Документ взят из кэша поисковой машины. Адрес оригинального документа : http://xray.sai.msu.ru/~polar/html/publications/icga/icga.ps Дата изменения: Wed Dec 19 16:29:39 2001 Дата индексирования: Sat Dec 22 04:41:22 2007 Кодировка: Windows-1251

Young close isolated compact objects 1
YOUNG CLOSE ISOLATED COMPACT OBJECTS
S.B. Popov 1 and M.E. Prokhorov 2
Sternberg Astronomical Institute, Universitetski pr. 13, 119899 Moscow.
M. Colpi 3
Universit
a degli Studi di Milano Bicocca, Piazza della Scienza 3, I-20126 Milano.
A. Treves 4
Universit
a degli Studi dell'Insubria, via Vallegio 11, 22100 Como.
R. Turolla 5
Universit
a degli Studi di Padova, via Marzolo 8, 35131 Padova.
We suggest that the seven radio-quiet isolated neutron stars observed with ROSAT are young cooling objects
associated to recent near-by supernova explosions which formed runaway stars and the Local Bubble, affecting the
topology of the interstellar medium in the vicinity of the Sun (within a few hundred parsecs). In the aftermath of
these explosions, a few black holes might have been formed, according to the local initial mass function. We thus
discuss the possibility of determining approximate positions of close-by isolated black holes using data on runaway
stars and simple calculations of binary evolution and disruption.
Молодые близкие одиночные компактные объекты
С.Б.Попов, М.Е.Прохоров, М.Колпи, А.Тревес, Р.Туролла
В этой статье мы обсуждаем различные наблюдательные проявления близких компактных объектов (ней-
тронных звезд и черных дыр), с учетом событий, происходящий при взрывах сверхновых. Мы предполагаем,
что семь радиотихих одиночных нейтронных звезд, обнаруженных с борта ИСЗ ROSAT, являются молодыми
остывающими объектами, которые связаны с недавними вспышками сверхновых в результате которых обра-
зовались убегающие звезды, сформировалась местная каверна (Local Bubble) и другие близкие структуры в
межзвездной среде. Мы обсуждаем возможность определения положения близких одиночных черных дыр,
используя данные о движении убегающих звезд и о механике распада двойных систем.
X-ray observations of neutron stars and black holes
of stellar origin have been probing, over the entire
Galaxy, properties of their interior and of their envi-
ronment. But the recent discovery of a number of iso-
lated neutron stars in the outskirts of the Sun provides
a new tool to study how they evolve and from where
they originate.
In this paper at first we discuss the properties of the
seven ROSAT radio-quiet isolated neutron star (INS)
candidates showing that the cooling hypothesis for their
X-ray emission is the most viable (see [19] and [36] for
recent reviews and [40] for data on the latest candidate).
Then we try to connect these seven young INSs with
different objects in the solar vicinity. Finally we dis-
cuss a possibility to observe close-by isolated black holes
(IBHs).
1 e-mail: polar@sai.msu.ru
2 e-mail: mike@sai.msu.ru
3 e-mail: Monica.Colpi@mib.infn.it
4 e-mail: treves@mi.infn.it
5 e-mail: turolla@pd.infn.it
1. Neutron stars
It is now widely believed that young INS can appear, in
the Galactic disk, as sources of different nature: radio
pulsars, soft-gamma repeaters (SGRs), anomalous X-ray
pulsars (AXPs), radio-quiet compact X-ray sources in
supernova (SNa) remnants.
Here we will focus on seven ROSAT INSs, for which
it was suggested that they are close-by sources showing
no sign of pulsar activity.
1.1. Accretion versus cooling
In the early 70's it was suggested by Shvartsman [29],
[30] and Ostriker et al. [21] that INSs and IBHs can be
observable due to accretion of the interstellar medium
(ISM). INSs can also be observed when they are young
and incandescent due to thermal emission of their hot
surface mainly in soft X-rays.
Soon after the discovery of the first of the seven
sources [38], it was clear that two different mechanisms
could equally explain the nature of the emission: ac-

2 Popov et al
cretion onto the (old) INS [35] or cooling of the young
INS (see [19], [20] for a recent discussion). Early in the
study of the detectability of INSs, it seemed plausible
that accretors by far outnumber coolers, since the stage
during which a young cooling INS is hot enough to be
observed in soft X-rays is very short, about 1 Myr or
less, against few Gyr lifetime of a typical accreting NS.
However, later it was recognized that accreting INSs
might be rare objects and this holds if NSs acquire at
birth very high peculiar velocities  200 km s 1 (as ob-
served in the pulsar sample [5], [17]). This would imply
an exceedingly low accretion rate, and in turn a very
low intrinsic luminosity, undetectable with present ca-
pabilities. Population synthesis studies [24] have show
that with the inclusion of a velocity distribution func-
tion that accounts for the observed high kick velocities,
most of the INSs are now in the stage of ejector (see
[15] for detailed explanation of NS evolutionary stages
or [4] for a brief description). It is worth noting that
if strongly magnetized INSs with very high spatial ve-
locities are situated in low density ISM regions then it
is possible that most of them spend significant part of
their lives as georotators, because the light cylinder ra-
dius, R l = c=! , is larger than the radius of gravitational
capture, RG = 2GM=v 2 . Such INSs leave the ejector
stage earlier than those that live in a normal ISM [34].
The periods P of the observed objects (for four of the
ROSAT INSs) have been found to fall in a rather narrow
interval of 5-23 seconds, and they could be explained in
both hypothesis. In the accretion scenario one has to
allow for magnetic field decay [14], [39], because at a
constant field B  10 12 G, the periods of an old accre-
tor is much longer [16], [26]. Under the cooling hypoth-
esis there are no restrictions on the period P , though
its value can provide clues on the magneto-rotational
evolution of the objects. Indeed, in the case of young
cooling objects periods as long as such observed can be
explained if the INSs are magnetars [6], i.e. has very
strong magnetic field. The large braking implied by a
high field would lead to an "accelerated"slow-down of
the pulsar. But to explain the clustering observed, the
magnetic field later has to decay (as customarily advo-
cated in magnetars for explaining the high quiescent lu-
minosity of SGRs and AXPs [33]). This would lead to a
saturation of P occurring when the decay is advanced
enough to almost freeze the action of braking torques
[3]. If ROSAT INSs are magnetars or are their close rel-
atives (or descendant) than the fraction of magnetars
among all NSs can be higher than previously estimated.
Is the cooling hypothesis, which seem preferable on the-
oretical ground, testable from observation? Clues would
come from any direct determination of the peculiar ve-
locity for the seven sources. Accretion can be certainly
excluded when the velocity exceeds 40-60 km s 1 under
the most favorable conditions in the ISM.
With the determination of the proper motion of one
of the seven sources, RX J1856-3754, (that implies a ve-
locity at birth of 200 km s 1 [37]) it is now rather clear
that the cooling hypothesis holds for this sources, and it
is tempting to consider the possibility that the emission
of all the seven ROSAT sources have the same nature.
This calls again for a statistical study. To this purpose,
in [25] we compared both hypothesis (accretors and
coolers) in an attempt to explain the observed Log N 
Log S distribution for the ROSAT INSs. We obtained
that at low fluxes (< 10 13 erg cm 2 s 1 ) accretors
outnumber coolers. But at brighter fluxes, where the
seven ROSAT INSs are situated, coolers can be more
abundant. Interestingly, we found that to explain these
seven objects in terms of cooling INSs it is necessary
to assume that the spatial density of NSs is about half
an order of magnitude higher than what inferred from
radio pulsars statistics. Are we then living in an over-
dense region or is it a global feature? If global, we have
to assume that most young NSs do not pass through the
stage of a radio pulsar, or that this stage is shorter than
we have estimated. Such hypothesis finds support in re-
cent observations [12], but the fraction of radio-quiet
INSs in the whole galactic population of NSs is a number
quite uncertain. Most probably it is not as high as it is
necessary to interpret the "magnificent seven"ROSAT
INSs. So it is more probable that such an over-density
is a local phenomenon, both in space (< 1 kpc) and in
time (< 100 Myr), and we next we will explore in more
details the neighborhoods of the Sun.
1.2. The ROSAT sources and the Gould Belt
In connection with young INSs and IBHs we are mostly
interested in massive stars and recent star formation
activity close to the Sun (in our future calculations of
coolers we plan to include the realistic distribution of
young stellar complexes around the Sun). In that sense
the solar vicinity is dominated by the Gould Belt and re-
lated OB-associations (see [23] for a detailed description
of the Gould Belt).
The Gould Belt is a disk-like structure. Its inclina-
tion to the Galactic plane is about 18 Ж ; the diameter
is about 750-1000 pc, and its center is located about
150-250 pc from the Sun.
The age of the Gould Belt is estimated to be about
30-70 Myr. This implies that the most massive stars are
now about 7-10 M and that recently there was a period
of frequent explosions of massive stars, producing NSs
and BHs, with nearly constant rate [18]. The SNa rate
in the Gould Belt is about 20-30 per Myr [13].
We see direct consequences of these explosions in
the form of runaway stars. In 700 pc around the Sun
56 runaway stars are known [10]. Only few of them can
result from star-star interactions. Others are products
of SNa explosions in binary systems. For some of them,
the corresponding compact objects have been identified
or suggested [11], [13], [37].
Clearly the seven ROSAT INSs can be the out-
come of recent SNae in the Gould Belt and related

Young close isolated compact objects 3
OB-associations. In that case we can easily explain the
increased rate of NS formation around the Sun, and it
is not necessary to assume very a high fraction of radio-
silent young INSs overall in the Galaxy. Also one can
predict the existence of more radio-quiet INSs within
1 kpc around the Sun, which are remnants of > 50
recent SNae. For objects with age  1 Myr, the corre-
sponding X-ray dim sources can be discovered (see for
example objects HIP 22061, 29678 in table 5 in [10]), but
as long as the NSs receive large kick velocities at birth
it is very difficult to predict their present positions. For
BHs the situation can be opposite as their kicks could
be much lower.
All these explosions should leave their imprints on
the structure of the local ISM. Indeed, several local cav-
ities are observed. The most well known is the Local
Bubble [28]. It was suggested [31], [18] that the Local
Bubble is a result of 3-6 (or even more) recent SNa ex-
plosions. We argue that at least some of the "magnifi-
cent seven"sources are remnants of these recent explo-
sions.
2. Black holes and the Gould Belt
SNa explosions produce not only NSs, but also BHs.
Having dozens of SNae in the solar vicinity during the
last 10 Myr we can expect several BHs to have formed
during the same period in the solar neighborhood.
Usually it is accepted that BHs are one order of mag-
nitude less abundant than NSs. This estimate comes
from the critical mass for BH formation. If this mass
is about 35 M then the fraction of BHs is about 10%
(see discussion on BH fraction in [7], [8]). So we can
expect about 5 BHs correlated to 56 runaway stars. If
there are 20-30 SNae per Myr in the Gould Belt, than we
can expect  6-12 IBHs younger than few Myr. Kick
velocity for BHs is unknown, but it is reasonable to as-
sume, that it is much lower than for NSs. If it is so, all
these BHs still should be around us.
IBHs are not expected to be bright objects. Close-
by IBH can be observed due to accretion from the ISM
([30], [9]), or due to a micro-lensing effect ([1], see also
[22]). That is why it is important to know their posi-
tions on the sky. Close massive runaway stars give us a
chance to calculate an approximate positions of close-by
young IBHs. Among runaway stars we can distinguish
the most massive:  Cep,  Pup, HIP 38518 and  Per
[10]. Their masses are larger than  33 M . It means,
that the companion (actually the primary in the original
binary) was even more massive on the main sequence
stage. So, the most likely product of the explosion of
such a massive star should be a BH.
If the present velocities of runaway stars are known,
one can estimate their ages and places of birth. This
has been done by Hoogefwerf et al. [10]. To calculate the
present position of a BH we have to know the binary pa-
rameters, i.e., the masses of stars before the explosion,
the BH mass, the eccentricity of the orbit before the
explosion, the orbit orientation, and finally the kick ve-
locity of the BH. Some parameters can be inferred from
the observation of the secondary star. Also we can give a
zero kick to the BH and zero orbital eccentricity. Below
we briefly comment on that choice. Other parameters
should be varied within assumed ranges.
As we do not know the exact mechanism of SNa ex-
plosion any value of a BH kick velocity would be specu-
lative. But in all mechanisms BH kicks should be smaller
than those of NSs. In particular, in [27] the authors ar-
gue that the magneto-rotational mechanism of SNae ex-
plosion (suggested in 70's by Bisnovatyi-Kogan) is the
most favored from the point of view of mass distribution
of compact objects. In this mechanism fastly rotating
protoNS form increasing superstrong toroidal magnetic
field (up to 10 17 G), which drives the envelope ejection.
In this case BHs should receive kicks much smaller than
those of NSs. We assume BH kick to be zero (but note,
that in [8] it was suggested that in disrupted binaries one
has to expect mostly low mass BHs, M bh  3 M , which
receive a modest kick  50 km s 1 at their birth). In
the binary systems which are the progenitor of runaway
stars we can expect circularization of orbits due to tidal
interaction, i.e. eccentricity in such binaries should be
zero. As far as the present velocities of secondaries are
high we can expect close systems (a  1000 R ) with
nearly ideal circularization.
Given these parameters and still leaving undeter-
mined the orientation of the orbital plane of a binary,
what remains to discuss are the masses of the primary
component before the explosion and of the BH. In the
case of two massive companions and in the most proba-
ble situation, the mass of the primary before the explo-
sion is close to that of the secondary. These circularized
systems with equal mass stars will survive the explosion
if the kick is strictly zero, since less than one half of the
total mass will be swept out. So, for systems producing
run-away stars the most probable primary's mass is as
close as that requested for binary disruption, i.e.:
M 1 = M 2 + 2
M bh : (1)
That leads to restrictions on the orbital separation: it
should be larger than about 100 R to avoid mass
transfer. On the other hand it has to be smaller than
2000 R , to guarantee that the velocity of the sec-
ondary component after SNa explosion is in agreement
with observations.
Strong mass loss due to stellar winds also leads to
relatively low masses for BH progenitor (before the ex-
plosion), as far as more massive stars loose mass due to
stellar winds faster than low mass stars, so for two stars
with initially very different masses this difference will
become smaller before the explosion.
Results of [7], [8] suggest that massive stars with
M > 40M produce BHs without SNae, so for such

4 Popov et al
cases binaries will survive BH formation. This argument
once again brings us to require the lowest masses for the
primaries that are permitted according to the condition
of binary disruption. Masses should then fall within a
relatively narrow interval close to critical mass of BH
formation.
Masses of BHs are determined now for nearly a dozen
of candidates in binary systems, mainly with low-mass
companions. Most of these determinations are concen-
trated around  7-10 M , but the potentially cover a
relatively wide range from 3 up to 50 solar masses [2],
[27] (see the theoretical expectations of the BH mass
spectrum, which is different from the observed one due
to selection effects [8]).
Given these constrains we are in the process of study-
ing the dynamics of the runaway BHs in the solar prox-
imity and work is in progress.
3. Conclusions
We conclude that the seven radio-quiet ROSAT INSs
can be connected with recent SNa explosions, which pro-
duced nearby runaway stars and peculiar features in the
local ISM including the Local Bubble.
We suggest a way to find approximate positions of
close IBHs from knowledge of nearby runaway stars and
from calculations of binary disruptions. We estimate a
number of close IBHs as > 5 with ages < 3-4 Myr.
Acknowledgement
We thank Luca Zampieri for discussions. S.P. thanks
Universit
a degli Studi dell'Insubria, Universit
a degli
Studi di Padova and Universit
a degli Studi di Milano
Bicocca for hospitality and support.
The work of S.P. and M.P. was supported by RFBR
(grants 01-02-06265, 01-15(02)-99310).
Список литературы
[1] D.P. Bennet et al., Gravitational microlensing events
due to stellar mass black holes; astro-ph/0109467.
[2] A.M. Cherepashchuk, Phys. Usp. 39, 753 (1996).
[3] M. Colpi, U. Geppert and D. Page, ApJ 529, L29
(2000).
[4] M. Colpi, A. Possenti, S. Popov and F. Pizzolato, in:
Physics of Neutron Star Interiors, eds. D. Blaschke,
N.K. Glendenning and A. Sedrakian, SpringerVerlag,
Berlin, p.441, 2001; astro-ph/0012394.
[5] J.M. Cordes and D.F. Chernoff, ApJ 505, 315 (1998).
[6] R.C. Duncan and C. Tompson, ApJ 392, L9 (1992).
[7] C.L. Fryer, ApJ 522, 413 (1999).
[8] C.L. Fryer and V. Kalogera, ApJ 544, 548 (2001).
[9] Y. Fujita, I. Susumi, T. Nakamura, T. Manmoto and
K.E. Nakamura, ApJ 495, L85 (1998).
[10] R. Hoogerwerf, J.H.J. de Bruijne and P.T. de Zeeuw,
A&A 365, 49 (2001).
[11] R. Hoogerwerf, J.H.J. de Bruijne and P.T. de Zeeuw,
ApJ 544, L133 (2001).
[12] E.V. Gotthelf and G. Vasisht, in: Proceedings of IAU
Coll. 177, "Pulsar Astronomy  2000 and Beyond", eds.
M. Kramer, N. Wex and N. Wielebinski, ASP Conf.
Series 202, 699 2000.
[13] I.A. Grenier, A&A 364, L93 (2000).
[14] D.Yu. Konenkov and S.B. Popov, PAZh 23, 569 (1997);
astro-ph/9707318.
[15] V.M. Lipunov, Astrophysics of Neutron Stars,
SpringerVerlag, Berlin, 1992.
[16] V.M. Lipunov and S.B. Popov, AZh 72, 711 (1995).
[17] A.G. Lyne and D.R. Lorimer, Nature 369, 127 (1994).
[18] J. Ma
iz-Apell
aniz, ApJ 560, L83 (2001).
[19] C. Motch, in: Proceedings of X-ray Astronomy '999 
Stellar Endpoints, AGN and the Diffuse Background,
eds. G. Malaguti, G. Palumbo and N. White, Gordon
& Breach (Singapore), 2001 ; astro-ph/0008485.
[20] R. Neuhauser and J.E. Trumper, A&A 343, 151 (1999).
[21] J.P. Ostriker, M.J. Rees, and J. Silk, Astroph. Letters
6, 179 (1970).
[22] B. Paczynski, Can HST measure the mass of
the isolated neutron star RX J185635-3754?; astro-
ph/0107443.
[23] W. Poppel, Fund. Cosm. Phys. 18, 1 (1997).
[24] S.B. Popov, M. Colpi, A. Treves, R. Turolla, V.M.
Lipunov and M.E. Prokhorov, ApJ 530, 896 (2000).
[25] S.B. Popov, M. Colpi, M.E. Prokhorov, A. Treves and
R. Turolla, ApJ 544, L53 (2000).
[26] M.E. Prokhorov, S.B. Popov and A.V. Khoperskov,
The period distribution of old accreting isolated neu-
tron stars, to appear in A&A; astro-ph/0108503.
[27] M.E. Prokhorov, K.A. Postnov, Why NS and BH mass
distribution is bimodal?, to appear in Odessa Astr.
Publ.; astro-ph/0110176.
[28] D.M. Sfeir, R. Lallement, F. Crifo and B.Y. Welsh,
A&A 346, 785 (1999).
[29] V.F. Shvartsman, AZh 47, 824 (1970).
[30] V.F. Shvartsman, AZh 48, 479 (1971).
[31] R.K. Smith and D.P. Cox, ApJ Supp. 134, 283 (2001).
[32] T.M. Tauris et al., ApJ 428, L53 (1994).
[33] C. Thompson and R.C. Duncan, ApJ , 473, 322 (1996)
[34] O.D. Toropina, M.M. Romanova, Yu.M. Toropin and
R.V.E. Lovelace, Propagation of Magnetized Neutron
Stars Through the Interstellar Medium, to appear in
ApJ ; astro-ph/0105422.
[35] A. Treves and M. Colpi, A&A 241, 107 (1991).
[36] A. Treves, R. Turolla, S. Zane and M. Colpi, PASP
112, 297 (2000).
[37] F.M. Walter, ApJ 549, 433 (2001).
[38] F.M. Walter, S.J. Wolk and R. Neuhauser, Nature 379,
233 (1996).
[39] J. Wang, ApJ 486, L119 (1997).
[40] L. Zampieri et al., A&A 378, L5 (2001).