Pulsars and Supernova Remnants

Home , Pulsar, SN 1054, SN 386

1604–2004: SUPERNOVAE AS COSMOLOGICAL LIGHTHOUSES ASP Conference Series, Vol. 342, 2005 M. Turatto, S. Benetti, L. Zampieri, and W. Shea

Roger A. Chevalier Dept. of Astronomy, University of Virginia, P.O. Box 3818, Charlottesville, VA 22903, USA

Abstract. Massive star supernovae can be divided into four categories depending on the amount of mass loss from the progenitor star and the star’s radius. Various aspects of the immediate aftermath of the supernova are expected to develop in diﬀerent ways depending on the supernova category: mixing in the supernova, fallback on the central compact object, expansion of any pulsar wind nebula, interaction with circumstellar matter, and photoionization by shock breakout radiation. Models for observed young pulsar wind nebulae expanding into supernova ejecta indicate initial pulsar periods of 10 − 100 ms and approximate equipartition between particle and magnetic energies. Considering both pulsar nebulae and circumstellar interaction, the observed properties of young supernova remnants allow many of them to be placed in one of the supernova categories; the major categories are represented. The pulsar properties do not appear to be related to the supernova category.

1. Introduction

The association of SN 1054 with the Crab Nebula and its central pulsar can be understood in the context of the formation of the neutron star in the core collapse and the production of a bubble of relativistic particles and magnetic fields at the center of an expanding supernova. Although the finding of more young pulsars and their wind nebulae initially proceeded slowly, there has recently been a rapid set of discoveries of more such pulsars and nebulae (Camilo 2004). In a number of cases there is interaction with a circumstellar medium, which gives further diagnostics on the system. There is considerable variety in the properties of these systems, and the variety is enhanced by the finding of non-normal pulsar objects in the centers of supernova remnants, such as the central compact object in Cas A (Pavlov et al. 2000). The variety of collapsed objects and their surroundings brings up the question of what determines the outcome of a particular core collapse. An initial approach to this question is to investigate whether there is a relation of the type of core collapse supernova to the nature of the compact star aftermath. The classification of core collapse supernovae is currently confused because different classes depend on different observed aspects of the event. I will use 4 basic categories of events: IIP, IIL/b, Ib/c, and SN 1987A-like. The first three categories refer to the amount of H envelope that is lost before the supernova. In the Type IIP supernovae (SNe IIP), the ones with a plateau light curve, most of the H envelope is retained at the time of the explosion. In the SNe IIL/b, ones which show a linear light curve or a late transition to Type Ib, some of the H envelope is present but most of it has been lost as a result of mass loss. In 422 Pulsars and Supernova Remnants 423 the SNe Ib/c, which have no H lines in their spectra, the H envelope has been entirely lost. The progenitors of SN 1987A-like events are stars that retain most of their H envelopes, like the IIP, but they explode as blue supergiants, unlike the red supergiant explosions of SNe IIP. As described in section 2, these different types of events lead to different young supernova remnant properties and to different environments for a central compact object. In section 3, I describe how observed young SNRs might be placed into these categories. The details of these arguments and full references can be found in Chevalier (2005).

2. Properties of the Supernova Categories

The density and composition structure of a supernova is set up soon after the shock front has passed through the star and the gas approaches a state of free expansion. In this process, the presence of strong density gradients at composition interfaces (H/He, He/O, and O/Si) plays an important role. At the edge of a density drop, the flow tends toward free expansion, but is decelerated by the next layer out. The deceleration gives an enhanced pressure which drives a reverse shock wave toward the center of the star and gives rise to Rayleigh-Taylor instabilities at the interface. The instabilities only have a chance to develop fully if the denser layer is completely decelerated, i.e. at the H/He interface the H envelope must be mostly present. At the interface, heavy material is mixed out in plumes, and the lighter inner material is mixed back toward the center in bubbles. Although there have been a number of computer simulation of the instabilities (Kifonidis et al. 2003), the detailed theoretical aspects of the flows remain uncertain and inferences from observations of supernovae in the free expansion phase are necessary. In IIP and 1987A-like supernovae, the H is observed to be present down to low velocities (few 100 km s−1). This is not the case for IIL/b supernovae; for example, in the Type IIb SN 1993J, the H extends over the velocity range 8500 − 10, 000 km s−1 (Houck & Fransson 1996). In both SNe IIL/b and Ib/c, there is evidence for mixing between heavy element layers. In addition to the composition structure, mixing to the center may be important for setting up the surroundings of the central neutron star. Fallback onto the neutron star occurs because some material will be marginally bound to the neutron star. Neutrino losses are high for material that is accreted to the neutron star, so that it can become part of the star. Possible modes of fallback are either material that does not directly escape the gravitational potential and material that is brought back to the center by the reverse shock(s). In the first mode, there is a critical energy below which, or mass above which, there is a great deal of fallback and a black hole is the likely outcome. In the other limit, there may be little fallback. In the case of the reverse shock, the fallback depends on the conditions in the center of the explosion which depend on details of the explosion hydrodynamics. For SN 1987A, Chevalier (1989) estimated a fallback mass of ∼ 0.1 M⊙, but the more recent calculations of Kifonidis et al. (2003) indicate central conditions which imply a smaller mass. An interesting aspect of material brought back by the reverse shock is that it may have greater angular momentum than the mass that is initially close to the compact object, possibly spinning up the neutron star by fallback. 424 Chevalier

Another diﬀerence between the types of supernovae is the energy in ionizing radiation at the time of shock breakout. The energy primarily depends on the radius of the progenitor star. For an extended red supergiant progenitor (SNe IIP or IIL/b), the radiation can ionize all the mass lost from the star. For a compact progenitor (SNe Ib/c), only a fraction of a solar mass can be ionized.

3. Pulsar Nebulae in Young Supernova Remnants

Pulsars are born inside of supernovae, where their nebulae of relativistic particles and magnetic ﬁelds can sweep up the surrounding supernova gas. The expansion of the nebula can be described as the expansion of a relativistic bubble of gas. The pulsar is expected to spin down at some point, which reduces the power being produced by the pulsar. Fig. 1 shows the energies associated with the nebula in various forms. Here, it is assumed that the nebula remains within the −1.06 −3 inner part of the supernova density proﬁle, with a form ρsn ∝ (r/t) t based on Matzner & McKee (1999). The constant τ is the initial spindown timescale of the supernova and n is the braking index, assumed to be constant. It can be seen that the ratio Eint/E˙ t provides an estimate of the evolutionary status of the pulsar and nebula, where Eint is the internal energy, E˙ is the current power output of the pulsar, and t is the age. Information on Eint can be obtained from the minimum energy derived from synchrotron emission.

Figure 1. The evolution of pulsar nebula internal energy, Eint, kinetic energy of the swept-up shell, Ekin, and the current E˙ t divided by the initial rotational energy, Erot. The 2 models are characterized by the power law index of the supernova density proﬁle, m, and the pulsar braking index, n. The reference time τ is the initial spindown time of the pulsar (from Chevalier 2004). Pulsars and Supernova Remnants 425

The results of applying this model of pulsar nebula expansion are shown in Table 1 (Chevalier 2004). The ages and initial periods, P0, are generally found by taking the current radius of the object and finding a model that fits the radius, assuming expansion into the inner part of a typical supernova. After a model is calculated, it is possible to check that the pulsar nebula is within the inner, moderately flat part of the density profile. This is true for the nebulae discussed here. There are some cases, such as the Crab and 0540–69, for which there is independent information on the age. In general, the preferred models have an internal energy within a factor of a few of the minimum energy required for the synchrotron emission. An age much less than the characteristic pulsar age, P/2P˙ , indicates that the nebula is in an early evolutionary stage. This is the case for G11.2-0.3 and, to some extent, for G54.1+0.3 and 0540–69. The pulsars in G54.1+0.3 and G292.0+1.8 currently have similar properties (P and P˙ ) but the G54.1+0.3 nebula has a smaller radius and luminosity, suggesting that the pulsar in G54.1+0.3 is younger and less evolved. The range of initial periods deduced for the pulsars is narrower than the current range, with a typical value of 10’s of ms for P0. These estimates do not include any initial spindown which might occur without doing work on the surrounding supernova.

Table 1. Pulsars and their nebulae

PSR Supernova P/2P˙ Age PP0 Supernova Remnant (yr) (year) (ms) (ms) Type B0531+21 Crab 1240 950 33 20 IIP J0205+64 3C 58 5390 3500 66 40 IIP B0540–69 N158A 1660 800 50 40 Ib/c J1846–03 Kes 75 723 1000 325 30 Ib/c B1509–58 MSH 15–52 1700 1700 150 10 Ib/c J1124–59 G292.0+1.8 2890 3200 135 40 IIL/b J1811–19 G11.2–0.3 24,000 1600 65 60 IIL/b J1930+19 G54.1+0.3 2890 1500 137 100 IIP,Ib/c J1119–61 G292.2–0.5 1606 1700 408 ≪ 200 IIP,Ib/c

The age estimates make it possible to check on suggested supernova identifications with the young remnants. In the case of 3C58, the age is 3000 − 4000 yr, inconsistent with the identification of 3C58 as the remnant of SN 1181 (Green & Stephenson 2001). The larger age is indicated by the internal energy in the pulsar nebula, the expansion of the pulsar nebula, and the mass of swept-up thermal gas, and is consistent with the relatively low expansion observed at optical and radio wavelengths. Models for the pulsar nebulae in MSH 15-52 and G11.2-0.3 are consistent with their identification with SN 185 and SN 386, re- spectively. The age of MSH 15-52 is better specified by the models than that of G11.2-0.3 because of the strong pulsar spindown in the case of MSH 15-52. In the case of G11.2-0.3, the external wind interaction is also consistent with the supernova identification. Table 1 also gives plausible supernova type identifications for the young remnants. The interaction with circumstellar matter is an important discriminant for the supernova type. Type IIP and IIL/b supernovae are thought be the 426 Chevalier explosions of red supergiant stars, which have slow dense winds. However, radio and X-ray observations of supernovae have shown that the Type IIP have relatively low density winds compared to the IIL/b. The expansion of the lower density winds can be stopped by the surrounding pressure at a relatively small radius (< 1 pc). Outside of this region is a low density bubble created by the main sequence star wind. This situation has the possibility of explaining the lack of observed interaction regions around the Crab Nebula and 3C58. The supernova ejecta would have overrun the wind material at an early phase and the emission from the hot gas may have dropped below the detection level as the shock front moves into the low density bubble region. Another property of SNe IIP is the presence of H-rich ejecta material with a velocity ∼ 1000 km s−1. Both the Crab and 3C58 have optical filaments which provide evidence for such material. The case of G54.1+0.3 is similar in that it does not have an observed interaction outside the pulsar nebula. However, no ejecta are observed in this case, so there is no conclusive evidence for a SN IIP identification. The red supergiant wind associated with SNe IIL/b is denser and has more momentum than for the case of SNe IIP. For typical parameters, the wind can extend out 5–7 pc from the star and the supernova interaction with the wind can continue for 2000 − 3000 yr. The remnants G292.0+1.8, with a supernova remnant radius of 7 pc, and G11.2–0.3, with a radius of 3.3 pc, are both excellent candidates for young remnants of SNe IIL/b with the wind interaction going on. In the case of G292.0+1.8, there is also evidence for heavy element rich ejecta moving at ∼ 1000 km s−1, which gives further support for this identification. If a massive star becomes a Wolf-Rayet star, the wind velocity increases to ∼ 100 times its value in the red supergiant phase. For typical parameters, the Wolf-Rayet wind is able to sweep up the red supergiant wind and move it to larger radii (∼> 10 pc). Some dense clumps of the red supergiant wind might remain, but the interclump medium is expected to have a low density (< 10−2 cm−3). One signature of gas that has been part of a red supergiant wind is the presence of an overabundance of N. Such an overabundance has been observed in the dense circumstellar material around a number of supernovae (Fransson et al. 2004). The remnants MSH 15-52, 0540–69, and Kes 75 have all expanded to a large −1 radius in a short amount of time, implying mean velocities ∼> 10, 000 km s . The rapid expansion requires a low surrounding density, as might be produced around a Wolf-Rayet star. In the cases of MSH 15-52 and 0540–69 there is evidence for N-rich clumps at a large radius, as might also occur around the explosion of a Wolf-Rayet star. In addition, the ejecta with velocity ∼ 1000 km s−1 in 0540–69 may be H-poor, as expected for a SN Ib/c. The suggested supernova types can be compared to the pulsar properties. Neither the estimated initial period nor the pulsar magnetic field appear to be related to the supernova type. There is no signature of fallback in the results. A dependence of pulsar properties on supernova type might be expected if the various types result from single stars, as described by, e.g., Heger et al. (2003). In this view, SNe IIP come from stars of mass ∼ 9−25 M⊙, SNe Ib/c from mass ∼> 35 M⊙, and SNe IIL/b from the intermediate range. The rate of core collapse supernovae is strongly dominated by Type IIP events; the ratio of IIP to IIL/b rates is ∼ 10 and IIP to Ib/c ∼ 5. In addition, almost all of the SNe IIL/b and Pulsars and Supernova Remnants 427 a substantial number of SNe Ib/c undergo core fallback to a black hole and do not leave a neutron star. The apparent finding of a number of young remnants of SNe IIL/b and Ib/c with neutron stars like those in remnants of SNe IIP may not be consistent with the single star scenario, although the number of objects is small at present. Another possibility is that binaries play a role with SNe IIL/b and Ib/c, in which case there is an overlap of the core masses giving rise to the supernova events and there may be similarity of the neutron star remnants. Acknowledgments. I am grateful to the organizers for a stimulating and enjoyable meeting and to NASA grant NAG5-13272 for support.

References

Camilo, F. 2004, in Young Neutron Stars and Their Environments, IAU Symposium 218, Ed. F. Camilo & B. M. Gaensler (San Francisco: ASP), 97 Chevalier, R. A. 1989, ApJ, 346, 847 Chevalier, R. A. 2004, ApJ, 619, 839 Fransson, C., et al. 2004, ApJ, 622, 991 Heger, A., Fryer, C. L., Woosley, S. E., Langer, N., & Hartmann, D. H. 2003, ApJ, 591, 288 Houck, J. C., & Fransson, C. 1996, ApJ, 456, 811 Kifonidis, K., Plewa, T., Janka, H.-T., & Muller,¨ E. 2003, A&A, 408, 621 Matzner, C. D., & McKee, C. F. 1999, ApJ, 510, 379 Pavlov, G. G., Zavlin, V. E., Aschenbach, B., Trump¨ er, J., & Sanwal, D. 2000, ApJ, 531, L53