1:
2: %\documentclass[apjl]{emulateapj}
3:
4: \documentclass[12pt,preprint]{aastex}
5:
6: %% manuscript produces a one-column, double-spaced document:
7:
8: %% \documentclass[manuscript]{aastex}
9:
10: %% preprint2 produces a double-column, single-spaced document:
11:
12: % \documentclass[preprint2]{aastex}
13:
14: %% \documentclass[preprint2,longabstract]{aastex}
15:
16: \newcommand{\vdag}{(v)^\dagger}
17: \newcommand{\myemail}{tsh@ll.iac.es}
18: \newcommand{\target}{2S\,0921--630/V395~Car}
19: \newcommand{\kms}{$\rm km\,s^{-1}$}
20: \newcommand{\Msun}{$\rm M_{\odot}$}
21: \newcommand{\Rsun}{$\rm R_{\odot}$}
22: \newcommand{\Lsun}{$\rm L_{\odot}$}
23:
24: %% You can insert a short comment on the title page using the command below.
25:
26: \slugcomment{}
27:
28: \shorttitle{The massive neutron star or low-mass black hole in \target.}
29: \shortauthors{Shahbaz et al.}
30:
31: \begin{document}
32:
33: \title{The massive neutron star or low-mass black hole in \target.}
34:
35: \author{T. Shahbaz\altaffilmark{1},
36: J. Casares\altaffilmark{1},
37: C. Watson\altaffilmark{2},
38: P. A. Charles\altaffilmark{3},
39: R. I. Hynes\altaffilmark{4},
40: S. C. Shih\altaffilmark{3},
41: D. Steeghs\altaffilmark{5}
42: }
43:
44:
45: \altaffiltext{1}{Instituto de Astrof\'\i{}sica de Canarias, 38200 La Laguna,
46: Tenerife, Spain. tsh, jcv@iac.es}
47: \altaffiltext{2}{Department of Physics and Astronomy, University of Sheffield,
48: Sheffield, S3 7RH, England. c.watson@sheffeild.ac.uk}
49: \altaffiltext{3}{School of Physics and Astronomy,
50: The University of Southampton, Southampton, SO17 1BJ, UK.
51: pac, icshih@soton.ac.uk }
52: \altaffiltext{4}{McDonald Observatory and Department of Astronomy,
53: The University of Texas at Austin, 1 University Station C1400, Austin,
54: Texas 78712, USA. rih@astro.as.utexas.edu}
55: \altaffiltext{5}{Harvard-Smithsonian Center for Astrophysics,
56: 60 Garden Street, MS-67, Cambridge, MA 02138, USA. steeghs@cfa.harvard.edu }
57:
58: \begin{abstract}
59:
60: We report on optical spectroscopy of the eclipsing Halo LMXB \target,
61: that reveals the absorption line radial velocity curve of the K0{\sc
62: iii} secondary star with a semi-amplitude $K_{\rm 2}$=92.89$\pm$ 3.84~\kms~, a
63: systemic velocity $\gamma$=34.9$\pm$3.3~\kms and an orbital period
64: $P_{\rm orb}$ of 9.0035$\pm$0.0029 day (1-$\sigma$). Given the quality of
65: the data we find no evidence for the effects of X-ray irradiation.
66: Using the previously determined
67: rotational broadening of the mass donor, and applying conservative
68: limits on the orbital inclination, we constrain the
69: compact object mass to be 2.0--4.3~\Msun (1-$\sigma$), ruling out a
70: canonical neutron star at the 99\% level. Since the nature of the
71: compact object is unclear, this mass range implies that the compact
72: object is either a low-mass black hole with a mass slightly higher
73: than the maximum neutron star mass (2.9~\Msun) or a massive neutron
74: star. If the compact object in \target\ is a black hole, it confirms
75: the prediction of the existence of low-mass black holes, while if the
76: object is a massive neutron star its high mass severely constrains the
77: equation of state of nuclear matter.
78:
79: \end{abstract}
80:
81: \keywords{accretion: accretion discs -- binaries: close
82: stars: individual (\objectname{2S\,0921--630/V395\,Car})}
83:
84: \section{Introduction}
85:
86: A knowledge of the neutron star mass distribution provides a
87: fundamental test of theories of the equation of state of nuclear
88: matter, the applicability of General Relativity as the correct theory
89: of gravity and significant information on the evolutionary history of
90: the progenitor stars. To date, only studies of millisecond radio
91: pulsars have provided accurate mass determinations of neutron stars,
92: reflecting their mass at formation; 1.35$\pm$0.04~\Msun\
93: \citep{Ker01}. Although dynamical neutron star masses can also be
94: obtained from accreting X-ray pulsars, there are large
95: uncertainties. In high-mass X-ray binaries (HMXBs) there are
96: uncertainties due to non-Keplerian perturbations in the radial
97: velocity curves caused by effects such as stellar wind contamination
98: and X-ray heating. In low-mass X-ray binaries (LMXBs), the situation
99: is worse because the intense X-ray irradiation usually
100: suppresses the light from the donor \citep{Cha03}. It is only in a
101: few exceptional cases where the companion is evolved (and hence more
102: luminous) or during X-ray off-states, when dynamical information can
103: be extracted about the nature of the compact object.
104:
105: The SAS-3 X-ray source, 2S~0921-630, was identified with a
106: $\sim$16$^m$ blue star \citep{Li78}, V395~Car, whose optical spectrum
107: is dominated by HeII\,4686\AA\ and Balmer emission \citep{Bra83},
108: characteristics of LMXBs, with occasional superimposed late-type
109: stellar absorption features visible \citep{Tho79}. EXOSAT
110: observations showed a broad ($>$1~day), shallow X-ray eclipse, during
111: which the spectrum softened \citep{Mas87}, allowing an estimate of the
112: size of the accretion-disc corona (ADC). This was comparable to the
113: eclipsing companion star, thus accounting for the partial X-ray
114: eclipse, and with an inclination of $i$=70$^\circ$--90$^\circ$.
115: %The variability prior to the broad X-ray eclipse is reminiscent of
116: %2A\,1822-371 \citep{Whi85}, where obscuration of the ADC by structure
117: %on the disc rim is responsible for the shape of the lightcurve.
118: Only a handful of ADC sources are known in which the compact object
119: is permanently obscured from our line-of-sight by the accretion disc,
120: and requires the observed X-rays to be scattered in a hot corona
121: \citep{Whi95}. By implication, the intrinsic X-ray luminosity is
122: {\it much} higher.
123:
124: Limited optical photometry and emission line spectroscopy
125: \citep{Cow82,Bra83} shows that \target\ has a long orbital period
126: of 9.02\,day and deep ($\sim$1 mag) dips where
127: the $(B-V)$ colour reddens by up to 0.4 mags, which has been
128: interpreted as the eclipse of the disc by the late-type
129: secondary \citep{Che82}. High-resolution optical spectroscopy
130: revealed the donor star to be of spectral type K0{\sc iii} with a
131: rotational velocity of $v\sin\,i$=65$\pm$9\,\kms, contributing
132: $\sim$25\% to the observed flux at 6500\AA\ \citep{Sha99}. \target\
133: is one of those rare LMXBs in which the secondary is visible despite
134: the presence of a luminous disc.
135: %Kinematic properties imply that
136: %it is located in the Halo at a distance $\sim$10\,kpc \citep{Cow82}.
137:
138: There has been no detection of type {\sc i} X-ray bursts, so the
139: nature of the compact object is unclear. In this letter we present the
140: results of an intensive campaign to measure the dynamical mass of the
141: compact object in \target. We determine the secondary star's radial
142: velocity curve, investigate the possible effects of X-ray irradiation
143: and obtain firm constraints on the compact object mass.
144:
145:
146: \section{Observations and Data Reduction}
147:
148: Time resolved spectroscopic observations of \target\ were obtained on
149: the 1.9-m telescope at the South African Astronomical Observatory
150: (SAAO) during 2003 March 30 to April 18.
151: Grating \# 5 (1200 lines/mm) was used centered at 4900\AA\ and
152: covering a wavelength range of 4529--5377\AA with a dispersion of
153: 0.49\,\AA/pixel$^{-1}$.
154: A slit width of 2.0\,\arcsec gave a spectral resolution of
155: 1.1~\AA\ (=64~\kms\ at 5200\AA) as measured from the Cu-Arc arc
156: lines. In general the conditions were good, with the seeing varying
157: between 1\,\arcsec--3\,\arcsec. A total of 83 spectra of \target\
158: were obtained with an exposure time of 1800\,s. Each spectrum was
159: bracketed by an observation of the internal Cu-Ar arc lamp at the
160: position of the star. Template field stars of a variety of spectral
161: types were also observed.
162:
163: The images were flat-fielded using observations of a tungsten calibration
164: lamp, and a constant bias level was subtracted from all the
165: images. Extraction of the 1-D spectra from the images were performed
166: using optimal extraction
167: \citep{Hor86a} and calibration of the wavelength scale was achieved
168: using 4th order polynomial fits to the position measured arc-line
169: positions, which gave an rms scatter of $\sim$0.02~\AA.
170:
171: In addition, 34 spectra were also obtained at the Very
172: Large Telescope (VLT; Paranal, Chile), New Technology Telescope
173: (NTT; La Silla, Chile) the Anglo Australian Telescope (AAT; Siding Springs,
174: Australia) and at the Magellan telescope (Las Campanas, Chile)
175: as part of a long-term project on X-ray binaries.
176: We used the FORS2 Spectrograph attached to the 8.2m Yepun Telescope (VLT)
177: on the nights of 2003 May 18 and June 22-23.
178: A total of 9 spectra with exposure times in the range 300 to 900\,s were obtained.
179: The R1400V holographic grating was used in combination with a
180: 0.7\,\arcsec slit resulting in a wavelength coverage of 4514--5815\AA\ and a
181: resolution of 70\,\kms. Instrumental flexure was monitored by
182: cross-correlating the sky spectra and was found to be very small, always
183: within 6\,\kms.
184: The Magellan spectra were obtained on 2003 Dec 12--16 using
185: the Baade telescope at Las Campanas using the IMACS imaging spectrograph in
186: long-camera mode.
187: A total of 11 spectra with an exposure time of 300s were obtained.
188: The 600 grating together with a 0.7\,\arcsec long slit provided spectra
189: covering 3720--6830\AA\ with a dispersion of 0.75\,\AA/pixel$^{-1}$.
190: The AAT spectra were obtained using the RGO spectrograph on the nights of 2002 June 6 to 11.
191: A total of 10 spectra with an exposure time of 1800\,s were obtained.
192: The R1200B grating centered at 4350\AA\ together with a 1.0\,\arcsec slit provided spectra
193: covering 3500-5250\AA\ with a resolution of 70\,\kms.
194: The total of 4 NTT spectra were obtained with EMMI and grating \# 6 on the
195: nights of 2002 June 8 to 10 exposure times of 600 to 900\,s.
196: The spectra cover the spectral range 4400--5150\AA\ with a resolution of 75\,\kms.
197: The AAT and NTT data reduction details are given in \citet{Cas03}.
198:
199: \section{The radial velocity curve and spectrum}
200: \label{RadVel}
201:
202: To increase the signal-to-noise of the spectra prior to
203: cross-correlation, the individual spectra were variance
204: averaged into nighly means or groups of $\sim$3\,hrs, depending on
205: the quality of the individual spectra. A total of 31
206: absorption line radial velocities were measured by cross-correlation
207: \citep{Ton79} with a template star. Prior to cross-correlation, the
208: spectra were interpolated onto a constant velocity scale (32
209: \kms\,pixel$^{-1}$) and normalized by fitting a spline function to
210: the continuum. The region 5100--5300\AA\ common to all the spectra,
211: primarily containing
212: the Mg{\sc i} (5167.3\AA, 5172.7\AA\ and 5183.6\AA) absorption blend
213: was used in the analysis. The template star's radial velocity,
214: determined using the position of the Fe{\sc i} 4957.597\AA\ absorption
215: line, was then added to the radial velocities of \target. The
216: resulting radial velocity data (see Figure\,\ref{RVCURVE}) were then
217: fitted with a circular orbit (sinusoidal function) using the method
218: of least-squares. Using spectral type template stars in the range
219: G6--K4{\sc iii} we obtain $K_{\rm 2}$ values in the range 92--99~\kms
220: (see Table~\ref{RVFIT}). The secondary star's spectral type has
221: previously been determined from high-resolution spectroscopy to be
222: K0{\sc iii} \citep{Sha99}, and our work confirms this. Fitting the
223: radial velocities with a K0{\sc iii} template star, we find a minimum
224: reduced $\chi^2$ of 4.3 with the best fit parameters
225: $\gamma$ = 34.9$\pm$3.3 \kms\,
226: $K_{\rm 2}$ = 92.89$\pm$3.84 \kms\,
227: $P_{\rm orb}$=9.0035$\pm$0.0029 day and
228: $T_{\rm 0}$=HJD\,2453099.51$\pm$0.08,
229: where $P_{\rm orb}$ is the orbital period, $T_{\rm 0}$ is time at
230: phase 0.0 defined as inferior conjunction of the secondary star,
231: $\gamma$ is the systemic velocity and $K_{\rm 2}$ is the radial
232: velocity semi-amplitude (1-$\sigma$ errors are quoted with the error
233: bars rescaled so that the reduced $\chi^2$ of the fit is 1.)
234: Our ephemeris is consistent with that determined by \citealt{Mas87}.
235: We also
236: computed a $\chi^2$ periodogram of the data to investigate the
237: significance of other periods. No significant peaks (at $>$99.99 \%
238: level) were present other than the one at 9.0035\,day. Finally, to
239: check if the 9.0035\,day period is affected by an alias, we computed
240: the window function; no significant peaks were present at
241: 9.0035\,day. Fitting the radial velocity data with an eccentric
242: orbit did not yield a better fit; the eccentric fit is only
243: significant at the 38\% level.
244:
245: Using a distance of $\sim$10\,kpc \cite{Cow82} \target\ lies 1.6\,kpc
246: below the galactic plane and so belongs to the Halo population.
247: Note the systemic velocity is not consistent with the radial velocity due to
248: Galactic differential rotation, $\sim$60~\kms\ at a distance
249: of $\sim$10\,kpc \citep{Deh98}.
250: However, what is interesting is that the magnitude of the systemic velocity
251: of \target\ seems to be about a factor
252: of $\sim$4 lower compared to other neutron star Halo systems
253: (\citealt{Cas98,Cas02,Tor02}).
254:
255: In Figure\,\ref{SPECTRUM} we show the variance-weighted Doppler
256: averaged spectrum of V395~Car. The strongest features are the H, He
257: emission lines, which appear single peaked, even at the resolution of
258: these data. Weak absorption features of Mg {\sc i} 5175\AA\ and the
259: Ca/Fe blend at 5269\AA\ allow us to determine the fraction of light
260: from the secondary star, by optimally subtracting a scaled K0{\sc iii}
261: template spectrum from V395~Car \citep{Mar94}. As the rotational
262: broadening of the donor (65$\pm$9~\kms; \citealt{Sha99}) is similar to
263: our spectral resolution (64~\kms), we do not broaden the template
264: before subtraction. We find that the secondary star contributes $\sim
265: 10$\% to the observed flux at 5200\AA\ (see Table~\ref{RVFIT}).
266:
267:
268: \section{Irradiation of the secondary star - the mass of the compact object}
269: \label{Irrad}
270:
271: The observed X-rays are reflected into our line of sight by
272: scattering. Although the observed X-ray luminosity is low
273: $L_{\rm X}=2\times 10^{35}$\,erg\,s$^{-1}$ \citep{Kal03}, it is still
274: interesting to explore the possible effects that X-rays may have on
275: the secondary star's atmosphere and hence radial velocity.
276: We investigate these effects using the X-ray binary model described by
277: \citet{Phi99,Sha00,Sha03}. We first phase-folded the
278: radial velocities using the orbital ephemeris derived in
279: section\,\ref{RadVel} and binned into 31 orbital phase bins.
280: We then fit the resulting radial velocity curve with the X-ray binary model.
281: We considered two extreme cases for the X-ray irradiation: case (a)
282: where the X-rays produced near the compact object irradiate the
283: whole inner face of the secondary star and case (b) where the X-ray's
284: do not affect the secondary star. Note that invoking a flared
285: accretion disc will have intermediate solutions. The
286: variable model parameters are the compact object mass $M_{\rm 1}$,
287: secondary star mass $M_{\rm 2}$, X-ray luminosity $L_{\rm X}$ and the
288: inclination $i$. To determine which elements on the
289: star contribute to the absorption line radial velocity we use the
290: factor $f_{\rm X}$, which is the fraction of the external radiation
291: flux that exceeds the unperturbed flux.
292: For a given $i$ and
293: combination of $M_{\rm 1}$ and $M_{\rm 2}$ we fit the data and use
294: the constraints imposed by the measured value for $v\sin\,i$
295: (64$\pm$9\,\kms; \citealt{Sha99}) to
296: produce plots in the $M_{\rm 1}$--$M_{\rm 2}$
297: plane (Figure\,\ref{MASSES})
298:
299: Given the evidence for optical and X-ray partial eclipse, $i$ must lie
300: in the range 70$^\circ$--90$^\circ$. Therefore we consider cases (a)
301: and (b) for $i$=70$^\circ$ and more importantly $i$=90$^\circ$,
302: because it gives a firm lower limit to the compact object's mass. For
303: case (a) we fix $L_{\rm X}=2\times 10^{35}$\,erg\,s$^{-1}$ and find
304: that the best fit ($\chi^2$ = 122) is obtained when all the elements
305: on the inner face contribute to the radial velocity, i.e. $f_{\rm X}>$
306: 60\%, i.e. X-ray heating is not preferred. Note, however, that the
307: radial velocity curve will not be sinusoidal because there will still
308: be some irradiated elements that contribute to the radial velocity.
309: Even if we increase to $L_{\rm X}=10^{39}$\,erg\,s$^{-1}$ as could be
310: the case for an LMXB, the fits are significantly worse ($\chi^2$ = 190)
311: and the best fit still occurs when all the inner face contributes to the
312: radial velocity. For case (b) we set $L_{\rm X}=0$, implying no X-ray
313: heating and find that it gives a fit with
314: similar quality ($\chi^2$ = 119). We conclude that the best fit to
315: the radial velocity curve is the case where the X-rays do not have an
316: observable effect given the quality of the data.
317: For our range in $i$ our fits for case (a) and (b)
318: give 2.0 $< M_{\rm 1} <$ 4.3~\Msun\ (1-$\sigma$), which means we can
319: rule out a canonical 1.4\,\Msun\ neutron star at the 99\% level. Our
320: solutions for the compact object mass suggests
321: either a massive neutron star or a low-mass black hole, since the
322: maximum mass allowed for a neutron star is 2.9~\Msun\ \citep{Kal96}.
323:
324: Since the donor star fills its Roche lobe and is synchronized
325: with the binary motion, we use the secondary star's $v\sin\,i$
326: and $K_{\rm 2}$ to determine the binary mass ratio
327: $q$=($M_{\rm 2}/M_{\rm 1}$)=0.89$\pm$0.18 \citep{Hor86b}. Given the
328: measured masses and orbital period, we use Kepler's Third Law to
329: determine the semi-major axis 56.1$<a<$72.4\,\Rsun.
330: Eggleton's (1983) expression for the effective radius of the Roche-lobe
331: then determines the radius of the secondary 10.4$<R_{\rm 2}<$13.4\,\Rsun,
332: and the temperature inferred from the spectral type \citep{Gra92}
333: and Stefan-Boltzmann's law determines the luminosity
334: 52.5$<L_{\rm 2}<$87.1\,\Lsun (1-$\sigma$ ranges are quoted).
335:
336:
337:
338: \section{Discussion}
339:
340: \subsection{The evolutionary status of the system}
341: \label{Evol}
342:
343: The position of the secondary star of \target\ in an HR diagram corresponds to a
344: normal star that has crossed the Hertzsprung gap and now lies on the
345: Hayashi line. The evolution of the binary is dominated by the
346: evolution of the evolved secondary star. Since such a star no longer
347: burns hydrogen in its core, the mass transfer is early massive Case B.
348: For a binary with a secondary star near the onset of such mass
349: transfer, one requires $q<$1 throughout its history \citep{Kin99}.
350: The case for a 1.4~\Msun\ neutron star can be ruled out since it
351: requires $M_{\rm 2}<$1.4~\Msun, which is inconsistent with our
352: results; the position of a secondary on the HR diagram undergoing
353: early massive Case B mass transfer is close to that of a single star
354: of the same mass \citep{Kol98}. For a 3~\Msun\ compact object, $M_{\rm
355: 2}<$3.0~\Msun, which is consistent with our observations.
356:
357: It is interesting to note that the mass ratio we derive lies on the limit
358: of stable/unstabe mass transfer $q$=5/6 \citep{Kin99}. It is possible that \target\
359: is undergoing unstable mass transfer similar to the supersoft sources and other
360: X-ray binaries \citep{Kah97}.
361:
362: \subsection{The formation of the compact object}
363:
364: The distribution of measured neutron stars masses provides fundamental
365: constraints on the equation of state of nuclear matter and the
366: observational identification of a black hole which is based on the maximum
367: gravitational mass allowed for a neutron star; 2.9~\Msun\ \citep{Kal96}.
368: Studies of radio pulsars show that neutron star masses are clustered in a
369: narrow range. However, there is
370: some evidence that neutron stars with a mass in excess of 1.4~\Msun\ do
371: exist. LMXBs are expected to contain massive ($\sim$2~\Msun) neutron stars,
372: resulting from the accretion of a considerable amount of material over
373: extended (10$^8$~yrs) periods of time \citep{Zha97}.
374: However, while a neutron star mass of 1.78$\pm$0.23~\Msun\ for the LMXB
375: Cyg\,X-2 is reported \citep{Oro99}, \citet{Tit02} have recently proposed a
376: mass of 1.44$\pm$0.06~\Msun\ based on the spectral and temporal properties of
377: the type {\sc I} X-ray bursts. The analysis of optical data for the X-ray
378: pulsar Vela\,X-1 suggests that the pulsar has a mass of
379: 1.87$^{+0.23}_{-0.17}$~\Msun\ \citep{Bar02}. Probably the strongest case for a
380: massive neutron star or low-mass black hole is in the eclipsing HMXB
381: 4U\,1700--37 which contains a 2.44$\pm$0.27\,\Msun\ compact object
382: \citep{Cla02}. Such masses already test soft
383: nuclear equations of state \citep{Mil98a}.
384: We obtain a mass of 2.0--4.3~\Msun\ for the compact object in \target, which
385: lies between the range of masses observed for neutron stars and black holes.
386: Models for nuclear equations of state which include the effects of three
387: nucleon interactions and realistic models of nuclear forces limits
388: the maximum mass of neutron stars to be below 2.5~\Msun \citep{Akm98}.
389: However, it should be noted that unconventional forms of matter such as
390: Q-stars do allow the existence of extremely massive neutron stars \citep{Mil98b}
391: Although our results do not allow us to distinguish between a massive neutron
392: star or a low mass black hole, the existence of a $>$2.5~\Msun neutron star
393: would place strong constraints on high density nuclear matter.
394:
395: Theoretical predictions by \cite{Fry01} suggest that the death of a
396: 60~\Msun\ star in a close binary system can produce anything from a
397: low-mass 1.2~\Msun\ neutron star to a 10~\Msun\ black hole, depending
398: on the wind mass-loss rate during the Wolf-Rayet phase. As suggested
399: by \citet{Bro96} the formation of a low-mass black hole with a high
400: space velocity is possible via a two stage process involving
401: the formation of a neutron star.
402: However, if there is symmetric mass ejection in the supernova, then a kick at
403: the formation of a black hole is not required to explain the varied
404: range in the observed space velocities of the black hole LMXBs
405: \citep{Nel99}.
406:
407: The unusual mass for the compact object in \target\ and the fact that the
408: nature of the compact object is not clear, conjours up many formation
409: scenarios. If it is a
410: neutron star, then its present mass can be explained by either the
411: accumulation of matter (1--2~\Msun) by a canonical 1.4~\Msun\ neutron star,
412: accreted at a high mass accretion rate over a long period of time, or the
413: direct collapse of a massive star forming a massive neutron star. If the
414: compact object in \target\ is a black hole then its low mass can be explained
415: by the accretion induced collapse of a massive 2.9~\Msun\ neutron star or the
416: formation of a low-mass black hole \citep{Bro96}. Given the fact that we do
417: not observe a peculiar systemic velocity comparable to other
418: neutron-star LMXBs, this suggests that we can rule out canonical neutron star
419: formation scenarios where a kick is produced during a type {\sc II}
420: supernova. The most likely scenario for the formation of the compact object is
421: the direct formation of a massive neutron star or a low-mass black hole.
422:
423: \acknowledgments
424:
425: TS and JC acknowledge support from the Spanish Ministry of Science
426: and Technology under the programme Ram\'{o}n y Cajal.
427: RIH is currently supported by NASA through Hubble Fellowship grant
428: \#HF-01150.01-A\ awarded by the STSci, which
429: is operated by AURA, for NASA, under contract NAS 5-26555.
430: DS acknowledges a Smithsonian Astrophysical Observatory Clay Fellowship.
431: The data reduction and analysis was performed using the {\sc pamela}
432: and {\sc molly} routines of K.\,Horne and T.\,R.\ Marsh.
433: This paper uses observations made at the South African Astronomical Observatory
434: (SAAO).
435:
436: %Facilities: \facility{SAAO}, \facility{AAT}, \facility{NTT}, \facility{VLT},
437: %\facility{Magellan}
438:
439: \begin{thebibliography}{}
440:
441: \bibitem[Akmal et al.(1998)]{Akm98}
442: Akmal, A., Pandharipande, V.~R., \& Ravenhall, D.~G.\ 1998,
443: \prc, 58, 1804
444:
445: %\bibitem[Bailey(1981)]{Bai81}
446: % Bailey, J.\ 1981, \mnras, 197, 31
447:
448: \bibitem[Barziv et al.(2002)]{Bar02}
449: Barziv O., Kaper L., van Kerkwijk M.~H., Telting J.~H.,
450: van Paradijs J., 2001, \aap, 377, 925
451:
452: \bibitem[Branduardi-Raymont et al.(1983)]{Bra83}
453: Branduardi-Raymont, G., Corbet, R.~H.~D., Mason, K.~O., Parmar, A.~N.,
454: Murdin, P.~G., \& White, N.~E.\ 1983, \mnras, 205, 403
455:
456: \bibitem[Brown et al.(1996)]{Bro96}
457: Brown G.~E., Weingartner J.~C., Wijers Ralph A.~M.~J.,
458: 1996, ApJ, 463, 297.
459:
460: \bibitem[Casares et al.(1998)]{Cas98}
461: Casares, J., Charles, P.~A., \& Kuulkers, E.\ 1998, \apjl, 493, L39
462:
463: \bibitem[Casares et al.(2002)]{Cas02}
464: Casares, J., Dubus, G., Shahbaz, T., Zurita, C., \&
465: Charles, P.~A.\ 2002, \mnras, 329, 29
466:
467: \bibitem[Casares et al.(2003)]{Cas03}
468: Casares, J., Steeghs, D., Hynes, R.~I., Charles, P.~A., \&
469: O'Brien, K.\ 2003, \apj, 590, 1041
470:
471: \bibitem[Charles \& Coe(2003)]{Cha03}
472: Charles P.\,A., Coe M., in Compact Stellar X--ray Sources,
473: ed. W.~H.~G. Lewin \& M. van der Klis, Cambridge University Press,
474: astro-ph/0308020
475:
476: \bibitem[Chevalier \& Ilovaisky(1982)]{Che82}
477: Chevalier, C.~\& Ilovaisky, S.~A.\ 1982, \aap, 112, 68
478:
479: \bibitem[Chevalier \& Ilovaisky(1981)]{Che81}
480: Chevalier, C.~\& Ilovaisky, S.~A.\ 1981, \aap, 94, L3
481:
482: \bibitem[Clark et al.(2002)]{Cla02}
483: Clark, J.~S., Goodwin, S.~P., Crowther, P.~A., Kaper, L.,
484: Fairbairn, M., Langer, N., \& Brocksopp, C.\ 2002, \aap, 392, 909
485:
486: \bibitem[Cowley et al.(1982)]{Cow82}
487: Cowley, A.~P., Crampton, D., \& Hutchings, J.~B.\ 1982, \apj, 256, 605
488:
489: \bibitem[Dehnen \& Binney(1998)]{Deh98}
490: Dehnen, W.~\& Binney, J.\ 1998, \mnras, 294, 429
491:
492: \bibitem[Fryer \& Kalogera(2001)]{Fry01}
493: Fryer, C.~L.~\& Kalogera, V.\ 2001, \apj, 554, 548
494:
495: \bibitem[Gray(1992)]{Gra92}
496: Gray, D.~F.\ 1992, Cambridge Astrophysics Series, Cambridge:
497: Cambridge University Press, 1992, 2nd ed.
498:
499: \bibitem[Horne(1986)]{Hor86a}
500: Horne, K.\ 1986, \pasp, 98, 609
501:
502: \bibitem[Horne et al.(1986)]{Hor86b}
503: Horne, K., Wade, R.~A., \& Szkody, P.\ 1986, \mnras, 219, 791
504:
505: %\bibitem[Hynes et al.(2004)]{Hyn04}
506: % Hynes, R.~I., Steeghs, D., Casares, J., Charles, P.~A., \&
507: % O'Brien, K.\ 2004, \apj, 609, 317
508:
509:
510: \bibitem[Kahabka \& van den Heuvel(1997)]{Kah97}
511: Kahabka, P.~\& van den Heuvel, E.~P.~J.\ 1997, \araa, 35, 69
512:
513: \bibitem[Kallman et al.(2003)]{Kal03}
514: Kallman, T.~R., Angelini, L., Boroson, B., \& Cottam, J.\ 2003,
515: \apj, 583, 861
516:
517: \bibitem[Kalogera \& Baym(1996)]{Kal96}
518: Kalogera, V.~\& Baym, G.\ 1996, \apjl, 470, L61
519:
520: \bibitem[King \& Ritter(1999)]{Kin99}
521: King, A.~R.~\& Ritter, H.\ 1999, \mnras, 309, 253
522:
523: \bibitem[Li et al.(1978)]{Li78}
524: Li, F.~K., Clark, G.~W., Jernigan, J.~G., Laustsen, S.,
525: Zuiderwijk, E.~J., \& van Paradijs, J.~A.\ 1978, \nat, 276, 799
526:
527: \bibitem[Kolb(1998)]{Kol98}
528: Kolb, U.\ 1998, \mnras, 297, 419
529:
530: \bibitem[Marsh et al.(1994)]{Mar94}
531: Marsh, T.~R., Robinson, E.~L., \& Wood, J.~H.\ 1994, \mnras, 266, 137
532:
533: \bibitem[Mason et al.(1987)]{Mas87}
534: Mason, K.~O., Branduardi-Raymont, G., Codova, F.~A., \&
535: Corbet, R.~H.~D.\ 1987, \mnras, 226, 423
536:
537: \bibitem[Miller et al.(1998a)]{Mil98a}
538: Miller, M.~C., Lamb, F.~K., \& Psaltis, D.\ 1998a, \apj, 508, 791
539:
540: \bibitem[Miller et al.(1998b)]{Mil98b}
541: Miller, J.~C., Shahbaz, T., \& Nolan, L.~A.\ 1998b, \mnras, 294, L25
542:
543: \bibitem[Nelemans et al.(1999)]{Nel99}
544: Nelemans, G., Tauris, T.~M., \& van den Heuvel, E.~P.~J.\ 1999,
545: \aap, 352, L87
546:
547: \bibitem[Orosz \& Kuulkers(1999)]{Oro99}
548: Orosz J. A., Kuulkers E., 1999, \mnras, 305, 132.
549:
550: \bibitem[Phillips et al.(1999)]{Phi99}
551: Phillips, S.~N., Shahbaz, T., \& Podsiadlowski, P.\ 1999,
552: \mnras, 304, 839
553:
554: %\bibitem[Ramseyer(1994)]{Ram94}
555: % Ramseyer, T.~F.\ 1994, \apj, 425, 243
556:
557:
558: \bibitem[Shahbaz et al.(1999)]{Sha99}
559: Shahbaz, T., Kuulkers, E., Charles, P.~A., van der Hooft, F.,
560: Casares, J., \& van Paradijs, J.\ 1999, \aap, 344, 101
561:
562: \bibitem[Shahbaz et al.(2000)]{Sha00}
563: Shahbaz, T., Groot, P., Phillips, S.~N., Casares, J.,
564: Charles, P.~A., \& van Paradijs, J.\ 2000, \mnras, 314, 747
565:
566: \bibitem[Shahbaz et al.(2003)]{Sha03}
567: Shahbaz, T., Zurita, C., Casares, J., Dubus, G., Charles, P.~A.,
568: Wagner, R.~M., \& Ryan, E.\ 2003, \apj, 585, 443
569:
570: \bibitem[Thorstensen \& Charles(1979)]{Tho79}
571: Thorstensen, J.~R.~\& Charles, P.~A.\ 1979, \baas, 11, 721
572:
573:
574: \bibitem[Titarchuk \& Shaposhnikov(2002)]{Tit02}
575: Titarchuk, L.~\& Shaposhnikov, N.\ 2002, \apjl, 570, L25
576:
577: \bibitem[Tonry \& Davis(1979)]{Ton79}
578: Tonry, J.~\& Davis, M.\ 1979, \aj, 84, 1511
579:
580: \bibitem[Torres et al.(2002)]{Tor02}
581: Torres, M.~A.~P., Casares, J., Mart{\'{\i}}nez-Pais, I.~G., \&
582: Charles, P.~A.\ 2002, \mnras, 334, 233
583:
584: \bibitem[van Kerkwijk(2001)]{Ker01}
585: van Kerkwijk, M.~H. 2001, in Black Holes in Binaries and Galactic
586: Nuclei, ed. L. Kaper, E.~P.~J. van den Heuvel, \& P.~A. Woudt
587: (Berlin: Springer), 39
588:
589:
590: \bibitem[White \& Mason(1985)]{Whi85}
591: White, N.~E.~\& Mason, K.~O.\ 1985, Space Science Reviews, 40, 167
592:
593: \bibitem[White et al.(1995)]{Whi95}
594: White, N.~E., Nagase, F., Paramar, A.~N., 1995, X-ray Binaries,
595: eds.~W.H.G.~Lewin, J.~van Paradijs, and E.P.J.~van den Heuvel
596: (Cambridge: Cambridge Univ.~Press), p.~1
597:
598: \bibitem[Zhang et al.(1985)]{Zha97}
599: Zhang W., Strohmayer T.~E., Swank J.~H., 1997, ApJ, 482, L167.
600:
601: \end{thebibliography}
602:
603: \clearpage
604:
605: %% This example uses \plotone to include an EPS file scaled to
606: %% 80% of its natural size with \epsscale. Its caption
607: %% has been written to indicate that additional figure parts will be
608: %% available in the electronic journal.
609:
610:
611: %% Here we use \plottwo to present two versions of the same figure,
612: %% one in black and white for print the other in RGB color
613: %% for online presentation. Note that the caption indicates
614: %% that a color version of the figure will be available online.
615: %%
616:
617: \begin{table}
618: \begin{center}
619: \caption{Fits to the radial velocity curve of \target\ (1-$\sigma$ errors are given).
620: \label{RVFIT}}
621: \begin{tabular}{lccccccc}
622: \\
623: \tableline\tableline
624: Template & SpT & $\gamma$ & $P_{\rm orb}$ & $T_{\rm 0}$ $*$& $K_{\rm 2}$ & $\chi^2_{\nu}$ & $f$ \\
625: & & (\kms) & day) & & (\kms) & & \% \\
626: \tableline
627: HD122571 & G6III & 33.3 (3.8) & 9.0019(36) & 9.54(10) & 95.1(4.4) & 5.8 & 6.7 \\
628: HD63513 & G7III & 34.2 (3.6) & 9.0017(32) & 9.53(9) & 94.2(3.7) & 5.3 & 7.8 \\
629: HD40359 & G8III & 36.9 (3.1) & 9.0040(30) & 9.51(9) & 92.2(3.7) & 4.5 & 8.6 \\
630: HD71863 & G9III & 35.1 (3.3) & 9.0027(29) & 9.51(8) & 93.9(3.8) & 4.3 & 9.6 \\
631: HD82565 & K0III & 34.9 (3.3) & 9.0035(29) & 9.51(8) & 92.9(3.8) & 4.3 & 9.9 \\
632: HD39523 & K1III & 37.6 (3.2) & 9.0036(28) & 9.53(7) & 94.3(3.7) & 4.6 & 9.0 \\
633: HD61248 & K3III & 41.0 (3.5) & 9.0062(32) & 9.58(7) & 96.5(4.1) & 6.1 & 7.9 \\
634: HD40522 & K4III & 42.0 (3.7) & 9.0057(34) & 9.59(7) & 98.6(4.3) & 6.0 & 8.2 \\
635: \tableline
636: \end{tabular}
637: \end{center}
638: $*$HJD 2453099+
639: \end{table}
640:
641:
642:
643: \begin{figure*}
644: \includegraphics[angle=90,scale=.80]{f1.eps}
645: \caption{The heliocentric radial velocity curve of the secondary star. The
646: crosses show the SAAO data and the stars the VLT, NTT, AAT and Magellan data.
647: The solid line shows a sinusoidal fit to the data. The data have been
648: folded on the orbital ephemeris and 1.5 orbital cycles are shown for clarity.
649: \label{RVCURVE}}
650: \end{figure*}
651:
652: \begin{figure*}
653: \includegraphics[angle=90,scale=.80]{f2.eps}
654: \caption{
655: Top: the variance-weighted Doppler shifted average spectrum of \target\
656: in the rest frame of the secondary star.
657: The most noticeable features are the H and He emission lines arising
658: from the accretion disc.
659: Bottom: the K0{\sc iii}
660: spectral type template star HD82565. The spectra have been normalized and
661: shifted vertically for clarity.
662: \label{SPECTRUM}}
663: \end{figure*}
664:
665: \begin{figure*}
666: \includegraphics[scale=.35]{f3.eps}
667: \caption{
668: The effects of X-ray heating on the secondary star's radial velocity curve and
669: hence the allowed binary component masses. The left and right plots show the
670: case without and with X-ray heating respectively.
671: The solid and dashed
672: contours show the 68\% confidence level obtained by fitting the radial velocity
673: curve with the X-ray binary model using $i$=90$^\circ$ and 70$^\circ$
674: respectively (see section\,\ref{Irrad})
675: The dotted horizontal lines mark 1.4~\Msun\ and 2.9~\Msun, and
676: represent the canonical and maximum neutron star mass respectively.
677: \label{MASSES}}
678: \end{figure*}
679:
680: \end{document}
681:
682: %%
683: