A Possible Connection Between Fast Radio Bursts and Gamma-Ray Bursts

A Preprint typeset using LTEX style emulateapj v. 5/2/11 A POSSIBLE CONNECTION BETWEEN FAST RADIO BURSTS AND GAMMA-RAY BURSTS Bing Zhang Department of Physics and Astronomy, University of Nevada Las Vegas, NV 89154, USA ABSTRACT The physical nature of Fast Radio Bursts (FRBs), a new type of cosmological transients discovered recently, is not known. It has been suggested that FRBs can be produced when a spinning supra- massive neutron star loses centrifugal support and collapses to a black hole. Here we suggest that such implosions can happen in supra-massive neutron stars shortly (hundreds to thousands of seconds) after their births, and an observational signature of such implosions may have been observed in the X-ray afterglows of some long and short gamma-ray bursts (GRBs). Within this picture, a small fraction of FRBs would be physically connected to GRBs. We discuss possible multi-wavelength electromagnetic signals and gravitational wave signals that might be associated with FRBs, and propose an observational campaign to unveil the physical nature of FRBs. In particular, we strongly encourage a rapid radio follow-up observation of GRBs starting from 100 s after GRB triggers. 1. INTRODUCTION into a black hole when centrifugal support no longer holds Recently, a new type of cosmological transients, gravity. When the magnetic field “hair” is ejected as the dubbed Fast Radio Bursts (FRBs), was discovered event horizon swallows the neutron star, a strong elec- (Lorimer et al. 2007; Thornton et al. 2013). These radio tromagnetic signal in the radio band (which they call a bursts have a typical duration of several milli-seconds, “blitzar”) is released. This is an FRB. high Galactic latitudes, and anomalously high disper- Falcke & Rezzolla (2013) suggested that such a delayed sion measure (DM) values corresponding to a cosmolog- collapse would happen several-thousand to million years ical redshift z between 0.5 and 1 (Thornton et al. 2013). after the birth of the supra-massive neutron star. Here we propose that a small fraction of such implosions could The inferred total energy release is 1038 − 1040 ergs, and − also happen shortly (hundreds to thousands of seconds) the peak radio luminosity is ∼ 1043 erg s 1. No detected after the birth of the neutron star, and a signature of electromagnetic counterpart was claimed to be associated such implosions may have been observed in the early X- with FRBs. ray afterglow light curves of some GRBs. The physical nature of FRBs is unknown. GRBs may originate from two types of progenitor: col- Thornton et al. (2013) discussed several possibili- lapse of a massive star (e.g. Woosley 1993) and coeles- ties, and suggested that the event rate of FRBs ∼ −3 −1 −1 cence of two neutron stars (NS-NS merger) or one neu- (RFRB 10 gal yr ) is much higher than those of tron star and one black hole (NS-BH merger) (e.g. of gamma-ray bursts (GRBs) and compact star mergers, Paczýnski 1986; Eichler et al. 1989). A large angular but could be consistent with those of soft gamma-ray momentum and a strong magnetic field are essential repeater giant flares or core-collapse supernovae. Since to launch a jet (e.g. Rezzolla et al. 2011; Etienne et al. the announcement of the discovery, several proposals 2012). There are two types of plausible central engine: have been made to interpret FRBs, including delayed one is a promptly formed black hole (e.g. Popham et al. collapses of supra-massive neutron stars to black holes 1999), which accretes materials from the remnant with an (Falcke & Rezzolla 2013), special magnetar radio flares −1 extremely-high accretion rate (∼ (0.1 − 1)M⊙ s ); the (Popov & Postnov 2007), mergers of double neutron other is a strongly magnetized (with surface magnetic stars (Totani 2013), mergers of binary white dwarfs ∼ 15 (Kashiyama et al. 2013), and flaring stars (Loeb et al. field 10 G) neutron star which is spinning near the arXiv:1310.4893v2 [astro-ph.HE] 26 Nov 2013 2013). break-up limit (millisecond rotation period) (e.g. Usov 1992). Our FRB model invokes this latter central en- 2. FRBS AS IMPLOSIONS OF NEW-BORN gine. SUPRA-MASSIVE NEUTRON STARS Even without direct evidence, a magnetar central en- The milli-second duration τ points towards a small gine is inferred indirectly for some GRBs. A shallow de- 7 cay phase (or “plateau”, Fig.1 lower panel) in the early emission size for FRBs: rFRB ∼ cτ ∼ 3 × 10 cm (τ/ms). The source of emission has to be limited to very com- X-ray afterglow of most long GRBs may require contin- pact objects involving neutron stars or black holes (white uous energy injection into the blastwave (Zhang et al. dwarfs may be marginally accommodated, Kashiyama et 2006), which would be consistent with a spinning-down al. 2013). At such a small size, the brightness tempera- neutron star engine (Dai & Lu 1998; Zhang & Mészáros ture of radio emission is extremely high, so the radiation 2001). An alternative explanation does not invoke a long- mechanism must be coherent (Katz 2013). lasting central engine, but invokes a stratification of the Falcke & Rezzolla (2013) made a good case that a ejecta Lorentz factor (Rees & Mészáros 1998). The de- supra-massive neutron star collapsing into a black hole generacy between the two models was broken when the would be a likely source of FRBs. The supra-massive so-called “internal plateaus” were discovered in the early neutron star is initially sustained centrifugally by rapid X-ray afterglow lightcurves of some GRBs (Fig.1 upper rotation. As it gradually spins down, it would collapse panel) (Troja et al. 2007; Liang et al. 2007). These are 2 Zhang X-ray plateaus followed by an extremely steep decay, dius) with decay index steeper than -3, sometimes reaching RLC -9. By contrast, most “normal” plateaus are followed 2 2 −6 2 EB ≃ (Bp/8π) (r/R) 4πr dr by a decay with a decay index around -1, which is well ZR consistent with the external shock model of GRBs (e.g. ≃ (1/6)B2R3 =1.7 × 1047 erg B2 R3. (1) Gao et al. 2013b, for a recent review of the external shock p p,15 6 model of GRBs). The steepest decay allowed in the ex- This is much larger than the observed energy of FRBs. ternal shock model is defined by high-latitude emission Only a small amount of this energy is adequate to power of a relativistic ejecta (e.g. when the blastwave enters a an FRB. density void), which has a decay slope α = −2+ β (con- The conversion of a small fraction of this energy to α β vention Fν ∝ t ν ), which usually cannot be smaller radio emission energy invokes a poorly known coherent than -3 (Kumar & Panaitescu 2000). The very steep de- radio emission mechanism, such as coherent curvature cay following those internal plateaus therefore demands radiation through “bunches” (Ruderman & Sutherland an internal dissipation mechanism (rather than the exter- 1975; Falcke & Rezzolla 2013) or “maser”-like amplifica- nal shock emission) to account for the data (and hence, tions of plasma modes (e.g. Melrose et al. 2009). Here the plateaus gain their name). This demands that the we assume that a certain coherent mechanism can op- central engine lasts much longer than the burst dura- erate during the hair-ejection process, and the observed tion. The essentially constant X-ray luminosity dur- νobs = (1.2 − 1.5) GHz radio wave can escape the emis- ing the plateau requires a steady central engine output, sion region2. It can reach the observer if it is not and a spinning-down magnetar naturally accounts for the absorbed by the GRB blastwave in front of the FRB data. Later a systematic analysis revealed more internal emission region. The comoving electron number den- plateaus (Lyons et al. 2010). Surprisingly, such a signa- sity in the shocked ejecta region of the blastwave is ture was also found in a good fraction of short GRBs ′ ≃ × 5 −3 −2 −2 n 1.8 10 cm L52Γ2 r17 (L is the wind luminos- (Rowlinson et al. 2010, 2013). ity of the GRB, Γ bulk Lorentz factor, and r the blast- If one accepts that a millisecond magnetar is indeed wave radius), which gives a comoving plasma frequency operating in both long (Usov 1992; Bucciantini et al. ′ ′ 2 1/2 ≃ × 6 1/2 −1 −1 2009; Metzger et al. 2011) and short (Dai et al. 2006; νp = (n e /πme) 3.8 10 Hz L52 Γ2 r17 , much Fan & Xu 2006; Metzger et al. 2008; Kiuchi et al. 2012) smaller than the radio wave frequency in the comoving − × 7 GRBs1, then the very steep decay at the end of inter- frame νobs/Γ=(1.2 1.5) 10 Hz. The density in the nal plateaus suggests that the emission stops abruptly. shocked circumburst medium region is even lower. So It is difficult to turn off a rapidly spinning-down mag- the FRB emission can pass through the blastwave region netar unless it collapses into a black hole. Simulations and reach Earth. Overall, the magnetic bubble associ- show that a rapidly spinning neutron star can have a ated with this FRB ejection, with total energy described threshold mass (for collapsing into a black hole) that is by Eq.(1), would accelerate and convert the energy to larger by 30% - 70% (depending on equation-of-state) the kinetic form. This energy is however small compared than the maximum mass of a non-rotating neutron star with the GRB energy, so would not leave noticeable im- (Bauswein et al.

Load more