High Energy Emission Properties of Neutron Stars

As a result of observations with the satellite observatories Einstein, ROSAT, ASCA, BeppoSAX, Chandra, and XMM-Newton, 80 rotation-powered pulsars were detected at X-ray energies until the end of 2006. Thus, in 7 yrs of operation XMM-Newton and Chandra have more than doubled the number of detected X-ray pulsars compared to what was known at the end of the ROSAT mission in December 1999 [11]. Table 14.2 reflects the progress made in detecting pulsars of various categories at X-ray energies in recent years. This progress clearly goes along with the increase of sensitivity and angular resolution of the available X-ray telescopes. XMM-Newton with its super collecting power allows to obtain timing and spectral information even from faint and millions of years old pulsars, while Chandra stands for subarcsecond angular resolution, which made it possible to detect and study neutron stars located in source confused regions such as globular clusters and supernova remnants.

Table 14.2 Progress in detecting rotation-powered pulsars with X-ray observatories (status as of December 2006)

Pulsar age

Pulsar

Einstein

ROSAT

XMM-Newton

( years )

category

ASCA

Chandra

<104

Crab-like

3

5

9

104 -105

Vela-like

1

9

15

105 -106

Cooling NS

5

6

106 -108

Old and nearby

1

3

8

> 108

binary

1

3

ms-Pulsar

11

39

Z detected:

5

33

80

While Einstein had only the sensitivity to see pulsed X-ray emission from the youngest and brightest pulsars and to detect a few others at the limit of its sensitivity, ROSAT/ASCA and XMM-Newton/Chandra allowed for the first time to study the emission mechanisms of rotation-powered pulsars based on a broader sample and of various categories

While Einstein had only the sensitivity to see pulsed X-ray emission from the youngest and brightest pulsars and to detect a few others at the limit of its sensitivity, ROSAT/ASCA and XMM-Newton/Chandra allowed for the first time to study the emission mechanisms of rotation-powered pulsars based on a broader sample and of various categories

Fortunately, with the increase in sensitivity of todays observatories a growing number of neutron stars are detected in more than just one waveband (e.g., at radio, optical, EUV, X- and gamma-rays), making it possible for the first time to carry out multiwavelength studies of the pulsar emission. This is a big advantage as the physical processes that cause the emission in different wavelength bands are obviously related to each other. Multiwavelength studies thus provide a much broader view into the physical processes operating in the neutron star magnetosphere than interpretating emission properties observed in a single wave band only.

14.3.1 Young Neutron Stars in Supernova Remnants

X-ray observations allow us to find both supernova remnants (SNRs) and the compact objects that may reside within them. In fact, neutron stars and neutron star candidates have been found in a small fraction of the 265 known galactic SNRs3 [40]. About 35 of these compact stellar remnants in SNRs are radio pulsars, others are radio-silent (or, at least, radio-quiet) neutron stars, which were found as faint pointlike X-ray sources near to the geometrical center of their supernova remnant.

Being in orbit for more than half of their nominal lifetime almost all young radio pulsars have been observed and detected by either XMM-Newton and/or Chandra (cf. Fig. 14.1). The young rotation-powered pulsars can be divided in two groups, Crab-like and Vela-like pulsars, according to somewhat different observational manifestations apparently associated with the evolution of pulsar properties with age. The radio-silent neutron stars include anomalous X-ray pulsars (AXPs), soft gamma-ray repeaters (SGRs), and "quiescent" neutron star candidates in SNRs. There is growing evidence that AXPs and SGRs are indeed magnetars (see [109] for a review). Magnetars are neutron stars with an ultra strong magnetic field (B > 1014G), which is supposed to be the source of the detected high energy radiation. A common property of these objects is that their periods are in a narrow range of 5 - 12 s, substantially exceeding typical periods of radio pulsars. Although no gamma-ray emission has been detected from AXPs, SGRs occasionaly emit soft gamma-ray bursts of enormous energy (up to 1042-1044 erg), a property, which gave this sources their name.

14.3.1.1 Crab- and Vela-Like Pulsars

On July 4, 1054 A.D., Chinese astronomers noted a guest star in the constellation Taurus. As we know today, this event marked the arrival of light from the deadth of a massive main sequence star that underwent a core collapse when its internal thermal energy produced by the nuclear fusion processes was not sufficient anymore to counteract the gravitational force against the star's collapse. The cloud of gas that

3 http://www.mrao.cam.ac.uk/surveys/snrs/

we observe today at the position of this guest star is the Crab supernova remnant. In the optical band the nebula has an extent of 4 x 6arcmin, corresponding to ^7 x 10 light years for a distance of 2kpc.

What we observe from the Crab nebula in X-rays is not the thermal emission from the ejecta-driven blast wave of the supernova, though, but the emission from charged particles that emit synchrotron radiation as they move along magnetic field lines. In X-rays the nebula has the form of a torus with jets, wisps and a counter-jet, having an overall extent of 2 x 2 arcmin in the sky. For the 2 kpc distance, the radius of the torus is 0.38 pc, that of the inner ring is 0.14 pc [107,108]. A composite image showing the Crab nebula and pulsar in the radio, optical and X-ray band is shown in Fig. 14.3.

In studying this system it became clear very early that the observed nonthermal emission required a continuous input of energetic charged particles to keep the nebula emitting. It was the question of the Crab nebula's central engine that caused

Fig. 14.3 A composite image of the Crab nebula and its central pulsar as produced by emission detected in the radio (green), optical (red), and X-ray (blue) wavebands

Pacini (1967) a few month before the discovery of radio pulsars to propose that a fast spinning and strongly magnetized neutron star could be the required source that supplies the energy into the nebula.

Indeed, the 33 ms pulsar in the Crab supernova remnant, PSR B0531+21, was the first rotation-powered pulsar from which high energy radiation was detected. Being the strongest rotation-powered pulsar with the highest spin-down energy it was considered - until recent years - to be the proto type for all young neutron stars of age 103 - 104 years. Because of this and its favorable brightness, it was studied in all frequency bands and by almost every observatory suitable to do so. The pulsar's characteristic double peaked pulse profile and its energy spectrum have been measured in detail throughout almost the entire electromagnetic spectrum. A compilation of pulse profiles as observed from the radio to the X-ray bands is shown in Fig. 14.4.

Despite this strong interest and a wealth of data that have been taken from the pulsar since its discovery, it only recently became clear in deep Chandra observations that the X-ray emission from the Crab pulsar is actually 100% pulsed [95] and that the radio, optical, and X-ray pulses are not fully phase aligned as suggested by

Fig. 14.4 The Crab pulsar's characteristic pulse profiles as observed at various frequency bands (from [71]). The phases of low (LFC) and high frequenc (HFC) radio pulse components are indicated. The dashed lines indicate the phase of the main pulse component and the interpulse. The profiles have been arbitrarily aligned by the peak of the main pulse

Fig. 14.4 The Crab pulsar's characteristic pulse profiles as observed at various frequency bands (from [71]). The phases of low (LFC) and high frequenc (HFC) radio pulse components are indicated. The dashed lines indicate the phase of the main pulse component and the interpulse. The profiles have been arbitrarily aligned by the peak of the main pulse the high-energy emission models. Indeed, the X-ray pulses lead the optical pulses by the small amount of <60 |s and the optical the radio pulses by 260 |s [6]. Mapping these pulse arrival time differences to photon travel-time differences means that the Crab's X-ray pulses may come from locations in the pulsar magnetosphere closer to the neutron star's surface than the radio emission, which then might be emitted at higher altitudes.

As far as the pulsar's emission mechanisms are concerned, it is very well established that magnetospheric emission from charged particles, accelerated in the neutron star magnetosphere along the curved magnetic field lines, dominates the radiation not only from the Crab pulsar but from almost all young rotation-powered pulsars with ages < 5000yrs (cf. 14.2.2). Accordingly, the radiation of Crab-like pulsars is characterized by a power-law spectrum, dN/dE ^ E-a, as the energy distribution of the particles that emit this radiation follows a power-law in a broad energy range. For the Crab pulsar, the slope of its flux spectrum slowly increases with photon energy - the photon index varies from a = 1. 1at E < 1 keV to a = 2.1 at E < 1010 eV.

Only recently a pulsar was detected, which seems to conflict with this empirical evidence of nonthermal dominated emission in young rotation-powered pulsars. PSR J1119-6127, which is located in the SNR G292.2-0.5, has an age of < 1600 yrs and a deduced magnetic field strength of B < 4.1 x 1013G. The latter is close to the quantum critical field of Bqed = m^/eh = 4.4 x 1013G and close to the magnetar range. There is strong evidence that its spectrum is dominated by thermal radiation corresponding to a temperature of <2.4 x 106 K and an emitting radius of <3.4km while its pulsed fraction in the 0.5 - 2.0keV band is as high as <74% ± 14%. Although it still requires better data to finally identify the X-rays of this pulsar as being of thermal origin, it is an interesting question of whether the presence of the strong magnetic field causes a completely different emission scenario than observed in other young and Crab-like pulsars.

Pulsars with a spin-down age of <104 - 105 yrs are often referred to as Vela-like pulsars, because of their apparent similar emission properties. Among the 14 pulsars of this group that have been detected in X-rays, four of them (the Vela pulsar PSR B0833-45, PSR B1706-44, B1046-58, and B1951+32) are gamma-ray pulsars, and only the Vela pulsar has been detected in the optical band. In some respects, these objects appear to be different from the Crab-like pulsars. In particular, their optical radiation is very faint compared with that of the very young pulsars, and the overall shape of their high-energy spectra looks different. For instance, the closest (d « 300 pc) and, hence, best-investigated Vela pulsar has an optical luminosity four orders of magnitude lower than the Crab pulsar [75], whereas its rotation energy loss is only a factor of 65 lower. Its pulse profile at various wavelength is very complex and difficult to associate with the various possible emission mechanisms [64]. The pulsed fraction in the soft X-ray range, «7%, is much lower than that observed from Crab-like pulsars.

In contrast to the young Crab-like pulsars, the soft X-ray spectrum of the Vela pulsar has a substantial thermal contribution with an apparent temperature

Fig. 14.5 The Vela pulsar and its plerion as observed by the Chandra ACIS-I detector. A torus and jets are seen similar as in the Crab plerionic nebula. The symmetry axis is almost aligned to the pulsar's proper motion direction of «106 K [79]. On the other hand, the spatial structure of the Vela plerion strongly resembles the inner Crab nebula - it also has a torus-like structure, an inner ring and jets (see Fig. 14.5). The symmetry axis of the nebula, which can be interpreted as the projection of the pulsar's rotation axis onto the sky plane, is roughly coaligned with the direction of proper motion. This is similar as observed in the Crab pulsar although the missalignement there is 26° ± 3°. The idea of a torus configuration formed by a shock-confined pulsar wind was first introduced by Aschenbach and Brinkmann [2] as a model to explain the shape of the inner Crab nebula. The discovery of a similar torus-like structure in the Vela synchrotron nebula indicated that this model may be applicable to many young pulsars. According to this model, the torus-like structure and its geometrical orientation with respect to the direction of the pulsar's proper motion arise because the interaction of the postshock plasma with the ambient medium compresses the plasma and amplifies the magnetic field ahead of the moving pulsar. This, in turn, leads to enhanced synchrotron emission with the observed torus-like shape.

Thus, young rotation-powered pulsars are in general surrounded by pulsar-powered nebulae (plerions) and/or supernova ejecta. Presumably, their magne-tospheric emission extends from at least the infrared to gamma-ray energies, with typical photon indices varying between a « 1 — 2 (about 1.4 — 1.7 in the soft X-ray range). As the plerionic emission is synchrotron radiation its spectrum is a power law. In the Crab plerion, Willingale et al. [108] found that the shape of the spectrum changes as a function of distance from the pulsar. He fitted the power law slope of the torus (a = 1.8 ± 0.006), the jet (a = 2.1 ± 0.013), and the outer nebula regions (a = 2.34 ± 0.006). Similar results were obtained by Chandra, measuring the hardness ratio distribution throughout the nebula [107]. For the pulsar, a photon spectral index of a = 1.63 ± 0.09 is observed. The spectral difference between the jet and the torus is found to be likely due to an intrinsically steeper electron spectrum of the jet. The outer regions of the nebula show the steepest spectrum, which is likely to be due to enhanced synchrotron losses of the electrons during their ride from the pulsar to the outskirts.

14.3.1.2 Central Compact Objects in Supernova Remnants

For many years, it has been generally believed that all young neutron stars have similar emission properties as those observed in Crab- and Vela-like pulsars, i.e., emitting strongly pulsed radiation caused by nonthermal emission processes in the neutron star's magnetosphere. Several recent observations of compact X-ray sources in supernova remnants, however, suggest that this picture is incomplete and indeed no longer justified: it has been shown that there are other manifestations of young neutron stars, e.g., as anomalous X-ray pulsars, soft gamma-ray repeaters or simply as faint point-like X-ray source in a supernova remnant. Most of these sources were identified by their high X-ray to optical flux ratios, others simply by their locations near to the expansion centers of supernova remnants, strongly suggesting that they are indeed the compact stellar remnants formed in the supernova events.

The group of SNRs, which are known to host a radio-quiet but X-ray bright central compact object (CCO), is listed in Table 14.3.

Whether this group of CCOs forms a homogenous class of sources such as the rotation-powered pulsars is currently an open question and is actually difficult to answer in view of the small number of known objects. All sources in common is that (1) they are located in supernova remnants of age < 104 yrs, (2) their X-ray luminosities are all in the braked 1032 — 1034 erg s—1, (3) down to an extent of <1 arcs none of them has been seen to maintain a plerionic X-ray nebula such as the Crab, Vela, or other young pulsars, and (4) no radio or optical counterpart could be detected from any CCO by now.

None of these properties is distinctive enough to justify the interpretation that all these sources form an own class of objects (e.g., Geminga is radio-silent as well). Interestingly, though, is that all CCOs share very similar spectral properties and those are markedly different from what is observed in young rotation-powered pulsars. The X-ray spectra of virtually all CCOs are very well modeled by either one or two component blackbody model with Tbb = (2 — 7) x 106k and a size of the projected emitting area in the range Rbb — (0.3 — 5) km. Alternatively, spectral models consisting of a blackbody and a power law provide valid fits as well. The inferred slope

Table 14.3 List of X-ray detected radio-quiet and optically dim central compact objects in supernova remnants (status: January 2007)

CCO

Hosting

Age

d

P

log Lx

Ref.

SNR

(kyr)

(kpc)

(ergs-1)

CXOU J232327.9+584843

Cas-A

-0.3

-3

-32.94

[33,93]

RX J0852.0—4622

Vela-Jr

-2

-2.2

-32.83

[3,7]

RX J1713.7—3946

G347.3-0.5

-10

-6

-32.63

[63,87]

1E 1613-5055

RCW 103

-2

-3.3

6.67 h

variable

[32,101]

RX J0822-4300

Pupis-A

-2

-2.2

0.22* s

-33.69

[53,54,86]

1E 1207.4-5209

PKS 1209-51

-10

-2.1

0.424 s

-33.39

[48,110]

CXOU J185238.6+004020

Kes 79

-9

-7

0.105 s

-33.07

[39,92]

The X-ray luminosity is computed for the energy band 0.5 —10keV and the specified distances *The periodicity of RX J0822—4300 awaits confirmation

Fig. 14.6 (a) Composite ROSAT HRI image of the Puppis-A supernova remnant. The blue ring indicates the 30 arcmin central region, which has been observed by XMM-Newton. (b) XMM-Newton MOS1/2 false color image of the central region of Puppis-A (red: 0.3 — 0.75 keV, green: 0.75 — 2keV, and blue: 2 — 10keV). The central source is the CCO RX J0822—4300. The inset shows the squared region as observed by the Chandra HRC-I. (Image from [53])

Fig. 14.6 (a) Composite ROSAT HRI image of the Puppis-A supernova remnant. The blue ring indicates the 30 arcmin central region, which has been observed by XMM-Newton. (b) XMM-Newton MOS1/2 false color image of the central region of Puppis-A (red: 0.3 — 0.75 keV, green: 0.75 — 2keV, and blue: 2 — 10keV). The central source is the CCO RX J0822—4300. The inset shows the squared region as observed by the Chandra HRC-I. (Image from [53])

of the power law component, though, is ^4 — 5, which is steeper than the photonindex r = 1 — 3 observed for rotation-powered pulsars [13]. However, as the true nature of the CCOs and their emission mechanisms are unknown it might not be justified to use the steepness of the power law to reject these models as unphysical. Worth to mention in this context is that the spectra observed from AXPs also require a blackbody plus power law model of similar properties.

Despite the common spectral emission properties, there are distinct differences in the temporal emission properties of some CCOs. There is strong evidence that 1E 1613—5055 in RCW 103 and 1E 1207.4-5209 in PKS 1209-51/52 are actually binaries composed of a compact object and a low-mass star in an eccentric orbit [10, 32,110]. For 1E 1613-5055 in RCW 103, a strong periodic modulation at 6.67 ± 0.03 h has been found in long XMM-Newton observations along with changes in the X-ray flux by factors 10-100 while 1E 1207.4-5209 in PKS 120951/52 shows erratic behavior in both its pulsed fraction and period derivative [114]. X-ray pulsations have also been observed from CXOU J185238.6+004020 in Kes 79 [39] and RX J0822-4300 in Puppis-A [53] and both results do not support a scenario of steady spin-down. For RX J0822-4300, e.g., the period time derivative calculated from the separation of the epochs of the two available XMM-Newton data sets is P = (2.112 ± 0.002) x 10-10s s-1, which is among the largest spin-down rates in the neutron star population. The largest known P was inferred for SGR 1806-20, P =(8 - 47) x 10-11s s-1, [60,111]. If the identifications of P and P for RX J0822-4300 are correct, it implies a nonsteady spin-down behavior. It should be noted that these results lend evidence to the interpretation that CCOs, AXPs, and SGRs are all magnetars. There are several AXPs and few SGRs that have been associated with supernova remnants [34,67]. On the basis of Spitzer and ground-based ^s-band images, Krause et al. (2005) suggested that a mid-twentieth century flare from CXOU J232327.9+584843 in Cas-A could have been the source of the apparent 24 jm light echo filaments observed ^20 arcmin north and south of the Cas-A remnant. However, with the available data the relation of CCOs with magnetars cannot be concluded without certainty.

Recently, it became possible for the first time to measure the proper motion of a CCO and to confirm that its back projected birth place is in agreement with the remnants explosion center, thus providing the first confirmation that CCOs are indeed the compact remnants formed in their hosting supernova. Using two Chandra data sets, which span an epoch of 1952 days, Hui and Becker [54] found the position of RX J0822-4300 in Puppis-A (c.f. Fig. 14.6) different by 0.57 ± 0.18 arcs, implying a proper motion of j = 107 ± 34 mas/yr. For a distance of 2.2 kpc, this proper motion is equivalent to a recoil velocity of 1120 ± 360 km s-1. Since both the magnitude and direction of the proper motion are in agreement with the birth place of RX J0822-4300 being near to the optical expansion center of the supernova remnant.

It is finally worth to mention that, by now, the relatively small number of discovered members of this class might be due to observational selection effects only. From the observers point of view, it is much easier to detect and identify active pulsars than these quiet compact sources observable only in the soft X-ray band. Also, once a supernova remnant disappears after about 105 yrs, it is almost impossible to find and identify its left over CCO. It is, therefore, very plausible that, in fact, CCOs may be more common than young Crab- and Vela-like radio pulsars.

0 0

Post a comment