our value of the effective temperature.

524: 

525: \subsection{Stellar Abundances}

526: \begin{deluxetable}{lccccc}

527: \tabletypesize{\scriptsize}

528: \tablecaption{Element Abundance Ratios in \mbox{XTE J1118+480}}

529: \tablewidth{0pt}

530: \tablehead{\colhead{Element} & $\mathrm{[X/Fe]}_{\rm XTE1118}$ &

531: $\Delta^{\tablenotemark{\star}}_{\rm [X/Fe],XTE1118}$ & $\mathrm{[X/Fe]}_{\rm stars}$ & $\sigma_{\rm

532: stars}$ & $\Delta_{\sigma,{\rm stars}}$} 

533: \startdata

534: Mg   &  0.17 & 0.22 &  0.02 & 0.08 & 0.02 \\

535: Al   &  0.42 & 0.15 &  0.14 & 0.10 & 0.02 \\

536: Si   &  0.19 & 0.17 & $-$0.03 & 0.04 & 0.01 \\

537: Ca   & $-$0.03 & 0.23 & $-$0.13 & 0.05 & 0.01 \\

538: Ti   &  0.14 & 0.22 &  0.05 & 0.08 & 0.01 \\

539: Ni   &  0.12 & 0.12 & $-$0.02 & 0.05 & 0.01 \\

540: \enddata

541: 

542: \tablenotetext{\star}{Uncertainties in the element abundance ratios

543: ($\mathrm{[X/Fe]}$) in the secondary star in \mbox{XTE J1118+480}.}

544: 

545: \tablecomments{$\mathrm{[X/Fe]}_{\rm stars}$ indicate the average

546: values of 24 stars with iron content in the range 0.01 to 0.35

547: corresponding to 1$\sigma$ in the $\mathrm{[Fe/H]}$ abundance of the

548: secondary star in \mbox{XTE J1118+480}, taken from Gilli et al.

549: (2006). The uncertainty in the average value of abundance

550: ratios in the comparison sample is obtained as $\Delta_{\sigma,{\rm

551: stars}} =\sigma_{\rm stars}/\sqrt{N}$, where $\sigma_{\rm stars}$ is

552: the standard deviation of the measurements and N is the number of stars.}    

553: 

554: \label{tbl3}

555: \end{deluxetable}

556: 

557: We inspected several spectral regions in the observed Keck/ESI spectrum 

558: of the secondary star, searching for suitable lines for a detailed

559: chemical analysis. Using the derived stellar parameters, we first 

560: determined the Fe abundance by comparing synthetic spectra with each

561: individual feature in the ESI spectrum (see Table~\ref{tbl2}). In

562: Fig.~\ref{fig4} (here) and Fig.~1 of Paper~I, we display some of the spectral

563: regions analyzed to obtain the Fe abundance. This figure also shows the

564: best synthetic spectral fit to the observed spectrum of a template

565: star (HIP 17420 with \teffo$=4801$ K, \loggo$ = 4.633$, and

566: [Fe/H]$=-0.14$ dex) using the stellar parameters and abundances

567: determined by Allende-Prieto et al. (2004). We only use as abundance

568: indicators those features which were well reproduced in the template

569: star. The chemical analysis is summarized in Table~\ref{tbl2}. The

570: errors in the element abundances show their sensitivity to the

571: uncertainties in the effective temperature

572: ($\Delta_{T_{\mathrm{eff}}}$), gravity ($\Delta{\log g}$), veiling

573: ($\Delta_{\rm veiling}$), and the dispersion of the measurements

574: from different spectral features ($\Delta_{\sigma}$). In

575: Table~\ref{tbl2} we also state the number of features analyzed for

576: each element.  

577: 

578: \begin{deluxetable}{lcccccccc}

579: \tabletypesize{\scriptsize}

580: \tablecaption{Supernova/Hypernova Model Parameters}

581: \tablewidth{0pt}

582: \tablehead{ $M_{\rm MS}$ & $M_{\rm He}$ & $E_{51}$ & 

583: \colhead{Z} & $M(Fe)$ & $M_{\rm cut}$ &

584: $M_{\rm fall}$ & 

585: $M_{{\rm BH},i}$\tablenotemark{a} & \colhead{Abundance Pattern}}

586: \startdata

587: 40   & 15.8 & 1  & 0.001 & 5.8982E-01 & 2.51 & 5.56 & 8.07 & Fig.~9a \\

588: 40   & 15.8 & 1  & 0.001 & 2.6581E-01 & 3.01 & 5.06 & 8.07 & Fig.~9a \\

589: 40   & 15.8 & 30 & 0.001 & 7.6394E-01 & 4.05 & 4.02 & 8.07 & Fig.~9b \\

590: 40   & 15.8 & 30 & 0.001 & 9.2867E-02 & 5.03 & 3.04 & 8.07 & Fig.~9b \\

591: 40   & 15.8 & 1  & 0.004 & 3.7283E-01 & 2.50 & 5.57 & 8.07 & Fig.~10a \\

592: 40   & 15.8 & 1  & 0.004 & 2.6649E-02 & 3.03 & 5.04 & 8.07 & Fig.~10a \\

593: 40   & 15.8 & 30 & 0.004 & 9.8590E-01 & 3.03 & 5.04 & 8.07 & Fig.~10b \\

594: 40   & 15.8 & 30 & 0.004 & 2.0354E-01 & 4.02 & 4.05 & 8.07 & Fig.~10b \\

595: 30   & 11   & 1  & 0.004 & 1.4667E-01 & 2.52 & 5.49 & 8.01 & Fig.~11a \\

596: 30   & 11   & 1  & 0.004 & 1.7806E-03 & 3.04 & 4.97 & 8.01 & Fig.~11a \\

597: 30   & 11   & 20 & 0.004 & 5.7203E-01 & 3.04 & 4.97 & 8.01 & Fig.~11b \\

598: 30   & 11   & 20 & 0.004 & 3.0448E-02 & 4.15 & 3.86 & 8.01 & Fig.~11b \\

599: 40   & 15.1 & 1  & 0.02 & 7.3775E-01 & 1.46 & 6.63 & 8.09 & Fig.~12a \\

600: 40   & 15.1 & 1  & 0.02 & 3.1602E-03 & 8.09 & 0    & 8.09 & Fig.~12a \\

601: 40   & 15.1 & 30 & 0.02 & 8.4572E-01 & 1.74 & 6.35 & 8.09 & Fig.~12b \\

602: 40   & 15.1 & 30 & 0.02 & 2.8515E-03 & 8.09 & 0    & 8.09 & Fig.~12b \\

603: 40   & 16 & 10 & 0.02 & 1.4337E-02 & 2.41 & 5.15 & 7.56 & Fig.~13a \\

604: 40   & 16 & 10 & 0.02 & 0	    & 7.56 & 0    & 7.56 & Fig.~13a \\

605: 40   & 16 & 10 & 0.02 & 5.1974E-01 & 2.41 & 5.15 & 7.56 & Fig.~13b \\

606: 40   & 16 & 10 & 0.02 & 1.2716E-01 & 7.56 & 0    & 7.56 & Fig.~13b \\

607: \enddata

608: 

609: \tablenotetext{a}{$M_{{\rm BH},i}$ is the mass of the remnant after the

610: explosion and before the secondary started to transfer matter onto the

611: compact object. $M_{{\rm BH},f}=8.53\pm0.60$ \Msun is the observed

612: mass of the black hole.}

613: 

614: \tablecomments{Supernova and hypernova explosion models used in

615: Figs.~9-13. The quantities shown are the main-sequence mass, the mass

616: of the He core, the explosion energy, $E_{51}=E_K/10^{51} \,{\rm

617: erg}$, the ejected Fe mass, the metallicity of the model, the

618: mass-cut, the mass of the fallback matter, and the final remnant mass.

619: The mass are in units of \Msuno.}

620: 

621: \label{tbl4}

622: \end{deluxetable}

623: 

624: The abundances of Ti and Si were mainly derived from several

625: lines in the spectral region 5920--5960 {\AA} where some telluric lines

626: are present. However, since 74 spectra with different radial

627: velocities in the range $\pm710$ \kms were combined to generate the

628: average spectrum of the secondary star, these telluric lines must have

629: been smoothed out. 

630: 

631: The Mg abundance was derived from one spectral line (see

632: Fig.~\ref{fig5}) and the error associated with the dispersion of the

633: measurements, $\sigma$, was assumed to be the average dispersion of  

634: Fe, Ca, and Ni abundances, and in this case, $\Delta_{\sigma}=\sigma$.

635: The same prescription was adopted for the analysis of Al and Li (see

636: Fig.~1 of Paper~I). The best fit to the \ion{Li}{1} 6708~{\AA} feature

637: provides an LTE abundance of $\log \epsilon(\mathrm{Li})_{\rm

638: LTE}=1.61 \pm 0.25$. We estimated the non-LTE abundance

639: correction\footnote{$\Delta_\mathrm{NLTE}=\log

640: \epsilon(\mathrm{X})_\mathrm{NLTE}-\log

641: \epsilon(\mathrm{X})_\mathrm{LTE}.$} for this element from the 

642: theoretical LTE and non-LTE curves of growth in Pavlenko \& Magazz\`u

643: (1996). We found $\Delta_\mathrm{NLTE}= 0.17$. Due to the weakness of

644: the absorption we consider this abundance estimate given in

645: Table~\ref{tbl2} as an upper limit. 

646: 

647: \begin{figure*}[ht!]

