astro-ph0002014/part1
1: 
2: %                                                                 aa.dem
3: % AA vers. 4.01, LaTeX class for Astronomy & Astrophysics
4: % demonstration file
5: %                                                 (c) Springer-Verlag HD
6: %-----------------------------------------------------------------------
7: %
8: \documentclass[referee]{aa} % for a referee version
9: %
10: \usepackage{epsfig,deluxe}
11: %
12: \begin{document}
13: 
14: \newcommand{\gsim}{\hbox{\rlap{$^>$}$_\sim$}}
15:   \thesaurus{06;  19.63.1}    
16: % A&A Section 6: Form. struct. and  evolut. of stars}
17: %  \thesaurus{06     % A&A Section 6: Form. struct. and evolut. of stars
18: %             (03.11.1;  % Cosmogony,
19: %              16.06.1;  % Planets and satellites: general,
20: %              19.06.1;  % Solar system: general,
21: %              19.37.1;  % Stars: formation of,
22: %              19.53.1;  % Stars: oscillations of,
23: %              19.63.1)} % Stars: structure of.
24: %
25:     \title{Soft Gamma-ray Repeaters and Anomalous X-ray Pulsars: 
26:            Magnetars or  Quark Stars?} 
27:     \author{Arnon Dar$^1$ and A. De R\'ujula$^2$}  
28:     \institute{1. Department of Physics, Technion, Haifa 32000, Israel\\ 
29:               2. Theory Division, CERN, CH-1211 Geneva 23, Switzerland\\} 
30:     \titlerunning{Soft Gamma-ray Repeaters ...} 
31:     \maketitle 
32: 
33:  
34: \begin{abstract}  
35: 
36: Recent measurements of the spin-down rates of soft gamma ray repeaters
37: (SGRs) and anomalous X-ray pulsars (AXPs) have been interpreted as
38: evidence that these objects are ``magnetars'': neutron stars spinning down
39: by magnetic dipole radiation, but with a magnetic field two orders of
40: magnitude larger than that of ordinary neutron stars. We discuss the
41: evidence disfavouring this interpretation. We argue that, instead, the
42: observations support the hypothesis that SGRs and AXPs are neutron stars
43: that have suffered a transition into a denser form of nuclear matter to
44: become, presumably, strange stars or quark stars.
45: %Their gravitational energy release 
46: %while they are cooling and contracting can power both their quiescent 
47: %X-ray emission and their star quakes which produce `soft' gamma ray 
48: %bursts.  
49: \end{abstract}
50: 
51: \section{Introduction} 
52: Consider a neutron star (NS) of radius R, with a magnetic field
53: ${\rm B_p}$  at the magnetic poles, spinning with a period P.
54:  If the star's magnetic
55: moment is misaligned with the spin axis by an angle $\alpha$,
56: electromagnetic energy is emitted at a rate 
57: (see, e.g., Shapiro \& Teukolsky 1983 and references therein.)  
58: \begin {equation}  
59: {\rm \dot E = -\,{B_p^2\, R^6\,\Omega^4 \, \sin^2\alpha\over 6\,c^3}}\; ,
60: \end{equation}  
61: with $\Omega=2\pi/{\rm P}$ the angular velocity.
62: If the NS is observed as a pulsar, its period can be measured as
63: a function of time. Let the NS have a moment of inertia I and 
64: let ${\rm \dot P}$ be the rate
65: at which its period decreases. The magnetic field required to explain
66: the slow-down rate by magnetic dipole radiation (MDR) is: 
67: \begin{equation}  
68: {\rm 
69: B_p=\left({6\,c^3I\over 4\,\pi^2\,R^6\,  
70: \sin^2\alpha}\right)^{1/2}~\sqrt{\dot P\,P}}\, .
71: \label{field}
72: \end{equation} 
73: For a ``canonical'' NS, 
74: ${\rm M\sim 1.4~M_\odot}$, R $\sim$ 10 km   and 
75: ${\rm I\sim 10^{45}~g~cm^2}$, so that   
76: ${\rm B_p=6.4\times 10^{19}}$ ${\rm \sin^{-1}\alpha\,\sqrt{\dot P P_s}}$ 
77: Gauss, with ${\rm P_s}$ the period in seconds.
78: Young radio pulsars such as the Crab pulsar are fast-rotating NSs 
79: with an estimated surface magnetic field ${\rm B_p\!\sim \!10^{13}}$  
80: Gauss. They fit the picture whereby their observed slow-down rates
81: are due to the MDR of their rotational energy${^{^1}}$.  
82: 
83: All observed slowly-rotating pulsars, of period ${\rm P\!>\!5}$ s, 
84: exhibit very peculiar 
85: properties. Four of them are known to emit occasional short bursts 
86: of radiation peaking at tens of keV energies,
87:  and are classified as ``Soft Gamma-ray Repeaters'' (SGRs).
88: Six other ``Anomalous X-ray Pulsars'' (AXPs)
89: persistently emit X-rays, but they are quiet at radio wavelengths.
90: Some of these objects'
91: observational parameters
92: are listed in Tables I and II.  Their rotational
93: energy loss  ${\rm \dot E_{rot}=I~\Omega~\dot\Omega}$ is insufficient  
94: to power
95: their observed radiation, suggesting the magnetic field energy 
96: ${\rm E_m\sim B_p^2~R^3/12}$ 
97: as an alternative source. 
98: Both SGRs and AXPs are located in directions close to those of
99: supernova remnants (e.g., 
100: Cline, et al. 1982;  Kulkarni \& Frail 1993; Vasisht et al. 1994;  
101: Hurley, K. et al. 1999;
102: Vasisht et al. 1997a; Vasisht et al. 1997b;
103: Gaensler, Gotthelf \& Vasisht 1999) 
104: that are observable for
105: only tens of thousands of years. The association with SNRs
106: would imply that these peculiar stars are
107: too young to have spun-down by MDR to their
108: current periods, if they were born rotating as fast as
109: the Crab pulsar, the Vela pulsar or any other young
110: pulsar (the ones also associated with SNRs),  and if they also have
111: characteristic radio-pulsar magnetic fields:  
112: ${\rm B\!\sim\! 10^{13}}$ Gauss.  
113: 
114: The values of P and ${\rm \dot P}$ for SGR 1806$-$20 and SGR 1900+14 were
115: recently measured (Kouveliotou et al. 1998; Kouveliotou et al. 1999; Woods
116: et al. 1999) and are listed in Table I. They imply, by use of
117: Eq. 2, magnetic fields in excess of $10^{15}$ Gauss. With the
118: information displayed in Table II, similarly large fields can be deduced
119: for AXPs (e.g., Gotthelf et al. 1999; Israel et al. 1999). Magnetic fields
120: of this enormous intensity can explain the rapid slow-down of SGRs and
121: AXPs and can store enough energy to power their emissions during their
122: active lifetime (Duncan and Thompson 1992; Thompson and Duncan 1995;
123: Thompson and Duncan 1996). Not surprisingly, the discovery of SGRs and
124: AXPs with fast spin-down rates was reported as the observational discovery
125: (Kouveliotou et al. 1998; Kouveliotou et al. 1999) of hypermagnetized
126: neutron stars, or {\it magnetars} (Duncan and Thompson 1992; Thompson and
127: Duncan 1995; Thompson and Duncan 1996). 
128:  
129: 
130: In this paper we discuss the observational evidence implying that SGRs and
131: AXPs are not magnetars (Dar 1999a; Marsden et al. 1999; Dar 1999b).  We
132: contend that the mechanism producing their observed rapid slow-down is not
133: MDR, but relativistic particle emission along the magnetic axis, be it in
134: the form of jets (Dar 1999b) or of winds (Harding et al. 1999). We argue
135: that the locations and estimated ages of SGRs, when compared to those of
136: their associated SNRs, strongly suggest that ``something'' happened to
137: these NSs well after they were born. Their inferred X-ray emitting surface
138: areas, significantly smaller than those of a ``canonical'' neutron star,
139: point in the same direction.  For the source of the emitted energy we do
140: not have an explicit model; we conjecture that the energy gained by steady
141: gravitational contraction can power both the quiescent X-ray emission and
142: the star quakes that produce ``soft'' gamma ray bursts (Ramaty et al.
143: 1980). A phase transition from a conventional neutron star into a strange
144: star or a quark star (Dar 1999a) can explain, we shall argue, all of the
145: properties of SGRs and AXPs. 
146: 
147: 
148: \section{Critique of the magnetar model} 
149:  
150: The magnetar model of SGRs cannot explain their ages, locations
151: and occasional increases in spin-down rate (Dar 1999a, Marsden et al. 1999).
152: The ages of SGRs, if estimated from their magnetic spin-down rate, are 
153: much smaller than the ages of the remnants of the supernovae in 
154: which they were born: an {\it age crisis}. The location of SGRs relative to
155: the centre of their associated remnant implies that they move with  
156: unacceptably
157: large peculiar velocities: a {\it separation crisis.} Sudden increases  
158: in ${\rm \dot P}$
159: require inexplicable jumps in the energy stored by the magnetic field: an
160: {\it energy crisis.}
161: 
162: A pulsar with initial spin period  ${\rm P_i}$
163: and constant moment of inertia, whose
164: rotational kinetic energy ${\rm E_{rot} =I~\Omega^2/2}$
165: powers the MDR, has an age, t, shorter than the ``characteristic  
166: age'', ${\rm \tau_s}$:
167: \begin{equation}  
168: {\rm  t = {P\over 2\,\dot P}\left(1-{P_i^2\over P^2}\right) 
169: \leq \tau_s \equiv {P\over 2\,\dot P}}\; . 
170: \label{age}
171: \end{equation}  
172: An age estimate independent from the above ``magnetic braking'' age 
173: is provided by the time elapsed since
174: the parent supernova event took place. The ejecta from 
175: SNe expand freely to a radius ${\rm R_{SNR}\propto t}$, as long as the 
176: mass of the swept-up ambient medium 
177: is smaller than the mass of the ejecta. When they are comparable, the
178: SNR enters a ``Sedov-Taylor phase'' during which ${\rm R_{SNR}\propto 
179: t^{2/5}}$. Finally, for a swept up mass superior to the ejected mass,  
180: the SNR 
181: cools radiatively and ${\rm R_{SNR}\propto t^{2/7}}$.  The expansion 
182: velocity and the size of SNRs, as well as their X-ray temperatures, are 
183: commonly used to estimate their ages (see, e.g., Shapiro \& Teukolsky 
184: 1983 and references therein.).  
185: 
186: In Table I we list the characteristic  ages $\tau_s$
187: of SGRs, and the ages of the SNRs where they were presumably born.
188: For the two SGRs whose slow-down rate ${\rm \dot P}$ has been measured  
189: (1806-20 and 1900+14), Eq.~\ref{age} results in $\tau_s\approx  
190: 1400,~1300$
191: y, respectively. This upper limit is significantly smaller than the  
192: estimated
193: age of their SNRs, which is larger than $5\times 10^3$ and $10^4$ y,
194: respectively. This age crisis of the magnetar model would recur
195: for SGR 0529-66 and 1627-41, if their slow-down rate is similar to
196: that of the other two known SGRs.
197:  
198: The characteristic ages of SGR 1806-20 and 1900+14 also imply a separation
199: crisis.  SGR 1806+20, if it was born at the centre of SNR G10.0-0.3
200: (Hurley et al. 1999b) and is less than ${\rm \tau_s\approx 1400~y}$ old,
201: must have travelled with a sky projected velocity larger than ${\rm
202: v_\perp\approx 5500~(D/14.5~kpc)~km~s^{-1}}$ to its present location
203: (Kulkarni et al. 1994). With its similar ${\rm \tau_s}$, SGR 1900+14, if
204: associated to SNR G42.8+0.6 (Vasisht et al. 1994), must have travelled at
205: ${\rm v_\perp\sim 27000~(D/7~kpc)~km~s^{-1}}$, or faster, to where it is
206: (Hurley et al. 1999c; Hurley et al. 1999d). The same argument, applied to
207: SGR 0526-66 and 1627-41 at their current locations (Cline et al. 1982;
208: Hurley et al. 1999a), would result in lower limits of ${\rm v_\perp\sim
209: 22000}$ and ${ \sim 4000}$ km s$^{-1}$, respectively, if their spin-down
210: rates and implied ages turned out to be akin to the measured ones. These
211: magnetar-model velocities are too large, compared to the mean observed
212: ${\rm v_\perp\sim 350\pm 70~km~s^{-1}}$ of pulsars (e.g., Lyne and Lorimer
213: 1994), and in particular of young pulsars such as the Crab pulsar 
214: (${\rm v_\perp\approx 170~km~s^{-1}}$, Caraveo and Mignani 1999) and the
215: Vela pulsar (${\rm v_\perp\approx 70~km~s^{-1}}$, Nasuti et al. 1997). 
216: 
217: 
218: In the magnetar model of SGRs the radiation-energy source 
219: is the magnetic field energy 
220: ${\rm E_M \sim B_p^2\,R^3/12}$ $\sim 10^{47}$ erg, for 
221: ${\rm B_p\!=\!10^{15}}$ Gauss and  ${\rm R\!=\!10}$ km.
222: Magnetic braking implies that the pulsar's surface field is  
223: ${\rm B_p^2\propto \dot P}$, an increase in ${\rm \dot P}$  implies a  
224: commensurate
225: increase in magnetic energy. The 
226: spin-down rate of SGR 1900+14 roughly doubled from  
227: ${\rm \dot P\sim 6\times 10^{-11}}$ to ${\rm \dot P\sim 13\times  
228: 10^{-11}}$  
229: around the time of its large flare on 27 August 1998
230: (Woods et al. 1999a, Marsden et al. 1999). 
231: How to explain a sudden doubling of a huge magnetic energy?
232: This is the energy crisis. As the magnetic energy is consumed
233: and the field weakens, the pulsar's spin-down rate should decrease,
234: countrary to observation: yet another problem for the magnetar
235: scenario (Marsden et al. 1999). 
236: 
237: For AXPs the magnetar model faces similar difficulties. AXPs 1709-40 and
238: 1E 1048-5937 have spin-down ages (9 ky and 4.6 ky) shorter than the
239: estimated age of their associated SNRs (20 ky and 10 ky), hinting at an
240: age crisis.  The projected sky velocities required to move these objects
241: from the centres of their associated SNRs (G346.6-0.2 and G287.8-0.5) to
242: their observed positions are 2100 km s$^{-1}$ and 2300 km s$^{-1}$, a
243: separation crisis. Observed jumps in the spin-down rate of AXPs
244: 1E1048.1-5937 and 1E2259+58, akin to the one in SGR 1900+14, entail an
245: energy crisis.  For AXP 1E 2259+586 (Corbet et al. 1995), the
246: magnetic energy inferred from its spin-down rate, ${\rm E_B\approx 2\times
247: 10^{45}}$ erg, is insufficient to power its steady X-ray luminosity, ${\rm
248: L_X\approx 8\times 10^{34}~erg ~s^{-1}}$, over its characteristic age,
249: ${\rm \tau_s\sim 1.5\times 10^5~y}$. Also, a magnetic field this large
250: would be inconsistent with the absorption features observed by ASCA in its
251: X-ray spectrum (Corbet et al. 1995), if interpreted as cyclotron lines. 
252:  
253: The magnetar model of SGRs and AXPs is not successful: alternatives are
254: called for.
255: 
256: 
257: 
258: 
259: \section{Spin-Down by Relativistic Jets}  
260:  
261: There is evidence for the emission of relativistic particles  
262: by SGRs. 
263: In the case of SGR 1806-20, the
264: non-thermal quiescent X-ray emission and the highly suggestive radio 
265: images (Vasisht et al. 1995; Frail et al. 1997) provide 
266: compelling evidence for steady relativistic particle emission,
267: perhaps in the form of relativistic jets.  A 
268: fading radio source is seen within the 
269: localization window of SGR 1900+14; it has 
270: been interpreted as a short-lived nebula 
271: powered by relativistic particles ejected during the intense high energy 
272: activity in late August 1998 (Frail et al. 1999). The 
273: emission of relativistic particles along the magnetic axis can be the 
274: dominant mechanism for the braking of slowly-rotating 
275: pulsars with normal magnetic fields (Dar 1999b, see also Harding et al.
276: 1999), for which magnetic braking is  
277: inefficient. Magneto-hydrodynamic calculations 
278: of pulsar braking by particle emission are a formidable task, but simple   
279: estimates (Dar 1999b) will suffice here.  
280:  
281: Let L$_{\rm RP}$ be a pulsar's luminosity in the form of relativistic
282: particles escaping from the magnetic poles along the open 
283: magnetic lines. The emitted particles co-rotate with the magnetic field
284: up to a radius ${\rm r_e \approx (3\,c\,B_p^2\,R^6/2\,L_{RP})^{1/4}}$,  
285: at which
286: their pressure (${\rm L_{RP}/[12\pi\,c\,r^2}]$) becomes comparable to  
287: the magnetic
288: pressure (${\rm B_p^2\,R^6/[8\pi\,r^6]}$ for a dipole field). 
289: Beyond this point a particle
290: of mass m and Lorentz factor $\gamma$ is no longer entangled
291: in the magnetic field and it escapes, carrying away an angular momentum
292: ${\rm \gamma\,m\,\Omega\,r_e^2\,\sin^2\alpha}$. The resulting rate of
293: rotational energy loss is 
294: \begin{equation}
295: {\rm \dot E_{rot}=I\,\Omega\,\dot\Omega 
296: \approx -  
297: \left( {3\,L_{RP}\over 2\,c^3}\right)^{1/2}\,B_p\,R^3\,\Omega^2 
298: \,\sin^2\alpha},
299: \end{equation} 
300: which yields an exponential 
301: decline, ${\rm E_{rot}(t)=E_{rot}(0)\,exp(-t/\tau_s)}$, with a 
302: characteristic time: 
303: \begin{equation}  
304: {\rm \tau_s={P\over 2\,\dot P}={I\over B_p\,R^3\,\  
305: sin^2\alpha}\left({c^3\over  
306: 6\,L_{RP}}\right)^{1/2}}.  
307: \end{equation} 
308: For a conventional ${\rm B_p=10^{13}}$ Gauss, ${\rm R=10}$ km,
309: ${\rm L_{RP}=10^{37}}$ erg s$^{-1}$, 
310: and $\sin\alpha\approx 1$, the characteristic slow-down time is  
311: ${\rm \tau_s\sim  2000~y}$, scaling as 1/R at fixed  ${\rm B_p\,R^2}$,
312: and consistent with the characteristic slow-down times of SGRs and AXPs.  
313: 
314: Relativistic particle emission may also be the dominant spin-down mechanism
315: in pulsars rotating faster than SGRs and AXPs.
316: A comparison between Eq.4 and Eq.1 shows that slowing-down by relativistic 
317: jets becomes faster than slowing-down by MDR when 
318: \begin{equation}
319: {\rm P\geq {(4\pi^2\,B_p\,R^3)^{1/2}\over (54\,c^3\,L_{RP})^{1/4}}}.
320: \end{equation}
321: For ${\rm B_p=10^{13}}$ Gauss, ${\rm R=10}$ km,
322: and ${\rm L_{RP}=10^{37}}$ erg s$^{-1}$, 
323: slowing-down by emission of relativistic jets is faster than by MDR if 
324: ${\rm P> 60~ms}$. But, for
325: ${\rm B_p=10^{13}~ Gauss}$, ${\rm L_{RP}<10^{37}~erg}$, 
326: and ${\rm P<100~ms}$, 
327: ${\rm r_e\,\sin\alpha}$ 
328: becomes larger than the radius of the light cylinder, ${\rm r_c=c/\Omega}$,
329: and  
330: relativistic particles of energy E that stop co-rotating with the pulsar 
331: at ${\rm r_c}$ carry away an angular momentum ${\rm E/\Omega}$,
332: so that the total rate of angular momentum loss by particle emission is
333: ${\rm \dot L\approx L_{RP}/\Omega} $, i.e., 
334: \begin{eqnarray}
335: {\rm \dot  E_{rot}\approx L_{RP}, \qquad}
336: {\rm \tau_s\approx E_{rot}/L_{RP}}.
337: \end{eqnarray}
338: %and 
339: %\begin{equation} 
340: %{\rm \tau_s\approx E_{rot}/L_{RP}}.
341: %\end{equation}
342: The relation ${\rm \dot E_{rot}=L_{RP}}$ is well  
343: satisfied, for instance, by the Crab pulsar,  for which 
344: ${\rm \dot E_{rot}=I\,\Omega\,\dot\Omega\approx}$ ${\rm 5\times 
345: 10^{38}~erg~s^{-1}}$, coinciding exactly with the estimated energy
346: input to the Crab nebula (Manchester and Taylor 1997), presumably 
347: supplied by relativistic particles from the pulsar.  
348:  
349: The gamma-ray bursts and radio flares of SGRs are presumably produced by
350: bursts of relativistic particles.
351: If relativistic particle emission induces
352: the observed spin-down, ${\rm \dot P}$ should increase during these periods
353: of activity.  Indeed, the
354: spin-down rate of SGR 1900+14 doubled during its intensive burst 
355: activity in 1998, after which it seems to resume
356:  its ``quiescent'' long-term value (Woods et al. 1999a), as shown in  
357: Fig.~1.  
358: 
359: 
360: \section{What powers SGRs and AXPs?} 
361: 
362: If, unlike in the magnetar model, the energy reservoir  of SGRs and AXP
363: is not magnetic, what can it be? 
364: A NS whose internal heat, magnetic field and/or angular momentum 
365: are diminishing as it radiates, may undergo a phase transition
366: (see, e.g., Shapiro \& Teukolsky 1983 and references therein) 
367: and collapse to a strange star (SS) or a quark star (QS).
368: Gravitational energy release during the subsequent
369: slow contraction of the cooling and spinning-down star may 
370: power  SGRs and AXPs (Dar 1999a).  
371: The equation of state of nuclear matter, or even that of quark matter at
372: supernuclear densities, has not yet been derived from first principles.
373: Yet, simple considerations indicate
374: that the possible phase transitions of NSs into SSs and QSs ought to  
375: be taken
376: seriously.
377: 
378: Naively approximate the pressure of cold nuclear matter at NS   
379: densities by that
380: of a non-relativistic degenerate Fermi gas of nucleons. Ignoring
381: general-relativistic corrections, the radius and central  
382: density ${\rho_c}$ of a self-gravitating gas of neutrons of  
383: total baryonic mass M  and zero angular momentum  
384: are then given by the polytropic Emden-Lane solution of the  
385: hydrostatic equation:
386: \begin{equation} 
387: {\rm R\approx 15.1 
388:        \left({M\over M_\odot}\right)^{-1/3}~km},  
389: \end{equation} 
390:  \begin{equation} 
391: {\rm \rho_c\approx 6\,\bar\rho \approx 0.83\times 10^{15} 
392:              \left({M\over M_\odot}\right)^2~g~cm^{-3}}.  
393: \end{equation} 
394: In this simplest of models, low mass NSs should indeed be made of neutrons,
395: but as M is increased past ${\rm 1.27\,M_\odot}$, 
396: ${\rm \rho_c}$ increases until the central Fermi energy
397: ${\rm E_F=(h^2/8\,m_n)(3\,\rho_c/\pi\, m_n)^{2/3}}$ exceeds 
398: ${\rm (m_\Lambda-m_n)\,c^2}$. At this point, it is favourable for the
399: strangeness changing weak process  ${\rm n\to\Lambda}$ (or ${\rm 
400: ud\rightarrow su}$) to start transforming neutrons at the top of the  
401: Fermi sea 
402: into (initially pressureless) $\Lambda$'s at the bottom of the sea.
403: This reduces the pressure, causes contraction and increases ${\rho_c}$,
404: initiating a run-away reaction that stops only as the n and $\Lambda$
405: chemical potentials equalize, i.e. 
406: until ${\rm E_F(n)-E_F(\Lambda)\approx}$ 
407: ${\rm  c^2\,(m_{_\Lambda}-m_n)(1-GM(r)/r)}$, 
408: where ${\rm M(r)}$ is the mass enclosed within r.
409: 
410: 
411: At the central densities of the strange stars of the previous paragraph,
412: the nucleons would be so snuggly packed that their ``individuality'' would
413: be in doubt.
414: But it has been argued (Alford et al. 1998; Berges et al. 1999
415: Rapp et al. 1999; Wilczek 1999; Li et al. 1999)  that cold 
416: nuclear matter,  
417: compressed to high nuclear densities, converts into a much denser
418: superfluid and superconducting Bose-condensate of spin zero diquarks.
419: Cooper pairing of quarks 
420: reduces their pressure and would trigger a gravitational collapse that,
421: if it does not proceed all the way to a black hole, would stop
422: only when the squeezed size of the pairs increases their internal 
423: energy above their binding energy.  
424: 
425: A neutron star may be born with a temperature, a magnetic field and/or
426: an angular momentum that prevent its transition to a strange- or  
427: quark-matter
428: state. As the star ages, it may reach a point at which a transition to
429: a denser state of matter is favourable. The collapse would
430: reheat the star to some extent; we conjecture that the gravitational energy
431: made available by its subsequent slow cooling and contraction 
432: can power SGRs and AXPs (estimates of the effect are difficult, since  
433: quantities
434: such as the heat conductivity are notoriously hard to predict).
435: For a pulsar which is mainly supported by the Fermi pressure of 
436: non-relativistic degenerate fermions contraction can power a total
437: luminosity:
438: \begin{equation}  
439: {\rm 
440: L\approx {2\over 7}\left({G\,M^2\over R}\right) {\dot R\over R}
441: }\; . 
442: \end{equation}  
443: For a canonical pulsar mass ${\rm M=
444: 1.4~M_\odot}$ and a radius ${\rm R=10}$ km, a contraction rate of  
445: ${\rm \dot 
446: R \sim 20~\mu m~ y^{-1}}$ (a tiny ${\rm \dot R/R\sim 2\times  
447: 10^{-9}~y^{-1}}$)
448: is sufficient to provide the inferred total luminosity of 
449: SGRs and AXPs, ${\rm L \leq 10^{37}~erg~s^{-1}}$ .  
450: 
451: \section{Extra evidence and hints in favour of collapsed NSs} 
452:  
453: The gravitational collapse of 
454: pulsars to strange or quark stars may offer explanations for some 
455: puzzling observations: the anomalously small effective surfaces of AXPs,
456: the origin of short duration gamma-ray bursts, the shape of some SNRs 
457: and the large peculiar velocities of old pulsars.
458: 
459: The X-ray spectra of SGRs and AXPs in quiescent periods have 
460: been interpreted as the Wein tail of black-body 
461: radiation from their surface. The Stefan-Boltzman law, ${\rm L_X=4\pi\, 
462: R^2\sigma\, T^4}$ (or ${\rm L_X \approx 1.3\times 10^{37}~erg~s^{-1}}$, 
463: for ${\rm R=10}$ km and ${\rm T=1}$ keV)
464: yields effective surface areas significantly 
465: smaller than expected for a  NS, ${\rm A_{NS}\approx 4\pi\times 10^2~km^2}$.
466: The AXP data are summarized in Table II, 
467: normalized to the measured distances  (analysis of the corresponding 
468: data for SGRs is complicated by their time variability and not well
469: determined temperatures).
470: All inferred areas are $\sim$20\% 
471: of the expectation. It would be difficult to attribute this
472: systematic effect to errors in the observations.
473: Effective areas smaller than expected 
474: may be due to non-uniform surface temperatures. But, more interestingly,
475: they can be real and reflect the small radii of SSs or QSs.  
476: 
477: Does the transition from a neutron star to a denser star
478: have directly observable signatures?
479: The answer 
480: may be guided by analogies with observed phenomena,
481: a detailed model would be very hard to develop. 
482: The gravitational binding energy release --of ${\cal O}(10^{53})$ ergs--
483: would be mainly emitted as a neutrino burst,
484: as in the Type II explosion that first begat the NS.
485: The collapse of a NS core
486: into a denser object should be accompanied by the ejection
487: of the outer layers, and be more similar to a Type I SN explosion
488: than to Type II, Ib or Ic events,  for which the ejected mass is much 
489: larger and consists mainly of light elements.
490: %A large fraction of the angular momentum of the original star may
491: %be carried away by the ejecta, so that the remnant would be left 
492: %rotating at a slower rate.
493: The ejecta should be mildly relativistic and deposit their
494: energy in the interstellar medium at a fast rate, giving rise to a  
495: short-lived
496: SNR, rich in Fe-group elements. 
497: The collapsing material may, as in active galactic nuclei,
498:  acquire an accreting toroidal structure and emit
499: highly relativistic and collimated jets. These jets, if they 
500: point in our direction, may produce 
501: %(short duration) 
502: gamma-ray bursts (Dar 1999a; Dar and Plaga 1999)
503: %(the long duration ones may be produced by highly 
504: %relativistic jets from the birth of black holes in gravitational collapse 
505: %of NSs due to mass accretion onto the proto-neutron star in supernovae 
506: %explosions, or onto NS in compact binary systems (Dar and DeR\'ujula, 
507: %under preparation)).
508: 
509: If collimated jets produce cosmological gamma-ray bursts,  
510: their kinetic energy must be ${\rm E_k\sim 10^{52}~erg}$, i.e. comparable 
511: to the kinetic energy of the SNR from the SN event in which the NS
512: was originally born. With that much energy, the jets may 
513: distort the first SNR in a  recognizable manner.  
514: Radio observations expose a vast range of SNR shapes
515: (see, e.g., Whiteoak and Green 1996). While very 
516: young SNRs have a simple expanding geometry, most older SNRs 
517: have a distorted and 
518: complicated appearance, which has been traditionally attributed 
519: to their expansion into an inhomogeneous interstellar medium. 
520: However, some SNRs have striking properties which require  
521: explanation either in terms of jets
522: (Manchester 1987; Rozyczka 1993; Gaensler 1998) or --if not due to
523: accidental superpositions-- in terms of  a second
524: explosion (Aschenbach  1998; Aschenbach et al. 1999; Gaensler 1999).
525: A second gravitational collapse, which produces 
526: a second bang and emits relativistic jets along the rotation axis may 
527: explain the puzzling morphology of many SNRs (Dar and De R\'ujula, in
528: preparation). 
529: 
530: In the collapse of a neutron star, an imbalance in the momenta of
531: oppositely ejected jets can impart a natal kick to the resulting SS or QS. 
532: This may explain the large observed velocities of ``old'', slowly-rotating
533: pulsars.  Millisecond pulsars and young pulsars have small velocities
534: (e.g., Toscano et al. 1999); their youth and large angular momentum may
535: have temporarily prevented their collapse (for millisecond pulsars this
536: may also be a selection effect: they are found in binary systems and only
537: with a small natal velocity could they remain bound and be spun up by mass
538: accretion). 
539:  
540: 
541: %The gravitational collapse to a 
542: %strange or a diquark star results in a mildly relativistic supernova-like 
543: %explosion and, possibly, in the ejection of highly relativistic jets.  
544: 
545: \section{Outlook} 
546: 
547: We have contended that SGRs and AXPs do not have the anomalously large
548: magnetic fields postulated in the magnetar model to be the cause of
549: their fast spin-down and the energy reservoir 
550: of their emitted radiation. 
551: Instead, we argued that the spin-down
552: is caused by relativistic particle emission and we conjectured that
553: the power supply is the  gravitational energy released by contraction,
554: resolving the conundrum associated with the large observed
555: jumps in spin-down rate.
556: 
557: Independently of the strength of their magnetic field, the well
558: measured SGRs and AXPs are truly puzzling: their spin-down ages
559: are much smaller than the age of the SNRs with which they are
560: associated, and the distance they must have travelled during their
561: lifetime implies an unacceptable velocity. The hypothesis that these
562: neutron stars have suffered a delayed transition to a denser type of  
563: constituency
564: resolves these problems: the measured spin-down ages  date back only to the
565: stars' ``second birth''. This hypothesis also explains why the star's  
566: surfaces,
567: as extracted from their X-ray emission, turn out to be smaller than expected
568: for a conventional NS, and why the morphology
569: of some of their associated 
570: SNRs hints at a double bang. 
571: If the second birth gives a new kick velocity to the NS, its direction  
572: should be
573: uncorrelated to the centre of the SNR, as in the Vela pulsar.
574: 
575: Most observed pulsars are not in binary systems and are not SGRs or AXPs.
576: These conventional pulsars have periods averaging to 1/2 s, significantly
577: shorter than the periods listed in Tables I and II. Their characteristic
578: spin-down ages, on the other hand, are longer, typically $10^7$ years. 
579: With such long
580: lifetimes, and if a good fraction of the rate of core-collapse supernova
581: (roughly one per century in our galaxy) results in pulsars,
582: one would expect to detect some $10^5$ of these objects, while only
583: about $10^3$ are actually observed. But, if within some $10^5$ years
584: a good fraction of these NSs --depending on their mass, rotation period
585: and magnetic field-- were to suffer a transition into a denser object, 
586: the observed numbers of supernovae, conventional pulsars, SGRs and
587: AXPs would fall into a consistent picture.
588: A star freshly reborn after a phase transition could be an SGR, whose  
589: longer period
590: is explained by rapid spin-down. In turn, SGRs could 
591: convert into AXPs after a period of bursting activity. 
592: As the AXPs cool and spin down, they should become slowly 
593: rotating, radio-quiet, X-ray-dim pulsars. Many
594: of these dim pulsars should still be present in the neighbourhood
595: of their SNRs.  Very sensitive X-ray searches are required to discover  
596:  their 
597: presence there.  
598: 
599: 
600: \begin{acknowledgements}
601: {\bf This research was supported in part by the Fund For 
602: Promotion Of Research At The Technion.} 
603: \end{acknowledgements}
604: \begin{thebibliography}{}
605: 
606: 
607: \bibitem{Alford}
608: Alford, M. et al. 1998, Phys. Lett. B422, 247 
609: \bibitem{Asaoka} 
610: Asaoka, I., \& Aschenbach, B. 1994, A\&A 284, 573   
611: \bibitem{Aschenbach} 
612: Aschenbach, B. 1998, Nature 396, 141   
613: \bibitem{Aschenbach2} 
614: Aschenbach, B., Iyudin, A. F. \& Sch\"onfelder, V. 1999, astro-ph/9909415
615: \bibitem{Atteia} 
616: Atteia, J. L. et al. 1987, ApJ. 320, L105 
617: \bibitem{Berges}
618: Berges, J. \& Rajagopal. 1999, Nucl. Phys. B538, 215 
619: \bibitem{Caraveo}
620: Caraveo, P. A. \& Mignani, R. P. 1999, A\&A 344, 367 
621: \bibitem{Cline} 
622: Cline, T. L. et al. 1982, ApJ 255, L45   
623: \bibitem{Corbet2} 
624: Corbet, R. H. D. \& Mihara, T. 1997, ApJ 475, L127 
625: \bibitem{Corbet}
626: Corbet, R. H. D. et al. 1995, ApJ 443, 786 
627: \bibitem{Dar} 
628: Dar, A. 1999a, A\&A Suppl. 138, 505  
629: \bibitem{Dar2} 
630: Dar, A. 1999b, sub. to ApJ.  astro-ph/9911473 
631: \bibitem{Dar3} 
632: Dar, A. and Plaga, R. 1999, A\&A {\bf 349}, 259 
633: \bibitem{Duncan} 
634: Duncan, R. C. \&  Thompson, C. 1992, ApJ. 392, L9  
635: \bibitem{Frail} 
636: Frail, D. A., Vasisht, G. \& Kulkarni, S. R. 1997, ApJ. 480, L129
637: \bibitem{Frail2} 
638: Frail, D. A. et al. 1999, Nature 398, 127 
639: \bibitem{Gaensler3} 
640: Gaensler, B. M. 1999, {\it Ph.D. Thesis,} Univ. of Sidney (unpublished)
641: \bibitem{Gaensler}
642: Gaensler, B. M., Gotthelf, E. V. \& Vasisht, G. 1999, ApJ 526, L37 
643: \bibitem{Gaensler 2} 
644: Gaensler, B. M., Green, A. J. \& Manchester, R. N. 1998,
645: M. N. R. A. S. 299,  812
646: \bibitem{Gotthelf2} 
647: Gotthelf, E. V. \& Vasisht, G. 1998, New Astr.  3, 293
648: \bibitem{Gotthelf}
649: Gotthelf, E. V., Vashisht, G. \& Dotani, T. 1999, ApJ 522, L49
650: \bibitem{Harding} 
651: Harding, A. K., Contopoulos, I. \& Kazanas, D. 1999, ApJ 525, L125
652: \bibitem{Hurley} 
653: Hurley, K. et al. 1999a, ApJ 519, L143 
654: \bibitem{Hurley0} 
655: Hurley, K. et al. 1999b, ApJ 523, L37 
656: \bibitem{Hurley3} 
657: Hurley, K. et al. 1999c, ApJ 510, L111 
658: \bibitem{Hurley2} 
659: Hurley, K.  et al. 1999d, ApJ. 510, L107
660: \bibitem{Israel}
661: Israel, G. L. et al. 1999, A\&A  346, 929  
662: \bibitem{Iwasawa} 
663: Iwasawa, K., Koyama, K. \& Halpern, J. P. 1992, PASJ  44, 9 
664: \bibitem{Kulkarni} 
665: Kulkarni, S. R. \& Frail, D. A. 1993, Nature 365, 33 
666: \bibitem{Kulkarni2} 
667: Kulkarni, S. R. et al. 1994, Nature 368, 129 
668: \bibitem{Kouveliotou3} 
669: Kouveliotou, C. et al. 1994, Nature 368, 125 
670: \bibitem{Kouveliotou2} 
671: Kouveliotou, C. et al. 1998, Nature 393, 235   
672: \bibitem{Kouveliotou} 
673: Kouveliotou, C. et al. 1999, ApJ 510, L115   
674: \bibitem{Kouveliotou33} 
675: Li, X. D., {\it et al.}  1999, Phys. Rev. Lett. 83, 3776
676: \bibitem{Lyne} 
677: Lyne, A. G. \& Lorimer, D. R. 1994, Nature 369, 127
678: \bibitem{Manchester2} 
679: Manchester, R. N. 1987, A\&A 171, 205  
680: \bibitem{Manchester} 
681: Manchester, R. N. \&  Taylor J. H. 1997, {\it Pulsars,} Freeman Eds., San 
682: Francisco, 1997    
683: \bibitem{Marsdena}
684: Marsden, D. et al.  1996, ApJ. 470, 513 
685: \bibitem{Marsden}
686: Marsden, D., Rothschild, R. E. \& Lingenfelter, R. E. 1999, ApJ 520, L107 
687: \bibitem{Mazets}
688: Mazets, E. P. et al. 1979, Nature, 282, 587
689: \bibitem{Mereghetti} 
690: Mereghetti, S., Belloni, T. \& Nasuti, F. P. 1997, A\&A 321, 835 
691: \bibitem{Mereghetti2} 
692: Mereghetti, S. \&  Stella, L. 1995, ApJ. 442, L17 
693: \bibitem{Murakami} 
694: Murakami, T. et al. 1994 Nature, 368, 127 
695: \bibitem{Nasuti}
696: Nasuti, F. P. et al. 1997, A\&A 323, 839  
697: \bibitem{Oosterbroek} 
698: Oosterbroek, T. et al. 1998, A\&A 334, 925  
699: \bibitem{Parmar}
700: Parmar, A. et al. 1998, A\&A, 330, 175 
701: \bibitem{Ramaty}
702: Ramaty, R. et al. 1980. Nature {\bf 287}, 122 
703: \bibitem{Rapp}
704: Rapp, R. {\it et al.} 1999, Phys. Rev. Lett. 81, 53   
705: \bibitem{Rozyczka} 
706: Rozyczka, M. et al. 1993, M.N.R.A.S. 261, 674
707: \bibitem{Shapiro} 
708: Shapiro, S. L.  \& Teukolsky, A. A. 1983, Black Holes, White Dwarfs 
709: and Neutron Stars, {\it John Wiley \& Sons Inc. (1983)},
710: \bibitem{Sonobe}
711: Sonobe, T. et al. 1994, ApJ. 436, L23 
712: \bibitem{Stella} 
713: Stella, L., Israel, G. L. \& Mereghetti, S. 1998, Adv. Sp. Res. 22 1025
714: \bibitem{Sugizaki} 
715: Sugizaki, M. et al. 1997,  PASJ, 49, L25 
716: \bibitem{Thompson}
717: Thompson, C. \& Duncan, R. C.  1995, M.N.R.A.S. 275, 255
718: \bibitem{Thompson2}
719: Thompson, C. \& Duncan, R. C.  1996, ApJ 473, 322
720: \bibitem{Torii}
721: Torii, K. et al. 1998, ApJ. 503, 843  
722: \bibitem{Toscano} 
723: Toscano, M. et al. 1999, M. N. R. A. S. 307, 925 
724: \bibitem{Vasisht4} 
725: Vasisht, G., Frail, D. A. \& Kulkarni, S. R. 1995, ApJ. 440, L65
726: \bibitem{Vasisht3} 
727: Vasisht, G \& Gotthelf, E. V. 1997, ApJ 486, L129  
728: \bibitem{Vasisht} 
729: Vasisht, G. et al. 1994, ApJ  431, L35 
730: \bibitem{Vasisht2} 
731: Vasisht, G. et al. 1997, ApJ 476, L43  
732: \bibitem{White} 
733: White, N. E. et al. 1996, ApJ 463, L83 
734: \bibitem{Whiteoak}
735: Whiteoak, J. B. Z. \& Green, A. J. 1996, A\&A Suppl. 118, 329 
736: \bibitem{Woods} 
737: Woods, P. M. et al. 1999a, ApJ 524, L55 
738: \bibitem{Woods2} 
739: Woods, P. M. et al. 1999b, ApJ  519, L139 
740: \bibitem{wilczek}
741: Wilczek, F. 1998, Nature 395, 220  
742: 
743: \end{thebibliography}{}
744: %\end{document}
745: 
746: 
747: \newpage
748: \footnotesize
749: \begin{deluxetable}{lcccccccc}
750: \tablefontsize{\tiny}
751: \tablecaption{Soft Gamma Ray Repeaters [SGRs]\vfill}
752: \label{tbl-1}
753: \tablehead{
754: \colhead{Pulsar} & \colhead{SNR} & \colhead{P} & \colhead{${\rm \dot P}$} &
755: \colhead{${\rm B_p}$} &
756: \colhead{${\rm \tau_s}$} &
757: \colhead{${\rm \tau_{_{SNR}}}$} &
758: \colhead{${\rm S_\perp}$}&\colhead{${\rm v_\perp}$}  \nl
759: \colhead{} & \colhead{} & \colhead{(s)}
760: & \colhead{}
761: & \colhead{(Gauss)}
762: & \colhead{(ky)} 
763: & \colhead{(ky)} 
764: & \colhead{(pc)}
765: & \colhead{(${\rm km~s^{-1}}$)} }
766: \startdata
767: SGR 1806-20$^b$ & G10.0-0.1  & 7.47  &$8.3\times 10^{-11}$
768: & $1.6\times 10^{15}$  & 1.4 & $>5$ 
769: & $8.3\,{\rm d_{15}}$ & $5500\,{\rm d_{15}}$ \nl
770: SGR 1900+14$^c$  & G42.8+0.6  & 5.16  &$6.1\times 10^{-11}$
771: & $1.1\times 10^{15}$  & 1.3 
772: &$>10$ & $36\,{\rm d_{7}}$ & $ 27000\,{\rm d_{7}}$ \nl
773: SGR 1627-41$^d$& G337.0-0.1  & 6.41 &\dots & \dots & \dots & $>5$
774: &$5.4\,{\rm d_{11}}$ &$5400\,{\rm d_{11}/\tau_3}$ \nl
775: SGR 0525-66$^e$ & N49    & $\sim 8$ &\dots & \dots 
776: &\dots & $>5$ &$>22$  & $>22000/\tau_3$\nl
777: \enddata
778: \tablenotetext{a}
779: {${\rm B_p}$ values are for sin$\alpha=1.$ Distances 
780: ${\rm d_x}$ are in units of x kpc. $\tau_3$ is
781: the age of the SGR in ky. \\
782: $^b$ Atteia et al. 1987; Kulkarni \& Frail 1993; Kouveliotou 
783: et al. Ref. 1994; Kulkarni et al. 1994; Murakami et al. 
784: 1994; Sonobe et al. 1994; Kouveliotou et al. 1998 and references therein.\\
785: $^c$ Kouveliotou et al. 1994; Vasisht et al 1994; Kouveliotou et al. 1999; 
786: Woods et al. 1999a; Hurley et al. 1999d and 
787: references therein.\\ 
788: $^d$ Woods et al. 1999b; Hurley et al. 1999a  and 
789: references therein.\\ 
790: $^e$ Mazets et al. 1979; Cline et al. 1982; Marsden et al. 1996 and 
791: references therein.} \end{deluxetable}
792: 
793: \begin{deluxetable}{lcccccccc}
794: \tablefontsize{\tiny}
795: \tablecaption{Anomalous X-Ray Pulsars [AXPs] \vfill}
796: \label{tbl-2}
797: \tablehead{
798: \colhead{Pulsar}& \colhead{P} &
799: \colhead{${\rm \dot P}$} & 
800: \colhead{${\rm B_p}$} &
801: \colhead{kT} & 
802: \colhead{${\rm L_X}$}&\colhead{${\rm A_s}$}  \nl
803: \colhead{}& \colhead{(s)}
804: & \colhead{}
805: & \colhead{(Gauss)}
806: & \colhead{(keV)} 
807: & \colhead{(${\rm 10^{35}~erg~s^{-1}}$)}
808: & \colhead{(${\rm 4\pi\times 10^2 km^2}$)} }
809: \startdata
810: 1E 1841$-$045$^b$ & 11.76 & $4.1\times 10^{-11}$
811: & $1.4\times 10^{15}$  
812: &0.55& $3\,{\rm d_{7}}^2$ & $0.25\,{\rm d_{7}^2}$  \nl
813: 1E 2259$+$586$^c$ & 6.98  & $\sim 5 \times 10^{-13}$
814: & $1.2\times 10^{14}$  &0.41
815: & $0.8 \,{\rm d_{4}^2}$ & $ 0.22\,{\rm d_{4}^2}$   \nl
816: 4U 0142$+$615$^d$ & 8.69  & $\sim 2\times 10^{-12}$ 
817: & $2.6\times 10^{14}$  & 0.39
818: & $0.7\,{\rm d_{4}^2}$ & $ 0.24\,{\rm d_{4}^2}$ \nl
819: 1E 1048$-$5937$^e$ & 6.44  & $1.5-4\times 10^{-11}$ 
820: & $>6.3\times 10^{14}$  & 0.64 
821: & $5 \,{\rm d_{10}^2}$ & $ 0.23\,{\rm d_{10 }^2}$ & \nl
822: RX J170849$-$4009$^f$ & 11.00 & $2\times 10^{-11}$
823: & $9.5\times 10^{14}$  & 0.40
824: & $10 \,{\rm d_{10}^2}$ & $0.17\,{\rm d_{10}^2}$ \nl
825: PSR J1844-0258$^g$ & 6.97  &    \dots 
826: & \dots  &0.64
827: & $3 \,{\rm d_{15}^2}$ & $ 0.15 \,{\rm d_{15}^2}$ \nl
828: \enddata
829: \tablenotetext{a}
830: {${\rm B_p}$ values are for sin$\alpha=1.$ 
831: All X-ray luminosities  are in the $\sim 1-10$ keV energy band
832: as corrected for absorption by Gotthelf and Vasisht 1998.
833: The distances ${\rm d_x}$ are in units of x kpc. \\
834: $^b$ Vasisht \& Gotthelf 1997\\ 
835: $^c$ Iwasawa  et al. 1992; Corbet et al. 1995; Parmar et al. 1998 
836: and references therein.\\
837: $^d$ Mereghetti \& Stella 1995; White  et al. 1996; 
838: Israel et al 1999 and references therein.\\
839: $^e$ Oosterbroek et al. 1998; Mereghetti et al. 1997; 
840: Corbet \& Mihara 1997 and references therein.\\
841: $^f$ Sugizaki et al. 1997\\
842: $^g$ Torii et al. 1998; Gaensler et al. 1999 and references therein.\\}
843: \end{deluxetable}
844: 
845: 
846:  
847: \newpage 
848: \begin{figure} 
849: \begin{center} 
850: %\vspace*{-1.6cm} 
851: %\hspace*{-1cm} 
852: \epsfig{file=sgr10.eps,width=14cm} 
853: \caption{The period of SGR 
854: 1900+14 as a function of time (Woods et al. 1999a), as measured by RXTE  
855: (squares), BeppoSAX (triangles), ASCA (circles) and BSA (crosses). The 
856: lines are linear fits to the X-ray periods of the SGR before June 
857: 9, 1998 (${\rm \dot P=6.1\times 10^{-11}}$), between June 9 - August 27, 
858: 1998 (${\rm \dot P=1.3\times 10^{-10}}$), and after August 27, 1998 (${\rm 
859: \dot P=6.1\times 10^{-11}}$).  Between June 9-August 28, the ``averaged''
860: spin-down rate has changed by a factor  
861: $\sim 2.2$ as a result of a continuous or sudden braking.}  
862: \vspace*{-0.5cm} 
863: \end{center} 
864: \end{figure} 
865: \end{document} 
866:   
867: 
868: