Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://xray.sai.msu.ru/~polar/html/publications/baranov/poster.ps Äàòà èçìåíåíèÿ: Tue Sep 14 14:16:50 1999 Äàòà èíäåêñèðîâàíèÿ: Sat Dec 22 04:55:58 2007 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: vela satellite

Nature of the compact
X­ray source in
supernova remnant
RCW103
and related problems
Sergei B.Popov
Sternberg Astronomical Institute, 119899,
Universitetskii pr.13, Moscow, Russia
e­mail: polar@xray.sai.msu.su
URL: http://xray.sai.msu.ru/~ polar/
Abstract
In this poster I discuss the nature of the compact X­ray
source in the center of the supernova remnant RCW 103.
Several models, based on the accretion onto a compact
object such as a neutron star or a black hole (isolated
or binary), are analyzed. I show that it is more likely
that the central X­ray source is an accreting neutron star
than an accreting black hole. I also argue that models of
a disrupted binary system consisting of an old accreting
neutron star and a new one observed as a 69­ms X­ray
and radio pulsar are most favored. Related problems of
neutron stars astrophysics are briefly discussed.
1

1 Introduction
Among all astrophysical objects neutron stars (NSs) and black holes (BHs)
attract most attention of physicists. NSs were initially predicted by Lev Landau
immediately after neutron was discovered in 1932. But as far as there were no
ideas how to observe so small (about 10 km in diameter) astronomical objects
they were not a subject of intense search, and were found as radio pulsars
occasionally in 1967 in the UK.
Now we know more than 1000 NSs as radio pulsars and more than 100 NSs
emmiting X­ or/and fl­rays (mostly X­ray sources are found in close binary sys­
tems, but several isolated ones are also observed). But the Galactic population
of these objects is much larger: about 10 8 -- 10 9 (the first number comes from
radio pulsar statistics, and the second one -- from chemical evolution of the
Galaxy). So, we observe only a tiny fraction of one of the most fascinating as­
trophysical objects. Studying of the rest of population is crucially important to
obtain initial parameters of NSs, and to understand their evolution (for exam­
ple, there are some argument for the proposition, that not all NSs pass through
the stage of radio pulsar in their young years; it can happen due to too long
initial periods, too high (magnetars) or too low magnetic fields).
It is generally accepted that most of NSs and BHs are the products of super­
nova (SN) explosions (although there is also a possibility of a ``quiet collapse'').
In most cases a supernova remnant (SNR) appears after a formidable explosion
of a massive star (with M ? 10 \Gamma 35M fi ). Although sometimes a young NS is
observed inside a SNR as a radio pulsar (e.g., Crab, Vela, etc.) or as a X­ray
source, in most cases no compact object is found inside a SNR, or an accidental
coincidence of the radio pulsar and the SNR is very likely (e.g., Kaspi 1996,
1998; Frail 1997).
It is possible, that about 50% of NSs are born with low magnetic fields,
so they never appear as radio pulsars and spend most of their lives on the
Ejector (but not radio pulsar!, they are below the death­line!) and Propeller
stages. These NSs with low magnetic fields can not spin­down significantly even
during the Hubble time, and so they can never come to the Accretor stage
and can't be observed as accretion­powered X­ray sources in correspondence
with the observations made by the ROSAT (Haberl et al. 1996; Walter et al.
1996; Treves & Colpi 1991), which showed that only a few old isolated accreting
NSs are observed. By the way, the ROSAT results also can be explained by
high average velocity of NSs which they obtain due to assymetry of the SN
explosion (the work on this topic is in progress now). For high­velocity NSs
the characteristic Ejector period is higher, and NSs spend most of their lives as
Ejectors. So radio pulsars, probably, are not the best representatives of the NS
population, and very old, mostly undetected at the present time, NSs can have
significantly different properties (probably even different initial properties: for
example longer periods or, most probably, lower magnetic fields).
Recently, Gotthelf et al. (1997) described a compact X­ray source in the
2

center of SNR RCW 103 with the X­ray luminosity L x ¸ 10 34 erg=s (for the
distance 3:3 kpc) and the black­body temperature about 0:6 keV . The source
flux has varied since previous observations (Petre et al. 1999). The nature of
the central compact source is unclear. No radio or optical compact counterpart
was observed. Also a 69­ms X­ray and radio pulsar with a characteristic age
about 8 kyr was discovered 7' from the center of the remnant (Kaspi 1998, Kaspi
et al. 1999), but the reality of the association of the pulsar with the SNR is
unclear (Dickel 1999). This result makes the situation around RCW 103 more
complicated and interesting.
In this poster (see also Popov 1998) I discuss possible models of that compact
central source and its possible connections with the 69­ms pulsar (the discussion
partly coincides with my previous note (Popov 1997)).
2 What is inside the RCW 103?
Gotthelf et al. (1997) discussed why the source cannot be a cooling NS, a plerion,
or a binary with a normal companion. The reader is referred to their paper for
the details. In the present analysis I assume that the X­ray luminosity of the
source is produced due to accretion of the surrounding material onto a compact
object (a NS or a BH). I analyze thus only models with compact objects, isolated
or with a compact companion (most probably the binary system was destroyed
after the second explosion, when the 69­ms pulsar was formed). Massive normal
companions are excluded by optical observations. If the companion is a low­
mass star, it is difficult to explain the X­ray luminosity as high as observed in
the RCW 103 because in low­mass systems accretion usually occurs after the
Roche lobe is overflowed with higher luminosities.
The main challenge for the models of accretion of the surrounding material
onto isolated compact object is to answer the question of where a NS or a
BH finds enough matter to accrete. I don't discuss it here, assuming that the
material is available in the surrounding medium (see, for example, Page et al.
1999).
3

10 5
10 7
10 9
10 11
10 13
.001 .01 .1 1 10 100
2
1
P, s
B,
G
Figure 1: Possible values of the magnetic field, B, and period, p, for the accreting
NS. The dotted line (1) corresponds to the equilibrium period, P eq , see eq.(4),
while solid line (2) corresponds to the accretor period, PA , see eq.(2).
4

2.1 Accreting isolated young black hole or accreting old
black hole in pair with a young compact object
An isolated BH accreting the interstellar medium can be, in principle, observed
by X­ray satellites such as ROSAT , ASCA etc (Heckler & Kolb 1996). To
achieve high X­ray luminosity, a compact object must move with a low velocity
relative the ISM:
—
M = 2ú
`
(GM ) 2 ae
(V 2
s + V 2 ) 3=2
'
; (1)
where V s is the speed of sound, V is the velocity of the compact object with
respect to the ambient medium, M -- the mass of the accreting star and ae is the
density of the accreting material. One can introduce the effective velocity, V eff ,
and rewrite eq. (1) as follows:
—
M = 2ú((GM ) 2 ae)=(V 3
eff ):
During the SN explosions a compact object can obtain an additional kick
velocity. At the present time the distribution of the kick velocity is not known
well enough (e.g., Lipunov et al. 1996). Although observations of radio pulsars
favor high kick velocities about 300 \Gamma 500 km=s (Lyne & Lorimer 1994), alter­
native scenarios in which the velocity of NSs significantly increases after the SN
explodes are also possible (Kaspi 1996; Frail 1997). We mark here, that if the
69­ms pulsar is a new born NS, and the central source is the older object, it
is not surprising, that the 69­ms pulsar is farther from the center of the SNR.
Because the new born NS received a high kick velocity (the required transverse
velocity is about 800 km/s (Kaspi 1998)), and the old one only saved its orbital
velocity, because the system survived in the first explosion. X­ray radiation of
the new born NS of course doesn't have the accretion nature.
To explain the observed X­ray luminosity of the compact object in the center
of RCW 103 the accretion rate, —
M , should be about 10 14 g=s. For all models
that consider accretion onto an isolated compact object, the density required to
obtain L x ¸ 10 34 erg=s is as high as 10 \Gamma22 g=cm 3 .
One can then estimate the size of the emmiting region, using observed lu­
minosity and temperature: L = 4ú \Delta R 2
emm oeT 4
For observed values of L x and T this equation gives Remm ¸ 1 km. For
BHs such a low value of Remm is very unlikely because the gravitational radius
is about RG ¸ 3 km (M=M fi ), and most of the present BH­candidates have
masses about 7 \Gamma 10M fi . This is probably the main argument against isolated
accreting BH as a model for the RCW 103. Also the efficiency of spherically
symmetric accretion onto a BH is very low resulting in a significantly higher
density required to achieve the same luminosity.
The same arguments can be used against models with a binary system (prob­
ably disrupted): BH+NS (NS was born in the recent SN explosion -- a 69­ms
pulsar).
5

2.2 Young isolated accreting neutron star
In the past few years isolated accreting NSs have become a subject of great
interest especially due to the observations with the ROSAT satellite (Treves
& Colpi, 1991; Walter et al. 1996; Haberl et al. 1996). In this subsection I
will present arguments that the compact X­ray source in RCW 103 can be an
isolated accreting NS and will estimate some properties of that NS.
There are four main possible stages for a NS in a low­density plasma: 1):
Ejector (a radio pulsar is an example of Ejector); 2): Propeller; 3): Accretor;
and 4): Georotator (Lipunov & Popov 1995; Konenkov & Popov 1997; Popov
& Konenkov 1998). The stage is determined by the accretion rate, —
M , the
magnetic field of the NS, B, and by the spin period of the NS, p.
If the NS is on the Accretor stage, then its period is longer than the so­called
Accretor period, PA :
PA = 2 5=14 ú (GM ) \Gamma5=7 (¯ 2 = —
M ) 3=7 sec; (2)
where ¯ = B \Delta R 3
NS is magnetic moment of the NS.
For the RCW 103 I use the following values: —
M = 10 14 g=s, M = 1:4 M fi ,
RNS = 10 6 cm which give:
B ¸ 10 10 \Delta p 7=6 G: (3)
If material is accreted from the turbulent interstellar medium, a new equi­
librium period can occur (Konenkov & Popov 1997; Popov & Konenkov 1998):
P eq ¸ 30 B 2=3
12
I 1=3
45
—
M \Gamma2=3
14
R 2
NS 6
V 7=3
eff 6
V \Gamma2=3
t 6
M \Gamma4=3
1:4
sec; (4)
where V t is the turbulent velocity (all velocities are in units of 10 km=s); M 1:4
is the mass of the NS in units of 1:4 M fi , B 12 is the magnetic field of the NS in
unites 10 12 G and RNS is the radius of the NS in units of 10 6 cm.
We then obtain:
B ¸ 8 \Delta 10 9 \Delta p 3=2 G: (5)
It is obvious that to explain the luminosity of the RCW 103 by an isolated
accreting NS, one must assume that the NS was born with extremely low mag­
netic field (see the remark above) or with unusually long spin period. The age
of the SNR RCW 103 is about 1000 years (Gotthelf et al. 1997), which means
that the magnetic field could not decay significantly (Konenkov & Popov 1997;
Popov & Konenkov 1998). The flux of the source is not constant (Petre et al.
1999), so the idea of cooling NS can be rejected. Thus, the model with isolated
young accreting NS is not a likely explanation for the data.
6

2.3 Accreting old neutron star in pair with a young neu­
tron star (or in the disrupted system)
Binary compact objects are natural products of binary evolution (Lipunov et
al. 1996). One can, therefore, discuss these scenarios as a viable alternative.
In the previous subsection I showed that accretion onto a young isolated NS
requires unusual initial parameters. However, there is a chance that we observe
a binary system (or a disrupted binary), where one component is an old NS
and the other component was formed in a recent SN explosion and appears as
a 69­ms pulsar.
In that case, the parameters determined by eqs.(3), (5) are not unusual:
old NS can have low magnetic fields and long periods (Lipunov & Popov 1995;
Konenkov & Popov 1997; Popov & Konenkov 1998). Due to the fact that
Gotthelf et al. (1997) did not find any periodic change of the luminosity, one
can argue that the field is too low to produce the observable modulation (the
accreting material is not channeled to the polar caps: B ! 10 6 G) or that the
period is very long (p ? 10 4 sec), which is possible for old NSs with ``normal''
magnetic fields (Lipunov & Popov 1995): P ú 500sec: The last opportunity is,
probably, better, as the emmiting area is not large ú 1km 2 .
The evolutionary scenario for such a system is clear enough (Lipunov et al.
1996). One can easily calculate it using the ``Scenario Machine'' WWW­facility
(http://xray.sai.msu.ru/sciwork/scenario.html; Nazin et al. 1996). For
example, two stars with masses 15 M fi and 14 M fi on the main sequence with
the initial separation 200 R fi , R fi -- the solar radius, after 14 Myr (with two
SN explosions with low kick velocities: about 60 km=s) end their evolution as a
binary system NS+NS. The second NS is ¸ 1 Myr younger. During ¸ 1 Myr the
magnetic field can decrease up to 1/100 of the initial value with a significant
spin­down (Konenkov & Popov 1997; Popov & Konenkov 1998). The binary
NS+NS is relatively wide: 20 R fi with an orbital period 5:8 d , so the orbital
velocity is not high (the orbital velocity of the accreting NS should be added to
V eff ).
The 69­ms X­ray pulsing source and it's radio pulsar counterpart that were
discovered near RCW 103 (Kaspi 1998, Kaspi et al. 1999) can be a new­born
radio pulsar. So, it means that the binary system was disrupted after the
second explosion. It means that in the first explosion the kick velocity was
small (about 50 km/s in the opposite case the system could be disrupted after
the first explosion and the older NS could leave the SNR before the second
explosion, but if the orbit was significantly eccentric, the kick velocity in the
first explosion could be high too) and in the second explosion it was as high as
750­800 km/s for the same initial parameters as in the previous example.
Of course other variants of the initial parameters are possible, and I showed
this one just as a simple example.
7

3 Conclusions
To conclude, I argued that the most likely model for the central compact X­ray
source of RCW 103 is that of an accreting old NS in a disrupted binary system
with a young compact object (the 69­ms pulsar) born in the recent SN explosion
that produced the observed supernova remnant (some ideas about a disrupted
binary in RCW 103 were discussed also in the article (Torii et al. 1998)). Such
systems are rare, but natural products of the binary evolution. Scenarios with
a single compact objects or with accreting BH are less probable.
Acknowledgments
I thank M.E. Prokhorov and V.M.Lipunov for helpful discussions, E.V. Got­
thelf and K. Torii for the information about RCW 103, A.V. Kravtsov for his
comments on the text and the ``Scenario Machine'' group for the WWW­version
of the program. The work was supported by the RFBR (98­02­16801), the IN­
TAS (96­0315) and NTP ''Astronomy'' (1.4.4.1) grants.
References
[1] Frail, D.A., 1997, in NATO Advanced Study Institute: ``The Many Faces of
Neutron Stars'', eds. A. Alpar, R. Buccheri, and J. van Paradijs
[2] Haberl F., Pietsch, W., Motch, C., Buckley, D.A.H., Circ. IAU, 1996, no.
6445.
[3] Heckler, A.F. & Kolb, E.W., 1996, ApJ 472, L85
[4] Gotthelf, E. V., Petre, R. & Hwang, U., 1997, ApJ, 487, L175 (astro­
ph/9707035)
[5] Kaspi, V.M., 1996, in IAU Colloquium 160 ''Pulsars: Problems and
Progress'' conference proceedings, editors S. Johnston, M. Bailes & M.
Walker, ASP Conference Series Vol. 105, p. 375
[6] Kaspi, V.M., 1998, astro­ph/9803026
[7] Konenkov D.Yu., & Popov, S.B., 1997, PAZh, 23, 569 (astro­ph/9707318)
[8] Page, D., Geppert, U, & Zannias, T, 1999, proceedings of the Workshop
on the relation between neutron stars and supernova remnants, Eds.: R.
Bandiera, E. Masini, F. Pacini, M. Salvati, L. Woltier
[9] Petre, R., Gotthelf, E.V., & Hwang, U., 1999, Ibid.
[10] Dickel, J.R., 1999, Ibid.
8

[11] Kaspi, V.M. et al., 1999, Ibid.
[12] Popov, S.B., & Konenkov D.Yu., 1998, Radiofizika, 41, 28 (astro­
ph/9812482)
[13] Popov, S.B., 1997, Astron. Astroph. Trans., 17, 35 (astro­ph/9708044)
[14] Popov, S.B., 1998, astro­ph/9806354
[15] Nazin, S.N., Lipunov, V.M., Panchenko, I.E., Postnov, K.A., Prokhorov,
M.E. & Popov, S.B., 1996, Grav. & Cosmol., 4, p.150 (astro­ph/9605184)
[16] Lipunov, V.M. & Popov, S.B., 1995, AZh, 71, 711 (see the English trans­
lation in Astronomy Reports, 1995, 39, 632 and also astro­ph/9609185)
[17] Lipunov, V.M., Postnov,K.A. & Prokhorov, M.E., 1996, Astroph. and
Space Phys. Rev. 9, part 4
[18] Lyne, A.G. & Lorimer, D.R., 1994, Nat 369, 127
[19] Torii, K. et al., ApJ 494, L207, 1998
[20] Treves, A. & Colpi, M., 1991, A & A, 241, 107
[21] Walter, F.M., Wolk, S.J., & Neuhauser, R., 1996, Nat , 379, 233
9