648: %\epsscale{.80}

649: %\plotone{f8.eps}

650: \centering

651: \includegraphics[width=16.cm,angle=0]{f8.eps}

652: \caption{\footnotesize{Chemical composition of the secondary

653: atmosphere contaminated with nucleosynthetic products of a 40~\Msun

654: spherically symmetric core-collapse SN explosion model ($M_{\rm He}

655: \approx 15$--16~\Msuno) for different metallicities and mass-cuts, $M_{\rm

656: cut}$.}} 

657: \label{fig8}

658: \end{figure*}

659: 

660: \section{Discussion}

661: 

662: As already discussed in Paper~I, the Fe abundance of the secondary

663: star is slightly higher than solar, but similar to that of many stars

664: in the solar neighborhood. The abundances of other elements listed in

665: Table~\ref{tbl2} relative to iron are compared in Fig.~\ref{fig7} with

666: the Galactic trends of these elements in the relevant range of

667: metallicities. Moderate anomalies are found only for Al. In

668: Table~\ref{tbl3} we show the element abundance ratios in the secondary

669: star in \mbox{XTE J1118+480} and the average values in stars with

670: iron content in the range $0.01 < \mathrm{[Fe/H]} < 0.35$, the

671: comparison sample, corresponding to a 1$\sigma$ uncertainty in the

672: iron abundance of the companion star. Whereas Ca and Ti are consistent

673: with the average values of the comparison sample, Ni and Si, at

674: 1$\sigma$, and especially Al, at 2$\sigma$, appear to be more abundant

675: than the average values of the stars in the comparison sample.        

676: 

677: The present location and space-velocity components ($U$, $V$) of the

678: system might suggest that the system belongs to the Galactic halo, but

679: the derived metallicity makes this possibility less likely. One could 

680: include the metallicity in the expression given in Bensby et al. (2005) to

681: estimate the relative likelihoods that a star belongs to the Galactic

682: thin disk, thick disk, and halo. The equations could be written as

683: follows:

684: 

685: \small

686: \begin{eqnarray}

687: P_{\rm thin-disk}=f_{\rm D}\frac{P_{\rm D}}{P},\nonumber\\

688: P_{\rm thick-disk}=f_{\rm TD}\frac{P_{\rm TD}}{P},\nonumber\\

689: P_{\rm halo}=f_{\rm H}\frac{P_{\rm H}}{P},\nonumber\\

690: P_i=K_i \times \exp\left[-\frac{U2_{\rm

691: LSR}}{2\sigma2_{U_i}}

692: -\frac{(V_{\rm LSR}-V_{{\rm asym},i})2}{2\sigma2_{V_i}}\right]\nonumber\\

693:  \times \exp\left[-\frac{W2_{\rm LSR}}{2\sigma2_{W_i}}

694: -\frac{({\rm [Fe/H]}-{\rm [Fe/H]}_{{\rm asym},i})2}

695: {2\sigma2_{\rm [Fe/H]_i}}\right],\nonumber\\

696: P=f_{\rm D}P_{\rm D}+f_{\rm TD}P_{\rm TD}+f_{\rm H}P_{\rm H},\nonumber\\

697: ~~{\rm and}~~ K_i=\frac{1}{(2\pi)2\sigma_{U_i}\sigma_{V_i}\sigma_{W_i}\sigma_{{\rm [Fe/H]}_i}},\nonumber  

698: \end{eqnarray}

699: \normalsize

700: 

701: \noindent

702: where the subscript $i$ indicates the three populations $D$

703: (thin disk), $TD$ (thick disk), and $H$ (halo). The total probability,

704: $P$, takes into account the fraction of stars belonging to each

705: population in the solar neighborhood ($f_{\rm D}=0.94$, $f_{\rm TD}=0.06$, 

706: and $f_{\rm H}=0.0015$; Bensby et al. 2003). The velocity distributions of each

707: population with respect to the local standard of rest (LSR) are

708: centered at zero except for the component $V_{\rm LSR}$, whose

709: center is displaced according to $V_{{\rm asym},i}$ (with $V_{{\rm

710: asym},D}=-15$\kmso, $V_{{\rm asym},TD}=-46$ \kms, and $V_{{\rm

711: asym},H}=-220$ \kmso). The metallicity distributions have been

712: characterized with ${\rm [Fe/H]}_{\rm asym,D}=-0.1$, ${\rm

713: [Fe/H]}_{\rm asym,TD}=-0.7$, and ${\rm [Fe/H]}_{\rm asym,H}=-1.4$, 

714: as well as with $\sigma_{{\rm [Fe/H]}_{\rm D}}=0.2$, 

715: $\sigma_{{\rm [Fe/H]}_{\rm TD}}=0.24$, and

716: $\sigma_{{\rm [Fe/H]}_{\rm H}}=0.5$, according to Allende Prieto et al.

717: (2004, 2006). Thus, the probability that a star with the Galactic

718: space velocity components of this system (Mirabel et al.\ 2001;

719: $U=-105\pm16$ {\kms}, $V=-98\pm16$ {\kms}, $W=-21\pm10$ {\kms}) 

720: and metallicity $\mathrm{[Fe/H]} = 0.18$ belongs to the Galactic halo 

721: is less than 0.1\%. Moreover, the kinematics alone suggest thick-disk

722: rather than halo membership, although the high metallicity of the

723: secondary star favors thin-disk membership. However, the system could

724: also have originated in a satellite galaxy. In particular, its

725: galactocentric orbit (Mirabel et al. 2001) is marginally consistent

726: with the equatorial orbit of the stream of the dwarf galaxy in Canis

727: Major, although its present Galactic latitude differs by more than

728: $40^\circ$ from the $l-b$ distribution of the remnant of this dwarf

729: galaxy, accreted by the Milky Way (Martin et al. 2004). 

730: 

731:  The system could also have originated as a consequence of an

732: encounter of an ancient black hole of the Galactic halo with a 

733: single star or a binary of two solar-type stars of the Galactic disk. 

734: However, this possibility is very unlikely due to the extremely low 

735: density of stars in the disk ($\sim$0.006 stars/pc$^3$; Mihalas

736: \& Binney 1981). The orbit of the system integrated backward in time

737: never crossed the Galactic plane through the inner 2~kpc (Gualandris

738: et al. 2005), so high-density regions near the Galactic center are

739: discarded. Portegies Zwart et al. (1997b) have modeled the encounter

740: of black holes in high-density systems. In high-density systems, where

741: the density of stars is $\sim4 \times 10^6$ stars/pc$^3$ ($10^9$ times higher 

742: than in the Galactic disk), a black hole spends 1.5~Gyr before it suffers 

743: a tidal capture by a main-sequence star. In addition, the cases where a

744: black hole can capture an isolated star or one star of a binary require

745: very stringent constraints on the closest approach and impact velocity

746: (Benz \& Hills 1992; Hills 1991). All these reasons make this

747: possibility very unlikely.

748: 

749: Therefore, in conclusion, the present location, velocity, and

750: metallicity of the secondary star in \mbox{XTE J1118+48} suggest that

751: the black hole formed in a supernova/hypernova explosion that occurred

752: within the binary system. This explosive event must have either

753: provided a kick to the system if it was formed in the thin disk, or

754: enriched significantly the atmosphere of the secondary star if the

755: system formed in the thick disk or halo. 

756: 

757: The present orbital separation between the compact object and the

758: secondary star has been estimated to be $a_{\rm c} \approx

759: 2.67$ {$R_\odot$} (Gelino et al. 2006). Thus, the secondary star could

760: have captured a significant amount of the ejected matter in the

761: SN explosion that formed the compact object. The chemical

762: composition of the secondary star may provide information on the

763: chemical composition of the progenitor of the compact object, and

764: therefore on the formation region (thin disk, thick disk, or halo) of

765: the binary system. We will now discuss the possibility that the SN

766: explosion of the massive progenitor enriched the secondary star from

767: different initial metallicities.

768: 

769: \begin{figure*}[ht!]

770: %\epsscale{.80}

771: %\plottwo{f9a.eps}{f9b.eps}

772: \centering

773: \includegraphics[width=8.cm,angle=0]{f9a.eps}

774: \includegraphics[width=8.cm,angle=0]{f9b.eps}

775: \caption{\footnotesize{Left panel: Observed abundances (filled circles

776: with error bars) in comparison with the expected abundances in the

777: secondary star after having captured 17.5\% of the matter ejected  

778: within the solid angle subtended by the secondary from a metal-poor

779: ($Z=0.001$) 40~\Msun spherically symmetric supernova explosion ($M_{\rm

780: He} = 15.8$ \Msuno) with $E_K = 10^{51}$ erg for two different

781: mass-cuts, $M_{\rm cut} = 2.51$~\Msun (solid line with open circles)

782: and $M_{\rm cut} = 3.01$~\Msun (dashed-dotted line with open circles).

783: The initial abundances of the secondary star were adopted for the

784: average abundances of halo stars with [Fe/H] $ = -1.4 \pm 0.2$, 

785: and the initial orbital distance was $a_{c,i} \approx 6$ \Rsuno. Right

786: panel: same as left panel but for a spherically symmetric

787: hypernova explosion ($E_K = 30 \times 10^{51}$ erg) for two different

788: mass-cuts, $M_{\rm cut} = 4.05$~\Msun (solid line with open circles)

789: and $M_{\rm cut} = 5.03$~\Msun (dashed-dotted line with open circles).}}

790: \label{fig9}

791: \end{figure*}

792: 

793: \subsection{Spherical Explosion}

794: 

795: Gelino et al. (2006) derived a current black hole mass of $M_{{\rm

796: BH},f}= 8.53\pm 0.60$~\Msun and secondary mass of $M_{2,f}= 0.37\pm

797: 0.03$~\Msuno. Using near-UV spectroscopic observations of the 

798: accretion disk, Haswell et al. (2002) suggested that the material

799: accreted onto the compact object is substantially CNO processed,

800: indicating that the initial mass of the secondary star could have

801: been as high as $\sim 1.5$~\Msuno. Hereafter we will adopt an initial

802: secondary mass of $M_{2,i} = 1$~\Msun and a black hole mass of $M_{{\rm

803: BH},i} = 8$~\Msuno. 

804: 

805: A binary system such as \mbox{XTE J1118+480} will

806: survive a spherical SN explosion if the ejected mass $\Delta

807: M=M_{\rm He}-M_{{\rm BH},i} \le (M_{\rm He}+M_{2,i})/2$ (Hills 1983).

808: This implies a mass of the He core before the SN explosion of $M_{\rm

809: He} \le 17$ \Msuno. Using the expressions given by Portegies Zwart et

810: al. (1997a, and references therein), we inferred a He core radius of

811: $R_{\rm He} \approx 2-3$ \Rsun for He core masses in the range $M_{\rm He}

812: \approx 8.5-17$ \Msuno. We will assume that the post-SN orbital

813: separation after tidal circularization of the orbit is in the range

814: $a_{c,i} \approx 4-6$ \Rsuno, since the secondary must have experienced

815: mass and angular momentum losses during the binary evolution until

816: reaching its present configuration, with $a_{c,f} \approx 2.67$ \Rsuno.

817: 

818: Assuming a pre-SN circular orbit and an instantaneous spherically

819: symmetric ejection (that is, shorter than the orbital period), one can

820: estimate the pre-SN orbital separation, $a_0$, using the relations

821: given by van den Heuvel \& Habets (1984): $a_0=a_{c,i}\mu_f$, where

822: $\mu_f=(M_{{\rm BH},i}+M_{2,i})/(M_{\rm He}+M_{2,i})$. We find $a_0

823: \approx 3$--5, essentially depending on the adopted values of $M_{\rm He}$ 

824: and $a_{c,i}$. At the time of the SN explosion ($\sim$5--6 Myr; Brunish \&

825: Truran 1982), a 1~\Msun secondary star, still in its pre-main-sequence

826: evolution, has a radius $R_{2,i} \approx 1.3$

827: \Rsun and a convective zone of mass $M_{\rm cz} \approx 0.652$ \Msun

828: (D'Antona \& Mazzitelli 1994). Thus, the amount of

829: mass deposited on the secondary can be estimated as $m_{\rm

830: cap}=\Delta M (\pi R_{2,i}^2/4 \pi a_02)f_{\rm cap}$ \Msuno, where

831: $f_{\rm cap}$ is the fraction of mass, ejected within the solid angle

832: subtended by the secondary star, that is eventually captured. We

833: assume that the captured mass, $m_{\rm cap}$, is efficiently mixed with

834: the mass of the convective zone, $M_{\rm cz}$.

835: 

836: We compute the expected abundances in the atmosphere of the secondary

837: star after the pollution from the progenitor of the compact

838: object as in Gonz\'alez Hern\'andez et al. (2004). We used 40~\Msun

839: spherically symmetric core-collapse explosion models ($M_{\rm 

840: He} \approx 15.1$--16.1~\Msuno) for different metallicities ($Z=$ 0, 0.001,

841: 0.004, 0.02) and explosion energies (Umeda \& Nomoto, 2002, 2005;

842: Tominaga, Umeda \& Nomoto 2007). These models imply $\Delta M \approx 

843: 7$--8~\Msun and need small capture efficiencies of $f_{\rm

844: cap} \la 0.1$ (i.e., 10\%) to increase significantly the metal content of

845: the secondary star. On the other hand, the use of 30~\Msun models

846: ($M_{\rm He} \approx 8.5$--11.2~\Msuno) would require $f_{\rm cap} \approx

847: 0.9$--1. These models would also provide a different mass fraction of

848: each element at each value of the mass-cut (the mass that initially 

849: collapsed forming the compact remnant). For more details in the models, 

850: see Tominaga, Umeda \& Nomoto (2007).

851: 

852: The explosion energy, $E_K = 1 \times 10^{51}$ erg and $E_K = 

853: (20-30) \times 10^{51}$ erg for the supernova (SN) and hypernova (HN)

854: models (respectively), is deposited instantaneously in the central

855: region of the progenitor core to generate a strong shock wave. The

856: subsequent propagation of the shock wave is followed through a

857: hydrodynamic code (Umeda \& Nomoto 2002, and references therein). In

858: our simple model, we have assumed different mass-cuts, fallback

859: masses, and a mixing factor of 1 which assumes that all fallback

860: matter is well mixed with the ejecta. The amount of fallback, $M_{\rm

861: fall}$, is the difference between the final remnant mass, $M_{{\rm

862: BH},i}$, and the initial remnant mass of the explosion, $M_{\rm cut}$.

863: We should recall here the ejected mass, $\Delta M$, which is equal to

864: $M_{\rm He}-M_{{\rm BH},i}$, where $M_{\rm He}$ is the mass of the He

865: core.

866: 

867: \begin{figure*}[ht!]

868: %\epsscale{.80}

869: %\plottwo{f10a.eps}{f10b.eps}

870: \centering

871: \includegraphics[width=8.cm,angle=0]{f10a.eps}

872: \includegraphics[width=8.cm,angle=0]{f10b.eps}

873: \caption{\footnotesize{Left panel: Observed abundances (filled circles

874: with error bars) in comparison with the expected abundances in the

875: secondary star after having captured the 15.5\% of the matter ejected  

876: within the solid angle subtended by the secondary from a metal-poor

877: ($Z = 0.004$) 40~\Msun spherically symmetric supernova explosion ($M_{\rm

878: He} = 15.8$ \Msuno) with $E_K = 10^{51}$ erg for two different

879: mass-cuts, $M_{\rm cut} = 2.50$~\Msun (solid line with open circles)

880: and $M_{\rm cut} = 3.03$~\Msun (dashed-dotted line with open circles).

881: The initial abundances of the secondary star were adopted for the

882: average abundances of thick-disk stars with [Fe/H] $= -0.7 \pm 0.2$,

883: and the initial orbital distance was $a_{c,i} \approx 6$~\Rsuno.

884: Right panel: same as left panel but for a spherically symmetric 

885: hypernova explosion ($E_K = 30 \times 10^{51}$ erg) for two different

886: mass-cuts, $M_{\rm cut} = 3.03$~\Msun (solid line with open circles)

887: and $M_{\rm cut} = 4.02$~\Msun (dashed-dotted line with open circles).}}

888: \label{fig10}

889: \end{figure*}

890: 

891: We use SN/HN models to provide us with the yields of the 

892: explosion before radiative decay of element species. We then 

893: compute the integrated, decayed yields of the ejecta by adopting 

894: a mass-cut and by mixing all the material above the mass-cut. 

895: Finally, we calculate the composition of the matter captured by

896: the secondary star, and we mix it with the material of its convective

897: envelope.

898: 

899: In Fig.~\ref{fig8} we show the expected abundances of the secondary 

900: star after contamination from the nucleosynthetic products of the SN

901: explosion ($E_K = 1 \times 10^{51}$ erg) of $M_{\rm He} \approx 

902: 15$--16~\Msun progenitor stars. The initial abundances of the 

903: secondary star have been estimated from the average abundances 

904: of halo stars with a metallicity of [Fe/H] $\approx -2.2$ (from 

905: Cayrel et al. 2004) and [Fe/H] $\approx -1.4$ (from

906: Jonsell et al. 2005), thick-disk stars with [Fe/H] $\approx -0.7$ (from

907: Jonsell et al. 2005), and thin-disk stars with [Fe/H] $\approx 0$ and

908: [Fe/H] $\approx 0.18$ (from Gilli et al. 2006). For each simulation at 

909: each metallicity, we have fixed the $f_{\rm cap}$ factor at a value that

910: allows us to approximately match the observed aluminum abundance. 

911: 

912: We should remark that for a given model at a given metallicity, once the

913: capture efficiency is fixed, the aluminum (Al) abundance in the secondary

914: star hardly depends on the mass-cut. Thus, the fact that the Al

915: abundance in Fig.~\ref{fig8} is almost constant with metallicity is

916: because we have adopted different $f_{\rm cap}$ factors for different

917: metallicities. These $f_{\rm cap}$ factors are higher for lower

918: metallicities since greater amounts of captured material are needed to

919: achieve the observed Al abundance in the secondary star. For other

920: figures in this paper, the $f_{\rm cap}$ factor was changed until an

921: abundance of [Al/H]$ \approx 0.5$ dex was obtained, compatible with

922: the observed abundance within the uncertainties. 

923: 

924: \begin{deluxetable}{lcccccccccccc}

925: \centering

926: \tabletypesize{\scriptsize}

927: %\rotate

928: \tablecolumns{7}

929: \tablecaption{Metal-Poor Supernova/Hypernova Explosion Models in \mbox{XTE

930: J1118+480}}

931: \tablewidth{0pt}

932: \tablehead{ & & & \multicolumn{4}{c}{${\rm [X/H] \: EXPECTED}$\tablenotemark{d}}\\

933: \cline{4-7}\\

934:  & & & \multicolumn{2}{c}{Supernova} &

935: \multicolumn{2}{c}{Hypernova} \\

936: \cline{4-5} \cline{6-7} \\

937: ELEMENT & ${\rm [X/H]\:\rm OBSERVED}$\tablenotemark{a} & ${\rm [X/H]}_{0}$\tablenotemark{b} &

938: $M_{\rm cut, low}$\tablenotemark{c} & $M_{\rm cut, up}$ & $M_{\rm cut, low}$ & $M_{\rm cut, up}$}

939: \startdata

940:  & & & \multicolumn{4}{c}{40 \Msun explosion model of $Z=0.001$} \\

941: \noalign{\smallskip}

942: \tableline

943: \noalign{\smallskip}

944: Mg & 0.35 & -1.08 &  0.83 &  0.85 &  0.75 &  0.79  \\

945: Al & 0.60 & -1.27 &  0.52 &  0.53 &  0.39 &  0.42  \\

946: Si & 0.37 & -1.13 &  0.88 &  0.89 &  1.08 &  1.09  \\

947: Ca & 0.15 & -1.12 &  0.74 &  0.75 &  1.09 &  1.01  \\

948: Ti & 0.32 & -1.15 &  0.14 &  0.11 &  0.54 &  0.12  \\

949: Fe & 0.18 & -1.46 &  0.47 &  0.14 &  0.63 & -0.22  \\

950: Ni & 0.30 & -1.48 &  1.09 & -0.52 &  0.14 & -0.81  \\

951: O  & \nodata & -0.76 &  0.75 &  0.77 &  0.69 &  0.73  \\

952: C  & \nodata & -0.86 & -0.54 & -0.53 & -0.60 & -0.58  \\

953: \noalign{\smallskip}

954: \tableline

955: \noalign{\smallskip}

956:  & & & \multicolumn{4}{c}{40 \Msun explosion model with $Z=0.004$}  \\

957: \noalign{\smallskip}

958: \tableline

959: \noalign{\smallskip}

960: Mg & 0.35 & -0.41 &  0.58 &  0.59 &  0.61 &  0.64 \\

961: Al & 0.60 & -0.46 &  0.51 &  0.52 &  0.41 &  0.44 \\

962: Si & 0.37 & -0.50 &  0.62 &  0.60 &  0.90 &  0.93 \\

963: Ca & 0.15 & -0.54 &  0.57 &  0.42 &  0.86 &  0.87 \\

964: Ti & 0.32 & -0.50 &  0.16 & -0.18 &  0.56 &  0.27 \\

965: Fe & 0.18 & -0.75 &  0.26 & -0.52 &  0.67 &  0.07 \\

966: Ni & 0.30 & -0.73 &  0.08 & -0.53 &  0.43 & -0.38 \\

967: O  & \nodata & -0.26 &  0.71 &  0.73 &  0.67 &  0.69  \\

968: C  & \nodata & -0.35 &  0.10 &  0.11 &  0.06 &  0.08  \\

969: \noalign{\smallskip}

970: \tableline

971: \noalign{\smallskip}

972:  & & & \multicolumn{4}{c}{30 \Msun explosion model with $Z=0.004$} \\

973: \noalign{\smallskip}

974: \tableline

975: \noalign{\smallskip}

976: Mg & 0.35 & -0.41 &  0.60 &  0.62 &  0.57 &  0.62  \\

977: Al & 0.60 & -0.46 &  0.52 &  0.54 &  0.40 &  0.45  \\

978: Si & 0.37 & -0.50 &  0.77 &  0.54 &  1.02 &  0.95  \\

979: Ca & 0.15 & -0.54 &  0.72 &  0.17 &  1.03 &  0.82  \\

980: Ti & 0.32 & -0.50 &  0.18 & -0.41 &  0.66 &  0.01  \\

981: Fe & 0.18 & -0.75 &  0.14 & -0.71 &  0.71 & -0.32  \\

982: Ni & 0.30 & -0.73 & -0.43 & -0.54 &  0.40 & -0.40  \\

983: O  & \nodata & -0.26 &  0.75 &  0.78 &  0.69 &  0.74  \\

984: C  & \nodata & -0.35 &  0.02 &  0.03 & -0.04 & -0.01  \\

985: \enddata

986: \tablenotetext{a}{Observed abundances of the

987: secondary star in \mbox{XTE J1118+480}.}

988: 

989: \tablenotetext{b}{Initial abundances assumed for

990: the secondary star in \mbox{XTE J1118+480}, see text. The initial

991: C and O abundances of the thick-disk model (${\rm [X/H]}_{0}=-0.7$)

992: were adopted from Ecuvillon et al. (2004, 2006).}

993: 

994: \tablenotetext{c}{$M_{\rm cut, low}$ and $M_{\rm cut, up}$ are the

995: lower and upper mass-cut adopted in the model computation. See the

996: exact value in the captions of Figs.~9-11.} 

997: 

998: \tablenotetext{d}{Expected abundances of the secondary star.}

999: 

1000: \tablecomments{Expected abundances in the secondary atmosphere

1001: contaminated with nucleosynthetic products of metal-poor explosion

1002: models for two different explosion energies and mass-cuts, presented

1003: in Figs.~9-11.} 

1004: 

1005: \label{tbl5}

1006: \end{deluxetable}

1007: 

1008:    From Fig.~\ref{fig8} one can see that while Mg and Al remain 

1009: significantly enhanced (above solar abundances) for all mass-cuts at 

1010: all metallicities, Si, Ca, Ti, Fe, and Ni are quite sensitive to the

1011: mass-cut of the model. Thus, for metallicities below $-1.4$ dex, only

1012: $M_{\rm cut} \la 3$~\Msun is able to enhance sufficiently the Ti, Fe, and

1013: Ni, whereas Ca and Si remain quite overabundant for $M_{\rm cut} \la 

1014: 4$~\Msun and 5~\Msun (respectively) at all metallicities. We should note

1015: that for metallicities [Fe/H] $\la -1.4$ dex (i.e., $Z \la 0.001$), the

1016: expected abundances in the secondary star do not depend on the initial

1017: abundances but on the SN yields. Thus, we found different expected

1018: abundances of the secondary between [Fe/H] $\approx -1.4$ dex and

1019: [Fe/H] $\approx -2.2$ dex because we used the SN $Z = 0.001$ model and 

1020: the SN $Z = 0$ model, respectively. 

1021: 

1022: \subsection{Formation in the Halo}

1023: 

1024: The kinematics of the system, at least $U$ and $V$, resemble those

1025: of halo stars, and therefore a significant kick during the black

1026: hole formation process appears unnecessary. Thus, a spherically

1027: symmetric SN explosion would provide the desired kick velocity to

1028: match the current velocity components of the system from initial

1029: velocities similar to those of halo stars. However, due to the 

1030: extremely low metallicities of halo stars, it is required that the 

1031: secondary captured enough matter from the ejecta to reach the current 

1032: abundances. 

1033: 

1034: \begin{figure*}[ht!]

1035: %\epsscale{.80}

1036: %\plottwo{f11a.eps}{f11b.eps}

1037: \centering

1038: \includegraphics[width=8.cm,angle=0]{f11a.eps}

1039: \includegraphics[width=8.cm,angle=0]{f11b.eps}

1040: \caption{\footnotesize{Left panel: Observed abundances (filled circles

1041: with error bars) in comparison with the expected abundances in the

1042: secondary star after having captured 90\% of the matter ejected  

1043: within the solid angle subtended by the secondary from a metal-poor

1044: ($Z = 0.004$) 30~\Msun spherically symmetric supernova explosion 

1045: ($M_{\rm He} = 11$ \Msuno) with $E_K = 10^{51}$ erg for two different

1046: mass-cuts, $M_{\rm cut} = 2.52$~\Msun (solid line with open circles)

1047: and $M_{\rm cut} = 3.04$~\Msun (dashed-dotted line with open circles).

1048: The initial abundances of the secondary star were adopted for the

1049: average abundances of the thick-disk stars with [Fe/H] $= -0.7 \pm 0.2$,

1050: and the initial orbital distance was $a_{c,i} \approx 6$ \Rsuno.

1051: Right panel: same as left panel but for a spherically symmetric

1052: hypernova explosion ($E_K = 20 \times 10^{51}$ erg) for two different

1053: mass-cuts, $M_{\rm cut} = 3.04$~\Msun (solid line with open circles)

1054: and $M_{\rm cut} = 4.15$~\Msun (dashed-dotted line with open circles).}}

1055: \label{fig11}

1056: \end{figure*}

1057: 

1058: In Fig.~\ref{fig9} we compare the observed abundances

1059: with the expected abundances in the secondary star for two different

1060: explosion energies and mass-cuts using a metal-poor 40~\Msun

1061: progenitor model. In these simulations, the mass-cut is sampled

1062: according to the mass bins given in the explosion models, from

1063: $\sim 1.5$~\Msun to 8~\Msun in steps of $\sim 0.5$ and 1~\Msun. The

1064: adopted mass-cuts shown in Fig.~\ref{fig9} were selected from those

1065: which provided better fits to the observed abundances. The initial

1066: abundances of the secondary have been estimated from the average 

1067: abundances of halo stars with [Fe/H] $\approx -1.4$ from Jonsell et 

1068: al. (2005). The value of $f_{\rm cap}$ has

1069: been adjusted until the observed aluminum abundance was roughly

1070: reproduced, providing $f_{\rm cap} \approx 0.17$. The parameters of

1071: the explosion models used in Fig.~\ref{fig9}~--~\ref{fig13} are 

1072: given in Table~\ref{tbl4}. In Table~\ref{tbl5} we show the

1073: expected abundances of the secondary star after contamination from

1074: metal-poor SN/HN explosion models.

1075: 

1076: With a supernova model, shown in the left panel of Fig.~\ref{fig9}, 

1077: it is not possible to recover the observed abundances because 

1078: these models produce too much Mg and Ca regardless of the Fe 

1079: abundance obtained; in addition, the Ni abundance is

1080: strongly dependent on the mass-cut. The hypernova model, displayed in

1081: the right panel, makes it even more difficult to fit the observed

1082: abundances since this model creates too much Ca relative to Fe and

1083: Ni. As inferred from Fig.~\ref{fig8}, a model with initial abundances 

1084: at [Fe/H] $\approx -2.2$ (Cayrel et al. 2004) would also not be

1085: successful in reproducing the observed abundances due to the

1086: tremendous and different sensitivity of each element to the mass-cut. 

1087: 

1088: In conclusion, neither of these very metal-poor models  

1089: is able to reproduce the observed abundances. We also try to fit

1090: the observed abundances using a metal-poor 30~\Msun explosion model,

1091: but the agreement is even worse than for 40~\Msun models, again due 

1092: to the strong sensitivity of the each element abundance to the

1093: mass-cut. The comparison of the observed abundances with SN yields

1094: allows us to rule out a Galactic halo origin for this black hole

1095: binary. 

1096: 

1097: \subsection{Formation in the Thick Disk}

1098: 

1099: \begin{deluxetable}{lcccccccccccc}

1100: \centering

1101: \tabletypesize{\scriptsize}

1102: %\rotate

1103: \tablecolumns{7}

1104: \tablecaption{Metal-Rich Supernova/Hypernova Explosion Models in \mbox{XTE

1105: J1118+480}}

1106: \tablewidth{0pt}

1107: \tablehead{ & & & \multicolumn{4}{c}{${\rm [X/H] \: EXPECTED}$\tablenotemark{d}}\\

1108: \cline{4-7}\\

1109: ELEMENT & ${\rm [X/H]\:\rm OBSERVED}$\tablenotemark{a} & ${\rm [X/H]}_{0}$\tablenotemark{b} &

1110: $M_{\rm cut, low}$\tablenotemark{c} & $M_{\rm cut, up}$ & $M_{\rm cut, low}$ & $M_{\rm cut, up}$}

1111: \startdata

1112:  & & & \multicolumn{4}{c}{Spherical explosion model of $Z=0.02$} \\

1113: \noalign{\smallskip}

1114: \tableline

1115: \noalign{\smallskip}

1116:  & & & \multicolumn{2}{c}{Supernova} &

1117: \multicolumn{2}{c}{Hypernova} \\

1118: \noalign{\smallskip}

1119: \cline{4-5} \cline{6-7} \\

1120: \noalign{\smallskip}

1121: Mg & 0.35 & 0.17 &  0.32 &  0.30 &  0.32 &  0.30  \\

1122: Al & 0.60 & 0.29 &  0.50 &  0.50 &  0.48 &  0.50  \\

1123: Si & 0.37 & 0.12 &  0.22 &  0.14 &  0.28 &  0.14  \\

1124: Ca & 0.15 & 0.02 &  0.09 &  0.02 &  0.15 &  0.02  \\

1125: Ti & 0.32 & 0.20 &  0.24 &  0.21 &  0.34 &  0.21  \\

1126: Fe & 0.18 & 0.18 &  0.30 &  0.18 &  0.32 &  0.18  \\

1127: Ni & 0.30 & 0.13 &  0.35 &  0.15 &  0.32 &  0.15  \\

1128: O  & \nodata & 0.18 &  0.33 &  0.33 &  0.33 &  0.33  \\

1129: C  & \nodata & 0.11 &  0.16 &  0.16 &  0.15 &  0.16  \\

1130: \noalign{\smallskip}

1131: \tableline

1132: \noalign{\smallskip}

1133:  & & & \multicolumn{4}{c}{Aspherical explosion model with $Z=0.02$}  \\

1134: \noalign{\smallskip}

1135: \tableline

1136: \noalign{\smallskip}

1137:  & & & \multicolumn{2}{c}{Angle\tablenotemark{e} $=0^\circ-15^\circ$}

1138:  & \multicolumn{2}{c}{Angle $=0^\circ-90^\circ$} \\

1139: \noalign{\smallskip}

1140: \cline{4-5} \cline{6-7} \\

1141: \noalign{\smallskip}

1142: Mg & 0.35 & 0.17 &  0.34 &  0.32 &  0.34 &  0.35 \\

1143: Al & 0.60 & 0.29 &  0.51 &  0.48 &  0.50 &  0.51 \\

1144: Si & 0.37 & 0.12 &  0.24 &  0.18 &  0.35 &  0.26 \\

1145: Ca & 0.15 & 0.02 &  0.05 &  0.02 &  0.22 &  0.11 \\

1146: Ti & 0.32 & 0.20 &  0.21 &  0.20 &  0.41 &  0.35 \\

1147: Fe & 0.18 & 0.18 &  0.18 &  0.18 &  0.28 &  0.22 \\

1148: Ni & 0.30 & 0.13 &  0.13 &  0.13 &  0.34 &  0.23 \\

1149: O  & \nodata & 0.18 &  0.37 &  0.35 &  0.38 &  0.39  \\

1150: C  & \nodata & 0.11 &  0.12 &  0.13 &  0.12 &  0.13  \\

1151: \enddata

1152: \tablenotetext{a}{Observed abundances of the

1153: secondary star in \mbox{XTE J1118+480}.}

1154: 

1155: \tablenotetext{b}{Initial abundances assumed for

1156: the secondary star in \mbox{XTE J1118+480}, see text. The initial

1157: C and O abundances of the metal-rich models were adopted from

1158: Ecuvillon et al. (2004, 2006).}

1159: 

1160: \tablenotetext{c}{$M_{\rm cut, low}$ and $M_{\rm cut, up}$ are the

1161: lower and upper mass-cut adopted in the model computation. See the

1162: exact value in the captions of Figs.~12 and 13.} 

1163: 

1164: \tablenotetext{d}{Expected abundances of the secondary star.}

1165: 

1166: \tablenotetext{e}{Angular range, measured from the equatorial plane,

1167: in which all the ejected material in the explosion has been completely

1168: mixed for each velocity point.} 

1169: 

1170: \tablecomments{Expected abundances in the secondary atmosphere

1171: contaminated with nucleosynthetic products of metal-rich explosion

1172: models for two different mass-cuts and symmetries, presented

1173: in Figs.~12 and 13.} 

1174: 

1175: \label{tbl6}

1176: \end{deluxetable}

1177: 

1178: The space-velocity components of the system are comparable to those of

1179: thick-disk stars, and its present location 1.6~kpc above the Galactic

1180: plane is slightly higher that the scale height of thick-disk stars

1181: ($\sim 0.8$--1.3 kpc; Reyl\'e \& Robin 2001; Chen 1997); thus, a 

1182: strong kick during the SN explosion would not be required. A

1183: spherically symmetric SN explosion of a 15.8~\Msun and 11~\Msun He core

1184: provides an impulse of $\sim 60$ and $\sim20$ \kms, respectively. 

1185: However, an enrichment from the typical abundances of thick-disk stars

1186: would have been necessary. In the simulations, the initial abundances

1187: were assumed to be the average values of thick-disk stars with

1188: [Fe/H] $\approx -0.7$ from Jonsell et al. (2005). 

1189: 

1190:   In Fig.~\ref{fig10} we compare the expected abundances from a

1191: 40~\Msun explosion model for two energies and mass-cuts (see also

1192: Table~\ref{tbl5}). The left panel shows the expected abundances

1193: from a SN model which seems to better approach the observed

1194: abundances than those of halo-like metallicities. As in the previous

1195: figure, the $f_{\rm cap}$ factor was changed until an abundance of

1196: [Al/H]$ \approx 0.5$ dex was obtained, compatible with the observed

1197: abundance within the uncertainties. For the mass-cut of 

1198: 2.50~\Msuno, the model roughly fits the observed abundances except for

1199: Ca, which appears enhanced in the model by a factor of 2.6. Better

1200: agreement with Ca could be obtained if the $f_{\rm cap}$ factor is

1201: lowered, but then Al, Ni, and Ti would not match their observed

1202: abundances. If we increase the mass-cut up to 3.03~\Msuno, the Ca

1203: abundance only decreases 0.15 dex whereas Ti, Fe, and Ni decrease by a

1204: factor of 2, 6, and 4 (respectively), which makes the model unable to fit

1205: the observed abundances. The hypernova case offers worse results since

1206: Si and Ca are no longer reproduced, and Fe and Ni cannot be fitted at

1207: the same time.

1208: 

1209: For this metallicity we have also explored the 30~\Msun explosion

1210: model. The results are displayed in Fig.~\ref{fig11}. In the left panel

1211: the SN model is compared with the observations. For the lower

1212: mass-cut, at 2.52~\Msuno, the model provides too high Si and Ca

1213: abundances and too low Ni abundance, whereas for a mass-cut of 

1214: 3.04~\Msuno, the observed Ti, Fe, and Ni abundances are too high in

1215: comparison with the model predictions. In the hypernova case (right

1216: panel of Fig.~\ref{fig11}), at these low mass-cuts the model produces

1217: too large amounts of Si and Ca; in addition, Ti, Fe, and Ni are too 

1218: sensitive to the location of the mass-cut.

1219: 

1220: Despite the fact that none of the explosion models explored is able to

1221: fairly reproduce the observed abundances in the secondary star, all of

1222: the models require vigorous mixing between the fallback matter and

1223: the final ejected matter (Kifonidis et al.\ 2000). For instance, in

1224: the 15.8~\Msun He core model, all of the material above the mass-cut

1225: placed at 2.50~\Msun should be well mixed in order to convey heavy

1226: elements like Fe and Ni to the outer layers of the explosion, which

1227: might make more unlikely a Galactic thick-disk origin for

1228: \mbox{XTE J1118+480}.

1229: 

1230: \begin{figure*}[ht!]

1231: %\epsscale{.80}

1232: %\plottwo{f12a.eps}{f12b.eps}

1233: \centering

1234: \includegraphics[width=8.cm,angle=0]{f12a.eps}

1235: \includegraphics[width=8.cm,angle=0]{f12b.eps}

1236: \caption{\footnotesize{Left panel: Observed abundances (filled circles

1237: with error bars) in comparison with the expected abundances in the

1238: secondary star after having captured the 5\% of the matter ejected  

1239: within the solid angle subtended by the secondary from a metal-poor

1240: ($Z = 0.02$) 40~\Msun spherically symmetric supernova explosion ($M_{\rm

1241: He} = 15.1$~\Msuno) with $E_K = 10^{51}$ erg for two different

1242: mass-cuts, $M_{\rm cut} = 1.46$~\Msun (solid line with open circles)

1243: and $M_{\rm cut} = 8.09$~\Msun (dashed-dotted line with open circles).

1244: The initial abundances of the secondary star were adopted for the

1245: average abundances of thin-disk stars with [Fe/H] $=0.0 \pm 0.2$.

1246: Right panel: same as left panel but for a spherically symmetric 

1247: hypernova explosion ($E_K = 30 \times 10^{51}$ erg) for two different

1248: mass-cuts, $M_{\rm cut} = 1.74$~\Msun (solid line with open circles)

1249: and $M_{\rm cut} = 8.09$~\Msun (dashed-dotted line with open circles).}}

1250: \label{fig12}

1251: \end{figure*}

1252: 

1253: \subsection{Formation in the Thin Disk}

1254: 

1255: In this scenario, the system must have acquired an impulse during the

1256: formation of the black hole that pushes it up from a Galactic plane

1257: orbit to the current halo orbit. The system should have been

1258: accelerated to a peculiar space velocity of $\sim 180$ {\kms}

1259: (Gualandris et al.\ 2005) to reach its present location, requiring 

1260: an asymmetric kick. It has been suggested that such kicks can be

1261: imparted during the birth of nascent neutron stars, due to asymmetric

1262: mass ejection and/or an asymmetry in the neutrino emission (Lai et

1263: al.\ 2001, and references therein). 

1264: 

1265: Podsiadlowski et al. (2002)

1266: proposed that the black hole in \mbox{Nova Sco 1994} could have formed

1267: in a two-stage process where the initial collapse led to the

1268: formation of a neutron star accompanied by a substantial kick and the

1269: final mass of the compact remnant was achieved by matter that fell

1270: back after the initial collapse. However, the black hole mass in 

1271: \mbox{Nova Sco 1994} is estimated to be $\sim 5.4$~\Msun (Beer \&

1272: Podsiadlowski 2002), while the black hole in \mbox{XTE J1118+480} has

1273: a mass of $\sim 8$ \Msun which would require a fallback mass of

1274: $\sim 6.6$ \Msun if we assume $\sim 1.4$~\Msun for a canonical

1275: neutron star. This scenario might take place in the context of

1276: collapsar models where the black hole would be formed in a mild

1277: explosion and substantial fallback (up to $\sim 5$~\Msun is expected) as 

1278: proposed by MacFadyen et al. (2001). Asymmetric mass ejection would

1279: relax this requirement, providing enough impulse to the system to be

1280: launched into its present orbit from the Galactic plane.

1281: 

1282: \begin{figure*}

1283: %\epsscale{.80}

1284: %\plotone{f13.eps}

1285: \centering

1286: \includegraphics[width=8.cm,angle=0]{f13a.eps}

1287: \includegraphics[width=8.cm,angle=0]{f13b.eps}

1288: \caption{\footnotesize{Left panel: Observed abundances (filled circles

1289: with error bars) in comparison with the expected abundances in the

1290: secondary star after having captured the 5\% of the matter ejected  

1291: within the solid angle subtended by the secondary from a

1292: non-spherically symmetric supernova explosion of $E_K =

1293: 10^{52}$ erg for two different mass-cuts, $M_{\rm cut} = 2.41$~\Msun

1294: (solid line with open circles) and $M_{\rm cut} = 7.56$~\Msun

1295: (dashed-dotted line with open circles). This model corresponds to the

1296: matter ejected in the equatorial plane of the primary where we assumed

1297: that the secondary star is located (for more details, see Gonz\'alez

1298: Hern\'andez et al. 2005b). Right panel: same as left panel but in this model

1299: we have assumed complete lateral mixing, where all the material within

1300: given velocity bins is completely mixed. Two simulations

1301: are shown for two different mass-cuts, $M_{\rm cut} = 2.41$~\Msun (solid

1302: line with open circles) and $M_{\rm cut} = 7.56$~\Msun (dashed-dotted

1303: line with open circles).}}  

1304: \label{fig13}

1305: \end{figure*}

1306: 

1307: 

1308: In Fig.~\ref{fig12}, we compare the expected abundances from the

1309: explosion of a 15.1~\Msun He core with the observed abundances of the

1310: secondary star (see also Table~\ref{tbl6}). As in the previous

1311: figures, the $f_{\rm cap}$ factor 

1312: was changed until an abundance of [Al/H]$ \approx 0.5$ dex was obtained,

1313: compatible with the observed abundance within the uncertainties. The

1314: initial abundances were assumed to match the average values of disk

1315: stars with similar iron content, which are provided in Table~\ref{tbl3}. 

1316: For both the SN model (left panel) and the hypernova model 

1317: (right panel), the observed element abundances can be reproduced for

1318: all mass-cuts. This means that neither substantial fallback nor an

1319: efficient mixing process is needed. For mass-cuts above 

1320: $\sim 3$~\Msuno, very little Si, Ca, Ti, Fe, and Ni is ejected; 

1321: thus, the expected abundances of the model essentially reflect

1322: the initial abundances of the secondary star. In contrast, Mg and

1323: Al are not sensitive to the mass-cut and are slightly enhanced due to

1324: the capture of enriched material in the SN explosion. 

1325: 

1326: We also investigate a model with solar initial abundances for the

1327: secondary star, i.e., ${\rm [X/H]}_{0}=0$, and the same explosion

1328: model of solar metallicity ($Z=0.02$), which

1329: would fit all the element abundances if the mass-cut is low enough

1330: ($M_{\rm cut} \la 3$~\Msuno) and mixing is so efficient that

1331: significant amounts of elements which form in the inner layers of the

1332: explosion (such as Ti, Ni, and Fe) are present in the ejecta.  

1333: 

1334: \subsection{Non-Spherical Explosion: Formation in the Thin Disk}

1335: 

1336: In this section, the thin-disk scenario is studied using

1337: non-spherically symmetric explosion models from Maeda et al. (2002).

1338: The chemical composition of the ejecta in a non-spherically symmetric

1339: SN explosion is strongly dependent on direction. In particular, 

1340: if we assume that the jet is collimated perpendicular to the orbital

1341: plane of the binary, where the secondary star is located, elements

1342: such as Ti, Fe, and Ni are mainly ejected in the jet direction, while

1343: O, Mg, Al, Si, and S are preferentially ejected near the equatorial

1344: plane of the helium star (Maeda et al.\ 2002). 

1345: 

1346: In Fig.~\ref{fig13} we compare the predicted abundances 

1347: in the secondary star after pollution from an aspherical explosion 

1348: model of a metal-rich progenitor having a 16~{\Msun} He core (see also

1349: Table~\ref{tbl6}). The initial abundances of the secondary were

1350: extracted from the average values of stars of the solar neighborhood

1351: with similar iron content (see Table~\ref{tbl3}). The left panel

1352: reflects the composition of the material ejected in the equatorial

1353: plane while in the right panel we have considered complete lateral

1354: mixing (Podsiadlowski et al.\ 2002) --- that is, the ejected matter is

1355: completely mixed within each velocity bin. This mixing process is

1356: done with the decayed yields of the model, and after that, all the

1357: material above the mass-cut is mixed and used to calculate the  

1358: composition of the ejected matter. The observed abundances are better

1359: reproduced if complete lateral mixing is considered since this process

1360: tends to enhance all of the element abundances at all mass-cuts. 

1361: However, the equatorial model (left panel) also provides good fits to the

1362: observed abundances. It should be noted that in this model, only the

1363: material ejected in the equatorial plane is captured and therefore,

1364: only Mg, Al, and Si are significantly enhanced with respect to the

1365: initial abundances. A model with solar initial abundances for the

1366: secondary star, i.e. ${\rm [X/H]}_{0}=0$, and the same explosion model of

1367: solar metallicity ($Z=0.02$) was also inspected, providing the

1368: same result except for the equatorial model which produces too low

1369: abundances of Ti, Fe, and Ni in comparison with the observations.

1370: 

1371: Both aspherical explosion models with initial abundances equal to the

1372: average values of thin-disk stars show little dependence on the

1373: mass-cut, and since the equatorial model can reproduce the observed

1374: abundances, extensive mixing processes are not required. In addition,

1375: the non-spherically symmetric ejection of the mass in the explosion

1376: could provide the kick that the system needs to change its orbit from

1377: the Galactic plane to the current orbit. Therefore, the element

1378: abundances in the secondary star and the kinematics of this system

1379: strongly suggest that the binary system \mbox{XTE J1118+480} formed in

1380: the Galactic disk (probably the thin disk) and it was then kicked

1381: towards the Galactic halo, most probably by asymmetric mass ejection

1382: in an asymmetric supernova/hypernova explosion that gave rise to the

1383: black hole in this system.

1384: 

1385: The elements O and C, for which we provide the expected abundances

1386: in Tables~\ref{tbl5} and~\ref{tbl6}, could be studied in future

1387: investigations, probably from CO and OH bands in the near-IR. This

1388: would help to recognize the operation of the CNO cycle on the surface

1389: the secondary star, proposed by Haswell et al. (2002), and

1390: possible processes of rotation-induced mixing during the evolution of

1391: the massive star. 

1392: 

1393: \section{Conclusions}

1394: 

1395: We have presented Keck~II/ESI medium-resolution spectroscopy of the

1396: black hole binary \mbox{XTE J1118+480}. The individual spectra of the

1397: system allowed us to derive an orbital period of $P = 0.16995 \pm

1398: 0.00012$~d and a radial velocity semiamplitude of the secondary star

1399: of $K_2 = 708.8 \pm 1.4$ \kmso. The implied updated mass function is

1400: $f(M) = 6.27 \pm 0.04$~\Msun, consistent with (but more precise than)

1401: previous values reported in the literature. Inspection of the

1402: high-quality averaged spectrum of the secondary star provides a

1403: rotational velocity of $v~\sin~i = 100^{+3}_{-11}$ \kmso, and hence a

1404: binary mass ratio $q = 0.027 \pm 0.009$. The derived radial velocity,

1405: $\gamma = 2.7 \pm 1.1$ \kmso, of the center of mass of the system

1406: agrees, at the 3$\sigma$ level, with the results of previous studies.

1407: 

1408: We have performed a detailed chemical analysis of the secondary star.

1409: We applied a technique that provides a determination of the stellar

1410: parameters, taking into account any possible veiling from the

1411: accretion disk. We find $T_{\mathrm{eff}} = 4700 \pm 100$~K, $\log 

1412: [g/{\rm cm~s}^2] = 4.6 \pm 0.3$, $\mathrm{[Fe/H]} = 0.18 \pm 0.17$, 

1413: and a disk veiling (defined as $F_{\rm  disk}/F_{\rm total}$) of

1414: $\sim$40\% at 5000~{\AA}, decreasing toward longer wavelengths.  

1415: 

1416: We have provided further details on the abundances of Mg, Al,

1417: Ca, Fe, Ni, and Li already reported by Gonz\'alez Hern\'andez et al.

1418: (2006), and we determined new element abundances of Si and Ti. The chemical

1419: abundances are typically higher than solar, and in some cases they 

1420: are slightly enhanced (e.g., Mg, Al, and Si) in comparison with the

1421: abundances of these elements in stars of the solar neighborhood 

1422: having similar iron content.

1423: 

1424: The present location and kinematics of this binary system had suggested

1425: that it could have originated in the Galactic halo. However, the

1426: chemical abundances strongly indicate that the black hole formed as a

1427: consequence of a supernova/hypernova explosion that occurred within

1428: the binary system. This explosive event must have either provided a

1429: kick to the system if it was formed in the thin disk or enriched

1430: significantly the atmosphere of the secondary star if the system

1431: formed in the thick disk or halo. 

1432: 

1433: We have explored a variety of supernova/hypernova explosion models for

1434: different metallicities, He core masses, and geometries. We compared

1435: the expected abundances in the secondary star after contamination from

1436: nucleosynthetic products from different initial metallicities of the

1437: secondary star ($-2.2 < \mathrm{[Fe/H]} < 0.2$), to investigate the formation 

1438: region in the Galactic halo, thick disk, or thin disk. Metal-poor

1439: explosion models ($Z = 0$ and $Z = 0.001$) were not able to fit the 

1440: observed abundances since they produce inappropriate ratios between

1441: $\alpha$-elements and iron-peak elements, and they are extremely

1442: sensitive to the adopted mass-cut. This comparison probably rules out 

1443: an origin in the Galactic halo for this black hole binary. 

1444: 

1445: For the thick-disk scenario, we carefully inspected the model

1446: predictions, and although they provide better fits to the observed

1447: abundances, they require substantial fallback (up to 5.5~\Msun) and

1448: very efficient mixing processes between the inner layer of the

1449: explosion and the ejecta. We thus conclude that this scenario is

1450: unlikely. 

1451: 

1452: Metal-rich spherically symmetric models for the thin-disk scenario

1453: were able to fairly reproduce the observed abundances, although they do

1454: not easily provide the energy required to launch the system from

1455: the Galactic plane to its current halo orbit. 

1456: 

1457: Finally, non-spherically symmetric models produce excellent agreement

1458: with the observed element abundances in the secondary star without

1459: invoking extensive fallback and mixing. In addition, asymmetric

1460: mass ejection would naturally provide the kick to expel this binary

1461: system from its birth place in the Galactic thin disk, which seems to

1462: be the most plausible explanation for the origin of this halo black hole

1463: X-ray binary.

1464: 

1465: \acknowledgments

1466: 

1467: We thank Keiichi Maeda for providing us with his aspherical explosion

1468: models, and for helpful discussions. We are grateful to Tom Marsh for

1469: the use of the MOLLY analysis package. The W. M. Keck Observatory is

1470: operated as a scientific partnership among the California Institute of

1471: Technology, the University of California, and NASA; it was made

1472: possible by the generous financial support of the W. M. Keck Foundation. 

1473: This work has made use of the VALD database and IRAF facilities. J.I.

1474: acknowledges support from the EU contract MEXT-CT-2004-014265

1475: (CIFIST). Additional funding was provided by Spanish Ministry project 

1476: AYA2005--05149, as well as by US National Science Foundation grants

1477: AST--0307894 and AST--0607485 to A.V.F.

1478: 

1479: \begin{thebibliography}{}

1480: 

1481: \bibitem[Allende Prieto et al., 2004]{all04}

1482: Allende Prieto, C., Barklem, P. S., Lambert, D. L., \& Cunha,

1483: K. 2004, \aap, 420, 183

1484: 

1485: \bibitem[Allende Prieto et al., 2006]{all06}

1486: Allende Prieto, C., et al. 2006, \apj, 636, 804 

1487: %Allende Prieto, C., Beers, T. C., Wilhelm, R. {\it et al.} 2006, \apj,

1488: %636, 804 

1489: 

1490: \bibitem[Al-Naimiy 1978]{aln78}

1491: Al-Naimiy, H. M. 1978, \apss, 420, 183

1492: 

1493: \bibitem[Beer \& Podsiadlowski 2002]{bap02}

1494: Beer, M. E., \& Podsiadlowski, Ph. 2002, \mnras, 331, 351

1495: 

1496: \bibitem[Bensby  et al. 2003]{ben03}

1497: Bensby, T., Feltzing, S., Lundstr\"om, I., \& Ilyin, I. 2003, \aap,

1498: 410, 527

1499: 

1500: \bibitem[Bensby et al., 2005]{ben05}

1501: Bensby, T., Feltzing, S., Lundstr\"om, I., \& Ilyin, I. 2005, \aap,

1502: 433, 185

1503: 

1504: \bibitem[Benz \& Hills 1992]{bah92}

1505: Benz, W., \& Hills, J. G. 1992, \apj, 433, 185

1506: 

1507: \bibitem[Brunish \& Truran 1982]{bat82}

1508: Brunish, W. M.,\& Truran, J. W. 1982, \apjs, 49, 447 

1509: 

1510: \bibitem[Cayrel et al. (2004)]{cay04} 

1511: Cayrel, R., et al. 2004, \aap, 416, 1117 

1512: %Cayrel, R., Depagne, E., Spite, M. {\it et al.} 2004, \aap, 416, 1117 

1513: 

1514: \bibitem[Chen 1997]{che97} 

1515: Chen, B. 1997, \apj, 491, 181 

1516: 

1517: \bibitem[D'Antona \& Mazzitelli 1994]{dam94}

1518: D'Antona, F., \& Mazzitelli, I. 1994, \apjs, 90, 467 

1519: 

1520: \bibitem[Ecuvillon et al. 2004]{ecu04}

1521: Ecuvillon, A., Israelian, G., Santos, N. C., Mayor, M., Villar, V., \&

1522: Bihain, G. 2004, \aap, 426, 619 

1523: 

1524: \bibitem[Ecuvillon et al. 2006]{ecu06}

1525: Ecuvillon, A., Israelian, G., Santos, N. C., Shchukina, N. G., Mayor,

1526: M., \& Rebolo, R. 2006, \aap, 445, 633

1527: 

1528: \bibitem[Frontera et al., 2001]{fro01}

1529: Frontera, F., et al. 2001, \apj, 561, 1006

1530: %Frontera, F., Zdziarski, A. A., Amati, L. {\it et al.} 2001, \apj, 561,

1531: %1006

1532: 

1533: \bibitem[Gelino et al., 2001]{gel01}

1534: Gelino, D. M., Harrison, T. E., \& Orosz, J. A. 2001, \apj, 122, 2668

1535: 

1536: \bibitem[Gelino et al. 2006]{gel06}

1537: Gelino, D. M., Balman, \c{S}., Kililo\u{g}lu, \"U., Yilmaz, A.,

1538: Kalemci, E., \& Tomsick, J. A. 2006, \apj, 642, 438

1539: 

1540: \bibitem[Gilli et al., 2006]{gua06}

1541: Gilli, G., Israelian, G., Ecuvillon, A., Santos, N. C., \&

1542: Mayor, M. 2006, \aap, 449, 723

1543: 

1544: \bibitem[Gonz\'alez Hern\'andez et al. 2004]{gon04}

1545: Gonz\'alez Hern\'andez, J. I., Rebolo, R., Israelian, G., Casares, J.,

1546: Maeder, A., \& Meynet, G. 2004, \apj, 609, 988

1547: 

1548: \bibitem[Gonz\'alez Hern\'andez et al. 2005]{gon05}

1549: Gonz\'alez Hern\'andez, J. I., Rebolo, R., Israelian, G., Casares, J.,

1550: Maeda, K., Bonifacio, P., \& Molaro, P. 2005, \apj, 630, 495

1551: 

1552: \bibitem[Gonz\'alez Hern\'andez et al. 2006]{gon06}

1553: Gonz\'alez Hern\'andez, J. I., Rebolo, R., Israelian, G., Harlaftis, E.

1554: T., Filippenko, A. V., \& Chornock, R. 2006, \apj, 644, L49

1555: 

1556: \bibitem[Grevesse et al. 1996]{gre96}

1557: Grevesse, N., Noels, A., \& Sauval, A. J. 1996, in Cosmic Abundances, 

1558: ed. S. S. Holt \& G. Sonneborn (San Francisco: ASP, Conf. Ser. Vol. 

1559: 99), 117

1560: 

1561: \bibitem[Gualandris et al., 2005]{gua05}

1562: Gualandris, A., Colpi, M., Portegies Zwart, S., \& Possenti,

1563: A. 2005, \aap, 618, 845

1564: 

1565: \bibitem[Haswell et al., 2002]{has02}

1566: Haswell, C. A., Hynes, R. I., King, A. R., \& Schenker, K. 2002,

1567: \mnras, 332, 928

1568: 

1569: \bibitem[Hills 1983]{hil83}

1570: Hills, J. G. 1983, \apj, 267, 322

1571: 

1572: \bibitem[Hills 1991]{hil91}

1573: Hills, J. G. 1991, \apj, 102, 2

1574: 

1575: \bibitem[Hynes et al. 2005]{hyn05}

1576: Hynes, R. I., Robinson, E. L., \& Bitner, M. 2005, \apj, 630, 405

1577: 

1578: \bibitem[Israelian et al., 1999]{isr99}

1579: Israelian, G., Rebolo, R., Basri, G., Casares, J., \&

1580: Mart{\'\i}n, E. L. 1999, \nat, 401, 142

1581: 

1582: \bibitem[Jonsell et al. 2005]{jon05}

1583: Jonsell, K., et al. 2005, \aap, 440, 321 

1584: %Jonsell, K., Edvardsson, B., Gustafsson, B. {\it et al.} 2005, \aap, 440,

1585: %321 

1586: 

1587: \bibitem[Kifonidis et al., 2000]{kif00}

1588: Kifonidis, K., Plewa, T., Janka, H.-Th., \& M\"{u}ller, E. 2000, 

1589: \aap, 531, L123 

1590: 

1591: \bibitem[Kurucz et al., 1984]{kur84}

1592: Kurucz, R. L. ATLAS9 Stellar Atmospheres Programs and 2 \kms

1593: Grid. (CD-ROM, Smithsonian Astrophysical Observatory, Cambridge,

1594: 1993).

1595: 

1596: \bibitem[Kurucz et al., 1984]{kur84}

1597: Kurucz, R. L., Furenild, I., Brault, J., \& Testerman, L. 1984, Solar 

1598: Flux Atlas from 296 to 1300 nm, NOAO Atlas 1 (Cambridge: Harvard Univ. 

1599: Press)

1600: 

1601: \bibitem[Lai et al 2001]{lai01}

1602: Lai, D., Chernoff, D. F., \& Cordes, J. M. 2001, \apj, 549, 1111

1603: 

1604: \bibitem[MacFadyen et al., 2001]{mwh01}

1605: MacFadyen, A. I., Woosley, S. E., \& Heger, A. 2001, \apj, 550, 410

1606: 

1607: \bibitem[Maeda et al. 2002]{mae02}

1608: Maeda, K., Nakamura, T., Nomoto, K., Mazzali, P. A., Patat, F., \&

1609: Hachisu, I. 2002, \apj, 565, 405

1610: 

1611: \bibitem[Marsh et al. 1994]{mah94}

1612: Marsh, T. R., Robinson, E. L., \& Wood, J. H. 1994, \mnras, 266, 137

1613: 

1614: \bibitem[Martin  et al. 2004]{mar04}

1615: Martin, N. F., Ibata, R. A., Bellazzini, M., Irwin, M. J., Lewis, G. F.,

1616: \& Dehnen, W. 2004, \mnras, 348, 12

1617:  

1618: \bibitem[McClintock et al., 2001]{mck01}

1619: McClintock, J. E., Garcia, M. R., Caldwell, N., Falco, E. E.,

1620: Garnavich, P. M., \& Zhao, P. 2001, \apj, 551, L147

1621: 

1622: \bibitem[Mihalas \& Binney 1981]{mab81}

1623: Mihalas, D., \& Binney, J. 1981, Galactic Astronomy: Structure

1624: and Kinematics, 2nd ed. (New York: W. H. Freeman \& Company)

1625: 

1626: \bibitem[Mirabel et al., 2001]{mir01}

1627: Mirabel, I. F., Dawan, V., Mignani, R. P., Rodrigues, I., \&

1628: Guglielmetti, F. 2001, \nat, 413, 139

1629: 

1630: \bibitem[Orosz et al. 2001]{osz01}

1631: Orosz, J. A., et al. 2001, \apj, 555, 489 

1632: %Orosz, J. A., Kuulkers, E., van der klis, M. {\it et al.} 2001, \apj, 555,

1633: %489 

1634: 

1635: \bibitem[Pavlenko \& Magazzu 1996]{pam96}

1636: Pavlenko, Ya. V., \& Magazz\`u, A. 1996, \aap, 311, 961

1637: 

1638: \bibitem[Piskunov et al. 1995]{pis95}

1639: Piskunov, N. E., Kupka, F., Ryabchikova, T. A., Weiss, W. W., \& Jeffery, C.

1640: S. 1995, \aaps, 112, 525

1641: 

1642: \bibitem[Podsiadlowski et al., 2002]{pod02}

1643: Podsiadlowski, P., Nomoto, K., Maeda, K., Nakamura, T., Mazzali, P., \&

1644: Schmidt, B. 2002, \apj, 567, 491 

1645: 

1646: \bibitem[Portegies Zwart et al., 1997a]{poz97}

1647: Portegies Zwart, S. M., Verbunt, F., \& Ergma, E. 1997a, \aap, 321, 207

1648: 

1649: \bibitem[Portegies Zwart et al., 1997b]{poz97}

1650: Portegies Zwart, S. M. et al. 1997b, \aap, 328, 143

1651: 

1652: \bibitem[Remillard et al. 2000]{rem00}

1653: Remillard, R., Morgan, E., Smith, D., \& Smith, E. 2000, IAU Circ. No. 7389

1654: 

1655: \bibitem[Reylé \& Robin 2001]{rar01}

1656: Reyl\'e, C., \& Robin, A. C. 2001, \aap, 373, 886

1657: 

1658: \bibitem[Sadakane et al. 2006]{sad06}

1659: Sadakane, K., et al. 2006, \pasj, 58, 595 

1660: %Sadakane, K., Arai, A., Aoki, W., Arimoto, N. {\it et al.} 2006, \pasj, 58,

1661: %595 

1662: 

1663: \bibitem[Sheinis 2002]{she02}

1664: Sheinis, A. I., et al. 2002, \aap, 114, 851

1665: %Sheinis, A. I., Bolte, M., Epps, H. W. {\it et al.} 2002, \aap, 114, 851

1666: 

1667: \bibitem[Sneden 1973]{sne73}

1668: Sneden, C. 1973,  PhD Dissertation, Univ. of Texas at Austin

1669: 

1670: \bibitem[Tominaga et al. (2007)]{tom07} 

1671: Tominaga, N., Umeda, H. \& Nomoto, K. 2007, \apj, 660, 516 

1672: 

1673: \bibitem[Torres et al. 2004]{gon04}

1674: Torres, M. A. P., et al. 2004, \apj, 612, 1026

1675: %Torres, M. A. P., Callanan, P. J., Garcia, M. R. {\it et

1676: %al.} 2004, \apj, 612, 1026

1677: 

1678: \bibitem[Umeda \& Nomoto 2002]{uan02}

1679: Umeda, H.,\& Nomoto, K. 2002, \apj, 565, 385

1680: 

1681: \bibitem[Umeda \& Nomoto 2005]{uan05}

1682: Umeda, H.,\& Nomoto, K. 2005, \apj, 619, 427

1683: 

1684: \bibitem[van den Heuvel \& Habets 1984]{hah84}

1685: van den Heuvel, E. P. J., \& Habets, G. M. H. J. 1984, \nat, 309, 598 

1686: 

1687: \bibitem[Wagner 2001]{wag01}

1688: Wagner, R. M., et al. 2001, \apj, 556,42 

1689: %Wagner, R. M., Foltz, C. B., Shahbaz, T. {\it et al.} 2001, \apj, 556,42 

1690: 

1691: \end{thebibliography}

1692: 

1693: \end{document}

1694: 

1695: 

1696: 

1697: