Collision between Neutron Stars and Asteroids as a Mechanism for Fast Radio Bursts
aa r X i v : . [ a s t r o - ph . H E ] D ec Collision between Neutron Stars and Asteroids as a Mechanism forFast Radio Bursts
Y. F. Huang, , and J. J. Geng , School of Astronomy and Space Science, Nanjing University, Nanjing 210046,China; [email protected] Key Laboratory of Modern Astronomy and Astrophysics (Nanjing University),Ministry of Education, Nanjing 210046, China
Abstract.
As a new kind of radio transient sources detected at ∼ . ∼ — 10 sky − day − . We suggest that fast radio bursts can be producedby the collisions between neutron stars and asteroids. This model can naturally explainthe millisecond duration of fast radio bursts. The energetics and event rate can alsobe safely accounted for. Fast radio bursts thus may be one side of the multifaces ofthe neutron star-small body collision events, which are previously expected to lead toX-ray / gamma-ray bursts or glitch / anti-glitches.
1. Introduction
Fast radio bursts (FRBs) are short bursts of radio emission from the sky. They werediscovered recently (Lorimer et al. 2007; Thornton et al. 2013), with about only tenevents being reported as definite detections till July 2015 (Keane et al. 2011; Burke-Spolaor & Bannister 2014; Spitler et al. 2014; Ravi et al. 2015; Keane & Petro ff ∼ . ∼ ff et al. 2015). The absence of counterparts leads to1 Y.F.Huang & J.J.Genggreat di ffi culties in understanding the nature of these enigma bursts. From the largeDM values, many authors deduced that FRBs should be at cosmological distances, typ-ically with redshift of 0 . < z <
1. The characteristic radio luminosity ( L FRB ) is then10 — 10 erg s − and the isotropic energy release ( E FRB ) will be 10 — 10 erg.Also, the quick variability of FRBs indicates that the emission region should be verysmall. Based on these facts, various models have been suggested. Some authors arguedthat FRBs are connected with the giant flares of magnetars (Popov & Postnov 2007;Lyubarsky 2014; Kulkarni et al. 2014; Pen & Connor 2015), others believed that theycorrespond to the collapse of hypermassive neutron stars into black holes (Zhang 2014;Falcke & Rezzolla 2014; Ravi & Lasky 2014). They were also suggested to be causedby planetary companions around pulsars (Mottez & Zarka 2014) or double neutron starmergers (Totani 2013). Other authors even proposed that FRBs come from flare stars(Loeb et al. 2014), binary white dwarf mergers (Kashiyama et al. 2013), or evapora-tion of small black holes(Barrau et al. 2014). In many of these models, a very strongelectro-magnetic outburst (something like a gamma-ray burst) would be triggered andshould be observed to associate with the FRB event. Multi-band afterglows are alsoexpected after the FRBs. However, these phenomena have not been observed. Also,some of the models have di ffi culties in accounting for the event rate of FRBs.Recently, we proposed a new model for FRBs (Geng & Huang 2015). We arguedthat they could be produced by the collision of asteroids with neutron stars. The modelcan account for many of the observational characteristics of FRBs, such as the duration,the energetics, the event rate, etc. In this review article, we will first summarize theobservations of FRBs (Section 2). We then present a detailed description of the collisionmodel (Section 3). In Section 4, conclusions and some brief discussion are presented.
2. Observed Features of FRBs
Till the end of July 2015, only 10 FRBs are reported in the literature. They are generallydiscovered through single-pulse search methods by using archive data of wide-fieldpulsar surveys at the multi-beam Parkes telescope and the Arecibo telescope. FRBs areusually detected at GHz frequency. The earliest event occurred in 2001, and the latesthappened in 2014. In Table 1, we list the key parameters of the 10 FRBs observedso far. Of these events, only FRB 121102 is detected by the Arecibo telescope, andall other FRBs are detected by the Parkes telescope. It thus would be very helpfulif FRBs could also be detected by other telescopes, especially by the future ChineseFive-hundred-meter Aperture Spherical radio Telescope (FAST), which will be readyfor observations in late 2016 (Nan et al. 2011).From Table 1, we see that 6 FRBs concentrate in the Galactic latitude range of40 o < | b | < o , But several FRBs also occur at low and middle Galactic latitudes. Amost distinct characteristic of FRBs is their high DM values. The directly measuredDMs are in the range of 375 — 1104 cm − · pc. They are much higher than the corre-sponding Galactic DMs (i.e. DM G in Table 1). After subtracting the contribution fromour Galaxy, the DMs left (i.e. DM E in Table 1) are still in the range of 223 — 1072cm − · pc, with a mean value of 602 cm − · pc. DM is a measure of the column densityof free electrons along the line of sight to the source. Many authors believe that DMscan be used to estimate the distance of these FRBs, but note that in the practice, weare troubled by the unknown contributions from the host galaxy and the local environ-ment of the FRB. In Table 1, we have subtracted a value of 100 cm − · pc from DM E RBsfromNeutron Star-Asteroid Collisions 3
Table 1. Key parameters of the 10 FRBs observed till July 2015. Note that thename of “010125” has been corrected from “011025” of Burke-Spolaor & Bannister(2014); The name of “010724” has been corrected from “010824” of Lorimer et al.(2007); The name of “110626” has been corrected from “110627” of Thornton et al.(2013). “010621” is the source of J1852-08 of Keane et al. (2011). The redshift isestimated by us from z = ( DM E (cm − · pc) − / − · pc is the assumed DM contribution from the host galaxy and thelocal environment of the FRB (Thornton et al. 2013). The luminosity distance iscalculated by assuming a flat Universe with H =
71 km / s / Mpc, Ω M = .
27, and Ω Λ = .
73. The references are: [1] Burke-Spolaor & Bannister 2014; [2] Keane& Petro ff ff et al. 2015.FRB name 010125 010621 010724 110220 110626 l o b o -20.0 -4.0 -41.8 -54.7 -41.7 DM (cm − · pc) 790 746 375 944 723 DM G (cm − · pc) 110 523 45 35 48 DM E (cm − · pc) 680 223 330 909 675Estimated redshift z D L (Gpc) 2.7 0.45 0.92 4.0 2.7Duration τ (ms) 10.3 8.3 20 6.6 1.4 S peak (Jy) 0.55 0.52 1.58 1.11 0.63 F obs (Jy · ms) 5.6 4.3 31.5 7.3 0.9 E radio (10 ergs) ∼ . ∼ . ∼ . ∼ . ∼ . l o b o -59.0 -66.2 -0.2 -21.9 -54.6 DM (cm − · pc) 1104 553 557 779 563 DM G (cm − · pc) 32 32 188 69 35 DM E (cm − · pc) 1072 521 369 710 528Estimated redshift z D L (Gpc) 5.1 1.8 1.1 2.9 1.9Duration τ (ms) 3.9 1.2 3.0 2.0 2.8 S peak (Jy) 0.45 0.62 0.4 2.0 0.47 F obs (Jy · ms) 1.8 0.8 1.2 4.0 1.3 E radio (10 ergs) ∼ . ∼ . ∼ . ∼ . ∼ . and derived the redshift for each event by using z = ( DM E (cm − · pc) − / . ≤ z ≤ .
81. The mean value of the estimated redshift is z = .
42. Consequently, themean luminosity distance is D L = . & J.J.GengFRBs are very short events. Their durations ( τ ) are in the range of 1.2 — 20 ms,with the mean value being 6.0 ms. However, they are very strong radio flashes. Theobserved peak flux density is 0 . < S peak < S peak = .
83 Jy).As a result, the radio fluence is 0 . · ms < F obs < . · ms (mean value: F obs = . · ms). Assuming a beaming solid angle of 1 sr, we then can give a rough estimateof the emitted radio energy, which is 0 . × ergs < E radio < . × ergs. Notethat the mean energy released is E radio = . × ergs.
3. Neutron Star-Asteroid Collision Model
Various models have been proposed for FRBs. However, many of the models cannotsatisfactorily explain the basic features of FRBs, such as the short duration, the strongintensity, the event rate, the absence of intensive emission of γ -rays, the event rate,etc. The nature of FRBs is thus still highly controversial. We argue that FRBs may beinduced by neutron star-asteroid collision events(Geng & Huang 2015). Our model cannaturally explain the millisecond duration of FRBs. It can also well account for variousother aspects of FRBs. We assume the neutron star has a mass of M = . ⊙ and its radius is R NS = cm.The mass and radius of the asteroid are designated as m and r , respectively. For sim-plicity, it is assumed to be of Fe-Ni composition, so that the density is ρ ∼ − .When the asteroid falls toward the neutron star, it will be broken up at the tidal breakupradius given by R b = (cid:16) ρ r MG / s (cid:17) / , where G is the gravitational constant and s is theshear strength. The subsequent free fall velocities of the leading and lagging fragmentswill be di ff erent (Colgate & Petschek 1981). As a result, the disrupted material is highlyelongated. The duration of the final collision of the material with the neutron star canbe estimated as (Colgate & Petschek 1981; Geng & Huang 2015): ∆ t a = r GMR b ! − / = . × − m / s − / ρ − ! − / M . M ⊙ ! − / s . (1)For an asteroid with m ∼ g, the collision will be completed on ms timescale. Sothe short durations of FRBs can be safely ensured.We assume a small portion ( η R ∼ − ) of the potential energy is emitted in radioband. Then from f E FRB = η R GMm / R NS , we can derive the asteroid mass as m = . × η − , − f − E FRB , R NS , M − . M ⊙ g , (2)where f ∼ × − is the estimated beaming factor of radio emission in our model(Geng & Huang 2015). Note that the isotropic FRB energy is E FRB = π E radio . Just before the final collision, the material is compressed to a dense thin sheet by themagnetic field of the neutron star. The collision will launch a rapidly expanding plas-moid fireball from the surface, leading to a fan of field lines filling with relativistic elec-trons. The thickness of the fireball is ∆ ≈ c τ . The emission volume is V emi ≈ π f ∆ r ,RBsfromNeutron Star-Asteroid Collisions 5where the emission radius r emi can be determined from the balance between the plasmapressure and the magnetic energy density. Electrons radiate coherently in patches withtypical volume given by V coh = (cid:16) /γ (cid:17) r × ( c /ν c ), where ν c is the characteristicfrequency of curvature emission which naturally falls in the radio band. The typicalLorentz factor of electron ( γ ) is estimated as several hundred. The total curvature lu-minosity can then be derived as (Kashiyama et al. 2013) L tot ≈ (cid:16) P e N (cid:17) × N pat , where P e is the emission power of a single electron, N coh is the number of electrons in eachcoherent patch, and N pat is the number of the patches. The collision will lead to a hot region on the neutron star surface, whose temperaturecan reach keV range initially. The hot region then cools down slowly, emitting X-raysin a relatively long period. In fact, much of the released potential energy will be emittedas X-rays. Previous studies indicate that the thermal radiation from such a hot region islikely to decay as a power-law function of time (Lyubarsky et al. 2002). In our study,we have calculated the X-ray afterglow of an FRB by using F X ≈ σ T R / d . Foran FRB occurring at cosmological distance, the decaying thermal emission is usuallymore than 12 magnitudes too dim to be detected by current X-ray detectors. However,note that for other FRB models such as the collapse of hypermassive neutron stars toblack holes (Zhang 2014), the energy released is much larger and X-ray afertglows dueto external shocks are detectable in some optimistic cases (Yi et al. 2014). For a typical neutron star, a strong asteroid collision event may happen every ∼ − years (Mitrofanov & Sagdeev 1990). In the volume of z ≤
1, there are about ∼ late-type galaxies, and each galaxy may contain ∼ neutron stars. As a result, we estimatethat the observable FRBs caused by collision events are ∼ − f − sky − day − .This rate is roughly consistent with that deduced from observations, i.e., ∼ —10 sky − day − (Thornton et al. 2013; Keane & Petro ff
4. Conclusions and Discussion
As a special kind of phenomena that are characterized by very short timescale and highintensity, the newly discovered FRBs are still highly controversial for their burstingmechanism. We suggest that the collision between asteroids and neutron stars canreasonably explain many of the observed features of FRBs, such as the millisecondduration, the energetics, the event rate, etc.Although many authors believe that FRBs are at cosmological distances, the possi-bility that FRBs are actually at non-cosmological distances or even are Galactic eventsstill cannot be excluded yet (Katz 2014; Maoz et al. 2015; Connor et al. 2015). Inter-estingly, our collision model can also explain these non-cosmological FRBs (Geng &Huang 2015). In this case, the asteroids involved will be much smaller.Neutron stars / pulsars are very intriguing objects. They can act as natural labora-tories for some special conditions (Weltevrede et al. 2011; Wang et al. 2014; Chen-namangalam et al. 2015; Gao et al. 2015). For a long time, people have expectedthat the collision of small bodies with neutron stars can give birth to some kinds ofX-ray bursts or some special γ -ray / soft γ -ray bursts (e.g. Colgate & Petschek 1981; Y.F.Huang & J.J.GengCordes & Shannon 2008; Campana et al. 2011). Recently, Huang & Geng (2014) fur-ther suggested that such collisions can also lead to glitches or even anti-glitches. Thein-spiralling merge of a strange quark star with a strange quark planet can even producea very strong gravitational wave burst, which can be used as a unique probe for theexistence of strange quark matter in the Universe (Geng et al. 2015). As shown in thisstudy, FRBs may also be a kind of phenomena associated with neutron star collisionevents. We hope the Chinese FAST telescope (Nan et al. 2011) would be a powerfultool to study the multidimensionality of such collisions.
Acknowledgments.
This study was supported by the National Basic Research Pro-gram of China with Grant No. 2014CB845800, and by the National Natural ScienceFoundation of China with Grant No. 11473012.
References
Barrau A., Rovelli C., & Vidotto F. 2014, Phys. Rev. D, 90, 127503Burke-Spolaor S., & Bannister K. W. 2014, ApJ, 792, 19Campana S., et al. 2011, Nature, 480, 69Chennamangalam J., et al. 2015, New Astronomy, 34, 245Colgate S. A., & Petschek A. G. 1981, ApJ, 248, 771Connor L., Sievers J., & Pen U.-L. 2015, arXiv:1505.05535Cordes J. M., & Shannon R. M. 2008, ApJ, 682, 1152Falcke H., & Rezzolla L. 2014, A&A, 562, A137Gao Z. F., et al. 2015, MNRAS submitted (arXiv:1505.07013)Geng J. J., & Huang Y. F. 2015, ApJ, 809, 24Geng J. J., Huang Y. F., & Lu T. 2015, ApJ, 804, 21Huang Y. F., & Geng J. J. 2014, ApJL, 782, L20Inoue S. 2004, MNRAS, 348, 999Ioka K. 2003, ApJL, 598, L79Kashiyama K., Ioka K., & Mészáros P. 2013, ApJL, 776, L39Katz J. I. 2014, Phys. Rev. D, 89, 103009Keane E. F., et al., 2011, MNRAS, 415, 3065Keane E. F., & Petro ff E. 2015, MNRAS, 447, 2852Kulkarni S. R., et al., 2014, ApJ, 797, 70Loeb A., Shvartzvald Y., & Maoz D. 2014, MNRAS, 439, L46Lorimer D. R., et al., 2007, Science, 318, 777Lyubarsky Y. 2014, MNRAS, 442, L9Lyubarsky Y., Eichler D., & Thompson C. 2002, ApJL, 580, L69Maoz D., et al. 2015, MNRAS, submitted (arXiv:1507.01002)Mitrofanov I. G., & Sagdeev R. Z. 1990, Nature, 344, 313Mottez F., & Zarka P. 2014, A&A, 569, A86Nan R. D., et al. 2011, International Journal of Modern Physics D, 20, 989Pen U.-L., & Connor L. 2015, arXiv:1501.01341Petro ffff