1: \documentstyle[11pt,newpasp,twoside,epsf]{article}
2: \markboth{Freire, Gupta, Ransom and Ishwara-Chandra}{The PSR~J0514$-$4002A binary system}
3: \pagestyle{myheadings}
4: \nofiles
5:
6: % Some definitions I use in these instructions.
7:
8: \def\emphasize#1{{\sl#1\/}}
9: \def\arg#1{{\it#1\/}}
10: \let\prog=\arg
11:
12: \def\edcomment#1{\iffalse\marginpar{\raggedright\sl#1\/}\else\relax\fi}
13: \marginparwidth 1.25in
14: \marginparsep .125in
15: \marginparpush .25in
16: \reversemarginpar
17:
18: \begin{document}
19: \title{The PSR~J0514$-$4002A binary system in NGC 1851}
20: \author{P. C. Freire}
21: \affil{Arecibo Observatory, HC 3 Box 53995, Arecibo, PR 00612, USA}
22: \author{Y. Gupta}
23: \affil{National Centre for Radio Astrophysics, P.O. Bag 3, Ganeshkhind, Pune 411007, India}
24: \author{S. M. Ransom}
25: \affil{McGill University, 3600 University St., Montreal, QC H3A 2T8
26: Canada}
27: \author{C. H. Ishwara-Chandra}
28: \affil{National Centre for Radio Astrophysics, P.O. Bag 3, Ganeshkhind, Pune 411007, India}
29:
30: \begin{abstract}
31: Using the Giant Metrewave Radio Telescope (GMRT), we have discovered
32: PSR~J0514$-$4002A, a binary millisecond pulsar in the globular cluster
33: NGC~1851. This pulsar has a rotational period of 4.99 ms and the most
34: eccentric pulsar orbit yet found: $e = 0.89$. The orbital period
35: is 18.8 days, and companion has a minimum mass of 0.9~M$_{\odot}$; its
36: nature is presently unclear. After accreting matter from a low-mass
37: stellar companion, this pulsar exchanged it for its more massive
38: present companion. This system presents the strongest evidence to date
39: of such a process.
40: \end{abstract}
41:
42: \section{Introduction}
43:
44: Since 1987, several globular cluster surveys (see review by F. Camilo
45: and papers by A. Possenti, S. Ransom and J. Hessels, these proceedings) have
46: confirmed that most of the binary millisecond pulsars (MSPs) in
47: globular clusters (GCs) have low-mass white dwarf companions and
48: nearly circular orbits, as observed in the Galactic disk. This is an
49: important confirmation of the evolutionary scenarios proposed by
50: Alpar~et~al.~(1982) for the formation of MSPs.
51:
52: In GCs, {\em exchange encounters}, which only have a significant
53: probability of occurring in dense stellar environments, occasionally
54: exchange one of the components of a binary system with a typically
55: more massive star. The exchanges may occur during encounters with
56: either other binaries or with isolated stars. In GCs such encounters
57: can place isolated neutron stars into binaries with a main sequence
58: (MS) star which eventually evolves, ``recycles'' the neutron star, and
59: finally forms a MSP$-$WD binary system. Such a process explains the
60: anomalously large numbers of MSPs in GCs (by mass) when compared to
61: the Galaxy. If the companion to the low-mass MS star is a previously
62: recycled neutron star, we observe ``irregular'' eclipsing binary
63: pulsars (see review by P. Freire).
64:
65: In this paper we present some preliminary results on a new, unique
66: binary millisecond pulsar, PSR~J0514$-$4002A (henceforth NGC~1851A) in
67: the globular cluster NGC~1851. This has been found in a
68: new, sensitive 327-MHz survey of globular clusters carried out using
69: the Giant Metrewave Radio Telescope (GMRT), at Khodad near Pune, India.
70:
71:
72: \begin{figure}
73: \setlength{\unitlength}{1in}
74: \begin{picture}(0,2.8)
75: \put(0.1,-2.2){\special{psfile=./fig1.eps hscale=70 vscale=70 angle=0}}
76: \end{picture}
77: \caption{Discovery observation for NGC~1851A, in the globular cluster
78: NGC~1851. Pulsed emission is persistent throughout the whole scan. The
79: pulse profile is rather narrow (right), corresponding to the time
80: resolution of the system for this DM.}
81: \end{figure}
82:
83: \section{The GMRT 327-MHz survey and the discovery of NGC1851A}
84:
85: Our use of a low radio frequency for the present pulsar survey is to
86: be contrasted with most recent pulsar surveys, which are carried at
87: higher radio frequencies. This survey is aimed at faint pulsars
88: with steep spectra, which are unlikely to be detectable in a
89: high-frequency survey. This particular survey benefits from the
90: large gain of the central array of the GMRT, 4.6~K/Jy when used in the
91: phased array mode of the Array Combiner (Gupta et al. 2000, Prabu
92: 1997). This produces a beam on the sky with a diameter of about 3
93: arcminutes. The number of spectral channels across the available
94: 16-MHz band is 256. The sampling time used is 258$\,\mu$s, each
95: observation consisting of a pair of 72-minute scans containing
96: 2$^{24}$ samples. After this time, the 14-antenna central array is
97: re-phased using a reference source.
98:
99: We observed a set of 16 GCs in February 2003. The data were written to
100: tape at the GMRT and taken to McGill University, where they were
101: processed using the BORG (a 104-node Beowulf cluster available there
102: for pulsar processing) running the PRESTO software package (Ransom 2001).
103: One of the GCs observed was NGC~1851. Its distance ($D$) from the Sun
104: is about 12.6~kpc ({Cassisi}, {De Santis}, \&
105: {Piersimoni} 2001), and its Galactic coordinates are
106: $l = 244.51^\circ, b = -35.04^\circ$ (Harris 1996)\footnote{See
107: http://physwww.physics.mcmaster.ca/$\sim$harris/mwgc.dat
108: for an updated version of the table of globular cluster parameters
109: presented in this paper}. It is
110: a relatively bright globular cluster ($M_v\,=\,-8.33$) with a very
111: condensed core ($c = log(r_t/r_c) = 2.32$, where $r_t$ and
112: $r_c$ are the tidal and core radii). It is among the ten clusters in
113: the Galaxy with the highest central luminosity density
114: ($\rho_0\,\simeq\,2 \times 10^5 \rm L_{\odot} pc^{-3}$).
115:
116: \begin{figure}
117: \setlength{\unitlength}{1in}
118: \begin{picture}(0,3.5)
119: \put(-0.3,-0.5){\special{psfile=./fig2.ps hscale=70 vscale=70 angle=0}}
120: \end{picture}
121: \caption{The measured barycentric rotational periods of NGC~1851A as a function of MJD.}
122: \end{figure}
123:
124: In the first scan for NGC~1851, taken on the 10 of February 2003, we
125: detected a clear pulsed signal with a period of 4.991 ms (see Figure
126: 1) and a DM of 52.15(10)~cm$^{-3}\,$pc. Analysis of the rotational
127: periods using {\sc tempo}\footnote{http://pulsar.princeton.edu/tempo}
128: proved most surprising: the best-fit model (see Figure 2) indicates
129: $e\,=\,0.889(2)$, the most eccentric orbit of any known binary pulsar
130: and many orders of magnitude more eccentric than the typical MSP orbit
131: (for this and following quantities, the number in parenthesis
132: indicates the $1-\sigma$ uncertainty, which we conservatively estimate
133: to be ten times the formal value computed by {\sc tempo}, as there is
134: still no phase-coherent timing solution for this pulsar). The
135: orbital period $P_b$ is 18.7850(8) days, and the
136: semi-major axis of the orbit projected along the line-of-sight ($x$) is
137: $36.4(2)$ light seconds. This implies a minimum companion mass of
138: 0.9~M$_{\odot}$, assuming a pulsar mass of 1.35~M$_{\odot}$ (for the
139: median of expected inclinations, 60$^\circ$, the companion mass is
140: 1.1~M$_{\odot}$). The epoch of periastron $T_0$ is MJD =
141: 52984.46(2), the longitude of periastron $\omega$ is
142: $82(1)^\circ$ and the rotational period $P$ is 4.990576(5) ms.
143: Imaging with the GMRT, made during the search and
144: confirmation observations, has shown that the pulsar is located very
145: near the centre of the cluster (see Figure 3), just ouside the
146: 0.06-arcminute core. The estimated flux density at 327~MHz is about
147: 3.4$\pm$0.4 mJy.
148:
149: \begin{figure}
150: \setlength{\unitlength}{1in}
151: \begin{picture}(0,5.0)
152: \put(-0.1,-1.0){\special{psfile=./fig3.ps hscale=70 vscale=70 angle=0}}
153: \end{picture}
154: \caption{A radio image of NGC~1851 at 327
155: MHz, made by combining the data from 4 different epochs of
156: observations in December 2003. The pulsar is the faint source near
157: the center of the cluster (indicated by the cross). The brighter
158: source to the north-east is not the pulsar, when the interferometer
159: was pointed at it no pulsed emission was detected. The small ellipse
160: in the lower right corner indicates the dimensions of the synthesized
161: interferometer beam, this interferometer includes {\em all} the
162: antennas in the array (the pulsation survey used only the central
163: square, which produces a 3-arcminute beam).}
164: \end{figure}
165:
166: \section{Formation and nature}
167: \label{sec:formation}
168:
169: All known eccentric ($e > 0.1$) binary pulsars in the disk of the
170: Galaxy have relatively massive companions. This varied set of
171: companions includes blue giants, other neutron stars and heavy WDs.
172: Blue giants live only a few Myr which is probably not long enough to
173: allow the sustained mass accretion required to spin up a pulsar
174: companion to millisecond spin periods. This is in accordance with the
175: observations, where the pulsars with massive companions have
176: rotational periods of tens or hundreds of milliseconds.
177: In addition, the second supernova event, where the giant becomes a
178: massive compact object, is likely to make the orbit significantly
179: eccentric, presuming the binary survives. Since both stars are now
180: compact, tidal circularization is henceforth impossible. The
181: prolonged episode of stable mass accretion needed to spin up a neutron
182: star to millisecond periods is only possible from evolved lower-mass
183: MS stars. Such large timescales allow effective tidal orbit
184: circularization as well. This process likely created the MSP currently
185: in NGC~1851A, although with a presently unknown low-mass WD companion.
186:
187: Some MSP $+$ low-mass white dwarf systems found in GCs, like
188: PSR~B1802$-$07 ($P = 23.1$\,ms, $P_b = 2.62$\,days, $e = 0.212$,
189: D'Amico et al. 1993) can become mildly eccentric due to interactions with
190: other objects in the cluster (Rasio and Heggie, 1995).
191: This is almost certainly not
192: the origin of the present NGC~1851A binary system since its
193: eccentricity is probably too high to be explained by this mechanism.
194: We are therefore lead to the conclusion
195: that after recycling, NGC~1851A exchanged its former low-mass
196: companion with a more massive object. This is the first system
197: presenting clear evidence of such a process. A massive star (the
198: pulsar's present companion) passed within a distance smaller than
199: about four times the separation of the components of the previous
200: binary system, a not uncommon event in the dense environment of a GC.
201: The most likely outcome from such an event is the formation of a
202: slightly tighter eccentric binary system containing the two more
203: massive objects (Hut 1996).
204:
205: The companion of NGC~1851A has a minimum mass of 0.9\,M$_{\odot}$ and
206: its nature is as yet unclear, it could be either a compact or extended
207: object. In NGC~1851, a cluster with an age of $\sim$9\,Gyr (Salaris \&
208: Weiss 2002)
209: where 1-M$_{\odot}$ stars are now leaving the main sequence, such
210: objects should be readily detectable by the HST. Because of the
211: lengthy episode of MSP recycling that preceded its formation, the
212: present NGC~1851A binary system is very likely to be a few Gyr younger
213: than the cluster in which it lies. Mathieu, Meibom and Dolan
214: (2004) have determined that, for the open cluster
215: NGC~188 (with an age of 7 Gyr and a stellar population similar to that
216: of GCs), binary systems containing MS stars with orbital periods
217: larger than 15 days have not yet had time to circularize. Therefore,
218: the observed eccentricity of the NGC~1851A system does not rule out
219: the possibility of the companion being an extended object. In fact,
220: partial and/or irregular ``eclipses'' from an extended object such as
221: a MS star may explain the apparent flux variability from NGC~1851A.
222:
223: \section{Conclusion}
224:
225: We have discovered a remarkable 5-ms binary pulsar, the first to be
226: found in the GC NGC~1851 and the first pulsar to be discovered with
227: the GMRT. Its orbit is the most eccentric known for any system
228: containing a pulsar, while its rotational period is much shorter than
229: that of any other pulsar in an eccentric binary system. This
230: combination of characteristics indicates that, after becoming an MSP
231: by accreting matter from a low-mass companion star, this neutron star
232: has almost certainly exchanged it for its present, significantly more
233: massive companion. This is the most effective way of forming a binary
234: system containing a millisecond pulsar and a black hole, provided
235: black holes exist in GCs. Follow-up studies of this object will allow
236: us to determine the nature of the companion and hopefully
237: measure the masses of both components of this binary system.
238:
239: \begin{thebibliography}{}
240: \bibitem[Alpar {et al.} 1982]{acrs82}
241: {Alpar}, M.~A., Cheng, A.~F., Ruderman, M.~A., \& Shaham, J. 1982, Nature, 300,
242: 728
243:
244: \bibitem[{Cassisi}, {De Santis}, \& {Piersimoni} 2001]{csp01}
245: {Cassisi}, S., {De Santis}, R., \& {Piersimoni}, A.~M. 2001, MNRAS, 326, 342
246:
247: \bibitem[D'Amico {et al.} 1993]{dbl+93}
248: D'Amico, N., Bailes, M., Lyne, A.~G., Manchester, R.~N., Johnston, S.,
249: Fruchter, A.~S., \& Goss, W.~M. 1993, MNRAS, 260, L7
250:
251: \bibitem[{Gupta} {et al.} 2000]{ggj+00}
252: {Gupta}, Y., {Gothoskar}, P., {Joshi}, B.~C., {Vivekanand}, M., {Swain}, R.,
253: {Sirothia}, S., \& {Bhat}, N.~D.~R. 2000, in { ASP Conf. Ser. 202: IAU
254: Colloq. 177: Pulsar Astronomy - 2000 and Beyond}, 277
255:
256: \bibitem[Harris 1996]{har96}
257: Harris, W.~E. 1996, Astron. J., 112, 1487
258:
259: \bibitem[{Hut} 1995]{hut96b}
260: {Hut}, P. 1995, in { ASP Conf. Ser. 72: Millisecond Pulsars. A Decade of
261: Surprise}, 46
262:
263: \bibitem[Mathieu, Meibom, \& Dolan(2004)]{mmd04} Mathieu,
264: R.~D., Meibom, S., \& Dolan, C.~J.\ 2004, ApJ, 602, L121
265:
266: \bibitem[Prabu 1997]{pra97}
267: Prabu, T. 1997.
268: \newblock Master's thesis, Indian Institute of Science.
269: \newblock see also http://www.rri.res.in/$\sim$dsp\_ral/gac/gacmain.htm
270:
271: \bibitem[Ransom 2001]{ran01}
272: Ransom, S.~M. 2001.
273: \newblock PhD thesis, Harvard University
274:
275: \bibitem[Rasio \& Heggie 1995]{rh95}
276: Rasio, F.~R. \& Heggie, D.~C. 1995, ApJ, 445, L133
277:
278: \bibitem[{Salaris} \& {Weiss} 2002]{sw02}
279: {Salaris}, M. \& {Weiss}, A. 2002, \aap, 388, 492
280:
281:
282: \end{thebibliography}
283:
284: \end{document}
285: