Arxiv:2002.11975V2

Home , Dwarf star

arXiv:2002.11975v2 [astro-ph.HE] 9 Apr 2020 rglci ngas a ( far so enigma tragalactic 1 2 2013 2016 2020 ( bursts repeating 2016 into divided nomenally ( rarely peaks 2016 multiple or mainly luminosity peak peak the with of pulses limits radio upper coherent average millisecond-duration are by discovered were they since 2018 ua ta.2019 al. et Kumar al. et Keane atojcshv enpooe ( com- proposed in involved been have models literature progenitor objects the of pact dozens in Catalogue FRB ena, reported the been in collected have and FRBs 100 × × 1 atrdobrt FB)hv eanda ex- an remained have (FRBs) bursts radio Fast yee sn L using 2020 Typeset 10, April version Draft http://www.frbcat.org 10 2 1 10 .Mawie oepanti ai phenom- radio ; this explain over to Meanwhile, date, To ). bursts. non-repeating and ) ; ; ; e aoaoyo oenAtooyadAtohsc (Nanji Astrophysics and Astronomy Modern of laboratory Key colo srnm n pc cec,NnigUniversity, Nanjing Science, Space and Astronomy of School 42 aa ta.2018 al. et Farah 42 erffe l 2019 al. et Petroff ihnalrernei oa niomnsadhs galaxies. host Keywords: F and environments and local 180924-like in FRB range of large properties a NSWD within observational This the channels. hos both BNS/BWD/AIC 180924- and of explain FRB environments, those of local resemble subset mergers ejecta, a NSWD the least since at me magnetars, NSWD produce flaring If a could field. rotation objects fossil differential such a to of then scenario t due conservation in mechanism flux result magnetic dynamo could the the transfer via mass field unstable with magnetic mergers NSWD that find NW)mres oiae yFB102-iefs ai uss( bursts radio by fast suggested channels 180924-like BNS/BWD/AIC in FRB prod born be by also motivated could mergers, magnetars whether (NSWD) investigate accretion-ind we or paper, mergers, this (BWD) dwarf In white binary or (BNS) star antr rmNurnSa–ht wr egr:Applicat Mergers: Dwarf Star–White Neutron from Magnetars ace&Rzol 2014 Rezzolla & Falcke − HM/R olbrto ta.2019a al. et Collaboration CHIME/FRB r ( erg ti ieyblee htmgeascudb oni oeclas sup core-collapse in born be could magnetars that believed widely is It 8 × ( 2012 hn 2018 Zhang 10 A T 1. E 44 X ,and ), (1339) ; opc iaysas(8) rvttoa ae 68;Magnet (678); waves Gravitational (283); stars binary Compact INTRODUCTION r s erg twocolumn HM/R olbrto tal. et Collaboration CHIME/FRB ; − ; hrtne al. et Thornton rcak ta.2019 al. et Prochaska 1 ,caatrzdb single a by characterized ), ods&Catre 2019 Chatterjee & Cordes n energy and az2018 Katz tl nAASTeX63 in style oie tal. et Lorimer ; ybrk 2014 Lyubarsky ahym tal. et Kashiyama h-igZhong Shu-Qing hmine al. et Champion 1 E ; ( ( ∼ 2013 erffe al. et Petroff pte tal. et Spitler oo tal. et Popov 7 × .They ). ,phe- ), ( 10 L 2007 p 39 ABSTRACT , b ∼ − ), ) aglte al. et Margalit ; ; ,2 1, ajn 103 hn;szoghtalcm [email protected] [email protected], China; 210093, Nanjing gUiest) iityo dcto,Nnig209,Chi 210093, Nanjing Education, of Ministry University), ng and enpooe ogv iet o-eetn rre- or non-repeating ( to respectively have rise flares give bursts, clustered to peating or proposed single been with magnetar flaring dr eid”btenbrt n lsee us ar- burst clustered and bursts between periods” “dark measure rotation polarization, as properties such observational (RM; all FRBs nearly of explain cessfully a onepr ( counterpart cal ( 2019 2017 al. et Kumar al. et Platts 2019 Beloborodov 2017 al. et Metzger 2016 2015 Huang & Geng 2017 2016 2018 2016 eetn Rs ulmdlls a ee to refer can list model full A FRBs. repeating esl ta.2019 al. et Hessels iGoDai Zi-Gao 2 n neetn ru fmdl eeatt young a to relevant models of group interesting One http://frbtheorycat.org ,cru-us iprinmaue(M,adthe and (DM), measure dispersion circum-burst ), suc- to developed been has them of One ). ; ; ; ; ihlie l 2018 al. et Michilli aglt&Mtgr2018 Metzger & Margalit eoooo 2017 Beloborodov age l 2016 al. et Wang ( hn 2016 Zhang 2019 aaiso h nlrmat from remnants final the of galaxies t cdclas AC fwiedwarfs. white of (AIC) collapse uced cdfo eto trwiedwarf star–white neutron from uced ieFB ihntefaeokof framework the within FRBs like ( ,2 1, .B rlmnr aclto,we calculation, preliminary a By ). gr a nedcet magnetars, create indeed can rgers 2019 B101.05+5lk FRBs 180916.J0158+65-like RB Rs osbyfo magnetars from possibly FRBs) eN curn nultra-strong an acquiring NS he htejee l 2017 al. et Chatterjee roa Se,bnr neutron binary (SNe), ernovae ,acutn o o-eetn and/or non-repeating for accounting ), dcneto rpsil via possibly or convection nd ) ; 2 o oFs ai Bursts Radio Fast to ion ,pritn ai oreadopti- and source radio persistent ), . ; u&Kmr2018 Kumar & Lu hne sas bet well to able also is channel , ; eoooo 2017 Beloborodov r 92;Rdobursts Radio (992); ars 2017 ; , ,feunydwwr drift downward frequency ), a ta.2016 al. et Dai hn 2014 Zhang 2019 ; ybrk 2014 Lyubarsky ; ; ytkve l 2016 al. et Lyutikov eze ta.2017 al. et Metzger eze ta.2019 al. et Metzger ; edla tal. et Tendulkar ; ; eze tal. et Metzger ; uaee al. et Murase ; ege al. et Deng ue al. et Gu .cn na ; Katz ; ; ; 2 Zhong & Dai rival times appearing in FRB 121102 and its hosted low tron star–white dwarf (NSWD) mergers. This channel metallicity dwarf star forming galaxy and its surround- has also been briefly mentioned and/or discussed by ing dense, highly magnetized, and dynamic plasma envi- Liu (2018, 2020), Khokhriakova & Popov (2019), and ronment (Chatterjee et al. 2017; Tendulkar et al. 2017) Beloborodov (2019) previously. To answer this question, within the framework of synchrotron maser emission we need to study (1) whether this channel can form mag- from decelerating relativistic blast waves produced by netars or not, and (2) if it can, whether the formed mag- flare ejecta from young magnetars (Metzger et al. 2019). netars can account for the observations of FRB 180924- In this framework, repeating FRBs similar to FRB like bursts in the flaring magnetar framework. If this 121102 are arise from young and very active millisec- channel can indeed form magnetars, it could increase ond magnetars quite possibly connected with superlumi- the magnetar event rate to some extent and contribute nous supernovae (SLSNe) and long-duration gamma-ray to at least a subset of FRBs similar to FRBs 180924, bursts (LGRBs; Metzger et al. 2017). Therefore, young 190523, 181112, and even 180916.J0158+65. In the fol- magnetars giving rise to FRB 121102-like FRBs might lowing, we organize the structure of the paper: §2 intro- be formed during the core-collapse of massive stars as- duces the possibility and speculation that NSWD merg- sociated with SLSNe or LGRBs (SLSNe/LGRBs chan- ers could form magnetars; whether or not the NSWD nels). channel can explain the observations of FRB 180924-like On the other hand, young millisecond magne- cases is discussed in §3; and a summary and discussion tars could also be born in binary neutron star are presented in §4. (BNS) mergers (Rosswog et al. 2003; Price & Rosswog 2006; Giacomazzo & Perna 2013), binary white dwarf 2. MAGNETARS FROM NSWD MERGERS (BWD) mergers (King et al. 2001; Yoon et al. 2007; The explosive outcomes of NSWD mergers have Schwab et al. 2016), or accretion-induced collapse been explored in the literature (Metzger 2012; (AIC) of white dwarfs (WD; Nomoto & Kondo 1991; Margalit & Metzger 2016, 2017; Zenati et al. 2019a,b; Tauris et al. 2013; Schwab et al. 2015). These mag- Fernández et al. 2019), but the final remnants netars could produce FRBs analogous to FRB 180924 of these events have been little investigated (see (Bannister et al. 2019), as suggested by Margalit et al. Paschalidis et al. 2011a,b; Margalit & Metzger 2016). (2019). Compared with FRBs created from magnetars Generally, there are two evolutionary pathways for born in SLSNe/LGRBs channels, the FRBs produced NSWD binaries, which depend on the critical mass from magnetars born in BNS/BWD/AIC channels could ratio q = M /M , where M is the crit- have a distinct observational properties. Just like FRBs crit WD,crit NS WD,crit ical WD mass and M is the NS mass. The first 180924 and 190523 (Ravi et al. 2019), likely as well as NS pathway is that the WD fills its Roche lobe and its FRB 181112 (Prochaska et al. 2019), they host an old matter undergoes stable mass transfer to the NS if massive galaxy with a relatively low rate of star for- q < q , evolving into an ultra-compact X-ray bi- mation and relatively high metallicity, lie in a large crit nary. The second pathway is that the WD is tidally spatially offset location relative to the central contain- disrupted by the NS via unstable mass transfer on a ment region of the galaxy, and have low DM and RM rather short dynamical timescale for q > q , lead- contributions from the host galaxy and no bright per- crit ing to an NSWD merger (Hjellming & Webbink 1987; sistent radio source (Bannister et al. 2019; Ravi et al. Hurley et al. 2002). The critical mass ratio is related 2019). If this is the case, FRBs could be divided into two to the critical WD mass M , which is found to populations: FRB 121102-like bursts stem from young WD,crit be M = 0.37 M (van Haaften et al. 2012) or magnetars born in SLSNe/LGRBs channels, while FRB WD,crit ⊙ M = 0.2 M (Bobrick et al. 2017). Thus an 180924-like cases come from those young magnetars WD,crit ⊙ NSWD merger with q > q is in the case of un- born in BNS/BWD/AIC channels. Additionally, most crit stable mass transfer. Toonen et al. (2018) pointed FRB 180924-like bursts should also be repeating in out that over 99.9% of semi-detached NSWD systems the flaring magnetar framework due to the event rate would merge when M =0.2 M , which indicates comparison between magnetars and total FRB events WD,crit ⊙ that the NSWD merger is a prevalent fate of semi- (Nicholl et al. 2017; Margalit et al. 2019; Ravi 2019), detached NSWD binaries. After an NSWD merger, as which is supported by FRB 171019 followed by faint shown by Paschalidis et al. (2011a,b), the final rem- bursts (Kumar et al. 2019). nant both in the inspiraling case and in the head-on In this paper, we investigate whether or not FRB case is a spinning Thorne-Zytkow-like object (TZIO; 180924-like bursts are also likely to be generated by Thorne & Zytkow 1977) surrounded by a massive ex- magnetars born in an alternative possible channel: neu- tended hot disk composed of WD debris without ex- FRBs from Magnetars Formed from NSWD Mergers 3 plosive outcomes. Considering the disk winds and nu- ical timescale ∼ 3 × 104 s. In this case, the accretion clear burning, Margalit & Metzger (2016) suggested an of the final NS surrounded by a massive disk may let NSWD merger likely forms an isolated millisecond pul- it differentially and rapidly rotate, as possibly shown sar surrounded by an accretion disk at the final stage. by the simulation results of Paschalidis et al. (2011b), Whether or not these final remnants evolve into mag- even if Paschalidis et al. (2011b) did not take into ac- netars has received less attention. How the magnetic count explosive outcomes. Moreover, during this short fields of the final remnants evolve remains unknown. dynamical timescale, transient ohmic dissipation of the Fortunately, it has been suggested that the ultra-strong final NS could be ignored, see an estimate in Sun et al. magnetic fields in magnetars may result from two main (2019) and Equation (9) of Urpin & Konenkov (1997). scenarios (for a review see Turolla et al. 2015). Al- Therefore, we just need to focus on the magnetic field though these two scenarios are mainly used for nascent amplification of the final NS that accretes the WD de- NSs born in SLSNe/LGRBs/BNS/BWD/AIC channels, bris material from the disk during the merger. Owing we guess that they might also be used for “renascent” to the lack of previous studies of the magnetic field (magnetic field undergoes amplification) NSs formed in evolution of NSWD mergers with unstable and rapid the NSWD channel. mass transfer, we perform only a preliminary analysis on the α − ω dynamo induced by possible differential 2.1. α–ω Dynamo rotation and/or convection that can amplify the NS’ magnetic field during the NSWD mergers. In the fi- The first scenario that we consider is magnetic field nal paragraph of this subsection, we would also discuss amplification by a vigorous dynamo action at the the dynamo induced by the magneto-rotational instabil- early, highly convective stage of the NSs after merg- ity (MRI; Balbus & Hawley 1998) in the disk that also ers: the α dynamo arising from the coupling of con- might contribute to the NS’ magnetic field amplification. vective motions and rotation, and the ω dynamo driven We assume that the final NS with differential rotation by differential rotation (Duncan & Thompson 1992; induced by accretion has two components: the core and Thompson & Duncan 1993). For an NSWD merger, ∼ the shell divided by a boundary at the radius Rc = 0.5R the magnetic field of the NS remnant surrounded by a (where R is the NS radius), as done by Dai et al. (2006), massive disk may either increase via a dynamo winding- its accretion can be generally determined by the Alfvén up process (as suggested by Paschalidis et al. 2011b) or radius decrease through an enhanced ohmic dissipation of ac- 3 4/7 −1/7 −2/7 creted matter in the NS’ crust (Urpin & Konenkov 1997; rm = (BsR ) (GM) M˙ Konar & Bhattacharya 1997; Cumming et al. 2004), 4/7 12/7 −1/7 Bs R M somewhat similar to the finding of Sun et al. (2019). =5.3 1012 G 12 km 1.4 M We discuss whether this finding is true in the follow- ⊙ −2/7 ing. Bobrick et al. (2017) showed that NSWD merg- M˙ × km, (1) ers exhibit an exponentially increasing rate of mass −6 −1 10 M⊙ s ! transfer during different phase transitions (see Figure

12 in Bobrick et al. 2017), and the NSWD mergers where Bs, R, M, and M˙ are the surface magnetic dipole in which the WDs have a higher mass would have a field strength, radius, mass, and accretion rate of the shorter dynamical timescale between the onset of sig- NS, respectively, and the corotation radius nificant mass transfer and the final merger (e.g., only 1/3 −3 −5 4 2 1/3 2 tdyn ∼ 10 (10 ) yr ∼ 3 × 10 (3 × 10 ) s for WD GM GMPs rc = = mass ∼ 0.75(1.2) M , see Figure 13 in Bobrick et al. Ω2 4π2 ⊙ s 2017). However, it is not true that the maximal rate 1/3 2/3 3 M Ps of disk accretion onto the final NS in Figures 11 and =7.8 × 10 km, (2) 1.4 M⊙ 10 s 12 of Bobrick et al. (2017) is limited by the Edding- ton rate since the disk accretion of NS can be highly where Ωs = 2π/Ps and Ps are the angular velocity and super-Eddington. If the majority of mass is lost via a spin period of the NS’ shell, respectively. One additional disk wind or possible a jet in the mass transfer pro- key radius is the light cylinder radius, cess and only 0.05 M can be accreted onto the NS, ⊙ P as discussed in Margalit & Metzger (2016) for a WD r = c/Ω =4.8 × 105 s km. (3) lc s 10 s with mass 0.6 M⊙ (close to 0.75 M⊙), the NS would accrete the disk material onto its surface with an aver- One expects that in the case of rm < rc < rlc for a −6 −1 12 age rate M˙ ∼ 10 M⊙ s during the short dynam- normal NS in an NSWD merger with Bs = 10 G, 4 Zhong & Dai

M = 1.4 M⊙, R = 12 km, and Ps = 10 s, as well where Ic is the moment of inertia of the core. Through −6 −1 as an accretion rate M˙ = 10 M⊙ s , disk mate- Equations (4), (8), and (9), we can solve Ωs, Ωc, ∆Ω, rial is column-accreted onto the NS’ surface and leads and Bφ as illustrated in Figure 1 via numerical cal- to the NS’ shell to spin up (Frank et al. 1992). Follow- culation, combining Equations (5) and (7). To obtain ing Piro & Ott (2011) and Dai et al. (2006), the time- these results, we have also employed: (1) typical values dependent angular velocity for the NS’ shell can be for a normal NS in an NSWD merger: M = 1.4 M⊙, 12 solved by R = 12 km, initial period Ps,0 = 10 s, Bs = 10 G, I =∼ I ∼ 1045 g cm2, ǫ = 0.3 (Dai et al. 2006), as dΩs s c I = N − N − N , (4) ˙ −6 −1 s dt acc dip mag well as an accretion rate M = 10 M⊙ s , such that the term N can be ignored in comparison with N where I is the moment of inertia of the shell, and (1) dip acc s even if P possibly reaches down to its break-up limit N is the accretion torque described by, when r < R s acc m P = 0.96 ms (Lattimer & Prakash 2004); (2) ini- for a normal NS from Equation (1), s,min tial conditions: Ωs,0 = 2π/Ps,0, Ωc,0 = (1+ A0)Ωs,0 Ωs (the initial angular velocity of the core Ω should be N = 1 − (GMR)1/2M,˙ (5) c,0 acc Ω larger than that of the shell Ω for a normal NS), K s,0 A = 10−3 related to a small residual differential ro- 3 1/2 0 where ΩK = GM/R ; (2) Ndip is the magnetic 8 tation, and Bφ,0 ∼ 10 G; (3) boundary conditions: dipole radiation torque for accreting NSs with rm < rlc, Ωs ≤ Ωs,max = 6541 due to Ps,min = 0.96 ms, Ωc < enhanced over the standard dipole torque by a factor Ω = c/R , and B < B = 1017 G because of 2 c,max c φ φ,max of (rlc/rm) > 1 due to the enhanced open magnetic the buoyancy effect. From Figure 1, we can acquire the field lines via the compression of the magnetosphere following: (Parfrey et al. 2016; Metzger et al. 2018), read as

2 6 3 2 • The top panel shows that the angular velocity 2B R Ω rlc N = s s of the shell gradually increases up to its limit at dip 3c3 r m about 200 s and lasts this lasts until the end of 2Ωs 3 6/7 2/7 4/7 = (BsR ) (GM) M˙ ; (6) the dynamical process. While the angular velocity 3c of the core reverses (rotating in an opposite direc- (3) Nmag is the magnetic torque acting between the core tion) at about 50 s and its absolute value rapidly 4 and shell, written as rises to a very large value c/Rc at around 1000 2 s. This is because the toroidal magnetic field Bφ N = R3B B , (7) mag 3 c r φ reverses, as in the bottom panel. The evolution of the differential angular velocity ∆Ω follows the where Br = Bs/ǫ (here, ǫ is defined by the ratio of the effective surface dipole field strength to the radial field angular velocity of the core. strength) and Bφ are the radial magnetic field compo- • nent and toroidal field component, respectively. They In the bottom panel, the toroidal magnetic field can be related to each other by Bφ rapidly declines to zero and then reverses before 0.1 s, its absolute value continues going up to dB 17 φ = (∆Ω)B ≡ (Ω − Ω )B , (8) the buoyancy limit 10 G at about 150 s. dt r c s r where ∆Ω is the differential angular velocity and Ωc is In short, these results manifest the toroidal magnetic the NS’ core angular velocity. On the other hand, the field can be enhanced during the dynamical timescale time-dependent angular velocity for the NS’ core can be as long as the initial remnant NS in an NSWD merger solved by3 has a small residual differential rotation. Additionally, during the field amplification, the spin-down torque Ndip dΩc Ic = Nmag, (9) responds to the magnetic dipole radiation luminosity dt

Ldip = NdipΩs 3 Please note that the right term of Equation (9) has no negative 2 sign, differing from Equation (3) in Dai et al. (2006) because the 2Ω = s (B R3)6/7(GM)2/7M˙ 4/7, (10) magnetic field is amplified via the differential angular velocity 3c s resulting from the angular momentum transport of the accreting material onto the NS’ shell rather than the angular momentum 4 loss of the core for an accreting NS system we consider here. The The core angular velocity Ωc that exceeds the break-up limit of same reason that the right term of Equation (4) in this paper the shell is reasonable since this break-up limit should not be that differs from Equation (2) in Dai et al. (2006). of the core angular velocity. Instead, its limit should be c/Rc. FRBs from Magnetars Formed from NSWD Mergers 5

1 1 1 1 1 1 Δ

11 1 1 1 1 1 1 1 1

1 11 L 1 1

1 1

1 1 1 11 1 11 1 1 1 1 11 1 11 1 1 1

t t

Figure 2. Magnetic dipole radiation evolution of the NS in 1 an NSWD merger during the dynamical timescale.

1 This type of convection usually occurs in a nascent NS left behind the collapse of a massive star, a BNS/BWD merger, or an AIC of a WD with a negative radial

1 entropy gradient from the interior to the outer layers (Thompson & Duncan 1993), but should not occur in 1 the old NS in an NSWD system. However, for the ac- 1 creting NS in an NSWD merger, its surface should be 11 covered by an accreting envelope with tidal WD de-

11 bris via magnetically channeled accretion. Under this condition, the accretion flow can produce heat radi- 11 1 11 1 11 1 1 1 ation due to the shock heating (see the appendix in t 9 Piro & Ott 2011), its temperature is Tsh ∼ 8 × 10 K, if M = 1.4 M , R = 12 km, B = 1012 G, and Figure 1. Evolution of the angular velocities of the shell ⊙ s ˙ −6 −1 and the core and the differential angular velocity ∆Ω (top M = 10 M⊙ s are considered (for a detailed deriva- panel), and the torodial magnetic field Bφ (bottom panel) of tion please refer to Equation (A1) in Zhong et al. 2019). the NS in an NSWD merger during the dynamical timescale Therefore, the shock heat cannot escape via neutrino 4 tdyn = 3 × 10 s. cooling (since a low temperature cannot induce neutrino emission) or photon diffusion (since photons are which follows the evolution of the angular velocity of the trapped and advected due to the high accretion rate). shell, as displayed in Figure 2. This Poynting flux could Throughout the large and radiatively inefficient accret- generate a high energy (X-ray/γ-ray) transient lasting ing envelope, convection may be an important source hundreds to thousands of seconds via magnetic dissi- of outward energy transport (e.g. Quataert & Gruzinov 46 −1 pation with brightness up to Ldip,peak ∼ 10 erg s , 2000). In this case, the energy flux due to convection which is likely similar to an X-ray transient source in the absence of bulk of motion or angular momentum named CDF-S XT2 discovered by Xue et al. (2019). transport should be from gravitational potential energy ˙ 3 Besides the differential rotation, Duncan & Thompson flux Fc ∼ GMM/(4πR ). If the magnetic field of the (1992) and Thompson & Duncan (1993) suggested NS can be enhanced by this convection, the magnetic that a key parameter for the success of α–ω dy- field could reach to its buoyancy value 1017 G (please 2 ˙ 3 namo is the Rossby number RO relevant to convec- note that Bc /(8π) < Fctdyn ∼ [GMM/(4πR )]tdyn and 20 tion (the ratio of the rotation period to the convec- thus Bc < 1.0 × 10 G) under the parameter values ˙ −6 −1 tive overturn time). An efficient dynamo result needs M = 1.4 M⊙, R = 12 km, M ∼ 10 M⊙ s , and 39 −2 −1 1/3 4 RO ∼ 10(P/10 ms)(F/10 erg cm s ) ≤ O(1) tdyn = 3 × 10 s, although the real circumstance could (where F is the entropy-driven convection heat flux). be more complex. 6 Zhong & Dai

Differing from the differential rotation and convection suitable at least for a small fraction (a few times 0.1%) processes that occur in the accreting NS of the NSWD of NSWD mergers, since the supernova producing NS merger, an alternative potential process, MRI, possibly precedes the NSWD merger (in which the WD forms be- occurs in the disk due to the presence of shear, like fore the NS for the majority of NSWD mergers) by less the disk dynamo in common envelope events for the than 100 years, as suggested in Toonen et al. (2018). formation of highly magnetic WDs (Tout et al. 2008; Whether or not these scenarios can enhance the mag- Nordhaus et al. 2011). This process amplifies the disk netic fields of final NSs post NSWD mergers still lacks field, and the field then would be conveyed to the surface evidence both in numerical simulations and observa- of the NS along with the radiatively inefficient magnet- tions. Accordingly, for the problem of the fate of the ically channeled accretion flow. The disk dynamo for remnants’ magnetic fields post NSWD mergers, deep WDs in the super-Eddington regime (Nordhaus et al. and complex exploration and magnetohydrodynamics 2011) should also be suitable for the accretion disk in the simulations for the magnetic field evolution of the rem- NSWD merger. Accordingly, based on Blackman et al. nants are required. (2001) and Nordhaus et al. (2011), the mean toroidal field via the MRI in the disk at radius r is estimated by 3. EXPLANATIONS OF OBSERVATIONAL

1/2 PROPERTIES OF FRB 180924-LIKE BURSTS M˙ ΩK r B¯φ ∼ If FRB 180924-like bursts can be produced from mag- r H ! netars born in NSWD mergers, this NSWD channel 1/2 should be able to explain all of the observational prop- M˙ M 1/4 =6 × 1012 erties of this FRB population, as well as the event 10−6 M s−1 1.4 M ⊙ ! ⊙ rate, host galaxy and offset, and circumburst environ- r −5/4 r/H 1/2 ment. Due to the similarities between the NSWD chan- × G, (11) 12 km 2 nel and BNS/BWD/AIC channels, we mainly follow Margalit et al. (2019) to analyze and discuss this NSWD 3 1/2 where ΩK = (GM/r ) is the Keplerian rate the disk channel, which is shown as follows. orbits, H is the isothermal scale height of the disk. This (1) Active Lifetime. The mass of magnetars formed process can enhance the magnetic field of the NS but from NSWD mergers, Mmag, could be smaller than or could not result in a magnetar-like field. close to the maximal mass of a non-rotating NS MTOV, given a critical WD mass MWD,crit = 0.37 M⊙ or Fossil Field 2.2. MWD,crit = 0.2 M⊙ for unstable mass transfer, and a The second scenario that we consider is the canonical NS mass MNS =1.4 M⊙. However, these mag- magnetic flux conservation–fossil field scenario netars should have a lower mass than those born in the (Ferrario & Wickramasinghe 2006, 2008), which sig- BNS channel, but likely a higher mass than those born in nifies that the magnetic fields of the NSs and/or the SLSNe/LGRBs or BWD/AIC channels. Moreover, the WDs in NSWD binaries should be stronger than normal mass Mmag should also exceed or approach to the thresh- NSs and/or WDs. Although it is hard to imagine that old mass for the onset of direct or modified URCA neu- the progenitors in binary compact stars possess a very trino cooling (Beloborodov & Li 2016). As pointed out strong magnetic field since such strong magnetic fields by Margalit et al. (2019), such magnetars may possess likely decay on much shorter timescales ∼ 104−5 yr sufficiently high central densities (or high temperatures) (Heyl & Kulkarni 1998; Harding & Lai 2006) than the to activate URCA cooling in their cores. Otherwise, merger lifetime. However, there may be some spec- their magnetic dissipation in the core is caused predom- ulating clues. For instance, the precursor of GRB inately by ambipolar diffusion (Goldreich & Reisenegger 090510 is likely related to a magnetar-like magnetic 1992; Thompson & Duncan 1996), which is sensitive to field (B > 1015 G in Troja et al. 2010) of the NS in the core temperature. Since the core temperature de- the progenitors if the precursor stems from the mag- pends on its URCA cooling at early times of mag- netospheric interaction of the NSs. Furthermore, the netar formation, their magnetic activity timescale in merger lifetimes can also be much shorter than the in- the direct URCA cooling (high-mass NS) and mod- spiral times for a sizable fraction of double NS mergers ified URCA cooling (normal-mass NS) can be esti- −1 3/2 −1.2 1.6 in some population synthesis models (Belczynski et al. mated as tmag ∼ B16 L5 20 yr (700B16 L5 yr) for 2002, 2006). After the merger lifetimes, the magnetic high-mass NS (normal-mass NS; see Equation (33) of fields should have decayed by only a factor of a few, as Beloborodov & Li 2016). This would correspond to a illuminated in Troja et al. (2010). This result should be magnetic energy dissipation with an average luminos- FRBs from Magnetars Formed from NSWD Mergers 7

40 3 −1 39 3.2 −1 ity Lmag ∼ 5 × 10 B16 erg s (10 B16 erg s ) (Margalit et al. 2019) for high-mass NS (normal-mass NS; see Equation (2) of 3M Margalit et al. 2019), which is just lower than the peak ej −3 −2 −2 DMej ≈ 2 ≈ 5 pc cm Mej,−1βej tyr , (12) luminosities of FRBs by two to four orders of magnitude 8πmp (vej t) (Zhang 2018). where Mej,−1 = Mej/0.1 M⊙, βej = vej/c and tyr = During the active lifetimes of magnetars, they there- t/1 year. For ejecta post NSWD mergers, Mej ∼ 0.01 − fore produce ∼ 100 repeating bursts resulting from rel- −3 −2 0.1 M⊙ and vej ∼ 0.1c, so DMej ∼ 50−500 pc cm tyr . ativistic blast waves caused by giant flares with lumi- If the radio frequency is transparent after one month, nosity higher than Lmag by several orders of magni- 3−4 −3 DMej would decrease to 5 × 10 pc cm . For tude, enough to satisfy the event rate of FRBs (Ravi the case FRB 180924, it has a mean contribution by 2019). Because the active lifetimes are several tens to −3 host galaxy DMhost 30 − 81 pc cm (Bannister et al. several hundreds of years, the “dark period” between 2019). The contribution by its ejecta should be bursts can average years to several tens of years; there smaller than DMhost. If so, FRB 180924 should es- should be have different local DMs between bursts even cape from the ejecta at least one year after the mag- though they stem from the same source. Under this con- netar is born in the NSWD channel, in which time dition, we might regard them as different bursts from the radio frequency is already transparent. Note different sources rather than repeating bursts from the that the most repeating FRBs have a nearly invari- same source, e.g., two possible cases: FRBs 110220 and able DM for long-term observations (Spitler et al. 140514 (Piro & Burke-Spolaor 2017), and FRBs 160920 2016; CHIME/FRB Collaboration et al. 2019a,b; and 170606. Therefore, it is easy to understand that Kumar et al. 2019; CHIME/FRB Collaboration et al. repeating bursts of FRB 180924 have not been detected 2020) signifying that the DM contribution from their during a relatively short follow-up observation. This is ejecta has declined close to zero several years to several also supported by the bright FRB 171019 followed by hundreds of years after their magnetar formation. faint bursts (Kumar et al. 2019). (4) RM. According to Margalit et al. (2018), the max- Burst Transparency. (2) In the framework of flaring imal contribution to the RM is primarily caused by magnetars, bursts can escape only when the surround- a nebula in which the cooled electrons and magnetic ing material is free-free transparent for radio frequency field injected by magnetar flares in the distant past ∼ GHz. Similar to BNS/BWD mergers and AIC, there and confined by the SN ejecta (Beloborodov 2017; should also be ejecta surrounding the “renascent” mag- Margalit & Metzger 2018). It is given by netars for the NSWD channel, which may also give rise to observable explosive transients. The ejecta consists e3 3e3 N B λ 1/2 RM = n B ds ≈ e n , of the WD debris disk and the accretion-driven outflow 2πm2c4 e k 8π2m2c4 R2 R 4 −1 e e n n with velocity extending up to ∼ 3 × 10 km s =0.1c, Z (13) with overall low mass of 0.01 − 0.1 M⊙ (Zenati et al. where the total number of electrons in the nebula 2019b). If the free-free optical depth of ejecta for which the temperature and ionization state are gov- Ne = ξEmag, (14) erned by photo-ionization due to spin-down of the “renascent” magnetar, as handled in Margalit et al. (2019), the magnetic field strength in the nebula 2/5 −1 the free-free transparency time could be tff ∝ M v 1/2 (1−α)/2 ej ej 6ǫBEmagabs(α − 1) t for a fixed ionization fraction (see Equation (18) in Bn ≈ , (15) R3 t Margalit et al. 2018). This result should be compa- n mag rable to BNS mergers, e.g., Mej ∼ 0.05 M⊙ and and the nebula size Rn is set by the outer ejecta radius vej ∼ 0.2c inferred from kilonova emission accompany- ing GW170817 (Cowperthwaite et al. 2017; Kasen et al. Rn = vejt, (16) 2017; Villar et al. 2017). Hence, FRBs can pass through and λ is correlation length-scale of the magnetic field the ejecta quickly and escape in just about a few weeks in the nebula, ǫ is the ratio of the magnetic energy in to months post magnetar formation, compared with B the nebula to the magnetic energy injected in relativistic t ∼ 10 − 100 yr for SLSNe (Margalit et al. 2018). ff particles over an expansion time t, α is the decay index (3) Circumburst DM. The circumburst DM contri- related to the average magnetic luminosity of the mag- bution of ejecta to the burst should be akin to that netar, and ξ is the average ratio of the number of ejected of BNS/BWD/AIC channels, can be calculated by baryons to the released magnetic energy (Beloborodov 3 −1 2017). If given ǫB = 0.1, ξ = ξmax ≈ 4 × 10 erg , 8 Zhong & Dai

49 Emag ≈ 3 × 10 erg (Beloborodov & Li 2016), vej = (7) Event Rate. The volumetric event rate of NSWD 4 −3 −1 0.1c, λ ∼ Rn, and α = 0, motivated by magnetic- mergers is in a range of (0.5 − 1) × 10 Gpc yr dissipation-powered FRB models (Margalit et al. 2018) in the local universe, which is ∼ 2.5 × 103 − 1 × for a magnetar formed from the NSWD channel, its RM 104 times more than that of the observed LGRBs −2 −1 is just ∼ 12 rad m at the time t ∼ 10 tmag, as- (Thompson et al. 2009) but roughly lower than that of suming tmag ∼ 100 yr. Moreover, its RM decreases FRBs by one order of magnitude (Nicholl et al. 2017). with time. These results are generally consistent with Khokhriakova & Popov (2019) obtained a roughly con- Margalit et al. (2019) and the RM observation of FRB sistent result that the total rate of NSWD mergers is 180924 (Bannister et al. 2019). ∼ 850 sky−1day−1 using cosmic star formation his- (5) Persistent Radio Source. The persistent radio tory from Madau & Dickinson (2014), which is approxi- emission arising from the NSWD channel should also mately lower than the rate of FRBs ∼ 103−4 sky−1day−1 be analogous to that from BNS/BWD/AIC channels, (Cordes & Chatterjee 2019) by one order of magnitude. due to the similar properties of their nebula and ejecta. However, the fraction of NSWD mergers generating Accordingly, there is no evidence for persistent radio magnetars is very unclear. If this fraction is compa- emission in FRB 180924 that can be easily understood rable to that of BNS mergers, i.e., 3%, as estimated in using synchrotron radiation in the nebula confined by Nicholl et al. (2017), the rate of magnetars formed from the ejecta, based on Figure 3 of Margalit et al. (2019). NSWD mergers is approximately 150 − 300 Gpc−3yr−1 (6) Host Galaxy. Metzger (2012) suggested that which is comparable to the overall rate of millisecond NSWD mergers involving pure-He WDs could be related magnetars born in SLSNe/LGRBs and SGRBs—BNS to faint type Ib Ca-rich SNe, which mostly explode in channels, i.e., few 10 − 100 Gpc−3yr−1 in Nicholl et al. early-type galaxies and old environments (Perets et al. (2017). If this is the case, magnetars formed from the 2010; Kasliwal et al. 2012; Lyman et al. 2013). On NSWD channel can contribute to at least a subset of the contrary, those mergers relevant to C/O or hybrid FRB 180924-like bursts. Due to a large uncertainty C/O/He WDs are likely associated with the transients of the magnetar formation rate in the NSWD channel, most similar to SNe Ic (Toonen et al. 2018; Zenati et al. magnetars formed from this channel are also required to 2019b). Moreover, Toonen et al. (2018) showed that emit several bursts over their lifetimes, especially their only a small fraction are expected to be found in early- active lifetimes, if all of FRB 180924-like bursts result type elliptical/S0 galaxies, while a large subset of NSWD from magnetars, based on Ravi (2019). mergers are most likely to be found in late-type, disk, and star-forming galaxies since the delay time distribu- 4. SUMMARY AND DISCUSSION tion peaks at early times (< 1 − 2 Gyr). This is because Assuming there are two FRB populations: FRB they argued that hybrid WD mergers are more com- 121102-like bursts arise from magnetars born in mon than pure He WD mergers. They also obtained SLSNe/LGRBs channels while FRB 180924-like bursts that the offsets of NSWD mergers, depending on the arise from magnetars born in BNS/BWD/AIC channels, stellar density of host galaxies, could range from small we have investigated whether FRB 180924-like bursts offsets due to a low escape velocity for those in dwarf could also arise from magnetars formed from the NSWD galaxies to very large offsets of up to a few hundred channel, i.e., (1) whether magnetars can be formed from kiloparsecs for those including NS natal kicks. How- NSWD mergers with unstable mass transfer, and (2) ever, it is still possible that some Ca-rich transients orig- if they can indeed, whether flaring magnetars formed inate from He WD mergers, while more massive NSWD from this channel can explain the observations of FRB mergers give rise to some kind of fast-evolving Ic-like 180924-like bursts such as their own characteristics, lo- transients (Margalit & Metzger 2016; Fernández et al. cal environments, host galaxies, and event rate. We ex- 2019). Therefore, we can see that the host galaxies plored the first question and speculated that there are of NSWD mergers are in a large range. In this case, two possible scenarios to produce strongly magnetized there should be a subset of galaxies hosted by NSWD “renascent” NSs from NSWD mergers. The first sce- mergers to satisfy the properties of host galaxies of FRB nario is magnetic field amplification by a vigorous α– 180924-like bursts. Thanks to the large range in local en- ω dynamo acting on the accreting NS surrounded by vironments and host galaxies, however, this channel can a massive extended hot disk composed of WD debris also account for the properties of local environment and during the mass transfer process. We performed a pre- host galaxy of FRB 180916.J0158+65, i.e., this FRB lo- liminary calculation and showed that the magnetic field cates at a star-forming region in a massive spiral galaxy of the final NS could be enhanced via the dynamo in- (Marcote et al. 2020). duced by differential rotation and convection in/on the FRBs from Magnetars Formed from NSWD Mergers 9 accreting NS, as well as the MRI in the disk. The sec- FRB 180924 such as circumburst DM, RM, and per- ond scenario is magnetic flux conservation of a fossil sistent radio source because of the similar ejecta, but field. This scenario could contribute to a small frac- also its local environment and host galaxy differing from tion of NSWD binaries in which the NSs are strongly FRB 180924. magnetized and remain their magnetic fields before co- In the future, an evident association between FRBs alescence. Whether or not these scenarios can give rise and magnetars formed from NSWD mergers should need to magnetars post NSWD mergers still requires evidence an association of transients most similar to faint type from both numerical simulations and observations. As Ib Ca-rich SNe (Metzger 2012) or SNe Ic (Zenati et al. a result, the magnetic field evolution of the remnants 2019b), gravitational waves from NSWD during the in- requires some deep and complex exploration and mag- spiral and merger phase detected by eLISA or even netohydrodynamics simulations. aLIGO/Virgo (Paschalidis et al. 2009), and FRBs, if In any case, if the NSWD channel can create magne- such bursts are indeed produced from flaring magnetars. tars, it could produce FRB 180924-like bursts and account for their properties over an active lifetime, burst transparency, circumburst DM, RM, persistent radio ACKNOWLEDGMENTS source, host galaxy, and event rate within the frame- We would like to thank the referee for his/her very work of flaring magnetars because the ejecta, local envi- careful and helpful comments and suggestions that ronments, and host galaxies of the final remnants from have allowed us to improve the presentation of this this channel resemble those of BNS/BWD/AIC chan- manuscript significantly. This work was supported by nels. Otherwise, within a large range in local environ- the National Key Research and Development Program ment and host galaxy, this channel can also account for of China (grant No. 2017YFA0402600) and the Na- the observational properties of FRB 180916.J0158+65 tional Natural Science Foundation of China (grant No. (Marcote et al. 2020): not only its properties similar to 11833003).

REFERENCES

Balbus, S. A., Hawley, J. F. 1998, Reviews of Modern CHIME/FRB Collaboration, Andersen, B. C., Bandura, K., Physics, 70, 1, doi: 10.1103/RevModPhys.70.1 et al. 2019b, ApJL, 885, L24, Bannister, K. W., Deller, A. T., Phillips, C., et al. 2019, doi: 10.3847/2041-8213/ab4a80 Science, 365, 565, doi: 10.1126/science.aaw5903 CHIME/FRB Collaboration, Fonseca, E., Andersen, B. C., Belczynski, K., Kalogera, V., & Bulik, T. 2002, ApJ, 572, Bhardwaj, M., et al., 2020, arXiv e-prints, 407, doi: 10.1086/340304 arXiv:2001.03595. https://arxiv.org/abs/2001.03595 Belczynski, K., Perna, R., Bulik, T., et al. 2006, ApJ, 648, Cordes, J. M., & Chatterjee, S. 2019, ARA&A, 57, 417, 1110, doi: 10.1086/505169 doi: 10.1146/annurev-astro-091918-104501 Beloborodov, A. M. 2017, ApJL, 843, L26, Cowperthwaite, P. S., Berger, E., Villar, V. A., et al. 2017, doi: 10.3847/2041-8213/aa78f3 ApJL, 848, L17, doi: 10.3847/2041-8213/aa8fc7 —. 2019, arXiv e-prints, arXiv:1908.07743. Cumming, A., Arras, P., & Zweibel, E. 2004, ApJ, 609, 999, https://arxiv.org/abs/1908.07743 doi: 10.1086/421324 Beloborodov, A. M., & Li, X. 2016, ApJ, 833, 261, Dai, Z. G., Wang, J. S., Wu, X. F., & Huang, Y. F. 2016, doi: 10.3847/1538-4357/833/2/261 ApJ, 829, 27, doi: 10.3847/0004-637X/829/1/27 Blackman, E. G., Frank, A., Welch, C. 2001, ApJ, 546, 288, Dai, Z. G., Wang, X. Y., Wu, X. F., & Zhang, B. 2006, doi: 10.1086/318253 Science, 311, 1127, doi: 10.1126/science.1123606 Bobrick, A., Davies, M. B., & Church, R. P. 2017, MNRAS, Deng, C.-M., Cai, Y., Wu, X.-F., & Liang, E.-W. 2018, 467, 3556, doi: 10.1093/mnras/stx312 PhRvD, 98, 123016, doi: 10.1103/PhysRevD.98.123016 Champion, D. J., Petroﬀ, E., Kramer, M., et al. 2016, Duncan, R. C., & Thompson, C. 1992, ApJL, 392, L9, MNRAS, 460, L30, doi: 10.1093/mnrasl/slw069 doi: 10.1086/186413 Chatterjee, S., Law, C. J., Wharton, R. S., et al. 2017, Falcke, H., & Rezzolla, L. 2014, A&A, 562, A137, Nature, 541, 58, doi: 10.1038/nature20797 doi: 10.1051/0004-6361/201321996 CHIME/FRB Collaboration, Amiri, M., Bandura, K., et al. Farah, W., Flynn, C., Bailes, M., et al. 2018, MNRAS, 478, 2019a, Nature, 566, 235, doi: 10.1038/s41586-018-0864-x 1209, doi: 10.1093/mnras/sty1122 10 Zhong & Dai

Fernández, R., Margalit, B., Metzger, B. 2019, MNRAS, Kumar, P., Shannon, R. M., Oslowski, S., et al. 2019, 488, 259, doi: 10.1093/mnras/stz1701 ApJL, 887, L30, doi: 10.3847/2041-8213/ab5b08 Ferrario, L., & Wickramasinghe, D. 2006, MNRAS, 367, Lattimer, J. M., & Prakash, M. 2004, Science, 304, 536, 1323, doi: 10.1111/j.1365-2966.2006.10058.x doi: 10.1126/science.1090720 —. 2008, MNRAS, 389, L66, Liu, X. 2018, Ap&SS, 363, 242, doi: 10.1111/j.1745-3933.2008.00527.x doi: 10.1007/s10509-018-3462-3 Frank, J., King, A., & Raine, D. 1992, Camb. Astrophys. Liu, X. 2020, arXiv e-prints, arXiv:2002.03693. Ser, Accretion power in astrophysics., Cambridge https://arxiv.org/abs/2002.03693 University Press, Cambridge, England, Vol. 21 Lorimer, D. R., Bailes, M., McLaughlin, M. A., Narkevic, Geng, J. J., & Huang, Y. F. 2015, ApJ, 809, 24, D. J., & Crawford, F. 2007, Science, 318, 777, doi: 10.1088/0004-637X/809/1/24 doi: 10.1126/science.1147532 Giacomazzo, B., & Perna, R. 2013, ApJL, 771, L26, Lu, W., & Kumar, P. 2018, MNRAS, 477, 2470, doi: 10.1088/2041-8205/771/2/L26 doi: 10.1093/mnras/sty716 Goldreich, P., & Reisenegger, A. 1992, ApJ, 395, 250, Lyman, J. D., James, P. A., Perets, H. B., et al. 2013, doi: 10.1086/171646 MNRAS, 434, 527, doi: 10.1093/mnras/stt1038 Gu, W.-M., Dong, Y.-Z., Liu, T., Ma, R., & Wang, J. 2016, Lyubarsky, Y. 2014, MNRAS, 442, L9, ApJL, 823, L28, doi: 10.3847/2041-8205/823/2/L28 doi: 10.1093/mnrasl/slu046 Harding, A. K., & Lai, D. 2006, Reports on Progress in Lyutikov, M., Burzawa, L., & Popov, S. B. 2016, MNRAS, Physics, 69, 2631, doi: 10.1088/0034-4885/69/9/R03 462, 941, doi: 10.1093/mnras/stw1669 Hessels, J. W. T., Spitler, L. G., Seymour, A. D., et al. Madau, P., & Dickinson, M. 2014, ARA&A, 52, 415, 2019, ApJL, 876, L23, doi: 10.3847/2041-8213/ab13ae doi: 10.1146/annurev-astro-081811-125615 Heyl, J. S., & Kulkarni, S. R. 1998, ApJL, 506, L61, Marcote, B., Nimmo K., Hessels J. W. T., et al. 2020, doi: 10.1086/311628 Nature, 577, 9, doi: 10.1038/s41586-019-1866-z Hjellming, M. S., & Webbink, R. F. 1987, ApJ, 318, 794, Margalit, B., Berger, E., & Metzger, B. D. 2019, ApJ, 886, doi: 10.1086/165412 110, doi: 10.3847/1538-4357/ab4c31 Hurley, J. R., Tout, C. A., & Pols, O. R. 2002, MNRAS, Margalit, B., & Metzger, B. D. 2016, MNRAS, 461, 1154, 329, 897, doi: 10.1046/j.1365-8711.2002.05038.x doi: 10.1093/mnras/stw1410 Kasen, D., Metzger, B., Barnes, J., Quataert, E., & —. 2017, MNRAS, 465, 2790, doi: 10.1093/mnras/stw2640 Ramirez-Ruiz, E. 2017, Nature, 551, 80, —. 2018, ApJL, 868, L4, doi: 10.3847/2041-8213/aaedad doi: 10.1038/nature24453 Margalit, B., Metzger, B. D., Berger, E., et al. 2018, Kashiyama, K., Ioka, K., & Mészáros, P. 2013, ApJL, 776, MNRAS, 481, 2407, doi: 10.1093/mnras/sty2417 L39, doi: 10.1088/2041-8205/776/2/L39 Metzger, B. D. 2012, MNRAS, 419, 827, Kasliwal, M. M., Kulkarni, S. R., Gal-Yam, A., et al. 2012, doi: 10.1111/j.1365-2966.2011.19747.x ApJ, 755, 161, doi: 10.1088/0004-637X/755/2/161 Metzger, B. D., Berger, E., & Margalit, B. 2017, ApJ, 841, Katz, J. I. 2016, ApJ, 826, 226, 14, doi: 10.3847/1538-4357/aa633d doi: 10.3847/0004-637X/826/2/226 Metzger, B. D., Beniamini, P., Giannios, D. 2018, Appl. —. 2018, Progress in Particle and Nuclear Physics, 103, 1, Phys. Lett., 857, 95, doi: 10.3847/1538-4357/aab70c doi: 10.1016/j.ppnp.2018.07.001 Metzger, B. D., Margalit, B., & Sironi, L. 2019, MNRAS, Keane, E. F., Stappers, B. W., Kramer, M., & Lyne, A. G. 485, 4091, doi: 10.1093/mnras/stz700 2012, MNRAS, 425, L71, Michilli, D., Seymour, A., Hessels, J. W. T., et al. 2018, doi: 10.1111/j.1745-3933.2012.01306.x Nature, 553, 182, doi: 10.1038/nature25149 Khokhriakova, A. D., & Popov, S. B. 2019, Journal of High Murase, K., Kashiyama, K., & Mészáros, P. 2016, MNRAS, Energy Astrophysics, 24, 1, 461, 1498, doi: 10.1093/mnras/stw1328 doi: 10.1016/j.jheap.2019.09.004 Nicholl, M., Williams, P. K. G., Berger, E., et al. 2017, King, A. R., Pringle, J. E., & Wickramasinghe, D. T. 2001, ApJ, 843, 84, doi: 10.3847/1538-4357/aa794d MNRAS, 320, L45, doi: 10.1046/j.1365-8711.2001.04184.x Nomoto, K., & Kondo, Y. 1991, ApJL, 367, L19, Konar, S., & Bhattacharya, D. 1997, MNRAS, 284, 311, doi: 10.1086/185922 doi: 10.1093/mnras/284.2.311 Nordhaus, J., Wellons, S., Spiegel, D. S., Metzger, B. D., Kumar, P., Lu, W., & Bhattacharya, M. 2017, MNRAS, Blackman, E. G. 2011, PNAS, 108, 3135, 468, 2726, doi: 10.1093/mnras/stx665 doi: 10.1073/pnas.1015005108 FRBs from Magnetars Formed from NSWD Mergers 11

Parfrey, K., Spitkovsky, A., Beloborodov, A. M. 2016, ApJ, Thompson, C., & Duncan, R. C. 1993, ApJ, 408, 194, 822, 33, doi: 10.3847/0004-637X/822/1/33 doi: 10.1086/172580 Paschalidis, V., Etienne, Z., Liu, Y. T., & Shapiro, S. L. —. 1996, ApJ, 473, 322, doi: 10.1086/178147 2011a, PhRvD, 83, 064002, Thompson, T. A., Kistler, M. D., & Stanek, K. Z. 2009, doi: 10.1103/PhysRevD.83.064002 arXiv e-prints, arXiv:0912.0009. Paschalidis, V., Liu, Y. T., Etienne, Z., & Shapiro, S. L. https://arxiv.org/abs/0912.0009 2011b, PhRvD, 84, 104032, Thorne, K. S., & Zytkow, A. N. 1977, ApJ, 212, 832, doi: 10.1103/PhysRevD.84.104032 doi: 10.1086/155109 Paschalidis, V., MacLeod, M., Baumgarte, T. W., & Thornton, D., Stappers, B., Bailes, M., et al. 2013, Science, Shapiro, S. L. 2009, PhRvD, 80, 024006, 341, 53, doi: 10.1126/science.1236789 doi: 10.1103/PhysRevD.80.024006 Toonen, S., Perets, H. B., Igoshev, A. P., Michaely, E., & Perets, H. B., Gal-Yam, A., Mazzali, P. A., et al. 2010, Zenati, Y. 2018, A&A, 619, A53, Nature, 465, 322, doi: 10.1038/nature09056 doi: 10.1051/0004-6361/201833164 Petroﬀ, E., Hessels, J. W. T., & Lorimer, D. R. 2019, Tout, C. A., Wickramasinghe, D. T., Liebert, J., Ferrario, A&A Rv, 27, 4, doi: 10.1007/s00159-019-0116-6 L., Pringle, J. E. 2008, MNRAS, 387, 897, Petroﬀ, E., Barr, E. D., Jameson, A., et al. 2016, PASA, 33, doi: 10.1111/j.1365-2966.2008.13291.x e045, doi: 10.1017/pasa.2016.35 Troja, E., Rosswog, S., & Gehrels, N. 2010, ApJ, 723, 1711, Piro, A. L., & Burke-Spolaor, S. 2017, ApJL, 841, L30, doi: 10.3847/2041-8213/aa740d doi: 10.1088/0004-637X/723/2/1711 Piro, A. L., & Ott, C. D. 2011, ApJ, 736, 108, Turolla, R., Zane, S., & Watts, A. L. 2015, Reports on doi: 10.1088/0004-637X/736/2/108 Progress in Physics, 78, 116901, Platts, E., Weltman, A., Walters, A., et al. 2019, PhR, 821, doi: 10.1088/0034-4885/78/11/116901 1, doi: 10.1016/j.physrep.2019.06.003 Urpin, V., & Konenkov, D. 1997, MNRAS, 284, 741, Popov, S. B., Postnov, K. A., & Pshirkov, M. S. 2018, doi: 10.1093/mnras/284.3.741 Physics Uspekhi, 61, 965, van Haaften, L. M., Nelemans, G., Voss, R., Wood, M. A., doi: 10.3367/UFNe.2018.03.038313 & Kuijpers, J. 2012, A&A, 537, A104, Price, D. J., & Rosswog, S. 2006, Science, 312, 719, doi: 10.1051/0004-6361/201117880 doi: 10.1126/science.1125201 Villar, V. A., Guillochon, J., Berger, E., et al. 2017, ApJL, Prochaska, J. X., Macquart, J.-P., McQuinn, M., et al. 851, L21, doi: 10.3847/2041-8213/aa9c84 2019, Science, 366, 231, doi: 10.1126/science.aay0073 Wang, J.-S., Yang, Y.-P., Wu, X.-F., Dai, Z.-G., & Wang, Quataert, Eliot & Gruzinov, Andrei. 2000, ApJ, 539, 809, F.-Y. 2016, ApJL, 822, L7, doi: 10.1086/309267 doi: 10.3847/2041-8205/822/1/L7 Ravi, V. 2019, Nature Astronomy, 3, 928, Xue, Y. Q., Zheng, X. C., Li, Y., et al. 2019, Nature, 568, doi: 10.1038/s41550-019-0831-y 198, doi: 10.1038/s41586-019-1079-5 Ravi, V., Catha, M., D’Addario, L., et al. 2019, Nature, Yoon, S. C., Podsiadlowski, P., & Rosswog, S. 2007, 572, 352, doi: 10.1038/s41586-019-1389-7 MNRAS, 380, 933, doi: 10.1111/j.1365-2966.2007.12161.x Rosswog, S., Ramirez-Ruiz, E., & Davies, M. B. 2003, Zenati, Y., Bobrick, A., & Perets, H. B. 2019a, arXiv MNRAS, 345, 1077, e-prints, arXiv:1908.10866. doi: 10.1046/j.1365-2966.2003.07032.x https://arxiv.org/abs/1908.10866 Schwab, J., Quataert, E., & Bildsten, L. 2015, MNRAS, Zenati, Y., Perets, H. B., & Toonen, S. 2019b, MNRAS, 453, 1910, doi: 10.1093/mnras/stv1804 486, 1805, doi: 10.1093/mnras/stz316 Schwab, J., Quataert, E., & Kasen, D. 2016, MNRAS, 463, Zhang, B. 2014, ApJL, 780, L21, 3461, doi: 10.1093/mnras/stw2249 doi: 10.1088/2041-8205/780/2/L21 Spitler, L. G., Scholz, P., Hessels, J. W. T., et al. 2016, Nature, 531, 202, doi: 10.1038/nature17168 —. 2016, ApJL, 827, L31, Sun, S., Li, L., Liu, H., et al. 2019, PASA, 36, e005, doi: 10.3847/2041-8205/827/2/L31 doi: 10.1017/pasa.2018.51 —. 2017, ApJL, 836, L32, doi: 10.3847/2041-8213/aa5ded Tauris, T. M., Sanyal, D., Yoon, S. C., & Langer, N. 2013, —. 2018, ApJL, 867, L21, doi: 10.3847/2041-8213/aae8e3 A&A, 558, A39, doi: 10.1051/0004-6361/201321662 Zhong, S.-Q., Dai, Z.-G., & Li, X.-D. 2019, PhRvD, 100, Tendulkar, S. P., Bassa, C. G., Cordes, J. M., et al. 2017, 123014, doi: 10.1103/PhysRevD.100.123014 ApJL, 834, L7, doi: 10.3847/2041-8213/834/2/L7