Multimessenger signals from black hole-neutron star mergers without significant tidal disruption
William E. East, Luis Lehner, Steven L. Liebling, Carlos Palenzuela
DDraft version February 1, 2021
Typeset using L A TEX twocolumn style in AASTeX63
Multimessenger signals from black hole-neutron star mergers without significant tidal disruption
William E. East, Luis Lehner, Steven L. Liebling, and Carlos Palenzuela Perimeter Institute, 31 Caroline St, Waterloo, ON N2L 2Y5, Canada Long Island University, Brookville, New York 11548, USA Departament de F´ısica & IAC3, Universitat de les Illes Balears, Palma de Mallorca, Baleares E-07122, Spain
ABSTRACTWe study the multimessenger signals from the merger of a black hole with a magnetized neutron starusing resistive magneto-hydrodynamics simulations coupled to full general relativity. We focus on acase with a 5:1 mass-ratio, where only a small amount of the neutron star matter remains post-merger,but we nevertheless find that significant electromagnetic radiation can be powered by the interactionof the neutron star’s magnetosphere with the black hole. In the lead-up to merger, strong twisting ofmagnetic field lines from the inspiral leads to plasmoid emission and results in a luminosity in excessof that expected from unipolar induction. We find that the strongest emission occurs shortly aftermerger during a transitory period in which magnetic loops form and escape the central region. Theremaining magnetic field collimates around the spin axis of the remnant black hole before dissipating,an indication that, in more favorable scenarios (higher black hole spin/lower mass ratio) with largeraccretion disks, a jet would form. INTRODUCTIONMergers of compact binary systems composed of atleast one neutron star are particularly exciting becauseof the prospect of observing sufficiently nearby eventsthrough gravitational wave (GW), electromagnetic, andeven neutrino signals. The presence of matter due tothe neutron star(s), along with the fact that neutronstars (NSs) can support intense magnetic fields withinarguably the strongest gravitational regime, offers thepotential of some of the strictest tests yet of fundamentalphysics.The observations by GW detectors and conventionalelectromagnetic telescopes of the binary NS mergerknown as GW170817 provided a spectacular inaugura-tion of this nascent enterprise (Abbott et al. 2017a,b).This golden event (which was located nearby) has en-abled a variety of groundbreaking results including: con-straining the equation of state of high density matter,establishing that the mergers of NSs give rise to shortgamma ray bursts, and demonstrating the productionof heavy elements from such mergers through detailedobservations of the subsequent kilonova. At the sametime, non-vacuum binaries present challenges to currentGW observatories. In the case of binary NS systems,the total mass M T is rather low [ (cid:46) M (cid:12) , which impliesa “collision” GW frequency f ≈ . M T /M (cid:12) ) / kHz]. Adopting here for concreteness a NS radius of 12 km, regardlessof mass.
As a consequence, the most relativistic regime, whichmost strongly probes fundamental physics, occupies thehigh frequency range where current detectors rapidlylose sensitivity.Black hole-neutron star (BHNS) binaries also have thepotential to source multimessenger signals. These sys-tems can potentially have much larger masses than bi-nary NSs, and thus their merger is, in principle, easier todetect, as it takes place at lower frequencies where cur-rent detectors are more sensitive [ f (cid:38) M (cid:12) /M T )Hz]. Importantly, for high enough binary mass ratios,the NS might be devoured by the black hole (BH) with-out being disrupted. In such a scenario, the GWs willbe essentially indistinguishable from those sourced by abinary BH with the same masses.For a sufficiently low mass ratio (where the exactthreshold depends on the BH spin and the propertiesof the high density equation of state), the star is tidallydisrupted by the BH outside the effective innermost sta-ble orbit of the binary, resulting in the ejection of mate-rial from the system and the formation of an accretiondisk that may produce an electromagnetic counterpart,in particular a short gamma ray burst. This scenariowould be ideal for producing signals that would probethe strong gravitational field (and arguably the strongestcurvatures) as well as the properties of high density mat-ter.However, observations of stellar mass BHs in binariesobtained to date (both in electromagnetic observationsof LMXBs and GW events by LIGO/Virgo) suggest a a r X i v : . [ a s t r o - ph . H E ] J a n dearth of such low mass ratio systems (Abbott et al.2020a; Corral-Santana et al. 2016). Instead, the evi-dence suggests a prevalence of higher mass BHs ( (cid:38) M (cid:12) ) with low spins. While this may be a result of ob-servational bias, and the companions in these binarieswere not NSs (with one or two possible exceptions), thissuggests that disruption or other tidal effects might notbe strong enough to manifest in the GWs from mostBHNS with high signal-to-noise, or to lead to an accre-tion powered transient. Thus, it is important to under-stand other potential electromagnetic emission, whichcould break the degeneracy between BHNS and binaryBH mergers.For instance, the recent event GW190814 (Abbottet al. 2020b), with mass ratio of ≈ ≈ . − . M (cid:12) . The GW signaldoes not indicate whether the companion was a NS ora BH, and its mass does not favor either one in par-ticular unless further (potentially biased) assumptionsare made. Thus, this object was either the most mas-sive NS, or the lightest BH, yet observed. Either answerwould have profound consequences for astrophysics andnuclear physics.The unknown nature of the secondary object inGW190814 demonstrates the importance of understand-ing potential electromagnetic signals that could breaksuch degeneracies. Because the NS in a BHNS is likelymagnetized (Spruit 2008), its interaction with a BH—which intensifies as the orbit tightens—could give riseto electromagnetic counterparts. Several authors havediscussed how this might occur within a unipolar induc-tion (UI) model (Hansen & Lyutikov 2001; McWilliams& Levin 2011; Piro 2012; Lai 2012; D’Orazio et al. 2016),with the BH acting as a battery in a DC circuit withthe NS. However, whether this simple steady-state pic-ture is accurate remains an open question. In particular,one interesting possibility that has not been explored indetail is that the continuous twisting of magnetic fieldlines leads to more complicated reconnection and plas-moid ejection, as well as emission channeled throughthe development of a current sheet. The twisting an-gle is related to the BH velocity relative to the NS as ζ φ ≈ v rel / ( πc ) (Lai 2012), and thus one expects theseeffects, and the departure from the UI model, to bestrongest around merger.Several studies of the dynamics of magnetospheric in-teractions in binary NS mergers have demonstrated howthe binary’s kinetic energy can be converted into elec-tromagnetic radiation (Palenzuela et al. 2013b,a; Ponceet al. 2014) through UI and accelerating magnetic dipoleeffects (Carrasco & Shibata 2020), as well as how the twisting of magnetic flux tubes can produce periodicflaring (Most & Philippov 2020).The BHNS case has been studied in the force-free ap-proximation assuming a helical Killing vector (Pascha-lidis et al. 2013; Carrasco et al. 2021), which would ap-proximately hold during the early inspiral. Here, weconcentrate on the final stages of a 5:1 mass-ratio BHNSmerger using full GR simulations. This allows us to ex-plore the most dynamical part of the merger, when theinteraction of the BH with the NS’s magnetosphere willbe strongest. It also allows us to study the post-mergerdissipation of the magnetosphere, which may also sourcean associated electromagnetic transient (Lehner et al.2012; Lyutikov & McKinney 2011; Pan & Yang 2019).We treat the plasma dynamics with a resistive magneto-hydrodynamics (MHD) approach (Palenzuela 2013) thatcan interpolate between the fluid pressure dominatedregions inside the NS, and the magnetically dominatedregime in the tenuous plasma surrounding the binary.We find that the interaction of the BH with the mag-netic field sourced by the NS as the binary merges leadsto significant electromagnetic emission. The continualtwisting of the magnetic field produces current sheetswith a complex configuration, occurring both in thevicinity of the BH and also at larger distances. Thesecurrent sheets, where charges can be effectively acceler-ated, result as field lines are stretched, forming X-pointswhere reconnection takes place. Plasmoids, isolated re-gions of closed field lines, are produced from the recon-nection, carrying field energy away from the system, andit is their emission which results in a level of electromag-netic emission stronger than that estimated by UI.As the merger ensues, the magnetic field collimatesaround the final rotating BH, but quickly dissipates dueto the lack of matter to anchor it, in part through plas-moid emission. In more favorable conditions, where anaccretion disk is formed, this collimation could launch ajet. However, we find that even in a case not conduciveto the latter scenario, the magnetosphere of a mergingBHNS binary can power electromagnetic emissions thatcan potentially be seen to tens of Mpc. SETUPTo study the system of interest, we employ the
Had computational infrastructure ( had home page2010) and implement the general relativistic, resistiveMHD equations, as described in Palenzuela (2013);Palenzuela et al. (2009), coupled to Einstein gravity inthe CCZ4 formulation (Alic et al. 2012; Bezares et al.2017). Thus, we capture the behavior of both resis-tive, magnetized matter, and its interplay with the dy-namical spacetime. We next describe the most relevantdetails involved in the description of the system; for athorough description of the implementation we refer thereader to the aforementioned works. Unless otherwisestated, in the following we use Lorentz-Heaviside unitswith G = c = 1.The magnetized star is described by the total stress-energy tensor T µν = [ ρ (1 + (cid:15) ) + p ] u µ u ν + pg µν + F µλ F νλ − g µν F λα F λα (1)where F µν is the Faraday tensor which can be decom-posed in terms of the electric E µ and magnetic B µ fields . Here u a is the fluid four velocity, ρ is the restmass density, (cid:15) the internal energy, and p is the pressure.Here we use a Γ = 2 equation of state p = ρ(cid:15) .The evolution of the magnetized matter must obeyboth the Maxwell equations and the conservation of to-tal stress-energy tensor. Going beyond the ideal MHDlimit, which treats the fluid as a perfect conductor, re-quires a prescription for the electric current to close thesystem of equations, called resistive MHD. The idealMHD and the force-free limits can be captured withthe phenomenological current introduced in Palenzuela(2013), which includes the isotropic conductivity and(some of) the anisotropic Hall terms J i = q (cid:2) (1 − H ) v i + Hv id (cid:3) + σ ζ (cid:20) E i + ζ B ( E k B k ) B i (cid:21) (2)where v id = (cid:15) ijk E j B k /B is the drift velocity, v i = u i /W is the Eulerian velocity with associated Lorentz factor W = αu t (in terms of the lapse α ), and we have intro-duced the shorthand E i = W (cid:2) E i + (cid:15) ijk v j B k − ( v k E k ) v i (cid:3) . (3)The kernel function H is defined such that it smoothlyvaries with density from zero inside the star, to unityoutside, with a very high isotropic conductivity σ =2 × s − and an anisotropic ratio ζ = Hσ . In our par-ticular scenario, in the interior of the star (i.e., H ≈ H ≈
1) the anisotropic terms dominate andeffectively enforce the force-free condition.We adopt initial data describing a non-spinning bi-nary consisting of a BH with mass M BH = 7 M (cid:12) anda NS with mass M NS = 1 . M (cid:12) , constructed with the Notice that a factor 1 / √ π has been absorbed in the definitionof the electromagnetic fields. Lorene library (Gourgoulhon et al. 2010). The NS obeysa polytropic equation of state p = Kρ Γ with adiabaticindex Γ = 2 and has a radius of R NS = 11 .
62 km. Theseparameters are chosen so as to be intermediate betweenthose giving significant tidal disruption, and those giv-ing no disruption at all. This allows us to make suit-able connections with both lower and higher mass ratiocases. Choosing a mass ratio that allows a small amountof matter to remain bound but outside the BH for sometime enables us to study the development of magneticfield structures post-merger that would be more marked,and relevant for observations, in cases with lower massratios/higher BH spins.This binary system is initialized in a quasi-circularconfiguration with an orbital frequency Ω = 890 Hz,roughly 2 . (cid:126)µ that describesa dipolar magnetic field (cid:126)B in the co-moving frame ofthe star. This magnetic moment is aligned with the or-bital angular momentum. The magnitude of the dipolemoment is related to the radial magnetic field at thepole of the star B ∗ , by the relation µ = B ∗ R . Theelectric field is obtained from the ideal MHD condition (cid:126)E = − (cid:126)v × (cid:126)B , where the velocity in the star is given bythe orbital motion, and we assume that the magneto-sphere is initially at rest. In our simulations, we take B ∗ = 3 × G, though at this low value the magneticfield does not have any significant effect on the space-time (Ioka & Taniguchi 2000), nor on the hydrodynam-ics, except perhaps for the low density material in thevicinity of the post-merger BH. Hence our results canbe approximately scaled to arbitrary NS magnetic fieldvalues B := B ∗ / (10 G) within this regime.The gravitational equations are discretized withfourth order accurate finite difference operators, whilehigh resolution shock capturing methods based on theHLL flux formula with PPM reconstruction are usedto discretize the resistive MHD equations (Palenzuela2013). The time evolution is performed through themethod of lines using a third order accurate Implicit-Explicit Runge-Kutta integration scheme (Pareschi &Russo 2005) in order to deal with the stiffness of theresistive equations (Palenzuela et al. 2009). In our pro-duction run, the adaptive mesh refinement criteria tol-erance is chosen to guarantee that the star is covered by84 points in each direction. The computational domain[ − , km is discretized with 7 refinement lev-els (with a 2:1 refinement ratio and with the coarsestgrid having a grid spacing of ∆ x = 30 km). We adopta Courant parameter ∆ t/ ∆ x = 0 .
25 in each refinementlevel. To check the consistency of our results, we alsoevolve: (i) the same computational set-up with one lessrefinement level for the full inspiral and merger; and(ii) another with 7 refinement levels, but in which thecomputational domain extends only to ±
600 km and thecoarsest resolution is ∆ x = 20 km, for roughly the firstorbit. The results of these additional simulations sug-gest that our production run is in the convergent regime. RESULTSThe BHNS binary undergoes roughly 2 orbits beforethe outermost layer of the star gets stripped away andthe bulk of the star is swallowed by the BH. This resultsin a final BH with dimensionless spin a/M (cid:39) .
4. Only asmall fraction of the matter remains outside the BH aftermerger ( ≈ M NS ), and it is slowly accreting onto theBH at a rate ˙ M ∝ t − / (i.e., there is negligible ejecta).While most of this behavior had been understood pre-viously (e.g. Chawla et al. 2010), our main focus here ison examining the behavior of the electromagnetic fieldand the potential electromagnetic signals induced by themerger. Previous attempts to estimate these have reliedon a UI model (Hansen & Lyutikov 2001; McWilliams& Levin 2011; Lai 2012). While this model can capturethe broad strokes of certain features, it provides a rathersimple view of the process, and does not capture, for ex-ample, the transitory plasma dynamics nor the detailsof the dissipative processes which we find here.Of particular relevance is the structure of currentsheets associated with the system, which not only bearsa strong correlation with the dynamics of the binary (asalready indicated in Palenzuela et al. 2013b,a; Ponceet al. 2014; Most & Philippov 2020), but also with thecharacteristics of the compact objects involved. Fig-ure 1 illustrates this structure, showing current sheetsthat have developed both on and off the orbital plane attwo representative times prior to merger. Roughly, wecan understand the development of these current sheetsas due to two different effects. On the one hand, fieldlines emanating from the NS get sufficiently bent, as thestar orbits, to seed a current sheet some distance away,even though the NS is not spinning as in previous bi-nary NS studies. Analogous to the light-cylinder radiusof an isolated spinning NS, we expect this to occur at alengthscale L (cid:39) Ω − (where Ω is the angular velocity ofthe binary). On the other hand, field lines sufficientlyclose to the BH get twisted to such a large degree as toseed a current sheet in the wake of the BH’s trajectory,even when the BH is not spinning [a behavior also pre-viously observed in simulations (Palenzuela et al. 2010;Neilsen et al. 2011)]. As evident in Fig. 1, this lattercurrent sheet is smaller scale, comparable to the radiusof the BH. Because current sheets are the site of reconnection,their dynamics is key to understanding the electromag-netic output of the system. As the orbit proceeds,field lines are more rapidly wound, and the magneticfield strength in the strongly gravitating vicinity of theBH increases. In this low density region, our resistivescheme approaches the force-free limit. With sufficientwinding, X-point reconnection occurs and leads to closedfield loops which propagate away at near-luminal speeds,with loops forming near the BH having greater fieldstrength than those produced further away . Previousstudies of reconnection in the force-free approximationhave found that the process is fast (relative, say, to thatin regimes closer to the ideal MHD limit), with relativis-tic speeds roughly v rec ≈ . ∼ R BH /v rec ≈ . ∼ Ω − /v rec ≈ v rec ≈ . ≈ . . The familiar chirpis present, and the post merger, quasinormal ringing isconsistent with the expectation for a remnant BH with M (cid:39) . M (cid:12) and a/M (cid:39) . Greater field strength of loops formed near the BH has also beenobserved in ongoing studies of force-free solutions sourced by aneffective NS orbiting in a fixed Kerr BH spacetime (Carrascoet al. 2021).
Figure 1.
Snapshots at times t = 6 . . ρ = 3 × g/cm ), the BH (black), magnetic fieldlines (blue, integrated over some seeds centered on the star)and current sheet structure (defined as the region where | (cid:126)E | > | (cid:126)B | ; grey, semi-transparent). The system orbitscounter-clockwise. The snapshots demonstrate the signifi-cant dynamics of the electromagnetic field, particularly thestructure of the current sheets. A current sheet forms behindthe high curvature region of the BH, and trails the BH asit orbits. Another current sheet is correlated with the mo-tion of the NS, but at some distance from the star. As theorbit proceeds, significant deformation of field lines leads tomore current sheets. These sheets support closed magneticfield loops that arise from magnetic reconnection and thatare transported away from the system. from simple arguments (Foucart 2012; Buonanno et al.2008)].Interestingly, the peak of the electromagnetic emissionoccurs a couple of milliseconds later than that of thegravitational radiation peak. While the GW peak cor-responds roughly to the maximum rate of change of thesystem’s quadrupole moment, the electromagnetic emis-sion arises from the reconnection of magnetic field lines.Thus the time delay of the electromagnetic emission isa consequence of the magnetic reconfiguration forced bythe formation of a common horizon for the BH and NS.We also note that the delay is commensurate with thespin period of the remnant BH, roughly 2.5 ms.An important result of this paper is the compari-son of this luminosity with that predicted by the UImodel L UI , which we also include in Fig. 2. Recall thatthe UI model prediction for the binary studied here is L UI ≈ × ( v rel /c ) B (100 km /d ) erg/s. For thefigure, we use the Keplerian expressions for { v, r } interms of the orbital frequency Ω. We estimate this lat-ter one in terms of the GW frequency f ≈ Ω π (hence L UI ∝ f / ), thus capturing the more rapid rate atwhich { v, r } change due to the increasingly strong grav-itational effects as the final plunge approaches. Thesimulations consistently show a larger luminosity thanthat of the UI model at earlier times (lower frequen-cies), as that model does not capture the complex phe-nomena associated with reconnections and the role ofcurrent sheets. Close to the coalescence, UI becomes sufficiently strong as to match the measured values .Furthermore, notice that a slightly weaker Poynting fluxis measured at larger extraction spheres, reflective ofthe energy dissipated at current sheets located betweenthe spheres (a behavior recently pointed out in Most& Philippov 2020). Post-merger, this difference is moresignificant, though this is likely, at least in part, becausethe larger scale of the post-merger electromagnetic fieldstructures means that one must go to larger radii to bein the wavezone and free of finite extraction radius ef-fects. In the following discussion of the luminosity, wetake the results from the largest radius shown.The significant post-merger emission evident in Fig. 2is of the order of magnitude one would calculate forthe Blandford-Znajek luminosity (Blandford & Znajek1977) using the initial magnetic field strength of the NS.However, as discussed further below, the magnetic fieldin the vicinity of the final BH is significantly lower, andthis emission is actually powered by the radiation ofmagnetic field loops. Previous studies of the collapse ofa magnetized NS in force-free or ideal MHD have foundthat the resulting BH sheds its magnetosphere within atimescale ≈ M BH ≈ . M BH / . M (cid:12) ) (Lehneret al. 2012; Lyutikov & McKinney 2011) or less. Thisis of the same order, if somewhat shorter than, thetimescale over which the post-merger luminosity, asmeasured at the largest extraction radius, decreases.One factor contributing to this longer timescale com-pared to NS collapse is the asymmetry of the mergerscenario which leads to magnetic field loops gathered toone side, instead of forming an axisymmetric, equatorialcurrent sheet.The luminosity is still significant at the end of thesimulation, but cannot be long lived if it is not poweredby the BH and/or accretion. We can obtain an upperbound on how long it could last at this level if we use thetotal magnetic energy stored in the NS’s initial dipoleas an estimate of the available energy (ignoring the factthat a significant amount is captured by the BH) toobtain U dip /L EM ≈
100 ms.To gain a quantitative sense of the directional depen-dence of this emission, we compute the luminosity withinspecific ranges of the polar angle, normalized by theirangular size. In Fig. 3 we can see that there is an in-crease in luminosity with polar angle, indicating greateremission near the orbital plane. The emission close tothe orbital plane is also less variable than the polar emis-sion. Near merger, the luminosity at all angles increases Note that such a match should be taken with care as the as-sumptions made in the derivation of the UI estimate becomeincreasingly suspect at merger. t [ms] L E M . ( B ) [ e r g / s] R R UI Figure 2.
Poynting flux luminosity measured at extractionsurfaces with radii R e = 510 km and 750 km as a functionof retarded time, along with the UI estimate (dashed) nor-malized to a surface field strength of B = 10 G. The insetdisplays the plus polarization of the GWs (indicated by thereal part of the l = m = 2 mode of Ψ ). The vertical lineindicates the time at which the GW peaks (i.e., when thebulk of the star crosses the effective light ring). Around thistime, the electromagnetic luminosity grows at a steep rateuntil it also reaches a peak. similarly rapidly. Postmerger, the emission from theequatorial region strongly dominates, and the luminos-ity at all angles then decreases as the small amount ofremaining matter is accreted and magnetic field is shedor swallowed by the BH (see also Lehner et al. 2012).The angular dependence is further illustrated in Fig. 4,which shows several snapshots of the angular depen-dence of the electromagnetic luminosity leading up to,and following merger. In addition to the stronger emis-sion in the vicinity of the equator, we can also see thestrong non-axisymmetric nature of the radiation.Finally, it is interesting to examine the large scalemagnetic field in the post-merger regime in which onehas a spinning BH with a small amount of matter re-maining outside. In Fig. 5, we display three snapshotsshowing select magnetic field lines at times ≈ . , . , and 9 . t [ms] L Ω [ e r g / s] o - 15 o o - 30 o o - 60 o o - 90 o Figure 3.
Luminosity at an extraction radius of R = 750km through specified ranges of the polar angle (normalizedby their angular size) with 0 ◦ corresponding to the directionof the total angular momentum of the system. During theinspiral phase, all regions show short timescale modulations,but the polar regions (in the direction transverse to the orbit)exhibit an overall increase through the inspiral that contrastswith the flatter level close to the equatorial (orbital) plane.At merger, the luminosity at all angles increases sharply,but subsequently decreases postmerger as the BH accretesthe little remaining matter, and the magnetic field lines areshed away or swallowed by the BH. tion propagate away, still producing sizeable, thoughmore transient, emission. DISCUSSIONThis study focuses on a BHNS system sitting roughlyat the boundary between two regimes: one in whichthe NS fails to disrupt with essentially no electromag-netic emission powered by post-merger matter , and theother in which the star fully disrupts leading to signif-icant electromagnetic emission due to the presence ofan accretion disk and/or ejecta. The disruption of thestar, leading to the formation of an accretion disk, isexpected for systems with a sufficiently low mass ratioand/or high BH spin—a scenario so far seemingly dis-favored by GW observations (assuming the binary BHobservations made so far are representative of BHs ingeneric BHNSs). For such a disk, a standing questionhas been the degree and timescale over which the mag-netic field re-orders and gives rise to a configuration fa-voring a potential jet (see e.g. Paschalidis et al. 2015;Ruiz et al. 2018; Christie et al. 2019b; Foucart 2020, forrecent efforts examining various aspects of this).In our case, where only a little material remainstemporarily outside the BH, we indeed see such a re-ordering. Our evolutions suggest the formation post-merger of poloidal structure which is a requirement ofmany models of BH jets (e.g. Hawley et al. 2015; Christieet al. 2019a). Furthermore, we note that in principle
Figure 4.
Snapshots of the Poynting flux at look-back times t − r = 6 .
4, 12.7, and 16.2 ms (from top to bottom) at anextraction sphere at radius r = 750 km. Figure 5.
Snapshots of the magnetic field at times t = 13 . B G to 10 B G, with all panels using thesame colormap. The field topology in the vicinity of the BHgradually collimates along the BH spin axis. the amount of material in such a disk need not be toolarge. As argued in Lee & Ramirez-Ruiz (2007), as littleas 10 − M (cid:12) can anchor magnetic fields with strengthsup to 10 G. For the physical parameters studied here,such an amount of matter however is accreted in a rel-atively short timescale. That the magnetic field is suc-cessfully collimated even with such a small amount ofmaterial remaining is suggestive of what would happenin more favorable configurations where the NS is dis-rupted earlier and the amount of post-merger matter islarger. In such a scenario, this matter would form a diskanchoring the magnetic field as magnetic winding andthe magneto-rotational instability increase the magneticstrength. The resulting increased field strength wouldpotentially arrest accretion (Tchekhovskoy et al. 2011),and help power strong outflows from the black hole re-gion.In the opposite direction (i.e. no spin, higher massratio), one would be left only with the type of electro-magnetic counterparts that would be induced by the fea-tures elucidated in this work: development of a complexcurrent sheet, strong twisting of field lines, and forma-tion of X-points; all favoring the emission of plasmoids.As we point out, the energy associated with such struc-tures is enhanced due to the spacetime dynamics, thusimproving the observational prospects.We comment briefly on the potential for observing sig-nals from the system. LIGO/Virgo is able to observeGWs from such binaries at distances in excess of 100Mpc, and this horizon will improve with further up-grades, including dramatically with planned third gen-eration detectors (e.g. Sathyaprakash et al. 2019). Onthe electromagnetic front, at high energies, and takingthe Burst Alert Telescope (BAT) at Swift as an exam-ple, its sensitivity to 100 keV photons (Barthelmy et al.2005) would allow for a detection up to a distance of D L (cid:39) B ) Mpc . Such an estimate assumes per-fect conversion efficiency; however, we note realistic es-timates are less optimistic by a factor of (cid:39) −
10% (e.g.Palmer et al. 2019). Considering the Fermi Gamma-rayBurst Monitor instead would yield luminosity distancesof the same order of magnitude (Meegan et al. 2009).At lower frequencies, coherent radio emission has beensuggested as a more likely observational prospect. As al-ready discussed in Most & Philippov (2020), both mag-netic reconnection in the current sheet and synchrotronmaser emission induced by magnetized bubbles shocking A study of the potential skymaps induced by the system in highenergy radiation will be presented elsewhere (Ortiz & et al. 2021). the ambient plasma are potential mechanisms for suchemission.Finally, it is interesting to compare the BHNS systemstudied here with magnetized binary NS systems. Inour study, the strong gravity of the BH plays a signif-icant role in twisting the magnetic field and increasingits strength, and we therefore expect that reconnectioncan be more ubiquitous and energetic in BHNS systems.In addition, the high mass ratio of a BHNS system en-sures low baryon pollution at all stages of the merger, incontrast to binary NSs. Thus, even if the BHNS config-uration is unfavorable for stellar disruption and its con-comitant emission, the system nevertheless offers boththe strongest gravitational field with a mostly pristineenvironment for electromagnetic emission from recon-nection. We also note that the inclusion of either BH orNS spin could enhance the electromagnetic output. Ex-ploring this, as well as the effect of the misalignment ofthe stellar magnetic dipole (instead of being transverseto the orbital plane), are interesting directions for futurework. ACKNOWLEDGMENTSWe thank Federico Carrasco and Alexander Philip-pov for discussions. LL thanks the CCA at the FlatironInstitute for hospitality during an early stage of thiswork. SLL is supported by the NSF under grants PHY-1912769 and PHY-2011383. CP acknowledges supportfrom the Spanish Ministry of Economy and Competitive-ness grants AYA2016-80289-P and PID2019-110301GB-I00 (AEI/FEDER, UE). WE and LL are supported inpart by NSERC through a Discovery Grant, and LLalso thanks CIFAR for support. Computations wereperformed on XSEDE resources and the Niagara super-computer at SciNet. SciNet is funded by: the CanadaFoundation for Innovation; the Government of Ontario;Ontario Research Fund - Research Excellence; and theUniversity of Toronto. Research at Perimeter Instituteis supported by the Government of Canada and by theProvince of Ontario through the Ministry of Research,Innovation and Science.REFERENCES
Abbott, B. P., et al. 2017a, Phys. Rev. Lett., 119, 161101,doi: 10.1103/PhysRevLett.119.161101—. 2017b, Astrophys. J. Lett., 848, L12,doi: 10.3847/2041-8213/aa91c9Abbott, R., et al. 2020a. https://arxiv.org/abs/2010.14533—. 2020b, Astrophys. J., 896, L44,doi: 10.3847/2041-8213/ab960fAlic, D., Bona-Casas, C., Bona, C., Rezzolla, L., &Palenzuela, C. 2012, PhRvD, 85, 064040,doi: 10.1103/PhysRevD.85.064040Barthelmy, S. D., et al. 2005, Space Sci. Rev., 120, 143,doi: 10.1007/s11214-005-5096-3Bezares, M., Palenzuela, C., & Bona, C. 2017, Phys. Rev.D, 95, 124005, doi: 10.1103/PhysRevD.95.124005Blandford, R. D., & Znajek, R. L. 1977, Monthly Not.Royal Ast. Soc, 179, 433Buonanno, A., Kidder, L. E., & Lehner, L. 2008, Phys.Rev. D, 77, 026004, doi: 10.1103/PhysRevD.77.026004Carrasco, F., & Shibata, M. 2020, Phys. Rev. D, 101,063017, doi: 10.1103/PhysRevD.101.063017Carrasco, F., Shibata, M., & Reula, O. 2021Chawla, S., Anderson, M., Besselman, M., et al. 2010,Phys. Rev. Lett., 105, 111101,doi: 10.1103/PhysRevLett.105.111101Christie, I., Lalakos, A., Tchekhovskoy, A., et al. 2019a,Mon. Not. Roy. Astron. Soc., 490, 4811,doi: 10.1093/mnras/stz2552 Christie, I. M., Lalakos, A., Tchekhovskoy, A., et al. 2019b,Monthly Notices of the Royal Astronomical Society, 490,4811, doi: 10.1093/mnras/stz2552Corral-Santana, J. M., Casares, J., Mu˜noz-Darias, T., et al.2016, A&A, 587, A61, doi: 10.1051/0004-6361/201527130D’Orazio, D. J., Levin, J., Murray, N. W., & Price, L. 2016,Phys. Rev. D, 94, 023001,doi: 10.1103/PhysRevD.94.023001Foucart, F. 2012, Phys. Rev. D, 86, 124007,doi: 10.1103/PhysRevD.86.124007—. 2020, Frontiers in Astronomy and Space Sciences, 7, 46,doi: 10.3389/fspas.2020.00046Gourgoulhon, E., Grandclement, P., & Novak, J. 2010 had home page. 2010Hansen, B. M., & Lyutikov, M. 2001, Mon. Not. Roy.Astron. Soc., 322, 695,doi: 10.1046/j.1365-8711.2001.04103.xHawley, J. F., Fendt, C., Hardcastle, M., Nokhrina, E., &Tchekhovskoy, A. 2015, Space Sci. Rev., 191, 441,doi: 10.1007/s11214-015-0174-7Ioka, K., & Taniguchi, K. 2000, Astrophys. J., 537, 327,doi: 10.1086/309004Lai, D. 2012, Astrophys. J. Lett., 757, L3,doi: 10.1088/2041-8205/757/1/L3Lee, W. H., & Ramirez-Ruiz, E. 2007, New J. Phys., 9, 17,doi: 10.1088/1367-2630/9/1/017home page. 2010Hansen, B. M., & Lyutikov, M. 2001, Mon. Not. Roy.Astron. Soc., 322, 695,doi: 10.1046/j.1365-8711.2001.04103.xHawley, J. F., Fendt, C., Hardcastle, M., Nokhrina, E., &Tchekhovskoy, A. 2015, Space Sci. Rev., 191, 441,doi: 10.1007/s11214-015-0174-7Ioka, K., & Taniguchi, K. 2000, Astrophys. J., 537, 327,doi: 10.1086/309004Lai, D. 2012, Astrophys. J. Lett., 757, L3,doi: 10.1088/2041-8205/757/1/L3Lee, W. H., & Ramirez-Ruiz, E. 2007, New J. Phys., 9, 17,doi: 10.1088/1367-2630/9/1/017